Petroleum Biotechnology - Developments and Perspectives

Studies in Surface Science and Catalysis 151 PETROLEUM BIOTECHNOLOGY Developments and Perspectives This page is inten...

Author: Rafael Vazquez-Duhalt | Rodolfo Quintero-Ramirez

64 downloads 1548 Views 50MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Studies in Surface Science and Catalysis 151 PETROLEUM BIOTECHNOLOGY Developments and Perspectives

This page is intentionally left blank

Studies in Surface Science and Catalysis Advisory Editors: B. Delmon and J.T. Yates Series Editor: G. Centi

Vol. 151

PETROLEUM BIOTECHNOLOGY Developments and Perspectives Edited by Rafael Vazquez-Duhalt Institute of Biotechnology National University of Mexico Morelos, Mexico and Rodolfo Quintero-Ramirez Mexican Petroleum Institute Colonia San Bartolo Atephehuacan, Mexico

ELSEVIER Amsterdam - Boston - Heidelberg - London - New York - Oxford - Paris - San Diego San Francisco - Singapore - Sydney - Tokyo

ELSEVIER B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam The Netherlands

ELSEVIER Inc. 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA

ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK

ELSEVIER Ltd 84 Theobalds Road London WC1X 8RR UK

© 2004 Elsevier B.V. All rights reserved. This work is protected under copyright by Elsevier B.V., and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier's Rights Department in Oxford, UK: phone (+44) 1865 843830, fax (+44) 1865 853333, e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage (http:// www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier's Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.

First edition 2004 Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress. British Library Cataloguing in Publication Data A catalogue record is available from the British Library. ISBN: ISSN:

0 444 51699 9 0167 2991 (Series)

@ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

V

PREFACE Without a doubt, historians will describe 20th and 21st centuries as the oil-based society. One hundred years ago oil exploitation began, first as a source of energy and later to include oil as a source of raw material. In addition to the 1 trillion barrels that have already been harvested, recent estimations shows that about 3 trillion barrels of oil remain to be recovered worldwide, half from proven reserves and half from undeveloped or undiscovered sources. Oil production is expected to peak sometime between 2010 and 2020, and then fall inexorably until the end of this century. After the production peak, the more expensive fuel sources will come into production. These include hard-to-extract oil deposits, tarry sands, and Synfuels from coal that requires alternative or complementary to conventional oil refining technologies. Our society has an inexorable challenge: to increase the production of goods and services for people, using new process technology that should be energetically efficient and environmental friendly. This also will be the case for the petroleum industry. Improvements in conventional oil refining processes such as cracking, hydrogenation. isomerization, alkylation. polymerization, and hydrodesulfurization, certainly will occur. Nevertheless, nonconventional biotechnological processes could be implemented. In contrast to the available processes, biological processing may offer less severe process conditions and higher selectivity for specific reactions. Biochemical processes are expected to be low demand energy processes and certainly environmentally compatible. The primary target of the petroleum industry is to enhance and maintain a continuous oil production. Preconceived ideas and misconceptions about biotechnology continue to limit the applications of biological processes in the chemical industry. Nevertheless, there are biotechnological processes that have been demonstrated to be industrially successful and that are shown to be sufficiently stable, productive and economic for commercial applications. Even if wastewater treatment and soil bioremediation are common biotechnological applications in the oil industry, petroleum biotechnology is still in its infancy. Doubtless, though, biotechnology will play an increasingly important role in future industrial processes. In this book, experts from 11 countries critically discuss the developments and perspectives of biotechnological processes for the petroleum industry. An integrated approach into the possibility of using petroleum biotechnology throughout the value chain of an oil company is presented. The authors discuss the evaluation of biotechnology as a general toolbox for solving some of the technology problems of today and future possibilities to implement new refinery processes. Petroleum refining could be enhanced by biochemical reactions in which the specificity exceeds by far these of chemical reactions. The selective removal of sulfur, nitrogen, and metals from petroleum by biochemical reactions performed by microorganisms and/or enzymes is discussed. Increasing supply of heavy crude oils and bitumens has increased the interest in the conversion of the high-molecular weight fractions of these materials into refined fuels and petrochemicals. This upgrading has typically been accomplished either with high-temperature and expensive processes thermal conversion (cracking or coking) or by catalytic hydroconversion. In contrast to the available processes, biological processing may offer less severe process conditions and higher selectivity to specific reactions. Enzymatic transformations of asphaltenes in nonconventional media, and biological upgrading to improve the quality of certain crude oils and liquid fuels could be envisaged, using biocatalysts to decrease aromaticity and sensitize aromatic heterocycles to subsequent heteroatom removal. Bioprocessing would complement conventional refining technologies and result in improved fuel quality at lower capital and operating costs and with reduced environmental impact.

vi

Innovative new processes could be explored, such as methanol production from methane. Methane monooxygenases are unique among known catalytic systems in their ability to convert methane to methanol under ambient conditions using dioxygen as the oxidant. The unusual reactivity and broad substrate profiles of methane monooxygenases suggest many possible applications in the petrochemical industry. In addition, the ability of anaerobic bacteria to convert petroleum into methane and thereby generate useful energy is a very interesting alternative. On the other hand, biological production of hydrocarbons by bacteria is revisited and its potential is explored, not only as an environmentally-friendly fuel supply, but also as a renewable source for basic petrochemicals. Microbial colonization of metal surfaces drastically changes the classical concept of the electrical interface commonly used in inorganic corrosion. Corrosion is a leading cause for pipe failure, and is a main component of the operating and maintenance costs of gas and oil industry pipelines. The cost of corrosion to the gas and oil industries was estimated in 2001 to be about $13.4 billion/yr and of this as much as $2 billion/yr may be due to microbiallyinduced corrosion. In order to moderate the economic importance of corrosion in the oil industry, molecular tools are used to study its microbial complexity. The current knowledge of the indigenous deep subsurface microbial community in petroleum reservoirs shows an enormous physiological diversity and constitutes a complex ecosystem with an active biogeochemical cycling of carbon and minerals. "Souring" of oil reservoirs by the formation of hydrogen sulfide has been a problem since the beginning of commercial oil production. Sulfate-reducing bacteria are the culprits that produce this noxious gas, leading to souring. This microbial process in wastewaters and oil field waters can be controlled by another group of microbes, known as nitrate-reducing bacteria. The use of nitrate to control microbially-produced sulfide in oil fields is a proven biotechnology that is grossly under-used by the petroleum industry. Its effectiveness has been demonstrated in many laboratory investigations and in some field studies. Nitrate has replaced biocides in some of the oil fields in the North Sea, and the results have been very positive. It is now very clear that land-based oil field operators should seriously consider using this proven biotechnology to control, and possibly eliminate, microbially-induced souring and the problems associated with H2S formation. Environmentally-related biotechnological processes were pioneered in the petroleum industry. Oil spill bioremediation technologies epitomize modern environmental techniques, working with natural processes to remove spilled oil from the environment while minimizing undesirable environmental impacts. The application of biological wastewater treatment in the frame of a process integration treatment technology will hopefully close the water cycle allowing "zero discharge" in the petroleum industry. Nowadays, water should be considered as one of the main raw materials of the petroleum industry and its treatment and reuse with advanced treatment technology should be applied. On the other hand, phytoremediation is an emerging technology that is based on sound ecological engineering principles, and that has developed into a more acceptable technology for the remediation of soils and groundwater polluted with residual concentrations of petroleum hydrocarbons. The advantages of using phytoremediation include cost effectiveness, aesthetic advantages, and long-term applicability. Finally, biological air treatment systems are among the established technologies that can be applied to control volatile organic compounds and odor emissions, and they are applicable for a wide range of volatile pollutants found in the petroleum industry. Biological treatment of polluted air emissions results from the competence of active microorganisms, including bacteria, yeast, and fungi, to transform certain organic and inorganic pollutants into compounds with lower health and environmental impact. Their applications are growing continually based on scientific and technological developments.

vii

The powerful tools of molecular biochemistry can be used to improve the enzyme stability and efficiency. These techniques may be applied to the particular needs of the petroleum industry. In addition, the enzymes isolated from extremophilic microorganisms are extremely thermostable and generally resistant to non-conventional conditions such as organic solvents and extreme pH. Thus, many enzymes and enzymatic proteins are still to be discovered. Rafael Vazquez-Duhalt The only way to discover the limits of the possible is to go beyond them into the impossible. (Arthur C. Clarke).


ix

Table of Contents Preface List of Contributors

Chapter 1 Use of Petroleum Biotechnology throughout the value chain of an oil company: An integrated approach. H.Kr. Kotlar, O.G. Brakstad, S. Markussen and A. Winnberg Statoil ASA. Trondheim, Norway

v xiii

1

Chapter 2 Petroleum biorefining: the selective removal of sulfur, nitrogen, and metals J.J. Kilbane II" and S. Le Borgneb a Gas Technology Institute, Illinois U.S.A. b Instituto Mexicano del Petroleo, Mexico

29

Chapter 3 Enzymatic catalysis on petroleum products M. Ayala" and R. Vazquez-Duhaltb a lnstituto Mexicano del Petroleo. Mexico b Instituto de Biotecnologia, UNAM, Mexico

67

Chapter 4 Prospects for biological upgrading of heavy oils and asphaltenes K.M. Kirkwood, J.M. Foght, and M.R. Gray University of Alberta, Canada

I 13

Chapter 5 Whole-cell bio-processing of aromatic compounds in crude oil and fuels J.M. Foght University of Alberta, Canada

145

Chapter 6 Biocatalysis by methane monooxygenase and its implications for the petroleum industry T.J. Smith" and H. Dalton 3 a University of Warwick, United Kingdom Sheffield Hal lam University, United Kingdom

177

X

Chapter 7 Biocorrosion H.A. Videla" and L.K. Herrerah a University of La Plata, Argentina University of Antioquia, Colombia,

193

Chapter 8 Molecular tools in microbial corrosion X. Zhu and J.J. Kilbane II Gas Technology Institute, Illinois U.SA.

219

Chapter 9 Potential applications of bioemulsifiers in the oil industry H. Bach" and D.L. Gutnick1' b Tel-Aviv University, Tel-Aviv, 69978, Israel a Taro Pharmaceuticals New York, U.S.A.

233

Chapter 10 Anaerobic hydrocarbon biodegradation and the prospects for microbial enhanced energy production J.M. Suflita", I.A. Davidova\ L.M. Gieg", M. Nanny" and R.C. Prince1' "University of Oklahoma, U.S.A. b ExxonMobil Research and Engineering Co., U.S.A.

283

Chapter 11 Using nitrate to control microbially-produced hydrogen sulfide in oil field waters R.E. Eckford and P.M. Fedorak University of Alberta, Edmonton, Canada

307

Chapter 12 Regulation of toluene catabolic pathways and toluene efflux pump expression in bacteria of the genus Pseudomonas J.L. Ramos, E. Duque, M.T. Gallegos, A. Segura and S. Marques Estacion Experimental del Zaidin, CSIC, Granada, Spain

341

Chapter 13 Bacterial hydrocarbon biosynthesis revisited B. Valderrama Instituto de Biotecnologia, UNAM. Mexico

373

Chapter 14 The microbial diversity of deep subsurface oil reservoirs N.-K. Birkeland University of Bergen, Norway

385

xi

Chapter 15 Biotechnological approach for development of microbial enhanced oil recovery technique K. Fujiwara11, Y. Sugai1', N. Yazawa1', K. Ohno\ C.X. Hong" and H. Enomoto1 a Chugai Technos Co. Ltd., Japan Akita University, Japan c Japan National Oil Corporation, Japan PetroChina Company Limited, China c Tohoku University, Japan

405

Chapter 16 Phytoremediation of hydrocarbon-contaminated soils: principles and applications R. Kamath, J. A. Rentz, J. L. Schnoor and P. J. J. Alvarez University of Iowa, U.S.A.

447

Chapter 17 Biological treatment of polluted air emissions S. Revah* and R. Auria" a Universidad Autonoma Metropolitana-lztapalapa, Mexico. b Universite de Provence, France

479

Chapter 18 Bioremediation of marine oil spills R. C. Prince and J. R. Clark ExxonMobil Research & Engineering Co.

495

Chapter 19 Biotreatment of water pollutants from the petroleum industry E. Razo-Flores, P. Olguin-Lora, S. Alcantara and M. Morales-Ibarria Institute Mexicano del Petroleo, Mexico

513


xiii

List of Contributors S. Alcantara Institute Mexicano del Petroleo Eje Central Lazaro Cardenas 152. C.P. 07730, Mexico D.F. P. J. J. Alvarez Department of Civil and Environmental Engineering, Seamans Center University of Iowa, Iowa City, Iowa, U.S.A. - 52242 R. Auria Laboratoire 1RD de Microbiologie, Universite de Provence CESB/ESIL, Case 925, 163 Avenue de Luminy 13288, Marseille Cedex 9 France M. Ayala Institute Mexicano del Petroleo. Eje Central Lazaro Cardenas 152, San Bartolo Atepehuacan 07730 Mexico DF, Mexico H. Bach Department of Molecular Microbiology and Biotechnology, Tel-Aviv University Tel-Aviv, 69978, Israel N.-K. Birkeland Department of Biology, University of Bergen, Box 7800, N-5020 Bergen, Norway O.G. Brakstad Sintef Materials and Chemistry, Trondheim, Norway .1. R. Clark ExxonMobil Research & Engineering Co. Annandale, NJ 08801 H. Dalton Department of Biological Sciences, University of Warwick Coventry CV4 7AL, United Kingdom I.A. Davidova Institute for Energy and the Environment and Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA. E. Duque Estacion Experimental del Zaidin. CS1C C / Profesor Albareda 1, 18008 Granada, Spain

xiv

R.E. Eckford Department of Biological Sciences, University of Alberta Edmonton, Alberta, Canada T6G 2E9 H. Enomoto Department of Geoscience and Technology, Graduate School of Environmental Studies, Tohoku University, Aramaki, Aoba-ku, Sendai 980-0845, Japan P.M. Fedorak Department of Biological Sciences, University of Alberta Edmonton, Alberta, Canada T6G 2E9 J. M. Foght Department of Biological Sciences, University of Alberta Edmonton, Alberta Canada T6G 2E9 K. Fujiwara Chugai Technos Co. Ltd. 9-20 Yokogawa-Shinmachi Nisi-ku Hiroshima City 733-0013, Japan M.T. Gallegos Estacion Experimental del Zaidin, CSIC C / Profesor Albareda 1, 18008 Granada, Spain L.M. Gieg Institute for Energy and the Environment and Department of Botany and Microbiology, University of Oklahoma, Norman, OK 73019, USA. M.R. Gray Department of Chemical and Materials Engineering, University of Alberta Edmonton, Alberta, Canada T6G 2G6 D.L. Gutnick Present address, Biotechnology Research Laboratories. Taro Pharmaceuticals U.S.A., 3 Skyline Drive, Hawthorne, New York, 10532, U.S.A. L.K. Herrerab Faculty of Engineering, University of Antioquia, Medellin, Colombia C.X. Hong PetroChina Company Limited, Jilin Oilfield Company Jilin province, China R. Kamath Department of Civil and Environmental Engineering, Seamans Center University of Iowa, Iowa City, Iowa, U.S.A. - 52242

XV

J.J. Kilbanell Gas Technology Institute, 1700 S. Mt. Prospect Rd.. Des Plaines 1L 60018 K.M. Kirk wood Department of Chemical and Materials Engineering, University of Alberta Edmonton, Alberta, Canada T6G 2G6 H.Kr. Kotlar Statoil ASA, R & D Center, Postuttak, N-7005 Trondheim, Norway S. Le Borgne 1 Institute Mexicano del Petroleo, Eje Central Lazaro Cardenas 152, Col. San Bartolo Atepehuacan, 07730 Mexico D.F., Mexico S. Markussen Department of Marine Environmental Technology, Trondheim, Norway S. Marques Estacion Experimental del Zaidin, CSIC C / Profesor Albareda 1, 18008 Granada, Spain M. Morales-lbarria Instituto Mexicano del Petroleo Eje Central Lazaro Cardenas 152, C.P. 07730, Mexico D.F. M. Nanny Institute for Energy and School of Civil Engineering and Environmental Science, University of Oklahoma, Norman, OK 73019, USA. K. Ohno Technology Research Center, Japan National Oil Corporation 1-2-2 Hamada, Mihama-ku, Chiba 261-0025, Japan P. Olguin-Lora Instituto Mexicano del Petroleo Eje Central Lazaro Cardenas 152, C.P. 07730, Mexico D.F. R. C. Prince ExxonMobil Research & Engineering Co. Annandale. NJ 08801 J.L. Ramos Estacion Experimental del Zaidin, CSIC C / Profesor Albareda 1, 18008 Granada, Spain

xvi

E. Razo-Flores, Institute Potosino de Investigation Cienti'fica y Tecnologica Camino a la Presa San Jose 2055,. C.P. 78216, San Luis Potosi, SLP, Mexico. .1. A. Rentz Department of Civil and Environmental Engineering, Seamans Center University of Iowa, Iowa City, Iowa. U.S.A. - 52242 S. Revah Department of Process Engineering, Universidad Autonoma Metropolitana-Iztapalapa (UAM-I). Apdo. Postal 55-534, 09340 Mexico D.F., Mexico J. L. Schnoor Department of Civil and Environmental Engineering, Seamans Center University of Iowa, Iowa City, Iowa, U.S.A. - 52242 A. Segura Estacion Experimental del Zaidin, CSIC C / Profesor Albareda 1, 18008 Granada, Spain T.J. Smith Biomedical Research Centre, Sheffield Hallam University Howard Street, Sheffield SI 1WB, United Kingdom .I.M. Suflita Institute for Energy and the Environment and Department of Botany and Microbiology, University of Oklahoma, Norman. OK 73019, USA. Y. Sugai Akita University Venture Business Laboratory 1-1 Tegatagakuen-cho Akita City ,010-8502, Japan B. Valderrama Departamento de Ingenieria Celular y Biocatalisis, Universidad Nacional Autonoma de Mexico. AP 510-3. Cuernavaca, Morelos, 62250, Mexico. R. Vazquez-Duhalt Instituto de Biotecnologia, UNAM. Apartado Postal 510-3 Cuernavaca, Morelos 62250 Mexico H.A. Videla Department of Chemistry. College of Pure Sciences, IN1FTA, University of La Plata, Argentina A. Winnberg Department of Biotechnology, N7465 Trondheim, Norway

xvii

N. Yazawa Technology Research Center, Japan National Oil Corporation 1-2-2 Hamada. Mihama-ku, Chiba 261-0025, Japan X. Zhu Gas Technology Institute, 1700 S. Mt. Prospect Rd., Des Plaines 1L 60018


Studies in Surface Science and Catalysis 151 R. Vazquez-Duhalt and R. Quintero-Ramirez (Editors) © 2004 Elsevier B.V. All rights reserved.

l

Chapter 1

Use of Petroleum Biotechnology throughout the value chain of an oil company: An integrated approach. H.Kr. Kotlara, O.G. Brakstad", S. Markussenc and A. Winnbergc. a

Statoil ASA, R & D Center, Postuttak, N-7005 Trondheim, Norway

Sintef Materials and Chemistry, bDept. Marine Environmental Technology, c Dept. Biotechnology, N7465 Trondheim, Norway

1. INTRODUCTION TO AN INTEGRATED APPROACH The history of biotechnology goes thousands of years back in time. One of the very first written statements of biotechnology is found in the Bible, telling that Lot was drinking wine, made through fermentation around 2000 B.C.E. In modern time Antoni van Leeuwenhoeck was the first to observe a microorganism in a primitive microscope in 1684. Louis Pasteur discovered how to protect against diseases by vaccination, using heat-inactivated organisms, around 1863. In 2002 the gene sequence of the human genome was completed. Biotechnology is continuously expanding, and will play an increasingly important role in future industrial process. Petroleum biotechnology is a very young and exiting part of these industrial possibilities It is well established that petroleum reservoirs contain active and diverse populations of microorganisms. Microbial growth within oil reservoirs has traditionally been associated with biofouling and souring. Furthermore, the potentials for microbial improved oil recovery (MIOR) have been investigated for many decades (see chapter 15)[1]. Recently, nitrate injection was introduced as a method for curing reservoirs "contaminated" by sulphate-reducing prokaryotes (see chapter 11)[2]. However, petroleum biotechnology possesses several other opportunities besides MIOR and nitrate injection. This chapter will focus on some of these issues. The primary target of the petroleum industry is to enhance and maintain a continuous oil production. In 1998/1999 Statoil initiated an R&D program

2

looking into the possibility of using petroleum biotechnology as an integrated approach throughout the value chain of the oil company. There were three main objectives: 1: Evaluation of biotechnology as a general toolbox for solving some of the technology problems of today. 2: Investigate future possibilities; e.g. to start refinery processes in the reservoir using dedicated microorganisms. 3: To generate a resource base for new genetic information achieved from the organisms in the reservoir. These objectives may be achieved through focusing on biotechnology as a new business concept of interest to the company. Coverage of all aspects of biotechnology would be an enormous task. However, the enhanced in-house understanding of reservoir microbiology has served as a basis for the few selected areas described below: . New techniques in exploration and production: Application of molecular biology techniques as new tools for specific identification and characterization of hydrocarbon sources during exploration and production. Samples may come from drill cuttings from exploration wells; produced oil and formation water; sediments from sea floor seep zones; etc. • Biological well treatments (preventive medication): Clogging of wells by scaling, hydrates, etc. may be prevented by applying environmentally friendly biological produced chemicals. This may be achieved by developing self-sustained, natural existing or bioengineered microbial populations placed inside the reservoir. The target is to produce biological substances that can replace traditional chemicals, and that this remediation will increase treatment lifetime to ensure a continuous oil production. • Bioreactors: Low energy biological processes for up-grading of oil to improve quality and thereby reduce penalty pricing. Various types of bioreactors and enzyme systems can replace traditional catalysts for certain chemical reactions, waste handling or the production of bio-energy. • New application of extremophiles: New thermophilic and piezophilic enzyme system can enable new bioengineering processes and products for applications in the above-mentioned areas, or give rise to entirely new products and business opportunities. Combined approaches of microbiology, biochemistry and DNA technology are used to obtain microorganisms with specifically designed metabolic

3

functions. Such organisms can be applied in reservoirs for the production of various treatment products or enzymes in situ. Thermophilic enzymes may also be employed to overcome possible fundamental problems related to the growth characteristics of these microorganisms. Additionally, the "gene-pool" of the indigenous microbial assemblages of the reservoir have direct implication to the success of the product in the above suggested business areas. Environmental aspects/public awareness: Apart from providing technical solutions, the outcome of this program will have a great impact on meeting the environmental challenge of the future. The Norwegian authorities consider many of the production chemicals applied in the fields today as harmful, and in the Norwegian sector of the North Sea there is a program for phasing out such chemicals, replacing them with more environmentally acceptable alternatives. Biotechnology may provide us with more environmentally friendly alternatives. Value generation: This program will contribute to increasing and maturing the reserve base (upstream), as well as creating business opportunities or increasing market shares downstream. The Fig. 1 below illustrates the potential influence of biotechnology throughout the entire value chain within an oil company.

Fig. 1. Biotechnology throughout the value chain.

4

The main challenges are related to: • The biological activities in a reservoir are still poorly understood. Growth control of reservoir microbes, and the knowledge to achieve this control, will be crucial. In bioreactor-type processes, however, this will be possible. . There are fundamental questions related to energy pathways and reaction rates that need to be resolved. Direct use of tailor-made enzyme system might bypass some of these obstacles. • In bioreactors, the main challenge is to achieve sufficient reaction rates that are required for a commercial process. This is not a challenge from the microbiological aspect only, but also from a chemical engineering point of view. Acquiring new knowledge: In order to balance the beneficial and detrimental effects of microbial growth in the reservoir, new knowledge is required. Growth and possible excretion of products under different reservoir conditions are not well known. To date, various types of chemicals are injected into the reservoir in order to maintain or restore oil production, e.g. to counteract or minimize the influence of scaling, hydrate and asphaltene precipitation. Occasionally, chemicals and antibiotics are injected to prevent microbial growth. Some of these chemicals are known to serve as energy source for the microorganisms, i.e. nitrogen, phosphor and carbon sources. [3-4]. Reservoir conditions vary significantly, and thus, the microbial communities will respond differently depending on this external influence. It is imperative to acquire in depth understanding of the growth and production of microbial products under the different reservoir conditions. In this respect modeling tools may be used to simulate how the changes will influence on the indigenous microorganisms. Joint efforts from internal experts and external collaborators are vital to the success of this type of projects. Much knowledge on microbial technologies already exists but the molecular biology approach represents a bold and important step forward. The nature of this research requires long-term commitment and support from the R&D management. A thorough understanding and awareness of the ethical implications is needed for all involved. 2. MICROBIAL DNA FINGERPRINT TECHNIQUES IN EXPLORATION AND PRODUCTION

Several studies have documented microbial communities in hot oil reservoirs (see chapter 14)[5-9]. Indigenous microbial communities have also been detected in core samples and water saturated regions of reservoirs [10]. Members of indigenous reservoir communities may include strictly anaerobic sulfate-reducing prokaryotes [5, 11-12] and methanogens [13-15], as well as

5

other microbes [9, 15]. Thus, one would expect to find genetic markers of microbial activities both during exploration, drilling and production. Statoil has filed a patent application for utilization of DNA technologies as a tool for identification and characterization of hydrocarbon sources during drilling or sampling from sea floor seep zones. Drill cuttings from exploration wells, sediments from sea floor seep zones or other specimens could be analyzed with a selection of specific DNA probes/markers. These specific DNA probes are taken from microbes found to be linked to different oil producing fields in the North Sea and other sources. The energy sources for these organisms will be constituents of the oil, gas or others, specific for the reservoir zones and conditions of the particular field This genetic tool may give valuable information on possible migration routes of the hydrocarbon from the source rock. Specific recognition patterns might also be used in monitoring different reservoir zones during production, and further indicate the individual contribution of the particular zone to the overall production. Possibly, sweep efficiency pattern could be calculated. Detection of DNA from drill cuttings, sediments, or core samples during explorative drilling may result in defined species pattern, resulting in indications of potential hydrocarbon bearing zones (Fig. 2).

Fig. 2. System for characterization of microbes in exploration cores by culture-dependent and -independent approaches, based on 16S rRNA gene sequencing. The sequences are used for the generation of DNA probes to be used for screening of cores.

6

2.1. Microbial diversity in oil reservoirs It is essential to establish databases of the microbial ecology in petroleum reservoirs. Genetic tools for exploration and production can then be developed. The knowledge of the in situ microbial activities should be improved through an interdisciplinary collaboration between specialists in petroleum exploration and production, chemists and microbiologists. Understanding the interactions between the biosphere and the geosphere is essential. The microbial diversity of two North Sea reservoirs (termed reservoir A and B) has been studied in some detail [16-17]. Both a culture collection and a 16S rDNA library have been established for these reservoirs. 2.1.1. Culture-independent methods Culture-independent methods have recently been used for the characterization of microbial communities in some oil reservoirs [9-10]. In these studies, DNA was extracted directly from reservoir samples (produced water, core samples, drill cuttings etc.) This approach was used for the comparison of microbial assemblages in some North Sea reservoirs with different reservoir characteristics and production histories. In our studies microbial communities differed significantly between the reservoirs (Fig. 3). Sequence studies of 16S rDNA clones from reservoir A showed that 32 % of the clones aligned to the sulfide reducing thermophile Archaeoglobus fulgidus, while bacterial clone inserts aligned to a variety of types, including Sphingomonas, Herbaspirillum, Nevskia, Aquabacterium, Alcanivorax, Bacillus and Acetobacterium. Clones from reservoir B were dominated by sequences aligning to the a-proteobacteria Erythrobacter, the sulfide-oxidizing e-proteobacteria Arcobacter, the halotolerant y-proteobacterium Halomonas, and the thermotogales Geotoga. Several of the microbial genes detected in our studies have been found in produced fluids or enrichment cultures from oil reservoirs in the Pacific Ocean or Canada [9, 18]. The differences in the assemblage compositions between oil reservoirs and other subsurface structures may reflect the geochemical influences on the community structures [19-20]. Biodegraded oils dominate the world's petroleum inventory, and microbial activities play an essential role in most oil reservoirs [21]. Recent studies have emphasized the impact of an active potentially indigenous subsurface community [19]. 2.1.2. Culture-based methods Most studies of reservoir communities have been conducted by culturebased methods [7-8, 22-23]. As a supplement to the culture-independent characterization of the two North Sea oil reservoirs, culture-based methods were used to study the diversity of the cultivable microbes in produced fluid from the reservoirs. Enrichment media for fermentatives, methanogenes, sulfideoxidizers, sulphate-reducers and acetogenes were designed, and cultures from

7

the two reservoirs showed dominance of small rods, single or in short chains, and sheathed rods (Thermotogales like). Pure isolates were obtained from only one of the reservoirs, reservoir A. Even though the enrichments from the other reservoir, reservoir B, showed a variety of organisms, it was not possible to obtain any pure isolates from these. The 16S rDNA clones from these enrichments aligned to Thermosipho japonicus, Bradyrhizobium and Aquabacterium. 16S rDNA clones from isolates from reservoir A, showed dominance of Archaeobglobus fulgidus, Methanococcus thermolithotrophicus, Thermococcus sibiricus and Thermosipho japonicus. Several of the sequences abundant in the cultures were not found in the clone library from the cultureindependent approach (2.1.1). This is in accordance with other studies [9], and suggests that several of the predominant members of the enrichment cultures (e.g. Thermosipho) are not the predominant member of the reservoir communities, but show fast-growing characteristics in several of the culture media. Other cultures included a-, P-, s- and y-Proteobacteria Sphingomonas, Stenotrophomonas, Halomonas meridiana, and Geospirillum, and the Grampositive bacterium Thermoanaerobacter ethanolicus.

Fig. 3. DGGE analysis of PCR-amplified 16S rDNA sequences from two North Sea oil reservoirs, reservoir A (1, 2) and reservoir B (3, 4, 5). Only sample 2 contained fluids with seawater penetration.

8

Thermophilic species of Thermotogales, Archaeoglobus, Thermoanaerobacter, Methanococcus and Thermococcus have been reported from hightemperature oil reservoirs [6-9, 14]. Several of these microbes are typical sulfurutilizers, being active in desulphurization of crude oil. These microbes may be the predominant sources for H2S generation rather than typical sulphatereducing bacteria, and interestingly several of them were enriched in culture media designed for SRB. 2.1.3. Detection of specific microbes Monitoring of microbes in the oil reservoir has traditionally been accomplished by culture methods, e.g. MPN methods for quantification of viable sulphate-reducing bacteria (SRB), as recommended by the American Petroleum Institute [24]. Some commercial techniques have also been introduced, for instance a commercialized immunoassay for semi-quantification of the SRB-specific enzyme APS reductase [25]. Monitoring may also include molecular biology methods. Currently, two RNA-based methods are investigated, fluorescence in-situ hybridization (FISH) and nucleic acid sequence-based amplification (NASBA). By using RNA detection mainly the metabolic active cells are assessed. The FISH methods include fluorescencelabeled DNA probes for the targeting of specific microbes. An example is given in Fig. 4 where bacteria, archaea, Archaeoglobus, Arcobacter and Erythobacter are enumerated in production fluids from two reservoirs. These methods may be further refined for offshore analysis by using field equipment, e.g. the Microcyte fluorescence cell counter. NASBA is an isothermic alternative to PCR [26]. Real-time miniaturized lab-on-a-chips systems are currently under development with the NASBA technology as basis [27]. 2.1.4. Characterization ofmicrobial dynamics by microarrays Nucleic acid microarrays have recently been introduced for phylogenetic identification in microbial ecology. Basically, microarrays consist of series of specific DNA probes (grabber probes) that are printed on glass slides. Sample nucleic acids are extracted and labeled (e.g. by fluorescence) and incubated on the slides, followed by recording. Labeled detector probes may be used for detection as alternatives or supplements to labeled target DNA [28]. The microarrays are made quantitative by employing reference DNA to normalize variations in spot size and hybridization (29). The methods provide a powerful tool for parallel detection of 16S rRNA genes [30-31] and may be particularly useful for environmental studies of phylogenetically diverse groups. Although most arrays are based on the PCR amplification of target genes prior to array hybridization, systems have also been described where direct profiling of extracted rRNA from environmental samples have been used [32]. Printed slides may be brought offshore and target genes quantified directly on the platforms by

9

portable devices. Arrays have also been established for the assessment of functional gene diversities and distribution, for instance with genes from the nitrogen cycling [33-34]. For offshore conditions the sulphur and nitrogen cycles may be addressed during curing of biological souring by nitrate injection. 3. BIOREACTOR: POTENTIAL USE OF BIOCATALYSTS IN CRUDE OIL UP-GRADING AND REFINING Until recently, research within oil biotechnology mainly focused on biodegradation and bioremediation in connection with clean up after oil spills, and less on the application of microbial systems in industrial processes. However, the interest in the latter has been growing the last years, addressing problems like asphaltenes, high sulfur content, the poor transportability of heavy crudes due to high viscosity, the presence of heavy metals and polyaromatic/ heterocyclic compounds (see chapters 2, 3, 4 and 5). The aim of our activity is to use biotechnological processes in up-grading of "problem" oils/heavy oil and refinery fractions. The overall scope is to define microbial/biotechnological technologies along the crude oil value chain that will give the potential highest cost-benefits, competing with or being superior to existing methods, or even better, provide solutions where no acceptable methods exist. In the current program there has been focused on:

Fig. 4. FISH enumeration of the total concentrations of cells (DAPI), bacteria (EUB338), archaea (ARCH915), Arcoglobus (ARGLO605) and thermotogales (THERSI672) in produced fluids from two North Sea reservoirs, Reservoir A and Resevoir B wl and w2.

10

Reduction of the viscosity of heavy crudes through partial degradation of waxes and/or asphaltenes, thereby increasing the transportability. Microbial or enzymatic ring opening of polyaromatic hydrocarbons in refinery distillates in order to increase the fraction of aliphatic components. Removal of heavy metals such as nickel and vanadium from crude oils through microbial sequestering, thereby simplifying the subsequent refining of the crude. Although chemical means to tackle the above problems exist, they are often relatively expensive and may lead to pollution of the environment. Biotechnological processes may represent new and more environmentally friendly alternatives for value enhancement of heavy oils and partially distilled petroleum products. 3.1. Pre-refining Up-grading of crude oils by biocatalytic processes may take place anywhere from down-hole to the refinery; in the reservoir, at the wellhead, during tanking, transport and storage. The pre-refining opportunity is to utilize the time slot from the start of drainage in the reservoir to the crude reaches the refinery stage. At any of these stages, a specially designed biocatalyst could be introduced (see Fig. 1). Although there will be considerable differences between traditional crude oils and the heavy crudes in physical handling as well as refinery processes, the chemistry of the compounds that need to be bio-converted could be close relatives within the same classes. 3.1.1. Increased transportability by biocatalytic cleavage of heavy compounds Extraction, transportation and handling of heavy oils often represent a problem due to high viscosity. Several classes of molecules are important in building viscosity. These are asphaltenes, waxes and the more heavy fractions of polyaromatics. Controlled biodegradation of asphaltenes and waxes in heavy crudes are highly desirable, as these processes could lead to a substantial economical gain (see chapter 4). Wax is degraded by several bacterial species that use the degradation products for their metabolic pathways [35-36]. Efficient methods for isolation of wax-utilizing microorganisms with the help of selective media, bacteriophages, and paraffin wax baiting system have been developed [37-38]. Although the enzymology of the wax degradation is not understood, some clues have been obtained through studies of wax biosynthesis by certain bacteria, such as Acinetobacter spp. [39-40].

11

Biodegradation of asphaltenes seems to represent a more challenging problem - very few publications is found on this subject. However, several studies have shown that biodegradation of asphaltenes occurs in nature [41], and that certain bacteria, such as Acinetobacter and Providencia, proliferate in environments containing high amounts of asphaltenes [42]. Fungi capable of "erosion" of hard coal due to the cleavage of asphaltenes have also been reported [43], as well as combined steam/bacteria treatment of asphaltene depositions [44]. In addition, biodegradation of bitumen has been observed [45], and bacteria like Pseudomonas, Flavobacterium, Acinetobacter, and Caulobacter growing on bitumen-contaminated surfaces have been described [46]. Potential processes are not limited to the natural occurring microorganisms and their native enzymes. By gene technology it is possible to improve key enzymes by rational engineering and by use of "gene shuffling" techniques. These methods make it possible to rapidly "adapt" a given enzyme to new substrates, or dramatically change the enzyme's properties such as Km, pH and temperature optimum [47]. The modified enzyme(s) may then be introduced into the appropriate microorganism(s) and its over-production, may greatly enhance the ability of this microbe(s) to reduce the viscosity of heavy oils. 3.1.2. Demineralization - Biosorption of heavy metals Demineralization of heavy oils that contain considerable amounts of Ni and V is an important issue for oil industry due to refinery stage catalyst poisoning. Several reports describing the use of microorganisms for bioremediation of environments polluted with heavy metals, suggest that the use of microbes for demineralization of heavy oils is possible [48-49]. Six mechanisms for microbial resistance to heavy metals have been described: exclusion by a permeability barrier, intra- and extra-cellular sequestration, active transport by efflux pumps, enzymatic detoxification, and reduction of sensitivity of cellular targets to metal ions [50]. For demineralization of heavy oils, sequestration and enzymatic detoxification seem to be the most relevant mechanisms to study. In our current work we have just entered this particular field of research. 3.2. Biocatalytic refining, distillate quality improvements Perio-refining or post-refining technologies might also be of interest. Although some of these areas have been addressed elsewhere in this book, we would like to convey some of our own work (see chapters 2, 3, 4 and 5).

12

3.2.1. Selective ring opening The mechanisms, the biochemical pathways, and the genetics of degradation and bioconversion of hydrocarbons in general, and polycyclic aromatic hydrocarbons in particular have been extensively studied [51-53]. The research has mainly concentrated on biodegradation and bioremediation in connection with cleanup after oil spills etc., and less on the application of these systems in processes. However, the interest in the latter has been growing the last years. In the petroleum industry there is a desire for products with a larger fraction of aliphatic components, and thus a higher H/C-ratio, and microbial/enzymatic ring opening of aromatics may be used to achieve this (see chapter 5). Development of biocatalysts for aromatic- and heterocyclic ring opening, including nitrogen compounds such as the polycyclic compound carbazole is of particular interest. Middle distillate fractions from thermochemical conversion of heavy oils contain di- and tricyclic aromatics with low fuel value. These are currently upgraded by expensive high pressure-high temperature chemical hydrogenation. A Canadian research group [54-55] has suggested an alternative to thermochemical cracking: "microbial cracking" - a two-step process where the aromatic rings first are cleaved enzymatically by a blocked mutant under "near ambient conditions", followed by hydrogenation of the oxygenated product under mild chemical conditions. Our group is currently engaged in a project, "Upgrading of crude oils and refined products" involving selective ring opening of aromatic distillates. In this work, a blocked mutant of Sphingomonas is used for studies of bioconversion of aromatic distillates in a bioreactor [56]. Bioconversion of aromatic compounds in a real feedstock from crude oil in a bioreactor system. The content of polyaromatic hydrocarbons (PAH's) in the diesel fuels contribute to low cetane numbers and particle emissions from combustion. The present study focuses on the use of a continuous bioreactor system for up-grading of light gas oil (LGO) feed stock from the refinery as a potential industrial process. This is done by biocatalytic ring opening of the PAH's to generate a more paraffmic diesel fuel. Two different bacterial strains, Sphingomonas yanoikuyae N2 and Pseudomonas fluorescence LP6a 21-41 (donated by Dr. Julia Foght, University of Edmonton Canada), and a mixed blend of six different strains were compared for biocatalysis of the PAH's in the LGO feed stock using a fed batch reactor/semi-continuous reactor.

13

Fig. 5. Schematic outline of the procedure for making blocked mutants with an inactive enzyme by gene disruption.

The P. fluorescence LP6a 21-41 was obtained by transposon mutagenesis and its genetic background remains unknown (see chapter 5). The PAH degradation pathway of S. yanoikuyae was genetically engineered in order to obtain a recombinant strain accumulating one of the intermediates, 2hydroxychromene-2-carboxylate. Thus, the degradation of PAH would terminate after the ring opening. This is important for keeping the octane number of the hydrocarbon fraction, and this was achieved by inactivating the gene encoding the specific hydratase-aldolase enzyme, (NahE), by gene disruption (Fig. 5). The mixed blend consisted of six different strains obtained from commercial culture collections and isolates from mud samples collected at a water purification plant. The organisms in this combined blend were not genetically modified to terminate the degradation of PAH's after the ringopening step. The LGO feed stock did not have any toxic effects in concentrations up to 50 vol%. In these studies, a continuous feed of 20 vol% was used. Comparison of the N2 and Lp6a 21-41 mutants show that the two strains have different uptake mechanisms and different preferences for certain PAH's. The N2 strain shows the highest conversion of the least substituted aromates (Fig. 6), while Lp6a 21 41 show a somewhat broader specificity range (data not shown).

14

Fig. 6. Bioconversion of light gas oil by the specially designed Sphingomonas spp. N2.

In order to apply the concept to a real industrial process, higher degrees of conversion of the more substituted aromatic compounds are necessary. The enzyme systems in the PAH degrading pathway of N2 were found to be too specific. Using the mixed biocatalytic blend a broader range of substrate conversion was observed. More than 30 % of both the di- and the tri aromatic compounds were removed from the LGO feedstock; in addition, approximately 30 % of the sulfur containing substrates was removed (Fig. 7). As already mentioned, the mixed blend had not been genetically modified to terminate the degradation of PAH's after the ring-opening step. The further uses of this mixed biocatalytic blend with respect to developing an industrial process; will demand genetic modification of the strains The results achieved in the fed batch reactor are now being verified in a continuous bioreactor to mimic a potential industrial process. Figure 10 shows the schematic outline of the continuous bioreactor. In conclusion, microorganisms with biocatalytic pathways that will selectively convert aromatic compounds in a crude hydrocarbon mixture without degrading aliphatic compounds exist. Such strains have been used as model systems for studies of bioconversion of aromatic distillates (LGO) from the refinery in a bioreactor system.

15

Fig. 7. Efficient bioconversion by a mixed biocatalyst.

The PAH degradation pathway of Sphingomonas yanoikuyae DSM 6900 have been genetically modified in order to obtain a recombinant strain that terminates the PAH degradation after the ring-opening. The LGO feedstock from the refinery has been shown to have no toxic effects on the tested organisms, S. yanoikuyae mutant N2 and Pseudomonas fluorescence LP6a mutant 21-41, in concentrations up to 50 vol%. This is of vital significance, because in an overall technological process it will be of importance to keep the water volumes as low as possible. The uptake mechanism and also the substrate specificity differ between the two strains. The substrate specificity seems to be rather narrow for each (both) of the strains, non- or mono substituted PAH's were the preferred substrates. Importantly, no C is lost by breaking the C - C chains in the blocked mutants. The organisms are not gaining energy by the reaction. It is of value that neither the fuel properties nor the cetan number are lost. A broader range of substrate specificity was observed with a mixed biocatalytic blend. More than 30% of both the di- and tri aromatic

16

compounds and approximately 30 % of the sulfur containing substrates were removed from the LGO feedstock in a continuous bioreactor system. In future refinery processes this might replace the energy-expensive distillation processes. These results suggest that bioreactor systems have the potential for up-grading of hydrocarbon refinery fractions, heavier distillates and possibly crude oils. In the years to come governmental regulations will be very strict on both PAH and sulfur content in the diesel fuel. These preliminary studies are thought as initial steps in a process of making a more environmental acceptable diesel fuel with dramatic reduction in both PAH's and sulfur content, while still maintaining adequate fuel combustion values (Fig. 8). This will be a bio-upgraded environmental friendly diesel.

Bioconversion for more environmental friendly diesel fuel

Fig. 8. Bio-reactor for conversion of PAH's in a real feedstock from crude oil

17

Study of pure enzyme vs. whole cell based biocatalysts. In future investigations this will include "the aromatic ring opening dioxygenase system". The Sphingomonas yanoikuyae N2 will be used as a model system for comparing enzyme and whole cell biocatalysts. In many instances it is an advantage to use pure enzyme systems instead of whole cells as biocatalysts (see chapter 3). Enzyme reactions are specific and easy to control, they can be carried out in non-aquatic environments, and enzymes, as other chemical catalysts, will not consume carbon i.e. the carbon content in the fuel will be preserved. The opening of the aromatic ring (e.g. naphthalene, Fig. 9) is a four step enzymatic process starting with a dioxygenase reaction, then a dehydrogenation followed by a second dioxygenase reaction and finally an isomerization. The first oxygenation requires NADH, but the formed NAD+ is recycled to NADH in the dehydrogenation reaction. The challenge is to develop a system where this multistep enzyme reaction could proceed efficiently in a cell free system.

Fig. 9: Metabolic pathway of naphthalene showing the enzymes involved See reference [57],

18

3.2.2. Bioreactors Bioconversion of refinery fractions may take place using growing or resting cells, "dead" cells, or immobilized cells or enzymes as biocatalysts. Aromatic ring-opening involves a multistep metabolic pathway. Multistep enzymatic reactions often require co-factors and/or reducing power (NAD (P) H) that has to be regenerated or supplied for the enzymatic reaction to take place. Thus, whole cells, rather than pure enzymes, are often required. The biocatalysts are usually contained in the aqueous phase and the reaction take place either in this phase or at the interface between the aqueous and the organic/oil phase. The components in the refinery fraction that are being up-graded usually show low water solubility, while the converted products usually are more soluble in the aqueous phase than in the organic/oil phase. Mass transfer of substrates and products between the water and oil phase is a major challenge. To achieve adequate mass transfer, reactors capable of generating a large interface between oil and water should be chosen. Various types of bioreactors have been employed by others [58], including stirred tank reactors, airlift reactors, emulsion phase contactors reactor and fluidized bed reactors. The current investigation has used stirred tank reactors run in batch, fed-batch and continuous mode with free growing or resting cells. However, immobilized cells and enzymes are included in the next phase of studies.

Fig. 10. Schematic of a bioreactor for continuous feed of LGO.

19

Continuous processes are well suited for multiphase processes. In the continuous bioreactor based on a stirred tank reactor in fig. 10, a continuous stream of substrate (oil phase) is run through the reactor while the biocatalyst (in the water phase) is recycled. Recycling of the biocatalysts reduces the amount of water needed in the process. The overall economy of the process is also dependent upon the lifespan of the biocatalyst and their stability in water/oil media. In a continuous reactor it is possible to regenerate or boost the biocatalyst. In the current studies, problems have been encountered connected to formation of stable emulsions. The emulsion increases mass transfer, but the stable emulsions made phase separation problematic. Currently, different approaches are explored to solve this problem. Enzymes or cells may be immobilized by binding or adsorption to membrane surfaces or beads, or by entrapment in a matrix. In a continuous reactor with an immobilized biocatalyst, it is possible to have a higher biocatalyst concentration, little or no water in the reactor, and the product separation is easy. Reactions with purified enzymes might be easier to control compared to whole cell biocatalysts (see chapter 3). Whole cells may contain different metabolic pathways and could lead to production of several unwanted by-products. By co-immobilization of series of enzymes in the water phase of the reactor, it might be possible to run multi step enzymatic reactions. Realistic cost of developing new technology. New technologies are often met with obstructive arguments. Sentences like "it cannot be done" and "it is impossible" are customary. Such arguments are "progress killers", and within the oil industry, new techniques will have to compete with traditional technology that has been optimized for the last 50 years. A lesson can be learned from the Canadians. None of their syncrude technologies for mining bitumen would have been available today if they had listened to the "wise guys" 14 years ago. At that time the operational cost of the technology was more than 30 US$/bbl, today the operational cost is down to around 10 US$/bbl. The OPEX (operational expenditure) profile (Fig. 11) illustrates the cost developments in developing new technology for mining bitumen. This curve profile is believed to be quite universal for most new technology implementations. 4. WELL TREATMENTS TO SECURE CONTINUOUS PRODUCTION BY PREVENTIVE MEDICATION. MICROBES AS SELF-GENERATING SYSTEMS Preventive medication could be defined as intelligent treatment concepts performed in advance during the complementation phase, before the impairment

20

in productivity occur in the well. The preventive actions are to avoid the onset of these predicted situations. With the advance in drilling and completion, increasing number of complex and expensive wells are being installed, e.g. multilateral, multi-zones, sidetrack and horizontal. The infrastructures that are in place, such as flow lines and platforms, also enable the targeting and drainage of the additional reserves found near the exiting fields. Very often these additional oil and/or gas are produced via tieback and satellite facilities. Successful treatments of stimulation, scale squeeze and tubing deposit removal in these wells can no longer rely on the traditional method of bullheading. Special tools such as coil tubing and inflatable plug will be needed to place the chemicals accurately down-hole. Intervention in these wells will be prohibitory expensive due to tools hire, personnel and extended period of deferred oil production (tools run). It is important to realize that for certain type of completion, well re-entry is almost impossible despite accepting the financial penalty. There is clearly a need to develop an intervention free system for these wells that allow the flow of oil unhindered and preferably with the chemicals pre-delivered down-hole.

Syncrude Canada OPEX

Fig. 11. OPEX profile in developments of new technology for mining bitumen. The curve shows the measured cost until 1998, then the further projection. The bars in 99, 00 and 01 are the actual cost. (Maurice B. Dusseault, personal communication).

21

4.1. Preventive treatment: Increased productivity by self-generating- or more environmental friendly treatment system/processes (scale, hydrate, asphaltenes, wax, etc.) The generation of effective production chemicals could be achieved using a self-sustained, natural existing or bio-engineered, microbial population. This will protect and free the well from most other intervention treatment and could be of great economical interest to an oil company, enhancing both well recovery and well productivity. This will imply the search for microbes that have the genetic machinery to produce certain treatment chemicals (i.e. organic acids, enzymes, surfactants, antifreeze-proteins etc). Alternatively, genetic engineering could be used to introduce this capability to the organisms. Such organisms could be introduced to the near well bore area by various means (i.e. squeezed with/without solid support, immobilized, combined with nutrients, etc), to produce the treatment chemicals. If the organism is not fit for life under the reservoir conditions, the bacteria can be used in bioreactors to produce the desired product. Bypassing the problems of placements: Correct placement of the treatment fluids is of crucial importance to the overall treatment success. Numerous treatments have failed due to poor placement. Nonetheless, in many wells, especially in gravel packed wells, uniform placement is difficult to achieve. With this new technology placement should no longer be the problem. The strategies of this new technology are illustrated in fig. 12 and include: Placement of the treatment during the completion stage. This can be done either by bullheading the specially designed organism together with nutrients into the formation, or by coiled tubing (CT) deployment. Use of porous particles soaked with the product placed inside the gravel packs at the completion face. Use of micro encapsulation, with the desired microorganism together with nutrition inside the capsules. Inject far beyond the critical matrix in the well. If successful, this concept constitutes the only possible self sustained and lasting method by which production chemicals can be produced in situ and to allow wells to operate free of most interventions.

22

In situ production of treatment chemicals

Fig. 12. Schematic view of in situ production of treatment chemicals.

4.2 Green treatment products In order to prove the basic concept of the above-mentioned technology of preventive medication for secured production, the following approaches have been made: A synthetic gene, coding for polyaspartate (polyAsp), has been cloned in E.coli. In a construct with 75 basepairs, coding for 25 amino acids, with a fusion protein included, the polyAsp polypeptide was expressed in the host cell. Most service companies in the oil industry are supplying polyAsp as a combined scale - and corrosion inhibitor. Recently, polyAsp has also proved to be an efficient bridging agent, boosting the squeeze lifetime of traditional scale inhibitor jobs. PolyAsp is classified as a green treatment product, being more than 60 % biodegradable and non-toxic. From 2005 the Norwegian government, through chart 12 and the Norwegian Pollution Authorities, SFT, will implement a "zero harmful discharge" policy for the Norwegian sector of the North Sea. This will focus the search for more environmental friendly treatment products. On shore bioreactor: E. coli will not survive during reservoir conditions. However, the bacteria can be used in bioreactors to produce the desired product. Bioreactor production of PolyAsp might prove to be economically feasible.

23

Down hole: Work is in progress, introducing the corresponding synthetic gene construct into a vector, compatible with extremophiles. This is a first step towards down hole application 5. NEW APPLICATION OF EXTREMOPHILES IN OIL RELATED INDUSTRY 5.1. Bioprospecting of the gene pools Oil quality may be linked to microbial growth in oil reservoirs. This has been substantiated in fields with biodegraded heavy oils. Although biogenic reservoir processes seem to be slow [21] oil is utilized as carbon source and water as a source of inorganic nutrition. The reservoir microbes, acting at high temperature and pressure, have preferences or tolerance for these extreme conditions. Enzymes from extremophilic microbes may be tailor-made for industrial systems run at high temperature and pressures, i.e. systems in which enzymes from mesophilic microbes will not function. Such enzyme systems may be utilized inside the reservoir, in bioreactors, in waste handling or in energy processes. DNA technology may be used to link appropriate enzyme systems to microbes growing at relevant temperature and/or pressure conditions. An immediate prerequisite for the utilization of microbes and enzymes from the hot oil reservoirs will be to perform surveys of the genetic pools within the reservoirs. The knowledge about microbial species in these environments is constantly increasing, but the understanding of the interactions between the microbes and their environments is still limited. It will be essential to characterize active enzyme systems in the reservoirs. Complete genomes have been sequenced for several microbes detected in oil reservoirs, including Archaeoglobus fulgidus and Methanococcus jannaschi [59-60]. Recent progress in molecular microbial ecology has revealed that traditional culturing methods fail to represent microbial diversity in nature, since only a small proportion of viable microorganisms in most environmental samples are recovered by culturing techniques. Methods to investigate the full extent of microbial genomes in nature include the use of BAC (bacterial artificial chromosome) vectors or random shotgun sequencing techniques [61-62]. These approaches also have potentials for characterization of the complete genomic structures in oil reservoirs. Besides explaining microbial structure-function relationships in the reservoirs, the genomic libraries may be excellent tools for prospecting of novel biocatalysts [63]. 5.2. Thermophilic/extremophilic enzymes New application of extremophilic/thermophilic enzyme systems: The concept is to investigate the commercial utilization of thermophiles. These organisms have enzyme systems working at high temperature, and often at high pressure.

24

Such enzymes are tailor-made as catalysts in industrial processes performed at extreme conditions. Enzymes from most mesophilic microbes will not function as the high temperature will denaturate their proteins (e.g. the enzymes). Such enzyme systems will work placed either inside the reservoir, in bioreactors, in waste handling or in energy processes. 5.3. Future prospective The petroleum biotechnology is still in its infancy and will play an increasingly important role in the future industrial processes. Within the oil company it will have a substantial economical impact throughout the value chain. This will influence on the development of: New techniques in exploration and production Biological well treatments (Preventive medication) Biocatalytic up-grading of oil New application of extremophiles Acknowledgement The authors would like to thank Statoil for the permission to publish this book chapter and for their support in the "Applied Biotechnology" program. Many thanks to our special adviser, Hakon Rueslatten, for valuable help and discussions. REFERENCES [1]

M.J. Mclnerney and K.L. Sublette, In: C.J Hurst, G.R Knudsen, MJ Mclnerney, L.D Stetzenbach, and M.V. Walter (eds.), Manual of Environmental Microbiology, ASM Press, Washington, D.C., 1997, pp. 600-607 [2] M. Nemati, TJ. Mazutinec, G.E. Jenneman, G. Voordouw, J. Ind. Microbiol. Biotechnol. 26 (2001) 350. [3] D. Lee, D. Lowe and P. Grant, 47th Annu. Cim. Petrol. Soc. Tech. Mtg. (Calgery), Vol.2, Pap. no. cim. 96- 09. [4] R.S. Bryant, SPE/DOE - 35356 1 (1996) 127. [5] J.T. Rosnes, T. Torsvik, and T. Lien, Appl. Environ. Microbiol. 57 (1991) 2302. [6] K.O. Stetter, R. Hubert, E. Blochl, M. Kurr, R.D. Eden, M. Fielder, H. Cash and I. Vance, NATURE 365 (1993) 743. [7] S. L'Haridon, A.L. Reysenbach, P. Glenat, P. Prieur and C. Jeanthon, NATURE. 377 (1995)223. [8] Y. Takahata, M. Nishijima, T. Hoaki and T. Marauyama, Appl. Environ. Microbiol. 66 (2000) 73. [9] VJ. Orphan, L.T. Taylor, D. Hafenbradl and E.F. DeLong, Appl. Environ. Microbiol. 66 (2000) 700. [10] I. Spark, I. Patey, B. Duncan, A. Hamilton, C. Devine and C. McGovern-Traa, Clay Minerals 35 (2000) 5.

25

[11] C. Tardy-Jackuenod, P. Caumette, R. Matheron, C. Lanau, O. Arnauld and M. Magot, Can J. Microbiol. 42 (1996) 259. [12] J. Beeder, R.K. Nilsen, T. Thorstensen and T. Torsvik, Appl. Environ. Microbiol. 62 (1996)3551. [13] T.K. NG, P.J. Weimer and L.J. Gawel, Geomicrobiol. J. 7 (1989) 185. [14] R. Nilsen and T. Torsvik, Appl. Environ. Microbiol. 62 (1996) 728. [15] G.S. Grassia, K.M. McLean, P. Glenat, J. Bauld and A.J. Sheehey, FEMS Microbiol. Ecol. 21 (1996)47. [16] O.G. Brakstad, S. Ramstad, G. Eidsaa, B.M. Hustad and H.K. Kotlar, 99th Annual Meeting of the American Society for Microbiology, Chicago, (1999) [17] O.G. Brakstad, K. Bonaunet and H.K. Kotlar, Proceedings to the Oil & Gas Science and Technology Conference on Microbiology of hydrocarbons: state of the art and perspectives, Paris, June 6-7 (2002). [18] G. Voordouw, S.M. Armstrong, M.F. Reimer, B. Fouts, A.J. Telang, Y. Shen and D. Gevertz, Appl. Environ. Microbiol. 62 (1996) 1623. [19] V.J. Orphan, S.K. Goffredi, E.F. Delonga and J.R. Boles, Geomicrobiol. J. 20 (2003) 295. [20] K. Takai, M.R. Mormile, J.P. McKinley, F.J. Brockman, W.E. Holben, W.P. Kovacik and J.K. Fredrickson, Environ. Microbiol. 5 (2003) 309. [21] I. M. Head, D.M. Jones and S.R.Larter, NATURE 426 (2003) 344. [22] M. Magot, B. Ollivier and B. Patel, Antonie van Leeuwenhoek 77 (2000) 103. [23] R.K. Nilsen, J. Beeder, T. Thorstenson and T. Torsvik, Appl. Environment. Microbiol. 62 (5) (1996) 1793. [24] API, (1975) [25] Gawel et al, Eur. Pat. Appl., No. 0272916 Al (1987) [26] M. Uyttendaele, R. Schukking, B. Vangemen and J.Debevere, J. Appl. Bacteriol. 77 (1994) 694. [27] A. Gulliksen, L. Solli, F. Karlsen, H. Rogne, E. Hovig, T. Nordstrom and R.Sirevag, Anal. Chem. 76 (2004) 9. [28] A. Spiro, M. Lowe and D. Brown, Appl. Envir. Microbiol. 66 (2000) 4258. [29] J.C. Cho and J.M. Tiedje, Appl. Environ. Microbiol. 68 (2002) 1425. [30] D.Y. Guschin, B.K. Mobarry, D. Proudnikov, D.A. Stahl, B.E. Rittmann and A.D. Mirzabekov, Appl. Envir. Microbiol. 63 (1997) 2397. [31] W.T. Liu, A.D. Mirzabekov and D.A. Stahl, Environ. Microbiol. 3 (2001) 619. [32] S. El Fantroussi, H. Urakawa, A.E. Bernhard, J.J. Kelly, P.A. Noble, H. Smidt, G.M. Yershov and D.A. Stahl, Appl. Envir. Microbiol. 69 (2003) 2377. [33] L.Wu, D.K. Thompson, G. Li, R.A. Hurt, J.M. Tiedje and J. Zhou, Appl. Envir. Microbiol. 67 (2001) 5780 [34] G. Taroncher-Oldenburg, E.M. Griner, C.A. Francis and B.B. Ward, Appl. Envir. Microbiol. 69(2003)1159. [35] H.M. Alvarez, O.H. Pucci and A. Steinbuchel, Appl. Microbiol. Biotechnol. 47 (1997) 132. [36] F. Kawai, M. Shibata, S. Yokoyama, S. Maeda, K. Tada and S. Hayashi, Macromolecular Symposia 144 (1999) 73. [37] A.R. Massengale, R.A. Ollar, S.J. Giordano, M.S. Felder and S.C. Aronoff, Diagn. Microbiol. Infect. Dis. 35 (1999) 177. [38] F. McKenna, K.A. El-Tarabily, S. Petrie, C. Chen and B. Dell, Lett. Appl. Microbiol. 35 (2002) 107. [39] S. Reiser and C. Somerville, J. Bacteriol. 179 (1997) 2969.

26

[40] T. Ishige, A. Tani, Y. Sakai and N. Kato, Appl. Environ. Microbiol. 66 (2000) 3481. [41] Z. Wang, M. Fingas, E.H. Owens, L. Sigouin and C.E. Brown, J. Chromatogr. A. 926 (2001) 275. [42] D.R. Kadavy, B. Plantz, C.A. Shaw, J. Myatt, T.A. Kokjohn and K.W. Nickerson, Appl. Environ. Microbiol. 65 (1999) 1477. [43] M. Hofrichter, F. Bublitz and W. Fritsche, Fuel. Proc. Technol. 52 (1997) 43. [44] A.Y. Zekri and R. El-Mehaideb, J. Petrol. Sci. Eng. 37 (2003) 123. [45] T.L. Potter and B. Duval, Environ. Sci. Technol. 35(2001)76. [46] W. Dott and D. Schoenen, Zentralbl. Bakteriol. Mikrobiol. Hyg. 180 (1985) 436. [47] K.A. Powell, S.W. Ramer, S.B. Del Cardayre, W.P. Stemmer, M.B. Tobin, P.F. Longchamp and G.W. Huisman, Angew. Chem. Int. Ed. Engl. 40 (2001) 3948. [48] D.L. Gutnick and H. Bach, Appl. Microbiol. Biotechnol. 54 (2000) 451. [49] A.C. Greene, B.K. Patel and A.J. Sheehy, Int. J. Syst. Bacteriol. 47 (1997) 505. [50] M.R. Bruins, S. Kapil and F.W. Oehme, Ecotoxicol. Environ. Saf. 45 (2000) 198. [51] T. Bugg, J.M. Foght, M.A. Pickard, M.R. Gray, Appl. Environm. Microbiol. 66 (12) (2000) 5387. [52] D.T. Gibson and R.E. Parals, Curr. Opinion in Biotechnol. 11 (2000) 236. [53] D.R. Boyd, N.D. Sharma and C.C.R. Allen, Curr. Opinion in Biotechnol. 12 (2001) 564. [54] P.M. Fedorak, M.A. Pickard, M.R. Gray and J.M. Foght, Prepr. Symp. Am. Chem. Soc. Div. Fuel Chem. 43 (3) (1998) 515. [55] J.M. Foght, P.M. Fedorak, M.A. Pickard and M.R. Gray, 48. Ann. Tech. Meet. Petr. Soc. (Calgary), Paper 97-13:1-9. (1997) [56] H.K. Kotlar, K. Rasmussen, K. Grande, M. Ramstad, S. Markussen, A. Winnberg, S. Zotchev and M. Gimmestad, In proceedings of 225th ACS National Meeting, New Orleans, LA, March 23-27 2003. [57] The University of Minnesota Biocatalysis/Biodegradation Database http://umbbd.ahc.umn.edu/naph/naph_image_map.html [58] B.L. McFarland, DJ. Boron, W. Deever, J.A. Meyer, A.R. Johnson and R.M. Atlas, Crit. Rev. Microbiol. 24 (2) (1998) 99. [59] H.P. Klenk, R.A. Clayton, J.F. Tomb, O. White, K.E. Nelson, K.A. Ketchum, RJ. Dodson, M. Gwinn, E.K. Hickey, J.D. Peterson, D.L. Richardson, A.R. Kerlavage, D.E. Graham, N.C. Kyrpides, R.D. Fleischmann, J. Quackenbush, N.H. Lee, G.G. Sutton, S. Gill, E.F. Kirkness, B.A. Dougherty, K. McKenney, M.D. Adams, B. Loftus, S. Peterson, C.I. Reich, L.K. McNeil, J.H. Badger, A. Glodek, L.X. Zhou, R. Overbeek, J.D. Gocayne, J.F. Weidman, L. McDonald, T. Utterback, M.D. Cotton, T. Spriggs, P. Artiach, B.P. Kaine, S.M. Sykes, P.W. Sadow, K.P. DAndrea, C. Bowman, C. Fujii, S.A. Garland, T.M. Mason, G.J. Olsen, CM. Fraser, H.O. Smith, C.R. Woese and J.C. Venter, NATURE 390 (1997) 364. [60] C.J. Bult, O. White, G.J. Olsen, L.X. Zhou, R.D. Fleischmann, G.G. Sutton, J.A. Blake, L.M. FitzGerald, R.A. Clayton, J.D. Gocayne, A.R. Kerlavage, B.A. Dougherty, J.F. Tomb, M.D. Adams, C.I. Reich, R. Overbeek, E.F. Kirkness, K.G. Weinstock, J.M. Merrick, A. Glodek, J.L. Scott, N.S.M. Geoghagen, J.F. Weidman, J.L. Fuhrmann, D. Nguyen, T.R. Utterback, J.M. Kelley, J.D. Peterson, P.W. Sadow, M.C. Hanna, M.D. Cotton, K.M. Roberts, M.A. Hurst, B.P. Kaine, M. Borodovsky, H.P. Klenk, CM. Fraser, H.O. Smith, C.R. Woese and J.C. Venter, SCIENCE 273 (1996) 1058. [61] M.R. Rondon, P.R. August, A.D. Bettermann, S.F. Brady, T.H. Grossman, M.R. Liles, K.A. Loiacono, B.A. Lynch, LA. MacNeil, C. Minor, CL. Tiong, M. Gilman, M.S. Osburne, J. Clardy, J. Handelsman and R.M. Goodman, Appl. Envir. Microbiol. 66 (2000)2541.

27

[62] G.W. Tyson, J. Chapman, P. Hugenholz, E.E. Allen, RJ. Ram., P.M. Richardson, V.V. Solovyev, E.M. Rubin, D.S. Rokshar and J.F. Banfield, NATURE 428 (2004) 37. [63] S. Voget, C. Leggewie, A. Uesbeck, C. Raasch, K.-E. Jaeger and W.R. Streit, Appl. Envir. Microbiol. 69 (2003) 6235.


Studies in Surface Science and Catalysis 151 R. Vazquez-Duhalt and R. Quintero-Ramirez (Editors) ©2004 Published by ElsevierB.V.

29

Chapter 2

Petroleum biorefining: the selective removal of sulfur, nitrogen, and metals J.J. Kilbane IIa and S. Le Borgne" a

Gas Technology Institute, 1700 S. Mt. Prospect Rd., Des Plaines, Illinois 60018

b

Instituto Mexicano del Petroleo, Eje Central Lazaro Cardenas 152, Col. San Bartolo Atepehuacan, 07730 Mexico D.F., Mexico

1. INTRODUCTION The quality of petroleum is progressively deteriorating as the highest quality petroleum deposits are preferentially produced. Consequently the concern about the concentrations of compounds/contaminants such as sulfur, nitrogen, and metals in petroleum will intensify. These contaminants not only contribute to environmental pollution resulting from the combustion of petroleum, but also interfere with the processing of petroleum by poisoning catalysts and contributing to corrosion. The selective removal of contaminants from petroleum while retaining the fuel energetic value is a difficult technical challenge. New processes are needed and bioprocesses are an option. Existing thermo-chemical processes, such as hydrodesulfurization, can efficiently remove much of the sulfur from petroleum but the selective removal of sulfur from compounds such as dibenzothiophene, the removal of organically bound nitrogen, and the removal of metals cannot be efficiently accomplished using currently available technologies. The specificity of biochemical reactions far exceeds that of chemical reactions. The selective removal of sulfur, nitrogen, and metals from petroleum by biochemical reactions performed by microorganisms and/or enzymes has been demonstrated. However, further research is needed before biorefining technology can be commercialized. This chapter reviews the status of biorefining and discusses topics requiring further research.

30

The geochemical conversion of organic matter into petroleum is a slow and inefficient process. It is estimated that 23.5 tonnes of plant material/biomass are required to form a single liter of petroleum during geological periods of time [1]. Moreover, the current rate of energy consumption is 400 times greater than the capacity of the planet to produce biomass. It behooves us to utilize our fossil fuel legacy as efficiently as possible while avoiding environmental damage. Environmental regulations limit the amount of sulfur oxides emitted to the atmosphere by the combustion of fossil fuels by regulating the concentration of sulfur in these fuels. In particular, transportation fuels are severely regulated. For example, the permissible concentration of sulfur in diesel has been progressively decreased over the past decade from 500 ppm to 10 to 15 ppm [2]. Environmental regulations do not specify concentration limits for nitrogen and metals in transportation fuels, but such regulations may be forthcoming as these compounds inhibit the catalytic converters used to cleanse exhaust gases from vehicles. The future use of petroleum products to power fuel cells may provide a further impetus to decrease the sulfur, nitrogen, and metal content of petroleum derived fuels as reforming catalysts and fuel cell electrodes are sensitive to impurities at ppm levels [3]. Sulfur, nitrogen, and heavy metals in petroleum not only contribute to environmental pollution when oil is burned, they also decrease the efficiency of catalytic cracking and hydrotreating processes in oil refineries by poisoning the involved catalysts [4-6]. Although the focus on biologically upgrading petroleum has mainly been on sulfur, the ability of certain microorganisms to metabolize organonitrogen compounds may be particularly important because organonitrogen compounds are associated with the majority of metals in petroleum [4, 7]. Thus, by metabolizing nitrogen-containing compounds, it may be possible to simultaneously achieve the selective removal of nitrogen and heavy metals, mainly nickel and vanadium, from petroleum. Furthermore, nitrogen compounds contribute to the instability of petroleum byproducts [7-9]. The selective removal of sulfur, nitrogen and heavy metals from oil would be highly desirable, but the existing physicochemical processes are not completely effective and moreover they are not environmentally friendly. 1.1 Composition of crude oils The sulfur content of crude oil can vary from 0.03 to 7.89 wt% [2, 4, 7, 10]. Sulfur is present in crude oil almost exclusively as organic sulfur. While there are multiple types of organosulfur compounds such as mercaptans, sulfides, disulfides and thiophenes, the most abundant form of sulfur in petroleum is usually thiophenic [7, 10]. Thiophenic sulfur often comprises 50% to 95% of the sulfur in crude oil and derived fractions, and alkylated derivatives

31

of dibenzothiophene are the most common organosulfur compounds typically found in crude oil and fractions used to produce diesel. Alkylated derivatives of benzothiophene are the most abundant organosulfur compounds in gasoline [11]. The sulfur, nitrogen, and metal content is preferentially associated with the higher molecular weight components of crude oils and, consequently, heavy crude oils typically have higher sulfur, nitrogen, and metals content than light crude oils. Similarly, when crude oil is refined, the sulfur, nitrogen, and metals concentrate into the high molecular weight fractions. Nitrogen compounds typically found in crude oil consist of heterocycles such as quinoline and carbazole, which are examples of basic and non-basic organonitrogen compounds, respectively [4, 8, 9]. The total nitrogen content of crude oil is typically about 0.3%, but it can be as high as 5%. Basic organonitrogen compounds in petroleum usually comprise 25 to 30% of the total nitrogen and include compounds such as quinoline and pyridine [12]. The majority of basic organonitrogen compounds in crude oils are actually alkylated derivatives of quinoline and pyridine. Quinolines and related nitrogen heterocycles have relatively high solubility in water and can be significant environmental contaminants. Quinoline and related compounds have been shown to be hepatocarginogens in rodents, and mutagens in toxicity tests [4, 1319]. Non-basic organonitrogen compounds typically comprise 70 to 75% of the total nitrogen in crude oils and alkylated derivatives of carbazole are the most typically found non-basic organonitrogen compounds. Basic compounds like quinoline are generally more reactive in the inactivation of catalysts than nonbasic compounds. However, non-basic compounds can potentially be converted to basic compounds during the refining/catalytic cracking process. The inactivation of catalytic cracking and hydrotreating catalysts decreases the efficiency of operation of a refinery and results in lower yields of transportation fuels [5]. The heavy metals found in greatest abundance in crude oil are nickel and vanadium, which are both potent inhibitors of refining catalysts. These heavy metals are typically associated with nitrogen compounds [20]. 1.2 The need for new technologies to upgrade crude is intensifying The standard practices used within the petroleum industry over decades are not capable of treating heavy oils and residuum, so the need for new technologies is intensifying (see chapter 4). The sulfur, nitrogen, and heavy metal content as well as the average molecular weight of available crude oils in the U. S. and in the world has increased significantly in recent years and will continue to increase due to the progressive depletion of light crude oils reservoirs [2, 6, 11, 21-24]. Multiple factors including physics, chemistry, environmental concerns and market forces contribute to this trend of increasing the heteroatom and metal content as well as the molecular weight of presently

32

available crude oils. However, the bottom line is that light crude oil is more readily recovered, and more readily processed/refined, than heavy oil. Consequently, deposits of light, low heteroatom content oil are preferentially brought into full production while known deposits of heavy/high heteroatom content petroleum are produced at less than full capacity or may even be idle. Moreover, as a deposit of light oil is harvested, the lighter fractions are preferentially removed such that after primary or secondary production mainly heavy oil remains. Because of this irreversible trend, the time when available crude oil is predominantly or exclusively heavy with a high heteroatom and heavy metals content is not far off. The chief concern is for the sulfur content of petroleum, but the nitrogen and metal content of petroleum is also of concern due to environmental, processing and corrosion concerns [23]. In North America, over 3 trillion barrels of known petroleum reserves are largely untapped or underutilized because of their high sulfur content and viscosity [22]. It is well known in petroleum chemistry that sulfur and heavy metals are preferentially associated with the higher molecular weight fractions of oil [4, 7]. So, not only is light oil easier to produce because of its physical properties, but it also contains significantly less undesirable impurities in comparison with heavy oils. Sulfur, nitrogen and heavy metal impurities are of great environmental concern since they originate acid rain as a consequence of sulfur and nitrogen oxides emissions from the combustion of petroleum derived fuels, and potential health effects due to high concentrations of heavy metals on combustion ashes [2, 10, 24, 25]. Some sulfur and nitrogen heterocycles are suspected carcinogens [8, 12] and sulfur compounds in oil have been implicated in the corrosion of pipelines and refinery equipment [7, 24, 26]. Heavy metals content, mainly nickel and vanadium, contributes to the poisoning of catalysts used in hydrodesulfurization or in catalytic cracking [5, 20, 23]. In addition to catalysts poisoning by heavy metals, sulfur and nitrogen in heterocyclic compounds are capable of poisoning catalysts by causing electronic modifications in Pd, Pt, Ni, and Ru compounds. The poisoning of catalysts exasperates the problems associated with the processing of heavy oils and residuum by interfering with the methods employed to reduce the heteroatom content and molecular weight, i.e. hydro treatment and cracking. The quantity of heavy oils to be processed is increasing not only due to the depletion of light oils but also to the increasing demand for cleaner transportation fuels and other low molecular weight products, so the importance of technologies capable of dealing with heavy oils and residuum has increased [11, 27]. This increased demand for lower molecular weight petroleum products seems incompatible with the use of heavy oils as primary feedstocks because of their metals, nitrogen and sulfur content that increases the production of coke and gas and accelerates catalyst deactivation [6]. But at the same time, the need to obtain greater quantities of gasoline, diesel and aviation fuels from each

33

barrel of oil demands increased attempts to further process residuum. Therefore, the petroleum industry is clearly forced to process increasing quantities of heavy crude oils and heavy residuum which have increased levels of environmental contaminants and are not efficiently treated by existing technologies. In the years 1972 through 1985, the U.S. petroleum industry spent approximately $1.4 billion in capital and operating expenditures for dealing with pollution abatement. The National Petroleum Refineries Association estimated that meeting Clean Air Act regulations, i.e. achieving a sulfur content of 0.05% for diesel fuel in 1994, had cost about $3.3 billion in capital expenditures and $1.2 billion in annual operating costs [22, 24, 28]. Similar estimations are not available for the desulfurization, denitrogenation and/or demetallation of heavy crude oils but considering that diesel fuel is far easier to desulfurize and handle than heavy oils and residuum, it is possible to predict that the costs associated with the upgrading of heavy oils and residuum would be correspondingly higher. Because of the quantity of petroleum consumed in the U.S., the price differential between high and low sulfur fuels and the opening or reopening of markets for high sulfur petroleum reserves, it is estimated that the value of an alternative desulfurization technology as biodesulfurization, which is capable of upgrading existing production, is in excess of $10 billion annually in the U.S [22]. Clearly then, there is a need for alternative technologies to deal with heavy oils and residuum and there is ample economic incentive for the development of appropriate technologies that are immune from the technical problems that limit existing technologies. Biorefining, which is defined here as the application of biotechnology to the upgrading of petroleum, is one such technology. 2. BIODESULFURIZATION Effective technologies for the treatment of heavy crude oils have been and continue to be a topic of keen interest. Hundreds of processes related to the desulfurization of heavy oils have been described in the patent literature, and the interest in such processes has steadily increased [2, 29]. Hydrodesulfurization can be used to desulfurize heavy oils and residuum, but does not lead to a significant decrease in molecular weight. The predominant method for upgrading heavy oils to decrease molecular weight and increase the yield of transportation fuels is the use of fluid catalytic cracking (FCC) [4, 30]. However, FCC cannot achieve desulfurization. Moreover hydrodesulfurization and FCC catalysts are poisoned in the process of treating heavy oils because of the presence of sulfur and nitrogen heterocycles and heavy metal contaminants [5, 31]. Therefore, a combination of technologies is needed to address both the removal of heteroatoms and the decrease in molecular weight in order to mitigate environmental problems and to get the greatest yield of value added

34

products. Of all of the chemical forms of sulfur in crude oil, the most recalcitrant to hydrodesulfurization is the thiophenic sulfur in thiophene and dibenzothiophene derivatives. Because of the abundance of alkylated dibenzothiophenes in crude oil and the recalcitrance of these compounds to hydrodesulfurization, there is a high level of interest in technologies that can effectively desulfurize dibenzothiophenes [2]. Researchers have been examining the possibility of biodesulfurization of petroleum or other fossil fuels for over 4 decades [32]. Presently, there is no commercial operation for biodesulfurization of fossil fuels, however several economic studies indicate a favorable prospect of developing such a technology [10, 24, 25, 27, 33]. Numerous microorganisms have been described in the literature that are capable of utilizing dibenzothiophene (DBT) as sole source of carbon, energy and sulfur. However, the complete degradation of organosulfur compounds is not beneficial for upgrading crude oils and derived fuels. The selective cleavage of carbon-sulfur bonds in DBT and derivatives is preferred, this way sulfur is selectively removed and the calorific value of the treated fuel remains intact. The first microorganism that was shown to be capable of selectively cleaving carbon-sulfur bonds in crude oil, coal, and a wide range of model compounds, resulting in the selective removal of sulfur and the retention of carbon and calorific value, is Rhodococcus erythropolis IGTS8 (ATCC 53968) [34]. Subsequently, numerous other bacteria capable of selectively cleaving carbon-sulfur bonds in DBT were isolated and characterized. The biochemical pathway used by these aerobic microorganisms to desulfurize DBT compounds has been termed the 4S pathway due to the progressive oxidation of sulfur that occurs through 4 steps [2, 35]. The selective removal of sulfur from DBT and from crude oil by anaerobic bacteria has also been reported. Sulfate-reducing bacteria such as Desulfovibrio desulfuricans have been shown to metabolize DBT to H2S and biphenyl [36]. The desulfurization of oil under anaerobic conditions avoids costs associated with aeration, and has the advantage of liberating sulfur as a gas. However, an anaerobic biodesulfurization process has not been developed due to low reaction rates, safety and cost concerns, and the lack of identification of specific enzymes and genes responsible for anaerobic desulfurization. Consequently, aerobic biodesulfurization has been the focus of the majority of research [2, 32]. 2.1. Substrate range of desulfurization A desulfurization competent moderate thermophile was recently isolated, Mycobacterium phlei GTIS10, that metabolizes DBT by the same pathway as R. erythropolis IGTS8 [37]. A comparison of the capabilities of these two microorganisms that have optimum growth temperatures of 30°C and 50°C

35

respectively, reveals a great deal about biodesulfurization. The range of substrates used as sulfur sources by M. phlei GTIS10, as shown in Table 1, is quite broad and essentially the same as reported for R. erythropolis IGTS8 [34]. The majority of bacterial cultures isolated based on their ability to metabolize DBT are reported to be unable to metabolize benzothiophene and/or thiophene [38-40]. While R. erythropolis IGTS8 has been reported to be unable to utilize thiophene and/or benzothiophene [38, 40], we find that M. phlei GTIS10, as well as R. erythropolis IGTS8, grew well with benzothiophene or thiophene as sole sulfur sources [37]. The utilization of benzothiophene as a substrate for the desulfurization enzymes of R. erythropolis IGTS8 has also been reported by others [41], so that substrate utilization of desulfurization competent cultures is somewhat controversial. This is probably due to the fact that adaptation by repeated subculturing in media containing thiophene, benzothiophene, or mixtures of thiophene or benzothiophene plus DBT are often required to establish good growth for cultures originally isolated based on their ability to metabolize DBT. The end product of DBT metabolism by M. phlei GTIS10 is 2-HBP, which has been reported to be bactericidal, bacteriostatic, and an inhibitor of the desulfurization enzymes [37]. Concentrations of 200 uM of 2-HBP are reported to be inhibitory and 400 uM of 2-HBP has been reported to completely prevent growth and desulfurization activity of Corynebacterium, Rhodococcus and Gordona cultures [27, 42-44]. It has been stated that bacterial strains with increased tolerance for 2-HBP are needed for a viable petroleum biodesulfurization process [19, 27]. M. phlei GTIS10 appears to be more resistant to 2-HBP than other previously reported desulfurizing bacteria. The 4S pathway for the desulfurization of DBT is shown in Fig. 1 [35, 4547]. The dszC gene encodes the dibenzothiophene monooxygenase that catalyzes the conversion of DBT into DBT sulfone (DBTSO2). The dszA gene encodes the dibenzothiophene-5,5-dioxide monooxygenase that catalyzes the conversion of DBTSO2 into 2-hydroxybiphenyl-2-sulfinate (HBPSi). The dszB gene encodes 2-hydroxybiphenyl-2-sulfinate sulfinolyase that catalyzes the conversion of HBPSi into 2-hydroxybiphenyl (2-HBP) and sulfite [43, 48]. The dszABC genes are transcribed as an operon found on a large plasmid, pSOX [49, 50]. An unlinked fourth gene, the dszD gene encoding a NADH-FMN oxidoreductase, is an accessory component of the desulfurization pathway and allows the regeneration of the cofactors needed for the monooxygenase reactions catalyzed by DszC and DszA [51, 52]. The enzymology of the 4S desulfurization pathway has been firmly established using purified enzymes from several desulfurization competent bacterial species and from the results of genetic analyses.

36

Table 1. Range of organosulfur substrates used as sole source of sulfur for growth byMp«ezGTIS10

37

An alternative desulfurization pathway has been found in Gordona sp. strain 213E [38] that converts benzothiophene to 2-(2'-hydroxyphenyl) ethan-1al but is unable to metabolize DBT. The limited substrate range of this pathway and a comparative lack of biochemical and genetic information have resulted in the majority of research concerning biodesulfurization being focused on the 4S pathway [2].

Fig. 1. The 4S metabolic pathway for DBT desulfurization. DszC is the DBT monooxygenase, DszA the DBT sulfone monooxygenase, DszB the HPBSi desulfmase and DszD is a flavin reductase. I, DBT; II, DBT sulfoxide; III, DBT sulfone; IV, hydroxyphenylbenzenesulfmate; V, 2-hydroxybiphenyl.

38

Table 2. Concentrations of metabolites of dibenzothiophene produced by M. phlei GTIS10 and R. erythropolis IGTS8

1 = All concentrations of metabolites are expressed as |j.g/ml of the ethyl acetate extract (1 ml total) derived from each culture grown with 1,440 (ig DBT as the sole sulfur source. Dihydroxybiphenyl #1 and #2 have identical molecular formulas, but could not be assigned specific molecular structures based on data available.

2.2. Range of metabolites produced by R. erythropolis IGTS8 versus M. phlei GTIS10 from DBT Because both R. erythropolis IGTS8 and M. phlei GTIS10 metabolize DBT via the 4S pathway, but have optimum temperatures of 30°C and 50°C respectively, the metabolites produced from DBT by these two cultures were compared. Of all of the DBT metabolized by R. erythropolis IGTS8, 95% was converted into 2-HBP; whereas, for M. phlei GTIS10 only 65% of the DBT metabolized was converted to 2-HBP [37]. The lack of quantitative conversion of DBT to 2-HBP has been noted in reports concerning other desulfurization cultures where as little as 54% mole of DBT metabolized could be accounted for as 2-HBP [53]. M. phlei GTIS10 produced greater levels of dibenzothiophene sulfoxide and dihydroxybiphenyls than did R. erythropolis IGTS8. A significant difference between data for M. phlei GTIS10 and R. erythropolis IGTS8 are the results obtained for dihydroxybiphenyls (Table 2). About 3.3% of the DBT metabolized by M. phlei GTIS10 was converted to dihydroxybiphenyls, while no dihydroxybiphenyls were observed in the metabolites produced by R. erythropolis IGTS8. While in the experiment reported in Table 2 no dihydroxybiphenyls were observed as metabolites produced by R. erythropolis

39

IGTS8, occasionally trace amounts of dihydroxybiphenyls were observed as metabolites of DBT produced by R. erythropolis IGTS8 [35]. This difference in the relative abundance of DBT metabolites produced by R. erythropolis IGTS8 versus M. phlei GTISIO may be due to differences in the thermotolerance of the desulfurization enzymes in these two cultures. To further explore the thermostability of the desulfurization enzymes, the R. erythropolis IGTS8 dszA, dszB and dszC genes were cloned individually and overexpressed in E. coli to determine the thermostability of corresponding enzymes [37]. Each desulfurization gene was cloned into the E. coli expression vector pQE80 and the DszA, B, and C proteins containing polyhistidine residues at their N-termini were produced and purified by affinity chromatography. Polyacrylamide-SDS gel electrophoresis was performed to confirm the presence of proteins of the correct molecular weight in the cell lysates. Samples containing the DszA, DszB and DszC proteins were pre-incubated at 30°, 37°, 45°, 52°, 60°, 65°, and 72°C for 30, 60, and 120 minutes. The enzyme solutions were assayed by adding the appropriate substrate and/or cofactors, incubated at 30°C for an hour and then the amount of product formed was determined by HPLC analysis. The results of these analyses indicate that DszB activity (conversion of HBPSi into 2-HBP) is not inhibited by pre-incubation at 30° or 37°C, but little or no activity is seen when samples are pre-incubated at temperatures of 45°C or higher even with exposure times as brief as 30 minutes. The half-life of dibenzothiophene monooxygenase, DszC, was 1 hour at 45°C and the half-life of DszA was 1 hour at 60°C. We examined the M. phlei GTISIO FMN oxidoreductase (DszD) for thermal inactivation and it was found to function with little inactivation up to 45 °C but was progressively inactivated by exposure to higher temperatures [37]. Data obtained here regarding the thermostability of desulfurization enzymes derived from M. phlei GTISIO are in agreement with results obtained for the purified desulfurization enzymes isolated from Rhodococcus cultures [43, 48, 54-56]. The DszC enzyme from R. erythropolis Dl was inactivated by preincubation at 45°C for 30 minutes and DszA purified from R. erythropolis Dl was inactivated by pre-incubation at 60°C for 30 min. Since DszB catalyzes the last step in the desulfurization pathway and is responsible for the release of sulfur from DBT it must be functional in order to allow cultures to grow and utilize DBT as sole sulfur source and it is required to allow the production of 2HBP. Therefore, it is surprising that the DszB enzyme appears to be thermally inactivated in vitro by exposure to temperatures as low as 45°C yet some activity is detected in whole cells at temperatures of 57°C [37] . While the DszB enzyme may be the most thermolabile enzyme in the desulfurization pathway, thermal inactivation is not instantaneous and it apparently has enough residual activity even at 57°C to allow the accumulation of 2-HBP to be detected. The more rapid thermal inactivation of purified desulfurization enzymes as

40

compared with desulfurization activity detected in whole cells requires further investigation. Table 2 shows that dihydroxybiphenyls were observed in the supernatant of M. phlei GTIS10 cultures grown at 45°C. Oldfield et al [46] describes a minor desulfurization pathway in which the dibenzothiophene-5,5dioxide monooxygenase (dszA product) produces dihydroxybiphenyls from biphenylene sulfone. Biphenylene sulfone can be formed from non-enzymatic oxidation of 2-hydroxybiphenyl-2-sulfmate (HBPSi). Perhaps at higher temperatures, the DszB enzyme is inactive and the minor desulfurization pathway catalyzed by DszA is active to provide sulfur to the cells. The dibenzothiophene monooxygenase encoded by dszC is responsible for the conversion of DBT into dibenzothiophene sulfoxide and then to dibenzothiophene sulfone. The elevated levels of dibenzothiophene sulfoxide detected in M. phlei GTIS10, as compared with R. erythropolis IGTS8 (Table 2), might suggest that DszC is temperature sensitive and is less active at 45 °C than at 30°C. Also the faster growth of M. phlei GTIS10 with DBTSO2 versus DBT suggests that the DszC activity may be thermolabile. However, the results of in vitro experiments do not support this conclusion as DszC exhibits activity at 45 °C while DszB is the most thermolabile desulfurization enzyme [37]. The DszC enzyme requires other cofactors/substrates for proper functioning: FMNH2 and oxygen. The FMN oxidoreductase encoded by dszD is responsible for providing FMNH2 to the reactions catalyzed by DszC [51]. The accumulation of DBTSO in the supernatant of the M. phlei GTIS 10 45°C growing cell experiment might indicate that the FMN oxidoreductase DszD was significantly inactivated in vivo at 45 °C and that the DszC catalyzed reaction did not go to completion due to a lack of cofactors. However, this explanation is inconsistent with the fact that DszD activity in M. phlei GTIS 10 extracts showed very little inactivation at 45°C. Moreover, the half-life of the DszD enzyme from R. erythropolis Dl was determined to be about 17 minutes at 50°C, slight activity was detectable even after incubation at 72°C for 60 minutes [51]. These data demonstrate that the thermostability of enzymes determined in vitro is not necessarily a good predictor of the functional range of an enzyme in vivo, and that the same operon can yield metabolic pathways with different rate limiting steps and different yields of metabolites in different hosts. 2.3. Comparison of the desulfurization activity of R. erythropolis IGTS8

and Paenibadllus sp. All-2 The isolation of the first desulfurization-competent thermophilic bacteria, Paenibadllus sp. All-2, was reported in 1997 [57]. However, the specific desulfurization activity of this culture (0.008 uM 2-HBP/hr/g dry cell weight {DCW}) was low in comparison to previously characterized mesophilic cultures (0.083 to 1.23 uM 2-HBP/hr/g DCW) [11, 27, 39, 52, 57-62].

41

The entire thermophilic desulfurization operon (tdsABC) of Paenibacillus sp. All-2, which shows substantial homology to the dsz ABC operon of R. erythropolis IGTS8, was cloned into E. coli and expressed with a maximum activity of 0.155 uM 2-HBP/min/g DCW [61, 63]. The tds genes showed 61% to 73% homology to the corresponding dsz genes and the Tds enzymes showed 51.5% to 64.5% homology to the corresponding Dsz enzymes [61]. Methyl, ethyl and propyl DBT derivatives were metabolized differently by Paenibacillus Al 1-2 compared to R. erythropolis KA2-5-1 (which is essentially identical with R. erythropolis IGTS8) [64]. These asymmetric substrates can yield two desulfurized products depending upon which C-S bond is cleaved first by DszA/TdsA. The ratios of the isomers produced by each strain were influenced both by the position and size of the alkyl groups, but the two strains generally showed opposite preferences for the reaction pathway. Desulfurization activity generally decreased with increasing size of the alkyl substitutent in both Paenibacillus and Rhodococcus. However, the substrate range of the two desulfurization pathways differ [64]. Thiophenic sulfur compounds that are recalcitrant to hydrodesulfurization include 4,6-dialkyl DBTs and 7-alkyl benzothiophenes. The Tds enzymes showed more tolerance for substituted DBTs than the Dsz enzymes and the Dsz enzymes did not metabolize 7-alkyl BTs whereas the Tds enzymes did. The substrate range of cell-free systems/lysates was broader than for the whole cell biodesulfurization catalysts, which suggests that the transport or the bioavailability of organosulfur compounds was limiting in both desulfurization catalysts [65]. While the enzymatic rates of the Tds enzymes were substantially less than with the Dsz enzymes, the different substrate range and thermotolerance of Tds versus Dsz enzymes make them valuable resources for possible future directed evolution experiments aimed at developing improved desulfurization enzymes. 2.4. Role of the desulfurization trait in nature The cleavage of carbon-sulfur bonds in molecules such as DBT liberates sulfur making it available as a nutrient to support the growth of bacteria. The widespread occurrence of desulfurization-competent bacteria in samples obtained from diverse environments and geographic locations indicates that the ability to obtain sulfur from organic substrates is an important and fairly common survival strategy for some bacterial species [2, 66]. Apparently, microenvironments exist, even in soils where inorganic sulfur is relatively abundant, where sulfur is a growth limiting nutrient. Clearly then, there is a selective advantage in many natural environments for bacteria that can utilize organic compounds such as DBT to obtain sulfur. The importance of the 4S pathway to the survival of some bacteria in nature is also illustrated by the fact that the dsz genes are found on conjugal plasmids and located in the proximity of insertion sequences [49, 67]. While laboratory data demonstrating the

42

conjugal transfer of plasmids containing the dsz genes or the transposition of dsz genes is sparse, the distribution of dsz genes in bacterial cultures strongly support the hypothesis that these genes are commonly subjected to horizontal transfer in nature. Indeed the DNA sequences of dsz genes from numerous bacterial cultures isolated in geographically distinct locations have been found to be nearly identical. Various Rhodococcus [68], Mycobacterium [37], Gordona [59], Corynebacterium [69], Arthrobacter [70], Enterobacter [39], Stenotrophomonas [39], Klebsiella [39], Bacillus [71], and Nocardia [72] species have been isolated that possess dsz gene sequences that are identical or highly homologous to the DNA sequence of the dsz gene of R. erythropolis IGTS8. However, some variation in the sequences of the dsz genes has been observed. The dsz genes of the moderate thermophile Paenibacillus sp. A 11-2 and Nocardia asteroides are only 52-65% and 89% homologous to R. erythropolis IGTS8, respectively [61, 73]. Moreover, PCR amplification of dsz genes from soil samples revealed relatively few variations in dsz gene sequences, with the majority of variations found in dszA, and even then homology to the R. erythropolis IGTS8 dszA sequence was 95% or more [66]. It is interesting to note that while several bacterial genera apparently participate in horizontal transfer of dsz genes in nature, and laboratory studies demonstrate that dsz genes and enzymes function well in Pseudomonas and E. coli strains, a naturally occurring desulfurization-competent Pseudomonas sp. is rarely encountered [74] and a desulfurization competent E. coli isolate has never been reported [2]. The reasons for the restricted range of distribution of dsz genes in nature are currently unknown, but one factor may be the ability of bacterial species to withstand exposure to substrates such as petroleum. Laboratory studies indicated that Pseudomonas sp. containing dsz genes could efficiently metabolize DBT in aqueous culture or DBT added in a solvent such as hexadecane. However, the ability of these same Pseudomonas cultures to metabolize DBT in diesel oil or other petroleum product is much reduced [75, 76]. Naturally occurring desulfurization-competent bacterial cultures are almost exclusively gram positive or gram variable and it may be that the cell wall/membrane structure of gram negative bacterial species is less able to tolerate exposure to petroleum compounds and solvents. There are many thousands of gram positive and gram variable bacterial species, yet the observed occurrence of dsz genes in naturally occurring desulfurization-competent bacterial isolates is restricted to only a few species. Clearly, more remains to be learned about the role dsz genes play in microbial ecology and the functioning of desulfurization enzymes in different bacterial hosts [77]. Other topics regarding biodesulfurization that are not well understood are the access of desulfurization enzymes to insoluble and high molecular weight substrates, and the mechanism by which the sulfur liberated from organosulfur substrates by the desulfurization enzymes is subsequently incorporated into

43

biomass. When R. erythropolis IGTS8 was first isolated, it was obtained from an enrichment culture growing in a defined mineral salts medium devoid of inorganic sulfur [78, 79]. All essential nutrients were present in abundance with the exception of sulfur, which was supplied in the form of coal or DBT creating an environment where any bacterial species that could utilize organically bound sulfur had a strong selective advantage. The mixed culture that grew with DBT as the sole source of sulfur was streaked onto a variety of agar plates allowing pure cultures to be obtained from each of the types of colonies present. Then each pure culture was tested individually to determine if it could utilize DBT as a sole source of sulfur for growth. It soon became clear that none of the pure cultures most readily isolated from the desulfurization-competent mixed culture were capable of utilizing DBT as a sole sulfur source. Perseverance in investigating this desulfurizationcompetent mixed culture eventually led to the isolation of a relatively slow growing pure culture that was demonstrated to utilize DBT as a sole source of sulfur, and this culture was subsequently identified as R. erythropolis IGTS8 [34]. R. erythropolis IGTS8 was present at low abundance in the original desulfurization-competent mixed culture and even when pure cultures of R. erythropolis IGTS8 and a desulfurization-deficient, but faster growing, bacterial culture such as Enterobacter cloacae, were combined in various ratios and used to inoculate growth experiments in sulfur-limited media where DBT was the sole source of sulfur, R. erythropolis IGTS8 invariably emerged as the least abundant species in the resulting culture [34]. These results cannot be explained if DBT is taken up into the cytoplasm of/?, erythropolis IGTS8 and only then is DBT converted to 2-HBP and sulfite, unless it is also hypothesized that sulfite is then excreted. While intracellular metabolism of DBT by R. erythropolis IGTS8 is stated to occur [46] there is no evidence for DBT transport/uptake in desulfurization competent Rhodococcus cultures [25], nor is there evidence for mass transfer limitations in DBT metabolism [25, 80]. If sulfur is liberated intracellularly within R. erythropolis IGTS8 and sulfur is the growth limiting nutrient, it seems unlikely that sulfite would be excreted extracellularly unless intracellular oxidation of sulfite to sulfate were not possible. Sulfate has been demonstrated to be the form of inorganic sulfur that is utilized by R. erythropolis IGTS8 [81] while sulfite has been demonstrated to be the form of sulfur obtained as a product of the desulfurization of DBT [43]. Further research into sulfite and sulfate metabolism by desulfurization competent cultures is warranted. One hypothesis that is consistent with the observation that faster growing desulfurization-deficient bacterial species can dominate mixed cultures when R. erythropolis IGTS8 is the only desulfurization competent culture present, is that desulfurization of DBT occurs in association with the external surface of R. erythropolis IGTS8 cells. The Dsz proteins are known to have membrane-

44

spanning domains [47, 82] so that the desulfurization pathway may function in association with the cell membrane such that extracellular substrates and intracellular cofactors can both be accessed. A further consideration regarding the localization of the Dsz enzymes is the size of some of the substrates that can be metabolized. Solid coal particles and high molecular weight coal derived polymers can be effectively desulfurized [83-85], yet there has never been a report documenting the intracellular uptake of substrates such as coal by any bacterial species. Moreover, the size of coal particles vastly exceeds the size of bacterial cells in experiments where biodesulfurization has been demonstrated to remove 72% of organic sulfur without otherwise altering the composition of the coal [85]. There is no evidence whatsoever that desulfurization enzymes are excreted from R. erythropolis IGTS8 cells, but the size of substrates metabolized and the ability of other bacterial species to successfully compete for sulfur liberated from organosulfur substrates by R. erythropolis IGTS8 make it likely that desulfurization does not occur intracellularly, but in association with the external surface of cells. A consequence of the fact that desulfurization-deficient bacterial species can successfully compete for sulfur liberated from organosulfur substrates by R. erythropolis IGTS8 is that a selective pressure favoring the evolution of a high specific activity for desulfurization enzymes is created. In a mixed culture environment where sulfur is the growth limiting nutrient and R. erythropolis IGTS8 is the only desulfurization-competent culture, this bacterium must liberate many times more sulfur than it needs to meet its own nutritional requirements because competition from other bacteria leaves only a fraction of the utilizable sulfur actually available for use by R. erythropolis IGTS8 [34]. If this dynamic typified the natural environment for R. erythropolis IGTS8 and other desulfurization competent cultures it would be reasonable to expect that a high level of desulfurization activity would have evolved in such cultures. However, that is not the case and even when grown as pure cultures, all naturally occurring desulfurization competent cultures have levels/activities of desulfurization enzymes that are growth limiting rather than capable of supplying sulfur in excess of the needs of the culture [2]. This further illustrates that we have much to learn about the role of Dsz enzymes in nature and the characterization of the microenvironment occupied in nature by R. erythropolis IGTS8 and other desulfurization-competent bacteria. Nevertheless, it is worth considering that enrichment cultures and directed evolution experiments designed to obtain cultures with higher levels of desulfurization activity may benefit from the intentional use of mixed cultures. 2.5. Influence of the bacterial host on biodesulfurization M. phlei GTIS10 appears to be highly similar to R. erythropolis IGTS8 as regards to biodesulfurization capability except that the maximum growth

45

temperature for R. erythropolis IGTS8 is about 33°C whereas M. phlei GTIS10 can grow up to temperatures of 52°C [37]. Figure 2 illustrates that R. erythropolis IGTS8 showed maximal activity at 30°C and a progressive loss of activity at 37°C and 45°C, while no activity is observed at temperatures of 52°C or above. On the other hand, M. phlei GTIS10 exhibits activity over the temperature range of 25°C to 57°C with maximal activity at 45°C to 50°C, and no activity at 62°C. Since these two cultures show such different temperature ranges for growth it was anticipated that the sequences of the dsz genes in M. phlei GTIS10 versus R. erythropolis IGTS8 would be different or contain mutations conferring thermostability; however, both cultures were found to contain the pSOX plasmid encoding dszABC genes having identical DNA sequences [37]. It is likely then that M. phlei GTIS 10 acquired the dsz genes by the conjugal transfer of the pSOX plasmid from R. erythropolis IGTS8 or other desulfurization-competent bacteria. Taken altogether the results shown in Fig. 2, plus the knowledge that DNA sequence analysis showed that both cultures contain identical dsz genes, indicate that the ability of the dsz enzymes to function is greatly influenced by the bacterial host strain. However, enzymes other than, or in addition to, those encoded by the dsz operon may contribute to the desulfurization activity and range of metabolites of DBT produced by M. phlei GTIS 10. Both M. phlei GTIS 10 and R. erythropolis IGTS8 exhibit maximal desulfurization activity corresponding to the optimum growth temperature of each culture, 50° and 30°C respectively, and then desulfurization activity declines in concert with decreasing cell viability at higher temperatures. The desulfurization pathway requires NADH, FMNH2, and oxygen in order to complete the conversion of DBT to 2-HBP. The host must supply these factors so that the functional temperature range of the desulfurization pathway is seen to be different in two different bacterial hosts possibly reflecting the ability of each bacterial species to provide cofactors and reaction substrates at various temperatures. Transport of substrates and products may also contribute to desulfurization activity. The observation that the dsz operon had two apparent temperature maximums in two different bacterial hosts perhaps suggests that if the dsz operon could be expressed in a thermophilic bacterial host the desulfurization enzymes may function at even higher temperatures. Other researchers have also reported obtaining desulfurization competent cultures that have identical dsz gene sequences, but exhibit different phenotypes. Several research groups reported that even though multiple desulfurization competent cultures isolated and examined were Rhodococcus species they exhibited different specific activities for DBT, yields of 2-HBP, activity with 4,6-dimethyl DBT, and sensitivity to hexadecane [86, 87]. The reasons for these phenotypic differences among highly similar bacterial cultures that possess

46

identical desulfurization genes are unknown. However, it is clear that the host contributes to the functioning of the desulfurization pathway in yet uncharacterized ways so that the manipulation of the dsz (or tds) genes alone may be insufficient to yield bacterial cultures with substantially higher desulfurization activity, such as would be required for a commercial biodesulfurization process. 2.6 Desulfurization activity of various cultures The maximum specific desulfurization activity for M. phlei GTIS10 was 1.1 ±0.07 umole 2-HBP/min/g DCW [37]. The maximum specific desulfurization activity for R. erythropolis IGTS8 observed at the Gas Technology Institute (1.2 ± 0.08 umole 2-HBP/min/g DCW) is higher than previous studies where other researchers reported specific desulfurization activity values ranging from approximately 0.6 to 5.8 umole 2-HBP/min/g DCW [46, 57]. The reason why our culture of R. erythropolis IGTS8 showed higher desulfurization specific activity than previously reported may be due to the continuous culturing (> 10 years) of this bacteria in our laboratory under conditions where DBT, or other organosulfur compounds, served as sole sulfur source for growth. The highest desulfurization specific activity reported for the thermophilic Paenibacillus sp. strain A l l - 2 was approximately 0.08 umole 2HBP/min/g DCW [57]. Bacterial cultures containing cloned desulfurization genes from Paenibacillus (tdsABC) were reported to have a maximum desulfurization specific activity of 0.16 umole 2-HBP/min/g DCW [61], while strains containing cloned Rhodococcus desulfurization genes (dszABC) were reported to have a maximum desulfurization specific activity of 4.7 umole 2HBP/min/g DCW [52, 58]. The moderate thermophile Mycobacterium phlei WU-F1 was described as having greater desulfurization activity than Paenibacillus sp. strain Al 1-2, but no specific activity data was reported [53]. 2.7. Genetic modifications to increase desulfurization activity Largely due to the interest in biodesulfurization and the high percentage of desulfurization competent bacteria that are Rhodococcus species as well as other potential applications of this genus, there has been a lot of research on the genetics of Rhodococci [88, 89]. Multiple cloning and shuttle vectors are available and genetic manipulation of Rhodococcus can now be conveniently and reliably performed [90-92]. However, one area of genetic research that has received comparatively little attention is gene expression in Rhodococcus. Overexpression of genes in Rhodococcus has been reported [89, 90, 92, 93]; however, an array of gene expression vectors and a knowledge of the consensus sequences of transcriptional promoters in Rhodococcus is lacking.

47

The coordinated expression of the desulfurization genes has been shown to be important in obtaining maximum desulfurization activity as well as the yields of the pathway intermediates. The flavin oxidoreductase that supplies the cofactors needed by DszC and DszA/(TdsC and TdsA) has been the subject of several investigations. The DszA, B, and C proteins are active in E. coli, but the level of flavin oxidoreductase is low in comparison with Rhodococcus [52, 94, 95], so a flavin oxidoreductase gene from Vibrio harveyi was cloned and expressed in an E. coli strain containing the dszABC genes [52]. This resulted in increasing FMN oxidoreductase levels from 0.03 umol/min/mg protein to 1.1 umol/min/mg protein. Co-expression of the dsz and flavin genes resulted in the highest rates of DBT transformation (51 mg/hr/g DCW), but accumulation of intermediates, mainly DBTSO2, rather than full conversion to 2-HBP, was observed suggesting that DszB is the rate-limiting enzyme in the desulfurization pathway. Other studies of the 4S pathway confirmed that DszB was limiting the global desulfurizing activity [11, 27].

Fig. 2. Resting cells of M. phlei GTIS10 exhibit specific desulfurization activity at higher temperatures than resting cells of R. etythropolis IGTS8. The amount of 2-HBP produced by the conversion of DBT by each culture after incubation for 24 hours at various temperatures was quantified by HPLC analysis. Rate of change in 2-HBP concentration was calculated from the linear portion of the curve, generally the first 4 hours of the incubation. The specific desulfurization activity values recorded are averages of three replicate samples from three separate experiments for a total of nine data points. Standard deviation was less than 10 %. O, M. phlei GTIS10; • , R. erythropolis IGTS8.

48

The highest rate of 2-HBP formation was obtained without the cloned FMN oxidoreductase, perhaps because the accumulation of intermediates inhibited the Dsz enzymes. The maximum amount of 2-HBP produced was 0.2 mM regardless of the amount of DBT added or the incubation time, which suggests that inhibition by 2-HBP is also a key factor limiting biodesulfurization efficiency in E. coli, and probably in other bacterial hosts as well [52]. Oshiro et al. [94] screened 80 bacterial and 20 yeast cell extracts to find flavin reductases with the best ability to support DszC and DszA activity. A flavin oxidoreductase from Paenibacillus polymyxa was found to be the best allowing 3.5 to 5-fold better activity of DszC and DszA as compared with the Rhodococcus flavin oxidoreductase. Rhodococcus strains containing increased copies of dszABC genes on plasmids or integrated into the chromosome have resulted in higher DBT conversion rates, but also in the accumulation of pathway intermediates as DBTSO and DBTSO2 [58, 96]. When the copy number of the dszD gene was increased in Rhodococcus erythropolis KA 2-5-1 cultures containing their natural complement of dszABCD genes, then DBTSO and DBTSO2 accumulation occurred. However, if the copy number of all of the dszABCD genes was increased then the accumulation of intermediates was avoided, but only when the correct balance between dsz genes was achieved [58]. Derivatives of R. erythropolis KA 2-5-1 originally had a specific desulfurization activity of 0.05 mmol/g DCW/hr while derivative cultures that, in addition to their natural complement of dszABCD genes, contained a plasmid with one copy of the dszABC genes had a specific desulfurization activity of 0.14 mmol/g DCW/hr; 0.19 mmol/g DCW/hr with dszABCD genes on a plasmid, and 0.28 mmol/g DCW/hr with two copies of dszABC genes and one dszD gene on a plasmid. Derivative cultures that contained additional copies of the dszABC operon or the dszD gene did not yield cultures with higher enzymatic activity. Similar results were obtained for the expression of various combinations of dsz and tds genes in Rhodococcus, E. coli, and Pseudomonas hosts [96, 97], demonstrating that in order to obtain bacterial cultures with the highest possible desulfurization activities it is necessary to obtain the proper ratio of desulfurization enzymes and cofactors. The results of experiments in which different copy numbers, combinations of dsz/tds genes and promoters were used also revealed that a limit in desulfurization activity is reached that can not be overcome by increasing the amount of the Dsz/Tds enzymes in cells [24, 76, 93, 95, 98]. There are other factors affecting desulfurization activity and/or the intrinsic properties of the Dsz/Tds enzymes needs to be improved if higher desulfurization activity is to be achieved. A way of improving the intrinsic properties of enzymes is directed evolution.

49

Rational protein engineering studies have been a powerful tool, enabling the modification of some enzymes to increase their thermostability, shift their pH optima and alter their substrate specificity [2, 73]. However, in order to do this, a fair amount of information is required such as the amino acid sequence, the three dimensional structure and the location of the active site of the protein. Even when all of this information is available, protein engineering is an uncertain undertaking that can be expensive and time consuming. It is more straightforward to use site specific mutagenesis to increase the hydrophobicity of proteins, however even this requires uncertain theoretical predictions. The amino acid sequences of the three desulfurization enzymes can be inferred from their DNA sequences; however, detailed knowledge of their active sites or three dimensional structures of these proteins is currently unavailable [43, 48, 51, 54]. To obtain this information would take a considerable amount of time and resources and would not be sufficient to guarantee success in developing enzymes with improved activity or thermostability. In addition, while there are some general rules/trends that have emerged regarding the thermostabilization of proteins (such as increasing the hydrophobicity of proteins by placing proline residues at beta-turns in proteins and adding disulfide bonds), these are very general rules that often do not hold true and are difficult to implement. Other methods to increase the thermotolerance of enzymes such as immobilization or post-production modification techniques are not available since biodesulfurization requires the intervention of three enzymes and their associated cofactors, requiring the need to use intact cells rather than immobilized enzyme systems. The thermostabilization of the three-enzyme desulfurization pathway (and the fourth enzyme which supplies cofactors) would be a daunting if not an overwhelming task to accomplish using protein engineering given the current state of knowledge. Fortunately, this information is not necessary in order to employ directed evolution to obtain thermostable desulfurization enzymes. Through the use of mutagenesis combined with natural selection, or screening for enzymatic activity, a directed evolution process can be employed to obtain thermostable derivatives of the desulfurization enzymes. In this method, no prior knowledge of the enzymes three-dimensional structure or even complete amino acid sequence is required. The method mimics nature's own protein engineering system: evolution. All that is needed is a powerful screen or selection so that the desired enzyme traits can be identified. The process of natural selection can then be accelerated in the laboratory to evolve the desired traits. In a short period of time, researchers have modified many enzyme traits such as thermostability, pH optima, substrate specificity and organic solvent tolerance [99-103]. Evolution is accomplished through the combined action of mutation, recombination and selection. Both general and site-specific mutagenesis of targeted genes can be used. In a typical experiment, part or all of a gene or

50

operon is subjected to a general mutagenesis method such as error-prone PCR [100]. Gene shuffling techniques can also be used to increase recombination frequencies and speed the evolutionary process. The gene is then transferred into the host organism and the selection or screen is applied. Candidate mutants with the desired phenotype are then identified and analyzed. These mutants can then be used for further rounds of mutagenesis so that a stepwise approach can be used to gradually evolve the trait of interest. Such a process has been used to improve the desulfurization activity of the DszC enzyme [73]. The dszC genes of R. erythropolis IGTS8 and Nocardia asteroides A3H1 are 89% homologous. More specifically, these two variants of the dszC gene contain alterations at 127 locations resulting in proteins that differ at 38 amino acid positions. This diversity of dszC genes was used in the DNA shuffling/RACHITT technology developed by Enchira Corp. (formerly Energy Biosystems Corporation) to yield a DszC derivative with a 70% higher enzymatic activity [73]. Similar experiments to obtain improved derivatives of other desulfurization enzymes have not been reported. The results obtained with improved derivatives of DszC are promising, but improved derivatives of all of the desulfurization genes are required in order to achieve high levels of improvement in desulfurization activity for the complete 4S pathway. However, since there are other, yet to be identified, cellular components that contribute to the functioning of the 4S pathway, improvements of desulfurization genes only may not be sufficient to achieve very high desulfurization activity compatible with a commercial application [24]. 2.8. Development of processes for the biodesulfurization of diesel and crude oil A cell-free biodesulfurization process would be impractical. The requirement of a three-enzyme pathway along with cofactors will prohibit the use of purified enzyme systems for a practical biodesulfurization process. A commercial biodesulfurization process will have to employ intact bacterial cells as biocatalysts [24, 104, 105]. Because of this, the use of immobilization or other post-production modification techniques to enhance the thermotolerance of enzymes would not be practical for use in an industrial process. It has been proposed however, that individual enzymes that do not require cofactors may be used. Chloroperoxidase and cytochrome C could be used to catalyze the oxidation of DBT to DBTSO2, as well as oxidize petroporphyrins resulting in the release of metals (see chapter 3)[2, 20]. It is necessary to add hydrogen peroxide and water to these peroxidase reactions, but high rates of conversion of DBT to DBTSO2 have been reported. Since DBTSO2 has a higher boiling point than DBT, it has been proposed that distillation can be used to obtain sulfur-free petroleum fractions following peroxidase treatment [2]. This approach has not

51

yet been tested in pilot scale experiments so the costs and efficiency of such a process are not yet known. Energy BioSystems Corporation (EBC) conducted a comprehensive evaluation, particularly as regards crude oil and fractions, of the biodesulfurization technology originally developed by the Institute of Gas Technology (IGT) [now known as the Gas Technology Institute (GTI)] under a program funded by the U.S. Department of Energy. Encouraged by their experimental results and feedback from the petroleum industry, EBC licensed the technology, and assembled a team of executives, engineers, and scientists from the petroleum industry, committed to the commercialization of biodesulfurization technology. The development of bioprocesses for biodesulfurization of petroleum have been almost exclusively focused on the use of biocatalysts that are derivatives of, or related to, R. erythropolis IGTS8, and diesel has been the target for the development of the first biodesulfurization processes [24, 25]. EBC was the first organization to seriously attempt the development of a commercial biodesulfurization process. They chose diesel fuel desulfurization as the target for initial process development efforts because environmental regulations mandating a reduction of the maximum permissible concentration of sulfur in diesel to 50 ppm had been proposed and existing refinery processes were not able to efficiently and economically meet this requirement [25]. The most abundant organosulfur compounds in diesel includes DBT and its derivatives which are recalcitrant to traditional hydrodesulfurization but are good substrates for biodesulfurization [2]. The application of any technology, chemical or biochemical, to the treatment of petroleum requires a highly efficient process as the resulting products are low priced commodities [105]. Moreover, the volume of petroleum processed, even at a small refinery, dwarfs the scale of bioprocesses typically used in the pharmaceutical and biotechnology industries. To address these process concerns, EBC claimed to have achieved a 200-fold improvement in the specific activity of the R. erythropolis IGTS8 biocatalyst using a combination of medium improvement, reaction conditions and genetic engineering [24]. Moreover, process engineering research increased the volumetric reaction rate (oil/water ratio), biocatalyst life and solved separations issues. Specific details about EBC's biodesulfurization process and the results achieved were not published, but a desulfurization rate of 20 umole DBT/min/g DCW was stated as a target for a commercially successful process [25]. The literature contains a large amount of information regarding the use of genetic engineering to achieve higher desulfurization rates as previously discussed in this chapter. It has been shown that R. erythropolis IGTS8 biocatalysts are capable of functioning at 9to-1 oil-to-water ratios [106], and maximum cell yields in fed batch fermentation

52

were reported to be 92 g DCW/liter [107]. Maintaining high cell densities and catalytic rates was accomplished by EBC by employing a cell recycling and regeneration step in their process [24]. The R. erythropolis IGTS8 cells were not used in a process mode in which they were required to grow in the presence of diesel that served as a sole source of sulfur. Instead, separation techniques, typically a hydrocyclone oil/water separator, were used to recover cells so that they could be cleansed of reaction products, regenerated by aeration and nutrients, and recycled in a concentrated form back to the biodesulfurization process. An air lift reactor was used in the EBC process to minimize energy costs in the process, and cell suspensions were found to be more efficient than immobilized cell preparations [24, 25]. A schematic representation of the diesel biodesulfurization process initially envisioned by EBC is shown in Fig. 3 [24]. Initially crude oil would be distilled to yield products that include diesel. Hydrodesulfurization (HDS) would first be used to reduce the sulfur content of the diesel, followed by biodesulfurization (BDS) to further reduce the sulfur content, ultimately yielding a product having 50 ppm sulfur or less. This process scenario takes advantage of the fact that hydrodesulfurization and biodesulfurization complement each other. Hydrodesulfurization can be operated more efficiently and at lower cost by treating only those compounds that are most reactive. Similarly, the biodesulfurization process operates more efficiently by receiving a feed that has a lower total sulfur content and contains almost exclusively only those organosulfur compounds that can be treated most effectively by biodesulfurization. The biodesulfurization process operates at lower temperatures and pressures than hydrodesulfurization. This allows hydrodesulfurization to operate at lower temperatures and pressures than if hydrodesulfurization was the only desulfurization treatment step. This scenario has advantages of 70-80% lower CO2 emissions and energy consumption, and safer operating conditions as compared with hydrodesulfurization alone [108]. Apparently even these improved operating parameters were not sufficient for a commercially viable process for the biodesulfurization of diesel. As discussed previously, the rate limiting step in the 4S pathway is the cleavage of the second/final carbon-sulfur bond catalyzed by DszB. Rather than, or in addition to, obtaining further improvements in the catalytic rate of DszB, an alternative biodesulfurization process scheme is illustrated in Fig. 4 [24]. The oxygenated sulfinate byproduct could be recovered as a value added byproduct, since it has surfactant properties, that would then improve the economics of the biodesulfurization process. However, even this version of a biodesulfurization process for diesel was not commercially successful and EBC went out of business.

53

Fig. 3. Overview of an integrated hydrodesulfurization and biodesulfurization process for diesel oil.

The key reasons that EBC did not succeed in developing a commercially viable biodesulfurization process included changes in the environmental regulations and improvements in hydrodesulfurization technologies. When EBC began process engineering efforts to develop a commercial process for the biodesulfurization of diesel, the environmental regulations specified a maximum total sulfur content of 50 ppm and the existing hydrodesulfurization processes could not efficiently achieve that goal. However, while EBC was involved in the challenging task of implementing the first bioprocess in the petroleum industry (other than waste remediation), stricter environmental regulations were proposed decreasing the maximum permissible sulfur content in diesel to 10 to 15 ppm. Additionally, during this same time frame, improvements were made in hydrodesulfurization technology that allowed these lower sulfur levels to be reached [109]. Integrating a biodesulfurization process into a refinery is the only way to treat a product such as diesel, but this requires a substantial modification of current operations in a refinery and requires that the biodesulfurization process operate at the same speed and reliability as other refinery processes so as not to disrupt normal refining operations. It is very challenging for any new technology to be embraced by a conservative industry such as the petroleum industry so that employing biodesulfurization as a component of refinery operations met with understandable opposition. However, alternative ways of implementing a biodesulfurization process exist (see chapter 4)[104].

54

Fig. 4. Overview of the Energy Biosystems Corporation process for the simultaneous biodesulfurization diesel oil and the production of a sulfinate/surfactant byproduct.

Biorefining can complement existing technologies by specifically addressing compounds/contaminants refractory to current petroleum refinery processes. Heteroatoms such as nitrogen, metals and sulfur can poison the catalysts used in catalytic cracking and hydrotreating processes [5, 109]. Existing refineries are not capable of operating efficiently with heavy crude oils and residuum that have high heteroatom content [110]. Bioprocesses could be used to pre-treat oil reducing the heteroatom content allowing the use of heavy crude oils that could not otherwise be treated with existing refinery processes [104]. Biorefining processes can also be used in conjunction with existing processes to meet the increasingly stringent environmental requirements for contaminant reduction. The development of a bioprocess to selectively remove sulfur, nitrogen, and associated metals, from crude oil and residuum will allow existing refineries to process lower quality oils that they could not otherwise accept. The reduction of sulfur, nitrogen and metals in petroleum would allow refineries to operate more efficiently, decrease costs and protect the environment.

55

A potentially attractive means of implementing a petroleum biodesulfurization process could be to treat heavy oil on site in the production field prior to the initial separation of petroleum from produced water [104]. Since biodesulfurization would require a water wash/separation step to remove the biocatalysts and liberate the sulfur from treated petroleum and since a water separation step is a normal component of petroleum production, performing biodesulfurization on produced oil would minimize the required processing steps. The produced water could be re-injected on site for secondary recovery/water flooding operations and could therefore eliminate the need for wastewater treatment [4]. The reinjection of sulfur-laden water into a petroleum field generates concerns about the potential souring of wells, but strategies for the prevention and control of H2S formation are routinely used in existing petroleum production operations and could be employed to deal with water resulting from biodesulfurization treatments. A schematic illustration of a biorefining process used to treat crude oil in association with production operations is shown in Fig. 5. This scenario could greatly improve the economics of a biodesulfurization process and could fit into existing petroleum production procedures with minimal modifications. This approach would require biodesulfurization reactors to be present at petroleum production sites; however, it is possible that conventional petroleum storage tanks can be modified for use in biodesulfurization treatments. The main concern about this biodesulfurization approach would be the temperature of the produced oil and the thermal tolerance of biocatalysts. Desalting and dewatering processes for petroleum are normally performed at temperatures of from 60°C to 100°C, and the use of elevated temperatures will be increasingly important in the treatment of heavy crude oils and residuum in order to deal with the viscosity of these heavy oils [4, 23, 110]. If thermophilic cultures could be used to desulfurize heavy oils in conjunction with desalting and dewatering processes, then both the viscosity and sulfur content of heavy oils could be simultaneously reduced allowing upgraded lighter oils to be sent to refineries for subsequent processing. The hydrodesulfurization and FCC of these upgraded oils would be easier [30, 110]. Thus, biodesulfurization of heavy crude oils would fit well within current practices of the petroleum industry, but the biodesulfurization should be performed at high temperatures (60°C to 100°C). Microbial cultures that can selectively desulfurize petroleum have already been identified, but cultures that will efficiently desulfurize petroleum at thermophilic temperatures are not yet available [37]. Performing biorefining processes at higher temperatures is not only more compatible with existing industry practices, but would also result in higher catalytic rates and the reduced viscosity of petroleum at higher temperatures would allow lower processing

56

costs. Thermophilic microorganisms have not been well studied and no systematic examination of thermophilic cultures for possible use in biorefining has been reported [111, 112]. Biotechnology may one day solve many problems confronting the petroleum industry today, but a biorefining process will have to operate on a far greater scale and at less cost than any current biotechnology process. For any process to be viable in the petroleum industry, it must not only be capable of treating the complex mixture of chemicals that comprise petroleum but it must also treat very large volumes in a cost effective way. Many enzymes catalyze reactions relevant to biorefining goals but they must be improved in numerous ways before practical and economical bioprocesses can be developed. Metabolic engineering and directed evolution approaches are possible means for the development of bioprocesses relevant to the petroleum industry. Specific development needs include: developing cultures capable of performing biodesulfurization at thermophilic temperatures (60° to 100°C), developing cultures with higher levels of desulfurization activity, and process development research for the biodesulfurization of heavy oils.

Fig. 5. Overview of a process for the biodesulfurization of crude oil.

57

3. BIODENITROGENATION OF PETROLEUM The removal of organically bound nitrogen from crude oil, without the loss of significant calorific value, requires the selective cleavage of carbon-nitrogen bonds. The selective cleavage of carbon-sulfur bonds in crude oil using biocatalysts has been demonstrated [2, 24, 113] and it may be possible to selectively cleave carbon-nitrogen bonds using biocatalysts developed from microorganisms capable of metabolizing compounds such as quinoline and carbazole [114]. The cleavage of carbon-nitrogen bonds resulting in the conversion of quinoline to 8-hydroxycoumarin and ammonia has been demonstrated [115], and the genes that encode the enzymes participating in the quinoline degradation pathway have been identified and sequenced [116]. The removal of nitrogen from crude oil by a quinoline degrading culture, Pseudomonas ayucida IGTN9m, has also been demonstrated [115]. However, the abundance of quinoline relative to other organonitrogen compounds in crude oil is low and existing quinoline degradation enzymes have a narrow substrate range. Consequently, even though removal of 68% of quinoline from crude oil was demonstrated the total nitrogen content was reduced by only 5%. An appropriate topic for future research is the development of cultures that express higher levels of quinoline degrading enzymes that have wider substrate ranges, but it is also important to develop biocatalysts that can remove nitrogen from other compounds typically found in petroleum such as carbazole.

Fig. 6. Carbazole Degradation Pathways. The top pathway illustrates the existing carbazole degradation pathway that results in overall degradation, whereas the bottom pathway illustrates a potential pathway for the selective removal of nitrogen from carbazole that could be developed using metabolic engineering.

58

Carbazole is a good model compound that is representative of the nitrogen-containing compounds present in the greatest abundance in many petroleum fractions [12, 117]. For developing a biological process for the removal of nitrogen from petroleum, none of the presently known carbazoledegrading cultures are particularly appropriate because nitrogen is only removed in the course of complete degradation [114, 118]. A variety of carbazole-degrading microorganisms have been reported in the literature including Sphingomonas, Pseudomonas, Mycobacterium, Ralstonia and Xanthomonas species [114, 119-125]. Insofar as biodegradation pathways have been investigated, these differing species of carbazole degraders follow a similar carbazole degradation pathway that begins with the oxidative cleavage of the hetorocyclic nitrogen ring of carbazole to form 2'aminobiphenyl-2,3-diol. This compound is then oxidized through meta cleavage yielding 2-hydroxy-6-oxo-6-hexa-2e,4z-dienoate. The next metabolic steps result in the degradation of one of the aromatic rings releasing carbon dioxide. In existing pathways nitrogen is released from carbazole only after substantial carbon degradation. Figure 6 illustrates the carbazole degradation pathway employed by currently known carbazole utilizing cultures as well as a potential pathway for selective removal of nitrogen from carbazole that could be created using metabolic engineering by combining the CarA enzyme from carbazole degraders such as Sphingomonas sp. GTIN11 with a suitable deaminase. Some carbazole-degrading cultures, like Sphingomonas sp. CB3 [124], have been found to contain carbazole dioxygenases that are related to biphenyl oxidases while other cultures, like Pseudomonas resinovorans CAIO [121], contain carbazole dioxygenases that show no close relationship to other characterized oxidases. CARDO consists of three components: a dioxygenase, ferredoxin, and ferredoxin reductase. In some carbazole-degrading cultures, like Pseudomonas resinovorans CAIO [121], the dioxygenase is a single protein, but in other cultures, like Sphingomonas sp. CB3 [124], the dioxygenase is comprised of two subunits. The arrangement of genes encoding CARDO also differs significantly as the four genes in Sphingomonas sp. CB3 (carAa, carAb, car Ac and car Ad) are contiguous and arranged in that order [124], while the genes encoding CARDO in Pseudomonas resinovorans CAIO are not contiguous [126]. Additionally, the carbazole dioxygenase enzyme of Sphingomonas sp. CB3 has a rather narrow substrate range and does not metabolize naphthalene, dibenzothiophene, phenanthrene, or fluorene unlike Pseudomonas resinovorans CAIO [121]. While the arrangement of genes encoding the enzymes involved in carbazole degradation in Sphingomonas sp. GTIN11 is similar to the order found in Pseudomonas resinovorans CAIO, the carA, carB, and carC genes of Sphingomonas sp. GTIN11 do not show significant homology to the car genes present in either Pseudomonas resinovorans CAIO or Sphingomonas sp. CB3. Moreover, while several

59

carbazole-degrading microbial cultures are known, the ability of these cultures to selectively remove nitrogen from crude oil has only been tested for Sphingomonas sp. GTIN11 [118]. Therefore several bacterial cultures are known that can utilize carbazole as a sole nitrogen source, but no culture is known that can selectively cleave both C-N bonds in carbazole while leaving the rest of the molecule intact. Sphingomonas sp GTIN11 [118] was demonstrated to metabolize carbazole, and to a lesser extent Cl and C2 derivatives of carbazole, from petroleum. As much as 95% of the carbazole present in crude oil was removed as a consequence of biotreatment by Sphingomonas sp. GTIN11. However, the reduction in the total nitrogen content of the crude oil was relatively modest. This highlights the need for developing of improved cultures that contain appropriate biochemical pathways for the cleavage of carbon-nitrogen bonds in organonitrogen compounds and that have broad substrate ranges that encompass the diverse mixture of organonitrogen compounds typically found in crude oil. The genes encoding the carbazole degradation pathway of Sphingomonas sp. GTIN11 have been cloned and sequenced [118]. The reaction catalyzed by CarA converts carbazole to 2'-aminobiphenyl-2,3-diol accomplishing the cleavage of the first C-N bond in carbazole. There are no known deaminases or amidases that can metabolize 2'-aminobiphenyl-2,3-diol and accomplish the cleavage of the final C-N bond [114, 118]. If a deaminase that will recognize 2'aminobiphenyl-2,3-diol as a substrate could be found then the gene encoding this enzyme could be combined with the carA genes (carAa, carAc, and carAd encoding for the carbazole dioxygenase, ferredoxin and ferredoxin reductase respectively) from Sphingomonas sp. GTIN11 and thereby construct a synthetic operon for the selective removal of nitrogen from carbazole. A preferred bacterial strain would lack the carB and carC genes so that complete biodegradation of carbazole would be avoided and the final product would be 2',2,3-trihydroxybiphenyl (or a similar compound). There are reports of mixed cultures of thermophilic bacteria that selectively reduce the nitrogen content of crude oil by as much as 45% [112]. However, other investigators have not verified these results, and the identity of the microbial species responsible for carbon-nitrogen bond cleavage is unknown. Consequently, nothing is known about the biochemistry or the genetics of this nitrogen removal phenomenon, so that there is no straightforward means of furthering and improving this process. While the selective removal of nitrogen from crude oil has been demonstrated and information about the biochemistry and genetics of quinoline and carbazole degradation is available, it is clear that much remains to be done before a commercially viable process for the biodenitrogenation of petroleum can be considered. The chief need is for enzymes with broader substrate ranges.

60

4. BIOPROCESSES FOR THE REMOVAL OF METALS FROM PETROLEUM The use of biotechnology to reduce the concentration of metals in petroleum is the least studied topic in biorefining research [2]. Chloroperoxidase and cytochrome C peroxidase could be used to catalyze the oxidation of petroporphyrins resulting in the release of metals (see chapter 3). The removal of 53% nickel and 27% vanadium from crude oil has been reported using chloroperoxidase [20]. Peroxidase reactions can be accomplished using enzymes rather than whole cells because a single enzyme, rather than a multi-step pathway, is involved and no cofactors are required. Water and hydrogen peroxide must be provided in order to promote this reaction, but a practical bioprocess for the removal of metals from petroleum may be possible. Since most metals in petroleum are associated with organonitrogen compounds it may be possible that improved biodenitrogenation processes will simultaneously reduce the nitrogen and the metal content, thereby avoiding the need for a separate metal removal process. 5. CONCLUSIONS AND FUTURE RESEARCH PRIORITIES Heavy crude oils and residuum constitute a significant, and constantly increasing portion of world reserves. These heavy oils possess high calorific content yet have comparatively low market values mainly because of high sulfur and metals content and high viscosity/molecular weight. The sulfur and nitrogen content is of environmental concern due to potential sulfurous and nitrous emissions from petroleum combustion. Metals, and to a lesser extent sulfur and nitrogen, present in heavy crude oils can poison catalysts used in hydrodesulfurization therefore limiting the effectiveness of current technologies to remove sulfur and nitrogen from these oils. Furthermore, the catalytic cracking process used to convert crude oil to lower molecular weight products is negatively affected by the sulfur, nitrogen and metal content. The high viscosity/high molecular weight of these oils limits the amount of higher value petroleum byproducts such as gasoline, aviation fuel, and diesel fuel that can be obtained as well as causes increased operating costs. These problems associated with heavy oils have prompted the preferential utilization of light crude oils. As light crude oils are consumed at a disproportional/high rate the amount of heavy oil as a percentage of remaining world petroleum reserves continues to increase. New technologies capable of dealing with heavy oils to mitigate environmental concerns and increase byproduct yields in a cost effective manner are needed: biorefining may provide an answer.

61

Perhaps the best way of implementing a biorefining process is to integrate it into existing industry practices to the greatest degree possible and in such a way as to complement existing industry practices. It is believed that biorefining can be particularly useful in the treatment of heavy crude oils as the technology can remove sulfur, nitrogen and metals and simultaneously reduce the viscosity/molecular weight of oil as a consequence of carbon-sulfur and carbonnitrogen bond cleavage, and can tolerate heavy metal and salt concentrations typically found in heavy oils and produced/formation waters. If biorefining can be used in conjunction with desalting and dewatering steps in oil production operations and reduce the sulfur, nitrogen, and metal content of crude oil prior to the oil being sent to refineries then existing refining technologies could be used with an expanded range of low quality oils that could not otherwise be treated. While the potential of biorefining has been demonstrated, more development is needed before the technology can be successfully commercialized for the treatment of heavy oils. Specific development needs include: developing cultures capable of performing biorefining at thermophilic temperatures (60° to 100°C), cultures with higher levels of enzymatic activity, new biochemical pathways for the selective cleavage of carbon-nitrogen bonds, new enzymes with broader substrate ranges, and process development research for the biorefining of heavy oils and residuum. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] II1] [12] [13] [14]

B. Mason, Nature October 29, (2003), www.nature.com/nsu/031027/031027031023.html. S. Le Borgne, and Quintero R., Fuel Processing Technol. 1641 (2003), 1. H. Kim, J. M. Vohs, and R. J. Gorte, Chem. Commun. (Camb) (2001) 2334. L. J. Drew, Kirk-Othmer Encyclopedia of Chemical Technology (1996) (Kroschwitz, J. I., and Howe-Grant, M., Eds.), pp. 342-476. L. L. Hegedus, and McCabe, R. W., Catalyst Rev. 23 (1981) 377. S. Reeson, Energy World 235 (1996) 9. J. G. Speight, The Chemistry and Technology of Petroleum, Marcel Dekker Inc, New York 1980. C. S. Hsu, K. Qian, and W. K. Robbins, J. of High Resolution Chromatography 17 (1994)271. C. S. Creaser, F. Krokos, K. E. O'Neill, M. J. C. Smith and P. G. McDowell, J. Am. Soc. Mass Spectrometry 4 (1993), 322-326. Shennan-J-L, J. Chem. Technol. Biotechnol. 67 (1996) 109. B. L. McFarland, D. J. Boron, W. Deever, J. A. Meyer, A. R. Johnson, and R. M. Atlas, Crit. Rev. Microbiol. 24 (1998) 99. G. W. Mushrush, E. J. Beal, D. R. Hardy, and J. M. Hughes, Fuel Processing Technology 61 (1999) 197. A. Donetti, E. Cereda, A. Ezhaya, and R. Micheletti, J. Med. Chem. 32 (1989) 957. E. V. Brown, and R. Isbrandt, J Med Chem 14 (1971) 84.

62

[15] J. Jacob, A. Schmoldt, C. Augustin, G. Raab, and G. Grimmer, Toxicology 68 (1991) 181. [16] K. G. Kropp, and P. M. Fedorak, Can. J. Microbiol. 44 (1998) 605. [17] F. A. Leighton, Fundam. Appl. Toxicol. 12 (1989) 787. [18] T. Maruyama, L. L. Wotring, and L. B. Townsend, J. Med. Chem. 26 (1983) 25. [19] T. McFall, G. M. Booth, M. L. Lee, Y. Tominaga, R. Pratap, M. Tedjamulia, and R. N. Castle, Mutat. Res. 135 (1984) 97. [20] L. Mogollon, R. Rodriguez, W. Larrota, C. Ortiz, and R. Torres, Appl. Biochem. Biotecchnol. 70-72 (1998) 765. [21] Monticello-D-J, and W. R. Finnerty, Ann. Rev. Microbiol. 39 (1985) 371. [22] P. Kassler, Energy exploration and exploitation (1996) (Jenkins, G., Ed.), pp. 229-242 Multi-Science Publishing Co, Berkshire, United Kingdom. [23] P. O'Connor, L. A. Gerritsen, J. R. Pearch, P. H. Desai, S. Yanik, and A. Humphries, Hydrogen Processing 11 (1991) 76. [24] M. A. Pacheco, E. A. Lange, P. T. Pienkos, L-Q. Y. M. P. Rouse, Q. Lin, and L. K. Linguist, National Petrochemical & Refiners Association (1999), pp. AM-99-27, San Antonia, Texas. [25] D. J. Monticello, Curr. Opin. Biotechnol. 11 (2000) 540. [26] Thomas-J-A, G. Ganapathi, and E. L. Stover, Res. J. Water Pollution Control Fed. 63 (1991)475. [27] B. L. McFarland, Curr. Opin. Microbiol. 2 (1999) 257. [28] Hart's Diesel Fuel News (1998). [29] Kilbane-J-J, Trends in Biotechnology 7 (1989) 97. [30] J. R. Harris, Hydrocarbon Processing 75 (1996) 63. [31] T. Isoda, S. Nagao, X. Ma, Y. Korai, and I. Mochida, Enegy & Fuels 10 (1996) 1078. [32] J. J. Kilbane, Trends in Biotechnology 7 (1989) 97. [33] Sandhya-S, Indian Journal of Microbiology 36 (1996) 1. [34] K. J. Kayser, B. A. Bielaga-Jones, K. Jackowski, O. Odusan, and J. J. Kilbane, J. Gen. Microbiol 139 (1993) 3123. [35] J. R. Gallagher, E. S. Olson, and D. C. Stanley, FEMS Microbiol. Lett. 107 (1993) 31. [36] Kim-T-S, H. Y. Kim, and B. H. Kim, Biotechnol. Lett. 12 (1990) 757. [37] K. J. L. C. Kayser, H.-S. Park, J.-H. Kwak, A. Kolhatkar, and J. J. Kilbane II, Appl. Microbiol. Biotechnol. 59 (2002) 737. [38] S. C. Gilbert, J. Morton, S. Buchanan, C. Oldfield, and A. McRoberts, Microbiology 144(1998)2545. [39] M. Kobayashi, T. Onaka, Y. Ishii, J. Konishi, M. Takaki, H. Okada, Y. Ohta, K. Koizumi, and M. Suzuki, FEMS Microbiol. Lett. 187 (2000) 123. [40] C. Oldfield, N. T. Wood, S. C. Gilbert, F. D. Murray, and F. R. Faure, Antonie Van Leeuwenhoek 74 (1998) 119. [41] T. Matsui, T. Onaka, Y. Tanaka, T. Tezuka, M. Suzuki, and R. Kurane, Biosci. Biotechnol. Biochem. 64 (2000) 596. [42] J. H. Chang, Y.J. Kim, B.H. Lee, K.S. Cho, H.W. Rhu, Y.K. Chang, and H.N. Chang, Biotechnology Progress 17 (2001) 876. [43] N. Nakayama, T. Matsubara, T. Oshiro, Y. Moroto, Y. Kawata, K. Koizumi, Y. Hirakawa, M. Suzuki, K. Muruhashi, Y. Izumi, and R. Kurane, Biochem. Biophys. Acta 1598 (2002) 122. [44] S. Nekodzuka, T. Nakajima-Kambe, N. Nomura, J. Lu, and T. Nakahara, Biocatalysis Biotrans. 15 (1997) 17.

63

[45] K. A. Gray, O. S. Pogrebinsky, G. T. Mrachko, L. Xi, D. J. Monticello, and C. H. Squires, Nat. Biotechnol. 14 (1996) 1705. [46] C. Oldfield, O. Pogrebinsky, J. Simmonds, E. S. Olson, and C. F. Kulpa, Microbiology 143(1997)2961. [47] C. S. Piddington, B. R. Kovacevich, and J. Rambosek, Appl. Environ. Microbiol. 61 (1995)468. [48] L. M. Watkins, R. Rodriguez, D. Schneider, R. Broderick, M. Cruz, R. Chambers, E. Ruckman, M. Cody, and G. T. Mrachko, Arch. Biochem. Biophys. 415 (2003) 14. [49] C. Denis-Larose, D. Labbe, H. Bergeron, A. M. Jones, C. W. Greer, J. al-Hawari, M. J. Grossman, B. M. Sankey, and P. C. Lau, Appl. Environ. Microbiol. 63 (1997) 2915. [50] C. Denis-Larose, H. Bergeron, D. Labbe, C. W. Greer, J. Hawari, M. J. Grossman, B. M. Sankey, and P. C. Lau, Appl. Environ. Microbiol. 64 (1998) 4363. [51] T. Matsubara, T. Oshiro, Y. Nishina, and Y. Izumi, Appl. Environ. Microbiol. 67 (2001) 1179. [52] D. S. Reichmuth, H. H., H. W. Blanch, and J. D. Keasling, Biotechnol. Bioeng. 67 (2000) 72. [53] T. Furuya, K. Kirimura, K. Kino, and S. Usami, FEMS Microbiol Lett 204 (2001) 129. [54] B. Lei, and S. C. Tu, J. Bacteriol. 178 (1996) 5699. [55] T. Ohshiro, K. Suzuki, and Y. Izumi, J. Ferment. Bioeng. 83 (1997) 233. [56] T. Ohshiro, T. Koshima, K. Torii, H. Kawasoe, and Y. Izumi, J. Biosci. Bioeng. 88 (1999)610. [57] J. Konishi, Y. Ishii, T. Onaka, K. Okumura, and M. Suzuki, Appl. Environ. Microbiol. 63(1997)3164. [58] K. Hirasawa, Y. Ishii, M. Kobayashi, K. Koizumi, and K. Maruhashi, Biosci. Biotechnol. Biochem. 65 (2001) 239. [59] S. K. Rhee, J. H. Chang, Y. K. Chang, and H. N. Chang, Appl. Environ. Microbiol. 64 (1998)2327. [60] M. Kobayashi, K. Horiuchi, O. Yoshikawa, K. Hirasawa, Y. Ishii, K. Fujino, H. Sugiyama, and K. Maruhashi, Biosci. Biotechnol. Biochem. 65 (2001) 298. [61] Y. Ishii, J. Konishi, H. Okada, K. Hirasawa, T. Onaka, and M. Suzuki, Biochem. Biophys. Res. Commun. 270 (2000) 81. [62] J. Konishi, T. Onaka, Y. Ishii, and M. Suzuki, FEMS Microbiol. Lett. 187 (2000) 151. [63] Y. Ishii, J. Konoshi, M. Suzuki, and K. Maruhashi, J. Biosci. Bioeng. 90 (2000) 591. [64] T. Onaka, J. Konishi, Y. Ishii, and K. Maruhashi, J. Biosci. Bioeng. 92 (2001) 193. [65] J. Konishi, H. Okada, K. Hirasawa, Y. Ishii, and K. Maruhashi, Biotechnol. Lett. 24 (2002) 1863. [66] G. F. Duarte, A. S. Rosado, L. Seldin, W. de Araujo, and J. D. van Elsas, Appl. Environ. Microbiol. 67 (2001) 1052. [67] S. A. Denome, and K. D. Young, Gene 161 (1995) 33. [68] T. Matsui, K-I. Noda, Y. Tanaka, K. Maruhashi, and R. Kurane, Curr. Microbiol. 45 (2002) 240. [69] T. Omori, L. Monna, Y. Saiki, and T. Kodama, Appl. Environ. Microbiol. 58 (1992) 911. [70] L. Serbolisca, F. de Ferra, and I. Margarit, Appl. Microbiol. Biotechnol. 52 (1999) 122. [71] K. Kirimura, T. Furuya, Y. Nishii, Y. Ishii, K. Kino, and S. Usami, J. Biosci. Bioeng. 91 (2001) 262. [72] J. H. Chang, S. K. Rhee, Y. K. Chang, and H. N. Chang, Biotechnol. Prog. 14 (1998) 851.

64

[73] W. M. Coco, W. E. Levinson, M. J. Crist, H. J. Hektor, A. Darzins, P. T. Pienkos, C. H. Squires, and D. J. Monticello, Nat. Biotechnol. 19 (2001) 354. [74] J. D. Van Hamme, and O. P. Ward, Appl. Environ. Microbiol. 67 (2001) 4874. [75] J. J Arensdorf, A.K. Loomis, P.M. DiGrazia, D.J. Monticello, and P. T. Pienkos, Appl. Environ. Microbiol. 68 (2002) 691. [76] K. Watanabe, K. Noda, Y. Ohta, and K. Maruhashi, Biotechnol. Lett. 24 (2002) 897. [77] S. Krawiec, Develop. Ind. Microbiol. 31 (1990) 103. [78] J. J. Kilbane II, Resource Conservation & Recycling 3 (1990) 69. [79] J. J. Kilbane, Trends in Biotechnology 7 (1989) 97. [80] E. N. Kaufman, J. B. Harkins, and A. P. Borole, Appl. Biochem. Biotechnol. 73 (1998) 127. [81] L. Setti, P. Farinelli, S. D. Martino, S. Frassinetti, G. Lanzarini, and P. G. Pifferi, Appl. Microbiol. Biotechnol. 52 (1999) 111. [82] S. A. Denome, C. Oldfield, L. J. Nash, and K. D. Young, J. Bacteriol. 176 (1994) 6707. [83] D. L. Stoner, J. E. Wey, K. B. Barrett, J. G. Jolley, R. B. Wright, and P. R. Dugan, Appl. Environ. Microbiol. 56 (1990) 2667. [84] J. J. Kilbane, and K. Jackowski, Biotechnol. Bioeng. 40 (1992) 1107. [85] G. P. Huffman, N. Shah, F. E. Huggins, L. M. Stock, K. Chatterjee, J. J. Kilbane, and M. M. Chou, Fuel 74 (1995) 549. [86] S. Abbad-Andaloussi, C. Lagnel, M. Warzywoda, and F. Monot, Microbial Technology 32 (2003) 446. [87] G. Castorena, C. Suarez, I. Valdez, G. Amador, L. Fernandez, and S. Le Borgne, FEMS Microbiol. Lett. 215 (2002) 157. [88] E. R. Dabbs, Antonie-Leeuwenhoek-Journal of Microbiology 74 (1998) 155. [89] M. J. Larkin, R. DeMot, L. A. Kulakov, and I. Nagy, Antonie-Leeuwenhoek-Journal of Microbiology 74 (1998) 133. [90] M. Vesely, M. Patek, J. Nesvera, A. Cejkova, J. Masak, and V. Jirku, Appl. Microbiol. Biotechnol. 61 (2003) 523. [91] R. DeMot, I. Nagy, A. DeSchrijver, P. Pattanapipitpaisal, G. Schoofs, and J. Vanderleyden, Microbiology 143 (1997) 3137. [92] M. Arenskotter, D. Baumeister, R. Kalscheuer, and A. Steinbuchel, Appl. Environ.Microbiol. 69 (2003) 4971. [93] E. Franchi, F. Rodriguez, L. Serbolisca, and F. de Ferra, Oil & Gas Science & Technol. Rev. IFP 58 (2003) 515. [94] T. Ohshiro, Y. Aoi, K. Torii< and Y. Izumi, App. Microbiol. Biotechnol. 59 (2002) 649. [95] P. Galan, E. Diaz, and J. L. Garcia, Environ. Microbiol. 2 (2000) 687. [96] T. Matsui, K. Hirasawa, K. I. Koizumi, K. Maruhashi, and R. Kurane, Biotechnol. Lett. 23(2001)1715. [97] Y. Ishii, T. Ohshiro, Y. Aoi, M. Suzuki, and Y. Izumi, J. Biosci. Bioeng. 90 (2000) 220. [98] K.-I. Noda, K. Watanabe, and K. Maruhashi, J. Biosci. Bioeng. 95 (2003) 504. [99] J. C. Moore, and F. H. Arnold, Nat. Biotechnol. 14 (1996) 458. [100]F. H. Arnold, L. Giver, A. Gershenson, H. Zhao, and K. Miyazaki, Ann. N Y Acad. Sci. 870 (1999) 400. [101] J. Hoseki, T. Yano, Y. Koyama, S. Kuramitsu, and H. Kagamiyama, J. Biochem. (Tokyo) 126 (1999) 951. [102]T. Yano, S. Oue, and H. Kagamiyama, PNAS 95 (1998) 5511. [103]T. Yano, and H. Kagamiyama, PNAS 98 (2001) 903. [104]R. G. Shong, Division of Fuel Chemistry, American Chemical Society 44 (1999), 1-4. [105]J. L. Shennan, J. Chem. Technol. Biotechnol. 67 (1996) 109.

65

[106] S. Patel, J. J. Kilbane, and D. A. Webster, J. Chem. Technol. Biotechnol. 69 (1997) 100. [107]O. Yoshikawa, Y. Ishii, K-I. Koizumi, T. Ohshiro, Y. Izumi, and K. Marahashi, J. Biosci. Bioeng. 94 (2002) 447. [108]L. Linguist, and M. Pacheco, Oil & Gas Journal (1999), pp. 45-48. [109] S. T. Oyama, and Y-K. Lee, American Chemical Society, Fuel Chemistry Division 48 (2003), 173-174. [110]S. Reeson, Energy World 235 (1996) 9. [111]M. S. Lin, T. Premuzic, J. H. Yablon, and W. M. Zhou, Appl. Biochem. Biotechnol. 5758(1996)659. [112]E. T. Premuzic, M. S. Lin, M. Bohenek, and W. M. Zhou, Enegy & Fuels 13 (1999) 297. [113] J. J. Kilbane II, Final Report, Energy BioSystems Project No. 40308-02 (1992). [114]M. J. Benedik, P. R. Gibbs, R. R. Riddle, and R. C. Wilson, Trends in Biotechnology 16 (1998)390. [115] J. J. Kilbane, R. Ranganathan, K. J. Kayser, L. Cleveland, C. Ribiero and M. M. Linhares, Appl. Environ. Microbiol. 66 (2000) 688. [116]U. Frerichs-Deeken, B. Goldenstedt, R. Gahl-Janben, R. Kappl, J. Huttermann, and S. Fretzner, European J. Biochem. 270 (2003) 1567. [117]S. Mitra-Kirtley, O. C. Mullins, J. van Elp, S. J. George, J. Chen, and S. P. Cramer, J. Am. Chem. Soc. 115 (1993) 252. [118] J. J. Kilbane II, A. Daram, J. Abbasian, and K. J. Kayser, Biochem. Biophys. Res. Comm. 297 (2002) 242. [119]H. Habe, Y. Ashikawa, Y. Saiki, T. Yoshida, H. Nojiri, and T. Omori, FEMS Microbiol. Lett. 211(2002), 43. [120]K. Kirimura, H. Nakagawa, K. Tsuji, K. Matsuda, R. Kurane, and S. Usami, Biosci. Biotechnol. Biochem. 63 (1999) 1563. [121]H. Nojiri, J. W. Nam, M. Kosaka, K. I. Morii, T. Takemura, K. Furihata, H. Yamane, and T. Omori, J. Bacteriol. 181 (1999) 3105. [122]N. Oichiyama, T. Omori, and T. Kodama, Biosci. Biotech. Biochem. 57 (1993) 455. [123] J. Schneider, R. J. Grosser, K. Jayasimhulu, W. Xue, B. Kinkle, and D. Warshawsky, Can. J. Microbiol. 46 (2000) 269. [124]J. M. Shepherd, and G. Lloyd-Jones, Biochem. Biophys. Res. Comm. 247 (1998) 129. [125]R. R. Riddle, P. R. Gibbs, R. C. Wilson, and M. J. Benedik, J. Ind. Microbiol. Biotechnol. 30 (2003) 6. [126]S. I. Sato, J. W. Nam, K. Kasuga, H. Nojiri, H. Yamane, and T. Omori, J. Bacteriol. 179 (1997)4850.


Studies in Surface Science and Catalysis 151 R. Vazquez-Duhalt and R. Quintero-Ramirez (Editors) © 2004 Elsevier B.V. All rights reserved.

67

Chapter 3

Enzymatic catalysis on petroleum products M. Ayala" and R. Vazquez-Duhaltb a

Instituto Mexicano del Petroleo. Eje Central Lazaro Cardenas 152, San Bartolo Atepehuacan 07730 Mexico DF, Mexico b

Instituto de Biotecnologia, UNAM. Apartado Postal 510-3 Cuernavaca, Morelos 62250 Mexico

1. INTRODUCTION The contemporary society is highly dependent on oil supply for energy, transportation, food production, and in general, industrial production. A century ago the oil exploitation began, first as a source of energy and now as a source both of energy and raw material. Thus, history will describe our time as the oil based society. Nature took 500 millions years to accumulate the world's oil; nevertheless, the world's petroleum could be consumed in two centuries [1]. The inexorable production peak is estimated to occur sometime between 2010 and 2020 and then the oil resources will be drastically reduced at the end of this century (Fig. 1) [2]. When the world's oil reserves become scarce, the more expensive fuel sources as hard-to-extract oil deposits, tarry sands, and synfuels from coal will be brought to the front of production.

Fig. 1. Petroleum availability estimation from World Resource Institute (1996)

68

New technologies should improve the refining efficiency in terms of the consumed energy and the environmental impact of the processes. Developments and implementation of new technologies for conventional processes, such as cracking, hydrogenation, isomerization, alkylation, polymerization, and hydrodesulfurization could be expected. Nevertheless, the introduction of nonconventional technologies, representing potential substitutes or complementary processes to traditional oil refining, may happen. Significant progress has been made in the last decades in technologies such as membrane separation [3], supercritical extractions [4-6], and many others. Biotechnology is among the new fields that might be introduced to the oil refining industry. The potential application of biochemical catalysis in the petroleum refining industry has been recently reviewed in a prospective analysis of the available data on the microbial and enzymatic modification of oil products [7]. The proposed biotechnological processes should be considered either alternative or complementary to conventional oil refining technologies. The introduction of such novel nonconventional techniques in the petroleum industry may improve its energetic efficiency and reduce its environmental impact. The first biotechnological processes applied in the oil industry were environmental processes, such as wastewater treatment and soil bioremediation. However, there are other potential uses for biotechnological processes in the oil industry. It is important to point out that so far no enzymatic or biochemical processes exist in the oil refining industry, thus this chapter shows a prospective analysis of data on enzymatic transformations of petroleum products and their derivatives, in order to evaluate the possible introduction of biotechnological processes in the petroleum refining industry. Preconceived ideas and misconceptions about enzymes continue to limit the applications of enzymes on biotransformations in the chemical industry. Nevertheless, the fact is that there are industrially-successful examples of biocatalytic process that show enzymes to be sufficiently stable, productive and economic for commercial applications. Enzymes have huge breadth of scope in the types of reactions that may be catalyzed and the chemical nature of compounds that may be transformed. Biocatalysis is no panacea, but still it is in its infancy and significant progress may be expected. Enzymes are far more efficient than chemical catalysts: high specificity, low substrate concentration, and mild reaction conditions are the most interesting properties of enzymes in this regard. Enzyme-catalyzed reactions usually display characteristically high turn-over numbers, with rate accelerations approaching or exceeding 108. In terms of productivity and if we consider single step enzymatic transformations, productivities of tens or even hundreds of grams

69

of product per liter per hour have been achieved. Biocatalytic reactions that have been successfully applied in the industry, at large scale, include: production of high fructose corn syrup, fatty acids and triglyceride oils, aspartame, acrylamide, antibiotic precursors, amino acids, S-2-chloropropionic acid, polylactic acid and cyclodextrins [8]. At first glance, enzymes seem to be more expensive than chemical catalyst. Enzyme prices ranges from $ 100 per kilogram, as for crude preparations of amylase, to $ 100,000 per kilogram, as for lactic dehydrogenase. However, the key cost to consider in biocatalysis should be not the cost of the enzyme itself, but rather the cost-contribution of the final product. This cost contribution could be as low as $ 0.10 per kilogram as in the case of aspartase in the L-aspartic acid production. When compared with the cost of other catalyst, especially those with similar selectivity, the prices of enzymes are not very different (Table 1). Still the industrial use of enzymatic catalysts is limited by their instability under harsh conditions, which are usually found in large-scale processes. Nevertheless, chemical and genetic modifications of enzymes to improve both activity and stability, together with solvent engineering and new catalytic activities from extremophiles microorganisms will provide better biocatalysts for the specific needs of the petroleum industry. Non-conventional uses of enzymatic transformation are still in their infancy. Non-aqueous systems, high temperatures and hydrophobic substrates are the three main characteristics of oil industry that represent the most important challenges for the enzymatic catalysis to be applied in the petroleum refining industry. The success of biocatalysis in the petroleum industry depends on the development of biocatalysts able to perform transformations of oil products in non-aqueous systems and stable under the conditions usually found in the refineries. Table 1

Bulk enzyme and chemical catalyst pricesa.

a

Adapted from Rozzell [8].

70

Large amounts of water are incompatible with the refining processes, as demonstrated in the case of microbial desulfurization process (see chapter 2) [9]. Since petroleum is a hydrophobic material, it is suitable to speculate that new enzymatic processes for the oil industry should be carried out in non-aqueous systems. The use of reaction mixtures containing organic solvents reduces mass transfer limitations, promoting the establishment of productive interactions between the enzyme and the hydrophobic substrates (oil-derived compounds). Fortunately, it is possible to have enzymatic activity in non-aqueous systems with very low water content, almost anhydrous. Biocatalysis in non-aqueous media has increased significantly the range of practical applications of enzymes [10]. The abundant information on enzymatic activity in hydrophobic solvents, such as hexane, toluene and many other organic solvents has been extensively reviewed [10, 11], so it is reasonable to expect enzymatic activity in petroleum or petroleum fractions. In addition, a biocatalyst placed in a non-aqueous medium shows interesting properties, such as improved thermostability, higher substrate accessibility, adjustable selectivity, and high storage stability [12, 13]. Enzymatic reactions can be performed at more than 120°C in organic solvents, even if the enzyme is not thermostable in aqueous media [14]. Any enhancement of the thermal stability of an enzyme would confer significant operational advantages such as higher reaction rates, increased substrate solubility in lower viscosity media, productive shifts in thermodynamic equilibrium and reduced risks of microbial contamination. Genetic approaches have yielded significant results in obtaining more thermostable enzymes, with higher temperature optima through directed evolution [15-17]. A combination of rational and random mutagenesis has been used to obtain a fungal peroxidase 174-times more thermostable than the wild type protein [18]. However, although directed evolution seems to be a powerful tool to enhance biocatalytic performance, its main drawback is that the knowledge of the properties gained during site directed mutagenesis or evolution processes can not be used as a general method to be applied to other proteins. On the other hand, chemical modification seems to be a more general method to improve intrinsic properties of proteins such as stability and activity without a deeper knowledge of the gene or protein structure. Several chemical methods have been employed to obtain more stable protein derivatives, including plastic conjugates [19, 20], crosslinked enzyme crystals (CLECs) [21-24], attachment to polysaccharides [25, 26], and chemical modification with amphiphilic polymers [27-30]. Catalytically active and stable enzymes at temperatures higher that 80°C are considered hyperthermophilic enzymes. Only three enzymatic preparations have been shown to be active and stable at temperatures higher than 100°C in aqueous systems, two hydrolytic enzymes: a pegylated trypsin [27] and a dextran-glycosylated amylase [26], and a pegylated cytochrome c performing peroxidase-like catalysis [30].

71

This work is a prospective analysis of data on enzymatic transformations of petroleum products and their derivatives, in order to evaluate the possible impact of biotechnological processes on the petroleum refining. This prospective analysis emphasizes enzymatic biodesulfurization and enzymatic asphaltene upgrading, but promising carbon-carbon bond enzymatic activation is also discussed. 2. DESULFURIZATION Petroleum contains sulfur compounds, which structure and concentration varies depending on the crude oil source. The combustion of these compounds produces sulfur oxides that have a negative environmental impact. Sulfur oxides combined with water in the atmosphere are the principal source of acid rain; moreover, these oxides poison catalytic converters in cars, leading to increased hydrocarbons emissions to the air. In order to reduce the environmental damage caused by these oxides many countries have regulated its release, for example by lowering the sulfur content in fuels [31]. Currently, sulfur content is reduced via hydrodesulfurization, a chemical process in which organic sulfur is converted to hydrogen sulfide in the presence of an inorganic catalyst. During this process, crude oil is reacted with hydrogen at high pressures (150 to 3000 psi) and high temperatures (290° to 455°C). While the sulfur in thiols, sulfides and thiophenes present in the lighter fractions of crude oil are readily removed by hydrodesulfurization, in the heavier fractions a significant amount of sulfur is present within polynuclear aromatic molecules such as benzothiophenes (BT) and dibenzothiophenes (DBT). These larger and more complex molecules, particularly those with alkyl substitutions near the sulfur atom, are not easily hydrodesulfurized [9]. Regulatory agencies require today around 500 ppm of sulfur in fuels, a specification that can be accomplished with the current desulfurization technology. However, by the end of this decade the required sulfur content in fuels is expected to be less than 15 ppm [31]. As the availability of light oil decreases, there is a need to process heavier oil. Energy and capital costs of hydrotreating would substantially increase in order to achieve a deep level of desulfurization. Therefore, there is a significant interest in low-cost desulfurization technologies that might complement hydrotreating. The fact that some developing technologies, such as oxidation or acid treatment, promote secondary reactions affecting other components in crude oil is a drawback to its use as a desulfurization strategy [32]. Amongst the more promising alternatives is biodesulfurization, due to its low-energy requirements and high specificity [9, 33].

72

2.1 Microbial desulfurization Microbial desulfurization is a process based on the removal of organic sulfur from petroleum mainly by bacterial cell systems. After three decades of efforts focused on the use of metabolically active bacterial cells for fuel desulfurization, and even pilot plant trails, the process shows to be limited by several factors: large amounts of water needed for the process, mass transfer limitations in the two phase reactions, and large time of batch reaction [7, 9]. This subject is extensively reviewed in chapter 2. In order to develop a commercially feasible technology, any designed biocatalyst must perform under the conditions found in refinery processes. A major limitation for large-scale application of available biocatalysts is their low activity and stability. The complexity of the microbial process for desulfurization, involving several enzymes and cofactors, seems to make the use of whole cells the only choice. However, the use of whole cells for desulfurization implies some bottlenecks such as mass transfer problems, product inhibition and metabolic repression. Some of these problems have been addressed by genetic manipulation of the system [34]. Hydrodesulfurization process removes from 70 to 3000 mg of sulfur (g catalyst)"1 h'1 [35], so that enzymes that are both more active and recognize a larger set of substrates are needed. Novel genetic strategies such as gene shuffling might provide such enzymes [36]. Furthermore, nearly all described microbial desulfurization processes take place by mixing a cell suspension with oil [37,38]. It is desirable to avoid the formation of a stable water-in-oil emulsion in order to facilitate oil recovery, so that the oil/water volumetric ratio should be carefully adjusted. Otherwise, the large-scale energetic cost of separating the multiphase system could strongly impact the economics of the whole process. Recently an intermittent process with immobilized cells was described by the Petroleum Energy Center of Japan [39]. By entrapping cells of Rhodococcus erythropolis KA2-5-1, it was possible to devise a two-phase system (immobilized cells and oil) to desulfurize a model oil containing 100 ppm of DBT. According to this report, the biocatalyst was easily recovered, reactivated and reused during 900 h. However, the desulfurizing activity in the two-phase system was lower than in the three-phase system, probably due to interference of the support with the diffusion of substrates and products. 2.2 Enzymatic desulfurization Besides the microbial option, enzymatic desulfurization represents a promising alternative for biotechnological processes applied to the oil industry. The advantages and disadvantages of enzymatic desulfurization when compared with metabolic desulfurization are shown in Table 2. On one hand, the use of microbes requires the maintenance of the entire cellular machinery in order to regenerate the cofactors and carry out desulfurization. Furthermore microbes

73

need an aqueous phase to accomplish sulfur removal, while enzymes are able to function in media containing very low water content. Thermodynamic water activity (aw) influences both enzyme activity and stability, as water acts as a lubricant altering the flexibility of enzyme molecules. Protein mobility and therefore protein unfolding is restrained in a low water content medium. However, a certain amount of protein-bound water is essential to allow enough molecules flexibility to execute catalysis [40]. Thus it is possible to optimize enzyme performance in hydrophobic media by controlling aw [10, 41-42]. In addition, it has been shown that in certain organic media enzymes are active and more thermostable than in aqueous media [12, 13, 43], and it is possible to perform enzymatic transformations at temperatures higher than 100°C. Although sulfur elimination might not be achieved by a single enzymatic step, the enzyme-mediated transformation of sulfur-containing compounds may facilitate its removal. An enzymatic procedure to reduce the sulfur content from straight-run diesel has been described [44]. A fungal chloroperoxidase from Caldariomyces fumago was able to oxidize the sulfur-containing fraction of untreated diesel containing 1.6% sulfur, in the presence of low concentrations of hydrogen peroxide. Figure 2 shows gas chromatograms with both Flame Ionization (FID, general) and Flame Photometric (FPD, sulfur selective) detectors. The distribution of compounds in straight-run diesel fuel before and after oxidation with chloroperoxidase, are shown in panel a and b, respectively. The oxidation is clearly detected by the increase of boiling point (retention time) of these compounds on the gas chromatogram monitored by the sulfur selective detector (FPD). The higher boiling point of the oxidized compounds allowed its removal by a distillation step. Microdistillation of both chloroperoxidase-oxidized and untreated diesel fuels monitored by FID and FPD (Fig. 3) shows that the hydrocarbon distillation profile changes slightly after enzymatic treatment. In contrast, the sulfur selective detector (FPD) shows a significant change of the distillation profile, in which most of organosulfur compounds were effectively oxidized and their boiling points increased after enzymatic treatment. Table 2 Process characteristics of enzymatic and metabolic desulfurization. Enzymatic desulfurization Activity in low water systems Activity at temperature higher than 100°C Activity in toxic systems Activity only on organosulfur compounds Life-time depending on molecule stability

Metabolic desulfurization Activity in aqueous phase Inactivation at high temperature Sensitive to toxics Needs carbon source Self-producing catalytic system

74

Fig. 2. Gas chromatograms of straight ran diesel fuel before (a) and after (b) chloroperoxidase treatment [44].

Oxidized sulfur compounds can then be removed by a distillation process (Table 3). After distillation, the sulfur content in the enzymatically oxidized diesel fuel is only 0.27%, while for the untreated fuel is 1.27%. The distillation of the straight-run diesel fuel (1.6% sulfur) to a final distillation point of 325°C produced a distillate containing 66% of the total sulfur, while if the diesel fuel was previously oxidized with chloroperoxidase, the obtained distillate contained only 12% of the total sulfur. Thus, by using an enzymatic oxidation with chloroperoxidase coupled with a distillation process it is possible to obtain a diesel fuel with 6-times lower sulfur concentration than straight-run diesel fuel. Few hydrocarbons are transformed during the enzymatic treatment, and after distillation an additional 12% of them remain in the residue (Table 3).

75

Fig. 3. Microdistillation of both chloroperoxidase-oxidized and untreated diesel fuels monitored by FID (general detection) and FPD (sulfur selective detection) [45].

Chloroperoxidase was able to oxidize a wide range of sulfides, benzothiophene and dibenzothiophene [45]. However, a drawback for this procedure is the potential modification of aromatic hydrocarbons due to the enzymatic treatment. The presence of chlorinated hydrocarbons was detected when individual reactions were performed with chloroperoxidase [46]. The generation and combustion of such compounds is environmentally undesirable, as it would deteriorate the fuel value and the air quality.

Table 3 Sulfur content of straight-run diesel fuel after enzymatic oxidation with chloroperoxidase from Caldariomyces fumago followed by a distillation to 325°C as final distillation point. Distillation TPH (%) Sulfur (%) Destillate 83 1.27 Residue 17 3.21 TPH. Total petroleum hydrocarbons.

Enzymatic -H distillation TPH (%) Sulfur(%) 71 0.27 29 5.51

76

Nevertheless it is expected that in a complex mixture, sulfur compounds would be preferentially oxidized due to the higher affinity and activity of chloroperoxidase towards these substrates [46]. Other enzymes are known to catalyze sulfoxidation, but chloroperoxidase shows higher activity and broader specificity [47-50]. Some of the sulfur-containing substrates transformed by chloroperoxidase are listed in Table 4. Enzymatic desulfurization shares some of the challenges of microbial desulfurization: the stability and activity of the biocatalyst must be appropriate in order to develop a commercially competitive process. Several factors influence these properties. One of them is enzyme preparation. Usually enzymes are immobilized on an inert support in order to avoid product contamination and simplify biocatalyst recovery. Immobilization might improve stability of the enzyme by multiple point attachment [21]. Moreover, the chemical and physical characteristics of the support determine the microenvironment of immobilized enzymes, therefore altering its activity as well as the diffusion of substrates and products [51]. Another factor influencing biocatalysts performance is water content in nonaqueous media. As mentioned earlier, the thermodynamic water activity influences both enzyme activity and stability in hydrophobic media. Depending on the solvent system and on the enzyme, optimal water content must be found in order to enhance enzyme performance [52]. It is know that some polar organic solvents interact unfavorably with proteins, displacing water molecules that may be essential for activity and accelerating protein unfolding [53, 54]. Hence sometimes enzyme folding is better preserved in the presence of nonpolar organic solvents and even thermostability might improve. Finally, substrate partition between bulk solvent and the active site of the enzyme may limit the reaction rate. It has been demonstrated that the partition of hydrophobic substrates to the active site of an enzyme is less favorable in the presence of organic solvents [55, 56]. The KM value offers an indication of the unfavorable partition since it increases with the organic solvent content [57]. Though stabilization and immobilization of chloroperoxidase for use in organic solvents have been accomplished, substrate availability needs additional study. Thermodynamic understanding of substrate partition and the influence of active site environment might provide the basis for solving this problem. The most important challenge for the scale-up of the enzymatic process using chloroperoxidase is, doubtless, the stability of the enzyme against the peroxide inactivation. The present and future commercial uses of peroxidases have been limited, mainly, by the low stability of peroxidases in the presence of their natural substrate, hydrogen peroxide. All hemeproteins, including peroxidases, are inactivated in the presence of catalytic concentrations of hydrogen peroxide [58]. This inactivation process is specially important in the absence of reducing substrates and its mechanism has not been fully elucidated.

77

Table 4 Sulfur-containing substrates of chloroperoxidase.

In order to overcome the described shortcomings, several tools may be used to enhance enzyme desulfurization, such as protein engineering, solvent engineering and immobilization procedures. Biocatalyst design should be conducted contemplating parallel increments in activity and stability. It has been observed that improvement of one characteristic might result detrimental for other biocatalyst properties [18]. Thus by selecting for simultaneous enhancements of different properties, more suitable biocatalysts might be obtained [36, 18].

78

Thus, enzymatic biodesulfurization is a promising alternative to achieve low sulfur levels in oil-derived streams. Some enzymes able to perform modifications on sulfur-containing compounds that facilitate its removal have been identified. It has been estimated that refineries investment for achieving deep desulfurization would be considerably reduced (by 50%) if combined hydrodesulfurization-biodesulfurization processes are implemented [9]. While microbial desulfurization has been well studied for the last ten years and significant advances have been achieved, enzymatic desulfurization is a less explored field with great potential as it lacks some of the important drawbacks of microbial desulfurization. In order to represent an economically viable alternative, biodesulfurization processes must adequate to the conditions found in a refinery. Specifically, both activity and stability of current biocatalyst must be enhanced. Additional effort must be done to design an appropriate biocatalyst considering all stages of the process. Importantly, preparation and recuperation strategies of biocatalyst should not be neglected. 3. ENZYMATIC TRANSFORMATION OF ASPHALTENES Asphaltenic and viscous heavy oils from bituminous deposits are a huge energy reserve to be exploited in next decades. More than 70 countries possess bituminous deposits. In Canada only, the oil reserve considered to be technically recoverable is estimated to be 280-300 Gb (billion of barrels), larger than the Saudi Arabia oil reserves estimated at 240 Gb [59]. These highly asphaltenic resources must be rigorously treated in order to convert them into an upgraded crude oil before them can be refined to produce gasoline and other fuels. Asphaltene, the highest molecular weight fraction of petroleum, is a dark amorphous solid specially rich in heteroatoms (S, O, N), and metals (Fe, Ni, V) [60-62]. Many problems associated with either recovery, separation or processing of heavy oils and bitumens, are related to the presence of high concentration of asphaltenes. This fraction is thought to be largely responsible for other adverse oil properties such as high viscosity and the propensity to form emulsions, polymers and coke. The molecular structure of asphaltenes has been an enigma for seven decades [62]. From numerous investigations there are indications that asphaltenes are condensed aromatic cores containing alkyl and alicyclic moieties. Heteroatoms, such as nitrogen, oxygen and sulfur are present as non- and heterocyclic groups. A significant amount of porphyrins (petroporphyrins) can be found containing mainly nickel and vanadium. A hypothetical asphaltene molecule is shown in Fig. 4. The complexity of the asphaltene chemical nature is evident by the difficulty of analysis of both their molecular weight and structure. The asphaltenic fraction is recognized as the most recalcitrant fraction of oil. So far, there is no clear evidence that asphaltenes are degraded by microbial

79

activity (see chapters 1 and 4). Some reports on oil biodegradation claim the degradation of asphaltenic fraction by mixed bacteria [63, 64]. However, none of these reports described the analytical results of extractable materials recovered from appropriate sterile controls. On the other hand, although microorganisms have been found associated with bitumens containing high amounts of asphaltenes [65], the asphaltenic fraction did not support bacterial growth and no changes in asphaltene content could be found after bioconversion of heavy oils and asphaltenes [66, 67]. Because the asphaltene content was usually determined gravimetrically after n-alkane precipitation, the reported changes could be attributed to the disruption of the asphaltenic matrix by the production of surfactants during bacterial growth, liberating trapped hydrocarbons. Therefore, most of the asphaltene losses during microbial activity could be considered to be abiotic losses [68]. Nevertheless, a clear experimental evidence that enzymes are able to modify asphaltene molecules has been reported [69]. Chloroperoxidase from the fungus Caldariomyces fumago was able to transform petroporphyrins and asphaltenes, and this modification was significantly higher in systems containing organic solvent than in aqueous systems [69, 70]. Asphaltenes and petroporphyrins are highly hydrophobic materials, thus mass transfer limitations are expected in aqueous reactions. The biocatalytic oxidation of a petroporphyrin rich-fraction of asphaltenes in the ternary solvent system and in the presence of hydrogen peroxide was performed. Chloroperoxidase catalyzed reaction produced notable spectral changes in the petroporphyrin rich-fraction of asphaltenes (Fig. 5). The destruction of petroporphyrins by chloroperoxidase in the presence of hydrogen peroxide leads to the removal of Ni and V from asphaltene molecules, as in the case of synthetic nickel and vanadium porphyrins (Table 5). On the other hand, a doubly modified cytochrome c (PEG-Cyt-Met) was able to catalyze the oxidation of a petroporphyrin rich-fraction of asphaltenes in the ternary solvent system and in the presence of 100 mM of tert-butyl hydroperoxide [71]. As chloroperoxidase, the PEG-Cyt-Met catalyzed reaction produced spectral changes in the petroporphyrin rich-fraction of asphaltenes (Fig. 5). The oxidative porphyrin ring disruption entails the simultaneously release of metal. The biocatalytic process with PEG-Cyt-Met removed 95% of the vanadium and 74 % of the nickel (Table 5). The destruction of the petroporphyrin molecules is conformed by the Soret band loss and metal removal.

80

Fig. 4. Asphaltene molecule proposed by Strausz et al. [62].

Figure 5. Absorption spectra of the petroporphyrin rich-fraction of asphaltenes after biocatalytic treatment. Control without treatment (a); control without biocatalyst and in the presence of hydroperoxide (b); reaction with one addition of biocatalyst (c); and reaction after a second addition of biocatalyst (d). [69, 71].

81

Table 5 Nickel and Vanadium removal from petroporphyrin rich fractions of asphaltenes by chloroperoxidase-mediated reaction. Heavy metal Nickel Vanadium

Chloroperoxidase [69]

Chloroperoxidase [70]

20% 19%

57% 52%

Chemically modified cytochrome c [71] 74% 95%

Petroporphyrin-rich fractions from asphaltenes with and without biocatalytic treatement were analyzed by Fourier transform infrared spectroscopy (FTIR) (Fig. 6). Significant differences could be detected mainly as an increased proportion of oxygen containing groups, such as hydroxyl (3310 cm"1), carboxyl (1300 cm"1, 1770 cm"', 1710 cm"1), aldehydes (1730 cm"'), sulfoxides (1040 cm"1), sulfones (1130 cm"1), and sulfonates (1160 cm"1, 1260 cm"1). Sulfur is, after carbon, the most important element in asphaltene molecules (Fig. 4), and most of it is contained in thiophenes and organic sulfides moieties. According to the FTIR spectrum, PEG-Cyt-Met also catalyzes the oxidation of carbon atoms from asphaltenes molecules. Biocatalytic cracking, or biocracking, is probably the most interesting biotechnology target for heavy oil upgrading (see chapter 4). Gel permeation chromatography (GPC) monitored by a diode array detector of both untreated and biocatalytically oxidized petroporphyrin are shown in Fig. 7 [71]. The isoabsorbance contours corresponding to the Soret band (maximal absorbance at 403 nm and elution time of 10.77 min) disappeared after the biocatalytic oxidation as expected, and in agreement with the spectra shown in Fig. 6. Nevertheless, gel permeation chromatograms show higher molecular weight distribution of oxidized petroporphyrin than the control distribution. However, these results should be taken cautiously because the oxidation process, which introduces polar groups in the molecules, may affect the asphaltenes aggregation state. Asphaltenes are a very complex mixture and are defined only by their solubility properties: the asphaltenic fraction is insoluble in short-chain nalkanes, specially pentane. As mentioned above, due to their complexity, the order of magnitude of asphaltene molecular weight is still controversial [72, 73, 74]. The 'H NMR analysis of control and treated samples showed the following regions (Fig. 8): Hy (y+ CH3) from 0.5 to 1.0 ppm; Hp (P+ CH2): HR from 1.0 to 1.6 ppm, HN from 1.6 to 2.0 ppm; H a (a CH2) from 2.0 to 4.0 ppm and Har (CH aromatic) from 6.0 to 9.0 ppm. Here, HR and HN refers to P-protons in aliphatic chains and naphthenic rings, respectively. The spectrum from untreated fraction (control) was similar to those found in other asphaltenes samples [76]. However,

82

the spectra from both untreated and treated samples showed an important 5.29 ppm signal, probably due to protons from non aromatic (C = C) double bonds, which are not detectable in whole asphaltenes fractions. The main differences of enzyme treated samples when compared with the untreated fraction appeared in the saturated hydrocarbon region: a quartet placed on 2.28 ppm, a triplet placed on 2.49 ppm, and a singlet placed on 3.63 ppm. These signals seem to be originated from the hydrocarbon chains of polar compounds, may be from oxygenated compounds as 13C spectrum shows (see below). The singlet shift can be attributed to ether or alcohol groups. The ester-amide signal (4.3-4.36 ppm) was very important in the oxidized sample, while was minor in the control petroporphyrins. The 13C NMR analysis showed the presence of 58.78 ppm and 46.11 ppm shifts in the control, which are attributed to the C-N bond (Fig. 9). These signals disappeared in the oxidized petroporphyrins. Signals between 10 ppm and 60 ppm are usually assigned to the hydrocarbon chains. The NMR spectrum from oxidized petroporphyrins showed a more intense terminal methyl (-CH3) signal than in the untreated sample. The (-CH2-) / (-CH3) intensities ratio was lower in the treated asphaltenes fraction than in the untreated ones, which could be attributed to the presence of shorter alkyl chains or more branched chains. Thus, this lower ratio could be the consequence of molecule cracking. The aromatic region of the spectra (110 ppm to 160 ppm) showed significant differences. The control showed a signal-hill between 133 ppm and 146 ppm, which include the carbon atoms corresponding to heteroatom moieties (S, N, O), aromatic carbons bonded to alkyl moieties, and aromatic carbons bonded to other aromatic carbon. This signal-hill disappeared in the oxidized fraction, suggesting a loss of heteroatoms or alkyl derivatives in the aromatic molecules. A reduction in the number of substituted aromatic carbons and an increase of the number of aromatic carbons bonded to hydrogen are observed. The enzymatic treatment of asphaltenes is an interesting alternative for the removal of heavy metals in order to reduce catalyst poisoning in hydrotreatment and cracking processes. On the other hand, enzymatic cracking of asphaltenes molecules should not be excluded. The enormous amount of energetic resource found as asphaltenes-rich deposits justify the exploration of alternative upgrading technologies.

83

Fig. 6. FTIR spectra of untreated and biocatalytically treated porphyrin-rich fractions from asphaltenes. FTER was performed using the film-spreading technique [71].

Fig. 7. Absorbance contours from gel permeation chromatography (GPC) of untreated and biocatalytically treated porphyrin-rich fractions from asphaltenes [71].

84

Fig. 8. H NMR analysis of control and enzymatically treated petroporphyrin-rich fraction of asphaltenes.

Fig. 9. C NMR analysis of control and enzymatically treated petroporphyrin-rich fraction of asphaltenes.

85

4. OXIDATION OF AROMATIC HYDROCARBONS The ability of microorganisms (bacteria and fungi) to modify polycyclic aromatic hydrocarbons (PAHs) by oxidation is well known, and a number of comprehensive reviews have been written on the microbial metabolism of PAHs [75-79]. The controlled partial oxidation of aromatic hydrocarbons by molecular oxygen (dioxygen) is both highly desirable, from an environmental point of view, and difficult [80]. Oxidation of organic matter to carbon dioxide and water by dioxygen is a thermodynamically highly favourable process. Fortunately for biological systems, the kinetic barrier is large. The main problem in biological systems is how to promote the reaction whilst at the same time limiting the damage caused by indiscriminate attack of dioxygen. Nature achieves this by using metalloenzymes, many of which contains iron porphyrin groups (hemoenzymes) as active-site responsible for activating the dioxygen. The next section is an overview focused on the biocatalytic oxidation of PAHs in vitro by using enzymatic and non enzymatic proteins. 4.1. Lignin peroxidase Lignin is the most abundant renewable organic material next to cellulose and it is an aromatic polymer. Lignin is mainly decomposed by higher basidiomycetous fungi that cause the white-rot wood decay. Because insolubility and complexity of this substrate, ligninolytic microorganisms have evolved to secrete multiple lignin peroxidase enzymes. These enzymes act as non-specific, diffusable oxidative catalysts that serve to degrade lignin. Lignin peroxidases from Phanerochaete chrysosporium are the most extensively studied. Twelve years ago, Sanglard et al. [81] reported the first evidence of enzymatic oxidation of benzo(a)pyrene by lignin peroxidase. The reaction mixture contained crude lignin peroxidase and a H2O2 generating system with glucose oxidase and glucose. Purified enzyme is able to oxidize PAHs with ionization potential (IP) lower than 8 eV in the presence of H2O2 [82, 83]. In addition, the specific activity was correlated with the IP of PAHs (Fig. 10). In most of cases, oxidation products were identified as quinones, although hydroxylated compounds were also detected (Table 6). Lignin peroxidases may also participate on the oxidation of aromatic intermediates. Phenanthrene (IP = 8.03 eV) is not substrate for lignin peroxidase. Nevertheless, its oxidized metabolite 9-phenanthrol is transformed to phenanthrene-9,10-quinone by lignin peroxidase in vitro [86]. Lignin peroxidase is able to catalyze other environmentally high-risk compounds, such as polychlorinated phenols [87], sulphur and nitrogen organocompounds [85, 88], and industrial dyes [89, 90].

86

Fig. 10. The influence of ionization potential of PAHs on the specific activity of lignin peroxidase oxidation [82].

4.2. Manganese peroxidase Ligninolytic microorganisms also produce extracelluar manganesedependent peroxidases. As in the case of lignin peroxidases, manganese peroxidases are a family of isoenzymes produced by ligninolytic fungi [91, 92]. Manganese peroxidases are heme glycoproteins that require Mn(II) for its activity. Mn(II) is oxidized to Mn(III), which behaves as a low-molecular-weight mediator that diffuse to remote regions of the lignin molecule and initiate the oxidation process. The oxidation of Mn(II) to Mn(III) is dependent on the presence of chelating agents, such as lactate, succinate or malonate. This fungal peroxidase is, so far, the only known enzyme system that utilizes soluble Mn(II)/Mn(III) as an obligatory redox couple. Manganese peroxidase is able to oxidize some xenobiotics including PAHs [93]. Enzymatically generated Mn(III) oxidize low redox potential methoxybencenes during lignin degradation [94]. During lipid peroxidation, this enzyme is able to oxidize phenanthrene [95], which has an IP higher than 8 eV and is not oxidized by lignin peroxidase [85]. The slow oxidation of phenanthrene to 2,2'-diphenic acid supported by manganese peroxidase requires Mn(II), oxygen, and unsaturated lipids. Fluorene, a PAH which again is not a substrate for lignin peroxidase, is oxidized by manganese peroxidase in a lipid peroxidationdependent reaction [96]. The product of fluorene is 9-hydroxyfiuorene via 9fluorenone, and this reaction is inhibited by free-radicals scavengers. Manganese peroxidase could also be involved in the oxidation of another PAH with high IP, chrysene [97]. Table 7 shows the slow oxidation of PAHs in vitro with manganese peroxidase from Phanerochaete chrysosporium during lipid peroxidation.

87

Table 6 Products from in vitro oxidation of polycyclic aromatic hydrocarbons by lignin peroxidase from Phanerochaete chrysosporium. PAH Anthracene Acenaphthene 1 -Methylanthracene 2-Methylanthracene 9-Methylanthracene

Pyrene Benzo(a)pyrene

Products 9,10- Anthraquinone 1-Acenaphthenol 1 - Acenaphthenone 1 -Methyl-9,10-anthraquinone 2-Methyl-9,10-anthraquinone 9-Methyl-9,10-anthraquinone 9-Methyl-10-anthranone 9,10-Anthraquinone 1,8-Pyrenedione 1,6-Pyrenedione 1,6-Benzo(a)pyrenedione 3,6-Benzo(a)pyrenedione 6,12-Benzo(a)pyrenedione

Ref. 83,84

83 83 83 83 83 83 83 83,84

82 87 87 87

PAH oxidation by this lipid peroxidation-mediated system with manganese peroxidase showed to be significantly slower than the lignin peroxidase oxidation. Nevertheless, manganese-dependent lipid peroxidation with extracelluar extracts from Phanerochaete chrysosporium oxidize PAH in complex mixtures, such as creosote, and the oxidation rate is correlated with the ionization potential [98]. In addition the catalytic activity seems to be very stable and constant activity could be found during more than 48 hours. Extracellular crude extracts and semipurified manganese peroxidase from Bjerkandera sp. are able to transform anthracene to anthraquinone [83, 99]. Ligninolytic fungi Phanerochaete leavis HHB-1625 which produces high levels of manganese peroxidase and no lignin peroxidase, is able to transform PAH in liquid culture [100]. Extracelluar extract from this strain was able to oxidize anthracene, phenanthrene, benz(a)anthracene, and benzo(a)pyrene in the presence of Mn(II) and hydrogen peroxide. 4.3. Versatile peroxidase Active lignin-degrading strains of Pleurotus eryngii were shown to produce a peroxidase, different from P. chrysosporium peroxidases, that can efficiently oxidize Mn(II) to Mn(III), but can also carry out Mn(II)-independent activity on aromatic substrates [101]. This novel manganese-lignin peroxidase hybrid enzyme, now called versatile peroxidase (VP), was also described in Bjerkandera sp. BOS55. This enzyme is able to oxidize various phenolic and nonphenolic substrates such as 2,6-dimethoxyphenol, guaiacol, ABTS, and veratryl alcohol, in the absence of Mn(II) [102], Similar hybrid peroxidases have been reported in Pleurotus eryngii [103-106], Pleurotus pulmonarius [107],

88

Pleurotus ostreatus [108], as well as in Bjerkandera adusta [102, 103, 109, 110]. In the VP from B. adusta the oxidation of Mn(II) to Mn(III) proceeds optimally at pH 4.5, while the LiP-like activity requires more acidic conditions, showing maximum rates at pH 3.0 [104]. VP seems to have a long range electron transfer pathway similar to that postulated for LiP [111]. The spectroscopic characterization of VP by EPR and electronic absorption techniques showed a protein centered radical in the presence of an excess of hydrogen peroxide which was assumed to be a tryptophanyl radical [112]. In addition the enzyme shows high identity with LiP (58-60%) and MnP [55%] from Phanerochaete chrysosporium [91]. The heterologous expression of VP in Aspergillus nidulans confirmed the ability of this hybrid enzyme to oxidize both Mn(II) and also different aromatic compounds in the absence of manganese [106]. The ability of the VP to oxidize PAHs was recently reported by Wang et al [113]. Oxidation of PAHs was examined by a purified VP isoenzyme in the presence and absence of Mn(II). PAH oxidation was reduced by the presence of Mn(II) and the inhibition kinetics were shown to be partially noncompetitive. The substrates were anthracene and its methyl derivatives, pyrene and benzo(a)pyrene, with IP of 7.43 eV or lower (Table 8). The PAH metabolites of the Mn-independent reaction were identified as the corresponding quinones. The pH optimum of the Mn-independent oxidation was around pH 4, while for the Mn-dependent reaction it was pH 3. The kinetic constants for the Mnindependent oxidation of 2-methylanthracene at pH 4 were determined, and the values we obtained were a kcat of 145 min"1, KM,apP for the aromatic substrate of 23.8 mM, and KM,app for hydrogen peroxide of 0.2 mM. 4.4. Cytochromes P450 Cytochromes P450 form a superfamily of hemoenzymes that were originally named as a pigment having maximum absorbance at 450 nm in the presence of CO and unknown function. Biologically occurring substrates for cytochrome P450 include fatty acids, steroids, eicosanoids, lipid hydroperoxides, retinoids, arginine, acetone and acetol. Interestingly, there is a very large number of xenobiotic compounds that are substrates for cytochrome P450. These xenobiotic substrates includes drugs, procarcinogens, antioxidants, solvents, anesthesics, dyes, pesticides, petroleum products, alcohols, flavorants and odorants. Ironically, in addition to its beneficial roles in metabolism, biosynthesis and detoxification, cytochromes P450 are implicated as the activators of many chemical carcinogens.

89

Table 7 In vitro PAHs oxidation with manganese peroxidase in lipid peroxidation reactions [96]. . , Aromatic compound r Fluorene Benz(a)antrhracene Benzo(a)pyrene Anthracene Dibenz(a,c)anthracene Benzo(e)pyrene Diphenylmethane Benzo(c)phenanthrene Benzo(b)fluoranthene Fluoranthene Phenanthrene

Oxidation rate , ... (nmol/h) 3.10 1.08 0.96 0.93 0.60 0.31 0.30 0.21 0.19 0.14 0.06

Table 8 Oxidation of aromatic compounds by versatil peroxidase at pH 4.0 in the absence ofMn(II)[113]. Specific activity (min"1) PAH Ionization potential (eV)a 9-Methylanthracene 7.25 52 (±2.7) b C 1 -Methylanthracene NA 22.4±(1.7) 2-Methylanthracene 7.37 12.4 (±1.1) Anthracene 7.41 2.5 (+0.01) Benzo(a)pyrene 7.41 0.32 (±0.04) Pyrene 7.42 0.008 (±0.01) Benzo(e)pyrene 7.43 NRd Chrysene 7.60 NR Carbazol 7.68 2.4 (±0.05) 1 -Methylphenanthrene 7.70 NR Acenaphthene 7.76 NR Phenanthrene 7.91 NR Dibenzothiophene 7.93 NR Fluoranthene 7.95 NR Naphthalene 8.15 NR a Photoelectron spectroscopy values (http://webbook.nist.gov) b Values in parentheses are standard deviations. C NA, not available d NR, no reaction

90

Cytochromes P450 are widely distributed in living organisms and have been found, for instance, in mammals, fish, yeast, bacteria, and plants [114]. Cytochromes P450 are part of multienzymatic systems called monooxygenases which catalyze the activation of dioxygen and the transfer of one oxygen atom to substrates, with the simultaneous consumption of NAD(P)H. The monooxygenase cycle of cytochrome P450 has been well characterized [49]. It has been observed that cytochromes P450 are able to carry out oxidations with exogenous single-oxygen-atom donors like H2O2, alkyl-hydroperoxides, iodosobenzene, amine oxides, and peracids. These oxidations are observed in vitro with cytochrome P450 alone, without consumption of NAD(P)H. The oxidation of benzo(a)pyrene catalyzed by cytochrome P450 in the presence or absence of hydroperoxides yield different products. While the major products formed in the presence of cumene hydroperoxide are quinones [117], only phenols are formed with NAD(P)H in the absence of the hydroperoxide. After the first measurement of benzo(a)pyrene hydrolase activity based in a direct fluorometric method [118], several studies have been carried out in order to determine the kinetic constants of both microsomal and purified preparations [115, 116, 119, 120]. However, because the large number of P450 enzymes that can be present in a single organism, and because multiple species with distinct conformations and substrate recognition profiles coexist in a biological membrane, data form microsomal preparations should be considered cautiously. Interaction of these complex systems with PAHs could be resolved by using rapid kinetic technique [121]. Data from highly purified preparations from yeast with different single-oxygen-donors are showed in Table 9. Benzo(a)pyrene has been extensively used as substrate, nevertheless other PAHs are oxidized by P450 enzymes in vitro, such as anthracene and alkylanthracenes [122], phenanthrene [123], and the most potent carcinogen among all PAH, dibenzo(a,l)pyrene [124].

Table 9 Kinetic constants of purified cytochrome P448 cerevisiea for oxidation of benzo(a)pyrene [115, 116]. System Reconstituted with NADPH Cumene hydroperoxide Hvdrosen oeroxide in situ

kcat (min"1) 33 125 200

from

Saccharomyces

KM ( U M )

0.017 0.022 0.034

91

Fig. 11. Engineering cytochrome P450 BM-3 for oxidation of polycyclic aromatis hydrocarbons [127].

In order for cytochrome P450 to be an effective catalyst, the enzyme must efficiently bind substrates. Thermodynamic studies on the substrate binding to the active site of rat liver cytochrome P450 by using a series of aromatic hydrocarbons [125] showed that substrate hydrophobicity is an important driving force that determines substrate affinity. The predominant force involved in binding is the ability of the active site to draw the aromatic hydrocarbon from the aqueous phase [126]. Using hydrophobic interaction analysis to design a new biocatalyst, sitedirected mutagenesis has been used to modify a bacterial cytochrome P450 [128]. This was the first report of rotational redesign of cytochrome P450 in which by changing only one active-site residue the affinity for different substrates was changed. The modified substrate pocket allowed tight binding of a novel substrate, diphenylmethane. Heme domain also influences substrate affinity, as showed when native prosthetic group is replaced by an heme dimethyl ester [129]. Esterification of the heme propionates groups removes the negative charges from the vicinity of substrate-binding site, increasing its hydrophobic nature, and thus increasing the substrate affinity. Significant work has been performed on these enzymes using molecular techniques. Laboratory evolution of cytochrome P450 from Pseudomonas putida for peroxide-mediated hydroxylation of naphthalene has been performed [130]. The obtained mutants showed, in the absence of cofactors through the "peroxide shunt" pathway, more than 20-fold higher activity than the native enzyme for

92

naphthalene hydroxylation. Moreover, cytochrome P450 has been enginnered into a catalyst for the oxidation of PAHs [127, 131]. Compared with the activities of wild type, those of the mutants improved by up to 4 orders of magnitude (Fig. 11). 4.5. Cytochrome c Cytochromes c are part of the energy-conserving electron transport systems. In living systems no catalytic activity of cytochrome c has been described. The ability of cytochrome c to act as catalyst in vitro has been reviewed [132]. Lipid peroxidation, hydroperoxide cleavage, hydroxylation of 4nitrophenol and oxidation of 2-keto-4-thiomethyl butyric acid in the presence of hydrogen peroxide have been reported. Peroxidase activity of cytochrome c has been also demonstrated by the oxidation of various electron donors including ABTS (2-2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) and 4-aminoantipyrine. In addition, cytochrome P450-like oxidative reactions such as N- and O-demethylations, S-oxidations and olefin epoxidation are catalyzed by free and immobilized cytochrome c in the presence of hydrogen peroxide or other organic hydroperoxide [132, 133, 134]. It has also been observed that aromatic substrates of cytochrome c interact with the heme group as ligand rather than as a substrate [133, 135].

Table 10 Specific activity of yeast cytochrome c on aromatic compounds [135]. Aromatic compound Dibenzothiophene Anthracene Pyrene Benzothiophene Carbazole Acenaphthene Chrysene Fluoranthrene Fluorene Phenanthrene Triphenylene NR, no reaction detected.

Product Dibenzothiophene sulfoxide 9,10-Anthraquinone 1,8-Pyrenodione Benzothiophene sulfoxide Unknown

Specific activity (min"1) 3.2 2.1 1.3 1.0 0.9

(±0.1) (±0.1) (±0.3) (±0.2) (±0.1)

NR NR NR NR NR NR

93

Table 11 Kinetic constants of wild-type and variants of yeast iso-1cytochrome c for pyrene oxidation [135]. kcat

Variant Ala86;ThrlO2 Phe67;ThrlO2 Ala72;ThrlO2 Ala52;ThrlO2 Ala73;ThrlO2 Ala87;ThrlO2 Phe82;CyslO2(WT) Ala79;ThrlO2

(s'1) 0.17 0.10 0.13 0.18 0.28 0.39 0.31 3.28

K M ,app

kcat/KM,app

(mM) 4.0 3.3 4.0 4.7 7.5 3.9 9.7 101.8

(s"1 M"1)

33 32 33 39 38 99 32 32

Table 12 Oxidation of polycyclic aromatic hydrocarbon by unmodified- and methylated poly(ethylene)glycol-modified-cytochrome c [137].

Aromatic compound 7,12-Dimethylbenzanthracene 1,2:3,4-Dibenzanthracene Azulene 3-Methylcholanthrene 7-Methylbenzo(a)pyrene 1,2:5,6-Dibenz anthracene Triphenylene Dibenzothiophene Anthracene Thianthrene Pyrene Fluoranthene Acenaphthene Benzo(a)pyrene Fluorene Phenanthrene Chrysene 9,10-Dimethylanthracene Naphthalene Biphenyl NR. No reaction detected

Specific activity (min"1) PEG-Cyt-Met Unmodified 80.33 (±3.83) 24.59 (±1.52) 16.60 (±2.24) NR 14.32 (±0.57) 2.26 (±0.29) 10.96 (±0.54) 1.88 (±0.07) 7.56 (±0.42) NR 5.70 (±0.31) NR 5.27 (±1.05) NR 0.67 (±0.06) 4.73 (±0.05) 3.09 (±0.32) 0.33 (±0.06) 1.41 (±0.08) 0.49 (±0.06) 0.51 (±0.05) 0.97 (±0.03) 0.65 (±0.09) NR NR 0.40 (±0.01) 0.22 (±0.02) 0.39 (±0.06) 0.22 (±0.01) NR 0.17 (±0.02) NR NR NR NR NR NR NR NR NR

94

The first oxidation of an aromatic hydrocarbon with cytochrome c in the presence of hydroperoxide was reported by Akasaka et al. [136]. Hydroxylation of benzene was carried out in organic solvent with less than 5% of water, and with immobilized protein. Free cytochrome c was unable to perform this reaction. The capacity of yeast cytochrome c to perform oxidations of PAHs has also been reported [135]. Biocatalytic activities on 11 aromatic compounds were tested in a system containing 10% acetonitrile and 1 mM H2O2. The specific activities for the oxidation of these compounds are shown in Table 10. Anthracene, pyrene, dibenzothiophene, benzothiophene and carbazole were oxidized by the catalytic activity of yeast cytochrome c. No correlation between substrate IP and specific activity was observed when using cytochrome c as catalyst. Site-directed mutagenesis has been performed on yeast cytochrome c; it was observed that Phe82 substitution significantly altered the kinetic behaviour of the protein. The Gly82:ThrlO2 variant showed 10 times more catalytic activity and a ten-fold catalytic efficiency than the wild-type iso-1-cytochrome c [135]. For the oxidation of pyrene, the different variants of yeast cytochrome c showed different catalytic constants (Table 11). Lysine 79 residue is placed at the edge of the solvent access to the heme group, and its substitution by alanine produced a protein with higher kcat but also higher KM, resulting in similar catalytic efficiency. These results show that site-directed mutagenesis could be a tool for the design of a better biocatalyst for PAHs oxidation. In addition to genetic techniques, chemical modification has been performed on horse heart cytochrome c [137]. Free amino and carboxylic groups of horse heart cytochrome c were modified by chemical reaction with poly(ethylene)glycol (PEG) moieties. As a consequence of the chemical modification the heme environment in the active site was altered. Cytochrome c with a double modification: PEG on free amino groups and methyl esters on carboxylic groups (including propionates of heme), was able to oxidize 17 aromatic compounds from 20 tested, while the unmodified protein was only able to oxidize 8 compounds (Table 12). Thus, chemical modification of biocatalyst could be also a tool for the design of new biocatalyst with environmental proposes. As mentioned above, cytochromes c are very stable proteins, and it is possible to perform on it a large variety of chemical reaction without a negative effect on the activity. 4.6. Hemoglobin In the presence of hydrogen peroxide, hemoglobin has been reported to oxidize aniline [138], lipids [139], S- and N-heterocycles [134, 140, 141] and other organic substrates [141]. This protein could be considered as an antioxidant in red blood cells [142]. Biocatalytic activity of hemoglobin on PAHs has been tested with 12 compounds in the presence of hydrogen peroxide [143]. Among the aromatic compounds tested, 6 were oxidized (Table 13). As in the case of

95

cytochrome c, and in contrast with lignin peroxidase, no correlation between the extent of oxidation by hemoglobin and the ionization potential of the substrates was found. Interestingly, hemoglobin is able to oxidize fluorene, while with lignin peroxidase and cytochrome c no reaction was detected [85, 135]. Reaction products were identified as quinones, and are the same that those obtained with lignin peroxidase and cytochrome c. The product from carbazole oxidation was not reported; however, this product could be a polymer such as in oxidation with lignin peroxidase [88]. The catalytic mechanism of hemoglobin seems to be similar to that of other hemoproteins. Early studies using ESR trapping techniques showed free radical involvement in the oxidative reactions by organic hydroperoxides and erythrocytes [144]. ESR studies, as for cytochrome c, detected peroxyl and alkoxyl radicals produced by reaction of hydroperoxides and hemoglobin [145]. These experiments suggested that the formation of free radicals involves highvalence-state iron complexes [139]; molecular oxygen could be involved, in part, in the oxidation reactions [141]. 4.7. Chloroperoxidase Chloroperoxidase from Caldariomyces fumago (CPO) is a 42 kDa extracellular heme glycoenzyme containing ferriprotoporphyrin IX as prosthetic group [146]. CPO exhibits a broad spectrum of chemical reactivities; even though in vivo it functions mainly as a peroxide-dependent chlorinating enzyme, it also catalyzes peroxidase-, catalase- and cytochrome P450-type reactions of dehydrogenation, H2O2 decomposition and oxygen insertion, respectively, in vitro [147]. Table 13 Biocatalytic oxidation of aromatic compounds by hemoglobin and hydrogen peroxide [143]. Compound 9-Hexylanthracene Anthracene Carbazole Pyrene Dibenzothiophene Fluorene Acenaphthene Chrysene Dibenzofuran Fluoranthene Phenanthrene

Reacted substrate (%) 100 91 (±4) 84 (±28) 85 (±10) 74 (± 1) 49 (±30) NR NR NR NR NR

Product 9,10-Anthraquinone 9,10-Anthraquinone Unknown 1,8-Pyrenodione Dibenzothiophene sulfoxide 9-Fluorenone

96

Table 14 Specific activity of chloroperoxidase from Caldariomyces fumago against aromatic compounds. Aromatic compound 9-Methylanthracene Azulene Anthracene 2-methylanthracene 7,12-Dimethylbenzanthracene Benzo[a]pyrene 7-Methylbenzo [ajpyrene Acenaphthene Pyrene Benzo[ghi]perylene Perylene Biphenylene Phenanthrene Fluoranthene Fluorene Triphenylene Naphtalene Biphenyl Dibenzofuran Anthrone NR: no reaction detected.

Specific activity (min"1) 758 (± 27) 676 (+ 34) 134 (±14) 107 (±8) 87 (± 8) 84 (± 6) 81 (±7) 65 (± 8) 53 (± 6) 45 (+ 7) 25 (± 10) 10 (± 0.5) 7 (±0.1) 3 (± 0.2) 1.9 (±0.13) 0.8 (± 0.09) 0.6 (±0.01) NR NR NR

Chloroperoxidase was able to chlorinate 17 of 20 aromatic hydrocarbons assayed in the presence of hydrogen peroxide and chloride ions [46] (Table 14). Reaction rates varied from 0.6 min"1 for naphthalene to 758 min"1 for 9methylanthracene. Mono-, di- and tri-chlorinated compounds were obtained from the chloroperoxidase-mediated reaction on aromatic compounds. Chlorination of aromatic hydrocarbons could be interesting for the production of fine chemicals from petroleum products. Figure 12 shows the correlation between IP values and specific activity for PAHs. Because ionization potential could be defined as the energy involved in taking out one electron from the substrate molecule, this correlation suggest a one-electron mechanism with a free radical-mediated reaction. Only PAH's with IP lower than 8.52 eV were halogenated (Table 14). In general, the lower the IP of the PAH, the higher the specific activity of the chloroperoxidase for that substrate. The IP value of 8.52 eV appears to be a threshold, as none of the compounds tested having higher ionization potentials were transformed by chloroperoxidase [45]. This threshold value is significantly higher than those

97

reported for other peroxidases. Lignin peroxidase is able to oxidize PAH's and form quinones up to a PAH's IP of 8.0 eV [85] and manganese peroxidase from P. chrysosporium shows a threshold value for PAH's substrates of 8.1 eV [98]. 4.8. Laccase Laccases (EC 1.10.3.2) are copper-containing enzymes widespread in white rot fungi which catalyze the oxidation of a variety of aromatic phenols and anilines, reducing oxygen to water. Their characteristics have been comprehensively reviewed [148, 149]. While the substrate range for laccase is normally limited to phenolic substrates, it can be extended to nonphenolic compounds with the addition of mediating substrates such as ABTS and HBT [150-155]. In vitro oxidation of PAH's has been demonstrated by purified fungal laccases [156-160]. The rate of oxidation of several PAH's has been shown to be enhanced by the addition of the cooxidant ABTS [158-161]. Purified laccase of C. gallica transformed 7 of 10 PAHs examined in the presence of ABTS (Table 15). Benzo[a]pyrene, 9-methylanthracene, 2-methylanthracene, anthracene, biphenylene, acenaphthene, and phenathrene were oxidized by laccase [160]. The synthetic or natural mediating substances acts as free-radical mediators. These mediators are sbstrates for laccase and tranformed into free radicals by one electron subtraction, and then they diffuse and oxidize the aromatic compound prducing, as peroxidases, mainly quinones. Unlike peroxidases, no clear relationship between the substrate ionization potential and first-order rate constant could be detected.

Fig. 12. Influence of the PAH ionization potential on the chloroperoxidase activity.

98

Table 15 First rate constants of reactions of C. Gallica laccase with polycyclic aromatic hydrocarbons [160]. PAH 9-Methylanthracene Benzo[a]pyrene Acenaphthene Anthracene 2-Methylanthracene Biphenylene Phenanthrene Pyrene Fluoranthene Azulene NR, no reaction detected. NEO, nonenzymatic oxidation.

Rate constant (h"1) 240 83 10 5.2 4.9 3.8 0.8 NR NR NEO

Ionization potential (eV) 7.23 7.12 7.7 7.55 7.42 7.58 8.03 7.72 7.76 7.43

The effects of mediating substances were examined by using anthracene as the substrate (Fig. 13). The presence of 1 mM of hydroxybenzotriazine (HBT) induced an oxidation rate of anthrecene of 2.4 h"1, while 1 mM of ABTS showed a rate constant of 5.2 h"1, but the 1 mM ABTS plus lmM HBT increased the oxidation rate to 45 h"1, nine fold compared with the oxidation rate in the presence only of ABTS

Fig. 13. Effects of the free-radical mediators HBT and ABTS on the anthracene oxidation by purified laccase from C. gallica [160].

99

5. ENZYMATIC C-C BOND ACTIVATION Short-chain alkanes such as methane, ethane and propane are the most abundant and cheapest hydrocarbons available. Nevertheless, they are not used directly as raw material for production of more valuable products, mainly because C-H bonds in alkanes are not easily activated. Thus, alkanes are used mainly as fuels to produce energy. However, the direct alkane activation to produce valuable petrochemicals would exploit an inexpensive hydrocarbon feedstock. Based on commercial and process viability, some of the most promising routes for direct alkane activation have been identified [162]. Table 18 shows some of the potential routes to produce important petrochemicals as well as the conventional industrial feedstocks and currently used processes. It has been estimated that the alternative technologies could represent cost savings of up to USD 380/metric ton over the conventional processes [162]. Efficient production of petrochemicals by direct activation of alkanes remains a challenge. Particularly, oxidation of alkanes into useful products is one of the major issues in catalysis research. Yields must be kept low when using metal-based catalysts in order to keep selectivity [163]. Besides, as products are more reactive than substrates, subsequent oxidation of partially oxidized alkanes leads to undesirable or low-value products. There are several reports in the literature regarding the transformation of saturated hydrocarbons by microorganisms. There have been found microorganisms able to mineralize or degrade Q to C44 alkanes. Table 19 lists the enzymes identified to catalyze the most common transformation, usually an hydroxylation, of different alkanes. This section will briefly describe substrate specificity, activities and limitations of the most representative enzymatic systems for alkane oxidation. Table 18 Processes for production of basic and intermediate petrochemicals [162]. Potential Alkane Feedstock Methane

Ethane

Propane

Usual Feedstock Methane Hydrocarbons Methanol Ethylene Methanol Ethylene Ethylene Propylene Propylene Propylene

Conventional Process

Product

Reforming Steam cracking Oxidation/dehydrogenation Oxychlorination Carbonylation Oxidation Oxidation Ammoxidation Oxidation Chlorhydrination, epoxidation

Methanol Ethylene Formaldehyde Vinyl chloride Acetic acid Acetaldehyde Ethylene oxide Acrylonitrile Acrylic acid Propylene oxide

100

5.1. Methane monooxygenase The methane monooxygenase (MMO) system is expressed in microorganisms able to use methane as an energy source (methanotrophs) (see chapter 6). These gram-negative microbes are able to use methane as energy and carbon source. However they are not able to grow on larger alkanes. Two forms of the enzymatic system have been described: a membrane associated, or particulate, methane monooxygenase (pMMO) and a soluble cytoplasmatic methane monooxygenase (sMMO) [164]. While pMMO is a membrane protein produced by all known methanotrophs, sMMO is expressed only by a subset of them. Moreover, sMMO shows a wide range of substrate specificity, an it is able to catalyze the oxidation of alkenes, aromatic, alicyclic and heterocyclic compounds, whereas pMMO only mediates the oxidation of a small group of short-chain alkanes and alkenes [164]. pMMO and sMMO are evolutionarily unrelated. It has been demonstrated that pMMO is an iron copper protein, which is produced only under conditions of copper sufficiency; regarding its mode of action, it has been related to ammonia monooxygenases [165]. sMMO, on the other hand, is an iron-containing enzyme produced only under copper-limiting conditions. It has been suggested that sMMO may provide a competitive advantage in copper-depleted sites, enabling the methanotrophs to colonize a wider range of environments. sMMO is comprised by three components: an oxygenase, a reductase and a coupling protein [166]. This system has been extensively characterized. The NADH-dependent oxidation reaction catalyzed by sMMO is depicted in Figure 14. So far pMMO has been poorly characterized, mainly due to its unstability to dioxygen exposure in cell-free fractions. Purification procedure is usually long and require strict anaerobic conditions in order to maintain its activity [167, 168]. Furthermore, most purified pMMO preparations show low activity; in vitro rates represent only about 1-5% of physiological rates. Table 19 Enzymes able to catalyze the oxidation of alkanes. Enzyme (EC number)

Paraffinic substrate

Active site

p-Methane monooxygenase (1.14.13.25) s-Methane monooxygenase (1.14.13.25) Alkane hydroxylase (1.14.15.3) P-450 monooxygenase (1.14.14.1)

Methane, Ci to C5 linear alkanes Methane, C5 to C7 linear and branched alkanes C5 to C24 alkanes

Trimeric copper 164 cluster Diiron cluster 164 Diiron cluster

Cyclohexane, C5 to C§ alkanes Heme

Ref.

169-171 172

101

Fig. 14. Steps involved in the oxidation reaction catalyzed by alkane hydroxylase and methane monooxygenase (A) and cytochrome P450 (B).

5.2. Alkane hydroxylase The alkane hydroxylase system (Alk) present in Burkholderia cepacia and Pseudomonas, Acinetobacter and Rhodococcus strains is a threecomponent monooxygenase, comprising an hydroxylase, a rubredoxin and a rubredoxin reductase [173]. The hydroxylase component is membrane-bound, while the rubredoxin and rubredoxin reductase components are soluble, cytoplasmic proteins. This enzyme is able to oxidize medium and long-chain linear alkanes using reducing equivalents from NADH or NADPH as shown in Fig. 15. The most studied system is the enzyme from Pseudomonas putida Gpol, which is able to oxidize C5 to C12 alkanes [169]. Enzymes from other organisms are able to oxidize larger alkanes, such as the enzyme from Acinetobacter sp. strain ADP1 (C]2 to C]8) [174], the enzyme from Rhodococcus sp. (C]2 to C16) [175], the two enzymes from Acinetobacter sp strain M-l (C]6 to C22 and >C22) [176] and the two enzymes from Pseudomonas aeuroginosa PAO1 (C12 to C20 and Ci5 to C24) [171]. Apparently, microorganisms have enzymes with different specificities that are expressed depending on the available substrate. Even though the organisms producing both sMMO and pMMO can catalyze the oxidation of several alkanes in addition to methane, they are unable to grow on any of them. On the other hand, organisms producing Alk use medium-chain alkanes as energy and carbon source. The alkane is oxidized to alcohol by Alk, while further oxidation to aldehyde and carboxylic acid is catalyzed by different enzymes. The carboxylic acid then enters the fatty acid degradation pathway and is used as an energy source. In order to make use of the Alk system for the production of oxidized intermediates, the metabolic reactions must be interrupted, such that the desired product accumulates.

102

Table 20 Relative activities of alkane hydroxylase for the oxidation of medium-chain alkanes [177]. Substrate

Major product

R elative

n-Hexane n-Heptane n-Octane n-Nonane n-Decane n-Undecane n-Dodecane

None n-Heptanol n-Octanol n-Nonanol n-Decanol n-Undecanol None

0 0. 58 1 0.83 0. 16 < 0.05 0

rate (k substrate/ k n-octane)

An example of these has been demonstrated by Bosetti and coworkers [177]. A plasmid containing the three components of the Alk systems was constructed and introduced to a Pseudomonas strain lacking the alcohol dehydrogenase. This enzyme catalyzes the second step on the metabolism of alkanes, that is, the oxidation of alcohol to aldehyde. As a result, the reeombinant bacteria was able to oxidize alkanes to alcohols, without oxidizing them further. As the reeombinant bacteria cannot use the alkane for growth, a carbon source must be supplied to the microorganism. Following this strategy, the bacteria was able to transform C7 to Cn alkanes to their corresponding alcohols (see Table 20). 5.3. Cytochrome P450 Another type of enzymes capable of oxidizing alkanes belong to the cytochrome P450 family. These enzymes, unlike MMO and Alk, are heme proteins. They catalyze the oxidation of substrates in the presence of NAD(P)H and usually the system consists of two components: an hydroxylase and a reductase. Cytochrome P450 present in yeast are able to convert alkanes to oxidized products. In particular, Candida sp. are able to convert >Ci2 alkanes to a,co-dicarboxylic acids that are secreted to the medium. The first reaction is the ©-oxidation of the alkane to the corresponding alcohol and it is catalyzed by a cytochrome P450. Further oxidation to the acid is catalyzed by a fatty alcohol oxidase and a fatty aldehyde deshydrogenase. The fatty acid is oxidized again by the same enzymes, to produce the diacid [178]. However, these cytochromes P450 are usually membrane-bound and have a multicomponent nature, which makes them difficult to produce in large quantities using reeombinant techniques.

103

Table 21 Alkane hydroxylation by different enzymatic systems [172] Enzymatic system

Substrate

Maximum rate (min"1)

P450 BM-3 139-3 P450BM-3 Alk sMMO

Hexane Hexane Octane Methane

3800 182 210 222

The most promising example in this family is the unspecific bacterial cytochrome P450 BM-3 from Bacillus megaterium. This enzyme is better suited for a potential industrial application as it has several advantages over other enzymes of the kind, such as being a soluble (not membrane-bound) single polypeptide chain that is readily expressed in E. coli [179]. A very attractive characteristic of cytochrome P450 BM-3 is that, unlike MMO, Alk or other cytochromes P450, the hydroxylase and reductase domain are comprised in the same polypetide. The natural activity of the enzyme is to catalyze the hydroxylation of C12 to C18 fatty acids with the concomitant consumption of dioxygen and NADPH. This enzyme has been engineered by Arnold and coworkers using laboratory evolution techniques to produce mutant 139-3, which acquired the capability of catalyzing the oxidation of medium-chain alkanes [178]. This mutant is the fastest known enzyme for alkane hydroxylation, as it is more than seventeen times faster than the MMO or Alk enzymatic systems already described. Table 21 shows the activity of mutant 139-3 compared with other systems for the catalytic oxidation of alkanes. Moreover, several mutants have been obtained from 139-3 mutant that catalyze the regio- and enantioselective production of alcohols from smaller alkanes. Some of the mutants are from 1.1 to 4.5 faster than 139-3, when as shown on Table 22.

6. OPORTUNITIES AND CHALLENGES In this century all industries, including the petroleum industry, should apply energetically efficient production processes with reduced environmental impact: this is their main challenge. In addition to the expected improvements of conventional processes, the use of new and non conventional techniques for petroleum refining should be evaluated. Doubtless, biotechnology is among the non conventional techniques to be explored. Enzymatic catalysis with high transformation efficiency, high specificity and mild reaction conditions offers a wide range of possibilities. The analysis of the available data on microbial and enzymatic transformations of oil products shows

104

several opportunities for some sectors of the petroleum industry, such as deep desulfurization and denitrogenation, and asphaltene upgrading. The powerful tools of molecular biochemistry can be used to improve the enzyme stability and efficiency. These techniques may be applied to the particular needs of the petroleum industry. Protein engineering is generally understood as the use of site-directed or random mutagenesis to alter the properties of a protein or enzyme. In addition, the enzymes isolated from extremophilic microorganisms are extremely thermostable and generally resistant to non conventional conditions such as organic solvents and extreme pH. Thus, many enzymes and enzymatic proteins are still to be discovered. In addition, over the past two decades people have seen many examples of the improvement of biocatalysts by chemical and genetic techniques. Still, there is not any enzymatic process ready to by applied in the petroleum refining industry, and three main research fields may be suggested to obtain an enzyme catalysts to be used in the petroleum industry: 1) The search of new enzymatic activities upon petroleum products, specially from extreme environments. New microorganisms are currently discovered from extreme environments such as thermal vents in the ocean deep and fossilized salt rocks. The enzymes isolated from extremophilic microorganisms are extremely thermostable and generally resistant to organic solvents and extreme pH. Enzymes from these microorganisms working in non-aqueous systems at temperatures higher as 200°C (operating conditions found in refineries) could be expected. Moreover at high temperatures, the hydrocarbons bioavailability and solubility is increased. All these unknown organisms are a potential source of new enzyme forms with different physicochemical properties: the potential source of biocatalytic activity for the oil industry could be there. 2) The improvement of the enzymatic activities in very low water systems, in order to increase the transformation rates using petroleum fractions without further addition of water. Since petroleum is a hydrophobic material, it is suitable to speculate that new enzymatic processes for the oil industry should be carried out in non-aqueous systems. The use of reaction mixtures containing organic solvents reduces mass transfer limitations, promoting the establishment of productive interactions between the enzyme and the hydrophobic substrates (oil-derived compounds). In addition, a biocatalyst placed in a non-aqueous medium shows interesting properties, such as improved thermostability, higher substrate accessibility, adjustable selectivity, and high storage stability. The study of the relationship between the solvent properties and the enzyme activity seems to be essential to understand and to improve the biocatalytic processes.

105

Table 22 Improvement of cytochrome P450 BM-3 for the catalysis of small alkanes oxidation [179] Enzyme mutant Mutant 139-3

Substrate

Propane Octane Mutant J Propane Octane Propane Mutant 9- 10A Octane Mutant 9-10A-A82L Propane Octane

Major product n-propanol 2-octanol n-propanol 2-octanol n-propanol 2-octanol n-propanol 4-octanol

Relative rate (k mutam/ k i39.3) 1 1 2.5 1.37 1.91 1.12 4.5 1.10

3) Finally, the enzyme design by genetic and chemical methods. The economical and technical feasibility of a large scale enzymatic process depends mainly on the activity and stability of the biocatalyst under the actual conditions found in the petroleum refining industry. Molecular biochemistry efforts should be directed to improve enzyme activity and stability in petroleum fractions with almost no addition of water and at the temperatures found usually in refineries. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15]

R.A. Kerr, Science 281 (1998) 1128. C.J. Campbell and J.H. Laherrere, Sci. Am. 278 (1998) 78. W.C. Lai and K.J. Smith, Fuel 80 (2001) 1121. D.S. Scott, D. Radlein, J. Piskorz, P. Majerski and T.J.W. DeBruijn, Fuel 80 (2001) 1087. M.N. Dadashev and G.V. Stepanov, Chem. Technol. Fuels Oils 36 (2000) 8. S.J. Park, C.J. Kim and B.S. Rhee, Ind. Eng. Chem. Res. 39 (2000) 4897. R. Vazquez-Duhalt, E. Torres, B. Valderrama and S. Le Borgne S., Energy Fuel 16 (2002) 1239. J.D. Rozzell, Bioorg. Med. Chem. 7 (1999) 2253. B.L. McFarland, D.J. Boron, W. Deever, J.A. Meyer, A.R. Johnson and R.M. Atlas, Crit. Rev. Microbiol. 24 (1998) 99. A. M. Klibanov, Nature 409 (2001) 241. A. Schmid, J.S. Dordick, B. Hauer, A. Kiener, M. Wubblolts and B. Witholt, Nature 409 (2001) 258. J. S. Dordick, Enzyme Microb. Technol. 11(1989) 192. J. S. Dordick, Y. L. Khmelnitsky and M. Sergeeva, Curr. Opin. Microbiol. 1 (1998) 311. A. Zaks and A.M. Klibanov, Science 224 (1984) 1249. S. Akanuma, A. Yamagishi, N. Tanaka, and T. Oshima, Protein Sci. 7 (1998) 698.

106

[16] L. Giver, A. Gershenson, P.O. Freskgard and F.H. Arnold, Proc. Natl. Acad. Sci. USA 95 (1998) 12809. [17] H. Zhao and F.H. Arnold, Protein Eng. 12 (1999) 47. [18] J. R. Cherry, M.H. Lamsa, P. Schneider, J. Vind, A. Sevendsen, A. Jones and A.H. Pedersen, Nat. Biotechnol. 17 (1999) 379. [19] A. Aksoy, H. Tumturk and N. Hasirci, J Biotechnol. 60 (1998) 37. [20] P. Wang, M.V. Sergeeva, L. Lim and J.S. Dordick, Nat. Biotechnol. 15 (1997) 789. [21] C.P. Govardhan, Curr Opin. Biotechnol. 10 (1999) 331. [22] N.L. St Clair and M.A. Navia, J. Am. Chem. Soc. 114 (1992) 7314. [23] S.S. Wong and C.L.J. Wong, Enzyme Microb. Technol. 14 (1992) 866. [24] M. Ayala, E. Horjales, M.A. Pickard and R. Vazquez-Duhalt, Biochem. Biophys. Res. Comm. 295 (2002) 828. [25] J.P. Lenders and R.R. Crichton, Biotechnol. Bioeng. 26 (1984) 1343. [26] R.A.K. Srivastava, Enzyme Microb. Technol. 13 (1991) 164. [27] H.F. Gaertner and A.J. Puigserver, Enzyme Microb. Technol. 14 (1992) 150. [28] D. Garcia, F. Ortega and J.L. Marty, Biotechnol. Appl. Biochem. 27 (1998) 49. [29] M.J. Hernaiz, J.M. Sanchez-Montero and J.V. Sinisterra, Enzyme Microb. Technol. 24(1999)181. [30] H. Garcia-Arellano, B. Valderrama, G. Saab-Rincon and R. Vazquez-Duhalt, Bioconjugate Chem. 13 (2000) 1336. [31 ] EPA (2000) EPA420-F-00-057 [32] A.M. Aitani, M.F. Ali and H.H. Al-AIi, Petrol. Sci. Technol. 18 (2000) 537. [33] L. Linguist and M. Pachecho, Oil Gas J. Feb 22 (1999) 46. [34] M.A. Kertesz and C. Wietek, Appl. Microbiol. Biotechnol. 57 (2001) 460. [35] J. Klein, D.E.A. Catcheside, R. Fakoussa, L. Gazso, W. Fritsche, M. Hofer, F. Laborda, I. Margarit, H.J. Rehm, M. Reich-Walber, W. Sand, S. Schacht, H. Schmiers, L Setti and A. Steinbuchel, Appl. Microbiol. Biotechnol. 52 (1999) 2. [36] W.M. Coco, W.E. Levinson, M.J. Crist, H.J. Hektor, A. Darzins, P.T. Pienkos, C.H. Squires ad D.J. Monticello, Nat. Biotechnol. 19 (2001) 354. [37] M.J. Grossman, M.K. Lee, R.C. Prince, K.K. Garrett, G.N. George and I.J. Pickering. Appl. Environ. Microbiol. 65 (1999) 181. [38] M.J. Grossman, M.K. Lee, R.C. Prince, V. Minak-Bernero, G.N. George and I.J. Pickering, Appl. Environ. Microbiol. 67 (2001) 1949. [39] M. Naito, T. Kawamoto, K. Fujino, M. Kobayashi, K. Maruhashi and A. Tanaka, Appl. Microbiol. Biotechnol. 55 (2001) 374. [40] J.A. Rupley and G. Carreri, Adv. Protein Chem. 41 (1991) 37. [41] A.M. Klibanov, Trends Biotechnol. 15 (1997) 97. [42] P.J. Hailing, Curr. Opin. Chem. Biol. 4 (2000) 74. [43] M. Tuena de Gomez-Poyou and A. Gomez-Poyou, Crit. Rev. Biochem. Mol. Biol. 33 (1998)53. [44] M. Ayala, R. Tinoco, V. Hernandez, P. Bremauntz and R. Vazquez-Duhalt, Fuel Process Technol. 57 (1998) 101. [45] M. Ayala, N.R. Robledo, A. Lopez-Munguia and R. Vazquez-Duhalt, Environ. Sci. Technol. 34 (2000) 2804. [46] R. Vazquez-Duhalt, M. Ayala and F.J. Marquez-Rocha, Phytochemistry 58 (2001) 929. [47] M.P.J. van Deurzen, F. van Rantwijk and R.A. Sheldon, Tetrahedron 53 (1997) 13183.

107

[48] S. Colonna, N. Gaggero, C. Richelmi and P. Pasta, Trends Biotechnol. 17 (1999) 163. [49] J.H. Dawson and M. Sono, Chem. Rev. 87 (1987) 1255. [50] F. van de Velde, F. van Rantwijk and R.A. Sheldon, Trends Biotechnol. 19 (2001) 73. [51] W. Tischer and V. Kasche, Trends Biotechnol. 17 (1999) 326. [52] F. Secundo, S. Spadaro, G. Carrea and P.L.A. Overbeeke, Biotechnol. Bioeng. 62 (1999) 554. [53] R. Jaenicke, J. Biotechnol. 79 (2000) 193. [54] Y.L. Khmelnitsky and J.O. Rich, Curr. Opin. Chem. Biol. 3 (1999) 47. [55] J.L. Schmitke, C.R. Wescott and A.M. Klibanov, J. Am. Chem. Soc. 118 (1996) 3360. [56] E. Torres, R. Tinoco and R. Vazquez-Duhalt, J. Biotechnol. 49 (1996) 59. [57] M.P.J. van Deurzen, I.J. Remkes, F. van Rantwijk and R.A. Sheldon, J. Mol. Cat. A Chemical 117(1997)329. [58] B. Valderrama, M. Ayala and R. Vazquez-Duhalt, Chem. Biol. 9 (2002) 555. [59] Government of Alberta, Canada, 2002. Department of Energy. (http://www.energy. gov.ab.ca). [60] J.W. Bunger and N.C. Li (eds.), Chemistry of Asphaltenes, American Society for Advanced Chemistry Series 195, 1981. [61] J.G. Speight, The Chemistry and Technology of Petroleum, Marcel Dekker Inc., New York, 1998, pp. 412-467. [62] O.P. Strausz, T.W. Mojelsky and E.M. Lown, Fuel 71 (1992) 1355. [63] J.C. Bertrand, E. Rambeloarisoa, J.F. Rontani, G. Giusti and G. Mattei, Biotechnol. Lett. 5(1983)567. [64] J.F. Rontani, F. Bosser-Joulak, E. Rambeloarisoa, J.C. Bertrand, G. Giusti and R. Faure, Chemosphere 14 (1985) 1413. [65] R.C. Wyndham and J.W. Costerton, Appl. Environ. Microbiol. 41 (1981) 791. [66] E. Premuzic, M.S. Lin, M. Bohenek and W.M. Zhou, Energy Fuels 13 (1999) 297. [67] G. Thouand, P. Bauda, J. Oudot, G. Kirsh, C. Sutton and J.F. Vidalie, Can. J. Microbiol. 45 (1999) 106. [68] D.J. Lacotte, G. Mille, M. Acquaviva and J.C. Bertrand, Chemosphere 32 (1996) 1755. [69] P.M. Fedorak, K.M. Semple, R. Vazquez-Duhalt and D.W.S. Westlake, Enzyme Microb. Technol. 15 (1993) 429. [70] L. Mogollon, R. Rodriguez, W. Larrota, C. Ortiz and R. Torres, Appl. Biochem. Biotechnol. 70-72 (1998) 765. [71] H. Garcia-Arellano, E. Buenrostro-Gonzalez and R. Vazquez-Duhalt, Biotechnol. Bioeng. (2004) (In press) [72] H. Groenzin and O.C. Mullins, J. Phys. Chem. A 103 (1999) 11237. [73] H. Groenzin and O.C. Mullins, Energy Fuels 14 (2000) 677. [74] E. Buenrostro-Gonzalez, S.I. Andersen, J.A. Garcia-Martinez and C. Lira-Galeana, Energy Fuels 16 (2002) 732. [75] H. Habe and T. Omori, Biosci. Biotechnol. Biochem. 67 (2003) 225. [76] R.M. Atlas, Microbiol. Rev. 45 (1981) 180. [77] R.M. Atlas, Petroleum microbiology, MacMillan Publishing Co., New York,1984. [78] R.R. Colwell and J.D. Walker, Crit. Rev. Microbiol. 5 (1977) 423. [79] T.D. Gibson, Science 161 (1968) 1093.

108

[80] R.L. Farrell, K.E. Murtagh, M. Tien, M.D. Mozuch and T.K. Kirk, Enzyme Microb. Technol. 11(1989)322. [81] D. Sanglard, M.S.A. Leisola and A. Fiechter, Enzyme Microb. Technol. 8 (1986) 209. [82] K.E. Hammel, B. Kalyanaraman and T.K. Kirk, J. Biol. Chem. 261 (1986) 16952. [83] J.A. Field, R.H. Vledder, J.G. van Zelst and W.H.Rulkens, Enzyme Microb. Technol. 18(1996)300. [84] S.D. Haemmerli, M.S.A. Leisola, D. Sanglard and A. Fiechter, J. Biol. Chem. 261 (1986)6900. [85] R. Vazquez-Duhalt, D.W.S. Westlake and P.M. Fedorak, Appl. Environ. Microbiol. 60 (1994) 459. [86] M. Tatarko and J.A. Bumpus, Lett. Appl. Microbiol. 17 (1993) 20. [87] K.E. Hammel and PJ. Tradone, Biochemistry 27 (1988) 6563. [88] R. Vazquez-Duhalt, D.W.S. Westlake, and P.M. Fedorak, Appl. Microbiol. Biotechnol.42(1995)675. [89] M.H. Gold, J.K. Glenn and M. Alic, Methods Enzymol. 161 (1988) 74. [90] C. Cripps, J.A. Bumpus and S.D. Aust, Appl. Environ. Microbiol. 56 (1990) 1114. [91] E.A. Pease and M. Tien, J. Bacteriol. 174 (1992) 3532 . [92] M.J.J. Kotterman, R. A. Wasseveld and J.A. Field, Appl. Environ. Microbiol. 62 (1996) 880. [93] J.A. Field, E. de Jong, G. Feijoo-Costa and J.A.M. de Bont, Trends Biotechnol. 11 (1993)44. [94] J.L. Popp and T.K. Kirk, Arch. Biochem. Biophys. 288 (1991) 145. [95] M.A. Moen and K.E. Hammel, Appl. Environ. Microbiol. 60 (1994) 1956. [96] B.W. Bogan, R.T. Lamar and K.E. Hammel, Appl. Environ. Microbiol. 62 (1996) 1788. [97] B. W. Bogan, B. Schoenike, R.T. Lamar and D. Cullen, Appl. Environ. Microbiol. 62 (1996)2381. [98] B.W. Bogan and R.T. Lamar, Appl. Environ. Microbiol. 61 (1995) 2631. [99] M.J.J. Kotterman, R.A. Wasseveld and J.A. Field, Appl. Environ. Microbiol. 62 (1996) 880. [100] B.W. Bogan and R.T. Lamar, Appl. Environ. Microbiol. 62 (1996) 1597. [101] M.J. Martinez, F.J. Ruiz-Duenas, F. Guillen and A.T. Martinez. Eur. J. Biochem. 237 (1996) 15412. [102] T. Mester and J.A. Field, J. Biol. Chem. 273 (1998) 15412. [103] A. Heinfling, M.J. Martinez, A.T. Martinez, M. Bergbauer and U. Szewzyk, FEMS Microbiol. Lett. 165 (1998) 43. [104] A. Heinfling, J. Ruiz-Duenas, M.J. Martinez, M. Bergbauer, U. Szewzyk, and A.T. Martinez, FEBS Lett. 428 (1998) 141. [105] F.J. Ruiz- Duefias, M.J. Martinez and A.T. Martinez, Mol. Microbiol. 31 (1999) 223. [106] F.J. Ruiz- Duefias, M.J. Martinez and A.T. Martinez, Appl. Environ. Microbiol. 65 (1999)4705. [107] S. Camarero, B. Bockle, M.J. Martinez and A.T. Martinez, Appl. Environ. Microbiol. 62(1996)1070. [108] S. Sarkar, A.T. Martinez and M.J. Martinez, Biochim. Biophys. Acta. 1339 (1997) 23. [109] Y. Wang, R. Vazquez-Duhalt and M.A. Pickard, Can. J. Microbiol. 47 (2001) 277.

109

[110] Y. Wang, R. Vazquez-Duhalt and M.A. Pickard, Curr. Microbiol. 43 (2002) 77. [111] S. Camarero, S. Sarkar, F.J. Ruiz-Duefias, M.J. Martinez and A.T. Martinez, J. Biol. Chem. 274 (1999) 10324. [112] M. Ayala, M.C. Baratto, R. Basosi, R. Vazquez-Duhalt and R. Pogni, J. Mol. Catalysis B: Enzymatic 16 (2001) 159.. [113] Y. Wang, R. Vazquez-Duhalt and M.A. Pickard, Can. J. Microbiol. 49 (2003) 675. [114] P.R. Ortiz de Montellano, Cytochrome P450, Structure, Mechanism and Biochemistry, Plenum Press, New York, 1986. [115] DJ. King, M.R. Azari and A. Wiseman, Xenobiotica 14 (1984) 187. [116] M.R. Azari and A. Wiseman, Enzyme Microb. Technol. 4 (1982) 401. [117] J. Capdevilla, R.W. Estabrook and R.A. Prough, Arch. Biochem. Biophys. 200 (1980) 186. [118] W. Dehnen, R. Tomingas and J. Roos, Anal. Biochem. 53 (1973) 373. [119] S.L. Kelly, D.C. Lamb, B.C. Baldwin and D.E. Kelly, Biochem Biophys. Res. Comm. 197(1993)428. [120] S. Masaphy, D. Levanon, Y. Henis, K. Venkateswarlu and S.L. Kelly, Biotechnol. Lett. 17(1995)969. [121] A.P. Koley, J.T.M. Buters, R.C. Robinson, A. Markowitz and F.K. Friedman, Arch. Biochem. Biophys. 336 (1996) 261. [122] P. Anzenbacher, T. Niwa, L.M. Tolbert, S.R. Sirimanne and F.P. Guengerich, Biochemistry 35 (1996) 2512. [123] A.D. Rahimtula, P.J. O'Brien, H.E. Seifreid and D.M. Jerina, Eur. J. Biochem. 89 (1978) 133. [124] M. Shou, K.W. Krausz, F.J. Gonzalez and H.V. Gelboin, Carcinogenesis 17 (1995) 2429. [125] W.L. Backes, M. Hogaboom and WJ. Canady, J. Biol. Chem. 257 (1982) 4063. [126] W.L. Backes, G. Cawley, C.S. Eyer, M. Means, K.M. Causey and WJ. Canady, Arch. Biochem. Biophys. 304 (1993) 27. [127] O-S. Li, J. Ogawa, R.D. Schmid and S. Shimizu, Appl. Environ. Microbiol. 67 (2001)5735. [128] S.M. Fowler, P.A. England, A.C.G. Westlake, D.R. Rouch, D.P. Nickerson, C. Blunt, D. Braybrook, S. West, L.L. Wong and S.L. Flitsch, J. Chem. Soc. Chem. Commun. (1994)2761. [129] S. Modi, W.U. Primrose, L.Y. Lian and G.C.K. Roberts, Biochem J. 310 (1995) 939. [130] H. Joo, Z. Lin and F. H. Arnold, Nature 399 (1999) 670. [131] C.F. Harford-Cross, A.B. Carmichael, F.K. Allan, P.A. England, D.A. Rouch and LL. Wong, Protein Eng. 13 (2000) 121. [132] R. Vazquez-Duhalt, J. Mol. Cat. B: Enzymatic 7 (1999) 241. [133] R. Vazquez-Duhalt, D.W.S. Westlake and P.M. Fedorak, Enzyme Microb. Technol. 15 (1993)494. [134] N.L. Klyachko and A.M. Klibanov, Appl. Biochem. Biotechnol. 37 (1992) 53. [135] E. Torres, J.V. Sandoval, F.I. Rosell, A.G. Mauk and R. Vazquez-Duhalt, Enzyme Microb. Technol. 17 (1995) 1014. [136] R. Akasaka, T. Mushino and M. Hirobe, J. Chem. Soc. Perkin Trans. 1 (1994) 1817. [137] R. Tinoco and R. Vazquez-Duhalt, Enzyme Microb. Technol. 22 (1997) 8. [138] JJ. Mieyal, R.S. Ackerman, J.L. Blumer and L.S. Freeman, J. Biol. Chem. 251 (1976) 3436. [139] Y. Yoshida, K. Kashiba andE. Niki, Biochim. Biophys. Acta 1201 (1994) 165.

110

[140] J.C. Alvarez and P.R. Ortiz de Montellano, Biochemistry 31 (1992) 8373. [141] G.L. Kedderis, D.E. Rickert, R.N. Pandey and P.F. Hollenberg, J. Biol. Chem. 261 (1986) 15910. [142] C. Giulivi and K.J.A. Davies, J. Biol. Chem. 265 (1990) 19453. [143] M. Ortiz-Leon, L. Velasco and R. Vazquez-Duhalt, Biochem. Biophys. Res. Comm. 215(1995)968. [144] PJ. Thornalley, RJ. Trotta and A. Stern, Biochim. Biophys. Acta 759 (1983) 16. [145] MJ. Davies, Biochem. Biophys. Acta 28 (1988) 28. [146] M. Sundaramoorthy, J. Terner and T.L. Poulos, Structure 3 (1995) 1367. [147] X. Yi, M. Mroczko, K.M. Manjol, X. Wang and L.P. Hager, Proc. Nat. Acad. Sci. 96 (1999) 12412. [148] L. Gianfreda, F. Xu and J-M. Bollag. Bioremediation J. 3 (1999) 1. [149] T.F. Thurston, Microbiology 140 (1994) 19. [150] R. Bourbonnais and M. G. Paice, FEBS Lett. 267 (1990) 99. [151] R. Bourbonnais, M. G. Paice, I. D. Reid, P. Lanthier and M. Yamaguchi, Appl. Environ. Microbiol. 61 (1995) 1867. [152] R. Bourbonnais, M. G. Paice, B. Freiermuth, E. Bodie and S. Borneman, Appl. Environ. Microbiol. 63 (1997) 4627. [153] R. Bourbonnais, D. Leech and M.G. Piace, Biochim. Biophys. Acta 1379 (1998) 381. [154] K. Li, F. Xu and K.E.L. Eriksson, Appl. Environ. Microbiol. 65 (1999) 2654. [155] F. Xu, J.J. Kulys, K. Duke, K. Li, K. Krikstopaitis, H.J.W. Deussen, E. Abbate, V. Galinyte and P. Schneider, Appl. Environ. Microbiol. 66 (2000) 2052. [156] S. Bohmer, K. Messner and E. Srebotnik, Biochem. Biophys. Res. Commun. 244 (1998)233. [157] P.J. Collins, M. J. J. Kotterman, J. A. Fiel, and A. D. W. Dobson, Appl. Environ. Microbiol. 62(1996)4563. [158] C. Johannes, A. Majcherczyk and A. Huttermann. Appl. Microbiol Biotechnol. 46 (1996)313. [159] A. Majcherczyk, C. Johannes and A. Huttermann, Enzyme Microb. Technol. 22 (1998)335. [160] M.A. Pickard, R. Roman, R. Tinoco and R. Vazquez-Duhalt, Appl. Environ. Microbiol. 65 (1999) 3805. [161] C. Johannes and A. Majcherczyk, Appl. Environ. Microb. 66 (2000) 524. [162] Chem Systems. Alkane Activation: Petrochemical Feedstocks of the Future, 1999. [163] J.A. Labinger and J.E. Bercaw, Nature 417 (2002) 507. [164] J.C. Murrell, B. Gilbert and I.R. McDonald, Arch. Microbiol. 173 (2000) 325. [165] H.T. Nguyen. S.J. Elliot, J.H. Yip, S.L. Chan, J. Biol. Chem. 273 (1998) 7957. [166] B.G. Fox, W.A. Froland, J.E. Dege and J.D. Lipscombs, J. Biol. Chem. 264 (1989) 10023. [167] D.W, Choi, R.C. Kunz, E.S. Boyd, J.D. Semrau, W.E. Antholine, J.I. Han. J.A. Zahn, J.M. Boyd, A.M. de la Mora and A.A. DiSpirito, J. Bacteriol. 185 (2003) 5755. [168] P. Basu, B. Katterle, K.K. Andersson and H. Dalton, Biochem. J. 369 (2003) 417. [169] A. Schmid, B. Sonnleitner and B. Witholt, Biotechnol. Bioeng. 60 (1998) 10. [170] T.H.M. Smits, S.B. Balada, B. Witholt and J.B. van Beilen, J. Bacteriol. 184 (2002) 1733. [171] N.N. Marin, L. Yuste, F. Rojo, J. Bacteriol. 185 (2003) 3232.

Ill

[172] M.W. Peters, P. Meinhold, A. Glieder and F.H. Arnold, J. Am. Chem. Soc. 125 (2003) 13442. [173] J. Shanklin, C. Achim, H. Schmid, B.G. Fox and E. Miinck, Proc. Natl. Acad. Sci. USA 94 (1997) 2981. [174] A. Ratajczak, W. GeiBdorfer and W. Hillen, J. Bacteriol. 180 (1998) 5822. [175] L.G. Whyte, T.H.M. Smits, D. Labbe, B. Witholt, C.W. Greer and J.B. van Beilen, Appl. Environ. Microbiol. 68 (2002) 5933. [176] A. Tani, T. Ishige, Y. Sakai and N. Kato, J. Bacteriol. 183 (2001) 1819. [177] A. Bosetti, J.B. van Beilen, H. Preusting, R.G. Lageveen and B. Witholt, Enzyme Microb. Technol. 14 (1992) 702. [178] D.L. Craft, K.M. Madduri, M. Eshoo and C.R. Wilson, Appl. Environ. Microbiol. 69 (2003) 5983. [179] A. Glieder, E.T. Farinas and F.H. Arnold, Nat. Biotechnol. 20 (2002) 1135.


Studies in Surface Science and Catalysis 151 R. Vazquez-Duhalt and R. Quintero-Ramirez (Editors) © 2004 Elsevier B .V. All rights reserved.

113

Chapter 4

Prospects for biological upgrading of heavy oils and asphaltenes K.M. Kirkwood", J.M. Foght", and M.R. Gray" department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2G6 b

Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada, T6G 2E9

1. INTRODUCTION Increasing supply of heavy crude oils and bitumens, mainly from Canada, Mexico and Venezuela, has increased the interest in transportation and conversion of the high-molecular weight fractions of these materials into refined fuels and petrochemicals. The high viscosity of these crudes requires addition of a solvent in order to allow pipelining over a significant distance. The cost of suitable solvents, such as naphtha or natural gas condensate, has led to study of new methods to reduce the viscosity of heavy crudes. Once they enter a refinery, processing of heavy crudes and bitumens requires conversion of the vacuum residue components, including the asphaltenes, into distillable oils. This upgrading has typically been accomplished with either thermal conversion (cracking or coking) or by catalytic hydroconversion. Thermal processing can range from mild cracking, to reduce viscosity, to severe cracking with attendant formation of coke. These high-temperature processes require expensive investment in process equipment and supporting infrastructure for supply of hydrogen and treatment of hydrogen sulfide in cracked off-gases. In contrast to the available processes, biological processing may offer less severe process conditions and higher selectivity to specific reactions. This chapter reviews the characteristics of the molecules in the vacuum residue fraction of crude oils, and examines the prospects for using biological processes to improve the value of these materials.

114

2. MOLECULES OF INTEREST Heavy crude oils pose new upgrading challenges, in addition to the upgrading needs common to lighter crudes. These problems are related to two types of high molecular weight molecules present in these oils: waxes and asphaltenes. Waxes are long-chain paraffinic molecules, or alkanes, which typically cause operational problems if longer than 40 carbon atoms [3]. Asphaltenes, on the other hand, are not classified by structure, but are defined as a solubility class, including material that is soluble in toluene but not in «-pentane (or alternatively «-heptane). There are two different views on the molecular structure of asphaltenic material. The first represents asphaltenes as having a single large condensed polycyclic aromatic core, with aliphatic chains attached on the periphery (Fig. la) [1, 4-6]. This type of structure, however, does not account for all of the physical and chemical properties of asphaltenes. The second

Fig. 1. Representative models of asphaltene molecules showing either (a) a single large condensed polycyclic aromatic core [1] or (b) multiple smaller polycyclic aromatic cores with aliphatic bridges [2].

115

representation describes asphaltenes as having multiple smaller polycyclic aromatic cores (2-4 rings) linked by aliphatic bridges of varying lengths (Fig. lb) [2, 7-10]. Sulfides, ethers, and esters have been identified as common linking structures in the aliphatic bridges found in asphaltenes [11]. This type of structure accounts for the observed reactivity of asphaltenes. These large molecules are problematic for biological transformation. Transformation rates are limited by the mass transfer of target molecules to the biocatalyst and, in the case of whole cells, across the cell membrane (reviewed in Ref. [12]). Interfacial mass transfer can be improved through emulsification, increasing the interfacial contact area, however emulsification is of limited value in overcoming the barrier of transport into biological cells unless appropriate uptake mechanisms are available. Despite these difficulties, there is evidence in the literature for bacterial transformation of complex, high molecular weight substrates. Rhodococcus erythropolis strain IGTS8, for example, was originally isolated from an enrichment culture with the ability to use coal as its sole source of sulfur. This mixed culture was able to remove over 90% of the organic sulfur from coal in a continuous flow reactor [13]. This sulfur would have been covalently bound within the coal matrix, primarily in thiophenic structures. The matrix of vulcanized rubber consists of carbon chains crosslinked by sulfide, disulfide, and polysulfide linkages. Bacterial attack appears to be limited to sulfur exposed at the surface of solid rubber particles, which is oxidized to sulfoxides and sulfones, and eventually released as sulfate [14]. The alkane-degrading bacterium R. erythropolis ATCC 13260 (originally reported as Nocardioides simplex) is able to degrade a high molecular weight fraction of crude oil [15]. This fraction contains 14.7% sulfur, and was known from prior work to have hydrocarbon subunits linked by sulfide bridges. R. erythropolis ATCC 13260 degraded sulfur-bound linear alkanes and steranes in this oil fraction, leaving oxidized sulfur-bound species such as carboxylic acids. Sulfur-specific oxidation to sulfones was also observed, but no carbon-sulfur bond cleavage or desulfurization was reported [15]. In some reports, treatment of heavy crude oils with thermophilic bacteria led to an apparent enrichment in the lighter fractions of the oil [16-18]. This shift in composition was attributed to depolymerization of the asphaltene fraction of the oils, which was defined as the dissociation of small molecules either physically associated with or weakly chemically bound to asphaltenes. No significant quantitative change in the asphaltene content was measured. In addition, uniform removal of the range of sulfur compounds present in the oil was reported, which is not consistent with known chemical or biological conversion processes. Problems with sample recovery could account for some of the observed changes in oil composition, however in the absence of appropriate

116

controls allowing complete material or sulfur balances, definite conclusions cannot be drawn from this work. Although these reports are mostly encouraging, and suggest that conversion of high-molecular weight components of crude oil may be achieved by biological means, practical experience with asphaltic materials suggests that conversion rates may be very low. Asphalt (or bitumen in Europe) is widely used in paving materials, building materials, and waterproofing of foundations and roofs because of its resistance to degradation by natural organisms in soil and water, and by photooxidation. In this chapter, we first consider the chemical structure of the high-boiling components of crude oil, then examine the reactions that would enhance the value of these materials and consider the evidence for achieving such transformations by adapting natural biological processes. 3. UPGRADING NEEDS AND OPPORTUNITIES The chemical goals of heavy oil upgrading encompass molecular weight reduction of residue fractions to distillate materials, hydrogenation to increase the hydrogen to carbon (H/C) ratio, and removal of heteroatoms, in particular sulfur and nitrogen [19]. We will define the potential scope of biological oil upgrading more broadly, to include all activities which make the material easier to produce and transport, as well as the chemical changes which increase the value of the oil. These activities could therefore be applied to in situ treatment, production, transportation, and processing of crude oils. Five key areas of heavy oil upgrading where biological treatment could have an impact are viscosity reduction, composition improvement, deposition control, de-emulsification, and naphthenic acids removal. 3.1. Viscosity reduction Heavy oils are currently diluted with light hydrocarbons to reduce viscosity and allow transportation by pipeline to processing facilities. Natural gas condensate is typically used as the diluent, and is currently available as a steady supply. Precipitation of asphaltenes in the pipeline can occur due to the aliphatic nature of the diluent, but this approach readily achieves the viscosity reduction needed and is generally accepted. The production of heavy oil is expected to increase over the next several years and will exceed the availability of the diluent, so an alternative or supplemental treatment will be required. The viscosity of heavy oil is a result of interactions among the heaviest molecules in the oil, the asphaltenes. These interactions include entanglement of the alkane chains [19] as well as more ordered interactions between the aromatic clusters leading to structure formation throughout the oil [20]. One potential biotechnological approach to viscosity reduction is emulsification of the oil

117

using bioemulsifiers such as emulsan. This is discussed in further detail in Chapter 9. Breaking the asphaltenes into smaller molecules should also reduce molecular interactions leading to a reduction in viscosity. Thermal processing (mild thermal cracking, or visbreaking) can achieve some reduction in viscosity by breaking up some of the aliphatic structures in the asphaltenes, but the products can be unstable in downstream processing operations [19]. Microbial cleavage of aliphatic sulfides to reduce molecular weight and viscosity is the subject of research in our laboratory and this approach will be discussed further later in this chapter. 3.2. Composition improvement Many of the upgrading needs of traditional crude oils are also applicable to heavy crude oils. These include removal of sulfur, nitrogen, and metals, aromatic ring cleavage, and hydrogenation. Molecular weight reduction is also required to improve the fractional composition and value of heavy crude oils. 3.2.1. Heteroatom removal Sulfur, nitrogen, and metals present in crude oils are problematic for refining operations since they are poisonous to the catalysts used. Sulfur and nitrogen removal is also required to meet governmental emissions regulations when the refined fuels are burned. The application of biotechnology to the removal of these elements is discussed in Chapters 3 and 4. 3.2.2. Aromatic ring cleavage The presence of aromatic hydrocarbons has adverse effects on production and processing of petroleum, and combustion of fuels rich in aromatic hydrocarbons contributes to soot formation and poor combustion characteristics (for example, in diesel engines). Aromatics are commonly cracked during conventional upgrading by high temperature, high pressure catalytic hydrogenation to saturate and break the aromatic rings, but this is a costly process in terms of operation and capital. A proposed biological alternative would employ whole cell biocatalysts and two-phase (oil-water) reactions to specifically oxidize one or more rings of the aromatic substrates present in crude oil or middle distillate fractions. Enzymatic ring cleavage without carbon loss would produce polar compounds soluble in the water phase [21-24]. These would be recovered for chemical hydrogenation under mild conditions to yield alkylaromatics with improved combustion characteristics compared to the parent compounds. Alternatively, enzymatic hydroxylation of the aromatics with subsequent chemical hydrogenation and hydrogenolysis in the aqueous phase [25] would yield cycloalkylaromatics sensitive to further thermochemical bond cleavage. It is proposed that the cost savings of conducting such processes under

118

near-ambient temperature and pressure would make biologically assisted aromatic ring opening an economically feasible adjunct to conventional upgrading technology. This potential treatment is reviewed in Chapter 5. 3.2.3. Hydrogenation Typical H/C ratios for bitumens and residues range from 1.4-1.6 mol/mol. Hydrogenation is required to increase the H/C ratio of these feeds to a level suitable for transportation fuels (diesel and jet fuels, around 1.8 mol/mol) [19]. The primary target is the aromatics, including the heterocyclic sulfur and nitrogen species. The use of microorganisms specifically for aromatic ring hydrogenation has not been explored, although ring hydrogenation has been observed in the biodegradation pathways of some aromatic compounds. The explosive 2,4,6-trinitrotoluene (TNT) is subject to biotransformation in a variety of anaerobic and aerobic bacterial systems, as well as fungal systems (reviewed in Ref. [30]). In some aerobic bacteria, the initial reaction is hydrogenation of the ring, forming hydride- and dihydride-Meisenheimer complexes (Fig. 2a) [26, 27, 31]. Hydride-Meisenheimer complexes are similarly formed in the biodegradation of picric acid (2,4,6-trinitrophenol) [32, 33]. For the better-characterized picric acid system, these reactions are catalyzed by a hydride transferase enzyme, with NADPH serving as the hydride source via reduced coenzyme F-420 [34, 35].

Fig. 2. Examples of hydrogenation reactions in the biodegradation of (a) TNT [26, 27] and (b) naphthalene [28, 29].

119

Ring saturation is also observed in the biodegradation of aromatic hydrocarbons under anaerobic conditions. In the metabolism of benzoate through the benzoyl-CoA pathway (reviewed in Ref. [36]), stepwise saturation of the ring precedes ring cleavage and mineralization. Under sulfate-reducing conditions, naphthalene is activated by carboxylation to form 2-naphthoic acid [29]. Before ring cleavage and mineralization, 2-naphthoic acid is hydrogenated starting with the unsubstituted ring, eventually forming decahydro-2-naphthoic acid (Fig. 2b) [28]. Water is used as the source of protons for these reactions. Phenanthrene is similarly activated through carboxylation [29], but further reaction steps have not been identified. Neither of these systems has been studied for the specific goal of ring hydrogenation. They are presented here to illustrate that this type of reaction does occur. Unanswered questions in the existing literature are whether hydride transferase enzymes exist that are active towards hydrocarbons, as opposed to nitroaromatics, and whether the anaerobic hydrogenation of activated naphthalene also occurs in larger or alkylated ring systems. The ability of bacteria to transfer protons from water to aromatic hydrocarbon derivatives is a sharp contrast to the hydrogen-using catalytic and thermal processes used in traditional upgrading, and deserves further research attention. 3.2.4. Molecular weight reduction Molecular weight reduction is required to convert the residue fraction of heavy crude oils (materials boiling at temperatures over 525°C) to distillates (boiling at temperatures under 525°C). Cracking of aliphatic C-S bonds contributes to molecular weight reduction, but cracking of the C-C bonds found in alkyl bridges is necessary to achieve the full reduction required. For primary upgrading of heavy crude oils, thermal treatment is used for cracking operations, using temperatures over 420°C. The usefulness of chemical catalysts in primary upgrading is limited due to excessive catalyst fouling and poisoning [19]. Chemically, alkanes are the least reactive of the hydrocarbons. Nevertheless, aerobic bacterial biodegradation of n-alkanes is well known [37], and has been documented for chain lengths from Ci (methane) to at least C36 (hexatriacontane) [38]. Branched isoprenoid alkanes such as pristane (2,6,10,14tetramethylpentadecane) are also biodegradable [37]. The most common mechanism involves activation of the alkane through addition of molecular oxygen to the terminal methyl group by a monooxygenase enzyme to form a primary alcohol (Fig. 3a). Subsequent oxidation to a carboxylic acid allows further degradation through central fatty acid metabolic pathways. This method of degradation of alkyl chains therefore requires a free methyl terminus, and CC bond cleavage only occurs two carbons away from the end of the molecule through P-oxidation [37].

120

Fig. 3. Representative biodegradation pathways of alkanes showing initial activation and transformation to a carboxylic acid, (a) Aerobic monooxygenation of H-dodecane [37]. (b) Anaerobic addition ofn-dodecane to fumarate [39]. (c) Aerobic oxidation and cleavage of cyclododecane [40].

Anaerobic bacteria do not have molecular oxygen available for activation of non-functionalized hydrocarbons. Alternate activation and biodegradation mechanisms have been the subject of intense research over the past 15 years (as reviewed in Refs. [41-44]). Sulfate, nitrate, or iron-reducing bacteria may activate hydrocarbons through carboxylation or by addition of a C-H from the hydrocarbon across the double bond of fumarate to form a substituted succinate. The latter reaction is well-established for toluene, and may also occur for w-xylene, /?-isopropyltoluene, and ethylbenzene [43]. This reaction has also been observed as the activation route for «-alkanes including C4-C8 n-alkanes

121

under nitrate-reducing conditions [45, 46] and w-dodecane [39] under sulfatereducing conditions. The addition of fumarate to alkanes does not occur at a terminal methyl C-H, but rather at either a C2 or C3 subterminal methylene C-H [39, 46], producing a branched dicarboxylic acid (Fig. 3b), which is degraded through fatty acid metabolism. As with the aerobic pathway described above, anaerobic alkane degradation therefore proceeds from the terminus of the molecule. A subterminal attack on long-chain «-alkanes may occur in some aerobic bacterial cultures. A mutant Rhodococcus strain, designated KSM-B-3M, accumulated c«-unsaturated metabolites of «-hexadecane, 1-chlorohexadecane, and heptadecanonitrile, which were not growth substrates [47]. (The first two compounds did support growth of the wild-type strain.) In all three compounds the unsaturation was at position 9. Unsaturated products were also detected for 1-hexadecanol, 1,2-epoxyhexadecane, hexadecyl benzene, and hexadecyl chloroformate. The mutant had likely lost the ability to cleave the alkane chain at the unsaturated bond, resulting in the inability to grow on these substrates [47]. The degradation of phytanyl octadecyl ether by a mixed soil culture and by Rhodococcus ruber (DSMZ 7512) also showed evidence of an initial subterminal dehydrogenation [48]. Degradation occurred initially on the linear side chain, and initial degradation products were the phytanyl ethers of C2 to C8 primary alcohols. The corresponding carboxylic acids appeared next as the alcohols disappeared from the cultures. The final products were the phytanyl ethers of acetic acid and propanoic acid. One other metabolite was observed, phytanyl octadec-9-enyl ether. The formation of unsaturated products was proposed to be analogous to the dehydrogenation of Cig fatty acids in the cell membrane, which also occurs at position 9. The alternate degradation pathway proposed starts with the observed internal dehydrogenation, followed by a hypothesized olefinic oxidation to a secondary alcohol, oxidation to a ketone, Baeyer-Villiger oxidation to an ester, ester cleavage, and P-oxidation [48]. A subterminal attack of this type has not been shown for a diterminally substituted alkyl chain, but a similar pathway has been shown as the mechanism of degradation for both small and large cyclic alkanes. Both cyclohexane (reviewed in Ref. [49]) and cyclododecane [40] are oxidized via an alcohol to a cyclic ketone. The ketones are oxidized to lactones by Baeyer-Villiger monooxygenases, followed by ester cleavage to an co-hydroxycarboxylic acid (Fig. 3c) and oxidation to a dicarboxylic acid [40] that can be degraded through central metabolic pathways. The Baeyer-Villiger monooxygenases appear to have fairly narrow substrate specificities. Cyclododecanone monooxygenase from R. ruber strain CD4 could also oxidize cyclopentadecanone, but not cyclohexanone or cyclooctanone [40]. In growth assays, R. ruber strain SCI, isolated on cyclododecanone, could also grow on Ci5, C13, C n , and C10 cyclic

122

ketones, but not on C8, C7, or C6 cyclic ketones [50]. Conversely, cyclohexanone monooxygenases are known to favour shorter chain cyclic ketones. For example, two enzymes from Brevibacterium sp. strain HCU could oxidize C4-C7 cyclic ketones, but not C 8 -C| 2 compounds [51]. Molecular weight reduction in the residue fraction of heavy oils requires cleavage of alkyl bridges, where both ends of the carbon chain are blocked by attachment to aromatic groups. The more common aerobic and anaerobic bacterial alkane-degradation pathways are not appropriate for molecular weight reduction in crude oil, because they only activate the free end of the molecule to create a fatty acid for central metabolic pathways. More relevant research has been done with long-chain «-alkanes and cycloalkanes. This work shows that an alkyl chain can be cleaved through bacterial attack in the absence of a terminal methyl group. This reaction is more directly analogous to the alkyl bridges found in high molecular weight crude oil components, and is a promising avenue for further work. 3.3. Deposition control Both asphaltenes and waxes may cause deposition problems in the reservoir, pipelines, and storage and processing equipment. Asphaltenes deposit due to an increase in the aliphatic content of the oil, while waxes crystallize and precipitate due to a drop in temperature (e.g. from the reservoir to the surface, [52]) or an increase in aromaticity of the bulk oil. Changes in solvency occur due to dilution or to blending of different oils [19]. Both types of compounds may co-precipitate, through entrapment of one type in a deposit of the other. Generally, a wax content greater than 2% by weight is found to lead to wax deposition problems [3]. Deposition prevention is accomplished through chemical treatment to maintain the molecules in solution, as well as through temperature and flow control. Waxes are a valuable feedstock for refinery operations, so prevention of wax deposition is important to preserve the value of the oil as well as to avoid operational problems. Existing deposits are removed through circulation of hot water, hot oil, solvents, and surfactants, or through "pigging" of transfer lines [53]. In the case of asphaltenes, treatments include addition of aromatic streams to dissolve the deposits, or the addition of dispersants to prevent flocculation of the asphaltenes into particles that subsequently deposit on surfaces. Although biological treatments for deposition control are commercially available, little scientific literature is available in this area [54, 55]. Three modes of biological activity are conceivably relevant to deposition control: production of metabolites (from carbon sources other than the oil) which improve the solubility of either waxes or asphaltenes, biotransformation of waxes and asphaltenes to more soluble products (through molecular weight reduction or functionalization), and biodegradation to remove the problematic compounds

123

either from the oil or from existing deposits. The ability of bacteria to degrade solid alkanes is limited by mass transfer rates. For instance, liposome encapsulation was required to achieve biodegradation of hexatriacontane (n-C^) by a Pseudomonas isolate which did not grow on the crystalline compound [38]. The usefulness of biological treatments for removal of deposits may therefore be limited to the production of solubilizing agents rather than direct transformation or degradation of the crystallized molecules. Isolated bacteria and consortia from paraffin deposits, hydrocarbon contaminated soils and waters, and brine have been shown to produce biosurfactants, as well as to degrade hydrocarbons from samples of paraffin deposits and paraffinic oils [56]. In a flow system, a consortium of these bacteria decreased the paraffin content of a heavy oil. The treated oil also had a lower freezing point and a decreased low temperature viscosity, but the effect of these changes on deposition in the flow system was not reported [56]. To the extent that microorganisms adsorb wax or asphaltic material, then bacteria could serve to disperse the deposits and prevent deposition on surfaces, however, no systematic research has been conducted in this area. 3.4. Emulsion behaviour and de-emulsiflcation Water-in-oil (W/O) and oil-in-water (O/W) emulsions occur throughout oil production, transportation, and processing. The water may be from the formation or may be added through water or steam injection to improve oil recovery, or addition of wash water in desalting operations. Emulsions may be produced incidentally through handling or deliberately to improve flow properties for enhanced oil recovery and transportation [57]. Desirable emulsions produced for pipeline transportation are O/W emulsions, usually containing around 30% aqueous phase [58]. Undesirable O/W emulsions are typically found in waste waters from the oil industry. Although deemulsification does recover some oil, treatment is generally driven by environmental concerns rather than economic incentive. On the other hand, resolution of W/O emulsions improves the quality of the oil, and is therefore economically driven [53]. Problems associated with water in oil include corrosion, scale formation, sludge accumulation in storage tanks, altered viscosity and flow properties, and reduced distillation efficiency [53]. Regardless of the source, emulsions must be resolved at some point before refining. This is accomplished through heating, settling, centrifugation, filtration, electrical dehydration, and chemical treatment. The pipeline specification includes both solids and water, and is typically a maximum of 0.5% bottom solids and water (BS&W) [58]. Emulsions are formed from two immiscible phases through mixing to produce a fine dispersion of droplets of one phase in the other, where the interface between the two phases is stabilized by emulsifying components. The

124

energy added through mixing is essential, since the emulsified state is not usually thermodynamically stable. Emulsifying agents associate at the interface of the two phases and impart kinetic stability to the emulsion, either through reduction of interfacial tension (chemical stabilization), or by providing a barrier to coalescence (physical stabilization). Resolution of emulsions, or deemulsification, proceeds via two steps: flocculation or aggregation of droplets, and coalescence of droplets to form a continuous second phase. De-emulsifiers may promote one or both of these phenomena [58]. Crude oil emulsions are complex, and vary from location to location. The emulsifying agents may be amphiphilic molecules from the oil, especially the resin fraction, including naphthenic acids. Many crude oil emulsions are stabilized by fine solids, including clays, scale, or wax crystals [59], or bacteria themselves [60], which present a barrier to droplet coalescence. Asphaltenes are especially important in heavy crude oil emulsions. After association with the interface, asphaltenes agglomerate to form a skin, which prevents coalescence of droplets. Resins are also believed to play a part in stabilizing this skin [58]. Complex emulsion structures, such as water-in-oil-in-water emulsions, have also been observed [61]. De-emulsification in the oil industry is challenging due to the variety of possible emulsion properties, and treatments are currently tailored to each site and adapted over time [59]. Biological de-emulsification has been studied using a variety of microorganisms. Whole bacterial cells have received the most research [62-69], but Streptomyces spores [70], bacterial metabolites [71], and yeast cells [64] have also been studied. The organisms and emulsion systems used are summarized in Table 1. The majority of studies have examined model, chemically stabilized emulsions consisting of water, hydrocarbon, and a commercial surfactant. This research has allowed some assessment of the mode of action. De-emulsification ability appears to be associated with the surface of the bacterial cells. Depending on their hydrophobicity, cells may aggregate at the oil-water interface, promoting flocculation and coalescence of droplets [72]. Differences in hydrophobicity may account for changes in effectiveness of microbes in different growth phases, as well as for differing abilities to resolve O/W or W/O emulsions. In general, it appears that more hydrophilic cells are required to treat W/O emulsions, while relatively more hydrophobic cells are able to resolve O/W emulsions [62, 65, 66, 69, 70]. The ability of bacterial cells to de-emulsify both model and oilfield O/W and W/O emulsions has been demonstrated, but the potential for treating the true spectrum of real crude oil emulsions has not been rigorously tested. As with chemical treatments, no single biological treatment will likely be effective for all chemically stabilized crude oil emulsions. Biological products may still be a valuable complement to existing chemical technologies.

125

Table 1 Biological systems shown to de-emulsify oil-water emulsions Organism Emulsion system Comments Nocardia amarae strain LL-Se6 (ATCC 27808)

Refs.

O/W emulsions: Alkanes / water Kerosene / water Oilfield emulsions W/O emulsions: Water / kerosene Oilfield emulsions

• Older, more hydrophobic cultures more effective

[62, 65, 66, 69]

Corvnebacterium petrophilum (ATCC 21404)

W/O emulsions: Oilfield emulsions

• Younger cultures (exponential growth phase) more effective

[64, 691

Micrococcus sp.

O/W emulsions: Kerosene / water W/O emulsions: Water / kerosene

• More effective for O/W emulsions • Solvent washing increased O/W, decreased W/O de-emulsification

[631

Mixed aerobic bacterial culture

W/O emulsions: Water / kerosene Oilfield emulsions

• More effective when grown on crude oil or motor oil than on carbohydrates

[67, 68, 731

Streptomvces sp. strain AA8321

O/W emulsions: Kerosene / water Alkanes / water Diesel / water Gasoline / water Paraffin oil / water Soybean oil / water

• Only aerial spores were effective • Effectiveness increased with culture age and hydrophobicity

[701

Bacillus subtilis

W/O emulsions: Water / crude oil

• Free acetoin in medium identified as active component

[711

Torulopsis bombicola (ATCC 22214)

W/O emulsions: Oilfield emulsions

• Rate increased with cell concentration

[641

• Younger cultures (exponential growth phase) more effective

The applicability of biotechnology to asphaltene- or solids-stabilized emulsions has not been studied. Biocatalysis or biologically produced chemicals may be effective in removing or dispersing asphaltenes or wax crystals, particularly in combination with suitable cell-surface properties to aid in dispersion of the solids or in aiding flocculation as appropriate.

126

3.5. Naphthenic acids Naphthenic acids are found in varying concentrations in crude oils worldwide. They are a family of carboxylic acids defined by the formula CBH2n.z02, where n is the carbon number and Z is related to the number of rings. Characteristic features of naphthenic acids, illustrated by the examples in Fig. 4, are saturated rings containing five or six carbon atoms, a carboxyl group separated from the rings by at least one methylene group, and an alkyl substitution [75]. Naphthenic acids contribute to the total acidity of a crude oil, typically expressed as the total acid number, or TAN (mg KOH required to neutralize 1 g of oil). TAN values of 0.5 mg or greater are generally correlated to high corrosivity, although the total corrosivity of an oil is also affected by factors such as sulfur content, flow conditions (velocity and turbulence), and temperature. Naphthenic acid corrosion typically occurs at processing temperatures between 220°C and 400°C, which corresponds to the boiling range of these compounds. The effects of corrosive oils are usually addressed through careful selection of materials of construction [76]. In oil sands processing, hot caustic solutions are used to separate the bitumen from the sand. Due to the high pH, naphthenic acids preferentially partition to the aqueous phase, and are discharged with the water into the tailings ponds. Naphthenic acids are believed to account for the high acute toxicity of the tailings waters [77]. Research into biodegradation of naphthenic acids has been pursued out of interest in remediation and reclamation of the tailings ponds. Early research looked at aerobic degradation of simple model compounds, including cyclohexane and cyclopentane carboxylic acids, 1- and 2methylcyclohexane carboxylic acids, cyclohexane pentanoic acid, 4pentylcyclohexane carboxylic acid, and decahydro-2-naphthoic acid [75, 78, 79]. These studies showed that mineralization of these simple compounds was possible, starting with the carboxylated side chain but also including the ring structures [78]. Alkylated compounds, representative of true naphthenic acids, were more resistant to degradation [75, 78]. Direct analysis of the naphthenic acids found in the tailings waters is challenging, due to the complexity of the mixture. Biodegradation of naphthenic acids extracted from tailings water was shown under aerobic conditions through both CO2 production and a reduction in acute toxicity [78]. Under methanogenic conditions, model compounds (cyclohexyl propanoic, butanoic, and pentanoic acids, and 6-phenyl hexanoic acid) were substrates for methanogenesis. Extracted and commercial naphthenic acids mixtures, however, delayed the onset of methanogenesis, and were not apparently methanogenic substrates [80]. More recently, gas chromatography with electron impact mass spectrometry has been used to resolve distinct ion fragments from naphthenic acid mixtures. The abundances of these ions can be allocated to specific carbon and Z-numbers, giving a 3-dimensional representation of the distribution of

127

naphthenic acids in a sample. This allowed direct observation of the differences among samples, and showed changes occurring through biodegradation [81]. Other recent advances in naphthenic acids research include a statistical method to show significant differences among samples analyzed using the mass spectrometric technique [74] and a high performance liquid chromatography method developed for quantitation of total naphthenic acids [82]. These methods were used to definitively show biodegradation of commercial naphthenic acids mixtures by aerobic enrichment cultures, accompanied by growth, CO2 production, and elimination of acute toxicity [83]. These commercial mixtures differ from the naphthenic acids found in oils and tailings primarily in a lack of compounds with carbon numbers of 22 or greater. It remains to be shown the extent of biodegradation possible in authentic tailings waters, and whether the larger naphthenic acids will be affected. The aim of bioremediation research differs from biological upgrading for naphthenic acids. Bioremediation requires the complete removal of these compounds. Biological upgrading would ideally involve biotransformation of naphthenic acids to compounds that are easier to handle and more valuable. The reactions observed in the biodegradation of naphthenic acids (side chain and ring oxidation and mineralization to CO2) are therefore not directly applicable to oil upgrading. The work done clearly shows that despite their toxicity, naphthenic acids are ready targets for microbial attack. Future upgrading research needs to look for more suitable reactions and for systems capable of catalyzing those reactions.

Fig. 4. Representative naphthenic acids structures (based on Ref. [74]). (R - alkyl group; m number of carbons in the side chain excluding the carboxyl group)

128

3.6. Opportunities and research gaps The ability of cells to physically interact with crude oil components at the interface raises an important issue in studies of bioprocessing. How can we ensure that proper measurements of oil composition are taken to ensure accurate assessment of conversion, if oil components can interact strongly with the cell surfaces by physical processes? The most rigorous measurements will rely on a material balance on the compound of interest, where the fate of the constituents in the oil is accurately accounted for in the reacting mixture. This approach works well with 14C labelled compounds in mineralization studies, but it is less applicable to complex fractions of heavy crudes. Balances on sulfur provide one approach, where the disappearance of sulfur compounds from the oil is confirmed by the appearance of other sulfur compounds in solution, such as sulfate. Similar balances can be attempted for the other reactions discussed above, to confirm not only that the substrate is disappearing, but also that its transformed product is detected stoichiometrically. The lack of such rigorous controls is likely at the heart of some claims for bacterial conversion of crude oil components under a range of conditions [16-18]. Microbes have the potential to aid in processing or transportation of heavy crude oils if interesting reactions can be harnessed, or if the surface properties of the microbes can be used to aid in flocculating emulsions or dispersing components. The natural environment of bacteria is either at the oil/water interface or in the bulk water phase. The most attractive opportunities for biological upgrading may therefore be in dealing with surface active components such as naphthenic acids or in flocculating oil-in-water emulsions. Interesting reactions such as hydrogenation and ring-opening deserve further study, but are likely to be extremely slow if the substrate consists of large, complex molecules such as vacuum residues or asphaltenes. 4. VISCOSITY AND ALIPHATIC SULFIDE CLEAVAGE 4.1. Viscosity correlations Experimental data show that the viscosity of oil is correlated to the average molecular weight of the material. Fig. 5a shows data compiled from different sources, including whole oils, bitumen, distillates, and residues. The observable trends are towards higher viscosity in heavier samples and towards lower viscosity at higher temperatures. The scatter in Fig. 5a indicates that there are factors involved other than molecular weight and temperature. Some published models include properties such as specific gravity to account for this variability [84, 85]. Although viscosity models fit the data used to generate them, they are often difficult to extend to other samples due to these other contributing effects. The general correlation to molecular weight appears to be sound and can be used as a basis for further analysis of viscosity.

129

Fig. 5. Viscosity data for whole oils, residuals, and distillates [88], oil sand bitumens and topped crudes [84], oil fractions [85], synthetic crude oils [86], and crude oils and natural bitumens [20] showing correlation to (a) average molecular weight and (b) asphaltene content. (Analysis temperatures are indicated in the legends)

130

The correlation of viscosity and molecular weight indicates that there should also be a correlation between the viscosity and the fractional composition of the oil. Fig. 5b shows that viscosity increases with the asphaltene content (weight %), the asphaltenes being among the heaviest molecules in the oil. This type of correlation has been used to formulate viscosity models based on logarithmic mixing rules, assigning "pseudo-viscosities" to the different fractions of the oil and applying a weighting factor to each one [86, 87]. Sulfides, ethers, and esters have been identified as common linking structures in the aliphatic bridges found in asphaltenes [11]. Nickel boride desulfurization was used to specifically cleave aliphatic sulfide bonds in two asphaltene fractions, giving a 4-fold reduction in the molecular weight of the higher molecular weight fraction. This indicated that aliphatic sulfides were involved in the linking structures of the molecules, including linkages between aromatic cores and to smaller structures like alkanes and steranes. The total sulfur content of asphaltenes includes the sulfide bridges (aliphatic), cyclic sulfides (aliphatic heterocycles, found as substituted thiolanes and thianes [89]), and thiophenic sulfur (aromatic sulfur heterocycles). Only cleavage of the sulfide bridges leads to a reduction in molecular weight, since removal of the cyclic sulfides and thiophenes leaves the carbon backbone intact.

Table 2 Selected organic sulfur compounds successfully used for enrichment of microorganisms able to use the compounds as sole sulfur source under sulfurlimited conditions Compound Procedures used Refs. Ametryne prometryne (herbicides)

and • Culture maintenance alternated between [90] selective liquid and non-selective solid media

Naphthalenesulfonic acids Benzenesulfonic acids (detergents)

• Substrate purification by high pressure liquid [91] chromatography • "Scrupulously clean glassware" (procedure not given)

Organic sulfur in coal

• Effluent from reactor mutagenized and [92] reinoculated to accelerate strain evolution

Endosulfan (an insecticide)

• Use of an Escherichia coli culture to [93] scavenge sulfate from medium, followed by filter sterilization to produce a sulfur-free medium

131

4.2. Biological sulfur requirements Sulfur is an essential element for bacterial growth, although total sulfur requirements are low. Sulfur limitation has been successfully applied to the enrichment of microorganisms active towards a variety of organosulfur compounds. Several examples are given in Table 2. All but one of the studies listed mentioned special procedures used to minimize the effect of contaminant sulfur. Aliphatic sulfides have not been studied in oil desulfurization research. The relatively low carbon-sulfur bond strength results in easy cleavage of these bonds under thermal treatment compared to thiophenic sulfur, which is only removed during catalytic hydro treating [19]. Correspondingly, research on biological desulfurization of oil has focused on the recalcitrant thiophenic compounds rather than the aliphatic sulfides (see chaper 2). A selection of the bacterial strains known to desulfurize dibenzothiophene (DBT) is given in Table 3. The substrate ranges of these organisms for organic sulfur sources frequently include some alkylated species, and sometimes extend to compounds with aliphatic sulfide bonds as well. The biodegradation of some aliphatic sulfides not related to oil has been studied, including dimethyl sulfide (DMS) and analogues of sulfur mustard (2,2'-dichlorodiethyl sulfide). Biodegradation of some larger sulfides has been reported as well. These will be presented here to illustrate the types of biological activity possible with aliphatic sulfides, and to show how they may be relevant to oil upgrading. 4.3. High molecular weight sulfides Phytanyl octadecyl sulfide was used as a model compound for sulfurbound hydrocarbons found in heavy oil macromolecules [15]. Biodegradation as a carbon source by R. erythropolis ATCC 13260 occurred only on the linear octadecyl chain, and not the branched phytanyl chain. Six chain degradation metabolites were identified (Fig. 6), which suggested two different mechanisms [15]. Metabolites with an even number of carbon atoms in the linear side chain were proposed to result from terminal oxidation followed by |3-oxidations removing two carbon atoms at a time. Metabolites with an odd number of carbon atoms cannot arise solely from P-oxidations, and an initial mid-chain oxidation was proposed to occur as well. Oxidation of the sulfur atom to a sulfone was observed both in the parent compound and in the degraded metabolites, indicating that sulfur oxidation was independent of the chain degradation pathways. No evidence of carbon-sulfur bond cleavage was reported [15].

132

Table 3 Selected dibenzothiophene desulfurizine bacteria and alternate sulfur sources

Bacterial use of a high molecular weight aliphatic sulfide as a sulfur source has only recently been reported. Rhodococcus sp. strain JVH1 is capable of using the novel compound fe-(3-pentafluorophenylpropyl) sulfide (PFPS) as its sole sulfur source for growth [104]. PFPS was specifically designed using ring fluorination to block any terminal attack on the molecule, necessitating a subterminal attack to support growth. The desulfurization pathway proposed is shown in Fig. 7. PFPS is first oxidized to the corresponding sulfoxide and sulfone (PFPSO and PFPSO2). Carbon-sulfur bond cleavage then yields the primary alcohol 3-pentafluorophenylpropan-l-ol, which is further oxidized to 3pentafluorophenylpropanoic acid. The second product of PFPSO2 cleavage was proposed to be a sulfmate, analogous to the 4S pathway for DBT desulfurization, but was not directly observed. Release of the sulfur as sulfite

133

was also hypothesized, but not directly observed. JVH1 was shown to use a variety of compounds with aliphatic carbon-sulfur bonds as sulfur sources (including dialkyl sulfides, thiacycloalkanes, and aryl-terminated sulfides), but not thiophenic compounds. This selective ability to cleave compounds with aliphatic carbon-sulfur bonds is extremely interesting for research into biological viscosity reduction in heavy crude oils.

Fig. 6. (a) Structure of phytanyl octadecyl sulfide. (b) Metabolites produced by Rhodococcus erythropolis ATCC 13260 and proposed reactions in the degradation of phytanyl octadecyl sulfide [15].

134

Fig. 7. Proposed pathway of PFPS metabolism in Rhodococcus sp. strain JVH1 [104]. Compounds in brackets were not directly observed. (PFPP-OH - 3-pentafluorophenylpropanl-ol; PFPP-acid - 3-pentafluorophenylpropanoic acid).

For high molecular weight aliphatic sulfides, the mechanism of attack appears to depend on the substituent groups. Sulfur-bound alkyl chains are subject to aerobic degradation, apparently through the same pathway as «-alkanes. Sulfur oxidation occurred independently, but did not prevent chain degradation. A 4S-like pathway has been reported for the fluorinated compound PFPS. Interestingly, DBT-desulfurizing strains only produced PFPSO2, being apparently unable to cleave the aliphatic carbon-sulfur bonds in PFPS [104]. This illustrated that sulfur-specific desulfurization of aliphatic sulfides and thiophenes may occur through analogous mechanisms, but that the desulfurization systems are not necessarily interchangeable.

135

4.4. Dimethyl sulfide DMS is part of the global sulfur cycle. It is formed in marine sediments from degradation of dimethylsulfoniopropionate, produced by marine algae and plants. Some DMS is released from the oceans to the atmosphere where it is involved in cloud formation, while most is degraded by a variety of marine microorganisms [105]. The pathway most studied, shown in Fig. 8a, occurs under aerobic conditions in a variety of species of Hyphomicrobium and Thiobacillus [106]. Initial cleavage to methanethiol and formaldehyde is catalyzed by a NADH-dependent monooxygenase. Methanethiol is then cleaved to a second molecule of formaldehyde and hydrogen sulfide by methanethiol oxidase. Sulfide is further oxidized to sulfate. Formaldehyde is oxidized to formate, then to carbon dioxide by formaldehyde dehydrogenase and formate dehydrogenase. Thiobacillus sp. strain ASN-1 can degrade DMS under both aerobic and anaerobic (nitrate-reducing) conditions [107]. Methanethiol, but not formaldehyde, was produced under aerobic conditions, suggesting a novel pathway for DMS degradation in this organism. The same degradative pathway was proposed for this organism under both aerobic and anaerobic conditions (Fig. 8b), with the terminal electron acceptors being oxygen and nitrate, respectively. Each methyl group is removed by a methyltransferase and oxidized to formate. The sulfur is first released as sulfide, which is oxidized to sulfate. This work was extended to larger sulfides, and it was shown that Thiobacillus sp. strain ASN-1 could grow on diethyl sulfide, dipropyl sulfide, dibutyl sulfide, dimethyl disulfide, and dibutyl disulfide [108]. Growth on butanethiol, as well as on acetate, propionate, and butyrate, suggested that similar reaction mechanisms were used for the larger compounds as for DMS. Lag periods were observed when transferring cultures from one sulfide compound to another, but not from a sulfide to the corresponding thiol, indicating that different enzymes are used for the degradation of the different sulfides [108]. Anaerobic degradation of DMS has been observed in marine sediments. Methanogenic consortia from these environments release methane from DMS, as well as from methanethiol and dimethyl disulfide. Ethane release from the analogous compounds diethyl sulfide and ethanethiol has also been observed in marine sediment samples. This release was not observed in killed controls, or in the presence of 2-bromoethanesulfonic acid, which inhibits methanogenic bacteria [109].

136

Fig. 6. Dimethyl sulfide biodegradation pathways, (a) aerobic marine thiobacilli and hyphomicrobia [106]. (b) Thiobacillus sp. strain ASN-1 [107]. (c) Rhodococcus sp. strain SY1 [95]. (X -cobalamin carrier of methyltransferase)

The terrestrial bacterium Rhodococcus sp. strain SY1, originally isolated for its ability to use DBT as a sole sulfur source [95], can also degrade DMS as a sulfur source [96]. A pathway analogous to the sulfur-specific 4S pathway for DBT was proposed based on the observed products from growth on successive intermediates (Fig. 8c). The sulfur atom in DMS is first oxidized giving dimethylsulfoxide and dimethyl sulfone. Growth on dimethylsulfoxide also showed release of dimethyl sulfone, as well as methanol and methane. Release of the sulfur as sulfite, which is spontaneously oxidized to sulfate under aerobic conditions, was proposed but never directly observed. For the smallest aliphatic sulfide, DMS, degradation primarily occurs without oxidation of the sulfur atom. This has been observed under aerobic, nitrate-reducing, and methanogenic conditions. Some larger sulfides may be subject to similar reactions, with butyl sulfide being the largest compound tested. A 4S-like oxidative pathway has also been reported, allowing use of DMS as a sulfur source by a DBT-desulfurizing organism. This suggests that the DBT-desulfurizing enzymes of some bacteria may also catalyze aliphatic C-S bond cleavage. 4.5. Sulfur mustard analogues Sulfur mustard (2,2'-dichlorodiethyl sulfide) was first used as a chemical warfare agent during World War I. The United States military is currently

137

interested in degradation strategies for this compound for disposal of stockpiles at American military bases [110]. Biodegradation of sulfur mustard is not studied directly because of its high toxicity. Proposed disposal processes start with chemical neutralization followed by treatments such as incineration, chemical oxidation, or biodegradation. In the neutralization step, sulfur mustard is hydrolysed to less toxic compounds, primarily the dechlorinated compound thiodiglycol (TDG) [110], which has been the focus of most biodegradation studies. Alcaligenes xylosoxydans ssp. xylosoxydans SH91, a soil bacterium, can use TDG as its sole source of carbon and energy [111]. TDG is oxidized via [(2hydroxyethyl)thio] acetic acid to thiodiglycolic acid in two oxygen-dependent steps catalyzed by butanol dehydrogenase (Fig. 9a). Thiodiglycol sulfoxide also accumulates as a dead-end metabolite. Further metabolites have not been shown analytically, but growth on TDG implies incorporation into cell mass and mineralization to CO2 and sulfate. A second strain of A. xylosoxydans, designated PGH10, is able to transform TDG in minimal medium with citrate or fructose [112]. Degradation is proposed to follow the same initial pathway as for A. xylosoxydans ssp. xylosoxydans SH91, based on observation of the same two initial metabolites, [(2hydroxyethyl)thio] acetic acid and thiodiglycolic acid. Although this strain can transform TDG, there is no evidence for the release of carbon or sulfur from the metabolites. It appears that this strain may not be able to cleave the carbonsulfur bond, resulting in its inability to use TDG as a carbon source. Fungal degradation of TDG has also been studied. The white-rot fungus Coriolus versicolor (IFO 30340) degraded TDG most effectively in a highcarbon, low-nitrogen medium, and the brown-rot fungus Tyromyces palustris (IFO 0507) degraded TDG in sulfur-free medium [113]. Degradation was measured only as loss of TDG, not as the formation of any metabolites. Both organisms were also able to degrade the brominated sulfur mustard analog 2,2'dibromodiethyl sulfide, again measured as loss of the starting compound. One further study has investigated a sulfur-specific detoxification strategy for sulfur mustard. The DBT-desulfurizing bacterium R. erythropolis strain IGTS8 can use the sulfur mustard analog 2-chloroethyl ethyl sulfide as a sole sulfur source for growth [114]. Two metabolites were found using gas chromatography and mass spectrometry. Growing cultures of R. erythropolis strain IGTS8 accumulated 2-chloroethanesulfinic acid, while 2-chloroethanol was found in resting cell cultures (Fig. 9b). Neither compound was observed in killed cell controls.

138

Fig. 7. Metabolites formed in the biodegradation of the sulfur mustard analogues (a) thiodiglycol [111] and (b) 2-chloroethyl ethyl sulfide [114]. Products in brackets were not directly observed.

For the sulfur-mustard analogue TDG, both sulfur oxidation and terminal carbon oxidation were observed in a bacterium using the compound as a carbon source. These reactions were apparently independent and mutually exclusive with the sulfoxide produced accumulating as a dead-end metabolite. Carbonsulfur bond cleavage was assumed to occur subsequent to terminal carbon oxidation, but only in the absence of sulfur oxidation. Some fungal strains were also able to degrade sulfur mustard analogues, although metabolites were not identified. As with DMS, 2-chloroethyl ethyl sulfide was subject to sulfurspecific degradation by a DBT-desulfurizing strain, demonstrating again that the desulfurization enzymes may have a sufficiently broad substrate specificity to allow attack on both thiophenes and aliphatic sulfides. 5. CONCLUSIONS Known interactions between microbes and the high molecular weight components of crude oils include oxidation of aliphatic and aromatic carbon groups, oxidation of naphthenic acids, and oxidation and desulfurization of aromatic and aliphatic sulfur groups. Hydrogenation and dehydrogenation reactions have been demonstrated only on lower-molecular weight components. All of these reactions are of potential interest for upgrading heavy crude oils and bitumens, but a major barrier is the transport of reactants to the active site of reaction, particularly for intracellular enzymes in bacteria. Although membranes may give significant barriers for bioprocessing of heavy hydrocarbons, the

139

interactions of cell membranes with oil/water interfaces may be of interest in de-emulsifying oil and in dispersing asphaltenic material to prevent deposition. Acknowledgments. The authors would like to thank Dr. P.M. Fedorak and Dr. J.D. Van Hamme for access to and information on works in press. Funding was provided by Alberta Energy Research Institute under the COURSE program and by the Natural Sciences and Engineering Research Council of Canada. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

H. Groenzin and O.C. Mullins, Energy Fuels, 14 (2000) 677. L. Artok, Y. Su, Y. Hirose, M. Hosokawa, S. Murata, and M. Nomura, Energy Fuels, 13 (1999)287. N.X. Thanh, M. Hsieh, and R.P. Philp, Org. Geochem., 30 (1999) 119. J.G. Speight, Fuel, 49 (1970) 134. T.F. Yen, Prepr. - Am. Chem. Soc, Div. Pet. Chem., 17 (1972) F102. T.F. Yen, J.G. Erdman, and S.S. Pollack, Anal. Chem., 3 (1961) 1587. O.P. Strausz, T.W. Mojelsky, F. Faraji, E.M. Lown, and P. Peng, Energy Fuels, 13 (1999)207. O.P. Strausz, T.W. Mojelsky, E.M. Lown, I. Kowalewski, and F. Behar, Energy Fuels, 13(1999)228. P. Peng, A. Morales-Izquierdo, E.M. Lown, and O.P. Strausz, Energy Fuels, 13 (1999) 248. O.P. Strausz, T.W. Mojelsky, and E.M. Lown, Fuel, 71 (1992) 1355. P. Peng, A. Morales-Izquierdo, A. Hogg, and O.P. Strausz, Energy Fuels, 11 (1997) 1171. D.C. Bressler and M.R. Gray, Int. J. Chem. React. Eng., 1 (2003) R3. J.J. Kilbane, Resour. Conserv. Recycl., 3 (1990) 69. O. Hoist, B. Stenberg, and M. Christiansson, Biodegradation, 9 (1998) 301. A. Jenisch-Anton, P. Adam, W. Michaelis, J. Connan, D. Herrmann, M. Rohmer, and P. Albrecht, Geochim. Cosmochim. Acta, 64 (2000) 3525. E.T. Premuzic, M.S. Lin, and B. Manowitz, Fuel Process. Technol., 40 (1994) 227. E.T. Premuzic and M.S. Lin, J. Pet. Sci. Eng., 22 (1999) 171. E.T. Premuzic, M.S. Lin, M. Bohenek, and W.M. Zhou, Energy Fuels, 13 (1999) 297. M.R. Gray, Upgrading Petroleum Residues and Heavy Oils, New York, NY, 1994. A.N. Ratov, Pet. Chem., 36 (1996) 191. J.M. Foght, Q. Wu, P.M. Fedorak, M.A. Pickard, and M.R. Gray, in the Proc. of the Joint meeting of The Petroleum Society (48th Annual Technical Meeting) and BIOMINET, Calgary, AB, 1997, Paper 97-13. Q. Wu, P.M. Fedorak, M.A. Pickard, M.R. Gray, and J.M. Foght, Prepr. Pap. - Am. Chem. Soc, Div. Fuel Chem., 43 (1998) 515. Q. Wu, P.M. Fedorak, M.A. Pickard, M.R. Gray, and J.M. Foght, Prepr. Pap. - Am. Chem. Soc, Div. Fuel Chem., 44 (1999) 10. Q. Wu, M.R. Gray, M.A. Pickard, P.M. Fedorak, and J.M. Foght, Prepr. - Am. Chem. Soc, Div. Pet. Chem, 48 (2003) 47.

140

[25] C.L. Coyle, M. Siskin, D.T. Ferrughelli, M.S.P. Logan, and G. Zylstra, Biological activation of aromatics for chemical processing and/or upgrading of aromatic compounds, petroleum coal, resid, bitumen and other petrochemical streams, US Patent No. 6 156 946(2000). [26] C. Vorbeck, H. Lenke, P. Fischer, and H.-J. Knackmuss, J. Bacteriol., 176 (1994) 932. [27] C. Vorbeck, H. Lenke, P. Fischer, J.C. Spain, and H.-J. Knackmuss, Appl. Environ. Microbiol., 64(1998)246. [28] X. Zhang, E.R. Sullivan, and L.Y. Young, Biodegradation, 11 (2000) 117. [29] X. Zhang and L.Y. Young, Appl. Environ. Microbiol., 63 (1997) 4759. [30] A. Esteve-Nunez, A. Caballero, and J.L. Ramos, Microbiol. Mol. Biol. Rev., 65 (2001) 335. [31] C.E. French, S. Nicklin, and N.C. Bruce, Appl. Environ. Microbiol., 64 (1998) 2864. [32] P.-G. Rieger, V. Sinnwell, A. Preuss, W. Francke, and H.-J. Knackmuss, J. Bacteriol., 181 (1999)1189. [33] C. Behrend and K. Heesche-Wagner, Appl. Environ. Microbiol., 65 (1999) 1372. [34] S. Ebert, P.-G. Rieger, and H.-J. Knackmuss, J. Bacteriol., 181 (1999) 2669. [35] G. Heiss, K.W. Hofmann, N. Trachtmann, D.M. Walters, P. Rouviere, and H.-J. Knackmuss, Microbiology, 148 (2002) 799. [36] C.S. Harwood, G. Burchhardt, H. Herrmann, and G. Fuchs, FEMS Microbiol. Rev., 22 (1999)439. [37] R.M. Atlas and R. Bartha, Microbial Ecology: Fundamentals and Applications, 4th ed., Menlo Park, CA, 1998. [38] R.M. Miller and R. Bartha, Appl. Environ. Microbiol., 55 (1989) 269. [39] K.G. Kropp, LA. Davidova, and J.M. Suflita, Appl. Environ. Microbiol., 66 (2000) 5393. [40] J.D. Schumacher and R.M. Fakoussa, Appl. Microbiol. Biotechnol., 52 (1999) 85. [41] J. Heider, A.M. Spormann, H.R. Beller, and F. Widdel, FEMS Microbiol. Rev., 22 (1999) 459. [42] A.M. Spormann and F. Widdel, Biodegradation, 11 (2000) 85. [43] F. Widdel and R. Rabus, Curr. Opin. Biotechnol., 12 (2001) 259. [44] J.D. Van Hamme, A. Singh, and O.P. Ward, Microbiol. Mol. Biol. Rev., 67 (2003) 503. [45] H. Wilkes, S. Kiihner, C. Bolm, T. Fischer, A. Classen, F. Widdel, and R. Rabus, Org. Geochem., 34 (2003) 1313. [46] R. Rabus, H. Wilkes, A. Behrends, A. Armstroff, T. Fischer, A.J. Pierik, and F. Widdel, J. Bacteriol., 183 (2001) 1707. [47] K. Koike, K. Ara, S. Adachi, H. Takigawa, H. Mori, S. Inoue, Y. Kimura, and S. Ito, Appl. Environ. Microbiol., 65 (1999) 5636. [48] A. Jenisch-Anton, A. Mikolajczak, A. Rabenstein, J. Klindworth, U. Fischer, and W. Michaelis, Biodegradation, 10 (1999) 383. [49] Q. Cheng, S.M. Thomas, and P. Rouviere, Appl. Microbiol. Biotechnol., 58 (2002) 704. [50] K. Kostichka, S.M. Thomas, K.J. Gibson, V. Nagarajan, and Q. Cheng, J. Bacteriol., 183(2001)6478. [51] P.C. Brzostowicz, K.L. Gibson, S.M. Thomas, M.S. Blasko, and P. Rouviere, J. Bacteriol., 182(2000)4241. [52] P. Singh, A. Youyen, and H.S. Fogler, AIChE J, 47 (2001) 2111. [53] J.R. Becker, Crude Oil Waxes, Emulsions, and Asphaltenes, Tulsa, OK, 1997. [54] M.M. Santamaria and R.E. George, in the Proc. of the SPE Annual Technical Conference & Exhibition, Dallas, TX, 1991, pp. 351-353.

141

[55] F.G. Brown, in the Proc. of the SPE Permian Basin Oil and Gas Recovery Conference, Midland, TX, 1992, pp. 251-259. [56] I. Lazar, A. Voicu, C. Nicolescu, D. Mucenica, S. Dobrota, I.G. Petrisor, M. Stefanescu, and L. Sandulescu, J. Pet. Sci. Eng., 22 (1999) 161. [57] D.P. Rimmer, A.A. Gregoli, J.A. Hamshar, and E. Yildirim, in L.L. Schramm (ed), Emulsions: Fundamentals and Applications in the Petroleum Industry, Washington, DC, 1992, pp. 295-312. [58] L.L. Schramm, in L.L. Schramm (ed), Emulsions: Fundamentals and Applications in the Petroleum Industry, Washington, DC, 1992, pp. 1-49. [59] R. Grace, in L.L. Schramm (ed), Emulsions: Fundamentals and Applications in the Petroleum Industry, Washington, DC, 1992, pp. 313-339. [60] L. Dorobantu, Stabilization of oil/water emulsions by hydrophobic bacteria, M.Sc. thesis, University of Alberta, Edmonton, AB, 2004. [61] R.J. Mikula, in L.L. Schramm (ed), Emulsions: Fundamentals and Applications in the Petroleum Industry, Washington, DC, 1992, pp. 79-129. [62] W.L. Cairns, D.G. Cooper, J.E. Zajic, J.M. Wood, and N. Kosaric, Appl. Environ. Microbiol, 43(1982)362. [63] M. Das, Bioresour. Technol., 79 (2001) [64] Z. Duvnjak and N. Kosaric, Biotechnol. Lett., 9 (1987) 39. [65] N.C.C. Gray, A.L. Stewart, W.L. Cairns, and N. Kosaric, Biotechnol. Lett., 6 (1984) 419. [66] N.C.C. Gray, A.L. Stewart, W.L. Cairns, and N. Kosaric, in C. Ratledge, P. Dawson, and J. Rattray (eds), Biotechnology for the Oils and Fats Industry, USA, 1984, pp. 255268. [67] N. Nadarajah, A. Singh, and O.P. Ward, Process Biochem., 37 (2002) 1135. [68] N. Nadarajah, A. Singh, and O.P. Ward, World J. Microbiol. Biotechnol., 18 (2002) 435. [69] A.L. Stewart, N.C.C. Gray, W.L. Cairns, and N. Kosaric, Biotechnol. Lett., 5 (1983) 725. [70] S.H. Park, J.-H. Lee, S.-H. Ko, D.-S. Lee, and H.K. Lee, Biotechnol. Lett., 22 (2000) 1389. [71] K.L. Janiyani, HJ. Purohit, R. Shanker, and P. Khanna, World J. Microbiol. Biotechnol., 10 (1994) 452. [72] W.L. Cairns, R. Rumble, and N. Kosaric, in C. Ratledge, P. Dawson, and J. Rattray (eds), Biotechnology for the Oils and Fats Industry, USA, 1984, pp. 223-239. [73] O. Ward, A. Singh, and J. Van Hamme, J. Ind. Microbiol. Biotechnol., 30 (2003) 260. [74] J.S. Clemente, N.G.N. Prasad, M.D. MacKinnon, and P.M. Fedorak, Chemosphere, 50 (2003) 1265. [75] D.C. Herman, P.M. Fedorak, and J.W. Costerton, Can. J. Microbiol., 39 (1993) 576. [76] E. Slavcheva, B. Shone, and A. Turnbull, Br. Corros. J., 34 (1999) 125. [77] M.D. MacKinnon and H. Boerger, Water Poll. Res. J. Canada, 21 (1986) 496. [78] D.C. Herman, P.M. Fedorak, M.D. MacKinnon, and J.W. Costerton, Can. J. Microbiol., 40(1994)467. [79] J.W.S. Lai, LJ. Pinto, E. Kiehlmann, L.I. Bendell-Young, and M.M. Moore, Environ. Toxicol. Chem., 15 (1996) 1482. [80] F.M. Holowenko, M.D. MacKinnon, and P.M. Fedorak, Water Res., 35 (2001) 2595. [81] F.M. Holowenko, M.D. MacKinnon, and P.M. Fedorak, Water Res., 36 (2002) 2843. [82] J.S. Clemente, T.-W. Yen, and P.M. Fedorak, J. Environ. Eng. Sci, 2 (2003) 177.

142

[83] J.S. Clemente, M.D. MacKinnon, and P.M. Fedorak, Environmental Science & Technology, (in press) [84] T. Wakabayashi, Fuel, 76 (1997) 1049. [85] M.R. Riazi and T.E. Daubert, Oil Gas J., 85(52) (1987) 110. [86] A. Werner, F. Behar, J.C. de Hemptinne, and E. Behar, Fluid Phase Equilib., 147 (1998) 343. [87] A. Werner, J.C. de Hemptinne, F. Behar, E. Behar, and C. Boned, Fluid Phase Equilib., 147(1998)319. [88] P.J. Closmann and R.D. Seba, J. Can. Pet. TechnoL, 29(4) (1990) 115. [89] J.D. Payzant, D.D. Mclntyre, T.W. Mojelsky, M. Torres, D.S. Montgomery, and O.P. Strausz, Org. Geochem., 14 (1989) 461. [90] A.M. Cook and R. Hiitter, Appl. Environ. Microbiol, 43 (1982) 781. [91] D. Ziirrer, A.M. Cook, and T. Leisinger, Appl. Environ. Microbiol., 53 (1987) 1459. [92] JJ. Kilbane, in D.L. Wise (ed), Bioprocessing and Biotreatment of Coal, New York, NY, 1990, pp. 487-506. [93] T.D. Sutherland, I. Home, M.J. Lacey, R.L. Harcourt, R.J. Russell, and J.G. Oakeshott, Appl. Environ. Microbiol., 66 (2000) 2822. [94] KJ. Kayser, B.A. Bielaga-Jones, K. Jackowski, O. Odusan, and JJ. Kilbane, J. Gen. Microbiol., 139(1993)3123. [95] T. Omori, L. Monna, Y. Saiki, and T. Kodama, Appl. Environ. Microbiol., 58 (1992) 911. [96] T. Omori, Y. Saiki, K. Kasuga, and T. Kodama, Biosci. Biotechnol. Biochem., 59 (1995)1195. [97] Y. Izumi, T. Ohshiro, H. Ogino, Y. Hine, and M. Shimao, Appl. Environ. Microbiol., 60 (1994)223. [98] S.-K. Rhee, J.H. Chang, Y.K. Chang, and H.N. Chang, Appl. Environ. Microbiol., 64 (1998)2327. [99] J.H. Chang, S.-K. Rhee, Y.K. Chang, and H.N. Chang, Biotechnol. Progr., 14 (1998) 851. [100]M. Kobayashi, T. Onaka, Y. Ishii, J. Konishi, M. Takaki, H. Okada, Y. Ohta, K. Koizumi, and M. Suzuki, FEMS Microbiol. Lett., 187 (2000) 123. [101] J. Konishi, Y. Ishii, T. Onaka, K. Okumura, and M. Suzuki, Appl. Environ. Microbiol., 63(1997)3164. [102]T. Furuya, K. Kirimura, K. Kino, and S. Usami, FEMS Microbiol. Lett., 204 (2001) 129. [103]H.Y. Kim, T.S. Kim, and B.H. Kim, Biotechnol. Lett., 12 (1990) 761. [104] J.D. Van Hamme, P.M. Fedorak, J.M. Foght, M.R. Gray, and H.D. Dettman, Appl. Environ. Microbiol., (2004, in press) [105]H. Fuse, O. Takimura, K. Murakami, Y. Yamaoka, and T. Omori, Appl. Environ. Microbiol., 66 (2000) 5527. [106]D.P. Kelly and N.A. Smith, Adv. Microb. Ecol., 11 (1990) 345. [107]P.T. Visscher and B.F. Taylor, Appl. Environ. Microbiol., 59 (1993) 3784. [108]P.T. Visscher and B.F. Taylor, Appl. Environ. Microbiol., 59 (1993) 4083. [109]R.S. Oremland, M.J. Whiticar, F.E. Strohmaier, and R.P. Kiene, Geochim. Cosmochim. Acta, 52(1988)1895. [110]D.A. Irvine, J.P. Earley, D.P. Cassidy, and S.P. Harvey, Water Sci. TechnoL, 35 (1997) 67.

143

[111]T.-S. Lee, Biodegradation and Biotransformation of Thiodiglycol, the Main Hydrolysis Product of Sulfur Mustard, Ph.D. thesis, University of Maryland, College Park, MD, 1998. [112] V. Garcia-Ruiz, L.E. Martin-Otero, and A. Puyet, Biotechnol. Progr., 18 (2002) 252. [113]N. Itoh, M. Yoshida, T. Miyamoto, H. Ichinose, H. Wariishi, and H. Tanaka, FEBS Lett., 412(1997)281. [114] J.J. Kilbane and K. Jackowski, J. Chem. Technol. Biotechnol., 65 (1996) 370.



145

Chapter 5

Whole-cell bio-processing of aromatic compounds in crude oil and fuels J. M. Foght Department of Biological Sciences, University of Alberta Edmonton, Alberta Canada T6G 2E9 1. INTRODUCTION This review discusses the potential for biological upgrading to improve the quality of certain crude oils and liquid fuels, using whole cell biocatalysts to decrease aromaticity and sensitize aromatic heterocycles to subsequent heteroatom removal. Some specific examples of research directed towards future applications are presented; however, much of this concept is hypothetical and requires further research to prove its potential. This chapter presents information about the aerobic bacteria currently being tested as aromatic ring opening biocatalysts and speculates on the application of their activities to biological petroleum upgrading ("bio-processing"). The review does not deal with reactions achieved with biological desulfurization (BDS), nor with purified enzymes. For treatments of those topics, see Chapters 2 and 3, respectively. 1.1. Problems posed by aromatic compounds in crude oil and fuels Aromatic hydrocarbons and heterocycles adversely affect several stages of petroleum production, handling, and processing. Homologous series of aromatic hydrocarbons and heterocycles occur in varying proportions in crude oils and their refined products, depending on the source of the oil and the refining process applied. Aromatic compounds influence the persistence and toxicity of oils spilled in the environment [1] and have poor combustion characteristics in diesel engines, including low cetane number and high particulate matter (soot) formation. During refining, nitrogen heterocycles (e.g., carbazoles) inactivate chemical catalysts, interfere with catalytic hydrodesulfurization and consume large amounts of H2 [2, 3]. As well, combustion of fuels containing S and N heteroatoms produces SOX and NOX in emissions implicated in acid rain. Effective and cost-efficient conversion of aromatic hydrocarbons and

146

heterocycles in crude oils and fuels therefore is desirable from an environmental viewpoint and is of interest to the refining industry. During conventional refining, expensive high-pressure hydrogenation and chemical catalysis are used to saturate and "crack" aromatic fused-ring structures. This thermochemical technology has several problems [4, 5] including unfavorable reaction kinetics, high consumption of thermal energy and hydrogen (which contributes to greenhouse gases and other emissions), and production of less desirable side-products such as gaseous hydrocarbons through non-specific reactions. The cycloalkane products have better combustion properties than aromatics, but the quality of their characteristics is still below those of linear alkane compounds. 1.2. Biological oxidation of aromatic compounds In contrast to chemical catalysis, biological (enzyme-mediated) reactions are usually highly substrate-specific and occur at near-ambient temperatures without the need for expensive high-pressure vessels and hydrogen. They do, however, require water for activity, necessitating biphasic reaction mixtures of oil and an aqueous suspension of microbial cells. In one proposed application, a two-stage process of "Biological Aromatic Ring Cleavage" (BioARC) is envisioned: the first stage is ring opening of the aromatic compounds by oxidative enzymatic reactions achieved by pre-grown whole cell biocatalysts. The two product streams would be a primary stream of BioARC-treated oil having decreased aromaticity, and a secondary stream of polar, water-soluble aromatic ring cleavage products (forming, for example, Compounds IV and V, Fig. 1) that would be recovered by solvent extraction or reversible sorption. In stage two, chemical hydrogenation of the polar ring cleavage products under milder conditions than the conventional process would yield alkyl-aromatics having more favorable combustion characteristics than either the parent compound or the conventional cycloalkane catalytic products. Potential side-benefits to the process would be sensitization of S- and Nheterocycles to subsequent desulfurization and denitrogenation, although there is no experimental work to support this prediction at present. A second process described in US Patent # 6,156,946 [6] as "biological activation of aromatics" uses whole cell biocatalysis to hydroxylate aromatic and heterocyclic substrates in petrochemicals (e.g., enzymatic steps A and B, forming Compounds II and III, Fig. 1), followed by aqueous chemical hydrogenation and hydrogenolysis to produce cyclic alkylaromatics. The hydrogenated products are proposed to be more susceptible to subsequent thermochemical cleavage than the parent compounds. This is essentially an "aromatic activation" process in which the substrates are sensitized to subsequent hydrogenation or ring cleavage.

147

Fig. 1. Classical "upper pathway" for phenanthrene ring cleavage (adapted from Refs. [7, 8]) showing stepwise oxidation and metabolites from the "aromatic activation" [6] and BioARC [9] processes. Theoretical products after hydrogenation are indicated. Compounds: I, phenanthrene; II, cw-3,4-dihydroxy-3,4-dihydrophenanthrene; III, 3,4-dihydroxyphenanthrene; IV, 2-hydroxy-2#-benzo[h]chromene-2-carboxylic acid; V, fra«.s-4-(l-hydroxynaph-2-yl)-2-oxobut-3-enoic acid; VI, l-hydroxy-2-naphthoaldehyde. Enzymatic steps: A; aromatic dioxygenase; B, dihydrodiol dehydrogenase; C, extradiol dioxygenase; D, isomerase; E, hydratase-aldolase.

148

Several species of organic nitrogen found in crude oil must be removed during refining because they form nitrogen oxides when combusted in fuels. The carbazoles in particular are resistant to removal by conventional catalytic hydrodenitrogenation, interfere with hydrodesulfurization, contribute to corrosion of refinery metals, and poison refinery catalysts [3]. Research has been initiated into biological denitrogenation, focusing on carbazole as the model compound because it is the most abundant species and the most recalcitrant to hydrotreatment. A theoretical approach to bio-denitrogenation is angular dioxygenation of carbazole, cleaving the heterocycle to yield an aromatic amine with side groups (e.g., compounds III and IV, Fig. 2). Depending on the extent of subsequent metabolism, the product(s) would either be polar enough to be water-washed from the oil, thus reducing its nitrogen content and generating a metabolite stream for separate processing, or would remain associated with the oil phase presumably to be subjected to hydrogenation as a less problematic feedstock. Key to all of these hypothetical approaches is complete recovery of the carbon skeleton without loss to either CO2 or biomass. This can be achieved by using a pre-grown biocatalyst with a truncated enzymatic pathway so that the biocatalyst cells do not use petroleum hydrocarbon for biomass or oxidize it to CO2. These three processes are described more completely in Section 5 and factors influencing their application to biological upgrading are discussed in general terms in Sections 1.3 to 4. Bio-processing of aromatic constituents of petroleum may be a feasible adjunct process, not a replacement, for conventional upgrading processes. It is possible that some bio-processing applications would be suitable for use in the field as crude oil is recovered and when it is already in intimate contact with production water, or applied at the refinery in holding tanks as long as rapid throughput was not a requirement. It would produce value-added fuels from crude oils or decreased aromaticity middle distillates at a reduced environmental cost by lowering the energy expenditure per cubic metre of feedstock. It may also yield products with a lower nitrogen and sulfur content by sensitizing heterocycles to subsequent hydrogenation. These improvements would reduce the contribution of fuel processing and combustion to greenhouse gas emissions and acid rain, and could be economically feasible if applied appropriately. 1.3. Substrates for aromatic bio-processing 1.3.1. Aromatic hydrocarbons and heterocycles Crude oils and most refined products contain complex mixtures of aromatic hydrocarbons and heterocycles. Unsubstituted aromatics include benzene, naphthalene and phenanthrene, representing the mono-aromatics and di- and

149

tricyclic polyaromatic hydrocarbons (PAH) respectively, as well as higher molecular weight PAH. Homologous series of alkyl-substituted aromatics are also present, such as toluene and the isomers of xylene, methyl- and ethylnaphthalenes [10]. Aromatic heterocycles with N-, S- and O-substitutions are represented by carbazole, dibenzothiophene and dibenzofuran, among others, and these also occur as families of alkyl homologues.

Fig. 2. Classical angular attack on carbazole showing main pathway (adapted from Ref. [11]) and some minor products [12]. Compounds: I, carbazole; II, postulated intermediate; III, 2'aminobiphenyl-2,3-diol; IV, 2-hydroxy-6-oxo-6-(2'aminophenyl)-hexa-2,4-dienoic acid; V, 2-hydroxypenta-2,4-dienoic acid; VI, anthranilic acid. Other metabolites selected from Ref.

150

[12].

Currently, aromatic bio-processing has focused on the di- and tricyclic homologues in crude oil and middle distillate fuels (like diesel) for biological reasons. The monoaromatics can present severe toxicity problems for biocatalyst cells at quite low concentrations [13]. At the other extreme, aromatics larger than three fused rings tend to be quite recalcitrant to biological treatment [1] and therefore are unfavorable substrates for bio-processing. The chemical reason for targeting di- and tricyclic aromatics is that conventional hydrotreatment yields cycloalkanes which still have low fuel value compared with straight chain alkanes; opening one or more aromatic rings to produce side-chain alkyl groups should improve the fuel value of the products. The alkyl-substituted homologues exhibit varying susceptibilities to biological oxidation, with the general rule that increasing molecular weight and substitution decrease susceptibility to biological attack (e.g., see Refs. [14, 15] and discussion in Ref. [16] on alkyl substitution of dibenzothiophenes). A practical biocatalyst must achieve ring opening of many or all of these substituted aromatics in addition to the unsubstituted parent compounds (i.e. have a broad aromatic substrate range), but not oxidize non-target hydrocarbons in the feedstocks such as alkanes. 1.3.2. Suitable feedstocks An ideal feedstock for aromatic bio-processing has low viscosity and a high content of di- and tri-cyclic aromatic hydrocarbons with low alkyl substitution. Experimentally, this has been achieved by preparing "model oils" comprising pure substrates (such as phenanthrene and dibenzothiophene) dissolved in an aliphatic carrier such as «-hexadecane, heptamethylnonane [17] or light mineral oil [18], or by dissolving model compounds in authentic feedstocks such as middle distillates [19] or crude oil [20]. However, most authentic feedstocks are less than ideal for bio-processing due to factors including viscosity, average molecular weight and degree of alkyl and heteroatom substitution. Haus et al. [21] determined the kinematic viscosity of aliphatic base oils containing a wide range of aromatic and polar compounds, and found that biodegradability was closely related to PAH content and decreased with increasing kinematic viscosity (the absolute viscosity of a liquid divided by its density). Bioprocessing would certainly be affected in the same manner. Conversely, low viscosity feedstocks with a high content of low molecular weight hydrocarbons, such as «-alkanes CH3OH + H2O Thus, MMOs activate the kinetically inert O2 molecule to react at 30-45 °C, 1 atmosphere pressure with the unreactive hydrocarbon methane (C-H bond energy = 1 1 9 kcal mol"1) to produce methanol with stoichiometric yield and high turnover (up to 6.0 s"1 [1]). In addition to this important transformation, they also catalyse the oxygenation of a diversity of adventitious substrates, including numerous petroleum constituents and by-products, which have led to intense interest in their exploitation for biocatalyis and bioremediation. In this chapter, we review the biological distribution of MMOs, their structural and catalytic properties and then discuss their applications and potential for transformation of petroleum-related products. Lastly, we discuss how MMOs may inspire the synthesis of biomimetic catalysts to facilitate methane to methanol conversion and other commercially important transformations.

178

2. BIOLOGICAL DISTRIBUTION AND CLASSIFICATION OF METHANE MONOOXYGENASES MMOs have been found only in methanotrophic bacteria, which are Gramnegative organisms that group within the a - and y-proteobacteria and can utilise methane as their sole source of carbon and energy. Methanotrophic bacteria, and therefore the MMOs that they produce, are widespread in the environment; they are found in mesophilic and extreme (e.g. as low as 4 °C [2] and up to 72 °C [3]) conditions and are estimated to be responsible for oxidation of up to 60 % of the 1188 Tg of methane produced globally each year by natural and anthropogenic sources [4]. Two forms of MMO are known, particulate (pMMO) and soluble (sMMO). Many methanotrophs, such as Methylomicrobium album BG8 and Methylomonas methanica, produce only the membrane-associated pMMO. Others, such as Methylococcus capsulatus (Bath) and Methylosinus trichosporium OB3b can elaborate either form [5], dependent on the copper-tobiomass ratio of the culture. Such organisms produce the copper-containing pMMO at high copper-to-biomass ratios and the nonheme iron-containing sMMO at low copper-to-biomass ratios [6]. It is relatively straightforward, during cultivation in laboratory- or industrial-scale bioreactors, to control the medium copper concentration and culture density to obtain either form of MMO. An easy colorimetric test, based on the oxidation of naphthalene [7], which is a substrate of sMMO alone, allows rapid confirmation of the form of MMO that the culture is expressing. Although sMMO and pMMO both oxygenate methane to methanol, they show no similarity in the amino-acid sequences of their protein components, their requirements for metal cofactors or their location within the cell. They also differ markedly in their substrate profiles and requirements for electron donors. It is remarkable that the particulate and soluble MMOs represent evolutionarily unrelated molecular solutions to a single chemical problem that are frequently found in the same bacterial cells.

3. SOLUBLE METHANE MONOOXYGENASE Thanks to more than 20 years of effort with diverse techniques in several research groups, sMMO is now probably the best characterised non-heme iron monooxygenase system. Two sMMO systems, from Me. capsulatus (Bath) [8] and Ms. trichosporium OB3b [1], have been studied in detail. These enzyme systems show different optimal temperatures (45 and 30 °C, respectively), which correspond to the optimal growth temperatures of the two species of bacteria. In their structure, substrate range and mechanistic properties, the two sMMOs are very similar. This section refers primarily to the Me. capsulatus enzyme,

179

although in many cases almost indistinguishable results have been obtained with the Ms. trichosporium system. 3.1. sMMO structure sMMO comprises three components (Fig. 1): (1) a hydroxylase with an (a|3y)2 structure in which the a, |3 and y subunits have masses of 61, 45 and 20 kDa, respectively; (2) a 39-kDa NAD(P)H-dependent reductase; (3) a third component known as protein B or the coupling/gating protein, which comprises a single polypeptide of 16 kDa [8]. The a subunits of the hydroxylase each contain a u-(hydr)oxo-bridged binuclear iron site [10-12] that is coordinated by four glutamate and three histidine sidechains and is the site of O2 activation. Xray crystallography of the hydroxylases from Me. capsulatus (PDB accession codes 1MM0, 1MTY) and Ms. trichosporium (1MHY, 1MHZ) has shown that the hydroxylase is a predominantly a-helical structure in which the binuclear iron centres reside within the a subunits, in a hydrophobic pocket that is almost certainly critical in binding substrates [13-15]. Indeed, energy minimisation calculations have suggested that the most favourable binding site for methane and other small substrates lies inside this pocket, within 3 A of the binuclear iron centre (Fig. 2) [16]. The reductase component, which is required for the supply electrons from NADH or NADPH, contains flavin adenine dinucleotide (FAD) and Fe2S2 prosthetic groups. Protein B has no prosthetic groups. NMR structural analysis (PDB accession codes 1CKV [17] and 2MOB [18]) has shown that it has a core ot/p structure with highly mobile regions at the N- and C-termini.

Fig. 1. Components of sMMO. The individual polypeptides of the enzyme are named according to the genes that encode them [9].

180

Fig. 2. Active site structure of sMMO based on crystallographic data in Ref. [14]. The iron atoms are shown as large grey spheres and the residues that ligate them as a grey ball-andstick representation. The residues forming the hydrophobic active site pocket [ 16] are shown in white and two solvent molecules that bridge the dinuclear iron site in black.

3.2. sMMO mechanism In the resting state of the sMMO enzyme, the binuclear iron centre is in the diferric (Fe m 2 ) oxidation state [10], and must be reduced to the diferrous (Fe"2) form to allow O2 to bind. The two electrons required for this reduction are provided from NAD(P)H, passing first via the FAD and then the Fe2S2 prosthetic group of the reductase, which acts as a two-electron/one-electron transformase, to allow the two-electron oxidation of NAD(P)H to be used to feed electrons singly into the binuclear iron centre of the hydroxylase [19]. After binding of O2, the intermediates O, P*, P and Q (Fig. 3) can be detected even in the absence of substrate. Compound O, whose existence has been inferred from the independence of formation of compound P on O2 concentration, appears to be a species in which O2 is bound to the hydroxylase but not covalently attached to the binuclear iron centre [20]. Formation of compound P* is accompanied by loss of the g = 16 electron paramagnetic resonance (epr) signal due to the highspin Fe"2 state of the binuclear iron centre. In compound P* the binuclear iron site may be in the FeII[2 or mixed valent (Fe11 Fe m ) state [21]. At this stage dioxygen is probably covalently bound to the binuclear iron centre in the form of an unprotonated bridging peroxo species. The compound P* to P transition requires a proton, probably to protonate the peroxo species before the 0 - 0 bond scission that occurs upon the decay of compound P. The binuclear iron centre of compound P has been shown by means of Mossbauer spectroscopy to be in the epr-silent Fe in 2 form. Compound P decays spontaneously to compound Q, which is also epr-silent but is bright yellow in colour (absorption maxima 350 and 420 nm) [20]. The binuclear iron centre of compound Q appears based on Mossbauer spectroscopy to be in the diferryl (FeIV2) oxidation state. Interatomic

181

distance constraints obtained from extended X-ray absorption fine structure (EXAFS) data are consistent with the bis-|i-oxo bridged 'diamond core' structure that Lipscomb and co-workers have proposed for the binuclear iron centre of compound Q (Fig. 3), although other possible structures also fit the data [22]. Compound Q is the quintessence of sMMO catalysis. It is kinetically competent to oxygenate methane and other substrates and yet, in the absence of substrate, is astoundingly stable (ti/2 14 s at 4°C) [23] for so powerful an oxidant in aqueous solution. When methane is the substrate, decay of compound Q exhibits a nonlinear Eyring plot with a point of inflection at 17°C. Other evidence suggests that this discontinuity is due to a change in rate-determining step and so the reaction of compound Q with the substrate must comprise at least two distinct steps [23]. In a study of the Ms. trichosporium enzyme, use of the chromogenic substrate nitrobenzene allowed detection of an enzyme-bound product species, termed compound T [24]. It is, however, the preceding events of carbon-hydrogen bond scission and substrate oxygenation that are fundamental to understanding how sMMO can activate unreactive hydrocarbons such as methane, and it is the mechanism of these steps that remain the greatest contention in the sMMO reaction cycle. In another study of the sMMO from Ms. trichosporium, reaction of compound Q with methane showed a very large primary deuterium isotope effect (ca. 50), which was too high to be accounted for solely by the difference in vibrational zero point energies between CH4 and CD4. One possibility is that quantum mechanical tunnelling of the hydrogen nucleus is required for in breakage of the inert C-H bond of methane; indeed, consistent with this reasoning, the primary deuterium isotope effects observed with less inert substrates are more modest (e.g. 1 in the case of ethane) [25]. There are three possible mechanisms for C-H bond cleavage: (1) cleavage may be homolytic, leading to a radical mechanism; (2) cleavage may be heterolytic, leading to a mechanism in which a carbanion intermediate may be stabilised by coordination to one of the active-site iron atoms [26]; (3) methane and an iron-oxygen species may react via a concerted mechanism of bond breakage and formation [27]. The current evidence and theories are reviewed in Refs. 28-30. The available data, derived principally from radical clock substrates and theoretical studies, are complex and not afford a single clear conclusion at the present time; indeed it is possible that multiple reaction pathways exist and that different mechanisms operate with different substrates. What is clear, however, is that sMMO generates within its active site one of the most reactive oxidising agents in nature that is, as described below, able to oxygenate a range of petroleum-related substrates.

182

Fig. 3. Principal intermediates during the sMMO catalytic cycle. References are given in the text.

3.3. Substrate profile of sMMO In addition to its remarkable ability to convert the unreactive methane molecule to methanol under ambient conditions, sMMO is remarkable in its diversity of adventitious substrates, often termed cosubstrates, that the enzyme can oxidise even though the resultant oxidation products cannot be used as nutrient sources by methanotrophic bacteria. The astonishing range of substrates known to be oxygenated by sMMO is summarised in Table 1, together with the principal products obtained. Broadly speaking, singly oxygenated products predominate with all substrates. Alkanes are hydroxylated, in the case of aliphatic compounds almost exclusively at the terminal and subterminal positions. Ring hydroxylation of aromatics occurs primarily at the meta position, along with a comparable amount of substituent hydroxylation when an alkyl substituent is present. sMMO oxygenates alkenes to epoxides with retention of stereochemistry around the C=C double bond. Ethers are cleaved oxidatively to yield mixtures of alcohols and aldehydes, and pyridine undergoes Noxygenation. It appears that the initial oxygenated products formed from some halogenated substrates decompose rapidly via nonenzymic pathways that result in the loss of halogen substituents [31]. It is certain that there are many substrates of sMMO that have simply never been tested with the enzyme. Indeed, the list of small organic compounds that have been investigated and found not to be effective sMMO substrates is very short, but includes tetrachloromethane, iodomethane, trimethylamine [32] and tetrachloroethene [31]. In addition, terminal alkynes are potent turnover-dependent inhibitors of

183

sMMO; they are probably oxidised to highly reactive ketenes which then attack catalytically essential groups on the enzyme [33]. 4. PARTICULATE METHANE MONOOXYGENASE 4.1. Structure and mechanism of pMMO Although pMMO appears to be responsible for the bulk of methane oxidation catalysed by methanotrophic bacteria in the environment, its instability during purification procedures has meant that a consensus about its composition and catalytic properties has been slow to emerge. pMMO is well established as a copper-containing enzyme. The intracytoplasmic membranes of methanotrophic bacteria that contain it are produced only during high-copper growth [5] and, in methanotrophs that can express either pMMO or sMMO, pMMO is induced in response to high copper-to-biomass ratio in the culture [6]. In addition, Cu2+ ions stimulate pMMO activity in vitro and the copper content of methanotroph membranes has been observed to be directly proportional to the pMMO activity [36]. One of the chief problems with purification of pMMO has been solubilising the enzyme without permanent loss of activity. In 1989, it was observed in our laboratory that solubilisation of pMMO by dodecyl-P-Dmaltoside was reversible, because activity could be restored by adding phospholipids back to the solubilised enzyme [37]. Dodecyl-(3-D-maltoside appears to be the most effective detergent tested to date for solubilisation of pMMO and to the authors' knowledge has formed the basis for all recent studies of purified pMMO. Shiemke and coworkers [38] observed that pMMO activity could be detected in solubilised pMMO if duroquinol or decyl plastoquinol were used as electron donor in lieu of NADH, suggesting that the direct electron donor into pMMO is a quinol. Protocols for purification of pMMO typically comprise preparation of the membrane fraction, washing with buffer containing up to 1 M monovalent salt and then one or more chromatographic separations, usually including at least one anion exchange column [39-43]. Some workers have found anaerobic conditions favourable to retention of activity [43], whilst others have found the presence of oxygen tolerable [40] or beneficial [41]. Active preparations of pMMO have generally contained three polypeptides, of about 49, 27 and 22 kDa [37], among which the 27-kDa subunit is most likely to contain the active centre because it becomes labelled by the inhibitor acetylene [33, 39]. The 49, 27 and 22-kDa components are encoded by the genes pmoA, B and C, respectively, which are multicopy genes [44, 45] induced in response to high copper-tobiomass ratio [46, 47].

184

Table 1 Principal oxidation reactions catalysed by sMMO Substrate Major detected product(s); relative molar proportions of multiple products are shown in parentheses. Alkanes Methane Methanol Ethane Ethanol Propane Propan-1-ol (39); propan-2-ol (61) Butane Butan-1-ol (54); butan-2-ol (46) Pentane Pentan-1-ol (28); pentan-2-ol (72) Hexane Hexan-1-ol (63); hexan-2-ol (37) Heptane Heptan-1-ol (22); heptan-2-ol (78) Octane Octan-1-ol (9); octan-2-ol (91) 2-Methylpropane 2-Methylpropan-2-ol (70); 2methylpropan-1-ol (30) 2-Methylpropane 2-Methylpropan-2-ol (70); 2methylpropan-1-ol (30) 2,3-Dimethylpentane 3,4-Dimethylpentan-2-ol Alkenes Ethene Propene But-1-ene czs-But-2-ene Zrarc.s-But-2-ene

Epoxyethane Epoxypropane 1,2-Epoxybutane cw-2,3-Epoxybutane (47); cis-2buten-1-ol (53) ?ra«5-2,3-Epoxybutane (27); trans-2buten-1-ol (73)

Al icy die hydrocarbons Cyclohexane 3 Cyclohexanol Methylene 1-Cyclohexane-1-methanol (13.7); cyclohexane methylene cyclohexane oxide (75.8); 4-hydroxymethylene cyclohexane (10.5) 6,6-Dimethylbicyclo[3.1.1 ]hept-2p-Pinene ene-2-methanol (72.3); P-pinene oxide. Adamantane 1-Adamantol (50); 2-adamantol (50) czs-1,4-Dimethyl l-cw-4-Dimethylcyclohexanol (35); cyclohexane 1 -rra«s-4-dimethylcyclohexanol (61); m-2,5-dimethylcyclohexanol (4) c/5-l,3-Dimethyl 3,5-Dimethylcyclohexanol (80); 1cyclohexane cw-3-dimethylcyclohexanol (14); 1trans-3 -dimethylcyclohexanol (6); Halogenated aliphatics Vinyl chloride 3 Trichloroethene 3 Formate (35); CO (53); glyoxylate (5); dichloroacetate (5); chloral (6)

Specific activity (nmol of product min~ mg" )

Reference (type of assay) 2

84 68 69 77 73 40 27 9 33

32 (SE) 32 (SE) 32 (SE) 32 (SE) 32 (SE) 32 (SE) 32 (SE) 32 (SE) 26 (PP)

33

26 (PP)

20

26 (PP)

148 83 49 33

32 32 32 26

39

26 (PP)

25

34 (SE) 26 (PP)

(SE) (SE) (SE) (PP)

26 (PP)

3 1.3

26 (PP) 26 (PP)

0.5

26 (PP)

748 682

31 (PP) 31 (PP)

185

Substrate

l,l-Dichloroethene J Trifluoroethene

3

Chlorotrifluoroetheylene 3 Tribromoethylene 3 Monoaromatics Benzene Toluene Ethylbenzene 3 Styrene 3 Pyridine Diaromatics Naphthalene Biphenyl 3 2-Hydroxybiphenyl 3 2-Methylbiphenyl 3 2-Chlorobiphenyl 3 2-Bromobiphenyl 3 2-Iodobiphenyl 3

Major detected product(s); relative molar proportions of multiple products are shown in parentheses. Glycolate (80); dichloroacetaldehyde (3) Glyoxylate (53); difluoroacetate (43); fluoral (5) Oxalate

Specific activity (nmol of product mirf'mg" 1 ) 1 648

Reference (type of assay) 2 31 (PP)

79

31 (PP)

17

31 (PP)

Formate (80); bromal (5)

9

31 (PP)

Phenol Benzyl alcohol; 4-cresol. 1-Phenylethanol (30); 4hydroxyethylbenzene (70) Styrene oxide Pyridine N-oxide

74 53 18.7

34 (SE) 32 (SE) 34 (SE)

82 29

34 (SE) 32 (SE)

1-Naphthol, 2-naphthol 2-Hydroxybiphenyl (9); 3hydroxybiphenyl (1); 4hydroxybiphenyl (90) Dihydroxybiphenyls Ring (56) and sidechain (44) hydroxylated products Hydroxychlorobiphenyls Hydroxybromobiphenyls (41); 2hydroxybiphenyl (59) Hydroxyiodobiphenyls (18); 2hydroxybiphenyl (82)

Substituted methane derivatives Chloromethane Dichloromethane Chloroform Bromomethane Nitromethane Methanethiol Methanol

7 (W) 35 (W)

35 (W) 35 (W) 35 (W) 35 (W) 35 (W)

84 82 35 66 45 64 246

32 32 32 32 32 32 32

(SE) (SE) (SE) (SE) (SE) (SE) (SE)

Others Diethyl ether Ethanol (47); acetaldehyde (53) 32 (SE) Carbon monoxide Carbon dioxide 61 32 (SE) Specific activities are as reported in the original publications. Caution is advised when comparing values from different studies, since various protein preparation and assay procedures were used by the different authors. Type of enzyme preparation used for each assay: PP, purified protein; SE, soluble extract; W, whole cells. 3 sMMO of Ms. trichosDorium OB3b: other entries refer to Me. capsulatus (Bath).

186

Active preparations of purified pMMO have been found to contain between 2 [41] and 15 [40] equivalents of copper per a(5y protomer. There is agreement that whilst the polypeptides of the intracytoplasmic membranes largely comprise pMMO subunits, most of the copper that is associated with them is relatively loosely bound and can be removed by repeated washing of membranes before solubilisation, increased detergent concentration or additional chromatographic separation of solubilised material [41-43]. The loosely associated copper may be chelated by a low molecular-mass copper-binding cofactor [39]. This cofactor may be part of a high-affinity copper-uptake system [48] and its complex with copper may protect pMMO by quenching superoxide radicals [43]. Chan and coworkers have classified the copper associated with pMMO into catalytic 'C-copper' and electron transfer, 'E-copper' and have proposed a model in which the E-copper corresponds to the less tightly associated copper component [40]. Epr spectroscopy reveals a type 2 copper centre [36, 39], which accounts for only a fraction of the copper in active pMMO preparations, at 1 [41] or 2 [36, 42] coppers per apy protomer. It is possible that much or all of the epr-silent copper is in the +1 oxidation state [36, 49] or could be magnetically coupled in a cluster [41]. Recent EPR results from our laboratory and Rosenzweig's are best accounted for if the epr-active copper is ligated by 3 to 4 nitrogen ligands with square planar geometry [41, 42] rather than the trinuclear copper cluster suggested by Chan's group [36]. There is a correlation between loss of copper from the enzyme and loss of function; the type 2 copper (II) centre appears to be the most tightly bound copper associated with the enzyme [41]. Chan's group have reported an NADHdependent preparation that contained no significant protein components except the a, (3 and y subunits of pMMO, and they put loss of NADH-driven activity in most preparations down to loss of the 'E-copper'. In addition, a 36-kDa FADcontaining NADH:quinone oxidoreductase has been found in Me. capsulatus (Bath) [50] and observed to co-purify with pMMO under certain conditions [43]. Thus, whilst the identity of the molecular components required to deliver electrons from NADH into the pMMO active site remains controversial, there is consensus that there are components specifically required for this electron transfer and that, when these are missing, electrons can still be delivered via duroquinol. Three out of four groups actively involved in research into the structural properties of pMMO consider that, in addition to copper, iron forms an important part of the functional enzyme, at one [41, 42] or two equivalents of per aPy protomer [43]. Based on EPR spectral properties a tyrosine-ligated Fe3+ ion has recently been proposed [41]. Little is currently known about the catalytic cycle of pMMO; however, the retention of stereochemistry observed during oxygenation of cryptically chiral alkanes strongly suggests a primarily concerted

187

mechanism of C-H bond breakage rather than one involving radical or cation intermediates [51]. 4.2 Substrate profile of pMMO The substrate profile of pMMO (Table 2) is considerably narrower than that of sMMO. It includes methane and linear short-chain hydrocarbons but excludes aromatics (benzene, ethyl benzene and styrene), the alicyclic hydrocarbon cyclohexane [34] and the branched aliphatic 2-methylpropane [52], all of which are substrates of sMMO. Thus it appears that access to the active site of pMMO is sterically more restricted than in the soluble enzyme. Consistent with this, acetylene is a potent suicide substrate of both soluble and particulate enzymes [33], whereas the bulkier phenylacetylene is much more effective against the sMMO than pMMO [53]. 5. METHANE MONOOXYGENASE IN BIOCATALYIS As indicated above MMO-containing bacteria are able to catalyse the oxidation of alkenes to epoxides. Since there is no direct, commercially viable, chemical route to epoxide formation from alkenes attempts have been made to exploit the MMO-catalysed reaction on a large scale. In a comprehensive study to produce epoxypropane from propene we have shown how a two stage system using Me. capsulatus (Bath) can be used to give high yields of product in a continuous operation [55]. This methanotroph was chosen because it grows readily at 45°C and at this temperature the epoxypropane is released into the gas phase. The product could then be readily condensed from the gas phase at a lower temperature. A major problem that had to be overcome was the inherent in vivo toxicity of the epoxides to the cells. In fact epoxypropane was shown to be an inhibitor of MMO and so caused an inactivation of the cells if grown in a single stage operation [56]. The problem was solved by using a two stage bioreactor system in which the first stage was used to catalyse the oxidation of propene to the epoxide and a second reactor was fed with cells from the catalytic reactor to reactivate inactivated cells. In the first reactor methanol was used as the electron donor for the whole cell reaction which had to be fed as a growth-limiting substrate to minimise its interference in MMO catalysis (methanol is a substrate for sMMO [see Table 1] but it has a much lower affinity for the enzyme than propene). In the second stage reactivator, methane was supplied as the growth substrate along with all nutrients necessary to allow resynthesis of inactivated cells. Critical to the success of the system were the operating parameters that included the ratio of the size of the reactors such that a continuous operation was possible that permitted active cells to be fed back to the first stage to maintain high catalytic productivity. The continuous recycling of cells from the reactor vessel to the reactivator (at an aspect ratio of 1:10) gave rates of 20-30 g L"1

188

day"1 over a three week period. Similar experiments with Methylocystis parvum (OBBP) gave rates of up to 90 g L"' day"1 over a one week period. In a feasibility study [55] using cells at 30 g L"1 and a production rate of 250 g L"1 day"1 the total cost of epoxypropane production was estimated at $1.26 kg"1. As such it was not competitive with the commercial oxirane process which values the product at around this price; no account in the biological; process was taken for storage, transport or profit which need to be added to this cost. Consequently, at the present time this system has not been commercialised although patents for the process have been filed worldwide. In principle such an operation could be adapted to produce any of the sMMO products listed in Table 1 although the separation and purification of the product will dictate the precise mode of operation. The process has been evaluated with other substrates including 1-butene and ethane [57] to produce epoxybutane and ethanal respectively. pMMO shows moderate stereoselectivity with some reactions (up to 80 % enantiomeric excess of the R-enantiomer of pentan-2-ol product formed from pentane) [52] and so may be suitable for eventual development as an enantioselective catalyst. Interestingly, whilst neither pMMO nor sMMO shows a high stereoselectivity in epoxide-generating reactions (enantiomeric excesses for epoxypropane generation by the two enzymes are 18.5 % S [52] and 21 % R [58], respectively), they show opposite enantioselectivity and so may be suitable for future genetic development into a pair of enantiocomplementary biocatalysts. Table 2 Principal oxidation reactions catalysed by pMMO Reference Substrate Product(s); relative molar proportions of multiple products are shown in parentheses. Alkanes Methanol Methane Ethane 52 Ethanol Propane 52 Propan-2-ol (ca. 100); propan-1-ol (trace) Butan-2-ol (95); butan-1-ol (5) 52 Butane Pentane Pentan-2-ol (95); pentan-1-ol (5) 52 Alkenes 52 Propene Epoxypropane 52 But-1-ene 1,2-Epoxybutane (58); but-3-en-2-ol (42) 1,3-Butadiene 1,2-Epoxybut-3-ene 52 «s-But-2-ene 2,3-cw-Epoxybutane 52 fr-ans-But-2-ene 2,3-/ra«^-Epoxybutane 52 Chlorinated aliphatics Trichloroethene' Carbon dioxide 54 sMMO of Ms. trichosporium OB3b; other entries refer to Me. capsulatus (Bath).

189

Both sMMO and pMMO oxidise the major pollutant trichloroethene. With sMMO, rates comparable to the oxidation of methane have been observed [31], whilst with pMMO the rate of trichloroethene breakdown is low (1.24 nmol min"1 mg of protein"1 in the pMMO-only methanotroph Mlb. album BG8 [54]). However, the detected products of trichloroethene breakdown by pMMOexpressing methanotrophs do not include chloral, a pollutant that may be more harmful than trichloroethene and has been detected in sMMO-mediated trichloroethene breakdown [31]. Hence bioremediation of trichloroethene by pMMO may have some advantages. During trials of methane-enhanced bioremediation of groundwater [59], both sMMO and pMMO genes were detected and it is likely that both were expressed to some extent in the population of methanotrophs present. 6. METHANE MONOOXYGENASE BIOMIMETICS The unique reactivity of MMOs, particularly their ability to convert methane to methanol under mild conditions, has led to intense interest in the development of small-molecule mimics of the enzyme active site that may be able to perform similar reactions. Whilst there is currently no commercially viable biocatalyst for industrial-scale conversion of methane to methanol, a low molecular-weight biomimetic that did not further oxidise the methanol produced could be used to upgrade the gaseous fuel methane to the more easily stored and transported methanol. In principle, such a mimic could be based on pMMO or sMMO. However, since there is currently high-resolution structural information for sMMO alone, this has been the overwhelming focus of attention. The literature abounds with reports of binuclear iron and other transition metal complexes and reports of sMMO-like properties. Whilst oxidation of methane to methanol by a biomimetic that uses dioxygen as oxidant has not been achieved, a number of important advances have been made. For instance, Jacobsen's group have prepared a carboxylate bridged binuclear iron complex that efficiently catalyses the hydrogen peroxide-driven epoxygenation of a range of alkenes [60]. A number of di-iron complexes have been described that catalyse O2-dependent oxidation of their own ligands, such as a di-iron complex that can oxidatively de-N-alkylate additional nitrogen-containing ligands [61]. A di-iron complex in which the ligands included an N3 ring catalysed O2-dependent oxidation of triphenylphosphine to the corresponding oxide [62]. Conversion of methane to methanol catalysed by di-iron complexes under has been reported, albeit at high pressure with low conversion (< 2 %) and significant side reactions [63]. Future studies in this area may not only deliver useful novel catalysts for oxygen-driven reactions but will also cast light on the reaction mechanism of sMMO.

190

7. FUTURE PROSPECTS The unusual reactivity and broad substrate profiles of MMOs suggest many possible applications in synthetic chemistry and bioremediation for the enzymes and biomimetics based on them. Our recent development of a system for sitedirected mutagenesis of the soluble enzyme [64] opens the way for fine-tuning of the catalytic versatility of sMMO for more precise and profitable biotransformations than are possible with the wild-type enzyme. Coupled with the sequencing of the Me. capsulatus genome, which is currently being undertaken by the University of Bergen and The Institute for Genomic Research, genetic technology may shortly enable metabolic engineering of novel pathways incorporating engineered methane monooxygenases for the synthesis of valuable Pharmaceuticals and other products using methane and other inexpensive starting materials. Acknowledgements We gratefully acknowledge research funding from the Biotechnology and Biological Sciences Research Council (UK), British Gas (UK), British Petroleum (UK), Idemitsu (Japan) and the Gas Research Institute (GRI) (Chicago, IL).

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13]

B.G. Fox, W.A. Froland, J.E. Dege and J.D. Lipscomb, J. Biol. Chem., 264 (1989) 10023. J.P. Bowman, S.A. McCammon and J.H. Skerratt, Microbiology, 143 (1997) 1451. L. Bodrossy, K.L. Kovacs, I.R. McDonald and J.C. Murrell, FEMS Microbiol. Lett., 170 (1999)335. W.S. Reeburgh, S. C. Whalen and M. J. Alperin, in J.C. Murrell and D.P. Kelly (eds.), Microbial growth on Cl compounds, Intercept, Andover UK, 1993, pp. 1-14. R.S. Hanson and T.E. Hanson, Microbiol. Rev., 60 (1996) 439. S.H. Stanley, S.D. Prior, D.J. Leak and H. Dalton, Biotechnol. Lett., 5 (1983) 487. G.A. Brusseau, H.-C. Tsien, R.S. Hanson and L.P. Wackett, Biodegradation, 1 (1990) 19. J. Colby and H. Dalton, Biochem. J., 171 (1978) 461. A.C. Stainthorpe, V. Lees, G.P. Salmond, H. Dalton and J.C. Murrell, Gene, 91 (1990) 27. M.P. Woodland, D.S. Patil, R. Cammack and H. Dalton, Biochim. Biophys. Acta, 873 (1986)237. A. Ericson, B. Hedman, K.O. Hodgson, J. Green and H. Dalton, J. Am. Chem. Soc. 110(1988)2330. J.G. DeWitt, J.G. Bentsen, A.C. Rosenzweig, B. Hedman, J. Green, S. Pilkington, G.C. Papaefthymiou, H. Dalton, K.O. Hodgson and S.J. Lippard, J. Am. Chem. Soc, 113 (1991)9219. A.C. Rosenzweig, C.A. Frederick, S.J. Lippard and P. Nordlund, Nature, 366 (1993) 537.

191

[14] A.C. Rosenzweig, H. Brandstetter, D.A. Whittington, P. Nordlund, SJ. Lippard and C.A. Frederick, Proteins, 29 (1997) 141. [15] N. Elango, R. Radhakrishnan, W.A. Froland, BJ. Wallar, C.A. Earhart, J.D. Lipscomb and D.H. Ohlendorf, Protein Sci., 6 (1997) 556. [16] A.R. George, P. C. Wilkins and H. Dalton, J. Molec. Catal. B, 2 (1996) 103. [17] K.J. Walters, G.T. Gassner, S.J. Lippard and G. Wagner, Proc.Natl.Acad.Sci.USA, 96 (1999)7877. [18] S.L. Chang, P.J. Wallar, J.D. Lipscomb and K.H. Mayo, Biochemistry, 38 (1999) 5799. [19] J. Lund and H. Dalton, Eur. J. Biochem., 147 (1985) 291. [20] K.E. Liu, A.M. Valentine, D. Wang, B.A. Salifoglou and S.J. Lippard, J. Am. Chem. Soc, 117(1995)10174. [21] B. Brazeau and J.D. Lipscomb, Biochemistry, 39 (2000) 13503. [22] L. Shu, J.C. Nesheim, K. Kauffmann, E. Miinck, J.D. Lipscomb and L. Que Jr., Science, 275(1997)515. [23] A.M. Valentine, S.S. Stahl and S.J. Lippard, J. Am. Chem. Soc, 121 (1999) 3876. [24] S.-K. Lee, J.C. Nesheim and J.D. Lipscomb, J. Biol. Chem., 268 (1993) 21569. [25] B.J. Brazeau, B.J. Wallar and J.D. Lipscomb, J. Am. Chem. Soc, 123 (2001) 10421. [26] J. Green and H. Dalton, J. Biol. Chem., 264 (1989) 17698. [27] K. Yoshizawa, T. Yamabe and R. Hoffmann, New J. Chem., 21 (1997) 151. [28] Y. Jin and J.D. Lipscomb, Biochim. Biophys. Acta, 1543 (2000) 47. [29] M.H. Baik, M. Newcomb, R.A. Friesner and S.J. Lippard, Chem. Rev., 103 (2003) 2385. [30] R.J. Deeth and H. Dalton, J. Biol. Inorg. Chem., 3 (1998) 302. [31] B.G. Fox, J.L. Bourneman, L.P. Wackett and J.D. Lipscomb, Biochemistry, 29 (1990) 6419. [32] J. Colby, D.I. Stirling and H. Dalton, Biochem. J, 165 (1977) 395. [33] S.D. Prior and H. Dalton, J. Gen. MicrobioL, 131 (1985) 155. [34] K.J. Burrows, A. Cornish, D. Scott and I.J. Higgins, J. Gen. MicrobioL, 130 (1984) 3327. [35] A.S. Lindner, P. Adriaens and J.D. Semrau, Arch. MicrobioL, 174 (2000) 35. [36] H.-H.T. Nguyen, K.H. Nakagawa, B. Hedman, S.J. Elliott, J.H. Yip, S.J. Jacobs, B.J. Hales, M.E. Lidstrom and S.I. Chan, J. Biol. Chem., 269 (1994) 14995. [37] D.D.S. Smith and H. Dalton, Eur. J. Biochem., 182 (1989) 667. [38] A.K. Shiemke, S.A. Cook, T. Miley and P. Singleton, Arch. Biochem. Biophys., 321 (1995)421. [39] J.A. Zahn and A.A. DiSpirito, J. Bacteriol., 178 (1996) 1018. [40] H.-H.T. Nguyen, S.J. Elliott, J.H.-K. Yip and S.I. Chan, J. Biol. Chem., 273 (1998) 7957. [41] P. Basu, B. Katterle, K.K. Andersson and H. Dalton, Biochem. J., 369 (2003) 417. [42] R.L. Lieberman, D.B. Shrestha, P.E. Doan, B.M. Hoffman, T.L. Stemmler and A.C. Rosenzweig, Proc Natl. Acad. Sci. USA, 100 (2003) 3820. [43] D.-W. Choi, R.C. Kunz, E.S. Boyd, J.D. Semrau, W.E. Antholine, J.-I. Han, J.A. Zahn, J.M. Boyd, A.M. de la Mora and A.A. DiSpirito, J. Bacteriol., 185 (2003) 5755. [44] S. Stolyar, A.M. Costello, T.L. Peeples and M.E. Lidstrom, Microbiology, 145 (1999) 1235. [45] B. Gilbert, I.R. McDonald, R. Finch, G.P. Stafford, A.K. Nielsen and J.C. Murrell, Appl. Environ. MicrobioL, 66 (2000) 966. [46] A.K. Nielsen, K. Gerdes and J.C. Murrell, Molec MicrobioL, 25 (1997) 399. [47] S. Stolyar, M. Franke and M.E. Lidstrom, J. Bacteriol., 183 (2001) 1810.

192

[48] A.A. DiSpirito, J.A. Zahn, D.W. Graham, H.J. Kim, C.K. Lerive, T.S. Derrick, CD. Cox and A. Taylor, J. Bacteriol, 180 (1998) 3606. [49] H.-H.T. Nguyen, A.K. Shiemke, S.J. Jacobs, BJ. Hales, M.E. Lidstrom, K.O. Hodgson and S.I. Chan, J. Am. Chem. Soc, 118 (1996) 12766. [50] S.A. Cook and A.K. Shiemke, Arch. Biochem. Biophys., 398 (2002) 32. [51] S.S.-F. Yu, L.-Y. Wu, K.H.-C. Chen, I.I. Luo, D.-S. Huang and S.I. Chan, J. Biol. Chem., 278 (2003) 40658. [52] S.J. Elliott, M. Zhu, L. Tso, H.-H.T. Nguyen, J.H.-K. Yip and S.I. Chan, J. Am. Chem. Soc, 119(1997)9949. [53] S. Lontoh, A.A. DiSpirito, C.L. Krema, M.R. Whittaker, A.B. Hooper and J.D. Semrau, Environ. Microbiol., 2 (2000) 485. [54] S. Lontoh, J.A. Zahn, A.A. DiSpirito and J.D. Semrau, FEMS Microbiol. Lett., 186 (2000) 109-113. [55] A.O. Richards, S.H. Stanley, M. Suzuki and H. Dalton, Biocatalysis, 8 (1994) 253. [56] S.H. Stanley, A.O. Richards, M. Suzuki and H. Dalton, Biocatalysis, 6 (1992) 177. [57] M. Suzuki, H. Dalton, A.O. Richards and S.H. Stanley, Canadian patent # 1,322,734. (1993). [58] S. Slade, T.J. Smith and H. Dalton, unpublished observations. [59] P.W. Baker, H. Futamata, S. Harayama and K. Watanabe, 38 (2001) 87. [60] M.C. White, A.G. Doyle and E.N. Jacobsen, J. Am. Chem. Soc, 123 (2001) 7194. [61] D. Lee and S.J. Lippard, J. Am. Chem. Soc, 123 (2001) 4611. [62] E.Y. Tshuva, D. Lee, W. Bu and S.J. Lippard, J. Am. Chem. Soc, 124 (2002) 2416. [63] P.-P.J.H.M. Knops-Gerits and W.P. Goddard, Catal. Today, 81 (2003) 263. [64] T.J. Smith, S.E. Slade, N.P. Burton, J.C. Murrell and H. Dalton, Appl. Environ. Microbiol, 68 (2002) 5265.


193

Chapter 7

Biocorrosion H.A. Videlaa and L.K. Herrerab department of Chemistry. College of Pure Sciences, INIFTA, University of La Plata, Argentina b

Faculty of Engineering, University of Antioquia, Medellin, Colombia

1. INTRODUCTION Microorganisms influence corrosion by changing the electrochemical conditions at the metal/solution interface. These changes may have different effects, ranging from the induction of localized corrosion, through a change in the rate of general corrosion, to corrosion inhibition [1]. Any biological effect that either facilitates or impedes one of the anodic or cathodic reactions of the corrosion process, or permanently separates anodic and cathodic sites, will increase corrosion. For instance, stimulation of the anodic reaction by acidic metabolites or the cathodic reaction by microbial production of a cathodic reactant like hydrogen sulfide, the breakdown of protective films or the increase in conductivity of the liquid environment will enhance corrosion. Although the electrochemical nature of corrosion remains valid for microbial corrosion, the participation of the microorganisms in the process induces several unique features, mainly the modification of the metal/solution interface by biofilm formation. Biofilms affect the interaction between metal surfaces and the environment, not only in biodeterioration processes like corrosion, but also in several biotechnological processes applied to materials recovery and handling [2]. Thus, the key to the alteration of conditions at a metal surface, and hence the enhancement or inhibition of corrosion is the formation of a biofilm [3]. This can be considered as a gel containing 95% or more water, made of a matrix of exopolysaccharidic substances (EPS) in which microbial cells, and inorganic detritus are suspended [4]. Biofilm formation is the result of an accumulation process -not necessarily uniform in time or space [5]- that starts immediately after metal immersion in the aqueous environment. A thin film (approximately 20-80 nm

194

thick), due to the deposition of inorganic ions and high relative molecular mass organic compounds, is formed in a first stage. This initial film is able to alter the electrostatic charges and wettability of the metal surface facilitating its further colonization by bacteria. In a short time, (minutes or hours according to the aqueous environment where the metal is immersed), microbial growth and EPS production results in the development of a biofilm. This biofilm is a dynamic system and the different transport processes and chemical reactions occurring at the biofouled interface will take place from now on, through the biofilm thickness [6] (Fig. 1). Microbial colonization of metal surfaces drastically changes the classical concept of the electrical interface commonly used in inorganic corrosion. Important changes in the type and concentration of ions, pH values and oxidation-reduction potential are induced by the biofilm, altering the passive or active behavior of the metallic substratum and its corrosion products, as well as the electrochemical parameters used for assessing corrosion rates [7]. Simultaneously with the biological changes that lead to biofilm accumulation, a sequence of inorganic changes takes place at the metal surface immediately after its immersion in an aggressive aqueous medium. This sequence involves the process of metal dissolution and corrosion product formation.

Fig. 1. Diagrammatic representation for biofilm development (from Ref. [8]).

195

Both biological and inorganic processes occur within the same time period, but following opposite directions at the metal/solution interface. Whereas corrosion and corrosion product accumulation occur from the metal surface towards the solution, biofilm formation is the result of accumulation processes directed from the bulk towards the metal surface (Fig. 2) [8]).

Fig 2. Sequence of biological and inorganic processes at a biologically and electrochemically conditioned metal/solution interface (from Ref. [8]).

Thus, a very active interaction between the corrosion product layers and the biofilms can be expected. The consequent corrosion behavior of the metal will vary according to the degree of this reciprocal interaction and a concept of a new biologically conditioned interface must be kept in mind [9]. The approach for a sound interpretation of any microbial corrosion case must then be interdisciplinary, and include a thorough process analysis combined with well defined microbiological and electrochemical methodologies. Biocorrosion has been focusing increasing attention from different research areas in the last two decades as an answer to the demand of a wide variety of industries. Fortunately an increasing intellectual and technical crossfertilization of ideas between researchers from different disciplines like microbiology, electrochemistry and materials science has allowed a considerable improvement in the understanding of biocorrosion to be reached. 2. SRB INDUCED CORROSION OF STEEL Biocorrosion of carbon steel in anaerobic environments involving the presence of SRB have been the focus of most biocorrosion research. Starting

196

with the cathodic depolarization theory (CDT) of von Wolzogen Kuhr & Van der Vlugt [10] (Fig. 3), a copious list of papers and reviews on the anaerobic corrosion of iron has been published [11-13]. Bacterial biofilms may develop anaerobic regions, even in aerobic bulk water environments [2], thus allowing SRB a very favorable environment for growth. The final result of these processes within biofilms is to produce a wide variety of sites on the metal surface that are markedly different from neighboring sites from a physicochemical standpoint, thus facilitating the initiation of localized corrosion processes.

Fig. 3. Sequence of reactions of the Cathodic Depolarization Theory. The three elements of biocorrosion (metal/solution/microorganisms) are involved in different reactions of the whole process.

As the localized corrosion and breakdown process is strongly dependent of several experimental factors such as the type and concentration of aggressive anions present in the medium and the passivating film characteristics, the effect of sulfur anions has been studied in a series of laboratory experiments using alkaline [14] and neutral buffered [15] solutions as well as SRB cultures in saline media [16] under well defined experimental conditions. Taking into account the results of these studies, a bioelectrochemical interpretation of the biocorrosion process of carbon steel in anaerobic environments may be summarized as follows [17]: i) biogenic sulfides effects on carbon steel passivity breakdown is similar to that of abiotic sulfides. The characteristics and intensity of sulfide effects on the corrosion behavior of carbon steel is narrowly related with the nature of the passive film already present on the metal surface (Fig. 4); ii) in neutral media sulfide ions lead to the formation of a poorly protective film of mackinawite; iii) the anodic breakdown of passivity would be the first stage of the corrosion process.

197

Fig. 4. Scheme of the bioelectrochemical interpretation of the biocorrosion process of carbon steel in anaerobic environments (from Ref. [17]).

Thus, the role of SRB may be indirect through the production of aggressive species either as final (sulfides, bisulfides or hydrogen sulfide) or intermediate metabolic compounds (thiosulfates, polythionates). Physicochemical characteristiscs of the liquid environment (pH, ionic composition, oxygen levels) can modify the SRB effects which could eventually change from corrosive to passivating; iv) cathodic depolarization effects attributed to SRB hydrogenase activity or to iron sulfide films would be developed later than passivity breakdown while corrosion process is in progress; v) the action of biogenic sulfides can be enhanced by other aggressive anions already present in the environment (e.g. chlorides) [18] or through microbial consortia within adhered biofilms on the metal surface [19]. Further studies developed in our laboratory [20-23] allowed us to conclude that SRB influenced corrosion of steel is markedly affected by the nature and the structure of the sulfide films produced during the metal dissolution. The environmental characteristics of the metal/biofilm/medium interface and its surroundings (pH, ionic composition, oxygen levels, EPS

198

distribution) control the chemical and physical nature of corrosion product layers and may change their effects on the metal behavior from corrosive to protective. In marine environments, the impact of sulfur compounds on corrosion is enhanced by other aggressive anions, such as the widely distributed chlorides, already present in the medium. The entrance of oxygen into the anaerobic environment accelerates corrosion rate, mainly through a change in the chemical nature of iron sulfides and elemental sulfur production. Both chemical species can provide additional cathodic reactants to the corrosion process, acting as electron carriers between the metal and the oxic interface within the biofilm. 2.1. Differences between biotic and abiotic media Recently, several surface analysis techniques, electrochemical experiments and microscopy observations were employed to clarify the role of biotic and abiotic sulfide films in the corrosion behaviour of steel in saline media [21, 23]. Microbiological experiments were performed under controlled laboratory conditions using a strain of Desulfovibrio alaskensis (D. alaskensis), known for its ability to produce EPS, isolated from a soured oil reservoir in Alaska [24]. Atomic force microscopy (AFM) was applied to image the SRB biofilms (Fig. 5), current transients were measured to determine the electrochemical behaviour of steel, and SEM observations coupled with ED AX, as well as XPS, XRD and electron microprobe analysis (EPMA) were carried out to examine the structure and composition of biotic and abiotic sulfide films on the surfaces of steel specimens. The main conclusions obtained from these studies have been: i) both abiotic and biotic iron sulfide films are related to the formation of tubercles on the steel surface. However, for biotic solutions FeS (mackinawite) predominates, whereas in abiotic media FeS2 (pyrite) is the major iron sulfide present (Fig. 6); ii) the structure of the outer crust of iron sulfide acts as a barrier for the diffusion of ions towards and from the pit cavity; iii) biotic films are more adherent to the metal surface, while abiotic films are flaky and loosely attached; iv) the previous history of the sulfide film may play a relevant role in the corrosion behaviour of steel. Depending on the sulfide concentration in the medium and on the presence or absence of biofilms and EPS, the protective characteristics of the corrosion product may change. Biogenic layers of corrosion products can offer enhanced protection by improving the adherence of the film to the metal, but can also increase corrosion, inducing the presence of heterogeneities at the metal surface.

199

Fig. 5. AFM image of a SRB biofilm on AISI 316 SS (from Ref. [21]).

The differences between biotic and abiotic media containing identical levels of corrosive compounds (i.e. iron sulfides) may be mainly attributed to the presence of extracellular polymers and to the heterogeneities created at the metal surface by the formation of a biofilm. In addition, the physicochemical parameters of the corrosion product layers may change in the absence or a presence of a biofilm, rendering these layers either protective or corrosive. This feature can explain the observed differences in corrosion behavior of steel between abiotic and biotic environments with respect not only to localized pitting, but also to the hydrogen attack developed as embrittlement and crack growth. 2.2. Hydrogen effects and SRB Although it is widely accepted that SRB can oxidize molecular hydrogen using their hydrogenase system, the involvement of this enzyme in the corrosion process remains unclear. The present state of knowledge does not support the hypothesis that the bacterial uptake of hydrogen is the rate-controlling step for SRB-influenced corrosion of mild steel [25].

200

Fig. 6. XPS spectra of a biogenic sulfide film (top) and of an abiotic sulfide film (bottom) (from Ref. [23]).

Hydrogen can have a considerable effect on susceptible alloys when it enters into the atomic structure of the metal and causes embrittlement. This phenomenon, including the acceleration of crack formation and growth, is frequently found in industrial systems (i.e. off-shore oil platforms). In addition, hydrogen, generated from cathodic protection can be a considerable problem in areas of stress corrosion or corrosion-fatigue [26]. The amount of hydrogen available to enter the steel and the ease with which it moves into the metal, is influenced by the nature of the environment in which the steel is placed (Fig. 7).

201

Fig. 7. Hydrogen permeation through 50D steel, cathodically protected or unprotected and coated or uncoated, exposed to open seawater and embedded in marine mud. 1 = Grit blasted plus cathodic protection. 2 = CTE coating plus cathodic protection. 3 = anti-fouling paint. 4 = grit basted only. 5 = As received (with mill scale, uncoated) (from Ref. [29]).

In seawater, where structures are cathodically protected from corrosion, hydrogen is available at the metal surface, particularly in anaerobic environments. Biological activity at the surface, especially bacterial activity, can enhance the entry of hydrogen into the metal. This action can be direct, by producing metabolites like sulfides which encourage the entry of hydrogen into the steel, and indirect, by both disrupting any surface films and due to the need of increasing cathodic protection when bacteria are present [27]. SRB are the main bacteria responsible for the enhancement of hydrogen entry as they are active in the areas where hydrogen will be generated under cathodic protection, i.e. anaerobic environments such as marine muds (Fig. 7). It is indisputable that a biologically active environment is able to induce conditions conductive to the enhancement of corrosion-fatigue crack growth and hydrogen embrittlement by the activity of SRB [28-30]. A "biologically active environment" involves not just the presence of microorganisms such as SRB or a single metabolite i.e. sulfide, but the effect of all the activities of the bacteria and their interactions with other components of the environment {e.g. degradation by large fouling organisms, production of EPS, interactions of

202

bacteria and the metal). Indeed, the local environment surrounding the metal surface is very different from that without bacteria, even if the same levels of sulfide are detectable. These areas are particularly important, as pipelines and the bottoms of the legs of offshore platforms can be buried in marine muds where anaerobic conditions are predominant and SRB activity is intense (Fig. 8). Moreover, sour environments, such as those frequently found in oil production activities, are particularly aggressive due to high levels of hydrogen available at the metal surface or in a crack, as a consequence of sulfide poisoning of the recombination reaction at the cathode [31]. In such habitats hydrogen effects can be altered by the presence of organic molecules on the metal surface and the existence of a biofilm with its EPS matrix. This feature can explain the differences between general embrittlement effects (as measured by hydrogen flux) and crak tip effects (as measured by crack growth). Embrittlement results in the general lowering of strength of the material causing it to fail in a catastrophic way at a lower load as it has been reported in high sulfide biological environments more than in low sulfide abiotic environments, even though the rest of the crack growth curve is very similar (Fig. 9) [28]. EPS and other organic molecules related to biofilms hinder dissolution and dissociation reactions and adsorption processes in the crack. Even under low frequency cyclic loading the crack tip opens and closes rapidly. Thus, any effect of the environment must occur fast and organic material dragged into the crack could have a blocking impact. The results referred here serve to illustrate the complex nature of the interactions between SRB biofilm and the steel. In many cases, bacterial metabolism within the biofilm generates sulfides, and consequently, this is the main cause of the corrosiveness of the environment. However, microbial metabolic activity is also responsible for the release of EPS which may have a blocking effect on hydrogen entry into the metal. 3. BIOCORROSION OF ALUMINUM ALLOYS IN FUEL/WATER SYSTEMS Microbial contamination of hydrocarbon fuels is the main cause of serious problems concerning the quality of maintenance of the product, as well as the corrosion of metals used during the processes of extraction, production, distribution, and storage of the fuel.

203

Fig. 8. Composite diagram of an offshore structure showing the main sites of biodeterioration problems: 1. marine fouling; 2. drill cuttings around legs; 3. oil storage and transport; 4. water-filled legs; 5. production system; 6. seawater injection system; 7. downhole pipework; 8. reservoir problems (from Ref. [31]).

Microoganisms usually present in droplets or thin films are sufficient to allow microbial growth and the development of biocorrosion. Because of the electrochemical nature of biocorrosion an aqueous environment is also required. Fungal and bacterial growth can also occur on side walls not necessarily adjacent to large waterbottoms. Hyphal growth of fungi usually ramify over fuel/water interface becoming a site for further water entrapment. As a result of this microbial propagation action, penetration and breakdown of fuel tank coatings and passive films occurs on the metal, leading to the onset of localized corrosion processes generally of the type of pitting attack. Due to its economic and technological importance, the biodeterioration of jet fuel and subsequent biocorrosion of aluminum alloys used in the aircraft industry has been extensively studied [32-35]. Since the end of the 70's several of our publications have been devoted to elucidate the mechanisms affecting the whole biocorrosion process [36-39].

204

Fig. 9. Crack growth rates of a RQT 501 steel in biologically active H2S seawater environments; solid line: crack growth rate in seawater; dotted line: crack growth rate in 520 ppm H2S in abiotic natural seawater (from Ref. [28]).

Different types of microorganisms usually present in soils and natural waters utilize paraffmic hydrocarbons in the range C,0-CIg (kerosene fraction) more easily than in the range C5-C9 (gasoline fraction). In a water-free fuel the microbial detrimental activity is nil. Conversely, as soon as water becomes available, the growth of microorganisms is possible by utilizing the hydrocarbons as a source of carbon for their metabolic activity. Chemical contaminants present in the fuel as well as in bilge water provide some kind of nitrogen source and the necessary trace elements for growth. Therefore, the major requirement for microbial activity is the presence of water in the storage tank. On this respect, during the processing of hydrocarbon products is practically impossible to avoid the existence of water. Although good housekeeping procedures could minimize the amount of water accompanying the fuel, the microorganisms are able to generate their own water phase for further proliferation [40]. Microbial metabolism generate in this way, small ecosystems which retain water in certain areas of the storage tank such as tubercles, slime deposits or even biofilms on the metal surface. The microbial ecology of aircraft fuel tanks and storage systems correspond to a wide diversity

205

of microorganisms, although in almost all cases the microbial contaminants reported were fungi and bacteria [41]. The fungus Hormoconis resinae (formerly Cladosporium resinae), and some species of the genera Aspergillus, Penicillium and Fusarium have been isolated in jet fuels and fuel storage systems [42]. H. resinae has also been reported as a contaminant of fuel oils for marine and terrestrial turbine engines [43]. Due to the high rate of consumption of kerosene fuel by jet engines, even small amounts of microbial sludges in the fuel becomes hazardous. The combination of microorganisms and their associated water creates major problems of fuel contamination, filter plugging, fuel gauge malfunction and fuel tank corrosion. The corrosion of the tank wall and subsequent leakage of hydrocarbon fuel can cause important economic losses, as well as different troubles related with soil and underground water contamination [44]. The corrosion attack is generally located at the tank bottom where there is an active microbial population associated with free water. To allow microbial growth, the environment must provide the basic elements (carbon, oxygen, hydrogen, nitrogen, phosphorous) and some other elements needed in much smaller amounts, but nevertheless essential for a normal metabolism. Although carbon and hydrogen are abundantly supplied by the hydrocarbon chains of the fuel (usually 95% of the product), an essential element for their biodegradation is oxygen. At low concentrations of oxygen, the rate of hydrocarbon oxidation is lower, although in aircraft fuel tanks there is a periodically replenishment of oxygen during tank refuelling. Usually, the nitrogen and phosphorous availability in kerosene fuel is the limiting factor for microbial growth. These elements are generally present as nitrates and phosphates, either in the water phase or in the additives present in the fuel. The action of fungal contaminants of jet fuels in corrosion can be accomplished through: a) Local increase in the proton concentration, derived from organic acidic metabolites, b) Greater oxidizing characteristics of the medium favoring pitting attack, c) Metabolite production decreasing the surface energy of the interface passive film/electrolyte d) Microbial adhesion processes enhancing metal dissolution, and e) Microbial uptake of corrosion inhibitors (mainly nitrates and phosphates). A simplified scheme to illustrate the mechanisms influencing biocorrosion of aluminum alloys in fuel / water systems is shown in Figure 10 [45]. A protective action of some contaminating bacteria (e.g. S. marcescens), can be developed when an active degradation of hydrocarbon chains is not accompanied by important reduction in pH levels. In those circumstances, the organic anions enhance the passive behavior of aluminum [1].

206

Fig. 10. Simplified scheme of the initiation of pitting attack on aluminum alloys in fuel/water systems (from Ref. [45]).

4. AEROBIC CORROSION OF IRON Metal-oxidizing microorganisms create environments for the accumulation of chloride ions, forming acidic ferric chloride and manganic chloride, which are corrosive to steel. However, the main mechanism to explain the action of these microorganisms in biocorrosion is the formation of differential aeration cells associated with tubercles. A typical aeration cell of this kind is formed when the tubercle structure at its outer area acts as a cathode due to the continuous oxygen supply from the water. Oppositely, an anode (anaerobic area ) is formed at the bottom of the formation, where access of oxygen is limited. In all cases corrosion acceleration is mainly due to the presence of SRB in the inner area of the tubercles where anaerobic conditions are created. Thus, it is highly feasible that the fastidious iron and manganese depositing bacteria require other organisms to create suitable conditions for their growth. One important feature of this case of biocorrosion is that once the tubercle (and the electrochemical cell) is formed, even the dead of the microorganisms does not

207

extinguish the differential aeration mechanism, since a substantial barrier to the intake of oxygen has been established. Thus, biocorrosion of cast iron in low flow or stagnant regions of water distribution pipelines is mainly due to tubercles formed by iron-oxidizing and manganese-oxidizing bacteria. These bacteria are able to oxidize soluble ferrous compounds to less soluble ferric compounds (e.g. ferric hydroxide) which precipitate at the inner wall of the pipelines under the form of tubercles (discrete hemispherical mounds). Tubercles create environments suitable for the growth of other hazardous microorganisms like the SRB which are generally found at the inner (anaerobic) regions of the deposits. Near the tubercles other aerobic microorganisms like slime-forming bacteria (e.g. Pseudomonas) are frequently found. In addition, tubercles are able to impede the penetration of biocides or corrosion inhibitors diminishing their effectiveness. Finally, iron-oxidizing bacteria are able to concentrate chlorides and manganese ions forming acidic ferric chloride and manganic chloride which are highly corrosive. 5. MICROBIAL INHIBITION OF CORROSION Corrosion inhibition is the slow down of the corrosion reaction usually performed by substances which, when added in small amounts to an environment, decrease the rate of attack by this environment on a metal (corrosion inhibitors). Microorganisms are able to drastically change the electrochemical conditions at the metal/solution interface. These changes can range from the induction or acceleration of corrosion to corrosion inhibition. There are several examples of microbial effects that could enhance corrosion: i) stimulation of the anodic reaction by acidic metabolites or the cathodic reaction by microbial production of a new alternative cathodic reactant (e.g. H 2 S); ii) the microbial breakdown of protective films or iii) the increase in conductivity of the liquid environment. However, microbial effects causing corrosion inhibition has been seldomly mentioned in the literature [46]. Microorganisms can aid to achieve corrosion inhibition according to some of the following mechanisms: i) neutralizing the action of corrosive substances present in the environment; ii) forming protective films or stabilizing a preexisting protective film on a metal; iii) inducing a decrease in the medium corrosiveness. As general key features of microbial inhibition of corrosion the following points can be summarized [47]: - Biocorrosion and its counter-process, microbial inhibition of corrosion, are rarely linked to a single mechanism of to a single species of microorganisms.

208

- Either the corrosive or the inhibitory action of bacteria are developed at biofilmed metal surfaces where complex biofilm/protective films interactions occur. Biological activity leads to important changes in the type and concentrations of ions, pH and oxygen levels inducing significant variations in the physical and chemical characteristics of the environment as well as in the electrochemical parameters used to measure the corrosion rate. - The main mechanisms of corrosion inhibition by bacteria are always linked to a marked modification of the environmental conditions at the metal/solution interface by biological activity. - Microbial corrosion inhibition is frequently accomplished through: i) a decrease in the cathodic rate by microbial consumption of a cathodic reactant (e.g. oxygen consumption by respiratory activity); ii) decreasing the medium aggressiveness in restricted areas of the metal solution interface (e.g. by neutralizing acidity); iii) providing or stabilizing protective films on the metal (e.g. biofilm exopolymers with metal binding capacity). - It must be stressed that in practical situations, the inhibitory action of bacteria can be reversed to a corrosive action within bacterial consortia structured in biofilm thickness. A proper understanding of the identity and role of microbial contaminants in the specific environment of a metal surface may be used to induce corrosion inhibition by bacteria as a useful tool to prevent frequent biodeterioration effects encountered in practice. 6. ELECTROCHEMICAL INTEPRETATION OF BIOCORROSION The basic concepts of electrochemical corrosion are valid in biocorrosion and could be used to interpret the acceleration of the corrosion process by microorganisms in different aqueous media both under anaerobic or aerobic conditions. Biocorrosion is rarely interpreted by a single mechanism or rarely caused by single species of microoganisms [7]. Hence, it is important to be cautious in the interpretation of data supplied by electrochemical methods. Frequently, these techniques have been used in complex media where the characteristics and properties of passive films are not well understood. For instance, the presence of complex deposits of corrosion products, metabolites and EPS may dramatically reduce the usefulness of some electrochemical results. Moreover, it has to be kept in mind that microbial colonization of passive metals can drastically change their resistance to breakdown by changing locally the type and concentration of ions, pH values, oxygen gradients and inhibitor levels. These changes should result in important alterations in the electrochemical behavior of the metal and in the electrochemical parameters measured in laboratory experiments. In

209

conclusion, whenever electrochemical techniques will be used for the evaluation of biocorrosion, the actual condition of the metal surface must be considered [9]. In classical electrochemical studies, the interface between a metal and the surrounding electrolyte has been characterized by a certain distribution of electrical charges giving rise to the so-called electrical double layer [8]. The knowledge about the structure of the double layer at the metal/solution interface is mainly based on experimental data obtained with the dropping mercury electrode. Thus, the behavior of the interface between mercury and several aqueous electrolyte solutions could be considered approximately equal to that of an ideally polarizable interface. However, at the light of the present knowledge on biocorrosion, it can be easily inferred that this behavior is markedly different from that corresponding to the complex metal/solution interface associated with biocorrosion [9]. Consequently, electrochemical concepts used for inorganic corrosion analysis will have to be adapted to the characteristics of the biologically conditioned interface. A detailed description of the electrochemical methods for evaluating biocorrosion is out of the scope of this chapter. A wide variety of electrochemical techniques such as corrosion and redox potentials measurements, Tafel and potentiodynamic polarization, linear polarization and electrical resistance probes, and several modern electrochemical techniques like alternating current methods or electrochemical noise have been reviewed by several authors in relation to their use in biocorrosion evaluation [48-50]. A careful use of several electrochemical concepts and methods should be coupled when possible, with visual inspection, microscopy, innovative surface analysis techniques and a sound microbiological analysis to reach an adequate characterization of the causative microorganisms, their interactions and the effects of their metabolic activities and products on corrosion. 7. PREVENTION AND CONTROL One of the classic concepts for maintaining an industrial system free of biodeteriorating effects is "to keep the system clean". Although this is a very difficult task to accomplish in practice, there are several general methods that can be used. These methods can be broadly classified in: i) physical and, ii) chemical. Among the former, flushing is perhaps the most simple, although of limited efficacy. A special case is the use of flushing supported by cleaners or jointly with chemical agents that induce biofilm detachment. Abrasive or nonabrasive sponge balls are frequently employed in the industry. The former could present problems related with protective passive films that can be damaged, and the second is not very effective with thick biofilms. With reference to chemical methods, the most common approach for controlling biofouling problems in industrial water systems, is the use of

210

biocides. These substances can be either oxidizing or non-oxidizing toxicants. Chlorine, ozone and bromine are three typical oxidizing agents of industrial use. Non-oxidizing biocides are claimed to be more effective than oxidizing biocides for an overall control of algae, fungi, and bacteria. They have a greater persistence and many of them are pH independent. Often a combination of oxidizing and non-oxidizing biocides or two non-oxidizing biocides is used to optimize microbiological control of industrial water systems. Typical biocides of the second type are formaldehyde, glutaraldehyde, isothiazolones and quaternary ammonium compounds. Increasing legislative requirements and the necessity for greater environmental acceptability, have contributed to restrict the use of some traditional biocides and to develop either new compounds or carefully selected blends of existing biocides.

Table 1 Biocides used in industrial water systems. Properties and usual concentrations. Chlorine: effective against bacteria and algae; oxidizing; pH dependent; concentration range: 0.1-0.2 mg/1 (continuous treatment) Chlorine dioxide: effective against bacteria, in a lesser extent against fungi and algae; oxidizing; not dependent on the pH; concentration range: 0.1-1.0 mg/1 Bromine: effective against bacteria and algae; oxidizing; wide pH range; concentration range: 0.05-0.1 mg/1 Ozone: effective against bacteria and biofilms; oxidizing; pH dependent; concentration range: 0.2-0.5 mg/1 Methylene-bis-thiocyanate: effective against bacteria; non-oxidizing; hydrolyses at pH higher than 8.0; concentration range: 1.5-8.0 mg/1 Isothiazolones: effective against bacteria, algae and biofilms; non-oxidizing; not dependent on the pH; concentration range: 0.9-10 mg/1 QUATS: effective against bacteria and algae; non-oxidizing; surface activity; concentration range: 8-35 mg/1 Glutaraldehyde: effective against bacteria, algae, fungi and biofilms; non-oxidizing; wide pH range; concentration range: 10-70 mg/1 THPS (tetra kis-hydroximethil phosphonium): effective against bacteria, algae and fungi; low environmental toxicity; specific action againts BRS.

211

Taking into account environmental concerns, the use of ozone for different types of industrial water systems, presents several advantages with respect to other biocides [51]. The unique combination of high toxicity of ozone during treatment, with no toxicant discharge, could make ozone the biocide of choice for the present decade, if an appropriate balance between positive effects and costs is reached. Several publications from our laboratory on ozone biocidal action on sessile and planktonic bacteria, its mechanisms of disinfection and the optimization of its use can be found in the literature [52]. Among the most promisory non-oxidizing biocides, the THPS (tetra kishydroximethil phosphonium) is a new compound of wide spectrum: effective on bacteria, fungi and algae. It is being widely used in the oil industry due to its capacity of dissolving the ferrous sulfide. Its main advantange is its low environmental toxicity. 8. MONITORING BIOCORROSION Monitoring programs for biofouling and biocorrosion have been mainly focused in the assessment of planktonic populations in water samples, and generalized corrosion by using corrosion coupons or some kind of resistance or polarization resistance probes. The main objections to these monitoring programs are: i) the planktonic population does not properly reflect the type and number of organisms living in the biofilm and causing biodeterioration problems; ii) susceptibility of planktonic microorganisms to antimicrobial agents markedly differs from that of sessile microorganisms within the biofilm, mainly because of the protective action of their EPS. Thus, the monitoring methods adopted must provide information of well-established biofilms like those developed in system water. From the corrosion side, the electrical resistance method is appropriate for indicating a change in the general corrosion rate, but the results are difficult to interpret in the presence of localized corrosion like pitting, the most frequent form of attack found in biocorrosion cases [48]. If biofilms or localized corrosion are present, the polarization resistance will reveal that something is happening, but may not give an accurate measure of the corrosion rate. Only the use of any of these techniques jointly with other electrochemical methods or parameters assessing localized corrosion hazard can provide valuable data for monitoring the deleterious effects of biocorrosion and biofouling. Owing to the variables of dissimilar nature involved in biofouling and biocorrosion, an effective monitoring program, either for the laboratory or the field, must necessarily supply information on water quality, corrosive attack, sessile and planktonic bacteria populations, biofilms characteristics, and chemical composition of inorganic and biological deposits [53].

212

Fig. 11. RENAprobe

sampler device.

Sampling devices for monitoring biocorrosion and biofilms can be simultaneously used to assess corrosion attack after the removal of biological and inorganic deposits, giving a wider and more useful information. Sampling devices fall into two main types: a) directly implanted and, b) side-stream implanted. Metal coupons, generally made with the same structural material of the system, present a known surface area, to enable an accurate count of sessile bacteria per square cm. after biofilm detachment. Coupons are mounted in holding assemblies which are inserted in the pipework of the laboratory or industrial system. Two practical cases of monitoring biocorrosion in industrial waters developed in our laboratory have been described in the literature [54, 55]. 8.1. Monitoring biocorrosion in chemical industry cooling water systems A monitoring program based on laboratory and field measurements for assessing biodeterioration on mild steel and stainless steel in recirculating cooling water systems has been reported [54]. This program was based on: i) water quality control; ii) corrosion monitoring in the field (weight loss and linear polarization probe); iii) laboratory corrosion tests (polarization techniques and corrosion potential vs. time measurements), and iv) use of a multipurpose sampling device that allows monitoring of sessile populations, biofilms, corrosive attack, morphology and intensity, and biological and inorganic deposits analysis. The side stream sampling device (RENAprobe™) allows realiable biofilm and corrosion sampling without introducing important modifications in the system (Fig. 11).

213

8.2. Oilfield water injection system Among the several factors related to biocorrosion hazard in secondary oil recovery operations are: i) velocity, temperature, oxygen level and redox potential of the injection water; ii) chemical composition, pH amount of organic matter and depth of pumping of the injection water and; iii) effectivenes of the biocide to gain access to the biofilms. Taking into account these factors a multiple monitoring program for biofouling and biocorrosion was implemented in an oild field in Argentina [55]. The goal of the program was to perform a rapid evaluation of the bacteriological status of the system, and to implement an effective biocide treatment for sessile bacteria. 9. NEW DEVELOPMENTS IN BIOCORROSION Recent improvements in analytical, microbiological, electrochemical and microscopical techniques and instrumentation have allowed the development of new methods for laboratory and field assessment of biocorrosion in industrial systems. Chemical analysis within the biofilm by means of microsensors is one of the most exciting advances in instrumentation [56]. Biofilm systems have been considered to be diffusion limited [57]. As a consequence, chemical conditions at the surface and within biofilms can vary dramatically over a distance of a few micrometers. Thus, the information obtained from bulk water analysis has a limited value and must be closely analysed before any conclusions are drawn about the behavior of the metal/solution interface. Direct measurements inside biofilms are restricted by: i) the small thickness of the biofilm; ii) the diffusion limitation of concentration profiles across the biofilm; iii) the heterogeneous nature of the biofilm. The latter aspect is specially important not only in relation to microbial coverage of the surface but also with respect to biocorrosion. An example of microsensor technology applied to evaluate vertical profiles of chemical species in biofilm systems has been reported [58]. Equally important tools for the study of biofilm structure are the fiber optic microprobe or optrode used for finding the location of the biofilm/bulk water interface, and the mapping of electric fields [59] by means of the scanning vibrating microprobe (SVM). Advanced microbiological techniques such as DNA probes have also been applied to biocorrosion and biofouling research [60]. Although these techniques are restricted to the laboratory at present, their joint utilization with microbiological field measurements can be highly useful for monitoring biocorrosion. Recent developments in microscopy such as the environmental scanning electron microscope (ESEM), the confocal laser microscope (CSL) and the

214

atomic force microscope (AFM) permit biofilm observation in real time and without intoducing distortion of the samples. There is an increasing number of references using these innovative technologies in recent biocorrosion literature [61-63]. A combination of CSL and microelectrode techniques allowed correlation of oxygen concentration profiles with biofilm structure [64]. CSL facilitates the visualization of biofilm structures by eliminating the interference arising from out of focus objects [65]. Observations performed under flow conditions and using physiologically active biofilms, provided information to construct a new conceptual model of biofilm structure. On the corrosion side, new electrochemical test methods for the study of localized corrosion phenomena in biocorrosion analysis and monitoring have been reported [66]. As an example, an electrochemical sensor for monitoring biofilms on metallic surfaces in real time has been recently presented [67]. The system provides an immediate indication of the condition of biological activity on probe surfaces and it is a powerful tool to optimize biocide treatment (Fig. 12). 10. CONCLUDING REMARKS Biocorrosion is rarely linked to a single mechanism or to a single species of microorganisms. Biofilms mediate the interaction between metal surfaces and the liquid environment. This interaction leads to an important modification of the metal/solution interface drastically changing the types and concentrations of ions, pH and oxygen levels. Biofilm/corrosion product layers interactions at the metal/solution interface condition the electrochemical behavior of metal surfaces in biological media. Acknowledgement H.A. Videla acknowledges the financial support of the Agencia de Promocion Cientifica y Tecnologica of Argentina through the project PICT/99 6782 on Biodeterioration of materials.

215

Fig. 12. Scheme of an electrochemical sensor for monitoring biofilms (from Ref. [67])

REFERENCES [1] [2] [3] [4] [5]

[6] [7] [8] [9]

H.A. Videla, Corrosion/96, paper No. 223, NACE International, Houston, TX (1996). H.A. Videla and W.G. Characklis, Int. Biodeterior. Biodegr., 29 (1992) 195. R.G.J. Edyvean and H.A. Videla, Interdisc. Sci. Rev., 16 (1991), 267. G.G. Geesey, Am. Soc. Microbiol. News, 48, (1982) 9. W.G. Characklis and K.C. Marshall, Biofilms: A Basis for an interdisciplinary approach. In: Biofilms, W.G. Characklis and K.C. Marshall (eds.) John Wiley and Sons Ltd, New York, 1990, pp. 3-15. W.G. Characklis, Biotech. Bioeng., 23, (1981) 1923. H.A. Videla, Metal Dissolution/redox in Biofilms. In: Structure and Function of Biofilms, W.G. Characklis and P.A. Wilderer (eds.) John Wiley & Sons, Chichester, U.K, 1989, pp. 301-320. H.A. Videla, Microbially induced corrosion: An updated overview. In: Biodeterioration and Biodegradation 8, H.W. Rossmoore (ed.) Elsevier Applied Science, London, UK, 1991, pp. 63-88. H.A. Videla, Electrochemical aspects of biocorrosion. In: Bioextraction and biodeterioration of metals, C.C. Gaylarde and H.A. Videla (eds.) Cambridge University Press, Cambridge, UK, 1995, pp. 85-127.

216

[10] G.A.H. von Wolzogen Kiihr and L.R. Van der Vlugt, Water (den Haag), 18, (1934) 147 (Translation in Corrosion 17, (1961) 293). [11] J.D.A. Miller, Metals, in: Microbial biodeterioration, A.D. Rose (ed.), Academic Press, New York, 1981, pp. 149-202. [12] R.C. Salvarezza and H.A. Videla, Corrosion, 36 (1980) 550. [13] G. Gragnolino and O.H. Tuovinen, International Biodeterioration, 20 (1984) 9. [14] R.C. Salvarezza, H.A. Videla and AJ. Arvia, Corros. Sci., 22 (1982) 815. [15] R.C. Salvarezza, H.A. Videla and AJ. Arvia, Corros. Sci., 23 (1983) 717. [16] H.A. Videla, Corrosion of mild steel induced by sulfate-reducing bacteria -a study of passivity breakdown by biogenic sulfides. In: Biologically Induced Corrosion, S.C. Dexter (ed.), NACE-8, Houston, TX, 1986, pp. 162-171. [17] H.A. Videla, Electrochemical interpretation of the role of microorganisms in corrosion. In: Biodeterioration 7, D.R. Houghton, R.N. Smith and H.O.W. Eggins (eds.) Elsevier Applied Science, London, UK, 1988, pp. 359-371. [18] C.A. Acosta, R.C. Salvarezza, H.A. Videla and A.J. Arvia, Electrochemical behaviour of mild steel in sulfide and chloride containing solutions, Passivity of Metals and Semiconductors, M. Froment (ed.) 1984, pp. 387-392. [19] J.W. Costerton and G.G. Geesey, The microbial ecology of source colonization and of consequent corrosion. In: Biologically Induced Corrosion, S.C. Dexter (ed.) NACE-8, Houston, TX, NACE, 1986, pp. 223-232. [20] H.A. Videla, C.L. Swords, M.F.L. de Mele, R.G. Edyvean, P,Watkins and I.B. Beech, Corrosion/98, paper No. 289, NACE International, Houston, TX (1998). [21] H.A. Videla, R.G. Edyvean, C.L. Swords, M.F.L. de Mele and I.B. Beech, Corrosion/99, paper No. 163, NACE International, Houston, TX (1999). [22] H.A. Videla, Biofouling, 15 (2000) 37. [23] H.A. Videla, C.L. Swords and R.G. Edyvean, Corrosion/2002, paper No. 02557, NACE International, Houston, TX (2002). [24] I.B. Beech, V. Zinkevich, R. Tapper and R. Avci, Journal of Microbiological Methods, 36(1999)3. [25] W. Lee, Z. Lewandowski, P.H. Nielsen and W.A. Hamilton, Biofouling, 8 (1995) 165. [26] Turnbull (ed) Hydrogen transport and cracking in metals, London Institute of Materials, UK, 1995. [27] S.G. Gomez de Saravia, M.F.L. de Mele, H.A. Videla and R.G.J. Edyvean, Biofouling, 11 (1997) 1. [28] CJ. Thomas, R.G.J. Edyvean and R. Brook, Biofouling, 1 (1988) 65. [29] R.G.J. Edyvean, J. Benson, CJ. Thomas, I.B. Beech, and H.A. Videla, Corrosion/97, paper No. 206, NACE International, Houston, TX (1997). [30] R.G.J. Edyvean, J. Benson, I.B. Beech and H.A. Videla, Microbiological influences on hydrogen effects on steels. In: Hydrogen Energy Progress XII, vol. 3, J.C. Bolcich and T.N. Veziroglu (eds.) 1998, pp. 1755-1762. [31] R.G.J. Edyvean, International Biodeterioration, 23 (1987) 199. [32] E.C. Hill, Biodegradation of petroleum products. In: Petroleum Microbiology, R.M. Atlas (ed.), Chap. 15, MacMillan Publishing Co., New York, 1984, p. 579-617. [33] P. Me Kenzie, A.S. Akbar and J.D. Miller, Fungal Corrosion of Aircraft Fuel Tank Alloys, Technical paper, The Institute of Petroleum, London, UK (1977) 37. [34] H.G. Hedrick, Materials Protection, 9 (1970) 27. [35] D.G. Parbery, Material und Organismen, 6 (1971) 161. [36] M.F.L. de Mele, R.C. Salvarezza and H.A. Videla, International Biodeterioration Bulletin, 15 (1979) 39.

217

[37] R.C. Salvarezza, M.F.L. de Mele and H.A. Videla, International Biodeterioration Bulletin, 15 (1979) 125. [38] R.C. Salvarezza, M.F.L. de Mele and H.A. Videla, Br. Corros. J., 16 (1981) 162. [39] R.C. Salvarezza and H.A. Videla, Electrochemical behavior of aluminum in Cladosporium resinae cultures, in: Biodeterioration 6, S. Barry, D.R. Houghton, G.C. Llewellyn and C.E. O' Rear (eds.) CAB International Mycological Institute, London, UK, 1986, p. 212-218. [40] RJ. Watkinson, Hydrocarbon degradation, in: Microbial Problems and Corrosion in Oil and Oil Products Storage, E.C. Hill (ed.) The Institute of Petroleum, London, UK, 1984, pp. 50-56. [41] R.A. Neihof and M. May, International Biodeterioration Bulletin, 19 (1983) 59. [42] H.A. Videla, P.S. Guiamet, S.M. do Valle and E.H. Reinoso, Corrosion/88, paper No. 91, NACE, Houston, TX (1988). [43] D. Allsopp and K.J. Seal, Introduction to Biodeterioration, Edward Arnold, London, 1986. [44] S. Holmes, Microbiology of hydrocarbon fuels. In: Proceedings 2nd International Conference on Long-Term Storage Stabilities of Liquid Fuels, Southwest Research Institute, San Antonio, TX, 1986, 336. [45] H.A. Videla, Practical Cases. In: Manual of Biocorrosion, Chapter 7, Boca Raton, Florida: CRC Lewis Publishers, 1996, pp. 179-220. [46] H.A. Videla, Corrosion inhibition in the presence of microbial corrosion. In: Reviews on Corrosion Inhibitor Science and Technology, Vol 2, A. Raman and P. Labine (eds.), Chapter IX, NACE International, Houston, TX, 1996, pp. 1-11. [47] H.A. Videla, Corrosion Inhibition by Bacteria. In: Manual of Biocorrosion, Chapter 5, Boca Raton, Florida: CRC Lewis Publishers, 1996, pp. 121-135. [48] S.C. Dexter, D.J. Duquette, O.W. Siebert and H.A. Videla, Corrosion/89, paper No. 616, NACE International, Houston, TX (1989). [49] D.J. Duquette, Electrochemical techniques for evaluation of Microbiologically Influenced Corrosion processes. Advantages and disadvantages In: Argentine-USA Workshop on Biodeterioration (CONICET-NSF), H.A. Videla (ed.) Aquatec Quimica S.A., Sao Paulo, Brazil, 1996, pp. 15-32. [50] F. Mansfeld and B J. Little, Corrosion/90, paper No. 108, NACE International, Houston, TX (1990). [51] J.M. Brook and P.R. Puckorius, Corrosion/91, paper No. 212, NACE International. Houston. TX (1991). [52] H.A. Videla, M.R. Viera, P.S. Guiamet and M.F.L. de Mele, Corrosion/99, paper No. 186, NACE International, Houston, TX (1999). [53] H.A. Videla, Detection, Identification and Monitoring. In: Manual of Biocorrosion, Chapter 6, Boca Raton, Florida: CRC Lewis Publishers, 1996, pp. 137-178. [54] H.A. Videla, F. Bianchi, M.M.S. Freitas, C.G. Canales and J.F. Wilkes, Monitoring biocorrosion and biofilms in industrial waters: a practical approach. In: Microbiological Influenced Corrosion Testing, J.R. Kerns and B.J. Little (eds.) ASTM Publications STP 1232, American Society for Testing and Materials, Philadelphia, PA, 1994, pp. 128137. [55] H.A. Videla, P.S. Guiamet, O.R. Pardini, E. Echarte, D. Trujillo and M.M.S. Freitas, Corrosion/91, paper No. 103, NACE International, Houston, TX (1991). [56] Z. Lewandowski, Dissolved oxygen gradients near microbially colonized surfaces. In: Biofouling and biocorrosion in industrial water systems, G.G. Geesey, Z. Lewandowski and H.C. Flemming (eds.) Lewis Publishers, Boca Raton, FL, 1994, pp. 175-188.

218

[57] W.G. Characklis, Biofilm processes, in: Biofilms, W.G. Characklis and K.C. Marshall (eds.) John Wiley & Sons, New York, 1990, pp. 195-231. [58] Z. Lewandowski, T. Funk, F. Roe and B. Little, Spatial distribution of pH at mild steel surfaces using an iridium oxide microelectrode. In: Microbiologically Influenced Corrosion Testing, J.R. Kearns and BJ. Little (eds.) ASTM Publications STP 1232, Philadelphia, PA, 1994, pp. 61-69. [59] Z. Lewandowski, F. Roe, T Funk and D. Chen, Proc. NSF-CONICET Workshop, Biocorrosion and Biofouling: Metal/Microbe Interactions, Buckman Laboratories International, Inc., Memphis, TN, 1993, pp. 52-61. [60] S. Le Borgne, J. Jan, J.M. Romero and M. Amaya, Corrosion/2002, paper No. 02461, NACE International, Houston, TX, (2002). [61] A. Steele, D.T. Goddard and I.B. Beech, Int. Biodet. Biodegrad. 34 (1994) 35. [62] A. Steele, I.B. Beech and D.T. Goddard, Proc. 1995 International Conference on Microbialy Influenced Corrosion, American Welding Society-NACE International, Houston, TX, 1995, 73/1-13. [63] J.W. Costerton, Structure of biofilms. In: Biofouling and Biocorrosion in Industrial Water Systems, G.G. Geesey, Z. Lewandowski and H.C. Flemming (eds.) Lewis Publishers, Boca Raton, FL, 1994, pp. 1-14. [64] Z. Lewandowski, P. Stoodley and F. Roe, Corrosion/95, paper No. 222, NACE International, Houston, TX (1995). [65] D.E. Caldwell, D.R. Korber and J.R. Lawrence, Adv. Microbial Ecol., 12 (1992) 1. [66] H.A. Videla, Fundamentals of electrochemistry. In: Manual of Biocorrosion, Chapter 4, Boca Raton, Florida: CRC Lewis Publishers, 1996, pp. 73-120. [67] G.J. Licina, Corrosion/2001, paper No. 01442, NACE International, Houston, TX (1995).


219

Chapter 8

Molecular tools in microbial corrosion X. Zhu and J.J. Kilbane II Gas Technology Institute, 1700 S. Mt. Prospect Rd., Des Plaines IL 60018 1. INTRODUCTION Corrosion is a leading cause for pipe failure, and is a main component of the operating and maintenance costs of gas and oil industry pipelines [1-5]. Quantifying the cost of corrosion generally in the gas and oil industry, and more specifically the cost associated with microbial corrosion, is not easily done and is controversial. Corrosion was estimated in 2001 to cost the gas and oil industry about $13.4 billion/yr and of this as much as $2 billion/yr may be due to MIC [6]. Basic research to increase our understanding of the microbial species involved in microbial corrosion and their interaction with metal surfaces and with other microorganisms will be the basis for the development of new approaches for the detection, monitoring, and control of microbial corrosion. A thorough knowledge of the causes of MIC and an efficient and effective means of detecting and preventing corrosion is lacking. It is well recognized that microorganisms are a major cause of corrosion of metal pipes, but despite decades of study it is still not known with certainty how many species ofmicroorganisms contribute to corrosion, how to reliably detect their presence prior to corrosion events, or how to rapidly assess the efficacy of biocides/mitigation procedures [2-4, 7-12]. Investigations of microbial species present in gas and oil industry pipelines have traditionally relied upon using samples obtained from pipelines to grow bacterial cultures in the laboratory [11]. Laboratory growth medium cannot accurately reflect the true conditions within pipelines and microbiologists have recognized that the vast majority of microbial species cannot currently be grown in the laboratory [13-17]. Thus, culture-dependent approaches underestimate the biocomplexity of microbial communities. Genetic methods can be used to overcome the difficulties associated with the laboratory cultivation of bacteria and provide a direct analysis of samples. In the past decade the use of genetic techniques to detect, identify, and quantify bacteria in

220

the medicine, food, and cosmetic industries has largely replaced microbial growth tests. These modern biotechnology methods are only beginning to be employed in the gas and oil industry for problems related to microbiologically influenced corrosion but it is likely that genetic techniques will be the methods of choice for monitoring MIC in the future. Initial efforts to introduce the use of genetic techniques for monitoring MIC or other environmental samples have involved a type of DNA hybridization test called reverse sample genome probing (RSGP). There are other hybridization-based genetic techniques including whole cell in situ fluorescent hybridization, DNA amplification followed by hybridization (dot-blot hybridization) or gel electrophoresis (denaturing gradient gel electrophoresis) [17-21] that could also be used to examine MIC samples, but this has not yet occurred. Another type of genetic test method that could be used to investigate MIC samples is based on DNA amplification using the polymerase chain reaction (PCR). PCR-based approaches include quantitative competitive PCR (cPCR), quantitative real-time PCR (q-PCR), and reverse transcriptase PCR (RT-PCR) [22-25]. This chapter summarizes the status of genetic tests to monitor MIC and discusses possible applications of genetic monitoring techniques for the future. 2. DNA HYBRIDIZATION TECHNIQUES TO MONITOR MICROBIOLOGICALLY INFLUENCED CORROSION One of the most commonly used genetic approaches to the monitoring of bacteria associated with corrosion is reverse sample genome probing (RSGP) [26-29]. This technique is based upon the hybridization of whole-community DNA obtained from environmental samples to the DNA of various characterized bacterial species (reference cultures) that are spotted on the master membrane. Hybridization experiments take advantage of the fact that DNA is normally double stranded so that if DNA is denatured (the strands are separated) complementary strands/DNA sequences will come back together. The additive nature of the hydrogen bonds formed by complementary nucleic acid bases provides a strong thermodynamic force favoring the formation of double stranded DNA by complementary DNA sequences. In RSGP various pure cultures, most frequently sulfate reducing bacteria (SRB), are isolated from environments of interest and then grown in the laboratory. About thirty cultures of SRB have been cultivated from gas and oil production operations so that the DNA of these cultures can be used in RSGP [30]. RSGP technique was also used to identify and quantify bacteria communities capable of degrading various aromatic hydrocarbons in contaminated soil [31-34]. Chromosomal DNA is purified from each species of bacterial culture that was cultivated in the laboratory. The DNA is then denatured and added as discrete drops to specific locations on a membrane, typically a nitrocellulose membrane. The DNA is then

221

chemically /thermally bound to the membrane, which then serves as the master grid in a subsequent hybridization experiment. To perform RSGP an environmental sample of interest, such as biomass obtained from a gas and oil production pipelines, is processed to obtain DNA from the mixture of bacterial species present. This mixed DNA sample is subjected to a biochemical labeling process that results in the addition of fluorescent or radioactive compounds to the DNA mixture [32]. When the labeled DNA mixture is denatured and added to the master grid complementary DNA strands form double stranded DNA, a washing step then removes any DNA not bound to the membrane. The amount of DNA in the mixture that corresponds to each species of bacteria present on the master grid is determined by quantifying the amount of fluorescence or radioactivity associated with each spot/location on the master grid. The quantity of signals in each spot on the master grid reflects the abundance of each reference organism in the environmental sample. This technique allows the quantification of many different microorganisms simultaneously. It can be readily appreciated that there are several drawbacks to RSGP. The technique requires specialized training and equipment and involves multiple steps so that several days are typically needed to obtain results. However, a more important limitation is that the only species of bacteria quantified by RSGP are those whose genomic DNA is spotted onto the master grid. Since it is generally accepted that less than 1% of bacterial species in nature can be cultivated in the lab [35, 36], only a small subset of bacterial species are available with which to prepare master grids. In microbial corrosion research, RSGP has been applied almost exclusively to the quantification of SRB. While SRB are unquestionably capable of causing metal corrosion it is also unquestionably true that several other types of bacterial groups such as acid producing bacteria, iron respiring bacteria, denitrifying bacteria, sulfur oxidizing bacteria, and methanogenic bacteria also can cause metal corrosion (see chapter 7). It would be extremely cumbersome to quantify all of these bacterial types using RSGP, but it may be possible to develop such tests in the future using microarray technology. Genetic methods can indeed provide more accurate data more quickly than microbial growth tests, but methods more convenient than RSGP are needed. The genetic method of choice for monitoring food and cosmetic products for microbial contamination, for the detection and identification of infectious microorganisms in medicine and bio-warfare monitoring is quantitative real-time PCR. Quantitative PCR has recently been adapted for use in monitoring microbial populations in gas and oil pipelines, and may become the most convenient and reliable method to monitor MIC in the future.

222

3. CHARACTERIZATION OF BACTERIAL AND ACHAEAL COMMUNITIES IN GAS PIPELINES BY DGGE AND 16S rRNA GENE SEQUENCING Before accurate and reliable genetic tests can be developed to quantify microbial species in a given environment it is helpful to first determine the composition of the microbial community as completely as possible. Molecular analysis of bacterial community structure of environmental samples has become a useful means of examining microbial communities. Molecular analysis involves extraction and purification of nucleic acids from environmental samples, PCR amplification of nucleic acids, followed by cloning of PCR products into a vector and sequencing of cloned PCR products, or followed by community fingerprinting techniques such as DGGE to separate the amplified PCR products [14, 17]. Each DGGE band is presumably representative of a specific bacterial population and the number of distinctive bands is indicative of total community richness. In addition, separated DDGE bands can also be cloned and sequenced. The comparison of DNA sequences of 16S rRNA genes with DNA sequence databases such as GenBank allows the identity of the species of microorganisms present in environmental samples to be determined. A greater understanding of the full range of bacterial species present in gas and oil production operation environments where corrosion is occurring will improve our understanding of the problem and our ability to detect and control it. Genetic techniques have been used to characterize the microbial communities present in natural gas pipelines and a thorough report has recently been published by the authors [18]. The important observations obtained from that study were that while the composition of microbial communities from different pipelines varies significantly, there were some commonly encountered species or types of microorganisms. Denitrifiers (bacteria that utilize nitrate and nitrite), such as Comamonas denitrificans, were found to be the most commonly encountered type of microorganism in gas pipelines. The presence of denitrifying bacteria in gas pipelines has not previously been reported in the literature and it is not routinely monitored in microbiological testing of gas pipeline samples. However, the frequent occurrence of denitrifiers in gas pipelines, and the proven ability of denitrifiers to contribute to corrosion [37, 38], suggests that denitrifiers should be monitored and that nitrate plays a key role in metabolism of biofilm microorganisms present in gas pipelines. This is particularly interesting because gas pipeline liquids do not typically contain significant levels of nitrate (< 5 mg/L). Even though the absolute concentration of nitrate may be low, the ability of different members in a microbial community to oxidize as well as to reduce nitrogen compounds may lead to continuous cycling of nitrogen within microbial communities, similar to what has been

223

found for sulfur [39]. Moreover, many sulfate reducing bacteria can also utilize nitrate [38-40]. Other results obtained from characterizing the microbial populations within gas pipeline samples were that methanogens were frequently present in pipeline and biofilm samples, and that sulfate reducers were often present at lower levels than indicated by microbial growth tests. The presence of methanogens in gas pipelines was unexpected, but it is significant because any microbial process, such as methanogenesis, that consumes hydrogen is capable of accelerating corrosion by cathodic depolarization (a process which pulls the cathodic reduction of protons by removal of the product and thereby accelerates anodic metal dissolution) [8, 41, 42]. These results highlight the fact that the composition of microbial communities in gas pipelines has not been thoroughly investigated and prior to the genetic studies described here the entirety of our knowledge about microorganisms in gas pipelines was restricted to information gathered by growing bacteria in various media under laboratory conditions. There are many types of bacteria, and testing for all types of bacteria using growth experiments would be very tedious, and in fact it has never been done. We only have data concerning those types of bacteria that have been tested for, so our view of the microbial ecology of gas pipelines is biased indeed. A commonly held belief regarding microbial corrosion is that SRB are the most important contributors to corrosion. Since traditional microbial growth tests most frequently test for SRB, and very few other types of microorganisms, it is not surprising that this belief is widely held. However, our genetic tests of gas pipeline samples were not biased in looking for any particular type of microorganism, and SRB were not among the most abundant microorganisms in any of the pipeline samples characterized. However, when these same pipeline samples were used to inoculate bacterial growth tests using SRB medium, SRB were invariably found. The differences in the number of particular types of bacteria detected by microbial growth tests and by genetic tests were investigated further, and it was found that the species of SRB that grew in laboratory media inoculated with gas pipeline samples were not the same species of SRB detected when the gas pipeline samples were examined directly. In other words, by growing bacteria in a particular medium an artificial environment is created where nutrient concentrations, and other factors we have yet to fully appreciate, differ significantly from the conditions that are present in gas pipelines. Thus, the composition of microbial communities detected in laboratory growth experiments differs profoundly from the composition of microbial communities actually present in a gas pipeline.

224

4. LACK OF SPECIFICITY OF MICROBIAL GROWTH TESTS The traditional means of quantifying bacterial populations present in gas and oil production samples is to perform microbial growth tests. The growth media used are not highly selective, and they may allow a range of types of bacteria to grow and not just the type of bacteria that is intended to be quantified in a given growth test. However, the accuracy of these microbial growth tests had not previously been tested to quantify what percentage of bacteria was actually the type of bacteria targeted. Accordingly tests were performed using traditional microbial growth media intended for the quantification of specific types of bacteria, and pure cultures of known bacterial types were then tested to determine how selective these growth media were. Typical results are shown in Table 1. Each bacterial culture tested in Table 1 was a pure bacterial culture isolated from a gas pipeline sample. It can be seen from the results shown in Table 1 that bacterial growth media that are intended to support the growth of a particular type of bacteria are not completely selective and other types of bacteria grow. This is particularly problematic when the purpose of the microbial growth test is to determine the quantity of a specific type of bacteria present in a sample. None of the microbial media tested here, which include all those growth media typically used in evaluating MIC in the gas industry, allow for the exclusive growth of the intended type of bacteria. Moreover, only a very small number of pure bacterial species were tested here, and the results obtained using complex mixed cultures typical of gas pipelines could show even more widespread growth. The implication of these results for monitoring MIC is that quantitative results using microbial growth tests can be misleading as growth observed is not always due to the type/species of microorganisms that is supposedly being quantified. Table 1 Growth of pure bacterial cultures in various types of microbial growth media after 6 days incubation at 30 °C Growth medium N B HAB MET IRB SRB APB D

Microbial type Methanogen

Cultures Methanoarcina

Iron-reducing

Shewamlla

Sulfate reducing

Desulfovibrio

Beta-Proteobacteria

Acidovorax

Denitrifier

Commomonas denitrificans +++

+

+

biack

++

+++

.

black

+++

++

+

+++

+ ++

++

+++

++

+

+++

_

++-).

++

+++

.

++

+++

225

5. DEVELOPING AN IMPROVED GENETIC METHOD TO QUANTIFY BACTERIA This section describes the development of a quantitative PCR (q-PCR) method to replace traditional microbial growth tests by providing rapid and more accurate data concerning the quantity of various types of microorganisms present in gas and oil industry pipeline samples. Microbial growth tests are rather slow and provide potentially misleading/inaccurate data. For examples, the previous studies indicated that the numbers of viable SRB in marine sediments can be underestimated at least 1,000-fold when standard most-probable-number (MPN) techniques are used with synthetic growth media [43-45], One of the goals of research at GTI has been the development of a genetic method to quantify bacteria in gas pipeline samples. We used the DNA sequence data for 16S rDNA genes from bacterial and archaeal species found in greatest abundance in actual gas pipeline samples and designed PCR primers and hybridization probes that can be used in quantitative or real-time PCR to quantify bacteria and archaea. Similarly we also developed PCR primers that allow us to quantify SRB, denitrifiers, and methanogens by targeting the dsrAB [40, 46, 47], nirS [48, 49], and mcrA [50, 51] genes respectively. These genes encode enzymes that play crucial roles in the metabolic pathways of sulfate reduction, denitrification, and methanogenesis, respectively, and thus constitute highly specific targets to allow a more precise quantification of these types of bacteria than can be achieved using microbial growth tests. The PCR primers used in this study were designed on the basis of DNA sequences of dsrAB, nirS, and mcrA genes most typically found in gas and oil production operation samples. This was necessary because of the vast diversity of the microbial kingdom. For example, only a small subset of all bacterial species that possesses dsrAB genes are typically present in gas and oil production samples. Therefore, PCR primers designed based on all of the DNA sequences known for dsrAB genes, such as are found in the GenBank database, will have to include many degenerate positions in their sequences, and hence complicate the measurement of effective primer concentrations in the PCR reactions, and reaction itself. Frequently, primers with many degenerate positions failed in q-PCR analysis of gas and oil production operation samples, or yielded results that were not as accurate as when PCR primers specifically designed to quantify this microbial community were used. To investigate the use of these genetic tests to quantify bacteria obtained from gas pipeline samples used to inoculate various types of microbial growth media, quantitative PCR was performed to quantify the number of bacteria, archaea, SRB, denitrifying bacteria, and methanogens present in various samples. The results shown in Table 2 indicate that archaea are generally present at very low levels even in samples cultivated in methanogen growth medium.

226

The fact that quantification of methanogens by quantifying mcrA gene sequences uniformly yields higher values than results obtained quantifying archaea with PCR primers targeting 16S rDNA genes illustrates that currently available PCR primers that are supposed to be "universal" for archaeal species are not sufficiently inclusive and many methanogens, particularly Methanococcus species, are not detected by these archaeal primers [52-54]. Thus more accurate data is obtained by monitoring the mcrA gene that encodes an enzyme that is essential to methanogenesis [50, 51]. The data in Table 2 make it clear that even when the mcrA gene is used to monitor methanogens, microorganisms other than methanogens can grow in methanogen growth media. In several instances there were no detectable methanogens in samples where significant growth in methanogenic growth media occurred, and in no case did the percentage of methanogens exceed 30% of the total microorganisms detected. This confirms the results shown in Table 1 that microbial growth media is not uniquely selective for the type of bacteria of interest, and that quantitative results based on growth in microbial growth media can be misleading. This point is further illustrated by examining the data in Table 2 regarding denitrifying bacteria. When quantitative PCR was performed using primers that target the nirS gene, which encodes an enzyme that is essential to denitrification in the majority of denitrifying bacteria [48, 49], the percentage of denitrifiers growing in denitrifying bacteria media ranges from undetectable to 11 %. These results again indicate that results obtained from microbial growth tests are not specific and that more accurate results can be obtained using genetic tests. Similar results were obtained using primers targeting dsrAB genes to quantify SRB (data not shown). 5.1. Verifying the accuracy of genetic testing methods To further investigate the accuracy and reliability of genetic tests to quantify bacteria and types of bacteria in gas pipeline samples, several tests were performed using pipeline samples that had been spiked with known concentrations of certain types of bacteria. The results of some of these tests are shown in Table 3 where seven different gas pipeline samples were analyzed with and without spiking of a known amount of DNA purified from cultures of bacteria (Pseudomonas aeruginosa PAO-1), archaea (Archaeoglobus fulgidus), SRB (Desulfovibno vulgaris), denitrifying bacteria {Pseudomonas aeruginosa PAO-1), and a methanogen (Methanococcus jannaschii). The first main column in Table 3 is the gene copy number detected by quantitative PCR after spiking with known copies of genes in the PCR reaction, and the second column represents the gene copy number calculated by adding the gene copies spiked to the reaction to the detected copy number from the un-spiked reaction. It is clear from Table 3 that the genetic tests were able to accurately quantify target DNA

227

spiked into gas pipeline samples. This is best seen by inspecting the ratio between the actually detected gene copy number from the spiked PCR reaction and the calculated copy number per reaction (copy number detected from unspiked reaction, plus the copies spiked into the reaction). If all of the DNA added in spiked samples was accurately quantified the ratio of detected and calculated would be 1; values above 1 means an overestimate of the actual concentration, and values below 1 means an underestimate of spiked gene copies. The average ratios of the detected and calculated were 1.02 ± 0.09, 0.69 ± 0.1, 1.2 ± 0.15, 0.87 ± 0.09, and 1.22 ± 0.16 for bacteria, archaea, SRB, denitrifiers, and methanogens, respectively, which are considered very accurate for the analysis of complex environmental samples [24]. The data in Table 3 illustrate that genetic tests employing quantitative PCR techniques provide accurate and reliable data concerning the quantity of various types of bacteria that may be present in gas pipeline samples. 5.2. Applying the genetic testing methods to quantify various groups of microorganisms present in pipeline samples Seven pipeline samples were collected from various natural gas companies at various geographical locations. The biomass was centrifuged down and used for DNA extraction using FastDNA SPIN Kit for Soil (Qbiogene, Carlsbad, CA). The extracted DNA was further purified by phenol/chloroform extraction to remove inhibitory substances commonly present in this type of samples. The spiking experiment showed that the extra purification step was sufficient to eliminate inhibitory effect of PCR amplification (Table 3). The purified genomic DNA was then amplified and quantified using quantitative real-time PCR with DNA standards of corresponding target genes. The results were summarized in Table 4. The results confirmed our previous observation [18] using DGGE analysis of 16S rRNA gene sequences that denitrifying bacteria and methanogens are two types of organisms which were commonly present in relatively high abundance in gas pipeline samples. Denitrifying bacteria were detected in all seven samples, and the concentration in the sample was as high as 7.95 x lO^/ml; methanogens were detected in five out of seven samples, and the concentration was as high as 3.7 x 10^/ml; SRB were detected in six samples, but at lower concentration in most of samples. The lower archaea concentration detected using 16S rRNA gene than methanogens based on the detection of mcrA gene is due to archaea primers, which will exclude many methanogens, especially Methanococcus species [52-54].

Table 2 Real-time PCR quantification of gas pipeline samples grown in various microbial growth media Growth Sample ID Medium 1 3 6 8 9 2 4 5

MET MET MET MET MET MET MET MET

Cone, of various microbes in growth medium (/ml) Methanogen Denitrifier Bacteria Archaea 4.20E+09 6.75E+04 8.15E+O5 2.04E+07 1.61E+05 2.81E+06 1.04E+06 1.00E+10 6.75E+04 5.65E+08 3.51E+1O 1.52E+11 1.02E+11 6.65E+09 4.35E+10 6.92E+07 ND ND 1.22E+07 ND ND 7.24E+06 ND ND

8 9 4 2

DNB DNB DNB DNB

7.84E+09 3.80E+10 1.06E+08 7.28E+05

ND = not detected MET = methanogenic bacteria DNB = denitrifying bacteria

2.35E+05 1.55E+06 ND

4.26E+07 4.38E+09 8.50E+05

ND

ND

Percentage of total microbes (%) Methanogen Denitrifier Archaea 0.00 0.02 0.69 12.02 0.00 0.01 0.30 18.74 4.36 28.52 ND ND ND

0.00 0.00 ND ND

ND ND ND

0.54 11.53 0.80 ND

229

Table 3 Accuracy of real-time PCR quantification of bacteria, archaea, SRB, denitrifiers, and methanogens in natural gas pipeline samples Bacteria detected (copy/rnx) Bacteria calculated (copy/rnx) w/ spiking w/o spiking + spiked copies 1 7.32E+08 7.97E+08 2 1.04E+08 1.16E+08 3 4.56E+06 4.54E+06 1.05E+08 4 1.14E+08 5 6.73E+O5 6.40E+05 8 1.05E+O9 9.76E+08 9 2.35E+09 2.61E+09 Spiking: 5.36E+05 copies of 16S rRNA genes of'/'seudomonas aeruginosa PAO-1 per reaction.

Ratio det:cal 0.92 1.12 1.01 1.08 0.95 0.93 1.11

Archaea detected (copy/rnx) Archaea calculated (copy/rnx) w/ spiking copy# w/o spiking + (3.88E+05) 1 3.07E+05 3.93E+O5 2 8.36E+O5 5.3OE+O5 3 2.35E+05 3.88E+O5 4 3.25E+O5 3.93E+O5 5 3.88E+O5 2.19E+05 2.13E+07 8 1.6E+07 9 2.29E+O6 1.5E+06 Spiking: 3.88E+05 copies of 16S rRNA genes of Archaeoglobus julgidus per reaction.

Ratio det:cal 0.78 0.63 0.61 0.83 0.56 0.75 0.64

SRB detected (copy/rnx) SRB calculated (copy/rnx) w/ spiking copy# w/o spiking + (2.24E+05) 1 2.47E+05 3.O2E+O5 2 2.29E+05 2.85E+05 3 2.24E+O5 3.09E4O5 4 3.18E+05 2.56E4O5 5 2.94E+05 2.24E+05 8 2.37E+06 2.47E+06 9 3.66E+O5 3.49E+05 Spiking: 2.24E+05 copies oidsrAB genes oi Desulfovibrio vulgaris per reaction.

Ratio det:cal 1.22 1.24 1.38 1.24 1.31 0.96 1.05

Denitrifier detected (copy/mx) Denitrifier calculated (copy/mx) w/ spiking copy# w/o spiking + (1.34E+05) 1 2.43E+07 3.19E+07 2 2.78E+05 2.96E+05 3 1.27E+O5 1.5OE+O5 4 4.73E+05 5.20E+05 5 1.38E+05 1.35E+05 8 6.66E4O5 7.91E4O5 9 2.02E+07 2.60E+07 Spiking: 1.34E+05 copies oinirS genes off. aeruginosa PAO-1 per reaction.

Ratio det:cal 0.76 0.94 0.85 0.91 1.02 0.84 0.78

Methanogen detected (copy/rnx) Methanogen calculated (copy/rnx) w/ spiking copy# w/o spiking + (5.10E+05) 1 7.01E+05 5.45E+05 2 1.48E+06 1.16E+06 3 5.63E+05 5.1OE+O5 4 6.37E-H)5 5.22E+05 5 5.1OE+O5 5.88E+05 8 1.12E+08 1.74E+08 9 1.03E-+O7 1.23E+07 Spiking: 5.1E+O5 copies of mcrA genes oiMethanococcus jannaschii per reaction.

Ratio det:cal 1.29 1.27 1.10 1.22 1.15 1.56 1.20

Sample ID

Sample ID

Sample ID

Sample ID

Sample ID

230

Table 4 Quantification of bacteria, archaea, SRB, denitrifiers, and methanogens in natural gas pipeline samples (/mL) Sample ID Bacteria Archaea SRB Denitrifier Methanogen 1 1.59E+03 7.95E+06 1.18E+04 1.99E+08 8.12E+03 2 2.58E+07 1.49E+05 4.05E+04 2.18E+05 1.98E+03 3 1.00E+06 5.93E+01 5.29E+01 4.00E+03 ND 4 2.75E+07 1.75E+03 9.51E+O3 1.02E+05 4.14E+03 5 3.42E+04 1.96E+02 ND 4.67E+01 ND 8 2.62E+08 6.97E+06 6.75E+05 1.64E+05 3.70E+07 9 5.88E+08 6.33E+05 6.48E+06 3.25E+06 4.11E+04

ND = not detected 6. CONCLUSIONS Quantifying various types of bacteria that may be present in gas and oil production operation samples is a difficult challenge. Traditional tests employ microbial growth media of various types that are intended to foster the growth of particular types of microorganisms. Unfortunately, microbial growth media are not uniquely selective and microorganisms other than the intended type often grow in the test medium. In some cases, even though significant growth occurs the concentration of the target population of bacteria is below detection limits. Thus, if the results of growth in microbial test media are used to determine the quantity of various types of bacteria present in gas and oil production operation samples the results obtained can be very misleading. Traditional microbial growth tests require weeks of incubation, are not accurate, and do not provide information about what microbial species were actually present in the environment. An improved means of quantifying various types of bacteria present in gas pipeline samples is using genetic techniques, such as RSGP and qPCR. Other industries such as medicine, food, and cosmetics share with gas and oil industry the need to detect, identify, and quantify microorganisms. These other industries have largely abandoned microbial growth tests in favor of genetic methods. Hybridization test methods such as RSGP are not typically used in characterizing environmental samples, but future improvements in microarray technology may change this situation. q-PCR has been adopted by other industries as the method of choice for rapid quantification of microorganisms, but this technique has not yet been in the gas and oil industry. In this chapter data was presented that demonstrated that q-PCR can obtain data within a few hours that specifically and accurately quantifies bacterial types present in gas pipeline samples. The gas pipeline samples are analyzed directly without any cultivation or other manipulation in the laboratory that would alter the composition of the microbial community. Therefore, q-PCR provides data

231

concerning the actual microbial community present in the gas pipeline. GTI has developed an accurate and reliable method to quantify bacterial types present in gas pipeline samples and is now offering this service to the industry. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [16] [17]

E. Buck, G. C. Maddux, and R. L. Sullivan, GRI-96/0056 document number 96-1466 (1996). S. Farthing, Pipeline and Gas Industry Oct '97 (1997) 43. J. W. Graves, E. H. Sullivan, Materials Protection 5 (1996) 33. V. P. Kholodenko, S. K. Jigletsova, V. A. Chugnov, V. B. Rodin, V. S. Kobelev, S. V. Karpov, Appl. Biochem. Microbiol. 36 (2000) 594. B. G Pound, (1998) Gas Research Institute. G. H. Koch, M. P. H. Brongers, N. G. Thompson, Y. P. Virmani, and J. H. Payer, FHWA-RD-01-156, published online, Federal Highway Administration, Washington D.C. (2001). K. M. E. Emde, D. W. Smith, and R. Facey, Water Res. 26 (1992) 169. P. Angell, Curr. Opin. Biotechnol. 10 (1999) 269. J. F. Batista, R. F. Pereira, J. M. Lopes, M. F. Carvalho, M. J. Feio, M. A Reis,, Biodegradation 11 (2000)441. D. H. Pope, T. P. Zintel, B. A. Cookingham, R. G. Morris, D. Howard, R. A. Day, J. R. Frank, and G E. Pogemiller, Corrosion '89 (1989), NACE paper 192. D H Pope, R. M. Pope, (1998) Gas Research Institute. L. N. Strickland, R T. Fortnum, B. W. DuBose, Corrosion '96 (1996), NACE paper 297. B. L. Maidak, J. R. Cole, T. G Lilburn, C. T. Parker, Jr., P. R Saxman, J. M. Stredwick, G. M. Garrity, B. L. Li, G. J. Olsen, S. Pramanik, T. M. Schmidt, and J. M. Tiedje, Nucleic Acids Res. 28 (2000) 173. G. Muyzer, E. C. de Waal, and A. G Uitterlinden, Appl. Environ. Microbiol. 59 (1993) 695. A. L. Reysenbach, G S. Wickham, and N. R. Pace, Appl. Environ. Microbiol. 60 (1994) 2113. W. A. Williams, J. H. Lobos, and W. E. Cheetham, Int. J. Syst. Bacteriol. 47 (1997) 207. X. Y. Zhu, T. Zhong, Y. Pandya. R. D. Joerger, Appl. Environ. Microbiol. 68 (2002) 124.

[18] X. Y. Zhu, J. Lubeck, J. J. KilbaneH, Appl. Environ. Microbiol. 69 (2003) 5354. [19] X. Y. Zhu, and R. D. Joerger, Poult. Sci. 82 (2003) 1242. [20] E. F. DeLong, L. T. Taylor, T. L. Marsh, C. M. Preston, Appl. Environ. Microbiol. 65 (1999)5554. [21] S. J. Giovannoni, M. S. Rappe, K. L. Vergin, N. L. Adair, Proc. Natl. Acad. Sci. USA 93(1996)7979. [22] S. Y. Lee, J. Bollinger, D. Bezdicek, A. Ogram, Appl. Environ. Microbiol. 62 (1996) 3787. [23] A. Felske, A. D. L. Akkermans, and W. M. de Vos, Appl. Environ. Microbiol. 64 (1998) 4581. [24] J. R. Stults, O. Snoeyenbos-West, B. Methe, D. R. Lovley, D. P. Chandler, Appl. Environ. Microbiol. 67 (2001) 2781.

232

[25] C. F. Brunk, and N. Eis, Appl. Environ. Microbiol. 64 (1998) 5064. [26] G. Voordouw, Y. Shen, C. S. Harrington, A. J. Telang, T. R. Jack, D W. S. Westlake, Appl. Environ. Microbiol. 59 (1993)4101. [27] G. Voordouw, J. K. Voordouw, T. R. Jack, J. Foght, P. M. Fedorak, D. W. S. Westlake, Appl. Environ. Microbiol. 58 (1992) 3542. [28] G. Voordouw, J. K. Voordouw, R. R. Karkhoff-Schweizer, P. M. Fedorak, D. W. S. Westlake, Appl. Environ. Microbiol. 57 (1991) 3070. [29] D. W. S. Westlake, J. M. Foght, P. M. Fedorak, G. Voordouw, T. R. Jack, Corrosion control for low-cost reliability: 12th international corrosion congress 5B (1993) 3794. [30] A. J. Telang, S. Ebert, J. M. Foght, D. W. S. Westlake, G. E. Jenneman, D. Gevertz, G. Voordouw, Appl. Environ. Microbiol. 63 (1997) 1785. [31] E. A. Greene, J G. Kay, K. Jaber, L. G. Stehmeier, G. Voordouw, Appl. Environ. Microbiol. 66 (2000) 5282. [32] Y. Shen, L. G. Stehmeier, and G. Voordouw, Appl. Environ. Microbiol. 64 (1998) 637. [33] G. Voordouw, S. M. Armstrong, M. F. Reimer, B. Fouts, A J. Telang, Y. Shen, D. Gevertz, Appl. Environ. Microbiol. 62 (1996) 1623. [34] C. Hubert, Y. Shen, and G. Voordouw, Appl. Environ. Microbiol. 65 (1999) 3064. [35] V. Torsvik, J. Goksoyr, and F. L. Daae, Appl. Environ. Microbiol. 56 (1990) 782. [36] M. S. Osburne, T. H. Gnossman, P. R. August, 1. A. MacNeil, ASM News 66 (2000) 411. [37] M. Nemati, G. E. Jenneman, and G. Voordouw, Biotechnol. Prog. 17 (2001) 852. [38] B. A. Till, L J. Weathers, and P. J. J. Alvarez, Environ. Sci. Technol. 32 (1998). [39] C. M. Santegoeds, T. G. Ferdelman, G. Muyzer, D. de Beer, Appl. Environ. Microbiol. 64(1998)3731. [40] A. Dhillon, A. Teske, J. Dillon, D. A. Stahl, and M. L. Sogin, Appl. Environ. Microbiol. 69(2003)2765. [41] D. Thierry, and W. Sand, In, Corrosion Mechanisms in Theory and Practice: 2nd edition (P. Marcus ed.) Marcel Dekker, New York (2002) 563. [42] J. Horn, and D. Jones, Manual of environmental microbiology (2002) (Hurst, C. J., Crawford, R. L., Knudsen, G. R, Mclnerney, M. J., Stetzenbach, L. D., Ed), pp. 10721083 ASM Press, Washinton, DC. [43] F Vester, and K. Ingvorsen, Appl. Environ. Microbiol. 64 (1998) 1700. [44] N. B. Ramsing, H. Fossing, T. G. Ferdelman, F. Andersen, B. Thamdrup, Appl. Environ. Microbiol. 62 (1996) 1391. [45] G. R. Gibson, R. J. Parkers, and R. A. Herbert, J. Microbiol. Methods 7 (1987) 201. [46] R. Karkhoff-Schweizer, D. Huber, and G. Voordouw, Appl. Environ. Microbiol. 61 (1995)290. [47] M. Wagner, A. J. Roger, J. L. Flax, G. A. Brusseau, D. A. Stahl, J. Bactenol. 180 (1998) 2975.

[48] G. Braker, A. Fesefeldt, and K.-P. Witzel, Appl. Environ. Microbiol. 64 (1998) 3769. [49] G. Braker, H. L. Ayala-del-Rio, A. H. Devol, A. Fesefeldt, J. M. Tiedje, Appl. Environ. Microbiol. 67(2001)1893. [50] P. E. Luton, J. M. Wayne, R. J. Sharp, P. W. Riley, Microbiology 148 (2002) 3521. [51] T. Lueders, M. W. Fnedrich, Appl. Environ. Microbiol. 69 (2003) 320. [52] A.-L. Reysenbach, K. Longnecker, and J. Kirshtein, Appl. Environ. Microbiol. 66 (2000), 3798-3806. [53] K. Takai, andK. Honkoshi, Appl. Environ. Microbiol. 66 (2000), 5066-5072. [54] M. T. Suzuki, L. T. Taylor, and E. F. DeLong, Appl. Environ. Microbiol. 66 (2000) 4605.


233

Chapter 9

Potential applications of bioemulsifiers in the oil industry H. Bacha and D.L. Gutnick Department of Molecular Microbiology and Biotechnology, Tel-Aviv University, Tel-Aviv, 69978, Israel a

Present address, Biotechnology Research Laboratories, Taro Pharmaceuticals U.S.A., 3 Skyline Drive, Hawthorne, New York, 10532, U.S.A. 1. INTRODUCTION

Virtually hundreds of different chemicals, both synthetic and naturally occurring materials, exhibit properties of specific and preferential interaction at surfaces and interfaces. This generally arises from the presence, in such molecules of both hydrophilic and hydrophobic moieties leading to the ability of these materials to orient themselves at the interface. The effects of such materials in modifying the particular surface and consequently affecting its properties represent its surface activity, and such materials are generally referred to as surfactants. Surfactants vary widely in structure and function and thus are used in virtually all major industries. Surfactants are very versatile and have found uses as detergents lowering interfacial tension, emulsifiers, dispersants, deemulsifiers, wetting agents, foam retardants, stabilizers, gelling agents etc. They are also used widely in consumer products as well major constituents in the household cleaning and soap industry [1], Moreover, the market for chemical surfactants continues grow. For example, the world wide consumption for surfactants was 18 million tons in 1970, 25 million tons in 1990, and about 40 million tons in the year 2,000 [2]. It has been forecasted that the surfactant market in the U.S. and Europe will increase at about 2.5% per year between 2000 and 2005 [3]. These surfactants, generally products of the petrochemical industry are used in almost all modern industrial processes. Chemical surfactants are generally produced as by-products of the petrochemical industry and consist primarily of alkylbenzene sulfonates, alkyl phenol ethoxylates, synthetic fatty alcohols and their derivatives. These products are believed to account for 7075% of the surfactant consumption in the industrialized countries [4].

234

While chemical surfactants are both inexpensive and efficient, they may have very negative effects on the environment. Increasing awareness on the part of the consumer, coupled with the potential for legislation governing excessive use of such chemicals offers new opportunities for biotechnological alternatives. The potential advantages of such products include, 1. biodegradability resulting in lower levels of pollution, 2. selectivity and specificity towards hydrocarbon substrates, 3. potential for using recombinant DNA technology to engineer changes in surfactant structure and function, 4. compatibility with chemical products leading to novel formulations, 5. natural products may have unique characteristics which cannot be produced by simple chemical synthesis. One such product, which is used in the oil industry is xanthan gum, a microbial polymeric exopolysaccharide with unique, sheer thinning rheological properties [5]. In this Chapter we will focus primarily on microbial biosurfactants as bioemulsifiers and their potential applications in the petroleum industry. 2. LOW MOLECULAR WEIGHT BIOSURFACTANTS As is the case with chemical surfactants of non-biological origin, biosurfactants exhibit a wide variety of surface activities ranging from reduction of surface tension, formation of water-in-oil and oil-in-water emulsions, adsorption to and coating of surfaces, surface wetting, flocculation of solids, foaming and defoaming at air/liquid interfaces and emulsion breakage. Moreover, many of these compounds can exhibit more than a single activity. Since there is a very wide range of structures and producing organisms it is often useful to consider the lower molecular mass biosurfactants as a group distinct from the polymeric bioemulsifiers [6-9]. In this section we will refer to the biosurfactants as low molecular mass detergents whose primary activity is to lower interfacial tension, while bioemulsifiers are generally higher molecular mass polymers. Biosurfactants generally are smaller and exhibit molecular masses in the range of a few thousands of Daltons. As detergents, their primary surface activity is to lower interfacial tension at the air/water interface. As is the case with chemical surfactants, biosurfactant formulations may include biosurfactants in combination with chemical or biological co-surfactants, solvents, biocides and other ingredients. Their production can be the result of denovo synthesis [10], or in some cases the result of biotransformations resulting from transacylation or other post synthetic processing [11]. From the point of view of large-scale production, microbial systems have been considered as the major sources of biosurfactants largely owing to the availability of fermentation technology for their production. Moreover, in certain cases, it is possible to manipulate the biosynthetic pathways and their regulation using modern recombinant DNA technology [12]. Since, in recent years a number of original papers and excellent

235

reviews dealing with low molecular weight biosurfactants have been published [7,9,13-16], the subject will be treated only briefly in this section. 2.1. Structure-function features of low molecular weight biosurfactants 2.1.1. Chemical characterization of biosurfactants As illustrated in Table 1A and IB, biosurfactants exhibit a wide variety of structures including fatty acids and alcohols, phospholipids, lipopeptides, which can be either cyclic or linear, and glycolipids consisting of disaccharides linked via esters or ethers to fatty acids. While most of these compounds are extracellular, an interesting exception has been reported, in which the cells themselves exhibit emulsifying activity [13]. Lipopeptides. In the lipopeptides, produced by several species of Bacilli and other organisms, the hydrophilic moiety is represented either by a cyclic oligopeptide, as in surfactin [17] or a 2, 4-diaminobutyric acid intercalated cyclic oligopolypeptide, as in polymyxin [18]. Another class of biodetergents was isolated from a variety of microorganisms belonging to the genera Pseudomonas. They share a common cyclic oligopeptide structure represented by charged amino acids such as serine, aspartic acid and glutamine, while noncharged amino acids as leucine, and isoleucine are present as well [19]. Interestingly, amino acids which are normally not found in proteins, such as Lnorvaline, L-norleucine, L-allylglycine, etc. have been found in some lipopeptide biodetergents strongly suggesting that NRPS's (Non-Ribosomal Protein Synthases) play a role in the generation of these amphipathic peptides [20]. The amino acid number that constitutes the oligopeptide varies among the species, but they share the common lactonized structure formed as a result of an ester bond. For example, serrawetin W2 contains 5 amino acids, surfactin is a heptapeptide; WLIP, viscosinamide, massetolides and viscosin contain 9 amino acids, while tensin, hodersin, pholipeptin, lokisin, arthrofactin and amphisin contain 11 amino acids (Table IB). In all of these compounds, a hydrophobic 3hydroxydecanoyl moiety serves as the lipophilic component [19]. Glyco- and flavolipids. Certain microorganisms produce biodetergents consisting disaccharides as the hydrophilic moieties and fatty acids linked via ester or ether linkages. In the case Pseudomonas aeruginosa the sugar residue is rhamnose to form a rhamnolipid [21], while Arthrobacter parqffineus produces a trehalose lipid [22]. Similarly, the yeast Torulopsis produces a biodetergent consisting of two molecules of glucose. Interestingly, here the acyl moiety is introduced by biotransformation of an exogenously added source of fatty acid [31]. Another fungal glycolipid contains sophorose as the hydrophilic moiety. Recently a new product was discovered during a screen for surfactant producers from soil samples in the Southwestern desert of the U.S. [24]. The new surfactant, produced by a Flavobacteria contains a hydrophilic group consisting

236

of citric acid and two cadaverine molecules. The hydrophobic moiety contains two acyl groups of from 6-10 carbons.

237

238

239

2.1.2. Surface activity of low molecular weight bio surfactants As is the case with chemical surfactants, low molecular weight biosurfactants generally act to lower the surface tension at the air/water interface, and in some cases lower the interfacial tension at the interface between two immiscible liquids (water/oil). The biosurfactant reduction in surface tension at the air/water interface is generally a reduction from 70 down to below 30 mN/m. Of course an additional feature is the critical micelle concentration (CMC), which is the highest effective concentration of the surfactant. Many of the biosurfactants exhibit CMC's in the order of hundreds of milligrams per liter, while significantly lower concentrations are required for the lowering of surface tension. As pointed out previously [7, 9, 13-14], there is not necessarily a correlation between the capacity for surface tension reduction and the capacity for forming and stabilizing emulsions. Part of this may relate to hydrocarbon substrate specificity associated with emulsification, and the affinity of the particular surfactant for the hydrocarbon, water interface. For example, the flavolipids lower surface tension and stabilize hydrocarbon in water emulsions [24], while many of the glycolipids are less effective at emulsification.

240

241

242

243

244

245

246

2.2. Searching for low molecular weight biosurfactants 2.2.1. Screening methods Most lower molecular weight biosurfactants are microbial in nature and are produced primarily by bacteria and fungi [9]. Since the growth of microorganisms on oil is almost always accompanied by the emulsification of the oil in the aqueous media, the most common mode of enrichment for such surfactant-producing microorganisms is to search for organisms capable of growth on either pure or crude petroleum products. In addition, the ability to grow on a variety of renewable waste products has been employed to isolate surfactant producers [21]. The advantage of this approach is that lowers the cost of the growth substrate and consequently the cost of production. Since the growth of hydrocarbon-degrading microorganisms is generally accompanied by the emulsification of the hydrocarbon growth substrate in the aqueous growth media, it was initially assumed that the natural role of the surfactant is to enable the efficient assimilation of the hydrophobic growth substrate by the surfactantproducing microorganism [49]. In some cases this was indeed demonstrated suggesting a potential application of biosurfactants in stimulating hydrocarbon bioremediation [16]. However, many of the surfactant producers such as Acinetobacter radioresistens or Acinetobacter sp. A2 cannot utilize hydrocarbons as growth substrates. This is also the case for many of the cyclic lipopeptides produced by the Bacilli. The results clearly suggest that detergency

247

and/or capacity for emulsification need not be directly related to the natural role of the surface-active molecule. A standard measure for surface activity is the reduction of surface tension. This method may vary somewhat from instrument to instrument and is somewhat cumbersome since frequently measurements must be made at several concentrations. Generally, a good candidate for a surfactant lowers the surface tension from 70mN/m to below 30mN/m. An additional screen based on surface activity was to search for organisms, which produced materials which lyse eukaryotic cells [50-53]. The problem with this approach was that it also enriched for microbial pathogens. Moreover, it was shown to have the weakest correlation with other modes of screening and to yield the highest number of false positives [14]. Another useful and effective screen involves examining the capacity of a drop of culture broth from a putative surfactant producer to collapse an aqueous droplet formed on a hydrophobic surface [54-55]. The surfactant in this case increases the contact angle and droplet collapse can be estimated as a function of surfactant concentration and related to standard materials. The droplet collapse method is rapid and yields relatively low numbers of false positives [14]. The oil spreading technique involves placing a droplet of a surfactant containing solution on a surface coated with a liquid hydrocarbon. The surface activity causes oil spreading leaving a clear zone at the point of application the diameter of which is a qualitative measure of surface activity [56]. Recently the three screening methods, lysis of blood agar, droplet collapse and oil spreading were compared with surface tension measurements for 205 natural isolates. The oil spreading technique appeared to give the highest correlation with the surface tension lowering, although there was strong negative correlation between clear zone diameter and droplet collapse suggesting that the two procedures measured similar activities and could be correlated well with surface tension measurements. It was suggested that an effective protocol for screening natural isolates is to use the droplet collapse method and subsequently employ the oil spreading technique for more quantitative preliminary evaluations [14]. The wide variability in structures (see Table 1A and IB) and the production and secretion of bio surfactants by organisms, which do not grow on hydrocarbons indicates that there is probably no general biological role for all biosurfactants [57]. Moreover, unless the search is for a closely related analog whose synthesis may be catalyzed by similar gene products, it is unlikely that modern methods in molecular ecology will be productive in identifying new organisms or products. Recently, natural isolates from arid soil samples from the Southwestern U.S. were screened for their ability to produce extracellular materials in the culture broth, which lowered surface tension at the air/water interface [12], According to this screen over two percent of the isolates produced surfactants when grown on a rich medium. Interestingly, biosurfactant producers were found in both uncontaminated soils as well as in soils showing

248

elevated levels of hydrocarbons or heavy metal ions. About 45 biosurfactant producing strains were isolated. In one case, a new organism a Flavobacterium was isolated which produced a new biosurfactant with a unique structure (Table 1A). 2.2.2. A role for some glycolipids The idea that bio surfactants increased the surface to volume ratio of the carbon source by physically breaking the oil down into smaller droplets has been proposed as a biological role for the surfactants in oil-degrading microorganisms [7]. However, this might be a strategy for organisms, which grow at the oil/water interface, but it cannot be true for those organisms which produce surfactants but which don't grow on hydrocarbons [12]. Rhamnolipids. Although a general role for bio surfactants in general microbial behavior is unlikely, a biological role for a specific glycolipid, rhamnolipid, has been implicated in biofilm formation by the cystic fibrosis pathogen, Pseudomonas aeruginosa [58]. Mutants defective in rhamnolipid biosurfactant production can adhere to the surface of tissues, but cannot organize themselves into a structured biofilm [59]. The biosynthesis of rhamnolipid proceeds via two gene products, RhlA and Rhl B, the genes for which are present in an operon whose transcription is cell-density dependent. The control of the operon is via a transcriptional regulator termed RhlR whose activity is controlled by an autoinducer, butanoyl homoserine lactone to activate transcription. The cell density dependence is due to a process termed quorum sensing in which the autoinducer is produced in low levels, and a threshold concentration is required to interact with RhlR, thereby converting the protein into a transcriptional activator. In the absence of the autoinducer, the RhlR protein appears to function as a repressor of the operon [60]. Serrawettin W2. Serratia liquifaciens MG1 exhibits a swimming motility on low agar (3xlO5 calcoaceticus MM5 Bacillus sp. IAF343 Bacillus cereus IAF 346 Halomonas euhhalina H28 9xl0 5 Acinetobacter radioresistens K53 Acinetobacter 51400 calcoaceticus A2 Acinetobacter venetianus lxlO6 RAG-1 Candida lypolytica ATCC 27600 8662

Alasan Biodispersan Emulsan Liposan

15

20

[70]

44 44 35

2 2 4 20

[71] [71] [72] [73]

70

30

[74]

70

15

83

17

12

[75]

[76]

Lipids Rhodococcus sp. Q15 Saccharomyces uvarum

14-18 18

[77] [78]

15 10 7

[79] [19] [56\ [34] [80] [80] [38]

Lipopeptides Amphisin Arthrofactin Bacitracin Hodersin Lokisin Lychesin A Serrawettin Surfactin Tensin Viscosin Viscosinamide Glycolipids Pentasaccharide Rhamnolipids Sophorose

Bacillus liqueniformis JF-2 Pseudomonas sp. DSS73 Pseudomonas sp. MIS 38 Bacillus liqueniformis Pseudomonas sp. Pseudomonas sp. DSS41 Bacillus liqueniformis BAS50 Serratia liquefaciens MG1 Bacillus subtilis ATCC 21332 Pseudomonas fluorescens 96.578 Pseudomonas viscosa Pseudomonas fluorescens DR54

1035 1395 1354 1600 1409 1355 1030

10 10 12-17

732 923

8 12

[42] [43]

1410

10

[45]

1126 1126

10 10

[46] [47]

Nocardia corynebacteroides SMI Pseudomonas strains

750

18-20

800

10

Candida bombicola ATCC 22214 Candida bogoriensis

1084

Aerobacter aerogenes 621

1371

[27]

[27] [28]

22

[81]

6

[23]

Other biosurfactants Aerobactin Mannoprotein +

Saccharomyces cerevisiae §

44

17

[82]

in Dalton. Carbohydrates in %. Proteins in %. * Fatty Acids in linear carbon length.

251

The activity of proteins surfactants is due in part to their backbone, which allows the polypeptide to be flexible and disordered in aqueous solution. Accordingly, particular domains present in a protein can rearrange at the interface to form an ordered structure while other domains remain unassociated with the interface. In the latter case, three dimensional loops and tails may be generated [90-93]. The association of specific domains in some ordered arrangement at the interface can be detected in a variety of ways including direct examination using physical and chemical analyses [94]. Many of the proteins, which associate with the interface, appear to be in their native state. For example, even an unstructured protein such as P-casein may not conform to the simple picture of loops and tails occurring in the interfacial area. With globular proteins such as lysozyme or fS-lactoglobin, so much secondary and even tertiary structure is retained that the adsorbed monolayer has been considered as a twodimensional system of "interacting deformable particles" [95], Although the situation in the interfacial region is dynamic, the overall effect is generally considered as an irreversible adsorption. This is because segments of the protein molecules can attach to the interface at many contact points, but it is unlikely that the protein will detach from all these points of contact at the same time or to the same extent [96]. If additional polymer is available this may also move towards the interface and associate with the free loops and tails of the bound material. In addition to proteins, there have been several reports of the emulsification properties of peptides obtained by enzyme hydrolysis [97]. These peptides were obtained either by synthesis [98] or by enzymatic proteolysis (e.g. elastin digested by either pepsin [84] or papain [99]; P-casein digested by chymosin [100]; BSA digested by trypsin [101], and gelatin digested by papain [102]. 3.1.2. Polysaccharides at oil-water interfaces Polysaccharides are widely used in food products to modify texture [103], to control water mobility [104], and to improve moisture retention [63]. The contribution of polysaccharides to the stability of emulsions has traditionally been related primarily to their rheological effects on the continuous phase [64]. Polysaccharides are known to have less surface activity in comparison to proteins [63]. This is due to their relative rigidity and low flexibility and to the repetition of the monomer units in the backbone [105]. Polysaccharides can be either stabilizers or emulsifiers depending on their behavior at liquid-liquid interfaces. Their activity as emulsifiers is evident when they adsorb onto the oilwater interface. Such property is characteristic, in particular, of semi-rigid-chain polysaccharides having an elongated chain conformation, for example, alginates and pectins [106]. These anionic polysaccharides possess surface activity resulting in an increase in the negative charge on the oil droplets giving rise to electrostatic repulsion [107-110]. According to a different explanation, it is

252

assumed that the surface activity of these polysaccharides is mainly caused by a combination of non-hydrated moieties such as certain carboxylates and methoxy carbonyl groups, with hydrated moieties such as a polydioxypyranosyl main chain [109]. Several stabilizing mechanisms have been proposed to explain the functionality of polysaccharides as stabilizing agents. One of them is their hydrophilicity accompanied by their high molecular weight, which leads to formation of a macromolecular barrier in the aqueous medium between dispersed droplets. The gums of Portulaca oleracea [111], and Opuntia ficus [112] are representatives of this group. Another mechanism of stabilization derives from the high intrinsic viscosity of certain polysaccharides in water [64] as in the case of xanthan gum, produced by Xanthomonas campestris [113]. These remarkably viscous polymers actually retard phase separation by enhancing the viscosity of the aqueous phase [114]. 3.1.3. Polysaccharide-protein interactions Interactions between surfactant molecules and synthetic or natural polymers have been studied extensively over the past several decades. These interactions have received considerable attention because of their ability to impart significant changes to the interfacial, rheological, and physicochemical properties of polymer systems, with important implications in various pharmaceutical, biomedical, food processing, and photographic applications [115-118]. In recent years, the study of the interaction between proteins and polysaccharides has grown in importance. The reason for this is that protein-polysaccharide solutions may exhibit special properties, which provide abundant topics for scientific investigation as well as the possibility of practical applications. Different classes of interactions are found between polysaccharides and proteins. Such interactions depend directly on the nature, pH, ionic strength, temperature, concentration and other minor variables, which generate attraction or repulsion between the polysaccharide and the protein in solution. In systems containing protein-coated oil droplets, an attractive proteinpolysaccharide interaction may enhance emulsion stability by forming a thicker and/or strong steric stabilizing layer, or it may act inversely destabilizing the emulsion by forming polymer bridges between flocculated droplets [119-121]. A repulsive protein-polysaccharide interaction may stabilize the emulsion by immobilizing protein-coated droplets in a polysaccharide gel network. This phenomenon is due to the formation of hydrogen bridges and local electrostatic interactions (dipoles) of oppositely charged groups [106]. Incompatibility of proteins with all types of polysaccharides becomes more pronounced with an increase in salt concentration. Presumably, this is mainly due to intensification of protein self-association (aggregating with itself) [122].

253

Protein and polysaccharide molecules might be brought together at an emulsion droplet surface by two distinct pathways: 1) covalent coupling by chemical reaction to form a protein-polysaccharide hybrid molecule [119, 123129] ; or 2) non-covalent association to form a protein-polysaccharide complex [94,96, 105, 121, 130-132]. 1) Covalent coupling: this is the case in which two kinds of biopolymer molecules are inseparably linked together. The main advantage of a covalent protein-polysaccharide hybrid over a non-covalent complex is the retention of molecular integrity and solubility over a wide range of solution conditions [94]. A covalent linkage between protein and polysaccharide is an attractive interaction, which can be specific, strong, and relatively permanent. This interaction might be observed using either natural or artificial hybrids. Natural hybrids are represented by galactomannans [133] and by gum arabic, which contains 2% covalently bound protein [134]. Examples of synthetic covalent hybrids are the interactions between gelatin [135] or whey proteins [105] with propylene glycol alginate under mildly alkaline conditions; ovalbumin-dextran under dry-heat storage [124] or casein-carrageenan in mixed gelling system [105, 136]. 2) Non-covalent coupling: In this case the association may be relatively weak and hence readily reversed. A non-covalent complex may dissociate or precipitate on changing temperature or pH. Weak interactions between proteins and polysaccharides can be repulsive or attractive. Repulsive interactions are always non-specific and are of transient duration. They usually arise from excluded volume effects and/or electrostatic interactions. Net repulsive proteinpolysaccharide interactions are most likely to be found in mixtures of proteins with non-ionic polysaccharides or with anionic polysaccharides at a pH above the protein isoelectric point [137]. This interaction is observed in systems containing mixtures of gelatin-dextran [138] or casein-guar gum [137]. Other non-covalent hybrids between a specific genetic recombinant protein or its peptides and various polysaccharides have been found in the case of the recombinant form of the cell surface esterase from Acinetobacter venetianus RAG-1 [139-140]. Attractive biopolymer interactions may be strong or weak. Strong attractive interactions may occur between positively charged proteins (pHp/) and polysaccharides [141]. Alternatively, weak attractive interactions arise as a result of averaging over a multitude of individual specific chemical interactions between groups on the biopolymers. Examples include 1) ionic interactions in BSA-dextran sulfate [95, 123] or BSA-K-carrageenan [142]; 2) van der Waals interactions as in the case of lysozyme with different polysaccharides [143]; 3) hydrogen bonding as in gelatin complexed with

254

different polysaccharides [131]; and 4) hydrophobic interactions as in poly methacrylic acid [144] or in mixed gels of gelatin with sodium alginate [145]. The protein-polysaccharide interactions may change from net repulsive to net attractive, or vice-versa, on changing the temperature [144] or the solvent conditions (pH, ionic strength) [146]. Therefore, many polysaccharides carrying negatively charged carboxyl or sulfate groups, can form strong electrostatic association, so called "complex coacervation" [147] between the two kinds of polyelectrolytes of opposite net charge at pH values below the isoelectric point of proteins (typically pH~5). Nevertheless, the same polyelectrolytes fail to form strong electrostatic complexes at neutral pH. In this case local electrostatic interactions may still occur between anionic polysaccharide molecules and positively charged patches on proteins [105, 148]. Such interactions can be reinforced by non-ionic attraction (e.g. hydrogen bonding) leading to proteinpolysaccharide complex formation. 3.2. Microbial sources of bioemulsifiers Table 3 A and 3B list a number of bacterial and fungal strains, which have been shown to produce extracellular bioemulsifiers. The emulsification process causes droplet formation of one of the phases and a subsequent increase in emulsion turbidity, which is easily monitored. In addition, the bioemulsifiers also stabilize the emulsions by retarding droplet coalescence. 3.3. The emulsan paradigm Arguably, the most extensively studied polymeric microbial bioemulsifier is emulsan, the amphipathic, polyanionic polysaccharide produced by the oildegrading Acinetobacter venetianus RAG-1 [65, 149-151]. In addition, this product represents an interesting case study encompassing all stages of development, technology transfer and product development. Although many of the features of this model system have been presented in other reviews [6-8, 151-153], newer developments regarding the genetics, biosynthesis and regulation of its production are only now emerging, and will be discussed here. Similarly, the development of novel approaches to generating new amphipathic bioemulsifier formulations [140] has now been developed and will also be presented. 3.3.1. Physical and chemical characteristics Emulsan was initially isolated from the cell-free broth of a stationary phase culture of A. venetianus RAG-1 grown on a light crude oil. Subsequently, the non-dialyzable polymer was produced on a variety of water soluble carbon sources with optimum yields obtained on ethanol as sole source of carbon and energy [65, 75, 154-156]. The bioemulsifier (106 Da) consists of complex

255

between a polyanionic heteropolysaccharide (75%) and a non-covalently associated protein mixture (25%) [75]. As shown in Fig. 1, the linear backbone of the polysaccharide consists of a trisaccharide subunit consisting of a 1:1 ratio of the amino-sugars, Dgalactosamine, D-galactosamine uronic acid and 2,5-dideoxy, 2,5-diamino glucose. The amphipathic properties of the biopolymer arise from the presence of about 20% by weight fatty acids present in the form of amides and esters. The fatty acid composition can be modified by including fatty acids in the growth media [161-164], although the commercial product produced on ethanol minimal media consists of a mixture of 2, and 3 hydroxy dodecanoic acids, palmitic acid, 2-hydroxybutyric acid and acetic acids [7(55]. The apparent pK of the polymer is about 3.05 owing to the galactosamine uronic acid. In addition to the lipoheteropolysaccharide composition of the bioemulsifier, the extracellular emulsan complex also contains between 10 and 25% by weight protein. A portion of the protein mixture can be removed by any of a number of procedures including sepharose gel filtration chromatography, polymer precipitation in the presence of the quaternary ammonium detergent cetyl, trimethyl ammonium bromide [766], hot phenol extraction [75] and proteolysis. As discussed below, the deproteinized polymer, termed apoemulsan [154] retains a portion of the emulsifying activity towards some of the substrates, but is generally inactive in the emulsification of less polar materials such as pure n-alkanes such as hexadecane. The role of this protein in mediating emulsification by apoemulsan will be discussed in a different section below.

Fig. 1. The trisaccharide structure of the emulsan subunit.

256

Table 3A Production and surface activity of bioemulsifiers Organism C-source Compound Exopolysaccharides

Yield*

Acinetobacter calcoaceticus BD4

glucose

0.6

Acinetobacter calcoaceticus MM5

tetradecanc

0.06

ST

Reference [157]

\70]

Bacillus cereus IAF 346

sucrose

0.5-1.2

53

Bacillus sp. IAF343

sucrose

0.5-1.2

28

Halomonas eurihalina H28

crude oil

Alasan

Acinetobacter radioresistens K53

ethanol

2.2

[73]

Biodispersan

Acinetobacter calcoaceticus A2

ethanol

4

[74]

Emulsan

Acinetobacter venetianus RAG-1

ethanol

15

[158\

Liposan

Candida lypolytica ATCC 8662

hexadecane

Rhodococcus sp. Q15

glucose, acetate

36

[77]

Nocardia erythropolis ATCC 4277

29

[159\

Saccharomyces uvarum

hexadecane, kerosene n-dccane

20

[78]

Bacillus liqueniformis JF-2

glucose

25-34 0.35

\7l] [71] [72]

[76]

Lipids

Lipopeptides 30

[160]

glucose

27

[80]

L-broth

24

[56]

27

[34]

glucose

27

[80]

glucose

27

[S01

28

[38]

28

[42]

27

[43]

glucose

27

ISO]

glucose

26.5

[19\

glucose

27

[80]

Cory'nebacterium lepus

kerosene

Amphisin

Pseudomonas sp. DSS73

Arthrofactin

Pseudomonas sp. MIS 38

Bacitracin

Bacillus liqueniformis

Hodersin

Pseudomonas sp.

Lokisin

Pseudomonas sp. DSS41

Lychesin A

Bacillus liqueniformis BAS50

glucose

Scrrawettin

Serratia liquefaciens MG1

glucose

Surfactin

Bacillus subtilis ATCC 21332

glucose

Tcnsin

Pseudomonas fluorescens 96.578

Viscosin

Pseudomonas viscosa

Viscosinamide

Pseudomonas fluorescens DR54

ing/I. Surface Tension in mN/m.

[79]

0.16

3-4

257

The molecular weight of emulsan is about 10 Da [75]. Nevertheless, the polymer exhibits a rather low reduced viscosity in water of about 570 cc/gram [150] and appears to be a long rod of axial ratio about 50:1. The low viscosity of the biopolymer complex is due in part to the presence of the associated-protein. Emulsan is Newtonian in its flow properties, despite the rather high molecular weight, a feature which makes it easier to produce as a fermentation product. 3.3.2. Emulsan activity As an emulsifier emulsan does not dramatically affect either surface tension or interfacial tension between two immiscible phases. Nor does it form micelles, which is characteristic of detergents. Rather its activity appears to depend on its high affinity for the oil/water interface. Being a water-soluble polymer emulsan both forms and stabilizes oil-in-water emulsions resulting from the orientation of the biopolymer at the oil droplet surface [169]. This orientation was first inferred from the binding characteristics of the positively charged rhodamine cation, which bound emulsan in an emulsion, but did not bind to the polyanionic polymer when emulsan was in solution. This binding of cations to emulsan preferentially at the oil/water interface was observed with several metal ions such as C d \ Zn+ , UO2 , Mn , and a host of others [169170]. In many cases the binding was more extensive than anticipated from the charge density of the polymer suggesting that the polymer had assumed a different conformation when oriented at the interface. Under conditions of high energy input, when the formation of small oil droplets is the result of vigorous agitation, the high affinity of emulsan for the interface results in its coating of the droplets preventing coalescence due to charge repulsion of the negatively charged uronic acids. This stabilization does not require the presence of protein, and can be achieved with apoemulsan at correspondingly low ratios of emulsifier to oil between 1:100 to 1:1000 parts of emulsan to oil. The resulting emulsions are very stable, and withstand high-speed centrifugation. In fact, rather than breaking the emulsion into two phases, the centrifugation causes formation of a cream layer which itself is an oil-in-water emulsion consisting of a bulk aqueous phase which constitutes only about 30-50% of the cream, depending on the ratio of emulsan to oil. This cream layer, termed an emulsanosol, is itself an oil-in-water emulsion readily dispersible in the aqueous phase. Emulsanosols or apoemulsanosols can be prepared with a variety of pure and cruder oils, and are stable for years. Their potential applications will be discussed later in the Chapter. In addition to stabilization of oil/water emulsions, emulsan also forms emulsions at lower rates of agitation. In these systems the presence of proteins plays a major role since emulsion formation is quite specific for the hydrocarbon substrates [150], In general, emulsan has been shown to emulsify relatively polar mixtures of hydrocarbons such as those containing both an aliphatic and an

258

aromatic substrate. Hexadecane alone is a very poor substrate, and in the absence of protein, emulsan is completely ineffective with non-polar materials [139]. Interestingly, activity towards hexadecane and other non-polar waxes and sludges can be reconstituted in the presence of a single recombinant protein, the cell-surface esterase of A. venetianus RAG-1. The results in Table 4 illustrate the effect of addition of the cell surface esterase of RAG-1 on the emulsification of a variety of substrates, none of which is emulsified very well by emulsan itself. Interestingly, the addition of the recombinant esterase to emulsan itself had a dramatic effect on the emulsification towards very hydrophobic substrates [139]. Table 3B Production and surface activity of bioemulsifiers Compound Organism Glycolipids Rhodococcus erythropolis

Yield' 32 8 0.1

26

[24]

2.8

26

[27]

1

29

[167]

70

31

[168]

43

[31]

Flavolipid

Flavobacterium sp MTN11

Mihagol L Mihagol S glucose

Pentasaccharide

Nocardia corynebacteroides

nCi4_ 15

DSM 43215

Reference

C-source

ST

[25]

SMI

Rhamnolipids

Pseudomonasputida 21BN

Sophorose

Candida bombicola ATCC 22214 Torulopsis petrophilum Candida bogoriensis

hexadecane, glucose glucose, safflower oil glucose glucose

2

[SI]

Others biosurfactants Aerobactin Mannoprotein

Aerobacter aerogenes 621 Saccharomyces cerevisiae

glucose glucose

1 8"

Synthetic surfactants Cetyl Triethyl Ammonium Bromide (CTAB) Linear alkylbenzene sulfonate Sodium dodecyl sulphate Tween 20

30 47 37 30

Water

72

ing/1. Surface Tension in mN/m. ± in gr/ wet cell gr.

[23] [82]

259

Table 4 Enhancement of apoemulsan activity on different hydrophobic substrates by recombinant esterase Hydrophobic substrate

Emulsifying Activity Ratio

Anthracene Crude oil Dicyclohexane Diesel oil Eicosane Fluoranthene Heptadecane Immersion oil 2-Methyl Naphthalene Mineral oil Octadecane Petroleum refinery sludge Pyrene Soya oil Squalene Tetracosane

966 4.3 8.9 5.7 1800 593 28 10.4 1984 6.6 2250 2 420 1260 600 506

Apoemulsan emulsifying activity in presence of recombinant esterase: emulsifying activity.using emulsan alone.

3.3.3. Emulsan as a microbialproduct Physiology of production. As a fermentation product emulsan is produced primarily on a minimal medium containing ethanol as a sole source of carbon and energy [65]. During early exponential growth the emulsan capsule is present on the cell surface as a minicapsule, which is subsequently released from the cell surface as the cells approach stationary growth. Turnover experiments using radioactive substrate indicated that product release is accompanied by de-novo synthesis [10, 171-172]. Moreover, polymer release is accompanied by a prior release of a protein complex including the cell surface esterase mentioned above. When the cells were starved for carbon in the presence of chloramphenicol the esterase was released but the emulsan remained associated with the cell surface. Addition of fresh nutrients did not give rise to polymer release under these conditions, presumably because the esterase could not be made in the presence of chloramphenicol. The results strongly implicated cell surface esterase in the release process. Interestingly, when emulsan production was studied in a resting cell system in which the cells were immobilized on a celite column, emulsan was shown to accumulate on the surface of the cells in response to increased shear forces [173].

260

Emulsan biosynthesis. In order to study the biosynthesis of emulsan, an insertional plasmid was used to generate mutants-defective emulsan production. Such mutants are visualized either by virtue of their translucent colonial morphology (Fig. 2) or according to their resistance to particular RAG-1 specific phage, ap3, which uses the cell bound emulsan minicapsule as a phage receptor [174]. A mutant blocked in emulsan biosynthesis is thus resistant to phage ap3, but sensitive to a second phage, ap2, which does not require the emulsan receptor. In this regard, the phage ap3 does not bind to the released, water soluble emulsan complex, but does bind to the concentrated emulsanosol, again supporting the idea that the emulsan polymer undergoes a conformational change at the oil/water interface [174]. These data also support the important notion that the conformation of the biopolymer on the cell surface is similar to its conformation at the oil/water interface. Natural role for emulsan. Cells defective in emulsan do not grow on crude oil in liquid culture, although they do grow on these hydrocarbons when they are provided in the vapor phase [174]. Gutnick and Shabtai showed that the cell bound polymer protected cells of RAG-1 from the toxic effects of catalytic surfactants such as cetyl-trimethyl ammonium bromide [166]. In fact, among mutants resistant to the cationic detergent were those, which actually produced more biopolymer than the corresponding wild-type RAG-1. Based on findings that cells lacking emulsan on their surface more avidly to hydrophobic materials [175-177], it is possible that the natural role of the biopolymer is to aid in actually removing the cells from the oil, when utilizable carbon is no longer available to the organism. This would expand the versatility of the organism by allowing it to release itself and search for new more productive surfaces for metabolism. In support of this hypothesis was the finding that emulsan could be used to remove other organisms from hydrophobic surfaces [178]. Another potential role for emulsan relates to its role as a microbial capsule prior to its release from the cell surface. Ophir and Gutnick [179] showed that capsule producing organisms exhibit at least a ten fold higher resistance to desiccation than do isogenic strains lacking the capsule. This resistance depended on the adherence of the capsule to the cell surface and could not be restored by addition of fresh biopolymer from the outside. Emulsan producers were shown to exhibit such resistance to desiccation. Emulsan biosynthetic pathway. The genes encoding the biosynthetic pathway for apoemulsan have recently been localized to a 27 kbp cluster termed the wee regulon [180]. The entire cluster was sequenced and shown to encode 23 putative open reading frames arranged in two divergent operons separated by a non-translated region (Fig. 3). Mutations in any of the genes resulted in a defect in emulsan production. These defects could all be complemented by a wild-type allele. Interestingly, most of the genes for emulsan biosynthesis were homologous to genes discovered in the biosynthesis of most polysaccharides in

261

spite of the fact that A. venetianus RAG-1 is thus far the only natural isolate which has been shown to produce emulsan [180]. According to convention, the specific genes for emulsan biosynthesis were termed wee; the first letter signifies that the gene is a biosynthetic gene from a polysaccharide biosynthetic cluster, the second land third letters identifying the specific product (emulsan, exopolysaccharide). In accordance with convention, some gene products exhibit similar functions in all organisms, and thus are allowed to retain their original names. Figure 3 summarizes a hypothetical biosynthetic pathway for apoemulsan, with, the rightward operon encoding proteins involved in precursor synthesis and activation, aminoglycosyl transferases for assembling the trisaccharide subunit on the inner side of the cytoplasmic membrane, a polymerase, decorating enzymes for the acylation of the aminosugars, a translocase which moves the polymer from the cytoplasmic face of the membrane to the outer or periplasmie face, and enzymes involved in subsequent translocation of the polymer through a specific channel or porin to the outer surface of the cell [180]. Regulation and the production of viscoemulsan. Located in the intercistronic region are two putative d promoters. The leftward operon consists of three repeating frames wza, wzb and wzc, which encode a porin, a protein tyrosine phosphate phosphatase and a protein tyrosine kinase, respectively [180-181]. Knockout mutants in any of these genes resulted in defects in emulsan production. Both Wzc and Wzb proteins of RAG-1 were cloned and over-expressed in E. coli. The Wzc Ptk was shown to be an autophosphyorylase in which a tyrosine (s) in the C-terminal portion of the protein is phosphorylated and subsequently dephosphorylated by the phosphatase [182]. Similarly, the phosphotyrosine of Wzc from RAG-1 was shown to be dephosphorylated by Wzb [181]. According to other reports, the phosphorylated form of Wzc is expected to negatively regulate polymer export through the porin Wza. Elevated levels of extracellular biopolymer production would then be initiated with the activity of Wzb, the phosphatase, which removes the phosphates, permiting the enhanced export. Consistent with the hypothesis was the finding that knockout mutants in the phosphatase were also emulsan deficient. However, the results did not explain why knockouts in Wzc would be emulsan deficient as well. Apparently there is a requirement for the Wzc protein even in its non-phosphorylated state. The Wzc protein contains a series of five tyrosine residues in close proximity to each other at the C terminus. When these tyrosines were deleted, the resulting protein was made but could be phosphorylated and surprisingly, a high molecular mass polysaccharide, termed viscoemulsan, was produced [181]. This product appears to contain the same constituents as emulsan, but is not active as an emulsifier (Nakar, In preparation). The introduction of a wild-type allele of wzc gave rise to the production of a wild-type allele of emulsan suggesting that the protein

262

tyrosine kinase may act to control the size of the exported polymer. It is also of interest that the Wzc protein is required for viscoemulsan production even though it cannot be phosphorylated [181] suggesting that there is an additional role for the protein. A model to describe the role of phosphorylation and dephosphorylation is shown in Fig. 4 [181]. According to this model Wzc, Wzb, Wza proteins and others interact in a multienzyme complex to control the export of the exopolysaccharide. The process is initiated by dephosphorylation of the protein relaxing the control on the porin diameter and enabling larger amounts of polymer to be translocated to the external surface of the cell. Under conditions of rapid growth and high ATP, all of the tyrosine residues are phosphorylated and polymer production is low. In fact, emulsan production does not take place in rich media, although its biosynthesis has been shown to occur. The polymer can be detected immunologically. The manipulation of the export process coupled with the modifications of the biosynthetic genes offers new approaches to the generation of new and novel products and is currently in progress. 3.3.4. Engineering novel derivatives of emulsan The production of new viscous derivatives such as viscoemulsan represents one approach to engineering new biopolymers. Kaplan and coworkers have used nutritional modification to modify fatty acid composition and surface active properties of the resulting derivatives of emulsan [161-164]. Moreover, as described above, the surface activity of apoemulsan-containing formulations can be enhanced by the addition of a particular cell surface enzyme, the cell surface esterase of RAG-1 [139].

Fig. 2. Colonial morphology of parental emulsan-producing RAG-1 and a translucent, emulsan-defective mutant, TR3.

263

Fig. 3. The wee cluster for the biosynthesis of emulsan. The scale of the cluster size is in kilobases. The black arrows represent putative orf sequences. White arrows represent partially sequenced orf s. Putative promoter sites are indicated with thin black arrows. The names of the genes are shown below the corresponding orf s. Orf s labeled solely with capital letters are putative pathway specific genes encoding Wee A-K respectively.

Novel surface-active breakdown products. In order to stabilize water/oil emulsions emulsan must be a polymer [176]. The evidence supporting this comes from the activity of a particular emulsan depolymerase isolated from a bacterial isolate capable of using galactosamine as a sole source of nitrogen. When apoemulsan was incubated with this crude enzyme for different periods of time, it was found that cleavage of less than five percent of the glycosidic bonds were necessary to inactivate the biopolymer. The results support the idea that the interaction of biopolymer at the oil/water interface is a weak, and that the stabilization is brought about by many points of weak interactions at the oil/water interface. Interestingly, subjecting apoemulsan to exhaustive digestion by the depolymerase generated a series of small acylated aminooligosaccharides consisting of between three and six aminosugars [183]. These materials were all found to act as small molecular weight detergents, although they did not effectively stabilize emulsions. They were found to be active towards more hydrophobic substrates such as hexadecane. Emulsification enhancing proteins and peptides (EEPs). As described above, emulsification of oils by emulsan and apoemulsan is strongly enhanced by the addition of a cell surface recombinant esterase from RAG-1 [139]. Potential principles governing this type of interaction have been presented above. The cloning of the enzyme in E. coli was first accomplished by selecting recombinants of E. coli, which could grow on simple triglycerides as sole sources of carbon and energy [184-186]. The clone was required to generate metabolizable substrates such as glycerol and simple fatty acids such as acetate in order to grow. Cloning, sequencing, over-expression and mutagenesis experiments demonstrated that the esterase is a serine protease [187]. However, using a threading program in which primary sequence was related to predicted structures it was found that the protein in fact more resembled an a, P hydrolase enzyme such as acetyl cholinesterase [187]. Interestingly, this was also found

264

for an esterase from another member of the genus Acinetobacter, the strain A. calcoaceticus BD4 [188] and its miniencapsulated derivative BD413. While this enzyme shows strong sequence and structural homology to the RAG-1 enzyme, it did not display any emulsification enhancement when added to apoemulsan [139]. Specificity towards hydrocarbons. As shown in Table 4 the recombinant esterase protein enhances emulsification of apoemulsan towards a variety of pure and crude hydrophobic substrates. EEP activity was observed with mutants of the esterase defective in catalytic activity, suggesting a role for the protein other than as an enzyme. Esterase exhibits EEP activity towards other polysaccharides. Surprisingly, the interaction of the recombinant RAG-1 esterase with the water soluble, rhamnose-containing exopolysaccharide from A. calcoaceticus BD4 led to the formation of a new bioemulsifier complex. In sharp contrast, the esterase from BD4 did not enhance emulsifying activity of apoemulsan towards hydrophobic substrates [140]. Remarkably, the recombinant esterase from RAG1 exhibited EEP activity with over 25 different natural biopolymers, none of which exhibited any emulsifying activity in the absence of the protein. In these cases, the enhancement was not dependent on catalytic activity of the recombinant protein (Bach and Gutnick, in preparation). The results point to a new approach to generation of amphipathic emulsifiers, which is no longer dependent on fermentation to produce the polymer emulsifier. Among the inexpensive materials, which can be converted into bioemulsifiers using this unique formulation with the RAG-1 esterase are cellulose, dextran, starch, xanthan, alginic acid, and a variety of plant and bacterial polysaccharides including the inactive viscoemulsan described above (Table 5). The mode of action of the EEP remains to be elucidated although evidence is discussed below demonstrating that there is a unique motif in the RAG-1 esterase, which is missing from other homologues. Mapping the EEP domain. Initial observations showed that limited proteolysis of the recombinant esterase yielded a fragment of about 10 kDa, which retained the ability to enhance emulsification of hydrophobic substrates such as hexadecane. Accordingly, a series of site directed mutants were generated and over-expressed to produce different fragments of the esterase. Since the fragments were rapidly degraded even in strains of E. coli lacking Clp or Lon proteases, fragments were prepared which were fused in frame to the Cterminus of the maltose binding protein [189]. The various constructs are shown in Fig. 5. The each over-expressed fusion was tested with apoemulsan using the model hydrophobic substrate, hexadecane as a substrate for emulsification. Virtually all the enhancing activity was localized to the C-terminal third of the esterase. It was of interest that the maltose binding protein itself exhibited no EEP activity. Moreover the fusion protein containing the active polypeptide was

265

no less active than the intact enzyme. Removal of the terminal 15 amino acids from the C-terminus completely abolished the EEP activity. Sequence analysis showed that this 15 amino acid C-terminal peptide is unique to the RAG-1 esterase and probably accounts for the unique characteristics of this protein. However, as shown in Table 2, many organisms produce emulsifiers consisting of protein/polysaccharide complexes [7, 9, 190]. In most cases the protein requirement has yet to be clarified and it is possible that there are other proteins or peptides, which exhibit unique EEP activity. Regardless, EEP technology offers a new approach to bioemulsifier production and paves the way for new families of inexpensive, non-toxic, amphiphiles.

Fig. 4. Hypothetical model for the role of protein tyrosine kinase (Wzc) and protein tyrosine phosphatase (Wzb) in emulsan export. 1. Dephosphorylated Wzc allows for polymerization and translocation of emulsan. 2. Phosphorylation of Wzc halts the process, thereby determining the size of the exported polymer. 3. Emulsan release and beginning of a new round of polymerization, translocation and release. Wza-translocation channel; Wzb-protein tyrosine phosphatase; Wzc-Protein tyrsoine kinase; Wzx-polymerase; Wzy-translocase.

266

3.4. Other polymeric bioemulsifiers 3.4.1. Alasan Acinetobacter radioresistans radioresistens KA53 produces a bioemulsifier complex (10' kDa) consisting of three proteins and a polysaccharide [73]. The emulsifying activity was associated primarily with the AlnA protein. Interestingly, the N-terminal sequence of a recombinant form of the AlnA protein produced in E. coli showed strong homology to the outer membrane protein, OmpA [191]. The recombinant form of AlnA was more active as an emulsifier than the complex. It was also shown to solubilize polyaromatic hydrocarbons and at higher concentrations of the substrate, to form hexamers [192]. The crude alasan complex also formed alkane/water emulsions at an optimum pH of 5. This activity was significantly enhanced after heating at 100°C. Interestingly, the alasan producing strain does not grow on hydrocarbons or on oil substrates and the biological role of this complex remains to be elucidated.

Table 5 Enhancement of the emulsifying activity of different polysaccharides by recombinant esterase in the presence of hexadecane _ , , , Polysacchande Agarose Alginic acid Apoemulsan BD-4 exopolysaccharide Carrageenan Cellobiose Cellulose Chitin Colamc acid Dextran Emulsan Ficoll 400 Gum Arabic Pectin Polyvinyl Pyrrolydone Potato starch Pullulan Stewartan Xanthan Xylan

Emulsifying activity ,.., , , , . . (U/mg polysacchande/mg esterase) 963 496 5430 3396 3345 626 766 540 2050 583 6752 263 1895 1830 1950 544 3400 1196 2720 1854

267

Recently another strain of A. radioresistans SI3, which grows on aromatic substrates was analyzed for changes in membrane protein composition in response to changes in the growth substrates [193]. Two-dimensional gel electrophoresis of protein extracts from S13 revealed elevated levels in an Omp A-like alasan ortholog in response to growth on phenol. 3.4.2. Liposan Liposan is a polymeric bioemulsifier produced by the yeast Candida lipolytica ATCC 8662 [76]. The protein polysaccharide complex consists of 83% polysaccharides and 17% protein. When grown on hexadecane organism appeared to colonize the hexadecane droplets. Liposan emulsified alkanes with a chain length between C6 and C18 with the emulsifying activity increasing with increasing chain length. Liposan has also been shown to emulsify various crude oils such olive and corn oils, gas oil, kerosene, paraffin, halowax 1000 and a series of aliphatic and aromatic hydrocarbons. 3.4.3. Biodispersan This polymer is produced by Acinetobacter product exhibited a molecular mass of 51,400. This is a dispersant, which disperses limestone and aids for grinding limestone to form a powder, which is 194].

sp. A2. The extracellular polyanionic polysaccharide lowers the energy required an ingredient in paper [74,

Fig. 4. Generated esterase constructs fused in frame to the C-terminus of the maltose binding protein.

268

3.4.4. Exopolysaccharide-protein complex from Acinetobacter calcoaceticus BD4 A.calcoaceticus BD4 produces a thick rhamnose-containing exopolysaccharide protein complex [157, 195]. The same biopolymer is produced by a strain, which produces similar quantities of extracellular polysaccharide even though it accumulates lower amounts of the capsular polysaccharide on the cell surface. The protein-polysaccharide complex was shown to emulsify mixtures of alkanes and aromatic substrates, but was inactive against pure alkanes. Interestingly, the complex was not particularly active in the presence of crude oil. Of interest, however, was the finding, that unlike the emulsan complex, which retained partial emulsifying activity even after removal of the protein fraction, the BD4 product was completely dependent on both the protein and the polysaccharide fractions. The polysaccharide could be prepared without the protein by physically shearing it from the cell surface, and the activity restored by the addition of the protein fraction [196]. The protein fraction was specific for the BD4 polysaccharide, and was able to enhance the emulsifying activity of apoemulsan (H.Bach, in preparation). However, as was the case with a host of bacterial exopolysacchride [140], the recombinant esterase from RAG-1 was able to reconstitute emulsifying activity to the polysaccharide from A. calcoaceticus BD4. 4. POTENTIAL APPLICATIONS 4.1. General comments The petroleum industry consumes millions of tons of surfactants each year in a large number of applications (see chapters 4 and 19). Surfactants are used in oil field applications (drilling muds, enhanced oil recovery), environmental and equipment clean-up and maintenance, viscosity reduction and oil transportation, emulsion breakage and dewatering of crude oil prior to refining, and more recently, water/oil based fuels. Generally, the surfactant packages include a combination of surface-active agents, and frequently include compatible solvents and specialized chemicals depending on the quality of the specific oils and sludges. As described in this Chapter, biological products and processes can be employed in all of these applications. In this section we will consider only a few larger scale experiments using biosurfactants. The successful trials indicate that at least in terms of product efficacy, these materials have potential. However, their profitability has yet to be unequivocally demonstrated. In this section we will discuss a few large- scale field trials pointing to the possibility of employing biosurfactants and emulsifiers in the oil industry. Unlike biotechnology products for other sectors such as health care or Pharmaceuticals, where the cost of development and even the cost of obtaining

269

approval from the regulatory agencies, are offset by the high prices and profitability of the product, the cost of applications in the oil industry must be kept relatively competitive with products of the chemical industry. This is a particularly difficult constraint considering that production of many of the biotechnological products may involve large-scale fermentation processes, which exert a considerable impact on the cost of the product, particularly if extensive downstream processing is required. Several approaches may be used to enhance the cost effectiveness of biosurfactants. 4.1.1. Searching for the "world beater" Occasionally, the search for natural products yields a compound with unique properties unlike others either natural or developed by the chemical industry. Such products are termed "world beaters" because their properties are unique and unmatched by products of chemical synthesis. Antibiotics represent a classical example of natural materials of major chemotherapeutic importance whose activities are unmatched by chemical synthesis, although they are generally modified by various chemical transformations [ 197]. Another example of a microbial product with unique properties is the exopolysaccharide product of the plant pathogen Xanthomonas campestris, xanthan, is a major polymer whose sheer thinning properties and high reduced viscosity make it a major industrial product in foods as a thickening agent, in drilling muds, and as a gelling agent for use in oil field fracturing programs [5]. Moreover, xanthan viscosity is exploited in oil field flooding during enhanced oil recovery, since the extraordinary high viscosity enables it to actually "push" the released oil out of the well. In most cases, however, the natural biosurfactant exhibits characteristics, which are promising but not necessarily unique. 4.1.2. Cutting the cost of production The cost of production of biosurfactants is frequently a function of the cost of fermentation and subsequent downstream processing. This is particularly true for production of products in which the carbon source must be a hydrocarbon, which, is often the case with low molecular weight glycolipids [9, 16, 53, 198199]. A key approach in this system involves enhancing the product yield by upgrading and optimizing the fermentation [21]. In the latter case the fermentation of rhamnolipids has been upgraded such that lOOg/liter was produced from 160g of soybean oil as a carbon source, a remarkable conversion of substrate to product. In addition, production on various industrial waste products has also lowered the cost of production [200-202]. Assuming that the product biosurfactant is sufficiently active, this presents a way of upgrading the waste material by producing a product of higher added value. This approach may be particularly advantageous in the oil industry, since the crude product

270

need not be purified to any significant extent and may not require expensive downstream processing [203]. Enhanced productivity can in principle be obtained by transferring biosynthetic genes into an organism such E. coli K-12 which is easy to grow and which utilizes a "friendlier" source of carbon and energy [204-205]. At least in pilot scale, most biosurfactants are produced in batch fermentations. Cooper and co-workers developed a semi-continuous approach to producing biosurfactants via self-recycling system [206]. Similarly, emulsan was produced in a similar protocol in which the product was allowed to grow and accumulate in early stationary phase followed by the removal of 90% of the cells and emulsan, which was harvested downstream. The fermentor was filled with fresh media and the culture again entered exponential and early stationary growth, the major portion of the emulsan recovered and the cycles repeated in the same fermentor for several semi-continuous production runs. The cost of production is thus significantly reduced (Cooper, D., Personal communication). Another way to cut the cost of production is to upgrade the producing strain in order to enhance overall productivity [8, 199], This approach has been used in the case of the emulsan producing strain A. venetianus RAG-1 [166]. The positive selection for emulsan overproducers was based on the fact that the emulsan polyanion binds the toxic cation cetyl-trimethylammonium bromide (CTAB). Among the mutants of RAG-1 resistant to CTAB, were those such as strain A. venetianus CTR49, which overproduce the extracellular polyanion and are thus significantly more resistant to the CTAB than the parent. In the laboratory, mutants of this variety produced up to twice as much emulsan per gram of ethanol carbon source than the wild-type. 4.1.3. Upgrading the product Metabolic engineering of new products. Another approach to generating a viable technology employing biosurfactants is to employ physiology, formulation and/or recombinant DNA technology to generate modified products with improved properties. One such approach involved preparing emulsan from RAG-1 cells grown in the presence of various fatty acids [161-164] in order to modify the nature of the acyl groups present in the side chains of the bioemulsifier. We have carried out similar experiments and have found that the emulsan produced is significantly more active towards hydrophobic substrates such as hexadecane alone. Of course, the enhanced efficacy still needs to be weighed against the increased cost of the fermentation due to the inclusion of fatty acids in the media. Formulation packages. Emulsifiers and surfactants are generally incorporated into surfactant packages, which include a mixture of surfactants designed to lower interfacial tension between water and oil phases. Also, in cleaning applications, the formulations may also contain a biopolymer to

271

stabilize the emulsion [207-209] and prevent coalescence of the phases, and they can also contain a compatible solvent. The solvent need not be a water based solvent, but it should be able to dissolve both the low molecular weight surfactants as well as the biopolymer. Materials such as pine oil, liquid terpenoids, dimethyl sulfoxide, and various light crude oils have all been included in various surfactant packages [210]. In addition to solubilizing all of the components into a pumpable mixture, the solvent addition has also been shown to enhance the cleaning of oil contaminated tanks by removing the last remnants of sludge and other flammable materials from the walls of the container rendering the tank not only clean, but also gas-free. The choice of suitable components for various surfactant packages must also take into account other potential components, which must be included. For example, if routine cleaning operations include rinses with anticorrosive materials, the formulation package must be designed on the basis of compatibility with such components. Similarly, emulsion based fuels may need to be formulated together with materials which lower sulfur emissions. Specially designed surfactant formulations may also require compatibility with a variety of materials including flame retardants, biocides etc. 4.1.4. EEP technology In section 3.3.2 the ability of a recombinant cell surface esterase from RAG-1 to enhance the emulsification of apoemulsan towards a variety of hydrophobic substrates was described [139]. The remarkable feature of this system was the finding that the RAG-1 esterase and several of its derivatives were able to interact with a host of polysaccharides to generate a series of amphipathic complexes, which exhibited strong emulsifying activity [140]. This surprising activity, has paved the way for the generation of a whole suite of bioemulsifiers, in which the polymeric component need not be produced as a fermentation product. In fact, it may be possible to upgrade waste materials such as crude celluloses, starch, pectins, etc. to bioemulsifier when combined with a specific peptide derived from the esterase. This peptide can be produced as an over-expressed protein fusion following cloning in a suitable vector [Bach and Gutnick, in preparation], or it can be generated via proteolysis of the esterase itself. The feasibility of employing EEP technology for emulsification in oil industry applications will become clearer once larger scale field trials are conducted and evaluated. 4.2. Bioemulsification, cleaning and sludge recovery 4.2.1. Tank clean ing Oil storage containers accumulate enormous quantities of sludges and bottom sediments. Previous work from this and other laboratories have

272

described bioremediation techniques in which the container is basically turned into a fermentor and microbial biodegradation is used to clean the tanks [211212]. Bioremediation of oily wastes is covered in other chapters in this volume. Another approach is to generate stable oil/water emulsions using a surfactant or biosurfactant package in order to recover the sludge waste material in a homogenous, combustible form. With this approach the costs of the tank cleaning may be partially offset by the added value of the recovered emulsion. Petroferm U.S.A. developed an emulsan-based biosurfactant package designed to clean oil tanks and render them gas-free. Tanks in the order of several to tens of thousands of cubic meters were cleaned using this system after first designing several pumping and mixing devices. The oil-in-water emulsions generated in these large-scale field trials were stable and could withstand high-speed centrifugation to generate emulsanosols [207]. As will be discussed below, these crude emulsanosols containing up to 30% water can be used as an emulsionbased fuel. Another emulsan-based application was developed by Dr. Mary Ann Jones at the research station of the U.S. Navy in Washington. This application is designed to rapidly and efficiently clean the filters normally used in the engine rooms of ships for sludge removal from bilges. The emulsan-based formulation includes, in addition to crude emulsan, light crude oil as a solvent. The system is not designed for oil recovery, but rather for rapid filter cleaning. A rough estimate suggests that the emulsan-based cleaning process could cost as little as $300 per ship cleaning [Jones, Personal communication]. 4.2.1. Emulsion-based fuels Stable oil-in-water or water-in-oil emulsions, if sufficiently homogeneous can be burned for energy, provided that the water content does not exceed about 30%. In fact, this is the water content of a stable emulsanosol prepared on an oil such as a light Texas or Iranian crude oil. A large scale experiment was conducted in which about fifty barrels of an emulsanosol was prepared from a n emulsion consisting of about 70% high vacuum residuals emulsified in 30% by weight water. The combustion of this material was tested at an experimental laboratory in MIT. The combustion was identical to the combustion of a light fuel oil in terms of burn temperature, efficiency and light off. Interestingly, the emulsion of nitrogen gases was slightly lower suggesting that in the presence of water there may be less NOx emission. The results of this large-scale field trial [207-210, 213] have paved the way for a potential application in which sludge from storage tanks can be emulsified, and the emulsion (i.e. the emulsanosol) can be recycled as a source of energy. The system allows for the upgrading for otherwise unusable materials such as vacuum residuals to be recovered as a source of energy offsetting the cost of the tank cleaning. In addition, the fact that the oil phase consists of small droplets suggests that the burn efficiency may be

273

somewhat higher than with a regular hydrocarbon fuel since the surface area is larger. What is even more interesting is that the quality of the burn was indistinguishable from that of a light high quality fuel oil suggesting that emulsion based fuels can be a viable alternative for some applications [207210,213]. Interestingly, these larger scale burn experiments were performed with a biosurfactant containing formulation. However, at Petroferm U.S.A. specialty chemical formulations were developed some of which did not contain the emulsan, but was composed of chemical surfactants, a solvent and other components. Emulsion based fuels have become more and more popular, because they permit the efficient combustion of various hydrocarbons which are normally difficult to burn. Arguably, the best studied system is the waterbitumen emulsion system, termed Oriemulsion which has been commercialized and is currently exported from Venezuela throughout the world. The emulsion is chemical based, and resembles the initial Petroferm formulations. 4.3. Viscosity reduction and oil transportation As discussed above, the stability of emulsan-based oil-in-water emulsions results from the coating of the oil droplets with emulsan in an oriented conformation; hydrophobic moieties coming into contact with the oil surface, and the hydrophilic components oriented towards the aqueous phase. This results in a homogeneous suspension in which the viscosity of the oil component is significantly reduced at room temperature. The extent of the viscosity reduction is a function of the water composition of the bulk phase, which for an emulsanosol can be as high as 30%. Under these conditions the viscosity of a high viscosity oil from the Orinoco basin in Venezuela, (Boscan crude) was reduced from >50,000 Cp to about 85 Cp in the form of an emulsanosol generated with an emulsan based surfactant package [207-210, 213] at a ratio of 1 part surfactant to 500 parts oil. About fifty barrels of this emulsion was transported through an experimental pipeline of 1.25 inches for 96 h during which the mixture was subjected to over 500 pump transits. There was no effect on the low viscosity, although the shear forces on the emulsion might have been expected to produce an inversion from an oil-in-water to a water-in-oil emulsion. Moreover, even after the system was shut down and the emulsion allowed to stand undisturbed for 48 h, there was no breakage or inversion of the emulsion and the low viscosity was maintained through the pipeline for an additional 48 h. The results support the use of surfactant packages to generate oil-in-water emulsions for pipelining highly viscous oils. In fact, this is the basis of the Orimulsion technology.

274

5. CONCLUDING REMARKS There is little doubt that biosurfactants and bioemulsifiers exhibit characteristics and activities, which are applicable in the oil industry. As noted, in some cases product efficacy has been tested in large scale and successful trials have been recorded. However, the impact of such materials on the oil industry is likely to be far less than in other industrial sectors, where the price of the final product is high relative to the costs of production. Ongoing efforts to isolate new biosurfactants, genetically modify existing ones, enhance productivity of the producing organism and otherwise lower the cost of production, and formulate new and more effective biosurfactant packages should yield a host of products and applications in the future. Moreover, for some applications such as enhanced oil recovery or bioremediation, the biotechnology associated with in-situ biosurfactant production accompanying the growth of microorganisms, discussed elsewhere in this book, can be a useful and economically competitive strategy. Finally, the incorporation of biosurfactants in novel formulations suitable for sludge liquification and viscosity reduction leading to economically feasible techniques for waste recovery and recycling.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11 ] [13] [14]

C.A. Houston, (2002) Colin A. Houston & Associates, Inc, NY. J.L. Salanger, Surfactants-Types and Uses, Laboratorio FIRP, Merida, Venezuela, (1999). R. Modler and Y. Ishikawa, Surfactants, Specialty Chemicals Update Program - SRI Consulting: CA. (2001). G. Bognolo, Sustainable Surfactants, Renewable Feedstocks for the 21 st Century, Competitive Industrial Materials from Non-Food Crops, Sand Hutton, York, UK, (1998). I.W. Sutherland, Biotechnol. Adv., 12 (1994) 393. D.L. Gutnick and W. Minas, Biochem. Soc. Trans., 15(1987) Suppl. 22S. E. Rosenberg and E.Z. Ron, Appl. Microbiol. Biotechnol., 52 (1999) 154. Y. Shabtai, O. Pines and D.L. Gutnick, Develop. Indust. Microbiol., 26 (1985) 291. J.D. Desai and I.M. Banat, Microbiol. Molec. Biol. Rev., 61 (1997) 47. C. Rubinovitz, D.L. Gutnick and E. Rosenberg, J. Bacteriol., 152 (1982) 126. H.J.A. Asmer, S. Lang, F. Wagner and V. Wray, J. Am. Oil Chem. Soc, 65 (1988) 1460. [12] A.A. Bodour, K.P. Drees and R.M. Maier, Appl. Environ. Microbiol., 69 (2003) 3280. A. A. Bodour and R.M. Miller-Maier, Biosurfactants: types, screening methods, and applications., Encyclopedia of Environmental Microbiology, Wiley, NY, 2002, pp.75O770. N.H. Youssef, K.E. Duncan, D.P. Nagle, K.N. Savage, R.M. Knapp and M.J. Mcmerney, J. Microbiol. Meth., 56 (2004) 339.

275

[15] A. Fiechter, Trends in Biotechnol., 10 (1992) 208. [16] I.M. Banat, R.S. Makkar and S.S. Cameotra, Appl. Microbiol. Biotechnol., 53 (2000) 495. [17] K. Arima, A. Kakinuma and G. Tamura, Biochem. Biophys. Res. Commun., 31 (1968) 488. [18] T. Suzuki, K. Hayashi, K. Fujikawa, and K. Tsukamoto, J. Biol. Chem., 57 (1965) 226227. [19] D. Sorensen, Cyclic Lipopeptides from Pseudomonas spp., The Marine Chemistry Section, Department of Chemistry, Faculty of Science, University of Copenhagen, Copenhagen, 2002, pp. 1-61. [20] MA. Marahiel, T. Stachelhaus and H.D. Mootz, Chem. Rev., 97 (1997) 2651. [21] S. Lang and D. Wullbrandt, Appl. Microbiol. Biotechnol., 51 (1999) 22. [22] T. Suzuki, K. Tanaka, J. Matsubara and S. Kmoshita, Agric. Biol. Chem., 33 (1969) 1619. [23] F. Gibson and D.I. Magrath, Biochim. Biophys. Acta, (1969) 65. [24] A.A. Bodour, C. Guerrerobarajas, B.V. Jiorle, ME. Malcomson, A.K. Paull, A. Somogyi, L.N. Trinh, R.B. Bates and R.M. Maier, Appl. Environ. Microbiol., 70 (2004) 114. [25] A. Kretschmer, H. Bock and F. Wagner, Appl. Environ. Microbiol., 44 (1982) 864. [26] K. Brandenburg, B. Lindner, A. Schromm, M.H.J. Koch, J. Bauer, A. Merkli, C. Zbaeren, J. Gwynfor Davies and U. Seydel, Eur. J. Biochem., 267 (2000) 3370. [27] J. Kim, M. Powalla, S. Lang, F. Wagner, H. Lunsdorf and V. Wray, J. Biotechnol., 13 (1990)257. [28] S. Itoh and S. Inoue, Appl. Environ. Microbiol., 43 (1982) 1278. [29] P. Rapp and H.E. Gabriel-Jurgens, Microbiology, 149 (2003) 2879. [30] http://www.cyberlipid.org. [31] D.G. Cooper and D A Paddock, Appl. Environ. Microbiol., 46 (1983) 1426. [32] D. Sorensen, T.H. Nielsen, C. Christophersen, J. Sorensen and M. Gajhede, Acta Crys., C57 (2001) 1123. [33] M. Morikawa, H. Daido, T. Takao, S. Murata, Y. Shimonishi and T. Imanaka, J. Bacteriol., 175(1993)6459. [34] J.D. Epperson and L. Ming, Biochemistry, 39 (2000) 4037. [35] M. Moormann, U. Zahringer, H. Moll, R. Kaufinann, R. Schmid and K. Altendorf, J. Biol. Chem., 272 (1997) 10729. [36] M. Doi, S. Fujita, Y. Katsuya, M. Sasaki, T. Taniguchi and H. Hasegawa, Arch. Biochem. Biophys., 395 (2001) 85. [37] J.A. Tnschman, PR. Jensen and W. Fenical, Tetrahedron Lett., 35 (1994) 5571 [38] MM. Yakimov, K.N. Timmis, V. Wray and H.L. Frederickson, Appl. Environ. Microbiol, 61 (1995)1706. [39] M. Fridkin, I. Ofek, H. Tsubery and S. Cohen, http://www.weizmann.ac.il. [40] H. Ui, T. Miyake, H. Iinuma, M. Imoto, H. Naganawa, S. Hatton, M. Hamada, T. Takehuchi, S. Umezawaand K. Umezawa, J. Org. Chem., 62 (1997) 103. [41] I. Kuiper, L. Lagendijk, R. Pickford, J.P. Derrick, EM. Lamers, J.E. Thomas-Oates, JJ. Lugtenberg and G.V. Bloemberg, Molec. Microbiol., 51 (2004) 97. [42] P.W. Lindum, U. Anthoni, C. Christophersen, L. Eberl, S. Molin and M. Givskov, J. Bacteriol., 180(1998)6384. [43] M. Kowall, J. Vater, B. Kluge, T. Stein, P. Franke and D. Ziessow, J. Colloid Interface Sci., 204 (1998)1.

276

[44] A. Segre, R.C. Bachmann, A. Ballio, F. Bossa, I. Grgurina, N.S. Iacobellis, G. Marino, P. Pucci, M. Simmaco and J.Y. Takemoto, FEBS Lett., 255 (1989) 27. [45] T.H. Nielsen, C. Thrane, C. Christophersen, U. Anthoni, and J. Sorensen, J. Appl. Microbiol., 89 (2000) 992. [46] T.R. Burke, M. Knight and B. Chandrasekhar, Tetrahedron Lett, 30 (1989) 519. [47] T.H. Nielsen, C. Christophersen, U. Anthoni, and J. Sorensen, J. Appl. Microbiol., 86 (1999)80. [48] R.J. Mortishire-Smith, J.C. Nutkins, L.C. Packman, C.L. Brodey, P.B. Rainey, K. Johnstone and D.H. Williams, Tetrahedron, 47 (1991) 3645. [49] J.E. Zajic and W. Seffens, CRC Critical Rev. Biotechnol., 1 (1984) 87. [50] P.G. Carnllo, C. Mardaraz, S.J. Pitta-Alvarez and A.M. Giuletti, World J. Microbiol. Biotechnol., 12(1996)82. [51] I.M. Banat, Biotechnol. Lett, 15 (1993) 591. [52] H. Yonebayashi, S. Yoshida, K. Ono and H. Enomoto, Sekiyu Gakkaishi, 43 (2000) 59. [53] C.N. Mulligan, D.G. Cooper and R.J. Neufeld, J. Ferment. Technol., 62 (1984) 311. [54] D.K. Jain, D.L. Collins-Thompson, H. Lee and J.T. Trevors, J. Microbiol. Methods., 13 (1991)271. [55] A.A. Bodour and R.M. Miller-Maier, J. Microbiol. Meth., 32 (1998) 273. [56] M. Morikawa, Y. Hirata and T. Imanaka, Biochim. Biophys. Acta, 1488 (2000) 211. [57] E.Z. Ron and E. Rosenberg, Environ. Microbiol., 3 (2001) 229. [58] C. Fuqua, MR. Parsek and E.P. Greenberg, Ann. Rev. Gen., 35 (2001) 439. [59] D.G. Davies, MR. Parsek, J.P. Pearson, B.H. Iglewski, J.W. Costerton and E.P. Greenberg, Science, 280 (1998) 295. [60] G. Medina, K. Juarez, B. Valderrama and G. Soberon_Chavez, J. Bactenol., 185 (2003) 5976. [61] L. Eberl, S. Molin and M. Givskov, J. Bacteriol., 181 (1999) 1703. [62] E. Dickinson (ed.), Food Emulsions and Foams, The Royal Society of Chemistry, London, 1987, pp. 1-29. [63] G.O. Philips, D.J. Wedlock and P A Williams (eds), Gums and Stabilizers for the Food Industry, IRL Press, Oxford, 1988, pp. 249-263. [64] A. Stephen (ed.), Food Polysaccharides and their Applications, Marcel Dekker, N.Y. 1995, pp. 501-515. [65] A. Reisfeld, E. Rosenberg and D.L. Gutnick, Appl. Microbiol., 24 (1972) 363. [66] A. Oberbremer, R. Miller-Hurting and F. Wagner, Appl. Microbiol. Biotechnol., 32 (1990)485. [67] Y. Zhang and R.M. Miller, Appl. Environ. Microbiol., 58 (1992) 3276. [68] M.F. Van Dyke, P. Couture, M. Brauer, H. Lee and J.T. Trevors, Can. J. Microbiol., 39 (1993)1071. [69] M.A. Providenti, C.A. Flemming, H. Lee and J.T. Trevors, FEMS Microb. Ecol.,17 (1995) 15. [70] M. Marin, A. Pedregosa and F. Laborda, Appl. Microbiol. Biotechnol., 44 (1996) 660. [71] D.G. Cooper and B.G. Goldenberg, Appl. Environ. Microbiol., 53 (1987) 224. [72] C. Calvo, F. Martinez-Checa, F.L. Toledo, J. Porcel and E. Quesada, Appl. Microbiol. Biotechnol., 60 (2002) 347. [73] S. Navon-Venezia, Z. Zosim, A. Gottlieb, R. Legmann, S. Carmeli, E.Z. Ron and E. Rosenberg, Appl. Environ. Microbiol., 61 (1995) 3240. [74] E. Rosenberg, C. Rubinobitz, R. Legmann and E.Z. Ron, Appl. Environ. Microbiol., 54 (1988)323.

277

[75] A. Zuckerberg, A. Diver, Z. Peeri, D.L. Gutnick and E. Rosenberg, Appl. Environ. Microbiol.,37(1979)414. [76] M.C. Cirigliano and G.M. Carman, Appl. Environ. Microbiol., 50 (1985) 846. [77] G. Whyte, S.J. Slagman, F. Pietrantonio, L. Bourbonniere, S.F. Koval, J.R. Lawrence, W.E. Inniss and C.W. Greer, Appl. Environ. Microbiol., 65 (1999) 2961. [78] S.J. Hamza, A.S. Abbas and A.A. Halob, J. Islamic Academy Sci., 7 (1994) 1. [79] S. Lin, M.A. Minton, M.M. Sharma and G. Georgiou, Appl. Environ. Microbiol., 60 (1994)31. [80] T.H. Nielsen, D. Sorensen, C. Tobiasen, J.B. Andersen, C. Christophersen, M. Givskov and J. Sorensen, Appl. Environ. Microbiol., 68 (2002) 3416. [81] A.J. Cutler and R.J. Light, J. Biol. Chem., 254 (1979) 1944. [82] D.R. Cameron, D.G. Cooper and R.J. Neufeld, Appl. Environ. Microbiol., 54 (1988) 1420. [83] M. Shimazu, M. Saito and K. Yamauchi, Agric. Biol. Chem., 49 (1985) 189. [84] M. Hatton and K. Takahashi, Food Hydrocol., 7 (1993) 325. [85] C.W. Cumperand AE. Alexander, Trans. Faraday Soc, 46 (1950) 235. [86] E. Dickinson, B.S. Murray, G. Stamsby and D.M.W. Anderson, Food Hydrocol., 2 (1988)477. [87] M. Shimazu, T. Kamiya and K. Yamauchi, Agric. Biol. Chem., 45 (1981) 2491. [88] IE. Kinsella, Cnt Rev. Food Sci. Nutr, 21 (1984) 197. [89] R.D. Wamska and J.E. Kinsella, J. Agric. Food Chem., 33 (1985) 1143. [90] E. Dickinson and M. Lai, Adv. Molec. Rel. Intern. Processes, 17 (1980) 1. [91] J.A. Defeyter and J. Benjamins, J. Colloid Interf. Sci., 9 (1982) 289. [92] E. Dickinson and S.E. Euston, Molec. Phys., 68 (1989) 407. [93] E. Dickinson and S.E. Euston, J. Chem. Soc. Faraday Trans., 86 (1990) 805. [94] E. Dickinson (ed), Food Polymers, Gels and Colloids, The Royal Society of Chemistry, Cambridge, 1991, pp. 113-122. [95] G.O. Philips, D.J. Wedlock and P. A. Williams (eds.), Gums and Stabilizers for the Food Industry, Oxford University Press, Oxford, 1992, pp. 351-362. [96] N.S. Hettiarachchy and G.R. Ziegler (eds.), Protein, Functionality in Food System, Marcel Dekker, N.Y., 1994, pp. 225-259. [97] K. Watanabe, H. Toyokawa, A. Shimada and S. Arai, J. Food Sci., 46 (1981) 1467. [98] M. Saito, M. Ogasawara, K. Chikuni and M. Shimizu, Biosci. Biotechnol. Biochem., 59 (1995)388. [99] Z. Mozaffar and Z.U. Haque, Food Hydrocol., 5 (1994) 573. [100] S.W. Lee, M. Shimizu, S. Kaminogawa and K. Yamauchi, Agric. Biol. Chem., 51 (1987) 161. [101] M. Saito, K. Chikuni, M. Monma and M. Shimizu, Biosci. Biotechnol. Biochem., 57 (1993)952. [102] M. Watanabe, N. Fujn and S. Arai, Agric. Biol. Chem., 46 (1982) 1587. [103] E. Dickinson and D. Lorient (eds.), Food Macromolecules and Colloids, The Royal Society of Chemistry, Cambridge, 1995 pp. 1-19. [104] G. Williamson, C.B. Faulds, J.A. Matthews, V.J. Morris and GJ. Brownsey, Carbohyd. Polymers, 13(1990)387. [105] J.R. Mitchell and D.A. Ledward (eds), Functional properties of food macromolecules, Elsevier Applied Science, London, 1986, pp. 315-3 54. [106] Y.A. Antonov, N.P. Lashko, Y.K. Glotova, A. Malovikova and O. Markovich, Food Hydrocol., 10(1996)1. [107] V.B. Tolstoguzov and E.E. Braudo, J. Disp. Sci. Technol., 6 (1985) 575.

278

[108] K.D. Schwenke and R. Mothes (eds.), Food Proteins. Structure and Functionality, VCH, Weinheim, 1993, pp. 203-209. [109] G.O. Philips, P. A. Williams and D.J. Wedlock (eds.), Gums and Stabilisers for the Food Industry, IRL Press, Oxford, 1994, pp. 115-124. [110] V. Grinberg and V.B. Tolstoguzov, Food Hydrocol., 11 (1997) 145. [Ill] N. Garti, Y. Slavin and A. Aserin, Food Hydrocol., 13 (1998) 145. [112] N. Garti and M.E. Leser, Polym. Adv. Technol, 12 (2001) 123. [113] W. Liu and T. Sato, Carbohydr. Research, 160(1987)267. [114] J. Greener, B.A. Contestable and M.D. Bale, Macromolecules, 20 (1987) 2490. [115] E. Dickinson (ed.), Food Emulsions and Foams, The Royal Society of Chemistry, London, 1987, pp. 1-29. [116] G.O. Philips, P.A. Williams and D.J. Wedlock (eds.) Gums and Stabilizers for the Food Industry, IRL Press, Oxford, 1990, pp. 157-175. [117] G.L. Hasenhuettle and R.W. Hartel (eds.), Food Emulsifiers and their Applications, Thomson Publishing, 1997. [118] S.E. Hill, A.P. Ledward and J.R. Mitchell (eds.), Functional Properties of Food Macromolecules, Aspen Publisher, Maryland, 1998, pp. 252-277. [119] Y. Cao, E. Dickinson and D.J. Wedlock, Food Hydrocol., 5 (1991) 443. [120] E. Dickinson, V.B. Galazka and D.M.W. Anderson, Carbohyd. Polymers, 14 (1991) 373. [121] E. Dickinson and P. Walstra (eds.) Food colloids and polymers: stability and mechanical properties, The Royal Society of Chemistry, Cambridge, 1993, pp. 3-15. [122] V. Ya, V. Grinberg and V.B. Tolstoguzov, Food Hydrocol., 11 (1997) 145. [123] A. Kato, T. Sato and K. Kobayashi, Agric. Biol. Chem., 53 (1989) 2147. [124] A. Kato, Y. Sasaki, R. Furuta and K. Kobayashi, Agric. Biol. Chem., 54 (1990) 107. [125] S. Nakamura, A. Kato and K. Kobayashi, J. Agric. Food Chem., 40 (1992) 735. [126] M. Hattori, S. Imamura, K. Nagasawa and K. Takahashi, Biosci. Biotech. Biochem., 58 (1994) 174. [127] M. Hattori, S. Imamura, K. Nagasawa and K. Takahashi, J. Agric. Food Chem., 42 (1994)2120. [128] E. Dickinson and V.B. Galazka, Food Hydrocol., 5 (1995) 281. [129] K. Nagasawa, K. Takahashi and M. Hattori, Food Hydrocol., 10 (1996) 63. [130] J.M.V. Blanshard and JR. Mitchell (eds.) Polysaccharides in Food, Butterworths, London, 1979, pp. 205-217. [131] J.R. Mitchell and D.A. Ledward (eds.), Functional Properties of Food Macromolecules, Elsevier Applied Science, London, 1986, pp. 385-415. [132] M. Ahmed and E. Dickinson, Colloid Surface, 47 (1990) 353. [133] N. Garti and D. Reichman, Food Hydrocol., 8 (1994) 155. [134] R.C. Randall, G.O. Philips and P.A. Williams, Food Hydrocol., 2 (1988) 131. [135] J.E. Mckay, G. Stainsby and E.L. Wilson, Carbohyd. Polymers, 5 (1985) 223. [136] K. Nishinari and E. Doi (eds.) Food Hydrocolloid-structures, properties and functions,. Plenum Press, N.Y., 1982, pp. 211-224. [137] Y.A. Antonov, J. Lefevre and J. Doublier, Appl. Polymer Sci., 71 (1999) 471. [138] P. Harris (ed.), Food Gels, Elsevier Applied Science, London, 1990, pp. 291-359. [139] H. Bach, Y. Berdichevsky and D. Gutnick, Appl. Environ. Microbiol., 69 (2003) 2608. [140] D.L. Gutnick and H. Bach. US Patent No. 6 512 014 (2003). [141] S. Ohashi, F. Ura, M. Takeuchi, H. Lida, K. Sakaue, T. Ochi, S. Ukai and K. Hiramatsu, Food Hydrocol., 4 (1990) 323. [142] E. Dickinson and K. Pawlowsky, Food Hydrocol., 12 (1998) 417.

279

[143] K. Maenaka, G. Kawai, K. Watanabe, F. Sunada and I. Kumagai, J. Biol. Chem., 269 (1994)7070. [144] J.W. Goodwin (ED), Colloidal Dispersions, The Royal Society of Chemistry, London, 1982, pp. 99-128. [145] M.A. Muchkin, E.S. Wajnermann and V.B. Tolstoguzov, Nahrung, 20 (1976) 313. [146] R.G. Chilvers and V. J. Morris, Carbohyd. Polymers, 7 (1987) 111. [147] H.R Kruyt (ed.), Colloid Science: reversible systems, Elsevier, Amsterdam, 1949,pp. 232-258. [148] D.A. Imeson, A.P. Ledward and JR. Mitchell, J. Sci. Food Agnc, 28 (1977) 661. [149] E. Rosenberg, A. Zuckerberg, C. Rubinovitz and D.L. Gutnick, Appl. Environ. Microbiol.,37(1979)402. [150] E. Rosenberg, A. Perry, D.T. Gibson and D.L. Gutnick, Appl. Environ. Microbiol., 37 (1979)409. [151] D. Gutnick, Biopolymers, 26 (1987) S223. [152] N. Kosaric (ed), Biosurfactants and Biotechnology, Marcel Dekker, N.Y., 1987, pp. 211-246. [153] K.J. Towner, E. Bergogne-Berezin and C. A. Fewson (eds.) The Biology of Acinetobacter, Plenum Press, N.Y., 1991, pp. 411-441. [154] D.L. Gutnick, E. Rosenberg and Y. Shabtai. US Patent No. 4 234 689 (1980). [155] D.L. Gutnick, E. Rosenberg, I. Belsky and Z Zosim. US Patent No. 4 311 830 (1982). [156] D.L. Gutnick, E. Rosenberg, I. Belsky and Z. Zosim. US Patent No. 4 395 3 54 (1983). [157] N. Kaplan, E. Rosenberg, B. Jann and K. Jann, Eur. J. Biochem., 152 (1985) 453. [158] D.L. Gutnick, Personal Communication, (2004). [159] C.R. Macdonald, D.G. Cooper and J.E. Zajic, Appl. Environ. Microbiol., 41 (1981) 117. [160] D.G. Cooper, J.E. Zajic and D.F. Gerson, Appl. Environ. Microbiol., 37 (1979) 4. [161] A. Gorkovenko, J. Zhang, R.A. Gross, A.L. Allen and D.L. Kaplan, Can. J. Microbiol., 43(1997)384. [162] A. Gorkovenko, J. Zhang, R.A. Gross, D.L. Kaplan and A.L. Allen, Carbohydr. Polym., 39(1999)79. [163] A.K. Johri, W. Blank and D.L. Kaplan, Appl. Microbiol. Biotechnol., 59 (2002) 217. [164] J. Zhang, A. Gorkovenko, R.A. Gross, A.L. Allen and D. Kaplan, Int. J. Biol. Macromol., 20 (1997) 9. [165] I. Belsky, D.L. Gutnick and E. Rosenberg, FEBS Lett, 101 (1979) 175. [166] Y. Shabtai and D.L. Gutnick, Appl. Environ. Microbiol., 49 (1985) 192. [167] B.K. Tuleva, G.R Ivanov and N.E. Chnstova, Naturforsch, 57C (2002) 356. [168] A.M. Davila, R. Marchal and J.P. Vandecasteele, Appl. Microbiol. Biotechnol., 47 (1997)496. [169] Z. Zosim, D.L. Gutnick and E. Rosenberg, Biotechnol. Bioeng., 25 (1983) 1725. [170] D.L. Gutnick and H. Bach, Appl. Microbiol. Biotechnol., 54 (2000) 451. [171] S. Goldman, Y. Shabtai, C. Rubinobitz, E. Rosenberg and D. Gutnick, Appl. Environ. Microbiol., 44(1982)165. [172] Y. Shabtai and D.L. Gutnick, J Bacteriol., 161 (1985)1176. [173] S.D. Wang and D.I.C. Wang, Biotechnol. Bioeng., 36 (1990) 402. [174] O. Pines and D. Gutnick, J. Bacteriol., 157 (1984) 179. [175] O. Pines, Y. Shoham, E. Rosenberg and D.L. Gutnick, Appl. Microbiol. Biotechnol., 28 (1988)93. [176] Y. Shoham, M. Rosenberg and E. Rosenberg, Appl. Environ. Microbiol, 46 (1983) 573. [177] M. Rosenberg and E. Rosenberg, J. Bacteriol., 148 (1981) 51.

280

[178] M. Rosenberg, A. Perry, E.A. Bayer, D.L. Gutnick, E. Rosenberg and I. Ofek, Infect Immun., 33(1981)29. [179] T. Ophir and D. Gutnick, Appl. Environ. Microbiol., 60 (1994) 740. [180] D. Nakarand D.L. Gutnick, Microbiology, 147 (2001) 1937. [181] D. Nakarand D.L. Gutnick, J. Bactenol, 185 (2003) 1001. [182] Mijakovic, S. Poncet, G. Boel, A. Maze, S. Gillet, E. Jamet, P. Decottigmes, C. Grangeasse, P. Doublet, P. LeJVlarechal and J. Deutscher, Embo J., 22 (2003) 4709. [183] Y. Shoham, E. Rosenberg and D.L. Gutnick, US Patent No. 4 704 360 (1983). [184] P.G. Reddy, R. Allon, M. Mevarech, S. Mendelovitz, Y. Sato and D.L. Gutnick, Gene, 76(1989)145. [185] R. Alon, Esterase from the oil-degrading Acinetobacter lwoffii RAG-1: expression of the est gene and protein characterization. Ph.D. Thesis, Tel-Aviv University (1993). [186] R.N. Alon and D.L. Gutnick, FEMS Microbiol. Lett, 112 (1993) 275. [187] R.N. Alon, L. Mirny, J.L. Sussman and D.L. Gutnick, FEBS Lett, 371 (1995) 231. [188] R.G. Kok, J.J. Van Thor, I.M. NugterenJRoodzant, MB. Brouwer, M.R. Egmond, C.B. Nudel, B. Vosman and K.J. Hellingwerf, Mol. Microbiol., 15 (1995) 803. [189] H. Bach, Y. Mazor, S. Shaky, A. Shoham Lev, Y. Berdichevsky, D.L. Gutnick and I. Benhar, J. Mol. Biol., 312 (2001) 79. [190] N. Sar and E. Rosenberg, Curr. Microbiol., 9 (1983) 309. [191] A. Toren, E. Orr, Y. Paitan, E.Z. Ron and E. Rosenberg, J. Bactenol., 184 (2002) 165 [192] A. Toren, E.Z. Ron, R. Bekerman and E. Rosenberg, Appl. Microbiol. Biotechnol, 59 (2002) 580. [193] E. Pessione, M.G. Giuffrida, L. Prunotto, C. Barello, R. Mazzoli, D. Fortunate, A. Conti and C. Giunta, 3 (2003) 1070. [194] E. Rosenberg, C. Rubinobitz, A. Gottlieb, S. Rosenhak and E.Z. Ron, Appl. Environ. Microbiol., 54 (1988) 317. [195] W.H. Taylor and E. Jum, J. Bacteriol., 81 (1961) 688. [196] N. Kaplan, Z. Zosim and E. Rosenberg, Appl. Environ. Microbiol., 53 (1987) 440. [197] O. Abian, C. Mateo, G. FernandezLorente, J.M. Guisan and R. Fernandez_Lafuente, Biotechnol. Progr., 19 (2003) 1639. [198] F. Arendt, M. Hinsenveld, and W.J. van den Brink (eds), Contaminated soil '90, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1990, pp. 491-492. [199] C.N. Mulligan, T.Y.K. Chowand B.F. Gibbs, Appl. Microbiol. Biotechnol, 31 (1989) 486. [200] GL. Ghurye, C. Vipulanandan and R.C. Willson, Biotechnol. Bioeng., 44 (1994) 661. [201] M.E. Mercade, M.A. Manresa, M. Robert, M.J. Espuny, C. De Andres and J. Guinea, Bioresource Technol., 43 (1993) 1. [202] M.E. Mercade and M.A. Manresa, J. Am. Oil Chem. Soc, 71 (1994)61. [203] H.J. Daniel, R.T Otto, M. Binder, M. Reuss, and C. Syldatk, Appl. Microbiol. Biotechnol., 51 (1999)40. [204] U.A. Ochsner, A. Fiechter and J. Reiser, J. Biol. Chem., 269 (1994) 19787. [205] U.A. Ochsner, J. Reiser, A. Fiechter and B. Witholt, Appl. Environ. Microbiol., 61 (1995)3503. [206] W.C. Mccaffrey and D.G. Cooper, J. Ferment. Bioeng., 79 (1995) 146. [207] M.E. Hayes, K.R. Hrebenar, P.L. Murphy, L.E. Futch Jr., J.F. Deal III and P.L. Bolden Jr. US Patent No. 4 684 372 (1987). [208] M.E. Hayes, K.R. Hrebenar, P.L. Murphy, L.E. Futch Jr., J.F. Deal m and P.L. Bolden Jr. US Patent No. 4 793 826 (1988).

281

[209] ME. Hayes, K.R Hrebenar, PL Murphy, L.E. Futch Jr., J.F. Deal in and PL. Bolden Jr.US Patent No. 4 821 757 (1989). [210] ME. Hayes, K.R. Hrebenar, J.L. Minor and L.M. Woodworth, US Patent No. 4 886 519 (1989) [211] D.L. Gutnick and E. Rosenberg, Ann. Rev. Microbiol, 31 (1977) 379. [212] I.M. Banat, N. Samarah, M. Murad, R. Home and S. Benerjee, World J. Microbiol. Biotechnol, 7(1991)80. [213]M.E. Hayes, K.R. Hrebenar, J.F. Deal in and PL. Bolden Jr. US Patent No. 4 666 457 (1987).



283

Chapter 10

Anaerobic hydrocarbon biodegradation and the prospects for microbial enhanced energy production J.M. Suflitaab, LA. Davidovaab, L.M. Giegab, M. Nanny ac and R.C. Princed a

Institute for Energy and the Environment, bDepartment of Botany and Microbiology, cSchool of Civil Engineering and Environmental Science, University of Oklahoma, Norman, OK 73019, USA. d

ExxonMobil Research and Engineering Co., Annandale, NJ 08801, USA.

1. INTRODUCTION Hydrocarbon-based energy underpins the economic, social, and political fabric of the world and demand for oil is expected to grow unabated for the foreseeable future. It is forecast that global oil consumption will increase annually by an average of more than 4 x 107 barrels per day to eventually reach 4.3 x 1010 barrels per year by 2020 [1]. This is projected to be about a 58% increase over current usage by 2025 [2]. The U.S. Geological Survey recently predicted that about 3 trillion barrels of oil remain to be recovered worldwide (with 1 trillion barrels already harvested), half from proven reserves and half from undeveloped or undiscovered sources [3]. However, as proven reserves get exploited and develop into mature fields, secondary and tertiary recovery technologies will increasingly be relied upon to obtain the remaining residual oil. This is particularly true for the U.S. and other oil-importing countries that tend to rely more heavily on mature, domestic energy sources. With demand far surpassing energy production in the U.S., there is heightened interest in diversifying energy sources, tapping unconventional energy supplies and the development of new technology to more fully exploit domestic reserves. Although oil is expected to remain the dominant energy fuel in the next 20 years, the use of natural gas as a substantial energy source has risen significantly in the past 10 years [4]. In fact, natural gas is projected to be the fastest growing primary energy source and an increasingly important alternative to oil [2]. Natural gas, consisting mainly of methane (>95%) but also with small amounts of other short-chain hydrocarbons (C2 to C4), can be harvested from

284

large gas fields, sometimes associated with oil reservoirs, or be obtained from unconventional sources such as shale, coalbeds, or tight sands. The use of natural gas is becoming increasingly popular due to its abundance across the globe. Further, the lower price of natural gas relative to oil makes it an attractive energy source [5]. Natural gas is also a cleaner energy source than oil or coal, and thus can help reduce greenhouse gas emissions [6]. Natural gas combustion produces only about 56% and 71% of the CO2 associated with the equivalent amount of energy produced from coal or oil, respectively [Energy Information Administration (1999). Natural Gas 1998: Issues and Trends (http://www.eia. doe.gov/oil_gas/natural_gas/analysis_publications/natural_gas_1998_issues_and _trends/it98.html). Moreover, methane use results in less NOX, SO2, and particulates per equivalent amount of energy generated, relative to other sources. Despite the increasing use of natural gas and its attendant environmental advantages, world reliance on oil is unlikely to wane in the near future given the existing energy infrastructure and the aforementioned dependence of many societies on this energy form. However, there is a biotechnological link between oil and natural gas that is the product of the relatively recent recognition that many hydrocarbons are susceptible to anaerobic biodegradation and can be converted to methane and carbon dioxide [7-9]. Unlike the well-documented patterns of aerobic oil biodegradation [10], anaerobic hydrocarbon metabolism was essentially dismissed as ecologically insignificant for many years. This view has been completely altered in recent years with the growing appreciation for the metabolism of hydrocarbons coupled with the consumption of electron acceptors other than oxygen. Not surprisingly then, the majority of world oil reserves are believed to be biodegraded to at least some degree, but it was generally accepted that aerobic oxidation processes were largely responsible for such alterations [11, 12]. Recent evaluations of many petroliferous formations have convincingly argued that it is actually anaerobic processes that predominate in oil and gas reservoirs, sometimes leading to the production of biogenic methane [12-14]. Geological evidence has suggested that such methanogenic processes occur very slowly over millennia, and are most important in reservoirs shallower than 4 km and at temperatures of less than 80°C [12, 15, 16]. Microbial decay of oils in deep subsurface reservoirs can clearly reduce oil quality, and a better understanding of the microbial principles behind such decay will be important to help distinguish between degraded, low-value oils and untouched, high-value oils [11]. However, if methanogenesis continues to be identified as an important process in deep reservoirs worldwide, the recovery of methane gas as an alternate form of energy from otherwise unrecoverable or biodegraded sources might have far-reaching economic and environmental implications. The purpose of this chapter is to review evidence for anaerobic hydrocarbon biodegradation and to provide an overview of some of the more

285

generalizing metabolic features. We will explore whether these reactions can be predicted and identify some of the implications for the ability of anaerobes to convert hydrocarbons into methane and and thereby generate useful energy.

2. ANAEROBIC HYDROGEN METABOLISM Crude oils are enormously complex mixtures containing tens of thousands of individual components [17]. Even condensates and refined products include a dizzying array of constituent hydrocarbons. Once released in the environment, the relative concentrations of these chemicals change over time reflecting individual susceptibilities to various fate processes such as sorption, volatilization, dispersion and biodegradation [18]. One way of assessing the susceptibility of such complex mixtures to anaerobic biodegradation is to consider individual chemical classes of hydrocarbons and determine how metabolism varies with structural complexity (see below). Two decades ago, it was generally considered that aliphatic and aromatic hydrocarbons could only be mineralized in the presence of oxygen. Oxygen served as both a respiratory electron acceptor and as a co-substrate for mono- and dioxygenases catalyzing initial hydrocarbon activation steps [19, 20]. We now know that microbial metabolism is much more diverse than this and many classes of hydrocarbons are amenable to microbial attack under a variety of anaerobic conditions. Substrates include BTEX (benzene, toluene, ethylbenzene and xylenes) compounds [21], polycyclic aromatic hydrocarbons [22], saturated and branched alkanes [23, 24] and alicyclic hydrocarbons [25, 26]. Knowledge of the mechanisms used by anaerobes to catalyze such transformations are summarized here. Toluene has historically served as a model substrate to study anaerobic alkylbenzene biodegradation. Initial work with denitrifying strains of Thauera and Azoarcus [27, 28] showed that the first step of toluene degradation occurred by the addition of the aryl methyl carbon to the double bond of fumarate to yield benzylsuccinic acid. This remarkable reaction is catalyzed by a novel glycyl radical-containing enzyme, benzylsuccinate synthase [29, 30]. Another denitrifying strain EbNl, [31], the sulfate-reducing isolates Desulfobacula toluolica and strain PRTOL1 [31, 32], the iron reducer Geobacter metallireducens [33], a defined methanogenic consortium [34] and an anaerobic phototroph [35] have also been shown to carry out this initial toluene transformation reaction. Subsequent transformations of benzylsuccinate lead to the formation of benzoyl-CoA and succinyl-CoA [28, 36, 37]. Benzylsuccinate gets thioesterified to a CoA derivative in a succinyl-CoA-dependent reaction and the resulting product is then oxidized to is-phenylitaconyl-CoA. The latter compound presumably undergoes modified [3-oxidation yielding benzoyl-CoA and regenerates succinyl-CoA. The succinyl-CoA-(i?)-benzylsuccinate CoA-

286

transferase and (7?)-benzylsuccinyl-CoA dehydrogenase responsible for these bioconversion steps have recently been purified and characterized [37, 38]. Benzoyl-CoA, a central metabolite in the anaerobic oxidation of numerous aromatic compounds, is presumably further oxidized to acetyl-CoA and CO2 as described by Harwood et al. [39]. Another addition reaction for anaerobic toluene decay has also been reported. Toluene activation by the denitrifying strain Azoarcus tolulyticus Tol-4 occurs in two consecutive steps with a 2-carbon fragment (presumably acetylCoA) to ultimately form benzylsuccinate [40]. Evidence supporting this mechanism includes the detection of cinnamic acid in cultures receiving toluene and the production of radiolabeled benzylsuccinic acid from incubations amended with trans-cvcmsmic acid and 14C-acetate. Benzylsuccinate formation is believed to be preceded by the formation of hydrocinnamoyl-CoA and cinnamoyl-CoA intermediates. Studies on this toluene pathway are rare so speculation on how common it might be is difficult. Biodegradation of m-xylene and o-xylene has been observed in enrichments under sulfate-reducing [41-43], nitrate-reducing [44] and methanogenic conditions [45]. The anaerobic biodegradation of p-xylene has only been demonstrated under sulfate- [41, 43] and nitrate-reducing conditions [46]. A handful of nitrate- and sulfate-reducing bacteria capable of m-xylene and o-xylene biodegradation have been isolated [47-51]. Under denitrifying conditions, m- and o-xylene are activated by fumarate addition reactions by Azoarcus sp. strain T [28, 52]. Tentative identification of £-(3-methylphenyl)itaconyl-CoA and 3-methylbenzoate from /w-xylene suggested a pathway analogous to that of anaerobic toluene oxidation [52]. Biochemical, molecular, and genetic studies of benzylsuccinate synthase from Azoarcus sp. strain T proved that this enzyme catalyzed the initial reaction in anaerobic biodegradation of both toluene and m-xylene [52, 53]. Further, partially purified benzylsuccinate synthase transformed all three xylene isomers to their methylbenzylsuccinate analogs [54]. Under sulfate-reducing conditions, wholecell suspensions of the toluene-grown sulfidogenic culture PRTOLl transformed o-xylene to (2-methylbenzyl)succinate. However, the cell could not grow on oxylene suggesting the involvement of benzylsuccinate synthase [55] in the transformation of the parent molecule. In sulfate-reducing enrichments, 2-, 3-, and 4-methybenzylsuccinates were positively identified as metabolites of 0-, m-, and p-xylene, respectively [43] indicating that parent molecules were anaerobically attacked in a comparable fashion when sulfate served as a terminal electron acceptor. There are multiple pathways for ethylbenzene metabolism under anaerobic conditions. Three pure denitrifying cultures were isolated for their ability to completely mineralize ethylbenzene [49, 56]. In these organisms, ethylbenzene is initially activated via dehydrogenation to yield 1-phenylethanol

287

[31,49] by a remarkable molybdenum enzyme that shows clear sequence similarities to the aerobic dimethyl sulfoxide reductase family of molybdopterincontaining enzymes [57]. In this reaction the oxygen atom in the hydroxyl group originates from water [56]. 1-Phenylethanol is then further transformed to acetophenone, and ultimately to benzoyl-CoA [36]. Under sulfate reducing conditions, ethylbenzene was converted to the corresponding ethylbenzylsuccinic acid (3-phenyl-l,2-butane-dicarboxylic acid) by a putative fumarate addition reaction [43]. Recent studies with a pure sulfate-reducing bacterium confirmed this mode of ethylbenzene metabolism [58], with the fumarate addition occurring at the methylene carbon of the side chain rather than at the terminal methyl group. Benzene is the least reactive of all aromatic hydrocarbons and was believed to be entirely recalcitrant under anaerobic conditions. This too has proven to be incorrect. Benzene can be degraded under nitrate- [59- 61], sulfate[62-65], and Fe(III)-reducing conditions [66-68] and with coupling to methane production [69-71]. To date only two pure benzene-degrading strains, both affiliated with the Dechloromonas genus, have been described [60]. Both strains can oxidize benzene with nitrate as an electron acceptor. Physiological studies and 13C-labeling data suggested benzene activation by an initial methylation reaction to form toluene [72]. However, alkylation reactions are not entirely consistent with previous information on anaerobic benzene decay. For example, 13 C-phenol and 13C-benzoate have been detected as intermediates in an enrichment culture incubated with 13C-benzene under sulfate-reducing conditions. 13C-Benzoate was also found in comparable methanogenic and Fe(III)-reducing enrichments [73], suggesting that the hydroxylation of benzene to phenol is one of the initial steps in anaerobic benzene decay. The conversion of phenol to benzoate could then occur by the carboxylation of phenol to form/>hydroxybenzoate followed by the reductive removal of the hydroxyl group to form benzoate. 13Carbon and deuterium labeling studies confirmed that the carboxyl carbon of the benzoate intermediate is derived from one of the carbon atoms of benzene [73, 74]. Therefore, it seems plausible that multiple mechanisms for anaerobic benzene decay also exist amongst the anaerobes. The anaerobic biodegradation of w-alkanes has also been demonstrated recently. Several sulfate-reducing and denitrifying bacterial strains [42, 75-78] as well enrichment cultures [44, 79] are capable of the complete conversion of «-alkanes to carbon dioxide. A broad range of «-alkanes may be susceptible to anaerobic biodegradation, including long-chain alkanes ranging from Ci5 to C34 [80]. All of the sulfate-reducing alkane-degrading strains isolated to date are short oval-shaped rods belonging to -subclass of the Proteobacteria. None of them has been fully characterized and the complete metabolic pathway for anaerobic alkane decay remains speculative. Recently, a product resulting from the initial metabolic activation of a model alkane was identified. When

288

deuterated dodecane was added to an alkane-degrading sulfate-reducing culture, alkylsuccinic acid derivatives with complete deuterium retention were formed, demonstrating that the primary attack on the parent substrate occurred by addition to the double bond of fumarate [79]. The same mechanism was subsequently demonstrated for a denitrifying Azoarcus-like strain [81]. This bioconversion represents a remarkable reaction that superficially resembles the anaerobic biodegradation of toluene. However, alkanes are not activated at a terminal methyl position like toluene or the xylene isomers. Rather the succinyl moiety is attached subterminally at the C2 (and less frequently the C3) position of alkanes [79, 81, 82]. The enzymology of this reaction is still under investigation, though EPR spectroscopy suggests a radical mechanism comparable to that for toluene decay [81]. Another reported alkane activation mechanism involves direct carboxylation with inorganic bicarbonate. In studies with the sulfate-reducer strain Hxd3, So et al. [83] showed that 13C-bicarbonate was added to alkanes at the C-3 position, followed by the elimination of the two adjacent terminal carbon atoms yielding a fatty acid one carbon shorter than the parent alkane. As of this writing, the fumarate addition mechanism seems to be more widespread for «-alkane activation as it has now been shown for three disparate anaerobic cultures [79, 81, 84, 85]. The metabolic steps following the formation of alkylsuccinates are not yet completely clear, but an important contribution has recently been published [86]. Studies performed with the denitrifying strain HxNl grown with deuterated «-hexane or deuterated fumarate revealed the formation of a suite of fatty acids [86]. The identification of 4methyloctanoic acid with deuterium in the C-3 position suggested that the product of fumarate addition, hexylsuccinate (or [l-methylpentyl]succinate), could possibly undergo rearrangement of the carbon skeleton prior to further oxidation. Based on the identification of other transient metabolites, such as 4methyloct-2-enoic and 3-hydroxy-4-methyloctanoic acids, a hypothetical pathway was proposed which allowed for the regeneration of fumarate. Recent studies with 13C-hexane transformation in a sulfate-reducing culture provided evidence supporting the proposed model [82]. Mass spectral and nuclear magnetic resonance (NMR) data indicate that multiple 13C nuclei originating from l-13C-hexane become incorporated into a variety of metabolites. The labeling patterns argue that 13C-fumarate is produced and recycled during hexane biodegradation. 4-Methyloctanoic acid, an important metabolite of the proposed pathway, was positively identified in the organic extracts of 12C- and 13 C-hexane-amended culture supernatants by both mass spectral analysis and 13 C-NMR. Similarly, 3-hydroxy-4-methyloctanoic acid, was also tentatively identified [82]. Common metabolites formed during anaerobic alkane utilization, as well as the presence of multiple 13C-carbons in the metabolites, strongly suggest that this sulfate-reducing culture and a denitrifying strain, HxNl, not

289

only initiate alkane degradation via fumarate addition, but most probably share the entire degradation pathway. Polycyclic aromatic hydrocarbons (PAHs) are also susceptible to anaerobic decay. Naphthalene can be completely mineralized by pure cultures of sulfate-reducing and denitrifying bacteria [87, 88]. Enrichments from coal-tar contaminated sediments and garden soil were reported to mineralize [14C]naphthalene with soluble Fe(III) and insoluble FeOOH, although not more than 15% of added radioactive substrate was recovered as 14CO2 [89]. Anaerobic degradation of phenanthrene was also demonstrated in sediments [22, 90] and by a sulfate-reducing enrichment culture [91]. Studies with marine sediments also indicated the loss of 2 to 5-ringed PAHs under anaerobic conditions, with the smaller PAHs degrading more rapidly than the heavier molecular weight counterparts [90]. It has been shown that unsubstituted PAHs, such as naphthalene and phenanthrene, are initially attacked by carboxylation to form 2naphthoic acid and phenanthrenecarboxylic acid, respectively [91, 92]. The carbon in both cases arises from inorganic CO2. 2-Methylnaphthalene is converted to 2-naphthoic acid following the anaerobic oxidation of the methyl group [93]. A mechanism for the activation of 2-methylnaphthalene is the addition of fumarate to the methyl group [92, 94]. The product of this reaction, naphthyl-2-methyl-succinic acid, is subsequently oxidized to 2-naphthoic acid which further decomposes by ring reduction reactions to form the fully saturated decalin-2-carboxylic acid prior to ring cleavage and ultimate mineralization [91, 92, 95]. Alicyclic hydrocarbons can comprise a substantial fraction (often up to ~12% wt/wt) of the organic molecules in petroleum mixtures. Despite this quantitative importance, little is known about the metabolic fate of this class of materials. Recently, a study of the anaerobic metabolism of a model alicyclic hydrocarbon, ethylcyclopentane, revealed that it too was initially activated by fumarate addition to form ethylcyclopentylsuccinic acid [25]. Wilkes et al. [84] recently observed that when the denitrifying strain HxNl was incubated with crude oil, a series of C4 to C8 «-alkanes as well as cyclic alkanes, were activated to their corresponding alkylsuccinates and methyl-branched fatty acids. Further, cyclopentane, cyclohexane, and their methyl-and ethyl substituted congeners were rapidly consumed in live incubations of sulfate-amended anoxic sediment enrichments from a gas condensate-contaminated aquifer [26]. Though alicyclic biodegradation was more extensive under sulfate-reducing conditions, there was biodegradation of simpler alicyclic compounds under methanogenic conditions. In parallel methanogenic incubations, 90% of cyclopentene and methylcyclopentene was lost in 100 days [26]. Thus, this class of materials is also susceptible to methanogenic biodegradation.

290

3. CHEMICAL THEORY AND THE SUSCEPTIBILITY OF HYDROCARBONS TO FUMARATE ADDITION REACTIONS The section above attests to the diversity of anaerobic hydrocarbon biodegradation reactions. Given the complexity of even simple hydrocarbon mixtures, it is clearly impractical, if not impossible, to assay all component molecules for their susceptibility to anaerobic biodegradation. While relatively little is known of the enzymatic processes and reaction mechanisms responsible for such bioconversions, insight can be gleaned from a consideration of the molecular structure of the reported metabolites. A common theme among many classes of hydrocarbons is the importance of fumarate addition reactions that lead to metabolites containing a succinic acid functional group. The site of fumarate addition to the hydrocarbon is likely a key determinant in the susceptibility of the parent molecules to anaerobic destruction and to accurate predictions of the metabolites that might reasonably be anticipated. Thus, accurate prediction of the fumarate addition site based upon chemical principles is essential. In a very general sense, fumarate addition to a hydrocarbon substrate can be thought of as a four-step process. The first step involves complexation of the hydrocarbon substrate and enzyme, and is controlled by factors such as diffusion of the hydrocarbon substrate to the enzyme and the thermodynamics of the enzyme-substrate complex. The second step involves oxidation of the hydrocarbon substrate, hypothesized by us to occur via a hydrogen atom transfer (HAT). This is a one-electron oxidation of the substrate through abstraction of a hydrogen radical (H) transforming the hydrocarbon substrate into a free radical intermediate [96]. In the case of alkyl aromatic compounds, a second oxidation mechanism is also possible, one that occurs through electron transfer (ET) [96]. ET involves the removal of an electron from the -orbitals of the aromatic ring, resulting in an aromatic radical cation. The aromatic radical cation then looses a proton (H+) to the surrounding matrix and is transformed into the hydrocarbon radical intermediate. It is hypothesized that the position of the radical in the hydrocarbon is controlled by the stability of the hydrocarbon radical intermediate, and therefore the site of fumarate addition can be readily predicted based upon chemical rules governing free radical stability. The third step involves addition of fumarate to the hydrocarbon radical, leading to the next step which is the release of the newly formed metabolite from the enzyme. The ensuing discussion focuses on the hypothesis that the structure of succinic acid metabolites suggests that anaerobic fumarate addition reactions proceed via a radical mechanism; either by a hydrogen abstraction transfer or an electron transfer mechanism. As a result, metabolite structure can be predicted for alkane, alicyclic and alkyl aromatic hydrocarbons based upon the chemical rules that govern free radical stability.

291

As noted, fumarate addition to «-alkanes predominantly occurs at the subterminal C2 position and to a lesser degree at the C3 position. Fumarate addition to these positions, rather than the less sterically-hindered terminal methyl position, strongly suggests a HAT mechanism. In the HAT mechanism, the site of homolytic cleavage {reaction 1) is a function of the C-H bond dissociation energy. RH->R+H

(1)

The bond dissociation energy (AH2g8) is a function of the stability of the free radical intermediate (R*). For alkane compounds, the relative stability of radical intermediates is of the order: tertiary > secondary > primary [97]. Thus, the ease of abstracting a hydrogen radical from carbon atoms in an alkane follows the same relative trend as the bond dissociation energy as seen for alkanes in Table 1. The relative stability pattern in alkanes results from hyperconjugation, that is, delocalization involving a bonds. The greater the number of hyperconjugative forms that can be generated for a free radical intermediate, the greater the stability of that intermediate [97]. The bond dissociation energy data in Table 1 presents a 1 to 3 kcal mol"1 difference between n-alkane methyl and methylene groups. More importantly, however, is the fact that a 1 kcal mol"1 difference exists between the terminal methyl group and the subterminal C2 carbon for both pentane and hexane, thus illustrating the favorable reactivity of the subterminal C2 carbon relative the terminal methyl group. Observation of fumarate addition to the C3 carbon is not surprising since the difference between the bond dissociation energies of the C3 and C2 methylene carbons is expected to be minimal, at least much less than 1 kcal mol"1. The site of fumarate addition to alicyclic compounds should follow the HAT mechanism similar to w-alkanes, although a decrease in ring strain due to the loss of a hydrogen atom from the alicyclic ring will slightly lower the bond dissociation energies relative to the w-alkane analog. This decrease in bond dissociation energy is observed for cyclopentane which is 3.6 kcal mol"1 less than that of the C2 carbon in «-pentane. Alkylation of cyclopentane to form methyl- and ethylpentane produces a tertiary carbon in the ring at the site of alkyl attachment. As predicted by the HAT mechanism, the tertiary carbon is more stable as a free radical than the secondary ring carbons, displaying bond dissociation energies of 93.7 kcal mol"1. Thus, based upon the bond dissociation energies, and assumption of a HAT mechanism, it is predicted that similar carbons, i.e., secondary and tertiary carbons, will be more reactive in alicyclic alkanes as compared to the corresponding «-alkane. Thus, the most favorable site for fumarate addition to an alkylated alicyclic compound will be at the tertiary carbon followed by secondary carbons on the alicyclic ring. Based upon

292

the relatively higher bond dissociation energies, fumarate addition to the alkyl side chain of an alkylated alicyclic hydrocarbon is therefore unexpected. Table 1 Bond Dissociation Energies (AH298) at 298 K for various hydrocarbons in the reaction RH -> R* + H* Bolded hydrogen atom represents abstracted hydrogen. zl//2»«(kcal mol"1)

reference

104.99 +/- 0.03 101.1+/-0.4 98.6 +/- 0.4 98.2 +/- 0.5 96.5 +/- 0.4 100.2 99.2 99.0 98.0

[98] [99] [99] [99] [99] [100] [100] [100] [100]

95.6 +/- 1 93.7 93.7

[101] [102] [102]

112.9+/-0.5 89.8 +/- 0.6 85.4+/- 1.5 87.5

[103] [104] [105] [106]

86.7 83.5 98.7

[107] [107] [107]

112.2+/-1.3 111.9+/-1.4

[108] [108]

85.1 +/- 1.5 85.6

[105] [107]

Alkanes CH3-H (methane) CH3CH2-H (ethane) (CH3)2CH-H (propane) CH3CH2CH2CH3 (H-butane) (CH3)3C-H (wo-butane) fl-CjHu-H («-pentane) CH3CH2(CH2)2CH3 («-pentane) «-C6H]3-H («-hexane) CH3CH2(CH2)3CH3 («-hexane) Alicvclic Alkanes CP-H (cyclopentane) CPH(CH3) (methylcyclopentane) CPH(CH2CH3) (ethylcyclopentane) Alkyl Aromatics C6H5-H (benzene) C6H5CH2-H (toluene) C6H5CH2 CH3 (ethylbenzene) C6H5CH2 CH2CH3 (n-propylbenzene) Y-C6H5CH(CH3)2 (zso-propylbenzene - substituted) Y = 2,5 dimethyl Y = 4-;-butyl C6H5C(CH3)2CH2-H(?-butylbenzene) Naphthalene-H (Ci position) (C2 position) Naphthalene-CH2-H (CH3 at Ci position) (CH3 at C2 position)

293

For alkyl aromatic compounds, the benzylic hydrogen atoms, i.e., hydrogen atoms bonded to the carbon atom directly attached to the aromatic ring, require the least energy to abstract due to resonance stabilization created by delocalization of the free radical within the aromatic 7t-orbital system of the aromatic ring. Table 1 illustrates how resonance stabilization decreases the bond dissociation energy of the benzylic hydrogen to a range of 83.5 - 89.8 kcal mol"1 for a variety of alkyl aromatic compounds relative to the bond dissociation energy for hydrogen atoms bonded directly to the aromatic ring (113 kcal mol"1) or to other carbons present in the alkyl functional group (e.g., 98.7 kcal mol"1 for /-butylbenzene). Therefore, in light of a radical mechanism, it is predicted that fumarate will add to the benzylic carbon atom (as long as the benzylic carbon is not quaternary and a benzylic hydrogen is available for abstraction) regardless of the alkyl functional. Moreover, in consideration of the stability afforded through hyperconjugation in the alkyl group, the C-H bond dissociation energy will be lower for a tertiary benzylic carbon compared to a secondary benzylic carbon. Such stabilization is demonstrated in the 0.8 to 4 kcal mol"1 decrease in the bond dissociation energy of various substituted z-propylbenzene compounds relative to n-propylbenzene. The metabolites of the TEX hydrocarbons have been observed to contain the succinic acid functional group at the benzylic carbon; no addition of fumarate to the methyl group of ethylbenzene has been detected. In light of a radical mechanism, the lower reactivity of benzene relative to the TEX compounds is supported by the relative bond dissociation energies. In fact, the lack of detection of succinic acid benzene metabolites and the detection of phenol and benzoate as intermediates (above), suggests that alternative mechanisms exist for oxidizing benzene for less energy than the 112.9 kcal mol"1 required for hydrogen radical abstraction from the aromatic ring. Similarly for naphthalene and alkylnaphthalene, the bond dissociation energy for abstracting a hydrogen radical from an unsubstituted naphthalene is relatively high, 111.9 to 112.2 kcal mol"1, while the comparable reaction from the methyl group of methylnaphthalene is 85 kcal mol"1. These differences in bond dissociation energies may account for the fact that naphthyl-2-methyl-succinic acid has been detected in cultures and in the field (above) but the succinic acid metabolites of naphthalene have not. 4. GEOCHEMICAL INDICATORS OF METHANOGENIC OIL BIODEGRADATION As discussed above, there is ample evidence that anaerobic microbial processes occur under reservoir conditions. There is even evidence, albeit indirect, that such processes are occurring in situ. The most widely used indicator for

294

biological methanogenesis comes from the carbon isotopic abundance signature of the methane in natural gas deposits (e.g. Hunt [109]). Most methane is thought to arise from thermogenic decomposition of biomass, kerogen, and oil [110], but biological processes are also important. Bacteria prefer the lighter 12C isotope over 13C, thus microbially-produced methane is isotopically lighter (C13 of -110 to -60 %o) than thermally-produced gas (C13 of -60 to -15 %o). The microbial process has classically been thought to occur in the relatively shallow subsurface, and to be from relatively recently buried biomass rather than from material that has undergone burial and catagenesis to petroleum. However, several reservoirs have now been found to have methane with isotopic signatures suggestive of a biogenic origin [14, 111], and this is certainly consistent with microbial methanogenesis from petroleum at depth. Unfortunately, ready interpretation of isotopic enrichment, already complicated by the likely mixing of thermogenic, biogenic, and abiogenic [112] sources, is further confounded by the discovery of anaerobic methane oxidation [113], a microbial activity in which the lighter methane isotope is clearly preferred [114]. It is thus clear that supporting evidence is needed to confirm a microbial origin for methane in many cases. This evidence might come from the oil itself. The consideration above indicates that anaerobes prefer to transform some hydrocarbons relative to others. For example, an anaerobic microbial consortium was able to degrade dimethyl-cyclopentanes and cyclohexanes under sulfate-reducing but not under methanogenic conditions and the activity under the former conditions was limited to specific isomers [26]. Perhaps the results of such preferences can be identified in oils from candidate reservoirs? Alternatively it may be possible to detect by-products of anaerobic biodegradation in waters associated with oil reservoirs, or in the oil itself. The former is proving very useful in identifying anaerobic biodegradation in contaminated aquifers, where succinate derivatives of w-alkanes, cyclic alkanes, and alkylaromatic hydrocarbons as well as naphthoic acids have been detected [115-117]. Detecting these compounds in produced waters would be good evidence that anaerobic hydrocarbon biodegradation was proceeding underground. Are there compounds in the oil that may act of fingerprints of biodegradation? Crude oils often contain naphthenic acids, carboxylic acids with one or more saturated ring structures, and at least some are believed to be the results of partial biodegradation of oil components [17]. Electrospray ionization mass spectrometry is proving to be an excellent tool for determining the molecular identity of naphthenic acids [118-120], and as more potential biodegradation intermediates are identified it will be important to see whether such compounds are present in crude oils. Dicarboxylic acids, such as the succinate derivatives indicative of anaerobic hydrocarbon metabolites, have not

295

yet been identified in oils, but they may be so polar that they primarily partition to the aqueous phase.

5. PROSPECTS FOR HYDROCABON METHANOGENESIS With notable exceptions, it is becoming increasingly clear that fumarate addition reactions represent an important mechanism for the initial activation of structurally-diverse hydrocarbons by anaerobic microorganisms. Indeed, recent surveys for such anaerobic metabolites at hydrocarbon-impacted sites identified a variety of alkylbenzylsuccinates and alkylsuccinates in situ, as well as putative PAH metabolites such as naphthoic acids and tetrahydronaphthoic acids [43, 115-117, 121, 122]. Based on such observations, one can envision that the same type of biochemical reactions might occur in oil reservoirs. However, it has long been accepted that the microbial food web in oil fields is based on aerobic hydrocarbon-oxidizing bacteria [109, 123-125]. According to this "aerobic" model, low molecular weight polar compounds such as fatty acids, organic acids, and alcohols resulting from aerobic hydrocarbon decay serve as substrates for fermentative, acetogenic, and sulfate-reducing bacteria. Further metabolic transformations of these compounds produce H2 and acetate, that can then be used by methanogenic bacteria to produce methane. While this aerobicanaerobic successional model of oil decomposition in reservoirs dominated popular thinking for many years, a reevaluation is needed in light of new knowledge. Recent geochemical considerations and microbiological data strongly indicate that oil biodegradation in the deep terrestrial subsurface proceeds mainly through anaerobic metabolism [11, 12, 16]. Biodegraded oils in deep anoxic horizons are often accompanied by hydrocarbon gasses of biological origin [14, 126]. Accordingly, isotopically light methane with 813C from -45% to -59 % indicative of a biological origin and in situ rates of methane production in the range from 1.3 to 80 nmol liter "' day"1 were observed in oil fields under various environmental conditions [127-130]. In the latter studies, methane precursors were considered to be low molecular weight compounds that originated from aerobic oil decomposition and migrated to anoxic layers. Recent studies have now shown that petroleum hydrocarbon biodegradation can be directly coupled to methane production. For example, the production of methane from the decay of toluene, o-xylene, benzene, alkanes, and some alicyclic compounds has been documented [26, 34, 45, 61, 70, 71] . In incubations of gas condensate-contaminated sediments amended with artificially weathered oil, the entire «-alkane fraction (Q3-C34 range) was completely consumed under both sulfate-reducing and methanogenic conditions. In the sulfate-free incubations, «-alkane degradation was accompanied by methane accumulation [9]. In other studies, individual alkanes such as hexadecane and

296

pentadecane were converted to CH4 by enrichment cultures and in sediment incubations [7, 8]. Based on our current understanding of methanogens, the conversion of hexadecane to CH4 might require as many as three groups of microorganisms: acetogenic (or syntrophic) bacteria converting hexadecane to acetate and H2, and acetoclastic and hydrogenotrophic archaea producing CH4 from acetate or H2 and CO2, respectively. Molecular characterization of a hexadecane-degrading methanogenic community confirmed this possible composition. It revealed three clones closely related to syntrophic bacteria of the genus Syntrophus, one clone closely related to the genus Methanosaeta, an acetoclastic methanogen, and two clones related to Methanospirillum and Methanoculleus, which comprise hydrogenotrophic methanogens [7]. Similarly, Watanabe et al. [131] found a substantial diversity of methanogens in the groundwater under an oil storage cavern in Japan. Though the most often described alkane-degrading bacteria are the sulfate-reducing bacteria, they can conceivably participate in methane production from hydrocarbons even in the absence of sulfate. These bacteria are known to couple with methanogens to form syntrophic associations wherein electron transfer occurs between the bacteria. In effect, the methanogen serves as the electron acceptor for the sulfate reducers. Thus, phylogenetic analysis of two alkane-degrading sulfate-reducing bacteria revealed that they were closely related to Syntrophobacter (from 92 to 95% identity), a genus that is known to degrade fatty acids in syntrophic co-culture with methanogens [132]. In consistent fashion, a defined co-culture of one of these organisms cultivated with Methanospirillum hungateii in the absence of sulfate could produce methane from dodecane (unpublished results). It is therefore not unreasonable to presume that similar microbial associations can exist in petroliferous subsurface formations and catalyze hydrocarbon conversions to methane and CO2. Of course, the rate of bioconversion is an extremely important when considering the prospects for microbial enhanced energy recovery. As noted, some researchers believe that such reactions, while clearly possible, take geologic time due to the limited diffusion of nutrients. While this may be true along oil migration paths, evidence to the contrary in other locales suggests that the rates need not be slow. For instance, it has been demonstrated that subsurface bacteria from oil-bearing sediments could convert hexadecane to methane quite rapidly and without a lag. Thus 10% of added 14C-hexadecane was converted to I4CH4 in about 15 d [8]. The in situ rates of methanogenesis can also be quite high in deep high temperature oil reservoirs. The rates of methanogenesis measured in formation waters of the Jurassic horizon (2299 m deep; 84°C) exceeded 80 nmol of CH4 liter "' day"1. Hybridization of 16S rRNA obtained from formation water with group-specific phylogenetic probes revealed the presence of thermophilic methanogens and heterotrophs [130]. Laboratory incubations of formation waters and raw production fluids from two deep high-

297

temperature petroleum reservoirs in California demonstrated active methane production at in situ temperatures (70-83°C). Total community DNA analysis revealed archaeal phylotypes closely related to thermophilic methanogens and sulfidogenic archaea as well as bacterial thermophiles such as Thermatoga sp., Thermococcus sp., Thermoanaerobacter sp. and Desulfothiovibrio sp. [133]. These findings contrast with the belief of low metabolic activity in the deep hot subsurface and the cessation of oil biodegradation due to the paleosterilization of formations that have at some time experienced temperatures greater than 80°C [12, 134]. The bulk of the accumulated microbiological evidence suggests that oildegrading subsurface microbial communities can be quite metabolically versatile. However, it is unreasonable to presume that the same community structure exists in all subsurface locales. The environmental conditions during oil diagenesis may have effectively eliminated critical bacterial components of obligate consortia responsible for oil methanogenesis. Clearly, the presence of hydrocarbons in the terrestrial subsurface attests to the fact that such consortia are far from ubiquitous in distribution. 6. HYDROCARBON METHANOGENESIS AND IMPLICATIONS FOR ENERGY RECOVERY Although oil is the dominant source of energy on a global scale, conventional oil production technologies are only able to recover about onethird of oil in reservoirs [135]. As a result, large quantities of residual oil remain trapped in reservoir rock pores, mainly due to capillary or subterranean forces in the vicinity of a well bore [135, 136]. Thus, enhanced oil recovery (EOR) methods have been developed to help overcome these forces and make oil move (see chapter 15). These technologies may be based on thermal, chemical, gasmiscible, or microbial technologies. It is estimated that EOR strategies can potentially add up to 60 billion barrels of oil in the near term though the increased use of existing domestic fields [137]. Understanding the multiphase flow properties of subsurface reservoir rocks and the forces that entrap oil is key for successful EOR and will help determine which technique may apply best for a given reservoir. The processes involved are complex and have been reviewed [135]. It has long been recognized that gasses dissolved in oil lower its viscosity and cause swelling. This is a major driving force for oil mobilization. In fact, gas-based EOR processes have been touted as the current, most profitable technology for recovering the large amounts of remaining oil in mature fields [135, 136]. Carbon dioxide has long been used effectively to drive enhanced oil recovery, and represents about 25% of EOR operations in the U.S. [6, 136, 138]. A secondary outcome in the use of CO2 to recover oil has far-reaching

298

environmental implications too; CO2 can be stored in reservoirs to help in the reduction of greenhouse gas emissions [6, 136, 138]. Since the combustion of fossil fuels is the largest contributor of greenhouse gas emissions, the recycling (capturing and subsequent sequestering) of anthropogenic CO2 into spent or even active reservoirs offers a promising way to both decrease the potential for global warming and increase oil recovery and profits. It has been estimated that fossil fuel reservoirs can store up to 900 billion metric tonnes of CO2 worldwide [138]. As outlined in the Introduction, natural gas is abundant worldwide, but like oil, natural gas fields can only be harvested to residual amounts or pressures making further gas unrecoverable. Carbon dioxide can also serve as an EOR gas for natural gas recovery by way of re-pressurization of reservoirs [139]. The use of CO2 as a cushion gas for natural gas storage is also being considered [140]. Of course, CO2 sequestration into natural gas fields for either recovery or as a cushion gas also offers the environmentally-friendly advantage of reducing greenhouse gas emissions [138] Although gas-based EOR with CO2 is best-understood and most widely used, the viscosity lowering of a crude by other gasses including nitrogen, flue gas, and dissolved methane and their relevance for EOR has also been considered [136, 141-143]. Indeed methane gas associated with oil can potentially help reduce its viscosity and thus enhance its recovery [141, 142]. In previous sections, we have discussed the prospect that methane gas found associated with oil reservoirs can be present as a by-product of anaerobic, microbial consumption of oil produced over millennia. In fact, there is evidence suggesting that many "dry gas" fields have arisen due to the microbial degradation of oil [13, 14]. In fields characterized by light hydrocarbons, C2 to C5 alkanes are presumably biodegraded to methane, helping to re-establish a "gas cap" [12, 15]. In theory, such "biogenic gas" could feasibly reduce oil viscosity to the point where it can be more easily recovered. In practice, gas pressure accumulations over geological time-scales have no doubt aided in conventional oil recovery but of course it remains unclear whether these gasses were thermally- or biologically-produced. Given the success of gas-based energy recovery, and the recent discovery that microorganisms can convert hydrocarbons into methane gas at substantial rates (i.e. faster than geological time scales), one could envision combining the principles of microbial- and gas-based-EOR to help recover residual oil in mature fields. Although not yet widely used in the oil industry, advances in microbial-EOR technologies have proved promising to recover residual oil (see chapter 15)[135]. Although too numerous to describe here, some MEOR technologies have explored the use of bacterial inoculation into wells to produce gaseous by-products which can help mobilize trapped oil [135]. By analogy, spent reservoirs might be inoculated with the appropriate microbial communities

299

to produce methane gas that could help decrease the viscosity of oil and aid in further recovery. What if such an inoculation procedure resulted in at least some fraction of the available energy being recovered as usable methane gas? Such speculative technology is quite far from being addressed or realized, especially from an economic point of view, but initial laboratory experimentation on this topic has been promising (Fig. 1). Samples (10 g) taken from a field in Nowata, OK that had undergone secondary oil recovery procedures (water flooding) were used to test the importance of a methane-producing oil-degrading inoculum enriched from a gas-condensate contaminated aquifer [9]. When residual oil core samples were ground or broken into small portions, the oil-degrading inoculum was effective in stimulating methanogenesis relative to a variety of controls. The latter included a heat-inactivated preparation, an oil-unamended control, and production water from the same field that received the inoculum ( Fig. 1). Interestingly, the rate of methanogenesis was much greater with the residual oil core samples than that observed when a standard oil or even when the formation (Nowata) crude alone served as a substrate for the inoculum. While the reasons for this result are under investigation, it is clear that such inocula may play a potential role for the enhanced recovery of methane from oil trapped in mature reservoirs.

Fig. 1. Methane production from residual oil in core samples inoculated with a methanogenic bacterial enrichment capable of anaerobic hydrocarbon metabolism. Symbols: Oil unamended control (•); Nowata crude oil (•); Production water (X); An artificially weathered Alaska north slope oil standard (A); Crushed core (o); Pebbled core (•). Heat inactivated and uninoculated controls are not depicted, but were uniformly negative.

300

7. MICROBIAL ENHANCED ENERGY RECOVERY AND CARBON DIOXIDE Figure 1 clearly indicates that a hydrocarbon-degrading methanogenic bacterial inoculum can attack oil deposited in rocks and covert it to natural gas. In fact, this metabolism is much faster than comparable incubations amended with an equivalent amount of oil from the same formation (estimated amount of oil in core was 0.0 lg oil/g rock based on 30-40% residual saturation). These observations lead to numerous questions that center on the rate and efficiency of oil bioconversion, the role of inocula in the process, the nutritional environment presented by petroliferous formations, the diversity of hydrocarbons susceptible to microbial attack, the biotechnological control of such bioconverstions and many other fundamental and practical considerations. Careful exploration of these issues in the future will help define the utility of enhanced energy recovery efforts at a time when the need for such considerations is particularly acute. Tomes have been written on the eventual transitioning of global energy use patterns and their potential impact on the environment. Yet, it seems clear that any energy form will have an impact on the environment and that fossil fuel use will remain the predominant energy form for decades to come. Global climate change concerns are forcing worldwide reductions in atmospheric CO2 emissions. Since methane consumption produces a fraction of the CO2 per BTU generated relative other fossil fuels, a greater reliance on methane will help reduce the rate of increase in global carbon dioxide emissions. The biotechnological link between the consumption of hydrocarbons for the production of methane may be a way of enhancing the recovery of energy in an environmentally responsible fashion, mostly from mature domestic reserves that are otherwise unprofitable or too technically difficult to exploit. It is our hope that this article helps spur such considerations. REFERENCES [1]

[2] [3] [4] [5] [6]

National Research Council (NRC). (2003). Committee on Oil in the Sea: Inputs, Fates, and Effects, Ocean Studies Board and Marine Board, Divisions of Earth and Life Studies and Transportation Research Board, NRC. In: Oil in the sea III: Inputs, fates, and effects. The National Academies Press, Washington, D.C. Energy Information Administration (EIA). (2004). www.eia.doe.gov/oiaf/ieo/index.html. C. Hall, P. Tharakan, J. Hallock, C. Cleveland, and M. Jefferson. Nature 426 (2003) 318. J.B. Curtis and S.L. Montgomery. AAPG Bull. 86 (2002) 1671. S.M. Al-Fattah and R.A. Startzman. SPE Journal, May (2000) 62-72. S. Bachu and S. Stewart. J. Can. Petrol. Technol. 41 (2002) 32.

301

[7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

[22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

Zengler, K., H.H. Richnow, R. Rosello-Mora, W. Michaelis, and F. Widdel. Nature 401 (1999)266. Anderson, R.T. and D.R. Lovley. . Nature 404 (2000) 722. Townsend, G.T., R.C. Prince, and J. M. Suflita. . Environ. Sci. Technol. 37 (2003) 5213. Prince, R. C. (2002). Petroleum and other hydrocarbons, biodegradation of. hi Encyclopedia of Environmental Microbiology; Bitton, G. Ed.; John Wiley, New York, pp. 2402-2416. Roling, W.F.M., I.M. Head, and S.R. Larter. Res. Microbiol. 154 (2003) 321. Head, I. M., D.M. Jones, and S.R. Larter. Nature 246 (2003) 344. Sakata, S., Y. Sano, Maekawa, T., and Igari, S.-I. Org. Geochem. 26 (1997) 399. Pallasser R.J. Org. Geochem. 31 (2000) 1363. Wenger, L.M., C.L. Davis, G.H. Isaksen. (2001). Multiple controls on petroleum biodegradation and impact on oil quality. SPE paper 71450. Larter, S., A. Wilhelms, I. Head, M. Koopmans, A. Aplin, R. DiPrimo, C. Zwach, M. Erdmann, N. Telnaes. Org. Geochem. 34 (2003) 601. Tissot, B.P. and D.H. Welte. 1984. Petroleum Formation and Occurrence. SpringerVerlag, Berlin. Prince, R. C , R.M. Garrett, R.E. Bare, M.J. Grossman, G.T. Townsend, J.M. Suflita, K. Lee, E.H. Owens, G.A. Sergy, J.F. Braddock, J.E. Lindstrom, and R.R. Lessard. Spill Sci. Technol. Bull. 8 (2003) 145. Britton, L. 1984. Microbial degradation of aliphatic hydrocarbons, hi D. T. Gibson (Ed.), Microbial degradation of organic compounds, Marcel Dekker, Inc., New York, pp. 89-129. Gibson, D.T. and V. Subramanian. 1984. Microbial degradation of aromatic hydrocarbons, hi T.D. Gibson (ed.), Microbial degradation of organic compounds, Marcel Dekker, Inc., New York, pp.181-252. Krumholz, L.R., M.E. Caldwell, and J. M. Suflita. (1996). Biodegradation of "BTEX" hydrocarbons under anaerobic conditions, hi Bioremediation Principles and Applications. (R.L. Crawford and D. L. Crawford, Eds.) Cambridge University Press, Great Britain, pp. 61-99. Coates, J.D., R.T. Anderson, and D.R. Lovley. Appl. Environ. Microbiol. 62 (1996) 1099. Widdel, F., and R. Rabus. Curr. Opin. Biotechnol. 12 (2001) 259. Spormann, A.M. and F. Widdel. Biodegradation 11 (2000) 85. Rios-Hernandez, L.A., L.M. Gieg, and J.M. Suflita. Appl. Environ. Microbiol. 69 (2003) 434. Townsend, G.T., R.C. Prince, and J.M. Suflita. FEMS Microbiol. Ecol. (2004). (in press). Biegert, T., G. Fuchs, and J. Heider. Eur. J. Biochem. 238 (1996) 661. Beller, H.R. and A.M. Spormann. J. Bacteriol. 179 (1997) 670-676. Leuthner, B., C. Leutwein, H. Schulz, P. Horth, W. Haehnel, E. Schiltz, H. Schagger, and J. Heider. Mol. Microbiol. 28 (1998) 615. Krieger C.J., W. Roseboom, S.P.J. Albracht, and A.M. Spormann. J. Biol. Chem. 276 (2001) 12924. Rabus, R. and J. Heider. Arch. Microbiol. 170 (1998) 377. Beller, H.R. and A.M. Spormann. Appl. Environ. Microbiol. 63 (1997) 3729. Kane, S.R., H.R. Beller, T.C. Legler, and R.T. Anderson. Biodegradation 13(2002) 149.

302

[34] Beller, H.R. and E.A. Edwards. Appl. Environ. Microbiol. 66 (2000) 5503. [35] Zengler, K., J. Heider, R. Rosello-Mora, and F. Widdel. Arch. Microbiol. 172 (1999) 204. [36] Heider, J., A.M. Spormann, H.R. Beller, and F. Widdel. FEMS Microbiol.Rev. 22 (1999) 459. [37] Leutwein C. and J. Heider. Arch. Microbiol. 178 (2002) 517. [38] Leutwein, C. and J. Heider. J. Bacteriol. 183 (2001) 4288. [39] Harwood, C.S., G. Burchhardt, H. Herrmann, and G. Fuchs. FEMS Microbiol.Rev. 22 (1999)439. [40] Chee-Sanford, J.C., J.W. Frost, M.R. Fries, J. Zhou, and J. Tiedje. Appl. Environ. Microbiol. 62 (1996) 964. [41] Edwards, E.A., L.E. Wills, M. Reinhard, and D. Grbic-Galic. Appl. Environ. Microbiol. 58(1992)794. [42] Rueter, P., R. Rabus, H. Wilkes, F. Aeckersberg, F. A. Rainey, H.W. Jannasch, and F. Widdel. Nature 372(1994)455. [43] Elshahed, M.S., L.M. Gieg, MJ. Mclnerney, and J.M. Suflita. Environ. Sci. Technol. 35(2001)682. [44] Rabus, R., H. Wilkes, A. Schramm, G. Harms, A. Behrends, R. Amann, and F. Widdel. Environ. Microbiol. 1 (1999) 145. [45] Edwards, E.A. and D. Grbic-Galic. Appl. Environ. Microbiol. 60 (1994) 313. [46] Haner, A., P. Hohener, and J. Zeyer. Appl. Environ. Microbiol. 61 (1995) 3185. [47] Dolfing J., J. Zeyer, P. Binder-Eicher, and R.P. Schwarzenbach. Arch. Microbiol. 154 (1990)336. [48] Fries, M.R., J. Zhou, J. Chee-Sanford, and J.M. Tiedje. Appl. Environ. Microbiol. 60 (1994)2802. [49] Rabus, R. and F. Widdel. Arch. Microbiol. 163 (1995) 96. [50] Hess, A., B. Zarda, D. Hahn, A. Haner, D. Stax,, P. Hohener, and J. Zeyer. Appl. Environ. Microbiol. 65 (1997) 2136. [51] Harms, G., K. Zengler, R. Rabus, F. Aeckersberg, D. Minz, R. Rossello-Mora, and F. Widdel. Appl. Environ. Microbiol. 65 (1999) 999. [52] Krieger, C.J., H.R. Beller, M. Reinhard, and A.M. Spormann. J. Bacteriol. 181 (1999) 6403. [53] Achong, G.R., A.M. Rodriguez, and A.M. Spormann. J. Bacteriol. 183 (2001) 6763. [54] Beller, H.R. and A.M. Spormann. (1999). FEMS Microbiol. Lett. 178 147. [55] Beller, H.R., A.M. Spormann, P.K. Sharma, J.R. Cole, and M.Reinhard. Appl. Environ. Microbiol. 62 (1996) 1188. [56] Ball, H.A., H.A. Johnson, M. Reinhard, and A.M. Spormann. J. Bacteriol. 178 (1996) 5755. [57] Johnson, H.A., D.A. Pelletier, and A.M. Spormann. J. Bacteriol. 183 (2001) 4536. [58] Kniemeyer, O., T. Fischer, H.Wilkes, F.O. Glockner, and F. Widdel. Appl. Environ. Microbiol. 69 (2003) 760. [59] Burland, S.M. and E.A. Edwards. (1999). Appl. Environ. Microbiol. 65 529. [60] Coates, J.D., R. Chakraborty, J.G. Lack, S.M. O'Connor, K.A. Cole, K.S. Bender, and L.A. Achenbach. Nature 411 (2001) 1039. [61] Ulrich, A C. and E.A. Edwards. Environ. Microbiol. 5 (2003) 92. [62] Edwards, E.A. and D. Grbic-Galic. Appl. Environ. Microbiol. 58 (1992) 2663. [63] Lovley, D.R., J.D. Coates, J.C. Woodward, and E.J.P. Phillips. Appl. Environ. Microbiol. 61 (1995)953.

303

[64] Coates, J.D., R.T. Anderson, J.C. Woodward, E.J.P. Phillips, and D.R. Lovley. Environ. Sci. Technol. 30 (1996) 2784. [65] Weiner, J. and D.R. Lovley. Appl. Environ. Microbiol. 64 (1998) 775. [66] Lovley, D.R., J.C. Woodward, and F.H. Chapelle. Nature 370 (1994) 128. [67] Anderson, R.T., J.N. Rooney-Varga, C.V. Gaw, and D.R. Lovley. Environ. Sci. Technol. 32(1998)1222. [68] Caldwell, M.E., R.S. Tanner, and J.M. Suflita. Anaerobe 5 (1999) 595. [69] Grbic-Galic, D. and T. Vogel. Appl. Environ. Microbiol. 53 (1987) 254. [70] Kazumi, J., M.E. Caldwell, J.M. Suflita, D.R. Lovley, and L.Y. Young. Environ. Sci. Technol. 31(1997)813. [71] Weiner, J. and D.R. Lovley. Appl. Environ. Microbiol. 64 (1998) 1937. [72] Coates, J.D., R. Chakraborty, and M.J. Mclnerney. Res. Microbiol. 153 (2002) 621. [73] Caldwell, M.E. and J.M. Suflita. Environ. Sci. Technol. 34 (2000) 1216. [74] Phelps, CD., X. Zhang, and L.Y. Young. Environ. Microbiol. 3 (2001) 600. [75] Aeckersberg, F., F. Bak, and F. Widdel. Arch. Microbiol. 156 (1991) 5-14. [76] Aeckersberg, F., F. Rainey, and F. Widdel. Arch. Microbiol. 170 (1998) 361. [77] So, CM. and L.Y. Young. Appl. Environ. Microbiol. 65 (1999) 2969. [78] Ehrenreich, P., A. Behrends, J. Harder, and F. Widdel. Arch. Microbiol. 173 (2000) 58. [79] Kropp, K.G., LA. Davidova, and J.M. Suflita. Appl. Environ Microbiol. 66 (2000) 5393. [80] Caldwell, M.E., R.M. Garrett, R.C. Prince, and J.M. Suflita. Environ. Sci. Technol. 32 (1998)2191. [81] Rabus, R., H. Wilkes, A. Behrends, A. Armstroff, T. Fischer, and F. Widdel. J. Bacteriol. 183(2001)1707. [82] Davidova, I., L. Gieg, K. Kropp, M. Nanny, J. Suflita. (2004) (submitted for publication) [83] So, CM., CD. Phelps, and L.Y. Young. Appl. Environ. Microbiol. 69 (2003) 3892. [84] Wilkes, H., S. Ktihner, C. Bolm, T. Fischer, A. Classen, F. Widdel, and R. Rabus. Org. Geochem. 34(2003)1313. [85] Callaghan, A.V., L.M. Gieg, K.G. Kropp, J.M. Suflita, and L.Y. Young. (2003). Fumarate addition during hexadecane degradation by the sulfate-reducer AK-01. American Society for Microbiology 103-rd General Meeting, Washington, D.C, Abstract Q-038, p. 521. [86] Wilkes, H., R. Rabus, T. Fischer, A. Armstroff, A. Behrends, and F. Widdel. Arch. Microbiol. 177(2002)235. [87] Galushko, A., D. Minz, B. Schink, and F. Widdel. Environ. Microbiol. 1 (1999) 415. [88] Rockne, K.J., J.C. Chee-Sanford, R.A. Sanford, B.P. Hedlund, J.T. Staley, and S.E. Strand. Appl. Environ. Microbiol. 66 (2000) 1595. [89] Ramsay, J.A., H. Li, R.S. Brown, and B. Ramsay. Biodegradation 14 (2003) 321. [90] Rothermich, M.M., L.A. Hayes, and D.R. Lovley. Environ. Sci. Technol. 36 (2002) 4811. [91] Zhang, X. and L.Y. Young. Appl. Environ. Microbiol. 63 (1997) 4759. [92] Annweiler, E., W. Michaelis, and R.U. Meckenstock. Appl. Environ. Microbiol. 68 (2002) 852. [93] Sullivan, E.R., X. Zhang, C. Phelps, and L.Y. Young. Appl. Environ. Microbiol. 67 (2001)4353. [94] Annweiler, E., A. Materna, M. Safinowski, A. Kappler, H.H. Richnow, W. Michaelis, and R.U. Meckenstock. Appl. Environ. Microbiol. 66 (2000) 5329. [95] Zhang X, Sullivan E. R, and L.Y Young. Biodegradation. 11 (2000) 117.

304

[96] Baciocchi, E., F. D'Acunzo, C. Galli, and O. Lanzalunga. J. Chem. Soc. Perkin Trans. 2.2(1996)133. [97] March, J. 1968. Advanced Organic Chemistry Reactions, Mechanisms, and Structure. McGraw-Hill Book Company, New York, New York. [98] Ruscic, B., M. Litorja, and R. Asher. J. Phys. Chem. A. 103 (1999) 8625. [99] Saekins, P.W., M.J. Pilling, J.T. Niiranen, D. Gutman, and L.N. Krasnoperov. J. Phys. Chem. 96(1992)9847. [100] Pedley, J.B, R.D. Naylor and S.P. Kirby. 1986. Thermodynamic Data of Organic Compounds. 2nd ed., Chapman and Hall, New York [101] Castelhano, A.L. and D. Griller. J. Amer. Chem. Soc. 104 (1982) 3655. [102] Tumanov, V.E. and E.T. Denisov. Neftekhimiya. 41 (2001) 109. [103] Wenthold, P.G., and R.R. Squires. J. Am. Chem. Soc. 116 (1994) 6401. [104] Ellison, G.B., G.E. Davico, V.M. Bierbaum, and C.H. DePuy. Int. J. Mass Spectrum. Ion Processes. 156 (1996) 109. [105] McMillen, D.F. and D. M. Golden. Ann. Rev. Chem. 33 (1982) 493. [106] Denisov, E.T. and T.G. Denosova. (2000). Handbook of Antioxidants. CRC Press, New York. [107] Kromkin, E.A., V.E. Tumanov, and E.T. Denisov. Neftekhimiya. 42 (2002) 3. [108] Reed, D.R. and S.R. Kass. J. Mass Spectrom. 35 (2000) 534. [109] Hunt, J.M. 1979. Petroleum Geochemistry and Geology, 2nd Ed. W.H. Freeman and Company, New York, pp. 413. [110] Domine, F., R. Bounaceur, G. Scacchi, P.-M. Marquaire, D. Dessort, B. Pradier, and O. Brevart. Org. Geochem. 33 (2002) 1487. [ I l l ] Sassen, R., A.V. Milkov, E. Ozgul, H.H. Roberts, J.L. Hunt, M.A. Beeunas, J.P. Chanton, D.A. DeFreitas, and S.T. Sweet. Org. Geochem. 34 (2003) 1455. [112] Sherwood Lollar, B., T.D.Westgate, J.A. Ward, G.F. Slater, and G. LacrampeCouloume. Nature 416 (2002) 522. [113] Nauhaus, K., A. Boetius, M. Kriiger, and F. Widdel. Environ. Microbiol. 4 (2002) 296. [114] Orphan, V.J., C. H. House, K.U. Hinrichs, K. D. McKeegan and E. F. DeLong. Science 293 (2001) 484. [115] Beller, H.R. Biodegradation. 11 (2000) 125. [116] Gieg, L.M. and J.M. Suflita. Environ. Sci. Technol. 36 (2002) 3755. [117] Griebler, C, M. Safinowski, A. Vieth, H.H. Richnow, and R.U. Meckenstock. Environ. Sci. Technol. 38 (2004) 617. [118] Barrow, M.P., L.A. McDonnell, X. Feng, J. Walker, and P.J. Derrick. Anal. Chem. 75 (2003) 860. [119] Gabryelski, W. and K.L. Froese. Anal. Chem. 75 (2003) 4612. [120] Lo, C.C., B.G. Brownlee, andNJ. Bunce. Anal. Chem. 75 (2003) 6394. [121] Phelps, C D , J. Battistelli, and L.Y. Young. Environ. Microbiol. 4 (2002) 532. [122] Martus, P. and W. Puttman. Sci. Total Environ. 307 (2003) 19. [123] Belyaev, S.S., K. Laurinavichus, A.Y. Obraztsova, S.N. Gorlatov, and M.V. Ivanov. Microbiologiya. 51 (1982) 997. [124] Nazina, T.N, E.P. Rozanova, and S.I. Kuznetsov. Geomicrobiol. J. 4 (1985) 103. [125] Palmer, S.E. (1993). Effect of biodegradation and water washing on crude oil composition. In Engel M.H, Macko S.A. (Eds.), Organic Geochemistry. Plenum Press, New York, pp. 511-533. [126] James, A.T. and B.J. Burns. Bull. Am. Assoc. Petrol. Geol. 68 (1984) 957.

305

[127] Ivanov, M.V., S.S. Belyaev, A.M. Zyakun, V.A. Bondar, and K.K. Laurinavichus. Geokhimiya. 11 (1983) 1647. [128] Borzenkov LA., S.S. Belyaev, Y.M. Miller, LA. Davydova, and M.V. Ivanov. Microbiology 66 (1997) 104. [129] Nazina, T,N., A.E. Ivanova, LA. Borzenkov, S.S. Belyaev, and M.V. Ivanov. Geomicrobiol. J. 13 (1995) 181. [130] Bonch-Osmolovskaya, E.A., M.L. Miroshnichenko, A.V. Lebedinsky, N.A. Chernyh, T.N. Nazina, V.S. Ivoilov, S.S. Belyaev, E.S. Boulygina, Y.P. Lysov, A.N. Perov, A.D. Mirzabekov, H. Hippe, E. Stackebrandt, S. L'Haridon, and C. Jeanthon. Appl. Environ. Microbiol. 69(2003)6143. [131] Watanabe, K., Y. Kodama, N. Hamamura, and N. Kaku. Appl. Environ. Microbiol. 68 (2002) 3899. [132] Davidova, I.A., K.G. Kiopp, K.E. Duncan, and J.M. Suflita. (2002). Anaerobic biodegradation of n-alkanes by sulfate-reducing bacterial cultures. Abstracts of the International Symposium on Subsurface Microbiology. Copenhagen, Denmark, September 8-13. [133] Orphan,, V.J., S.K. Goffredi, E.F. Delong, and J.R. Boles. Geomicrobiol. J. 20 (2003) 295. [134] Connan, J. 1984. Biodegradation of crude oils in reservoirs. In J. Brooks, and D.H. Welte (Eds.), Advances in Petroleum Geochemistry. Academic Press, London, pp. 89129. [135] Mclnerney, M.J. and D.W.S. Westlake. 1990. Microbial enhanced oil recovery. In: Microbial Mineral Recovery, H.H.L. Ehrlich and C.L. Brierley (Eds.), McGraw-Hill, NY, pp. 409-445. [136] Rao, D. J. Can. Petrol. Technol. 40 (2001) 11. [137] Aycaguer, A.-C, M. Lev-On, and A.M. Winer. Energy & Fuels 15 (2001) 303. [138] Oldenburg, CM., K. Pruess, and S.M. Benson. Energy & Fuels 15 (2001) 293. [139] Oldenburg, CM. Energy & Fuels 17 (2003) 240. [140] Killesreiter, H. Erdoel und Kohle, Erdgas, Petrochemie 38 (1985) 405. [141] Frauenfeld, T.W.J., R.K. Ridley, R.K., and D.M. Nguyen. J. Petrol. Technol. 40 (1988)333. [142] Mayne, CJ. and R.W. Pendleton. Soc. Petrol. Eng. AIME 1 (1986) 131. [143] Alvarez, M.R, M.F. Hilton, and H.L. Oil & Gas J. 82 (1984) 95.



307

Chapter 11

Using nitrate to control microbially-produced hydrogen sulfide in oil field waters R.E. Eckford and P.M. Fedorak Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2E9

1. INTRODUCTION The presence of hydrogen sulfide (H2S) in oil fields can be the result of abiotic or biotic processes. In the later case, sulfate-reducing bacteria (SRB) are the culprits that produce this nocuous gas, leading to "souring" that is defined as the process whereby petroleum reservoirs experience an increase in the production of H2S during the economic production life of the field [1]. The increase in H2S content leads to a decrease in the economic value of the gas and oil, as well as operational problems associated with the H2S. This microbial process in wastewaters and oil field waters can be controlled by another group of microbes, known as nitrate-reducing bacteria (NRB). Their metabolic activities stop sulfate reduction by SRB, and in many cases the NRB can actually consume sulfide, thus decreasing H2S concentration in the waters. Jenneman et al. [2] have referred to these sulfide-consuming bacteria as "sulfide bioscavengers". Hitzman and Sperl [3] used the term "biocompetitive exclusion" to describe the microbial process in which NRB use volatile fatty acids and out-complete SRB to prevent or decrease sulfide production, and enhance oil recovery. This chapter will review (a) H2S in the petroleum industry, (b) the metabolism of SRB leading to sulfide production, (c) the occurrence, types and activities of NRB that might be found in oil field waters, (d) some laboratory studies that have elucidated the mechanisms by which NRB control sulfide produced by SRB, (e) some oil field experiences with nitrate injection to control sulfide in wastewaters, surface waters and oil field waters, and (f) some of the U.S. patents that apply to this microbial process.

308

Although nitrite, rather than nitrate, addition has been studied, this chapter focuses solely on the use of nitrate to control sulfide in oil field waters. This is a proven biotechnology that is under-utilized by the petroleum industry. 2. H2S AND THE PETROLEUM INDUSTRY 2.1. Formation of H2S Kerogen is the organic source material from which petroleum is formed and released [4-5]. The formation of petroleum occurs in the deeper subsurfaces as burial continues and temperature and pressure increase [5]. First oil, then gas is expelled from kerogen as the maturation process continues. Significant oil generation occurs between 60° and 120°C, and significant gas generation occurs between 120° and 2v25°C [5]. During the maturation process, H2S is also released. Machel [6] wrote, "The association of dissolved sulfate and hydrocarbons are thermodynamically unstable in virtually all diagenetic environments. Hence, redox-reactions occur, whereby sulfate is reduced by hydrocarbons either bacterially (bacterial sulfate reduction) or inorganically (thermochemical sulfate reduction)." Temperature is the major factor determining which process occurs. The microbiological process is common at temperatures for 0 to 60 or 80°C, whereas, the thermochemical process occurs at temperatures greater that 100° to 140°C [6]. Because temperature increases with burial depth, H2S found at shallow depths is usually the result of bacterial sulfate reduction whereas, H2S found at greater depths is the result of thermochemical sulfate reduction [7]. However, there are shallow pools that contain higher than expected concentrations of thermochemically generated sulfide [8]. These are believed to be the result of thermochemical sulfate reduction occurring downdip and migrating upward to a shallow reservoir [8]. At the time of discovery, the H2S concentration in an oil field depends upon its maturation history and/or the migration of H2S into the oil field. However, during oil recovery from some oil fields, an increase in H2S concentration (souring) can occur as a result of pressurizing the formation by injecting water into the reservoir. This process, know as waterfiooding, is discussed in section 3. Three well-documented examples of oil field souring are given in the following paragraphs. Cochrane et al. [9] describe the souring of the Ninian field in the North Sea. This field was discovered in 1974, and after several years of operation, injection of sea water was used to maintain the production rate. This was followed by an increase in sulfide production attributed to bacterial sulfate reduction. The reservoir temperature was initially between 100° to 120°C, but in the areas adjacent to the injection well bores, the temperature was cooled to as low as 40°C, which was conducive to bacterial sulfate reduction.

309

Frazer and Boiling [10] described the souring of the Kuparuk River field on the North Slope of Alaska. The field was initially sweet, but after injection of Beaufort Sea water, detectable levels of H2S began to appear at the producing wells. The connate water contained essentially no sulfate. However, the sulfate in the sea water stimulated bacterial sulfate reduction in the reservoir that had a temperature of about 70°C. The Skjold oil field in the North Sea soured upon the onset of waterflooding [11]. Oil and gas production began from this field in 1982 and sea water injection began in April 1985. In September 1985, the first recorded H2S production was measured to be 1.8 ppm in the gas phase. In 2002, the concentrations varied from 10 to 1000 ppm [11]. In late 1999, this field produced 1150 kg H2S d"1. These examples clearly demonstrate that waterflooding can stimulate bacterial sulfate reduction, leading to souring. Although these examples refer to offshore oil fields, souring also occurs in land-based oil fields using waterflooding [12-15]. As a result of the bacterial production of toxic H2S, the value of the oil decreases as the oil field sours. 2.2. H2S toxicity and properties H2S is a very dangerous gas, even though it occurs in nature. Its characteristic rotten egg smell is generally obvious at 0.13 ppm by volume and quite noticeable at 4.6 ppm [16]. Unfortunately the smell sense becomes quickly fatigued and can fail to warn of higher concentrations. Collapse, coma and death from respiratory failure may occur within a few seconds after one or two inspirations of the undiluted H2S [17]. The U.S. Occupational Safety and Health Administration has established the acceptable ceiling concentration of 20 ppm (by volume) for H2S with an acceptable maximum peak above the acceptable ceiling concentration of 50 ppm for an 8-h shift [16]. The specific gravity of H2S is 1.19; therefore it will collect in low places and accumulate under poorly ventilated conditions [18]. H2S is soluble in water and oil. It is a weak acid existing in aqueous solutions as H2S, HS~, or S~ (pKa values of 7.04 and 11.96). Aqueous solutions of H2S absorb O2 leading to the formation of elemental sulfur [17]. 2.3. Detrimental effects of H2S Besides its toxicity, H2S is a nuisance in the petroleum industry because it contaminates gas and stored oil, it corrodes iron in the absence of air (anaerobic corrosion), and it precipitates as amorphous ferrous sulfide (FeS), plugging and diminishing the injectivity of water injection wells [18]. In addition, fluids with water and H2S, may cause sulfide stress cracking of susceptible metals. This is affected by metal composition, pH, H2S concentration, total pressure, total tensile stress, temperature and time [19].

310

Two types of cracking known to occur in wet H2S environments are sulfide stress corrosion cracking and hydrogen-induced cracking (see chapters 7 and 8). The former occurs in steels of relatively high strength and in welds of welded steel structures. A crack propagates under working stress or residual stress vertically to the stress axis [20]. This type of corrosion is most damaging to drillpipe and well production facilities [21]. Hydrogen-induced cracking occurs parallel to the surface when no external stress is applied. It is also known as hydrogen blistering because of the blisters that appear on the surface of the metal [20]. General corrosion attack by H2S is influenced by the presence of CO2, O2 and brine, [18, 21]. It is related to the alloy composition and strength of steel [21]. H2S forms FeS scale, which is cathodic to the metal, promoting localized attack under the scale, as well as the penetration of H2 into the metal [21-22]. Figure 1 shows the process whereby an anode and cathode pair are generated by the action of SRB acting on sulfates in the presence of iron. The cathode is depolarized as the SRB consume H2. At the anode, iron (Fe) is oxidized to Fe2+ which combines with H2S produced by the SRB, giving FeS. This process results in a loss of structural material. Heterotrophic SRB also play a role in the deposition of FeS (Fig. 1).

Fig. 1. Iron metal corrosion mediated by SRB in a biofilm. The process is caused by the consumption of H2 causing cathodic depolarization. Adapted from Ref. [18].

311

Removal of dissolved gases (O2, H2S and CO2) from drilling and produced fluids is necessary to minimize corrosion damage. H2S in oil base drilling fluid is removed by gas separators and vacuum degassers, and then neutralized. Controlling corrosion in H2S-containing environments requires proper selection of materials, including the use of low-hardness steels, application of inhibitors and complete exclusion and removal of O2 from water used in petroleum production [21]. Clearly, the presence of H2S greatly increases the cost of exploration for oil and natural gas, and the cost of production and storage of petroleum. Plugging (or biofouling) of injection wells is also caused by SRB. The sulfide they produce, precipitates soluble iron in the injection or formation water forming colloidal FeS [23]. This colloidal material becomes associated with bacterial cells and oil, forming a gummy mass that can clog reservoirs and plug injection wells. The activities of SRB can also produce calcite (CaCO3) that can add to the plugging problem. 3. OIL RECOVERY AND WATERFLOODING Under primary oil recovery, typically less than 30% of the original oil is produced, so that improved or enhanced methods are used to recover some of the remaining oil [24]. These processes, known as secondary and tertiary recovery methods, include the addition of energy into the reservoir and are accomplished by injecting some type of fluid through injection wells. This is referred to as enhanced oil recovery and involves water injection, gas injection, steam injection, combustion, miscible fluid displacement and polymer injection [24]. In this paper, only water injection or waterflooding will be discussed. Waterflooding involves pumping water into the reservoir to stimulate production. The injected water provides pressure to force the oil out of the rock and to sweep it toward producing wells as shown in Fig. 2. Waterflooding has been attempted in almost every type of reservoir, with its greatest success in relatively homogenous reservoirs having sufficient permeability to allow water injection at a reasonable rate [24]. Up to 60% of the oil can be recovered with waterflooding [5]. Water handling can become a major operational procedure. For example, in some western Canadian oil fields, the proportion of water in the oil-water emulsion brought to the surface can be 95% by volume [15]. That is, the volume of water handled is 19 times greater than the volume of oil produced. Water used as injection water can be of three types: formation water, sea water or fresh water. Formation water is subsurface brackish or brine water produced from a petroleum or non-petroleum producing formation. Sea water may also include water from a salty (non-potable) lake. Fresh water, containing

312

less than 2000 ppm dissolved solids, is primarily water that can be made potable by flocculation, filtration and chlorination [25]. Because oil field reservoir rocks are porous, they are susceptible to plugging by solids suspended in or precipitated from an injection fluid [26]. This makes water quality testing necessary to determine parameters such as: amount and composition of suspended solids, clay sensitivities, presence of bacteria, compatibility of two or more waters, and compatibility of the injection solution with reservoir rock. An example of incompatible waters occurs when sulfate scales, such as barium sulfate, calcium sulfate or strontium sulfate are formed by mixing waters containing sulfate with waters containing barium, calcium or strontium ions [26]. As well, the gases O2, H2S and CO2 found in injection waters and implicated in corrosion [25-26], must be monitored. Water quality testing, should be continued after the enhanced oil recovery operation hasstarted, to ensure that the system is maintained at optimum conditions [25]. Water treatment methods are outlined by Rose et al. [27].

Fig. 2. A simple waterflooding operation. Oil, gas and water are collected from the production wells and the produced water is separated from the oil and gas. The produced water is combined with source water and injected into the oil-bearing rock to pressurize the formation and sweep the oil to the producing wells.

313

Water should be free of bacteria that can cause corrosion [25-26], or plugging of equipment and injection wellbores [25]. The presence of bacteria can be problematic because they reproduce rapidly over wide ranges of pH, temperature, pressure and anoxia in the reservoir. Bacteria found in oil field injection waters that cause problems are SRB, iron-reducing bacteria and slimeformers [25, 27]. Of special concern are the SRB. Source waters used in waterflooding can increase the activities of SRB souring for several reasons [1]. The source water, especially sea water, may contain sulfate to serve as a terminal electron acceptor and may introduce SRB, nutrients such as short chain fatty acids and ammonium into the reservoir. Large volumes of source water may reduce the salinity and temperature in the formation near the injection well, providing an environment that is more conducive to the growth of SRB and oil field souring. 4. SULFATE-REDUCING BACTERIA Ask any person who works in the oil field or who is involved with the transport or storage of crude oil to name some bacteria, and most will immediately respond "sulfate-reducing bacteria" or "SRB". These bacteria are well-known, and in the oil field environment, they are a nuisance because their metabolic activities produce H2S that can sour reservoirs, create plugging through FeS formation and induce corrosion [28]. SRB have the unique ability to utilize sulfate as a terminal electron acceptor. This is an anaerobic respiratory process used to generate energy for the biosynthetic reactions involved in cell growth and maintenance [29]. The SRB are a diverse group of prokaryotes that are found in many anaerobic environments. These bacteria have been the subject of several books [30-33] and countless articles. The phylogeny of SRB has recently been reviewed [34], and based on rRNA sequences, they fall into four groups: Gramnegative mesophiles, Gram-positive endospore-formers, thermophilic bacteria, and thermophilic Archaea. 4.1. Overview of the metabolism of SRB The dissimilatory H2S-producing SRB have little energy available to them. The upper limits of energy conservation from sulfate reduction are set by thermodynamics. For example, if a potent electron donor like H2 is oxidized, the free energy change of the overall reaction, under standard conditions at neutral pH, is -38 kJ (mole H2)A (reaction 1), which is 6-fold lower than with O2 as a terminal electron acceptor (reaction 2) [35]. 4H2 + SOzf + 2H+ -> H2S + 4H2O

G°' = -38 kJ (mol H2)"1

(1)

314

4H2 + 202 -> 4H2O

G°' = -237 kJ mol H2)"1

(2)

As late as the 1970's, only a few genera of SRB were recognized, and these were known to use only a few growth substrates, most notably lactate, pyruvate or H2. Now it is apparent that SRB are capable of using various compounds for electron donors. Based on their metabolic capabilities, heterotrophic SRB fall into two groups: those that cannot oxidize acetate, and those that carry out complete oxidation of acetate to C0 2 [36]. Reaction (3) illustrates the overall reaction of lactate-utilizing SRB that cannot oxidize acetate. One mol of acetate accumulates for each mol of lactate that is consumed. 2CH3CHOHCOO" + S04 = + 2H+ -> 2CH3COO" + 2H2O + 2CO2 + H2S G°' = -77 kJ (mol lactate)"1

(3)

The complete oxidation of acetate is given by reaction (4), showing that less energy is available per mol of acetate than per mol of lactate (reaction 3). CH3COO" + SOzf + 3H+ -» 2CO2 + H2S + 2H2O G°' = -41 kJ (mol acetate)"1

(4)

Increased understanding of the metabolic diversity of SRB now indicates that nearly 100 organic compounds can be used by various SRB [37]. These substrates include fatty acids up to C2o; aromatic hydrocarbons such as toluene, xylenes, ethylbenzene, and naphthalene; «-alkanes from (C6 to C2o); and simple oxidation products of hydrocarbons such as benzoate, phenol, and cresol [3840]. These substrates are present in native crude oils or partially degraded crude oils. Thus, if there is an ample supply of sulfate in water contacting crude oil in an anaerobic environment, there is the potential for SRB to actively produce H2S, using many different organic compounds (or H2) as an energy source. The ability to reduce sulfate links this diverse group of bacteria. However, it is now apparent that various SRB can reduce other chemical species including Fe(III), nitrate, some chlorinated aromatics, sulfur oxyanions and O2 [37]. Molecular oxygen can be reduced by most SRB. In this case, the stoichiometry (for example, 2H2 consumed per O2 reduced) indicates that O2 can be completely reduced to water. SRB are also capable of fermentative growth or utilization of other electron acceptors, such as sulfite, thiosulfate and elemental sulfur [12, 35] and tetrathionate [12]. Many SRB are able to ferment organic substrates in the absence of sulfate. For example, Desulfotomaculum orientis can carry out

315

fermentation using homoacetate. Also many SRB can perform a unique fermentation of inorganic sulfur compounds which are disproportionated to sulfate (a more oxidized compound) and sulfide (a more reduced compound). For example, thiosulfate is transformed to equal amounts of sulfate and sulfide, and sulfite is disproportionated to 3/4 sulfate and 1/4 sulfide [35]. Some species of SRB are able to utilize nitrate as an electron acceptor. When nitrate is used as an electron acceptor, SRB produce ammonia, but not N2, as an end product. Nitrite is formed as an intermediate of nitrate reduction and can be reduced by many sulfate reducers unable to reduce nitrate. In the presence of both sulfate and nitrate some SRB will preferentially use one or the other as an electron acceptor, and some SRB will reduce both concomitantly [35]. 4.2. Activities of SRB in anaerobic environments When microorganisms get into stagnant or closed water systems, dissolved O2 is quickly and completely consumed. Despite the absence of O2, organic matter may undergo biological decomposition by microbial activities, including fermentation. The degradation reactions by which most fermentative bacteria gain energy are disproportionations of the organic matter, part converted to CO2, and part converted to reduced products, such as fatty acids, H2, and alcohols [18]. If sulfate is abundant in these anaerobic environments, the fermentation products are used by SRB. Sulfate serves as the terminal electron acceptor, and the reducing power from the decomposed organic matter results in the formation of H2S. SRB grow in anaerobic muds found in fresh water or sea water environments [41]. They are also indigenous members of the microbial community in ground waters, marine environments, coastal sediments, marine hydrothermal vents associated with volcanic or tectonic activity, and hot springs [42]. SRB can flourish in environments wherever decomposable organic matter gets into anaerobic, sulfate-containing waters. Here H2S is produced and evidenced by visible blackening of the sediment when FeS forms from iron minerals [18]. Marine and estuarine saltmarsh sediments, saline and hypersaline lakes and ponds, as well as oil field waters with high sulfate content are the most permanent and significant habitats of SRB [43]. Large amounts of sulfate are required for this process, so that the consequence resulting from the growth of SRB is the dissemination of massive quantities of H2S [29]. Many SRB use simple, low molecular weight compounds, and therefore depend on fermentative bacteria to cleave and ferment complex organic matter. SRB convert only about 10% of the total substrate carbon to cellular material, so that the bulk of the substrate has to be decomposed for providing energy. Thus,

316

SRB, make themselves conspicuous by the formation of their metabolic product, H2S, rather than by formed cell mass [18]. How do SRB become so closely linked to oil recovery processes? Some think that SRB are imported with surface or ground waters. This hypothesis is illustrated by a gradual increase of sulfide production after the beginning of operations in oil fields [18]. Azadpour et al. [42] reported that SRB were absent in thirteen core samples of petroliferous formations obtained from a wide variety of geographical locations, depths and types of formations. Produced waters from six of the wells were also tested and five were positive for SRB. Acetateutilizing SRB of the genus Desulfobacter were found in an oil field sea water injection system [44]. In culture, they produced extensive biofilm and exhibited high levels of hydrogenase activity, which suggests a sessile habit and a role in the cathodic depolarization mechanism of microbially influenced corrosion. Others have suggested that deep terrestrial subsurface reservoirs contain active and diverse populations of microorganisms including SRB [12]. Thermophilic SRB isolated from oil field waters in the Norwegian sector of the North Sea were thought to be indigenous to the reservoir [45]. See chapter 14 and Ref. [46] for discussion of microorganisms and oil reservoirs. 4.3. Controlling SRB in oil fields using biocides Virtually all oil field water systems contain some bacteria [27], and biocides are widely used to kill or inhibit the activities of these microorganisms, including SRB. There are two general types of biocides: oxidizing and nonoxidizing. Typically, oxidizing biocides (such as chlorine, sodium hypochlorite, chlorine dioxide, chloroamines and bromine) are used in fresh water systems, whereas non-oxidizing biocides (including aldehydes, quaternary amines, halogenated organics, organosulfur compounds, and quaternary phosphonium salts) are used in many different types of water systems [47]. Biocide application in large waterflooding systems presents problems such as high cost, environmental risks [18], and worker safety. The use of biocides is most successful in controlling unwanted activities in surface facilities. When used to eliminate bacteria in injection water or kill SRB in the formation, the degree of difficulty and expense increases significantly [12]. Nonetheless, application of biocides is the most common method of controlling microbial activities in the oil field. Jack and Westlake [48] reviewed the control of SRB in the petroleum industry. 5. NITRATE-REDUCING BACTERIA 5.1. Types of NRB There are two major groups of bacteria that could be stimulated by the presence of nitrate in anaerobic environments. These are chemoorganotrophs

317

(heterotrophs) that use organic compounds as electron donors and as their carbon source for growth (Fig. 3), and chemolithotrophs (autotrophs) that typically use reduced inorganic sulfur species as electron donors and CO2 as their carbon source for growth (Fig. 4). The latter group is also known as the "colorless sulfur bacteria". Figures 3 and 4 show some of the characteristics of these NRB and their end products from nitrate reduction. These figures broadly represent the types of bacteria that might be stimulated by nitrate, although some, such as Thiobacillus denitrificans, and Paracoccus pantotrophus, (Fig. 4) do not appear to have been described as oil field bacteria. Pseudomonas stutzeri is given as an example of a heterotrophic NRB that might be stimulated by nitrate (Fig. 3). A nitrate-respiring bacterium, that has a 100% similarity to P. stutzeri, was isolated from an enrichment from water injectors in a North Sea oil field [49]. Among the heterotrophs in Fig. 3 are facultative anaerobes (such as some Pseudomonas and Bacillus species), that prefer to grow using O2 as their terminal electron acceptors, but will grow using nitrate as their terminal electron acceptor in the absence of O2. These are known as denitrifying bacteria, yielding N2 as the major endproduct of nitrate respiration. There have been countless studies of denitrifying bacteria in soils and wastewater treatment, but these bacteria have been largely ignored in oil field studies. Denitrifying bacteria have been shown to degrade a variety of hydrocarbons (for review see Refs. [39-40]), and with the abundant supply of dissolved hydrocarbons in produced waters, these heterotrophs may be stimulated by nitrate injection into a reservoir. Another group of heterotrophic, facultative anaerobes is the ammoniumproducing, NRB, such as Citrobacter spp. (Fig. 3), other members of Enterobacteriaceae, and a few other genera [50]. We have found no investigations that have described ammonium production in oil field waters by this group of facultative anaerobes. However, Telang et al. [51] mentioned an oil field isolate (designated NH15b) that was tentatively identified as a Citrobacter sp. or Salmonella sp. These would have the potential to reduce nitrate to ammonium. Using a MPN method with medium that is selective for heterotrophic, ammonium-producing, NRB, we have observed that these NRB were detected, but not abundant, in western Canadian oil field waters nor were their numbers greatly increased when nitrate was added to laboratory incubations of produced waters [Eckford and Fedorak, unpublished data]. Recently, the strictly anaerobic ammonium-producing, nitrate-reducing bacterium, Denitrovibrio acetiphilus was isolated from an oil reservoir model column, and it was shown to produce ammonium in medium that contained acetate and nitrate [52]. Some SRB (Desulfovibrio spp.) have also been included as heterotrophs that might be stimulated by the addition of nitrate (Fig. 3) because a few of these

318

reduce nitrate to ammonium [53-56]. In the presence of nitrate, some SRB will preferentially use nitrate, and some will use both concomitantly [54]. Thiobacillus denitrificans is listed as one of the chemolithotrophs in Fig. 4. In general, this species is not tolerant to high sulfide concentrations, but Sublette and Woolsey [57] enriched Thiobacillus denitrificans strain F that initially tolerated up to 1.75 mM sulfide, and later up to 2.5 mM sulfide [58]. This strain has been used in studies to demonstrate its ability to reduce H2S concentrations in porous rock cores [59-60] and in sour produced waters [58,61]. Gevertz et al. [62] described two novel bacterial isolates that are obligate chemolithotrophs, using nitrate as a terminal electron acceptor, and sulfide as an energy source. Both grow under anaerobic conditions. One isolate is a denitrifier that closely resembles Thiomicrospria denitrificans, and it has been called Thiomicrospria strain CVO (Fig. 4). The other isolate was called Arcobacter strain FWKO B, and it reduces nitrate to nitrite.

Fig. 3. Examples of some heterotrophic bacteria that could be stimulated by the presence of nitrate in anaerobic environments that contain suitable organic substrates.

319

Fig. 4. Examples of some chemolithotrophic bacteria that could be stimulated by the presence of nitrate in anaerobic environments. See text for details.

Injection of nitrate into an oil field might also stimulate the activity of bacteria similar to P. pantotrophus [63] (formerly Paracoccus denitrificans [64] and Thiosphaera pantotropha strain GB17 [65]). This bacterium was isolated from a denitrifying effluent treatment system. It is a facultative anaerobe and facultative autotroph (Fig. 4) that uses nitrate as an electron acceptor. It grows autotrophically with sulfide as an electron donor, or heterotrophically with a variety of organic compounds (including acetate which is commonly found in produced waters [66-67]) as electron donors [65]. We are not aware of any research that has detected facultative chemolithotrophs in oil field waters. The bacteria shown in Fig. 4 all have the capability of oxidizing sulfide while reducing nitrate. These are referred to as nitrate-reducing, sulfideoxidizing bacteria (NR-SOB). Greene et al. [68] compared the sulfide tolerance of four species of NR-SOB. In their liquid medium, sulfide was oxidized by Thiobacillus denitrificans strain F at concentrations less than 0.5 mM, by Thiomicrospira denitrificans and Arcobacter sp. strain FWKO B at up to 3 mM, and by Thiomicrospira strain CVO at up to 15 mM.

320

Although only a few NR-SOB have been identified in oil field waters, Loka Bharathi et al. [69] isolated over 100 strains of anaerobic colorless NRSOB from sea water and a sulfide-rich creek. Their data showed that different isolates oxidized sulfide at different rates. For example, one isolate oxidized all of the sulfide in the medium within 9 days, whereas another isolate oxidized only 2.9% of the sulfide in the same time. Thus, it is likely that different NRSOB in the produced water from oil fields would oxidize sulfide at different rates. 5.2. NRB in oil field waters The presence of NRB in oil field waters has not be studied extensively. This group of microorganisms was not even mentioned in a review entitled "Microbiology of petroleum reservoirs" [46]. Several investigations have enumerated NRB in oil field waters using most probable number (MPN) methods with different media formulations. Some of the results are summarized in Table 1, in chronological order. One of the first enumeration studies [70] used molasses or sucrose as electron donors in the media to count heterotrophic NRB in samples taken as near the wellheads as possible. Very low numbers ( 4 L"1) were found in these samples. Most of the other media formulations preferentially, but not exclusively, cultured autotrophs. For example, the medium used by Davidova et al. [14] (Table 1) contained only inorganic compounds except for yeast extract, with thiosulfate serving as the electron donor. This would preferentially grow microorganisms that are similar to Thiobacillus denitrificans. Other investigations in Table 1 used sulfide as the electron donor with filter-sterilized produced water from the oil field that was being studied [51, 71]. The filtered produced water undoubtedly contained some dissolved organic compounds, so it would support the growth of heterotrophic NRB and autotrophic NRB. The medium used by Telang et al. [72] in Table 1, contained only inorganic compounds except for acetate, with sulfide serving as the electron donor. Telang et al. [72] in Table 1 described the isolation and characterization of two autotrophic NR-SOB from an oil field in Saskatchewan, Canada. One was designated Thiomicrospira strain CVO (formerly Campylobacter strain CVO, [51 ]) and the other was designated Arcobacter strain FWKO B. The DNA from these two isolates has been used extensively with a method known as reverse sample genome probing (RSGP), first described by Voordouw et al. [73]. Using RSGP, Telang et al. [51] (Table 1), demonstrated that the abundance of strain CVO increased after the waterflooded oil field was treated with nitrate. This molecular technique corroborated the increase in NR-SOB numbers determined by the MPN method. The high specificity of the RSGP for NR-SOB precluded the detection of other NRB in samples from four additional oil fields

321

from western Canada and west Texas [72], although culture methods detected NRB (Table 1). Eckford et al. [74], in Table 1, surveyed five oil fields in western Canada for various types of NRB. Different media formulations were used to selectively enumerate thiosulfate-oxidizing NRB, heterotrophic NRB, or NR-SOB. None of the 18 water samples contained detectable numbers of thiosulfate-oxidizing NRB. As was observed by Adkins et al. [70], the numbers of NRB were very low or non-detectable near the wellheads [74]. However, NRB were detected in source and preinjection waters, and in samples from water storage tanks and free water knock out units. Although much of the work on NRB in oil field waters has neglected the heterotrophic NRB, the numbers of heterotrophic NRB were greater than the numbers of autotrophic NRB in 12 of the 15 samples compared. In one oil field, heterotrophic NRB were found, but no autotrophic NRB were detected (Ref. 74, Table 1). NRB were detected in biofilms on coupons in the anaerobic part of the water injection system of the Veslefrikk field in the North Sea [75], (Table 1). The medium used to enumerate these attached bacteria contained organic acids as carbon sources, providing counts of heterotrophic NRB. These numbers increased dramatically after nitrate injection (Table 1). The literature surveyed in Table 1 represents 15 different oil fields that have been examined for NRB. Each of the oil fields contained detectable numbers of NRB at one or more sampling locations. Thus, each field had a microbial community containing NRB with the potential to be stimulated by nitrate amendment. 6. CONTROLLING MICROBIAL PRODUCTION OF SULFIDE WITH NITRATE ADDITION 6.1. Microbial mechanisms leading to the control of sulfide concentrations after nitrate addition There appear to be five mechanisms by which sulfide concentrations can be controlled in the presence of nitrate and sulfate. The first involves the competition between heterotrophic NRB and SRB for a common electron donor. For example, acetate serves as an electron donor for NRB [76] and for several genera of SRB [34]. Equations (5) and (6) illustrate that if acetate is available, nitrate reduction yields more energy per mol of electron donor or acceptor than does sulfate reduction [77].

322

Table 1 Detection and enumeration of NRB in oil field waters. Refs.

Oil

70

fields

Methods

Comments

Oklahoma, USA

MPN with molasses and sucrose as electron donor

Samples collected near wellheads. Medium would detect heterotrophic NRB. MPN values were 4mL"\

71

Saskatchewan, Canada

Single-bottle MPN using filter-sterilized oil field water supplemented with nitrate

Oil field water contained about 120 mg sulfide L"1. Method likely selected for NR-SOB. Initial count, 104 ml/ 1 , Count after nitrate injected into reservoir, 108 mL"1.

51


Single-bottle MPN using filter-sterilized oil field water supplemented with nitrate

Oil field water contained about 100 mg sulfide L"1. Method likely selected for NR-SOB. Initial counts as low as 0 mL"1. Counts after nitrate injected into reservoir, as high as 108 mL"1.

51


RSGP

NR-SOB strain CVO became dominant community member after nitrate injection into reservoir.

72

Western Canada Single-bottle MPN and west Texas, using medium with USA sulfide, acetate and nitrate

Method likely selected for NR-SOB, but may have grown heterotrophic NRB. Counts from 102 mL"1 to 106 mL"1 in five samples examined.

72

Western Canada RSGP & west Texas, USA

NR-SOB strains CVO and FWKO B detected in only one of five samples examined.

14

Oklahoma, USA and Alberta, Canada

MPN with inorganic salts, yeast extract, and thiosulfate as the electron donor.

Method likely selected for thiosulfateoxidizing NRB, but may have grown heterotrophic NRB. Counts were typically IOHCO3" + 4N2 + 4H2O AG0' = -495 kJ (mol N O 3 ) " ' or AG°' = -792 kJ (mol acetate)"1 (5) CH3COO" + SO4= - • 2HCO3~ + HS" AG°' = -47 kJ (mol acetate or SO4= )"' (6) Thus, heterotrophic NRB out-compete heterotrophic SRB for electron donors, thereby suppressing sulfide production. Oil field waters contain dissolved organic compounds including short-chain fatty acid anions like acetate, propionate and butyrate [12, 46, 67], as well as aromatic compounds such as toluene and phenols that are substrates for heterotrophs. This mechanism would stop sulfide production, but it would not remove sulfide that is present in the reservoir or produced waters. A second mechanism results from the increased redox potential of an aqueous environment caused by the activities of denitrifying bacteria [78]. The production of N2O (and maybe NO), two oxidizing agents, raises the redox potential to above -100 mV, which is too high for the growth of SRB [31]. Nitrate reduction in laboratory experiments causes the redox indicator, resazurin, to turn from colorless to pink [78-79]. Resazurin is 50% oxidized at -51 mV [80]. This alteration of the redox potential in an aqueous environment inhibits sulfide production. A third mechanism results from the stimulation of NR-SOB in the presence of nitrate. Two processes come into play in this case. Some NR-SOB are denitrifiers and they produce N2O from nitrate, thereby elevating the redox potential of the medium [78]. In addition, the NR-SOB use sulfide as their electron donor, and oxidize it to elemental sulfur or sulfate [2]. Thus, these two processes combine to inhibit sulfate reduction and remove sulfide that is present in the aqueous environment. The activities of the NR-SOB have the potential to stop sulfide production, and to remove essentially all of the sulfide in the aqueous environment. A fourth mechanism is nitrate reduction by SRB. Some SRB reduce nitrate to ammonium [53-56]. The importance of this mechanism in controlling sulfide production is largely unexplored. Jenneman et al. [78] point out that when SRB reduce nitrate, ammonium is formed rather than N2O or N2. The formation of N2O would be detrimental to the SRB as discussed above. The fifth mechanism is the production and accumulation of nitrite during nitrate reduction. Myhr et al. [49] demonstrated that the activity of the dominant sulfate-reducing strain found in their laboratory experimental system was inhibited by 120 uM nitrite. However, some species of SRB contain nitrite

324

reductase which reduces nitrite to ammonium [81], thereby protecting these species from the nitrite produced by NRB [68]. 6.2. Control of sulfide in wastewaters Long before nitrate addition was considered for controlling sulfide in oil field waters, it was used to control odors in wastewaters and receiving surface waters. For example, in 1931, a combination of sodium nitrate and chlorinated lime was used to control odors from Coney Island Creek in New York [82]. This creek was described as "one of the vilest bodies of water in the United States" [82] as a result of receiving 6,000,000 gallons (23,000,000 L) of sewage and industrial wastewater. After the first day of chemical application, there was a marked decrease in odor. During the month-long treatment, 10 tons (9 Mg) of sodium nitrate were applied to the creek, and the sulfide concentrations in the water decreased sharply. Table 2 summarizes five studies, in chronological order, in which nitrate was used to control odors and sulfide production in wastewaters. The first four entries in Table 2 enhance the activities of native NRB by adding nitrate. Two of these were large scale projects that involved nitrate applications to a river [83] and to a sludge storage lagoon [84] for odor control. The other three studies were laboratory-scale investigations using sewage sludge [78], oily sludges from naval operations [85], and aqueous solutions of sulfide [58]. The latter report described work in which Thiobacillus denitrifwans was initially used to oxidize sulfide, and later Thiobacillus denitrificans strain F was used because of its tolerance to higher sulfide concentrations. Each of the attempts to control odor or sulfide production listed in Table 2 was successful. One of the studies [84] observed that nitrate amendment led to increased redox potential followed by a reduction in odor. The increased redox potential was observed in another study [78] and this was attributed to the microbial production of N2O. The increase in redox potential to above -100 mV would inhibit growth of SRB. 6.3. Laboratory studies using cores or columns Using nitrate to control sulfide production in a petroleum reservoir involves adding nitrate to the injection water and pumping it into the oil-bearing formation. To be effective, the nitrate must migrate into the reservoir and be consumed by NRB. The NRB may be present in the oil field or water handling system, or they might be deliberately added to the oil field to stimulate nitrate reduction. Several laboratory studies have been done to assess the effectiveness of this process using cores or a column of sand. Five of these studies are summarized in Table 3, in chronological order.

325

Table 2 Laboratory and field studies using nitrate to control sulfide production in wastewaters Ref.

Summary

83

Three pulp mills discharged sulfite wastes into the Androscoggin River in Maine U.S.A. This resulted in H2S production in the river and odor problems in nearby towns. In 1949, a total 641 tons (582 Mg) of NaNC>3 were added to the river. This controlled H2S production and odors. Most of the nitrate was reduced to ammonium.

84

To control odor, waste sodium nitrate liquor (containing both nitrate and nitrite) was added to a storage lagoon that held aerobically digested waste activated sludge. Initially, the redox potential of the water was near -lOOmv, but after several months of nitrate addition, it rose to near +300 mV. There was low odor potential when the redox was above +100. Acetate concentrations decrease in the lagoon, and N2 production from denitrification provided mixing within the sludge.

78

Laboratory studies were done with a 10-fold dilution of sewage sludge amended with 20 mM sulfate and one of three electron donors: glucose, acetate, or H2. The addition of 59 mM nitrate completely inhibited sulfide production. Nitrate, nitrite and N2O were detected in the inhibited samples, and the oxidation of the redox indicator, resazurin, was attributed to the presence of N2O. The numbers of SRB decreased with prolonged incubation of the oxidized medium.

85

Oily sludge from a settling tank at the U.S. Navy Craney Island Fuel Depot in Virginia was placed in serum bottles and amended with nitrate, stimulating indigenous NRB. Sulfate reduction was diminished with 50 mM nitrate, and sulfide accumulation was prevented with as little as 16 mM nitrate. Nitrite and nitrous oxide were products of nitrate reduction. Sulfide was oxidized to sulfur or sulfate. The results indicated that nitrate would be useful for preventing sulfide formation in oily wastes produced onboard marine vessels.

58

This paper reviewed bench-scale processes developed for the sulfide removal from gases and aqueous solutions by Thiobacillus denitrificans. When H2S was introduced to batch anoxic or aerobic cultures of T. denitrificans, the H2S was immediately metabolized. Oxidation of H2S to sulfate was accompanied by growth. T. denitrificans was immobilized by co-culture with floc-forming heterotrophs and this mixture was used to treat water that was contaminated with sulfide. The sulfide-active floe was stable for 5 months of operation with no external organic carbon required to support the growth of the heterotrophs. T. denitrificans strain F, which tolerates higher sulfide concentrations, was also used in some studies.

326

Table 3 Laboratory studies using nitrate to control sulfide production columns or cores Ref.

Summary

59

This study investigated the efficacy of nitrate and the sulfide-tolerant Thiobacillus denitrificans strain F in controlling H2S concentrations in cores of sandstone. Formation water from a gas storage facility in Redfield, Iowa, U.S.A. was injected into two core systems, with hydraulic retention times (HRTs) of 3.2 h and 16.7 h. With the addition of nitrate alone, no thiobacilli were cultured from the core system, but nitrate was consumed and the concentrations of sulfide in effluent decreased by about 40% in the core with the shorter HRT, and 98% with the longer HRT. Thus, an indigenous microbial community capable of oxidizing sulfide while using nitrate as the electron acceptor was present. Inoculation with strain F reduced the effluent sulfide by about 80% in the core with the shorter HRT.

60

The test materials for this study included core material from the St. Peter formation at Redfield, Iowa, U.S.A. and water from the same formation, supplemented with acetate and enriched with SRB to 107 cells ml/ 1 . The core material did not contain large numbers of organisms capable of using nitrate, and no strain F-like organisms were detected. When nitrate and strain F were injected into the core, sulfide concentrations decreased, demonstrating the ability of strain F to control sulfide in the core.

86

This work examined controlling microbial souring in anaerobic upflow columns containing crushed Beria sandstone maintained at 60°C. Produced waters from the Ninian North Sea and the Kuparuk North Slope oil fields were used as sources of microorganisms, and these gave similar results. A highly anaerobic medium that contained short-chain organic acids found in the produced waters was pumped through the columns. Nitrate injection stimulated indigenous microbes and inhibited souring at thermophilic temperatures. Initially, 3.6 mM nitrate was needed to inhibit souring but later 0.36 mM nitrate prevented further souring. Nitrate was reduced to nitrite, with no N2O, N2 or ammonium detected.

2

Brine from an oil field near Coleville, Saskatchewan, Canada was filtered, supplemented with phosphate and nitrate and pumped into a porous (1288 mD) ceramic core 19.1 cm long. When 5 mM nitrate was shut in the column, all of the sulfide was removed in 3 d and the numbers of NRB increased. Under various flow regimes, with sulfide-containing brine, sulfide removal was between 87 and 100%. Elemental sulfur, bacteria and CaCC>3 were produced, but there was no significant permeability changes across the core following all treatments.

49

Separate enrichments of aerobic oil-degrading bacteria, NRB, SRB and methanogens were inoculated into a 200-cm column packed with oil-soaked silica sand. The column was flooded with air-saturated synthetic sea water and operated under different influent regimes for nearly 1100 d. Injecting 0.5 mM nitrate led to the complete elimination of H2S. Inhibition of the SRB was attributed to the nitrite produced from nitrate reduction. Three strains of heterotrophic NRB were isolated from the column and none used H2S or S° as electron donor.

327

Four of the five studies in Table 3 detected NRB in the cores or produced waters used in the experimental systems. In the fifth study, [49] the investigators inoculated the column with a mixture of enrichment cultures, including NRB. Two of the studies, Refs. 59 and 60, focused on the activities of thiobacilli. None were detected in the cores or waters, similar to the findings of Eckford and Fedorak [74]. Inoculating these two cores with Thiobacillus denitrificans strain F stimulated sulfide reduction when nitrate was injected into the cores (Refs. 5960, Table 3). Two of the studies [2, 86], (Table 3) relied solely on the formation water as the source of NRB. One study supplemented the medium with short-chain organic acids [86], whereas the other study did not supplement with organic compounds [2]. Thus, these studies likely enriched for different nutritional types of NRB. Nonetheless, souring was inhibited in both studies. Indeed, sulfide production was controlled in each of the five studies summarized in Table 3. 6.4. Laboratory studies using natural microbial communities in produced waters Produced waters from various oil fields have been used as sources of planktonic microorganisms in studies of the ability of nitrate to control sulfide formation in these waters. Table 4 summarizes four of these investigations in chronological order. In each study, sulfide removal was stimulated by nitrate addition. In three of the reports, no organic supplementation was required to stimulate sulfide removal. However in one case [87], two of the four oil field waters did not respond to amendments with inorganic nutrients (nitrate and phosphate). Sulfide removal was only stimulated after the addition of acetate or formate plus vitamins or yeast extract, indicating that in some cases heterotrophic NRB play an important role in the process of sulfide removal. Eckford and Fedorak [15] demonstrated that heterotrophic NRB can be stimulated by simply adding nitrate. This is illustrated in Figs. 5 and 6. A produced water sample was collected from the free water knock out at the Coleville field that has a severe souring problem. This water was used for a serum-bottle microcosm study. Initially, the microcosm contained 2.7 mM sulfide which increased to 3.1 mM by day 1 and then dropped below detection by day 3 in the nitrate-amended microcosm (Fig. 5a). The sulfate concentration increased noticeably over the first 14 d of incubation, with a total increase of 3.5 mM by day 38, closely matching the 3.1 mM decrease in sulfide. The nitrite concentration was at a maximum of 1.8 mM on day 3 and then gradually decreased to 0.2 mM by day 38. Figure 5b shows the results of chemical analyses of a microcosm that was not supplemented with nitrate. The sulfide increased to 4 mM by day 5, and the sulfate remained fairly steady at from 0.68 mM to 0.54 mM throughout the testing period. Neither nitrate nor nitrite was detected in the microcosms. The huge increase in numbers of heterotrophic NRB

328

(Fig. 6a) during the time that sulfide was removed (Fig. 5a) suggests that these bacteria play a role in this process. However, their role has not be elucidated.

Table 4 Laboratory studies on controlling sulfide production in produced waters by adding nitrate to stimulate natural microbial communities. Ref.

Summary

71 & 2

Anaerobic enrichments were prepared by supplementing nitrate and phosphate to brine samples collected from an oil field near Coleville, Saskatchewan, Canada. Within 24 to 48 h after supplementation, complete oxidation of 3 to 4 mM sulfide was observed. Elemental sulfur was formed and the stoichiometry of the reaction was 5HS" + 2NO3~ + 7H+ -> 5S° + N2 + 6H2O.

87

Waters from four west Texas oil fields were used to determine which amendments were required to stimulate sulfide removal. In two of the samples, addition of 40 mM nitrate and phosphate was not sufficient to promote microbial removal of sulfide over a 28-d incubation. However, sulfide removal was observed when acetate or formate plus vitamins or yeast extract were added to these two waters that had been supplemented with nitrate and phosphate. These results illustrate the importance of heterotrophic activity in sulfide removal.

14

Two waterflooded, souring oil fields in Oklahoma, U.S.A. and Alberta, Canada were studied. SRB and NRB were found in produced waters from both oil fields. The majority of the sulfide production appeared to occur after the oil was pumped aboveground, rather than in the reservoir. Sulfide production was greatest in the water storage tanks in the Alberta field. Laboratory experiments showed that adding 5 and 10 mM nitrate to produced waters from the Oklahoma and Alberta oil fields, respectively, decreased the sulfide content to negligible levels and increased the numbers of NRB.

15

Produced waters from three sulfide-containing western Canadian oil fields were amended with nitrate only. In less than 4 d, the sulfide was removed from the waters from two of the oil fields (designated P and C), whereas nearly 27 d were required for sulfide removal from the water from the third oil field (designated N). Nitrate stimulated large increases in the numbers of the heterotrophic NRB and NR-SOB in the waters from oil fields P and C, but only the NR-SOB were stimulated in the water from oil field N. These data suggest that the stimulation of the heterotrophic NRB is required for rapid removal of sulfide from some oil field produced waters.

329

Fig. 5. Chemical analyses of microcosms that contained produced water from the Coleville oil field in Canada. Nitrate amended (a), unamended (b). From Ref. 15.

Bacterial enumerations were done on samples from the nitrate-amended and the unamended microcosm. The MPN results are shown in Fig. 6. Initially, the number of NR-SOB (2.1xlO5 ml/ 1 ) was much greater than the number of heterotrophic NRB (4.3x102 ml/ 1 ). There was no increase in the numbers of heterotrophic NRB (Fig. 6a) or NR-SOB (Fig. 6b) in the unamended microcosm. In contrast, there was a rapid increase in the numbers of heterotrophic NRB and NR-SOB by day 7 (Figs. 6a and 6b) in the nitrate-amended microcosm. The numbers of heterotrophic NRB and NR-SOB increased 22,000-fold and 440fold, respectively. These proliferations occurred during the time when nitrate consumption was the most rapid, and sulfide was depleted from the microcosm (Fig. 5a). At day 7, the numbers of heterotrophic NRB and NR-SOB were 9.3xl0 6 ml/ 1 and 9.3xl0 7 ml/ 1 , respectively. Over the remainder of the

330

incubation, the heterotrophic NRB numbers remained high, whereas the NRSOB numbers dropped to near their original count (Figs. 6a and 6b). The SRB numbers did not change in the nitrate-amended microcosm and showed a slight increase in the unamended microcosm with a maximum at day 7 (Fig. 6c).

Fig. 6. Heterotrophic NRB (a), NR-SOB (b) and SRB (c) counts is samples from microcosms that contained produced water from the Coleville oil field in Canada (Fig. 5). Error bars show 95% confidence intervals. From Ref. [15].

331

Laboratory studies have led to field application of nitrate or changes to field operations. For example, the work described in references [2, 71] (Table 4) preceded the experimental injection of nitrate into the Coleville field in Saskatchewan, Canada [13, 71, 88], and results from laboratory studies encouraged the implementation of nitrate injection in a North Sea oil field [75]. Based on laboratory investigations, Davidova et al. [14] observed that the rate of sulfide production was higher in aboveground samples than in samples collected from wellheads. At an Alberta oil field, they observed high sulfate reduction activity in water storage tanks that had retention times of 2 to 3 d, and they calculated that 80 kg of microbially-produced sulfide was injected into the reservoir daily from these storage tanks. Operators of this oil field have now eliminated the long retention time in the storage tanks, which has helped to reduce souring. 6.5. Laboratory studies using co-cultures of bacteria To assess the microbial dynamics and processes that occur when nitrate is added to communities containing NRB and SRB, Voordouw and co-workers have done several studies in which pure cultures of bacteria were mixed and monitored (Table 5). Their work focused on the activities of the autotrophic NR-SOB Thiomicrospira strain CVO and Arcobacter strain FWKO B. In all cases, the NR-SOB proliferated with the addition of nitrate, and in most cases, they removed sulfide from the medium and caused the cessation of sulfate reduction. However, sulfate reduction was not stopped in co-cultures in which the SRB produced nitrite reductase [68] (Table 5). Nitrite formed during nitrate reduction is inhibitory to SRB. However, the inhibition is only transient when SRB, that produce nitrite reductase, reduce nitrite to ammonium [81]. This work [68] (Table 5) clearly demonstrated that the activities of these NR-SOB cannot control sulfide production by all SRB, although the NR-SOB can oxidize the sulfide that is formed by the SRB. Rates of corrosion have also been studied in co-culture experiments (Table 5) [89]. The addition of strain CVO and nitrate to a culture of Desulfovibrio sp. strain Lac6 accelerated the corrosion rate to 0.07 mm y"1. Lacatena et al. [90] also measured corrosion rates, but they worked with an undefined, mixed enrichment culture in produced water. In the absence of nitrate in produced water, the corrosion rate was 0.46 mm y"1, but with nitrate in the produced water, the corrosion rate dropped sharply to 0.03 mm y"1. Data from nitrate injection into a North Sea oil field showed that prior to nitrate injection the corrosion rate was 0.7 mm y"1, but after 4 months of nitrate injection, the rate dropped to 0.2 mm y"1 [75]. Thus, the co-culture experiments (Table 5, Ref. 89) gave results that differed from those obtained with undefined mixed cultures [90] and full scale operations [75].

332

Table 5 Laboratory studies using co-cultures and nitrate to control sulfide production. Ref.

Summary

72

Mixtures of strains CVO and FWKO B were incubated in medium with different concentration of sulfide. Using RSGP, it was demonstrated that CVO dominated in co-cultures with low (1 mM) sulfide, but FWKO B dominated with high (15 mM) sulfide. CVO or FWKO B were co-cultured with Desulfovibrio strain Lac6. Sulfide drop from 1 mM to 0 mM in 24 h in the presence of CVO. Over a 277-h incubation, sulfide remained between 1 and 2 mM in the presence of FWKO B.

91

Strain CVO was added to cultures of Desulfovibrio strain Lac6 that were growing in various concentrations of nitrate or lactate. In pure culture, sulfate reduction by the Desulfovibrio sp. was unaffected by the nitrate concentrations up to 10 mM. Sulfide concentrations decreased rapidly after the addition of CVO. This effect was due to the increase in the redox potential of the medium, as indicated by the oxidation of resazurin.

89

The influence of nitrate-mediated control of sulfide production on metal corrosion was studied with strain CVO and a Desulfovibrio strain Lac6. The corrosion rate in cultures of the Desulfovibrio sp. without or with nitrate was 0.01 mm y"'. The addition of CVO to the nitrate-containing culture increased the corrosion rate to 0.07 mm y"1. The same trend was observed when CVO and nitrate were added to a consortium of SRB from a produced water. The increased rate of corrosion was attributed to the formation of thiosulfate and polysulfide during the oxidation of sulfide.

68

Strain CVO was grown in co-cultures with four different Desulfovibrio strains. Two of these did not have nitrite reductase, and their growth was stopped in the presence of CVO as it produced nitrite and elevated the redox potential of the medium. However, two of the strains had nitrite reductase, and they reduced the nitrite formed by strain CVO. The SRB decreased the redox potential and continued to produce sulfide. This illustrated that the action of strain CVO cannot inhibit SRB that possess nitrite reductase.

6.6. Oil field observations There have been few reports of field tests or full-scale application of nitrate injection to control sulfide. Six reports are summarized in Table 6 (in chronological order). Three of these focused on the extensive studies done on the Coleville oil field in Canada during two experimental injections [13, 51, 71, 88]. The microbial community in the Coleville oil field was extensively characterized using the RSGP method, and the produced water was the source of the well-studied NR-SOB, Thiomicrospira strain CVO and Arcobacter strain FWKO B. Results from nitrate addition to two oil fields in the North Sea have also been reported [11, 75] (Table 6). These include an 8-month study [11] and a

333

long-term application, with data reported after 32 months of operation [75]. Numbers of planktonic NRB were monitored in the first five studies listed in Table 6, and numbers of sessile NRB were reported in the last study given in Table 6. Three common observations were evident from the field studies summarized in Table 6. First, NRB were present in each of the oil field waters studied. Thus, no intentional inoculation of NRB was required to stimulate the beneficial activities of these bacteria. Second, nitrate injection stimulated the NRB and, in reports in which NRB were enumerated, their numbers increased 100- to 60,000-fold during the monitoring times. Third, nitrate injection controlled sulfide production. Each of these observations was completely predicable from laboratory studies summarized in Tables 3, 4, and 5. 6.7. U.S. Patents The ability to stop sulfide production in oil fields, or to remove sulfide from sour waters and petroleum are essential in petroleum recovery and processing. The inhibition of sulfate reduction decreases corrosion and other problems associated with SRB and provides huge cost saving to the oil field operators. Therefore, it is not surprising that several patents have been issued for the use of nitrate or NRB for sulfide removal or control. Table 7 lists some of the U.S. patents dealing with these processes. Patent no. 4,879,240 uses a mutant strain of Thiobacillus denitrificans that is tolerant to elevated concentrations of sulfide and glutaraldehyde (presumably strain F) to control sulfide in environments such as oil field injection waters, reservoirs, and waste treatment of materials that contain SRB. A sulfide-tolerant strain of Thiobacillus denitrificans is the microbial component of patent no. 4,880,542 used to remove H2S from sour waters originating from petroleum production, anaerobic sewage digestion or other industries. These autotrophic bacteria are co-immobilized with CaCO3 in alginate beads and placed in a column, through which the wastewater is pumped. Nitrate or O2 can serve as the terminal electron acceptor. The activities of heterotrophic denitrifying bacteria are stimulated by supplementing oil field waters (or other sulfide-containing waters) with nitrate and an organic compound, such as acetate (patent nos. 5,405,531 and 5,750,392; Table 7). This allows the NRB to out-compete the SRB for organic substrates. In addition, these patents include the addition of molybdate to further inhibit SRB. The use of the autotrophic NR-SOB Thiomicrospira (formerly Campylobacter sp.) strain CVO and Arcobacter strain FWKO B for the removal of sulfide from oil field brines is covered by patent nos. 5,686,293 and 5,789,236 (Table 7). The uses include aboveground treatment of sour waters or injection of these NR-SOB into subterranean formations. The waters are supplemented with nitrate and phosphate.

334

Table 6 Field studies and operations using nitrate to control sulfide production. Ref.

Summary

12

Ammonium nitrate (45 T) was injected into a souring oil field at the Southeast Vassar Verta Sand Unit in Oklahoma, U.S.A. At the time of injection, no nitrate was detected in three adjacent production wells. Forty-five days after injection, nitrate was detected at these wells, and the sulfide concentrations were reduced by 40 to 60%.

71

In 1994, a solution of NH4NO3 and NaH2PO4 was injected into three wells in the Coleville field in Saskatchewan, Canada. Prior to treatment, the produced waters from these wells contained between 52 and 160 mg sulfide L"1. After injection, there were shut-in periods of between 24 and 70 h before pumping resumed. The sulfide concentrations dropped by as much as 98% of the initial concentrations, with ranges between 40% and 60% being sustained for several hours. The numbers of NRB increased by 100- to 10,000-fold.

88 & 13

In 1996, a solution of NH4NO3 and NaH2PO4 was injected into two injection wells in the Coleville field for 50 d. Two producer wells were monitored for 90 d after the injection began. After 10 d, the sulfide in the producers decreased by as much as 50 to 60% of the initial concentrations of 60 and 40 mg L"1. The cumulative sulfide removal from the two producers were estimated to be 50 and 70 kg over the 90-d test period. The numbers of NRB increased at least 1,000-fold during the time of nitrate injection.

51

Samples were taken from the Coleville field in 1996. These were taken 8 d before and 20, 55, and 82 d after the injection of a solution of NH4NO3 and NaH2PO4 began. RSGP analyses, using 47 DNA standards, showed that strain CVO became the dominant community member immediately after injection. The abundance of CVO decreased within 30 d after completion of nitrate injection.

11

Studies were done in the Skjold oil field in the North Sea in 2000. Three injection strategies were used. In each case, the highest nitrate concentrations were used at the beginning of the treatment, then the concentration was decreased. First, nitrate (4.5 to 1.7 mM) was injected into one well for 1 month; second, nitrate (3.8 to 1.8 mM) was injected into this well plus another well for 2 months; third, nitrate (4.4 mM to a mean of 2.8 mM) was injected into all of the other wells for 3 months. Only one of the monitored production wells showed marked reduction in H2S. This well was in the highly fractured zone of the reservoir, and nitrate reached it within 24 h of the start of injection. The amount of H2S in the produced gas dropped from 240 ppm to between 30 to 60 ppm. After nitrate addition, the numbers of mesophilic NRB and NR-SOB increased about 10,000- and 1,000-fold, respectively.

75

Data were presented after 32 months of adding nitrate to water injected from the Veslefrikk platform in the North Sea. Glutaraldehyde injection was stopped in January 1999, and replaced by continuous 0.25 mM nitrate injection. Microbial counts in biofilms were monitored and corrosion was measured by weight loss from C-steel biocoupons. After 32 months, the numbers of SRB decreased 20,000-fold and after 18 months, the number of NRB increased 60,000-fold. Most of the NRB were heterotrophic facultative anaerobes. Sulfate-reducing activity (measured using 35S-sulfate) decrease 50-fold. Prior to nitrate treatment, the corrosion rate was 0.7 mm y"1. This fell to 0.02 mm y"1 after 4 months of nitrate injection.

335

Table 7 Examples of United States patents for the control of sulfide through the application of NRB. Patent no.

Inventors and year

Title

4,879,240

Sublette et al. 1989

Microbial control of hydrogen sulfide production by sulfate reducing bacteria

4,880,542

Sublette 1989

Biofilter for the treatment of sour water

5,405,531

Hitzman et al. 1995

Method for reducing the amount of and preventing the formation of hydrogen sulfide in an aqueous system

5,686,293

Jenneman et al. 1997

Sulfide-oxidizing bacteria

5,750,392

Hitzman et al. 1998

Composition for reducing the amount of and preventing the formation of hydrogen sulfide in an aqueous system, particularly in an aqueous system for oil field applications

5,789,236

Jenneman 1998

Process of using sulfide-oxidizing bacteria

6.8. Economics and advantages of using nitrate to control sulfide production Based on the trial injections at the Coleville oil field in Canada, Jenneman et al. [88] did a cost analysis for sulfide removal using different chemicals. They injected ammonium nitrate (cost US$0.31 kg"1) and monosodium phosphate (cost US$2.57 kg"1) to stimulate NRB in the reservoir. The combined cost of these chemicals was determined to be between US$0.76 and $1.19 kg"1 H2S removed. They compared this cost to reported costs for sulfide removal from wastewaters using hydrogen peroxide or sodium hypochlorite. With hydrogen peroxide, the estimated cost was between US$4.40 and $17.60 kg"1 H2S removed, and with sodium hypochlorite the estimated cost was between US$3.96 and $13.20 kg"1 H2S removed. With data from one well, Jenneman et al. [88] estimated the cost of using ammonium nitrate and monosodium phosphate to be $0,018 barrel"1, or $1.80 (100 barrels)"1, of produced water treated. Herbert [92] compared the costs of using nitrate with those of using the biocides glutaraldehyde or tetrakishydroxymethylphosphonium sulfate (THPS)

336

for offshore oil fields. The costs did not include the cost of transporting the chemicals. The estimated prices per litre of the chemicals were: US$0.25 for nitrate (as a 40% solution of CaNO3), $2.50 for glutaraldehyde (as a 50% solution), and $4.00 for THPS (as a 50% solution). Although the cost of nitrate was lower, the solution was continuously injected at a dose of 60 mg L"1. In contrast, the two biocides were injected for 1 h, twice per week at a dose of 500 mg L"1. Based on treating 200,000 barrels of produced water per d, the yearly costs for chemicals were US$575,000 for nitrate, $345,00 for glutaraldehyde, and $500,000 for THPS. Per 100 barrel of water treated, these costs become US$0.79, and $0.47, and $0.68, respectively. From these two cost analyses, the use of nitrate for sulfide control is competitive with other chemicals. The cost of treating 100 barrels of water calculated from the data given by Jenneman et al. [88] is higher than that reported by Herbert [92], because Jenneman et al. [88] also injected monosodium phosphate, which is 8 times as expensive as the ammonium nitrate. Herbert [92] used only calcium nitrate. The need to add a phosphate source to stimulate NRB would have to be evaluated for each oil field. Besides the cost, other factors must be considered when choosing chemicals for controlling sulfide in produced waters. Most notably, workers safety and potential environmental impact of spilled chemical must be considered. Nitrate salts are far less toxic than the biocides commonly used in oil fields, and therefore its use presents few safety issues for oil field workers. Spilled biocides have negative affects on the environment. In contrast, nitrate is widely used as an agricultural fertilizer, so spills on land present no major problem. Nitrate is listed as a substance that poses little or no risk to the marine environment [75]. However, caution must be used to avoid contamination of fresh surface waters or potable ground waters with nitrate (or any biocide). 7. CONCLUDING REMARKS The use of nitrate to control microbially-produced sulfide in oil fields is a proven biotechnology that is grossly under-used by the petroleum industry. Its effectiveness has been demonstrated in many laboratory investigations and in some field studies. The microbiology is adequately well-understood, although it is not clear whether heterotrophic or autotrophic NRB play the more important role. This may vary from oil field to oil field. Nonetheless, from the results in the literature, nitrate amendment (and in some cases phosphate or organic acid amendment) stimulates NRB in the oil field waters, and there appears to be little need to add an inoculum of NRB. Nitrate has replaced biocides in some of the oil fields in the North Sea, and the results have been very positive. Besides controlling sulfide levels, there is also preliminary evidence that corrosion rates are reduced [75]. In addition,

337

there are plans to use nitrate in the Gulf of Mexico when sea water injection begins in the near future (Stephen Maxwell, Commercial Microbiology Inc., personal communication). In contrast, there is little or no use of nitrate in landbased souring oil fields in North America. It is now very clear that land-based oil field operators should seriously consider using this proven biotechnology to control, and possibly eliminate, microbially-induced souring and the problems associated with H2S formation. REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] [II] [12] [13] [14] [15] [ 16] [17] [18]

G.B. Farquhar, Corros. Prev. Contr., 45(2) (1998) 51. G.E. Jermeman, D. Gevertz and M. Wright, In Proceedings of the Third International Petroleum Environmental Conference, Vol. II, Albuquerque, NM, 1996, pp. 693-704. D.O. Hitzman and G.T. Sperl, Paper SPE 27752, presented at the 9th Symposium on Improved Oil Recovery Tulsa, Oklahoma, USA, April 17-20, 1994. B.P. Tissot and D.H. Welte, Petroleum Formation and Occurrence 2nd edition, Springer-Verlag, Berlin, 1984. R.C. Selley, Elements of Petroleum Geology, 2nd edition, Academic Press, NY, USA, 1998,296. H.G. Machel, Sediment. Geol, 140 (2001) 143. H.G. Machel and J.M. Foght, In R.E. Riding and S.M. Awramic (eds.), Microbial Sediments, Springer-Verlag, Berlin, 2000, pp. 105-120. B.K. Manzano, M.G. Fowler and H.G. Machel, Org. Geochem., 27 (1997) 507. WJ. Cochrane, P.S. Jones, P.F. Sanders, D.M. Holt and M.J. Mosley, Presented at the SPE European Petroleum Conference, London UK, October 1988, SPE paper 18368, Society of Petroleum Engineers, Richardson, Texas, USA. L.C. Frazer and J.D. Boiling, Presented at the International Arctic Technology Conference, Anchorage Alaska, May, 1991, SPE paper 22105, Society of Petroleum Engineers, Richardson, Texas, USA. J. Larsen, Paper 02025 Proceedings of the NACE Expo 2002 Annual Conference and Exposition, Denver, Colorado, USA, April 7-11, 2002. M.J. Mclneraey, K.L. Sublette, V.K. Bhupathiraju, J.D. Coates and R.M. Knapp, In E.T. Premuzic and A. Woodhead. (eds.), Microbial Enhancement of Oil Recovery-Recent Advances, Elsevier Science Publishing, BV. Amsterdam, 1993, pp. 363-371. G.E. Jenneman, P.D. Moffitt, G.A. Bala and R.H. Webb, SPE Prod. Facil., 63 (1999) 219. I. Davidova, M.S. Hicks, P.M. Fedorak and J.M. Suflita, J. Ind. Microbiol. Biotechnol., 27 (2001) 80. R.E. Eckford and P.M. Fedorak, J. Ind. Microbiol. Biotechnol., 29 (2002) 243. American Petroleum Institute, Recommended Practices for Oil and Gas Producing and Gas Processing Plant Operations Involving Hydrogen Sulfide, 2nd edition, API Recommended Practice 55, Washington DC, USA, 1995. P.G. Stecher, Hydrogen Sulfide Removal Processes, Noyes Data Corp., New Jersey, USA, 1972, p. 1. R. Cord-Ruwisch, W. Kleinitz and F. Widdel, J. Petrol. Techno!., 39 (1987) 97.

338

[19] R.N. Tuttle and R.D. Kane (eds.), H2S Corrosion in Oil and Gas Production - A Compilation of Classic Papers, National Association of Corrosion Engineers, Houston, Texas, USA, 1981. pp. 1018-1038. [20] A. Ikeda and M. Kowaka, Chem. Econ. Eng. Rev., 10 (1978) 12. [21] T.A. Bertness, G.V. Chilingarian and M. Al-Basson, In G.V. Chilingarian, J.O. Robertson, Jr. and S. Kumar (eds.), Surface Operations in Petroleum Production II, Elsevier Science Publishing Co. Inc., New York, USA, 1989, pp. 283-317. [22] I.B. Beech, In G. Bitton (ed.), Encyclopedia of Environmental Microbiology, Wiley and Sons, New York, USA, 2002, pp. 465-475. [23] W.P. Iverson and G.J. Olson, In R.M. Atlas (ed.), Petroleum Microbiology, Macmillan Publishing Company, New York, USA, 1984, pp. 633-683. [24] F.A. Giuliano (ed.), Introduction to Gas and Oil Technology, 3rd edition, Prentice Hall, Inc., Englewood Cliffs, New Jersey, USA, 1989. [25] A.G. Collins and C.C. Wright, In E.C. Donaldson, G.V. Chilingarian and T.F. Yen. (eds.), Enhanced Oil Recovery I: Fundamentals and Analysis, Elsevier Science Publishing Co. Inc., New York, USA, 1985, pp. 151-221. [26] C.C. Wright and G.V. Chilingarian, In G.V. Chilingarian, J.O. Robertson, Jr. and S. Kumar (eds.), Surface Operations in Petroleum Production II, Elsevier Science Publishing Co., Inc. New York, USA, 1989, pp. 319-372. [27] S.C. Rose, J.F. Buckwalter and R.J. Woodhall (eds.), The Design Engineering Aspects of Waterflooding, Society of Petroleum Engineers, Richardson, Texas, USA, 1989. [28] T.R. Jack, In F.T. Premuzic and A. Woodhead (eds.), Microbial Enhancement of Oil Recovery - Recent Advances, Elsevier Science Publishing, BV. Amsterdam, 1993, pp. 7-16. [29] J.M. Akagi, In L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995, pp. 89-112. [30] J.R. Postgate, The Sulphate-Reducing Bacteria, Cambridge University Press, Cambridge, UK. 1979. [31] J.R. Postgate, The Sulphate-Reducing Bacteria, 2nd edition, Cambridge University Press, Cambridge, UK, 1984. [32] J.M. Odom and R. Singleton, Jr. (eds.), The Sulfate-Reducing Bacteria: Contemporary Perspectives, Springer-Verlag, New York, USA, 1992. [33] L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995. [34] H.F. Castro, N.H. Williams and A. Orgam, FEMS Microbiol. Ecol., 31 (2000) 1. [35] H. Cypionka, In L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995, pp. 151-184. [36] N. Pfennig, F. Widdel and H.G. Triiper, In M.P. Starr, H. Stolp, H.G. Triiper, A. Balows and H.G. Schegel (eds.), The Prokaryotes, A Handbook of Habitats, Isolation, and Identification of Bacteria, Vol. 1, Springer-Verlag, Berlin, 1981, pp. 926-947. [37] L.L. Barton and F.A. Tomei, In L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995, pp. 1-32. [38] T.A. Hansen, In J.M. Odom and R. Singleton, Jr. (eds.), The Sulfate-Reducing Bacteria: Contemporary Perspectives, Springer-Verlag, New York, USA, 1993, pp. 21-40. [39] F. Widdel and R. Rabus, Curr. Opin. Biotechnol., 12 (2001) 259. [40] J. Heider, A.M. Spormann, H.R. Beller and F. Widdel, FEMS Microbiol. Rev., 22 (1999)459. [41] D. White, The Physiology and Biochemistry of Prokaryotes, Oxford University Press Inc., New York, USA, 1995.

339

[42] A. Azadpour, L.R. Brown and A.A. Vadie, J. Ind. Microbiol., 16 (1996) 263. [43] G.D. Faugue, In L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995, pp. 217-242. [44] D.E. Brink, I. Vance and D.C. White, Appl. Microbiol. Biotechnol., 42 (1994) 469. [45] J.T. Rosnes, T. Torsvik and T. Lien, Appl. Environ. Microbiol., 57 (1991) 2302. [46] M. Magot, B. Ollivier and B.K.C. Patel, Antonie van Leeuwenhoek, 77 (2000) 103. [47] J. Boivin, Mater. Perform., 34(2) (1995) 65. [48] T.R. Jack and D.W.S Westlake, In L.L. Barton (ed.), Sulfate-Reducing Bacteria, Plenum Publishing Corp., New York, USA, 1995, pp. 265-292. [49] S. Myhr, B.-L. Lilleb0, E. Sunde, J. Breeder, and T. Torsvik, Appl. Microbiol. Biotechnol., 58 (2002) 400. [50] J. Tiedje, In A.J.B. Zehnder (ed.), Biology of Anaerobic Microorganisms, Wiley, New York, USA, 1988, pp. 179-244. [51] A.J. Telang, S. Ebert, J.M. Foght, D.W.S. Westlake, G.E. Jenneman, D. Gevertz and G. Voordouw, Appl. Environ. Microbiol., 63 (1997) 1785. [52] S. Myhr and T. Torsvik, Int. J. Syst. Bacteriol., 50 (2000) 1611-1619. [53] R.G.L. McCready, W.D. Gould, and R.W. Barendregt, Can. J. Microbiol., 29 (1983) 231. [54] G.J. Mitchell, J.G. Jones and J.A. Cole, Arch. Microbiol., 144 (1986) 35. [55] H.-J. Seitz and H. Cypionka, Arch. Microbiol., 146 (1986) 63. [56] T. Dalsgaard and F. Bak, Appl. Environ. Microbiol., 60 (1994) 291. [57] K.L. Sublette and M.E. Woolsey, Biotechnol. Bioeng, 34 (1989) 565. [58] K.L. Sublette, MJ. Mclnerney, A.D. Montgomery and V. Bhupathiraju, In C.N. Alpers and D. W. Blowes (eds.), Environmental Geochemistry of Sulfide Oxidation, American Chemical Society, Washington, DC, USA, 1994, pp. 68-78. [59] M.J. Mclnerney, V.K. Bhupathiraju and K.L. Sublette, J. Ind. Microbiol., 11 (1992) 53. [60] M.J. Mclnerney, N.Q. Wofford and K.L. Sublette, Appl. Biochem. Biotechnol., 57/58 (1996) 933. [61] K.L. Sublette, D.E. Morse and K.T. Raterman, Presented at the 68th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers Houston, Texas, SPE paper 26396, Society of Petroleum Engineers, Richardson, Texas, USA, October 3-6, 1993. [62] D. Gevertz, A.J. Telang, G. Voordouw and G.E. Jenneman, Appl. Environ. Microbiol., 66(2000)2491. [63] F.A. Rainey, D.P. Kelly, E. Stackebrandt, J. Burghardt, A. Hiraishi, Y. Katayama and A.P. Wood, Int. J. Syst. Bacteriol., 49 (1999) 645. [64] W. Ludwig, G. Mittenhuber and C.G. Friedrich, Int. J. Syst. Bacteriol., 43 (1993) 363. [65] L.A. Robertson and J.G. Kuenen, J. Gen. Microbiol., 129 (1983) 2847. [66] W.W. Carothers and Y.K. Kharaka, Am. Assoc. Petrol. Geol. Bull, 62 (1978) 2441. [67] T. Barth, Appl. Geochem, 6 (1991) 1. [68] E.A. Greene, C. Hubert, M.. Nemati, G.E. Jenneman and G. Voordouw, Environ. Microbiol, 5 (2003) 607. [69] P.A. Loka Bharatm, S. Nair and D. Chandramohan. J. Mar. Biotechnol, 5 (1997) 172. [70] J.P. Adkins, L.A. Cornell and R.S. Tanner, Geomicrobiol. J, 10 (1992) 87. [71] D. Gevertz , G.E. Jennemen, S. Zimmerman and J. Stevens, In R. Bryant (ed.), Proceedings of the Fifth International Conference on Microbial Enhanced Oil Recovery and Related Biotechnology for Solving Environmental Problems, Richardson, Texas, USA, 1995, pp. 295-309. [72] A.J. Telang, G.E. Jenneman and G. Voordouw, Can. J. Microbiol, 45 (1999) 905.

340

[73] G. Voordouw, J.K. Voordouw, R.R. Karkhoff-Schweizer, P.M. Fedorak and D.W.S. Westlake, Appl. Environ. Microbiol., 57 (1991) 3070. [74] R.E. Eckford and P.M. Fedorak, J. Ind. Microbiol. BiotechnoL, 29 (2002) 83. [75] T. Thorstenson, G. Bodtker, B.-L.P Lillebo, T. Torsvik, E. Sunde and J. Beeder, Paper 02033 Proceedings of the NACE Expo 2002 Annual Conference and Exposition, Denver, Colorado, USA, April 7-11, 2002. [76] E.G. Beauchamp, J.T. Trevors and J.W. Paul, Adv. Soil Sci., 10 (1989) 113. [77] R.K. Thauer, K. Jungermann and K. Decker, Bacteriol. Rev., 41 (1977) 100. [78] G.E. Jenneman, M.J. Mclnerney and R.M, Knapp, Appl. Environ. Microbiol., 51 (1986) 1205. [79] G.E. Jenneman, A.D. Montgomery and M.J. Mclnerney, Appl. Environ. Microbiol., 51 (1986)776. [80] H.-E. Jacob, In J.R.. Norris and D.W. Ribbons (eds.), Methods in Microbiology, Vol. 2, Academic Press Inc., New York, USA, 1970, 93-123. [81] S. Moura, C. Bursakov and J.J.G. Moura, Anaerobes, 3 (1997) 279. [82] W.T. Carpenter, Water Works Sewerage, 79 (1932) 175. [83] W.A. Lawrance, Sewage Ind. Wastes, 22 (1950) 820. [84] R.A. Poduska and B.D. Anderson, J. Water Pollut. Control Fed., 53 (1981) 299. [85] K.L. Londry and J.M. Suflita, J. Ind. Microbiol. BiotechnoL, 22 (1999) 582. [86] M.A. Reinsel, J.T. Sears, P.S. Stewart and M.J. Mclnerney, J. Ind. Microbiol., 17 (1996) 128. [87] M. Wright, G.E. Jenneman and D. Gevertz, Proceedings of the 4th International Petroleum Environmental Conference: Environmental Issues and Solutions in Exploration, Production and Refining. San Antonio, Texas, USA, (on CD-ROM), 1997. [88] G.E. Jenneman, P.D. Moffitt, G.A. Bala and R.H. Webb, Paper SPE 38768, presented at the 1997 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, October 5-8, 1997. [89] M. Nemati, G.E. Jennenman and G. Voordouw, BiotechnoL Prog., 17 (2001) 852. [90] R.M. Lacatena, S. Lunetto, A. Robertiello, M. Marzorati and F. de Ferra, Presented at 2nd International Conference on Petroleum Biotechnology, Mexico City, November 5-7, 2003. [91] M. Nemati, G.E. Jennenman and G. Voordouw, BiotechnoL Bioeng., 74 (2001) 424. [92] B. Herbert, Lecture presented at Reservoir Microbiology Forum 9: Use of Nitrates to Control Bacterial Problems, September 9, 2003, London, UK.


341

Chapter 12

Regulation of toluene catabolic pathways and toluene efflux pump expression in bacteria of the genus Pseudomonas J.L. Ramos, E. Duque, M.T. Gallegos, A. Segura and S. Marques Estacion Experimental del Zaidin, CSIC, C / Profesor Albareda 1, 18008 Granada, Spain

1. TOLUENE EXTRUSION AND DEGRADATION PATHWAYS INFLUENCE SURVIVAL IN PSEUDOMONADS Aromatic hydrocarbons have been present in the environment for millions of years since they are the product of the natural pyrolysis of organic material [1] and are widely distributed in natural environments. One ring aromatic compounds such as benzene, xylenes, ethylbenzene and toluene, which have a logPow (logarithm of its partition coefficient in «-octanol and water) between 2.5 and 3.5, are toxic for microorganisms and other living cells because they partition preferentially in the cytoplasmic membrane, disorganizing its structure and impairing vital functions [2]. The toxicity of these compounds depends not only on the inherent toxicity of the solvent but also on the intrinsic tolerance of the species and strains. Because living organisms have been in contact with these chemicals through long evolutionary periods of time, it is not surprising that microbes have developed the capability to degrade them. Many of the aromatic compound-degrading organisms are bacteria that belong to the Pseudomonadaceae. All Pseudomonas strains that use toluene as a carbon source have a series of mechanisms that allow them to cope with the stress imposed by toluene itself. Nonetheless, most Pseudomonas strains are highly sensitive to aromatic hydrocarbons such as toluene (logPow 2.5), styrene (logPow 3.6) orp-xylene (logPow 3.2); however, independent laboratories have isolated Pseudomonas putida strains tolerant to these toxic compounds [3-7] . A common theme in toluene tolerance in Pseudomonas is the change in cis/trans isomerization of unsaturated fatty acids [8]. The increase in trans isomers (which are directly synthesized from the cis isomers with no shift in the

342

position of the double bond) causes rigidification of the membrane to counteract the increase in membrane fluidity caused by the organic solvent [9-12]. These changes probably represent a first response that allows the cells to prepare for the de novo biosynthesis of other components involved in tolerance towards organic solvents. One such response was thought to be the metabolism of organic solvents, but some of the tolerant strains are not able to degrade toluene. For example, P. putida S12 is a toluene-tolerant strain that cannot degrade it [13], and a P. putida DOT-TIE mutant deficient in the tod (toluenedioxygenase) degradation pathway is as tolerant as the wild type to a sudden toluene shock [14]. These two observations suggest that degradation of the toxic compound is not a key factor in solvent tolerance. All the efflux pumps for organic solvents identified so far in gramnegative bacteria belong to the Resistance-Nodulation-Cell Division (RND) family. The functioning of these efflux pumps seems to be coupled to the proton motive-force via the TonB system, although the intimate mechanism of energy transfer remains elusive [15, 16]. Bacterial RND efflux pumps work together with a membrane fusion protein (MFP) and an outer membrane protein (OMP). These three components form a structure that expand both the inner and outer membranes [17-19]. The efflux pump transporter AcrB of the E. coli RND multidrug efflux system AcrAB-TolC was recently crystalized. The AcrB component is as a trimer with a 50-A transmembrane region and a 70-A part located in the periplasm that is thought to be involved in substrate recognition [18, 20]. The crystal structure of the outer membrane protein TolC -which forms a trimeric channel that penetrates the periplasm and contacts the efflux pump transporter has also been reported [17]. Finally, a lipoprotein anchored to the inner membrane which expands into the periplasmic space may serve as a bracket for the other two components [19, 21]. This structural organization allows substrates to be extruded into the external medium by passing the periplasmic space [17, 18, 22]. In spite of its toxicity and thermodynamic stability, toluene can be degraded by many microbes using a common strategy to weaken the aromatic ring prior to its cleavage: the introduction of two hydroxyl groups that destabilize the chemically stable resonant structure. However, bacteria have developed different molecular mechanisms to produce this dihydroxylated compound. So far, five aerobic pathways have been described for the bacterial degradation of toluene, all of them leading to catechol, methylcatechols or protocatechuate: the so-called TOL, TOD, TMO, TOM, and TBU pathways. However, only three of them, the TOL, TOD and TMO pathways, have been found in Pseudomonas species. The enzymes that carry out the first reaction, i.e. the direct insertion of one or two oxygen atoms in the toluene molecule, largely determine the pathway that is followed for degradation. The key steps involved in the pathways are briefly described below, and they are summarized in Fig. 1.

343

Fig. 1. Pathways for the aerobic degradation of toluene.

1.1. Toluene degradation pathways The TOL pathway, coded by the archetypical plasmid pWWO [23, 24], is a very well characterized pathway for toluene degradation from a biochemical and genetic point of view. This pathway is composed of two segments, an upper and a lower pathway. Through the upper pathway, the methyl group of toluene is sequentially oxidized to render benzoate (Fig. 1). The first enzyme of this upper pathway is a toluene monooxygenase that oxidizes the methyl group of toluene to yield benzyl alcohol. Subsequent oxidation of the side chain is accomplished in two steps: first benzyl alcohol dehydrogenase renders benzaldehyde, which is further oxidized to benzoate by a benzaldehyde dehydrogenase. It is of interest to note that the enzymes of the upper pathway also accept as substrates the corresponding compounds substituted with a methyl group at the meta and/or para position and ethyl group at the meta position. These compounds are, therefore, oxidized to 3-, 4-methylbenzoate, 3,4dimethylbenzoate and 3-ethylbenzoate, respectively. The aromatic carboxylic acids are then further metabolized through the meto-cleavage pathway, in which the benzoate, alkylbenzoate(s) is(are) then oxidized and decarboxylated to produce the corresponding catechol(s) (Fig. 1).

344

These compounds undergo meta fission to yield 2-hydroxymuconic acid semialdehyde or the corresponding alkyl derivatives. Metabolism of the semialdehydes occurs via a branched pathway that rejoins at a common intermediate, 2-oxopent-4-enoate. The semialdehyde produced from w-toluate is hydrolyzed, whereas the semialdehyde from benzoate and p-methylbenzoate is metabolized via the oxalocrotonate branch, which involves at least three enzymatic steps [25]. 2-Oxopent-4-enoate is further converted in 2-oxo-4hydroxypentonate, which eventually renders Krebs cycle intermediates. For a more detailed description of the characteristics of the pathway enzymes, the reader is referred to earlier reviews and original papers [24, 26-30]. A second degradation pathway found in Pseudomonas is the so-called TOD pathway, which was first described in P. putida strain Fl [31]. In this pathway, probably the best known in terms of the biochemistry involved, the first step is carried out by a three-component enzyme complex, the toluene 2,3dioxygenase (TOD), which renders czs-toluene dihydrodiol (Fig. 1). This compound then undergoes dehydrogenation to yield 3-methylcatechol. Ethylbenzene is also a substrate of this enzyme being converted into 3ethylcatechol. The alkylcatechols are then the substrates for ring fission in the meta position in a set of reactions similar to those described for the TOL metacleavage pathway. Finally, a third pathway was described in P. mendocina KR1, where the first step in toluene metabolism is carried out by the toluene-4-monooxygenase (TMO), which hydroxylates the aromatic ring in the para position to render pcresol (Fig. 1 and [32, 33]). In the subsequent steps the methyl group is transformed by a methyl hydrolase, first rendering the alcohol and then the aldehyde derivative, which is finally oxidized to /?-hydroxybenzoate by a dehydrogenase. The ring is further oxidized to 3,4-dihydroxybenzoate, which is the substrate for ring cleavage in the ortho position to enter the P-ketoadipate pathway. As mentioned above, two additional pathways for toluene degradation with different initial reactions have been described in strains that were originally considered Pseudomonas: the so-called toluene-3-monooxygenase pathway (TBU) of Ralstonia picketti, where toluene was proposed to be first oxidized to ra-cresol and then to 3-methylcatechol [34]. However, recent work in Tom Wood's laboratory has suggested that toluene-3-monooxygenase indeed functions as a hydroxilating enzyme at the para position, and that 90% of toluene is oxidized to p-cresol, which is subsequently oxidized to yield 4methylcatechol (Fig. 1 and [35]). The second pathway, not present in Pseudomonas, is the one known as the toluene 2-monooxygenase pathway (TOM) of Burkholderia cepacia where the first oxygen is inserted in the ortho position to render o-cresol, which is then oxidized to 3-methylcatechol (Fig. 1). In addition, another pathway for the degradation of toluene was recently

345

described in P. stutzeri OX1, a strain able to degrade o-xylene through an initial step that involves two successive monooxygenations of the aromatic ring carried out by the same enzyme, toluene o-xylene monooxygenase (ToMO) [36-38]. This enzyme is interesting because of its broad substrate specificity and its relaxed regioselectivity, which make it able to hydroxylate more than one position of an aromatic substrate [39]. A common feature to all toluene pathways from different bacteria is that the genes involved in the different reactions are organized as operons, which are either independent for the different segments of the pathways (i.e., the TOL pathway) or transcribed as a single unit (i.e. the tod operon). In all cases, these pathways are under the control of regulatory mechanisms, which are ultimately modulated by toluene or intermediate substrates. Below we review the regulatory networks controlling the expression of the different pathways below and the mechanisms developed to regulate toluene efflux pumps.

2. REGULATION OF THE TOL PATHWAY In P. putida mt-2, the genetic information that determines growth on toluene, xylenes and other alkyl derivatives is encoded by the xyl operons of the TOL plasmid pWWO [24]. The xyl genes are organized in four transcriptional units: the upper and the meta operons and the xylS and xylR genes (Fig. 2). The upper operon xylUWCAMBN codes for the enzymes necessary for the oxidation of toluene to benzoate, whereas the meta-operon xylXYZLTEGFJQKIHencodes the enzymes for the oxidation of benzoate, the ensuing ring cleavage and the degradation to TCA cycle intermediates. The xylS and xylR genes, which are transcribed divergently, are located close to the meta operon 3' end, and their products regulate the expression of the meta and upper catabolic operons, respectively. This pathway is no doubt the most extensively characterized regulatory system among the aromatic degradation pathways, and, as a whole, the TOL regulatory network can be considered a paradigm of integrated transcriptional regulation in prokaryotes: there are two o54-dependent promoters, each with unique features, one regulator belonging to the NtrC, one to the AraC family of regulators, a o32/o38-dependent promoter, and several o70-dependent promoters, all of them under superimposed global control. 2.1. Overview of the regulatory network A scheme of the regulatory network that operates in the TOL pathway is presented in Fig. 2. The model summarizes experimental evidence collected since the 1970s in different laboratories to explain the expression of the pathway enzymes in the presence of toluene, benzoate or their derivatives. Two different regulatory circuits operate, depending on the nature of the aromatic compound present in the culture medium [40]. When cells are growing in the absence of

346

any aromatic substrate, the xylS gene is expressed at low levels from the a70dependent promoter Ps2, ensuring the presence of basal levels of XylS protein. This protein as such is not able to activate transcription. When a substrate of the meta pathway, e.g. 3-methylbenzoate, is present in the growth medium, the XylS protein interacts with it and becomes active to promote transcription from the Pm promoter, which controls expression of the meta pathway. Expression from Pm requires RNA polymerase with either a32 in the early exponential phase or a38 thereafter. The XylR protein, which regulates its own transcription from two o70-dependent promoters, is synthesized in sufficient amounts under all growth conditions. When a substrate of the upper pathway, e.g. toluene, is present in the culture medium, the binding of this effector to the protein triggers a series of molecular events that result in the activation of transcription from two o5 dependent promoters: Psl for the xylS gene, and Pu, which drives expression of the upper pathway. This latter activation requires the integration host factor (IHF). As a consequence of Psl activation, the XylS protein is overproduced, and even in the absence of a meta pathway effector, transcription from Pm occurs. The current knowledge of the molecular biology of each step on the regulatory pathway is reviewed in detail below.

Fig. 2. The TOL pathway regulatory network. Elliptical boxes indicate the inactive form of the regulatory proteins. Shaded square boxes indicate the active form of the regulatory proteins. Lines represent the connections between regulatory proteins and promoters, where (+) is activation of transcription and (-) is inhibition of transcription; GR, global regulation. The dotted line indicates transcription activation of overproduced XylS in the absence of effector. The sigma factor(s) involved in transcription initiation are indicated above each promoter. Aromatic substrates of the pathways that act as effectors of the regulatory proteins are indicated. The regulatory circuits are explained in the text.

347

2.2. Regulation of the meta pathway Two main players are responsible for the expression of the meta pathway: the tandem Pm promoter/RNA polymerase, and the tandem XylS/effector molecule. Extensive genetic data have been obtained to determine the architecture of Pm promoter and the mechanisms through which the regulator becomes active to promote transcription from Pm. As stated above, basal levels of the XylS protein are guaranteed under any growth conditions by the activity of the constitutive o70-dependent promoter Ps2 (Fig. 2). 2.2.1. The XylS regulator XylS belongs to the AraC/XylS family of transcriptional regulators, which includes at least 284 different proteins [41-44]. Members of this family present two domains: a 100 amino acid conserved domain involved in DNA binding (the C-terminal domain in most of the proteins of the family, including XylS), and a nonconserved domain (the N-terminal domain in XylS) involved in effector binding and dimerization. Interactions between XylS and its effector have been studied by analyzing the ability of the protein to activate transcription in the presence of a wide range of substituted benzoates, as well as by selecting XylS mutants with altered effector specificity [16, 43, 45, 46]. Recognition of ring substituents strongly depends on the position and nature of the chemical substituent, with meta being the most permissive position in the aromatic ring (-CH3, -CH2H5, and -OCH3 groups and F, Cl, Br and I atoms are permissible substituents), whereas positions ortho and para pose some restricitions to substituents (-CH3 and F and Cl atoms are allowed but not -C2H5 and I atoms) [45]. Although disubstitutions involving positions o- and m-, and m- and p-, are permissible, other combinations are usually non permissible, which suggests that interactions between the effector and the regulator are nonsymmetrical. Ramos et al. [48, 49] and Michan et al. [46] isolated and sequenced a series of mutant regulators able to recognize substituted benzoate effectors that are not recognized by the wild-type regulator. Key residues clustered in two noncontiguous segments in the N-teminus end of XylS. Mutations were found to be clustered at positions 37-45, 88-92, 151-155, and around residues 256 and 288. These finding suggest that the recognition pocket for XylS effectors may be composed of two or more noncontiguous segments of its primary sequence. Arg-41 seems to be a critical residue for interaction(s) with effectors, as changes in this position result in many different phenotypes. For example, XylSArg41Gly is a mutant regulator whose ability to recognize o- and p-methylbenzoate was lost, although it retained its capacity to be activated by w-methylbenzoate. Substitution of Arg41 by Leu resulted in a mutant that was unable to respond to benzoate effectors [46]. Therefore,

348

transition from the inactive to the active form may be mediated by effector binding. Recent studies further pinpointed residues Asp 137 and His 153 as crucial for interactions with the effector molecule [50]. In addition to influencing effector specificity, these two residues were shown to contact specific residues in the RNA polymerase a subunit carboxy terminal domain (aCTD) [50] . XylS mutants such as XylSArg41Cys, XylSPro37Gly XylSGly44Ser, XylSSer229Ile, XylSAsp274Val, and XylSAsp274Glu mediated transcription from Pm in the absence of effectors [46, 47]. These results support the hypothesis that XylS exists in vivo in a dynamic equilibrium between an inactive and an active form, with respect to transcriptional stimulation. Within the family, some regulators such as MarA are present in solution as monomers, whereas most of the members of the family are found as dimers [51-54]. XylS is likely active as a dimer and in vivo and in vitro assays have shown that Leu 193 and Leul94 in XylS play a crucial role in dimerization [55]. It is predicted that the DNA binding domain of XylS consists of seven ahelix units which fold to assemble two helix-turn-helix (HTH) motifs that interact with two neighboring major grooves on one face of the target DNA. Involvement of the XylS C-terminal domain in DNA binding was first predicted after the finding of mutations in this domain that rendered mutant regulators able to promote high transcription levels in the absence of effectors [47, 48]. Mutation analysis of the predicted conserved positions of the HTH motifs of XylS showed that the most conserved positions in the family seem to be essential to preserve the structure of this domain [56]. Deletion of the 209 Nterminal residues of XylS rendered a C-terminal domain-protein able to bind Pm promoter and, when overproduced, able to activate transcription in vivo to levels similar to those in the wild type protein. However, activity was clearly reduced when the C-terminal fragment was synthesized at physiological levels. As expected, the truncated protein was not responsive to effector-mediated control [57]. 2.2.2. The Pm promoter XylS-mediated transcription activation from Pm requires a DNA fragment extending to position -70 upstream from the transcription start site. The DNA in this region exhibits a 40° bend centered between positions -41 and -46 [58]. The XylS binding site in the Pm promoter was first defined through site-directed mutagenesis [59-63] and further confirmed by in vitro and in vivo footprint assays [60, 64, 65]. The XylS binding site in Pm consists of two directed repeats (5'-TGCAN6GGNTA-3') spanning positions -34 to -68, and overlapping the RNA polymerase biding site by 1 bp [58, 60]. This overlap with the RNA polymerase binding site is also observed in several other members of the family [66-68].

349

In vivo transcription from Pm is mediated by two different RNA polymerases, depending on the growth phase. In the early exponential phase, RNA polymerase with a32 is necessary for transcription, whereas in late exponential growth phase and in stationary phases, a38 is required [65, 69, 70]. Despite the alternation in RNA polymerase, the same transcription start site is detected along the growth curve, suggesting that the same promoter is used by both forms of RNA polymerase. Transcription of o32-dependent promoters requires the stabilization of this sigma factor, which takes place through a series of events known as heat-shock response [71]. In the TOL pathway regulatory network, aromatic effectors are required not only because of their direct role in XylS activation to promote transcription, but also to trigger the heat-shock response and provide the appropriate RNA polymerase for transcription in the exponential phase [70]. The direct involvement of the two sigma subunits in Pm transcription was further supported by the finding that a mutant Pm promoter with an altered XylS binding site, combined with the mutant regulator XylSGly44Ser, was able to overcome the requirement of a38 for transcription in the stationary phase [69]. Moreover, the mutants XylSAspl37Glu and XylSHisl53Gln were able to stimulate transcription from Pm in the absence of a38 [50]. 2.3. Regulation of the upper pathway In the presence of toluene or a substrate of the upper pathway, P. putida ensures the coordinated expression of the two catabolic operons, so that the aromatic compound is totally degraded to TCA intermediates. The key regulator in this process is XylR, which is responsible for the coordinated expression of the o54-dependent promoters Pu and Psl. Pu drives transcription of the upper pathway and Psl increases the synthesis of XylS, responsible for meta pathway expression (Fig. 2). 2.3.1. The XylR regulator XylR protein belongs to the NtrC family of enhancer-binding proteins (EBP) [72-74]. It contains the four distinctive domains of this family: i) An Nterminal A-domain responsible for signal reception, i.e., interaction with the effector molecule (see below), ii) the A-domain is linked to the central domain, called domain C by the short B-domain (Q-linker), iii) the C-domain is involved in ATP binding and hydrolysis, and plays a major role in the isomerization of the o54-dependent promoters from close to open complexes, and iv) the Cterminal D-domain contains the HTH motif for DNA binding. XylR is activated by aromatic compounds with a wide variety of substitutions such as alkyl groups of different length or oxidized intermediates of the toluene methyl group, such as benzyl alcohol, benzaldehyde and derivatives [75, 76]. Early evidence indicated that the A-domain was the signal

350

receptor from the environment and the direct sensor of the aromatic molecule. This was surmised from the ability of the protein to activate transcription from the Pu promoter in the presence of a wide range of toluene derivatives, and by experiments with XylR mutants with altered effector specificity [75, 76, 77, 78]. These data, obtained in the heterologous host E. coli, led to the conclusion that XylR was directly activated via interaction with the effector. XylR is closely related to the DmpR regulator for phenol degradation in Pseudomonas sp. CF600, which recognizes phenol and derivatives, but not toluene, as an effector [79]. Further evidence for the direct interaction of the A-domain of these proteins with the effector molecule came from the construction of a chimeric protein in which the receptor domain of DmpR was replaced by the corresponding domain of XylR, resulting in a hybrid regulator that responded to toluene for activation of the Vo promoter of the phenol degradation pathway [80]. DNA shuffling assays to create hybrid A-domains between DmpR and XylR confirmed that the residues 110 to 186 of both proteins were responsible for the effector profile of these regulators [80]. The A-domain operates as an intramolecular repressor of the central activating domain of the protein [81, 82]. In fact, a XylR derivative in which the A domain has been deleted is able to activate Pu in the absence of an aromatic effector. The truncated derivative of XylR depleted of the A domain and therefore unable to respond to effector-dependent modulation showed intrinsic ATP binding and hydrolysis activity, located in the central activation domain (C-domain). This activity was stimulated by the presence of a DNA fragment containing the native XylR binding site in Pu (UAS) [83]. Furthermore, binding of ATP to this truncated protein alone was able to induce conformational changes in the protein. Initially, a cyclic model to explain XylR activation of Pu was proposed by Perez-Martin and de Lorenzo [83], according to which ATP binding to the XylR central domain led to multimerization of the regulator bound to its UAS in Pu, followed by ATP hydrolysis. This in turn triggered a54dependent transcription initiation in Pu, allowing the system to return to its initial disassembled state [83]. Recently, Shingler and co-workers studied the analogous regulator DmpR, and suggested an alternative mechanism to explain effector-dependent activation of a54-dependent promoters. According to their model, DmpR dimers are activated after binding of the effector molecule to the A domain, followed by a conformational change that allows ATP binding to the central domain and oligomerization to a hexameric conformation, probably required to promote transcription initiation. Finally, ATP hydrolysis leads to dissociation of the hexameric structure and dissociation of the effector [83]. 2.3.2. The Pu promoter Pu promoter belongs to the class of promoters dependent on the alternative sigma factor o54 (Fig. 2), and shows the typical architectural

351

organization typical of the promoters of this class. A -12/-24 sequence is responsible for the recognition of o54-RNAP [84, 85], and upstream activator sequences (UAS) ensure XylR binding between positions -120 and -175 [86, 87]. An IHF binding site between these two elements (positions -52 to -79) [87] is required to bring the regulator into contact with the RNA polymerase by looping out the intermediate sequence. Direct evidence that IHF causes Pu to form an open bend was obtained by atomic force microscopy, where an angle of 123° was measured between the UAS and the RNA polymerase binding site [88]. However, Bertoni et al. [89] recently found that a second upstream element, reminiscent of the so-called a-CTD-binding UP elements of o70dependent promoters, was important in Pu recognition by a5 -RNAP, and that IHF-binding played an additional role in Pu. This role consisted of a5 -RNAP recruitment to its promoter, determined by the correct positioning of the UP-like element with respect to the -12/-24 binding site after IHF-dependent DNA bending [39]. That IHF-mediated o54-RNAP recruitment to Pu was reproduced in vitro with the XylR regulator acting from solution, i.e., in the absence of UAS [90]. These authors have shown that RNA polymerase binding is an important rate-limiting step in Pu activation, that could become crucial when the enzyme is present at a low concentration [90]. 2.4. Expression of the regulatory genes The level of the XylR and XylS proteins is finely modulated in vivo. This fine regulation takes place in the 300-bp intergenic region between the xylR and the xylS genes, which contains four promoters: the two a70-dependent tandem promoters of xylR, Prl and Pr2, divergent from the two that drive transcription of xylS, the a54-dependent promoter Psl and the a70-dependent promoter Ps2. The binding sites for the different regulatory proteins in this short region totally or partially overlap; thus XylR UASs in Psl partly cover the two RNA polymerase binding sites of Prl and Pr2. In addition, two sequences with different affinity for IHF are found which overlap the Psl -12/-24 RNA polymerase binding site and one of the UAS [91-93]. As a consequence, the levels of expression normally observed in the wild-type strain for each promoter are far below maximum values, suggesting the involvement of repressive element(s) in the maintenance of appropriate levels of expression. In fact, as expected from the promoter architecture described above, XylR strongly represses its own synthesis [75, 89, 94-97]. Activation of the Psl promoter and autoregulation of XylR expression seem to be the consequence of the binding of XylR to the UASs for Psl that overlap the Prl and Pr2 promoters. This is in agreement with the finding that XylR is consistently bound to target sequences [75]. Thus, xylR promoters may be subjected to two levels of repression depending on the mechanism of XylR activation discussed above: an ATP-independent level resulting from non-cooperative interaction of

352

nonactivated XylR with the Psl UAS [94], and an ATP-dependent repression level resulting from the cooperative oligomerization of activated XylR at the UAS in Psl [89]. As soon as the protein is activated and the UAS is strongly bound by the regulator, XylR expression is minimized, thus limiting the period of time during which the Ps 1 and Pu promoters of the TOL plasmid are in an activated state [89]. The role of IHF in Psl expression deserves special attention. Analysis of Psl activity in isogenic IHF-plus and minus backgrounds showed that in the presence of toluene, the highest levels of expression were achieved in the absence of IHF [97]. This may reflect a better access of either XylR to its binding site or of o54-RNA polymerase to the -12/-24 region of Psl, or both. On the other hand, it may be the consequence of structural hindrance, as the DNA bending induced by IHF bound to two sites may give rise to a highly ordered structure that restricts the access of regulatory proteins to the corresponding promoters. The high level of expression from Psl in the IHF-minus background in the presence of effectors contrasts with the diminished expression from the TOL plasmid Pu promoter for the upper pathway in an IHF-deficient background. The most noticeable difference between the two promoters is the position of the IHF binding site, which in Pu lies between the UASs and the 12/-24 box. In addition to affecting Psl expression, the close proximity of the regulatory sequences in the intergenic region results is a high expression level from Ps2 in the absence of a54, i.e., when RNA polymerase is unable to bind to the Psl promoter [97]. In general, the physiological consequence of this organization is that in the absence of any effector in the culture medium, Ps2, Prl and Pr2 promoters are slightly repressed. In the presence of toluene, activation of Psl causes a stronger repression of both xylR promoters. As a result, the level of XylR decreases at approximately 30 monomers per cell [98], which are apparently sufficient to promote high expression of both xylS and the upper pathway. Under these conditions the XylS protein is overproduced, which allows induction of expression from the Pm promoter even in the absence of meta pathway substrates. Therefore in the presence of toluene or a substituted derivative, both the upper and the meta pathways are coordinately expressed to optimize total degradation of the aromatic (Fig. 2). 2.5. Integration in the bacterial metabolism The expression of the TOL pathways is tightly regulated according to the carbon sources available for growth [99-104]. The regulation is exerted mainly at the level of the two o54-dependent promoters Pu and Psl, and was first observed as a delay in the induction of expression from these promoters when cells were induced in a rich complex medium [102, 103]. Because both promoters were silent in this medium during rapid exponential growth, and

353

expression appeared only at the end of the exponential phase, this behavior was named exponential silencing [105]. However, it is worth noting that exponential silencing is only observed in rich medium; in defined minimal media with succinate (for example) as a carbon source, expression of both Psl and Pu is observed immediately after induction [102, 103, 106]. Growth rate as a determinant of Pu and Psl expression was ruled out through a series of continuous culture experiments that compared different growth rates controlled by different limiting substrates. The results led to the conclusion that repressive conditions correlated with a high energy status of the cells [99, 100]. In other words, in all conditions tested where excess carbon was available, the system was repressed. However, when oxygen was the growth-limiting substrate, a situation where carbon was also present in excess, a certain degree of derepression was observed although activity never reached maximum values (Fig. 3).

Fig. 3. Integration of cell signals that lead to modulation of the expression of the Pm and Pu promoters.

354

In minimal medium batch cultures amended with casamino acids, carbon sources such as glucose, gluconate or a-ketoglutarate expression from Pu was inhibited. In general this phenomenon reflects the preferential and sequential use of the different carbon sources present in a mixture; hence it could be considered a typical case of catabolite repression. However, catabolite repression in Pseudomonas seems to be exerted through mechanisms that differ greatly from the classical CRP-dependent phenomenon observed in E. coli. Similar physiological regulations of other catabolic operons has been observed in Pseudomonas [79, 107]. The molecular basis for the observed repression of Pu and Ps 1 expression remains unknown. Several alternatives have been envisaged and the current picture is compatible with the partial involvement of different systems in the global regulation response. Originally, the phenomenon known as exponential silencing was shown to be due neither to a late activation of XylR by the aromatic effector nor to changes in the intracellular levels of IHF during growth [105]. However, recent findings obtained with in vivo UV laser footprint technology have shown that IHF occupancy of its target site in Pu increases upon entry into the stationary phase, in parallel with an increase in IHF concentration in the cell. Therefore, this could explain the increase in Pu activity with growth phase in batch cultures. Nevertheless, these results do not preclude the integration of physiological repressive signals through additional mechanisms [108]. The a54 factor of RNA polymerase has also been considered a possible target of global regulation. Overproduction of the a54 factor allowed Pu to partially overcome exponential silencing, although not carbon source-dependent repression [105]. Because a54 protein levels remained approximately constant during growth under physiological conditions [109], exponential silencing of Pu may be caused ultimately by changes in the activity of the sigma factor itself. The ATP-dependent physiological protease FtsH, a member of the so-called AAA family of ATPases responsible for the stability of various transcription factors such as a32 [110], has been shown to play a key role in Pu expression. FtsH is required for XylR-mediated Pu transcription in a process that is not related to XylR or IHF, but which is rather exerted through a mechanism that involves the loss of a54 activity (Fig. 3). In fact, overproduction of a54 restored about 60% of Pu activity in the absence of FtsH [111]. Furthermore, the overproduction of FtsH partially relieved exponential silencing of Pu expression. In this connection, the target of FtsH activity seems to be an additional factor that downregulates a54 post-translationally, or that hinders the contacts of o54RNAP with the promoter or with the activator. Interestingly, E. coli FtsH levels are controlled in response to physiological signals and its proteolytic function is stimulated by the proton-motive force [112].

355

Finally, exponential silencing and carbon source-dependent repression can be distinguished at a genetic level. The rpoN gene, which codes for the o54 sigma factor required for Pu activity, is the first gene of an operon found in gram-negative bacteria that includes 4 additional ORFs. Two genes of this cluster, ptsN and ptsO, code for two proteins, IIANtr and NPr respectively that show similarity to phosphotransferases belonging to the phosphoenol pyruvate: sugar phosphotransferase system (PTS) family. To understand the putative role of the ORFs in this cluster in the global control of Pu, knock-out mutants were generated and analyzed. A mutant in the ptsN gene (which encodes IIANtr) relieves C source inhibition, but not the exponential silencing of Pu [113]. The ptsO gene together with ptsN operates in Pu regulation, where phosphorylation of the pteOencoded protein NPr is necessary for the normal response of Pu to glucose [114, 115]. NPr probably modulates IIANtr activity, promoting its dephosphorylation. This increases the concentration of unphosphorylated IIANtr, and as a result inhibition of Pu disappears. Interestingly, a site-directed ptsN mutant in the conserved phospho-acceptor His-68 residue made Pu unresponsive to the presence or absence of glucose, thus supporting the notion that phosphorylation of IIANtr mediates the C source inhibition of the promoter [116]. The observations reported above suggest that the global regulation of TOL pathway expression responds to several complex mechanisms that act at the level of o54-dependent promoters (Fig. 3). 3. REGULATION OF THE TOD PATHWAY Pseudomonas putida strains Fl and DOT-TIE use toluene and ethylbenzene as the sole carbon source [6, 117, 118]. The catabolic genes for the complete conversion of toluene/ethylbenzene to TCA cycle intermediates are clustered in a single unit, the tod operon, as todXFClC2BADEGIH [14, 119-123] and these genes are coordinately induced by toluene/ethylbenzene (Fig. 4) [12, 124]. Wang and coworkers [121] and Mosqueda and coworkers [14] identified a single promoter upstream from the todX gene, whose -10 and -35 regions showed homology with P. putida o70-dependent promoters [125]. Expression of the tod catabolic operon is regulated by todST gene products, which are located as a separate transcriptional unit downstream of todH, the last gene of the operon. Translational coupling between todS and todT ensures the balanced transcription of both genes [119]. TodS and TodT proteins belong to the family of two-component signal transduction systems. The regulation mechanism of two-component control systems is based on a histidine-aspartate phosphorelay circuit working between the two components. One of them is a sensor that autophosphorylates in response to an external signal, and the other one is the socalled response regulator, which receives the phosphate from the former and

356

activates transcription. TodS, the 108-kDa sensor protein, is a member of the hybrid class of histidine kinases and possesses multiple protein domains. The N terminus of TodS contains a motif characteristic of the basic region leucine zirjper (bZIP), that consists of a region with several basic residues which probably contact DNA, and an adjacent region containing a heptad repeat of leucine, the leucine zipper. Indirect evidence suggests that in TodS, the leucine zipper mediates dimerization, which is required for DNA binding [119]. A duplicated histidine kinase motif, each element of which is characterized by five short sequence blocks that are highly conserved, flanks a receiver domain of the response regulator located at the center of the protein adjacent to one set of a PAS domain, known as a signal sensor and found in various redox, light and hydrocarbon sensor proteins [126, 127]. The todT gene encodes a 206-residue protein with an estimated mass of 23 kDa. Analysis of the amino acid sequence of TodT shows significant similarity with response regulators of two-component signal transduction systems [128, 129]. TodT consists of an N-terminal receiver domain to accept the phosphoryl group from TodS, and a helix-turn-helix (HTH) DNA-binding domain. The TodT protein was shown to specifically bind to the todX promoter region at a 6-bp inverted repeat located 105 bp upstream from the transcription start site, and known as the todbox [119].

Fig. 4. Organization of the tod genes and its regulatory circuit. Top. The tod genes are organized in two operons expressed from sigma-70 dependent promoters upstream from todX and todS. The TodS protein (O) is synthesized in an active form that in the presence of toluene phosphorylates TodT ( • ) , which functions as the activator of the todX...

357

This two-component signal transduction system positively regulates the tod operon [119]. TodS is predicted to function as a hybrid kinase that uses ATP to autophosphorylate a specific histidine residue in response to toluene. TodT as a response regulator probably receives the phosphate at a conserved aspartate (Asp-56) and then mediates transcription activation at the todX promoter. Although biochemical evidence supporting this model is consistent [119], in vivo studies confirming the role of each member of the tandem are scarce. Furthermore, the physiological behavior of this system under different growth conditions and the precise mechanism through which the presence of toluene triggers changes in tod gene expression in vivo are not completely understood. On the other hand, the role of TodS as a sensor for directly detecting inducers has not been clearly demonstrated, nor has the role of these proteins in selfregulation been clarified. Expression from the todX promoter occurred in response to toluene, ethylbenzene, styrene, xylenes and other aromatic hydrocarbons, although the greatest level of expression was obtained with toluene. Expression from the todS gene was constitutive regardless of which aromatic was tested [14]. It is interesting to note that both TodS and TodT proteins are required for chemotaxis to aromatic hydrocarbons in P. putida Fl [130]. This observation indicates that both catabolism and chemotaxis are coordinately regulated at the transcriptional level. 4. REGULATION OF THE TOLUENE-4-MONOOXYGENASE PATHWAY Genes involved in toluene degradation in Pseudomonas mendocina KR1 are organized in an unusual manner: the five proteins that make up the multicomponent enzyme toluene-4-monooxygenase, which carries out the primary oxidation of toluene to />-cresol (Fig. 5), are coded by the tmoXABCDEF gene cluster [33, 131], where tmoX is homologous to the outer membrane protein that codes todX from P. putida Fl [132]. Complete toluene oxidation to TCA cycle intermediates requires another operon, pcuCAXB for pcresol oxidation to ^-hydroxybenzoate, and the pobA gene, which presents two alleles, pobAl and pobA2, for the further transformation of this compound into protocatechuate. The pcuCAXB and pobA genes are not clustered with the tmo genes on the P. mendocina chromosome. These operons and the pobA genes are also independently regulated [133]. The regulatory elements involved in the toluene degradation pathway in Pseudomonas mendocina KR1 are summarized in Fig. 5.

358

Fig. 5. Regulation of toluene degradation in P. mendocina KR1. The upper part of the figure summarizes the oxidation of toluene to protocatechuate and the genes needed for each reaction. The lower part shows a scheme of the cluster of tmolpculpob genes in this strain. tmoABCDEF, toluene-4-monooxygenase genes; c, putative cytochrome c gene; pcuRCAXB, /?-cresol utilization genes; pobRl and pobAl, regulator and p-hydroxybenzoate hydroxylase, respectively; tmoST, two-component signal transduction system; p-HBOH, />-hydroxybenzyl alcohol; />-HBHO, p-hydroxybenzyl aldehyde; p-HBA, p-hydroxybenzoate; PCA, protocatechuate. Reproduced with permission from [132].

4.1. Regulation of tmo operon expression The transcription initiation point from the tmo operon has been mapped and the sequence upstream has revealed strong identity with the promoter of the tod operon of P. putida Fl, including an inverted repeat located at position -100 and an almost identical P. putida F1 tod box. This suggested the involvement of a regulatory system similar to TodS-TodT for the transcriptional control of the P. mendocina toluene degradation pathway. In fact, a novel two-component signal transduction system was recently described in P. mendocina KR1 [132]. Transcription from the VtmoX promoter, which directs the expression of the tmoXABCDEF gene operon, is induced in the presence of toluene or/>-cresol by a two-component system made up of TmoS and TmoT, which are 83% and 85% identical, respectively, to the TodS and TodT proteins described above. Furthermore, transcription from P,mox and P,O-cresol to TCA cycle intermediates requires the catabolic operons pcuCAXB, pobA, and the gene necessary for ring opening and subsequent degradation. The pcuCAXB operon is required to transform the/?-cresol produced from toluene by toluene monooxygenase into /?-hydroxybenzoate. The pcuCAXB operon is regulated by the divergently transcribed pcuR gene, which belongs to the NtrC family. Expression analysis of the pcuCAXB operon using a reporter gene fused to the promoter showed that only substrates of the pathway, such as p-cresol, p-hydroxybenzyl alcohol or p-hydroxybenzyl aldehyde, were effectors of PcuR. Neither toluene nor />-hydroxybenzoate was able to induce expression of the pathway [134]. The pobA gene codes for/>-hydroxybenzoate hydroxylase, which converts /?-hydroxybenzoate into protocatechuate, the substrate for ring fission [135, 136]. It has been shown that in this strain, expression of the pobA 1 gene is under the control of the divergently transcribed pobRl gene. However, no data are available on the specific molecular mechanisms involved in this process. 5. OVERVIEW OF THE REGULATION OF THE TBU AND TOM PATHWAYS IN RALSTONIA PICKETTII AND BURKHOLDERIA CEPACIA. Ralstonia pickettii PK01 (formerly Pseudomonas pickettii PK01) is able to grow on toluene, phenol, and benzene as the sole carbon and energy source [34]. The pathway responsible for the degradation of these compounds to TCA cycle intermediates is called the toluene-3 -monooxygenase pathway (also tbu pathway for toluene benzene utilization) (Fig. 1). The genes that encode enzymes for the pathway are grouped in three operons. The tbuAlUBVA2C and tbuT operon encodes the initial toluene-p-monooxygenase and the transcriptional activator TbuT [137]. The tbuD operon encodes phenol/cresol hydroxylase [138, 139],

360

and the tbuWEFGKIHJ operon encodes enzymes of the /weta-cleavage pathway for the conversion of catechol and methylcatechols to tricarboxylic acid cycle intermediates [140]. In addition, tbuX codes for a putative facilitator of toluene entry into the cell [141], and is located immediately downstream from tbuT. In turn, TbuT, a member of NtrC family of transcriptional activators, controls transcription of each of these operons in response to aromatic effector compounds [137]. TbuT is activated by aromatic effectors (toluene, benzene and ethylbenzene) and trichloroethylene. Expression of tbuT auto-regulated, as suggested by the finding that it is linked to the expression of the tbuAlUBVA2C operon by read-through transcription of tbuT from the toluene-/>monooxygenase promoter. As a result, transcription of tbuT is low when the toluene-p-monooxygenase operon is not induced and high when expression of tbuAlUBVA2C is induced by toluene. Thus, the toluene-/>-monooxygenase promoter drives the cascade expression of both the toluene-p-monooxygenase operon and tbuT, resulting in a positive feedback circuit [137]. Upstream from the o54-dependent toluene-p-monooxygenase promoter (P/&«AI), a DNA region with dyad symmetry may serve as the TbuT-binding site [137, 142]. Two additional regulatory genes, TbuS and TbuR, are located upstream from tbuD. In the absence of effectors, TbuS represses transcription of tbu WEFGKIHJ. The current view of the regulatory mechanism suggests that in the presence of the effectors phenol or w-cresol, these compounds interact with TbuS and the effector-TbuS complex acts as a transcription activator of tbuWEFGKIHJ. In addition, the effectors that interact with TbuR form complexes able to activate transcription oftbuD [34]. Burkholderia cepacia G4, formerly Pseudomonas cepacia, degrades toluene through a unique initial step that involves toluene-o?t/j0-monooxygenase (TOM) (Fig. 1). This strain also contains a catechol 2,3-dioxygenase for the meta-cleavage of methylcatechol. Because of its unique substrate specificity, the biochemistry of the TOM enzyme has been studied extensively. However, little is known about the genetics and regulation of the pathway. The genes responsible for this pathway are located in a 70-100 kb megaplasmid present in this strain. It has been shown that expression of the pathway is constitutive [143]. Burkholderia sp. strain JS150 contains a plasmid that carries the genes encoding for a toluene-2-monooxygenase clustered in the operon tbmABCDEF and its NtrC-like regulator coding gene tmbR. Additionally, a toluene-4monooxygenase activity was assigned to an independent region of the same plasmid, which was shown to be regulated by TbmR [144]. Finally, crossactivation between toluene-3-monooxygenase and toluene-2-monoxygenase by regulators TbuT and TbmR has been reported [142].

361

6. TRANSCRIPTIONAL REGULATION OF THE ORGANIC SOLVENT EFFLUX PUMPS IN Pseudomonasputida In a recent study, several Pseudomonas putida strains were analyzed with regard to toluene tolerance [145]. Three of these strains have been classified as highly resistant (P. putida DOT-TIE, P. putida S12 and P. putida MTB6). In P. putida DOT-TIE, three efflux pumps are involved in solvent tolerance: TtgABC [146], TtgDEF [147] and TtgGHI [148]. The same three efflux pump operons are present in the P. putida MTB6 chromosome although their participation in organic solvent extrusion has not been studied in detail. Pseudomonas putida S12 contains two of these efflux pumps encoded by the arpABC genes (98% identical to ttgABQ [149], and the srpABC (99% identical to ttgGHI), although only one of these efflux pumps, SrpABC, has been involved in solvent tolerance in the S12 strain [151]. P. putida Fl has two efflux pumps ttgABC and sepABC [152] and is more tolerant to toluene that P. putida KT2440, which only has the ttgABC pump, but it is more sensitive than DOT-TIE. In P. putida DOT-TIE, solvent tolerance is an inducible process, as growth of P. putida DOT-TIE in the presence of toluene supplied in the gas phase has a clear effect on cell survival: the sudden addition of 0.3% (vol/vol) toluene to P. putida DOT-TIE pre-grown with toluene in the gas phase resulted in survival of almost 100% of the initial cell number, whereas only 0.01% of the cells pre-grown in the absence of toluene tolerated exposure to this aromatic hydrocarbon (Fig. 6) [146, 153]. The three efflux pumps in this strain should therefore work together in this strain to achieve the maximal level of solvent tolerance, and they are probably tightly regulated in order to produce an optimal response to solvent stress. Most of the regulatory genes that encode for proteins involved in control of the expression of the efflux pumps belonging to the RND family are located adjacent to the structural genes of the pump, divergently transcribed from the efflux pump operon [154]. 6.1. Regulation of the ttgABC efflux pump operon The TtgABC efflux pump was the first efflux pump identified in P. putida DOT-TIE as involved in solvent tolerance. Physiological experiments done with a ttgB knockout mutant suggested that this efflux pump was involved in the socalled intrinsic tolerance. This mutant (P. putida DOT-TIE-18) did not withstand the sudden toluene shock (0.3% vol/vol) at all, and only a small but significant fraction (about 1 out of 105 cells) survived if pre-exposed to low toluene concentrations [146]. On the basis of this observation the existence of other(s) efflux pump(s) involved in toluene extrusion was postulated. The fact that in P. putida DOT-TIE-18 cultures, no survival at all was observed after sudden toluene shock compared with the 0.01% cell survival observed in the

362

wild-type, pointed out to a high basal expression of the TtgABC pump responsible for this noninduced intrinsic resistance. This hypothesis was confirmed after studying the expression levels of the ttgABC operon. In fact, the operon is transcribed at relatively high levels in the presence and also in the absence of toluene in accordance with its physiological behavior [155, 156]. Interestingly, the analysis of the TtgABC efflux pump susbstrate range revealed that it not only extruded different organic solvents such as toluene, styrene or p-xylene, but also different antibiotics such as ampicillin, carbenicillin, tetracycline, nalidixic acid and chloramphenicol [146, 155, 157]. Gene fusion to lacZ and primer extension assays showed that some substrates of the TtgABC efflux pump, such as chloramphenicol or tetracycline, did increase (to a different extent) the expression of the pump operon, whereas others (nalidixic acid, streptomycin or carbenicillin) did not [156]. Therefore, the contribution of this solvent efflux pump to antibiotic resistance is probably a consequence of the broad substrate specificity of both the pump itself and the transcriptional regulator. Upstream of the ttgABC operon, there is an open reading frame (named ttgR) encoding a protein that shows 50-60% sequence similarity with a number of transcriptional repressors such as AcrR (repressor of the acrAB operon in Escherichia coli) or MtrR (regulator of the MtrCDE efflux pump in Neisseria gonorrhoeae [158, 159]). All these proteins belong to the TetR family of transcriptional regulators [160]. TtgR is a repressor of the TtgABC efflux pump: a ttgR knockout mutant (P. putida DOT-TIE-13) exhibited an increased expression (6-fold higher than the wild-type) of the ttgABC operon. However, this increase in the efflux pump transcription levels did not lead to a higher survival rate of the culture when shocked with 0.3% (vol/vol) toluene, but did increase resistance towards different antibiotics such as chloramphenicol, carbenicillin and tetracycline [155]. ttgR gene expression was very similar to that observed for its cognate efflux pump operon: both are induced by chloramphenicol and tetracycline but not by toluene [155, 156]. In the TtgR-deficient mutant strain the basal activity of the ttgR promoter was eight-fold higher than the wild-type background, indicating that TtgR also down regulates its own transcription. The transcription initiation points of ttgA and ttgR have been mapped. The -10 region, although not the -35 region, of both the ttgA and ttgR promoters exhibits a certain degree of similarity to promoters recognized by sigma-70 [155]. The location of the start sites indicates that the divergent promoter regions fully overlap (Fig. 6). This overlap could explain some of the great similarity in the expression pattern.

363

Fig. 6. Organization of the ttgABC (A), ttgDEF (B) and ttgGHI (C) operons and their respective regulatory genes. The regulatory regions of each gene cluster are zoomed. TtgR (A), TtgT (B) and TtgV (C) DNA binding regions, deduced from DNAsel footprinting, are shadowed. Putative palindromic (arrows) or symetric (bold and underlined) recognition sites for each repressor are indicated. The +1 and the direction of transcription are marked with small triangles for each promoter (except for ttgT one, which distance from indicated). The -10 and the -35 positions of each promoter are also shown.

364

A recombinant and functional His-tagged TtgR protein has been overexpressed, purified and used in several in vitro experiments carried out to elucidate its regulatory role in ttgABC and ttgR expression. Gel shift assays showed that TtgR binds specifically to a DNA fragment corresponding to the ttgR-ttgABC intergenic region. Teran and coworkers [156] demonstrated that the addition of chloramphenicol or tetracycline to the binding reaction led to a gradual dissociation of TtgR from the ttgR-ttgABC intergenic region, suggesting that TtgR is able to bind these two antibiotics, which triggers its release from the promoter regions with subsequent enhanced expression. DNAse I footprint assays allowed the identification of the TtgR operator within the ttgA-ttgR intergenic region: a 36-bp DNA segment that includes the -10 and -35 region of the ttgABC promoter and the -10 of the ttgR promoter. The TtgR binding site deduced from DNAse I footprint, revealed a particularly long inverted repeat (28 bp) comprising two 12-bp half sites separated by 4 bp [156]. This indicates that probably each half site would accept two monomers of TtgR, as reported for the TtgR homolog QacR (repressor of the qacA multidrug pump gene of Staphylococcus aureus), whose 3D structure bound to its operator was resolved [161]. The mepABC efflux pump of a toluene-resistant variant of P. putida KT2442 has also been implicated in solvent and antibiotic resistance, and its sequence is practically identical to those of the TtgABC and ArpABC efflux pumps. The corresponding regulatory protein is probably encoded by mepR, although its function as a regulator has not been investigated yet [162]. Although TtgABC-like efflux pumps are widespread in different Pseudomonas putida strains (P. putida MTB5, P. putida KT2440, P. putida SMO116, among others) with different levels of toluene tolerance [145]. The role of these efflux pumps in solvent or antibiotic resistance and their regulation remains unknown. 6.2. Regulation of the ttgDEF efflux pump operon As described above, solvent tolerance studies in a ttgB knockout mutant of P. putida DOT-TIE (P. putida DOT-TIE-18) suggested the existence of other(s) inducible solvent efflux pump(s). By sequencing downstream from the toluene dioxygenase (tod) operon of P. putida DOT-TIE, Mosqueda and Ramos [147] identified three open reading frames (ttgDEF) that encode for the three components of an efflux pump which shares homology with other efflux pumps of Pseudomonas. The transporter, named TtgE, shares 59% identity with the previously described TtgB and 75% identity with TtgH. The contribution of this efflux pump to solvent tolerance was studied in a ttgD knockout mutant (P. putida DOT-T1E-1). A culture of this mutant strain showed a survival rate identical to the wild type when shocked with 0.3% (vol/vol) toluene. However if the cells were pre-induced with toluene in the gas phase, the survival rate of the

365

mutant was 100 times lower than in the wild type. Expression studies of the ttgDEF operon at the transcriptional level revealed that this pump is not expressed during growth under normal laboratory conditions, and demonstrated its inducible character in the presence of organic solvents (toluene or styrene) (Table 1) [147]. The wild type multiple antibiotic resistance was not affected in a TtgDEFdeficient strain; moreover, no increase in antibiotic resistance was obtained by pre-inducing the culture with toluene [147], suggesting that the substrate specificity of this pump is limited to organic solvents. There was also no induction of the ttgDEF operon in the presence of several antibiotics in the culture media (Teran et al., unpublished). Upstream from the ttgDEF operon and divergently transcribed, there is an open reading frame whose product shares homology with several members of the IclR family of transcriptional regulators. This gene, called ttgT, encodes for a protein 70% identical to the SrpS-negative regulator of SrpABC solvent efflux pump of P. putida S12 (see below). A ttgT knockout mutant showed a small increase in ttgDEF expression under non-inducing conditions, suggesting its involvement in the negative regulation of this operon. The fact that in this mutant strain there was still a strong induction of the ttgDEF expression in the presence of organic solvents suggested that TtgT is not the only protein involved in the induction of this operon by organic solvents (Teran et al., unpublished). Differently from ttgR, ttgT gene expression remained unaltered regardless of the organic solvent present in the growth medium, which suggested that expression from ttgDEF and ttgT promoters was not coordinated. Moreover, in the TtgT-deficient mutant, the activity of the ttgT promoter was similar to that of the wild-type, indicating that TtgT does not regulate its own transcription (Teran et al., unpublished). Gel shift experiments showed that TtgT was able to specifically bind a DNA fragment containing the ttgT-ttgDEF intergenic region (Teran et al., unpublished). DNAse I footprint assays revealed a single binding site along the ttgT-ttgDEF intergenic region which covers only the ttgDEF promoter region (37 to +5 from the transcription start point) and not the ttgT one, consistent with the in vivo expression studies described above. Therefore TtgT is directly involved in ttgDEF operon repression, probably by competing with the RNA polymerase for access to the efflux pump promoter. Analysis of the operator sequence does not reveal the presence of a clear single inverted repeat. Recent results of our laboratory suggests the induction of this efflux pump in a ttgTdeficient background by organic solvents is mediated regulator by the TtgV regulator.

366

6.3. Regulation of the ttgGHI efflux pump operon The third efflux pump involved in solvent tolerance in P. putida DOTTIE is called TtgGHI. A knockout mutant in which the TtgGHI efflux pump of P. putida DOT-TIE is not functional (P. putida DOT-T1E-PS28) is not able to survive a sudden 0.3% (vol/vol) toluene shock regardless of the growth conditions [163]. In contrast, the TtgABC and TtgDEF knockout mutants were still able to survive the toluene shocks to a different extent. The extreme toluene sensitivity of the P. putida DOT-T1E-PS28 mutant strain (ttgH::Q.Sm) under induced or noninduced conditions suggested that this efflux pump is involved in intrinsic as well as inducible resistance to organic solvents. The ttgGHI operon is expressed from a single promoter PG2 at a certain basal level in the absence of solvents, and its expression increases several-fold in the presence of aromatic hydrocarbons such as toluene and styrene, aliphatic alcohols such as 1-octanol, but not in the presence of antibiotics [163, 164]. In P. putida DOT-TIE, two genes ttgV and ttgW were identified upstream from the ttgGHI operon. They are transcribed divergently from the efflux pump operon. TtgV showed an overall 50%-60% similarity with a number of transcriptional regulators belonging to the IclR family, whereas ttgW is probably a pseudogene and the protein encoded seems not to be functional. A TtgV knockout mutant was constructed and characterized [163]. This strain showed increased resistance toward toluene shocks under noninduced conditions when compared with the wild type. The fraction of cells that survived the sudden addition of 0.3% (vol/vol) toluene was the same (107 cells) under induced and noninduced conditions. Analysis of the expression of the ttgGHI operon in this genetic background showed that the level of transcription increased 4-fold in the absence of toluene. Taken together, these data clearly indicated that TtgV is a repressor that prevents expression of the ttgGHI operon. The transcription initiation point of the ttgVW operon was mapped in cells growing in the absence and in the presence of toluene. The operon was transcribed from a single promoter regardless of the growth conditions, but the level of expression in the presence of toluene was 3- to 4-fold higher than in the absence of the aromatic hydrocarbon. The ttgVW operon was also shown to be induced by several organic solvents but not by antibiotics. The ttgVW operon showed a pattern of inducibility similar to that of ttgGHI, probably because both promoters are regulated in the same way. Sequence analysis of the promoter region showed that the -10 and -35 boxes of the VG2 overlap with the -35 and 10 boxes of the ftgFpromoter (Fig. 6). P-galactosidase assays carried out with a transcriptional fusion of VtlgV promoter to lacZ in a TtgV-deficient background showed that expression was about 3-fold higher than in the wild type strain, also in the absence of toluene, indicating that TtgV negatively controls its own expression. Together, these results suggested that the TtgV protein is a repressor of its own synthesis as well as of ttgGHI operon expression [163]. It is

367

interesting to note that in TtgV-deficient background expression from ttgDEF operon promoter increased 3-fold suggesting the involvement of this regulator in the control of this operon. Sequence analysis of PHgG and VttgV showed that both promoters overlapped. The TtgV protein has been overexpressed and purified with an Nterminal histidine tag. In vitro gel mobility shift assays demonstrated the specific binding of TtgV to a 210-bp DNA fragment comprising the ttgG and ttgV intergenic region. The DNAse I-protected region extended a 40-bp covering the -10/-35 regions of the ttgG promoter and the divergently oriented ttgV promoter [163]. To gain insight into the mechanism of regulation of ttgGHI transcription by TtgV, in vitro transcription experiments were carried out using the purified protein and the ttgV-ttgG intergenic region on a supercoiled plasmid. When the TtgV protein and the plasmid were incubated before the addition of RNApolymerase, ttgGHI transcription was completely repressed. However, when TtgV was added after the formation of the RNA-polymerase-«gGif/ promoter open complex, the repression level became negligible [164]. These findings support the idea that TtgV binding to the intergenic region blocks the entry of RNA-polymerase to transcribe both operons. In vitro transcription assays were carried out in the absence and in the presence of increasing concentrations of 1-hexanol, a known inducer of ttgGHI operon. Addition of the inducer to the transcription reaction in the presence of TtgV led to transcriptional levels similar to those observed in the absence of TtgV repressor, resulting in VttgG expression. This suggests that 1-hexanol decreased TtgV binding to the intergenic region, as it has been demonstrated by gel shift assays [164]. Acknowledgments Work in our laboratory was supported by grants of the European Commission (QLK3-CT-1999-00041, QLK3-CT-2001-00435 and QLK3-CT2000-0170) and a grant from the Human Science Foundation (RGY0021/2001). We thank C. Lorente for reading the manuscript and improving the language. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8]

S. Dagley, Adv. Microb. Physiol, 6 (1971) 1. J. Sikkema, J.A.M. de Bont, and B. Poolman, Microbiol. Rev. 59 (1995) 201. D.L. Cruden, J.H. Wolfram , R.D. Rogers and D.T. Gibson , Appl. Environ. Microbiol. 58 (1992) 2723. A. Inoue and K. Horikoshi, Nature 338 (1989) 264. K. Kim, S. Lee, K. Lee, and D. Lim, J. Bacteriol. 180 (1998) 3692. J.L. Ramos, E. Duque, MJ. Huertas and A. Haidour, J. Bacteriol. 177 (1995) 3911. F.J. Weber, S. Isken and J.A.M. de Bont, Microbiology 140 (1994) 2013. H. Keweloh and H.J. Heipieper, Lipids 31 (1996) 129.

368

[9] HJ. Heipieper and J.A.M. de Bont, Appl. Environ. Microbiol. 60 (1994) 4440. [10] HJ. Heipieper, G.Meulenbeld, Q. van Oirschot and J.A.M. de Bont, Appl Environ Microbiol. 62(1996)2773. [11] B. Loffeld and H. Keweloh , Lipids31 (1996)811. [12] A. von Wallbrun, N.H.Richnow,, Neumann, G., Meinhardt, F., and Heipieper, H.J., (2003), J. BacterioL, 185 1730. [13] F.J. Weber and J.A.M. de Bont, Biochim. Biophys. Acta 1286 (1996) 225. [14] G. Mosqueda, M. I. Ramos-Gonzalez and J. L. Ramos, Gene, 232 (1999) 69. [15] P. Godoy, M.I. Ramos-Gonzalez and J.L. Ramos, J. BacterioL 18 (2001) 5285. [16] Q. Zhao, X.Z. Li, A. Mistry, R. Srikumar, L. Zhang, O. Lomovskaya and K. Poole, Antimicrob. Agents. Chemother. 42 (1998) 2225. [17] V.Koronakis, A. Sharfl, E. Koronakis, B. Luisi and C. Hughes, Nature 405 (2000) 914. [18] S. Murakami, R. Nakashima, E. Yamashita and A. Yamaguchi, Nature 419 (2002) 587. [19] H. I. Zgurskaya and H. Nikaido J. Mol. Biol. 285 (1999) 409. [20] E.W. Yu, G. McDermott, H.I. Zgurskaya, H. Nikaido and D.E. Jr. Koshland, Science 300(2003) 976. [21] H.I. Zgurskaya and H. Nikaido, J. BacterioL 182 (2000) 4264. [22] C.A. Elkins and H. Nikaido, Drug Resist. Update 6 (2003) 9. [23] S. J. Assinder and P.A. Williams, Adv. Microb. Physiol., 31 (1990) 1. [24] P.A. Williams, R.M. Jones and G. Zylstra, (2004), Genomics of catabolic plasmids. In Pseudomonas, vol. I, Chapter 6, pp. 165-196, Ed. J.L. Ramos, Kluwer Academic/Plenum Publishers. [25] S. Harayama, N. Mermod, M. Rekik, P.R. Lehrbach and K.N. Timmis, J. BacterioL, 169 (1987) 558. [26] S. Harayama and K.N. Timmis (1989), Catabolism of aromatic hydrocarbons by Pseudomonas. In Genetics of Bacterial Diversity, ed. D. Hopwood, K. Chater, 151-174. New York Academic. [27] C. Nakai, H. Kagamiyama and M. Nozaki, J. Biol. Chem., 258 (1983) 2923. [28] A. Polissi and S. Harayam, EMBO J. 12 (1993) 3339. [29] J.P. Shaw and S. Harayama, Eur. J. Biochem., 191 (1990) 705. [30] M. Suzuki, T. Hayakawa, J.P. Shaw, M. Rekik and S. Harayama, J. BacterioL 173 (1991) 1690. [31] D. T. Gibson, J. R. Koch and R. E. Kallio, Biochemistry, 7 (1968) 2653. [32] G. M. Whited and D. T. Gibson, J. BacterioL, 173 (1991) 3010. [33] K. M. Yen, M. R. Karl, L. M. Blatt, M. J. Simon, R. B. Winter, P. R. Fausset, H. S. Lu, A.A. Harcourt and K. K. Chen, J. BacterioL, 173 (1991) 5315. [34] R. H. Olsen, J. J. Kukor and B. Kaphammer, J. BacterioL, 176 (1994) 3749. [35] A. Fishman, Y. Tao and T.K. Wood, J. BacterioL, (2004) in press. [36] G. Baggi, P. Barbieri, E. Galli and S. Tollari, Appl. Environ. Microbiol., 53 (1987) 2129. [37] G. Bertoni, F. Bolognese, E. Galli and P. Barbieri, Appl. Environ. Microbiol., 62 (1996)3704. [38] G. Bertoni, N. Fujita, A. Ishihama and V. de Lorenzo, EMBO J., 17 (1998) 5120. [39] G. Bertoni, S. Marques and V. de Lorenzo, Mol. Microbiol., 27 (1998) 651. [40] J.L. Ramos, E. Duque, J.J. Rodriguez-Herva, P. Godoy, A. Haidour and J.L. Ramos, J. Biol. Chem. 272 (1997) 3887. [41] EganS.M., (2002) J. BacterioL, 184 5529. [42] M. T. Gallegos, R. Schleif, A. Bairoch, K. Hofmann and J. L. Ramos, Microbiol. Mol. Biol. Rev., 61 (1997) 393.

369

[43] J. L. Ramos, C. Michan, F. Rojo, D. Dwyer and K. Timmis, J. Mol. Biol., 211 (1990) 373. [44] R.Tobes and J. L Ramos, Nucleic Acids Res., 30 (2002) 318. [45] J. L. Ramos, A. Stolz, W. Reineke and K. N. Timmis, Proc. Natl. Acad. Sci. USA, 83 (1986) 8467. [46] C. Michan, L. Zhou, M. T. Gallegos, K. N. Timmis and J. L. Ramos, J. Biol. Chem., 267 (1992) 22897. [47] L. M. Zhou, K. N. Timmis and J. L. Ramos, J. Bacteriol., 172 (1990) 3707. [48] J. L. Ramos, F. Rojo, L. Zhou and K. N. Timmis, Nucleic Acids Res., 18 (1990) 2149. [49] J.L. Ramos, A. Wasserfallen, K. Rose and K.N. Timmis, Science 235 (1987) 593. [50] R. Ruiz and J. L. Ramos, J. Biol. Chem, 277 (2002) 7282. [51] N. LaRonde-LeBlanc and C. Wolberger, Biochemistry, 39 (2000) 11593. [52] C. A. Poore, C. Coker, J.D. Dattelbaum and H. L. Mobley, J. Bacteriol, 183 (2001) 4526. [53] S. Rhee, R. G. Martin, J. L. Rosner and D. R. Davies, Proc. Natl. Acad. Sci. USA, 95 (1998) 10413. [54] S. M. Soisson, B. MacDougall-Shackleton, R. Schleif and C. Wolberger, Science, 276 (1997)421. [55] R. Ruiz, S. Marques and J. L. Ramos, J. Bacteriol, 185 (2003) 3036. [56] M. Manzanera, S. Marques and J. L. Ramos, FEBS Lett, 476 (2000) 312. [57] N. Kaldalu, U. Toots, V. de Lorenzo and M. Ustav, J. Bacteriol, 182 (2000) 1118. [58] M. M. Gonzalez-Perez, S. Marques, P. Dominguez-Cuevas and J. L. Ramos, FEBS Lett, 519(2002) 117. [59] M. T. Gallegos, S. Marques and J. L. Ramos, J. Bacteriol, 178 (1996) 6427. [60] M. M. Gonzalez-Perez, J. L. Ramos, M. T. Gallegos and S. Marques, J. Biol. Chem, 274(1999)2286. [61] B. Kessler, V. de Lorenzo and K. Timmis, J. Mol. Biol, 230 (1993) 699. [62] B. Kessler, M. Herrero, K. N. Timmis and V. de Lorenzo, J. Bacteriol, 176 (1994) 3171. [63] B. Kessler, S. Marques, T. Kohler, J. L. Ramos, K. N. Timmis and V. de Lorenzo, J. Bacteriol, 176(1994)5578. [64] N. Kaldalu, T. Mandel and M. Ustav, Mol. Microbiol, 20 (1996)569. [65] K. Miura, S. Inouye and A. Nakazawa, Mol. Gen. Genet, 259 (1998) 72. [66] I. Artsimovitch, K. Murakami, A. Ishihama and M. M. Howe, J. Biol. Chem, 271 (1996) 32343. [67] S. Busby and R. H. Ebright, Mol. Microbiol, 23 (1997) 853. [68] A. Dhiman and R. Schleif, J. Bacteriol, 182 (2000) 5076. [69] S. Marques, M. T. Gallegos and J. L. Ramos, Mol. Microbiol, 18 (1995) 851. [70] S. Marques, M. Manzanera, M. M. Gonzalez-Perez, M. T. Gallegos and J. L Ramos, Mol. Microbiol, 31 (1999) 1105. [71] T. Yura and K. Nakahigashi, Curr. Opin. Microbiol, 2 (1999) 153. [72] S. Inouye, A. Nakazawa and T. Nakazawa, Gene, 66 (1988)301. [73] S. Kustu, E. Santero, J. Keener, D. Popham and D. Weiss, Microbiol. Rev, 53 (1989) 367. [74] A. K. North, K. E. Klose, K. M. Stedman and S. Kustu, J. Bacteriol, 175 (1993) 4267. [75] M. A. Abril, C. Michan, K. N. Timmis and J. L. Ramos, J. Bacteriol, 171 (1989) 6782. [76] A. Delgado and J. L. Ramos, J. Biol. Chem, 269 (1994) 8059. [77] A. Delgado, R. Salto, S. Marques and J. L. Ramos, J. Biol. Chem, 270 (1995) 5144.

370

[78] J. Garmendia, D. Devos, A. Valencia and V. de Lorenzo, Mol. Microbiol., 42 (2001) 47. [79] V. Shingler,, (2004), Transcriptional regulation and catabolic strategies of phenol degradative pathways. In Pseudomonas, vol. II, Chapter 16, pp. 451-478, Ed. J.L. Ramos, Kluwer Academic/Plenum Publishers. [80] E. Skarfstad, E. O'Neill, J. Garmendia and V. Shingler, J. Bacteriol, 182 (2000) 3008. [81] S. Fernandez, V. de Lorenzo and J. Perez-Martin, Mol. Microbiol, 16 (1995) 205. [82] J. Perez-Martin and V. de Lorenzo, J. Mol. Biol, 258 (1996) 575. [83] J. Perez-MartinandV.de Lorenzo, Cell, 86(1996)331. [84] R. Dixon, Mol. Gen. Genet, 203 (1986) 129. [85] S. Inouye, A. Nakazawa and T. Nakazawa, Gene, 29 (1984) 323. [86] M. A. Abril, M. Buck and J. L. Ramos, J. Biol. Chem, 266 (1991) 15832. [87] V. de Lorenzo, M. Herrero, M. Metzke and K. N. Timmis, EMBO J, 10 (1991) 1159. [88] G. H. Seong, E. Kobatake, K. Miura, A. Nakazawa and M. Aizawa, Biochem. Biophys. Res. Commun, 291 (2002) 361. [89] G. Bertoni, M. Martino, E. Galli and P. Barbieri, Appl. Environ. Microbiol, 64 (1998) 3626. [90] M. Carmona, V. de Lorenzo and G. Bertoni, J. Biol. Chem, 274 (1999) 33790. [91] A. Holtel, M. A. Abril, S. Marques, K. N. Timmis and J. L. Ramos, Mol. Microbiol, 4 (1990) 1551. [92] A. Holtel, D. Goldenberg, H. Giladi, A. B. Oppenheim and K. N. Timmis, J. Bacteriol, 177(1995)3312. [93] A. Holtel, K. Timmis and J. L. Ramos, Nucleic Acids Res, 20 (1992) 1755. [94] G. Bertoni, J. Perez-Martin and V. de Lorenzo, Mol. Microbiol, 23 (1997) 1221. [95] M. Gomada, S. Inouye, H. Imaishi, A. Nakazawa and T. Nakazawa, Mol. Gen. Genet, 233(1992)419. [96] S. Inouye, A. Nakazawa and T. Nakazawa, J. Bacteriol, 163 (1985) 863. [97] S. Marques, M. T. Gallegos, M. Manzanera, A. Holtel, J. Bacteriol, 180 (1998) 2889. [98] S. Fraile, F. Roncal, L. A. Fernandez and V. de Lorenzo, J. Bacteriol, 183 (2001) 5571. [99] W. A. Duetz, S. Marques, C. de Jong, J. L. Ramos and J. G. van Andel, J. Bacteriol, 176(1994)2354. [100] W.A. Duetz, S. Marques, B. Wind, J.L. Ramos and J.G. Van Andel, Appl. Environ. Microbiol, 62(1996)601. [101] A. Holtel, S. Marques, I. Mohler, U. J akubzik and K. N. Timmis, J. Bacteriol, 176 (1994) 1773. [102] N. Hugouvieux-Cotte-Pattat, T. Kohler, M. Rekik and S. Harayama, J. Bacteriol, 172 (1990)6651. [103] S. Marques, A. Holtel, K. N. Timmis and J. L. Ramos, J. Bacteriol, 176 (1994) 2517. [104] M. J. Worsey, and P. A. Williams, J Bacteriol. 124 (1975) 7. [105] I. Cases, V. de Lorenzo and J. Perez-Martin, Mol. Microbiol, 19 (1996) 7. [106] M. Carmona, M. J. Rodriguez, O. Martinez-Costa and V. de Lorenzo, J. Bacteriol, 182 (2000)4711. [107] F. Rojo and M.A. Dinamarca, (2004), Catabolite repression and physiological control. In Pseudomonas, vol. II, Chapter 13, pp. 365-388, Ed. J.L. Ramos, Kluwer Academic/Plenum Publishers. [108] M. Vails, M. Buckle and V. de Lorenzo, J. Biol. Chem, 277 (2002) 2169. [109] P. Jurado, L. A. Fernandez and V. de Lorenzo, J. Bacteriol, 185 (2003) 3379. [110] T. Tomoyasu, K. Yamanaka, K. Murata, T. Suzaki, P. Bouloc, A. Kato, H. Niki, S. Hiraga and T. Ogura, J. Bacteriol, 175 (1993) 1352.

371

[111] M. Carmona and V. de Lorenzo, Mol. Microbiol, 31 (1999) 261. [112] Y. Akiyama, Proc. Natl. Acad. Sci. USA, 99 (2002) 8066. [113] I. Cases, and V. de Lorenzo, J. Bacteriol., 182 (2000) 956. [114] I. Cases, J. A. Lopez, J. P. Albar and V. de Lorenzo, J. Bacteriol., 183 (2001) 1032. [115] I. Cases, F. Velazquez and V. de Lorenzo, J. Bacteriol., 183 (2001) 5128. [116] I. Cases, J. Perez-Martin and V. de Lorenzo, J. Biol. Chem., 274 (1999) 15562. [117] B. A. Finette, V. Subramanian and D. T. Gibson, J. Bacteriol., 160 (1984) 1003. [118] D. T. Gibson, G. H. Zylstra and S. Chauhan, (1990), Biotransformations catalyzed by toluene dioxygenase from Pseudomonasputida Fl, p. 121-132. In S. Silver, A.M. [119] P. C. Lau, Y. Wang, A. Patel, D. Labbe, H. Bergeron, R. Brousseau, Y. Konishi and M. Rawlings, Proc. Natl. Acad. Sci. USA, 94 (1997) 1453. [120] F. M. Menn, G. J. Zylstra and D. T. Gibson, Gene, 104 (1991) 91. [121] Y. Wang, M. Rawlings, D. T. Gibson, D. Labbe, H. Bergeron, R. Brousseau and P. C. Lau Mol. Gen. Genet., 246 (1995) 570. [122] G. J. Zylstra and D. T. Gibson, J. Biol. Chem., 264 (1989) 14940. [123] G. J. Zylstra, W. R. McCombie, D. T. Gibson and B. A. Finette, Appl. Environ. Microbiol., 54 (1988)1498. [124] B. A. Finette and D. T. Gibson, Biocatalyst 2 (1988) 29. [125] P. Dominguez-Cuevas, and S. Marques, (2004), Compiling sigma-70 dependent promoters. In Pseudomonas, vol. II, Chapter 11, pp. 319-344. Ed. J.L. Ramos, Kluwer Academic/Plenum Publishers. [126] B. L. Taylor and I. B. Zhulin, Microbiol. Mol. Biol. Rev., 63 (1999) 479. [127] I. B. Zhulin, B. L. Taylor and R. Dixon, Trends Biochem. Sci., 22 (1997) 331. [128] J. A. Hoch and T. J. Silhavy, (1995), Two-component signal transduction. American Society for Microbiology, Washington DC. [129] J. Reizer and M. H. J. Saier, Curr. Opin. Struct. Biol., 7 (1997) 407. [130] R. E. Parales, J. L. Ditty and C. S. Harwood, Appl. Environ. Microbiol., 66 (2000) 4098. [131] K. M. Yen and M. R. Karl, J. Bacteriol., 174 (1992) 7253. [132] M. I. Ramos-Gonzalez, M. Olson, A. A. Gatenby, G. Mosqueda, M. Manzanera, M. J. Campos, S. Vichez and J. L. Ramos, J. Bacteriol., 184 (2002) 7062. [133] A. Wright and R. H. Olsen, Appl. Environ. Microbiol., 60 (1994) 235. [134] A. Ben-Bassat, M. Cattermole, A.A. Gatenby, K.J. Gibson, M.I. Ramos-Gonzalez, J.L. Ramos and S. Sariaslani, (2003), Method for the production ofp-hydroxybenzoate in species of Pseudomonas and Agrobacterium. US Patent 6,586,229 [135] B. Entsch, Y. Nan, K. Weaich and K. F. Scott, Gene, 71 (1988) 279. [136] C. M. Wong, M. J. Dilworth, and A. R. Glenn, Microbiology, 140 (1994) 2775. [137] A. M. Byrne and R. H. Olsen, J. Bacteriol., 178 (1996) 6327. [138] J. J. Kukor and R. H. Olsen, J. Bacteriol., 172 (1990) 4624. [139] J. J. Kukor and R. H. Olsen, J. Bacteriol., 174 (1992) 6518. [140] J. J. Kukor and R. H. Olsen, J. Bacteriol., 173 (1991) 4587. [141] H. Y. Kahng, A. M. Byrne, R. H. Olsen and J. J. Kukor, J. Bacteriol., 182 (2000) 1232. [142] J. G. Leahy, G. R. Johnson and R. H. Olsen, Appl. Environ. Microbiol., 63 (1997) 3736. [143] M. S. Shields, M. J. Reagin, R. R. Gerger, R. Campbell and C. Somerville, Appl. Environ. Microbiol., 61 (1995) 1352. [144] G.R. Johnson and R.H. Olsen, Appl. Environ. Microbiol., 63 (1997) 4047. [145] A. Segura, A. Rojas, A. Hurtado, M.J. Huertas and J.L. Ramos, Extremophiles 7 (2003)371. [146] J.L. Ramos, E. Duque, P. Godoy and A. Segura, J. Bacteriol. 180 (1998) 3323.

372

[147] G. Mosqueda and J.L. Ramos, J. Bacteriol. 182 (2000) 937. [148] A. Rojas, E. Duque, G. Mosqueda, G. Golden, A. Hurtado, J.L. Ramos and A. Segura, J. Bacteriol., 183(2001)3967. [149] J. Kieboom and J.A.M. de Bont, Microbiology 147 (2001) 43. [150] J. Kieboom, J.J. Dennis, J.A.M. de Bont and G.J. Zylstra, J. Biol. Chem. 273 (1998) 85. [151] J. Kieboom, J.J. Dennis, G.J. Zylstra and J.A.M. de Bont, J. Bacteriol. 180 (1998) 6769. [152] P. Phoenix, A. Keane, A. Patel, H. Bergeron, S. Ghoshal and P.C.K. Lau, Env. Microbiol. 5 (2003) 1309. [153] J. L. Ramos, S. Marques and K. N. Timmis, Annu. Rev. Microbiol., 51 (1997) 341. [154] A. Segura, E. Duque, G. Mosqueda, J.L. Ramos and F. Junker, Environ. Microbiol. 1 (1999)191. [155] E. Duque, A. Segura, G. Mosqueda and J.L. Ramos, Mol. Microbiol. 39 (2001) 1100. [156] W. Teran, A. Felipe, A. Segura, A. Rojas, J.L. Ramos and M.T. Gallegos, Antimicrob. Agents Chemother. (2003), in press. [157] J.L. Ramos, E. Duque, M.T. Gallegos, P. Godoy, M.I. Ramos-Gonzalez, A. Rojas, W. Teran and A. Segura, Ann. Rev. Microbiol. 56 (2002) 743. [158] D. Ma, M. Alberti, C. Lynch, H. Nikaido and J.E. Hearst, Mol. Microbiol. 19 (1996) 101. [159] W. Pan and B.G. Spratt, Mol. Microbiol. 11 (1994) 769. [160] W. Hillen and C. Berens, Annu. Rev. Microbiol. 48 (1994) 345. [161] M.A. Schumacher, M.C. Miller, S. Grkovic, M.H. Brown, R.A. Skurray and R.G. Brennan, EMBO J. 21 (2002) 1210. [162] F. Fukumori, H. Hirayama, H. Takami, A. Inoue and K. Horikoshi, Extremophiles 2 (1998)395. [163] A. Rojas, A. Segura, M.E. Guazzaroni, W. Teran, A. Hurtado, M.T. Gallegos and J.L. Ramos, J. Bacteriol., 185 (2003) 4755. [164] M.E. Guazzaroni, W. Teran, X. Zhang, M.T. Gallegos, and J.L. Ramos, J. Bacteriol., (2004) in press.


373

Chapter 13

Bacterial hydrocarbon biosynthesis revisited B. Valderrama Departamento de Ingenieria Celular y Biocatalisis, Universidad Nacional Autonoma de Mexico, AP 510-3, Cuernavaca, Morelos, 62250, Mexico.

1. INTRODUCTION One of the greatest challenges faced by the modern world is the dissociation from the heavy dependency of the energy technologies upon the chemical bonds of hydrocarbons. The imminent exhaustion of conventional oil sources, ranging from a pessimistic ultimate recovery volume of 0.6 trillions of barrels to a highly optimistic volume of 3.9 trillions of barrels [4], results in a stringent requirement for the development of alternative technologies. It is important to note that the world is not to run out of hydrocarbons, given the substantial amount of lowgrade, hard-to-extract supplies such as the Canadian tar sands or the abundant heavy oil reservoirs in Venezuela and Mexico. Nevertheless, exploiting these reservoirs is likely to be much more expensive financially, energetically, politically and especially environmentally. Biotechnology has greatly impacted modern industry, from the now conventional production of goods by the use of fermentations to the novel synthesis of valuable fine chemicals using enzymes [5, 6, 7, 8]. Notwithstanding its enormous potential, the incorporation of biotechnological tools into the oil industry has faltered [9]. In particular, the search of alternative hydrocarbon sources through biotechnological media has not been assessed. Here, I compile information regarding the biological production of hydrocarbons by bacteria and explore its potential, not only as an environmentally-friendly fuel supply, but also as a renewable source for basic petrochemicals.

374

2. HYDROCARBON BIOSYNTHESIS IS A COMMON TRAIT AMONG BACTERIA The ability of organisms to synthesize hydrocarbons has been observed in all phyla [10]. Whereas alkanes are mainly involved in epicuticular wax biosynthesis in plants [11], in insects, their roles are more diverse, ranging from waterproofing of the cuticle to participation in sexual behaviors as aphrodisiac pheromones [12]. Furthermore, many marine animals, from invertebrates to whales, contain appreciable amounts of hydrocarbons as component of waxes, which appear to have a variety of functions, from serving as energy source to insulation, buoyancy and even echo-location [13]. The accumulation of nonvolatile hydrocarbons by microorganisms has been shown to occur in microalgae [14, 15, 16, 3], in bacteria [17, 18, 19] and in yeast [20, 21]. Originally, the ability of microorganisms to produce hydrocarbons was studied as part of the biogenic hypothesis for oil reservoir formation. This hypothesis was actively investigated between 1930 and 1960 by C.E. ZoBell, who proposed a significant bacterial role in the origin of petroleum [22, 23]. In 1950 he suggested that microbial modification of organic remains in sediments could contribute precursors for oil formation by lowering their oxygen and nitrogen content and increasing their carbon and hydrogen content, and by the direct production of methane and other higher hydrocarbons [24, 25]. ZoBell also suggested that hydrogenation of unsaturated fatty acids and their subsequent decarboxylation might have contributed to petroleum formation. Although a bacterial role in the initial processing of organic matter from which petroleum is derived has become generally accepted, there is no experimental evidence that bacteria were directly responsible for hydrocarbon production in significant quantities. Although ZoBell later abandoned the idea of direct microbial formation of hydrocarbons in large amounts from organic matter, he described several cases where cultures of marine bacteria presented substantial capacity for aliphatic hydrocarbon biosynthesis [26, 27, 23]. After the development of more powerful analytical procedures, the subject of bacterial hydrocarbon production resurged between 1960 and 1970 [1, 28, 19, 18]. Unequivocal evidence of hydrocarbon accumulation was observed in all the bacterial species tested (Table 1), including photosynthetic bacteria as well as in non-photosynthetic bacteria. The composition of bacterial hydrocarbons was complex, with length ranging from Ci5 up to C36, and including n-alkanes, alkenes, and branched hydrocarbons. In particular, nonphotosynthetic bacteria accumulate long-chain n-alkanes (C27-C29), whereas nalkanes with shorter chains (C|7-C2o) are more abundant in photosynthetic bacteria [29]. Photosynthetic bacteria, as well as anaerobic non-photosynthetic bacteria, are characterized by the presence of isoprenic units of pristane and phytane [30].

375

Table 1 Taxonomical distribution of eubacterial and archaeal species able to synthesize and accumulate hydrocarbons. Phyllum

Class

Genus

Reference

Proteobacteria

a Proteobacteria

Rhodopseudomonas sphaeroides Rhodospirillum rubrum Chromatium Escherichia coli Rhodimicrobium vannielii Vibrio furn issii Serratia marinorubrum Vibrio marinus Vibrio ponticus Vibrio furnissii Desulfovibrio desulfuricans Desulfovibrio Essex Desulfovibrio Hildenborough

[33,30] [33,30] [29] [33,30] [30] [17] [23] [34] [23] [17] [35] [30] [30]

Clostridium acidiurici Clostridium tetanomorphum Sarcinaflava Sarcina lutea Sarcina subflava Staphylococcus sp. Bacillus sp.

[30] [30] [18] [18,28] [18] [18] [36]

y Proteobacteria

8/E Proteobacteria

Firmicute

Clostridia

Bacilli

Actinobacteria

Actinobacteria

Micrococcus lysodeikticus Micrococcus sp. Mycobacterium sp. Corynebacterium sp. Arthrobacter sp.

[18,33,30] [36] [36] [36] [36]

Cyanobacteria

Nostococales Oscillatoriales

Nostoc muscorum Nostoc sp. Phormidium luridum

[30] [36] [30]

Chlorobi

Chlorobia

Chlorobium

[30]

Euryarchaeota

Thermoplasmata Methanomicrobia Halobacteria

Thermoplasma sp. Methanosarcina barkeri Halobacterium cutirubrum

[32] [32] [32]

Crenarchaeota

Thermoprotei

Sulfolobus sp.

[31]

This ability is not restricted to eubacteria. Some archeal species from the genus Sulfolobus, Thermoplasma, Methanosarcina and Halobacterium, have been demonstrated able to synthesize and accumulate hydrocarbons such as squalene and other acyclic isoprenoids (C20-C25). [31, 32]. Furthermore, individual species that produce hydrocarbons as major components have been isolated from mesophilic, thermophilic, psycrophilic, acidophilic, alkalinophilic,

376

and halophilic environments under aerobic or anaerobic, autotrophic or heterotrophic conditions. The environmental distribution of hydrocarbon producers follows no discernible pattern that can be used as a guide for finding prolific hydrocarbon producers. 3. BIOSYNTHETIC PATHWAYS Non-isoprenoid biological hydrocarbons were presumed to derive from fatty acids. Currently, there are two known pathways for the conversion of fatty acids to straight-chain hydrocarbons. The best known of them is the elongationdecarboxylation process (Fig.lA). In this case, a fatty acid precursor, such as oleic acid, is elongated by the continuous addition of a C2 unit derived from malonyl-CoA. The hydrocarbon produced is then cappedoff through a decarboxylation reaction when it reaches the designated length. The second mechanism involves the "head-to-head" condensation of two fatty acids (Fig. IB). In this path, one of the acid derivatives is specifically decarboxylated following the condensation step, while the total carbon chain of the other is incorporated into the hydrocarbon [3]. The commitment step in these pathways is the decarboxylation reaction. It has been well documented that CO2 elimination from carboxylic acids requires high energy and therefore has to be activated by a P-substituent able to stabilize the negative charge generated by CO2 release. Accordingly, it has generally been thought that activated fatty acid derivatives are the intermediates in the decarboxylation leading to hydrocarbons.

Fig. 1 Metabolic pathways for aliphatic hydrocarbon biosynthesis (Modified from [3] and [1]). LCFA - Long Chain Fatty Acids

377

Fatty acyl-CoA reductase activity has been identified in a variety of other organisms, from bacteria to animals [40, 41, 42, 38]. The gene encoding this enzyme has been recently cloned from the y-proteobacteria Acinetobacter calcoaceticus and Photobacterium leiognathi, as well as from seeds of jojoba (Simmondsia chinensis) [42, 43, 44]. As can be seen in Fig. 2, all of them harbor the residues conforming the catalytic triad observed in related dehydrogenases [45]. Interestingly, each one of these reference sequences represents independent groups, on the basis of sequence similarity (see Fig. 3). Groups I and II are rather selective, being all their members either plantsor bacteria, respectively. Group III has the sequence from Acinetobacter calcoaceticus as sole member. This sequence is similar to oxido-reductases with different substrate specificities from various sources. The aldehydes generated from fatty acid reduction in B. braunii are further reduced to hydrocarbons (alkanes). The initial observation that the resulting alkane had one less carbon than the aldehyde leaded to the proposal of a decarbonylation as the final step. Such an activity would yield one molecule of alkane and one molecule of CO as products. Two plant aldehyde decarbonylases (from pea and from B. braunii) have been studied in some detail [46, 47]. They are integral membrane proteins with the pea decarbonylase suggested to be located in the cuticular cell membrane and the algal decarbonylase in the microsomal membranes. Both use highly hydrophobic fatty aldehydes as substrate and need metal ions for their function. B. braunii decarbonylase is a cobalt-porphyrin enzyme able to convert a fatty aldehyde to hydrocarbon and CO without requiring any other cofactor under anaerobic conditions [48]. The partially purified decarbonylase from pea is merely known to depend on metal ions, probably copper, the activity being severely inhibited in the presence of metal ion chelators [47]. Unfortunately, none of their genes have been cloned to date. Genetic approaches produced mutants of Arabidopsis that have altered surface composition, including a decreased amount of hydrocarbons. One of these mutants, cerl, was postulated to be located in a decarbonylase because it is proportionally deficient only in alkanes (and alkane-derived metabolites) and accumulated fatty aldehydes [11]. The cerl gene was cloned and shown to bedeposits actively expressed in stem and in fruit tissue, which corresponded to the main of waxes. Interestingly, cerl affected pollen fertility. The CER1 protein has been related, on the basis of sequence similarity, to C-5 sterol desaturase and C-4 methyl oxidase, involved in cholesterol biosynthesis [49,50]. All these enzymes are integral membrane proteins containing many conserved histidine residues.

378

Fig. 2. Multiple alignment of fatty acyl CoA reductases from the proteobacteria Acinetobacter calcoaceticus [43] and Photobacterium leiognathi [44] and from the plant Simmondsia chinensis (jojoba) [42]. Residues conforming the predicted catalytic triad are highlighted.

The n-heptane tissue-specific biosynthetic pathway in Pinus jeffreyi proceeds through the polymerization of acetate via a tipical fatty acid synthase reaction sequence yielding a C8 thioester, which subsequently undergoes a twoelectron reduction to generate a free thiol molecule and octanal, the latter undergoes the direct loss of Ci to generate n-heptane [2] (Fig.4). Aside from the obvious generality of the aforementioned pathway, alternative reactions for alkane biosynthesis have been revealed recently. In insects, whereas aldehyde was the immediate precursor of alkanes, the enzyme involved in the deearbonylation step was proposed to be a P450 enzyme which required the presence of molecular oxygen and NADPH or NADH (less effectively) with the production of CO2 instead of CO [51]. In archaea, the

379

synthesis and cleavage of acetyl-CoA is catalyzed by the acetyl-CoA decarbonylase synthase (ACDS) complex, a completely different enzyme composed of five different subunits, [52, 53]. There is no information available regarding an equivalent acyl-CoA decarbonylation reaction in bacteria. 3.1. Bacterial biosynthesis In contrast to the heterogeneity of plant synthesized hydrocarbons [3], a very large proportion of bacterial hydrocarbons are less dispersed in length and structure. The best known example comes from the study of Sarcina lutea, whose most abundant hydrocarbon presented a branch methyl on both ends of the molecule, a double bond near the center, and contained a number of carbon atoms equal to one less than two times the average number of carbon atoms in the most abundant fatty acid [54] (Fig. 5). These structural characteristics, as well as the distribution of the carbon chains of iso-leucine, valine and acetate in the fatty acids and hydrocarbons synthesized in vivo, are consistent with a headto-head condensation of fatty acids mechanism [1, 54, 55]. A plausible precursor for this mechanism would be multiple methyl-branched fatty acyl-CoA molecules produced by the methyl-malonate driven elongation of fatty acid molecules, as has been observed in mycobacteria [56]. In the head-to-head condensation mechanism, one molecule of fatty acid undergoes decarboxylation. When S. lutea cells are grown with low acetate, approximately 70% of the incorporated palmitate molecules are decarboxylated, in contrast, when acetate was included in the growth medium, palmitate was incorporated without undergoing further decarboxylation, suggesting high specificity [57].

Fig. 3. Deduced Fatty Acyl CoA reductases are organized in three different groups on the basis of sequence similarity. Sequence identification numbers in parenthesis.

380

Fig. 4. Proposed pathway for formation of n-heptane from acetate in Pinusjeffreyi. Adapted from [2]

The in vitro incorporation of palmitate into hydrocarbons requires the addition of CoA, Mg++, ATP, NADPH and pyridoxal or pyridoxamine phosphate [58]. The requirement for the first three cofactors was consistent with the participation of acyl-CoA and this was confirmed by showing that in the absence of added CoA, palmitoyl-CoA was over 20 times better a precursor than the free acid. Under these conditions, palmitate was sometimes decarboxylated when added to the system. The specificity was originated from the source of palmitate, whether it was free or esterified. Approximately 30% of the free acid was decarboxylated while the methyl ester derivative was essentially 100% decarboxylated. Also, the CoA derivative of palmitate was extensively decarboxylated. Furthermore, the direct incorporation of palmitate required piridoxamine as a cofactor instead of CoA [58]. Although it is clear that biosynthesis occurs by a head-to-head condensation mechanism, the intermediates of the pathway have not yet been fully elucidated.

Fig. 5 Relationship between isoleucine and acetate, the anteiso C15 fatty acid and the C29 hydrocarbon with anteiso branch methyls in both ends. Modified from [1].

381

All this information leads to the idea that ability of bacteria to synthesize hydrocarbons is widespread and that it probably occurs by more than one mechanism which are significantly different compared to those described in plants. 4. DOWNSTREAM PROCESSING The isolation and refinement of bacterial hydrocarbons has not been approached at production scale. Nevertheless, there is abundant information regarding the operations developed for a similar procedure with B. braunii. The process of extracting hydrocarbons from these cells can be thought of as consisting of three major operations. The first is that of harvesting the cells from the growth medium. This involves the concentration or flocculation of cells from the liquid where it is grown. This operation can be achieved through a variety of means that include filtration, mechanical centrifugation or concentration, gravitational concentration or chemical flocculation. The most efficient method for largescale hydrocarbon recovery is chemical flocculation. For efficient extraction, the cells must be concentrated to a semi-dry paste. The second step is that of the actual physical extraction of the hydrocarbon fuel from the cells. Under suitable conditions, up to 70% of the total hydrocarbon content can be released by 30 min of contact with solvents. The selected solvent should be immiscible with water, with a density significantly different than water, should be non-toxic and reusable. In view of these considerations, hexane appears to be the solvent of choice [59]. Growth and hydrocarbon production are not affected by repeated extraction with hexane. In fact, a higher content of hydrocarbons has been observed in hexane-treated biomass relative to controls. Nevertheless, recovery yields are influenced by the physiological status of the culture. Scale up of the extraction can be difficult, given the algae propensity to aggregate. Extensive clumping shields a large fraction of the biomass from exposure to solvent. Alternative methods aimed to increase the oil extraction yield have been explored. Recovery yields are markedly increased relative to freely suspended controls when cells immobilized by adsorption in polyurethane foam were continuously extracted with hexane [60]. Supercritical fluid extraction is another technology that has been applied [61]. The third operation would be the collection and concentration of the hydrocarbon product. Although biosynthetic hydrocarbons can be directly used in internal combustion engines after extraction with hexane, its performance is improved by further modification. Cellular hydrocarbons can be converted to gasoline (60 to 70%), light cycle oil (10 to 15%), heavy cycle oil (2 to 8%) and coke (5 to 10%) after catalytic cracking [62]. The yield of gasoline obtained by

382

this method is comparable to the yield obtained from petroleum. Additionally, the gasoline produced has a sufficiently high octane number for direct use in automobiles. In principle, all or some of these operations might be directly applied for the extraction and upgrading of bacterial hydrocarbons. 5. FUTURE PROSPECTS One of the main features impelling the study of microalgae as an alternative source of hydrocarbons lies in their ability to fix carbon dioxide through photosynthesis. The possibility of reducing the atmospheric carbon load by direct recycling into fuels is appealing [63]. Despite all efforts, the production of algal hyodrocarbons is not competitive with petroleum derived fuels, mainly due to the slow growth rate of microalgae, the low extraction yields and the high viscosity of the cultures. The alternative approach of cloning the algal hydrocarbon synthesis genes into other microorganisms has proven to be difficult. Nevertheless, the necessity for an alternative source of hydrocarbons, not only as fuels, but also for the fine-chemicals industry is still there. As presented in this document, the ability to synthesize hydrocarbons is widely distributed among eubacteria. The biosynthetic pathways seem to be completely different compared to those observed in plants, involving a novel set of enzymatic activities. At the present time, two different strategies might be suggested in order to increase the accumulation of aliphatic hydrocarbons in bacteria. A direct one, involving the identification and cloning of the genes encoding the relevant activities involved, aimed at their subsequent expression in a suitable host might result in an important advance. However, this strategy might not be as straightforward as appears. It is well understood that the arbitrary modification of the cellular carbon fluxes might severely impair cell viability and performance. In this case, the natural deviation of the cellular carbon pool into a reserve compound (hydrocarbons) might not be gratuitous but the result of a finely tuned metabolic network. The deeper understanding of the physiology of the process might eventually provide the knowledge basis for the rational design of an imbalanced metabolism yielding the desired accumulation of hydrocarbons without compromising cell viability. The current availability of more powerful analytical as well as genetic tools enables us to face this challenge. Aknowledgements The author thanks Shirley Ainsworth for assistance during the bibliographical investigation. This work was supported by PEMEX grant 138.

383

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] II1] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

P.W. Albro and J.C. Dittmer, Lipids, 5 (1970) 320. T.J. Savage, M.K. Hristova and R. Croteau, Plant PhysioL, 111 (1996) 1263. P. Metzger, C. Largeau and E. Casadevall in W. Herz, G.W. Kirby, W. Steglich and Ch. Tamm (eds.) Progress in the Chemistry of Organic Natural Products, Springer-Verlag, New York (1991). C. Hall, P. Tharakan, J. Hallock, C. Cleveland and M. Jefferson, Nature, 426 (2003) 318. A.L. Demain, Biotechnol. Adv., 18 (2000) 499. J.D. Rozzell, Bioorg. Med. Chem., 7 (1999) 2253. W. Gerhartz (eds.), Enzymes in industry: production and applications, VCH, Germany, (1990). A. Liese and M.V. Filho, Curr. Opin. Biotechnol., 10 (1999) 595. R. Vazquez-Duhalt, E. Torres, B. Valderrama and S. Le Borgne, Energy Fuels, 16 (2002) 1239. Chemistry and Biochemistry of Natural Waxes, Elsevier, Amsterdam, (1976). M.G. Aarts, C.J. Keijzer, WJ. Stiekema and A. Pereira, Plant Cell, 7 (1995) 2115. M. Ashburner (eds.) Drosophila. A laboratory handbook, Cold Spring Harbor Laboratory Press, Cold Spring Harbor (2003). RJ. Hamilton (eds.), Waxes: Chemistry, molecular biology and functions, The Oily Press, Dundee, (2003). T. Rezanka, J. Zahradnik and M. Podojil, Folia Microbiol, 27 (1982) 450. R. Vazquez-Duhalt, Current Topics in Phytochemistry, 14 (1995) 69. A. Banerjee, R. Sharma, Y. Chisti and U.C. Banerjee, Crit. Rev. Biotechnol., 22 (2002) 245. M.-O. Park, M. Tanabe, K. Hirata and K. Miyamoto, Appl. Microbiol. Biotechnol., 56 (2001) 448. T.G. Tornabene, S.J. Morrison and W.E. Kloos, Lipids, 5 (1970) 929. E.G. Dediukhina and V.K. Eroshin, Usp. Sovrem. Biol, 76 (1973) 351. J. Baraud, A. Maurice and C. Napias, Bull. Soc. Chim. Biol., 52 (1970) 421. E.G. Dediujina, V.P. Zhelifonova, L.V. Andreev and B.K. Eroshin, Prikl. Biokhim. Mikrobiol, 9(1973)813. H.L. Ehrlich in C.R. Bell, M. Brylinsky and P. Johnson-Green (eds.) Microbial Biosystems: New Frontiers, Atlantic Canada Society for Microbial Ecology, Halifax, Canada (1999). R.W. Stone and C.E. ZoBell, Ind. Eng. Chem., 44 (1952) 2564. C.E. ZoBell, World Oil, 130 (1950) 128. C.E. ZoBell, J. Sediment Petrol., 22 (1952) 42. G.J. Jankowski and C.E. ZoBell, J. Bacteriol., 47 (1944) 447. C.E. ZoBell, Science, 102 (1945) 364. P.W. Albro and C.K. Huston, J. Bacteriol., (1964) J.G. Jones and B.V. Young, Arch. Mikrobiol., 70 (1970) 82. J. Han and M. Calvin, Proc. Natl. Acad. Sci. U. S. A., 64 (1969) 436. T.G. Tornabene, J. Mol. Evol, 11 (1978) 253. T.G. Tornabene, T.A. Langworthy, G. Holzer and J. Oro, J. Mol. Evol., 13 (1979) 73. J. Han, E.D. McCarthy, W. Van Hoeven, M. Calvin and W.H. Bradley, Proc. Natl. Acad. Sci. U. S. A., 59 (1968) 29.

384

[34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63]

J. Oro, T.G. Tornabene, D.W. Nooner and E. Gelpi, J. Bacteriol, 93 (1967) 1811. J.B. Davis, Chem. Geol., 3 (1968) 155. J.G. Jones, J. Gen. MicrobioL, 59 (1969) 145. J. Templier, C. Largeau and E. Casadevall, Phytochemistry, 23 (1984) 1017. X. Wang and P.E. Kolattukudy, FEBS Lett., 370 (1995) 15. J. Vioque, T. Sirakova and P.E. Kolattukudy, J. Phycol., 35 (1999) 121. X. Wang and P.E. Kolattukudy, Biochem. Biophys. Res. Commun., 208 (1995) 210. J. Vioque and P.E. Kolattukudy, Arch. Biochem. Biophys., 340 (1997) 64. K.D. Lardizabal, J.G. Metz, T. Sakamoto, W.C. Hutton, M.R. Pollar and M.W. Lassner, Plant Physiol, 122 (2000) 645. S. Reiser and C. Sommerville, J. Bacteriol., 179 (1997) 2969. J.W. Lin, Y.F. Chao and S.F. Weng, Biochem. Biophys. Res. Commun., 191 (1993) 314. J. Benach, S. Atrian, R. Gonzalez-Duarte and R. Ladenstein, J. Mol. Biol., 282 (1998) 383. M.W. Dennis and P.E. Kolattukudy, Arch. Biochem. Biophys., 287 (1991) 268. F. Schneider-Belhaddad and P.E. Kolattukudy, Arch. Biochem. Biophys., 377 (2000) 341. M. Dennis and P.E. Kolattukudy, Proc. Natl. Acad. Sci. U. S. A., 89 (1992) 5306. M. Bard, D.A. Bruner, C.A. Pierson, N.D. Less, B. Biermann, L. Frye, C. Koegel and R. Barbuch, Proc. Natl. Acad. Sci. U. S. A., 93 (1996) 186. B.A. Arthington, L.G. Bennett, P.L. Skatrud, C.J. Guynn, R.J. Barbuch, CD. Ulbright and M. Bard, Gene, 102 (1991) 39. J.R. Reed, P. Hernandez, G.J. Blomquist, R. Feyereisen and R.C. Reitz, Insect Biochem. Mol. Biol., 26(1996)267. S. Gencic and D.A. Grahame, J. Biol. Chem., 278 (2003) 6101. E. Kocsis, M. Kessel, E. DeMoll and D.A. Grahame, J. Struct. Biol., 128 (1999) 165. P.W. Albro, J. Bacteriol., 108 (1971) 213. P.W. Albro, T.D. Meehan and J.C. Dittmer, Biochemistry, 9 (1970) 1893. S. Kikuchi, D.L. Rainwater and P.E. Kolattukudy, Arch. Biochem. Biophys., 295 (1992) 318. P.W. Albro and J.C. Dittmer, Biochemistry, 8 (1969) 1913. P.W. Albro and J.C. Dittmer, Biochemistry, 8 (1969) 3317. J. Frenz, C. Largeau, E. Casadevall, F. Kollerup and A.J. Daugulis, Biotehnol. Bioeng., 34(1989)755. J. Frenz, C. Largeau and E. Casadevall, Enzyme Microb. Technol., 11 (1989) 717. R.L. Mendes, J.P. Coelho, H.L. Fernandes, IJ. Marrucho, J.M.S. Cabral, J.M. Novais and A.F. Palavra, J. Chem. Technol. BiotechnoL, 62 (1995) 53. H. Kitazato, S. Asaoka and H. Iwamoto, Sekiyu Gakkaishi, 32 (1989) 28. P. Pedroni, J. Davison, H. Beckert, P. Bergman and J. Benemann, J. Energy Environ. Res., 1 (2001) 136.


3 85

Chapter 14

The microbial diversity of deep subsurface oil reservoirs N.-K. Birkeland Department of Biology, University of Bergen, Box 7800, N-5020 Bergen, Norway

1. THE DISCOVERY OF MICROBIAL LIFE IN DEEP OIL WELLS "Souring" of oil reservoirs by the formation of hydrogen sulfide has been a problem since the beginning of commercial oil production (see chapter 11). Whether the sulfide is a result of abiotic chemical processes or due to microbiological activities has been debated for many decades. The first indications of an active role of sulfidogenic bacteria in this process were presented already in 1926 based on the observation that sulfate-reducing bacteria were widespread in oil-well production waters [1], Thermophilic sulfatereducing bacteria recovered from water produced from a North Sea oil-well was in 1991 found to be able to survive, multiply and actively produce sulfide under simulated reservoir conditions at temperatures and pressure up to 80°C and 4,500 psi, respectively [2, 3], thus demonstrating that sulfide formation can be caused by microorganisms even under the extreme physical conditions found in deep and hot petroleum reservoirs. Hyperthermophilic organisms able to grow at even higher temperatures were soon to be recovered from deep oil wells in Alaska [4] and the North Sea [5]. The question whether these bacteria were contaminants that had been introduced to the oil wells during drilling or through the repressurization by water injection into the oil-bearing strata, or whether they belonged to an indigenous community of subsurface microbes remained an open question. During the last decade, however, our perception of this has changed with the recovery from numerous oil wells of a large number of anaerobic microorganisms representing a wide range of different metabolic types, including sulfate- and iron-reducing bacteria, fermentative bacteria and methanogenic Archaea. A number of different thermophilic and hyperthermophilic organisms have now been recovered from both terrestrial and offshore

386

wells that have never been water-flooded [6, 7], strongly indicating an indigenous origin. The presence of an indigenous microbial community is further supported by the isolation of novel species that never has been recovered from any other sources, and by the physiological characteristics of some isolates indicating a close adaptation to the respective in situ reservoir conditions. Recovery of closely related strains from remote oil fields [7-9] also supports the existence of a widespread microbial biosphere in oil-bearing strata. However, the problems associated with recovery of biological samples from oil wells are extensive. Sampling from wellheads is the only way of collecting samples from petroleum reservoirs, and the possible sources of contamination are numerous. It is furthermore possible that exogenous mesophilic bacteria can propagate in top facilities of the oil field installations. Occasionally, aerobic and microaerophilic bacteria are recovered from produced oil-well water, but available chemical data suggest that oxygen is absent in oil reservoirs, and these isolates should thus not be considered as being truly indigenous to deep oil wells. In addition to SRB and fermentative bacteria, Voordouw et al. [10] detected several aero- and microaerophilic bacteria in a 600 m deep water flooded oil reservoir in Canada. Nitrate-reducing bacteria were recovered from a similar shallow oil field [11]. It is postulated that oxygen and nitrate is able to reach these shallow oil-bearing formations through diffusion or convection from surface layers, giving support to a limited community of bacteria respiring with nitrate or oxygen [10, 11]. The number of bacterial cells in water produced from oil reservoirs is highly variable. Total bacterial counts demonstrated the presence of more than 106 cells per ml in water from a non-water flooded reservoir in California [6], and from sulfide-rich production water from a German water-flooded petroleum reservoir up to 6.3 x 106 colony-forming units of sulfate-reducing bacteria per ml has been obtained [12]. Although only a few bacteria per ml have been detected in water produced from some oil wells, these results show that the bacterial density can be significant. In the present chapter our current knowledge of these microorganisms is reviewed. 2. SULFATE-REDUCING BACTERIA AND ARCHAEA (SRB) Sulfate-reducing prokaryotes constitute a diverse physiological group of sulfideproducing microorganisms able to carry out anaerobic respiration with sulfate as a terminal electron acceptor. They are widespread in anaerobic environments were sulfate is available. Typically, they oxidize organic acids either to acetate, or by complete oxidation of acetate or other acids to CO2, but autotrophic species using H2 and CO2 as energy and carbon-sources are also common. A wide range of organic acids, e.g. acetate, propionate, butyrate, pentanoate and hexanoate, at concentrations up to 20mM has been found in oil reservoirs [13, 14]. On the other hand, the concentration of sulfate seems to be rather low in

387

non-water-flooded reservoirs, and the abundance of SRB is therefore probably sulfate-limited. There are few reports on the quantification of SRB in produced water, but the few estimates that have been made show a considerable variation. As mentioned above, up to 6.3 x 106 colony-forming units of SRB per ml was obtained from wellhead water from a German sulfide-rich reservoir [11]. Acetate was used as a substrate. In a more recent survey, up to 4.5 x 104 SRB per ml were counted in wellhead water from low-temperature reservoirs using most probable number (MPN) technique and with lactate + acetate as substrates [15]. The number of SRB decreased with increasing in situ temperature, and from wells with an in situ temperature of 85° no SRB could be detected. These results are comparable to the results reported by Nazina et al. [16] and Rozanova and Nazina [17], who detected only very low populations of SRB in hightemperature reservoirs. A low number of SRB have also been detected in reservoirs of high salinity. In a study of sulfate reducers in a high-temperature oil field in the North Sea up to 2 x 104 thermophilic or hyperthermophilic SRB were detected directly in the produced water using fluorescent antibody technique with conjugated antibody directed against three specific SRB groups [18]. No correlation between the duration of seawater injection and the number of SRB in the water was observed, indicating that thermophilic and hyperthermophilic SRB are indigenous to this oil field. The salinity of oil-well water is very variable and is an important factor for the in situ microbial activity and diversity. Oil-well brines with salinity above 20% have been found, but most oil well formation waters have a moderate salinity (10). Empty bed retention time (EBRT) is generally between 30 seconds and 2 minutes. Due to the type of supports used, the height of the packed bed is generally about 0.8 to 1.2 m, making thus necessary to have a large footprint, which may be a disadvantage for situations where space is limited. 2.2.2. Biotricklingfilters (BT). In BT, the polluted air (Fig. 4) flows upflow or downflow through a packed column where liquid is continuously recirculated. The pollutant is first solubilized in the falling liquid film and then transferred to the biofilm developed on the support. The liquid provides moisture, nutrients, pH control to the biofilm and allows the removal of inhibiting products and excess biomass. Table 1. Classification of biological reactors Biomass Fixed on a support Fixed on a support

Liquid phase Stationary Flowing

Suspended Suspended or fixed Fixed on a membrane

Flowing Stationary Flowing

Reactor Biofilter, BF Biotrickling filter, BT Rotating contactors, RC Bioscrubber, BS Suspended growth, SR Membrane, MR

483

Fig. 3. Schematic representation of bio filter (BF).

Inert random supports or structured packing are used. Some examples include plastic corrugated structured PVC sheets, Raschig or Pall rings and saddles, lava rock and polyurethane foam [2]. To maintain low pressure drop and reduce clogging, the supports have low porosity and low specific surface (100 - 400 m2 m"3). EBRT are normally around 30 seconds but systems with EBRT as low as 2 seconds have been reported for low H2S concentration [9].

Fig. 4. Schematic representation of biotrickling filter (BT)

484

2.2.3. Bioscrubbers (BS). In bioscrubbers, the pollutant in the gas phase is first absorbed in a gasliquid contactor (Fig. 5). Subsequently, it is eliminated in a bioreactor and the liquid, containing the suspended microorganisms, is returned to the contactor. Nutrients and pH regulators can be added to maintain microbial activity and the excess of biomass and sub products can be controlled by purging. The gas-liquid contactors can be packed towers, venturi scrubbers or spray towers [10]. Bioscrubbers are designed to favor mass transfer with low pressure drop (< 3 cm H2O m"1). In the bioreactor, supplementary air is added to favor the oxidation of the pollutant. Water retention time in the reactor is calculated to eliminate the soluble pollutant, and the biomass concentration is generally about 5-8 g L"1 [11] to foster high volumetric rates while reducing clogging problems in the contactor. Bioscrubbers are used for hydrophilic pollutants (H esters > ketones > aromatics > alkanes t Biodegradability decreases with higher number of halogens

489

2.3.4. Performance parameters Three parameters are often used to compare the pollutant treatment efficiency in BATS. Inlet mass pollutant load (IMPL), is defined as the amount of pollutant introduced into the bioreactor normalized by its empty-bed volume. (6) Elimination capacity (EC), is the quantity of pollutant removed per bioreactor volume per time unit. (7) Removal efficiency (RE), is the fraction of the pollutant removed expressed as percentage. (8)

3. EXAMPLES OF TREATMENT OF VOLATILE PETROLEUM HYDROCARBONS BY BATS Petroleum products such as gasoline, fuel oils, and diesel fuels are among the most important water, soil and air pollutants. They are complex mixtures of organic compounds containing a significant volatile fraction. Hydrocarbons are composed of four main structural classes: 1) «-alkanes (linear saturated hydrocarbons), 2) isoalkanes (branched saturated hydrocarbons), 3) cycloalkanes (saturated cyclic alkanes) and 4) aromatics [25]. Moreover, in the case of gasoline, other oxygenated additives such as the ethers MTBE, ethyl tbutyl ether (ETDE) or /-amyl methyl ether (TAME) or alcohols such as ethanol, can be added to improve combustion and consequently air quality. Hydrocarbons may be released into the atmosphere by evaporation during production, transport and storage. Table 5 presents some examples of the degradation of volatile organic compounds from gasoline by BATS. The removal efficiency of volatile individual compound in the gasoline varies from 5 % to 99 %. Aromatics are generally removed more efficiently than n-alkanes. Light aliphatic compounds ( -»

SO 4 (S°) + Biomass H2O + CO2 + Biomass

516

Nitrogen, mainly as ammonium (NH4+), is one of the most abundant contaminants in the petroleum industry wastewaters. Ammonium can be biologically eliminated by means of a double process, nitrification and denitrification, producing molecular nitrogen. Nitrification is a strict aerobic process, litoautotrophic, where ammonium is the electron and nitrogen sources, and it is oxidized to nitrate. Denitrification is a reductive process either heterotrophic or litoautotrophic process, where nitrate is reduced to elemental nitrogen. In this chapter it will be presented some of the more recent developments in biological wastewater treatment technology with application to the petroleum industry. The topics that will be covered are: a) Anaerobic biodegradation of aromatic compounds like phenol, alkylphenols and terephthalate. b) Biotransformation of S- and N-bearing inorganic compounds. c) Methyl-tert-butyl ether (MTBE) biodegradation. MTBE is a high recalcitrant compound and a potential water contaminant that only in few cases can be treated with technology originally developed for biological wastewater treatment. Conventional biological treatment, like activated sludge, is out of the scope of this chapter. 2. ANAEROBIC BIODEGRADATION AND BIOTRANSFORMATION OF AROMATIC COMPOUNDS The implementation of anaerobic wastewater treatment in the petroleum industry was initially limited due to the presumed toxicity and biodegradability of aromatic compounds present in these waste streams. However, the treatment of chemical and petrochemical wastewater has lately become a reality, due to a better understanding of the microbial biodegradation process and the discovery of the methanogenic granular sludge structure, which plays a key role in the development of the so called high rate anaerobic processes. The granular sludge is an aggregation of several metabolic groups of bacteria living in synergism. The granules have a diameter between 0.5 to 3 mm and a biomass concentration of approximately 100 g dry matter I'1. 2.1. Toxicity and biodegradability of phenolic compounds Spent caustic is one of the refineries waste streams, rich in phenolic compounds. This effluent is produced from nonregenerative desulfurization processes that use caustic soda scrubbing in combination with air oxidation. This process is used to remove H2S and CH3SH from gasoline and to remove H2S, CO2 and HCN from sour condensate gas [2]. The effluent, although involves very small volumes and its contain high concentration of phenols and sulfide.

517

The average phenol and alkyl phenols concentrations are 30.5 g 1" and 28.2 g 1' , respectively. There are several reports about the toxic effect produced by phenolic compounds on the acetoclastic methanogenic activity (AMA) of the granular sludge. Table 3 shows the inhibitory concentrations that reduce in 50% (IC50) the AMA. In general, the susceptibility of a granular sludge to the inhibitory effect of phenolic compounds is affected by the impact of its "acclimation history". The phenol-degrading acid-forming bacteria are more susceptible to phenol inhibition than the methanogens [5]. Most granules have a layered structure that protects bacteria, particularly methanogens. In the case of a phenol-acclimated granular sludge, it is possible that a phenol-degraders layer develops in the external zone of the granules, preventing the inward diffusion of the toxic compounds. This outer layer can prevent the methanogens deactivation either by reducing the exposure level or by a partial or complete biotransformation into nontoxic intermediates such as volatile fatty acids [5]. The selection and multiplication of an acetoclastic flora more resistant to those toxic compounds might be another protection mechanism. The inhibitory mechanism of the phenolic compounds is governed by their hydrophobicity that increases the ability of these compounds to solubilize into the lipid bacterial membranes, altering the membrane functions, such as ion transports causing cellular lysis. High linear correlations of methanogenic toxicity data to the logarithm of octanol-water partition coefficients of phenolic compounds (log P) have been proposed as shown in Fig. 1. This simple model adequately estimates the IC50 values for anaerobic granular sludge in the presence of phenolic compounds. Table 3 Inhibitory concentrations that reduce in 50% (IC50) the acetoclastic methanogenic activity of granular sludge (phenol-acclimated and non-acclimated) in batch assays [6-8]. Compound Phenol o-cresol m-cresol p-cresol 3,4-dimethylphenol 2-ethylphenol 4-methylphenol 4-ethylphenol

IC50 (mg I"1) 470 - 7802 433 - 844 443 - 919 389-1525 329-378 195-207 657 289

518

Fig. 1. Relationship between IC50 of phenolic compounds and the octanol/water partition coefficient (Log P). Synthetic "spent-caustic phenols mixture" (X), data from reference [6] (•), data from reference [7] (•) and data from reference [8]. (A). (1), phenol; (2), 4-methylphenol; (3), 4-ethylphenol; (4), o-cresol; (5), m-cresol; (6), /)-cresol; (7), 3,4-dimethylphenol; (8), 2-ethylphenol; (9), "synthetic phenols mixture". Log (l/ICso) = 0.77 Log P - 2.28, r2 = 0.90

Phenol is a compound easily biodegradable under anaerobic conditions. The biodegradation is initiated by phosphorylation of hydroxyl group followed by carboxylation of the ring in the para position (benzoyl-CoA pathway). In the case of the three cresol isomers (p-, m- and o-cresol) there are differences in their anaerobic biodegradability pathways. Methanogenic consortia are capable of /?-carboxylate the m-cresol and the o-cresol to their methylbenzoic acids. After carboxylation, the main degradation mechanism is the oxidation of the methyl group and in case of the /?-cresol it leads to the formation of a metabolic intermediate, the p-hydroxybenzaldehyde. It has been reported that p-cresol degradation also initiates by fumarate addition to the methyl group, forming benzyl-succinate. There are few reports about o-cresol biodegradation, since this compound is considered hard to be degraded under anaerobic conditions. Biodegradation rates of a mixture of phenol, p- and o-cresol obtained in batch experiments, with an adapted granular sludge, were approximately two-orders of magnitude higher than those observed with non-adapted sludge [9]. From evaluating the interaction of substrates, it was observed that p- and o-cresol did not affect phenol biodegradation, however, both phenol and o-cresol negatively affected /?-cresol biodegradation at the concentrations tested [9]. In other study, degradation of />-cresol ceased when phenol was depleted. This suggests that degradation of the most refractory p-cresol also requires phenol as a co-substrate. However, after a period of acclimation to the phenol-free environment, the biomass was able to degrade p-cresol without any co-substrate [10]. So far, both xylenols and ethylphenol biodegradation has not been reported

519

under methanogenic conditions. A reversible reaction from 2-ethylphenol to 3hydroxy-4-ethylphenol seems to take place, but no further degradation has been described. 2.1.2. Anaerobic treatment systems for the biodegradation of phenol Anaerobic treatment of phenolic-bearing wastewaters produced from the petroleum industry is a viable option. The bioreactor system most commonly used for the anaerobic treatment of phenolics is the UASB, operating to a certain organic loading rate (OLR), usually referred to chemical oxygen demand (COD). Lab scale UASB reactors have been applied to treat single phenolic compounds at OLR as high as 6 and 7.2 g COD I"1 d"1 for phenol and />-cresol, respectively, showing high compound removal efficiencies [11, 12]. However, effluents from the petroleum industry are expected to contain mixtures of phenol and cresols as the main COD bearing fractions. Thus, a successful treatment of these effluents would require a simultaneous degradation of the major phenolic substrates. Table 4 shows some results of anaerobic treatment of phenolic compounds mixtures. Table 4. Continuous anaerobic treatment results of mixtures of phenolic compounds treated in upflow anaerobic sludge bed reactors. Mixture

OLR (g COD I ' d 1 )

COD removal (%)

Reference

Phenol p-Cresol

7

94

[9]

7.1

91

[9]

2.95

81.8

[9]

8.12

85

[10]

0.66

85

[13]

4.3

-

[15]

Phenol /?-Cresol Phenol jt?-Cresol

o-Cresol Phenol p-Cresol Phenol />-Cresol o-Cresol Phenol m-Cresol

520

The operational parameters of the UASB reactors have important implications on the biodegradation efficiency of the phenolic compounds. In general, the effect of OLR is more drastic in reactors with increased phenolic concentrations, than in reactors with a constant phenolic concentration, and a decreased hydraulic retention time [10]. In the same way, the phenol/cresols ratio has to be controlled to avoid inhibitory or toxic effects to the living biomass. Cresol concentrations higher than 600 mg 1" can cause severe inhibition on the activity of the granular sludge [9, 10]. In a typical reactor wih phenol is used as sole source of carbon and energy, granulation is reported to initiate after 3 months of the start-up operation and develops for 6 months, to become fully mature. Granular sludge cultivated has an average diameter of 1.8 mm and is highly settable with a settlement volumetric index (SVI) of 14 ml g"1 [5]. The removal of phenolic mixtures can be improved in an UASB reactor using bioaugmentation. This method not only improves the start-up time, but also the COD removal. The bioaugmentation can be performed by simple adsorption of the specific bacterial consortium onto the granules, to protect it from being washed-out [13]. An increase of the enrichment from 2 to 5% improved considerably the start-up of the reactors treating phenolic compounds [14]. It was until 1981 that the two first full scale reactors treating chemical wastes were built by Celanese Company in USA. A third reactor was built three years later, and by 1989, 19 full-scale reactors were in operation treating wastewater from the chemical and petrochemical industry. Since 1990, the rate of digesters construction for that industrial sector increased from 2.1 to 4.6 reactors per year. Although an UASB reactor has been in operation since 1986 to treat phenol-bearing wastewater, no other reactor has been built to treat the same type of effluents since then. This UASB reactor of 1280 m3 is treating a 30.5 g COD I"1 with an OLR between 9 to 12 g COD l"'d"' and a COD removal of 95% has been achieved [3]. 2. 2. Toxicity and biodegradability of terephthalic acid Phthalic acid isomers (benzene-dicarboxylic acid) are important constituents of polyester fibers, films, polyethylene terephthalate (PET) bottles and other plastics. During production of phthalic acids, an important volume of wastewater is generated, approximately 3-10 m3 per ton of purified terephthalic acid (PTA) containing 5-20 kg COD m"3 [16]. The main components in the wastewater are terephthalic acid, acetic acid, benzoic acid and p-toluic acid in decreasing order of concentration. After neutralization with NaOH, all acids are present as sodium salts. Due to the characteristics of these wastes, anaerobic

521

pretreatment has been generally recognized as beneficial for wastewater treatment. 2.2.1. Toxicity and biodegradation Terephthalate concentration of 5 g COD I"1 does not produce any substrate inhibition on its biodegradation and methanogenic activity [17]. Hydrogenotrophic methanogenesis inhibition by 4-carboxybenzaldehyde, ptoluate and terephthalate generates IC50 values of 0.8 g I"1, 4.6 g I"1 and 16.6 g I"1, respectively. Nonetheless, methane production can be inhibited by un-ionized terephthalic acid, in near colloid state, using a settled terephthalic acid wastewater (pH 4.5) or purified terephthalic acid (0.183 g g'VSS) adjusted to pH6.15 [18]. The very low specific growth rate of terephthalate-biodegrading bacteria (0.04 h"1) explains the long-lasting acclimation period and low loading rate applied in UASB reactors [17]. The use of co-substrates like sucrose, benzoic and acetic acids inhibits the terephthalate and p-toluate biodegradation. The addition of benzoate delays the terephthalate biodegradation, which resembles a diauxic inhibition [17]. The generally accepted metabolic pathway of terephthalate biodegradation is the benzoyl CoA pathway after a probable decarboxylation leading to the formation of benzoate. The decarboxylation step is thermodynamically favorable under standard conditions, while the conversion of benzoate to acetate is a highly endergonic process. The global conversion of terephthalate to acetate and H2 becomes exergonic only when acetate and H2 are at very low concentrations. The fermentation of co-substrate by methanogenic granular sludge results usually in the production of H2 and acetate, generating an increase in the AG 0 ' which, in turn, may limit the terephthalate biodegradation. Analysis of specific activities of terephthalate and benzoate biodegradation demonstrated that terephthalate biodegradation activity was lower with a 33.6 mg COD g"'VSS d"1 value versus 117 mg COD g"1 VSS d'1 value for benzoate activity. Thus, the initial conversion of terephthalate to benzoate seems to be the limiting step of the microorganisms involved in terephthalate anaerobic biodegradation. Three bacterial populations were involved: 1) a syntrophic organism similar to that described for Syntrophus buswellii [19] able to convert terephthalate into acetate, CO2 and H2; 2) an acetoclastic methanogen; and 3) a hydrogenotrophic methanogen. 2.2.2. Terephthalate anaerobic treatment at full scale Crude terephthalic acid wastewaters must fulfill some conditions to be successfully pre-treated with anaerobic process: Certain grade of effluent neutralization, limiting concentration of other substrates than terephthalate or benzoate or acetate, high biomass retention rate and low volumetric loading rate.

522

The best operation is obtained with a plug-flow process or staged reactor system, because no substrate toxicity has been reported in normal operation with neutralized effluent. The company Amoco Petrochemicals Inc. operates a 15200 m3 full-scale downflow fixed film reactor with an OLR of 4.0 kg COD m3 d"1, which demonstrated the real feasibility of such pre-treatment [20]. The use of an expanded granular sludge bed-type bioreactor allowed a terephthalate removal higher than 80% and steady COD removal of 60% at an upflow velocity of 10 m d"1; however, in such conditions, p-toluate appeared to be recalcitrant to degradation [21]. Up to date, there are more than 10 full scale reactors treating terephthalic acid, indicating that the anaerobic treatment has become a conventional treatment for this kind of wastewater. The used bioreactor configurations are UASB, expanded granular sludge bed and hybrid reactors [22]. 3. BIOTRANSFORMATION OF S- AND N-BEARING INORGANIC COMPOUNDS FROM SOUR STREAMS The microbial treatment of sour wastewater resulting from either oil production or refining and other fossil fuels has been subject of intensive worldwide studies. The term "sour" was originated to describe those wastes contaminated with sulfide [23]. In refineries, sour wastewaters are generated from sour steam condensates that have been in contact with petroleum products, specifically from thermal or hydrogen cracking operations, where a carrier steam is used for injection or aeration [24]. Common total sulfur contents in sour water are around 1194 mg I"1. Because of the high sulfide, ammonium and phenols content, sour wastewater must be treated before its release into the environment. Both, aerobic and anaerobic processes have been reported to treat sour waste streams. 3.1. Aerobic processes Aerobic Thiobacilli species, which oxidize reduced sulfur compounds to obtain their growth energy, have been studied to promote the sulfur production from partial sulfide oxidation as shown in Eq. (1) [25, 26, 27]. These bacteria are gram-negative rods of about 0.3 um in diameter and 1 to 3 um long and belong to the colorless sulfur bacteria. An important characteristic is their capacity to excrete elemental sulfur, in contrast to filamentous colorless sulfur bacteria, as Thiotrix sp., which accumulate it intracellularly. Sulfur production from partial oxidation of sulfide instead of a complete oxidation to sulfate has a significant relevance because elemental sulfur can be recovered from the medium closing the sulfur cycle. Additionally, lower energy consumption is required because the oxidation to sulfur requires 4-fold less oxygen that the complete oxidation to sulfate, as shown in Eq. (2).

523

(1) (2) The reactor configuration, to promote both sulfur formation and accumulation, was evaluated and reported by Janssen et al. [27] and Alcantara et al. [28]. The configuration consisted mainly in the separation of aeration process from the bioreactor. Thus the liquid saturated with oxygen from the aerator vessel is sent to the reactor (reaction vessel) at a specific rate, which allows the control of stoichiometric molar ratio between oxygen and sulfide (theoretical molar ratio, Rmt, O2/S2~). When Rmt is close to 0.5, the sulfide oxidation is driven to elemental sulfur formation, while a Rmt close to 2 promotes sulfate as the main product. The performance of the system reported by Alcantara et al. [28] was inoculated with a sulfoxidizing consortium and it is shown in Fig. 2. Sulfide oxidation was studied under different dilution rates at steady state conditions of 0.5, 1, 1.5, 2 and 3 d"1 (zones A, B, C and D, respectively), maintaining a constant sulfide concentration in the feed solution at 4.0 g I"1. Elemental sulfur was produced at dilution rates of 0.5, 1, 1.5 and 2. The maximum sulfur formation occurred at Rmt of 0.5, where 85% of the total sulfur added to the reactor as sulfide was transformed to elemental sulfur and 92% of it was recovered from the bottom of the reactor.

Fig. 2. Performance of the recirculation reactor system under different culture conditions. Capital letters corresponds to the following dilution rates (d"1): A, 0.5; B, 1; C, 2 and D, 3. Subtitle letters show the Rmt evaluated: 2: b, c; 1.5: d; 1, e, k; 0.75: f, 1: m; 0.5: a, g; 0.35: h; 0.25: i; 0.15: j . Sulfide influent (—), sulfate (•), elemental sulfur (A), thiosulfate (o) and sulfide effluent (A).

524

Elemental sulfur production was affected by the dilution rate applied to the system. When the system operated at Rmt for sulfur production (0.5 and 0.75) and dilution rates of 0.5, 1 and 2, the elemental sulfur produced was higher than 60%, while washout conditions were observed when the dilution rate was increased from 2 to 3, at a Rmt of 0.75. The Thiobacilli species are strict autotrophic bacteria, thus organic compounds negatively affect their growth. However, sulfoxidizing consortia have shown an adequate metabolism to oxidize reduced sulfur and organic sulfur (CS2 for example) compounds [28], in presence of organic matter. According to Sublette et al. [23] and Alcantara et al. [28, 29], the oxidation of sulfur compounds is carried out by autotrophic bacteria while organic compounds are used as energy and carbon source by heterotrophic microorganisms. Phenol, o-, m- and />-cresol were degraded in a chemostat at various organic loading rates by the consortium. Under all conditions sulfide was completely oxidized to sulfate. Microcosm experiments showed that carbon dioxide production increased under presence of phenols, suggesting that these compounds were oxidized and they may be used as carbon and energy source by heterotrophic microorganisms present in the consortium [28]. The expanded bed reactor reported by Janssen et al. [27] is actually builtin to a family of processes called THIOPAQ, which are applied for the treatment of wastewater containing sulfide. Also, this technology has been proposed for the treatment of similar streams from petrochemical industries e.g. spent sulfidic caustics and from liquefied petroleum gas (LPG) scrubbers [30]. 3.2. Anaerobic processes Thiobacillus denitrificans is a gram-negative, chemoautotroph and facultative anaerobic bacteria, which oxidizes reduced sulfur compounds to obtain its growth energy and it is able to use nitrate as electron acceptor. According to Cadenhead and Sublette [31], this microorganism shows clear advantages to oxidize sulfide over other Thiobacilli, such as Thiobacillus thioparus, Thiobacillus versutus and Thiobacillus thiooxidans. Sulfide is commonly oxidized to sulfate (Eq. 3) or elemental sulfur (Eq. 4) under anoxic conditions, and where nitrate is used as a terminal electron acceptor being reduced to elemental nitrogen. 1.25 S2" + 2 NO3" + 2 H+ -> 1.25 SO42" + N2 + H2O

(3)

5 S2~ + 2 NO3" + 6 H2O - > 5 S ° + N 2 + 12 OH"

(4)

Sour waste streams, including sour water, sour gases and refinery spentsulfidic caustics, have been successfully treated using Thiobacillus denitrificans. For instance, the organic compounds such as benzene, toluene and phenol are

525

biodegraded by heterotrophic bacteria grown in co-culture with Thiobacillus denitrificans [23, 32]. Sublette [23] identified some technical limitations to apply this technology for the full-scale treatment of sour wastes. These include: substrate inhibition (sulfide), product inhibition (sulfate), the need for septic operation, biomass recycle and recovery, mixed waste issues, and the need for large-scale cultivation of the organism for the process start up. T. denitrificans strain F is sulfide tolerant [33] and it was used to treat oilfield produced water containing sulfides under full-scale field conditions at Amoco Production Co. in Salt Creek Field in Midwest, WY. More than 800 m3 d"1 of produced water containing 100 mg I'1 sulfide and total dissolved solids of 4800 mg I"1 were successfully biotreated in an earthen pit (3000 m3) over a sixmonth period. Based on an average flow of 795 m3 d"1, sulfide influx to the pit was about 80 kg d"'. Complete removal of sulfides and elimination of associated odors were clearly observed. More recently, there has been an increased interest about the oxidation of reduced sulfur compounds in presence of organics under denitrifying conditions [34, 35]. The novelty of this approach is the integration of biological processes that frequently were studied and applied separately. The coupling of carbon, nitrogen and sulfur cycles implicates the oxidation of reduced forms of sulfur, organic compounds, as well as the reduction of nitrate [36, 37]. According to Betlach and Tiedje [38], the heterotrophic denitrification process uses many organic compounds as carbon and energy source; thus organic transformations were coupled to nitrate reduction and further to molecular nitrogen. In the case of autotrophic denitrification, reduced sulfur compounds are oxidized to non-toxic compounds and nitrate, which is used as final electron acceptor, is reduced to molecular nitrogen. Reyes-Avila et al. [36] reported that the critical parameters to steer the nitrate reduction to molecular nitrogen are the C:N and N:S ratios for either heterotrophic or autotrophic processes, respectively. The same authors reported that biological denitrification was used to eliminate carbon, nitrogen and sulfur in an anaerobic continuous stirred tank reactor. Acetate and nitrate at a C:N ratio of 1.45 were fed at loading rates of 0.29 Kg C m"3 d"1 and 0.2 Kg N m"3 d"1, respectively. Under steady state denitrifying conditions, the carbon and nitrogen removal efficiencies were higher than 90%. Under these conditions, sulfide (S2") was fed to the reactor at several sulfide loading rates (0.044 to 0.295 Kg S2" m" 3 d~'). The high nitrate removal efficiency of the denitrification process was maintained along the whole process, whereas the carbon removal was 65%, even at sulfide loading rates of 0.295 Kg S2" m^d"1. The sulfide removal increased up to 99% via partial oxidation to insoluble elemental sulfur (S°) which accumulated inside the reactor.

526

In the same way, a denitrifying fluidized bed reactor for effectively remove sulfide, acetate and nitrate was proposed by Gommers et al., 1988 [39]. The authors reported that the rate-limiting step was the oxidation of sulfur to sulfate, nevertheless, the biomass showed an overcapacity to oxidize sulfide to sulfur and to degrade the acetate, under most tested loads. However, in order to develop a denitrifying technology to treat wastewaters from the petroleum industry, more studies are needed to elucidate the effect of phenolic compounds on both sulfide oxidation and nitrate reduction. 4. OXYGENATED FUEL ADDITIVES Oxygenated gasoline additives have been used since mid-1970s to substitute toxic lead compounds. The most common oxygenated used is methyl tert-hu\y\ ether (MTBE), that became the fourth chemical produced in USA [40] because of their mixing properties, high octane level, low cost and good results in reducing toxic emissions. MTBE is manufactured from isobutene (isobutylene or 2-methylpropene), a byproduct of petroleum refining, and methanol. Therefore MTBE can be easily and inexpensively produced at refineries. The MTBE presence in refinery effluents is due to discharges from facilities as a byproduct of the reprocessing of contaminated or "out of spec" product from the refinery. The volume and type of waste processed by refineries varies greatly over time, resulting in order-of-magnitude variations in the MTBE discharges. Few studies have evaluated the impact of this specific compound in complex wastewater in refineries [41]. MTBE has been present as a pollutant in numerous water resources mainly groundwater, The MTBE environmental impact is enhanced by the high solubility in water, low retention on organic matter, low detection threshold (2.5 and 2.0 jig I"1, for odor and taste, respectively) and low biodegradability. In 1996, the first case of contaminated aquifers by MTBE was reported in Santa Monica, CA. and 250,000 leaking underground fuel tank sites showed different levels of MTBE contamination [40]. In Germany, traces of MTBE were detected in rivers and influents and effluents of wastewater treatment plants [42]. The MTBE half-life in groundwater systems is several years [43]. There are few reports in Mexico about MTBE occurrence in the environment. Air concentrations of 11.5 ppb [44] and 4.4 ppb were monitored at a service station [45] and emissions of on-road vehicles, respectively. Additionally, concentrations between 100-1500 mg kg soil"1 were found in soils at fuels distribution and storage stations [46]. Concentrations in the range of 487 mg I"1 were found in groundwater at the surroundings of gas stations.

527

Fortunately, MTBE was detected in none of the nearby 33 monitored drinking water wells [47]. Increasing reports of MTBE in groundwater produced great concern about the toxicity and the carcinogenicity of this compound. Toxicological studies classified the compound as a potential carcinogen for humans [48] and regulations about the maximal concentration in groundwater were established. An extreme case was adopted in California where MTBE phase out by 2003 was ordered. However, as long as the use of MTBE continues, the risk of its presence in refinery effluents and water resources will be latent and treatments will be required. Due to its unique above-mentioned physicochemical properties, the clean up using common techniques like air injection, activated carbon filtering, etc. are inefficient for MTBE removal. Thus, biological techniques are of particular interest. In this section a review of MTBE biodegradation and biotreatments is done, in order to consider the experience adquired in this area for the eventual treatment of wastewater polluted with MTBE. 4.1 MTBE biodegradation MTBE has become a challenge for elucidation of its low biodegradability and the scarcity of MTBE-degrading microorganisms using it as carbon and energy source. The relatively recalcitrance of MTBE to microbial attack is intrinsic to its structure containing a combination of an ether link and the branched moiety. Alkyl ethers are stable molecules (AG° of the ether bond formation is 360 kJ mol"1 [49]). The high-energy demand for MTBE degradation is reflected by the low efficiency of biomass production on MTBE. Fortin et al, [50] pointed out the low MTBE biomass yield obtained analyzing different consortia. Salanitro [51] suggested that the slow growth on MTBE might also be due to considerable feedback regulation metabolites on the oxygenase responsible for the ether bond cleavage. The necessity of regenerating cofactors, such as NADH, could also have an influence on the rate of MTBE degradation, since reduced cofactors are required for several oxidation steps. Although initial works showed the high recalcitrance of this compound, some authors have reported the biodegradation of MTBE as sole carbon source. Moreover, cometabolism was shown to be an important mechanism for MTBE biodegradation by microorganisms able to grow mainly on short-chain alkanes. Anaerobic MTBE degradation has been recently observed under methanogenic [52], nitrate [53] and Fe(III) reducing conditions [54] with longer adaptation and degradation times. As far as we know, the highest value of MTBE heterotrophic degradation rate of 454 mg g protein"' h"1 was reported for a strain Hydrogenophaga flava ENV735 [55]. For cometabolism, the highest value was

528

obtained by the strain Mycobacterium vaccae JOB5 with a MTBE degradation rate of 111 mg g protein"1 h"1 when hexane was used as growth source [56]. A metabolic pathway for MTBE degradation has been proposed (Fig. 3), where the MTBE ether bond is enzymatically cleaved yielding tert-butyl alcohol (TBA) and formaldehyde as the main metabolic intermediates. TBA has been shown to further biodegrade to 2-methyl-2-hydroxy-l-propanol and 2hydroxyisobutyric acid [57]. Suspected further intermediates of the MTBE degradation metabolic pathway include 2-propanol, acetone and hydroxyacetone. The complete understanding of poor MTBE biodegradability would require the isolation of specialized microorganisms as will as the characterization of genes and enzymes involved in the degradation and regulation. Although microorganisms are able to grow using MTBE as a sole carbon and energy source, we are still far from understanding all causes for its low biodegradability. A number of excellent reviews are available on aerobic biodegradation of MTBE [10, 43, 58, 59].

Fig. 3. Proposed metabolic pathway for aerobic MTBE biodegradation Adapted from Fayolle et al. [49] and Steffan et al. [57].

529

4.2. MTBE removal biotreatments Although MTBE can be removed from groundwater by physical technologies such as activated carbon adsorption and air stripping, the costeffectiveness of these technologies in removal of MTBE is approximately 10 times higher than their application for removal of hydrocarbons, such as benzene and toluene, in groundwater. In December 2003, the USEPA established a database of 356 MTBE polluted sites and treatment technologies [60] including 111 full-scale completed cases. The main technologies used were: soil vapor extraction (18%), pump and treat (17%), in situ bioremediation (21%), air sparging (14%) and other technologies (30%). Bioremediation is a common technology and its cost has been estimated [61]. There are two engineering challenges associated with the in situ aerobic bioremediation of MTBE. First, groundwater polluted with MTBE has very low dissolved oxygen, thus in all cases the addition of air/oxygen is a requirement for the treatment; and the second, is the introduction of microorganisms able to degrade it. Table 5 shows some of the reported cases for in situ treatments. Field treatment includes the formation of a reactive zone named biobarrier by introducing to the subsurface MTBE-degrading microorganisms, which is placed to avoid the advance of the MTBE plume. Oxygen is supplied to the subsurface either by pulse injecting oxygen gas, air or any oxygen release compound. MTBE-contaminated water flowing through the biobarrier will contact the microbes and be degraded to CO2 and water. Biobarriers that have been applied successfully through biostimulation in some field studies, suggest that native microorganisms can degrade MTBE through amendments of nutrients and oxygen. However, the bioaugmentation, by adding microorganisms already adapted to MTBE degradation, has probed to be a more feasible option mainly when time-reduction in the treatment is required. In Salanitro's work, a comparison between biostimulation and bioaugmentation was performed [62]. The author found a notable MTBE reduction in both cases, but there was a difference of approximately 150 days in the lag phase between the treatments, achieving the total bioremediation of the site in approximately 200 days. Other example of biostimulation versus bioaugmentation was performed by Wilson et al. [63]. After six months a noticeable decrease in MTBE was achieved in both inoculated (with PM1 strain) and non-inoculated zones. Polymerase chain reactions techniques showed that in non-inoculated zone there was the presence of PM-1 like bacteria [64].

Table 5 Technology Performance for MTBE biological removal Treatment

Scale

Microorganism

MTBE initial concentration

Treatment period or removal rate

Reference

Field Field

MC-100 Native microorganisms PM-1 ENV425 Native microorganisms

7 mg I"1 1.5 mgT 1

150-200 days 4 days

[62] [63]

320 mg r 1 19.6 mgl- 1

90 days 60 days

[65] [66]

Hydrogenophaga flava Mixed culture Cytophaga-Flexibacter-Bacteroides

1000 mgl"1 5 mg T1

42 mg I"1 h"1 2.5 mg I"1 rf1

[67] [68]

ENV735

10 mg T1

10 mg 1"' in 15 min

[65]

10-50 mgT 1

29 mg r 1 h '

[69]

9 mg r 1 If1

In situ treatment

Field Field Ex situ treatment Laboratory Membrane Laboratory Fluidized bioreactor

Laboratory

Laboratory Mixed culture, cometabolism isopentane

Biotrickling filter Biofilter Biofilter N.S. not specified

Field Laboratory Field

Mixed culture Mixed bacterial culture N. S.

9.6 mg r 1 8.25 mg I"1 10 mgl"' and 15mgr'TBA

4.5 g rf' MTBE and 6.2 g hf' TBA

Laboratory Laboratory Laboratory

F-consortium PM-1 P. aeruginosa

0.8 mg I"1 100 mg 1"' 1.1-12.3 mgl"1

50 mg r1 h"1 58 mg r1 h"1 l.Smgr'h' 1

1

ls.smgr'h"

[69] [70] [71] [72] [73] [74]

531

On the other hand, the addition of a cosubstrate (propane) (US patent 5,814,514, Sept 29, 1998 and US patent 6,194,197 Feb 27,2001) to promote the cometabolic biodegradation of MTBE was useful for groundwater in situ bioremediation [65]. The authors inoculated a propane oxidizing strain ENV425 to cleanup the polluted site by installing biosparging and propane injection systems, and obtained a reduction of 90% in 90 days of treatment. This treatment should be preferred when polluted sites present hydraulic problems or for sites where groundwater extraction is required to stop the migration of contaminant plumes toward neighboring receptors. Bioreactors performance has been studied at lab scale and in some field applications (see Table 5). Most of the investigated bioreactors use immobilized microorganisms including membrane and fluidized reactors. Membrane technology retains high biomass levels improving the volumetric performance and reducing the area for treatment. However, limitations of this technology are the economic cost associated with the capital investment, low service-life and moderated operating costs associated with the pressure-driven mechanism of separation. Membrane fouling can also be a cost factor depending upon feed water conditions that might require pretreatment. Table 5 shows some works using this technology. In fluidized bioreactors, the biomass is immobilized in a support material (granular activated carbon, GAC, is commonly used) and this particles are in continuos movement using an upward water flow. Fluidization significantly increases the specific surface area available for biomass and thus degradation of contaminants. Besides the use of GAC as the fluidizing bed medium also increases specific surface area available for microbial colonization. These reactors avoid the bed plugging problems associated with a fixed bed bioreactor, but special care with operational flows should be taken to avoid washout the bed. However, this type of bioreactor requires a higher degree of operator maintenance and process control than the other readily available treatment processes. Some of the fluidized bioreactor studies are shown in Table 5, including two field experiences. MTBE treatment in vapor phase emissions is necessary when any stripping technology (soil vapor extraction, air stripping, etc.) is used for cleaning up groundwater containing MTBE (see chapter 17). Basically two configurations have been proved: Biofilters and biotrickling filters (Table 5). Biofilters use organic (diatomeaceous earth) or inert (vermiculate or granular activated carbon) packing material to support the microorganisms with non-addition or sporadic nutrient addition. Biotrickling filters are

532

similar to biofilters, but they have an aqueous phase trickling over the packed bed. The liquid contains essential nutrients and it is usually recycled. Biotrickling filters are more complex than biofilters but are usually more effective, especially for the treatment of compounds that generate acidic by-products (see chapter 17). 5. PERSPECTIVES It is expected that more stringent environmental regulatory actions will be taken by governments, worldwide. As water is the most important resource for human, animal and plant life, holistic environmental wastewater management will continue to gain in importance with time [75]. During the last decade significant efforts were devoted to the development of technologies for process integration targeting energy conservation and waste reduction. Great efforts have been done in industries in order to increase the water conservation and reduce wastewater [76]. However, these integrated technologies will produce less and more concentrated wastewater whose characteristic would lead to a complete redesign of the biological wastewater treatment processes that are currently applied on the process industry. Consequently, facility upgrading, innovative and sustainable treatment technologies would reshape the petroleum industry. The anaerobic processes for the treatment of organic compounds in industrial wastewater offer important advantages over conventional aerobic processes. To date, less than 15% of the nearly 1600 full-scale anaerobic wastewater treatment systems are used by the chemical and petrochemical industry. However, as the range of compounds that are found to be biodegraded under anaerobic conditions has increased enormously lately, a large potential expansion seems possible in the future [22]. Thanks to a combination of a simple construction and a high volumetric treatment capacity, the UASB reactor is the dominant concept in the industrial anaerobic wastewater treatment and it probably will keep reigning in the future. Nonetheless, higher loaded expanded granular sludge bed reactors will gradually replace at least part of the UASB applications. In the case of wastewater streams rich in reduced sulfur compounds, the new sulfur biotechnology has allowed the development of reactor systems to remove sulfide producing elemental sulfur. This technology has been adapted for the sweetening of natural gas [30] and more recently for liquefied petroleum gas (LPG), which contains predominantly sulfide and lower alkylthiols [77]. The latter process involves three steps: 1) extraction of the sulfur compounds from the liquefied hydrocarbon phase to a mild

533

carbonate solution in an absorption column; 2) anaerobic conversion of alkylthiols to sulfide and methane in an UASB reactor; and 3) partial oxidation of sulfide into elemental sulfur. Noteworthy, biological processes developed specifically for wastewater treatment will play a key role in the treatment of gas streams from the petroleum industry. Additionally, it is expected that the combination of the biological carbon, nitrogen and sulfur cycles under anaerobic conditions would be a potential technology for the removal of such contaminants in a single step. In conclusion, the application of biological wastewater treatment in the frame of a process integration treatment technology will hopefully close the water cycle allowing the "zero discharge" in the petroleum industry as shown in Fig. 4.

Fig. 4. Schematic representation of the close water cycle in the petroleum industry.

534

REFERENCES [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] II1] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

M.R. Beychock, Aqueous Wastes from Petroleum and Petrochemical Plants, John Wiley & Sons, London (1967). F. Berne and J. Cordonnier, Industrial Water Treatment: Refining, Petrochemicals and Gas processing Techniques, Gulf Publishing, Houston (1995). H. Macarie, Water Sci. Technol., 42:5-6 (2000) 201. R.E. Speece, Anaerobic biotechnology for Industrial Wastewaters, Archaea Press, Nashville (1996) 25-68. J.H Tay, Y.X He and Y.G Yan, Water Environ. Res. 72 (2000) 189. P. Olguin-Lora, L. Puig-Grajales and E. Razo-Flores, Environ. Technol., 24 (2003) 999. R. Sierra-Alvarez and G. Lettinga, Appl. Microbiol. Biotech., 34 (1991) 544. B. Donlon, E. Razo-Flores, J. Field and G, Appl. Environ. Microbiol., 61 (1995) 3889. E. Razo-Flores, M. Iniestra-Gonzalez, J.A. Field, P. Olguin-Lora and L. PuigGrajales, J. Environ. Eng., 129 (2003) 999. H. Fang and G. Zhou, Water Sci. Technol., 42:5 (2000) 237. P. Hwang and S. Cheng, Water Sci. Technol., 24:5 (1991) 133. H. Fang, T. Chen, Y. Li and H. Chui, Water Res., 30 (1996) 1353. K. Tawfiki, K. Lepine, J. Bisaillon, R. Beaudet, J. Hawari and S. R Guiot, Biotechnol. Bioeng., 67 (2000) 419. S.R. Guiot, K. Tawfiki-Hajji and F. Lepine, Water Sci. Technol., 42:5-6 (2000) 245. G. Zhou and H. Fang, Bioresource Technol., 61 (1997) 47. R. Kleerebezem, J. Mortier, L.W. Hulshoff-Pol and G. Lettinga, Water Sci. Technol., 36:2-3 (1997) 237. R. Kleerebezem and G. Lettinga, Water Sci. Technol., 42:5-6 (2000) 259. H. Macarie, Ph. D. Thesis, Universite de Provence Marseille, France (in Freeh) 1992. D.O. Mountfort, W.J. Krumholz and M.P. Bryant, Int. J. System. Bacteriol, 134 (1984)216. S. Shelley, Chem. Eng., 98 (1991) 90. S.S. Cheng, C.Y. Ho and J.H. Wu. 8th Int. Conf. On Anaerobic Digestion. Sendai, Japan (1997) R. Kleerebezem and H. Macarie, Chem. Eng., April, (2003) 56. K. Sublette, R.M. Kolhatkar and K. Raterman, Biodegradation, 9 (1998) 259. C. Buisman, R. Post, P. Ijspeert, G. Geraats and G. Lettinga, Acta Biotechnol., 9 (1989)255. A. Janssen, R. Sleylter, C. van der Kaa, J. Jochemsen, J. Bontsema, S. Ma and G. Lettinga, Biotechnol. Bioeng., 47 (1995) 327. J. Visser, L. Robertson, H. Verseveld and J. Kuenen, Appl. Environ. Microbiol., 63 (1997)2300. A. Janssen, S. Ma, P. Lens and G. Lettinga, Biotechnol. Bioeng., 53 (1997) 32. S. Alcantara, A. Velasco, A. Munoz, J. Cid, S. Revah and E. Razo-Flores, Environ. Sci. Technol., 38 (2004) 918. S. Alcantara, I. Estrada, M. Vasquez and S. Revah, Biotechnol. Lett., 21 (1999) 81.

535

[30] B.J. Arena, H.N. Robson, A.L. de Vegt and C.J. Buisman, National Petroleum Refiners Asociation, Annual Meeting (1988). [31] P. Cadenhead and K. Sublette, Biotechnol. Bioeng., 35 (1990) 1150. [32] B. Rajganesh, M. Selvaraj, K. Sublette and C. Camp, Appl. Biochem. Biotechnol., 51/52(1995)735. [33] K. Sublette and M. Woolsey, Biotechnol. Bioeng., 34 (1989) 565. [34] E. W. Kim and J. H. Bae, Water Sci. Technol, 42:3-4 (2000) 233. [35] B. Krishnakumar and V. B. Manilal, Biotechnol. Lett., 21 (1999) 437. [36] J. Reyes-Avila, E. Razo-Flores and J. Gomez, Water Res., (2004) submitted. [37] F. Fdez-Polanco, M. Fdez-Polanco, N. Fernandez, M. Uruena, P. Garcia and S. Villaverde, Water Sci. Technol., 44: 4 (2001) 15. [38] M.R. Betlach and J.M. Tiedje, Appl. Environ. Microbiol., 42 (1981) 1074. [39] P.J. Gommers, W. Buleveld, F.J. Zuiderwijk and J.G. Kuenen, Water Res. 22 (1988) 1075. [40] R. Johnson, J. Pankow, D. Bender D., C. Price and J. Zogorsky, Environ. Sci. Technol., 32 (2000) 210. [41] J. Brown, S. Bay, D. Greenstein and W. Ray, Report Southern California Coastal Water Research Project. www.sccwrp.org/pubs/annrpt/99_00/abstl I_ar34.htm [42] C.Achten, A. Kolb and W. Puttmann, Environ. Sci.Technol, 36 (2002) 3652. [43] S. Fiorenza and H. Rifai, Bioremediation Journal, 7 (2003) 1. [44] G. Reyna, E. Vega, E. Reyes, V. Mugica, V. Chow, J. Watson and J. Arriaga. 94th Annual Conference and Exhibition of the Air and Waste Management Association. Orlando, FL (2001). [45] L. Manzanares, L. Mufioz, C. Romero, V. Nevarez, E. Ramirez, M. Delgado and A. Keer. 94th Annual Conference and Exhibition of the Air and Waste Management Association. Orlando, FL (2001). [46] R. Iturbe, R. Flores and L.Torres, Water Air Soil Poll., 14 (2003) 261. [47] C. Buenrostro and A. Dovali. DGCH. (2001) Study for determining and evaluating methyl tert-buryl ether (MTBE). Oficio GDF-SOS/01-523. In Spanish. [48] USEPA 1997, EPA822-F-97-009, Office of Water, 34. [49] F. Fayolle, J. Vandecasteele and F. Monot, Appl. Microbiol. Biotechnol., 56 (2001) 339. [50] N. Fortin, M. Morales, Y. Nakagawa, D. Focht and M. Deshusses, Environ. Microbiol., 3(2001)407. [51] J. Salanitro, Curr. Op. Biotechnol., 6 (1995) 337. [52] J. Wilson, J. Cho, B. Wilson and J. Vardy. U.S. Environmental Protection Agency. 2000. Natural attenuation of MTBE in the subsurface under methanogenic conditions. EPA/600/R-00/006. U.S. EPA Office of Research and Development: Washington, D.C. [53] P. Bradley, F. Chapelle and J. Landmeyer, Appl. Environ. Microbiol., 67 (2001) 1975. [54] K. Finneran and D. Lovley, Environ. Sci. Technol., 35 (2001) 1785. [55] R. Steffan, S. Vainberg, C. Condee, K. McClay and P. Hatzinger. In G. Wrickramayake, A. Gavaskar, B. Alleman, V. Magar (eds.) Bioremediation and phytoremediation of chlorinated and recalcitrant compounds. Batelle, Columbus, OH (2000) 165.

536

[56] M. Hyman and K. O'Reilly. In: Alleman B, Lesson A (eds.) In situ bioremediation of petroleum hydrocarbon and other organic compounds. Battelle, Columbus, OH. (1999)7. [57] R. Steffan, K. Me Clay, S. Vainberg, C. Condee and D. Zhang, Appl. Environ. MicrobioL, 63(1997)4216. [58] R. Deeb, K. Scow and L. Alvarez, Biodegradation, 11 (2000) 171. [59] R. Prince, Crit. Rev. MicrobioL, 26 (2002) 163. [60] USEPA 2003. MTBE treatment case study website, http://clu-in.org/products/mtbe/ December 2003. [61] B. Wilson and J. Wilson, Contaminated Soil Sediment and Water (2002) 47. [62] J. Salanitro, P. Johnson, G. Spinnler, P. Maner, H. Wisniewski and C. Bruce, Environ. Sci. TechnoL, 34 (2000) 4152. [63] R. Wilson R., K. Scow and D. Mackay, Environ. Sci. TechnoL, 36 (2001) 190. [64] K. Hrystova, B. Gebreyesus, D. Mackay and K. Scow, Appl.Environ. MicrobioL, 69(2003)2616. [65] R. Steffan, P. Hatzinger, Y. Farhan and S. Drew. NGWA/API Conference on Petroleum Hydrocarbons and Organic Chemicals in groundwater: Prevention, detection and Remediation. Westville, OH (2001). [66] J. Landmeyer, F. Chapelle, H. Herlong and P. Bradley, Environ. Sci. TechnoL, 35 (2001) 1118. [67] R. Steffan, J. Johnson and S. Drew. In Sublette K (eds.) Proceedings of the 7th International Petroleum Environmental Conference. IPEC, Albuquerque, NM (2000) 1-12. [68] J. Morrison, M. Suidan and A. Venosa, J. Environ. Eng., 128 (2002) 836. [69] W. Stringfellow and K. Oh, J. Environ. Eng., 128 (2002) 852. [70] S. Vainberg, A. Togna, P. Sutton and R. Steffan, J. Environ. Eng., 128 (2002) 852. [71] J. O'Connell and S. Zigan. In, E. Moyer, P. Kostecki (eds.) MTBE remediation Handbook. Amherst Scientific Publishers. Amherst Massachusets. (2003) [72] N. Fortin and M. Deshusses, Environ. Sci. TechnoL, 33 (1999) 2980. [73] J. Eweis, J. Scarano, B. Converse, D. Chang and E. Schroeder. Report American Petroleum Institute Contract # 97000-2577(1999). [74] D. Dupasquier, S. Revah and R. Auria, Environ.Sci.TechnoL, 36 (2002) 247. [75] W.W. Eckenfelder Jr. and A.J. Englande Jr., Water Sci. TechnoL, 34:10 (1996) 1. [76] R.F. Dunn and M.M. El-Halwagi, J. Chem. TechnoL BiotechnoL, 78 (2003) 1011. [77] J. Sipma, A.H. Janssen, L.W. Hulshoff Pol and G. Lettinga, BiotechnoL Bioeng., 82(2003)1.

537 STUDIES IN SURFACE SCIENCE AND CATALYSIS Advisory Editors: B. Delmon, Universite Catholique de Louvain, Louvain-la-Neuve, Belgium J.T. Yates, University of Pittsburgh, Pittsburgh, PA, U.S.A.

Volume 1

Volume 2

Volume 3

Volume 4

Volume 5

Volume 6

Volume 7

Volume 8 Volume 9

Volume 1 0

Volume 11

Volume 12

Volume 13 Volume 14

Volume 15

Preparation of Catalysts I. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the First International Symposium, Brussels, October 14-17,1975 edited by B. Delmon, P.A. Jacobs and G. Poncelet The Control of the Reactivity of Solids. A Critical Survey of the Factors that Influence the Reactivity of Solids, with Special Emphasis on the Control of the Chemical Processes in Relation to Practical Applications by V.V. Boldyrev, M. Bulens and B. Delmon Preparation of Catalysts ll.Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Second International Symposium, Louvain-laNeuve, September 4 - 7 , 1978 edited by B. Delmon, P. Grange, P. Jacobs and G. Poncelet Growth and Properties of Metal Clusters. Applications to Catalysis and the Photographic Process. Proceedings of the 32 nd International Meeting of the Societe de Chimie Physique, Villeurbanne, September 2 4 - 2 8 , 1979 edited by J. Bourdon Catalysis by Zeolites.Proceedings of an International Symposium, Ecully (Lyon), September 9 - 1 1 , 1980 edited by B. Imelik, C. Naccache,Y. BenTaarit, J.C.Vedrine, G. Coudurier and H. Praliaud Catalyst Deactivation. Proceedings of an International Symposium, Antwerp, October 1 3 - 15,1980 edited by B. Delmon and G.F. Froment New Horizons in Catalysis. Proceedings of the 7 th International Congress on Catalysis, Tokyo, June 30-July 4, 1980. Parts A and B edited by T. Seiyama and K.Tanabe Catalysis by Supported Complexes by Yu.l.Yermakov, B.N. Kuznetsov and V.A. Zakharov Physics of Solid Surfaces. Proceedings of a Symposium, Bechyne, September 29-October 3,1980 edited by M. Laznicka Adsorption at the Gas-Solid and Liquid-Solid Interface. Proceedings of an International Symposium, Aix-en-Provence, September 2 1 - 2 3 , 1981 edited by J. Rouquerol and K.S.W. Sing Metal-Support and Metal-Additive Effects in Catalysis. Proceedings of an International Symposium, Ecully (Lyon), September 1 4 - 1 6 , 1982 edited by B. Imelik, C. Naccache, G. Coudurier, H. Praliaud, P. Meriaudeau, P. Gallezot, G.A.Martin and J.C.Vedrine Metal Microstructures in Zeolites. Preparation - Properties - Applications. Proceedings of a Workshop, Bremen, September 2 2 - 2 4 , 1982 edited by P.A. Jacobs, N.I. Jaeger, P. Jirii and G. Schulz-Ekloff Adsorption on Metal Surfaces. An Integrated Approach edited by J. Benard Vibrations at Surfaces. Proceedings of the Third International Conference, Asilomar, CA, September 1-4, 1982 edited by C.R. Brundle and H.Morawitz Heterogeneous Catalytic Reactions Involving Molecular Oxygen by G.I. Golodets

538

Volume 1 6

Volume 17

Volume 1 8

Volume 19

Volume 20

Volume 21

Volume 22 Volume 23 Volume 24

Volume 25

Volume 26

Volume 27 Volume 28

Volume 29 Volume 30

Volume 31


Volume 35

Preparation of Catalysts III. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Third International Symposium, Louvain-la-Neuve, September 6 - 9 , 1982 edited by G. Poncelet, P. Grange and P.A. Jacobs Spillover of Adsorbed Species. Proceedings of an International Symposium, Lyon-Villeurbanne, September 12-16, 1983 edited by G.M. Pajonk,S.J.Teichner and J.E. Germain Structure and Reactivity of Modified Zeolites. Proceedings of an International Conference, Prague, July 9 - 1 3 , 1984 edited by P.A. Jacobs, N.I. Jaeger, P. Jiru, V.B. Kazansky and G. Schulz-Ekloff Catalysis on the Energy Scene. Proceedings of the 9 th Canadian Symposium on Catalysis, Quebec, P.Q., September 30-October 3, 1984 edited by S. Kaliaguine and A.Mahay Catalysis by Acids and Bases. Proceedings of an International Symposium, Villeurbanne (Lyonl, September 2 5 - 2 7 , 1984 edited by B. Imelik, C. Naccache, G. Coudurier.Y. Ben Taarit and J.C.Vedrine Adsorption and Catalysis on Oxide Surfaces. Proceedings of a Symposium, Uxbridge, June 2 8 - 2 9 , 1984 edited by M. Che and G.C.Bond Unsteady Processes in Catalytic Reactors by Yu.Sh. Matros Physics of Solid Surfaces I984 edited by J.Koukal Zeolites:Synthesis, Structurejechnology and Application. Proceedings of an International Symposium, Portoroz-Portorose, September 3 - 8 , 1984 edited by B.Drzaj.S. Hocevar and S. Pejovnik Catalytic Polymerization of Olefins. Proceedings of the International Symposium on Future Aspects of Olefin Polymerization, Tokyo, July 4 - 6 , 1985 edited by T.Keii and K.Soga Vibrations at Surfaces 1985. Proceedings of the Fourth International Conference, Bowness-on-Windermere, September 15-19, 1985 edited by D.A. King, N.V. Richardson and S. Holloway Catalytic Hydrogenation edited by L. Cerveny New Developments in Zeolite Science and Technology. Proceedings of t h e 7th International Zeolite Conference, Tokyo, August 17-22, 1986 edited by Y.Murakami,A. lijima and J.W.Ward Metal Clusters in Catalysis edited by B.C. Gates, L. Guczi and H. Knozinger Catalysis and Automotive Pollution Control. Proceedings of t h e First International Symposium, Brussels, September 8 - 1 1 , 1986 edited by A. Crucq and A. Frennet Preparation of Catalysts IV. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Fourth International Symposium, Louvain-laNeuve, September 1-4, 1986 edited by B. Delmon, P. Grange, P.A. Jacobs and G. Poncelet Thin Metal Films and Gas Chemisorption edited by P.Wissmann Synthesis of High-silica Aluminosilicate Zeolites edited by P.A. Jacobs and J.A.Martens Catalyst Deactivation 1987. Proceedings of the 4 t h International Symposium, Antwerp, September 29-October 1, 1987 edited by B. Delmon and G.F. Froment Keynotes in Energy-Related Catalysis edited by S. Kaliaguine

539

Volume 36

Volume 37

Volume 38

Volume 39

Volume 40

Volume 41


Volume 45 Volume 46

Volume 47 Volume 48

Volume 49

Volume 50

Volume 51 Volume 52

Volume 53

Volume 54

Methane Conversion. Proceedings of a Symposium on the Production of Fuels and Chemicals from Natural Gas, Auckland, April 2 7 - 3 0 , 1987 edited by D.M. Bibby, C.D.Chang, R.F.Howe and S.Yurchak Innovation in Zeolite Materials Science. Proceedings of an International Symposium, Nieuwpoort, September 1 3 - 1 7 , 1987 edited by P.J. Grobet, W.J.Mortier, E.F.Vansant and G. Schulz-Ekloff Catalysis l987.Proceedings of the 10 th North American Meeting of the Catalysis Society, San Diego, CA, May 17-22, 1987 edited by J.W.Ward Characterization of Porous Solids. Proceedings of the IUPAC Symposium (COPS I), Bad Soden a. Ts., April 26-29,1987 edited by K.K.Unger, j . Rouquerol, K.S.W.Sing and H. Krai Physics of Solid Surfaces I987. Proceedings of the Fourth Symposium on Surface Physics, Bechyne Castle, September 7 - 1 1 , 1987 edited by J.Koukal Heterogeneous Catalysis and Fine Chemicals. Proceedings of an International Symposium, Poitiers, March 1 5 - 1 7 , 1988 edited by M. Guisnet, |. Barrault, C. Bouchoule.D. Duprez, C. Montassier and G. Perot Laboratory Studies of Heterogeneous Catalytic Processes by E.G. Christoffel, revised and edited by Z. Paal Catalytic Processes under Unsteady-State Conditions by Yu. Sh. Matros Successful Design of Catalysts. Future Requirements and Development. Proceedings of the Worldwide Catalysis Seminars, July, 1 988, on the Occasion of the 30 th Anniversary of the Catalysis Society of Japan edited by T. Inui Transition Metal Oxides. Surface Chemistry and Catalysis by H.H.Kung Zeolites as Catalysts, Sorbents and Detergent Builders. Applications and Innovations. Proceedings of an International Symposium, Wurzburg, September 4 - 8 , 1 9 8 8 edited by H.G. Karge and j.Weitkamp Photochemistry on Solid Surfaces edited by M.Anpo and T. Matsuura Structure and Reactivity of Surfaces. Proceedings of a European Conference, Trieste, September 13-16, 1988 edited by C.Morterra.A. Zecchina and G. Costa Zeolites: Facts, Figures, Future. Proceedings of the 8th International Zeolite Conference, Amsterdam, July 10-14, 1989. Parts A and B edited by P.A. Jacobs and R.A. van Santen Hydrotreating Catalysts. Preparation, Characterization and Performance. Proceedings of the Annual International AlChE Meeting, Washington, DC, November 27-December 2, 1988 edited by M.L. Occelli and R.G.Anthony New Solid Acids and Bases.Their Catalytic Properties by KJanabe, M. Misono.Y. Ono and H. Hattori Recent Advances in Zeolite Science. Proceedings of the 1989 Meeting of the British Zeolite Association, Cambridge, April 1 7 - 1 9 , 1989 edited by J. Klinowsky and P.j. Barrie Catalyst in Petroleum Refining 1989. Proceedings of the First International Conference on Catalysts in Petroleum Refining, Kuwait, March 5 - 8 , 1989 edited by D.LTrimm.S.Akashah, M.Absi-Halabi and A. Bishara Future Opportunities in Catalytic and Separation Technology edited by M. Misono,Y.Moro-oka and S. Kimura

540

Volume 55

Volume 56

Volume 5 7A

Volume 57B

Volume 58 Volume 59

Volume 60

Volume 61

Volume 62

Volume 63

Volume 64 Volume 65

Volume 66

Volume 67

Volume 68

Volume 69

Volume 70

New Developments in Selective Oxidation. Proceedings of an International Symposium, Rimini, Italy, September 18-22, 1989 edited by G. Centi and F.Trifiro Olefin Polymerization Catalysts. Proceedings of the International Symposium on Recent Developments in Olefin Polymerization Catalysts, Tokyo, October 2 3 - 2 5 , 1989 edited by T.Keii and K.Soga Spectroscopic Analysis of Heterogeneous Catalysts. Part A: Methods of Surface Analysis edited by J.L.G. Fierro Spectroscopic Analysis of Heterogeneous Catalysts. Part B: Chemisorption of Probe Molecules edited by J.L.G. Fierro Introduction to Zeolite Science and Practice edited by H. van Bekkum, E.M. Flanigen and ].C. Jansen Heterogeneous Catalysis and Fine Chemicals II. Proceedings of t h e 2 n d International Symposium, Poitiers, October 2 - 6 , 1 990 edited by M. Guisnet, J. Barrault, C. Bouchoule.D. Duprez, G. Perot, R.Maurel and C. Montassier Chemistry of Microporous Crystals. Proceedings of the International Symposium on Chemistry of Microporous Crystals, Tokyo, June 2 6 - 2 9 , 1990 edited by T. Inui.S. Namba and T.Tatsumi Natural Gas Conversion. Proceedings of the Symposium on Natural Gas Conversion, Oslo, August 12-17, 1990 edited by A. Holmen, K.-j. Jens and S.Kolboe Characterization of Porous Solids II. Proceedings of t h e IUPAC Symposium (COPS II), Alicante, May 6 - 9 , 1990 edited by F. Rodriguez-Reinoso, j . Rouquerol, K.S.W. Sing and K.K.Unger Preparation of Catalysts V. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Fifth International Symposium, Louvain-la-Neuve, September 3 - 6 , 1990 edited by G. Poncelet, P.A. Jacobs, P. Grange and B. Delmon New Trends in CO Activation edited by L. Guczi Catalysis and Adsorption by Zeolites. Proceedings of ZEOCAT 9 0 , Leipzig, August 20-23, 1990 edited by G. Ohlmann, H. Pfeifer and R. Fricke Dioxygen Activation and Homogeneous Catalytic Oxidation. Proceedings of t h e Fourth International Symposium on Dioxygen Activation and Homogeneous Catalytic Oxidation, Balatonf tired, September 10-14, 1990 edited by L.I. Simandi Structure-Activity and Selectivity Relationships in Heterogeneous Catalysis. Proceedings of the ACS Symposium on Structure-Activity Relationships in Heterogeneous Catalysis, Boston, MA, April 22-27', 1990 edited by R.K. Grasselli and A.W.SIeight Catalyst Deactivation 1991. Proceedings of the Fifth International Symposium, Evanston, IL, June 2 4 - 2 6 , 1991 edited by C.H. Bartholomew and J.B. Butt Zeolite Chemistry and Catalysis. Proceedings of an International Symposium, Prague, Czechoslovakia, September 8 - 1 3 , 1991 edited by P.A. Jacobs, N.I. Jaeger, L.Kubelkova and B.Wichterlova Poisoning and Promotion in Catalysis based on Surface Science Concepts and Experiments by M. Kiskinova

541

Volume 71

Volume 72

Volume 73

Volume 74 Volume 75

Volume 76 Volume 77

Volume 78

Volume 79

Volume 80

Volume 81

Volume 82

Volume 83

Volume 84


Catalysis and Automotive Pollution Control II. Proceedings of t h e 2 n d International Symposium (CAPoC 2), Brussels, Belgium, September 10-13, 1990 edited by A. Crucq New Developments in Selective Oxidation by Heterogeneous Catalysis. Proceedings of the 3rd European Workshop Meeting on New Developments in Selective Oxidation by Heterogeneous Catalysis, Louvain-la-Neuve, Belgium, April 8 - 1 0 , 1991 edited by P. Ruiz and B. Delmon Progress in Catalysis. Proceedings of the 12th Canadian Symposium on Catalysis, Banff, Alberta, Canada, May 2 5 - 2 8 , 1992 edited by K.J. Smith and EX. Sanford Angle-Resolved Photoemission.Theory and Current Applications edited by S.D. Kevan New Frontiers in Catalysis, Parts A-C. Proceedings of the 10 th International Congress on Catalysis, Budapest, Hungary, 19-24 July, 1992 edited by L. Guczi, F. Solymosi and P.Tetenyi Fluid Catalytic Cracking: Science and Technology edited by J.S.Magee and M.M. Mitchell, Jr. New Aspects of Spillover Effect in Catalysis. For Development of Highly Active Catalysts. Proceedings of the Third International Conference on Spillover, Kyoto, Japan, August 17-20, 1993 edited by T. Inui, K. Fujimoto.T.Uchijima and M. Masai Heterogeneous Catalysis and Fine Chemicals III. Proceedings of the 3rd International Symposium, Poitiers, April 5 - 8, 1993 edited by M. Guisnet, J. Barbier, J. Barrault, C. Bouchoule.D. Duprez, G. Perot and C. Montassier Catalysis: An Integrated Approach to Homogeneous, Heterogeneous and Industrial Catalysis edited by J.A. Moulijn, P.W.N.M. van Leeuwen and R.A. van Santen Fundamentals of Adsorption. Proceedings of the Fourth International Conference on Fundamentals of Adsorption, Kyoto, Japan, May 17-22, 1992 edited by M. Suzuki Natural Gas Conversion II. Proceedings of the Third Natural Gas Conversion Symposium, Sydney, July 4 - 9 , 1993 edited by H.E. Curry-Hyde and R.F.Howe New Developments in Selective Oxidation II. Proceedings of t h e Second World Congress and Fourth European Workshop Meeting, Benalmadena, Spain, September 2 0 - 2 4 , 1993 edited by V. Cortes Corberan and S.Vic Bellon Zeolites and Microporous Crystals. Proceedings of the International Symposium on Zeolites and Microporous Crystals, Nagoya, Japan, August 2 2 - 2 5 , 1993 edited byT. Hattori and T.Yashima Zeolites and Related Microporous Materials: State of the Art I994. Proceedings of the 10th International Zeolite Conference, Garmisch-Partenkirchen, Germany, July 17-22, 1994 edited by J.Weitkamp, H.G. Karge.H. Pfeifer and W. Holderich Advanced Zeolite Science and Applications edited by J.C. jansen, M. Stocker, H.G. Karge and J.Weitkamp Oscillating Heterogeneous Catalytic Systems by M.M. Slinko and N.I. Jaeger Characterization of Porous Solids III. Proceedings of t h e IUPAC Symposium (COPS III), Marseille, France, May 9 - 1 2 , 1993 edited by j.Rouquerol, F. Rodriguez-Reinoso, K.S.W. Sing and K.K.Unger

542

Volume 88

Volume 89

Volume 90

Volume 91

Volume 92

Volume 93 Volume 94

Volume 95 Volume 96

Volume 97

Volume 98

Volume 99 Volume 100

Volume 101

Volume 102 Volume 1 03 Volume 104 Volume 105

Catalyst Deactivation 1994. Proceedings of the 6 th International Symposium, Ostend, Belgium, October 3 - 5 , 1994 edited by B. Delmon and G.F. Froment Catalyst Design for Tailor-made Polyolefins. Proceedings of t h e International Symposium on Catalyst Design for Tailor-made Polyolefins, Kanazawa, Japan, March 10-12, 1994 edited by K.Soga and M.Terano Acid-Base Catalysis II. Proceedings of the International Symposium on Acid-Base Catalysis II, Sapporo, Japan, December 2 - 4 , 1993 edited by H. Hattori.M. Misono and Y.Ono Preparation of Catalysts VI. Scientific Bases for the Preparation of Heterogeneous Catalysts. Proceedings of the Sixth International Symposium, Louvain-La-Neuve, September 5-8, 1994 edited by G. Poncelet, J.Martens.B. Delmon, P.A. Jacobs and P. Grange Science and Technology in Catalysis I994. Proceedings of t h e Second Tokyo Conference on Advanced Catalytic Science and Technology, Tokyo, August 2 1 - 2 6 , 1994 edited by Y. Izumi, H.Arai and M. Iwamoto Characterization and Chemical Modification of the Silica Surface by E.F.Vansant, P.Van Der Voort and K.C.Vrancken Catalysis by Microporous Materials. Proceedings of ZEOCAT'95, Szombathely, Hungary, July 9-13, 1995 edited by H.K. Beyer, H.G.Karge, i. Kiricsi and J.B.Nagy Catalysis by Metals and Alloys by V. Ponec and G.C.Bond Catalysis and Automotive Pollution Control III. Proceedings of the Third International Symposium (CAPoC3), Brussels, Belgium, April 2 0 - 2 2 , 1994 edited by A. Frennet and J.-M. Bastin Zeolites:A Refined Tool for Designing Catalytic Sites. Proceedings of the International Symposium, Quebec, Canada, October 15-20, 1995 edited by L. Bonneviot and S. Kaliaguine Zeolite Science 1994: Recent Progress and Discussions. Supplementary Materials t o t h e 10th International Zeolite Conference, Garmisch-Partenkirchen, Germany, July 17-22, 1994 edited by H.G. Karge and J.Weitkamp Adsorption on New and Modified Inorganic Sorbents edited by A.Dabrowski and V.A.Tertykh Catalysts in Petroleum Refining and Petrochemical Industries 1995. Proceedings of the 2nd International Conference on Catalysts in Petroleum Refining and Petrochemical Industries, Kuwait, April 22-26, 1995 edited by M.Absi-Halabi, J. Beshara, H. Qabazard and A. Stanislaus II th International Congress on Catalysis -40 t h Anniversary. Proceedings of the 11 t h ICC, Baltimore, MD, USA, June 30-July 5, 1996 edited by J.W. High tower, W.N. Delgass, E. Iglesia and A.T. Bell Recent Advances and New Horizons in Zeolite Science and Technology edited by H. Chon.S.I.Woo and S. -E. Park Semiconductor Nanoclusters - Physical, Chemical, and Catalytic Aspects edited by P.V. Kamat and D. Meisel Equilibria and Dynamics of Gas Adsorption on Heterogeneous Solid Surfaces edited by W. Rudzihski.W.A. Steele and G. Zgrablich Progress in Zeolite and Microporous Materials Proceedings of the 11 t h International Zeolite Conference, Seoul, Korea, August 12-17, 1996 edited by H. Chon.S.-K. Ihm and Y.S.Uh

543

Volume 1 06

Volume 1 07

Volume 108

Volume 109

Volume 110

Volume 111

Hydrotreatment and Hydrocracking of Oil Fractions Proceedings of the 1 s t International Symposium / 6 th European Workshop, Oostende, Belgium, February 17-19, 1997 e d i t e d b y G.F. Froment.B. Delmon and P. Grange Natural Gas Conversion IV Proceedings of the 4 th International Natural Gas Conversion Symposium, Kruger Park, South Africa, November 19-23, 1995 e d i t e d b y M. de Pontes, R.L. Espinoza, C.P. Nicolaides, J.H. Scholtz and M.S. Scurrell Heterogeneous Catalysis and Fine Chemicals IV Proceedings of the 4 th International Symposium on Heterogeneous Catalysis and Fine Chemicals, Basel, Switzerland, September 8-12, 1996 e d i t e d by H.U. Blaser, A. Baiker and R. Prins Dynamics of Surfaces and Reaction Kinetics in Heterogeneous Catalysis. Proceedings of the International Symposium, Antwerp, Belgium, September 15-17, 1997

edited by G.F. Froment and K.C.Waugh Third World Congress on Oxidation Catalysis. Proceedings of the Third World Congress on Oxidation Catalysis, San Diego, CA, U.S.A., 21-26 September 1997 e d i t e d b y R.K. Grasselli,S.T.Oyama, A.M. Gaffney and J.E. Lyons Catalyst Deactivation I997. Proceedings of the 7th International Symposium, Cancun, Mexico, October 5-8,

1997 Volume 11 2

Volume 113

Volume 114

Volume 11 5

Volume 116

Volume 11 7

Volume 118

Volume 11 9

Volume 12 0 A

e d i t e d b y C.H. Bartholomew and G.A. Fuentes Spillover and Migration of Surface Species on Catalysts. Proceedings of the 4 th International Conference on Spillover, Dalian, China, September 15-18, 1997 e d i t e d b y Can Li and Qin Xin Recent Advances in Basic and Applied Aspects of Industrial Catalysis. Proceedings of the 13th National Symposium and Silver Jubilee Symposium of Catalysis of India, Dehradun, India, April 2-4, 1997 e d i t e d b y T.S.R. Prasada Rao and G.Murali Dhar Advances in Chemical Conversions for Mitigating Carbon Dioxide. Proceedings of the 4 th International Conference on Carbon Dioxide Utilization, Kyoto, Japan, September 7-11, 1997 e d i t e d b y T. Inui, M.Anpo.K. liui.S.Yanagida and T.Yamaguchi Methods for Monitoring and Diagnosing the Efficiency of Catalytic Converters. A patent-oriented survey by M. Sideris Catalysis and Automotive Pollution Control IV. Proceedings of the 4 th International Symposium (CAPoC4), Brussels, Belgium, April 9-11, 1997 e d i t e d b y N. Kruse, A. Frennet and J.-M. Bastin Mesoporous Molecular Sieves 1998 Proceedings of the 1 s t International Symposium, Baltimore, MD, U.S.A., July 10-12, 1998 e d i t e d b y L.Bonneviot, F. Beland, C.Danumah, S. Giasson and S. Kaliaguine Preparation of Catalysts VII Proceedings of the 7 th International Symposium on Scientific Bases for the Preparation of Heterogeneous Catalysts, Louvain-la-Neuve, Belgium, September 1-4, 1998 e d i t e d b y B. Delmon, P.A. Jacobs, R. Maggi, J.A.Martens, P. Grange and G. Poncelet Natural Gas Conversion V Proceedings of the 5th International Gas Conversion Symposium, Giardini-Naxos, Taormina, Italy, September 20-25, 1998 e d i t e d b y A. Parmaliana, D. Sanfilippo, F. Frusteri, A.Vaccari and F.Arena Adsorption and its Applications in Industry and Environmental Protection. Vol I: Applications in Industry e d i t e d b y A. Dijbrowski

544

Volume 1 2 0 B

Volume 121

Volume 122

Adsorption and its Applications in Industry and Environmental Protection. Vol II: Applications in Environmental Protection edited b y A. Dqbrowski Science and Technology in Catalysis 1998 Proceedings of the Third Tokyo Conference in Advanced Catalytic Science and Technology, Tokyo, July 19-24, 1998 edited b y H. Hattori and K. Otsuka Reaction Kinetics and the Development of Catalytic Processes Proceedings of the International Symposium, Brugge, Belgium, April 19-21,

1999 Volume 1 23

Volume 1 24 Volume 125

Volume 1 26

Volume 127

Volume 1 28

Volume 129

Volume 1 30

Volume 131

Volume 1 32

Volume 1 33

Volume 134

edited b y G.F. Froment and K.C.Waugh Catalysis:An Integrated Approach Second, Revised and Enlarged Edition edited b y R.A. van Santen, P.W.N.M. van Leeuwen, J.A. Moulijn and BAAveriil Experiments in Catalytic Reaction Engineering by J.M. Berty Porous Materials in Environmentally Friendly Processes Proceedings of the 1 s t International FEZA Conference, Eger, Hungary, September 1-4, 1999 edited b y I. Kiricsi, G. Pal-Borbely, J.B.Nagy and H.G. Karge Catalyst Deactivation 1999 Proceedings of the 8th International Symposium, Brugge, Belgium, October 10-13, 1999 edited b y B. Delmon and G.F. Froment Hydrotreatment and Hydrocracking of Oil Fractions Proceedings of the 2nd International Symposium/7th European Workshop, Antwerpen, Belgium, November 14-17, 1999 edited b y B. Delmon, G.F. Froment and P. Grange Characterisation of Porous Solids V Proceedings of the 5th International Symposium on the Characterisation of Porous Solids (COPS-V), Heidelberg, Germany, May 30- June 2, 1999 edited b y K.K.Unger,G.Kreysa and J.P. Baselt Nanoporous Materials II Proceedings of the 2nd Conference on Access in Nanoporous Materials, Banff, Alberta, Canada, May 25-30, 2000 edited byA. Sayari.M. jaroniec and T.J. Pinnavaia 12 th International Congress on Catalysis Proceedings of the 12 t h ICC, Granada, Spain, July 9-14, 2000 edited byA. Corma, F.V. Melo.S. Mendioroz and J.L.G. Fierro Catalytic Polymerization of Cycloolefins Ionic, Ziegler-Natta and Ring-Opening Metathesis Polymerization By V. Dragutan and R. Streck Proceedings of the International Conference on Colloid and Surface Science, Tokyo, Japan, November 5-8,2000 25 th Anniversary of the Division of Colloid and Surface Chemistry, The Chemical Society of Japan edited b y Y. Iwasawa, N.Oyama and H.Kunieda Reaction Kinetics and the Development and Operation of Catalytic Processes Proceedings of the 3rd International Symposium, Oostende, Belgium, April 2225, 2001 edited b y G.F. Froment and K.C.Waugh Fluid Catalytic Cracking V Materials and Technological Innovations edited b y M.L. Occelli and P. O'Connor

545

Volume 135

Volume 136

Volume 137

Volume 138 Volume 139

Volume 140

Volume 141

Volume 142

Volume 143

Volume 144

Volume 145

Volume 146

Volume 147

Volume 148

Volume 149

Volume 150

Zeolites and Mesoporous Materials at the Dawn of the 21 st Century. Proceedings of the 13lh International Zeolite Conference, Montpellier, France, 8-13 July 2001 edited by A. Galameau, F. di Renso, F. Fajula ans J. Vedrine Natural Gas Conversion VI Proceedings of the 6th Natural Gas Conversion Symposium, June 17-22, 2001, Alaska, USA. edited by J J . Spivey, E. Iglesia and T.H. Fleisch Introduction to Zeolite Science and Practice. 2nd completely revised and expanded edition edited by H. van Bekkum, E.M. Flanigen, P.A. Jacobs and J.C. Jansen Spillover and Mobility of Species on Solid Surfaces edited by A. Guerrero-Ruiz and I. Rodriquez-Ramos Catalyst Deactivation 2001 Proceedings of the 9th International Symposium, Lexington, KY, USA, October 2001 edited by J.J. Spivey, G.W. Roberts and B.H. Davis Oxide-based Systems at the Crossroads of Chemistry. Second International Workshop, October 8-11, 2000, Como, Italy. Edited by A. Gamba, C. Colella and S. Coluccia Nanoporous Materials III Proceedings of the 3rd International Symposium on Nanoporous Materials, Ottawa, Ontario, Canada, June 12-15, 2002 edited by A. Sayari and M. Jaroniec Impact of Zeolites and Other Porous Materials on the New Technologies at the Beginning of the New Millennium Proceedings of the 2nd International FEZA (Federation of the European Zeolite Associations) Conference, Taormina, Italy, September 1-5, 2002 edited by R. Aiello, G. Giordano and F.Testa Scientific Bases for the Preparation of Heterogeneous Catalysts Proceedings of the 8th International Symposium, Louvain-la-Neuve, Leuven, Belgium, September 9-12, 2002 edited by E. Gaigneaux, D.E. De Vos, P. Grange, P.A. Jacobs, J.A. Martens, P. Ruiz and G. Poncelet Characterization of Porous Solids VI Proceedings of the 6lh International Symposium on the Characterization of Porous Solids (COPS-VI), Alicante, Spain, May 8-11, 2002 edited by F. Rodriguez-Reinoso, B. McEnaney, J. Rouquerol and K. Unger Science and Technology in Catalysis 2002 Proceedings of the Fourth Tokyo Conference on Advanced Catalytic Science and Technology, Tokyo, July 14-19, 2002 edited by M. Anpo, M. Onaka and H. Yamashita Nanotechnology in Mesostructured Materials Proceedings of the 3rd International Mesostructured Materials Symposium, Jeju, Korea, July 8-11, 2002 edited by Sang-Eon Park, Ryong Ryoo, Wha-Seung Ahn, Chul Wee Lee and Jong-San Chang Natural Gas Conversion VII Proceedings of the 7lh Natural Gas Conversion Symposium, Dalian, China, June 6-10, 2004 edited by X. Bao and Y. Xu Mesoporous Crystals and Related Nano-Structured Materials Proceedings of the Meeting on Mesoporous Crystals and Related Nano-Structured Materials, Stockholm, Sweden, 1-5 June, 2004 edited by O. Terasaki Fluid Catalytic Cracking VI: Preparation and Characterization of Catalysts Proceedings of the 6th International Symposium on Advances in Fluid Cracking Catalysts (FCCs), New York, September 7 - 1 1 , 2003 Edited by M. Occelli Coal and Coal-Related Compounds Structures, Reactivity and Catalytic Reactions edited by T. Kabe, A. Ishihara, W. Qian, I.P. Sturisna and Y. Kabe


Petroleum Biotechnology - Developments and Perspectives

Petroleum Geology (Developments in Petroleum Science)

Thermal Methods of Petroleum Production (Developments in Petroleum Science)

Abnormal Formation Pressures (Developments in Petroleum Science)

Paraffin Products (Developments in Petroleum Science)

Polymer Flooding (Developments in Petroleum Science)

Integrated Flow Modeling (Developments in Petroleum Science)

Cavernous Sinus: Developments and Future Perspectives

Homological Mirror Symmetry: New Developments and Perspectives

Developments in Psychotherapy: Historical Perspectives

Casing Design - Theory and Practice (Developments in Petroleum Science)

Casing Design - Theory and Practice (Developments in Petroleum Science)

PVT and Phase Behaviour of Petroleum Reservoir Fluids (Developments in Petroleum Science)

Handbook on Cyanobacteria: Biochemistry, Biotechnology and Applications (Bacteriology Research Developments)

Well Completion Design, Volume 56 (Developments in Petroleum Science)

Fundamentals of Reservoir Engineering (Developments in Petroleum Science)

Tracers in the Oil Field (Developments in Petroleum Science)

Fundamentals of Numerical Reservoir Simulation (Developments in Petroleum Science)

Microbial Enhanced Oil Recovery (Developments in Petroleum Science, No. 22)

Petroleum Economics and Engineering

Petroleum Formation and Occurrence

Biotechnology

Petroleum Geoscience

Petroleum Geoscience

Biotechnology

Biotechnology

Biotechnology

Petroleum Directory

Origin and Refining of Petroleum

Biotechnology

Petroleum Biotechnology - Developments and Perspectives

Petroleum Geology (Developments in Petroleum Science)

Thermal Methods of Petroleum Production (Developments in Petroleum Science)

Abnormal Formation Pressures (Developments in Petroleum Science)

Paraffin Products (Developments in Petroleum Science)

Polymer Flooding (Developments in Petroleum Science)

Integrated Flow Modeling (Developments in Petroleum Science)

Cavernous Sinus: Developments and Future Perspectives

Homological Mirror Symmetry: New Developments and Perspectives

Developments in Psychotherapy: Historical Perspectives

Casing Design - Theory and Practice (Developments in Petroleum Science)

Casing Design - Theory and Practice (Developments in Petroleum Science)

PVT and Phase Behaviour of Petroleum Reservoir Fluids (Developments in Petroleum Science)

Handbook on Cyanobacteria: Biochemistry, Biotechnology and Applications (Bacteriology Research Developments)

Well Completion Design, Volume 56 (Developments in Petroleum Science)

Fundamentals of Reservoir Engineering (Developments in Petroleum Science)

Tracers in the Oil Field (Developments in Petroleum Science)

Fundamentals of Numerical Reservoir Simulation (Developments in Petroleum Science)

Microbial Enhanced Oil Recovery (Developments in Petroleum Science, No. 22)

Petroleum Economics and Engineering

Petroleum Formation and Occurrence

Biotechnology

Petroleum Geoscience

Petroleum Geoscience

Biotechnology

Biotechnology

Biotechnology

Petroleum Directory

Origin and Refining of Petroleum

Biotechnology

Recommend Documents