Marine Bioprocess Engineering

Marine Bioprocess Engineering Vol. 14 (1978) Vol. 15 (1979) Vol. 16 (1982) Vol. 17 (1983) Vol. 18 (1983) Vol. 19 (198...

Author: J.G. Burgess | R.H. Wijffels

181 downloads 2347 Views 26MB 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

Marine Bioprocess Engineering

Vol. 14 (1978) Vol. 15 (1979) Vol. 16 (1982) Vol. 17 (1983) Vol. 18 (1983) Vol. 19 (1984) Vol. 20 (1984) Vol. 21 (1989) Vol. 22 (1986) Vol. 23 (1986) Vol. 24 (1986) Vol. 25 (1988) Vol. 26 (1989) Vol. 27 (1989) Vol. 28 (1993) Vol. 29 (1994) Vol. 30 (1994) Vol. 31 (1995) Vol. 32 (1995) Vol. 33 (1995) Vol. 34 (1995)

edited by M.J. Bull (1st reprint 1983) edited by M.J. Bull edited by M.J. Bull edited by M.E. Bushell Microbial Polysaccharides, edited by M.E. Bushell Modem Applications of Traditional Biotechnologies, edited by M.E. Bushell Innovations in Biotechnology, edited by E.H. Houwink and R.R. van der Meer Statistical Aspects of the Microbiological Analysis of Foods, by B. Jarvis Moulds and Filamentous Fungi in Technical Microbiology, by O. Fassatiovfi Micro-organisms in the Production of Food, edited by M.R. Adams Biotechnology of Animo Acid Production, edited by K. Aida. I. Chibata, K. Nakayama, K. Takinama and H. Yamada Computers in Fermentation Technology, edited by M.E. Bushell Rapid Methods in Food Microbiology, edited by M.R. Adams and C.F.A. Hope Bioactive Metabolites from Microorganisms, edited by M.E. Bushell and U. Gr~ife Micromycetes in Foodstuffs and Feedstuffs, edited by Z. Jesenskfi .4spergillus: 50 years on, edited by S.D. Martinelli and J.R. Kinghorn Bioactive Secondary Metabolites of Microorganisms, edited V. Betina Techniques in Applied Microbiology, edited by B. Sikyta Biotransformations" Microbial Degradation of Health Risk Compounds, edited by V.P. Singh Microbial Pentose Utilization. Current Applications in Biotechnology, by A. Singh and P. Mishra Culture Media for Food Microbiology, edited by J.E.L. Corry, G.D.W. Curtis and R.M. Baird (second impression 1999) . .

Marine Bioprocess I: ng=neerlng

Proceedings of an International Symposium organized under auspices of The Working Party on Applied Biocatalysis of the Eurpean Federation of Biotechnology and The European Society for Marine Biotechnology, Noordwijkerhout, The Netherlands, November 8-11, 1998 "~. .

Edited by R. Osinga

Food and Bioprocess Engineering Group, WageningenAgricultural University, Wageningen, The Netherlands

J. Tramper

Food and Bioprocess Engineering group, WageningenAgricultural University, Wageningen, The Netherlands

J. G. Burgess

Department of Biological Sciences, Heriott-Watt University, Edinburgh, United Kingdom

R.H. Wijffels

Food and Bioprocess Engineering group, WageningenAgricultural University, Wageningen, The Netherlands

progress in industrial microbiology

1999

ELSEVIER A m s t e r d a m - Lausanne - New Y o r k - O x f o r d - S h a n n o n - Singapore - Tokyo

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

9 1999 Elsevier Science B.V. All rights reserved.

This work is protected under copyright by Elsevier Science, 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 Science Rights & Permissions Department, PO Box 800, Oxford OX5 I DX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also contact Rights & Permissions directly through Elsevier's home page (http://www.elsevier.nl), selecting first 'Customer Support', then 'General Information', then 'Permissions Query Form'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (978) 7508400, fax: (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) 171 631 5555; fax: (+44) 171 631 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 Elsevier Science 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 Science Rights & Permissions Department, at the mail, 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.

Reprinted from Journal of Biotechnology, Vol. 70/1999

Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for.

ISBN: 0-444-50387-0

Q

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

Printed in The Netherlands.

Publication of the proceedings was possible thanks to a grant from the Directorate General Science, Research and Development of the European Commission (BIO4-CT98-4814)

This Page Intentionally Left Blank

Contents Cited in: Biological Abstracts; Biological & Agricultural Index; Cambridge Scientific Abstracts; Chemical Abstracts; Elsevier BlOBASE/Current Awareness in Biological Sciences; Current Contents (Agriculture, Biology & Environmental Sciences); EMBASE/Excerpta Medica; Index Medicus

Special Issue Marine Bioprocess Engineering

Editorial

1

Opening keynote

S.A. Pomponi (USA) The bioprocess-technological potential of the sea

Products

M.H.G. Munro, J.W. Blunt, E.J. Dumdei, S.J.H. Hickford, R.E. Lill, S. Li, C.N. Battershill and A.R. Duckworth (New Zealand) The discovery and development of marine compounds with pharmaceutical potential

15

J.G. Burgess, E.M. Jordan, M. Bregu, A. Mearns-Spragg and K.G. Boyd (UK) Microbial antagonism: a neglected avenue of natural products research

27

T. Matsunaga, H. Takeyama, T. Nakao and A. Yamazawa (Japan) Screening of marine microalgae for bioremediation of cadmium-polluted seawater

33

E.G. Vrieling, T.P.M. Beelen, R.A. van Santen and W.W.C. Gieskes (The Netherlands) Diatom silicon biomineralization as an inspirational source of new approaches to silica production

39

M. Turkiewicz, E. Gromek, H. Kalinowska and M. Zieliriska (Poland) Biosynthesis and properties of an extracellular metalloprotease from the Antarctic marine bacterium Sphingomonas paucimobilis

53

S. Kreitlow, S. Mundt and U. Lindequist (Germany) Cyanobacteria--a potential source of new biologically active substances

61

F. Sponga, L. Cavaletti, A. Lazzarini, A. Borghi, I. Ciciliato, D. Losi and F. Marinelfi (Italy) Biodiversity and potentials of marine-derived microorganisms

65

Z. CsSg6r, D. Melgar, K. Schmidt and C. Posten (Germany) Production and particle characterization of the frustules of Cyclotella cryptica in comparison with siliceous earth

71

G. Liebezeit, T.D. K(Jnnemann and G. Gad (Germany) Biotechnological potential of North Sea salt marsh plants--a review of traditional knowledge

77

A. Koulman, L.M.C. Pruijn, T.S.A. Sandstra, H.J. Woerdenbag and N. Pras (The Netherlands) The pharmaceutical exploration of cold water ascidians from the Netherlands: a possible source of new cytotoxic natural products

85

Energy J. Miyake, M. Miyake and Y. Asada (Japan) Biotechnological hydrogen production: research for efficient light energy conversion

89

i. Eroglu, K. Asian, U. GEmdEiz, M. YScel and L. TEIrker (Turkey) Substrate consumption rates for hydrogen production by Rhodobacter sphaeroides in a column photobioreactor

103

V. Sediroglu, i. Eroglu, M. YEicel, L. TElrker and U. GdndEIz (Turkey) The biocatalytic effect of Halobacterium halobium on photoelectrochemical hydrogen production

115

D.O. Yi~it, U. GSnd~z, L. T~rker, M. YScel and ]. Ero~lu (Turkey) Identification of by-products in hydrogen producing bacteria; Rhodobacter sphaeroides O.U. 001 grown in the waste water of a sugar refinery

125

Cultivation of marine organisms B. Rinkevich (Israel) Cell cultures from marine invertebrates: obstacles, new approaches and recent improvements

133

R. Osinga, P.B. de Beukelaer, E.M. Meijer, J. Tramper and R.H. Wijffels (The Netherlands) Growth of the sponge Pseudosuberites (aff.) andrewsi in a closed system

155

E. Armstrong, J.D. McKenzie and G.T. Goldsworthy (UK) Aquaculture of sponges on scallops for natural products research and antifouling

163

B.J.B. Wood, P.H.K. Grimson, J.B. German and M. Turner (UK, USA) Photoheterotrophy in the production of phytoplankton organisms

175

M.E. de Swaaf, T.C. de Rijk, G. Eggink and L. Sijtsma (The Netherlands) Optimisation of docosahexaenoic acid production in batch cultivations by Crypthecodinium cohnii

185

R.D. Bowles, A.E. Hunt, G.B. Bremer, M.G. Duchars and R.A. Eaton (UK) Long-chain n - 3 polyunsaturated fatty acid production by members of the marine protistan group the thraustochytrids: screening of isolates and optimisation of docosahexaenoic acid production

193

H. Helmholz, P. Etoundi and U. Lindequist (Germany) Cultivation of the marine basidiomycete Nia vibrissa (Moore & Meyers)

203

S. La Barre, S. Singer, E. Erard-Le Denn and M. Jozefowicz (France) Controlled cultivation of Alexandrium minutum and [33p] orthophosphate cell labeling towards surface adhesion tests

207

J.C. Ogbonna, S. Tomiyama and H. Tanaka (Japan) Production of c~-tocopherol by sequential heterotrophic-photoautotrophic cultivation of Euglena gracilis

213

K. Tsukahara, S. Sawayama, T. Yagishita and T. Ogi (Japan) Effect of Ca2 + channel blockers on glycerol levels in Dunaliella tertiolecta under hypoosmotic stress

223

P. Garcia-Jimenez, F.D. Marian, M. Rodrigo and R.R. Robaina (Spain) Sporulation and sterilization method for axenic culture of Gelidium canariensis

227

Design and scale-up of in vitro cultures E. Molina Grima, F.G.A. Fernandez, F. Garcia Camacho and Y. Chisti (Spain) Photobioreactors: light regime, mass transfer, and scaleup

231

A. S~nchez Mir6n, A. Contreras GOmez, F. Garcia Camacho, E. Molina Grima and Y. Chisti (Spain) Comparative evaluation of compact photobioreactors for large-scale monoculture of microalgae

249

M.M. Rebolloso Fuentes, J.L. Garcia Sanchez, J.M. Fernandez Sevilla, F.G. Aci~n Fernandez, J.A. S~nchez P~rez and E. Molina Grima (Spain) Outdoor continuous culture of Porphyridium cruentum in a tubular photobioreactor: quantitative analysis of the daily cyclic variation of culture parameters

271

J.C. Ogbonna, T. Soejima and H. Tanaka (Japan) An integrated solar and artificial light system for internal illumination of photobioreactors

289

G. Chini Zittelli, F. Lavista, A. Bastianini, L. Rodolfi, M. Vincenzini and M.R. Tredici (Italy) Production of eicosapentaenoic acid by Nannochloropsis sp. cultures in outdoor tubular photobioreactors

299

M.A. Borowitzka (Australia) Commercial production of microaigae: ponds, tanks, tubes and fermenters

313

M. Janssen, T.C. Kuijpers, B. Veldhoen, M.B. Ternbach, J. Tramper, L.R. Mur and R.H. Wijffels (Netherlands) Specific growth rate of Chlamydomonas reinhardtii and Chlorella sorokiniana under medium duration light/dark cycles: 13-87 s

323

D. Baquerisse, S. Nouals, A. Isambert, P.F. dos Santos and G. Durand (France) Modelling of a continuous pilot photobioreactor for microalgae production

335

P.C. Wright, C. Stevenson, E. McEvoy and J.G. Burgess (UK) Opportunities for marine bioprocess intensification using novel bioreactor design: frequency of barotolerance in microorganisms obtained from surface waters

343

N. Zou and A. Richmond (Israel) Effect of light-path length in outdoor flat plate reactors on output rate of cell mass and of EPA in Nannochloropsis sp.

351

K. Kotzabasis, A. Hatziathanasiou, M.V. Bengoa-Ruigomez, M. Kentouri and P. Divanach (Greece, Spain) Methanol as alternative carbon source for quicker efficient production of the microalgae Chlorella minutissima: Role of the concentration and frequence of administration

357

Product recovery L.A.M. van der Wielen and L.K. Cabatingan (The Netherlands, Philippines) Fishing products from the sea--rational downstream processing of marine bioproducts

363

S. Hirano, T. Nakahira, M. Nakagawa and S.K. Kim (Japan, South Korea) The preparation and applications of functional fibres from crab shell chitin

373

A. Robles Medina, L. Esteban Cerd~n, A. Gim6nez Gim~nez, B. Camacho Paez, M.J. Iba~ez Gonzalez and E. Molina Grima (Spain) Lipase-catalyzed esterification of glycerol and polyunsaturated fatty acids from fish and microalgae oils

379

L. Vandanjon, P. Jaouen, N. Rossignol, F. Qu~m~neur and J.-M. Robert (France) Concentration and desalting by membrane processes of a natural pigment produced by the marine diatom Haslea ostrearia Simonsen

393

Closing keynote O.R. Zaborsky (USA) Marine bioprocess engineering: the missing link to commercialization

403

Author Index Subject Index

409 413


JOUItNAL

OF

Biotecbnology ELSEVIER

Journal of Biotechnology 70 (1999) 1-3

Editorial

This special issue contains full papers of both oral and poster presentations of the international symposium 'Marine Bioprocess Engineering' which was held in Noordwijkerhout, The Netherlands, 8-11 November 1998. The symposium focused on the bioprocessing of marine natural products. Biotechnology is the application-oriented integration of biodisciplines such as cell biology, biochemistry, microbiology, molecular biology and molecular genetics, with engineering. Applications of biotechnology are found in health care, environmental remediation and the production of energy, food and fine chemicals. Bioprocess engineering has been the key to success in the commercialization of biotechnology, especially with respect to biopharmaceuticals. In marine biotechnology, both new and existing biotechnological techniques are developed and applied to organisms from marine sources. For marine biotechnology, bioprocess engineering represents the link between discovery and commercialization. The world's oceans represent one of the largest untapped biological resources. Appropriately, the United Nations has designated 1998 as the 'Year of the Ocean' in recognition of the importance of oceans to the well-being of this planet and the need for proper use of their resources for sustainable development. The diversity of marine life points to a myriad of new bioproducts waiting to be discovered and developed commercially through engineering research and a systems approach that spans activities from the identification of new marine bioproducts through production,

separation, formulation and delivery. Numerous potential pharmaceuticals (e.g. bryostatin, halichondrin), valuable biopolymers (e.g. chitin), compounds that can be used in the anti-fouling industry (as an alternative to toxic chemicals like organotins), products for the food industry (polyunsaturated fatty acids) and products for the cosmetic industry (e.g. phycobiliproteins) have been found in and isolated from marine organisms. In addition, marine microorganisms can be used to produce energy (biohydrogen). In order to stimulate research that leads to commercial exploitation of these natural resources, marine organisms need to be cultivated in vitro for the rational and sustainable exploitation of the sea as a source of food and natural products. While the importance of marine biotechnology has been recognized, the engineering component has not been fully integrated. At the symposium 'Marine Bioprocess Engineering' we have begun to bridge the gap between the isolation of products from marine organisms in the laboratory and industrial applications by focusing on the bioproLess-engineering aspects. Reviews and recent developments in product discovery, bio-energy production, cultivation of marine organisms, scale up and product recovery were presented and discussed during five sessions (oral and poster presentations):

I. Products

The focus was on the screening of marine or-

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 0 5 1 - 6

2

Editorial

ganisms for useful products, including novel screening strategies. In situ biochemical functions as well as biomedical properties of the target compounds were also discussed.

2. Energy

There is an increasing interest in the biotechnological production of hydrogen as an energy source. Biological aspects (organisms, metabolic processes), technological aspects as well as economic aspects of biological hydrogen production were discussed.

3. Cultivation of marine organisms

This topic included techniques for cultivating bacteria, fungi, algae, thraustochytrids and invertebrates, invertebrate-cell cultures, physiological barriers for in vitro cultivation and methods to stimulate the production of the target compounds by the cultured organisms.

4. Design and scale-up of in vitro cultures

Presentations in this session were focused on the possibilities and drawbacks of marine bioreactor design. This includes different types of bioreactors (for instance: photobioreactors), hydrodynamics and mass transfer.

5. Product recovery

Downstream processing is for many natural products the most expensive process step. The recovery of marine natural products may differ from conventional recovery techniques. Presentations that considered these differences or that illustrated new and emerging recovery techniques were given. The manuscripts presented in this special issue give an extensive and timely overview of the current research and applications in marine bioprocess engineering. We hope the manuscripts will

stimulate application-oriented research in the field of marine biotechnology and that co-operations between biologists and engineers will result in an integrated research-approach. Researchers from biological disciplines should already implement in their research strategy questions that will raise when the processes are scaled up (think big!). Similarly, industry and researchers applying or studying cultivation and product recovery should discuss scale-up aspects in an early stage of the research with researchers from biological disciplines (scale-up by scaling down!). The Guest Editors hope that engineering aspects of marine biotechnology will receive further attention in the future. Exploration of new bioproducts from the ocean should be followed up by a sustainable exploitation of these valuable resources. The organization committee of the international symposium 'Marine Bioprocess Engineering' acknowledges with gratitude the following organizations: Applikon Dependable Instruments, (The Netherlands), for sponsoring the session 'design and scale-up' of in vitro cultures; Shell Nederland by, for sponsoring the session 'energy'; The Foundation for Biotechnology in The Netherlands and The National Committee for International Cooperation and Sustainable Development, The Netherlands, who supported the participation of several participants. Publication of this special issue and a spin-off book was possible thanks to a grant from the Directorate General Science, Research and Development of the European Commission (BIO4CT98-4814). In addition, the Guest Editors are indebted to the following people, who served as the editorial board for the special issue by reviewing the submitted manuscripts: A. Ballesteros (Madrid, Spain); M.A. Borowitzka (Murdoch, Australia); G.B. Bremer (Portsmouth, UK); C. Bucke (London, UK); L. Dijkhuizen (Groningen, The Netherlands); M. Jaspars (Aberdeen, UK); Y. LeGal (Concarneau, France); F. Marinelli (Gerenzano, Italy); T. Matsunaga (Tokyo, Japan); M. Meiners (Emden, Germany); J. Miyake (Tsukuba, Japan); E. Molina Grima (Almeria, Spain); A. Muller-Feuga (Plouzan6,

Editorial

France); M.H.G. Munro (Christchurch, New Zealand); L.R. Mur (Amsterdam, The Netherlands); P.D. Nichols (Hobart, Tasmania, Australia); S.A. Pomponi (Fort Pierce, USA); B. Rinkevich (Haifa, Israel); G.L. Rorrer (Corvallis, USA); M.R. Tredici (Florence, Italy); and O.R. Zaborsky (Honolulu, USA). With the help of so many, we have been able to turn the international symposium 'Marine Bio-

3

process Engineering' into a high quality event, ultimately resulting in this special issue of the Journal of Biotechnology. The Guest Editors, R. Osinga, Wageningen, The Netherlands J. Tramper, Wageningen, The Netherlands J.G. Burgess, Edinburgh, United Kingdom R.H. Wijffels, Wageningen, The Netherlands


I

9

J O U R N A L

OF

Biotechnology ELSEVIER


The bioprocess-technological potential of the sea Shirley A. Pomponi * Division of Biomedical Marine Research, Harbor Branch Oceanographic Institution Inc., Fort Pierce, FL, USA

Received 19 October 1998; received in revised form 30 November 1998; accepted 22 December 1998

Abstract

Marine bioprocess engineers face a unique challenge for the millennium: designing methods for the sustainable development of known marine resources, as well as inventing a new generation of tools and processes that will enable a greater understanding of the ocean and its resources and lead to the discovery of new bioproducts for the future. The identification and application of novel, marine-derived pharmaceuticals, cosmetics, nutritional supplements, enzymes, and pigments have already been realized. The current and potential market value of these marine bioproducts is substantial. Continued discovery and development of marine resources will depend on a number of factors: identification of new bioproducts, sustainable use of the product, optimization of production, and efficient product recovery. Successfully addressing these challenges will require the integration and collaboration of mutidisciplinary teams of oceanographers, biologists, chemists, and engineers. 9 1999 Published by Elsevier Science B.V. All rights reserved. Keywords: Aquaculture; Invertebrate cell culture; Marine bioproducts; Marine pharmaceuticals; Sustainable use

I. Introduction

The marine environment has proven to be a rich source of both biological and chemical diversity. The oceans contain nearly 300 000 described species, but it is estimated that this number is only a small percentage of the total number of species that have yet to be discovered and described (Winston, 1988; Malakoff, 1997). Marine microorganisms represent the greatest percentage of undescribed marine species (Colwell, 1997). Marine bacteria alone could constitute as much as 10% of the total living biomass carbon of the biosphere (Parkes et al., 1994). * Fax: + 1-561-461-2221. E-mail address: [email protected] (S.A. Pomponi)

F r o m a relatively small number of these species that have been studied to date, thousands of chemical compounds have been isolated (Ireland et al., 1993), yet only a small percentage of these compounds has been studied for their potential as useful products. The oceans represent a virtually untapped resource for discovery of even more novel chemicals with potential as pharmaceuticals, nutritional supplements, cosmetics, agrichemicals, molecular probes, enzymes and fine chemicals. Each of these classes of marine bioproducts has a potential multi-billion dollar market value (BioScience, 1996). There are several marine-derived products currently on the market (Table 1). This discussion will focus on the current status and future poten-

0168-1656/99/$ - see front matter 9 1999 Published by Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00053-X

6

S.A. Pomponi / Journal of Biotechnology 70 (1999) 5-13

tial of bioprocess engineering related to the discovery, development, and sustainable use of marine-derived compounds with biomedical applications. Marine bioprocess engineers will be an integral part of the multidisciplinary approach to discovery and development of new marine-derived pharmaceuticals and molecular probes to study human disease processes. Novel tools and processes are required for identification of new products, sustainable use of the products, optimization of production, and efficient product recovery.

2. Identification of new products and development of novel screening technologies Marine organisms for drug discovery research have, to date, been collected using relatively routine methods, such as scuba diving, submersibles, dredging and trawling. Submersibles enable scientists to access unusual habitats, such as vent communities and deep sea benthic habitats (Fig. 1),

and some systems are equipped with specialized tools and chambers that allow samples to be maintained under ambient conditions, i.e. high pressure and low temperature. There is still a need for the development of versatile bioreactors that can be deployed and operated in extreme environments (e.g. hypersaline, vent, anoxic, and deep sea habitats). Such bioreactors could be used for collection, at-sea maintenance, and evaluation of both macroorganisms and microorganisms, so that their metabolites could be evaluated under physiological conditions that are as similar as possible to ambient conditions. Another approach to the identification of new products is the incorporation of miniaturized biosensors into both collecting tools and bioreactors for rapid, in situ analysis of wild and cultivated marine organisms for target molecules. A number of miniaturized biosensors and probes to study human disease processes are in development. Adaptation of these for in situ evaluation of marine-derived products would be an interest-

Table 1 Some examples of commercially available marine bioproducts Product

Application

Original source

Method of production

Ara-A

Antiviral drug

Marine sponge

Ara-C

Anticancer drug

Marine sponge

Microbial fermen- McConnell et al., tation of analog 1994 Chemical synthesis McConnell et al., of analog 1994 Cell culture Tachibana et al., 1981 Wild harvest of Glaser and sponge Jacobs, 1986 Recombinant Mattila et al., protein 1991 Cell culture ESPGAN, 1991

Molecular probe: phosDinoflagellate phatase inhibitor Molecular probe: phospholi- Marine sponge, Luffariella Manoalide variabilis pase A 2 inhibitor Deep sea hydrothermal Vent TM DNA polymerase Polymerase chain reaction vent bacterium enzyme Fatty acids used as additive Marine microalgae Formulaid | (Martek Biosciences, Columbia, MD) in infant formula nutritional supplement Bioluminescent jellyfish, Bioluminescent calcium Aequorin Aequora victoria indicator Bioluminescent jellyfish, Reporter gene Green Fluorescent Protein Aequora victoria (GFP) Conjugated antibodies used Red algae Phycoerythrin in ELISAs and flow cytometry Resilience | (Estre Lauder) 'Marine extract' additive in Caribbean gorgonian, Pseudop terogorgia skin creams

Okadaic acid

elisabethae

References

Recombinant protein Recombinant protein Cell culture

Badminton et al., 1995 Chalfie et al., 1994 Glazer, 1989

Wild harvest of gorgonian

Look et al., 1986


Fig. 1. The drug discovery research program at Harbor Branch Oceanographic Institution uses the Johnson-Sea-Link manned submersibles to conduct targeted, controlled collections of unique invertebrates from deep-water habitats such as rocky slopes and steep vertical walls that are not accessible by dredging and trawling. ( 9 Branch Oceanographic Institution, Inc.)

ing bioengineering challenge. One example is GeneChip | technology (Affymetrix, Santa Clara, CA), a rapidly developing field in which high-density oligonucleotide arrays on prefabricated solid chips are used to simultaneously analyze expression levels of thousands of genes (Fodor et al., 1991). The GeneChip probe arrays are manufactured using a photolithographic process similar to that used for making a computer microchip. A photo-protected glass substrate is selectively exposed by light passing through a photolithographic mask. The unmasked areas are photo-activated, the substrate is incubated with nucleosides, and chemical coupling occurs at the activated sites. This process of photo-activation and chemical coupling is repeated until the desired set of probes is synthesized. The GeneChip probe array can contain tens to hundreds of thousands of different oligonucleotide or cDNA probes, arranged in precise sequences and locations. A fluorescently labelled sample is analyzed by hybridization with the GeneChip probe array. The GeneChip probe array is then scanned by a laser, and the fluorescent pattern is analyzed for matches to known sequences. Although GeneChip probe arrays are currently used for diagnostics

7

and research primarily in mammalian systems, the technology could be adapted to identify new or known molecules from marine invertebrates, algae, and microorganisms for drug discovery research. The engineering challenge is to adapt the technology for rapid, in situ screening of marine organisms and their bioproducts. It would involve specialized tools for sampling the target organisms, preparing the samples for hybridization with the oligonucleotide arrays, and analyzing the arrays with modified scanners--all in situ. Potential applications are the identification of new or previously untested species, as well as analysis of gene expression that is specific to a particular disease or therapeutic area. The biological evaluation of marine-derived extracts and pure compounds has been based on assays developed for the 'high-throughput' screening of large libraries of synthetic compounds. They measure a number of end-points, such as activation or inhibition of enzymes or receptors involved in human disease processes, inhibition of growth of human pathogenic microorganisms, and toxicity against human cancer cells (Suffness et al., 1989; Ireland et al., 1993; McConnell et al., 1994; Munro et al., 1994). None of the assays used in major pharmaceutical drug discovery programs takes into account the role of marinederived compounds in nature, i.e. the in situ biochemical functions of both primary and secondary metabolites, and how those functions may be applied to the discovery of new drugs and probes to study human disease processes. Marine organisms as model systems offer the potential to understand and develop treatments for disease based on the normal physiological role of their secondary metabolites. For example, the mechanisms of action of Conus toxins are well-known (Hopkins et al., 1995; Shon et al., 1997), and are currently being applied to the development of new classes of drugs. Development of in situ biosensors would enhance our ability to probe the expression of secondary metabolites in response to various stimuli, lead to a better understanding of the role of the secondary metabolites in nature, and perhaps provide clues to the potential biomedical utility of these compounds.

8

S.A. Pomponi /Journal of Biotechnology 70 (I 999) 5-13

3. Sustainable use of marine resources

With the enormous potential for discovery, development, and marketing of novel marine bioproducts comes the obligation to develop methods by which these products can be supplied in a way that will not disrupt the ecosystem or deplete the resource. Supply of most marine-derived compounds is a major limiting factor for further pharmaceutical development. Often, the metabolite occurs in trace amounts in the organism, and a steady source of supply from wild harvest cannot provide enough of the target compound for preclinical studies. In general, the natural abundance of the source organisms will not support production based on wild harvest. Some options for sustainable use of marine resources are chemical synthesis, controlled harvesting, aquaculture of the source organism, in vitro production through cell culture of the macroorganism or microorganism source, and transgenic production. Each of these options has its advantages and limitations, not all methods will be applicable to supply of every marine bioproduct, and most of the biological supply methods are still in development. The approach to be used will be based on a number of factors: 3.1. Complexity of the molecule: can it be synthesized using an industr&lly feasible process7 Synthetic processes have been published for many marine bioproducts in development as pharmaceuticals (e.g. Kageyama and Tamura, 1990; Corey et al., 1996; Harried et al., 1997). Unfortunately, most of these are multi-step processes that are not amenable to economic, industrial-scale synthesis. The cost of reagents, affordable on a small scale, may be prohibitively expensive to use for large-scale synthesis. In addition, many natural products are complex molecules with several stereocenters, making it necessary to control the stereochemistry of the chemical reactions. If the compound is an enantiomer, and both isomers are synthesized, the process may require separation of the isomers by chromatographic techniques which could be costly on a large scale.

Other sources of bulk supply are generally required for early-stage drug development, during which time the pharmaceutical company may also optimize chemical synthesis at the industrial scale. There is thus an opportunity and a market for supply of marine bioproducts from aquaculture, in vitro production, and transgenic production, not only for long-term applications, but also for early-stage validation of their utility. 3.2. Abundance of the organism in nature: what do we know about the impact of collections on the habitat or species populations7 Prior to large-scale wild harvest of an organism for recovery of a bioproduct, harvesting feasibility studies should be conducted. These should define factors such as the standing stock of the organism, its growth rate and the factors that affect growth, and the impact of harvesting and postharvesting recovery of the target organism. These data could then be used not only to assess the potential of supply from wild harvest, but also to develop models for aquaculture and/or in vitro production. Unfortunately, this is rarely done. One notable exception is the survey of the New Zealand sponge, Lissodendoryx sp., which produces the halichondrins, potent antitumor compounds. The sponge is restricted in occurrence to one locality off the Kaikoura Peninsula. Research conducted by Battershill et al. (1998) demonstrated that harvesting by dredging significantly reduced the standing stock of the sponge, but that the population may recover rapidly as a result of asexual propagation of sponge fragments dispersed as a result of the dredging. This study suggested that harvesting is only feasible for smaller quantities of the sponge. On the other hand, the bryozoan Bugula neritina, the source of the anticancer compound bryostatin 1 (Pettit et al., 1982), is a common fouling organism found throughout the world in both temperate and tropical habitats. The US National Cancer Institute collected more than 12 000 kg wet weight of this 'nuisance' organism from docks and pilings with apparently little impact on the populations. Moreover, this single bulk collection was sufficient to supply about 18 g


of bryostatin 1--enough to conduct all preclinical and clinical trials (Schaufelberger et al., 1991). Recent studies indicate, however, that there are actually two 'chemotypes' of Bugula neritina: samples occurring deeper than 9 m generally contain bryostatin 1, whereas samples collected from depths shallower than 9 m do not. In addition, only the Pacific samples have consistently yielded bryostatin 1 (Dominick Mendola, CalBioMarine Technologies, personal communication). This type of information is not only necessary in planning controlled harvesting, but is also critical for selection of genetically superior brood stock for aquaculture and in vitro production.

3.3. Source of the compound: is it microbially produced? A significant number of marine bioproducts with pharmaceutical potential have been identified from heterotrophic marine microorganisms isolated from coastal sediments (Fenical, 1993; Kobayashi and Ishibashi, 1993; Davidson, 1995). In addition, some marine bioproducts originally isolated from macroorganisms, such as sponges, have been subsequently discovered to be localized in microbial associates (e.g. Bewley et al., 1996). If these symbiotic microorganisms can be isolated and cultured, optimization of production in marine microbial bioreactors may lead to an industrially feasible supply option. The large-scale, photoautotrophic production of polyunsaturated fatty acids for use as nutritional supplements (e.g. Formulaid | Martek Biosciences) demonstrates the commercial feasibility of photoautotrophic microbial culture for bulk supply. Bulk supply of bioproducts from both photoautotrophic and heterotrophic microbial culture presents some bioengineering challenges that are addressed by a number of papers in this volume. If the source of the compound is the macroorganism itself, development of in vitro production methods could provide bulk supply of the compound. Research in progress in our laboratory on in vitro production of sponge metabolites has resulted in the establishment of primary sponge cell cultures that can be stimulated by lectins and

9

other growth-regulating compounds to divide and continue to produce bioactive compounds after doubling (Fig. 2) (Pomponi et al., 1997, 1998). The objective of our research is to establish cell lines of bioactive marine invertebrates that can be used as models to study in vitro production of bioactive metabolites and the factors which control expression of production. This could ultimately lead to in vitro production of marine bioproducts, including production of compounds that may only be produced by an intact invertebrate-microorganism symbiotic association. More importantly, an understanding of the cellular and molecular processes that control production of these metabolites could be used to enhance upstream processing/culture optimization and to stimulate production of 'unnatural' natural products--i.e, chemicals that the organism would not produce under normal conditions, but which may be more potent than the 'natural' product. As discussed by Rinkevich (this volume), there are a few obstacles that still need to be overcome before marine invertebrate in vitro production can become a viable option for bulk supply of bio225

e,-

._Q m c (.1 c

8 O

.G

100

r162 c

0.00

u aays

36 hours

"firm Fig. 2. Concentration of stevensine per cell in cultures of the sponge Teichaxinella morchella incubated in phytohemagglutinin (PHA), expressed as percent of control. One population doubling occurred within 36 h in PHA-stimulated cultures; control cultures did not divide. After 8 days, PHA-treated cultures showed an increase in stevensine concentration per cell (relative units) (From Pomponi et al. (1997)).

10

S.A. Pomponi Journal of Biotechnology 70 (1999) 5-13

Fig. 4. Typical grow-out plate with aquacultured Bugula neritina colonies attached, following retrieval from the undersea structure by divers after 5 months in the sea. (Photo courtesy of CalBioMarine Technologies, Carlsbad, CA, USA.)

Fig. 3. Launching of CalBioMarine's prototype in-sea structure for aquaculture of Bugula neritina to yield bryostatin 1. (Photo courtesy of CalBioMarine Technologies, Carlsbad, CA.

USA.) products, but our data indicate that in vitro production of bioactive metabolites by marine invertebrate cells is feasible.

3.4. In situ growth conditions: is aquaculture an option for deep water organisms? Both in-the-sea and land-based aquaculture methods have been developed by CalBioMarine Technologies (Carlsbad, CA) for the bryozoan, Bugula neritina (Figs. 3 and 4), and for Ecteinascidia turbinata (Figs. 5 and 6) the ascidian from which the antitumor compound, ecteinascidin 743, has been isolated (Rinehart et al., 1990; Wright et al., 1990). These are both common, shallow-water organisms for which reproduction and growth have been studied, but the factors controlling production of the compounds are not yet completely known. The New Zealand deepwater sponge, Lissodendoryx sp., is the source of the antitumor compounds, the halichondrins. The sponge occurs at 85-105 m, but has been cultured successfully from cuttings on lantern arrays in shallower wa-

ter, and has maintained production of the halichondrins (Battershill et al., 1998). Current efforts are directed toward modification of metabolite production by altering the microenvironment (Battershill, personal communication). This indicates that aquaculture of some deep water sponges is feasible, however, species from deeper water may have more critical growth requirements, such as high pressure and low temperature. Although in-the-sea aquaculture is a cost-effective method of production, it still does not afford the opportunity for over-expression of production of the compounds or for complete control of environmental parameters. Development of

Fig. 5. Underwater close-up of aquacultured colony of Ecteinascidia turbinata on poly-line. (Photo courtesy of CalBioMarine Technologies. Carlsbad, CA, USA.)


Fig. 6. Lab-settled colonies of Ecteinascidia turbinata on polylines in an indoor aquaculture tank, Keys Marine Lab, Long Key, FL, USA. (Photo courtesy of CalBioMarine Technologies, Carlsbad, CA, USA.)

closed-system bioreactors for the culture of both shallow water and deep water organisms is a particularly challenging opportunity for marine bioprocess engineers. Research in progress by Osinga et al. (1997) (and this volume) demonstrates the feasibility of this approach for shallowwater sponges. A combination of cell culture and aquaculture techniques may be applied to development of novel, closed bioreactor systems. For example, three-dimensional matrices may be used to establish seed cultures in vitro from dissociated cells of biomedically important species. Once established, the seeded matrices could be transferred into larger, closed bioreactors for optimization, scaleup, and product recovery.

3.5. Biosynthetic pathway" is genetic engineering realistic for the compound7 If the biosynthesis of the target compound is understood, it may be possible to identify, isolate, clone, and express in a heterologous host the genes responsible for production of the metabo-

11

lite. In many cases, of course, biosynthesis of the product is not known, or it is a multi-step process involving several enzymatic reactions. For these cases, transgenic production is not a trivial process. Matsunaga (1998) succeeded in transgenic production of eicosapentaenoic acid (EPA) by cloning an EPA gene cluster, isolated from Shewanella putrefaciens, into the marine cyanobacterium Synechococcus. Alternatively, chemoenzymatic synthesis, by which marine bioproducts are synthesized in cellfree, enzyme-based systems, offers a complementary technique to in vitro and transgenic production methods. Kerr et al. (1996a,b) have demonstrated the conversion of basic biosynthetic building blocks to the target molecule for a number of marine bioproducts, including bryostatin from Bugula neritina and pseudopterosins from the gorgonian Pseudopterogorgia americana (Figs. 7 and 8).

4. Optimization of production Perhaps the area in which marine bioprocess engineering has had the greatest impact--and still has the greatest opportunity for novel developments--is in the design and optimization of bioreactors for marine metabolite production. Since bioreactor design is addressed by several other authors in this volume, suffice it to say that Secosteroid Production using Enzymes

from

Local C o r a l .

% ~ . o~o~,~~~,.t,, ,o~,~j

Fres~ty collected

~o.,

"

F,lter, po . . . . to

~, (.,0

9 , . t t t t l _~ S~lut,or' of

Punfied Enzymes

STEROID (e.g Cholesterol)

Enzyme Powder

SECOSTEROIO ANTI-INFLAMMATORY .AGENTS - efl~oent, inexpensive new production method

Fig. 7. The first step in chemoenzymatic synthesis of antiinflammatory secosteroid involves production of active enzyme precipitate (PAP) from the gorgonian Pseudopterogorgia americana. (Photo courtesy of Dr Russell Kerr, Florida Atlantic University, Boca Raton, FL, USA.)

S.A. Pomponi /Journal of Biotechnology 70 (1999) 5-13

12

Enzyme Column STEROID

-- (1 m g

r.holesterol 1100 mL buffer)

Im P A l )

9(11)-SECOSTEROID

(100%

yield cholesterol)

-~

Fig. 8. A steroid (e.g. cholesterol) is transformed into the anti-inflammatory secosteroid product by the active enzyme precipitate (PAP) immobilized and packed in a column. (Photo courtesy of Dr Russell Kerr, Florida Atlantic University, Boca Raton, FL, USA.)

a variety of bioreactor designs have been implemented, with varying degrees of success. The opportunity to produce new, bioactive structural analogs of known compounds via manipulation of culture conditions presents marine bioprocess engineers and their collaborators in marine natural products biochemistry, cellular and molecular biology with a unique challenge for new bioproduct discovery. Innovations in media development (chemical engineering), bioreactor design (bioprocess engineering), and transgenic production (molecular engineering), coupled with efficient downstream processing and product recovery, will be necessary to meet the needs of both discovery and bulk production of novel marine bioproducts. In summary, marine bioprocess engineers face a unique challenge for the millennium: designing methods for the sustainable development of known marine resources, as well as inventing a new generation of tools and processes that will enable a greater understanding of the ocean and its resources and lead to the discovery of new bioproducts for the future. Successfully addressing these challenges will require the integration and collaboration of multidisciplinary teams of marine biologists, pharmacologists, cell and molecular biologists, biochemists, and engineers.

References Badminton, M.N., Kendall, J.M., Sala-Newby, G., Campbell, A.K., 1995. Nucleoplasmin-targeted aequorin provides evidence for a nuclear calcium barrier. Exp. Cell Res. 216, 236-243. Battershill, C.N., Page, M.J., Duckworth, A.R., Miller, K.A., Bergquist, P.R., Blunt, J.W., Munro, M.H.G., Northcote, P.T., Newman, D.J.. Pomponi, S.A., 1998. Discovery and sustainable supply of marine natural products as drugs, industrial compounds and agrochemicals: chemical ecology, genetics, aquaculture and cell culture. In: Origin and Outlook: 5th International Sponge Symposium 1998, Book of Abstracts. Queensland Museum, Brisbane, Australia, p. 16. Bewley, C.A., Holland, N.D., Faulkner, D.J., 1996. Two classes of metabolites from Theonella swinhoei are localized in distinct populations of bacterial symbionts. Experientia 52, 716-722. BioScience, 1996. Marine Biotechnology Special Issue, 46. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W., Prasher, D.C., 1994. Green fluorescent protein as a marker for gene expression. Science 263, 802-805. Colwell, R.R., 1997. Microbial biodiversity and biotechnology. In: Reaka-Kudla, M.E., Wilson, E. (Eds.), Biodiversity II: Understanding and Protecting Our Biological Resources. Joseph Henry Press, Washington, DC, pp. 279-287. Corey, E.J., Gin, D.Y, Kania, R.S., 1996. Enantioselective total synthesis of ecteinascidin 743. J. Am. Chem. Soc. 118, 9202-9203. Davidson, B.S., 1995. New dimensions in natural products research: cultured marine microorganisms. Curr. Opin. Biotechnol. 6, 284-291. ESPGAN Committee on Nutrition, 1991. Comment on the content and composition of lipids in infant formulas. Acta Paediatr. Scand. 80, 887-896. Fenical, W., 1993. Chemical studies of marine bacteria: developing a new resource. Chem. Rev. 93, 1673-1683. Fodor, S.P.A, Read, J.L., Pirrung, M.C., Stryer, L., Lu, A.T., Solas, D., 1991. Light-directed, spatially addressable parallel chemical synthesis. Science 251, 767-773. Glaser, K.B., Jacobs, R.J., 1986. Molecular pharmacology of manoalide. Inactivation of bee venom phospholipase A2. Biochem. Pharmacol. 35, 449-453. Glazer, A.N., 1989. Light guides. Directional energy transfer in a photosynthetic antenna. J. Biol. Chem. 264, 1-4. Harried, G.Y., Strawn, M.A., Myles, D.C., 1997. The total synthesis of (-)-discodermolide: an application of the chelation-controlled alkylation reaction. J. Org. Chem. 62, 6098-6099. Hopkins, C., Grilley, M., Miller, C., Shon, K.J., Cruz, L.J., Gray, W.R., Dykert, J., Rivier, J., Yoshikami, D., Olivera, B.M., 1995. A new family of Conus peptides targeted to the nicotinic acetylcholine receptor. J. Biol. Chem. 38, 2236122367.

S.A. Pomponi/Journal of Biotechnology 70 (1999) 5-13

Ireland, C.M., Copp, B.R., Foster, M.D., McDonald, L.A., Radisky, D.C., Swersey, J.C., 1993. Biomedical potential of marine natural products. In: Attaway, D.H., Zaborsky, O.R. (Eds.), Marine Biotechnology. Pharmaceutical and Bioactive Natural Products, vol. 1. Plenum Press, New York, pp. 1-43. Kageyama, M., Tamura, T., 1990. Synthesis of bryostatin 7. J. Am. Chem. Soc. 112, 7404-7408. Kerr, R.G., Lawry, J., Gush, K.A., 1996a. In vitro biosynthetic studies of the bryostatins, anticancer agents from the marine bryozoan Bugula neritina. Tetrahedron Lett. 37, 8305-8308. Kerr, R.G., Rodriguez, L., Kellman, J., 1996b. A chemoenzymatic synthesis of 9(ll)-secosteroids using an enzyme extract of the marine gorgonian Pseudopterogorgia americana. Tetrahedron Lett. 37, 8301-8304. Kobayashi, J., Ishibashi, M., 1993. Bioactive metabolites of symbiotic marine microorganisms. Chem. Rev. 93, 17531769. Look, S.A., Fenical, W., Jacobs, R.S., Clardy, J., 1986. The pseudopterosins: anti-inflammatory and analgesic natural products from the sea whip Pseudopterogorgia elisabethae. Proc. Natl. Acad. Sci. USA 83, 6238-6240. Malakoff, D., 1997. Extinction on the high seas. Science 277, 486-488. Matsunaga, T., 1998. Screening of marine microalgae for fine chemicals and bioremediation. In: Wijffels, R.H., et al. (Eds.), International Symposium Marine Bioprocess Engineering, Wageningen Agricultural University, Wageningen, The Netherlands, Book of Abstracts, p. 14. Mattila, P., Korpela, J., Tenkanen, T., Pitkanen, K., 1991. Fidelity of DNA synthesis by the Thermococcus litoralis DNA polymerase - - an extremely heat stable enzyme with proofreading activity. Nucleic Acids Res. 19, 4967-4973. McConnell, O.J., Longley, R.E., Koehn, F.E., 1994. The discovery of marine natural products with therapeutic potential. In: Gullo, V.P. (Ed.), The Discovery of Natural Products with Therapeutic Potential. Butterworth-Heinemann, Boston, pp. 109-174. Munro, M.H.G., Blunt, J.W., Lake, R.J.U., Litaudon, M., Battershill, C.N., Page, M.J., 1994. From seabed to sickbed: what are the prospects? In: Van Soest, R.W.M., Van Kempen, T.M.G., Braekman, J.-C. (Eds.), Sponges in Time and Space, Proceedings of the 4th International Porifera Congress. A.A. Balkema, Rotterdam, pp. 473484. Osinga, R., Tramper, J., Wijffels, R.H., 1997. Cultivation of marine sponges. In: Marino, D., Willensky, D., Capone, D. (Eds.), 4th International Marine Biotechnology Conference, Abstracts. Stazione Zoologica 'Anton Dohrn', Naples, Italy, p. 210. Parkes, R.J., Cragg, B.A., Bale, S.J., Getliff, J.M., Goodman, K., Rochelle, P.A., Fry, J.C., Weightman, A.J., Harvey, S.M., 1994. Deep bacterial biosphere in Pacific Ocean sediments. Nature (London) 371,410-413.

13

Pettit, G.R., Herald, C.L, Doubek, D.L, Herald, D.L., 1982. Isolation and structure of bryostatin 1. J. Am. Chem. Soc. 104, 6846-6848. Pomponi, S.A., Willoughby, R., Kaighn, M.E., Wright, A.E., 1997. Development of techniques for in vitro production of bioactive natural products from marine sponges. In: Maramorosch, K., Mitsuhashi, J. (Eds.), Invertebrate Cell Culture: Novel Directions and Biotechnology Applications. Science Publishers, Inc, Enfield, New Hampshire, USA, pp. 231-237. Pomponi, S.A., Willoughby, R., Wright, A.E., Pecorella, C., Sennett, S.H., Lopez, J., Samples, G., 1998. In vitro production of marine-derived antitumor compounds. In: Le Gal, Y., Halvorson, H.O. (Eds.), New Developments in Marine Biotechnology. Plenum Press, New York, pp. 7376. Rinehart, K.L., Holt, T.G., Fregeau, N.L., Stroh, J.G., Keifer, P.A., Sun, F., Li, L.H., Martin, D.G., 1990. Ecteinascidins 729, 743, 745, 759A, 759B, and 770: potent antitumor agents from the Caribbean tunicate Ecteinascidia turbinata. J. Org. Chem. 55, 4512-4515. Schaufelberger, D.E., Koleck, M.P., Beutler, J.A., Vatakis, A.M., Alvarado, A.B., Andrews, P., Marzo, L.V., Muschik, G.M., Roach, J., Ross, J.T., Lebherz, W.B., Reeves, M.P., Eberwein, R.B., Rodgers, L.L., Testerman, R.P., Snader, K.M., Forenza, S., 1991. The large-scale isolation of bryostatin 1 from Bugula neritina following good manufacturing practices. J. Nat. Prod. 54, 12651270. Shon, K.J., Grilley, M., Jacobsen, R., Cartie, G.E., Hopkins, C., Gray, W.R., Watkins, M., Hillyard, D.R., Rivier, J., Torres, J., Yoshikami, D., Olivera, B.M., 1997. A noncompetitive peptide inhibitor of the nicotinic acetylcholine receptor from Conus purpurascens venom. Biochemistry 31, 9581-9587. Suffness, M., Newman, D.J., Snader, K., 1989. Discovery and development of antineoplastic agents from natural sources. In: Scheuer, P.J. (Ed.), Bioorganic Marine Chemistry, Vol. 3. Springer-Verlag, New York, pp. 131-168. Tachibana, K., Scheuer, P.J., Tsukitani, Y., Kikuchi, H., Van Engen, D., Clardy, J., Gopichand, Y., Schmitz, F.J., 1981. Okadaic acid, a cytotoxic polyether from two marine sponges of the genus Halichondria. J. Am. Chem. Soc. 103, 2469-2471. Winston, J.E., 1988. The Systematists' Perspective. In: Fautin, D.G. (Ed.), Biomedical Importance of Marine Organisms. California Academy of Sciences, San Francisco, CA, pp. 1-6. Wright, A.E., Forleo, D.A., Gunawardana, G.P., Gunasekera, S.P., Koehn, F.E., McConnell, O.J., 1990. Antitumor tetrahydroisoquinoline alkaloids from the colonial ascidian Ecteinascidia turbinata. J. Org. Chem. 55, 4508-4512.


JOURNAl.

OF

Biotechno,logy ELSEVIER


The discovery and development of marine compounds with pharmaceutical potential Murray H.G. Munro a,, John W. Blunt a, Eric J. Dumdei a, Sarah J.H. Hickford a, Rachel E. Lill a, Shangxiao Li ", Christopher N. Battershill b, Alan R. Duckworth b a Department of Chemistry, University of Canterbuo,, PB 4800, Christchurch, New Zealand b National Institute of Water and Atmospheric Research, PO Box 14-901, Wellington, New Zealand

Received 12 October 1998; received in revised form 1 December 1998; accepted 22 December 1998

Abstract An assessment of the current status of marine anticancer compounds is presented along with a case study on the aquaculture of Lissodendoryx n. sp. 1, a sponge that produces the antimitotic agents halichondrin B and isohomohalichondrin B. The use of polymer therapeutics to enhance the properties of marine natural products is considered. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Biodiversity; MarinLit; Sponge aquaculture; Anticancer; Antimitotic; Polymer therapeutics

1. The discovery phase In contrast to work on terrestrial natural products the first serious work on studying marine natural products started just 50 years ago with the pioneering work of Bergman (e.g. Bergrnan and Feeney, 1951). While the difficulties of collecting marine samples cannot be underestimated, a large number of easily accessible marine samples are available simply by shore-wading. That the opportunity was not seriously grasped until the * Corresponding author. Tel.: + 64-3-364-2434; fax: + 643-364-2110. E-mail address: [email protected] (M.H.G. Munro)

1940s is possibly a commentary on the difficulties of isolation and purification of marine natural products with the limited techniques available at that time. However, since the 1940s the field has blossomed and matured. In 1997 there were 713 papers published on marine natural products. This is out of a total of 10 311 papers recorded in MarinLit, a database dedicated to the marine natural products literature (MarinLit, 1998). At the time of the mid-year release of the 1998 version of MarinLit, 484 new papers had been included. From the marine literature it is the Porifera that have been the most studied phylum followed closely by the Cnidaria, Chromophycota, Rhodophycota, Mollusca, Chordata and the Echinodermata (MarinLit, 1998).

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00052-8

16

M.H.G. Munro et al. , Journal o f Biotechnology 70 (1999) 1 5 - 2 5

8,246

Microorganisms

Absolute number screened in itafics

Marine Plants

6,540

Marine Animals

Terrestrial Plants

18,293

Terrestrial animals

434

1

0

Percentage

2

Data by courtesy of Dr Peter Murphy, AIMS

Fig. 1. Distribution of samples with significant cytotoxicity in the NCI's preclinical screen.

Over the years some distinct trends have emerged in the study of marine natural products. One has been the emphasis on the discovery of new bioactive natural products. Initial work by Bergman was undoubtedly curiosity-driven, but it was his discovery of the biologically-active, pharmaceutically important and novel arabino-nucleosides from the sponge Cryptotethya crypta that sparked interest in marine natural products and served to highlight the biomedical potential of the field (Bergman and Feeney, 1951). With advances in chromatographic techniques for dealing with polar compounds along with better analytical and structural elucidation technology an increasing proportion of the compounds isolated have shown cytotoxic properties (suggestive of potential antitumour compounds). In an early review (Munro et al., 1987) covering the marine literature up to early 1986, 185 bioactive compounds were reported. In 1993 a review (Schmitz, 1994) covering the next 5 years commented on an additional 400 compounds. A survey of MarinLit reveals that this trend has continued with some 46% of all cytotoxic compounds in the database having been reported since 1993.

As a source of bioactive compounds with pharmaceutical potential how well does the marine environment compare with the more traditional areas such as terrestrial microorganisms and plants? The best comparative data is that published by Garson based on statistical data from the US National Cancer Institute (NCI) screening programme provided by Dr Peter Murphy. This clearly indicated that marine invertebrates are a preferred source due to the much higher incidence of significant cytotoxic activity (Garson, 1994) (Fig. 1). If those screening data for marine animals are in turn examined on a phylum basis certain phyla (e.g. Porifera, Bryozoa, Chordata) have a higher incidence of bioactivity with the trend becoming very obvious as species with very significant bioactivity (ICs0 < 2 lag ml-~) are selected (Fig. 2). As the data in Fig. 1 suggest, the sampling of oceanic life-forms enhances the probability of discovering species from natural sources with potential anticancer activities. This can be rationalised as a sampling strategy which accesses the widest range of phyla. Greater than 70% of all recorded living species belong to the animal kingdom.

M.H.G. Munro et al.//Journal of Biotechnology 70 (1999) 15-25

17

16- i

12-J 10-~! % Actives 8 -~ 4-~I

Z-i t

02

t-o

N

,

m

E

~

------.

10.8), the polymerization was observed to occur within 15 min after shifting the pH ( < 6); the primary particles (6-7 nm) finally formed large aggregates ( > 7 0 nm) in 2 h (Perry et al., 1991). The advantage of the low pH of the SDV for diatoms is that the formed siliceous parts are protected from continuous dissolution that may destroy fragile or important details such as processes, ribs and spiculae (Vrieling et al., 1999b).

5. Characterization of diatomaceous silica

In comparison to chemically derived silicabased materials, biogenic silica has been studied quite poorly. In the past diatomaceous earth has been examined, but this appeared to be less interesting for industrial applications since its characteristics (e.g. surface area, porosity, contamination) limit its use as alternative for synthetically derived silica materials. Minor attention has been paid to cultured or freshly harvested silica of diatoms and this most probably reflects a lack of interest or even a prejudice of chemists, because of the assumed close relation to highly aged diatomaceous earth. For the classification of diatoms electron microscopy is very important and even at low magnifications morphological alterations can easily be followed (Vrieling et al., 1999c). For detailed

46

E.G. Vrieling et al./Journal of Biotechnology 70 (1999) 39-51

analysis of the silica, high-resolution microscopy may give insight into crystalline phases (not observed yet in biogenic silica) or the size and arrangement of silica particles or spheres in amorphous systems. By combining microscopy with analytical methods such as energy dispersion Xray or microprobe analysis, the elemental composition of the SiO2-matrix can be determined (van Bennekom et al., 1989; Swift and Wheeler, 1992; Vrieling et al., 1999c). This approach, however, is most often not sensitive enough for elements that are present at concentration < 0.2%. In order to determine the coordination of silicon molecules in biogenic silica and silica-based materials 29Si N M R can be applied (Perry, 1989; Wijnen et al., 1993; Burkett and Davis, 1994). N M R is very sensitive and alterations such as replacements o f - O H groups by - O - b r i d g e s can be identified. Furthermore, N M R can discern silica samples that seem to be similar by powder diffraction methods (Perry 1989). For solids (as diatomaceous silica is) the outer boundaries of the particles can be studied and as a consequence the tetrahedral coordination of silicon can be determined; the diatom Navicula pelliculosa contains silicon as Si(OSi-)4, Si(OSi-)3OH and Si(OSi-=-)2(OH)2, but there is no evidence yet that S i - O - C or S i - O - N bonds are present in their frustule silica. In situ 29Si N M R measurements during dissolution of diatomaceous silica have not been carried out, but such observations may answer questions related to the bonds between organic matter (the potential nucleators or structure directing agents) and silicon. To determine the surface area or pore size range of silicas or zeolites, sorption methods have often been applied. For diatomaceous silica a number of measurements has been performed (Table 2), showing that the specific surface area differs among and even within species. As mentioned, the silica diatoms is covered by an organic coating before excretion to the aquatic environment. Apparently the removal of this matter is quite difficult and no efficient and reproducible procedures have been designed yet. Furthermore, the effect of temperature used to dry the samples (to remove remaining water from capillary cavities) and the often small sample size (in the order

of milligrams rather than grams) affect the estimation of the representative specific surface area as m 2 per gram material (Beelen, Dokter, Vrieling, unpublished results). As a consequence of uncertainties the measured specific surface area of the silica of diatoms has not been compared properly to areas of synthetically derived silica-based materials. Diffraction methods are useful to determine crystallinity of silica samples. Powder X-ray diffraction (referred to as XRD) examines domains over l0 nm and peaks in the spectra indicate presence of crystalline phases. Based on powder diffraction, diatomaceous silica of different origin (cultures, earth) seems to be comparable and based on the absence of clearcut peaks the amorphous character was determined. However, infrared spectroscopical analysis of the same samples showed that the absorption patterns and ratios of absorption intensities are different, indicating that the silica is not exactly identical among species (Kamantani, 1971; Perry, 1989). For diatoms, crystalline phases have not been reported (see above), but it was suspected that at domains < l0 nm (below the limit of XRD) crystalline phases exist (Pickett-Heaps et al., 1990). Heat-treated diatomaceous earth indeed revealed crystalline phases (Hurd, 1983), but recent analysis by wide angle X-ray scattering (WAXS) showed that at domains between 0.07 and 0.5 nm no crystalline phases are present in silica from recent freshly cultured and collected diatoms (Fig. 4; Vrieling et al., 1999b). Small angle X-ray scattering (SAXS) discerns between matter (in our case silica) and the surrounding (water, air). SAXS allows accurate analysis of silica-based materials with respect to the particle/pore sizes (distributed over the silica solids or in solutions) and organization of particles/pores in clusters or cluster sizes (Dokter, 1994; Beelen, 1996; de Moor, 1998; Vrieling et al., 1999a). Using SAXS, silica of a selection of cultured and field-harvested diatoms has now been examined, showing that at over a size domain of 5.0-35.0 nm the slopes of the scattering spectra of all species and field samples were comparable (Fig. 5). From these data it could be concluded that the scatterers (the pores in the silica of diatoms), over the measured size range, have a quite

E.G. Vrieling et al. /Journal of Biotechnology 70 (1999) 39-51

consistent dimension and architecture (Vrieling et al., 1999a). Detailed electron microscopical exami-

47

nation and SAXS revealed that pore size-range was determined to vary between 30 and 500 .~

Table 2 Specific surface area (m 2 g-1) for silica of cultured diatoms, field diatoms and diatomaceous earth determined by sorption analysis a Species

Specific surface area (m 2 g-1)

Method to remove organic matter

Sorption method

References

Coscinodiscus asteromphalus Lauderia borealis

89

Nitric acid

N2

Lewin (1961)

75-269 b

9

Navicula pelliculosa Nitzschia sigma N. sigma

123

Low temperature ashing Nitric acid

N2

van Beusekom and van Bennekom, personal communicaton Lewin (1961)

10.5-25.0 c 9.3

N2 N2

This contribution This contribution

N. sigma

9.0

N2

This contribution

N. sigma

12.0

N2

This contribution

Odontella sinensis

49

EGME d

van Bennekom et al. (1991)

Odontella sinensis

44

N2

This contribution

Rhizoselenia hebetata R. tesselata

60

Nitric acid H202, KMnO4, oxalic acid H2SO 4, KMnO4, oxalic acid K2S20 8, KMnO4, oxalic acid Low temperature ashing Low temperature ashing Nitric acid

N2

Kamantani and Riley (1979)

EGME d

van Bennekom et al. (1991)

Thalassiosira descipiens T. eccentrica

258

N2

Kamantani and Riley (1979)

9

T. nordenskirldii

16--127b

T. punctigera

15-43 b

Dutch Wadden Sea 1997 e Dutch Wadden Sea 1998 Dutch Wadden Sea 1998 Dutch Wadden Sea 1998 Dutch Wadden Sea 1998 Diatomaceous earth

87

Low temperature ashing Low temperature ashing Low temperature ashing H202, H2504

N2

van Beusekom and van Bennekom, personal communicaton van Beusekom and van Bennekom, personal communicaton van Beusekom and van Bennekom, personal communicaton This contribution

17-165 r

Nitric acid

N2

This contribution

84

H202, KMnO4, oxalic acid H2SO4, KMnO4, oxalic acid K2S208, KMnO4, oxalic acid Nitric acid

N2

This contribution

N2

This contribution

N2

This contribution

N2

Iler (1979)

50

19

69 48-52 22

Low temperature ashing Nitric acid

9 9

a Notes: b Surface area depended on culture condition with respect to additions of aluminium. c Surface area depended on the applied sorption instrumentation. d The ethylene glycol monoethylether technique of Heilman et al. (1965). e Natural populations of benthic diatoms harvested from tidal flats.


48

Overlap region of mean .lopee of Itm ~ samples tested by SAXS (Mmr97)

I !

Wldden Sm 2 Wmdden See 1 T.wliBIbgl T. ptmctlgem

m~l

ysis indicates that silica of cultured diatoms has its own characteristic features that are not found in industrial silica-based materials, but the translation of these data to a potential use of this type of silica has yet to be made.

"--

!

,

! ,,,

[

S. oonsldda S. coslalum N. lhefnlllil N. sigma N. dolltwJum N. lallnMlall N.CfylNOCqd~l O. #inenm L. bomells C. gnmit

I

_,

__

6. Conclusion and perspectives |

.

! I

A. )mx.~a

1

-1.00 10

-0.75 1

i

i

-0.50

-

-0.25

0.00

Log Q (nm"1)

1

1

1

i

Diatom silicon biomineralization is related to silica chemistry when chemical processes that are involved in the formation of the silica are consid-

-

I

1

9

0.25 1

i-

1

65 60 55 50 45 40 35 30 25 20 15 10 5

~

A

d (nm)

~ Fig. 5. Overview of the results obtained by SAXS of 14 diatoms and two natural populations of benthic species (indicated as Wadden Sea 1 and 2). The marked area represent the overlapping size region (5 < d < 35 nm) of the scattering pores in the silica; the SAXS spectra (see also Fig. 6) in this region are comparable for all species. Methods: SAXS was performed as described in Fig. 4, but scattering was now recorded as a function of scattering angle, applying a camera length of 3.50 m, by means of a multi-wire gas filled quadrant detector. Scattering spectra were presented as log scattering intensity (arbitrary unit) versus log Q (size measurement related to the scattering angle denoted as n m - 1) to determine the slope(s) as indicator for the structure, dimension and spatial relation/distance of pores in the silica (Vrieling et al., 1999a). For SAXS analysis the Q-axis was calibrated using a specimen of wet rat tail collagen.

=

-3,54

4

c"

B < 10")

6

and this is comparable to mesoporous systems (pores > 10 A), but not to microporous systems (pores < 10 A). In comparison to chemically derived silicas the SAXS spectra of the biogenic silica of diatoms reveal a porous structure rather than fractal aggregates. This structure is closely related to the shape and dimensions of the pores in the frustule and it can be altered by inducing changes in the silica by varying the salinity (the salt concentration, see above) or adding aluminium (cations, see above). Both parameters have effect on the scattering behaviour (the slopes of the SAXS spectra; Fig. 6A) and the preferential distribution of pores over the silica (the curvature of the SAXS spectra; Fig. 6B). Thus, SAXS anal-

4

-0.4

-o.2

o:o

0.2

-1

Log O (nm) Fig. 6. Effects of the salinity of the culturing medium and enrichments with A1CI3 on the SAXS spectra of the pennate diatom Navicula salinarum. (A) By lowering the salinity from PSU 28 (top spectrum) to PSU 20 (bottom 2 spectra) the structure or geometric dimensions of the pores in the silica (5 < d < 50 nm) change, resulting in a decrease of the slopes in SAXS-spectra. (B) By adding aluminium (Al/Si-ratios increase in spectra from top to bottom) the distribution of the pores and their preferential spacing is affected, resulting in variation in the curvatures of the spectra. Note: SAXS analysis was performed as described in Fig. 5.


ered. We have argued that the pH, the presence of cations and salts and the concentration of silica precursors play an important role in biosilicification of diatoms, apparently natural silicon biomineralization is quite conform to the rules known in 'classical' silica chemistry. The low contamination of 'fresh' diatomaceous silica with metals (e.g. A1, Fe, Ti) indicates that diatoms produce a nearly pure SiO2-matrix that does not require exhaustive purification, contaminants of this kind are often present in large-scale synthetic silica products. Although diatomaceous silica has hardly ever been characterized as to physicochemical properties, there are now indications that species or species complexes represent unique forms with respect to both morphology and molecular structure. Direct applications of silica obtained from cultured diatoms as alternatives for synthetically derived silica-based materials are not foreseen yet, but it is obvious that nature provides us with a wealth of new structures, many of them of interest within the field of industrial silica. Moreover, these new morphologies are an inspiration in the quest for new structure-directed synthesis methods. To exploit these new possibilities, more effort should be put into the characterization of diatomaceous silica with emphasis on: intra-species variability; the yield of silica by mass culturing; product optimalization (via specified growth conditions using e.g. altered salinity or adding cations); and the biological processes involved in diatom silicification. According to Ozin et al. (1997) and Yang et al. (1997) structure directing agents can be used to develop tailor-made mesoporous silica materials. By characterizing the organic molecules that are identified in the SiO2-matrix of diatoms, the role of these molecules in silicon biomineralization by diatoms can be determined. Eventually, such molecules can be useful in chemical syntheses to develop silica that mimicks the characteristic one of diatoms. The development of tailor-made silicas, with the variety of pore sizes observed in diatoms, provides not only alternatives to existing silica or zeolites, but has the potential to be used in new applications as well.

49

Acknowledgements We thank P.-P.E.A. de Moor, F. Giimiisburun and E. van Oers (Eindhoven University of Technology), B.U. Komanschek (Synchrotron Radiation Source, Daresbury Laboratory, UK) and W.H. van de Poll and M. Keijmel (Department of Marine Biology) for their assistance during various experiments, the results of which are presented in this contribution. A.J. van Bennekom (Netherlands Institute of Sea Research, NL) and J.E.E. van Beusekom (Biologische Anstalt Helgoland, Germany) kindly supplied data on the specific surface area of some diatoms. E.G.V. was supported by the Netherlands Technology Foundation (STW; grant GBi55.3883) which is subsidized by the Netherlands Organisation for the Advancement of Pure Research (NWO).

References Addadi, L., Moradian-Oldak, J., Weiner, S., 1991. Macromolecule-crystal recognition in biomineralization. Studies using synthetic polycarboxylate analogs. In: Sikes, C.S., Wheeler, A.P. (Eds.), Surface Reactive Peptides and Polymers; Discovery and Commercialization, vol. 444. ACS Symposium Series, Dallas, TX, pp. 13-27. Beelen, T.P.M., 1996. Inorganic particle gels. Curr. Opin. Coll. Intersurf. Sci. 1, 718-725. Bergna, H.E., 1994. The colloid chemistry of silica. Adv. Chem. Ser. 234, 517-531. Bhattacharyya, P., Volcani, B.E., 1980. Sodium-dependent silicate transport in the apochlorotic marine diatom Nitzschia alba. Proc. Natl. Acad. Sci. 77, 6383-6390. Brinker, C.J., Scherer, G.W., 1990. Sol-Gel Science. Academic Press, New York. Burkett, S.L., Davis, M.E., 1994. Mechanism of structure direction in the synthesis of Si-ZSM-5: an investigation by intermolecular IH-29Si CP MAS NMR. J. Phys. Chem. 98, 4647-4653. Conley, D.J., Kilham, S.S., Theriot, E., 1989. Differences in silica content between marine and freshwater diatoms. Limn. Oceanogr. 34, 205-213. Corma, A., 1997. Preparation and catalytic properties of new mesoporous materials. Top. Catal. 4, 249-260. Dokter, W.H., 1994. Transformations in silica gels and zeolite precursors. Ph.D. Thesis. Eindhoven University of Technology, The Netherlands. Gensemer, R.W., 1990. Role of aluminium and growth rate on changes on cell size and silica content of silica-limited populations of Asterionella ralfsii var. americana (Bacillariophyceae). J. Phycol. 26, 250-258.

50

E.G. Vrieling et al.//Journal of Biotechnology 70 (1999) 39-51

Gordon, R., Drum, R.W., 1994. The chemical basis of diatom morphogenesis. Int. Rev. Cyt. 150, 243-372. Harrison, C.C., 1996. Evidence for intramineral macromolecules containing protein from plant silica. Phytochemistry 41, 37-42. Hecky, R.E., Mopper, K., Kilham, P., Degens, E.T., 1973. The amino acid and sugarcomposition of diatom cell-walls. Mar. Biol. 19, 323-331. Heilman, M.D., Carter, D.L., Gonzalez, C.L., 1965. The ethylene glycol monoethyl-ether (EGME) technique for determining soil-surface area. Soil Sci. 100, 409-413. Hildebrand, M., Higgins, D.R., Busser, K., Volcani, B.E., 1993. Silicon-responsive cDNA clones isolated from the marine diatom Cylindrotheca fusiformis. Gene 132, 213218. Hildebrand, M., Volcani, B.E., Gassmann, W., Schroeder, J.I., 1997. A gene family of silicon transporters. Nature 385, 688-689. Hurd, D.C., 1983. Physical and chemical properties of siliceous skeletons. In: Aston, S.R. (Ed.), Silicon Geochemistry and Biogeochemistry. Academic Press, London, pp. 187-244. Iler, R.K., 1979. The Chemistry of Silica. Wiley, New York. Jansen, J.C., Strcker, M., Karge, H.G., Weitkamp, J., 1991. Advanced zeolite science and applications. Studies in Surface Science and Catalysis, vol. 85. Elsevier, Amsterdam. Kamantani, A., 1971. Physical and chemical characteristics of biogenous silica. Mar. Biol. 8, 89-95. Kamantani, A., Riley, J.P., 1979. Rate of dissolution of diatom silica in seawater. Mar. Biol. 55, 29-35. Lewin, J.C., 1961. The dissolution of silica from diatom walls. Geochim. Cosmochim. Acta 21, 182-198. Lewin, J.C., 1966. Silicon metabolism in diatoms. V. Germanium dioxide, a specific inhibitor of diatom growth. Phycologia 6, 1-12. Lobel, K.D., West, J.K., Hench, L.L., 1996. Computational model for protein-mediated biomineralization of the diatom frustule. Mar. Biol. 126, 353-360. de Moor, P-P.E.A., 1998. The mechanism of organic-mediated zeolite crystallization. Ph.D. Thesis. Eindhoven University of Technology, The Netherlands. Okita, T., Volcani, B.E., 1980. Role of silicon in diatom metabolism. X. Polypeptide labelling pattems during the cell cycle, silicate starvation and recovery in Cylindrotheca fusiformis. Exp. Cell Res. 125, 471-481. Ozin, G.A., Yang, H., Sokolov, I., Coombs, N., 1997. Shell mimetics. Adv. Mater. 9, 662-667. Paasche, E., 1973. Silicon and the ecology of marine plankton diatoms. 2. Silicate-uptake kinetics in five diatom species. Mar. Biol. 19, 262-269. Perry, C.C., 1989. Chemical studies of biogenic silica. In: Mann, S., Webb, J., Williams, R.J.P. (Eds.), Biomineralization: Chemical and Biochemical Perspectives. VCH, New York, pp. 223-256. Perry, C.C., Fraser, M.A., Hughes, N.P., 1991. Macromolecular assemblages in controlled biomineralization. In: Sikes, C.S., Wheeler, A.P. (Eds.), Surface Reactive Peptides and

Polymers; Discovery and Commercialization, vol. 444. ACS Symposium Series, Dallas, TX, pp. 316-339. Pickett-Heaps, J., S~hmid, A.M., Edgar, L.A., 1990. The cell biology of diatom valve formation. Progr. Phycol. Res. 7, 1-168. Robinson, D.H., Sullivan, C.W., 1987. How do diatoms make silicon biominerals. TIBS 12, 151-154. Romero, O., 1994. Morphological variation of Skeletonema costatum (Greville) Cleve (Bacillariophyceae) in Aysen fiord, Chile. Rev. Biol. Mar. 29, 1-21. Round, F.E., Crawford, R.M., Mann, D.G, 1990. The Diatioms. Biology and Morphology of the Genera. Cambridge University Press, Cambridge. Schmid, A.M.M., 1986. Wall morphogenesis in Coscinodiscus wailesii Gran et Angst. II. Cytoplasmic events of valve morphogenesis. In: Ricard, M. (Ed.) Proceedings of the Eight International Symposium on Living and Fossil Diatoms. O. Koeltz, Koenigstein, Germany, pp. 294-314. Swift, D.M., Wheeler, A.P., 1991. Some structural and functional properties of a possible organic matrix from the frustules of the freshwater diatom Cyclotella meneghiniana. In: Sikes, C.S., Wheeler, A.P. (Eds.), Surface Reactive Peptides and Polymers; Discovery and Commercialization, vol. 444. ACS Symposium Series, Dallas, TX, pp. 340353. Swift, D.M., Wheeler, A.P., 1992. Evidence of an organic matrix from diatom biosilica. J. Phycol. 28, 202-209. Simpson, T.L., Volcani, B.E., 1981. Silicon and siliceous structures in biological systems. Springer, Heidelberg. Van Bekkum, H., Flanigan, E.M., Jansen, J.C., 1991. Introduction to zeolite science and practice. Elsevier, Amsterdam. Van Bennekom, A.J., Buma, A.G.L., Nolting, R.F., 1991. Dissolved aluminium in the Weddel-Scotia Confluence and the effect of A1 on the dissolution kinetics of biogenic silica. Mar. Chem. 35, 423-434. Van Bennekom, A.J., Jansen, J.H.F., van der Gaast, S.J., van Iperen, J.M., Pieters, J., 1989. Aluminium-rich opal: and intermediate in the preservation of biogenic silica in the Zaire(Congo) deep-sea fan. Deep Sea Res. 36, 173-190. Van de Poll, W.H., Vrieling, E.G., Gieskes, W.W.C., 1999. Location and expression of frustulins in the pennate diatoms Cylindrotheca fusiformis, Navicula pelliculosa and Navicula salinarum. J. Phycol. (submitted for publication). Vrieling, E.G., Beelen, T.P.M., van Santen, R.A., Gieskes, W.W.C., 1999a. Combined small and wide angle X-ray scattering reveals nano-scaled uniformity of pore architecture in diatomaceous silica. J. Phycol. (submitted for publications). Vrieling, E.G., Gieskes, W.W.C., Beelen, T.P.M., 1999b. Silicon deposition in diatoms: control by the pH inside the silicon deposition vesicle. J. Phycol. 35 (in press). Vrieling, E.G., Poort, L., Beelen, T.P.M. and Gieskes, W.W.C., 1999c. Growth and silicon content of the diatoms Thalassiosira weissflogii and Navicula salinarum at different salinities and enrichments with aluminium. Eur. J. Phycol. (submitted for publication).

E.G. Vrieling et al./Journal of Biotechnology 70 (1999) 39-51 Wijnen, P.W.J.G., Beelen, T.P.M., Rummens, K.P.J., van Santen, R.A., 1993. The role of cations in the formation of aqueous silica gels. J. Non-Crystall. Sol. 152, 127-136. Yang, H., Coombs, N., Ozin, G.A., 1997. Morphogenesis of shapes and surface patterns in mesoporous silica. Nature 386, 692-695.

51

Zaremba, C.M., Stucky, G.D., 1996. Biosilicates and biomimetic silicate synthesis. Curr. Opin. Solid State Mater. Sci. 1,425-429. Zones, S.I., Nakagawa, Y., Lee, G.S., Chen, C.Y., Yuen, L.T., 1998. Searching for new high silica zeolites through a synergy of organic templates and novel inorganic conditions. Micropor. Mesopor. Mater. 21, 199-211.


JOURNAL

OF



Biosynthesis and properties of an extracellular metalloprotease from the Antarctic marine bacterium

Sphingomonas paucimobilis Marianna Turkiewicz *, Ewa Gromek, Halina Kalinowska, Maria Zielifiska Institute of Technical Biochemistry, Technical University of Lodz, 4/10 Stefanowskiego Street, Lodz 90-924, Poland Received 14 October 1998; received in revised form 26 November 1998; accepted 22 December 1998

Abstract An extracellular protease from the marine bacterium Sphingomonas paucimobilis, strain 116, isolated from the stomach of Antarctic krill, Euphausia superba Dana, was purified and characterized. The excretion of protease was maximal at temperatures from 5 to 10~ i.e. below the temperature optimum for the strain growth (15"C). The highly purified enzyme was a metalloprotease [sensivity to ethylenediaminetetraacetic acid (EDTA)] and showed maximal activity against proteins at 20-30~ and pH 6.5-7.0, and towards N-benzoyl-tyrosine ethyl ester (BzTyrOEt) at pH 8.0. At 0~ the enzyme retained as much as 47% of maximal activity in hydrolysis of urea denatured haemoglobin (Hb) (at pH 7.0), and at - 5 and - 10~ 37 and 30%, respectively. The metalloprotease was stable up to 30~ for 15 min and up to 20~ for 60 min. These results indicate that the proteinase from S. paucimobilis 116 is a cold-adapted enzyme. 9 1999 Elsevier Science B.V. All rights reserved.

Keywords: Psychrophilic enzymes; Metalloprotease; Antarctic marine bacteria

1. Introduction

Abbreviations: BSA, bovine serum albumin; BzTyrOEt, Nbenzoyl-tyrosine ethyl ester; CFU, colony forming unit; EDTA, ethylenediaminetetraacetic acid; Hb, haemoglobin; IAA, iodoacetic acid; PAGE, polyacrylamide gel electrophoresis; pCMB, p-chloromercuribezoate; PMSF, phenylmethylsulfonyl fluoride; Rm, relative mobility; TCA, trichloroacetic acid; Tris, tris(hydroxymethyl)aminomethane. * Corresponding author. Tel.: + 48-42-6366618; fax: + 4842-6313402. E-mail address: [email protected] (M. Turkiewicz)

The overwhelming majority of commercially important enzymes are produced by mesophilic microorganisms and applied in conditions resembling their environment in vivo. New biocatalysts active in unusual conditions are looked for a m o n g extremophilic microorganisms including psychrophiles and psychrotrophs (Adams et al., 1995). They constitute the most abundant group of organisms, since temperatures below 5~ occur in about 80% of biosphere and even more than 90% of marine environment (Margesin and Schinner, 1994; Brenchley, 1996).

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All fights reserved. PII: S0168-1656(99)00057-7

54

M. Turkiewicz et al./ Journal of Biotechnology 70 (1999) 53-60

Cold-adapted enzymes secreted by these microorganisms, also called psychrophilic enzymes or psychrozymes, show high catalytic efficiency at low temperatures, especially at 0-20~ i.e. a temperature range at which mesozymes are usually inactive, as well as reduced thermostability resulting in denaturation at moderate temperatures (Margesin and Schinner, 1994; Feller et al., 1996). Cold enzymes are potentially useful for numerous biotechnological processes, especially those which require a supply of exogenous energy, are exposed to higher risk of microbial contamination or temperature instability of reactants or products (Margesin and Schinner, 1994; Marshall, 1997). Another profit resulting from studies on microbial psychrozymes is an enrichment of pure culture collection by strains existing in permanently cold environments, e.g. marine polar regions (DeLong, 1997). Information acquired in this way are important for development of our still very limited knowledge of biodiversity.

rpm. The strain growth was observed by determination of CFUs on solid growth medium (0.3% of beef extract, 0.5% of bactotryptone and 5.51% of bacto-marine agar 2216, Difco) or by absorbance reading at 660 nm. All other chemicals employed in the studies were purchased from standard sources and were at least reagent grade.

2. Materials and methods

2.3. Protein determination

2.1. Organ&m and cultivation

Protein was assayed according to Lowry et al. (1951) using bovine serum albumin (BSA) as a standard, or spectrophotometrically at 280 nm in fractions eluted from chromatography columns.

Gram-negative rods Sphingomonas paucimobilis, strain no. 116, from the collection of the Department of Antarctic Biology of Polish Academy of Sciences in Warsaw, were used for the studies. This strain was isolated from a stomach of Antarctic krill, Euphausia superba Dana by Stuart Donachie at the Arctowski Polish Antarctic Station. The strain classification was done using API ZONE strips (Api System, Biomerieux, France) at 15~ as well as microscopic observations and Gram staining (Donachie, 1995). Shaken cultures of the strain were performed at pH 8.0 in a liquid medium containing synthetic marine water (Instant Ocean, Aquarium System Inc., France), 0.2% of bactopeptone and 1% of casein digest (Difco). One hundred and twenty ml of the medium in a 500 ml flask was inoculated with a standard inoculum (4% v/v, 1.5 x 105 CFU (colony forming units) ml-1) and cultivated at 10~ in a shaker (INFORS-Switzerland) at 160

2.2. Protease assay Proteolytic activity was determined according to Anson (1938) using urea denatured Hb (pH 7.0, 30~ reaction time 15 min) or spectrophotometrically according to Hummel (1959) using Nbenzoyl-tyrosine ethyl ester (BzTyrOEt) in 50 mM tris(hydroxymethyl)aminomethane (Tris)-HC1 buffer pH 8.0 at 20~ (increase of absorbance at 254 nm). In both methods, the activity was expressed in lamol of product (L-tyrosine present in trichloroacetic acid (TCA)-soluble hydrolysis products or N-benzoyl-L-tyrosine, respectively) liberated from respective substrate for 1 min in standard conditions.

2.4. Electrophoresis Polyacyrlamide gel electrophoresis (PAGE) was run in nondenatured conditions, according to Laemmli (1970), using 7.5% gels (0.6 x 8 cm) in 0.1 M borate buffer pH 9.0 and current intensity 2.5 mA per gel. Gels were fixed and stained with Coomassie brillant blue R-250. Proteolytic activity was assayed in 2 mm slices of nonstained gels against denatured Hb (2% v/w, 30~ pH 7.0) as a substrate.

2.5. Protease purification Purification of the protease was carried out at 4~ and in presence of Ca 2 + ions (0.5 mM) as a stabilizing factor. The protein was precipitated

M. Turkiewicz et al./Journal of Biotechnology 70 (1999) 53-60

from the centrifuged (5000 x g, 30 rnin) culture medium with ammonium sulphate (80% of saturation, 1 h, 0~ dissolved in 10 mM sodium phosphate buffer pH 7.0 and desalted on Sephadex G-25 (0.9 x 20 cm) equilibrated with the same buffer. Five ml protein fractions (A28o above 0.8) were pooled and applied to a bacitracin-Sepharose 4B column (0.9 x 10 cm, the buffer as above), prepared according to Stepanov and Rudenskaya (1983). The column was prewashed with the starting buffer (flow rate 20 ml h - ~) and the bound proteins were eluted using this buffer enriched with 25% (v/v) of isopropanol and 1 M of NaC1. Active fractions were pooled, dialysed against the starting buffer and stored in small aliquots at - 40~

3. Results

3.1. Effect of temperature on growth and protease secretion S. paucimobilis 116 was grown at temperatures from 0 to 30~ The highest proteolytic activity in the culture medium was observed in a range from 5 to 10~ (about 0.25-0.26 units ml-1, Fig. 1). At 15~ which is optimal for the strain growth, as well as at 0~ the activity was twice lower. At 30~ the strain did not excrete extracellular proteases. Above 30~ its growth was not observed. At 10~ which was chosen as a standard

55

6-

~ 0,30 i

4-" 5 i r,.)

7

~

0

,

2

5

"~

,

3

0,15

2

0,10

1

0,05

.~ >

"=

._o e~

o 0

1

2

3

4

:

~ -

i O,O0

5

6

7

cultivationtime(days) proteolyticactivity

--a- biomass +

Fig. 2. Relationship between the strain growth at 10~ and the proteolytic activity in the culture medium during 7 days of cultivation. Proteolytic activity was assayed in standard conditions using urea denatured Hb.

strain cultivation temperature, the protease synthesis occurred to run simultaneously to the growth and reached the maximum after 72 h, i.e. at a late logarithmic phase of bacterial growth (Fig. 2). Protein pattern of the third day supernatant of the culture medium shows Fig. 3. Two protein bands were active against denatured Hb. The proteolytic activity present in this supernatant was inhibited by 1 mM phenylmethylsulfonyl fluoride (PMSF) (34% of inhibition) and 1 mM EDTA (ethylenediaminetetraacetic acid) (38% of inhibition) which proves that S. paucimobilis 116 70 -

65,2

50

0,30 "~

40

0,25 30

.~ 0,20 ==

20

0,15

10

,* 0,10

0

.-1

4

7

10

13

I6

19 22

25 28

31

34

37

0,05 number o f 2 mm gel sfices

0,00 0

5

10

15

20 3O temperature ("C)

Fig. 1. Effect of S. paucimobilis 116 growth temperature on synthesis of extracellular proteases. Proteolytic activity was assayed in culture medium after 72 h of cultivation, in standard conditions, using urea denatured Hb.

Fig. 3. Electophoretic pattern of proteins and its relationship with proteolytic activity in the culture medium after 72 h of the strain growth.


56 A~so 0,40 ]-

(-

- 0,35

i

0,35 -i-

+ 0,30 -," - 0,25 .~

0,30 i 0,25

T 036

0,20

:~

o,15

0,15

0,10

0,10

0,05

0,05 0,00 I

6

11

---N- A~u

16

......... 21 26 31 ~activity

0,00 36

41

46

51 56 fraction number

(-) pooled fraction

Fig. 4. Elution profile of proteins (A280) and proteolytic activity in affinity chromatography of the strain proteases on bacitracin-Sepharose 4B. The arrow shows the start of elution with 25% (v/v) isopropanol.

S. paucimobil& 116 protease was most active at 25-30~ for 15 min, and at 15-20~ for 60 min, against urea denatured Hb (Fig. 6). At - 1 0 to 0~ the enzyme retained 30-47% of maximal activity. The apparent energy of activation for the hydrolysis of urea denatured Hb calculated from an Arrhenius plot was 24.3 kJ mol-~ The S. paucimobilis 116 metalloprotease is very thermolabile, since the complete stability was observed for 15 min up to 30~ and for 60 min up to 20~ (Fig. 7). At 45~ only 38% of the activity remained after 15 min, and after 60 min it was inactive. The metalloprotease digested Hb at pH from 5.0 to 10.5 with maximum at pH 6.5-7.0

produces extracellular serine proteases and metalloproteases, p-Chloromercuribezoate (p-CMB) and iodacetic acid (IAA) did not influence the activity.

, . ,i

3.2. Metalloprotease properties The highly purified metalloprotease was obtained using the procedure described in Section 2. The enzyme was isolated from 760 ml of the supernatant from 72 h S. paucimobilis 116 culture medium (Table 1). Specific activity of the pooled active fractions after affinity chromatography (Fig. 4) was about 20-fold higher in comparison to the starting supernatant, both against urea denatured Hb and BzTyrOEt. Although bacitracin-Sepharose 4B coupled all the proteolytic activity (Fig. 4), the fractions eluted from this column contained only 10% of it. The eluted proteolytic activity was in 92% inhibited by EDTA (final concentration 1 mM, enzyme-inhibitor incubation time 1 h) and not influenced by PMSF. It proves that only the metalloprotease was eluted from the column and the serine protease present in the starting material remained bound to bacitracin-Sepharose 4B. The yield of purification was low (3-4%, Table 1). PAGE of this highly purified enzymic preparation run in non-denatured conditions showed one strong protein band with relative mobility (Rm) 0.58 contaminated with trace amount of another protein with R m 0.83 (Fig. 5).

..

9

'

:ri!'

Fig. 5. PAGE of the highly purified metalloprotease from S. paucimobilis 116 (pooled active fractions after bacitracin-sepharose 4B column).

Table 1 Purification of a metalloprotease from S. paucimobilis 116 ~~

Purification step

Volume (ml)

Protein (mg)

Activity Specific (units mgof protein)

'

Yield ('5))

Purification (fold)

Hb

Hb

\

Total (units)

BzTyrOEt

BzTyrOEt

6 1:

4,

Culture supernatant Salted out protein Sephadex G-25 Bacitracin-Sepharose 4B

760 10

43 5.4

313.20 29.84 11.70 0.46

Hb

BzTyrOEt

Hb

BzTyrOEt

0.23 1.40 2.33 5.60'

0.04 0.15 0.28 0.72'

71.82 41.84 27.30 2.60

11.70 4.39 3.29 0.33

* The activity was inhibited by EDTA (in 92%) and not by PMSF.

a 100

59 38 3.6

100

-

-

38 28 3

6 10

4 8

24

19

58

M. Turkiewic: et al ..... Journal of Biotechnology 70 (1999) 53-60

100

100-

80 60 20 :

-10

-5

40

b

0

5

I0

15

20

25

30

35

40

45

20-

50

temperature (~

0

Fig. 6. Effect of temperature on the activity of the S. pauchnobilis 116 metalloprotease. The activity was assayed for two reaction times: (a) 15; and (b) 60 min, using urea denatured Hb pH 7.0.

(Fig. 8). Optimal pH in BzTyrOEt hydrolysis was 8.0 and at pH 10.5 the enzyme showed 30% of the maximal activity (Fig. 8). At 4~ the protease was stable at pH 6.0-8.0 for 60 min, and at pH 9.5 exhibited 91% of maximal activity (data not presented). The preferred substrate of the highly purified metalloprotease from S. paucimobilis 116 was denatured Hb (Table 2). Its activity against casein, fibrinogen and native Hb was 20-30% lower. BSA was digested to a small extent.

4. Discussion

A metalloprotease active against various proteins at neutral and slightly alkaline pH was isolated from extracellular proteases secreted by

80

a

5,0

6,0

7,0

8,0

9,0

10,0

11,0 pH

Fig. 8. Influence of pH on the activity of the S. paucimobilis 116 metalloprotease. The enzyme activity was assayed against urea denatured Hb (a) and BzTyrOEt (b).

S. paucimobilis strain 116, isolated from stomachs of Antarctic krill E. superba Dana. S. paucimobilis 116 synthesizes extracellular proteases with highest intensity at a temperature markedly lower (5-10~ than corresponding to its maximal growth rate (15~ similarly to many other coldadapted microorganisms, including Antarctic ones (Margesin and Schinner, 1994; Feller et al., 1996). It is presumed that intensification of enzyme biosynthesis at low temperatures is one of the main mechanisms of adaptation of these organisms to surrounding environment (Margesin and Schinner, 1994; Brenchley, 1996). Secretion of proteases by S. paucimobilis 116 starts already in the phase of adaptation and its maximum coincides with the end of logarithmic phase, falling on 72 h. This period of time is shorter than observed e.g. by Margesin and Schinner (1992a) for a metalloprotease from Alpine

~ Table 2 Activity of the S. paucimobilis 116 metalloprotease against selected proteins

60*

20,'

0

a

5

10

20

30

35

40

45

50

temperature (~

Fig. 7. Thermostability of the S. paucimobilis 116 metalloprotease. The enzyme was incubated at different temperatures (0-50~ for: (a) 15; and (b) 60 rain and the residual activity towards urea denatured Hb was determined in standard conditions (30~ pH 7.0).

Substrate

Activitya (units m1-1)

Hb denatured Hb native Casein Fibrinogen BSA

0.249 0.160 0.200 0.177 0.017

a Assay conditions: 30~ pH 7.0, 2% solutions of protein substrates, 15 min reaction.

M. Turkiewic: et al. /Journal of Biotechnolog3" 70 (1999) 53-60

strain of Pseudomonas fluorescens (96 h) or for subtilisin from Antarctic Bacillus TA 39 (150 h, Feller et al., 1996). Electrophoretic analysis and inhibitory tests proved that S. paucimobilis 116 produces at least two extracellular proteases, i.e. a metallo- and a serine enzyme. However we have managed to purify only the metalloenzyme. The serine protease coupled with bacitracin-Sepharose 4B but was immobile in a milieu of the eluent employed in our studies. According to Stepanov and Rudenskaya (1983) ligand from bacitracin-Sepharose possesses various amino acid residues and therefore binds not only serine proteases, but some other too which was revealed in model experiments on papain, subtilisin and B. subtilis metalloprotease. Kinetic properties such as very weak thermostability and low optimal temperature (15-30~ depending on reaction time) of the S. paucimobilis 116 metalloprotease point to its increased adaptation to low temperatures in comparison to many other enzymes, also originating from permanently cold, e.g. Antarctic habitats. Reported metalloproteases from other strains inhabiting cold environments usually exhibit higher optimal temperatures. Good examples are proteases from Alpine strain of P. fluorescens 114, most active at 40-45~ for 30 rain (Hamamoto et al., 1994) and from Xanthomonas maltophilia with maximal activity at 50~ for 15 rain (Margesin and Schinner, 1991). However, a protease isolated by K/irst et al. (1994) from an unidentified Antarctic bacterial strain was most active against proteins at 28~ Another important feature pointing to the psychrophilic character of the metalloprotease from S. paucimobilis 116 is its high activity at temperatures from 0 to 5~ (47-60% of maximal activity), resembling the strain physiological temperature range, as well as the ability to protein hydrolysis below 0~ (e.g. a t - 1 0 ~ the enzyme shows 30% of maximal activity against urea denatured Hb) which was not reported for other coldadapted proteases. For comparison, a metalloprotease from P. fluorescens 114 exhibited 20-30% of maximal activity at temperatures from 0 to 10~ (Hamarnoto et al., 1994), enzymes from Alpine strains showed 15-18% of maximal activ-

59

ity at 10~ (Margesin and Schinner, 1992b), the protease from X. maltophilia did not hydrolyse azocasein at 0~ and at 10~ retained only about 5% of maximal activity (Margesin and Schinner, 1991). The value of activation energy calculated for the S. paucimobilis 116 (24.3 kJ mol-1) is lower than obtained for enzymes from X. maltophilia (61.9 kJ tool-l, Margesin and Schinner, 1991), psychrophilic P. fluorescens strains (36.938.0 kJ mol-~, Margesin and Schinner, 1992a) or Antarctic psychrophile Bacillus TA41 (38.5 kJ mol- 1, Davail et al., 1994). In all the cases, values of activation energies of mesozymes tested under the same conditions were even higher (Margesin and Schinner, 1991, 1992a; Davail et al., 1994). The reduction of energy barrier displayed by the S. pauchnobilis 116 metalloprotease and correlated with its low temperature optimum and weak thermostability may result from more flexible conformation of the molecule in comparison to enzymes mentioned above. A potentially useful property of the S. paucimobilis 116 metalloprotease might be its relatively high activity towards native Hb, equal to about 70'/o of the activity against urea denatured Hb. Cold-adapted proteases may be applied in many fields, e.g. in production of detergents for washing at low temperatures, in tannery, in the food industry (haze removal from beer, bakery, cheese-making, production of fermented foods, meat tenderisation) or high protein waste degradation (Margesin and Schinner, 1994; Brenchley, 1996). Our studies demonstrate that proteases synthesized by marine Antarctic bacteria might become an interesting alternative to currently applied mesophilic enzymes.

References Adams. M.W.W.. Perler. F.B., Kelly, R.M., 1995. Extremozymes: expanding the limits of biocatalysis. Bio/' Technology 13, 662-668. Anson, M.L., 1938. Estimation of pepsin, trypsin papain and cathepsin with haemoglobin. J. Gen. Physiol. 22, 79-82. Brenchley, J.E.. 1996. Psychrophilic micro-organisms and their cold-active enzymes. J. Ind. Microbiol. 17, 432-437. Davail, S., Feller. G., Narinx, E., Gerday. C., 1994. Cold adaptation of proteins. J. Biol. Chem. 269, 17448-17453.

60


DeLong, E.F., 1997. Marine microbial diversity. Trends Biotechnol. 15, 203-207. Donachie, S.P., 1995. Ecophysiological description of marine bacteria from Admiralty Bay (Antarctica) and the digestive tracts of selected Euphausiidae, Philosophy Doctor Thesis, Polish Academy of Sciences, Warsaw. Feller, G., Narinx, E., Arpigny, J.L., Aittaleb, M., Baise, E., Genicot, S., Gerday, Ch., 1996. Enzymes from psychrophilic organisms. FEMS Microbiol. Rev. 18, 189-202. Hamamoto, T., Kaneda, M., Horikoshi, K., Kudo, T., 1994. Characterization of a protease from a psychrotroph Pseudomonas fluorescens 114. Appl. Environ. Microbiol. 60, 3878-3880. Hummel, B.C.W., 1959. A modified spectrophotometric method for determination of chymotrypsin and thrombin. Can. J. Biochem. Physiol. 37, 1393-1400. K/irst, U., Woehl, M., Czempinski, K., Schmid R.D., 1994. Characterization of extracellular hydrolases from marine psychrophilic bacteria. In: Seventh International Congress of Bacteriology and Applied Microbiology, Praha, Communicate MO122.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 165-275. Margesin, R., Schinner, F., 1991. Characterization of a metalloprotease from psychrophilic Xanthomonas maltophilia. FEMS Microbiol. Lett. 79, 257-262. Margesin, R., Schinner, F., 1992a. Production and properties of an extracellular metalloprotease from a psychrophilic Pseudomonas fluorescens. J. Biotechnol. 24, 207-210. Margesin, R., Schinner, F., 1992b. A comparison of extracellular proteases from three psychrotrophic strains of Pseudomonasfluorescens. J. Gen. Appl. Microbiol. 38, 209-225. Margesin, R., Schinner, F., 1994. Properties of cold-adapted micro-organisms and their potential role in biotechnology. J. Biotechnol. 33, 1-14. Marshall, C.J., 1997. Cold-adapted enzymes. Trends Biotechnol. 15, 359-363. Stepanov, V.M., Rudenskaya, G.N., 1983. Proteinase affinity chromatography on bacitracin-Sepharose. J. Appl. Biochem. 5. 420-428.

JOURNAL

OF



Cyanobacteria

a potential source of new biologically active substances

Susann Kreitlow *, Sabine Mundt, Ulrike Lindequist Institute of Pharmacy, Department of Pharmaceutical Biology, Ernst-Moritz-Arndt-University Greifswald, F.-L.-Jahn-Str. 15a, D- 17487 Greifswald, Germany


Abstract

Hydrophilic and lipophilic extracts of twelve cyanobacterial strains, isolated from flesh and brackish water, and two waterblooms, collected during the summer from the Baltic Sea, were investigated for their antibiotic activities against seven microorganisms. No inhibitory effects were found against the three Gram-negative bacteria Escherichia coli, Proteus mirabilis and Serratia marcescens and the yeast Candida maltosa. Of all cyanobacterial samples, extracts from seven species inhibited the growth of at least one of the Gram-positive bacteria Micrococcusflavus, Staphylococcus aureus and Bacillus subtilis. M. flavus proved to be the most sensitive bacterium in the agar diffusion test system. In particular, the hexane and dichlormethane extracts showed antimicrobial effects. But only one water extract, prepared from material of a natural waterbloom, was found to be active. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Cyanobacteria; Blue-green algae; Antibiotic; Antifungal; Agar diffusion method

Cyanobacteria (blue-green algae) provide a potential source of biologically active secondary metabolites (Teuscher et al., 1992). Investigations over the last decades have identified compounds with for instance cytotoxic, antifungal, antibacterial or antiviral activity. Most of the isolated substances belong to groups of polyketides, amides, alkaloids and peptides (Borowitzka, 1995; Falch, 1996; Kleinkauf and von D6hren, 1997). * Corresponding author. Tel.: + 49-3834-864865; fax: + 493834-864802. E-mail address: [email protected] (S. Kreitlow)

The blue-green algae are among the oldest photoautotrophic organisms. Their cultivation without organic substrates can be an economical advantage over other microorganisms. An optimised production of relevant compounds under controlled culture conditions is conceivable (Kulik, 1995). In view of the growing resistance of bacteria to c o m m o n antibiotics the search for new antimicrobially active compounds has become increasingly important (Trias and Gordon, 1997). In our screening programme we tested approximately fifty extracts from twelve different cyanobacterial strains and two waterblooms against different bacteria and one yeast. The re-


62

S. Kreitlow et al. 'Journal of Biotechnology 70 (1999) 61-63

sults of our study show the ability of cyanobacteria to produce compounds with antimicrobial effects. Ten filamentous cyanobacterial strains: Anabaena solitaria AS, Anabaena cylindrica 1611, Anabaena variabilis 1 and Anabaena sp. 7120 (Nostoccaceae); Oscillatoria tenuis 03, Oscillatoria sp. 022, Oscillatoria rubescens O16, Oscillatoria prolifica 02, Pseudanabaena catenata 154 and Limnothrix sp. 051 (Oscillatoriaceae) and the unicellular form Synechocystis aquatilis 428 (Chroococcaceae) were isolated from fresh water. The filamentous strain Nodularia spumigena 280 (Nostoccaceae) originated from brackish water. The waterblooms were collected from different regions of the Baltic Sea during the summer of 1997 and consist mainly of Anabaena lemmermannii (Waterbloom 97-7) and Microcystis ichthyoblabe, A. lemmermannii and Nodularia sp. (Waterbloom 97-8). The strains were cultivated batchwise in 2-1 flasks either in BG11- or Z1/2-medium (Meffert, 1971; Rippka et al., 1979). The cultures, which were illuminated continuously with warm-white light under room temperature (20_+ 2~ were harvested in their exponential growth phase by centrifugation (4~ 4000 x g). The freeze-dried cyanobacterial material (about 1 g) was taken for exhaustive extraction. Both lipophilic and hydrophilic extracts were screened for antibiotic activity. In general, the extraction was carried out successively with hexane, dichlormethane (or in one case: ethylacetate), methanol and water. After evaporation in vacuum, the extracts were stored at - 2 0 ~ until use. The antibacterial and antifungal activities of all samples were tested against the Gram-negative bacteria Escherichia coli, Serratia marcescens and Proteus mirabilis, the Gram-positive bacteria Staphylococcus aureus, Bacillus subtilis, Micrococcus flavus and the yeast Candida maltosa. The in vitro antimicrobial test utilised for the investigation is based on the diffusion method on agar plates (Collins et al., 1989). Therefore, nutrient agar was inoculated with a standardised quantity of a suspension of the respective organism. Sterilised paper disks (0 9 ram) containing the extract (2 mg) were transferred to the prepared agar

dishes. The antibiotic substance ampicillin was assayed for reference purposes. The inhibition zones (diameter minus diameter of the paper disk) were measured after 16 h incubation at 37~ (M. flavus at 22 _+ 2~ In order to increase the sensitivity of the test system a prediffusion for 4 h at 4~ was granted. Because of the small yields the dichlormethane extracts were not tested against all bacteria. None of the examined extracts has shown activity against Gram-negative bacteria. In addition, there were no inhibitory effects against the yeast C. maltosa. Among the water extracts only that from the natural waterbloom (97-8) was effective against S. aureus. The hexane extract from this sample gave the widest inhibition zones (3-4 ram) against B. subtilis. Two lipophilic extracts of the Limnothrix sp. 051 inhibited notably the growth of at least two of the tested Gram-positive bacteria. The dichlormethane extract of Oscillatoria sp. 022 showed the strongest activity against S. aureus and generally the largest inhibition zone (13 mm) as measured in our screening system. But no inhibitory effects of the other extracts of these species could be tested. M. flavus was the most sensitive bacterium against the methanol extract of O. rubescens 016. As reported in Table 1, which summarizes all extracts that showed activity against at least one of the test organisms, mainly hexane and dichlormethane extracts showed antimicrobial activities. The screening of the extracts was extended by testing their cytotoxicity against human fibroblast cells (Lindl and Bauer, 1994). In general, the lipophilic extracts showed stronger cytotoxic effects in concentrations of and above 100 ~tg m l (data not shown). However, there were no remarkable differences in cytotoxicity between extracts with antibiotic activities and extracts without antibiotic activities. In this early stage of the investigation, a differentiation between cytotoxic and antibiotic substances is not possible because the crude extracts contain a variety of different compounds. We were able to show that six out of twelve cyanobacterial strains cultivated under laboratory conditions produce substances with antibacterial activity. The ability of the blue-green algae to

S. Kreitlow et al./Journal of Biotechnology 70 (1999) 61-63

63

Table 1 Activity of the lipophilic and hydrophilic extracts of different cyanobacterial strains in a concentration of 2 mg paper d i s k - l a Zones of inhibition (in mm diameter) Cyanobacteria

Extract

S. aureus

B. subtilis

M. flavus

O. tenuis 03

Dichlormethane Methanol Methanol Hexane Hexane Dichlormethane Hexane Ethylacetate Water

+ + + + + -

+ + + + + + -

+ + + + + + + +

O. rubescens 016 Anabaena sp. 7120 A. solitaria AS Oscillatoria sp. 022 Limnothrix sp. 051 Waterbloom 97-8

+ + + +

+ + +

+

+ + +

+ + + + + + + + +

a--, No inhibitory effect; + , width < 1 mm; + + , width 1-8 mm; + + + , width > 8 mm.

produce antibiotic compounds could be an advantage for their survival in their natural environment (Teuscher et al., 1992). A large scale production to obtain sufficient cyanobacterial biomass is the precondition for the isolation and characterisation of the antibiotically and/or cytotoxically active compounds. Studies on this aspect are in progress.

Acknowledgements The authors thank Professor Kohl, Institute of Ecology, Humboldt-University, Berlin; Professor Pohl, Institute of Pharmacy, University Kiel; Dr Fulda, Botanical Institute, University Rostock and Dr Hfibel, Institute of Ecology, University Greifswald for supplying the cyanobacterial strains. The financial support by the government of Mecklenburg-Vorpommern is gratefully acknowledged.

References Borowitzka, M.A., 1995. Microalgae as sources of pharmaceu-

ticals and other biologically active compounds. J. Appl. Phycol. 7, 3-15. Collins, C.H., Lynes, P.M., Grange, J.M., 1989. Antimicrobial sensitivity and assay tests. In: Collins, C.H., Lynes, P.M. (Eds.), Collins and Lyne's microbial methods. Butterworth, London, pp. 155-168. Falch, B., 1996. Was steckt in Cyanobakterien? Pharmazie in unserer Zeit. 25, 311-321. Kleinkauf, H., von D6hren, H., 1997. Products of Secondary Metabolism. In: Rehm, H.J., Reed, G. (Eds.), Biotechnology, vol. 7. Weinheim, pp. 308-309. Kulik, M.M., 1995. The potential for using cyanobacteria (blue-green algae) and algae in the biological control of plant pathogenic bacteria and fungi. European J. Plant Path. 101, 585-599. Lindl, T., Bauer, J., 1994. Zell- und Gewebekultur. GustavFischer-Verlag, Stuttgart, p. 201. Meffert, M.E., 1971. Cultivation and growth of two planktonic Oscillatoria spec. Mitt. Internat. Verein. Limnol. 19, 189-205. Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M., Stanier, R.Y., 1979. Generic assignments, strain histories and properties of pure cultures of cyanobacteria taxonomy redefinition of the blue-green algae to conform to the bacteriological code. J. Gen. Microbiol. 111, 1-16. Teuscher, E., Lindequist, U., Mundt, S., 1992. Cyanobakterien, Quellen biogener Wirkstoffe. Pharm. Ztg. Wiss. 137, 57-69. Trias, J., Gordon, E.M., 1997. Innovative approaches to novel antibacterial drug discovery. Curr. Opin. Biotech. 8, 757762.




Biodiversity and potentials of marine-derived microorganisms Federica Sponga *, Linda Cavaletti, Ameriga Lazzarini, Angelo Borghi, Ismaela Ciciliato, D. Losi, Flavia Marinelli Biosearch Italia, Via R. Lepetit 34, 21040 Gerenzano (Va), Italy


Abstract

The marine environment is a prolific resource for the isolation of less exploited microorganisms. As a matter of fact, in the sea, untapped habitats exist with unique characteristics. In addition, the potential contribution of marine sources to the discovery of new bioactive molecules was recently recognized. Biosearch Italia possesses a collection of about 40000 microorganisms, isolated from different ecological niches. In the search of new bioactive entities, investigations were expanded to marine habitats including marine sediments and organisms. More then 800 microorganisms have been isolated. About half belong to fungal genera, the others being actinomycetes. The frequency of antibiotic activities produced by these marine strains has been determined. Initial data are encouraging: marine isolates produce antibiotic activities with frequencies comparable to terrestrial ones. These activities probably represent a mixture of novel metabolites and known products previously discovered from terrestrial isolates. Further investigations are ongoing to assess the novelty of these observed microbiological activities. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Biodiversity; Marine microorganisms; Antibiotic activities

1. Introduction

In the process of searching for useful natural products, it is generally agreed that a diverse and less exploited repertoire of microbes is essential to obtain a variety of novel metabolites. Biosearch Italia possesses a collection of more than 40000 microorganisms. In order to enlarge the collection with a novel array of microorgan* Corresponding author. Fax: + 39-2-96474238. E-mail address: [email protected] (F. Sponga)

isms, different ecological areas have been investigated as potential sources of microbial diversity. The marine environment, representing more than two thirds of our planet, is still under-explored and is considered a prolific resource for the isolation of less exploited microorganisms. Data based on 16S r R N A sequencing show that streptomycete strains of marine origin can be representative of peculiar species (Stackebrandt et al., 1991). As a matter of fact, the increasing impact of marine organisms on the discovery of new bioactive molecules has been recognized (Fenical and


66

F. Sponga et al. Journal o/" Biotechnology 70 (1999) 65-69

Jensen, 1994; Liberra and Linderquist, 1995: Bernan et al., 1997; Romero et al., 1997). For all these reasons, our isolation programs have been recently focused on marine sediments (Cuomo et al., 1995) and marine sponges (Hohnk and Ulken, 1979; Imamura et al., 1993; Abrell et al., 1994; Numata et al., 1997). We report here on the isolation of actinomycetes and filamentous fungi from these two sources. The capability of the isolates to produce antibiotic activities is compared with that of the previously observed from terrestrial strains. We have also evaluated the effect of the presence of sea water in the fermentation media on the antibiotic productivity by actinomycetes. 2. Materials and methods

2.1. Isolation from marine sediments A total of 38 marine sediments were collected in different areas (Mediterranean Sea, Red Sea, Atlantic Ocean), at different depths but not exceeding 20 m, and stored at 4~ until use. Soil dilution plate method and a modified soil plate method (Warcup, 1960) were used to isolate fungi. The isolates were cultivated in agar media into Petri dishes at 22~ under dark conditions. A number of media were tested to determine the optimal growth conditions. Petri dishes were inspected daily for up to 1 month for the development of colonies on the agar. The isolated fungi were identified to the level of genus by microscopic morphology. Nomenclature of fungi follows that listed in Ainsworth et al. (1973) and in Kohlmeyer et al. (1991). In the case of actinomycetes, 1 g of sediment was suspended in 10 ml of sterile sea water, vortexed for 1 min and decanted for 30 min. Supernatant (100 ill) was seeded on low and high nutrient agar isolation media, containing sea water. After 2-3 weeks of growth at 28~ single colonies were macro/microscopically observed for a preliminary classification at genus level based on morphology, picked up and transferred on oatmeal agar plates and allowed to grow for a further 2 weeks at 28~

2.2. Isolation from marine sponges Freshly withdrawn samples of sponge (30), belonging to ten different species, from Ligurian coastal areas at 3 - 4 m of depth were used. The sponges were collected and classified up to species by C. Cerrano, Institute of Zoologia, University of Genova (Italy). One sample from an unclassified alga was also included into the isolation program. To isolate filamentous fungi, fresh sponge material was rinsed in marine sterile water, soaked in ethyl alcohol for 5 min and then rinsed in sterile water for 5 rain. Small pieces of each sponge were deposited on two different agar media in Petri dishes: one rich medium and the same diluted 1:1 with marine water. Then the isolation followed the protocol as before. In the case of actinomycetes, sponges were rinsed in sterile sea water, cut into small pieces and treated as described for sediments. Moreover, the same pieces were also deposed directly onto a set of isolation plates. After 2 - 3 weeks of incubation at 28~ single colonies were microscopically observed for morphological classification, picked up, transferred onto oatmeal agar plates, and reincubated until growth.

2.3. Fermentation and sample preparation The fungal strains were inoculated in liquid media routinely used for the production of secondary metabolites in 500 ml Erlenmeyer flasks. Flasks were then incubated on a rotatory shaker at 180 rpm and 22~ for 4 days. Pure cultures of actinomycetes were used to seed two sets of flasks containing two liquid media: a rich medium routinely used for the production of secondary metabolites, and the same medium diluted 1:1 with marine water. Flasks were then incubated on a rotatory shaker at 200 rpm at 28~ for 5 days. Samples were prepared by resin absorption/elution of all fermented broth cultures, with the aim to eliminate impurities and enzymatic activities which could interfere with the biological assays.

F. Sponga et al. ,' Journal of Biotechnology 70 (1999) 65-69 2.4. Assay of antimicrobial activity

r Spongia virgultosa

The following test-strains were used: Enterococcus faecium D399, as a representative of Grampositive bacteria, Escherichia coli clinical isolate, as a Gram-negative representative, and Candida albicans clinical isolate, as a representative of yeasts. Microbial assays were performed in 96-well microtitre plates, in liquid medium inoculated 10 4 cfu m l - 1 with the opportune test-microorganism. Microbial growth was then detected at 620 nm after overnight incubation at 37~ in appropriate humidity conditions. A sample was defined 'active' when able to inhibit the growth of the testmicroorganism in a percentage higher than 80 with respect to controls.

Mieroor~misms mFUNOI r--tSTREPTO NRARE

r Ircinia r Acanthella acuta r Chlatrina cerebrum r Agelas oroides r Petrosia ficiformis r Ascandra falcata r Spongia officinalis r

reniformis r Axinella damicornis 0

3. Results

To evaluate the potential of marine sources for natural product discovery, different parameters were considered: abundance, variety of isolates per source, and their ability to produce diverse antibiotic activities. A total of 395 fungi were isolated either from sediments or from marine organisms (Table 1). The contribution of sediments was about seven strains per sample, while from sponges, nine different fungal strains were isolated per sample. When compared with the number of the isolates from soil samples, both marine sources (marine sediments and sponges) appear rich in fungal strains. Among these marine fungi, about 75% sporulated on the media currently used. These sporulating fungi could be assigned to the followTable 1 Number of sources and related isolates Source

Strain number

Type

N~

Fungi

Sediment Organism

38a 31

a

Rare

Stm

130

17

265

284

22 129

Only 18 were used to isolate fungi.

67

20

40 60 80 100 120 number of isolates

Fig. 1. Distribution of microbial population among sponge species. Value in brackets represents the number of individuals used. ing genera: Acremonium, Alternaria, Aspergillus,

Cephalosporium, Chaetomium, Cladosporium, Geotricum, Fusarium, Gliomastix, Humicola, Paecilomyces, Penicillium, Pestalotia, Phoma, Plectosphaerella, Scopulariopsis, Stachybotrys, Trichoderma. In the case of actinomycetes isolation, marine organisms yielded an average of ten morphologically different strains per single source: 31% being streptomycetes, and 69% belonging to rare genera mainly represented by micromonosporas, followed by nocardioforms and actinomaduras/microtetrasporas. Marine sediments yielded an average of one isolate per source: 50% streptomycetes and 50% rare actinos as shown in Table 1. Even among these rare actinos, Micromonospora was the more represented genus, in accordance with published studies (Jensen et al., 1991). It appeared of interest to investigate the distribution of the microbial population in the different species of sponges. The results are reported in Fig. 1.

68

F. Sponga et al. ,,'Journal of Biotechnology 70 (1999) 65-69

All strains were fermented in standard production liquid media and assayed for antimicrobial activity against three representative microorganisms: a G + , a G - and one yeast. The percentages of active strains are reported in Table 2. Marine actinomycetes have also been fermented and assayed for antimicrobial activity in the production medium diluted 1:1 with sea water. In most cases the antimicrobial activities were detected in both conditions, however four strains produced activities only when fermented in the presence of sea water.

4. Conclusion Most of the fungi isolated from these marine samples belong to genera usually considered soil inhabitants. Among the actinomycetes, a good variety of known genera was represented. A general conclusion is that marine organisms yielded a number of isolates per source that is comparable with that obtained from soils. On the other side, the number of isolates obtained from marine sediments is lower, particularly in the case of actinomycetes. Therefore, in terms of abundance of both types of microbial communities, sponges could be considered a more suitable source than sediments. The abundance of microorganisms isolated is different in the different species of sponges. Further investigation is needed to determine if the microbial population is specifically associated to each species. The frequency of antimicrobial activities displayed by marine fungi and actinomycetes is similar to that previously observed in terrestrial Table 2 Percentage of strains producing antimicrobial activities in our assays % Active strains

Fungi Rare Stm

Test-microorganisms E. faecium

E. coli

C. albicans

1.2 3.5 6

0.6 0 1.3

2.2 0.3 0.7

strains during the course of our screening programs. The ability shown by fungal isolates to germinate and produce secondary metabolites under the conditions generally used for the terrestrial ones make it difficult to define them as marine fungi. The presence of a few antibacterial activities expressed only when some actinomycete strains are incubated in the fermentation medium diluted with sea water, suggests that these isolates could be defined as 'adapted' to the marine environment. Structural characterization of the thus far discovered antibiotic activities, will provide additional information on the potential of marine environments as sources of novel metabolites.

Acknowledgements The authors wish to thank Maurizio Rogina for supplying marine sediments, and also the staff of both the Institute of Zoologia and the Institute of Biochimica of the University of Genova, Italy for supplying marine sponges and for their valuable support on the classification and evaluation of these.

References Abrell, L.M., Cheng, X.C., Crews, P., 1994. New nectriapyrones by salt water culture of a fungus separated from an Indo-Pacific sponge. Tetrahedron Lett. 35 (49), 9159-9160. Ainsworth, G.C., Sparrow, F.K., Sussmann, A.S., 1973. The Fungi, an Advanced Treatise, Academic, New York, San Francisco, London, vols. IVA and IVB. Bernan, V.S., Greenstein, M., Maiese, W.M., 1997. Marine microorganisms as a source of new natural products. Adv. Appl. Microbiol. 43, 57-90. Cuomo, V., Palomba, I., Perretti, A., Guerriero, A., D'Ambrosio, M., Pietra, F., 1995. Antimicrobial activities from marine fungi. J. Mar. Biotech. 2, 199-204. Fenical, W., Jensen, P.R., 1994. Strategies for the discovery of secondary metabolites from marine bacteria: ecological perspective. Annu. Rev. Microbiol. 48, 559-584. Hohnk, W., Ulken, A., 1979. Pilze aus marinen schwammen Veroff. Inst. Meeresforsch. Bremerh. 17, 199-204. Imamura, N., Nishijma, M., Adachi, K., Sano, H., 1993. Novel antimycin antibiotics, urauchimycins A and B, produced by marine actinomycete. J. Antibiot. 46 (2), 241246.

F. Sponga et al./Journal of Biotechnology 70 (1999) 65-69 Jensen, P.R., Dwight, R., Fenical, W., 1991. Distribution of Actinomycetes in near-shore tropical marine sediments. Appl. Environ. Microbiol. 57 (4), 1102-1108. Kohlenmeyer et al., 1991. Illustrated key to the filamentous higher marine fungi. Bot. Mar. 34, 1-61. Liberra, K., Linderquist, U., 1995. Marine fungi-a prolific resource of biologically active natural products? Pharmazie 50, 583- 588. Numata, A., Amagata, T., Minoura, K., Ito, T., 1997. Gymnnastatins, novel cytotoxic metabolites produced by a fungal strain from a sponge. Tetrahedron Lett. 32, 5675-5678.

69

R omero, F., Espliego, F., Perez Baz, J., Garcia de Quesada, T., Gravalos, D., De la Calle, F., Fernandez-Puentes, J.L., 1997. Thiocoraline, a new despsipeptide with antitumor activity produced by a marine Micromonospora. J. Antibiot. 50, 734-737. Stackebrandt, E., Witt, D., Kemmerling, C., Kroppenstedt, R., Liesack, W., 1991. Designation of streptomycete 16S and 23S rRNA-based target regions for oligonucleotide probes. Appl. Environ. Microbiol. 57, 1468-1477. Warcup, J.H., 1960. Methods for isolation and estimation of activity of fungi in soil. In: Ecology of Soil Fungi. Liverpool University Press, pp. 3-21.


Biotecbn,o,logy ELSEVIER


Production and particle characterization of the frustules of Cyclotella cryptica in comparison with siliceous earth Zsuzsa Cs6g6r *, Dyna Melgar, Karsten Schmidt, Clemens Posten 1 Institut fiir Mechanische Verfahrenstechnik und Mechanik, Universitiit Karlsruhe, D-76128 Karlsruhe, Germany Received 14 October 1998; received in revised form 3 December 1998; accepted 22 December 1998

Abstract Diatoms are p h o t o a u t o t r o p h i c micro-organisms that use inorganic carbon sources and light in photosynthesis.

Diatom frustules were characterized in terms of particle techniques and compared with siliceous earth, i.e. depositions of diatoms that have wide technical applications. To obtain enough biomass for frustule characterization Cyclotella cryptica has been cultivated in a 15 1 photobioreactor under controlled conditions. Native diatom frustules are characterized by a 1.5-fold lower density and 80-fold higher specific surface than siliceous earth. Therefore, native diatom frustules provides a material with novel properties which might be interesting for special technical application. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Diatoms; Cultivation; Photobioreactor; Frustules; Siliceous earth

I. Introduction

Inorganic materials with a complex pattern on a microscopic scale are of increasing research interest (Mann and Ozin, 1996). Possible technical applications could be the design of new types of catalyst supports, specific filter aid or chromatographic separation material, or biomedical implants with macroporosity. Siliceous earth is an inorganic material widely used in industrial scale because of its high poros* Corresponding author. Tel.: + 49-721-6086171; fax: + 49721-693965. E-mail addresses: [email protected] (Z. Cs6g6r), [email protected] (C. Posten) Tel.: + 49-721-6082410; fax: + 49-721-693965.

ity (e.g. as filtering aid) and its high surface area (e.g. as HPLC column filling material). Siliceous earth is a fossil deposition of diatoms. Diatoms are eukaryotic microalgae. The cells are enclosed in amorphous silica structures (frustules). The cells may adhere through organic polymers forming chains of different length. The chloroplasts are of brownish color because of the pigment fucoxanthin. Although siliceous earth consists of diatom frustules, the amorphous native particles may differ in various respects. Properties such as volumetric density, particle size, surface area, porosity may have changed after deposition. Particularly transport behavior, catalytic activity, and separation efficiency are influenced strongly by particle

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 0 6 0 - 7

72

Z. Cs6g6r et al./Journal of Biotechnology 70 (1999) 71-75

size and surface area. Thus, native diatom frustules may prove to be an interesting material for high quality application. The aim of the work presented in this paper was to cultivate diatoms, characterize their frustules from the point of view of particle techniques and to compare native diatom frustules with siliceous earth. Such a comparison has not been reported in the literature, yet.


2.1. Microorgan&ms and culture conditions Cyclotella cryptica Reimann, Lewin and Guillard 1070 was obtained from the G6ttingen Algal Culture Collection (Schl6sser, 1994). The algae were originally isolated from the highly polluted river Werra in Germany. The strain is free from algal or bacterial contamination. Preoculation was performed in 300 ml Erlenmeyer flasks containing 100 ml medium under aseptic conditions. The flasks were inoculated with 2 ml pregrown medium, stirred at 100 rpm on a rotary shaker and illuminated with lamps providing a photon flux density of 15 ~tE m - 2 s-1 at the surface of the cultures. They were illuminated continuously or under a light/dark regime of 14/10 h, respectively.

The basis for this medium optimization were the flesh water and sea water media for diatom cultivation given in the literature (Stein, 1973; Schl6sser, 1994). According to the high salt pollution in the river where the species has been isolated from, the optimization lead to a brackish water medium. It was found that in Erlenmeyer flasks best growth could be obtained with continuous illumination in media containing medium salt concentration ( ~ 18 g 1-I NaC1). Based on these findings, the medium described below was applied to cultivate C. cryptica in a photobioreactor. The macro-element solution contained 17.95 g 1-~ NaC1; 4.90 g 1-~ MgC12 6 H20; 1.19 g l Na2HPO4 2 H20; 0.87 g 1-1 KNO3; 0.109 g 1-1 Ca(NO3)2 4 H20; 0.093 g 1-~ NazSiO3 9 H20; 0.013 g 1- ~ NaHCO3; 0.001 g 1- i KHzPO4. This medium was supplemented with 1% micro-element solution containing 1.21 g 1-~ Fe-citrate; 0.29 g 1-1 H3BO3; 0.277 g 1-~ MnSO4 4 H/O; 0.028 g 1- 1 Z n S O 4 7 H20; 0.0127 g 1-~ C o ( N O 3 ) 6 6 H20; 0.010 g 1-~ CuSO4 5 H20; 0.0063 g 1NazMoO4 2 H20 and 1 ml 1-~ vitamin solution containing 0.05 g 1-~ vitamin B~, and 0.0005 g 1-~ vitamin BI2. All chemicals used were of analytical quality. The pH was adjusted to 7.5. Media were sterilized at 121~ for 20 min, microelement solution was sterilized separately, vitamin solution was filtered (0.2 lam) and added to the medium at room temperature.

2.3. Cultivation in photobioreactor 2.2. Growth medium In order to cultivate enough cells for particle characterization a medium had to be found that allows good growth of the diatoms. Preferably it had to be a defined medium to obtain reproducible results. The medium given in the literature for the cultivation of the C. cryptica species used (Schl6sser, 1994) does not support growth to a high cell density. Moreover it is a complex medium containing 1% soil extract. Therefore, a medium optimization study had been carried out prior to these experiments in to find a defined medium that allows good growth (Z. Cs6g6r et al., unpublished results).

A 15 1 stirred tank reactor (all glass) was illuminated continuously from outside with halogen lamps to give a photon flux density of 92 gE m - 2 s - l at the inner surface of the glass reactor. Ten 1 sterile defined medium was inoculated with 400 ml medium containing 5 x 109 cells ml-1 C. cryptica pregrown in Erlenmeyer flasks. The medium was bubbled with sterile air/CO2 mixture (1% CO2) at 1 1 m i n - l pH, pO2, temperature, and the concentrations of 02 and CO2 in the exhaust gas were measured on-line, the pH was maintained at 7.5 during the cultivation. Cell number and optical density (see below) were determined daily.

z. Cs6g6r et al. / Journal of Biotechnology 70 (1999) 71- 75 2.4. Determination o f growth and growth rate

The optical density was determined at 430 nm in a spectrophotometer (Unicam UV2). The cell concentration was determined in a Neubauer chamber to obtain the growth rate/z according to Cx(t) = Cx,o e ~' t, with cx determining the cell concentration and Cx.o the cell concentration at the beginning of the cultivation. The cultivation data from the exponential phase of growth were taken into account for the growth rate determination. The optical density correlated well with the cell concentration (r2= 0.98). The dry weight was determined according to Zhu and Lee (1997). Cell concentration, optical density, and dry weight were analyzed in duplicate, mean values are shown. 2.5. Preparation o f the frustules

For the determination of the specific surface and density, C. cryptica cultured in the 15 1 bioreactor was harvested and washed in demin, water, and separated by centrifugation (5 rain in a centrifuge BGH Hermle ZK 630 at 4~ and 2000 • g). The cells were broken with the protease mixture Novozym FM 2.0 L: 0.1 mg protease per 7.5 mg algal suspension was applied at 55~ for 5 h, then washed three times as described above. Chlorophylls and other organic material were removed through acetone extraction until the preparation lost colour (no extinction in the visible range of light). Finally the frustules were dried by lyophilization a t - 3 0 ~ and 0.375 mbar (Fa. Christ alpha 1-4), mechanically loosened with a spoon and stored in an exsiccator. For scanning electron microscopy the same procedure was performed. Subsequently the dried frustules were sputtered with palladium and platinum and scanned in a scanning electron microscope (Hitachi S 4500). 2.6. Characterization o f the frustules and siliceous earth

The diatom frustules cultivated as described above were characterized in terms of particle techniques and compared to siliceous earth (Merck

73

1.10601). The data shown are mean values of one sample analyzed in duplicate. A gaspycnometer (Micromeritics NP1305) was used for the measurements with helium as measuring gas. A gaspycnometer accomplishes the measurement of skeletal volumes by observing the reduction of gas capacity in the sample chamber caused by the presence of a sample. Since the measuring gas penetrates even the smallest pores and surface irregularities, the volume obtained permits computation of the ultimate theoretical density of the solid comprising the sample if there are no enclosed pores. For the determination of the size of the particles a monochromatic laser light diffractiometer was used (Sympatec Helos H9236). Ultrasound was applied for 2 min to achieve good dispersion before the measurements (Bransonic B12, 150 W, 48 kHz). The specific surface was determined according to the Brunauer-Emmett-Teller method (BET) which considers the adsorption of gas to the surface of the sample. The adsorption isotherm can be calculated by measuring one point only and assuming the ordinate origin as the second (one-point BET) or by measuring several points and calculating the isotherm (multiple-point BET). The gas sorption analyzer Nova 1200 (Quantachrome Co.) was used with nitrogen as measuring gas. The specific surface was determined with both one-point and multiple-point BET.

3. Results and discussion 3. I. Cultivation in photobioreactor

The cultivation of C. cryptica in the bioreactor resulted in a good growth (see Fig. 1). The growth rate determined in the exponential growth phase between day 4 and 9 was 0.55 d-1, this is about twice the best growth rates determined in Erlenmeyer flasks (data not shown). The higher growth rate is due to the higher concentration of CO2 (1% added in the reactor vs. air) and the better illumination in the reactor compared to cultures grown in Erlenmeyer flasks (92 laE m - 2 S--1 and 15 laE

Z. Cs6g6r et al./ Journal of Biotechnology 70 (1999) 71-75

74

m - 2 S - 1 , respectively). After 10 days the culture did not grow exponentially any more because it became light limited. The cells were harvested at a concentration of 9.3 • 107 cells ml-1. The cells were cleaned as described above and the frustules were taken for particle characterization.

3.2. Characterization of the frustules C. cryptica is a unicellular centric diatom (Fig. 2). The cells do not adhere like many other diatoms thus allowing detailed analysis of the particles. The cell wall is composed of amorphous, hydrated silica and organic components (Will6n, 1991). The frustules obtained in the 15 1 cultivation scale were purified from the organic components and characterized in terms of particle techniques in comparison with siliceous earth. The particle size distribution of both are given in Fig. 3. C. cryptica frustules are much smaller in

size than siliceous earth. The frustules are not equal in size as can also be seen in the picture in Fig. 2. The relatively wide distribution of siliceous earth is due to the fact that it contains a variety of different species. The density of both native diatom particles and siliceous earth were analyzed in a gaspycnometer. The results are given in Table 1. C. cryptica is characterized by a 1.5-fold lower density in comparison with siliceous earth. The latter consists of many different diatom species with different amounts of hydrate: SiO2 may contribute between 10 and 70% in native frustules (Will6n, 1991). The discrepancy may also be due to the fact that the native frustules are of amorphous structure whereas in siliceous earth a compression to a crystalline structure has taken place. The specific surface of both was determined, the results are given in Table 1. The one-point BET and the multiple-point BET give very similar re-

1,40E+10

1,20E+10

E L

9 1,00E+10 JD

E C ~ 8.O0E+09 U c o

m 6,00E+09 L c

o u C

0 4,00E+09 U 0

r

2,00E+09

O,OOE+O0

0

1

2

3

4

Cultivation

5

6

7

8

9

10

time [days]

Fig. 1. Growth of C. cryptica in a stirred 15 1 photo-bioreactor, illuminated from outside with halogen lamps to give an average photon flux densitiy of 92 laE m - 2 s - ] , bubbled with a 1% CO2/air mixture. Points are measured cell concentrations, line gives exponential approximation of growth between days 3 and 9.

Z. CsOg6r et al./Journal of Biotechnology 70 (1999) 71-75

75

100

....

90 t 3

: -=-~r -

~~~'~A='

70

~ 60

,0a0

Y -=

4-

20 [ 0

_~

J

i

i

..A&

I i.-:

cyc,o,o, cryptica [ "" ~''siliceous

I

. . . . . .

1

10 100 particle size x [IJm]

1000

Fig. 3. Mass particle size distribution of C. cryptica and siliceous earth after 2 min dispersion with ultrasound. Fig. 2. C. cryptica, scanning electron microscope picture; the dried frustules were sputtered with palladium and platinum (bar length is 5 ~tm).

suits. The specific surface of C. cryptica frustules is much higher than that of siliceous earth (80 fold). This is both due to the smaller particle size (about 10 and 100 lam, respectively) and to the higher porosity of the native fluso tules. As the comparison shows, the native frustules of C. cryptica clearly exceed siliceous earth in respect to surface and porosity, properties that have led to the wide use of the latter in many areas. Furthermore they are more uniform in shape and size because of the controlled cultivation conditions of unialgal and axenic cultures. Although the cultivation of diatoms is expensive, it may turn out that the frustules are of advantage for special application.

Table 1 Characterization of C. cryptica and siliceous earth in terms of particle techniques

Density (g cm- 3) Mean particle size (lam) Spec. surface (m 2 g-Z) (one-point BET) Spec. surface (m 2 g-1) (multiple-point BET)

C. cryptica

Siliceous earth

1.46 + 0.01 I0.1 26.21 + 0.48

2.22 +__0.02 92.5 0.35 +0.003

27.70 + 0.49

0.35 ___0.004

References Mann, S., Ozin, G.A., 1996. Synthesis of inorganic materials with complex form. Nature 382, 313-318. Schl6sser, U.G., 1994. SAG--Sammlung yon Algenkulturen at the University of G6ttingen Catalogue of Strains 1994. Bot. Acta 107, 113-186. Stein, J.R., 1973. Handbook of phycological methods. Cambridge University Press. Will6n, E., 1991. Planctonic diatoms-an ecological review. Alg. Stud. 62, 69-106. Zhu, C.J., Lee, Y.K., 1997. Determination of biomass dry weight of marine microalgae. J. Appl. Phycol. 9, 189-194.


|111 ii

IOUINAL

i|

oF



Biotechnological potential of North Sea salt marsh plants a review of traditional knowledge Gerd Liebezeit a,, Thorsten D. Ktinnemann b, Gunnar Gad b a Forschungszentrum Terramare, Schleusenstrafle 1, D-26382 Wilhelmshaven, Germany b FB Biologic, Universitdt Oldenburg, Postfach 2503, D-26111 Oldenburg, Germany Received 13 October 1998; received in revised form 1 December 1998; accepted 22 December 1998

Abstract

Extreme environments are characterised by wide variations in physical factors. Tidal flats and their adjacent salt marshes along the North Sea coast experience this type of variability on a daily, seasonal and annual basis. Plants and animals living in these extreme environments have to adapt to these, at times rapid, changes. This can be done by developing specific physiological responses which often encompass the synthesis of unusual chemicals. A number of salt marsh plants have traditionally been used for medical, nutritional or even industrial purposes. Here, examples are presented for these plants. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Extreme environments; Salt marsh plants; Semiterrestrial

1. Introduction Organisms thriving in extreme habitats have developed a number of mechanisms to cope with their physical environment. These may include a.o. the production of antifreeze compounds or heatshock proteins in fish and other organisms (Sicherl and Yang, 1995; Berger et al., 1996), specific enzyme systems capable of operating at temperatures as high as 113 ~ (Bl6chl et al., 1997) or at deep sea pressures (Huber et al., 1989; Gross and Jaenicke, 1994). Therefore, considerable biotechnological interest in these environCorresponding author. E-mail address: [email protected] (G. Liebezeit) *

ments has recently been expressed (Jensen, 1993; Pennisi, 1997; Gross, 1998). Extreme environments are usually defined as being extreme in one or several physical parameters such as temperature (e.g. hydrothermal vents, ice-covered regions), salinity (hypersaline lagoons, Dead Sea) or pressure (deep sea). However, in these systems the external variables remain relatively stable over time and are extreme only by human standards. On the other hand, marine systems with a high variability in temperature, salinity or radiation can also be found. These include a.o. salt marshes, tidal flats and mangrove ecosystems. Organisms thriving in saline environments have been found to produce a large number of sec-

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All fights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 0 6 1 - 9

78

G. Liebezeit et al./Journal of Biotechnology 70 (1999) 77-84

ondary natural products that might be of potential interest for pharmacological or industrial applications (Hay, 1996). Marine neurotoxins, for example, may be used to study ion channels and receptors (Myers et al. 1993). Bryostatins, isolated from bryozooans, are currently being investigated as powerful anti-leukemia agents (Olson, 1996). Salt marshes and tidal fiats experience regular (e.g. tides) and irregular (e.g. storms) external influences. Temperatures may vary from below -20~ in severe winters to > 40~ in extreme summers, while the salt regime may change from freshwater conditions after heavy rains to hypersaline after prolonged evaporation periods (Reineck and Flemming, 1990). To be able to deal with these variations in environmental conditions organisms inhabiting these ecosystems have developed a number of adaptation mechanisms. These include mechanisms for a.o. salt excretion or storage, osmoregulation (i.e. control of cellular ion concentrations) or desiccation protection. In this context it should be mentioned that out of 16 mangrove species tested for antiviral activity, 10 were found to contain active compounds (Padamakumar and Ayyakkannu, 1994), while of the marine macroalgae investigated only 30% behaved similarly. The documented production of saponins in seeds (Glenn et al., 1991) or other toxic compounds by salt marsh plants (Cooper, 1984) can be regarded as a mechanism to avoid or at least to minimise predation pressure. This suggests that semiterrestrial plants may be a particularly wealthy source for unusual chemical compounds. Marine biotechnology largely depends on the availability of information on the natural products chemistry of marine organisms (Rouhi, 1995; Colwell, 1983) which can be obtained in various ways. One of these is the installation of screening programmes, whereby organisms of interest are systematically investigated for usually a restricted number of biological activities. The latter may encompass antibacterial, antiviral or other compounds of pharmacological interest. A second approach is to use traditional knowledge as assembled in e.g. homeopathy text books or passed on by folklore. In these cases, generally no details on the structures of molecules of interest

will be made available if they are known at all. Rather, these sources can be used to obtain information on the biological activities displayed by different plants or parts thereof. In this review we will concentrate on this second approach, taking a closer look at selected plants from North Sea salt marshes, chiefly glasswort (Salicornia spp.), sea lavender (Limonium vulgare), arrowgrass (Triglochin maritima), sea wormwood (Artemisia maritima) and sea plantain (Plantago maritima). Other plants which have also been used traditionally will be mentioned briefly.

2. Methods

The information provided here is based on an intense literature search including World Wide Web sources, notably wwwl. In addition, a number of older, out-of-print text books on botany have been used, especially Hegi (1909, 1926). Furthermore, for information on homeopathic usage of plants the publications of Anton (no publication date); Alsamer (1901), Flamm and Kr6ber (1936) as well as Beckmann and Beckmann (1997) have been consulted.

3. Presentation of data

Information on several salt marsh plants and their traditional usage is summarised in Table 1.

3.1. Salicornia europaea Salicornia europaea (glasswort) (Fig. 1), an annual plant of the family Chenopodiaceae, is the first pioneer plant of saltmarshes colonising the border between tidal fiats and the salt marsh. It is dependent on the presence of salt and accumulates large amounts of salts including sodium, potassium, iodide and bromide compounds during growth. This is the main reason why its ash has been used in glass making (hence its popular name) as the high salt content of the ash lowers the melting point of glass. As it also contains sodium carbonate, it was used traditionally in France for washing purposes.


79

Table 1 Salt marsh plants with traditional usage Plant species

Trivial name

Compound(s)

Traditional usage

Salicornia sp. L. vulgare T. maritima

Glasswort Sea lavender Seaside arrowgrass Sea wormwood Sea plantain

Inorganic salts Triglochinin taxiphyllin a.o.

Ash for glass production and washing agents Antidiarrheticum Cabbage substitute

Prickly glasswort

Saponins

Scurvy grass

Ascorbic acid

A. maritima P. maritima Aster tripolium Salsola kali Cochlearia sp. Odontites rubra ssp. litoralis

Antihelminthicum; insecticide (moths, lice, fleas) Spinach substitute; soda (ancient Egypt) Vegetable (S Holland, China) Deadly to cows in large quantities; fresh plant juice as diureticum; soda (ancient Egypt) Scurvy treatment Cures toothache when cooked in wine

The plant is particularly conspicuous in autumn, when it displays a bright red discolouration. Although the pigment signature of S. europaea is relatively simple (Fig. 2), no attempts have been made so far to structurally characterise the compounds in question. HPLC analysis of aqueous acetone extracts indicates one or more carotenoid pigments to be possibly responsible for the red discolouration. These elute close to the column dead volume and are thus highly polar in nature. The precursor compounds eluting at 13, 15.5 and 23 min have so far not been identified. While the on line-absorption spectra (Fig. 2) indicate that these are carotenoids, a positive identification is not possible with these data alone. Carotenoids have biotechnological potential (Ausich, 1997). Incidentally, a yet unidentified pigment from Hasleria ostrearia, a benthic diatom, is employed in its native form for oyster colouring (Groth-Nard et al., 1994). Cooper and Johnson (1984) reported that among eight salt marsh plants tested S. europaea was the only one not showing limiting effects during growth in media containing up to 10 mM manganese sulphate. As this is far above the natural concentrations encountered in salt marsh soils this capability might point to the presence of a powerful complexing agent in Salicornia sp. In recent years, another Salicornia species, S. bigelovii, has received considerable interest as a cultivar in salt water irrigated agriculture (Glenn et al., 1991, 1998). Its seeds, besides containing

unidentified saponins, were found to be rich in proteins and lipid material, notably fatty acids. Weete et al. (1970) have suggested that the variability of the hydrocarbon and fatty acid composition of S. bigelovii is governed by external factors, i.e. differences in environmental ion concentrations.

Fig. 1. Salicornia europaea.


80 0.25 0.16

Z w I-

0.20

oI=
1.25 > 1.25

ND r

ND ~

Ascidiella aspersa

0.074

COLO32o GLC 4

0.96 0.80

> 1.25 > 1.25

0.45 0.23

Styela clara

0.166

COLO320 GLC 4

0.89 0.25

> 1.25 > 1.25

0.46 0.33

M. manhattensis

0.063

COLO32o GLC4

0.88 0.74

> 1.25 > 1.25

0.08 0.02

A. glabrum

0.152

COLO32 o GLC 4

0.19 0.15

> 1.25 > 1.25

0.08 0.07

D. lahillei

0.161

COLO32 o GLC4

0.25 0.20

> 1.25 > 1.25

0.23 0.19

a IC5o, concentration (in lag ml -~) giving 50% growth inhibition of the COLO32 o or GLC 4 cell line in the MTT-assay. As a reference the cytostatic cisplatin (Aldrich, Milwaukee, USA) gave the following IC5o values; COLO32 o. 0.8 lag m l - l (2.7 laM); GLCa: 0.3 lag m l - l (1.0 laM). b DW/FW, ratio of dry weight and fresh weight. r no detectable effect from the extract in the MTT-assay.

2.4. Bioguided isolation of active compounds The species with most active extracts were chosen (M. manhattensis, A. glabrum and D. lahillei ) for further fractionation over a silica 60 column using different solvents. The fractions were analyzed qualitatively by thin layer chromatography (TLC). For TLC, silica 60-F-254 (Merck) plates were used, with CHC13:MeOH (9:1) as the eluents. The spots were detected under UV light (254 and 366 nm) followed by spraying with anisaldehyde. Fractions with an identical TLC pattern were combined and tested for their cytotoxicity. For further purification this procedure was repeated with flash chromatography using a prepacked column (Merck) at a low pressure (4 atm).

3. Results

3.1. Screening for cytotoxic compounds The crude extracts of all species tested showed a

moderate cytotoxic acticity (Table 2). In comparison to known cytostatic cisplatin, the most active extracts in Table 2 were at least 103 times less cytotoxic. The DW/FW (Table 2) ratio shows large differences in the water content of the species, indicating that the concentrations of bioactive compounds in living tissue are not only low, but also highly variable. Based on their low IC5o values (Table 2) only the CHC13 extracts of M. manhattensis, A. glabrum and D. lahillei were used for the bioguided isolation of cytotoxic compounds.

3.2. Isolation of bioactive compounds A. glabrum yielded a very active fraction (IC5o = 5 lag ml-~), but TLC demonstrated that amounts of compounds were too small for further purification. M. manhattensis also gave a very active crude extract (IC50 = 8 lag ml-~), but only fractions with a moderate cytotoxicity were obtained by column chromatography.

88

A. Koulman et al./Journal of Biotechnology 70 (1999) 85-88

D. lahillei yielded an active, almost pure compound. To obtain this compound in sufficient quantities, the isolation was performed with 60 g freeze-dried material and finally 2 mg of an active compound was isolated. The ICso of the compound for COLO320 (33 ~tg ml-1) was lower than for GLC 4 (49 ~tg ml-1), which indicates a specific activity. Spectroscopic analysis (Infra red (IR), proton and carbon nuclear magnetic resonance (1H-NMR and 13C-NMR), ion electron mass spectrometry (IEMS) and chemical ionization mass spectrometry (CIMS)) of the compound however, showed impurities to be present. ElMS gave the following spectrum M:412 (26), 398 (13), 397 (15), 396 (43), 384 (100), 109 (14), 107 (15), 105 (12). The unknown compound is presumably a C25 compound (sesterterpene) giving large stable ions in ELMS. New material of D. lahillei collected (on the 19th of October 1997 at the same place) for a large scale isolation did not contain the active compound.

4. Conclusion The screening of the different ascidian species for the presence of cytotoxic compounds yielded fractions with an interesting activity of A. glabrum and D. lahillei, proving that in addition to tropical species, cold water ascidians might also form a source for new cytotoxic compounds. In these ascidians, the active compounds were present at very low concentrations. Large scale isolation will be necessary to obtain sufficient quantities of pure compounds for structure elucidation and further pharmacological evaluation. The already mentioned unknown compound from D. lahillei is of special interest because of its more specific cytotoxicity. However, there are probably seasonal variations in the secondary metabolite composition. The generally moderate activity of the majority of the extracts is remarkable because none of the presented species seem to suffer from predation. In addition, there is no relation between the cytotoxicity and the FW/DW ratio, which implies that

large differences will occur in the concentrations of bioactive compounds in living tissue of the different species. It is therefore likely that other factors, besides bioactive compounds, might play an additional role in the defense mechanism of the species investigated.

Acknowledgements The authors would like to thank Jaap Begeman and Dr Karen E. Voskamp for their help with the collection of the animal material. Dr J.K. Herreman, W.H. Kruizinga and Dr A. Kiewiet are gratefully acknowledged for the NMR and MS spectra.

References Beekman, A.C., Woerdenbag, H.J., Van Uden, W., Pras, N., Konings, A.W.T., Wikstr6m, H.V., Schmidt, T.J., 1997. Structure-cytotoxicity relationships of some helanolidetype sesquiterpene lactones. J. Nat. Prod. 60, 252-257. Buizer, D.A.G., 1983. De Nederlandse zakpijpen (manteldieren) en mantelvisjes. Wet. Med. K.N.N.V. 158. Nienhuis, P.H., Smaal, A.C., 1994. The Oosterschelde estuary, a case-study of a changing ecosystem: an introduction. Hydrobiology 282-283, 1-14. Knight-Jones, E.W., Ryland, J.S., 1995. Acorn-worms and sea squirts (phyla Hemichordata and Urochordata). In: Hayward, P.J., Ryland, J.S. (Eds.), Handbook of the Marine Fauna of North-West Europe. Oxford University Press, Oxford. Koulman, A., Proksch, P., Ebel, R., et al., 1996. Cytotoxicity and mode of action of aeroplysinin-1 and a related dienone from the sponge Aplysina aerophoba. J. Nat. Prod. 59, 591-594. Leewis, R.A., Waardenburg, H.W., Van Der Tol, M.W.M., 1994. Biomass and standing stock on sublittoral hard substrates in the Oosterschelde estuary (SW Netherlands). Hydrobiology 282, 397-412. Rinehart, K.A., Kishore, V., Bible, K.C., Sakai, S., Sullins, D.W., Kai-Ming, L., 1988. Didemnins and tunichlorin: novel natural products from the marine tunicate Trididirnnurn solidum. J. Nat. Prod. 51, 1-21. Scudiero, D.A., Shoemaker, R.H., Paull, K.D., et al., 1988. Evaluation of soluble tetrazolium/formazan assay for cell growth and drug sensitivity in culture using human and other tumor cell lines. Cancer Res. 48, 4827-4833. Watters, D.J., Van de Brink, A.U, 1993. Toxins from ascidians. Toxicon 31 (11), 1349-1372.

JOURNAL

OF



Biotechnological hydrogen production" research for efficient light energy conversion Jun Miyake a,b,, Masato Miyake b, Yasuo Asada b National Institute for Advanced Interdisciplinary Research, Tsukuba, Ibaraki 305-8562, Japan b National Institute of Bioseience and Human Technology, Tsukuba, lbaraki 305-8566, Japan


Abstract

The study of biological hydrogen production by using photosynthetic bacteria and cyanobacteria is described based on a national R&D project in Japan. We describe here the subjects examined in the research for photosynthetic bacteria: analysis of the relationship between the penetration of light to photobioreactor and hydrogen production, genetic engineering of photosynthetic bacteria to control the pigment content for making the light penetration easy. Examples of bench-scale reactors are shown. Genetic manipulation of hydrogenase to enhance the protein expression is also studied. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Hydrogen; Cyanobacteria; Photosynthetic bacteria; Renewable energy; Solar energy

I. Introduction

Ever aware of limitations to the amount of resources and hazardous effects to the environment of fossil energy, development of renewable energy is an urgent task. Solar energy is the most abundant amongst various renewable energy sources. Its radiation provides the biggest flow of energy on the earth, ~ 5.7 x 1024 J year-1. This is 10000 times more than the total energy consumed by human beings. However, the efficiency of converting solar energy into any of a number of possible renewable * Corresponding author. Tel." + 81-298-54-2558; fax: + 81298-54-2558. E-mail address: [email protected] (J. Miyake)

energy carriers is very limited, representing inefficient use of the available solar energy. In general, so-called renewable sources have a significant drawback in that we have to accumulate solar energy before supplying it to industrial applications. Collecting solar energy with photovoltaic cells or thermal solar collectors yields rather higher conversion efficiency, but because solar radiation at the surface of the Earth is very low power at 1 kW m -2, we need a huge area to collect a sufficient amount of energy. This defect prohibits a commercial application. Biological energy conversion system as photosynthesis might not to be so efficient in the solar energy utilization. However, it is a remarkable process in the viewpoint of energy accumulation. Plants and photosynthetic microorganisms grow


90

J. Miyake et al./Journal of Biotechnology 70 (1999) 89-101

by themselves and collect energy. Such spontaneous energy accumulation system is not obtainable in man-made industrial technologies. There are many types of energy carriers as hydrogen, methane, alcohol, and all other biomass, which could be produced biologically from renewable energy sources. Among them, hydrogen is a powerful energy source that does not produce carbon dioxide (Fig. 1). In this article, we describe briefly an approach for creating a safe and efficient energy production system. Although in the laboratory rather high efficiency of light to hydrogen conversion has been recorded, especially by photosynthetic bacteria, we have to improve the organisms and the reactor system further to realize the potentially high activity under applicational conditions. Not only the biological improvements, but also the engineering aspects are studied here. We should like to describe the basis of hydrogen biological production, genetic engineering and the reactor design

and field tests.

1.1. Thermodynamic advantage of biological energy conversion (Miyake, 1998a) In the case of organic substances, as waste water they are dissolved and diluted in the water, i.e. in a high entropy state, it is impossible to obtain their combustion enthalpy unless extracting these substances. The diluted carbohydrates contained in the wastewater of a tofu factory have a combustion enthalpy of 150 kJ 1-~ Evaporation of 1 1 of water requires about 2200 kJ 1-~, yet the available energy of the substance is far smaller. It is very difficult to obtain energy from dilute solutions by such a mechanical way. On the other hand, bacteria can convert such organic material in a dilute solution. Using methanogenic bacteria, we can obtain methane with an energy content of 145 kJ 1-~ (combustion enthalpy). By using photosynthetic bacteria and

BIOLOGICAL PRODUCTION OF HYDROGEN BY ENVIRONMENTALLY ACCEPTABLE TECNOLOGIES

Solarenergy CO

0

2 g 9

H2,CO21 ..._.~l Hydrogen

I etc"

Sea water,

I generation reactor

[ (Algae)

- ~ inorganic p | salts l [ (N.P.etc) ]

--

2

Hydrogengas

r

separator

o

r

~[]~

Raw materials

fixationf~176 carbondioxide, etc

H2

........................................... f

~

H2, CO 2 Hydrogen generetion

~

-

Fuelfor

motor-cars, rockets

reactor

(Bacterium)

f

L

Fuelcells = Waste water contaning organic materials

[Pretreatment device I_~ P

Bacterial cell separator

r

Separationand Production of usefulby-products

Fig. 1. Concept of biological hydrogen production and utilization.

Physiological active materials


illuminating the solution, we can obtain hydrogen that contains 185 kJ 1-l of chemical energy. In this way the dilute energy can be extracted in a usable form. It is, in a sense, an entropy reducing process which could not be realized by mechanical or chemical systems. Because of this special capability, we believe that renewable energy sources could be efficiently recovered by using a biological method as discussed in this article.

1.2. Mechanisms of biological hydrogen production (Miyake, 1998b) In nature, only bacteria and algae have the capability of hydrogen production. Amongst these organisms, those currently selected for research are anaerobic bacteria, photosynthetic bacteria, cyanobacteria and green algae. Green algae and cyanobacteria directly decompose water to hydrogen and oxygen with light energy: light

H20 ~ H2 + 1/20 2 This is the process by which the earth obtained oxygen in the distant past. The reaction requires only water and sunlight and is very attractive from the viewpoint of environment protection. However, naturally born organisms that carry out this process showed rather low efficiency of hydrogen production due to the complicated reaction system which remains to overcome the large free energy ( + 242 kJ mol-1 hydrogen). For future applications, many improvements are required. Photosynthetic bacteria do not utilize water as the starting compound for hydrogen production but use organic acids. An equation below shows the production of hydrogen from lactate by photosynthetic bacteria. light

C3H60 3 + 3H20 ~ 6H 2 + 3CO 2 In this reaction, free energy is + 8.5 kJ mol-1 hydrogen. Compared to the algal hydrolysis, less light energy is required to produce hydrogen in the case of photosynthetic bacteria when starting from organic substances. There are various kinds of photosynthetic bacteria and many kinds of organic substrates such as fatty acids, sugars,

91

starches, cellulose, etc. that could be used as starting materials for hydrogen production. Photosynthetic bacteria can decompose such organic substances completely. Hydrogen production could be combined with organic wastes treatment (waste fluid from food factories, pulp factories, etc.). Mechanisms of converting light to hydrogen by photosynthetic bacteria and cyanobacteria are shown in Fig. 2. In the project, due to differences in the mechanism of the hydrogen production, different methods of genetic engineering are employed for photosynthetic bacteria and cyanobacteria.


2. I. Cultivation techniques Screening in nature in the early stage of the project provided strains with high efficiency of hydrogen production. The strains used throughout this project are Rhodobacter sphaeroides RV (photosynthetic bacteria), Synechococcus sp. PCC7942 (cyanobacteria), and Clostridium butyricum (anaerobic bacteria). Screening of other strains was also done and some useful ones have been found. Synechococcus PCC7942 (a cyanobacterial host for genetic engineering) was grown at 30~ in liquid BG-11 medium (Rippka et al., 1979) with aeration, under continuous illumination provided by fluorescent lamps (14 W m-2). R. sphaeroides RV and its UV-mutant, P3, were grown in aSy medium (pH 6.8) containing 0.1% yeast extract, 10 mM ammonium sulfate, and 30 mM sodium succinate. The pH was adjusted with NaOH. The cells were cultivated with a 5% inoculure at 30~ for 24 h under anaerobic conditions with tungsten lamps (250 W m-2). The 50 ml of culture was transferred into a 200-ml bottle. Then the culture was filled with gL medium (pH 6.8) containing 50 mM sodium lactate and 10 mM sodium glutamate for hydrogen production. The culture was further incubated at 30~ for 18 h under anaerobic conditions with tungsten lamps (500 W m-2). Only the cultures evolving hydrogen at 10-15 ml h - l were used for experiments.

92


photosvstem

N2ase Electron pool

Plant type Photons Photons

PSI

PS II

f ' ~ ADP ATP 2H20 4H++O2

Bacterial type

N N,

4ADP

PSI . .

ADP

ATP

e

Organic ~ substrates

H§

ATP pool

Fig. 2. Mechanism of light to hydrogen conversion by phototrophic bacteria and cyanobacteria.

2.2. Measurement of hydrogen production Hydrogen evolution from cyanobacteria and photosynthetic bacteria was measured using a hydrogen electrode system (Biott, Tokyo). The reactor volume was 2.5 ml. The light illumination area was 1.77 cm 2. A halogen lamp (Bromo lamp delux, LPL, Tokyo) was used as the light source. The light intensity was determined by a radiometer (model 4090, SJI, Tokyo).

buffer (272 mM sucrose, 8 mM HEPES, pH 7.4). The cells were re-suspended in SH buffer at a concentration of 1 mg chlorophyll ml-~. Hydrogenase solutions of 50 ~tl (0.4 units) were quickly added into 350 lal cell suspensions. An electric pulse with a time-constant of 4.2 ms was immediately applied to the cell suspension (within 1 min) at room temperature (25-30~ under dim light conditions. The cells were washed with Tris buffer to remove extracellular protein and was used for studies.

2.3. Pseudotransformation techniques 2. 4. Genetic manipulation techniques Cells were harvested at 30~ by centrifugation at 10000 x g for 10 min and washed with SH

DNA cloning was carried out according to a

J. Miyake et al./Journal of Biotechnology 70 (1999) 89-101 laboratory manual (Sambrook et al., 1989). Genetic manipulation of cyanobacteria was carried out as described by Poter (1988). Genetic manipulation of photosynthetic bacteria was carried out as described by Simon et al. (1984). 2.5. Photobioreactor research The photobioreactor was composed of four equal compartments. Each compartment had a 0.5-cm light path and a 37.5-cm 2 irradiation area and was made of polyacryl resin of 0.5 cm thickness using transparent resin for the windows and black resin for the frame (Fig. 3). The reactor was immersed into a water bath to keep the temperature at 30~ The compartments were aligned along the light path with a 0.3-cm gap between compartments for the circulation of water to control the temperature. The light path was set normal to the parallel planes of the compartments to analyze hydrogen evolution at various depths in the reactor.

Hydrogen

93

A tungsten lamp of 150 W (color temperature of 2900 K) was used for the light source. Light was passed through 4 cm of water in the water bath and 0.6 cm of polyacryl resin of the water bath vessel before reaching the surface of the reactor. The intensity of the light at the surface of the reactor was adjusted to be either 360 or 720 W m - 2. Hydrogen gas was collected in glass syringes to measure the evolution rates as reported previously (Miyake and Kawamura, 1987). The wavelength-dependent light penetration into the reactor was measured by using a spectrophotometer (Hitachi, model 330, Tokyo) with an integrating sphere. Light energy was measured with a radiometer (YSI-Kettering, model 65A, USA). The emission light spectrum from the lamp was measured using a diode array spectrophotometer (Otsuka Electronic, model MCP-1100, Osaka). The cell density of the photosynthetic bacteria was measured with a spectrophotometer (Hitachi, model 110). The efficiency of the conversion of light energy to hydrogen was calculated using the equation below. Efficiency (%) = (combustion enthalpy of hydrogen) /(absorbed light energy)x 100

3. Results and discussion

3.1. Cyanobacteria i

I

Fig. 3. Photobioreactor with four compartments. The compartments were made with polyacryl resin of 0.5 cm thickness. Irradiation area was 37.5 c m 2 and the light path of each compartment was 0.5 cm. Spectra of light at the depth are illustrated.

3.1.1. Strategy for genetic breeding The mechanism of hydrogen evolution by cyanobacteria has been extensively studied for its potential use as a method of energy production from sunlight. The photosynthetic system produces the reducing power and ATP from light energy. They are transferred to the enzyme system for hydrogen production, hydrogenase or nitrogenase. However, the supply of reducing power and/or ATP for hydrogenase or nitrogenase is not always directly dependent on photosynthetic energy-pro-

J. Miyake et al. Journal of Biotechnology 70 (1999) 89-101

94

Hydrogenase system ,

_.

I compounds t_ I _.~n_._nl

~..

H2

-,

fermentation p a t h w a y

j /

~H2aseJ

7 H2

Bacterial hydrogenase system .o

-----, electronflow Nitrogenase system

H2

Fig. 4. Genetic strategy for hydrogen production by cyanobacteria. PS, photosystem, fd., ferredoxin; H2ase, hydrogenase; N2ase, nitrogenase.

duction but sometimes involves the decomposition of the carbohydrates that have been photosynthetically produced. The involvement of carbohydrate metabolism may be one of the factors which reduce the conversion efficiency of light to hydrogen. An alternative electron transporting pathway is the direct connection of the photosynthetic system to the hydrogen producing enzymes that is expected for efficient hydrogen production (Fig. 4).

3.1.2. Pseudotransformation system As the first step to see the effect of clostridial hydrogenase incorporated into cyanobacterial cells, we have directly introduced the enzyme protein into the cyanobacterial cells by electroporation. We have called the method 'pseudotransformation' (Miyake and Asada, 1997). The cells temporally showed clostridial hydrogenase and were designated as 'pseudotransformants', which retain clostridial hydrogenase without genetic manipulation. These cells were used to investigate in vivo coupling between the photosynthetic system and clostridial hydrogenase in cyanobacteria.

A considerable amount of the active hydrogenase was incorporated into the cyanobacterial cells after electroporation with clostridial hydrogenase at 9 kV cm-~ of electric field strength (Fig. 5). Incorporation of hydrogenase is dependent on strength of electric field, since low levels of hydrogenase activity were detected at low ( < 9 kV cm-1) electric field strengths. Clostridial hy-

>,.~ 4-J

12

~9o

10

~

8

~0 "0 m

4

~m Ib ,-

2 0 0 5 10 15 Electric field strength (kV cm 9 "1)

Fig. 5. Direct electroporation of clostridial hydrogenase protein into cyanobacterial cells.

J. Miyake et al. / Journal of Biotechnology 70 (I 999) 89-1 O1 (b)

(a)

(c)

t-

E c .~/o n -C E e- ~ 8 d

c

.2

-

c

~t-.- g

time 10min

light

~ 1~,, ~ Uru

o

~

L":'-E: .:-...:::- ".:-:. : :-:--.

d time

time

10min

10min

light

light

x x

DCMU

it.,

2

o8

e-

95

x

\ [X;MU H20 1 / 2 0 2

CO2 fixation

H201/202

2

2H +

H2

Q - clostridial hydrogenase

Fig. 6. Light-dependent hydrogen evolution by pseudotransformants. (a) Wild-type cells after electroporation without clostridial hydrogenase. (b) and (c) Pseudotransformants retaining clostridial hydrogenase.

drogenase inside the cells was detected by Western blot analysis (data not shown; see Miyake and Asada, 1997). Light-dependent H2 evolution by the pseudotransformants was measured without any artificial electron carriers. Wild-type cells produced no H2 (Fig. 6a). Induction of hydrogen evolution via light irradiation were observed by the pseudotransformation (91 ~tE m - 2 s - ] ) (Fig. 6b). Dichlorophenyldimethylurea (DCMU) inhibits electron transfer between PS II and PS I. The hydrogen evolution was inhibited by DCMU (10 nM) (Fig. 6c). These results strongly suggest that clostridial hydrogenase could couple with photosynthetic water splitting in vivo.

In order to get permanent incorporation of clostridial hydrogenase in the cyanobacterial cells, we have worked to develop a strong expression system for the non-nitrogen-fixing cyanobacterium, Synechococcus sp. PCC7942. A 262 bpDNA fragment, no. 9 (Miyake et al., 1996), showing strong promoter activity was cloned from Synechococcus PCC7942 by using a promoterprobe plasmid vector, pKE4, based on a promoter-less cat gene cartridge from pKK232-8 and a shuttle vector, pECAN8, which replicates in Escherichia coli and PCC7942. A sigma-70 type promoter (Fig. 7) similar to the other strong promoters was found in fragment no. 9. Tran-35

3.1.3. Expression of clostridial hydrogenase in cyanobacteria Genetic manipulation to improve the hydrogen evolution enzyme system has been studied as the second step. Clostridial hydrogenase I, the powerful enzyme for hydrogen production, was expressed in cyanobacteria. This enzyme is more efficient than nitrogenase which the bacteria usually uses for hydrogen production.

-10

+1

promoter origine

ql~ TAC_~ . . . . (17nt) . . . . TA GATT . . . . A

psbDI psbAIII psbAII

"]['I'GCGA . . . . (18nt) . . . . . TATA CT---G___T_T

No.9 promoter

"FrGACA

E col~ a 7~

" ~ G TAA . . . . (16nt) . . . . TAT C A T - - - - - A CACA . . . . (16nt) . . . . . TAT TAT . . . . A

TATAAT

Fig. 7. Alignment of sigma-70 type promoters in Synechococcus sp. PCC7942.

J. Miyake et al. / Journal of Biotechnology 70 (1999) 89-101

96

Ampicillin resistant gene

cat gene

s~ r

t

pKE4-9 8.50 Kb

I cyanobacterlal

i replication ori

l~;No.9 promoter

Fig. 8. Structure of a shuttle vector between E. coli and Synechococcus sp. PCC7942, pKE4-9.

scription of mRNA started 5-bases downstream of the -10 region of the promoter. The content of CAT protein expressed by the no. 9 promoter reached 20% of the soluble proteins in PCC7942 cells. The no. 9 promoter could be used to express a clostridial hydrogenase I gene in cyanobacteria. A plasmid, pKE4-9 (Fig. 8), which retained the no. 9 promoter, was used to express clostridial hydrogenase gene in Synechococcus PCC7942. The clostridial hydrogenase I gene was cloned from the downstream region of the no. 9 promoter of pKE4-9. E. coli and C. pasteurianum ribosomal binding sequences were designed upstream of the clostridial hydrogenase I gene to investigate the effect of ribosomal binding sequence variation. Clostridial hydrogenase I protein was expressed exclusively by using an E. coli ribosomal binding sequence but not by using clostridial sequence. These results suggest that alteration of ribosomal binding sequence is essential for efficient expression of clostridial hydrogenase I in cyanobacteria. Precise control of the gene expression and enzymatic activity is underway.

3.2. Photosynthetic bacter& 3.2. I. Strategy for genetic breeding A purple bacterium, R. sphaeroides RV is a strong hydrogen producer (Miyake and Kawamura, 1987) and is applied for photobioreactors (Tsygankov et al., 1994; Nakada et al., 1995). The energy conversion efficiency from light to hydrogen is approximately 7% under a solar simulator.

However, the efficiency varies under different light sources. It is believed that the hydrogen production by photosynthetic bacteria may depend on the spectral distribution, since the bacteria utilize the specific light wavelengths for photosynthesis. In the intracytoplasmic membrane of photosynthetic bacteria, two different light-harvesting (LH) complexes are involved in absorption of light quanta and transfer of excitation energy to the photosynthetic reaction center (RC) (Richter and Drews, 1991). R. sphaeroides shows the absorption peaks at 800, 850 and 875 nm, corresponding to the LH2-bacteriochlorophyll B800, B850 and LH 1-bacteriochlorophyll, respectively. An approach for the improvement of hydrogen production by photosynthetic bacteria is the control of photosynthetic protein expression to allow efficient absorption of light energy.

3.2.2. Rearrangement of light harvesting system A mutant, P3 strain, was obtained by UV-irradiation from R. sphaeroides RV. The amount of bacteriochlorophylls in LH1 were reduced to 30% and those of LH2 were enhanced to 140% in P3 compared to R. sphaeroides RV. The contribution of each LH complex to the hydrogen production was examined under monochromatic light (Fig. 9). At 850 and 800 nm (absorption maxima for LH2), the energy conversion efficiencies of P3 were 6.3 and 4%, respectively, whereas those of RV were 5.5 and 3.4%. The maximal production rates of P3 were also 1.4 and 1.5 times higher than those of R. sphaeroides RV. At 875 nm (absorption maximum for LH1), the energy conversion efficiency and the maximal hydrogen production rate of P3 were still high despite a decrease in the level of LH1. Genetic transformation of P3 with a plasmid carrying genes encoding LH1 pigment-binding proteins restored the wild-type absorbance (original LH1/ LH2 ratio) and hydrogen production characteristics of R. sphaeroides RV. We consider that the alteration of the LH1/ LH2 ratio induces the change of the energy pathway from LH pigments to the photosynthetic RC. The change in the amount of bacteriochlorophylls effected the light penetration profile in the photo-


~r-

0.8

~ -o E

0.6

" -5 ,-- E ~ E ~

0.4

97

I

0.2

0 t_

-r~'

0

I

0

9

I

o

I

,

I

,

I

20 40 60 80

,

I

,

0

I

-.

I

o

I

..-

I

9

20 40 60 80

0

I

I

20 40 60 80 100

Light intensity (W/m 2)

Fig. 9. Comparison of hydrogen evolution rates between R. sphaeroidesRV and the UV-mutant, P3, under different monochromatic light conditions. Black circle, R. sphaeroides RV; white circle, the mutant P3. bioreactor which also affected the hydrogen production. Optimization of pigment composition by genetic engineering would improve the hydrogen production by photosynthetic bacteria.

3.2.3. Promoter competition A method for the enhancement of the bacterial light-dependent hydrogen production is proposed by rearrangement of light harvesting systems. Using R. sphaeroides RV, variants have been obtained by the 'promoter competition method' (Fig. 10), thereby making sure of the effectiveness of this method of breeding improvement. Competitive inhibition of the expression of puf operon encoding LH1 and RC were investigated by using multicopy-plasmids, pRKM-LPPluc, which was constructed by insertion of the p u f promoter region (Nagamine et al., 1996) into a broad host vector, pRKM-415, as a control plasmid. R. sphaeroides RV retaining pRKM-LPPluc (RV-LPPluc) showed a weaker accumulation of bacteriochlorophyll than a transformant retaining pRKM-415 (RV-415) after transferring the cultures from 'aerobic dark' to 'anaerobic light' conditions. It was shown that after changing growth conditions the strain RV-LPPluc demonstrated a considerable lag-period before initiating photoheterotrophic growth in comparison with control strain RV-415. The luciferase activity of RV-LPPluc was increased and the maximum luc-activity was observed in the beginning of exponential growth. The existing additional copies of p u f promoter cloned into plasmid may lead to prolongation of

the lag-period prior to photoheterotrophic growth. More detailed investigations of this effect of promoter competition are necessary and continuous cultivation of the bacterium would be preferred.

3.2.4. Basis for the breeding: light penetration into the cell suspension R. sphaeroides RV shows characteristic absorption bands at around 400, 800 and 850 nm. The two major absorption bands at 800 and 850 nm are attributed to the absorption of bacteriochlorophylls. The absorption band around 400 nm is

puf promoter fragment

. ~ ~

chromosome

RNApolymerese

?Oo O j

pRKM-LPPItm

Fig. 10. Competitive inhibition of puf operon expression by using a plasmid, pRKM-LPPluc, which carries puf promoter sequence and a luciferase gene as a reporter gene.

98

J. Miyake et al. /Journal of Biotechnology 70 (1999) 89-101

/ / oe i/

0.9 >,

g6~ /

'

~

-i

.2

3 5 0 ~ 500

'+o+++%

0

20

++~,~+

Fig. 11. Spectra of light at various depths. Light spectrum and energy differ very much at various depths. Up to 20 mm. light energy is mostly adsorbed.

attributed to the absorption of bacteriochlorophylls and hemes (Soret band). Light of wavelength around the absorption maxima of the bacterium (at 800 and 850 nm) decreased rapidly with passage through the cell culture. The spectrum of the light was greatly altered in the deep part of the reactor where the light had maximum energy at 500-780 nm (Fig. 11). The portion of incident light energy absorbed in this first compartment was 69%. The second compartment absorbed 21%, the third 7.0% and the fourth 2.0%. The incident light energy only effectively reaches the first cell. Cells deep in the reactor are only poorly illuminated. Light reaching bacterial cells in the second or third compartments showed a much different spectrum to that of the incident light at the reactor surface. Energy of light of wavelength around the absorption maxima of the cells (800-850 nm) was mostly lost. Only light of wavelength in the region of 500-780 nm (mainly 600-80 nm) reached deep in the reactor. Hydrogen was produced from the energy of this light because the wavelength of which was outside the range of the absorption maxima of bacteriochlorophylls. However, light in the wavelength range was used for hydrogen production in the deep part of the reactor (Fig. 12).

Inhomogeneity of the light distribution in the reactor lowers the overall conversion efficiency. To enhance the total efficiency of light to hydrogen conversion in a photobioreactor, light energy should be equally distributed throughout the reactor. The spectrum of light in the reactor also affected hydrogen evolution. Light with 600-780nm wavelength reached the deep part of the reactor and contributed to the hydrogen evolution. The above conclusion leads us to control the level of pigments in the cells genetically. Cells with less pigment enables the following advantages: light penetration to the deep region of the reactor, moderate quantity of light energy is absorbed at the shallow region of the reactor. These may lead to the total efficiency of a reactor, though the improvement reduces the activity of individual cell. Fig. 13-1 shows a photobioreactor with light diffuser distributing light into the deep part of the reactor. It provided a high efficiency of light to hydrogen conversion (E1-Shishtawy et al., 1997). Solar light was collected and delivered to a reactor by using a fiber glass (Fig. 13-2). This type of reactor has defects in that the production cost is high and there is difficulty in cleaning. In application, a large-scale reactor is necessary as shown in Fig. 13-3. The engineering study on how to satisfy the above light distribution and the simple construction is yet to be examined.

m L

._o ---= u o > c

3

e-

t'13.}

.o_ 2 I.s~:

~t

o "o

1

"1"

1

2

Compartment

E!

3 number

Fig. 12. The relationship between the depth and hydrogen production efficiency. Efficiency at the low light (deep region of the reactor) is high but at high light is low.


T

2

/7,/3, ,J ,..) oa"d

. . . .

,o

I" I: I:

i:

i

o,,~ .~.

I

-o

I

- . I

ol

I

C

/

Fig. 13. Some examples of the reactors developed in RITE projects. (13-1) A reactor with wall-type light diffuser by Kashima Co. (13-2) A reactor illuminated intrinsic illuminating rod by Kubota. Co. Light is supplied from a solar collector (Himawari, La Fore Eng. Co.). (13-3) A large scale reactor (800 1:2 x 4 x 0.1 m) set on the sea surface by Ishikawajima Heavy Industry Co. Ltd.

3.3. Potential of biological hydrogen production The hydrogen production using the photosynthetic microorganisms is regarded as an environmentally acceptable (or friendly) technology. Wastes could be treated while producing hydro-

99

gen at the same time. Hydrogen gas is a clean fuel because it does not produce carbon dioxide. In principle, it dispenses with the use of fossil fuels and is useful in avoiding global warming. An attempt has been made to evaluate the 'environmental friendliness', with the carbon dioxide disposal ( L C C O 2 ) from fossil fuel used in all stages from the construction of the plant to its operation as the index. An important point of the application of solar energy is the efficiency of the conversion. It has been said that the plant photosynthesis is done with an energy conversion efficiency as low as 1%. In our study, R. sphaeroides showed a remarkable conversion ratio of up to 7%. It is comparable to the amorphous semiconductors especially taking into account the low energy cost of system production. Further improvement of the bacteria could provide higher efficiencies. One of the characteristic features of hydrogen production by use of photosynthetic bacteria is the ability to recover and concentrate the energy from high water content organic resources such as waste effluents and sludges from which energy cannot be recovered by combustion. Compared with methane fermentation, a similar energy recovering method, the energy recovery ratio from biomass is theoretically elevated by that increment corresponding to the light energy taken in by the reaction of converting organic acids into hydrogen. Furthermore, the use of the gas obtained thus as a fuel for fuel cells will testify to the superior total energy efficiency of hydrogen production by this system, for hydrogen has higher power generating efficiency than methane. Since hydrogen production from photosynthetic bacteria requires waste matters or effluents containing organic substances as the substrate material, the location of the plant should be in an area where such materials are available, and the plant scale will be regulated by the output of these materials. The requirements for achieving commercialization in the hydrogen production cost are, (1) a steep enhancement of the energy conversion efficiency of the photosynthetic bacteria, (2) a substantial reduction in the installation cost for facilities ranging from pretreatment plants to photosynthetic reactors, and (3) the payback expected

100


from the effluent treatment, the by-product excess microorganisms and so forth. A preliminary estimation has clearly pointed out cost reduction, although the calculation of hydrogen production cost has not yet been reliably made.

3.3.1. A research project for biological hydrogen production in Japan Research Institute of Innovative Technology for the Earth (RITE) launched a project of biological hydrogen production (Development of Environmentally Friendly Technology for the Production of Hydrogen) in 1991. The collaboration included six private industries, and a national institute join the project. The project consists of scientific and engineering research. Genetic improvement of bacteria to produce more hydrogen and design and operation of model reactors are simultaneously carried out in the project. For the attainment of this target, the following R&D activities are underway: 1. Search for and improve photosynthetic microorganisms which have a high hydrogen production capacity. 2. Large scale cultivation techniques for maximizing the efficiency of hydrogen production making use of microorganisms. 3. Techniques for effectively separating and refining the hydrogen formed. 4. Techniques for recovering useful materials including physiologically active materials, etc. 5. Construction of integrated systems for hydrogen production and technical evaluations.

4. Conclusion Use of fossil energy enhances the carbon dioxide concentration in the atmosphere leading to the global warming. New energy sources and the carder should be developed to prevent environmental problems. An energy source which is renewable and easily obtainable is desired. Solar radiation could be the most promising. It is huge in amount, but its energy density is low at the surface of the earth. As a cartier of solar energy, hydrogen is the most promising one because it does not cause air pollution as carbon dioxide

formation. The RITE biological hydrogen production project was established to develop methods suitable for this low density energy. Utilization of photosynthesis could provide a system to accumulate solar radiation. Photosynthetic microorganisms grow by themselves to cover a large area. The biological system does not require factory products. In the studies on the search for and breeding improvement of photosynthetic microorganisms, genetic manipulation methods have been established. R&D of mass cultivation techniques has been carried out to make an economically feasible system of hydrogen production coupled with wastewater treatment.

Acknowledgements This research was done in part as RITE biohydrogen project supported by NEDO/MITI.

References E1-Shishtawy, R.M.A., Kawasaki, S., Morimoto, M., 1997. Biological hydrogen production using noble light-induced and diffused photobioreactor. Biotechnol. Tech. (in press). Miyake, J., 1998a. The science of biohydrogen--background, research and potential. In: Zaborsky, O., Benemann, J.R., Miyake, J., Matsunaga, T., San Pietro, A. (Eds.), Biohydrogen, Plenum, New York, NY, in press. Miyake, J., 1998b. Biological solar energy conversion. In: Miyamoto, K. (Ed.), Renewable Biological Systems for Alternative Sustainable Energy Production. FAO Agricultural Services Bulletin 128, FAO, UN, pp. 7-17. Miyake, M., Asada, Y., 1997. Direct electroporation of clostridial hydrogenase into cyanobacterial cells. Biotechnol. Tech. 11, 787-790. Miyake, J., Kawamura, S., 1987. Efficiency of light energy conversion to hydrogen by the photosynthetic bacterium Rhodobacter sphaeroides. Int. J. Hydrog. Energy 12, 147149. Miyake, M., Yamada, J., Aoyama, K., Uemura, I., Hoshino, T., Miyake, J., Asada, Y.Y., 1996. Strong expression of foreign protein in Synechococcus PCC7942. J. Mar. Biotechnol. 4, 61-63. Nagamine, Y., Kawasugi, T., Miyake, M., Asada, Y., Miyake, J., 1996. Characterization of photosynthetic bacterium Rhodobacter sphaeroides RV for hydrogen production. J. Mar. Biotechnol. 4, 34-37.

J. Miyake et al./Journal of Biotechnology 70 (1999) 89-10I

Nakada, E., Asada, Y., Arai, T., Miyake, J., 1995. Light penetration into cell suspensions of photosynthetic bacteria, relation to hydrogen production. J. Ferment. Bioeng. 80, 53-57. Poter, R.D., 1988. DNA transformation. Methods Enzymol. 167, 703-712. Richter, P., Drews, G., 1991. Incorporation of light-harvesting complex I alpha and beta polypeptides into the intracytoplasmic membrane of Rhodobacter capsulatus. J. Bacteriol. 137, 5336-5345. Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M., Stanie, R.Y., 1979. Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J. Gen. Microbiol. 111, 1-61.

101

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning. A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Simon, R., Priefer, U., Puhler, A., 1984. A broad host range mobilization system for in vivo genetic engineering: Transposon mutagenesis in gram negative bacteria. Bio/Technology 1, 784-791. Tsygankov, A.A., Hirata, Y., Miyake, M., Asada, Y., Miyake, J., 1994. Photobioreactor with photosynthetic bacteria immobilized on porous glass for hydrogen photoproduction. J. Ferment. Bioeng. 77, 575-578.


iiii

JOURNAL

OF



Substrate consumption rates for hydrogen production by Rhodobacter sphaeroides in a column photobioreactor [nci Eroglu a

a,, Kadir Asian, Ufuk Giindiiz b Meral Yticel b Lemi Tiirker c

Department of Chemical Engineer&g, Middle East Technical University, 06531 Ankara, Turkey b Department of Biology, Middle East Technical University, 06531 Ankara, Turkey c Department of Chemistry, Middle East Technical University, 06531 Ankara, Turkey


Abstract

The effect of L-malic acid and sodium glutamate, which serve as the carbon and nitrogen source, respectively, on hydrogen production by Rhodobacter sphaeroides O.U.001 has been investigated in a batch water jacketed glass column photobioreactor (PBR), which has an inner volume of 400 ml. The PBR was operated at different carbon to nitrogen ratios at 32~ with a tungsten lamp at a light intensity of 200 W m-2. Carbon to nitrogen ratio was found to be an important parameter for bio-hydrogen production. Moreover, hydrogen gas production was dependent on certain threshold concentrations of sodium glutamate. L-malic acid consumption was found to be first order with respect to L-malic acid concentration, whereas sodium glutamate consumption was found to be second order with respect to glutamate concentration. It was concluded that there is a close relationship between the hydrogen production rate and substrate consumption rates. A kinetic model is developed, which relates hydrogen gas production per amount of biomass, L-malic acid, and sodium glutamate concentrations. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Hydrogen production; Photobioreactor; Rhodobacter sphaeroides; Instantaneous fractional yield; Hydrogen production factor; Dual substrate

1. Introduction

By using sunlight, some photosynthetic microorganisms produce hydrogen. Hydrogen is considered to be a promising, ideal, and renew-

* Corresponding author. Tel.: + 90-312-2102609; fax: + 90312-2101264. E-mail address: [email protected] (i. Eroglu)

able form of energy. Hydrogen is also used industrially in the synthesis of ammonia, in hydrogenation reactions, and in many other important applications. Bioproduction of hydrogen by photosynthetic bacteria from various substrates and from wastes has previously been investigated. Sasikala et al. (1991, 1992, 1995) produced hydrogen by Rhodobacter sphaeroides O.U.001. Recently, bioprocesses have been developed for hydrogen production by R. sphaeroides RV (Tsy-

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: SO 168-1656(99)00064-4

104

]. Eroglu et al./Journal of Biotechnology 70 (1999) 103-113

gankov et al., 1993; Minami, 1997; Kitajima et al., 1998). At the Middle East Technical University in Turkey, three different bio-systems were designed for hydrogen production: (1) by R. sphaeroides O.U.001, (2) by coupled systems of Halobacterium halobium and E. toll, and (3) photoelectrochemical hydrogen production by H. halobium. In previous work (Arik et al., 1996), a 150 ml glass column photobioreactor (PBR) was constructed for the production of hydrogen by R. sphaeroides O.U.001 under anaerobic conditions and with a fixed light intensity. The optimum hydrogen production conditions were determined to be as follows: pH between 7.3 and 7.8, temperature 31-36~ light intensity 200 W m -2, and cell concentration 1.6-1.8 g 1-1 (dry weight). The maximum hydrogen production rate obtained was 0.047 1 1-1 h-1 gas produced per unit volume of culture with 99% purity. Then, a continuous PBR was scaled up to 400 ml (Eroglu et al., 1998). It was found that both the growth medium and the cell concentration have an influence on photosynthetic hydrogen production. The effect of parameters such as cell concentration and dual substrate concentrations (g-malic acid and sodium glutamate) on hydrogen production rate should be determined. If the relationship between the hydrogen production rate and the substrate concentrations can be expressed in terms of kinetic models, it might be possible to achieve prolonged continuous hydrogen production in a larger column PBR. To date, no reference exists in the literature about the modeling of large scale hydrogen production. To achieve such modeling, primarily kinetic models that relate the consumption of the carbon source, the nitrogen source and the cell growth and hydrogen production rates are required. Therefore, the objective of the present study is to find relationships between the consumption rates of the two substrates--u-malic acid as a carbon source and sodium glutamate as a nitrogen s o u r c e ~ o n cell growth and hydrogen gas production. This paper presents an initial quantitative approach to describe kinetics of hydrogen production by photoheterotrophic bacteria.


R. sphaeroides O.U.001 (DSM 5648) was grown under anaerobic and sterile conditions in a minimal medium of Biebl and Pfenning (1981). The growth medium contained n-malic acid as the carbon source and sodium glutamate as the nitrogen source, and a vitamin solution (thiamin and niacin; 0.0005 g 1-1). Temperature was 32~ The growth medium was illuminated using a tungsten lamp at a light intensity of 200 W m -2. The initial pH of the growth medium was 7. Argon gas was used to create anaerobic conditions. The column PBR was made up of a glass cylinder that had an inner volume of 400 ml and was surrounded by a water jacket. At the top of the reactor, there was an inlet for the medium and outlets for the argon and for the hydrogen gas that was collected in a gas-measuring burette. Fresh medium was added from a reservoir that was placed above the PBR. Microorganisms were inoculated through the septum. At the bottom of the column PBR, there was an outlet for the culture and an inlet for argon gas. Experiments were carried out using media containing different initial amounts of the substrates n-malic acid and sodium glutamate. In the various experiments the initial concentrations of u-malic acid were 7.5, 15 or 30 mM, whereas the initial concentrations of sodium glutamate were 1, 2 or 10 mM. Samples were taken at 12 h time intervals while flushing the column with argon. Bacterial cell concentrations were measured as an increase in absorbance at 660 nm (Hitachi Spectrometer). The pH of the samples was also measured. In order to determine the consumption of the substrates, the samples were centrifuged and the supernatant was subjected to HPLC analysis (Shimadzu HPLC, BIORAD Aminex Ion Exclusion Column). The gas produced was analyzed by gas chromatography (Hewlett Packard 5890, Series II). We replicated each of the nine runs at least twice.


In a set of nine replicated experiments, different combinations of initial substrate concentrations of

105

I. Eroglu et al./Journal of Biotechnology 70 (1999) 103-113

L-malic acid (CLMAo) and sodium glutamate (CNGo) were tested. The experiments showed that hydrogen production had started at some time after the inoculation of the bacteria. Different hydrogen production starting times (to) were obtained in different runs. Table 1 summarizes the total volume of hydrogen gas evolved (VT), gas production starting time (to), duration of the gas production, and the maximum cell concentration measured in these runs. Fig. 1 illustrates total hydrogen gas produced with respect to time in Runs 4 - 6 where initial L-malic acid concentration was 15 mM in each run. Initial sodium glutamate concentrations were varied and set at 1, 2 and l0 mM, respectively. High initial concentration of sodium glutamate (10 mM) in Run 6 led to a 4-fold decrease in hydrogen production relative to Run 4. Fig. 2 shows the growth curves of R. sphaeroides O.U.001 obtained for the same set of runs. The maximum cell concentration obtained in Run 6 was almost two times greater than that of Run 4 and Run 5. It was observed that excess sodium glutamate enhanced the cell growth but inhibited the hydrogen production. The pH varied between 7 and 7.6 in these runs. That is, during the course of these experiments, a slight decrease is observed in the pH of the medium during the

cell growth period, but once the hydrogen was produced, the pH increased. A similar comparison was made between the results obtained in Run 1, Run 4 and Run 7. In all these experiments the initial sodium glutamate concentration was set at 1 mM, and the initial L-malic acid concentrations were varied to 7.5, 15 and 30 mM, respectively. The total volume of hydrogen gas evolved, the duration of the hydrogen production, and the maximum cell concentration were increased, as the initial L-malic acid concentration was increased. This might indicate that high L-malic acid concentration enhanced both the cell growth and the hydrogen production at low sodium glutamate concentrations (Run 1 and Run 4). The pH varied between 7 and 7.8 during those runs. The consumption of the substrates during the runs must also be taken into account. It was observed that hydrogen production ceased when the L-malic acid had been totally consumed in all those runs having low initial sodium glutamate concentration (Runs 1, 2, 4, 5, 7 and 8) (Eroglu et al., 1998). However, if high sodium glutamate was present in the system (Runs 3, 6 and 9), both substrates were not utilized completely and they were left in the culture when hydrogen production stopped. It should be emphasized that cell concen-

g ,,,,,,,

200

0

--4~--.. 1 mM

100

'-"

2mM

. - ~ k ~ l O mM "

0

-----

----t

30

~--

;-

f

60

,,~ 90

f 120

~ 150

,I 180

t

210

240

Time (hrs)

Fig. 1. Total hydrogen production at a substrate concentration of 15 mM L-malic acid and three different sodium glutamate concentrations (1, 2 and l0 raM).

Table 1 The results of the hydrogen production experiments

h n I\

Run I 2 3 4 5 6 7 8 9

Initial L-malic acid concentration (mmol I - ' )

7.5 7.5 7.5 15 15

15 30 30 30

Initial sodium glutamate concentration (mrnol 1- ') I 2

Total volume of H, produced (ml)

I 2

203 1 I7 52 312 3 14 86 512 477

10

101

10

1 2 10

Gas production starting time I, (h) 40 40 40 50 40 40 70 50 90

Duration of gas production (h) 91 73 55 174 97 50 229 287 87

Maximum cell concentration (g I - ] ) 3.6 4.5 8.5 5.0 5.6 10

7.4 6.4 9.0

6 3

%

s

107

]. Eroglu et al./Journal of Biotechnology 70 (1999) 103-113 12

- - e - - 1 mM

.J

~10

~2mM

3=

--~ir-- 10 rr~

~'8

"t3

~6 u C

uo 4 u2 o ~==='~'~ 0

i

~

t

i

50

100

150

200

Time(hr)

250

Fig. 2. Growth curves of bacteria at a substrate concentration of 15 mM L-malic acid and three different sodium glutamate concentrations (1, 2 and 10 mM). tration reached very high values in a short time in these runs. That might indicate the positive effect of high sodium glutamate concentration on cell growth, but would also indicate the inverse effect on hydrogen production.

4. Dual substrate consumption rates Integral method of analysis followed, to interpret the substrate consumption data and to find the consumption rate equations for L-malic acid and sodium glutamate. During this analysis, the volume of the PBR was assumed to be constant, since small amounts of samples (2 ml) were taken out. Significant variations of temperature and concentration were not expected in the reactor, since the reactor was small and argon gas bubbled during sampling. The cells were suspended in the column, but some of the cells--most probably the dead cells--sank to the bottom, which caused small fluctuations in cell concentration values. Various rate models m including Monod type of equations--have been tried using the Microsoft Excel 5 package program. The following rate equations gave the best fit: the first order consumption rate equation for L-malic acid (Fig. 3):

(1)

-- rLMA = --dCLMA/dt = k1CLMA and the second order consumption sodium glutamate (Fig. 4):

rate

for

- ryG = -- dCyG/dt = k 2 C 2 G

(2)

Table 2 summarizes the best fitting rate parameters of Eqs. (1) and (2), where n denotes the order of the rate equation, kl and k2 are the rate constants for the first order (h-1) and for the second order (1 mol-1 h - l ) rate equations, respectively. R 2 measures the dispersion of distribution from the mean. It varies between values 0 and 1, where 1 denotes the maximum agreement of the experimental data to calculated data. It should be noted that these fits were quite acceptable, considering the experimental difficulties and the scatter in the experimental data. No direct relationship was found between the rate constants and the concentration of the other substrate. Therefore, these rate equations were considered to be independent. The second order consumption rate of sodium glutamate may indicate that two molecules of glutamate are involved in the consumption reaction as would be the case if the enzyme had two binding sites for glutamate.

5. Cell growth without hydrogen production The growth curves obtained were examined in two phases. Phase 1 is the cell growth without hydrogen production, and Phase 2 is the hydrogen production. In Phase 1, between initial time and tc, substrates were mainly consumed by the

108

I. Eroglu et al. / Journal of Biotechnology 70 (1999) 103-113

perimental data. For convenience in evaluating cell concentration and substrate distribution, two yield terms are introduced: instantaneous fractional yield of cells, y, and overall fractional yield of cells, Y. Let y,V/LMA(I) be the ratio of the rate of new cells formed to the rate of consumption of L-malic acid (in terms of mass) at any time t. YX/LMA is called instantaneous fractional yield of cells with respect to L-malic acid. Similarly, the instantaneous fractional yield of cells with respect to sodium glutamate, y x / N G ( t ) , is the ratio of the rate of mass of new cells formed to the rate of consumption of sodium glutamate (in terms of mass). The L-malic acid consumption rate, rLMA, and sodium glutamate consumption rate, -rNG, are related to the growth rate, rG, by yield factors YX/LMA and YX/NG. From Eqs. (1) and (3) and similarly from Eqs. (2) and (3), the following relations are obtained:

growing cells. The growth rate of the bacteria, rG (the change in cell concentration relative to the change in time) is: rG = d X / d t

= mm](

(3)

where X is cell concentration (g 1-1), t is time (h), and m m is the specific growth rate (h-1). Eq. (3) should be valid for the exponential growth phase of the bacteria. In this phase, the cells adapt themselves to their new environment. After the adaptation period, cells should multiply rapidly, and cell mass and cell number density should increase exponentially with time. This is a period of balanced growth in which all components of the cell grow at the same rate. That is, the average composition of a single cell remains approximately constant during this phase of growth. During balanced growth, the specific growth rate determined from either cell number or cell mass would be the same. Since the nutrient concentrations are large in this phase, the growth rate is assumed not to be limited by the nutrient concentration, hence the exponential growth rate could be expressed as first order. Time required for the inoculation of the bacteria is neglected. A plot of In X versus time yields mm which is the slope of the best fitting line to represent the cell growth data obtained for each run. It was interesting to note that mr, varied between 0.022 and 0.217 h for different runs (Table 2). This result might indicate the significance of dual initial substrate concentrations on specific growth rate. In the present work, a new approach, the yield model, has been introduced to represent the ex2,50

l

1,50 1,00

(4)

yx/yG(t) = mmX/(keC2GMWNG)

(5)

The instantaneous fractional yields of cells with respect to each substrate depend on time. Therefore, they should be estimated by introducing the concentrations measured at that time. Table 3 lists the instantaneous yield factors estimated at to. Overall fractional yield which is the ratio of total mass of new cells formed until tc to total mass of L-malic acid consumed until tr is the mean of the instantaneous fractional yields between the initial time and t~ with respect to each substrate:

=oo gx I

y

-.- 2,00

yX/LMA(I) = m m X / ( k , C L M A M W L M A )

'~' 0,50 0,00

~-

0

~

t

20

"

'

?"

40

I

I

I '

I

60 80 timelhour)

'

1 ~

100

I w

120

140

Fig. 3. The first order consumption rate model for L-malic acid in the medium containing 7.5 mM L-malic acid and 1 mM sodium glutamate initially.

]. Eroglu et al./Journal of Biotechnology 70 (1999) 103-113

109

2000 1600 ~'

z

1200 y = 4,7357x + 951 RZ = 0,8825

800 400 0

20

40

60 80 time (hour)

100

120

140

Fig. 4. The second order consumption rate model for sodium glutamate in the medium containing 7.5 mM L-malic acid and 1 mM sodium glutamate initially. X(tc) - Xo

YX/LMA = ( C N ~ o -

CN~(to)MWLMA)

(6)

Similarly, the overall fractional yield of cells with respect to sodium glutamate is: X(tc) - ;Co

YX/N~ = ( CNGo - CN~(tr

(7)

where Xo is the initial cell concentration, which is neglected in the present work. Table 3 lists the overall fractional yield factors estimated in Phase 1. The instantaneous fractional yields estimated at tr were greater than the overall fractional yields. The yields for L-malic acid were much less than the yields for sodium glutamate. Those results indicated the importance of sodium glutamate for cell growth.

6. Hydrogen production Hydrogen production rate is defined in three different ways: 1. The average hydrogen production rate per culture volume, Rn2av, which is calculated by dividing total volume of gas produced by the volume of the reactor and by the duration of gas production, and has the unit of l 1-1 h-z. 2. The maximum hydrogen production rate per culture volume, RH, which is estimated from the total volume of~gas produced versus time data by simulation with a polynomial curve fitting and differentiation with respect to time. The plot of hydrogen production rate versus time gives a maximum at some time, tm. This is the time corresponding to the maximum

hydrogen production rate per unit volume of the culture, where the units are 1 1-1 h-1. 3. The maximum hydrogen production rate per biomass, R h2, is calculated by dividing the maximum hydrogen production rate, RH2, by cell concentration, X, measured at /m, where the unit is 1 g-~ h-~. The maximum hydrogen production rates were found to be quite close to the average hydrogen production rates (Table 3), which might indicate that the hydrogen production rate does not change too much during the gas's evolution. The rate of hydrogen production obtained in Run 5 was the highest compared to the other runs (0.01 1 1-~ h - t or 0.0024 1 g - ~ h - ~). These results are quite comparable with the rates obtained by Kitajima et al. (1998) who found hydrogen production rates as 0.042 1 1-~ h-~ or 0.0042 1 g-~ h-~ in plane type photosynthetic bioreactor. Minami (1997) reported 0.015 1 1-~ h-~ for R. s p h a e r o i d e s RV in 10 1 continuous PBR, and 0.011 1 1-~ h for 1.4 1 continuous PBR. The cells grow at a slower rate after tc during hydrogen production. The cell growth rate in Phase 2 is: !

ro = d X / d t = m m X

(8)

where m m is the specific growth rate during hydrogen production, m m values were obtained from the slope of the best fitting line of In X versus time, between tc < t < tm, and are listed in Table 2. m m values were found to be almost an order of magnitude less than the mm values for Phase 1. That could be attributed to the increase in the rate of death of the cells. In continuous experi-

110

]. Eroglu et al. / Journal of Biotechnology 70 (1999) 103- 113

ments the difference between m m and mm is very critical. If the dilution rate is not adjusted to mm, the cells will wash out. Instantaneous fractional yields in Phase 2 were estimated from Eqs. (4) and (5) by replacing m m with mm, which might also involve the consumption of the substrates due to the maintenance of the cells and hydrogen production.

YHJx = 3.17(X/CNG2) T M

(R2=0.74)

(b) The interactive model: hydrogen production factor depends on the yield of both of the substrates. That is:

Y n J x oc (Yx/NG)/(Yx/LMA)

(13)

where Y n J x is proportional to (CLMA/CNG2). The best fitting equation is:

Y n J x = 3.48(CLMA/CNG2) ~

7. Yield models In order to compare the results of different runs, a factor (the hydrogen production factor) is defined:

yH2/X = R'H2tm

(12)

(R 2 = 0.82)

(14)

It is concluded that hydrogen production is related to the ratio of the concentration of Lmalic acid to the square of the concentration of sodium glutamate.

(9)

yn2/x, by definition, has the unit volume of gas produced per dry weight of bacteria (1 g-~). Table 3 lists the instantaneous fractional yields and the hydrogen production factor estimated at t m . Two yield models are proposed for the estimation of the hydrogen production factor: (a) The non-interactive model: the hydrogen production factor depends on the yield of one of the substrates: either on the yield of L-malic acid or on the yield of sodium glutamate: yH2/X W.YX/LMA OCX~ CLMA

(1 O)

YH2/X OCYNG/X Ct?.X/ CNG2

(1 1)

The hydrogen production factor did not depend ( X / C L M A ) , hence, R 2 was very small. However, the hydrogen production factor depended on (X/ CNG:). The best fitting equation is: on

8. Hydrogen production models considering carbon to nitrogen ratio It was observed that the hydrogen production rate was affected by the ratio of substrate consumption rates, as explained in the previous section. Moreover, the ratio of carbon to nitrogen could be taken as a parameter instead of the substrate concentrations individually (Minami, 1997). That is to say, the hydrogen production rate is related with residence time, the carbon to nitrogen ratio, and cell concentration. Using these parameters in trial runs, it was found that the following models give the largest R 2 values:

yH2/X = 6.25[(C/N)X] ~

(R 2 = 0.56)

Y n J x = 22.6[(C4/N)X] ~

(15)

(R 2 = 0.75)

(16)

Table 2 Summary of the rate parameters Run

k 1 (h -l)

R2

k2 (1 mol -I h -l)

R2

mm (h-1)

mm (h-1)

1 2 3 4 5 6 7 8 9

0.0179 0.0316 0.0089 0.0109 0.0183 0.0141 0.0075 0.0096 0.0101

0.92 0.94 0.96 0.90 0.88 0.91 0.87 0.87 0.70

4.74 3.43 0.19 2.68 2.45 0.27 2.78 8.12 0.30

0.88 0.63 0.86 0.72 0.95 0.89 0.79 0.82 0.73

0.058 0.038 0.217 0.068 0.164 0.089 0.072 0.022 0.14

0.0066 0.0061 0.0085 0.0021 0.0014 0.0286 0.0034 0.0020 0.0055

c. Table 3 Yield factors and hydrogen production rates I

22 21 242 20 56 18 23 3 72

218 196 406 20 668 56 887 37 490

4.5 4.8 14.9 3.0 4.6 2.6

I44 24 64 41

1.1

I09

70 65 65 130 80 70 190 175 160

96

18

8.0 3.4 -~ ~

--

33 13 __

-

0.006 0.008 0.002 0.007 0.012 0.005 0.006 0.006 0.003

I

')

R112.1" (1

0.006 0.004 0.002 0.005 0.010 0.005 0.006 0.004 0.003

h-' 1

I)

R;IJ

h - ' g - I ) Yl12/X(LJ (1 g - ' )

0.00 19 0.0018 0.0003 0.0020 0.0024 0.0009 0.00 10 0.0010 0.0005

0.133 0.117 0.020 0.260 0. I92 0.063 0. I90 0.175 0.080

Y X / L M A (1,)

YXING

3.7 8.2 12.9

32.3 38.8 21.5 51.4 8.0 49.0 63.3 28.2 20.3

1.1

0.6 15.1 3.3 2.4 3.7

(Irn)

2 4 a

112

]. Eroglu et al./Journal of Biotechnology 70 (1999) I03-113

where C/N denotes the ratio of the total moles of carbon per the total moles of nitrogen present in the culture, whereas C4/N denotes the moles of carbon present in e-malic acid per the total moles of nitrogen present in the culture. It is interesting to note that hydrogen production depends on the moles of carbon present in L-malic acid more than on the total moles of carbon present in the medium.

9. Conclusions The following conclusions can be drawn from this study: 1. Both of the substrates (e-malic acid and sodium glutamate) are vital for hydrogen production. Moreover, hydrogen gas production is dependent on certain threshold concentrations of sodium glutamate. Interestingly, as the concentration of sodium glutamate reaches a certain upper limit, hydrogen production ceases. Therefore, the stress condition exerted by sodium glutamate to produce hydrogen is operative in a certain range of the concentration. 2. Hydrogen production starts after a lag period (40-80 h) and continues even after, the cell concentration has leveled out. The specific growth rate of bacteria in the hydrogen production phase is less than the specific growth rate of the bacteria in the exponential cell growth phase without hydrogen production. This finding contradicts the conclusions of previous researchers (Sasikala et al., 1992; Arik et al., 1996), who found that hydrogen was mainly produced in the exponential phase of growth in smaller sized PBRs. 3. The e-malic acid consumption rate was found to be first order with respect to the e-malic acid concentration, whereas the sodium glutamate consumption rate was found to be second order with respect to the sodium glutamate concentration. 4. There is a relationship between the cell concentration and the hydrogen production rate. However, the hydrogen production rate basically depends on the L-malic acid to sodium

glutamate ratio. The maximum hydrogen production rate is observed with the growth medium initially containing 15 mM L-malic acid and 2 mM sodium glutamate concentrations.

Acknowledgements This research has been supported by the Turkish Scientific Research Council (TUBITAK) Project number TBAG 1535, and the Middle East Technical University (METU) Research Fund, project number AFP-96-07-02-02.

Appendix A. Nomenclature

C C/N C4/N kl k2 mm t

mm

MW RG R RH 2

Rh2 RH2av R2 t tc tm

concentration (mol 1-1) ratio of the total moles of carbon per moles of nitrogen present in the culture ratio of the moles of carbon present in L-malic acid per moles of nitrogen in the culture first order reaction rate constant (h -1) second order reaction rate constant (1 mol-1 h-l) specific growth rate in Phase 1 (h-') specific growth rate in Phase 2 (h -1) Molecular weight (g mol -~) growth rate of the bacteria (g 1h -l) consumption rate (mol 1-~ h-l) maximum hydrogen production rate per culture (1 1-~ h -~) maximum hydrogen production rate per biomass (1 g-1 h-l) average hydrogen production rate per culture (1 1-~ h -~) goodness of fit time (h) hydrogen gas production starting time (h) residence time of maximum hydrogen production rate (h)

i. Eroglu et al./Journal of Biotechnology 70 (1999) 103-113

Vv X Xmax

total volume of hydrogen gas evolved (ml) cell dry weight concentration (g 1-1 ) maximum cell concentration (g 1-1)

yH2/X

hydrogen production factor (1 g-l)

y

instantaneous fractional yield of cells overall fractional yield of cells

Y

Subscripts H2 LMA NG o

X

hydrogen gas L-malic acid sodium glutamate initial cells

References

Arik, T., Gunduz, U., Yucel, M., Turker, L., Sediroglu, V., Eroglu, I., 1996. Photoproduction of hydrogen by Rhodobacter sphaeroides O.U.001, Proceedings of the 11th World Hydrogen Energy Conference, Stuttgart, Germany, vol. 3, 2417-2424.

113

Biebl, H., Pfenning, N., 1981. Isolation of member of the family Rhodosprillacaea. In: The Prokaryotes. SpringerVerlag, New York, pp. 267-273. Eroglu, I., Asian, K., Gfindfiz, U., Yficel, M., Tfirker, L., 1998. Continuos hydrogen production by Rhodobacter sphaeroides O.U.001. In: Zaborsky, O.R. (Ed.), BioHydrogen. Plenum press, New York, pp. 143-149. Kitajima, Y., E1-Shishtalwy, R.M.A., Ueno, Y., Otsuka, S., Miyake, J., Morimoto, M., 1998. Analysis of compensation points of light using plain type photosynthetic bioreactor. In: Zaborsky, O.R. (Ed.), BioHydrogen. Plenum Press, New York, pp. 359-367. Minami, M., 1997. Biohydrogen production using sewage sludge by photosynthetic bacteria. Paper presented in BioHydrogen '97, the International Conference on Biological Hydrogen Poduction, Kona, Hawaii, USA. Sasikala, K., Ramana, C.H.V., Rao, P.R., 1991. Environmental regulation for optimal biomass yield and photoproduction of hydrogen by Rhodobacter sphaeroides O.U.001. Int. J. Hydrogen Energy 16, 597-601. Sasikala, K., Ramana, C.H.V., Rao, P.R., 1992. Photoproduction of hydrogen from the waste water of a distillery by Rhodobacter sphaeroides O.U.001. Int. J. Hydrogen Energy 17, 23-27. Sasikala, K., Ramana, C.H.V., Rao, P.R., 1995. Regulation of simultaneous hydrogen photoproduction during growth by pH and glutamate in Rhodobacter sphaeroides O.U.001. Int. J. Hydrogen Energy 20, 123-126. Tsygankov, A.A., Hirata, Y., Miyake, M., Asada, Y., Miyake, J., 1993. Hydrogen evolution photosynthetic bacterium Rhodobacter sphaeroides RV immobilized on porous glass. New Energy Systems and Conversions, Universal Academy Press, pp. 229- 233.


i

)OUkI~AL

OF



The biocatalytic effect of Halobacterium halobium on photoelectrochemical hydrogen production Vedat Sediroglu a Inci Eroglu a Meral Yiicel b,* Lemi Tiirker r Ufuk Giindiiz a

b

Department of Chemical Engineering, Middle East Technical University (METU), 06531 Ankara, Turkey b Department of Biology, Middle East Technical University (METU), 06531 Ankara, Turkey c Department of Chemistry, Middle East Technical University (METU), 06531 Ankara, Turkey


Abstract Hydrogen gas can be produced electrochemically by leading a current through two electrodes immersed in a NaC1 solution. Bacteriorhodopsin (BR) a protein found in the purple membrane of Halobacterium halobium, is known to pump protons across the membrane upon illumination. In this study, the effect of BR on photoelectrochemical hydrogen production was investigated. A batch type bio-photoelectrochemical reactor was designed and constructed. The photoelectrochemical hydrogen production experiments were performed with free H. halobium packed cells or immobilised H. halobium cells. The cells were either immobilised in polyacrylamide gel (PAG) or on cellulose acetate membrane (CAM). Experiments were also performed with purple membrane fragments of H. halobium immobilised on cellulose acetate membrane. It was found that the presence of bacteriorhodopsin (BR) in the reactor enhances the hydrogen production rate upon illumination. Immobilisation increased the amount of hydrogen produced per mole of BR. Compared to control experiments without BR, the power requirement of the photoelectrochemical reactor per amount of hydrogen produced decreased fourfold when purple membrane fragments immobilised on CAM were used. The presence of BR regulates the pH of the system, increases the hydrogen production rate and causes light-induced proton dissociation, which lowers the electrical power requirement for the electrochemical conversion. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Hydrogen production; Bacteriorhodopsin; Halobacterium halobium; Cellulose acetate membrane; Immobilisation; Polyacrylamide gel; Photoelectrochemical reactor

1. Introduction

* Corresponding author. Fax: + 90-312-2101289. E-mail address: [email protected] (M. Y/icel)

Hydrogen, being a renewable and a clean energy source, is an i m p o r t a n t fuel and basic raw material for chemical industries. Rapid advancement in biotechnology in recent years has drawn


116

V. Sediroglu et al./Journal of Biotechnology 70 (1999) 115-124

attention to technologies for the production of hydrogen with micro-organisms and solar energy, as well as its conventional production methods such as steam reforming of hydrocarbons, coal gasification, and electrolysis (Shreve, 1984). There are many studies on photobiological hydrogen production by photosynthetic bacteria, green algae, cyanobacteria, and cell free systems (Hall et al., 1995; Markov et al., 1995). However, technical problems currently limit the yield and duration of these systems, and therefore further development of a practical photobiological hydrogen producing system is necessary. Halobacterium halobium is a photosynthetic bacteria that can survive in 4 M NaC1 solutions (Oesterhelt and Stoeckenius, 1971). The cell membrane of H. halobium is subdivided into five sections; red, brown, white, yellow, and purple membranes (Oesterhelt and Stoeckenius, 1974). Although they have different functions in H. halobium, 80% of the cell membrane is composed of purple fragments, which contain a light transducing retinal protein called bacteriorhodopsin (BR). BR uses light energy to pump protons across the cell membrane. The retinal protein BR is very stable and may retain its activity for several years even in isolated form. This property has attracted the attention of researchers (Birge, 1990; Trissl, 1990). In recent studies, systems containing H. halobium or separated BR have been used for the photo production of hydrogen. Although protons are generated upon illumination in these micro-organisms, protons cannot be converted to molecular hydrogen because they normally lack hydrogenase enzymes. A proton reduction system should be coupled to H. halobium to enable the hydrogen production. This has been achieved by coupling H. halobium to a salt tolerant Escherichia coli (Khan and Bhatt, 1989; Patel and Madamwar, 1994; Kaya et al., 1996; Khan and Bhatt, 1997). H. halobium packed cells coupled to an electrochemical system was also found to be very promising and advantageous, since nutrients are not required (Khan and Bhatt, 1990, 1992; Sediroglu et al., 1996a). In these systems NaC1, which is very cheap and abundant, is used as electrolyte. This system is

considered to have an industrial potential, but it needs further improvement. The main objective of this study is to investigate and compare the different process conditions for using Halobacterium halobium for photoelectrochemical hydrogen production. Also, new immobilisation techniques were developed in order to improve the stability of the system. The factors studied are sodium chloride concentration, concentration of BR, and novel immobilisation techniques.


2.1. Preparation of H. halobium packed cells and purple membrane fragments The materials used were all of reagent quality and were obtained from Merck (Germany). H. halobium S-9 strain was kindly provided by Professor Oesterhelt from the Max Planck Institute (Munich, Germany). H. halobium cells were grown as described by Yiicel et al. (1995). After 7 days of growth, the cells were collected by centrifugation at 6380 rpm for 20 min in a Sorvail centrifuge (GSA rotor). The precipitated cells were named as packed cells. These cells were no longer alive, but BR was active for a long period of time. In the experiments, these cells were resuspended in 4 M NaCI solution which prevents the osmotic rupture of the cells and to keep these cells in intact form. The purple membrane fragments were prepared according to the method as described in the literature (Yticel et al., 1995). 2.2. Measurement of photoactivity of H. halobium packed cells and PM fragments

The photoactivity of H. halobium packed cells and PM fragments were measured as pH versus time by using a combined pH electrode connected to a pH transmitter at 25~ as described by Yficel et al. (1995). Illumination was achieved by means of a 1000 W projector lamp located at a distance of 50 cm away from the vessel.

V. Sediroglu et al. / Journal of Biotechnology 70 (1999) 115-124

All the membranes were kept in 4 M NaC1 solution at 4~ in the dark for at least 24 h prior to the experiments. H. halobium packed cells were immobilised in polyacryamide gel (PAG) using the procedure originally developed by Eroglu et al. (1994) for the immobilisation of PM. The only difference was that instead of PM, H. halobium packed cells were added to gel solution. The concentration of BR changes with the amount of H. halobium packed cells added. The gel solution polymerised within the circular holes of the immobilisation assembly and nestled down in the niche by sticking to the periphery of the hole (Fig. 1). Polymerisation was completed in about 30 min. The gel constituted assembly were sunk into 4 M NaC1 solution to attain the uppermost swelling of the gel, and the assembly was kept in this solution in a refrigerator.

2.3. Immobilisation techniques

Cellulose acetate membranes (CAM) having a pore size and thickness of 0.2 gm and 0.15 mm, respectively, were placed into a novel immobilisation assembly (Fig. 1). The immobilisation assembly made of Plexiglas had nine circular sieved holes, each of which was 2.5 cm in diameter and 0.35 cm in depth. The holes in the sieves are each 1 mm in diameter. Phospholipid solution was prepared by dissolving egg-yolk phosphatidyl cholin (lecithin) in chloroform (10 g 1-1). According to the selected lecithin to BR ratio, the required amount of phospholipid solution was supplied to one side of the CAM and dried (Sediroglu et al., 1996b). Then, packed cells or purple membrane (PM) fragments with a known amount of BR were spread onto the lecithin impregnated cellulose acetate membrane surfaces.

[ i p n , T ]11

~ 9 ~ ~

~ 9 ~

IO01

12

~ 9176

9

''~176176176'.:

+

-

117

.

~ ~

~

"9

9

~176

9

x "~ "

"

~

1 ~

.

"'~176

i

T" 9

lo "~

...

4

5 13

2I

1

~

o

o

Fig. 1. Experimental setup: (1) light source; (2) magnetic stirrer; (3) water bath; (4) temperature probe; (5) pH electrode; (6) anode and cathode; (7) power supply; (8) analog digital converter; (9) pressure transducer; (10) immobilization assembly; (11) pH transmitter; (12) potentiometer; (13) black plate.

118


2.4. Photoelectrochemical hydrogen production The experimental design and the bio-photoelectrochemical reactor (Bio-PEC) are shown in Fig. 1. The Bio-PEC was a 1 1 glass tank placed in a constant temperature water bath (made of glass) and stirred with a magnetic stirrer at 100 rpm (Sediroglu et al., 1996a). The upper part of the reactor had four openings: inlets for the electrode assembly, the combined pH electrode, the temperature probe, and a H2 gas outlet. The electrode assembly contained a platinum electrode as the cathode (surface area was 0.8 cm 2) and a silver electrode as the anode (surface area was 4.5 cm 2) located 1 cm away from each other. The electrodes were connected to a constant voltage power supplier. Voltage and current were recorded by a computer connected to an analogue digital converter. The combined pH electrode attached to a pH transducer and a temperature probe connected to a potentiometer were employed. The hydrogen gas produced was collected in a 1.4 1 closed vessel and a pressure transducer (Cole Parmer, J-7352 Series) connected to an analogue digital converter, which was employed to record the increase in pressure. In control experiments the evolved gas was collected in a gas measuring burette by reversible displacement of water. The gas was sampled with a gas-tight syringe (Alltech) and analysed by gas chromatography (Hewlett 5890 Packard Series II). Control experiments did not involve any BR. The voltage applied was 0.88 V. Fresh NaC1 solution was used and both Pt and Ag electrodes were regenerated before each experiment. The experiments were repeated at least three times. Experiments were performed at 25~ under anaerobic conditions. Illumination was achieved at a light intensity of 1000 W m - 2 . During the experiments voltage, current, pressure change and pH were recorded with respect to time. In the control experiments, NaC1 concentrations were changed from 0.1 to 3.5 M. The ranges of total BR used were 0.098-1.76 ~tmol for free H. halobium packed cells, 0.033-0.098 ~tmol for the packed cells immobilised on CAM and 0.0980.294 lamol for the packed cells immobilised on PAG. Whereas, BR used was 0.067 jamol for PM fragments immobilised on CAM.

8.0

6.0

/ar z'

-~--. E 4.0

. . . . .

~ . . ~ : ~ ~ ' ~

,. . . , . ,

0.0 -.--0

i_r a

"' 200

400 Time(rain)

8.74 V

600

.

800

b)

8.34 ~ 7.94 7.54 7.14 -0

~ ........... 200

400 Tlme(mln)

600

800

c)

4.00 2.00 E [

0.00 200

L1

400

-2.00

& -4.00 Tlme(mln)

Fig. 2. Effect of NaC1 concentration on photoelectrochemical hydrogen production. (a) H2 produced as volume versus time. (b) Change of pH versus time. (c) Change of current versus time. (~ Power ON-light ON; T power OFF-light OFF; ( 0 ) 0.1 M, (11) 1 M, (e) 2 M, (A) 3.5 M).


3.1. Effect of NaCl concentration on the photoelectrochemical hydrogen production The effect of NaC1 concentration on the photoelectrochemical hydrogen production without BR was investigated by changing the NaCI concentration of the solution from 0.1 to 3.5 M in control experiments. As seen in Fig. 2a the rate of hydrogen production and the total amount of H2 produced were increased by increasing the NaC1 concentration of the solution. The following reac-

V. Sediroglu et al./'Journal of Biotechnology 70 (1999) 115-124

tions are expected to occur at the electrodes: At the cathode, 2H20 + 2e- ---,H 2 + 2 O H -

E ~ =

-

0.83

V

(1) At the anode, Ag + C1- ~ AgC1 + e -

E ~ = - 0.22 V

(2)

and/or 2Ag + 2 O H - ~ Ag20 + H 2 0 + 2eE ~ = - 0.34 V

(3)

In all experiments, however, hydrogen could not be produced below - 0 . 8 8 V. As it is known from the literature, the solubility of silver chloride in water is very small. However, its solubility increases with increasing NaC1 concentration. In addition to that alkali chlorides convert silver oxides into silver chloride (Mellor, 1963). Therefore, at high NaC1 concentrations, chloride ions covering the silver surface may dissolve silver oxide formed on the anode by the following reactions" Ag20 + 2C1- + H20 ~ 2AgC1 + 2 O H -

(4)

AgC1 + C1- *-~AgClz-

(5)

or

AgCI*-~ Ag + + C1 -

(6)

Meanwhile, the following photoreaction might help the regeneration of Ag electrode (Mellor, 1963). hv

2AgC1 + 2 O H - ~ H 2 0 2 + 2Ag + 2C1-

(7)

The surface analysis of silver electrodes by scanning electron microscopy showed the formation of Ag20 on the silver electrodes after all the photoelectrochemical hydrogen production experiments. The lower hydrogen production rate observed at low NaC1 concentrations was attributed to the accumulation of Ag20 on the silver electrode surface. Also the cessation of hydrogen production after 400 min might be attributed to Ag20 accumulation on the electrode. However, it should be emphasized that H 2 could not be produced by replacing the Ag electrode with a new one, unless the Pt electrode was also changed. All

119

these findings imply that a thin hydrogen layer adsorbed on the Pt electrode might cause a large resistance, thus stopping the hydrogen production. Fig. 2b illustrates the change of pH versus time. As H2 was produced, the pH of the solution increased because O H - ions were produced according to reaction (1). It is interesting to note that minimum pH increase was observed at 3.5 M NaC1 concentration, although the maximum H 2 production was achieved under this condition (even the initial pH of the solution was higher). This might be due to consumption of hydroxyl ions by Ag at the anode, as indicated in reaction (3). This might explain why the rate of pH increase was higher in low NaC1 concentration compared to that in high NaC1 concentration. In the light of the observations above, 3.5 M NaC1 solution has been selected for photoelectrochemical hydrogen production. Fig. 2c illustrates the change of current while the system was in operation. The current decreased during the first 200 min of H2 production and then remained almost constant. Current direction was reversed after power was turned off.

3.2. Effect of H. halobium on the photoelectrochemical hydrogen production The results of experiments carried out with free or immobilised H. halobium packed cells on CAM or the purple membrane immobilised on CAM were compared with the control experiment. As it is seen from Fig. 3a, the H2 production is enhanced by increasing the amount of BR. It seems that BR affects the pH of the solution as it is illustrated in Fig. 3b. The highest production of hydrogen was obtained with the immobilised system containing the highest amount of BR (0.098 lamol). The pH increase was the lowest in this system. Upon illumination BR pumps out protons and most probably this process is accompanied by deprotonation of BR (Yiicel et al., 1995). Consequently, BR has a buffering effect on the solution, thus enhancing the hydrogen production. BR was found to be more effective in the immobilised system than in the free system (which might be due to the synergetic behaviour of BR in the


120

immobilised system). The immobilisation might have increased the effective utilisation of light energy. In immobilised systems the membranes are directly and continuously illuminated because they are kept in place in the reactor. Whereas the free Halobacterium halobium cells in the suspension may be exposed to a less amount of light flux due to improper orientation while stirring. Photoelectrochemical hydrogen production was even higher with PM fragments immobilised on CAM.

27 25

a)

~,23 ~'19 17 150

'

*

i

200

400

600

&

800

'nn~min)

8.6

8.4 25.0

:~ 8.2

23.0 ~

b)

a)

"8.0

21.0

7.8

19.0

7.6,.

-

~

17.0

8 0

15.0 0

200 K

400

600

800

0.8

Time(min)

0.6

8.60

~

8.40

'0.0 ~ -0.2

8.20

~ .0.4

8.00

.0.6

_ _ .

~

c)

0.4 0.2 ,

) ~

8 0

Q.

.0.8 7.80

A

Time(rain)

7.60 0

100

A 0.80 0.60 0.40

200

300

400

500

600

700

Time(rata)

........................................................................ c)

0.20

Fig. 4. Effect of immobilized H. halobium packed cells in PAG (a) on rate of hydrogen production, in terms of pressure (b) pH change with respect to time (c) change of current versus time (T Power ON-light ON; ], power OFF-light OFF); (O) control; (M) PAG-Control; (@) PAG (BR = 0.098 lamol); ( 9 PAG (BR = 0.196 lamol); ( 9 PAG (BR = 0.294 lamol).

0.00

.~ .0.20 -0.40

-0.60 .0.80

Fig. 3. Effect of immobilization of BR on CAM. (a) On the rate of hydrogen production, in terms of pressure. (b) pH change with respect to time. (c) Change of current versus time. (up arrow) Power ON-light ON; (down arrow) power OFFlight OFF; (solid line) control; H. halobium packed cells immobilized on CAM (rectangle) BR = 0.033 lamol, (square) BR = 0.067 l~mol, (circle) BR = 0.098 lamol; PM immobilized on CAM (diamond) BR =0.067 ~tmol; free H. halobium packed cells (triangle) BR = 0.098 lamol.

Hydrogen production with H. halobium packed cells immobilised in PAG, having different amounts of total BR (ranged from 0.098 to 0.294 lamol), were compared with two control experiments. The light induced deprotonation of BR immobilised in PAG can be clearly seen in Fig. 4b. Although gas was produced at the beginning of the runs, a slight drop in pH was still observed. After some time, however, the pH increased. It was observed that the rate of hydrogen production was almost the same within the first 50 min for flesh and aged membranes of both types.


As the experiments proceeded, however, a slight decrease in the hydrogen production for the aged membranes compared to the fresh ones was observed, probably due to the loss of surface deprotonation capacity by stabilisation of biosynthetic membranes. Results are compared on the basis of the hydrogen production ratio (HPR), average and initial hydrogen production rate, activity and power requirements. The hydrogen production ratio is defined as the ratio of the difference between the total number of moles of H 2 produced in the presence and absence of BR within a certain period of time, divided by the number of moles of BR present in the system. As seen in Table 1, HPR obtained by using free H. halobium packed cells was lower than the immobilised systems containing the same amount of BR. The immobilised systems were apparently more effective for the utilisation of the light energy. The highest hydrogen production ratio was obtained with immobilised PM on CAM. Initial rates is defined as the number of moles of hydrogen produced per hr within the first 50 min, and average rate is defined as the number of moles of hydrogen produced per hour within 520 min. Activity is defined as the average rate per mg cell. As expected, the initial rate increased as the concentration of BR increased. However, for the free H. halobium packed cells an increase of approximately one order of magnitude in the concentration of BR resulted only in a slight increase of the initial rate. For the same BR concentration (0.098 gmol BR) the initial rate was considerably higher for the H. halobium packed cells immobilised on CAM compared to that of the free H. halobium packed cells. Immobilisation capacity of CAM was found to be limited between 0.033 and 0.098 gmol BR (Sediroglu et al., 1998). The maximal amount of packed cells immobilised in PAG system resulted in amounts of BR (0.294 gmol BR) that were three times higher than that of the CAM system. Higher BR amounts immobilised in PAG led to higher hydrogen production rates, as shown in Table 1. Since the rate of hydrogen production decreases with time, the average hydrogen production rate was smaller than the corresponding initial rate, as

121

given in Table 1. The change in the average rates with increasing BR amount showed a similar trend as the initial rates of the corresponding systems. For the same systems, the percentage change in the average values of the amount of BR was approximately the same as the percentage change shown in the initial values. The maximum activity obtained was almost 1.5 times higher than that obtained by Khan and Bhatt (1992). Power requirement per mole of hydrogen produced was calculated from the current and voltage data obtained during the experiments, on the basis of either the initial or the average rate of hydrogen production, as given in Table 1. Power requirements based on the average rate of hydrogen production was about three times that of the ones based on the initial rates. It was remarkably higher in the absence of BR and it decreased as the concentration of BR increased. It should be noted that the minimum power consumption was observed with PM fragments immobilised on CAM. The light induced protonation of BR had decreased the power consumption per mole of hydrogen produced from 4.5 to 1 W mol-~ h-1 when the control experiment was compared with the experiment carried out with PM fragments immobilised on CAM. The power requirement was decreased by 30% if PM was immobilised on CAM instead of packed cells containing the same amount of BR (BR = 0.067 gmol). This indicates the advantage of using PM instead of packed cells due to the following reasons; 1. packed cells contain all cellular components, however PM contain only lipids and BR. 2. the size of the packed cells are greater than PM. Therefore self orientation of PM on CAM might be much better than the packed cells, and utilisation of light energy might thus be better for PM fragments. 3. present experiments revealed that the immobilisation of PM fragments is easier than the immobilisation of packed cells on CAM. Although, long term stability of packed cells has not been reported yet, some experiments of ours implied that the stability of PM fragment is higher than the packed cells.

I

N

Table 1 Summary of the experimental results Conditions"

BR used (pmol)

HPR

Average rate (pmol h - I )

Initial rate (pmol h-I)

Activity (pmol h - ' mg-')

Power based on initial rate (w mol-' h-I)

Power based on average rate ( w mol-' h-I)

S

ia B

a

QQ

a

-

227.78

-

26.28

136.66

-

4.24

17.54

2

I .600 0.760 0.098

20 228 9840 1350

512.50 469.79 455.55

177.95 318.43 2324.18

59.13 54.20 52.56

324.58 281.87 256.25

0.016 0.029 0.190

1.94 2.16 2.18

9.96 I 1.45 8.65

3

0.098 0.067 0.033 0.067

I350 900 450 900

683.33 605.03 320.31 697.56

4648.46 5630.59 2803.93 701 1.64

78.84 69.81 36.95 80.48

427.08 375.83 196.45 504.87

0.315 0.417 0.438 0.588

1.35 1.35 2.78 1.01

6.45 6.15 10.92 4.47

529.30 495.40 410.00 230.62

0. I56 0.148 0.303

1.13

4.98

1.15 1.31 1.68

5.10 7.43 10.13

CON

-

BI B2 B3 CI c2 c3 C4' PI P2 P3 P4d

Cell wet weight Total H, proinitially usedh duced (pmol) (mg)

0.294 0. I96 0.098 -

4050 2700 I350 -

811.45 740.27 633.50 313.19

1985.27 2614.74 4140 -

93.62 85.41 73.09 36.13

-

a Where CON denotes control experiments, B denotes free H. holohiurn packed cells, C denotes immobilised packed cells on CAM and P denotes immobilised packed cells in PAG. Density of packed cell is I g ml- I . Denotes immobilised PM fragments on CAM. Denotes control experiments with PAG.

,

5% B

gn

20

6

2

2 +

-.\

2

h

s


However, packed cells are cheaper than PM fragments since PM fragments require sophisticated purification techniques that might be quite expensive and time consuming. Based on overall evaluation of these factors, further investigation is required as to whether PM fragments might be more economical than using the packed cells or not.

123

for the photoelectrochemical hydrogen production. It should be also noted that the regeneration possibility of the system, the stability of the biological membranes, the increase in hydrogen production rate and decrease in power requirement make the photoelectrochemical hydrogen production a promising new system. However, further investigations are needed for the optimisation of this system for any practical application.

4. Conclusion Acknowledgements The minimum voltage required for the hydrogen production was 0.88 V in the bioelectrochemical reactor designed and constructed for the present work. The presence of BR, either in the form of H. h a l o b i u m packed cells or as PM fragments, enhances the hydrogen production rate upon illumination. Hydrogen production also increased as the BR concentration increased. BR does not lose its photoactivity on the membrane surface of CAM and PAG for at least 2 months. Drastic improvement of the hydrogen production rate in the immobilisation systems compared to free H. h a l o b i u m packed cells in NaC1 solution has been observed. Larger amounts of BR can be immobilised in PAG than on CAM. One of the most promising results of the present work is the low power requirement per mole of hydrogen produced. The minimum power requirement is obtained with the photoelectrochemical system containing PM fragments immobilised on CAM. This power requirement is about one fourth of the control experiment, which did not contain H. h a l o b i u m . In the present system, NaC1 solution, which is cheap and abundant, is used instead of expensive alkaline solutions. These findings considerably increase the future potential of photoelectrochemical hydrogen production. It should be emphasised that the presence of BR does not only regulate the pH of the system, thus causing increased hydrogen production rate, but also causes light-induced proton dissociation, which may lower the required electrical power for the electrochemical conversion. The system presented in this study can be exploited in a device which can utilise light energy and electrical energy

This research is supported by M E T U Research Fund, project number AFP 96-07-02-02 and by Turkish Scientific and Research Council, project number TBAG 15 35.

References Birge, R.R., 1990. Nature of the primary photochemical events in rhodopsin and bacteriorhodopsin. Biochem. Biophys. Acta 1016, 293-327. Ero~lu, |., Aydemir, A., Tiirker, L., Yficel, M., 1994. Photoresponse of bacteriorhodopsin immobilised in polyacrylamide gel membrane. J. Membr. Sci. 86, 171-179. Hall, D.O., Markov, S.A., Watanabe, Y., Rao, K.K., 1995. The potential applications of cyanobacterial photosynthesis for clean technologies. Photosynth. Res. 46, 129-167. Kaya, B., Giindiiz, U., Ero~,lu, i., Tiirker, L., Yiicel, M., 1996. Hydrogen gas production using coupled enzyme systems from E. coli and H. halobium. In: Proceedings of l lth World Hydrogen Energy Conference, vol. 3. Stuttgart, Germany, pp. 2595-2600. Khan, M.M.T., Bhatt, J.P., 1989. Light dependent hydrogen production by H. halobium MMT22 coupled to E. coli. Int. J. Hydrog. Energy 14, 643-645. Khan, M.M.T., Bhatt, J.P., 1990. Photoelectrochemical studies on H. halobium or continuous production of hydrogen. Int. J. Hydrog. Energy 15, 477-480. Khan, M.M.T., Bhatt, J.P., 1992. Large scale photobiological solar hydrogen generation using H. halobium MMT22 and silicon cell. Int. J. Hydrog. Energy 17, 93-95. Khan, M.M.T., Bhatt, J.P., 1997. Photosensitized continuous production of hydrogen by Halobacterium halobium MMT22 coupled to Escherichia coli. Int. J. Hydrog. Energy 22, 995-997. Markov, S.A., Bazin, M.J., Hall, D.O., 1995. The potential of using cyanobacteria in photobioreactors for hydrogen production. Adv. Biochem. Eng. 52, 60-86. Mellor, J.W., 1963. Supplement to Mellor's Comprehensive Treatise on Inorganic and Theoretical Chemistry, vol. 2. Longrnans, London, p. 390.

124


Oesterhelt, D., Stoeckenius, J.W., 1971. Rhodopsin-like protein from the purple membrane of H. halobium. Nature New Biol. 233, 149-154. Oesterhelt, D., Stoeckenius, J.W., 1974. Isolation of the cell membrane of H. Halobium and its fraction into red and purple membrane. Methods Enzymol. 31, 667-671. Patel, S., Madamwar, D., 1994. Photohydrogen production from a coupled system of H. halobium and P. valderianum. Int. J. Hydrog. Energy 19, 733-738. Sediroglu, V., Gfindfiz, U., Yficel, M., Tfirker, L., Erofglu, i., 1996a. Photoelectrochemical H 2 production by H. halobium. In: Proceedings of l lth World Hydrogen Energy Conference, vol. 3. Stuttgart, Germany, pp. 2589-2594. Sediroglu, V., Giindiiz, U., Yiicel, M., Tiirker, L., Er@,lu, i., 1996b. Modeling of long term photo response of bacteri-

orhodopsin immobilized on cellulose acetate membranes. J. Membr. Sci. 113, 65-71. Sediroglu, V., Yficel, M., Gfindfiz, U., Tfirker, L., Ero~lu, i., 1998. The effect of Halobacterium halobium on the photoelectrochemical hydrogen production. In: Zaborsky, O.R. (Ed.), BioHydrogen, Plenum Press, New York, pp. 295-304. Shreve, N.R., 1984. Chemical process industries, fifth ed. McGraw-Hill, New York, pp. 98. Trissl, H.W., 1990. Photoelectric measurements of purple membranes. Photochem. Photobiol. 51 (6), 793-818. Yficel, M., Zabut, B.M., Er@,lu, [., Tfirker, L., 1995. Kinetic analysis of light induced proton dissociation and association of bacteriorhodopsin in purple membrane fragments under continuous illumination. J. Membr. Sci. 104, 65-72.

i

JOURNAL

i

i

i

O F



Identification of by-products in hydrogen producing bacteria; Rhodobacter sphaeroides O.U. 001 grown in the waste water of a sugar refinery Deniz Ozgfir Yi~,it

a

Ufuk Gfindfiz b , Lemi Tiirker c,, Meral Yiicel inci Ero~,lu d

b

Department of Biotechnology, Middle East Technical University, 06531 Ankara, Turkey b Department of Biology, Middle East Technical University, 06531 Ankara, Turkey c Department of Chemistry, Middle East Technical University, 06531 Ankara, Turkey d Department of Chemical Engineering, Middle East Technical University, 06531 Ankara, Turkey a


Abstract Rhodobacter sphaeroides O.U. 001 is able to produce hydrogen anaerobically upon illumination. The cells were screened for the presence of valuable by-products such as poly-13-hydroxy (PHB) butyric acid aiming to improve the feasibility of the system. Also waste water from a sugar refinery was used for bacterial growth to further increase the feasibility. Under aerobic conditions the standard growth media containing L-malic acid and sodium glutamate in 7.5/10 and 15/2 molar ratios and a medium containing 30% waste water from sugar refinery were used. In this case the maximum concentration of PHB produced were approximately 0.2 g 1-~ in both of the standard media whereas it was 0.3 g I-~ in medium containing 30% waste water. By using the medium containing 30% waste water, PHB and hydrogen productions were determined under anaerobic conditions. The maximum concentration of PHB produced was around 0.5 g 1-~ and the amount of gas collected was 35 ml in 108 h. From these results it can be concluded that PHB can be collected during hydrogen production. The use of waste water from sugar refinery increased the yield. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Rhodobacter sphaeroides; Hydrogen; Poly-13-hydroxy butyric acid; Waste water

I. Introduction

There are limited reserves of fossil fuels on earth so there is a strict need for d e v e l o p m e n t of * Corresponding author. Fax: + 90-312-2101-280. E-mail address: [email protected] (L. Tiirker)

renewable and clean energy sources. Since solar energy seems to be the m o s t suitable starting point m u c h attention has been given to the design of systems capable of t r a n s f o r m i n g solar energy into high energy c o m p o u n d s such as hydrogen, which is a clean and highly efficient fuel that results in water upon burning without generating CO2.


126

D.O. Yi~it et al. ,'Journal of Biotechnology 70 (1999) 125-131

Photobiological hydrogen production attracts much attention because biological transformation systems are simple in operation, stable, low energy consuming and renewable. Feasibility of a photobiological hydrogen production system can be increased by the use of cheaper substrates for the microorganism such as the waste water and the collection and recycling of useful by-products other than hydrogen. Rhodobacter sphaeroides O.U. 001 is able to produce hydrogen anaerobically in a photobioreactor (Ank et al., 1996). It was shown that waste water from sugar refinery can partially replace the C source of bacterial growth medium (Yeti~ et al., 1998). It was reported that R. sphaeroides O.U. 001 cells contain valuable by-products (Yamagishi, 1995). Poly-[3-hydroxy butyric acid (PHB) is one of the by-products which is a biodegradable thermoplastic that can be synthesized during unfavorable growth conditions by a wide range of bacteria (Byrom, 1987; Page, 1989; Kim et al., 1994). PHB has industrial applications; it can be used to construct biodegradable carriers for long term dosage of herbicides and insecticides, packaging containers, bottles and bags (Anderson and Dawes, 1990). PHB has also very important medical applications; it can be used as a biodegradable carrier for long term dosage of drugs and can be used to construct surgical pins, sutures, staples, swabs and it can also be used for wound dressing (Abe and Doi, 1992; Sang, 1996). The aim of the study was to test different parameters to optimize the conditions of a hydrogen production system for maximum feasibility. R. sphaeroides O.U. 001 cells were therefore screened for PHB accumulation. Bacterial growth was determined under aerobic conditions in standard growth media (Biebl and Pfennig, 1981) containing L-malic acid and sodium glutamate in 7.5/10 and 15/2 molar ratios and concentrations of PHB accumulated at different phases of growth were determined. It was shown before that bacterial growth was optimum at the ratio of 7.5/10 and hydrogen production was maximum at 15/2 (Ero~lu et al., 1997) so PHB production was determined at these ratios. To reduce the cost of growth medium; a medium containing 30% waste

Table 1 Contents of standard growth medium having different L-malic acid to sodium glutamate ratios L-malic acid/sodium glutamate (raM m M - 1 ) NB content (g 1-1) KH2PO 4 MgSO4.7H20 L-Malic acid Sodium glutamate NaC1 CaCI2.2H20 Niacin Thiamin Trace element solution 7 Ferric-citrate solution

7.5/10 0.5 0.2 1.0 1.8 0.4 0.05 0.0005 0.0005 1 ml

15/2 0.5 0.2 2.0 0.36 0.4 0.05 0.0005 0.0005 1 ml

5 ml

5 ml

water from sugar refinery was tested. Bacteria and PHB concentrations were determined and it was seen that the use of waste water increased the yield of PHB. When the experiment was repeated under anaerobic conditions it was observed that PHB was accumulated in a hydrogen producing system which can utilize waste water from sugar refinery so that the cost of growth medium can be decreased and valuable by-products can be collected to increase the feasibility.

Table 2 Contents of growth medium including 30% waste water from sugar refinery N B content

g 1-l

KH2PO 4 MgSO4-7H20 L-Malic acid Sodium glutamate NaC1 CaC12.2H20 Niacin Thiamin Trace element solution 7 Ferric-citrate solution Pretreated waste water

0.5 0.2 0.54 0.8 0.4 0.05 0.0005 0.0005 1 ml 5 ml 300 ml

127

D.O. Yi~it et al. /'Journal of Biotechnology 70 (1999) 125-131

1.8

1.6

1.6

1.4

1.4

- 1.2

E ~.2 t-. C:) 1 t,D (,C) ~ , 0.8 a

- 0.8 -0.6

O o.6 0.4

'--.~I,--OD (660 nm)

0.2

(a)

r"-

~Dry

0 20

40

60

TimeS~hr)

0,2 0,18 0,16

Wt (g/L)

0.4 0.2 0

loo

12o

140

160

0,19~.

~" o,14 ~ 0,12 rn

0,1

N.~ 0,08 0,06

0,04 0,02 (b) 0

~

v

,

0

9

20

,

40

i

,

60

80

,

100

',

120

Time (hr) Fig. 1. The growth curves and poly-13-hydroxy (PHB) production curves in standard growth medium containing L-malic acid and sodium glutamate having 7.5/10 and 15/2 molar ratios under aerobic conditions. (a) Bacterial growth curve based on both OD at 660 nm and dry weight in growth media having 7.5/10 k-malic acid and sodium glutamate molar ratio. (b) PHB production (g 1- 1) in growth media having 7.5/10 L-malic acid and sodium glutamate molar ratio. (c) Bacterial growth curve based on both OD at 660 nm and dry weight in growth media having 7.5/10 L-malic acid and sodium glutamate molar ratio. (d) PHB production (g 1-1) in growth media having 15/2 L-malic acid and sodium glutamate molar ratio.

2. Materials and methods R. sphaeroides O.U. 001 ( D S M 5648) was grown in 50 ml flasks and anaerobic jars both aerobically and under argon atmosphere (anaerobically) at 36~ under illumination at 200 W m - 2 . Both the standard growth media containing L-malic acid and sodium glutamate having 7.5/10 and 15/2 molar ratios and a medium containing 30% waste water were used. Waste water was

obtained from sugar refinery at Etimesgut, Ankara. It was pretreated by incubation at 95~ for 45 min followed by filtration through cheese cloth. Then it was centrifuged at 3000 rpm for 30 min and filtered through glass prefilters and 0.45 lam W h a t m a n n filter paper. Its p H was adjusted to 7.0 and sterilized by autoclaving. It was stored at - 20~ until it was added to the growth media. The c o m p o n e n t s of the growth media are shown in Tables 1 and 2. Initial p H of the growth media

128

D.O. Yi~it et al./Journal of Biotechnology 70 (1999) 125-131 1,2

1,8 1,6 1,4

E r

C) CO r E3

0

1,2

0,8

r

1

- 0,6

0,8 0,6

I"-

- 0,4

0,4 ---e--

0,2

(c)

OD (660 nm)

- 0,2

---II-- Dry Wt (g/L)

0

9

20

,

L

40

60

!

80

100

120

140

0 160

Time (hr)

...J O) ~,~ rn "r-9 Q. (d)

0,2 0,18 0,16 0,14 0,12 0,1

0,18 g/L

0,08 0,06 0,04 0,02 0

0

I

I

I

|

20

40

60

80

,,=

i

100

Time (hr) Fig. 1. (Continued) were adjusted to 6.8-7. Growth curves were constructed by OD and dry weight measurements. Concentration of bacterial PHB was determined by the method of Bowker (1981). After the given incubation period cells were collected by centrifugation, digested by sodium hypochlorite treatment, PHB was extracted by hot chloroform treatment and it was converted to crotonic acid by sulfuric acid addi-

tion. Their OD were measured using a spectrophotometer. Concentration and %(w/w) PHB was calculated using the standard curve calibrated.

3. Results and discussion The growth of R. sphaeroides O.U. 001 and

129

D.O. Yi~it et al./Journal of Biotechnology 70 (1999) 125-131 0,35 0,3 0,25 ._J

0,2 0,15

"!"

0,1

Q.

0,05 ,,

0

20

,

- ,

40

T

60

-

,

100

80

Time (hr) Fig. 2. Poly-13-hydroxy (PHB) production (g 1-1) in growth medium containing 30% waste water from sugar refinery under aerobic conditions. 0,6 0,50g/L

0,5 .._1

0,4

1:~ 0,3 nn "1" 0,2 v

13_

0,1 -~ 0

-

" t

20

-

t

t

40

60

"

Time (hr)

t

80

-..

~' 100

"

120

Fig. 3. Poly-13-hydroxy (PHB) production (g 1-1) in growth medium containing 30% waste water from sugar refinery under anaerobic conditions. 35 30 25 20

E ~ "1-

15

10 5

0 0

20

40

60

Time (hr)

80

I00

120

Fig. 4. Collected gas amount (ml) in growth medium containing 30% waste water from sugar refinery under anaerobic conditions.

130

D.O. Yi~it et al./Journal of Biotechnology 70 (1999) 125-131

Table 3 Poly-[3-hydroxy (PHB) accumulation in different media Growth medium

Time (h)

Max PHB concentration (g 1-1)

% PHB (w/w)

15/2 7.5/10 30% ww (aerobic) 30% ww (anaerobic)

96 96 84 96

0.2 0.2 0.3 0.5

19.8 16.8 52.4 70.4

PHB production in standard growth medium containing L-malic acid and sodium glutamate at two different molar ratios (7.5/10 and 15/2) under aerobic conditions are shown in Fig. 1. It is seen that the maximum amounts of PHB accumulated are0.19 g l - 1 (16.8%PHB w/w) a n d 0 . 1 8 g l - I (19.8% PHB w/w) for 7.5/10 and 15/2 molar ratios, respectively. In the growth medium, containing L-malic acid and sodium glutamate at molar ratio of 15/2, the percentage of PHB is higher compared to the growth medium having 7.5/10 molar ratio. In both media PHB production was observed in a stationary phase. It was reported that PHB is produced during the stationary phase of the growth as a storage material under stress conditions; to be used during starvation (Yamagishi, 1995). PHB production curve in growth medium containing 30% waste water under aerobic conditions is shown in Fig. 2. It is seen that the amount of PHB accumulated is 0.30 g 1-~ and 52.4% PHB (w/w). Growth medium containing 30% waste water has been adjusted to contain nearly the same C/N molar ratio as standard growth 4 3.5 3 2.5 O3

2 1.5 1 0.5 0

v

v

0

50

Time (hr)

~

T.

y

100

v

150

Fig. 5. Ratio of hydrogen to poly-J3-hydroxy (PHB) production in growth medium containing 30% waste water under anaerobic conditions

medium containing L-malic acid and sodium glutamate at 15/2 molar ratio; since waste water also contains carbon 1.46 g l-1 L-malic acid is saved. The maximum amount of PHB produced is higher when waste water is present in the growth media. The reason for this may be due to the stress caused during utilization of carbon from waste water. In addition waste water may also contain some components which may cause stimulation of PHB production. PHB production in growth medium containing 30% waste water under anaerobic conditions is shown in Fig. 3. It is seen that the maximum amount of PHB accumulated is 0.50 g 1-l which corresponds to 70.4% PHB (w/w). This shows that PHB is also produced under anaerobic conditions and the production is even higher than aerobic conditions. It was also reported that PHB is accumulated in large quantities inside the bacterial cellular material of photosynthetic bacteria when cultured under anaerobic conditions (Yamagishi, 1995). R. sphaeroides O.U. 001 produces hydrogen gas under anaerobic conditions. In Fig. 4 the amount of collected gas under anaerobic conditions at time intervals in growth medium containing 30% waste water is shown. It is seen that 34.8 ml gas was collected in 108 h. The hydrogen production rate is nearly 0.006 1 l - l h - l ; and yield of hydrogen is 0.46 1 g-~. When this hydrogen production rate is compared with the literature data it is seen that 0.047 1 1-~ h -~ hydrogen was produced under optimized conditions (Arlk et al., 1996). On the other hand 0.006 1 1-1 h-~ hydrogen was produced by the use of waste water (Sasikala et al., 1992) and 0.042 1 1-~ h - l hydrogen was produced when artificial waste was used (Kitjama et al., 1997). The ratio of hydrogen to PHB production in growth medium containing 30%

D.O. Yi~it et al./Journal of Biotechnology 70 (1999) 125-131

waste water under anaerobic conditions is shown in Fig. 5. From this figure, it is evident that PHB production competes with light dependent hydrogen production. The ratio suddenly decreases after 60th h which is the starting point of PHB production. The reason for this competition is that PHB production and hydrogen production use the same reducing power that results from the metabolism of organic acids (Yamagishi, 1995). PHB accumulation in different growth media is summarized in Table 3. When these results are compared with the literature data it is seen that Alcaligenes latus produces the highest amounts of polyhydroxy alkonates (88% w/w PHA) (Steinbtichel and Ffichtenbusch, 1998). R. sphaeroides produces 24% (w/w) PHB by using acetate as carbon source (Gross et al., 1992). The important point is that PHB can be collected from the hydrogen producing photobioreactor that utilizes waste water from sugar refinery to increase the feasibility of the total system. The cells and their growth medium would also be screened for other possible by-products such as carotenoids, bacteriochlorophyll and ubiquinone.

References Abe, H., Doi, Y., 1992. Controlled release of lastet, an anticancer drug, from poly(3-hydroxybutyrate) microspheres containing acylclycerols. Macromolec. Rep. A29, 229-235. Anderson, A.J., Dawes, E.A., 1990. Occurrence, metabolism, metabolic role, and industrial uses of bacterial poly hydroxy alkanolates. Microbiol. Rev. 54, 450-472. Ank, T., Gtindiiz, U., Yticel, M., Tiirker, L., Sedirofglu, V., Erofglu, I., 1996. Photoproduction of hydrogen by Rhodobacter sphaeroides O.U.00. Proceedings of the 1l th World Hydrogen Energy Conference, Vol. 3, Stuttgart, Germany, pp. 2417-2424. Biebl, H., Pfennig N., 1981. Isolation of member of the family Rhodosprillacaea. The Prokaryotes. Springer-Verlag, New

131

York, pp. 267-273. Bowker, R.R., 1981. Manual of Methods for General Bacteriology. American Society for Microbiology, Washington DC. Byrom, D., 1987. Polymer synthesis by microorganisms; technology and economics. Trend Biotechnol. 5, 246-250. Ero~lu, i., Asian, K., Gtindiiz, U., Yficel, M., Ttirker, L., 1997. Continuous hydrogen production by Rhodobacter sphaeroides OU 001. Paper presented at Biohydrogen 97 Int. Conference on Biological Hydrogen Production, Kona Hawaii, USA, pp. 143-149. Gross, A.R., Ulmer, H.W., Lenz, R.W., Tshudy, J.D., Udon, C.P., Brandt, H., Fiiller, R.C., 1992. Biodeuteration of poly ([3-hydroxybutyrate). Int. J. Biol. Macromol. 14, 3340. Kim, B.C., Lee, S.C., Lee, S.Y., Chang, H.N., Chang, Y.K., Woo, S.I., 1994. Production of poly(3-hydroxybutyric acid) by fed-batch culture of Alcaligenes eutrophus with glucose concentration control. Biotechnol. Bioeng. 43, 892-898. Kitjama, Y., E1-Shishtawy, R., Ueno, Y., Otsuka, S., Miyake, J., Moritomo, M., 1997. Analysis of compensation point of light using Plane-Type photosynthetic bioreactor. Proceedings of the Int. Conference on Biological Hydrogen Production, Kona, HI, USA, pp. 359-367. Page, W.J., 1989. Production of poly-13-hydroxybutyrate by Azotobacter vinelandii strain UWD during growth on molasses and other complex carbon sources. Appl. Microbiol. Biotechnol. 31, 329-333. Sang, Yup Lee, 1996. Bacterial Polyhydroxyalkanoates. Biotechnol. Bioeng. 49, 1-14. Sasikala, K., Ramana, C.H.V., Roa, P.R., 1992. Photoproduction of hydrogen from the waste water of a distillary by Rhodobacter sphaeroides OU 001. Int. J. Hydrogen Energy 17, 23-27. Steinbtichel, A., Fiichtenbusch, B., 1998. Bacterial and other biological systems for polyester production. TIBTECH 16, 419-426. Yamagishi, K., 1995. Interim Evaluation Report of Development of Environmentally Friendly Technology for the Production of Hydrogen. New Energy and Industrial Technology Development Organization, Japan. Yetis, M., Gtindiiz, U., Ero~lu, i., Yficel, M., Tiirker, L., 1998. Photoproduction of Hydrogen from waste water of a sugar refinery by Rhodobacter sphaeroides OU 001. Paper to be presented in Marine Bioprocess Engineering, Noordwijkerhout, The Netherlands.


Bio ELSEVIER

ology


Cell cultures from marine invertebrates" obstacles, new approaches and recent improvements Baruch Rinkevich * National Institute of Oceanography, Tel Shikmona, P.O. Box 8030, Haifa 31080, Israel


Abstract

The establishment of cell lines from marine invertebrates has been encountered with obstacles. Contrary to insects and arachnids where the development of a variety of cell lines has become routine, there is no single established cell line from marine invertebrates. This review examines the activity in the field of marine invertebrate cell cultures within the last decade (1988-1998). During this period, attempts (90 peer reviewed studies in addition to many other abstracts, chapters in books, symposia presentations and reports) were limited to a few species within only six phyla (Porifera, Cnidaria, Crustacea, Mollusca, Echinodermata, Urochordata: in addition to freshwater/terrestrial annelids and platyhelminths). These studies which are summarized here, on one hand indicated ubiquitous problems and on the other, unique characterizations to each phylum studied. Only one-third of the studies revealed cultures of 1 month or longer but most of these were long-term cultures found or suspiciously considered to be contaminated by other unicellular eukaryotic organisms, mainly by thraustochytrids. Three unique approaches/obstacles for marine invertebrate cell cultures (source of cell, cryopreservation and eukaryotic contaminants) are further discussed. The overall impact of recent improvements and developed protocols raises the suggestion for testing different, novel routes in the establishment of cell cultures from marine invertebrates. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Marine invertebrates; Cell lines; Platyhelminths

I. Introduction

Cells under in vitro conditions are used as exceptionally important tools in a variety of scientific disciplines including biological and medical sciences. In vitro applications may also be used as alternative tools for animal experimentation, for biotechnological applications and pathological in* Tel.: + 972-4-851-5202; fax: + 972-4-851-1911. E-mail address: [email protected] (B. Rinkevich)

vestigations. Invertebrates may be regarded as a major source for such applications. They comprise more than 95% of the animal species and, within them, marine invertebrates are thought to account for more than 30% of all animal species ( N a g a n u m a et al., 1994). The large group of marine invertebrates, which encompasses more than 20 different phyla, represent a rich source of cell and tissue types that significantly differ from one group to another. Many cell types from a variety of invertebrates possess extensive morpho-

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00067-X

134

B. Rinkevich /'Journal of Biotechnology 70 (1999) 133-153

genetic potentialities (multipotency, totipotency, including neoplasia; Rinkevich et al., 1994; Rosenfield et al., 1994; Rinkevich and Rabinowitz, 1997). This leads to high in vivo plasticity of shapes, structures, cell replacements, proliferation processes and cell lineage in different invertebrate taxa, sometimes differs significantly even between systematically related groups of organisms. Attempts to maintain and grow invertebrate cells in vitro were made quite early in the history of tissue culture, nearly 100 years ago (Gomot, 1971; Rannou, 1968, 1971). However, while there have been more than 200 cell lines established from insects and ticks (Mitsuhashi, 1989), all efforts to develop permanent and proliferative cell cultures from marine invertebrates have been unsuccessful (Rinkevich et al., 1994). Moreover, contrary to the considerable attention that has been given to the in vitro use of marine fish and plants, the literature points out to only limited activity directed towards marine invertebrate cell cultures (Rosenfield, 1993)--largely due to the fact that experimental failures are not suitable for publication in most scientific journals. Within recent years, there has been veritable activity of many scientists in marine invertebrate cell cultures/cell lines. This stems from the understanding that basically, all cells of different taxa within the kingdom Animalia are the same, having similar nutrient requirements, are controlled by the same developmental and physiologicalbiochemical pathways and are under the expression of identical genes. It is therefore plausible that techniques developed in one scientific discipline may be adapted (with some alterations) to other fields, so that the outcome will serve the new requirements. For example, mitogens that have important roles in cells derived from mammals have been found to alert similar activities in invertebrate cells in vitro (Peddie et al., 1995; Raftos and Cooper, 1996). Such approaches that allow creative novel designs for the development of cell cultures from marine invertebrates, gave new dimensions to this field, providing new results, ideas and improvements to this relatively stagnant field. The present communication aims in the summarizing of attempts for cell cultures from a

variety of marine invertebrates within the last l0 years (1988-1998). General trends in the results, common obstacles, the new approaches used and recent achievements are reviewed with an eye for formulating, if possible, a more ubiquitous framework for marine invertebrates cell culture technology.

2. Activity during the last decade During the last decade, attempts to develop cell cultures from marine invertebrates were limited to a few species within six phyla (Porifera, Cnidaria, Crustacea, Mollusca, Echinodermata, Urochordata; a short account for freshwater/terrestrial annelids and platyhelminths is also included below due to the possible adaptation of their developed techniques for the related marine taxa as well) out of > 20 invertebrate phyla available (Fig. 1). Within each studied phylum, only a limited number of scientific groups and species were studied for cell cultures. For example, no polychaete cells were employed in the phylum Annelida, only few groups of decapods were studied in the crustaceans, no single marine species from the flat worms, no schyphozoans in the Cnidaria, cephalopods, polyplacophores and opistobranchs in the Mollusca, are only few for the remarkable outcome of low diversity in the attempts to use marine invertebrates for in vitro approaches. A total of 90 publications (Fig. 1, 1988-1998; excluding abstracts or manuscripts published in symposia or journals which were unavailable) are included in the detailed examination of the literature dealing with tissue cultures from marine invertebrates. Only 27 publications (30.0%) describe cases where cultures were maintained for more than 1 month in vitro. In four phyla (Cnidaria, Crustacea, Mollusca, Echinodermata) < 29.0% of the publications described long-term studies, whereas in the Porifera and Urochordata, longterm studies composed 36.4 and 45.5%, respectively, of all publications. In 18 of the long-term reports (refer to the section on unicellular eukaryotic contaminations), it was confirmed or suspected that the developed cultures were

B. Rinkevich / Journal of Biotechnology 70 (1999) 133-153

contaminated by other unrelated cells. There is no single valid marine invertebrate cell line developed from any of the studies published in the last 10 years of research. This outcome further reveals the problematic configuration of this scientific discipline: many attempts (the above 90 publications, Fig. 1, > 50 additional abstracts that were never published in peer reviewed journals; other unpublished studies that failed to mature to the form of a scientific communication) but little success. Below are short accounts for the development of in vitro studies in the above-mentioned invertebrate phyla:

135

organisms, it is plausible to predict that sponges will be one of the most amenable sources for invertebrate cell cultures. However, all attempts to develop long-term, continuous cell lines from sponges (including the studies published during the last decade: Klautau et al., 1993, 1994; Pomponi and Willoughby, 1994; Imsiecke et al., 1995; Ilan et al., 1996; Pomponi et al., 1997; Leys, 1997, 1998; Rinkevich et al., 1998) have not been as rewarding and the methodology for establishing cell lines is still in its preliminary stages. Even the establishment of long-term cell cultures from several marine sponges (Klautau et al., 1993, 1994) has been found to be erroneous. Cells that were isolated and maintained in culture for years were identified as protozoan cells (Custodio et al., 1995; and unpublished data). This resulted in a suggestion to cultivate sponge fragments in bioreactors rather than sponge cell cultures, for biotechnological production of sponge metabolites (Osinga et al., 1988). Although simply by their structure, sponges exhibit a variety of obstacles for the development of cell culture, two of them unique to this phylum. (1) Many species (especially within the hexactinellid sponges) are syncytial animals, consisting of a

2.1. Porifera (marine and freshwater sponges) Sponge cell culture has been proposed as a model system for the study of a variety of cellular aspects. During the last few years, it was also evident that in vitro cultures of marine sponge cells may address the need for the commercial supply of novel biologically active chemical compounds with high pharmacological potential. Since sponges are considered to be the most simple metazoan group, and are frequently placed at the base of the evolutionary tree for multicellular 40 TOTAL PUBLICATIONS r - - ] LONG TERM CULTURES Z

30

r,..) ,..1

~ o

20

0Z 10

PORIFERA

CNIDAPdA

CRUSTACEA

__L

MOLLUSCA

ECHINODERMATA

UROCHORDATA

PHYLUM

Fig. 1. Marine invertebrate cell cultures: Total number of publications (closed bars; sorted from ASFA, MEDLINE and BIOSIS) and total number of publications describing long-term cell cultures ( > 1 month) open bars; during the last decade (1988-1998). Six marine invertebrate phyla are analyzed. Abstracts, some chapters in symposia/internal reports and publications in journals that are unavailable, are not included.

136

B. Rinkevich ,' Journal of Biotechnology 70 (I 999) 133-153

single multinucleated giant cell, the trabecular syncytium that is connected via perforated plugged junctions to cellular components. Under in vitro conditions, a dissociated tissue is aggregated by fusing the membranes to form a giant multinucleatal syncytium (Leys, 1998). Therefore, the basic cytological organization of this group differs substantially from that of other multicellular organisms including other sponge groups, raising questions regarding the amenability for hexactinellid sponges for cell cultures. (2) Unlike other metazoans, there are no areas of a sponge from which sterile, primary cultures can be obtained (Pomponi and Willoughby, 1994). A further complication of this situation is the fact that many sponges host endosymbiotic intracellular microorganisms. The use of antibiotics for the production of axenic conditions may be harmful or may have inhibitive effects on sponge cells as well. Without antibiotics, cultures are contaminated within 1-3 days (Pomponi and Willoughby, 1994). While there is yet no finite or continuous cell lines from sponges, primary cultures have been obtained in vitro for months (25 weeks for adultderived cultures, Ilan et al., 1996; 41 weeks for embryo-derived primary cultures, Rinkevich et al., 1998). Recent studies have revealed the positive effects of a variety of growth factors, such as epidermal growth factor (EGF), fibroblast growth factor (FGF), insulin, lectins (wheat germ agglutinin (WGA), Ulex europaeus agglutinin (UEA), phytohemagglutinin (PHA), concanavalin A (ConA)), exogenous cAMP, and other additives and substrates on cell division and behavior in primary cultures (Gaino and Magnino, 1994, 1996; Pomponi and Willoughby, 1994; Pomponi et al., 1997; Rinkevich et al., 1998). In the same manner, improved conditions for the shipment of living material (whole portions of adult sponges at 4~ Ilan et al., 1996; cryopreserved cells, Pomponi et al., 1997) as well as the use of 'preferable parts' (sponge fragments without the cortex layer, Ilan et al., 1996; embryos, Rinkevich et al., 1998) have progressed survivorship and cell divisions in primary cultures. The best dissociation protocols used are mechanical (Ilan et al., 1996: Rinkevich et al., 1998) and the combination of

mechanical/chemical dissociations (Pomponi and Willoughby, 1994; Pomponi et al., 1997). Several sponge cell culture media are used but it is evident that culture optimization is only in its primary stages. Improvement of culture conditions should be addressed by a number of experimentation procedures to evaluate and reveal better in vitro conditions specific to sponge cells.

2.2. Cnidaria (corals, sea anemones, hydrozoans) The Cnidaria, one of the most ancestral invertebrate phyla, is the lowest metazoan group possessing an organized body structure. The body plan of the cnidarians consists of only two cell layers, ectoderm and endoderm, with acellular intermediate gelatinous matrix, the mesoglea. The simplicity of cnidarians' structure and their relatively limited number of cell types provides an excellent opportunity for studying biological processes in vitro. Moreover, the symbiotic relationships between many cnidarians and unicellular algae (zooxanthellae, zoochlorella) exhibit an additional scientific challenge and several recent studies were engaged with these symbioses in vitro (i.e. Gates and Muscatine, 1992; Gates et al., 1992; Apte et al., 1996) and with dissociation protocols developed for cnidarian tissues (Gates and Muscatine, 1992: Gates et al., 1992; Greber et al., 1992; Frank et al., 1994; Weber 1995). However, in only one study (Frank et al., 1994), have long-term cell cultures been developed. In this study, continuous cell cultures from ten taxa of sedentary colonial marine cnidarians So'lophora pistillata, Porites lutea, Favia favus (Anthozoa, Madeporaria), Parerythropodium fulrum fidt'um, Dendronephthya hemprichi, Nephthya sp., Heteroxenia fitscescence (Anthozoa, Alcyonacea), Clathraria rubrinoides, Plexaura sp. (Anthozoa, Gorgonacea) and Millepora dichotoma (Hydrozoa) were established in vitro. Primary cultures of various cell types and sizes (5-20 ~tm) were obtained from colony fragments and/or planula larvae using three dissociation approaches: a mechanical approach, chemical approach and a novel approach, spontaneous dissociation (see below). Cells were cultured in a modified Leibowitz L15 medium, with 5-10%

B. Rinkevich / Journal of Biotechnology 70 (I 999) 133-153

heat-inactivated fetal bovine serum, diluted in seawater. Cell proliferation was observed in primary cultures within 7-20 days following dissociation. Several cell cultures were developed, maintained for approximately 1 year, subcloned and cryopreserved. However, frozen cell materials, when thawed, were dominated by eukaryotic unicellular organisms from the phylum Labyrinthulomycota (unpublished; see below). These organisms took over the cultures competitively, excluding the original cell types. Additional efforts to establish secondary cell cultures from corals (unpubl.) were not as rewarding and in most cases, cells remained in an arrested form for up to 1 year with no signs of proliferation. One of the interesting points, not yet explored in the phylum Cnidaria, is the topic of cell-substrate interactions. In some of the few investigations, a species-specificity has been revealed. Jellyfish tissues of striated muscle and endoderm adhered and spread on mesoglea in a species-specific manner (Schmid and Bally, 1988). The importance of the mesoglea as the preferable substratum was also documented for primary culture of neurons from a hydrozoan jellyfish (Przysiezniak and Spencer, 1989). In addition, a recent study (Frank and Rinkevich, in press) revealed that dissociated cells from a variety of cnidarians (including anthozoans and hydrozoans) adhered, spread and penetrated into the jellyfish mesoglea cubes while cells from other organisms, such as tunicates, responded to the same mesoglea as an intact substrate (unpubl). On the other hand, one of the most critical factors in Cnidaria cell culture is the dissociation method employed. Dissociation protocols have a significant impact on the survivorship of cells, tissue culture conditions and prolonged cell cultures (Frank et al., 1994) and several studies during the last decade have tested this approach (Schmid and Bally, 1988; Gates and Muscatine, 1992; Gates et al., 1992; Greber et al., 1992; Frank et al., 1994; Weber, 1995; Apte et al., 1996).

2.3. Platyhelmintes and Annelida Earlier studies on annelid neuron cultures (especially in the leech genus Hirudo) have already

137

been reviewed (Townsel and Thomas, 1987). Cell cultures from endoparasitic worms, with relation to human health, are also of great interest (reviewed in Toledo et al., 1997; Bayne, 1998). With regard to oligochaetes, while earthworms of the family Lumbricidae are of great interest in a variety of scientific disciplines including invertebrate immunology, environmental pollution and evolution, almost nothing has been added to our knowledge on annelid tissue cultures during the last three decades (following the reviews of Gomot, 1971; Rannou 1971). During the past 10 years, only a single report (Battaglia and Davoli, 1997) has studied in vitro conditions for Eisenia foetida cells. Earthworm cultures were maintained in vitro for > 1 year, possessing mainly three different cell types. One of the cell types, the non-adherent cells which are approximately 20 lam in diameter, appeared to be similar in morphology to the earthworm's coelomocytes. No cell proliferation has been recorded. Relatively longterm cultures of freshwater planarian neoblasts ( > 14 weeks) has also been recorded (Shiirmann and Peter, 1988).

2.4. Crustacea (crayfish, lobsters, shrimps) In vitro techniques in crustacean biology have become important and sometimes vital tools for the study of crustacean endocrinology and diseases of edible species. The in vitro study of endocrinology, especially in aspects of the endocrine regulation of molting (Brody and Chang, 1989; Chang, 1997), provides an applied tool for the characterization of this complex system. The intensive agricultural practices, which are associated with worldwide viral infections, represent an emerging problem that hampers culture of several commercially valuable species. Crustacean in vitro cell cultures that can support in vitro viruses and other parasitic replications are an needed urgently to study viral infections (Rosenthal and Diamant, 1990; Nadala et al., 1993; Hsu et al., 1995; Tong and Miao, 1996; Loh et al., 1997; Tapay et al., 1997). These are the two main reasons why cell cultures of edible crustaceans (also freshwater species) have gained new attention during the last decade (Brody and Chang, 1989; Chang and

138

B. Rinkevich /Journal of Biotechnology 70 (1999) 133-153

Brody, 1989; Chen and Kou, 1989; Chen et al., 1989; Cook et al., 1989; Itami et al., 1989; Hu, 1990; Ke et al., 1990; Rosenthal and Diamant, 1990; Ellender et al., 1992; Luedman and Lightner, 1992; Najafabadi et al., 1992; Loh, 1993; Nadala et al., 1993; Cancre et al., 1995; Ghosh et al., 1995; Hsu et al., 1995; Lu et al., 1995; Tapay et al., 1995, 1997; Frerichs, 1996; Tong and Miao, 1996; Toullec et al., 1996; Chang, 1997; Loh et al., 1997). As a result, much focus was placed on the establishment of cell lines derived from various tissues including epidermal origin (branchial walls, limb), hematopoietic and lymphoid tissues, gonads (ovary, testicular cells) nerve cells and hepatocytes. However, most studies concentrated on short culture periods and only few attempted to raise long-lasting cultures and cell lines ( < 3 weeks, Rosenthal and Diamant, 1990; 10 weeks to several months, Toullec et al., 1996; 3-11 months, Brody and Chang, 1989; > 90 passage, Hsu et al., 1995). Neither one of the above attempts has successfully matured into developing a cell line. The study on cell cultures from crustaceans is hampered by the fact that almost each laboratory has chosen a different species and other organs/ tissues for in vitro studies. It is very difficult to compare the studies and to draw a line of experimental steps for further consideration. For example, even the same laboratory may develop several dissociation protocols for different organs tested. While enzymatic dissociation was not beneficial for hepatopancreas (mechanical and explant spontaneous dissociation were superior), it was the preferable technical approach for epidermal tissues of penaeid shrimps (Toullec et al., 1996). In the same way, shrimp and Palaemon tissues responded differently to different media tested (Nadala et al., 1993; Cancre et al., 1995; Tong and Miao, 1996; Toullec et al., 1996) or to different growth factors (Cancre et al., 1995; Hsu et al., 1995). Cell culture optimization in crustacean studies is only in its primary steps and even within the same laboratories, long-term utilization and development of primary culture techniques was not the main topic for research, and therefore has not been advanced significantly (for example, comparison of methodology between Brody and Chang (1989) and Chang (1997)).

2.5. Mollusca (mussels, oysters, clams, snails) Mollusc cell culture is probably the most intensively studied group of marine invertebrates. Mollusc cell lines have been used in the study of neurobiology (Tamse et al., 1995; earlier studies reviewed in Townsel and Thomas, 1987), evaluation of xenobiotic effects on the cellular level and monitoring deleterious effects of pollution (Auffret and Oubella, 1997), the study of the immune system, especially of molluscs that are intermediate hosts of human parasites (Laursen et al., 1997; Davids and Yoshino, 1998) or those exhibiting a variety of tumors, including sarcomas and hematopoietic neoplasia (reviewed in Rosenfield, 1993; Rosenfield et al., 1994), the study for physiological/biochemical aspects of calcification in pearl oysters (Machii, 1988; Awaji, 1991, 1997; Samata et al., 1994), but above all, on a variety of aspects (biochemistry, mechanisms controlling development, metamorphosis, growth, pathogenesis, etc.) on molluscs which are of major commercial importance such as mussels, oysters, clams, snails, abalone (Ellis et al., 1985; Ellis and Bishop, 1989; Machii and Wada, 1989; Abbot, 1990, Kumazawa et al., 1990; Boulo et al., 1991; Holden and Patterson, 1991; Noel et al., 1991; Odintsova and Khomenko, 1991; Auzoux et al., 1993; Cornet, 1993, 1995; Odintsova et al., 1993, 1994; Wen et al., 1993; Domart-Coulon et al., 1994; Naganuma et al., 1994; Takeuchi et al., 1994, 1995a,b; Odintsova and Tsal, 1995; Renault et al., 1995; Kleinschuster et al., 1996; Lebel et al., 1996; Walker et al., 1996). Disease epidemics in edible molluscs have been of great interest and importance for more than three decades and the use of cell culture for studying mollusc pathogenesis has been of high priority (literature cited in Li et al., 1966; Brewster and Nicholson, 1979, Ellis et al., 1985). The great interest in cell culture from molluscs during the 1970s has revealed two major important outcomes: (1) oyster amebocytes were successfully maintained in vitro for long periods of up to 6 months (Brewster and Nicholson, 1979); (2) the most significant achievement, the development and characterization of the first and the only molluscan cell line from embryos of the freshwa-

B. Rinkevich / Journal of Biotechnology 70 (I 999) 133-153

ter snail Biomphalaria glabrata (Hansen, 1976). It is therefore of great disappointment that not much has been achieved regarding the development of cell lines from marine molluscs and, on the other hand, that so few researchers used the available single cell line for their studies (for example: Yoshino and Laursen, 1995; Laursen et al., 1997). Studies of mantle tissue culture from the pearl oyster are similarly disappointing. Pearl oyster cell culture started almost 25 years ago (Machii, 1974) with tissue explants. In this culture system, secretion of an organic substance was followed in vitro. A year later, the same laboratory documented mitoses in cell cultures from mantle tissue. Even with the importance of this industry of pearl oysters, recent studies (Awaji, 1991, 1997; and literature therein) do not show any significant improvement as compared to the 1970s cell culture studies. It is therefore even more evident that outside insects and ticks, primary cell or organ culture maintenance has rarely given subcultures of any cell type and almost no finite or continuous cell lines (summarized in Rosenfield, 1993). During the last 10 years, an impressive variety of organs and cells from molluscs have been cultured, including epithelial cells from embryos, gills and mantles (Machii, 1988; Awaji, 1991, 1997; Auzoux et al., 1993; Odintsova et al., 1994; Samata et al., 1994; Takeuchi et al., 1994; Cornet, 1995), neurons (Berdan et al., 1990; Tamse et al., 1995, and literature therein), digestive glands (Odintsova et al., 1994), muscles including heart cells (Ellis and Bishop, 1989; Odintsova and Khomenko, 1991; Wen et al., 1993; Domart-Coulon et al., 1994; Naganuma et al., 1994; Odintsova et al., 1994; Odintsova and Tsal, 1995; Takeuchi et al., 1995a,b; Kleinschuster et al., 1996) and the hematopoietic systems (Kumazawa et al., 1990; Boulo et al., 1991; Noel et al., 1991; Lebel et al., 1996; Auffret and Oubella, 1997; Davids and Yoshino, 1998). The results, however, are not so rewarding. For example, when focusing on Perkinsus disease, which causes mass mortality in oyster and clam cultures, Auzoux et al. (1993) planned to establish a suitable cell culture system for clam's gill for the purpose of studying host-pathogen relationships

139

and to screen in vitro anti-parasitic candidate agents. However, only relatively short (approximately 1 month) cultures have been established. The aim of culturing nudibranch neurons for the study of cellular physiology and behavior have revealed only short functional cultures. In this system, signs of cellular deterioration appeared as soon as days 10-12 (Tamse et al., 1995). Wen et al. (1993) clearly stated that a critical review of the efforts to culture adult tissues, especially heart cells from marine molluscs, resulted in difficulties which hampered further development. In the same manner, even neoplastic tissues from molluscs were not amenable for long, in vitro cultures (Noel et al., 1991; Rosenfield et al., 1994; Walker et al., 1996). Some of the studies that documented long-term cultivation for a variety of mollusc tissues for longer periods of times (Ellis and Bishop, 1989; Odintsova and Khomenko, 1991; Wen et al., 1993; Domart-Coulon et al., 1994; Odintsova et al., 1994; Takeuchi et al., 1995a,b) are also probably subject to contamination by thraustochytrid species (Ellis and Bishop, 1989; see below). It may therefore be concluded that despite great efforts to establish marine mollusc cell lines, this field is still in its preliminary stages. In several cases, however, the current state-ofthe-art regarding in vitro conditions was good enough for the specific questions asked. For example, hemocyte aggregation in the oyster Crassostrea gigas has been used as a sensitive bioassay for in vitro measurement of the effect of xenobiotics (Auffret and Oubella, 1997), and oyster cell cultures have been used to show that the American oyster is unable to synthesize sterols (Holden and Patterson, 1991). In vitro spreading and motility of circulating phagocytic cells (hemocytes) of snails was tested for the hypothesis that this behavior is mediated through RGD-binding integrin-like surface receptors (Davids and Yoshino, 1998) and mollusc cell cultures have been explored for synthesis of adhesive proteins (Abbot, 1990). In the same way, a functional study of burst respiratory activity has been successfully analyzed in vitro on scallop hemocytes as well as the analysis of interactions with protozoan and prokaryotic pathogens (Boulo et al., 1991). These in vitro model systems have applied the

140


accumulated knowledge of in vitro maintenance of mollusc cells, providing first-hand evidence for the potential of in vitro cell cultures in a variety of scientific needs. General in vitro conditions have been recently tested for further improvements, including cellular cryopreservation (Odintsova and Tsal, 1995), media supplementation (Wen et al., 1993; DomartCoulon et al., 1994; Lebel et al., 1996), substrate of choice (Odintsova et al., 1994) and the effects of dissociation protocols (Awaji, 1991; Takeuchi et al., 1995a,b). Therefore, all facets of in vitro conditions are being studied in order to obtain greater knowledge of mollusc cell cultures. There are still difficulties common to primary molluscan cultures, such as tissue sterilization, cell proliferation, cellular purification/characterization and thraustochytrid contamination (part of this list is summarized in Awaji (1997)). The above studies embraced the four culture situations uniquely characteristic of a variety marine molluscs: giant neurons routinely used in neurobiology (Tamse et al., 1995), pearl formation (Awaji 1991, 1997 and literature therein; Samata et al., 1994), mollusc neoplasia, frequently detected during epidemiological surveys of bivalve mollusc stocks (Noel et al., 1991; Rosenfield et al., 1994, and literature therein) and the appearance of many snails as intermediate hosts for human parasites (such as the blood fluke Schistosoma mansoni; Davids and Yoshino, 1998). These situations are seldom referred to, or are not applicable to other invertebrate phyla. They provide experimental in vitro challenges uniquely applied to the phylum Mollusca. 2. 6. Echinodermata Primary cultures of echinoderm cells during the last 10 years concentrated on a variety of cellular differentiation aspects. Studies were performed on sea urchin cells (Benson et al., 1990; Odintsova et al., 1994; Ermak and Odintsova, 1996), primary starfish cultures (Kaneko et al., 1995, 1997) and cultures from sea cucumbers (Blinova et al., 1993; Odintsova et al., 1994). Due to the questions asked, most studies concentrated on embryonic cells (Benson et al., 1990; Odintsova et al., 1994;

Kaneko et al., 1995, 1997; Ermak and Odintsova, 1996) or germ line cells (Poccia, 1988) and for relatively short periods of a few days to up to 40 days. There are two characteristics of primary echinoderm cultures which were not recorded as important in most other invertebrate cell cultures. (1) The impact of substrates on cellular differentiation. For example, two different cell types (epithelial or mesenchymal) may be developed 1 week after initiation of cell cultures from sea urchin larvae at the gastrula stage (Ermak and Odintsova, 1996). Epithelial cells predominated wells coated with polylysine (with a high level of [3H]thymidine incorporation), while the addition of fibronectin or oncoprecipitin A (oncoA, a specific glycoprotein isolated from an ascidian) resuited in syncytia of different shapes and sizes and conglomerates typical of primary mesenchyme cells were formed. (2) As shown years ago (Bertheussen and Seljelid, 1978), a variety of cell types in echinoderms tend to form aggregates and/or monolayer syncytia. As a result of the limited number of studies aimed at the production of continuous cell culture from echinoderms, there are almost no data for the response of echinoderm cells in long-term conditions. Dissociation protocols, media, general conditions and organs to be used as a source for culture requirements to be defined in future experimentation. 2. 7. Urochordata (solitary, colonial tunicates) Tunicates, the most primitive group of the phylum Chordata, may reveal critical answers in the phylogeny of several biological phenomena such as immunology, developmental biology, genetics, cellular biology and more. Following that, the importance of urochordates as model organisms has already been established in a variety of disciplines, reflecting as well the field of cell culture. During the past 10 years, most studies have attempted to develop cell cultures from solitary and colonial tunicate blood cells (Raftos et al., 1990, 1991; Raftos and Cooper, 1991, 1996; Rinkevich and Rabinowitz, 1993; Sawada et al., 1994; Peddie et al., 1995; Rinkevich et al., 1996) mainly

B. Rinkevich /Journal of Biotechnology 70 (I 999) 133-153

due to their importance in allorecognition and other immune responses. Few studies focused on epithelial cell cultures (Kawamura and Fujiwara, 1995; Rinkevich and Rabinowitz, 1997) or on embryo-derived cell cultures (Rinkevich and Rabinowitz, 1994). Some of the studies evaluated short-term experimental sets. However, relatively more frequent than in other phyla studied, longterm experiments ( > 2 months; Raftos et al., 1990; Rinkevich and Rabinowitz, 1993, 1994; Sawada et al., 1994; Kawamura and Fujiwara, 1995) have resulted in cell cultures from Botryllus schlosseri (Rinkevich and Rabinowitz, 1994) and Polyandrocarpa misakiensis (Kawamura and Fujiwara, 1995). At least in the B. schlosseri cultures, it has been evident that opportunistic organisms contaminated the cultures (thraustochytrids, see below). The in vitro development of long-term primary cultures from tunicate blood cells is faced with a significant challenge. Tunicate hemocytes have a rapid turnover and in vivo circulatory hemocytes survive for only several weeks (Raftos et al., 1990). Keeping this fact in mind, several studies used explant cultures of 'hematopoietic organs' (such as pharyngeal sinuses of solitary ascidians) as target sources for cell-line development (Raftos et al., 1990, 1991; Raftos and Cooper, 1991; Sawada et al., 1994). Where no hematopoiesis centers are known, blood cells were collected directly from blood vessels for in vitro approaches (Rinkevich and Rabinowitz, 1993; Peddie et al., 1995). In the first approach, resident hemocytes developed, multiplied, matured and migrated outside of the explants for more than 2 months, with clear indications that proliferation renewed the pool of hemocytes within the explant tissues (Raftos et al., 1990). Tunicate blood cells are one of the major cell types used in urochordate tissue cultures where several factors (also of mammalian systems) are found to influence cell proliferation and activation (reviewed in Peddie et al., 1995; Raftos and Cooper, 1996). Activities of tunicate defense-related hemocytes were found to be modulated by regulatory signals using mammalian cytokines and plant lectins. A detailed study on this phenomenon has resulted in the characterization of

141

proteins (designated as tunlLl~ and tunlL1 ]3; reviewed in Raftos and Cooper, 1996). These proteins have a variety of cellular effects in vitro other than mitogenesis. They increase in vitro phagocytosis and opsonization, working as extremely potent materials (being active at concentrations of as little as 150 ng/ml) and they activate hemocyte chemotaxis and metabolism (activation of incorporated amino acids and glucose). The use of different dissociation protocols (Rinkevich and Rabinowitz, 1994), a variety of mitogens (Rinkevich and Rabinowitz, 1993; Peddie et al., 1995), osmolarity (Kawamura and Fujiwara, 1995) and different media (Raftos et al., 1990) may have significant effects on cell viability and cellular proliferation. In vitro conditions for tunicate cells are still far from optimal. One example is the capacity for proliferation during regeneration of isolate Botrylloides ampullae (Rinkevich et al., 1996). While any peripheral isolated fragment of blood vessel can regenerate in vivo (within filtered seawater) into whole newly formed zooids in less than 2 weeks, in vitro conditions lead to a complete failure of this unique regeneration from totipotent blood cells. This example reveals the potentiality of the system and the lack of amenability of currently developed protocols.

3. Marine invertebrates: unique approaches/obstacles The lack of significant progress of marine invertebrate cell cultures may be related to an inappropriate comparison of invertebrate cell culture requirements with the culture conditions of vertebrate cell lines or may be due to special or unique requirements needed for invertebrate cell cultures (Rinkevich et al., 1994; Bayne, 1998). The brief discussion above, dealing with the scientific activity during the last decade, further reveals unique characteristics of cell cultures for even different invertebrate phyla. The syncytial cells of hexactinellid sponges (Leys, 1998) and the cells from sponge species harboring endosymbiotic intracellular microorganisms (Pomponi and Willoughby, 1994) clearly require different in vitro conditions

142


than, for example, mollusc neoplastic cells (Noel et al., 1991; Rosenfield et al., 1994) or urochordate blood cells (Rinkevich and Rabinowitz, 1993). The hypothesis that culture requirements for vertebrates and invertebrates are not comparable has been discussed in detail by Goodwin (1991), who advocated that invertebrate cell surface (lipidic and extracellular matrix) specificities differ significantly from those of the vertebrates. At the same time, significant genetic homologies were revealed between vertebrate and invertebrate bioactive peptides (Goodwin, 1991) a fact which indicates the opposite conclusion. On the other hand, it is clear that invertebrate cell cultures require unique approaches and bear obstacles not recorded in vertebrate cell cultures. Some of these points (aseptic conditions, media supplements and substrate cell types and classification; Rinkevich et al., 1994) and others (selection of species, selection of basic media and additives, the merit of feeder layers, and more; Bayne, 1998) were reviewed recently. A number of questions may be addressed when summarizing efforts of the last decade for invertebrate cell cultures. These questions include the investigation for unique approaches and obstacles which are common, or even ubiquitous, to only marine invertebrate cell cultures. In general, we do not even know why marine invertebrate cells never propagate in vitro. One of the striking examples is the survival budding processes in the colonial urochordate Botrylloides (Rinkevich et al., 1996). In this group of organisms, any separated minute fragment of peripheral blood vessels with 100-300 blood cells gives rise within a few weeks to a fully organized organism. However, all efforts to induce regeneration of isolated ampullae in vitro have failed (Rinkevich et al., 1996). Are blocks to propagation inadvertently introduced in procedure protocols or are they a natural part of in vivo versus in vitro conditions? There are several topics, however, that have emerged during the last decade, providing us with a new look, ubiquitous to marine invertebrate in vitro approaches.

3.1. Cell sources for better in vitro results

While all attempts to develop long-term, continuous cell cultures from marine invertebrates have not been rewarding and this field is still in its preliminary stages, there is one promising approach which has proven to be valid in several studies: the target organ/cell source for in vitro studies. Cells for in vitro studies were obtained from invertebrate epithelia (gills, mantle, pharynx, buds, outer epithelium), muscles (heart, a variety of other muscles), neurons, blood/hemolymph, internal organs (glands, hepatopancreas), gonads and others. Two of the most promising approaches are the amenability of larval/embryonic cells and the use of organ culture/spontaneous dissociation protocols. The use of larval and embryonic cells in vitro has been employed in a variety of freshwater and marine invertebrates, including sponges (Imsiecke et al., 1995; Rinkevich et al., 1998), crustaceans (Frerichs, 1996; Tong and Miao, 1996; Toullec et al., 1996), molluscs (Hansen, 1976; Brewster and Nicholson, 1979; Ellis et al., 1985; Ellis and Bishop, 1989; Machii and Wada, 1989; Odintsova and Khomenko, 1991; Odintsova et al., 1994; Naganuma et al., 1994; Odintsova and Tsal, 1995; Takeuchi et al., 1995a,b), cnidarians (Schmid, 1974; Frank et al., 1994), echinoderms (Benson et al., 1990; Odintsova et al., 1994; Kaneko et al., 1995; Ermak and Odintsova, 1996) and urochordates (Rinkevich and Rabinowitz, 1994; Kaneko et al., 1997). In the invertebrates, embryonic cell lines, which are restricted to insects (for example, Stiles et al., 1992) showed their amenability as an excellent tool for a variety of approaches and scientific questions. This source of cells is also less vulnerable to contamination from bacteria, yeast and mold (Frank et al., 1994; Rinkevich et al., 1998). This is of further importance because available tissue sources from marine invertebrates, particularly those that contain intracellular micro-organisms and crypto-organisms cannot be completely sterilized (Rinkevich et al., 1994). Moreover, when compared to adult primary cultures, embryonal cultures survived longer (Rinkevich and Rabinowitz, 1994; Rinkevich et al., 1998) and in many cases, consistently yielded

B. Rinkevich / Journal of Biotechnology 70 (1999) I33-153

cultures that were capable of at least limited growth in vitro (Ellis and Bishop, 1989). However, while cell lineages of gastropod embryos, for example, are well characterized (Refs. in Naganuma et al., 1994), this is not the case for other marine invertebrates such as sponges, cnidarians or even protochordates. This fact may provide an obstacle in characterization of cellular components developed in vitro from dissociated larvae/ embryos. On the other hand, larvae/embryos may provide large numbers of synchronously developing cell lineages and posses cell populations with high mitotic indices which are better candidates for primary and continuous cultures than any other organ or tissue. Organ cultures and spontaneously dissociated tissue fragments are two approaches which employ similar protocols for two different purposes: long-term cultivation of tissue fragments which consist of foci for cell proliferation that migrate from the explants into the medium during culture, and short-term cultivation of tissues that dissociate spontaneously (without the employment of any chemical or enzymatic treatment) and provide a wide scale of cell types for in vitro applications. Organ culture for sustained viability of proliferation of cells is mainly carried out on solitary tunicates where cultured pharyngeal fragment sites of hematopoiesis are used (Raftos et al., 1990; Sawada et al., 1994; Raftos and Cooper, 1996). In this system, pharyngeal tissues remained viable and proliferated. This maintained the pool of hemocytes within the explants and facilitated the migration of hemocytes from explants into the culture medium (Raftos et al., 1990). The same methodology has been employed for establishing primary epithelial cultures from tunicate buds (Kawamura and Fujiwara, 1995; Rinkevich and Rabinowitz, 1997). Cells spread out from the epithelial explant and proliferated (at least for a short period of 1-2 weeks) on the substrate. Spontaneous dissociation is based on inserting tissue fragments or whole larvae into tissue culture medium, where they are dissociated without employing any chemical/enzymatic or further mechanical treatment. This was successfully performed on earthworm and platyhelminth cultures (Battaglia and Davoli, 1997; Toledo et al., 1997),

143

shrimp cell cultures (Nadala et al., 1993; Tong and Miao, 1996; Toullec et al., 1996), bivalves (Auzoux et al., 1993; Wen et al., 1993; Samata et al., 1994) and cnidarian larvae (Frank et al., 1994). In two cases (Frank et al., 1994; Toullec et al., 1996) cultures of cells from explants were superior to cultures obtained from dissociated tissues and provided viable cells, sometimes even more cell types. This methodology may be explored further to find additional improvements which will reduce stress to cells during tissue dissociation.

3.2. Cryopreservation Cryopreservation has been widely used for long-term conservation of a variety of vertebrate cells, cell lines and insect cell cultures (Hink, 1979). Cryopreserved cells may provide a yearround supply of various cell types with similar qualifications and the same source for long-term comparative experimental studies. The establishing of the cryopreservation methodology is essential for holding and developing cell cultures and cell lines (Saxena et al., 1995). The cryopreservation methodology has been tested in marine invertebrate cell and tissue cultures as of 1992, by employing studies on two potentially different sources: cells already in vitro conditions and tissues amenable for future dissociation. Ellender et al. (1992) tried to store shrimp hemocytes at - 7 0 ~ for 2 months (10% DMSO (dimethylsulfoxide) solution). Poor cellular recovery ( < 26%) was recorded after vial thawing. On the other hand, Odintsova and Tsal (1995) have shown that cryopreservation of primary cell cultures of bivalves is not only a feasible goal (cell viability percentages after thawing the frozen materials were almost the same as in freshly prepared primary cultures) but that the use of 5-10% DMSO is favorable over 10% glycerol as cryoprotectant. No difference in viability was recorded after thawing primary cnidarian cell cultures (Frank et al., 1994) and embryo derived tunicate cells (Rinkevich and Rabinowitz, 1994) frozen in either 10% glycerol or 20% DMSO. Both protocols yielded high rates of cell viability. In the same

144

B. Rinkevich, Journal of Biotechnology 70 (1999) 133-153

manner Pomponi et al. (1997), routinely used cryopreserved sponge archaeocytes (15% DMSO) as a major source for the purpose of establishing primary cultures, monitoring cultures for production of bioactive metabolites and for microtiter plate assays. The above experiments, carried out on representative species from five phyla (Porifera, Cnidaria, Arthropoda, Mollusca, Urochordata), clearly indicate the feasibility of cryopreservation technology for future demands of cell supplies. Two studies also examined the feasibility of cryopreserved whole embryos from marine invertebrates for future studies. This was performed on bivalve trochophores and veligers (Odintsova and Tsal, 1995), and on sponge parenchymella larva (Rinkevich et al., 1998). With bivalve larvae, the veligers showed high cell viability of > 90% (in either 10% DMSO or 10% glycerol; similar to the control, unfrozen cells), while only < 35% of the trochophore cells survived the freezing-thawing protocol. In the sponge system, immediately after dissociation, cell viability was similar in both treatments (70_+ 7 . 8 ~ n = 30 for thawed embryos; 77.0_+ 1 3 . 0 ~ n - 50 for freshly collected embryos; P > 0.05) as were cell yields and cell type distributions (Rinkevich et al., 1998). However, for the cryopreserved embryos, 196 out of 200 wells (98%) were contaminated by bacteria along a 39-day observational period (140 wells, 70% were septic within 1 week). In the controls, during a total of 122 out of 300 wells (40.7%; P < 0.05) were contaminated, mainly by bacteria and thraustochythrids (see below), but also by fungi, cyanobacteria and amoebae. Most cases of thraustochythrid contamination developed after 3 weeks and those with bacteria, within the first 2 weeks. The reviving protocol for cryopreserved sponge embryos, while not affecting initial cell viability, results in increased bacterial contamination compared to controls. Bacteria are likely to take advantage of the fragile conditions imposed on embryo cells during the freezing-thawing cycle. This approach, has the advantage of reproducibility, high rates of cell viability and a potentially continuous supply of embryos, and deserves, therefore, further consideration (Rinkevich et al., 1998).

3.3. Unicellular eukaryotic contamination The morphologies of invertebrate cells in vitro may differ from those in vivo (Rinkevich et al., 1994). The additional high plasticity of shapes and structures of cells from a specific animal (Gomot, 1971) and the insufficient information for the classification of cell types/lineages in most invertebrates add to the confusion of in vitro cell identification. This may result in the culturing of alien unicellular eukaryotic contaminants in vitro as 'genuine' cells originating from the animal under investigation. One such example was recorded during the attempts to establish long-term cell cultures from several sponge species (Klautau et al., 1993, 1994). Cells that were isolated and maintained in culture for long periods (years) were subsequently identified as the protozoan Neoparamoeba aestuarina (Custodio et al., 1995, and unpublished data). Many other types of amoebas were also recorded in primary cell cultures from Negombata (sponge) adults and embryo preparations (Blisko, 1998). A much more common type of in vitro contamination in many cultures of marine invertebrate cells is the appearance of thraustochytrids (Fig. 2a-d), common marine and freshwater heterotrophic protists, that feed as saprophores, as parasites or as bacterivores (Porter, 1990; Raghukumar, 1992). Although they are very common in coastal waters (up to 5.6 x 10 4 cells 1- 1; Naganuma et al., 1998) and on a variety of living marine organisms such as sponges, corals, hydroids, bivalves, octopus, squids, nudibranches, echinoids, diatoms, sea grass and more (Cousserans et al., 1974; Raghukumar, 1988; Porter, 1990; Raghukumar and Balasbramanian, 1991; Bower, 1995, and Refs. therein) their evolutionary relationships and taxonomy are still poorly understood (Porter, 1990; Cavalier-Smith et al., 1994). For example, they were characterized recently as neither protozoa nor fungi, but as heterotrophic heterokontchromists (Cavalier-Smith et al., 1994). The interest in thraustochytrids as possible major contaminants of invertebrate cultures was initiated by Ellis and co-worker's work on molluscs cell cultures (Ellis et al., 1985; Ellis and Bishop, 1989). These thraustochytrids were encountered in

B. Rinke~'ich .. Journal of Biotechnology 70 (1999) 133-153

145

Fig. 2. Unicelluar eukaryotic contaminations in marine invertebrate primary cultures: (a) thraustochytrids in primary sponge cell culture (Negombata sp.). Large and dividing lumps of small cells (arrowheads) are the typical appearance for these contaminants ( x 400); (b) phase contrast of thraustochytrids from Porites lutea (a hermatypic coral), characteristic by the network of connecting filipodia between the small size cells ( x 400); (c) possible thraustochytrid infection in a tunicate (Boto'llus schlosseri) primary epithelial cell culture. Epithelial cells and thraustochytrids are both characterized by extended filipodia ( x 400); (d) unspecified unicellular eukaryotic contaminant from tunicate cell culture, characterized by dividing cells in the periphery of the cell mass

( • 200). approximately 30% of oyster cultures and about 27% of clam cultures, and appeared in a variety of forms as rapidly dividing round cells, round cells with filopods forming a stellate pattern around the cells, cells connected by net-like ectoplasmic processes or, as spherical-to-ellipsoid cells. Recently, the appearance of a variety of thraustochytrid organisms has been confirmed in cell cultures from sponges (Ilan et al., 1996; Blisko, 1998), corals (Frank et al., 1994), oysters (Awaji, 1997) and tunicates (Rinkevich and Rabinowitz, 1993, 1994, 1997; Table l, Fig. 2a-d). Moreover, detailed examination of the literature on marine invertebrate cell cultures during the last decade, especially the studies that described highly prolif-

erating cultures (Table 1), reveal that thraustochytrid protists are very common in these cultures from sponges, cnidarians, crustaceans, molluscs, echinoderms and tunicates. Possibly, many reports on invertebrate cell cultures describe the development and maintenance of thraustochytrids rather than the original animal cultures, further revealing a poorer yield of cell cultures from marine invertebrates (Table 1). The above list of 21 confirmed and suspicious cases for thraustochytrid contamination in marine invertebrate cell cultures (Table 1) points to the importance of these contaminants as an obstacle in the development of cell lines from a variety of marine organisms. Thraustochytrids usually de-


146

velop in cultures following other, more opportunistic parasitic forms. In sponges, (i.e. Fig. 3; Blisko, 1998) most of the thraustochytrids developed at days 19-39, while most bacteria, fungi and other protists appeared within the first 3 weeks. Similar results were obtained in colonial tunicate cell cultures. Records from 17 experiments where thraustochytrids were developed in primary Botryllus schlosseri cultures (from buds and blood cells) indicated that in most cases, they appeared > 12 days after culture initiation (up to 48 days; unpublished). There are several ways to identify thraus-

tochytrids in vitro; unfortunately, none of them is conclusive. One way of confirmation can be obtained by the use of electron microscopy (Ellis and Bishop, 1989; Ilan et al., 1996; Blisko, 1998). Contrary to cells originating from marine invertebrates, thraustochytrids have cell walls composed of overlapping plate elements. This material can, however, sometimes be very thin and not easily discernible even under an electron microscope. The mitochondrial cristae of thraustochytrids have a tubular profile, whereas mitochondria with lamellar cristae profiles are the usual animal pattern. A specialized structure unique to the thraus-

Table 1 Confirmed (C; as recorded by original authors) and suspicious (S) identifications (ID) of thraustochytrid protists contamination in marine invertebrate cell cultures during the last decade a Phylum

Reference

ID

Remarks

Porifera

Ilan et al. (1996) Blisko (1998)

C C

Cnidaria

Frank et al. (1994)

C

Crustacea

Itami et al. (1989)

S

Ke et al. (1990)

S

Hsu et al. (1995)

S-C

Toullec et al. (1996)

S

Mollusca

Ellis and Bishop (1989) Auzoux et al. (1993) Takeuchi et al. (1994) Lebel et al. (1996) Awaji (1997)

C S S S C

Echinoderrnata

Kaneko et al. (1995)

S-C

Ermak and Odintsova (1996)

S-C

Rinkevich and Rabinowitz (1993, 1994, 1997) Sawada et al. (1994)

C

Kawamura and Fujiwara (1995)

S-C

All proliferating cultures were identified as thraustochytrids. About 10% of all adult and embryo derived cultures were infected. Found in all ten soft and hard coral cultured species. All secondary cultures and proliferating cell lines were contaminated. Longer cultures (30-54 days) of shrimp lymphoid cells developed cell connections similar to thraustochytrids (their Table 1). Cultures of up to 5 m from shrimp hepatopancreas resemble thraustochytrid cultures by cell structures and by forming centers of multiplications (their Figs 1-4). Shrimp lymphoid tissue developed typical thraustochytrid cultures (their Figs. 6-8). Epidermal and hepatopancreas isolated cells emitted filopods typical to thraustochytrids. Developed in marine bivalve primary cultures. Suspicious thraustochytrid development (their Fig. 3 and text). Suspicious thraustochytrid development (their Fig. 3 and text). Suspicious thraustochytrid development (their Fig. 1). Thraustochytrid developed in pearl oyster mantle cultures (literature cited provided two more confirmed cases from 1988 to 1989 studies of same author). Suspicious thraustochytrid development (their Figs. 1-3 and text). Fine network of cell processes between cellular elements from starfish embryos. Dense network of filipodial processes between embryonic sea urchin cells. Thraustochytrids developed in primary cultures from blood ceils, embryos and epithelial cell of a colonial tunicate. An EM study or solitary tunicate hemocytes reveals cells with cell wall, typical to thraustochytrids (their Fig. 2b). Suspicious thraustochytrid development (their Fig. 4 and text).

Urodordata

S

a S-C refer to cases where the original figures in the published manuscripts depict thraustochytrids while authors were unaware of their presence.

B. Rinkevich /'Journal of Biotechnology 70 (1999) 133-153 a Total Bacteria Fungi Thraustochytrids --~ Others 20

0

,,,

7

~

.~_ E

51

N

~

14 21 28 Culture period (d)

35

42

o 50

40

3O

2O

8

lo

C)

"

0

7

14

21 28 Culture period (d)

J

35

42

Fig. 3. Contamination of primary sponge cultures (Negombata sp.) within the first 39 days of in vitro conditions. (a) Adult sponge cultures ( n - 1 9 0 wells); (b) embryo-derived cultures ( n - 300 wells). In both cases, thraustochytrids appeared in 9-10% of total wells following bacteria and fungi contaminations. In total, 90 wells (47.4%) of adult sponge cultures and 122 (40.5%) of embryos were contaminated during this period (Blisko, 1998).

tochytrids is the sagenogenetosome (Porter, 1990), usually diagnostic in electron microscopy. This structure can, however, be difficult to find in some thraustochytrids since there may only be one in a cell of up to 100 lam in diameter. Under light epifluorescence microscopy, the use of acriflavine hydrochloride, which stains the sulfated polysaccharide cell walls of these organisms, is highly recommended (Raghukumar and Schaumann, 1993). Another characteristic feature is the cytoplasmic extensions. In thraustochytrids, these extensions have no organellar inclusions and at the light microscope level often form anastomosing networks that rather resemble the web of an

147

orb-weaving spider. Under certain culture conditions, thraustochytrids may produce biflagellated zoospores. The zoospores are not produced in some genera of thraustochytrids, so they cannot be relied upon to identify contamination. In some cases, zoospore production is restricted to a short interval following cell feeding (Ellis, pers. commun.). The culture medium may also serve as an indicative parameter. Many thraustochytrids are capable of growth in very simple media. The basic requirements include pyridoxal phosphate, niacinamide, and vitamin B12 and a carbon and nitrogen source (Ellis, pers. commun.). Most animal cells would be unable to proliferate in this medium while thraustochytrids can. However, lack of growth cannot entirely ensure that the isolates are not thraustochytrids which are dependent on a specific carbon or nitrogen source. Many thraustochytrids, survive and grow on pollen grains (Raghukumar, 1992), a feature which is not characteristic to any animal cell types. Ellis and Bishop (1989) were also successful in screening bivalves and thraustochytrid cells by biochemical markers. Contrary to the thraustochytrids, in bivalves and a few other invertebrates, there is no L-lactate dehydrogenase; instead, they have opine or D-lactate dehydrogenases and this trait may serve as a useful marker in their cell cultures. Thraustochytrids may be distinguished from all animal cells by their typical 18S mRNA signatures (Cavalier-Smith et al., 1994). Additionally, many thraustochytrids may produce and release antibiotic products into the medium (Raghukumar, pers. commun.; Rinkevich and Rabinowitz, 1994, our unpublished results) or may feed directly on bacteria found in the medium (unpublished); but so might sponge cells. Since, usually one of the above criteria for the confirmation of thraustochytrid cells in culture is insufficient, it is highly suggested to combine several checks. Another important point for consideration is that thraustochytrid contaminations are sometimes inevitable. We (unpublished) found these cells circulating in the blood systems of Botryllus schlosseri, a colonial tunicate, and within/on embryos of corals (Frank et al., 1994), tunicates (Rinkevich and Rabinowitz, 1994) and sponges (Blisko, 1998).

148


4. Perspectives During the middle 1970s, there were 70 cell lines available from different species of invertebrates, insects and ticks (Hink, 1979). More than 20 years thereafter, and despite the substantial, pent-up demand for such cell cultures from a variety of taxa, and for numerous applications, there are several hundred insect-arachnid cell lines but no single cell line from marine invertebrates. During the last decade, numerous studies followed previous efforts and failures for the development of marine invertebrate cell cultures (90 peer-reviewed publications, > 50 abstracts and many unpublished reports). This continued failure to establish cell lines from these invertebrate taxa may suggest that we still lack vital information regarding invertebrate cell physiology/biochemistry and biology. For example, during the development of the only single cell line from freshwater snails, an extreme sensitivity of the molluscan cells to changes in vertebrate serum supplementation was evident by changing serum lots or by the use of different serum sources (Hansen, 1976). One obvious conclusion may be that the lack of progress in this discipline is related to inappropriate comparisons of marine invertebrate cell culture requirements with the culture requirements of the vertebrate cells (Goodwin, 1991). There are probably many more reasons and any attempt to anticipate the whole range of unique obstacles is somewhat problematic. Approaches and protocols that already have been developed have successfully yielded primary sterile cultures from many marine invertebrates. Others yielded artifacts (such as the appearance of thraustochytrids), but the problems are encountered with possible solutions. Any cell which is derived from a multicellular organism represents on one hand, the most elementary expression of the organism's genetic trait, and on the other, establishes in vitro a new entity, of its 'own life', without being influenced by any organized structure, humoral or neural factors (Rinkevich et al., 1994). Therefore, in the vertebrates as well as in insects, primary culture methodologies have opened up new approaches for studying in vitro cellular and molecular

events. Some of the vertebrate growth factors may have positive influences on a variety of marine invertebrate primary cultures (Rinkevich and Rabinowitz, 1993, 1994; Lebel et al., 1996; Pomponi et al., 1997) but others do not. Some basic media used for vertebrate cell cultures showed positive influences on marine invertebrate primary cultures as well but the same media were detrimental in other cases (reviewed in Bayne, 1998). The overall conclusion from the variety of studies on freshwater and marine invertebrate primary cultures indicate, however, that we need to consider the production of secondary and cloned cell lines. The literature of the last decade clearly points out that the methodologies for holding short-term, viable primary cultures have been established for a variety of marine organisms. Efforts should therefore be concentrated on such approaches as trangenesis (Rosenfield et al., 1994), mutagenesis by irradiation/chemicals, cell hybridization (Diekmann-Schuppert et al., 1989), replacement of vertebrate sera with lipids and other factors (Goodwin, 1991), the use of feeder layers and more. Such approaches many successfully replace the continuous use of the same methodologies for in vitro conditions of primary marine invertebrate cultures which probably have a limited benefit maintaining viable, but non-proliferating cells.

Acknowledgements This is part of the research was carried out at the Minerva Center for Marine Invertebrate Immunology and Development Biology and was also supported by the Joint German Israeli project on Biotechnology.

References Abbot, A., 1990. Bioadhesives: potential for exploitation. Sci. Prog. 74, 131- 146. Apte, S., Khoury, F., Roth, W., Schlichter, D., 1996. Transport of amino acids into freshly isolated cells from a sea anemone. Endocytobiosis Cell Res. 11, 129-146. Auffret. M., Oubella, R., 1997. Hemocyte aggregation in the oyster Crassostrea gigas: In vitro measurement of experi-

B. Rinkevich / Journal of Biotechnology 70 (1999) 133-153

mental modulation by xenobiotics. Comp. Biochem. Physiol. 118A, 705-712. Auzoux, S., Domart-Coulon, I., Doumenc, D., 1993. Gill cell cultures of the butterfish clam Ruditapes decussatus. J. Mar. Biotechnol. 1, 79-81. Awaji, M.S., 1991. Separation of outer epithelial cells from pearl oyster mantle by dispase digestion. In: Fraser, M.J. (Ed.), Eighth International Conference on Invertebrate and Fish Tissue Culture. Tissue Culture Association, Columbia, MD, pp. 43-46. Awaji, M., 1997. Primary culture techniques for the outer epithelial cells of pearl oyster mantle. In: Maramorosch, K., Mitsuhashi, J. (Eds.), Invertebrate Cell Culture, Novel Directions and Biotechnology Applications. Science Publishers, NH, USA, pp. 239-244. Battaglia, M., Davoli, C., 1997. Long-term culture of the earthworm Eisenia foetida) (Annelida, Oligochaeta). In: Maramorosch, K., Mitsuhashi, J. (Eds.), Invertebrate Cell Culture, Novel Directions and Biotechnology Applications. Science Publishers, NH, USA, pp. 261-268. Bayne, C.J., 1998. Invertebrate cell culture considerations: insects, ticks, shellfish, and worms. In: Mather. J.P., Barnes, D. (Eds.), Methods in Cell Biology, vol. 57. Academic Press, New York, pp. 187-201. Benson, S., Smith, L., Wilt, F., Shaw, R., 1990. The synthesis and secretion of collagen by cultured sea urchin micromeres. Exp. Cell. Res. 188, 141 - 146. Berdan, R.C., Hauser, G., Bulloch, A.G.M., 1990. Ultrastructure of an identified molluscan neuron in organ culture and cell culture following axotomy. J. Comp. Neurol. 296. 437-446. Bertheussen, K., Seljelid, R., 1978. Echinoid phagocytes in vitro. Exp. Cell Res. 111,401-412. Blinova, M.I., Goryunova, L.B., Leibson, N.L., 1993. Cell culture of regenerating tissues of the sea cucumber Stichopus japonicus. Biol. Morya 2, 84-91. Blisko, R., 1998. Reproduction Processes and Establishment of Cell Culture from Marine Sponges Negombata spp. M.Sc. dissertation, Tel Aviv University, Israel. Boulo, V., Hervio, D., Morvan, A., Bachere, E., Mialhe, E.. 1991. In vitro culture of mollusc hemocytes. Functional study of burst respiratory activity and analysis of interactions with protozoan and procaryotic pathogens. In: Fraser, M.J. (Ed.), Eighth International Conference on Invertebrate and Fish Tissue Culture. Tissue Culture Association, Columbia, MD, pp. 56-64. Bower, S.M., 1995. Parasitic diseases of shell fish. In: Woo. P.T.K. (Ed.), Fish Diseases and Disorders. Protozoan and Metazoan Infections, vol. I. Cab International, NY. USA, pp. 673-728. Brewster, F., Nicholson, B.L., 1979. In vitro maintenance of amoebocytes from the American oyster (Crassostrea virginica). J. Fish. Res. Bd. Can. 36, 461-467. Brody, M.D., Chang, E.S., 1989. Development and utilization of crustacean long-term primary cell cultures: ecdysteroid effects in vitro. Invertebrate Reprod. Dev. 16, 141 - 147.

149

Cancre. I., Van-Wormhoudt, A., LeGal, Y., 1995. Effects of cellular growth factors on crustacean hepatopancreas cell suspensions. J. Mar. Biotechnol. 2, 83-87. Cavalier-Smith. T., Allsopp, M.T.E.P., Chao, E.E., 1994. Thraustochytrids are chromists, not Fungi: 18s rRNA signatures of Heterokonta. Phil. Trans. R. Soc. London B. 346, 387-397. Chang, E.S., 1997. In vitro applications for research on crustacean endocrinology. In: Maramorosch, K., Mitsuhashi, J. (Eds.). Invertebrate Cell Culture, Novel Directions and Biotechnology Applications. Science Publishers, USA, pp. 245-251. Chang, E.S.. Brody, M.D.. 1989. Crustacean organ and cell culture. Adv. Cell Culture 7, 19-86. Chen, S.N., Kou, G.H., 1989. Infection of cultured cells from the lymphoid organ of Penaeus monodon Fabricius by mondon-type baculovirus (MBV). J. Fish Dis. 12, 73-76. Chen, S.N., Jong, K.J., Kou, G.H., 1989. Cell cultures derived from tissues of penaeid shrimp, Penaeus penicillatus, and hard clam, Meretrix lusoria. In: Mitsuhasi, J. (Ed.), Invertebrate Cell System Applications, vol. 2. CRC Press, Boca Raton. FL, pp. 253-262. Cook, I.M.. Graf, R.. Grau, S., Haylet, B., Meyers, D., Ruben, P.. 1989. Crustacean peptidergic neurons in culture show immediate outgrowth in simple medium. Proc. Natl. Acad. Sci. USA 86, 402-406. Cornet, M., 1993. A short-term culture method for chromosome preparation from somatic tissues of adult mussel (Mytilus edulis). Experientia 49, 87-90. Cornet, M.. 1995. A study of some factors affecting the efficiency of in vitro mussel tissue culture for chromosome preparation. Cytobiosis 84, 7-12. Cousserans, F., Bonami. J-R., Comps, M., Grizel, H., 1974. Developpment de protistes parasites en culture cellulaire de l'huitre japonaise Crassostrea gigas. CR Acad. Sci. (Paris) 279. 2119-2121. Custodio. M.R.. Imsiecke, G., Borojevic, M., Rinkevich, B., Rogerson. A.. Miiller, W.E.G., 1995. Evolution of cell adhesion systems: evidence for Arg-Gly-Asp-mediated adhesion in the protozoan Neoparamoeba aestuarina. J. Euk. Microbiol. 42. 721 - 725. Davids. B.J., Yoshino~ T.P., 1998. Integrin-like RGD-dependent binding mechanism involved in the spreading response of circulating molluscan phagocytes. Dev. Comp. Immunol. 22, 39-53. Diekmann-Schuppert, A., Ruppel, A., Burger, R., Frank, W., 1989. Echmococcus multilocularis: in vitro secretion of antigen by hybridomas from metacestode germinal cells and murine tumor cells. Exp. Parasitol. 68, 186-191. Domart-Coulon. I.. Doumenc, D., Auzoux-Bordenave, S., Le Fichant. Y.. 1994. Identification of media supplements that improve the viability of primarily cell cultures of Crassostrea ~igas oysters. Cyotechnology 16, 109-120. Ellender, R.D.. Najafabadi. A.K., Middlebrooks, B.L., 1992. Observations on the primary culture of hemocytes of Penaeus. J. Crustacean Biol. 12, 178-185.

150

B. Rinkevich Journal of Biotechnology 70 (1999) 133-153

Ellis, L.L., Bishop, S.H., 1989. Isolation of cell lines with limited growth potential from marine bivalves. In: Mitsuhashi, J. (Ed.), Invertebrate Cell System Applications, vol. 2. CRC Press, Boca Raton, FL, pp. 243-251. Ellis, L.L., Brodey, M.M., Bishop, S.H., 1985. Preparation of cells from marine molluscs for culture. In: Techniques in the Life Sciences, C1. Elsevier, Ireland, pp. 1-15. Ermak, A.V., Odintsova, N.A., 1996. Effects of adhesive factors on differentiation and growth of embryonic sea urchin cells in primary cultures. Russ. J. Mar. Biol. 22. 338-343. Frank, U., Rinkevich, B. Scyphozoan jellyfish's mesoglea supports attachment, spreading and migration of anthozoans" cells in vitro. Cell Biol. Internat. (in press). Frank, U., Rabinowitz, C., Rinkevich, B., 1994. In vitro establishment of continuous cell cultures and cells lines from ten colonial cnidarians. Mar. Biol. 120, 491-499. Frerichs, G.N., 1996. In vitro culture of embryonic cells from the freshwater prawn Macrobrachium rosenbergii. Aquacultue 143, 227-232. Gaino, E., Magnino, G., 1994. Various morphological aspects of dissociated sponge cells in relation to different biological substrata. In: van Soest, R.W.M., van Kempen, T.M.G., Braekman, J.C. (Eds.), Sponges in Time and Space. Balkema, Rotterdam, The Netherlands, pp. 365-370. Gaino, E., Magnino, G., 1996. Effects of exogenous cAMP on the morphology and behavior of dissociated cells of the sponge Clathrina cerebrum (Porifera, Calcarea). Eur. J. Cell Biol. 70, 92-96. Gates, R.D., Muscatine, L., 1992. Three methods for isolating viable anthozoan endoderm cells with their intracellular symbiotic dinoflagellates. Coral Reefs 11, 143-145. Gates, R.D., Baghdasarian, G., Muscatine, L., 1992. Temperature stress causes host cell detachment in symbiotic cnidarians: implication for coral bleaching. Biol. Bull. Mar. Biol. Lab. Woods Hole 182, 324-332. Ghosh, D., Dasmahaparta, A.K., Ray, A.K.. 1995. Primary culture of prawn hepatocytes in serum free media. In vitro Cell. Dev. Biol. 31 A, 811-813. Gomot, L., 1971. The organotypic culture of invertebrates other than insects. In: Vago, C. (Ed.), Invertebrate Tissue Culture. Academic Press, New York, pp. 41-136. Goodwin, R.H., 1991. Impediments to the culture of nonfibroblastoid cells from insects and other invertebrates, and their resolution. In: Fraser, M.J. (Ed.), Eighth International Conference on Invertebrate and Fish Tissue Culture. Tissue Culture Association, Columbia, MD. pp. 36-42. Greber, M.J., David, C.N., Holstein, T.W., 1992. A quantitative method for separation of living Hydra cells. R oux's Arch. Dev. Biol. 201, 296-300. Hansen, E., 1976. A cell line from embryos of Biomphalaria glabrata (Pulmonata): establishment and characteristics. In: Maramorosch, K. (Ed.), Invertebrate Tissue Culture: Research Applications. Academic Press, New York. pp. 75-97. Hink, W.F., 1979. Cell lines from invertebrates. In: Jakoby. W.B, Pastan, I.H. (Eds.), Methods in Enzymology: Cell Culture. Academic Press, New York, pp. 450-466.

Holden, M.J.. Patterson, G.W., 1991. Absence of sterol biosynthesis in oyster tissue culture. Lipids 26, 81-82. Hsu, Y.-L., Yang, Y-H., Chen, Y.-C., Tung, M.-C., Wu, J.-L., Engelking, M.H., Leong, J.C., 1995. Development of an in vitro subculture system for the oka organ (lymphoid tissue) of Penaeus monodon. Aquaculture 136, 43-55. Hu, K., 1990. Studies on a cell culture from the hepatopancreas of the oriental shrimp, Penaeus orientalis Kishinouge. Asian Fish. Sci. 3, 299-307. Ilan. M., Contini, H., Carmeli, S., Rinkevich, B., 1996. Progress towards cell cultures from a marine sponge that produces bioactive compounds. J. Mar. Biotechnol. 4, 145-149. Imsiecke, G.. Steffen, R., Custodio, M., Borojevic, R., Mfiller, W.E.G., 1995. Formation of spicules by sclerocytes from the freshwater sponge Ephydatia muelleri in short term cultures in vitro. In Vitro Cell. Dev. Biol. 31A, 528-535. Itami, T., Aoki, Y., Hayashi, K.I., Yu, Y., Takahashi, Y., 1989. In vitro maintenance of cells of lymphoid organ in Kuruma shrimp Penaeus japonicus. Nippon Suisan Gakkaishi. Bull. Jpn. Soc. Sci. Fish. 55, 2205. Kaneko, H., Kawahara, Y., Dan-Sohkawa, M., 1995. Primary culture of mesodermal and endodermal cells of the starfish embryo. Zool. Sci. 12, 551-558. Kaneko. H., Kawahara, Y., Okamato, M., Dan-Sohkawa, M.D., 1997. Study on the nature of starfish larval muscle cells in vitro. Zool. Sci. 14, 287-296. Kawamura, K.. Fujiwara, S., 1995. Establishment of cell lines from multipotent epithelial sheet in the budding tunicate, Polyandrocarpa misakiensis. Cell Struct. Funct. 20, 97-106. Ke, H., kiping. W., Yumei, D., Shuji, Z., 1990. Studies on a cell culture from the hepatopancreas of the oriental shrimp, Penaeus orientalis Kishinouye. Asian Fish. Sci. 3, 299-307. Klautau. M., Custodio, M.R., Borojevic, R., 1993. Cell culture of sponges (Clathrina and Polymastia). In Vitro Cell. Dev. Biol. 29A, 97-99. Klautau, M., Custodio, M.R., Borojevic, R., 1994. In vitro culture of primary cell lines from marine sponges. In: van Soest. R.W.U., van Kempen, T.M.G., Braekman, J.C. (Eds.), Sponges in Time and Space. Balkema, Rotterdam, The Netherlands, pp. 401-406. Kleinschuster, S.J., Parent, J., Walker, C.W., Farley, C.A., 1996. A cardiac cell line from Mya arenaria (Linnaeus, 1759). J. Shellfish Res. 15, 695-707. Kumazawa, N.H., Tanigawa, T., Tanaka, Y., Osatake, H., Tanaka, K., 1990. Preliminary study on culture and morphology of hemocytes of two gastropod molluscs, Clithon rectropictus and Nerita albicilla. Venus 49, 233-239. Laursen, J.R.. Liu, H-d., Wu, X-J., Yoshino, T.P., 1997. Heat-shock response in a molluscan cell line: characterization of the response and cloning of an inducible HSP70 cDNA. J. Invertebrate Pathol. 70, 226-233. Lebel, J-M., Giard. W., Favrel, P., Boucaud-Camou, E., 1996. Effects of different vertebrate growth factors on primary cultures of hemocytes from the gastropod mollusc, Haliotis tuberculata. Biol. Cell 86, 67-72.

B. Rinkevich Journal of Biotechnology 70 (1999) 133-153 Leys, S.P., 1997. Sponge cell culture: a comparative evaluation of adhesion to a native tissue extract and other culture substrates. Tissue Cell 29, 77-87. Leys, S.P., 1998. Fusion and cytoplasmic streaming are characteristics of at least two hexactinellids: examination of cultured tissue from Aphrocallistes vastus. In: Watanabe. Y., Fusetani, N. (Eds.), Sponge Sciences--Multidisciplinary Perspectives. Springer, Tokyo, pp. 215-226. Li, M.F., Stewart, J.E., Drinnan, R.E., 1966. In vitro cultivation of cells of the oyster, Crassostrea t'irginica. J. Fish. Res. Bd. Can. 23, 595-599. Loh, P.C., 1993. Primary culture of lymphoid, nerve, and ovary cell from Penaeus stylirostris and Penaeus vannamei. In Vitro Cell. Dev. Biol. 29A, 620-622. Loh, P.C., Tapay, L.M., Lu, Y., 1997. Quantal assay of shrimp viruses in primary shrimp lymphoid cell cultures. In: Maramorosch, K., Mitsuhashi, J. (Eds.), Invertebrate Cell Culture, Novel Directions and Biotechnology Applications. Science Publishers, NH, USA, pp. 253-259. Lu, Y., Tapay, L.M., Loh, P.C., Brock, J.A., Gose. R., 1995. Development of a quantal assay in primary shrimp culture for yellow-head baculovirus (YBV) of penaeid shrimp. Virmet 52, 231-236. Luedman, R., Lightner, D.V., 1992. Development of an in vitro primary cell culture from the penaeid shrimp, Penaeus stylirostris and Penaeus vannamei. Aquaculture 101, 205-211. Machii, A., 1974. Organ culture of mantle tissue of the pearl oyster Pinctata fucata (Gould). Bull. Nat. Pearl Res. Lab. 18, 2111-2117. Machii, A., 1988. Pearl formation studies in vitro. Commun. Appl. Cell Biol. 6, 1-9. Machii, A., Wada, K.T., 1989. Some marine invertebrates tissue culture. In: Mitsuhasi, J. (Ed.), Invertebrate Cell System Applications, vol. 2. CRC Press, Boca Raton, FL, pp. 225- 233. Mitsuhashi, J. (Ed.), 1989. Invertebrate Cell System Applications, vols. I, II. CRC Press, Boca Raton, FL. Nadala, E.C., Lu, Y., Loh, P.C., 1993. Primary culture of lymphoid, nerve and ovary cells from Penaeus stylirostris and Penaeus vannamei. In Vitro Cell. Dev. Biol. 29A, 620-622. Naganuma, T., Degan, B.M., Horikoshi, K, Morse, D.E., 1994. Myogenesis in primary cell cultures from larvae of the abalone, Haliotis rufescens. Mol. Mar. Biol. Biotech. 3, 131-140. Naganuma, T., Takasugi, H., Kimura, H., 1998. Abundance of thraustochytrids in coastal plankton. Mar. Ecol. Prog. Ser. 162, 105-110. Najafabadi, A.K., Ellender, R.D., Middlebrooks, B.L., 1992. Analysis of shrimp hemolymoph and ionic modification of a Penaeus cell culture formulation. J. Aquat. Anim. Health 4, 143-148. Noel, D., Moore, J., Miahle, E., Boulo, V, Elston, R., 1991. Study of neoplastic hemocytes in the bivalve mollusc Mytilus. In: Eighth International Conference on Invertebrate and Fish Tissue Culture. Tissue Culture Association, Columbia, MD, pp. 47-55.

151

Odintsova, N.A., Khomenko, A.V., 1991. Primary cell culture from embryos of the Japanese scallop Mizuchopecten yessoensis (Bivalvia). Cytotechnology 6, 49-54. Odintsova. N., Tsal, L.. 1995. Cryopreservation of primary cell cultures of bivalvia. Cryo-Letters 16, 13-20. Odintsova. N.A.. Nesterov, A.M., Korchagina, D.A., 1993. A growth factor from tissue of the mussel Mytilus edulis. Comp. Biochem. Physiol. 105A, 667-671. Odintsova, N.A., Ermak, A.V., Tsal, L.G., 1994. Substrate selection for long-term cultivation of marine invertebrate cells. Comp. Biochem. Physiol. 107A, 613-619. Osinga, R., Tramper, J., Wijffels, R.H., 1988. Cultivation of marine sponges for metabolic production: applications for biotechnology? Trends Biotechnol. 16, 130-135. Peddie, C.M., Richest, A.C., Smith, V.J., 1995. Proliferation of undifferentiated blood cells from the solitary ascidian, Ciona intestinalis in vitro. Dev. Comp. Immunol. 19, 377387. Poccia, D.. 1988. In vitro differentiation of male germ line cells from sea urchin testis. J. Exp. Zool. 246, 57-64. Pomponi, S.A.. Willoughby, R., 1994. Sponge cell culture for production of bioactive metabolites. In: van Soest, R.W.M., van Kempen, T.M.G., Braekman, J.C. (Eds.), Sponges in Time and Space. Balkema, Rotterdam, The Netherlands, pp. 395-400. Pomponi, S.A., Willoughby, R., Edward Kaighn, M., Wright, A.E., 1997. Development of techniques for in vitro production of bioactive natural products from marine sponges. In: Maramorosch, K.. Mitsuhashi, J. (Eds.), Novel Directions and Biotechnology Applications. Science Publishers, NH, USA. pp. 231-237. Porter, D.. 1990. Phylum Labryinthulomycota. In: Margulis, L., Corliss. J.O., Melkonian, M., Chapman, D.J. (Eds.), Handbood of Protoctista. Jones and Bartlett, Boston, pp. 388-398. Przysiezniak. J., Spencer. A.N., 1989. Primary culture of identified neurones from a cnidarian. J. Exp. Biol. 142, 97-113. Raftos, D.A., Cooper, E.L., 1991. Proliferation of lymphocyte-like cells from the solitary tunicate, Styela clara, in response to allogeneic stimuli. J. Exp. Zool. 260, 391-400. Raftos, D.A.. Cooper, E.L., 1996. Tunicate immunobiology: Factors that influence cell proliferation and activation. In: Maramorosch, K., Loeb, M.S. (Eds.), Invertebrate Cell Culture: Looking Toward the 21st Century. Society for In Vitro Biology, USA. pp. 11-20. Raftos, D.A.. Stillman. D.L., Cooper, E.L., 1990. In vitro culture of tissue from the tunicate Styela clava. In Vitro Cell. Dev. Biol. 26A, 962-970. Raftos, D.A. Cooper, E.L., Habicht, G.S., Beck, G., 1991. Invertebrate cytokines: tunicate cell proliferation stimulated by an interleukin l-like molecule. Proc. Natl. Acad. Sci. USA 88. 9518-9522. Raghukumar. S., 1988. Detection of the thraustochytrid protist Ulkenia visurgensis in a hydroid, using immunofluorescence. Mar. Biol. 97. 253-258.

152


Raghukumar, S., 1992. Bacterivory: a novel dual role for thraustochytrids in the sea. Mar. Biol. 113, 165-169. Raghukumar, S., Balasbramanian, R., 1991. Occurrence of thraustochytrid fungi in corals and coral mucus. Indian J. Mar. Sci. 20, 176-181. Raghukumar, S., Schaumann, K., 1993. An epifluorescence microscopy method for direct detection and enumeration of the fungilike marine protists, the thraustochytrids. Limnol. Oceanogr. 38, 182-187. Rannou, M., 1968. Formation de spicules dans des cultures cellulaires de cnidaire (gorgone). Vie Millieu 19, 53-57. Rannou, M., 1971. Cell culture of invertebrates other than molluscs and arthropods. In: Vago, C. (Ed.), Invertebrate Tissue Culture, vol. 1. Academic Press, New York, pp. 385-410. Renault, T., Flaujac, G., Le Deuff, R.-M., 1995. Isolation and culture of heart cells from the European fiat oyster, Ostrea edulis. Methods Cell Sci. 17, 199-205. Rinkevich, B., Rabinowitz, C., 1993. In vitro culture of blood cells from the colonial protochordate Botryllus schlosseri. In Vitro Cell. Dev. Biol. 29A, 79-85. Rinkevich, B., Rabinowitz, C., 1994. Acquiring embryoderived cell cultures and aseptic metamorphosis of larvae from the colonial protochordate Botryllus schlosseri. Invertebrate Reprod. Dev. 25, 59-72. Rinkevich, B., Rabinowitz, C., 1997. Initiation of epithelial cell cultures from palleal buds of Botryllus schlosseri, a colonial tunicate. In Vitro Cell. Dev. Biol. 33A, 422-424. Rinkevich, B., Frank, U., Gateno, D., Rabinowitz, C., 1994. The establishment of various cell lines from colonial marine invertebrates. In: Miiller, W.E.G. (Ed.), Use of Aquatic Invertebrates as Tools for Monitoring of Environmental Hazards. Gustav Fischer, Stuttgart, pp. 253-263. Rinkevich, B., Shlemberg, Z., Fishelson, L., 1996. Survival budding in colonial tunicate Botrylloides from the Mediterranean Sea: the role of totipotent blood cells. In: Maramorosch, K., Loeb, M.J. (Eds.), Invertebrate Cell Culture: Looking Towards the 21st Century. Society for In Vitro Biology, USA, pp. 1-9. Rinkevich, B., Blisko, R., Ilan, M., 1998. Further steps in the initiation of cell cultures from embryos and adult sponge colonies. In Vitro Cell. Dev. Biol. 34, 753-756. Rosenfield, A. (Ed.), 1993. Marine Invertebrate Cell Culture: Breaking the Barriers. NOAA Technical Memorandum NMFS-F/NEC-98, 25 pp. Rosenfield, A., Kern, F.G., Keller, B.J. (Eds.), 1994. Invertebrate Neoplasia: Initiation and Promotion Mechanisms. NOAA Technical Memorandum NMFS-NE-107, 31 pp. Rosenthal, J., Diamant, A., 1990. In vitro cell cultures from Penaeus semisculcatus. In: Perkins, F.O., Cheng, T.C. (Eds.), Pathology in Marine Science. Academic Press, New York, pp. 7-13. Samata, T., Somiya, H., Hotira, C., Akera, S., 1994. SEM Observation of microcrystals developed over black secretion on the cultured tissue of the pearl oyster Pinctada fucata. Fish. Sci. 60, 343-344.

Sawada, T., Zhang, J., Cooper, E.L., 1994. Sustained viability and proliferation of hemocytes from the cultured pharynx of Styela clara. Mar. Biol. 119, 597-603. Saxena, A.K., Ramchandani, S., Dwivedi, A., Sharma, R., Bajpai, V.K., Bhardwaj, K.R., Balapure, A.K., 1995. Simplified cryopreservation of mammalian cell lines. In Vitro Cell. Dev. Biol. 31A, 326-329. Schmid, V., 1974. Structural alternations in cultivated straited muscle cells from anthomedusae (Hydrozoa). Exp. Cell Res. 86, 193-198. Schmid, V., Bally, A., 1988. Species specificity in cell-substrate interactions in medusae. Dev. Biol. 129, 573-581. Shiirmann, W., Peter, R., 1988. Isolation and cultivation of planarian neoblasts by a novel combination of methods. In: Ax, P., Ehlers, U., Sopott-Ehlers, B. (Eds.), Free Living and Symbiotic Plathelminthes. Gustav Fischer, New York, pp. 111-114. Stiles, B., McDonald, I.C., Gerst, J.W., Adams, T.S., Newman, S.M., 1992. Initiation and characterization of five embryonic cell lines from the cotton boll weevil Anthonomus grandis in a commercial serum-free medium. In Vitro Cell. Dev. Biol. 28A, 355-363. Takeuchi, Y., Yamamoto, S., Odo, S., 1994. Primary and secondary cultures of larval cells of Pacific oyster, Crassostrea gigas. J. Mar. Biotechnol. 1, 171-175. Takeuchi, Y., Yamamoto, S., Odo, S., 1995a. Primary culture of larval cells of the mussel, Mytilus edulis galloprovincialis. J. Mar. Biotechnol. 2, 51-54. Takeuchi, Y., Yamamoto, S., Odo, S., 1995b. Primary culture of larval cells of the mussel, Mytilus edulis galloprovincialis. J. Mar. Biotechnol. 2, 51-54. Tamse, C.T., Collin, C., Kuzirian, A.M., Smith, P.J.S., 1995. Culture of Hermissenda crassicornis (Mollusca) neurons. In Vitro Cell. Dev. Biol. 3A, 653-656. Tapay, L.M., Lu, Y., Brock, J.A., Nadala, E.C., Loh, P.C., 1995. Transformation of primary cultures of shrimp (Penaeus stylirostris) lymphoid (Oka) organ with simian virus40 (T) antigen. Proc. Soc. Exp. Biol. Med. 209, 73-78. Tapay, L.M., Lu, Y., Gose, R.B., Nadala, E.C.B., Brock, J.A., Loh, P.C., 1997. Development of in vitro quantal assay in primary cell cultures for a non-occluded baculo-like virus of penaeid shrimp. J. Virol. Methods 64, 37-41. Toledo, A., Cruz, C., Fragosco, G., Laclette, J.P., Merchant, M.T., Hernandez, H., Sciutto, E., 1997. In vitro culture of Taenia crassiceps larval cells and cyst regeneration after injection into mice. J. Parasitol. 83, 189-193. Tong, S.-L., Miao, H.-Z., 1996. Attempts to initiate cell cultures from Panaeus chmensis tissues. Aquaculture 147, 151-157. Toullec, J.Y., Crozat, Y., Patrois, J., Porcheron, P., 1996. Development of primary cell cultures from the penaeid shrimps Penaeus vannamei and P. indicus. J. Crust. Biol. 16, 643-649. Townsel, J.G., Thomas, W.E., 1987. On the status of the study of invertebrate neurons in tissue culture phyla Mollusca and Annelida. Comp. Biochem. Physiol. 86, 199-208.

B. Rinkevich /Journal of Biotechnology 70 (1999) 133-153 Walker, C.W., Key, S.A., Mulkern, J.E., Verma, S., Jacobs, J.A., 1996. Expression of the tumor suppresser gene p53 in normal and leukemia clam blood cells in vivo and in vitro. J. Shellfish Res. 15, 520. Weber, J., 1995. Novel tools for the study of development, migration and turnover of nematocytes (cnidarian stinging cells). J. Cell Sci. 108, 403-412.

153

Wen, Ch.-M., Kou, G.-H., Chen, S.-N., 1993. Establishment of cell lines from the Pacific oyster. In Vitro Cell. Dev. Biol. 29A, 901-903. Yoshino, T.P., gaursen, J.R., 1995. Production of Schistosoma incensoni daughter sporocysts from mother sporocysts maintained in synxenic culture with Biomphalaria glabrata embryonic (Bge) cells. J. Parasitol. 81, 714-722.


lul

/OUIINAL

OF



Growth of the sponge Pseudosuberites (aft.) andrewsi in a closed system Ronald Osinga *, Peter B. de Beukelaer, Ellen M. Meijer, Johannes Tramper, Ren6 H. Wijffels Wageningen University, Food and Bioprocess Engineering Group, PO Box 8129, 6700 EV, Wageningen, The Netherlands


Abstract

Explants of the Indo-Pacific sponge Pseudosuberites aft. andrewsi were fed with the microalgae Chlorella sorokiniana and Rhodomonas sp. It was microscopically observed that these algae were ingested and digested by the sponge cells, suggesting that they were consumed by the sponges. The algae were further used for two growth experiments with five explants of P. aff. andrewsi and four explants of P. andrewsi. Growth was measured as the increase in projected body area. The explants showed considerable growth (up to 730% in 54 days for P. aft. andrewsi and up to 680% in 22 days for P. andrewsi), which is much higher than previously reported growth rates for sponges. Growth started after a stationary phase of 5-20 days in which the projected body area did not increase. The growth of P. aff. andrewsi appeared to be linear and was inhibited at the end of the experiment. Two explants of P. andrewsi showed exponential growth instead of linear growth. Hence, no general statements about the growth kinetics of these sponges can be made at this time. However, the high growth rates found in this study suggest a promising future for cultivation of sponges in closed systems. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Sponges; In vivo cultivation; Pseudosuberites (aft.) andrewsi; Growth

1. Introduction

Sponges (Porifera) are primitive multicellular aquatic animals. They live attached to solid substrata, such as rocks, corals or shells, and feed upon small particles (microalgae, bacteria and dead organic material), which they filter out of the surrounding water. Most of the sponge species * Corresponding author. E-mail address: [email protected] (R. Osinga)

described so far live in marine environments. The most characteristic feature of the sponge body is the so-called aquiferous system, a network of channels and chambers through which water flows continuously. This water current is generated by flagellated cells (choanocytes) that line the walls of the chambers. The space between the channels and chambers is called the mesohyl: a gelatinous matrix, containing free-floating cells (archaeocytes) and skeletal material. The choanocytes not only generate a water current, but these cells also trap and ingest the food


156

R. Osinga et al./Journal of Biotechnolog)' 70 (1999) I55-161

particles. Some of the particles are transferred to the archaeocytes, whose role it is, among others, to generate metabolic energy. During the past few decades, marine sponges have been recognised as potential candidates for biotechnology, due to the numerous potentially interesting novel compounds that have been isolated from this group of animals (Garson, 1994; Munro et al., 1994). However, the methodology to produce large amounts of sponge biomass is still in its infancy. Some progress has been made with sponge aquaculture for drug production (Battershill and Page, 1996; Duckworth et al., 1997), but techniques to produce sponge biomass under completely controlled conditions in bioreactors are not available at present (Osinga et al., 1998). Perhaps the single most important problem to address when developing an in vivo cultivation system is to provide a suitable food regimen. In situ, sponges feed rather unselectively upon the complex mixture of particles that is available in natural seawater. Selecting those food particles that provide the sponge with all their metabolic requirements may be the key to a successful in vivo sponge culture. In our laboratory, the Indo-Pacific demosponge species Pseudosuberites andrewsi (Kirkpatrick) is used as a model species for the development of such an in vivo cultivation method. This species was maintained in controlled systems for more than a year, using two microalgal species (the freshwater chlorophycean Chlorella sorokiniana and the marine cryptophycean Rhodomonas sp.) as a food source (Osinga et al., 1999). These algae have also been used in previous studies to feed cultures of the temperate marine sponge Halichondria panicea (Barthel and Theede, 1986; Thomassen and Riisggtrd, 1995). It is the aim of the present study to verify if the chosen food particles are indeed ingested by the sponge cells. In addition, we will quantify the growth rates and describe the growth kinetics of explants of P. andrewsi and Pseudosuberites aff. andrewsi cultivated under controlled conditions in a bioreactor using C. sorokiniana and Rhodomonas sp. as the food source.


2. I. Sponges Fresh material of P. andrewsi was obtained from Blijdorp Zoo (Rotterdam, The Netherlands), where it was growing in a large, shallow basin, in which a strong water current was generated to simulate an intertidal environment. More sponge material was obtained from Artis Zoological Garden, Amsterdam. Here, the sponges grow in the central filtration system of the tropical marine aquaria. This biofilter consists of small pebbles (grain size approximately 5 mm). The sponges grow on top and within the first 2 cm of this pebble layer. All the organic dirt from the aquaria (food remainders, organic wastes, bacteria) is deposited onto this filter, which is therefore a good locality for the sponges in terms of food availability. The sponge material obtained from Artis Zoo was classified as P. aff. andrewsi. The spicules of these sponges resembled the form of the spicules of P. andrewsi as described by Kirkpatrick (1900), but were considerably smaller ( ~ 200 ~m instead of 350 lam). In our laboratory, the sponges were held in 200 dm 3 air-lift bioreactors containing artificial seawater (using Instant Ocean Reef Crystals artificial sea salt) with a salinity of ~ 32%o. This water was replaced continuously ( D = 0 . 0 3 3 day-~). The temperature in the bioreactor varied between 25 and 29~ In order to provide the sponges with a source of silica, 0.25 mmole d m - 3 Na203Si-9H20 was added to the artificial seawater. Measurements of the silica concentration in the outflowing water showed that this addition was sufficient to cope with the sponges' demands. Non-axenic batch cultures of C. sorokiniana (average size ,-~ 3 ~m) and Rhodomonas sp. (average size ~ 6 ~m) were added as a food source for the sponges. Twice a week, 1 dm 3 of a culture of Chlorella sorokiniana was added, containing ~ 1 x 107 cells cm -3. In addition, 1 dm 3 of a culture of Rhodomonas sp., containing ~ 1 • 106 cells cm-3, was added weekly. The algae were cultured at a temperature varying between 17 and 20~ A light-dark cycle of 14 h light and 10 h darkness was applied. The growth media for the algae are

R. Osinga et al./Journal of Biotechnology 70 (1999) 155-161

given in Table 1. When the cultures were added to the sponges, the algae were usually near the end of their logarithmic growth phase.

2.2. Particle ingestion studies To study the ingestion of the supplied food particles by the sponges, small pieces of sponge tissue were incubated in 100 cm 3 aerated, stirred beaker glasses filled with artificial seawater. A few cm 3 culture of either C. sorokiniana or Rhodomonas sp. was added as a food source. Dissociated cells obtained from the incubated sponges were microscopically examined for the presence of algal material. This was done before feeding, 2 h after feeding and 24 h after feeding. Dissociated sponge cells were obtained from the sponges by putting the sponge tissue into a petri disc with Ca 2 + and Mg 2+ free Artificial Seawater (CMFS) and by cutting it into smaller pieces with Table 1 Growth media for the algae (a freshwater medium for C. sorokiniana and a seawater medium for Rhodomonas sp.) a Component

NaHCO 3 KNO 3 NaH2PO 4 Instant Ocean Reef Crystals artificial seasalt MgSOa'7H20 CaCI2"2H20 EDTANa2.2H20 FeC13 Na2B407" 10H20 ZnSO4"7H20 CuSO4"5H20 MnSO4"H20 Na2MoOa'2H20 NiSOa'6H20 NaVO 3 Thyamin-HCl Cyanocobalamin Biotin

Freshwater medium concentration 10.0 1.00 0.10

Seawater mediumconcentration 5.00 0.50 0.05 ~ 3 3 g dm -3

4.99 0.272 0.391 0.148 4.72 x 10 -2 3.13 x 10 -2 3.20 x 10 -2 3.59 • 10 -2 2.07 x 10 -2 2.85 • 10 -3 2.85 x 10 -3 5.93 • 10 -5 5.90 • 10 -6 1.64 • 10 -6

aThe freshwater medium was based on the A9 medium described by Lee and Pirt (1981). Concentrations are given in mM, unless indicated otherwise.

157

a razor blade. C M F S is used to prevent reaggregation of the dissociated sponge cells. Ca and Mg ions play a crucial role in the reaggregation process (Miiller, 1982). The cell and tissue suspension was sieved through a 70 ~tm mesh to obtain a cell suspension. This cell suspension was centrifuged for 5 min at 600 • g at a temperature of 10~ The pellet was resuspended in a few cm 3 of C M F S and this suspension was examined under a light microscope.

2.3. Growth experiment An experiment to determine the growth rate of P. aff. andrewsi was performed with five explants (small cuttings of sponge colonies). These explants were prepared using razor-sharp knifes and were obtained from different sponge colonies. However, they probably all originate from a single individual, coincidentally introduced into the aquaria of the zoo. The pieces of sponge tissue were tied onto glass slides with nylon fishing-line. The explants were placed in temperature controlled, 1.58 dm 3 bioreactors, equipped with a sparger for air supply and a magnetic stirrer to keep the food particles in suspension. A second, repetitive experiment was done with four explants from P. andrewsi, which were all obtained from a single sponge colony. The sponges were fed with C. sorokiniana (twice a week, 50 cm 3) and Rhodomonas sp. (once a week, 50 cm 3) using material from the batch cultures that were described in the previous section. Temperature and salinity in the bioreactor was kept constant at 25~ and 34 %0, respectively. The first experiment was run for a period of 54 days, and the second experiment for a period of 22 days. To determine growth, the size of the explants was followed during the experiment. Explant sizes were measured as a two-dimensional projections of body area, which were determined from photographs. A similar method was used by Ayling (1983) to measure in situ growth and regeneration of sponges. The advantage of this method is that the sponges used in experiments do not have to be removed from the water. Exposure to air can cause serious damage to sponge tissue (Foss~ and

158

R. Osinga et al. Journal of Biotechnology 70 (1999) 155-161

Fig. 1. Explants of P. aft. andrewsiplaced on the rack designed for photography of the body area. The black dots indicate a marked distance. Nilsen, 1996), and may therefore affect the growth rate that is to be determined. We have evaluated the pros and cons of this method, and compared it with another, novel method to measure sponge biomass: determination of the underwater weight. The results of this comparative study will be described elsewhere (Osinga et al., 1999). Photographs of the explants were taken on days 0, 5, 8, 19, 33, 39, 47 and 54 in the first experiment (P. aff. andrewsi) and on days 1, 7, 10, 20 and 22 in the second experiment (P. andrewsi ). The explants (always kept underwater) were placed onto a rack on which black dots were painted that indicated a known distance (Fig. 1). Photographs were taken under a straight angle with a digital camera (Hewlett Packard PhotoSmart Model C5340A). The digital images were printed, the areas of the sponges were cut out with scissors and these cuttings were weighed. The weights of these cuttings were converted to areas by comparing them with the weight of a cutting of a known area. The thus obtained values were converted to real surface areas using the marked distances on the rack as a reference.


3.1. Particle &gestion studies Usually, a change in colour of the outer appearance of the sponges was observed approximately 1 h after feeding. The colour changed from

yellow into green after feeding with the greencoloured C. sorok&iana, and from yellow into brown after feeding with the red-coloured Rhodomonas sp. This indicated that the food particles accumulated in the sponge body. The changes in colour were also observed regularly in the large bioreactor and may be used as an indication of the condition of the sponges. Sponges that change colour rapidly exhibit a strong pumping activity. The microscopic observations on dissociated cells showed that both Chlorella and Rhodomonas cells were taken up by the archaeocytes of P. andrewsi (Fig. 2B,C). Complete algal cells were detected in many sponge cells 2 h after feeding. Thus, ingestion by choanocytes and transfer to archaeocytes takes place within 2 h. This is comparable with the rate at which the freshwater algae Chlamydomonas re&hardtii was ingested and transferred by the freshwater sponge Spongilla lacustris (Imsiecke, 1994). After 24 h, some sponge cells still contained intact algae, but other cells only exhibited a green or red colour, indicating that the algae were being digested. It can therefore be concluded from the microscopic observations that P. andrewsi is able to feed upon C. sorokiniana and Rhodomonas sp. It remains to be studied further whether these algae are also the most suitable food source for this sponge species.

3.2. Growth experiment The obtained growth curves clearly show that the sponges P. aft. andrewsi (Fig. 3A) and P. andrewsi (Fig. 3B) are able to grow in a closed system. Unfortunately, in situ growth rates for P. andrewsi are not available. We compared our results with growth rates from other sponge species that were described in the literature (Table 2). In this table, the total growth of the explants in Fig. 3 during the whole experimental periods (54 and 22 days, respectively) is expressed as a percentage of the initial body size. The growth rates for P. (aft.) andrewsi obtained in our systems are much higher than the literature values in Table 2. The five explants of P. aff. andrewsi (Fig. 3A) all exhibited a similar growth pattern: a stationary

R. Osinga et al. ,'Journal of Biotechnology 70 (1999) 155-161

phase during the first 5-20 days, followed by a period in which the projected body area increased rather linearly. At the end of the experiment, the growth rate tended to decrease, sometimes even to zero. We have no explanation for the irregular pattern observed for explant 4. The peak at day 8 will not be further discussed. The explants of P. andrewsi in the second experiment (Fig. 3B) also showed a stationary phase of 7 or 10 days followed by a period of considerable growth. However, a somewhat different growth pattern was observed: explants 1 and 2 exhibited exponential growth, explant 3 did not grow at all, and explant 4 grew irregularly. A uniform description of the growth kinetics of these sponges cannot be deduced from these results. A specific growth rate expressed in percentage per day can only be calculated from the curves obtained for explants 1 and 2 in Fig. 3b, as only these data can be described by the function In X / X o =/ut (in which X is the amount of sponge biomass, X0 is the amount of sponge biomass at t = 0 and /~ is the growth rate in percentage per day). The calculated growth rates were 0.08 day-~ and 0.11 day-~ for explants 1 and 2, respectively. This is again higher than previously reported growth rates, which range from 0.01 to 0.058 day-~ (Thomassen and Riisg~.rd, 1995).

159

The stationary phase that was observed in both experiments could be a response of the sponge to the cutting procedure: the tissue has to rearrange into a functional sponge. By determination of the underwater weight, it was found that most explants even lost some weight during the stationary phase (Osinga et al., 1999), which indicates that the first days after the cutting procedure are indeed a period of adaptation. At the end of the experiment with P. aff. andrewsi, the growth rate decreased. There may be two explanations for this decrease. Sponges are typical modular organisms (Kaandorp 1991), i.e. they grow by repeated production of nearly identical multicellular structures (modules). In the species studied by Kaandorp (1991), these modules are determined by the skeletal structure: a new skeleton module is built upon the previous layer. Other examples of modular growth may be the formation of new channels and choanocyte chambers, or the formation of a new osculum. The modular growth may result in periods of growth and periods of tissue rearrangement. During the latter, the increase in sponge biomass will be lower, and this may have occurred at the end of the experiment. A second explanation for the decreased growth at the end of the experiment may be found in growth limitation. The food regimen was not

~ D

Fig. 2. Microscopic view of dissociated archaeocyte cells: (A) before feeding: (B) 2 h after feeding with C. sorokiniana; (C) 2 h after feeding with Rhodomonas sp.; (D) 24 h after feeding with Rhodomonas sp.

R. Osinga et al./Journal of Biotechnology 70 (1999) 155-161

160

with batch cultures will result in short periods of starvation, which may not be in favour of optimal growth. The fact that all explants stopped growing at the end of the experiment supports the suggestion of growth limitation. The growth of P. aff. andrewsi should therefore be further studied under conditions of continuous food supply, thus excluding the possibility of food limitation.

A: Pseudosuberites aft. andrewsi ~e~explantl --@--explant 2 -,W--explant 3 -"B---explant 4 --'0--explant 9 5

A 04

[ ] [ ]

< 6 E o v

1. The radial flow velocities may be estimated from radial dispersion coefficients in turbulent flow; hence DR URoc~

(51)

dr"

dk = d r - 2dL.

(43)

Thus, the volume of the dark zone per unit tube length is Dark volume = i t ( d r - 2dL) 2 9 4

(44)

If 0a is the maximum acceptable duration of the dark period between successive light periods, then

Radial dispersion coefficients can be measured using suitably designed experiments. Correlations would need to be established for such dispersion coefficients as functions of the linear flow Reynolds number. In practice, radial flow may be enhanced by deploying static mixers inside a tube (Chisti, 1998). These mixers should be minimally intrusive

244

E. Molina Grima et al./Journal of Biotechnology 70 (1999) 231-247

and should be confined to well within the dark core. One possibility is the use of a coaxially located rod with suitably spaced barbs or projections that have a somewhat flat profile and are suitably angled to direct the flow into the annular light zone. The projections should not extend into the light zone to prevent any loss of illumination in that zone. If for an algal species, the acceptable continuos dark time 0a is infinitesimally short, then scaleup without loss of productivity would be impossible by increasing the tube diameter. In theory, exceedingly high levels of turbulence and high radial velocities can be generated. In practice, an upper limit would be encountered at much lower than technically possible levels of turbulence in view of the known algal sensitivity to hydrodynamic shear and limited pressure tolerance of the typically employed transparent materials of construction. In addition, higher airlift devices would be necessary to generate higher flows through the tubes while compensating for the significant pressure drop due to static mixing elements (Chisti, 1989; Chisti et al., 1990).

6. Nomenclature a ai

b

CT r

D DR

( m 2 S-1)

dT

dks dL

5. Concluding remarks

Fco Of the many types of photobioreactors proposed for closed monoculture, tubular devices are amongst the more scaleable and suited to largescale production. Unlike the widely used open culture systems, the design of closed tubular photobioreactors is more complex. Irradiance levels in culture can be predicted as discussed. Similarly, carbon dioxide supply problems are relatively easily resolved. Difficulties arise with scaleup because relative volumes of light and dark zones change as the tube diameter increases; however, promising new leads in this area suggest that dependable scaleup based on fundamental principles should become feasible in the foreseeable future. At present the recommended method is to use a pipe diameter of no more than 0.1 m, and a continuous run length of about 80 m with a flow velocity of 0.3-0.5 m s -1. Multiple parallel run tubes originating and ending in common headers are apparently the best way to accommodate higher flows and volumes.

parameter in Eq. (8) path length parameter defined by Eq. (25) (m) parameter in Eq. (8) biomass concentration (kg m-3) Fanning friction factor total inorganic carbon in the liquid (tool m - 3) parameter in Eq. (8) dilution rate (s -1) radial dispersion coefficient

Fo FH20 FN f H

Hco

/40

Ho

diameter or hydraulic diameter (m) tube diameter (m) diameter of dark zone (m) diameter of dark zone at larger scale (m) diameter of dark zone at smaller scale (m) depth of light zone (m) photosynthetic efficiency of the solar radiation ( - 1.74 _+ 0.07 ~E J - ' ) carbon dioxide molar flow rate in the gas phase (mol s -1) oxygen molar flow rate in the gas phase (tool s-l) molar flow rate of water in the gas phase (tool s -l) nitrogen molar flow rate in the gas phase (tool s -l) scale factor total daily radiation on a horizontal surface (J m -2 day -1) daily direct radiation on a horizontal surface (J m -z day-l) Henry's law constant for carbon dioxide (mol m-3 a t m - l) daily diffuse radiation impinging on a horizontal surface (J m -2 day-1) daily global extraterrestrial solar radiation (J m -2 day-~) Henry's law constant for oxygen (mol m - 3 a t m - l)


h I

I.v

IBt(~, fO) IBt(ri, 09)

I,)

IDt IDt((-o) IDt(ri, (]9)

/max

Io Ir go K,

K~ K, Ks Kw (kLat)co 2

solar hour (h) hourly incident photosynthetic radiation on a horizontal surface (laE m-2 s-1) photosynthetically active hourly average irradiance inside culture (laE m-2 s-1) hourly PA direct irradiance on a horizontal surface (laE m-2 s - ' ) direct hourly PA irradiance on a vertical surface (~tE m-2 s-~) direct local hourly PA irradiance inside vertical column (~E m-2 s -1 ) hourly PA diffuse irradiance on horizontal surface (I~E m-2 s - ' ) hourly PA diffuse irradiance on inclined surface (~tE m-2 s - ' ) disperse hourly PA irradiance on vertical surface (~tE m - 2 S- I) local hourly disperse PA irradiance inside vertical column (~tE m-2 S-1) microalgal affinity for light (~tE m-2 S-1) saturation value of I (~tE m-2 S--I) solar PA irradiance impinging on the reactor's surface (IBt-+IDt) (~tE m - 2 S-1) hourly reflected PA irradiance on a surface (~tE m -2 s-l) absorption coefficient (m 2 g-l) first dissociation constant for H2CO 3 (mol m - 3) second dissociation constant for H2CO3 (tool m - 3) photoinhibition constant (I~E m-2 S-I) saturation constant (~tE m -2 S-1) dissociation constant for water (mol 2 m -6) volumetric gas-liquid mass transfer coefficient for carbon dioxide (s-1)

(kLaL)o2

245

volumetric gas-liquid mass transfer coefficient for oxygen

(S -1 ) L m N n PA p Pc%

Pdirect

tube length (m) exponent in Eq. (5) day of the year exponent in Eq. (6) photosynthetically active biomass productivity (kg m-3 s -I ) carbon dioxide partial pressure in the gas phase (atm) distance traveled by a direct incident ray from the tube's surface to any internal point (?'i, q))

(m) Pdisr~rse

Po2 PT P~,

QL QR R Rco 2 R02 r ri S UL UR

URL URs

u

distance traveled by disperse radiation from the tube's surface to any internal point (ri, ~0) (m) oxygen partial pressure in the gas phase (atm) total pressure in the system (arm) partial pressure of water vapor (atm) volumetric liquid flow rate (m3 s-') volumetric flow rate out of dark zone (m3s-') radius or hydraulic radius (m) carbon dioxide consumption rate (mol CO2 m - 3 s - 1 ) oxygen generation rate (mol 02 m -3 s -1) radial distance (m) distance in polar coordinates (m) cross-sectional area of the tube

(m2) superficial liquid velocity in the tube (m s-1) fluid interchange velocity (m s-l) fluid interchange velocity at larger scale (m s - ' ) fluid interchange velocity at smaller scale (m s -1) eddy velocity defined by Eq. (41) (m s-')

246


distance in the direction of flow (m) liquid phase molar concentration of species in brackets (mol m - 3) liquid phase equilibrium concentration of species in brackets (mol m - 3)

x

[] []*

6.1. Greek symbols

c ~G

0 O' Od

O. !

Oz 12 /.1 L

Pmax

7E

P PL

O9 (_~0S

N

parameter in Eq. (2) surface tilt angle relative to the horizontal surface azimuth angle ( - 180~< 7 < 180~ declination of the angular position of the Sun at solar noon with respect to the plane of the equator, north positive ( - 23.45 ~ < fi < 23.45~ defined by Eq. (11) angle shown in Fig. 3 fractional gas holdup angle of incidence defined by Eq. (21) angle 0 modified by refraction in the culture maximum duration of dark period (s) zenith angle of the Sun, defined by Eq. (22) zenith angle of the Sun modified by refraction in the culture specific growth rate (s-~) viscosity of liquid (Pa s-l) maximum specific growth rate (s -1) energy dissipation rate per unit mass (W kg -~) pi ground reflectivity density of liquid (kg m - 3) geographic latitude angular position in polar coordinates angle corresponding to the solar hour, defined by Eq. (16) hour angle at sunrise, defined by Eq. (10) atmospheric clarity index estimated at 0.74 + 9% universal solar constant (( = 1353 W m -2)

Acknowledgements Some of the work described was supported by the Comisi6n Interministerial de Ciencia y Tecnologia, C.I.C.Y.T., Spain (Project Bio 95-0652) and the European Union (Project BRPR CT970537).

References Aci6n Fernfindez, F.G., Garcia Camacho, F., Sfinchez P6rez, J.A., Fernfindez Sevilla, J., Molina Grima, E., 1997. A model for light distribution and average solar irradiance inside outdoor tubular photobioreactors for microalgal mass culture. Biotechnol. Bioeng. 55, 701-704. Aci6n Fernfindez, F.G., Garcia Camacho, F., S~inchez P6rez, J.A., Fernfindez Sevilla, J., Molina Grima, E., 1998. Modelling of biomass productivity in tubular photobioreactors for microalgal cultures. Effects of dilution rate, tube diameter and solar irradiance. Biotechnol. Bioeng. 58, 605611. Aiba, S., 1982. Growth kinetics of photosynthetic microorganisms. Adv. Biochem. Eng. 23, 85-156. Bannister, T.T., 1979. Quantitative description of steady state, nutrient-saturated algal growth, including adaptation. Limnol. Oceanogr. 24 (1), 76-96. Benemann, J.R., 1989. The future of microalgal biotechnology. In: Cresswell, R.C., Rees, T.A.V., Shah, N. (Eds.), Algal and Cyanobacterial Biotechnology. Longman, London, pp. 317-337. Borowitzka, M.A., 1996. Closed algal photobioreactors: design considerations for large-scale systems. J. Mar. Biotechnol. 4, 185-191. Camacho Rubio, F., Aci6n Fern/mdez, F.G., S~inchez P6rez, J.A., Garcia Camacho, F., Molina Grima, E., 1999. Prediction of dissolved oxygen and carbon dioxide concentration profiles in tubular photobioreactors for microalgal culture. Biotechnol. Bioeng. 62, 71-86. Chisti, Y., 1989. Airlift Bioreactors. Elsevier, London, p. 345. Chisti, Y., 1998. Pneumatically agitated bioreactors in industrial and environmental bioprocessing: hydrodynamics, hydraulics and transport phenomena. Appl. Mech. Rev. 51, 33-112. Chisti, Y., Kasper, M., Moo-Young, M., 1990. Mass transfer in external-loop airlift bioreactors using static mixers. Can. J. Chem. Eng. 68, 45-50. Chrismadha, T., Borowitzka, M.A., 1994. Effect of cell density and irradiance on growth proximate composition and eicosapentaenoic acid production of Phaeodactylum tricornutum grown in a tubular photobioreactor. J. Appl. Phycol. 6, 67-74. Duffle, J.A., Beckman, W.A., 1980. Solar Engineering of Thermal Processes. Wiley, New York, p. 762.

E. Molina Grima et al ..... Journal o[" Biotechnology 70 (1999) 231-247

Erickson, L.E., Lee, H.Y., 1986. Process analysis and design of algal growth systems. In: Barclay, W., McIntosh, R.P. (Eds.), Algal Biomass Technologies: An Interdisciplinary Perspective. Nova Hedwigia, Berlin, p. 197. Fernfindez Sevilla, J., 1995. Estudio del crecimiento simultaneamente fotolimitado y fotoinhibido de la microalga marina Isochrysis galbana. Productividad en ficidos grasos poliinsaturados n-3. Tesis doctoral. Universidad de Almeria, Spain. Frohlich, B.T., Webster, I.A., Ataai, M.M., Shuler, M.L., 1983. Photobioreactors: models for interaction of light intensity, reactor design and algal physiology. Biotechnol. Bioeng. Symp. 13, 331-350. Garcia Camacho, F., Contreras Gdmez, A., Aci6n Fern~indez, F.G., Molina Grima, E., 1999. Use of concentric tube air-lift photobioreactors for microalgal outdoor mass culture. Enzyme Microbial. Technol. 24, 164-172. Grobbelaar, J.U., 1994. Turbulence in algal mass cultures and the role of light/dark fluctuations. J. Appl. Phycol. 6. 331-335. Grobbelaar, J., Neddal, L., Tichy, V., 1996. Influence of high frequency light/dark fluctuations on photosynthetic characteristics of microalgae photo acclimated to different light intensities and implications for mass algal cultivation. J. Appl. Phycol. 8, 335-343. Incropera, F.P., Thomas, J.F., 1978. A model for solar radiation conversion to alga in shallow ponds. Solar Energy 2{). 157-165. Laws, E.A., 1980. Nutrient and light limited growth of Thalassiosirafluviatilis in continuous culture with implications for phytoplankton growth in the ocean. Limnol. Oceanogr. 25. 455-473. Lee, Y.K., 1986. Enclosed bioreactors for the mass cultivation of photosynthetic microorganisms: the future trend. Trends Biotechnol. 4, 186-189. Lee, Y.K., Low, C.S., 1991. Effects of photobioreactors inclination on the biomass productivity of an outdoor algal culture. Biotechnol. Bioeng. 38, 995-1000. Lee, Y.K., Low, C.S., 1992. Productivity of outdoor algal cultures in enclosed tubular photobioreactors. Biotechnol. Bioeng. 40, 1119-1122. Liu, B.Y.H., Jordan, R.C., 1960. The interrelationship and characteristic distribution of direct, diffuse and total solar radiation. Solar Energy 7, 53-65. Livansky, K., 1982. Effect of suspension temperature on mass transfer coefficient of carbon dioxide from algal suspension into air on a cultivation platform with baffles. Arch. Hydrobiol. Suppl. 63 (3), 363-367. Livansky, K., 1990. Losses of CO2 in outdoor mass cultures: determination of the mass transfer coefficient kl by means of measured pH course in NaHCO 3 solution. Algol. Stud. 58, 87-97. Livansky, K., Bartos, J., 1986. Relationship between pCO_, and pH in a medium for algal culture. Arch. Hydrobiol. Suppl. 73 (3), 425-431. Mfirkl, H., Mather, M., 1985. Mixing and aeration of shallow open ponds. Arc. Hydrobiol. Beih. 20, 85-93. Merchuk, J.C., Ronen, M., Giris, S., Arad, S., 1998. Light-

247

dark cycles in the growth of the read microalgae Porphyridium sp. Biotechnol. Bioeng. 59, 705-713. Molina Grima, E., Garcia Camacho, F., Sfinchez P6rez, J.A., Fernfindez Sevilla, J.. Aci6n Fernfindez, F.G., Contreras Gomez. A.. 1994. A mathematical model of microalgal growth in light limited chemostat cultures. J. Chem. Technol. Biotechnol. 61, 167-173. Molina Grima. E.. Fernandez Sevilla, J.M., Sanchez Perez, J.A.. Garcia Camacho, F.. 1996. A study on simultaneous photolimitation and photoinhibition in dense microalgal cultures taking into account incident and averaged irradiances. J. Biotechnol. 45, 59-69. Myers. J.E.. 1980. On the algae: thoughts about physiology and measurement of efficiency. In: Falkowski, P.G. (Ed.), Primary Productivity in the Sea. Plenum Press, New York, pp. 1-16. Philliphs, J.N.. Myers. J.. 1954. Growth rate of Chlorella in flashing light. Plant Physiol. 29, 152-161. Pulz. O.. Scheinbenbogen. K., 1998. Photobioreactors: design and performance with respect to light energy input. Adv. Biochem. Eng. Biotechnol. 59, 123-152. Quiang. H.. Gutterman. H., Richmond, A., 1996. A flat inclined modular photobioreactor for outdoor mass cultivation of photo autotrophs. Biotechnol. Bioeng. 51, 5160. Rabe. A.E.. Benoit. A., 1962. Mean light intensity. A useful concept in correlating growth rates of dense cultures of microalgae. Biotechnol. Bioeng. 4. 337-390. Ree. G.Y.. Gotham. I.J.. 1981. The effect of environmental factors on phytoplankton growth: Light and interaction of light with nitrate limitation. Limnol. Oceanogr. 26, 649659. Steele, J.H.. 1977. In kapidus, L., Amundson, N.R. (Eds.), Microbial Kinetics and Dynamics in Chemical Reactor Theory. Prentice-Hall. Englewood Cliffs, NJ, pp. 405-483. Tamiva. H.. Hase. E.. Shibata, K., Mituya, A., Iwamura, T., Nihei. T.. Sasa, T., 1953. Kinetics of growth of Chlorella, with special reference to its dependence on quantity of available light and on temperature. In: Burlew, J.S. (Ed.), Algal Culture from Laboratory to Pilot Plant. Carnegie Institution of Washington, Washington, DC, pp. 204-232. Terry. K.L.. 1986. Photosynthesis in modulated light: Quantitative dependence of photosynthesis enhancement on flashing rate. Biotechnol. Bioeng. 28, 988-995. Tredici. M.R.. Materassi. R., 1992. From open ponds to vertical alveolar panels: the Italian experience in the development of reactors for the mass cultivation of phototrophic microorganisms. J. Appl. Phycol. 4, 221-231. Van Oorschot, J.L.P., 1955. Conversion of light energy in algal cultures. Med van Lund. Wang. 55, 225-277. Weissman. J.C., Goebel, R.P., Benemann, J.R., 1988. Photobioreactor design: mixing, carbon utilization, and oxygen accumulation. Biotechnol. Bioeng. 31, 336-344. Yamaguchi. K.. 1997. Recent advances in microalgal bioscience in Japan, with special reference to utilization of biomass and metabolites: a review. J. Appl. Phycol. 8, 487-502.


i

JOURNAL

iiui

OF



Comparative evaluation of compact photobioreactors for large-scale monoculture of microalgae Asterio Sfinchez Mir6n, Antonio Contreras G6mez, Francisco Garcia Camacho, Emilio Molina Grima, Yusuf Chisti * Department of Chemical Engineering, University of Almeria, E-04071 Almeria, Spain Received 27 October 1998; received in revised form 17 November 1998; accepted 22 December 1998

Abstract Engineering analyses combined with experimental observations in horizontal tubular photobioreactors and vertical bubble columns are used to demonstrate the potential of pneumatically mixed vertical devices for large-scale outdoor culture of photosynthetic microorganisms. Whereas the horizontal tubular systems have been extensively investigated, their scalability is limited. Horizontal tubular photobioreactors and vertical bubble column type units differ substantially in many ways, particularly with respect to the surface-to-volume ratio, the amount of gas in dispersion, the gas-liquid mass transfer characteristics, the nature of the fluid movement and the internal irradiance levels. As illustrated for eicosapentaenoic acid production from the microalga Phaeodactylum tricornutum, a realistic commercial process cannot rely on horizontal tubular photobioreactor technology. In bubble columns, presence of gas bubbles generally enhances internal irradiance when the Sun is low on the horizon. Near solar noon, the bubbles diminish the internal column irradiance relative to the ungassed state. The optimal dimensions of vertical column photobioreactors are about 0.2 m diameter and 4 m column height. Parallel east-west oriented rows of such columns located at 36.8~ latitude need an optimal inter-row spacing of about 3.5 m. In vertical columns the biomass productivity varies substantially during the year: the peak productivity during summer may be several times greater than in the winter. This seasonal variation occurs also in horizontal tubular units, but is much less pronounced. Under identical conditions, the volumetric biomass productivity in a bubble column is ~ 60% of that in a 0.06 m diameter horizontal tubular loop, but there is substantial scope for raising this value. 9 1999 Elsevier Science B.V. All rights reserved.

Keywords: Microalgae; Photobioreactors; Scale-up; Photosynthetic culture; Eicosapentaenoic acid; Phaeodactylum tricornutum

I. Introduction

* Corresponding author: Tel.: + 34-950-215566; fax: + 34950-215484. E-mail address: [email protected] (Y. Chisti) 9 1999 Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00079-6

P h o t o b i o r e a c t o r s for large-scale m o n o c u l t u r e of microalgae have conventionally been designed as devices with large s u r f a c e - t o - v o l u m e ratios. Various types of tubular p h o t o b i o r e a c t o r s are examples of this a p p r o a c h (Lee, 1986; Borowitzka,

250

A. Shnchez Mir6n et al./'Journal of Biotechnology 70 (1999) 249-270

1996). These reactors occupy vast land areas: they are expensive to build; difficult to maintain; and only somewhat scaleable. Tubular photobioreactors can usefully satisfy only medium level production demands. Attempted, large- scale production in horizontal tubular loops has failed quite spectacularly in one case (Fig. 1); hence, other reactor configurations are needed for the production of larger quantities of pharmacologically active compounds that certain microalgae can potentially produce. The areal productivity, i.e. productivity per unit land area, is low for conventional tubular photobioreactors and in large units as well as modular designs, sterile operation to the levels demanded in the pharmaceutical industry is difficult. Some low surface-to-volume, pneumatically agitated photobioreactors can potentially overcome these significant disadvantages. Examples of the latter type are bubble columns and airlift bioreactors. Large scale culture of microalgae in these systems has not been investigated as it has always been assumed that small surface-to-volume ratios of these devices would make them ineffective. This need not be so as reported in this work which deals with comparative outdoor evaluation of pilot scale bubble column photobioreactors with respect to performance in horizontal tubular loops. Data are reported on three aspects of comparative characterization: (a) gas-liquid hydrodynamics and mass transfer; (b) internal irradiance levels as functions of Sun's location relative to the photobioreactors; and (c) performance during culture of the microalga Phaeodactylum tricorntum. Also reported are the effects of hydrodynamics on survival behavior of algal cells.

2. I. Hydrodynamics

2. Comparison of performance

where PL is the density of the liquid, g is the gravitational acceleration and Uc is the superficial gas velocity in the column. Eqs. (1) and (2) were established by Chisti (1989) for tap water and 0.15 M aqueous sodium chloride, respectively. The equations apply to the bubble flow regime, or UG values less than about 0.05 m s-1 (PG/VL "~ 500 W m-3) The obvious decline in the rate of increase of gas holdup with increasing specific

The vertical and the horizontal tubular photobioreactors differ in several significant ways including differences in light regimens, gas-liquid hydrodynamics and mass transfer behavior. Some of these factors--e.g, hydrodynamics and light regimen--are interrelated. Their impact on culture performance is discussed below.

The hydrodynamics of flow in horizontal tubes and vertical columns are generally quite different. The necessarily gas sparged bubble columns and airlift reactors tend to have substantially greater gas holdups than do horizontal tubular solar receivers. The latter are virtually free of gas and any bubbles present are localized to a narrow zone along the upper portion of the tubes; moreover, the bubbles are relatively small. In contrast, there are many more and larger bubbles in vertical photobioreactors and the gas-liquid flow is much more chaotic than the highly directional flow in a small-diameter horizontal pipe. Differences in gas holdup and the bubble size affect light penetration, gas-liquid mass transfer, mixing and shear stress levels. 2. I.I. Gas holdup

Gas holdup measurements in a bubble column photobioreactor confirm that holdup increases with specific power input in accordance with previously published data (Chisti, 1989). Thus, the holdup data in tap water (Fig. 2) closely followed the equation e = 3 . 3 1 7 x 10- \ V L j

'

(1)

whereas data in sea water (Fig. 2) agreed with the correlation e = 7.958x10

-

5(P-~L)I249.

(2)

In these equations the power input due to aeration is calculated as

eo_ p~u~, v~

(3)

A. Shnchez Mir6n et al./'Journal of Biotechnolgoy 70 (1999) 249-270

251

Fig. 1. A commercial horizontal tubular bioreactor facility that failed to perform to expectations and was abandoned. This facility was located in Cartagena, Spain and it was owned by Photobioreactors Ltd.

A. Shnchez Mir6n et al./'Journal of Biotechnology 70 (1999) 249-270

252

used in microalgal culture in vertical photobioreactors because of the cell damaging potential of intense turbulence (Silva et al., 1987; Suzuki et al., 1995; Contreras et al., 1998; Chisti, 1999a). Data in Fig. 2 reveal that the flow transition occurs earlier in tap water, around a power input of 280

power input values above about 400 W m - 3 (Fig. 2b) is due to a well-known change in the flow regime from bubble to churn turbulent flow (Chisti, 1989). Equations for estimating gas holdup in the latter regime have been published (Chisti, 1989), but that regime is not likely to be 100

_

'

_

i

i

u

i

i

i

i

|

u

i

t. . . . . .

i

u

u

i

I z

TAP WATER

_

o~

i

(a)

10

13. E3 .--I 0 "1oo < L9

1

.

-

.

.

_

0.1

I

I,

. 6:3.317x10-4(

L

I,

1

I

I

10

p

-~ O.97

G1

~,v,)

I

[

|

|

I

I

J

~

=

100

1000

PG/VL ( W m -3)

100

_

i

i

I

i

i

i

f

i

I

1

1

1

i

i

i

!

i.

(b) _

-

-_

SEA WATER / 0

10

Q.. E3 ..J O -r" oo
0.006 m) that rise rapidly. Radial flow into and out of the central dark core can be enhanced by ensuring that quite large spherical cap bubbles (dB ~ 0.02-0.04 m) rise through this zone intermittently (Fig. 11). These schemes are currently being experimented with. In summary, the low surface-to-volume photobioreactors can potentially approach the volumetric productivity of horizontal tubular loops, however, attaining this requires that: (i) the cumulative residence time of the cells in the light zone of the reactor should be comparable to that of the tubular device; and (ii) the frequency of light-dark transition cycle in the 'deeper' vessels should be comparable to that of tubular systems. These two objectives can be substantially achieved in bubble columns and airlift bioreactors by enhancing radial movement of cells. In airlift bioreactors, but not in batch bubble columns, radial flow may be enhanced by using static mixing elements.

259

2.2.2. Vertical photobioreactor emplacement The maximum number of the vertical column reactors that may be accommodated in a given area depends on the height of the column which, together with the position of the Sun, establishes the maximum extent of the column's shadow on the ground. The length of the shadow from the column's base is given by hc tan0e

L~ = ~ ,

(8)

where hc is the height of the column and Oi is the angle of incidence of the direct solar radiation. The angle of incidence--the inclination of the Sun from the normal to the vertical axis of the bubble columnmdepends on the geographic latitude ~b, the day of the year N, and the solar hour 1~; the angle of incidence is given as (Liu and Jordan, 1960): 0; = 90 ~ - cos- ~(cos~5 9cosq~ 9cosco + sine5 9sinq~) (9)

where 4~ is the geographic latitude. The angles co and c~ are related to the solar hour and the day of the year (Liu and Jordan, 1960), respectively, as follows:

4500

SEA W A T E R U G 0.051 m s -1

20

--

4000

10 E

w

v

3500

z~

uJ z 0

lad 0

,,

3000 A

9

n,, 2500

2000

I 10

i .

DIFFERENCE (%)(08-09-98) I

.o BUBBLES(0.9-9.,

9 BUBBLED(0909-98)

1 12

~J z LU el,' LU 11 LI. ~3

~

I

~'

.~

14

~ll

'~

LIIL ~R;

16

1 18

El-,

-10

-20

T I M E O F D A Y (h)

Fig. 9. Daily variation in internal irradiance in the bubble column in sea water. The gas holdup value was 0.135. The difference between the aerated and unaerated irradiance levels is also noted. The data are for 8 and 9 September, 1998.

260

A. Sdnchez Mir6n et al. ~!Journal of Biotechnology 70 (1999) 249-270

Fig. 10. The bubble dispersion on the left allows through transmission of light whereas the cloudy dispersion on the right blocks light. These pictures were taken in sea water at aeration power inputs of 154 and 518 W m - 3 , respectively. The perforated pipe sparger hole diameter was 1 mm in both cases.

co =- 1 5 ( 1 2 - h),

(lo)

and 9 //360(284 + N ) t g = 23.45. sm~ 365 / "

(11)

ing of 0.35 m between centers of adjacent columns of 0.2 m diameter. Close spacing within east-west rows has no impact on illumination, but it improves efficiency of land use.

2.3. Shear effects The loci of the maximum extent of the shadow of a 2 m tall bubble column are plotted in Fig. 12 for representative days in winter, spring, and summer seasons at the geographic location of Almeria (36.8~ 2.9~ The maximum extent of the shadow in January is about 7.4 m, whereas the maximum extent in July is about 1.3 m. These distances are measured north-south between parallel east-west lines passing through the base of the vertical column and the tip of the column's shadow. Ideally, parallel east-west rows of bubble columns should be spaced by at least the maximum length of the shadow in winter. This would assure that the reactors are never mutually shaded, however, a more optimal setup would place the rows of reactors closer, about midway between the high extremes of the shadow length in the summer and the winter. Consequently, there will be no mutual shading in the summer but some shading would occur during the winter. In a single east-west row of columns the columns could be spaced quite close together; e.g. a spac-

Shear stress has been implicated as an important factor in culture of several microalgae (Silva et al., 1987; Suzuki et al., 1995; Contreras et al., 1998; Chisti, 1999a). A cell damaging hydrodynamic environment is easily attained in bubble columns and airlift reactors (Silva et al., 1987; Suzuki et al., 1995; Contreras et al., 1998; Chisti, 1998a), but damage to algal cells has never been documented within tubes of tubular photobioreactors. This may suggest that the damage is somehow linked with the presence of bubbles in pneumatically agitated devices, but extensive studies prove otherwise. With most microalgae, increasing aeration rate up to quite high values improves culture productivity (Silva et al., 1987; Contreras et al., 1998), but damage occurs when the turbulence is so intense that the fluid microeddy size approaches cellular dimension. Only in one case--that of the commercially important but unusually fragile marine alga Dunaliella--has damage been associated directly with the bubbles

A. Sdnchez Mir6n et al./Journal of Biotechnolgoy 70 (1999) 249-270

(Silva et al., 1987). Survival of Dunaliella in aerated systems is improved by supplementing the culture with viscosity enhancers such as carboxymethyl cellulose and agar (Silva et al., 1987), but this approach may not be applicable universally (Chisti, 1999a). In one study with D. tertiolecta, culture in a bubble column was quite successful, but when the bubble column was converted to an airlift device by inserting a vertical baffle, the productivity declined (Suzuki et al., 1995). Under conditions that were earlier identified as optimal, no growth was observed in the airlift reactor whereas good

LIGHT ZONE BUBBLE COLUMN DARK ZONE

SPHERICAL CAP BUBBLE

SPHEROIDAL BUBBLE

Fig. 11. Use of spherical cap bubbles in the central dark core to enhance radial light-dark interchange of fluid in a bubble column. The bubbles in the peripheral light zone are predominantly ellipsoidal with a diameter of about 0.006 m.

261

growth occurred in the bubble column. Microscopic examination showed significant disruption of the cells in the airlift device (Suzuki et al., 1995). This was associated with the hydrodynamic stresses generated as the culture flowed over the upper edge of the baffle into the downcomer (Suzuki et al., 1995). This effect could have been avoided, or at least minimized, by hydrodynamic smoothing of the upper and lower parts of the baffle to prevent flow separation (Chisti, 1998a). In the bubble column, the growth was sensitive to aeration rate: growth rate increased with increasing superficial gas velocity until a velocity of about 0.6 m min 1, or a specific power input of 98 W m-3. Further increase in aeration rate reduced growth, apparently because of hydrodynamic stresses in the fluid (Suzuki et al., 1995). Under non-growth conditions (no light), the specific death rate in the bubble column was shown to increase with superficial gas velocity for velocities exceeding 0.6 m min 1 (Suzuki et al., 1995). At a fixed aeration velocity (UG = 1 m min 1), the specific death rate decreased with increasing height of the culture fluid in the column (Suzuki et al., 1995), probably because the specific power input and, hence, the turbulence intensity declined (Chisti, 1998a). Similar behavior has been reported with animal cells in bubble columns (Emery et al., 1987; Tramper et ali,, 1987a,b)i Improved growth with increasing aeration up to a limit has been documented for several algae including Dunaliella (Silva et al., 1987); P. tricornutum (Contreras et al., 1998)i and others (Chisti, 1999a). This effect has been observed invariably under light limited conditions and it is best explained by improved light-dark cycling due to improved agitation. Indeed, the effect occurs even in the absence of gas when mec used to improve light-dark int( photobioreactors that contain In indoor batch cultures of P. out in a draft-tube sparged co~ nal-loop airlift photobioreact01 mination of 1200 ~tE m - 2 S the reactor, Contreras et al. ( the maximum specific g r o w t h increasing aeration velocity in increase occurred until a velo

A. Sdnchez Mirdn et al./Journal of Biotechnology 70 (1999) 249-270

262

NORTH

DISTANCE (m)

8 -

JANUARY 01

6 -

5- !

/

LOCI OF MAXIMUM EXTENT OF SHADOW

SHADOW AT SUNRISE

APRIL 01 JUNE 21

I

i

I

-15

-10

-5

t,.# ~lr

0\

i

i

I

5

10

15

DISTANCE (m) BUBBLE COLUMN

WEST

EAST

SOUTH

DAILY PATH OF SUN

IsuNl

Fig. 12. Locus of the shadow of a 2 m tall bubble column during the day. The loci are shown for three representative seasons for 8 h (January) or 10 h (April, June) days, centered at midday.

s - ] , corresponding to a specific power input of 270 W m -3 Whether improved growth was due to enhanced fluid interchange between the light and dark zones was unclear and other factors may have contributed. The cultures were clearly light limited whenever the biomass concentration exceeded about 1 kg m-3. Increase in aeration rate beyond 0.055 m s -~ substantially reduced the specific growth rate of P. tricornutum cultures (Contreras et al., 1998). At

the critical aeration rate value of 0.055 m s - ] , the calculated Kolmogoroff microeddy scale was 45 gm, or comparable to the dimensions of the algal cells (up to 35 lam long and 3 gm wide (Lewin et al., 1958)). Except for robust microorganisms, damage to cells has generally been observed when the dimensions of the isotropic turbulence microeddies approach those of the cells (Chisti, 1999a), however, the assumption of isotropic turbulence is almost never justifiable under typical

263

A. Sdnchez Mir6n et al./Journal of Biotechnolgoy 70 (1999) 249-270

operation of bubble columns and airlift systems (Chisti, 1998a). Why has cell damage not been observed in horizontal tubular photobioreactor pipes is readily explained. For a 200 m straight run smooth horizontal tube of 0.06 rn diameter, the culture velocity would have to exceed 1.14 rn s-1 for the microscale of turbulence to approach 45 g m - - t h e value that damaged the P. tricornutum cells in the vertical airlift device. This velocity value is easily estimated from the relationship between the Kolmogoroff microeddy length scale (f) and the specific energy dissipation rate (Chisti, 1999a):

__ (/AL~3/4E- 1/4.

(12)

k,P L f

The specific energy dissipation rate E in (Eq. (12)) is related with the pressured drop (AP), i.e. E=

ULAP

(13)

pL L '

where L is the length of the tube and UL is the culture velocity. The pressure drop in (Eq. (13)) is obtained using the expression: L A P = ZCf -~ pLU 2 .

(14)

In (Eq. (14)), the Fanning friction factor (C f) is obtained from the Blasius equation:

CF = 0.0792(P~ULd) -°2s ,

(15)

where d is the diameter of the tubing. For the above calculations, the measured viscosity of the algal culture was virtually the same as that of water. In practice, because of low pressure ratings of typically used transparent materials of construction, a culture velocity as high as 1.14 m s cannot be attained in long-run continuous tubings. Typically, the culture velocity in tubular loops does not exceed 0.5 m s-1, or less than half the damaging threshold value. In theory, high culture velocities should benefit productivity by increasing the allowable length of a continuous run tube before oxygen concentration builds up to high inhibitory levels. In reality, pumping becomes difficult as the tube length increases. The

cell damaging potential of pumps is another consideration that is especially relevant to tubular photobioreactor culture. Damage in pumps is treated elsewhere (Chisti, 1999a). 2.4. Gas- liquid mass transfer

Microalgae generate oxygen during photosynthesis, hence, dissolved oxygen levels equivalent to several times air saturation are easily reached in closed cultures. Oxygen concentrations above air saturation generally inhibit photosynthesis in microalgae (Aiba, 1982). In studies with Chlorella vulgaris, M/irkl and Mather (1985) noted that the rate of photosynthesis increased by 14% when there was almost no dissolved oxygen. Saturation of the medium with pure oxygen reduced the photosynthesis rate by 35%. Accumulation of oxygen to inhibitory levels is a major problem in horizontal tubular photobioreactors and this problem becomes more severe as the length of continuous run tubing increases. With Spirulina platensis under intense artificial illumination, oxygen production rates have been estimated at 0.35 0.5 g 1-1 h - 1 for radiation intensity levels of 15002600 gE m - 2 s-1 (Tredici et al., 1991). In one case, a tubular photobioreactor designed as a commercial production unit with several kilometer long tube runs (Fig. 1) failed to produce partly because of oxygen accumulation. This device was apparently the largest horizontal tube photobioreactor ever constructed and it's failure is clear evidence of limited scalability of this type of bioreactor. Existing functioning tubular photobioreactors are typically a mere fraction of the size of the device shown in Fig. 1. Oxygen removal into the exhaust gas-phase is substantially greater in pneumatically agitated vertical reactors than in horizontal tubes. For example, for a power input value of 24 W m - 3 the lowest value in Fig. 2a the overall volumetric gas-liquid mass transfer coefficient (kLaL) in the bubble column was estimated at 0.0037 s-1 using the equation

4(PG~ 0"86

kLaL=2.39 X 10- \ V L ]

"

(16)

264

A. Sfinchez Mir6n et al./Journal of Biotechnology 70 (1999) 249-270

Eq. (16) applies to air-water systems (Chisti, 1989). The estimated kLaL in the column was about 4-fold the estimated value for a horizontal tube with a reported 3% maximum gas holdup (Camacho Rubio et al., 1999). In view of these results, vertical tubular photobioreactors such as bubble columns should easily maintain a dissolved oxygen level only a little higher than the air saturation value. Consequently, vertical reactors will experience little oxygen inhibition. Indeed, in one case the dissolved oxygen level did not exceed the air saturation value during outdoor culture of P. tricornutumin a draft-tube airlift reactor (Contreras, 1996). Because of a reduced level of dissolved oxygen, the photooxidation-associated loss of biomass and product metabolite will be lower in vertical reactors. Photooxidation occurs especially when high levels of dissolved oxygen combine with an intensely irradiated environment.

2.5. Overall productivity: the case of EPA Eicosapentaenoic acid (EPA) is a polyunsaturated fatty acid that is potentially medically significant in treatment of certain cancers and heart disease (Rambjor et al., 1996; Simonsen et al., 1998). EPA is produced by several microalgae and some other microorganisms. EPA occurs also in fish oil. Current annual demand of EPA is approximately 125 tonnes in Japan alone and a world-wide demand has been estimated at 300 tonnes per annum (Corden, personal communication). A facility producing just 20 tonnes EPA per annum from microalgal sources would need to generate 24 tonnes of wet biomass daily, operating continuously for 95% of the calendar year. This figure assumes an EPA content of about 2% (w/w, dry weight basis) in P. tricorntum, an optimistic 80% recovery of the EPA and 85% moisture in the biomass. For this production capacity, an optimally operated horizontal tubular photobioreactor (0.06 m tube diameter) would have a volume of 2350 m 3. This volume is based on an experimentally observed dry biomass productivity of 2.02 kg m - 3 d - 1 in June, or an average annual productivity of 1.535 kg m -3 d 1. The latter value accounts for the experimentally documented 8% per month decline in productivity going from

mid summer to mid winter (Aci6n Fern/mdez et al., 1998). Such a reactor will occupy a surface area of about 14.1 ha. This estimation of surface area assumes an optimal distance of 0.11 m between adjacent parallel horizontal tube runs. In theory, this distance could be reduced by half although some loss of productivity would occur due to mutual shading. A bubble column tank farm for the same 20 tonne annual EPA production would have a total volume of 8220 m 3 based on experimentally measured dry biomass productivity of 0.64 kg m -3 d - 1 in June, or a mean annual productivity of 0.486 kg m - 3 d 1. Such a tank farm will occupy an area 16.3 ha. Each bubble column would be 0.2 m in diameter, 2.1 m tall, with a culture volume of about 0.06 m 3. East-west oriented rows, 995 m long, will have a spacing of 0.35 m between column centers, inter-row spacing will be 3.4 m measured between column centers. Because a 3.4 m inter-row spacing results in mutual shading of columns for part of the year, the mean annual volumetric productivity of biomass declines to 0.439 kg m - 3 d - 1 , or about 90% of the unshaded productivity. The land use efficiency may seem lower for the bubble column tank farm, however, the column height used was only 2.1 m and efficiency improves with increasing column height as shown in Fig. 13. Thus, for a more realistic column height of 4 m, the total volume needed to for 20 tonnes per annum EPA capacity will be about 10 120 m 3. The mean annual volumetric productivity of biomass will decline further to 0.356 kg m - 3 d - 1 because of increased mutual shading of the taller columns, but the area demand will reduce to ~ 40% of that with 2.1 m columns. Because of the larger volume per column, fewer columns will be needed. The effect of bubble column height on mean annual volumetric biomass and EPA productivities is shown in Fig. 14 for a facility with 46 columns arranged in four east west oriented rows with 3.4 m interrow spacing. The volumetric productivity declines with increasing column height because of increased duration of the mutually shaded period. The increased production with increasing column height (Fig. 13) occurs because the culture volume depends directly on height and more volume is

265

A. Sdnchez Mir6n et al./Journid of Biotechnolgoy 70 (1999) 249-270

accommodated on a given land surface when taller columns are employed. Because taller columns have reduced volumetric productivity as a result of longer periods of mutual shading, increasing height does not increase the annual production linearly (Fig. 13). Beyond a height of 5 m, the annual production from a given area is no longer sensitive to column height (Fig. 13) because the decline in volumetric productivity exactly balances the effect of increased volume. The biomass productivity numbers used in these estimations were obtained in a 0.2 m 3 horizontal tubular reactor that occupied an area of 12 m 2. The reactor was located on a reflective surface with an albedo of 2 and it produced P. tricorntum biomass. The culture was carried out in Mann and Myers (1968) medium (Table 2) formulated in sea water (Table 1) at 3-fold concentration than recommended by Mann and Myers (1968). The optimal dilution rate during the summer (June) was ~ 0.05 h -1, giving a biomass productivity of 2.02 kg m -3 d -1 with 4 kg m - 3 biomass in the effluent. The maximum a real productivity of the reactor was a mere 0.034 kg m - 2 d - 1 . 800

I

I

I

I

As noted earlier, the experimental productivity in a single 2 m tall column, 0.2 m in diameter, was 0.64 kg m - 3 d-1 in June. The maximum areal productivity of the single reactor was 0.093 kg m - 2 d - 1, accounting for the 1.3 m shadow length and 0.3 m effective width of the column. The culture medium was a modified Ukeles (1961) medium (Garcia Sfinchez, 1994) made in sea water (Table 1) at three times the components concentrations noted in Table 2. Differences in media had no effect on culture behavior. Comparing superficially, the mean annual volumetric productivity of the vertical column was only 30% of that of the horizontal tubular loop, however, it needs emphasizing that the horizontal device was located on a reflective surface that enhanced the total incident radiation on it's surface by a factor of two. If the vertical column is placed on a similarly reflective background its productivity will increase minimally by a factor of 1.8; hence, comparing on an equal basis, the volumetric productivity of the vertical unit is about 57% of that in the horizontal device.

I

I

I

I

I~

-24

BIOMASS - 20

600

-

16

A

E eet~

-

l;

12

2~ v

'E 0 ~ 0.oo10 -

',,,,

13.

0.0000

6

8

10

12

Hour

14

16

18

20

Fig. 4. Daily variation of photosynthetic activity of the cells.

277

M.M. Rebolloso Fuentes et al. j Journal o f Biotechnology 70 (1999) 271-288 21.6 .Co 21.2 20.8

0.60

140

0.50 c

120

0.40 ~

100

-

- 8.0 7.8 p.,

L

0.30 = --O E

~ 20.4

0.20 (~

~ 20.0

0.10

.

19.6 0.0

0.5

J t

.

. 1.0

.

.

0.00

.

1.5 2.0 "13me, days

o

2.5

~"

80

,.~

60

o

m r

3.5

3.0

7.6 9

.

t

i]

9

"6 7.4 :z:

"

7.2

20

. 0.0

J

Fig. 5. Daily variation of gas exhaust composition during the quasi steady state.

solar irradiance is the highest, and again an increase during the afternoon up to 0.0012 mol O2 m - 3 s - 1, when solar irradiance decreases (Fig. 4). The gas exhaust composition varied according to the variations in dissolved oxygen. The oxygen molar fraction of the exhaust gas varied around the oxygen molar fraction in the air, ranging from 21.4% in the daylight period to 19.8% during the night (Fig. 5). The carbon dioxide molar fraction of the exhaust gas was always higher than in the air, ranging from 0.35% in the central hours of the day to ~ 0.50% during the night (Fig. 5). The high values of carbon dioxide molar fraction in the gas exhaust compared to air are due to the injection of pure CO2 bubbles for pH control. Therefore, when the CO2 injection is released inside the reactor, the broth becomes locally saturated with pure CO2. Then, when the culture is aerated in the airlift system, the carbon dioxide is stripped to air. This happened in both the daylight period and night period because the pH control system was not switched off at night. Moreover, the carbon dioxide generated during the night by respiration also increases the carbon dioxide molar fraction in the exhaust gas. The daily variation of solar irradiance also influenced the pH of the culture and the carbon losses. Although the pH of the culture was controlled at 7.6_+0.3 by on-demand injection of pure CO2, a cyclic daily variation was observed (Fig. 6). The average pH of the culture in the daylight period was higher, 7.75 _+ 0.08, than during the night, 7.51 _+ 0.05. In the daylight period,

,

40

9 Oxygen ....~ - - O ~ x ) n dioxide I

8.2

.

0.5

. 1.0

.

.

.

7.0

1.5

2.0 Time, days

F i g ....Carbonlosses

2.5 ,r

3.0

3.5

PHI

Fig. 6. Daily variation of carbon losses and pH of the culture during the quasi steady state.

the pH was above the set point due to the carbon consumption by photosynthesis, whereas during the night, the pH was under the set point due to CO2 generation by respiration. Although the pH of the culture was controlled, the carbon losses were high, ~ 100% during the night, dawn and dusk, due to the low carbon consumption rate. In the central hours of the day, the carbon losses were lower, equal to 20%, due to the high carbon consumption rate (Fig. 6). On the daylight period, the biomass concentration increased indicating that the growth rate was higher than the imposed dilution rate ( D - 0.049 h-1) and the biomass was accumulated in the reactor (Fig. 7). During the night, a fraction of the accumulated biomass was consumed, thus the 112 110

~

Day 1, iJ=0.0107 h1 BL=9% Day 2 IJ=0.0049 h"1 BL=20%

~.~

1 08 o r

106-

r

..j 1 0 2 1 O0 0.98 0.96

8

10

12

14 Hour

16

18

20

i O Day 1 IEIDay 2 A Day 3 I

Fig. 7. Daily variation of biomass concentration, additional growth rat in the daylight period, ~so~, and biomass losses, LB, during the quasi steady state.

278

M.M. Rebolloso Fuentes et al. Journal of Biotechnology 70 (1999) 271-288

biomass concentration at first hours in the morning being lower. The 3-day average quasi steady state biomass concentration was 3.50 g 1-1, this resulted in an average biomass productivity of 1.76 g 1-ld-~. In front, the biomass night losses were ~ 14% of the biomass concentration at the end of daylight time. The additional growth rate in the daylight period, ~tso~ (~t~o~= actual growth rate imposed dilution rate) was different for each day, the maximum, 0.0107 h-1, being obtained on day 1, and the minimum, 0.0049 h - ~, on day 2 (Fig. 7). In addition, the biomass night losses also varied for each day. A maximum loss was observed on day 2 (20%) while a minimum was obtained on day 1 (9%) (Fig. 7). Elemental analysis of the culture broth, cell free culture medium and liophylised biomass were carried out in order to determine the variation of metabolic routes of the cells during the day. The carbon content of the broth was in the range 0.14 + 0.02% , being higher in the morning and lower in the afternoon in spite of the higher biomass concentration (Fig. 8). This can only be due to either a carbon accumulation in the cell free culture medium (supernatant) during the morning that is eliminated in the afternoon, or to a decrease during the day of the carbon content in the biomass. The elemental analysis of the supernatant showed carbon accumulation during the morning in the liquid phase and elimination during the afternoon, as well as a reduction of nitrogen and sulphur content during the day (Fig. 8). Mean values ranges were 0.017 + 0.003, 0.025 + 0.005 and 0.024 + 0.004% DW of carbon, nitrogen and sulphur, respectively. With regard to elemental analysis of the biomass, the carbon and oxygen content increased during the daylight period (37.3 +_ 0.6 and 39.0 +_ 0.7% DW), whereas the nitrogen and sulphur content decreased (4.6 _+ 0.2 and 1.10 + 0.06%DW) and hydrogen content remained constant (5.5 _+ 0.3% DW) (Fig. 8). Variations of the culture conditions also determined changes in the biochemical composition of the biomass. The pigment content, as chlorophylls and carotenoids, slightly decreased during the daylight period, being 0.36 + 0.02% at the beginning and decreasing to 0.32 _+ 0.02% at the end of daylight time (Table 1). This trend caused a de-

0.20 ~

9

_ 0.15 j

9149 9

e

e-

[_,

0.05

~J

0.00 ~ 0.0

~

~176

mo(~

~-0.5

, 1.0

I 9Carbon

1.5 2.0 -time, days A Nitrogen

,

1

., ,

2.5

3.0

3.5

mSulphur I

0.040 T A

0.03o

O

O

A 9

,.a

0.010 ~-

~

0.000

9

0 u

~

0.5

0.0

~, ,,J ,,,,l

9

1.0

1.5

2.0

2.5

3.0

3.5

Time, days 9% Carbon ~ % Nitrogen o % Sulphur / J

7.0 6.o

"~ 42 +

gas, ~

oo

~" 32 ~ 0 r

3.0~" 2.0

ee*

-

0.5

0.0 r

ooo 1.0

oo

o

. %

1.5 2.0 2.5 3.0 -time, days o % Oxy n O/o Hyd

A % Nitrogen

o % Sulphur

10~"

3.5

,,,,I = 1

Fig. 8. Variation of elemental composition of broth, cell free culture medium and liophylised biomass during the quasi steady state reached.

crease of the absorption coefficient of the biomass, K, the mean values varying from 0.041 + 0.001 m 2 g-1 at the beginning of the day to 0.038 + 0.001 m 2 g-1 at the end of daylight time (Table 1), because Ka is a function of pigment content (Molina Grima et al., 1994a,b,c). The protein content also decreased in the daylight period, from a mean value at the beginning of the daylight time of 30.3 + 0.6% to 27.0 ___0.6% at the end of the afternoon (Table 1). With regard to fatty acids, only the main ones were considered,

i

t

P2

Table 1 Variation of the biochemical composition and extinction coefficient of the biomass during the quasi steady state Date

2413 2413 2413 2513 2513 2513 2613 2613 2613

Hour

9.5 12.5 18.0 9.5 12.5 18.0 9.5 12.5 18.0

Chlorop (‘XDW)

0.339 0.284 0.261 0.300 0.342 0.333 0.358 0.303 0.314

Carot..(‘L,DW)

0.031 0.014 0.023 0.022 0.013 0.029 0.020 0.047 0.020

Pigments (‘%DW)

0.370 0.298 0.284 0.322 0.355 0.362 0.378 0.350 0.334

K , (mZ g ’)

0.042 0.037 0.036 0.038 0.041 0.041 0.042 0.040 0.039

Proteins (‘AtDW)

30.144 27.274 25.769 30.206 29.885 29.141 30.638 28.873 26.094

2 Fatty acids (‘LIDW)

_

_

16:n

1~:2

20:4

20:5

31.666 31.554 32.395 31.341 3 1.739 30.932 30.884 30.323 31.372

12.928 12.959 13. I48 12.100 I I ,836 I 1.654

38.592 38.290 37.901 37.589

16.812 17.197 16.554 18.970 23.522 20.355 21.754 21.997 20.448

________~ i6:n

18:2

214

1.669 I531

0.682 0.629 0.631 0.665 0.639 0.657 0.660 0.649 0.565

2.034

1.555

1.720 1.691 1.742 1.844

1.784 1.587

20:s

0.886 0.834 0.795 1.043 2.065 1.084 1.986 1.147 2.088 1.299 2.168 1.294 2. I57 1 . ~ 7 3 I ,035 I.858 1.819

1 1.058 I I .02 I

11.156

3X.818

37.064 36.304 36.659 37.024

_

_

~

N

W 4

280

M.M. Rebolloso Fuentes et al. /Journal of Biotechnology 70 (1999) 271-288

these being 16:0, 18:2, 20:4 and 20:5. The araquidonic acid (20:4) is the main fatty acid with 2%DW, the next being the palmitic acid (16:0) with 1.7%DW and eicosapentaenoic acid (20:5) with 1.1%DW (Table 1). All of the fatty acids content decrease in the daylight period, from 1.74 _+ 0.05, 0.67 _+ 0.0, 2.09 _+ 0.06 and 1.08 _+ 0.06% of biomass DW at the first hours in the morning to 1.63 _+ 0.05, 0.62 _+ 0.02, 1.93 _+ 0.06 and 0.99 +0.06% of biomass DW at the last hours in the afternoon, for 16:0, 18:2, 20:4 and 20:5, respectively. However, no variation of the fatty acids profile (as total fatty acids) during the day was observed (Table 1). The polyunsaturated fatty acids, araquidonic and eicosapentaenoic, represent more than 50% of total fatty acids (37.5% of araquidonic and 19.2% of eicosapentaenoic), the rest being mainly palmitic acid with 31.3% of total fatty acids (Table 1).

4. Analysis of the experimental results

r 2 = 0.9766

I =/0 exp( - K a ' P " C)

(6)

Ia__~= exp( -- Ka 9Paverage " C ) = m = cte

(7)

Iw

Secondly, the cells tend to adapt to the average irradiance inside the culture. Thus, the higher the external irradiance and biomass concentration, the lower the extinction coefficient of the biomass must be to keep the average irradiance constant. Eq. (5) makes it possible to determine the average irradiance inside the culture, Iav at any time during the quasi steady state directly from the external irradiance, Iw. With regard to the oxygen generation rate, Ro, it appears to be a function of the average lrrad~ance inside the culture (Fig. 10). Vonshak et al. (1985) also observed this dependence as a hyperbolic variation of the photosynthesis rate with irradiance in the laboratory thin layer cultures of P. cruentum, while Jensen and Knutsen (1993) observed this same behaviour with S. platensis. However, in outdoor cultures, the existence of photoinhibition has been observed (Fig. 4). In the present study, the photosynthetic activity of the cells was very high at first hours in the morning, indicating a fast response of Porphyridium to light 9

Although in outdoor chemostat cultures of microalgae the pH, temperature and dilution rate are controlled, a daily cyclic variation of the culture parameters takes place mainly due to variations in the solar irradiance. In addition to solar irradiance, the average irradiance inside the culture, Iav, defined as the mean light intensity a cell randomly moving inside the culture intercepts, also varies during the day. The external irradiance, biomass concentration and pigment content of the biomass and thus the average irradiance inside the culture are modified during a day and between different days. A variance analysis indicates that the oxygen generation rate is influenced by the average irradiance inside the culture rather than by the irradiance on the culture surface. In this sense, a linear relationship between I~v and I,,. is observed (Fig. 9). Iav = 0.089Iw + 4.05

between Iw and lav implies the quotient Iav/Iw is a constant and therefore the product K a ' P " C is also a constant during the quasi steady state. Thus, considering a mean light path, Pa,,erage, from Eq. (7) means that the product Ka" C must also be a constant.

(5)

The linear dependence of light intensity inside fermentor and outdoor light intensity has been previously reported (Hirata et al., 1996). This linear variation of Iav with Iw brings about two important points. Firstly, a linear relationship

400 7

~

w 200

i

.

S

100 l

O' 0

1000

2000

Iw, pE/m2s

3000

4000

Fig. 9. Correlation between the average irradiance inside the culture, Iav, and the external irradiance on the culture surface, Iw during the quasi steady state reached.

M.M. Rebolloso Fuentes et al./Journal of Biotechnology 70 (1999) 271-288 0.0020 o.oo15 -~ oe

o.oolo

.~

-o

0.0005

0.0000

and regeneration of key components of the photosynthetic apparatus coexist and the available amount of functional photosynthetic pigments is the result of an equilibrium between that two processes.

# t.

*

9 -

9

-0.0005 ~ e 0

281

1O0

200

300

400

lav, pE/m2s Fig. 10. Variation of oxygen generation rate with the average irradiance inside the culture estimated by using Eq. (5). Line represents a smooth tendency curve.

availability. In the central hours of the day, when the solar irradiance is at the peak levels, the photosynthetic activity declined and then a slight increase again in the afternoon, when the solar radiation was reduced. This behaviour evidences the existence of photoinhibition that appears as a decrease in the efficiency of the culture to metabolize the available light when solar irradiance is increased, though the oxygen generation rate within the photobioreactor is not decreased because the loss in photosynthetic efficiency is compensated by a higher irradiance level. Different authors have referenced the existence of photoinhibition in outdoor cultures. Molina Grima et al. (1996a) observed the existence of photoinhibition in outdoor cultures in two ways. During the day, photoinhibition could be caused by an increase of the irradiance on the culture surface. With the position inside the culture, photoinhibition could arise caused by the strong irradiance gradient that exist in dense cultures. In this sense, Vonshak and Guy (1992) observed the existence of photoinhibition during the day in S. platensis cultures. Jensen and Knutsen (1993) demonstrated that photoinhibition is a reversible process in which degradation

5. D i s c u s s i o n

The influence of photoinhibition in the behaviour of outdoor cultures must always be considered. In outdoor cultures of P. tricornutum (Aci6n Ferndndez et al., 1998) and L galbana indoor cultures (Molina Grima et al., 1996b), the growth rate has been observed to be a function of average irradiance rather than external irradiance. Molina Grima et al. (1996a,b) report that the growth rate varies hyperbolically with the average irradiance inside the culture although the external irradiance might influence the parameters of the growth model. Additionally, Aci6n Fern~indez et al. (1998) proposed a mathematical growth model that considers the existence of photolimitation and photoinhibition, taking into account a decrease in the efficiency of the cells when external irradiance increased. Bearing these references in mind, the variation of oxygen generation rate with the average irradiance inside the culture has been fitted to a hyperbolic expression (Eqn. 8) for each single day.

Ro2

R~ = I~, + I~,, - R ~

(8)

Since the solar irradiance measured each day is different, the characteristic parameters of the model results are different (Table 2). The results show how the parameters of the model vary: n decreased while Ik increased linearly with the daily mean irradiance on the reactor surface, Iwm (Table

Table 2 Values of characteristics parameters of the model obtained by non-linear regression for each day during the quasi steady state

Day Iwm (laE 1 2 3

2000 1800 1600

m -2 s -1)

Ro2max (mol 0.002968 0.003238 0.001663

02 m -3 s -1)

n

Ik (laE m -2 s - l )

Ro2min (mol 02 m -3 s -1)

r2

0.837415 0.963006 1.325950

148.575 115.235 76.56

0.000482 0.000886 0.000082

0.99235 0.923979 0.959093

282

M.M. Rebolloso Fuentes et al./Journal of Biotechnology 70 (1999) 271-288

0.0020 1 Y = 0.91101x + 0.00005 "o 0.0015 t R2 = 0.93502

=

0.0005

.

99

*

1

o.

0.0000 1

-0.00__0.00050~40.00000.00050.00100.00150,0020 ROz experimental Fig. 11. Correlation between oxygen generation rate estimated by the model Eq. (9) and experimental oxygen generation rate.

2). Equation 8 can therefore be enhanced to take these linear relationships of n and lk with Iw. Doing so, the following model of instantaneous oxygen generation rate with external irradiance and average irradiance inside the culture is obtained. 9I( a + b / w )

Ro2max -av

R~ = (c + d" Iw)(~+ b. Iw) + I(~v+b. Zw)-

R~

(9)

Ro2max -- 0.003979 mol 02 m - 3 s - ~, a = 0.635, b = -0.000063 m2s ~tE-~, c = 706.65, laE m - 2 s-~, d = - 0.025471,

Ro2mi n

= 0.000317 mol 02 m - 3 S-1, r2._. 0.9317. The parameters of the model were obtained by non-linear regression, fitting to this equation all of the experimental data available (74 values). A representation of estimated oxygen generation rate Eq. (9) versus experimental oxygen generation rate show the adequacy of the model (Fig. 11). The maximum photosynthesis rate, 0.0040 mol O2 m - 3 s - 1 (600 ~tmol 0 2 h -~ mg Chlorophyl-~), is lower than the maximum value referenced for this strain, 1000 ~tmol 02 h-~ mg Chlorophyl-~ (Vonshak et al., 1985), this maximum being obtained under laboratory conditions. Ohta et al. (1992) observed a maximum CO2 consumption rate of 0.0044 mol m - 3 s - 1 in laboratory cultures Porphyridium under lightdark cycles. If an O2/CO2 ratio of one is accepted, the maximum photosynthesis rate obtained is the

same as predicted by the model. The minimum photosynthesis rate, 0.000317 mol O2 m - 3 s-1, is the respiration rate during the night, and although it has been considered to be constant, a certain influence of the daily culture conditions can be expected. In this sense, Torzillo et al. (1991a,b) and Molina Grima et al. (1994b,d) observed that the respiration rate is a function of the temperature and the irradiance during the daylight time. Moreover, the experimental results show that the higher the growth rates in the daylight time the lower the biomass losses during the night (Fig. 7). This same behaviour was observed in laboratory light-dark cycle cultures of Porphyridium (Ohta et al., 1992). Thus, the lower the growth rate the higher the carbohydrates synthesis and the lower the protein synthesis, the carbohydrates being consumed during the night by respiration. The obtained model Eq. (9) allows to determine the oxygen generation rate as a function of external and average irradiance inside the culture, thus being a useful tool in the design and scaleup of tubular photobioreactors. In this sense, one major problems in the scale up of tubular photobioreactors is to determine the acceptable maximum loop length, which will be limited by the maximum dissolved oxygen level that the cells can support without being affected by its toxicity. Therefore, for any given conditions of Iw and Iav, the model will estimate the oxygen generation rate. This, together with the liquid flow rate in the solar receiver, determines the maximum loop length admissible for a given maximum oxygen level. Alternatively, for a given length of the loop it is possible to calculate the liquid flow necessary to keep the dissolved oxygen under a given level and thus avoid its toxic effects. Moreover, by applying an oxygen mass balance to the overall reactor and by rejecting the accumulation term (Camacho Rubio et al., 1998), the model could be used to determine the mass transfer capacity necessary in the system in order to limit the dissolved oxygen level in the culture. O2 in - 02 out = 02 generated + 02 accumulated

(lo)

M.M. Rebolloso Fuentes et al./Journal of Biotechnology 70 (1999) 271-288 y = 41572x + 67.71

140 _

120

160 140 .'~ =

R2 = 0.8895

.

;~ 100

_ tO*****

9 ,=

20

.q

0.~

9

60 !

. 9

40

" ,W*~.,.,

"

0

-0.0005

100 i

0.0005

0.0010

N(~, m~m~

I 9Carbonlosses,._o Dissolv~

020

0.111115 0.0020

oxygen]

Fig. 12. Variation of dissolved oxygen and carbon losses in the culture with the oxygen generation rate.

d[O2] d----~

Kla([02] - [O~=]) = Ro2 +

(1 l) (12)

Kla[02] - K/a[O~ ] = R% 1

[02] = ~a ~

(13)

+ [0~']

A high photosynthesis rate causes a high dissolved oxygen level in the system. On the other hand, the higher the volumetric mass transfer, Kla, the lower the dissolved oxygen. In this sense, the dissolved oxygen linearly increases with the oxygen generation rate (Fig. 12), the slope of this line being the inverse of the volumetric mass transfer coefficient of the system. Thus, the Kla of the system can be determined from the value of this experimental slope, resulting as 0.0023 s - 1, which is very similar to that experimentally determined (Kla=O.O029 s -~) by Camacho Rubio et al. 0.0020 0.0015

E

0.0010

C~ 0.0005 0.0000 -0.0005

-0.0005

|

0.0000

,

0.0005

,

0.0010

i

0.0015

0.0020

RO=, mol/mSs

Fig. 13. Correlation between carbon consumption rate and oxygen generation rate during the quasi steady state.

283

(1998) for this same culture system. A linear relationship between oxygen generation and carbon consumption is also observed, the ratio being 1.004 mol CO2 per tool O2 (Fig. 13). This value is higher than others referenced for different strains like Chlorella pyrenoidosa (Myers, 1980) being 0.70 mol CO2 per mol O2. The difference can be attributed to the different metabolic routes of each strain. In particular, P. cruentum trends to accumulate carbohydrates and excrete polysaccharides, thus increasing the value of CO2 consumption to the 02 generation ratio. From the linearity between carbon consumption and oxygen generation, two conclusions can be drawn. First, the oxygen generation and carbon consumption are coupled, the ratio being constant and equal to one all the time and independent of cell metabolism or culture conditions. Second, the obtained model also allows to determine the carbon requirements of the system, thus becoming a tool useful to study how the carbon losses can be reduced by optimising the amount injected. In this sense, it is observed (Fig. 12) that the carbon losses are a function of oxygen generation rate. A high oxygen generation rate causes a high carbon consumption and low carbon losses. This fact points out how the injection of pure CO2 allows for reliable pH control but causes a high loss of CO2. In order to reduce these values, Camacho Rubio et al. (1998) proposed to modify the composition of the gas phase, reducing the COz molar fraction and thus reducing the carbon dioxide oversaturation in the culture. However, this mode of operation causes an increase of the pH in the culture, which lowers the carbon availability and could reduce the yield of the system (Molina Grima et al., 1996a). With regard to the biomass productivity, the observed day to day variations of culture conditions determine different growth rates and values of maximum oxygen generation rate. However, these values are interrelated, the higher the growth rate in the daylight period, the higher the maximum oxygen generation rate (Fig. 14). This same behaviour was observed in Spirulina cultures (Guterman et al., 1990). The most significant variable determining both the growth rate in the daylight period and the maximum oxygen genera-

284

M.M. Rebolloso Fuentes et al. ~Journal of Biotechnology 70 (1999) 271-288 0.0022 y = 0.1216x + 0.0005

0

R 2 = 0.9627

o.ools

E

~ 0.0014

C 0.0010

--

0.004

,

,

,

,

0.006

0.008

0.010

0.012

I~ol, h "1

Fig. 14. Variation of maximum oxygen generation rate with the additional growth rate in the daylight time, ~tso=for 3 days during the quasi steady state.

tion rate is the mean average irradiance inside the culture in the daylight period. Moreover, both the growth rate and the maximum oxygen generation rate linearly increase with the mean average irradiance inside the culture in the daylight period (Fig. 15). The higher the solar irradiance or the lower the biomass concentration, the higher the mean average irradiance and light availability, thus the growth rate and photosynthesis rate being highest. Aci~n Fernfindez et al. (1998) observed that the growth rate of P. tricornutum varies hyperbolically with the mean average irradiance inside the culture, this relationship being approximately linear until Iav values of 200 gE m -2 s-1, with the slope strongly decreasing for irradiances above this value. The linear relationship between growth rate and mean average irradiance observed agreed with the linear relationship observed for P. tricornutum at low 0.012

0.0030

y = 0.00028x - 0.03985

0.010

R= = 0 . 9 ~

,,,,,,` ' ~

O.0025

0.008 0

~.oo6

0.0020 E

0.004

0.0015 n,,

0.002 R 2 = 0.981510

0.000 160

I

1

,

165

170 lawn, pE/rn=s

175

0.0010 180

Fig. 15. Variation of additional growth rate in the daylight period, ~tso~, and maximum oxygen generation rate with the mean average irradiance inside the culture, Iavm.

values of Iavm, thus indicating that the same behaviour could be accepted in this case although further research is necessary at different Iav values. With regard to elemental analysis of culture, supernatant and biomass, the carbon content of the supernatant showed a variation parallel to the solar irradiance during the day, an increase during the morning and a decrease in the afternoon (Fig. 8). However, Camacho Rubio et al. (1998) previously demonstrated that cultures operated with pH control by on-demand CO2 injection had a lower carbon content in the central hours of the day, due to the consumption by the cells, than in the first and last hours of the daylight period, by a reduction in photosynthesis and respiration The carbon content of the supernatant changes in a opposing way. Thus, at noon the carbon content of the supernatant is 0.200 g 1-~ whereas Camacho Rubio et al. (1998) estimate that there must be 0.072 g 1-1 of total inorganic carbon in the culture. The difference corresponds to the organic carbon excreted to the medium by the cells, 0.128 g 1 - 1 consisting mainly of exopolysaccharides. Vonshak et al. (1985) observed exopolysaccharide concentrations of around 0.24 g 1-1, double than the values measured in the present study, however this could be due to the low dilution rate, 0.02 h-1, and low pH, 7.5, at which the cultures were operated. In the first and last hours of the day, the carbon content in the supernatant decreased, 0.13 g 1-1, while the total inorganic carbon estimated by Camacho Rubio et al. (1998) increased, 0.077 g 1-1. This difference indicates a lower exopolysaccharides concentration in these hours, 0.053 g 1-1, and reflects that exopolysaccharides synthesis modifies during the day increasing at noon due to the existence of adverse culture conditions by photoinhibition. Thus, Vonshak et al. (1985) found that Porphyridium excretes big amounts of exopolysaccharides under adverse conditions as nitrogen limitation, a low growth rate, etc. The cellular polysaccharides are originated by the continuous excretion of polysaccharides, which form an envelope around the outer cell wall. The outer layer of this envelope is gradually sloughed off, enriching the growth medium and constituting the exocellular polysaccharide fraction (Vonshak et al., 1985).

M.M. Rebolloso Fuentes et al./"Journal of Biotechnology 70 (1999) 271-288

With respect to nitrogen, the second macronutrient, the amount of nitrogen supplied daily to the system with the medium is 27.4 g d - 1. However, the elemental analysis of the produced biomass indicated that only 16.03 g d-~ of nitrogen is taken up, thus the yield of this nutrient being 58%. In addition, the nitrogen content of supernatant decrease during the day, being a function of the daily variation of the cellular metabolism and not of the oxygen generation rate. Therefore, the results of elemental analysis of the biomass showed how during the illuminated period the metabolism is mainly directed to the synthesis of carbohydrates and lipids--carbon, oxygen and hydrogen content of the biomass increasing 1.2, 1.0 and 0.3%, respectively--while during the dark period, the accumulated carbohydrates were consumed and the metabolism was directed to the synthesis of proteins--nitrogen and sulphur content of the biomass increasing 3.7 and 0.2%, respectively--and structural lipids. The nitrogen is mainly taken during the afternoon and night when the protein synthesis is active. In this sense, the results confirm the idea of Professor Gudin (personal communication) that the nitrogen should be better supplied to the system during the night to avoid nitrogen limitation. However. in this research nitrogen limitation did not exist due to the excess of nitrogen supplied. A daily variation of cellular metabolism was observed in outdoor cultures of P. tricornutum (Aci6n Fernfindez et al., 1998), S. platensis (Bocci et al., 1988; Torzillo et al., 1991a,b) and Oscillatoria agardhii (Van Liere et al., 1979) and it is attributed to a response of the cells to the daily solar cycle, because the cells store energy during the illuminated period which is metabolised in the night period. The mean elemental composition of the P. cruentum biomass was estimated as C1Hl.v6N0.1180.olO0.79, a value similar to that proposed by different authors for microalgal biomass (Borowitzka and Borowitzka, 1988). The daily variation of the cellular metabolism also caused the variations on the fatty acids content of the biomass. The content of all the fatty acids decreases during the daylight time, the decrease being sharper for polyunsaturated fatty acids, 18:2, 20:4 and 20:5 (8-9%), than for satu-

285

rated and monounsaturated fatty acids, 16:0 (6%) (Fig. 14). In P. tricornutum (Aci6n Fernfindez, 1996) and I. galbana outdoor cultures (Molina Grima et al., 1995), an increase in the fatty acids content during the daylight period was observed, the increase being higher for the short-chain than for the long-chain fatty acids. This behaviour can be attributed to the different roles of the fatty acids in the cell metabolism. Short-chain saturated and monounsaturated fatty acids were found to be the main components of neutral lipids (storage lipids) (Molina Grima et al., 1994a,b, 1995), and the initial step in the path of fatty acid synthesis (Hodgoson et al., 1991). Thus, the amount of these fatty acids is subject to the daily cyclic variation of environmental conditions. On the other hand, polyunsaturated fatty acids are structural lipids, mainly found in glycolipids and phospholipids (Molina Grima et al., 1994a,b), and therefore, their contents are more related to state of growth than to short-term environmental variations. However, in this research a decrease in all the fatty acids content was observed. This decrease can be attributed to the accumulation of carbohydrates during the daylight period and highlights that the synthesis of lipids in P. cruentum takes place mainly in the darkness, when the carbohydrates accumulated are metabolised. Moreover, the lower decrease in the fatty acid content of 16:0 than 18:2, 20:4 and 20:5 indicates that the behaviour observed in P. tricornutum and I. galbana is also reproduced in P. cruentum, although variation of fatty acid profile during the day was not significant. With regard to the fatty acids profile (as total fatty acid), although the biomass was obtained in quasi steady state, the differences in the additional growth rate in each day determined a variation on the fatty acids profile. The higher the growth rate the higher the polyunsaturated fatty acid fraction and the lower the 16:0 and 18:2 fatty acid fraction (Fig. 16). This behaviour can be attributed to the reduction in the duplication time when growth rate increases, that increases the requirements of structural biomolecules while the storage of energy as lipids is reduced. This phenomenon causes a reduction in the fraction of saturated and monounsaturated fatty acids, and an increase of

286

M.M. Rebolloso Fuentes et al. / Journal of Biotechnology 70 (1999) 271-288 -~ 0.40

2.1

'~ 1.8

0.30 ~.

"o

o~ 1.5

u

1.2

toO.9

"o

Z"

0.20 g

T

.T,

c

,,T.

O.lO .~

..=_

m 0.6 u.

n

l

0.3

v

,

,

,

l

10

12

14 Hour

16

18

9

I--n-- 16:0

~

18:2

o

20:4 --e--20:5 --e--Pigmefcs I

24

4O ai

38

=I-

20.

al.

,r ~534 d

0.00

12

r

,

w~

32 4

"- 30 28 0.004

0.006 I n

16:0

0.008 psol, h"1 e

20:4

~

0.010

""

0 0.012

18:2-~--20:51

Fig. 16. Variation of biochemical composition of the biomass with the additional growth rate, gsol, during the quasi steady state reached. polyunsaturated acids that are the main constituents of cellular membranes (Kates and Volcani, 1966).

6. C o n c l u s i o n s

In outdoor systems, daily variations of culture conditions take place due to the variation of solar irradiance. In order to respond to these variations, the cells adapt their behaviour. Some of these modifications are fast and are a direct function of culture conditions in a short time interval (seconds, minutes or a few hours). However, others are slower and are a function of culture conditions averaged over a long time interval (all the daylight time, all the night or all the day). Short term responses of oxygen generation rate, carbon consumption rate and exopolysaccharides production are observed whereas the growth rate and

metabolic routes show long-term responses. In this sense, in outdoor P. c r u e n t u m cultures, the oxygen generation rate is mainly determined by the average irradiance inside the culture, although the existence of photoinhibition, causing a reduction in the efficiency of the system to utilise the solar radiation, is observed. Considering this, a model for the photosynthesis rate as both oxygen generation rate and carbon dioxide consumption rate is obtained. This model allows to estimate the photosynthesis rate as a function of the solar irradiance--a function of system geometry and location of the factory--and average irradiance inside the culture--a function of the imposed dilution rate, biomass concentration and pigment content (Aci6n Fern~.ndez et al., 1 9 9 8 ) g a t which the cells are exposed to. Moreover, the linear relationship between the oxygen generation rate and carbon consumption observed allows the model to also estimate the carbon requirements of the system. The obtained model is therefore a useful tool in determining design parameters such as maximum solar receiver length of tubular external loops or the requirements of mass transfer in the system in order to avoid the deleterious effects of oxygen oversaturation. The elemental analysis of the biomass revealed how the daily variation of the cellular metabolism differs in the variation of the photosynthesis rate, indicating that the storage of energy by photosynthesis and its use by the cellular metabolism are not simultaneous. During daylight time, carbohydrates were accumulated by the biomass as energy storage, to be metabolised later during the night to synthesise proteins and structural lipids. This behaviour indicates that a supply of nitrogen during the night could be necessary, while the carbon injection can be switched-off in order to reduce the carbon losses.

Acknowledgements

This research was supported by the Comisi6n Interministerial de Ciencia y Tecnologia (CICYT) (BIO-95-0692) (Spain), and Plan Andaluz de Investigaci6n II, Junta de Andalucia.


References Aci6n Fernfindez, F.G., 1996. Operaci6n en continuo de fotobiorreactores tubulares externos con la microalga Phaeodactylum tricornutum. Tesis Doctoral. Universidad de Almeria. Aci6n Fern~indez, F.G., Garcia Camacho, F., S~inchez P6rez, J.A., Fern~indez Sevilla, J.M., Molina Grima, E., 1997. A model for light distribution and average solar irradiance inside outdoor tubular photobioreactors for the microalgal mass culture. Biotechnol. Bioeng. 55 (5), 701 - 714. Aci6n Fernfindez, F.G., Garcia Camacho, F., S~inchez P6rez, J.A., Fernfindez Sevilla, J.M., Molina Grima, E., 1998. Modelling of biomass productivity in tubular photobioreactors for microalgal cultures: effects of dilution rate, tube diameter and solar irradiance. Biotechnol. Bioeng. 58 (6), 605-616. Ahern, T.J., Katoh, S., Sada, E., 1983. Arachidonic acid production by the red alga Porphyridium cruentum. Biotechnol. Bioeng. 25, 1057-1070. Alfano, O.M., Romero, R.L., Cassano, A.E., 1986. Radiation field modeling in photoreactors-I Homogenious media. Chem. Eng. Sci. 41,421-444. Arad, S.M., Adda, M., Cohen, E., 1985. The potential of production of sulphated polysaccharides from Porphyridium. Plant Soil 89, 117-128. Arad S.M., 1987. Biochemicals from unicellular red algae. International Industrial Biotechnology, June/July, pp. 281-284. Bocci, F., Torzillo, G., Vincenzini, M., Materassi, R., 1988. Growth physiology of Spirulina platensis in tubular photobioreactor under natural light, Algal Biotechnology, Elsevier Applied Science, London, pp. 219-228. Borowitzka, L.J., Borowitzka, M.A., 1988. Micro-Algal Biotechnology. Cambridge University Press, Cambridge. Borowitzka, M.A., 1996. Closed algal photobioreactors: design considerations for large-scale systems. J. Mar. Biotechnol. 4, 185-191. Camacho Rubio, F., Aci6n Fern~indez, F.G., S~nchez P6rez, J.A., Garcia Camacho, F., Molina Grima, E., 1998. Prediction of dissolved oxygen and carbon dioxide concentration profiles in tubular photobioreactors for microalgal culture. Biotechnol. Bioeng. 62 (1), 71-86. Chini Zitelli, G., Tomasello, V., Pinzani, E., Tredici, M.R., 1996. Outdoor cultivation of Arthrospira platensis during autumn and winter in temperate climates. J. Appl. Phycol. 8, 293-301. Collos, Y., Slawyk, G., 1980. Nitrogen uptake and assimilation by marine phytoplankton. In: Falkowski, P.G. (Ed.), Primary productivity in the sea. Plenum Press, New York, pp. 195-211. Duffle, J.A., Beckman, W.A., 1980. Solar Engineering of Thermal Processes. Wiley & Sons, New York. Evers, E.G., 1991. A model for light limited continuous cultures: growth, shading and maintenance. Biotechnol. Bioeng. 38, 599-605. Garcia S~inchez, J.L., Molina Grima, E., Garcia Camacho, F.,

287

S~nchez P6rez, J.A., Gim6nez Gim6nez, A., 1993. Cuantificaci6n de ~icidos grasos a partir de biomasa microalgal. Grasas y Aceites 44 (6), 348-353. Gudin, C., Chaumont, D., 1983. Solar biotechnology study and development of tubular solar receptors for controlled production of photosynthetic cellular biomass. In: Palz, W, Pirrwitz, D. (Eds.), Proceedings of the Workshop and E.C. Contractor's meeting in Capri. Reidel Publishing Company, Dordrecht, pp. 184-193. Gudin, C., Therpenier, C., 1986. Bioconversion of solar energy into organic chemicals by microalgae. Adv. Biotechnol. Process. 6, 73-110. Gudin C., 1989. An overview of microalgae biotechnology for fine chemicals. The first International Marine Biotechnology Conference, 3-6 September, Tokyo, Japan. Guterman, H., Vonshak, A., Ben-Yaakov, S., 1990. A macromodel for outdoor algal mass production. Biotechnol. Bioeng. 35, 809-819. Hansmann, E., 1973. Pigment analysis. In: Stein, J.R. (Ed.), Handbook of Psychological Methods, Culture Methods and Growth Measurements. Cambridge University Press, London, pp. 359-368. Hemerick, G., 1973. Culture methods and growth measurements. In: Stein, J.R. (Ed.), Handbook of Physiological Methods. Cambridge University Press, Cambridge, UK, pp. 259-260. Hirata, S., Hayashitani, M., Taya, M., Tone, S., 1996. Carbon dioxide fixation in batch culture of Chlorella sp using a photobioreactor with a sunlight collection device. J. Ferment. Bioeng. 81 (5), 470-472. Hodgoson, P.A.. Henderson, R.J., Sargent, J.R., Leftley, J.W., 1991. Patterns of variation in the lipid class and fatty acid composition of Nannochloropsis oculata (Eustigmaphiceae) during batch culture. I. The growth cycle. J. Appl. Phycol. 3, 169-181. Jensen, S., Knutsen, G., 1993. Influence of light and temperature on photoinhibition of photosynthesis in Spirulina platensis. J. Appl. Phycol. 5, 495-504. Kates, M., Volcani, B.E., 1966. Lipid content of diatoms. Biochem. Biophys. Acta 116, 264-278. Lee. Y.K., 1986. Enclosed bioreactors for the mass cultivation of photosynthetic microorganisms: the future trend. TIBTECH 4. 186-189. Lee, Y.K., Low. C.S., 1992. Productivity of outdoor algal cultures in enclosed tubular photobioreactor. Biotechnol. Bioeng. 40, 1119-1122. Lee, Y.K., Low, C.S., 1993. Productivity of outdoor algal cultures in unstable weather conditions. Biotechnol. Bioeng. 41, 1003-1006. Lepage, G., Roy, C., 1984. Productivity of outdoor algal cultures in unstable weather conditions. J. Lipid Res. 25, 1391-1396. Little, A.D., 1953. Pilot plant studies in the production of Chlorella. In: Burlew, J.S. (Eds.), Algal culture from laboratory to pilot plant. Carnegie Institute of Washington, pp. 235-273.

288


Molina Grima, E., Garcia Camacho, F., Sfinchez P6rez, J.A., Aci6n Fernandez, F.G., Fernandez Sevilla, J.M., Vald6s Sanz, F., 1994a. Effect of dilution rate on eicosapentaenoic acid productivity of Phaeodactylum tricornutum UTEX 640 in outdoor chemostat culture. Biotechnol. Lett. 16 (10), 1035-1040. Molina Grima, E., Garcia Camacho, F., Sanchez Perez, J.A., Urda Cardona, J., Aci6n Fernandez, F.G., Fern~indez Sevilla, J.M., 1994b. Outdoor chemostat culture of Phaeodactylum tricornutum UTEX 640 in a tubular photobioreactor for the production of eicosapentaenoic acid. Biotechnol. Appl. Biochem. 20, 279-290. Molina Grima, E., Garcia Camacho, F., S~inchez P6rez, J.A., Fernandez Sevilla, J.M., Aci6n Fernfindez, F.G., Contreras G6mez, A., 1994c. A mathematical model of microalgal growth in light limited chemostat culture. J. Chem. Tech. Biotechnol. 61, 167-173. Molina Grima, E., S~nchez P6rez, J.A., Garcia Camacho, F.. Garcia S~nchez, J.L., Aci6n Fernandez. F.G., Lopez Alonso, D., 1994d. Outdoor culture of Isochrvsis galbana ALII-4 in a closed tubular photobioreactor. J. Biotechnol. 37, 159-166. Molina Grima, E., S~,nchez P6rez, J.A., Garcia Camacho. F.. Garcia S~nchez, J.L., Fernandez Sevilla, J.M., 1995. Variation of fatty acid profile with solar cycle in outdoor chemostat culture of Isochrysis galbana ALII-4. J. Appl. Phycol. 7, 129-134. Molina Grima, E., S~nchez P~rez, J.A., Garcia Camacho. F.. Fernfindez Sevilla, J.M., Aci6n Fernfindez, F.G, 1996a. Productivity analysis of outdoor chemostat culture in tubular air-lift photobioreactors. J. Appl. Phycol. 8, 369-380. Molina Grima, E., Fernfindez Sevilla, J.M., Sfinchez P6rez, J.A., Garcia Camacho, F., 1996b. A study on simultaneous photolimitation and photoinhibition in dense microalgal cultures taking into account incident and averaged irradiances. J. Biotechnol. 45, 59-69. Myers, J., 1980. On the algae: thoughts about physiology and measurements of efficiency. In: Falkowsky, P.G. (Ed.), Primary Productivity in the Sea. Plenum Press, New York, pp. 1-15. Nichols, B.W., Appleby, R.S., 1969. The distribution and biosynthesis of arachidonic acid in algae. Phytochemistry 8, 1907-1915. Ohta, S., Chang, T., Aozasa, O., Kondo, M., Miyata, H., 1992. Sustained production of arachidonic and eicosapentaenoic acids by the red alga Porphyridium purpureurn cultured in a light/dark cycle. J. Ferment. Bioeng. 74, 398-402. Percival, E., Foyle, R.A.J., 1979. The extracellular polysaccha-

rides of Porphyridium cruentum and aerugineum. Carbohydr. Res. 72, 165-176.

Porphyridium

Qiang, H., Richmond, A., 1996. Productivity and photosynthetic efficiency of Spirulina platensis as affected by light intensity, algal density and rate of mixing in a flat plate photobioreactor. J. Appl. Phycol. 8, 139-145. Ramus, F., 1972. The production of extraceUular polysaccharides by the unicellular red alga Porphyridium aerugineum. J. Phycol. 8, 97-111. Rebolloso Fuentes, M.M., Aci6n Fernandez, F.G., S~nchez P6rez, J.A.. Gil Garcia, M.D., Guil Guerrero, J.L., 1998. Variacion de la composici6n nutritiva (composici6n centesimal, elementos minerales y ~cidos grasos) de la biomasa microalgal de Porphyridium cruentum. Food Sci. Technol. Int. (in press). Richmond, A., Boussiba, S., Vonshak, A., Kopel, R., 1993. A new tubular reactor for mass production of microalgae outdoors. J. Appl. Phycol. 5, 327-332. Sukenik, A., Falkowski, P.G., Bennett, J., 1986. Potential enhancement of photosynthetic energy conversion in algal mass culture. Biotech. Bioeng. 30, 970-977. Torzillo, G., Sacchi, A., Materassi, R., 1991a. Temperature as an important factor affecting productivity and night biomass loss in Spirulina platensis grown outdoors in tubular photobioreactors. Biores. Technol. 38, 95-100. Torzillo, G., Sacchi, A., Materassi, R., Richmond, A., 1991b. Effect of temperature on yield and night biomass loss in Spirulina platensis grown outdoors in tubular photobioreactors. J. Appl. Phycol. 3, 103-109. Torzillo, G., Carlozzi, P., Pushparaj, B., Montaini, E., Materassi, R., 1993. A two plane tubular photobioreactor for outdoor culture of Spirulina. Biotechnol. Bioeng. 42, 891898. Tredici, M.R., Materassi, R., 1992. From open ponds to alveolar panels: the Italian experience in the development of reactors for the mass cultivation of phototrophic microorganisms. J. Appl. Phycol. 4, 221-231. Van Liere, L., Mur, L.R., Gibson, C.E., Herdman, M., 1979. Growth and physiology of Oscillatoria agardhii cultivated in continuous culture with a light-dark cycle. Arch. Microbiol. 123, 315- 318. Vonshak, A., Cohen, Z., Richmond, A., 1985. The feasibility of mass cultivation of Porphyridium. Biomass 8, 13-25. Vonshak, A., Guy, R., 1992. Photoresponse, photoinhibition and productivity in the blue-green alga, Spirulina platensis, grown outdoors. Plant Cell Environ. 15, 613-616. Whyte, J.C., 1987. Biochemical composition and energy content of six species of phytoplankton used in mariculture of bivalves. Aquaculture 60, 231-241.

iii

i

I O U I t N A L

OF



......

An integrated solar and artificial light system for internal illumination of photobioreactors James C. Ogbonna a

a,b,, Toshihiko Soejima a, Hideo Tanaka a,b

Institute of Applied Biochemistry, University of Tsukuba, I-I-1 Tennodai, Tsukuba 305-8572, Japan b CREST, Japan Science and Technology Corporation (JST), Tsukuba, Japan


Abstract Exploitation of photosynthetic cells for the production of useful metabolites requires efficient photobioreactors. Many laboratory scale photobioreactors have been reported but most of them are extremely difficult to scale up. Furthermore, the use of open ponds and outdoor tubular photobioreactors is limited by the requirement for large spaces and the difficulty in maintaining sterile conditions. In view of this, we have designed and constructed an internally illuminated stirred tank photobioreactor. The photobioreactor is simple, heat sterilizable and mechanically agitated like the conventional stirred tank bioreactors. Furthermore, it can easily be scaled up while maintaining the light supply coefficient and thus the productivity constant. A device was installed for collecting solar light and distributing it inside the reactor through optical fibers. It was equipped with a light tracking sensor so that the lenses rotate with the position of the sun. This makes it possible to use solar light for photosynthetic cell cultivation in indoor photobioreactors. As a solution to the problems of night biomass loss and low productivity on cloudy days, an artificial light source was coupled with the solar light collecting device. A light intensity sensor monitors the solar light intensity and the artificial light is automatically switched on or off, depending on the solar light intensity. In this way, continuous light supply to the reactor is achieved by using solar light during sunny period, and artificial light at night and on cloudy days. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Photobioreactor; Internal illumination; Indoor culture; Solar light; Artificial light

1. Introduction Solar light energy is one of the most a b u n d a n t natural resources on earth and since artificial light sources are expensive, it is desirable to utilize

* Corresponding author. Fax: + 81-298-534605. E-mail address: [email protected] (J.C. Ogbonna)

solar energy for photosynthetic cell cultivation. Thus, most commercial cultivation of photosynthetic cells is carried out in open ponds, utilizing solar light energy. However, the productivities of these o u t d o o r open ponds are very low (Terry and R a y m o n d , 1985; Laws et al., 1986) due to m a n y problems including the difficulties of controlling the culture conditions, c o n t a m i n a t i o n problems, and requirement for large area of land. Therefore,

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)0008 1-4

290

J.C. Ogbonna et al./Journal of Biotechnology 70 (1999) 289-297

there is a need to develop a compact and sterilizable photobioreactor which can be efficiently illuminated by solar light. In order to achieve this, we have investigated growth kinetics of photosynthetic cells (Ogbonna et al., 1995b) and developed an index for quantitative evaluation of light conditions inside photobioreactors (Ogbonna et al., 1995c). Based on these, a method for designing and scaling up of internally illuminated photobioreactor was proposed and a photobioreactor which can be internally illuminated by either artificial or solar light energy was constructed (Ogbonna et al., 1996; Ogbonna and Tanaka, 1997). Day-night cycles and diurnal variation in light intensity is a major problem with the use of solar light energy. Depending on the location and season, the number of hours per day when the light intensity is high enough to support photosynthetic cell growth can be very short. In the absence of light energy or some other metabolizable organic carbon source in the medium, cells metabolize the cell components to obtain maintenance energy, thus leading to a decrease in cell weight. Both the productivity and biochemical composition of the cells were affected by light/dark cycles during the cultivation of Chlorella pyrenoidosa (Ogbonna and Tanaka, 1996). Depending on the growth phase and cultivation conditions, the decrease in the total biomass concentration during the night may be as high as 17% (Ogbonna and Tanaka, 1996). Results of other investigations (Grobbelaar and Soeder, 1985; Torzillo et al., 1991a,b) have also shown that as high as 35% of the biomass produced during the day may be lost through respiration at night. Furthermore, prolonged bad weather (cloudy and/or rainy days) often lead to total process failure when only solar light energy is used for cultivation. As a solution to this problem, we have reported that cyclic autotrophic/heterotrophic cultivation (whereby controlled amount of organic carbon source is added during the night period), can be used not only to prevent night biomass loss but to achieve continuous cell growth even under light/dark cycles (Ogbonna and Tanaka, 1996, 1998). However, this method can only be used for cells which can

grow heterotrophically. Furthermore, during the dark heterotrophic growth phase, the cellular concentrations of the photosynthetic metabolites decrease and recover only during the subsequent autotrophic growth phase. Thus there is an optimum ratio of autotrophic to heterotrophic growth phase within a cycle and this method would not work well if the dark period is not followed by an appropriate period of good sunshine. Thus, it would not work well in cases of prolonged period of bad weather. In this work, we have developed a system for internal illumination of indoor stirred tank photobioreactor using solar light. In order to overcome the problems of diurnal variation in solar light intensity and prolonged periods of bad weather, an illumination system with integrated solar and artificial light sources was developed. Solar light is used for illumination during the day but when the solar light intensity decreases below a set value (during cloudy days and at night) the system switches automatically to artificial light source, thus ensuring continuous light supply to the reactor.


2.1. Photobioreactor A schematic diagram showing the concept and construction of the photobioreactor used in this study is shown in Fig. 1. The illumination capacity of a single light source is determined for the desired cell and process, using the reactor shown in Fig. I(A). The cells are cultivated in vessels of various diameters and the optimum light path (OL) which can be efficiently illuminated by the single light source is determined. Such a reactor with a single light source at the center is defined as a one-unit photobioreactor. A large photobioreactor with the same light condition as the single unit can be constructed by increasing the number of units in three dimensions (Ogbonna et al., 1996, 1997b). The optimum light path depends on the cell and the nature of product. Using a 4-W fluorescent lamp (light intensity = 163 ~tmol


291

•

!

Glass

i

Aeration E x h a u s t gas

:OLi

I ! |

Glass tube r

IlerI Light ator

'~"

"k II

! |

! I

Illlll ~

I

NaOH

Light switch

Fig. 1. Schematic diagrams of the set-up used to determine the optimum unit size (A) and a 4-unit photobioreactor (B) used in this study. Detailed description of this photobioreactor is given in the text.

m - 2 S-1), the optimum light path for production of Chlorella biomass was determined to be 2.3 cm and a 4-unit photobioreactor was constructed (Fig. 1B). By keeping the distance between two light radiators equal to 2 x OL (4.6 cm), the light condition (both the light energy supplied per unit volume and the light distribution inside the reactor) in the 4-unit photobioreactor was the same as that in the single unit. The working volume of the 4-unit photobioreactor is 3.5 1. Since the light radiators are not mechanically fixed to the reactor, the reactor can be heat-sterilized and illuminated after cooling by inserting the light radiators into the glass tubes. A modified impeller was installed to achieve good mixing and high rate of mass transfer inside the reactor. The glass tubes which house the light radiators also serve as baffle plates in breaking the gas bubbles and thus improving mass transfer and light refraction inside the reactor. Since the light radiators have no direct contact with the culture broth, there is no problem of cell adhesion on the light radiators.

2.2. Solar light collection and transmission into the photobioreactor Fig. 2 shows the schematic diagram of the device used for collecting solar light and distributing it inside the reactor. It comprised of solar light collectors each with 12 fresnel lenses of 105 mm in diameter. The total illumination surface for each ~Sun~

Lightcollectiondevice

,,-~-~ nb, r~

Photobioreactor

Fig. 2. Schematic diagram of the solar light collection device. The collected light is transmitted to the photobioreactor through the optical fibers.

292


solar light collector was 864 cm 2. It was covered by a transparent acryl dome which protects the lenses from dust and rain. The solar light collected by each lens was transmitted through an optical fiber (1 mm in diameter). The light collection device was equipped with a light tracking sensor so that the lenses rotate with the position of the sun. Approximately 38% of the collected photosynthetic active radiation was transmitted into the reactor. The solar light collection device is commercially available (Laforet Engineering and Information Service, Tokyo, Japan) under a trade name of 'Himawari'. In order to use this solar light collector for internal illumination of tank-type photobioreactor, six optical fibers (from six fresnel lenses) were made into a bundle and connected to a light radiator. Thus, there were two light radiators per solar light collector and thus two solar light collectors for the 4-unit photobioreactor. The light radiators were cylindrically shaped and made of vertically etched transparent glass or quartz. The light intensity is almost uniform throughout the illumination surface of the light radiators.

2.3. Integrated solar and artificial light illumination system Schematic diagram of the integrated solar and artificial light illumination system is shown in Fig. 3. A metal halide lamp was used as the artificial light source. Parabolic mirrors were used to produce parallel rays from the lamps which were then filtered and transmitted through the optical fibers to the light radiators inside the photobioreactor as described for the solar light. The same light radiators were used for both the solar light and metal halide lamp. A light intensity sensor monitors the solar light intensity and the artificial light is automatically switched on or off, depending on the solar light intensity. In this way, light energy is supplied continuously to the culture.

2.4. Micro-organism and medium composition Chlorella sorokiniana IAM C-212 formerly known as C. pyrenoidosa (Kessler and Huss, 1992) was obtained from the algal collection of the

Light tracking sensor

Fresnel lenses

Light collection device Control box

Optical fibers

Metal lamp

Fig. 3. A schematic diagram of the integrated solar and artificial light for internal illumination of the stirred tank photobioreactor.

Institute of Applied Microbiology, University of Tokyo, Japan. The medium was composed of (g 1-i) KNO3, 1.25; KHzPO 4, 1.25; MgSO4" 7H20, 1.25; CaC12, 0.04; FeSO4" 7H20, 0.002; and A5 solution, 1.0 ml 1-1. The composition of the A5 solution was given previously (Ogbonna et al., 1995a). The pH of the medium was adjusted to 6.8 before autoclaving at 121~ for 15 min.

2.5. Cultivation method Cells were activated by pre-cultivating them in 100-ml Roux bottles under continuous illumination at 36~ for 24 h. White fluorescent lamps were used as the light source and the light intensity at the surface of the bottle was 200 lamol rn- 2 s - 1. Aeration and mixing were achieved by sparging air enriched with 5% CO2 through a glass-ball filter, which was inserted to the bottom of the Roux bottle, at 0.3 vvm. The seed culture broth was used to inoculate the main culture in the photobioreactor. The agitation speed, aeration rate (with 5% CO2 in air) and cultivation temperature were 120 rpm, 0.3 vvm and 36~ respectively. Cultivation was done either with only solar light illumination or with the integrated solar and artificial light system. The artificial light was automatically switched on

J.C. Ogbonna et al. /'Journal of Biotechnology 70 (1999) 289-297

whenever the solar light intensity on the surface of the light radiators decreases below 50 lamol m - 2 s - l and off when the solar light intensity increases above this value. Cyclic autotrophic-heterotrophic cultivation was carried out as described in our previous papers (Ogbonna and Tanaka, 1996, 1998), using ethanol as the organic carbon source at night.

L_

:.

i

~" ~'~ ^ ! [~f~

:~

I

i!) [ii

INFRA-RED

100 ~ i, v ~ s ~ Y

~, 80~

,

293

t m Direct sunlight spectru

- -

surface~176

/

2.6. Analytical methods Cell dry weight (DW) determinations were made according to the method described in our previous papers (Ogbonna et al., 1997a). When the cell concentration was very low, it was estimated by measuring the optical density at 680 nm (Spectronic 20A, Shimadzu Scientific Instruments, Japan). In the latter case, the OD readings were converted to dry cell concentrations using predetermined calibration curves. The light intensities were measured by an analogue photometer (LI185B, Licor, Nebraska, USA).

3. Results 3.1. Wavelength distribution in the radiated light A comparison of the wavelength spectrum of the direct sunlight and that of the radiated light inside the photobioreactor is shown in Fig. 4. The distribution patterns are similar in the direct sunlight and the radiated light. However, most of the infrared ( ~ 90%) and ultraviolet ( ~ 99.1%) wavelengths were removed in the radiated light due to chromatic aberration in the fresnel lenses. Also, by using lenses to channel the metal halide light to the optical fiber, all the ultraviolet and a significant percentage of the infrared wavelengths were filtered out before transmission into the photobioreactor (Fig. 5).

20~ /:f i

\ i

280 500

An example of a time course of C. sorokiniana growth and the average light intensity during the day, when only the solar light was used for culti-

i ~

1000 1500 Wavelength (nm)

1

2000

~

2500

Fig. 4. Wavelength distribution of the direct solar light and the transmitted light on the surface of the light radiators.

vation, is shown in Fig. 6. In this example, the first two days of good weather was followed by a sunny morning but cloudy afternoon, two rainy days, a cloudy morning but sunny afternoon, another two rainy days and two days of good weather. During the sunny days, the average light intensity at the surface of the glass tubes housing the light radiators was ~ 450 gmol m - 2 S - 1. This decreased to ~ 200 gmol m - 2 s-~, when only a part of the day was sunny. On rainy/cloudy days, the solar light intensity could be as high as 400 gmol m -2 s - l , but the light intensity at the 100 80

._ 40

i

20

3.2. Cultivation of C. sorokiniana using only the solar light collection device

t

0

x,

__] 300

400

600 500 Wavelength (rim)

700

800

Fig. 5. Wave length spectrum of the metal halide lamp transmitted to the photobioreactor.

J.C. Ogbonna et al. :/Journal of Biotechnology 70 (1999) 289-297

294

500 ~

5

400 ~

~

300 ~

.o.

200 =

~:

4

e~ e~

~0.5

0 0

V

"L)

100 200 Cultivation time (h)

-9

o

100 ~

=

o

2

~' 1

0 300

Or 0

Fig. 6. Time course of C. sorokiniana growth using only solar light transmitted through the optical fibers.

surface of the light radiators was almost zero. This could be due to light diffusion/dispersion which reduces the efficiency of light collection by the solar light collectors during rainy/cloudy weather. The pattern of cell growth followed that of the solar light intensity. The average linear growth rate during the first two days of good weather was 0.3 g 1-~ d-~ which corresponds to CO2 fixation rate of 0.54 g - C O 2 1 - 1 d - 1 . There was no significant increase in the cell concentration during the following bad weather but when the good weather returned, the cell concentration increased rapidly with an average linear growth rate of 0.48 g 1-1 d-1 and C02 fixation rate of 0.864 g C02 1-1 d-1.

3.3. Cyclic autotrophic-heterotrophic cultivation We have already shown that cyclic autotrophic-heterotrophic cultivation can be used to achieve continuous cell growth under light/ dark cycles (Ogbonna and Tanaka, 1998). As shown in Fig. 7, when artificial light was used at a definite cycles of 10 h light and 14 h dark, very high cell concentration with high protein content (56%) was achieved by adding controlled amounts of ethanol at night. This method has also been shown to work well for ~-tocopherol production by Euglena gracilis (Ogbonna and Tanaka, 1998). However, it is applicable only to cells with heterotrophic metabolism. Furthermore, when this cyclic autotrophic-heterotrophic cultivation was done using only the solar light, the protein con-

150

50 100 Cultivation time (h)

Fig. 7. Time courses of C. sorokiniana cell growth during autotrophic cultivation under light/dark cycles (O) and during cyclic autotrophic-heterotrophic cultivation (0.3'%1).The values of August 28 are those of the laboratory culture used as inoculum. The analyses were done at the end of the light period.

t,l

s

Table 2 Variation of fatty acid content and profile ( X , of dry weight) of Nunnochloropsis sp. cultures grown under natural light-dark cycle conditions in September 1996" Fatty acid

Day of the month

19 September ~

14:O 16:O 16:lw7 18:lw9 I8:2w6 20:406 20:503 TFA

20 September _____

21 September -~

22 September

23 September

~-

24 September

25 September _ _ ~ _ _

-~

L

D

L

D

L

D

L

D

L

D

L

D

L

D

0.71 3.71 4.20 0.97 0.36 0.87 3.48 14.5

0.63 2.94 3.82 0.94 0.44 0.88 3.82 13.6

0.63 3.45 3.82 0.76 0.32 0.89 3.85 13.9

0.57 3.02 3.67 0.77 0.42 0.81 4.09 13.5

0.59 2.98 3.73 0.80 0.34 0.85 4.05 13.5

0.57 2.50 3.53 0.70 0.41 0.79 4.09 12.7

0.58 2.78 3.52 0.64 0.33 0.78 4.02 12.6

0.47 2.24 3.13 0.61 0.39 0.80 4.32 12.1

0.49 2.72 3.29 0.66 0.32 0.83 4.20 12.7

0.47 2.35 3.15 0.61 0.41 0.81 4.54 12.5

0.56 3.43 3.43 0.60 0.31 0.86 4.20 13.6

0.50 2.80 3.29 0.47 0.38 0.74 4.35 12.8

0.51 2.79 3.30 0.54 0.32 0.83 4.31 I12.8

0.52 2.43 3.32 0.49 0.41 0.83 4.73 12.9

"Only the major fatty acids are shown (>0.3'%)).The analyses were done at the end of the light period (L) and at the end of the dark period ( D ) v) h,

W 0 vl

G. Chini Zittelli et al./Journal of Biotechnology 70 (1999) 299-312

306

significantly during the night (on average, 14:0 by 8.3%, 16:0 by more than 16%, 16:1037 by 5.4% and 18:1039 by 8%). A m o n g the P U F A s , 20:4036 did not change significantly, while 18:2036 and EPA increased by more than 24% and by 7%, respectively. This increase of both EPA and 18:2036 was relative, since it was due to the night consumption of storage material (carbohydrates, and saturated and m o n o u n s a t u r a t e d fatty acids) and not due to new synthesis (data not shown).

3.3. Prolonged outdoor cultivation of Nannochlorops& sp. & N H T R reactors under the climatic conditions of central Italy: biomass and EPA productivity Nannochloropsis sp. was grown outdoors in different types of N H T R systems uninterruptedly from March to September 1997 under the climatic conditions of central Italy (Florence). F r o m March to June and in September, 1 N H T R units were used; during the summer three-tube (3 N H T R s ) and eight-tube reactors (8 N H T R s ) were used. Table 3 shows the mean monthly productivities attained, together with the mean minimum and m a x i m u m culture temperatures, and the mean solar irradiance values recorded. The

highest volumetric productivities (0.7-0.8 g 1-1 d a y - 1 ) were obtained in late spring. The slightly lower productivities (0.5-0.6 g 1-1 d a y - l ) obtained in M a r c h - A p r i l might be due either to the lower solar irradiance available to the cultures (13.5 MJ m - 2 d a y - 1 in March and 15.3 MJ m - z d a y - ~ in April) compared to the M a y - J u n e period (17-19 MJ m - 2 d a y - 1) or to the excessively low temperatures during the night and the early morning (mean minimum temperatures of 2 - 3 ~ were recorded in the cultures during the spring). Since a mean monthly productivity of 0.73 g 1-1 d a y - ~ had been obtained in the previous September, with a mean minimum temperature of 10~ and at irradiance levels lower than 13 MJ m -2 d a y - 1 , it is likely that the lower performance of early spring cultures was due to the low night temperatures. Similar productivities obtained in May 1997 with a mean irradiance of 19 MJ m - 2 d a y - 1 , and in September 1996 with only 13 MJ m - 2 d a y - 1 , also indicate that solar irradiances higher than 13 MJ m -2 d a y - 1 do not significantly improve the productivity of Nannochloropsis sp. cultures under the conditions adopted. In the summer the productivity of cultures in the 3 N H T R and 8 N H T R units declined regularly to a minimum of 0.45 g 1-~ d a y - 1 . The

Table 3 Productivity of Nannochloropsis sp. cultures grown outdoors in NHTRs from March to September 1997 under the climatic conditions of central Italy (Florence)a Month

March April May June July

Reactor type

Mean volumetric productivity (g 1-1 day-l)

1 NHTR 0.51 + 0.36 1 NHTR 0.60 + 0.31 1 NHTR 0.76 + 0.39 1 NHTR 0.69 + 0.30 3 NHTR 0.56 + 0.30 3 NHTR 0.60 + 0.23 8 NHTR 0.56 + 0.26 3 NHTR 0.45 + 0.17 3 NHTR shaded 0.41 + 0.24 1 NHTR 0.57 + 0.20 w

August September

September 1996 1 NHTR

0.73 + 0.37

Culturetemperature (~ Min.

Max.

2.0 + 3.3 3.0 + 4.0 10.6 + 3.8 15.3 + 2.2 14.5 + 2.5 13.9 + 3.0 13.6 + 3.0 14.2 + 2.0 15.3 + 2.0 10.5 + 3.6

26.2 + 0.6 25.2 + 3.8 28.1 + 1.3 28.9 + 1.6 29.7 + 1.5 28.3 + 1.1 28.9 + 1.1 28.3 + 0.5 27.6 + 1.0 27.3 + 1.3

10.2 + 4.2

28.8 + 1.4

_

Solar irradiance (MJ m -2 day- 1)

13.5+ 4.1 15.3+ 6.1 18.9+ 4.4 17.4+ 5.7 21.0+ 2.5

m

17.6+ 3.3 15.9+ 3.2 12.7+ 5.1

The minimum and the maximum culture temperatures and the solar irradiance are also reported. Data of September 1996 is shown for comparison. Mean monthly values + S.D. are shown. a

307


Fig. 3. A cluster of Nannochloropsis sp. cells. The exuded polysaccharide matrix was stained with alcian-blue (Crayton, 1982). Bar: 10 lam. Photomicrograph by C. Sili.

reason for this is not yet clear. Signs of stress, such as abundant exudation of polysaccharide and presence of numerous and very large cell clusters (Fig. 3), possibly due to high light intensity or to prolonged exposure to higher than optimal diurnal temperatures, were observed in summer cultures. An experiment carried out to compare shaded versus unshaded cultures did not give a clear response. The shaded cultures showed a better appearance (lower degree of aggregation and higher pigment content), but attained a lower productivity. In September, the productivity of Nannochloropsis sp. rose again to about 0.6 g 1-l

d a y - 1 . The fact that the productivity values of the previous September were not reached despite similar climatic conditions, suggests that the cultures had not yet fully recovered from the stress of the summer. The mean fatty acid content of the biomass and the mean monthly productivity of EPA for the entire cultivation season ( M a r c h - S e p t e m b e r 1997) are shown in Table 4. The culture used as inoculum, which had been grown in the laboratory as previously described, had a total fatty acid content of 12% and an EPA content of 3.6%. As observed in the previous autumn, the transfer from laboratory to outdoor conditions caused a significant increase of the total fatty acid content with a maximum of 26% in mid-April. The increase mainly involved 16:0, 16" 1037 and 18:1039. EPA content, except for a peak of 4.6% in June, was rather stable around 4% throughout the cultivation period. Exposure to low temperatures during the night and in the early morning (Table 3), after growth at optimal temperature in the laboratory, can again be considered the main cause of the phenomenon. In May, when minimum temperatures rose above 10~ the fatty acid profile and content showed values similar to those observed in the laboratory. EPA productivity essentially reflected culture productivity. A maximum of 32 mg 1-~ d a y - 1 and a minimum of 18 mg 1-1 d a y - 1 were achieved in M a y - J u n e and August, respectively.

Table 4 Fatty acid composition (% of dry weight) and EPA productivity (mg 1-~ day-~) of Nannochloropsis sp. cultures grown outdoors in NHTRs from March to September 1997 under the climatic conditions of central Italy (Florence)a

Fatty acid 14:0 16:0 16:1co7 18:lco9 18:2co6 20:4co6 20:5co3 TFA EPA productivity

March

April

May

June

July

August

September

0.75 3.53 3.97 1.31 0.58 0.86 3.90 15.3 19.9

1.06 7.46 8.05 3.38 0.52 0.85 3.95 25.8 23.8

0.91 5.10 5.31 1.29 0.38 0.60 4.22 18.2 31.5

0.82 4.71 5.01 1.22 0.32 0.68 4.64 17.9 32.0

1.03 4.12 4.32 0.61 0.70 1.11 4.18 16.6 23.4

0.90 3.83 3.44 0.54 0.56 0.90 3.96 14.5 17.8

1.20 4.63 4.63 0.70 0.74 0.89 3.98 17.1 22.0

a Only the major fatty acids are shown (>0.3%).

308


In September and October 1998, Nannochloropsis sp. was grown outdoors in pilot 8 NHTR units (Fig. 1). The mean productivity was 0.4 g 1d a y - ~ and the EPA content of the biomass was 4%, with a mean EPA productivity of 16 mg 1d a y - 1. This experiment is still in progress. Contamination by protozoa or other microalgae was not a major problem in our NHTRs once efficient filtration systems for both the air and for the culture medium were adopted. Several bacterial species were observed in Nannochloropsis sp. cultures, but in very low concentrations. Moreover, the composition of the bacterial population associated with Nannochloropsis sp. cultures was rather stable and did not change significantly even after transfer from laboratory to outdoor conditions (Pastorelli et al., 1998). A noticeable exception occurred in the summer when, along with the abundant exudation of polysaccharidic material and massive formation of cell clusters, the bacterial load of the cultures increased by two orders of magnitude (data not shown).

3.4. Effect of salinity on productivity and EPA content of Nannochloropsis sp. cultures To evaluate the effect of a salinity value lower than that of seawater on culture productivity and fatty acid composition, Nannochloropsis sp. was grown outdoors in two 1 NHTR units at salt concentrations of 20 and 33 g 1-1. Neither the productivity (the mean monthly productivities were 0.69 __+0.4 at 20 g 1-1 and 0.71 + 0.4 g 1- 1 d a y - t at 33 g 1- ~ salinity), nor the total fatty acid content of the biomass (about 16.8% at both salt concentrations) was influenced by salinity in the range tested, although the fatty acid profile was modified (Table 5). At a salinity of 20 g 1-1 saturated fatty acids decreased by about 4% and monounsaturated fatty acids by about 23%, while PUFAs increased by more than 13%. EPA content rose from 4.3 to 4.8% of the dry biomass. Since the productivity of the culture was the same at both salinity values, EPA productivity was slightly higher at 20 g 1-1 (33.1 mg 1-1 day-1) than at 33 g 1-~ (30.5 mg 1- 1 day-1).

Table 5 Influence of salinity on fatty acid content and profile (% of dry weight) of Nannochloropsis sp. cultures grown outdoors in I NHTR units in May 1997a Fatty acid

14:0 16:0 16:1037 18:1o39 18:2036 20:4036 20:5033 TFA a

Salinity (g 1-l) 20

33

0.81 4.26 4.42 0.80 0.47 0.84 4.77 16.70

0.81 4.45 4.82 1.15 0.37 0.65 4.32 16.90

Only the major fatty acids are shown (>0.3%).

4. Discussion

Due to its high EPA content, Nannochloropsis has been proposed as a potential source of EPA for human consumption (Seto et al., 1984; Sukenik, 1991, 1998). Although in Japan Nannochloropsis is currently cultivated year-round in large outdoor tanks to provide a food chain for the production of fish larvae (Okauchi, 1991), data from outdoor mass cultivation of these microalgae is scanty. Boussiba et al. (1987) cultivated Nannochloropsis salina outdoors in 1- and 2.5-m 2 raceway-type ponds; prevention of contamination by diatoms and of predation by ciliates appeared to be the crucial factors for the successful operation of the cultures. Outdoor cultivation of Nannochloropsis sp. in raceway ponds with surface area ranging from a few square meters to 3000 m 2 was carried out at the National Center for Mariculture, Israel Oceanographic and Limnological Research, and at the Nature Beta Technology algae production plant, both located in Eilat, Israel (Sukenik et al. 1993a; Sukenik, 1998). The work done by Sukenik and co-workers at Eilat facilities has elucidated some fundamental aspects of the mass cultivation of this microalga and has confirmed the difficulty of maintaining monoalgal cultures outdoors for prolonged periods (Sukenik, 1998). The present work shows that photobioreactors have the potential to overcome the main limita-


tions encountered in open ponds and may allow long lasting outdoor cultivation of this eustigmatophyte achieving a relatively high EPA productivity compared with other autotrophic systems. In fact, contamination by bacteria, protozoa or other microalgae has not been a major problem in the N H T R units used in our experiments. Cultivation of microalgae in photobioreactors is hampered by some problems, including overheating, fouling and accumulation of oxygen to toxic levels (Tredici and Materassi, 1992). Temperatures of up to 5~ over the optimal value (25~ were measured in our Nannochloropsis cultures during the hours of strongest irradiation, particularly in the summer. Prolonged exposure to supraoptimal diurnal temperatures, together with high irradiances at midday, were considered the main cause of the low performance of the cultures in July and August. Fouling and accumulation of oxygen to toxic levels have not been a particularly serious problem in the N H T R design. It must be pointed out, however, that during the summer, when maximum productivities were expected, the cultures performed poorly. Hence, it was not possible to demonstrate the ability of NHTR systems in avoiding build-up of toxic oxygen tensions during periods of high production rates for Nannochloropsis cultures. EPA productivity by Nannochloropsis sp. grown in 16 mm thick alveolar panels under laboratory conditions averaged 36 mg 1-1 24 h-1. Outdoors, in the NHTRs, mean monthly EPA productivity varied between 18 and 32 mg l-1 day-1 essentially following the trend of biomass productivity which in turn was dependent upon the season. This data is comparable to that obtained with the diatom P. tricornutum (Molina Grima et al., 1994a) and the freshwater eustigrnatophyte M. subterraneus (Hu et al., 1997). These two microalgae are considered among the most promising EPA producers under autotrophic conditions and are presently under investigation as part of the EC Brite-Euram project 'An integrated production system of highly purified eicosapentaenoic acid from microalgae' coordinated by Professor Molina Grima. The major fatty acids in Nannochloropsis are 14:0, 16:0, 16:1037 and EPA, whereas C18 fatty

309

acids and 20:4o36 are present in lower quantities (Teshima et al., 1983; Sukenik et al., 1989; Volkman et al., 1993). Fractionation of lipid extracts from Nannochloropsis sp. indicated that EPA is mainly associated with galactolipids (the major lipid constituents of the thylakoid membranes) and phosphatidyl glycerol, whereas the storage lipid triacylglycerol contains primarily 14:0, 16:0 and 16:1o37 (Sukenik et al., 1989, 1993b). The effect of environmental factors on lipid content and fatty acid composition has been investigated in various species of Nannochloropsis with the majority of the work conducted under laboratory conditions (Teshima et al., 1983; Seto et al., 1984; Sukenik et al., 1989, 1993a,b; Sukenik and Carmeli, 1990; Renaud et al., 1991; Sukenik, 1991). These studies have shown that light intensity and temperature, as well as nitrogen starvation, play a key role in regulation of the relative abundance of fatty acids and of the absolute quantity of EPA. The effect of temperature on EPA biosynthesis appears to be complex and has not yet been clarified. Seto et al. (1992) found that cells grown at 20~ contained 60% more EPA than cells grown at the optimal growth temperature of 25~ Similarly, Sukenik et al. (1993b) showed that cultures grown at a low temperature (20~ were characterized by a high rate of galactolipid synthesis and hence by a relatively high level of EPA, as compared with cells grown at a high temperature (30~ In contrast with these authors, Teshima et al. (1983) found maximal EPA synthesis at the optimal growth temperature of 25~ Although the effect of temperature is not easy to establish in outdoor cultures, where both temperature and irradiance vary during the day, the present work would suggest that minimum night temperatures play a key role in influencing the fatty acid level and composition. The main outcome of our work, in partial disagreement with conclusions drawn from most other studies, is that even when the fatty acid profile is deeply modified in response to an environmental stimulus (in our case temperature), the total content of EPA in the biomass is not significantly affected. Sukenik et al. (1989, 1993b) studied the influence of light intensity on lipid content and com-

310

G. Chini Zittelli et al. , Journal of Biotechnology 70 (1999) 299-312

position in Nannochloropsis sp. grown in the laboratory. Under light-limiting growth conditions, the cells were characterized by a high density of thylakoids with most of the lipid carbon allocated in the EPA-rich galactolipids. In contrast, cells grown under saturating light conditions accumulated carbohydrate and triacylglycerols as storage materials, resulting in a high content of 16:0 and 16:1037 and a reduction in the cellular EPA content of more than 50%. Similarly, Renaud et al. (1991) showed increasing saturation of the fatty acids of N. oculata with increasing irradiance using large-scale outdoor cultures. In the same organism, Seto et al. (1992) found no effect of high irradiances on EPA distribution and content. In our experiments, light intensity seems to play a minor role in fatty acid composition with respect to temperature. However, high irradiances might have been involved in causing excessive exudation and cell aggregation in Nannochloropsis sp. cultures. Sukenik and Carmeli (1990) investigated the synthesis of fatty acids in Nannochloropsis sp. grown in the laboratory under a 12:12-h lightdark cycle. These authors showed that storage triacylglycerol lipids are synthesized and accumulated during the light period and are rapidly consumed for cellular maintenance along with carbohydrates during the dark period. Galactolipids are also synthesized during the light period, but not turned over in the dark. Thus, the dark period is characterized by an increase of the relative proportion of EPA which is associated with the galactolipids. Our experiments confirm that Nannochloropsis sp. also uses fatty acids associated with storage lipid (mainly 14:0 and 16:0) during the dark period when cultivated under natural light-dark cycle conditions. Thus, as suggested by Sukenik and Carmeli (1990), the early morning is the preferred time for harvesting Nannochloropsis sp. biomass to obtain a high EPA content. Teshima et al. (1983) found no significant variation of the EPA content of N. oculata biomass with salinities between 4 and 30 g 1-1. Similarly, Renaud and Parry (1994) found that EPA content did not vary significantly with salinity in the range of 10-35 g 1-~ although

salinities from 20 to 25 g 1-~ were optimal for growth and production of EPA. Our results show that lowering salinity from 33 to 20 g 1slightly stimulates EPA accumulation and productivity. The effect of seasonal variations in temperature and light availability on fatty acid distribution was studied in outdoor mass cultures of Nannochloropsis sp. by Sukenik et al. (1993a). The highest cellular EPA content (3.8%) was observed during winter with water temperatures between 8 and 16~ and under low irradiances, i.e. in conditions of very low biomass productivity. In summer, when biomass productivity was highest, EPA content decreased to less than half the winter value. Maximum EPA productivity of our cultures was attained in late spring and early autumn, i.e. under relatively low irradiances and when prolonged exposure at supraoptimal diurnal temperatures could be prevented. The primary goals of algal biotechnology for production of EPA are high EPA productivity and high EPA content of the biomass. Most previous studies conducted with Nannochloropsis have shown that maximum EPA content is obtained under low light conditions and at temperature values slightly lower than the optimal temperature for growth, and have concluded that the conditions required to maximize EPA productivity are different from those needed to maximize EPA content (Sukenik, 1991). Our work has led us to a different conclusion: since the main environmental factors, such as temperature and irradiance, do not significantly influence the EPA content of the biomass in our photobioreactors, conditions for maximum EPA productivity are those required to maximize biomass production. Moreover, for establishing a reliable system for EPA production based on cultivation of Nannochlorops& in photobioreactors, high irradiances are not ideal or can even be detrimental, as is prolonged exposure to supraoptimal temperatures that cause a serious imbalance of algal growth. The potential of Nannochloropsis cultivation in closed systems would therefore be best realized in temperate countries.

G. Chini Zittelli et al../Journal of Biotechnology 70 (1999) 299-312

Acknowledgements We are deeply indebted to Drs Susan Blackburn and Laura Barsanti for critically reviewing the manuscript. This work was partly supported by EXENIA S.r.1. (Milano, Italy).

References Barclay, W.R., Meager, K.M., Abril, J.R., 1994. Heterotrophic production of long chain omega-3 fatty acids utilizing algae and algae-like microorganisms. J. Appl. Phycol. 6, 123-129. Bousfield, I.J., Smith, G.L., Dando, T.R., Hobbs, G., 1983. Numerical analysis of total fatty acid profiles in the identification of coryneform, nocardioform and some other bacteria. J. Gen. Microbiol. 129, 375-394. Boussiba, S., Vonshak, A., Cohen, Z., Avissar, Y., Richmond, A., 1987. Lipid and biomass production by the halotolerant microalga Nannochloropsis salina. Biomass 12, 37-47. Chrismadha, T., Borowitzka, M.A., 1994. Effect of cell density and irradiance on growth, proximate composition and eicosapentaenoic acid production of Phaeodactylum tricornutum grown in a tubular photobioreactor. J. Appl. Phycol. 6, 67-74. Cohen, Z., Vonshak, A., Richmond, A., 1988. Effect of environmental conditions on fatty acid composition of the red alga Porphyridium cruentum: correlation to growth rate. J. Phycol. 24, 328-332. Crayton, M.A., 1982. A comparative cytochemical study of Volvocacean matrix polysaccharides. J. Phycol. 18, 336344. Gill, I., Valivety, R., 1997. Polyunsaturated fatty acids, part 1: occurrence, biological activities and applications. TIBTECH 15, 401-409. Gonen-Zurgil, Y., Carmeli-Schwartz, Y., Sukenik, A., 1996. Selective effect of the herbicide DCMU on unicellular algae--a potential tool to maintain monoalgal mass culture of Nannochloropsis. J. Appl. Phycol. 8, 415-419. Guillard, R.R.L., Ryther, J.H., 1962. Studies of marine planktonic diatoms. I. Cyclotella nana (Hustedt) and Detonula confervacea (Cleve). Can. J. Microbiol. 8, 229-239. Hu, Q., Hu, Z., Cohen, Z., Richmond, A., 1997. Enhancement of eicosapentaenoic acid (EPA) and ~,-linolenic acid (GLA) production by manipulating algal density of outdoor cultures of Monodus subterraneus (Eustigmatophyta) and Spirulina platensis (Cyanobacteria). Eur. J. Phycol. 32, 81-86. Leman, J., 1997. Oleaginous microorganisms: an assessment of the potential. Adv. Appl. Microbiol. 43, 195-243. Molina Grima, E., Sanchez Perez, J.A., Garcia Camacho, F., Acien Fernandez, F.G., Fernadez Sevilla, J.M., Valdez Sanz, F., 1994a. Effect of dilution rate on eicosapentaenoic acid productivity of Phaeodactylum tricornutum UTEX 640

311

in outdoor chemostat culture. Biotechnol. Lett. 16, 10351040. Molina Grima, E., Sanchez Perez, J.A., Garcia Camacho, F., Fernadez Sevilla, J.M., Acien Fernandez, F.G., 1994b. Effect of growth rate on the eicosapentaenoic acid and docosahexaenoic acid content of Isochrysis galbana in chemostat culture. Appl. Microbiol. Biotechnol. 41, 23-27. Nordoy, A., 1991. Is there a rational use for n-3 fatty acids (fish oils) in clinical medicine.'? Drugs 42, 331-342. Okauchi, M., 1991. The status of phytoplankton production in Japan. In: Fulks, W., Main, K.L. (Eds.), Rotifers and Microalgae Culture Systems. Proceedings of a US-Asia Workshop. The Oceanic Institute, Honolulu, HI, pp. 247256. Pastorelli, R., Chini Zittelli, G., Tredici, M.R., 1998. Bacterial population associated with Nannochloropsis sp. mass cultures. Abstracts of the International Symposium 'Marine Bioprocess Engineering'. Noordwijkerhout, The Netherlands. November 8-11, 1998. Radwan, S.S., 1991. Sources of Cs0-polyunsaturated fatty acids for biotechnological use. Appl. Microbiol. Biotechnol. 35, 421-430. Renaud. S.M., Parry, D.L., 1994. Microalgae for use in tropical aquaculture II: Effect of salinity on growth, gross chemical composition and fatty acid composition of three species of marine microalgae. J. Appl. Phycol. 6, 347-356. Renaud. S.M., Parry, D.L., Luong-Van, T., Kuo, C., Padovan. A., Sammy. N., 1991. Effect of light intensity on the proximate biochemical and fatty acid composition of lsochrvsis sp. and Nannochloropsis oculata for use in tropical aquaculture. J. Appl. Phycol. 3, 43-53. Senzaki, H., Iwamoto, S., Ogura, E., Kiyozuka, Y., Arita, S., Kurebayashi, J., Takada, H., Hioki, K., Tsubura, A., 1998. Dietary effects of fatty acids on growth and metastasis of KPL-1 human breast cancer cells in vivo and in vitro. Anticancer Res. 18, 1621- 1627. Seto, A., Wang, H.L., Hesseltine, C.W., 1984. Culture conditions affect eicosapentaenoic acid content of Chlorella minutisshna. J. Am. Oil. Chem. Soc. 61, 892-894. Seto, A., Kumasaka, K.. Hosaka, M., Kojima, E., Kashiwakura, M., Kato, T.. 1992. Production of eicosapentaenoic acid by marine microalgae and its commercial utilization for aquaculture. In: Kyle, D.J., Ratledge, C. (Eds.), Industrial Application of Single Cell Oils. American Oil Chemists' Society, Champaign, IL, pp. 219-234. Sukenik, A., 1991. Ecophysiological considerations in the optimization of eicosapentaenoic acid production by Nannochloropsis sp. (Eustigmatophyceae). Bioresour. Technol. 35, 263-269. Sukenik, A., Carmeli, Y., 1990. Lipid synthesis and fatty acid composition in Nannochloropsis sp. (Eustigrnatophyceae) grown in a light-dark cycle. J. Phycol. 26, 463-469. Sukenik, A., Carmeli, Y., Berner, T., 1989. Regulation of fatty acid composition by irradiance level in the eustigrnatophyte Nannochloropsis sp. J. Phycol. 25, 686-692. Sukenik, A., Zmora, O., Carmeli, Y., 1993a. Biochemical quality of marine unicellular algae with special emphasis

312


on lipid composition. II. Nannochloropsis sp. Aquaculture 117, 313-326. Sukenik, A., Yamaguchi, Y., Livne, A., 1993b. Alteratiofis in lipid molecular species of the marine eustigrnatophyte Nannochloropsis sp. J. Phycol. 29, 620-626. Sukenik, A., 1998. Production of eicosapentaenoic acid by the marine eustigrnatophyte Nannochloropsis sp. In: Cohen, Z. (Ed.), Chemicals from Microalgae. Taylor and Francis, London, in press. Teshima, S., Yamasaki, S., Kanazawa, A., Hirata, H., 1983. Effects of water temperature and salinity on eicosapenatenoic acid level of marine Chlorella. Bull. Jpn. Soc. Sci. Fisheries 49, 805. Tredici, M.R., Chini Zittelli, G., 1998. Efficiency of sunlight utilization: tubular versus fiat photobioreactors. Biotechnol. Bioeng. 57, 187-197.

Tredici, M.R., Materassi, R., 1992. From open ponds to vertical alveolar panels: the Italian experience in the development of reactors for the mass cultivation of phototrophic microorganisms. Appl. Phycol. 4, 221-231. Tredici, M.R., Carlozzi, P., Chini Zittelli, G., Materassi, R., 1991. A vertical alveolar panel (VAP) for outdoor mass cultivation of microalgae and cyanobacteria. Bioresour. Technol. 38, 153-159. Volkman, J.K., Malcom, R.B., Dunstan, G.A., Jeffrey, S.W., 1993. The biochemical composition of marine microalgae from the class eustigmatophiceae. J. Phycol. 29, 69-78. Yongrnanitchai, W., Ward, O.P., 1989. Omega-3 fatty acids: alternative sources of production. Process Biochem. 24, 117-125.

Biotechno,logy ELSEVIER


Commercial production of microalgae" ponds, tanks, tubes and fermenters M i c h a e l A. B o r o w i t z k a * Algae Research Group, School of Biological Sciences and Biotechnology, Murdoch University, Murdoch, WA 6150, Australia


Abstract The commercial culture of microalgae is now over 30 years old with the main microalgal species grown being Chlorella and Spirulina for health food, Dunaliella salina for 13-carotene, Haematococcus pluvialis for astaxanthin and several species for aquaculture. The culture systems currently used to grow these algae are generally fairly unsophisticated. For example, Dunaliella salina is cultured in large (up to approx. 250 ha) shallow open-air ponds with no artificial mixing. Similarly, Chlorella and Spirulina also are grown outdoors in either paddle-wheel mixed ponds or circular ponds with a rotating mixing arm of up to about 1 ha in area per pond. The production of microalgae for aquaculture is generally on a much smaller scale, and in many cases is carried out indoors in 20-40 1 carboys or in large plastic bags of up to approximately 1000 1 in volume. More recently, a helical tubular photobioreactor system, the BIOCOIL TM, has been developed which allows these algae to be grown reliably outdoors at high cell densities in semi-continuous culture. Other closed photobioreactors such as fiat panels are also being developed. The main problem facing the commercialisation of new microalgae and microalgal products is the need for closed culture systems and the fact that these are very capital intensive. The high cost of microalgal culture systems relates to the need for light and the relatively slow growth rate of the algae. Although this problem has been avoided in some instances by growing the algae heterotrophically, not all algae or algal products can be produced this way. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Photobioreactors; Chlorella; Spirulina; Dunaliella

1. Introduction Microalgal culture is one of the modern biotechnologies. The first unialgal cultures were achieved by Beijerinck (1890), with Chlorella vul* Fax: + 61-8-93606303. E-mail address: [email protected] (M.A. Borowitzka)

garis, and the use of such cultures for studying plant physiology was developed by W a r b u r g in the early 1900s (Warburg, 1919). Mass culture of microalgae really began to be a focus of research after 1948 at Stanford (USA), Essen (Germany) and Tokyo and the classic book edited by Burlew (1953) summarises many of these early studies. Interest in applied algal culture continued, especially with studies on the use of algae as photo-


M.A. Borowitzka /Journal of Biotechnology 70 (1999) 313-321

314

synthetic gas exchangers for space travel and as microbial protein sources. The results of these studies can be found in several papers and books (Shelef and Soeder, 1980; Venkataraman and Becker, 1985; Wharton et al., 1988). Commercial large-scale culture of microalgae commenced in the early 1960s in Japan with the culture of Chlorella (Tsukada et al., 1977) followed in the early 1970s with the establishment of a Spirulina harvesting and culturing facility in Lake Texcoco, Mexico by Sosa Texcoco S.A. (Durand-Chastel, 1980). In 1977 Dai Nippon Ink and Chemicals Inc. established a commercial Spirulina plant in Thailand, and by 1980 there were 46 large-scale factories in Asia producing more than 1000 kg of microalgae (mainly Chlorella) per month (Kawaguchi, 1980) and in 1996 about 2000 t of Chlorella were traded in Japan alone (Lee, 1997). Commercial production of Dunaliella salina as a source of 13-carotene became the third major microalgae industry when production facilities were established by Western Biotechnology Ltd and Betatene Ltd in Australia in 1986. These were soon followed by other commercial plants in Israel and the USA. As well as these algae, the large-scale production of cyanobacteria (blue-green algae) commenced in India at about the same time (Venkatamaran, 1986). Thus in a short period of about 30 years the industry of microalgal biotechnology has grown and diversified significantly. The growth of commercial microalgae production is probably best illustrated by data on Spirulina production for which reasonably reliable figures are available (Fig. 1). The success of commercial large-scale production of microalgae depends on many factors, and one of these is the development of cost effective large-scale culture systems for the algae and the development of these has been, and continues to be, a gradual process. This paper will review some of the steps along the way and the future advances we can expect.

range in volume from about 102 1 (used for the production of organic compounds labelled with stable isotopes) to > 101~1 (used for the culture of D. salina). However, aside from the specialised small-scale ( < 1000 1) culture systems four types of culture systems predominate: large open ponds, circular ponds with a rotating arm to mix the cultures, raceway ponds and large bags. Other commercial large-scale systems include tanks used in aquaculture, the cascade system developed in Trebon, Czech Republic (Setlik et al., 1970) and heterotrophic fermenter systems used for the culture of Chlorella in Japan and Taiwan (Kawaguchi, 1980; Soong, 1980) and for the culture of Crypthecodinium cohnii in the USA (Kyle and Gladue, 1991; Kyle et al., 1998). Table 1 summarises the culture systems currently in use for commercial algal culture.

3000 Cuba Chile Myanmar

2500 -

[~ [~

raiwan Japan

2000 i

c c 0

c

0

India

Viemam China USA Thailand

Mexico

1500 -

0

lO0O

-

i

500

-

o

/

1970

1975

,l!iIill 1980

1985

1990

1995

2000

Year

2. Culture systems

Existing commercial microalgae culture systems

Fig. 1. Global production figures of Spirulina by country based on literature, company and trade information. Figures for 1997 and 1998 are estimates based on projected production figures provided by the producers.


315

Table 1 Commercial microalgae culture systems currently in use and the algal species cultured Culture system

Algae

Tanks Extensive open ponds Circular ponds with rotating arm Raceway ponds

Many species (for aquaculture)

Cascade system with baffles Large bags

Chlorella spp. Many species (used for aquaculture) Chlorella spp., Crypthecodinium

Fermenters (heterotrophic) Two-stage system (indoors in closed reactor and then outdoors in paddlewheel ponds) a

Dunaliella salina Chlorella spp. Chlorella spp., Spirulina spp. Dunaliella salina

Approximate maximum volume (1)a 1• 1• 1.5 • 3•

104 109 104 104

3 x 104 1 x 103 > 103

cohnii Haematococcus pluvialis

Location

World Wide Australia Taiwan, Japan Japan, Taiwan, USA, Thailand, China, India, Vietnam, Chile, USA, Israel, China Czech Republic, Bulgaria World wide Japan, Taiwan, Indonesia, USA USA

These are order of magnitude estimates only.

There are several considerations as to which culture system to use. Factors to be considered include: the biology of the alga, the cost of land, labour, energy, water, nutrients, climate (if the culture is outdoors) and the type of final product (Borowitzka, 1992). The various large-scale culture systems also need to be compared on their basic properties such as their light utilisation efficiency, ability to control temperature, the hydrodynamic stress placed on the algae, the ability to maintain the culture unialgal and/or axenic and how easy they are to scale up from laboratory scale to large-scale. These properties are compared in Table 2. The final choice of system is almost always a compromise between all of these considerations to achieve an economically acceptable outcome. A common feature of most of the algal species currently produced commercially (i.e. Chlorella, Spirulina and Dunaliella) is that they grow in highly selective environments which means that they can be grown in open air cultures and still remain relatively free of contamination by other algae and protozoa. Thus, Chlorella grows well in nutrient-rich media, Spirulina requires a high pH and bicarbonate concentration and D. salina grows at very high salinity (Soong, 1980; Borowitzka and Borowitzka, 1988; Belay, 1997).

Those species of algae which do not have this selective advantage must be grown in closed systems. This includes most of the marine algae grown as aquaculture feeds (e.g. Skeletonema, Chaetoceros, Thalassiosira, Tetraselmis and Isochrysis) and the dinoflagellate C. cohnii grown as a source of long-chain polyunsaturated fatty acids.

3. Open-air systems Although much of the early work on microalgal mass culture focused on closed culture systems (see Burlew, 1953) all very large commercial systems used today are open-air systems (Table 1). The reason for this is simple economics; closed culture systems are very expensive and many of them are difficult to scale up. Furthermore, most closed systems are operated indoors with artificial lighting and this results in high energy costs whereas open air systems can utilise sunlight. However, only a small number of algal species can be grown successfully in open air systems. The four major types of open-air systems currently in use (shallow big ponds, tanks, circular ponds and raceway ponds) all have their advantages and disadvantages. The selection of a partic-

Table 2 Comparison of the properties of different large-scale algal culture systems Reactor type

Mixing

Light utilisa- Temperature tion efficiency control

Gas transfer

Hydrodynamic Species constress on algae trol

Sterility

Scale-up

Reference

Unstirred shallow ponds

Very poor

Poor

None

Poor

Very low

Difficult

None

Very difficult

Tanks Circular stirred ponds Paddle-wheel Raceway Ponds Stirred Tank reactor (internal or external lighting) Air-Lift reactor Bag Culture

Poor Fair

Very poor Fair-good

None None

Poor Poor

Very low Low

Difficult Difficult

None None

Very difficult Very difficult

Fair-good

Fair-good

None

Poor

Low

Difficult

None

Very difficult

Largely uniform

Fair-good

Excellent

Low-high

High

Easy

Easily achievable

Difficult

Borowitzka and Borowitzka, 1989 Fox, 1983 Tamiya, 1957; Stengel, 1970; Soeder, 1981 Weissman and Goebel, 1987; Oswald, 1988 Pohl et al., 1988

Generally uni- Good form Variable Fair-good

Excellent

High

Low

Easy

Difficult

Juttner. 1977

Low-high

Low

Easy

Difficult

High

Low high

Easy

Easily achievable Easily achievable Achievable

Baynes et al., 1979 Hu et al., 1996; Tredici and Zitelli, 1997 Richmond et al., 1993; Torzillo, 1997 Borowitzka, 1996

Uniform

Excellent

Good (indoors) Excellent

Tubular reac- Uniform tor (Serpentine type) Tubular Reac- Uniform tor (Biocoil tYF4

Excellent

Excellent

Low-high

Low-high

Easy

Achievable

Reasonable

Excellent

Excellent

Low-high

Low-high

Easy

Achievable

Easy

Flat-Plate reactor

Difficult


ular system is also influenced by intrinsic properties of the alga as well as local climatic conditions and the costs of land and water. For example, the Australian producer of D. salina, Betatene Ltd, uses very large ponds of up to 250 ha in area at their plants at Whyalla and Hutt Lagoon, which are unmixed other than by wind and convection. This is possible since land costs are low, water (seawater) is free other than for pumping costs, and the climate is close to optimum so that production can be achieved all year round. Furthermore, the company has a very efficient harvesting system (Borowitzka, 1991; Schlipalius, 1991). On the other hand, the other Dunaliella producers in the USA (no longer in operation) and in Israel use paddle-wheel driven raceway ponds to achieve higher cell densities (Ben-Amotz, 1995). This is because the cost of land is relatively high at both of these sites and site preparation costs are also significant. As well as this they require the addition of NaC1 to the medium which is a significant additional cost (the cultures operate at between 15 and 25% NaC1). These factors mean that pond area and volume must be minimised and the cell density in the culture must be maximised in order to have an economical process. The culture of Spirulina and Chlorella also requires a well mixed system such as a raceway system to achieve high growth rates and to minimise the risk of overgrowth by other algae. The only successful extensive culture system for these algae was the 'Caracol' used by Sosa Texcoco in Mexico for the culture of Spirulina maxima. The Caracol was a giant spiral-shaped solar evaporator of 3200 m diameter with a surface area of 900 ha within Lake Texcoco where Spirulina grows naturally. This system had to be abandoned eventually due to the increased contamination resulting from urban and industrial development in Mexico City. The greater susceptibility to contamination by other algae and/or protozoa also means that Spirulina and Chlorella must be grown in batch or semi-batch mode with periodical reseeding of ponds with new inoculum (Kawaguchi, 1980; Belay, 1997). Furthermore, many of these algal plants are located in regions with climatic conditions which do not allow year round produc-

317

tion. This means that productivities have to be maximised to offset the high capital costs of the ponds and associated production systems such as harvesters and dryers. The problems with open air pond systems are that the productivity achieved is less than that theoretically possible and that it is difficult to control the culture environment which is exposed to the elements. The pond depth is a compromise between the need to provide adequate light to the algal cells (i.e. the shallower the more light is available to the cells) and the need to maintain an adequate water depth for mixing and to avoid large changes in ionic composition due to evaporation. Thus most paddle-wheel raceway ponds are between 20 and 30 cm deep and the very large (up to approx. 250 ha) unmixed ponds used in the culture of Dunaliella in Australia the depth may be up to 50 cm deep. This means that, in almost all cases, the algae are light limited and the maximum biomass achieved on a regular basis is between approx. 0.1 and 0.5 g dry weight I-~. The relationship between pond depth, culture density and productivity has been most extensively studied for Spirulina (cf. Vonshak, 1997) and is important for maximising the biomass output. The algae are also CO2 limited, however the addition of CO2 to these large ponds is usually inefficient and uneconomical except in the case of Spirulina culture where it is essential to maintain high alkalinity (Belay, 1997). The only open-air system which achieves significantly higher sustainable cell densities is the cascade system developed by Setlik et al. (1970) and which is still in use in Trebon, Czech Republic for the culture of Chlorella. In this system the culture depth is less than 1 cm and cell densities up to about 10 g 1-~ can be maintained while maintaining high growth rates (Doucha and Livansky, 1995). The Trebon system is very expensive with the base of the sloping culture surface made of glass. However, improved materials mean that a similar system could now be constructed at significantly lower cost (Doucha, personal communication). Furthermore the central European location of Trebon means that there is a relatively short annual culture period. Preliminary calculations indicate that this system would be very competi-

318


tive with paddle-wheel raceway systems if located in a sunnier, warmer climate and if constructed of lighter and cheaper materials. A similar system comprising a 0.5 ha sloping, plastic lined pond was used to produce Chlorella near Dongara, Western Australia, for several years. This system had an average productivity of about 25 g m - 2 day-1 and could be operated in semi-continuous mode for the whole year due to the near optimal Climate. Unfortunately technical problems with further scale-up eventually led to the closure of this plant.

4. Closed systems Despite the success of open systems, future advances in microalgal mass culture will require closed systems as the algal species on interest do not grow in highly selective environments. Furthermore, many of the new algae and algal products must be grown free of potential contaminants such as heavy metals and micro-organisms. The concept of closed systems has been around for a long time (Little, 1953; Jfittner, 1977; Pirt et al., 1983) however their high cost has largely precluded their commercial application until recently. Algae can be grown in closed systems either photoautotrophically, mixotrophically or heterotrophically. Heterotrophic cultivation on acetate or glucose as carbon sources has been used for some time for Chlorella (Kawaguchi, 1980; Soong, 1980) with approximately 550 t produced in Japan in 1996 (Lee, 1997). For a short period Tetraselmis was also grown heterotrophically in the UK (Gladue, 1991; Laing and Verdugo, 1991), however the production costs for the latter proved uneconomical and the spray dried product was inferior to photoautotrophically grown algae. More recently, heterotrophic culture is being used for the dinoflagellate, C. cohnii by Martek Inc. in the USA to produce long-chain polyunsaturated fatty acids (Kyle et al., 1992, 1998). Heterotrophic culture has several advantages: (i) fermentation systems are well understood and there is wide experience in their design and operation; and (ii) high cell densities of between 20 and 100 g 1-1 can be

achieved reducing harvesting costs and the capital costs of the cultivation vessels (Radmer and Parker, 1994; Running et al., 1994). However, there are disadvantages as well. The main disadvantages are that heterotrophic cultivation is not possible for all microalgae and that the chemical composition of the algae often changes under heterotrophic conditions. Closed photoautotrophic culture systems are widely used in the aquaculture industry for the production of a range of algal species. The most widely used large-scale (up to about 1000 1) system is the 'big bag' system (Baynes et al., 1979; Watson, 1979). These systems use large sterile plastic bags of about 0.5 m diameter fitted with a system for aeration. Although most of these systems are operated in batch mode, semi-continuous systems have also been developed. A variant on this system has been developed by Cohen and Arad (1989) using multiples of narrower bags; however there appears to be no commercial production as yet using this system. The main problem with the big bag system is that the system has to be operated indoors for temperature control and that the relatively large diameter of the bags and the need for artificial lighting results in light-limited cultures. The operation of these systems is also labour intensive and the cultures are generally inadequately mixed, leading to recurrent culture 'crashes'. This makes the algae expensive to produce. Costs are about $US 50 kg-~ of dry algal biomass (Fulks and Main. 1991) for a large, specialised oyster hatchery in the USA with many smaller hatcheries having costs of up to $US 300-600 kg-1 (Coutteau and Sorgeloos, 1992; Borowitzka, 1997). This is much greater that the estimated production costs for Chlorella, Spirulina and Dunaliella of between $US 9 and 25 kg -1 (Borowitzka, 1992; Tanticharoen et al., 1993). In recent years there have been several major advances in the design and operation of closed photobioreactors for algal culture and several systems are likely to be commercial realities in the near future. The two basic designs are the flat plate reactors (Pulz, 1994; Hu et al., 1996; Tredici and Zitelli, 1997) and the tubular photobioreactors (Miyamoto et al., 1988; Richmond et al.,


1993; Borowitzka, 1996; Torzillo, 1997). The fundamental principle in all of these designs is to reduce the light path and thus to increase the amount of light available to each cell. These reactors are also well mixed to ensure optimum light availability to the cells and to enhance gas exchange. The optimum thickness of the algal culture in these reactors is between 2 and 4 cm. Pilot scale units of up to about 1000 1 have been operated successfully for Spirulina, Chlorella and several marine microalgae, and at least one pilotscale system is in use in Potsdam, Germany for the production of Spirulina and another 6000 1 unit in Elbingerode, Germany is growing Chlorella using waste CO2 from a lime factory (Menz, 1998). These closed reactors have several advantages such as 'clean' algal culture, high light utilisation efficiency leading to high productivities as high sustainable biomass, temperature control and the ability to be used outdoors in natural daylight. This means that a much wider range of species can be cultivated as contamination is avoided and the reactors can be operated over a much wider climatic range that the open-air systems. There is also a greater ability to control the culture conditions so that the final product is of more consistent composition and quality (e.g. Chrismadha and Borowitzka, 1994). Finally, these systems are amenable to operation in continuous culture mode. Continuous culture and good control over the growth environment results in a consistent product quality and the higher operating cell densities also mean that harvesting costs are reduced and there is a smaller requirement for land. The challenge now is to reduce the construction costs of these systems further to make them more economically competitive. The BIOCOIL TM appears the most promising design at present (Robinson et al., 1988; Robinson and Morrison, 1992). The BIOCOIL is a helical tubular photobioreactor consisting of a photostage of small diameter clear plastic tubing (between 2.4 and 5 cm diameter) would helically around a tower (Borowitzka and Borowitzka, 1989). Several parallel bands of tubes are connected via manifolds to a pumping system which may be an airlift or pumps such as centrifugal,

319

diaphragm or lobe pumps, the type of pump used depends on the species of alga grown. If required a gas exchange tower may also be incorporated in the circuit. Temperature control is either by a heat exchanger or by evaporative cooling by flowing water over the photostage surface. We have grown a wide range of marine microalgae including Tetraselmis spp., Isochrysis galbana, Phaeodactylum tricornutum and Chaetoceros spp. as well as Spirulina at high productivities in pilot scale BIOCOILs of up to 700 1 for periods greater than 4 months in semi-continuous culture. The design of the BIOCOIL ensures uniform mixing and minimises adhesion of the algal cells to the inside of the tubes. The BIOCOIL can also be scaled up easily and the whole culture process can be automated thus minimising labour costs and improving reliability. Depending on need, the system can be designed to be operated as an axenic culture. Not all algal species are suitable for culture in this system. Some especially delicate species are damaged by the circulation system and other species may be too 'sticky' for normal operation. However, systems such as the BIOCOIL greatly expand the suite of algae which can be grown on a large scale and provide the opportunity to develop new algae and algal products such as bioactive molecules (Borowitzka, 1995). They will also serve to reduce the costs associated with the production of algae such as those species used in aquaculture.

5. Conclusions

Commercial microalgal culture is a well established industry. Most of the culture systems in use today are open air and relatively unsophisticated. However, over the last 50 years great advances have been made in our understanding of the biology of the algae and in the engineering requirements of large-scale algae culture systems. This has led to the development of several types of closed photobioreactors which will enable the commercialisation of new algae and algal products in the next decade.

320


References Arthur D. Little Inc., 1953. Pilot-plant studies in the production of Chlorella. In: Burlew, J.S. (Ed.), Algal Culture. From Laboratory to Pilot Plant. Carnegie Institution, Washington, DC, pp. 235-272. Baynes, S.M., Emerson, L., Scott, A.P., 1979. Production of algae for use in the rearing of larval fish. Fisheries Research Technical Report 53, 13-18. Beijerinck, M.W., 1890. Kulturversuche mit Zoochloren, Lichenengonidien und anderen niederen Algen. Bot. Ztg. 48, 725-785. Belay, A., 1997. Mass culture of Spirulina outdoors--The Earthrise Farms experience. In: Vonshak, A. (Ed.), Spirulina platensis (Arthrospira): Physiology, cell-biology and biotechnology. Taylor and Francis, London, pp. 131-158. Ben-Amotz, A., 1995. New mode of Dunaliella biotechnology: two-phase growth for 13-carotene production. J. Appl. Phycol. 7, 65-68. Borowitzka, L.J., 1991. Development of Western biotechnology algal beta-carotene plant. Bioresource Technol. 38, 251-252. Borowitzka, L.J., Borowitzka, M.A., 1989. Industrial production: methods and economics. In: Cresswell, R.C., Rees, T.A.V., Shah, N. (Eds.), Algal and Cyanobacterial Biotechnology. Longrnan Scientific, London, pp. 294-316. Borowitzka, M.A., 1992. Algal biotechnology products and processes: matching science and economics. J. Appl. Phycol. 4, 267-279. Borowitzka, M.A., 1995. Microalgae as sources of pharmaceuticals and other biologically active compounds. J. Appl. Phycol. 7, 3-15. Borowitzka, M.A., 1996. Closed algal photobioreactors: design considerations for large-scale systems. J. Mar. Biotechnol. 4, 185-191. Borowitzka, M.A., 1997. Algae for aquaculture: Opportunities and constraints. J. Appl. Phycol. 9, 393-401. Borowitzka, M.A., Borowitzka, L.J., 1988. Dunaliella. In: Borowitzka, M.A., Borowitzka, L.J. (Eds.), Microalgal Biotechnology. Cambridge University Press, Cambridge, pp. 27- 58. Burlew, J.S. (Ed.), 1953. Algae Culture. From Laboratory to Pilot Plant. Carnegie Institution of Washington, Washington, DC. Chrismadha, T., Borowitzka, M.A., 1994. Effect of cell density and irradiance on growth, proximate composition and eicosapentaenoic acid production of Phaeodactylum tricornutum grown in a tubular photobioreactor. J. Appl. Phycol. 6, 67-74. Cohen, E., Arad, S., 1989. A closed system for outdoor cultivation of Porphyridium. Biomass 18, 59-67. Coutteau, P., Sorgeloos, P., 1992. The use of algal substitutes and the requirement for live algae in the hatchery and nursery of bivalve molluscs: an international survey. J. Shellfish. Res. 11,467-476. Doucha, J., Livansky, K., 1995. Novel outdoor thin-layer high density microalgal culture system: Productivity and operational parameters. Algol. Stud. 76, 129-147.

Durand-Chastel, H., 1980. Production and use of Spirulina in Mexico. In: Shelef, G., Soeder, C.J. (Eds.), Algae Biomass. Elsevier/North Holland Biomedical Press, Amsterdam, pp. 51-64. Fox, J.M., 1983. Intensive algal culture techniques. In: McVey, J.P. (Ed.), CRC Handbook of Mariculture Crustacean Aquaculture. CRC Press, Boca Raton, FL, pp. 15-41. Fulks, W., Main, K.L., 1991. The design and operation of commercial-scale live feeds production systems. In: Fulks, W., Main, K.L. (Eds.), Rotifer and Microalgae Culture Systems. The Oceanic Institute, Honolulu, HI, pp. 3-52. Gladue, R., 1991. Heterotrophic microalgae production: potential for application to aquaculture feeds. In: Fulks, W., Main, K.L. (Eds.), Rotifer and Microalgae Culture Systems. The Oceanic Institute, Honolulu, HI, pp. 275-286. Hu, Q., Guterman, H., Richmond, A., 1996. A flat inclined modular photobioreactor for outdoor mass cultivation of photoautotrophs. Biotechnol. Bioeng. 51, 51-60. Jiittner, F., 1977. Thirty liter tower-type pilot plant for the mass cultivation of light- and motion-sensitive planktonic algae. Biotechnol. Bioeng. 19, 1679-1687. Kawaguchi, K., 1980. Microalgae production systems in Asia. In: Shelef, G., Soeder, C.J. (Eds.), Algae Biomass Production and Use. Elsevier/North Holland Biomedical Press, Amsterdam, pp. 25- 33. Kyle, D.J., Gladue, R.M., 1991. Eicosapentaenoic acids and methods for their production. World Patent 9,114,427. Kyle, D.J., Boswell, K.D.B., Gladue, R.M., Reeb, S.E., 1992. Designer oils from microalgae as nutritional supplements. In: Bills, D.D., Kung, S.D. (Eds.), Biotechnology and Nutrition. Butterworth-Heinemann, Boston, pp. 451-468. Kyle, D.J., Reeb, S.E., Sicotte, V.J., 1998. Dinoflagellate biomass, methods for its production, and compositions containing the same. USA Patent 5,711,983. Laing, I., Verdugo, C.G., 1991. Nutritional value of spraydried Tetraselmis suecica for juvenile bivalves. Aquaculture 92, 207-218. Lee, Y.K., 1997. Commercial production of microalgae in the Asia-Pacific rim. J. Appl. Phycol. 9, 403-411. Menz, K., 1998. Biotechnologische Nutzung von Kohlendioxid als Rohstoff mit Hilfe yon Mikroalgen. Forschung Technik und Innovation (Preussag) 23, 85-89. Miyamoto, K., Wable, O., Benemann, J.R., 1988. Vertical tubular reactor for microalgae cultivation. Biotech. Lett. 10, 703- 708. Oswald, W.J., 1988. Large-scale algal culture systems (engineering aspects). In: Borowitzka, M.A., Borowitzka, L.J. (Eds.), Microalgal Biotechnology. Cambridge University Press, Cambridge, pp. 357-394. Pirt, S.J., Lee, Y.K., Walach, M.R., Pirt, M.W., Balyuzi, H.H.M., Bazin, M.J., 1983. A tubular bioreactor for photosynthetic production of biomass from carbon dioxide: design and performance. J. Chem. Tech. Biotechnol. 33B, 35-58. Pohl, P., Kohlhase, M., Martin, M., 1988. Photobioreactors for the axenic mass cultivation of microalgae. In: Stadler, T., Mollion, J., Verdus, M.C., Karamanos, Y., Morvan,

M.A. Borowitzka /Journal of Biotechnology 70 (1999) 313-321 H., Christiaen, D. (Eds.), Algal Biotechnology. Elsevier, London, pp. 209-217. Pulz, O., 1994. Open-air and semi-closed cultivation systems for the mass cultivation of microalgae. In: Phang, S.M., Lee, K., Borowitzka, M.A., Whitton, B. (Eds.), Algal Biotechnology in the Asia-Pacific Region. Institute of Advanced Studies, University of Malaya, Kuala Lumpur, pp. 113-117. Radmer, R.J., Parker, B.C., 1994. Commercial applications of algae--opportunities and constraints. J. Appl. Phycol. 6, 93-98. Richmond, A., Boussiba, S., Vonshak, A., Kopel, R., 1993. A new tubular reactor for mass production of microalgae outdoors. J. Appl. Phycol. 5, 327-332. Robinson, L.F., Morrison, A.W., 1992. Biomass production apparatus. USA Patent 5,137,828. Robinson, L.F., Morrison, A.W., Bamforth, M.R., 1988. Improvements relating to biosynthesis. European Patent 261,872. Running, J.A., Huss, R.J., Olson, P.T., 1994. Heterotrophic production of ascorbic acid by microalgae. J. Appl. Phycol. 6, 99-104. Schlipalius, L., 1991. The extensive commercial cultivation of Dunaliella salina. Bioresour. Technol. 38, 241-243. Setlik, I., Veladimir, S., Malek, I., 1970. Dual purpose open circulation units for large scale culture of algae in temperate zones. I. Basic design considerations and scheme of pilot plant. Algol. Stud. (Trebon) 1, 11. Shelef, G., Soeder, C.J. (Ed.), 1980. Algae biomass. Production and use. Elsevier/North Holland Biomedical Press, Amsterdam. Soeder, C.J., 1981. Productivity of microalgal systems. U.O.F.S. Publ., Series C 3, 9-15. Soong, P., 1980. Production and development of Chlorella and Spirulina in Taiwan. In: Shelef, G., Soeder, C.J. (Eds.), Algae Biomass. Elsevier/North Holland Biomedical Press, Amsterdam, pp. 97-113. Stengel, E., 1970. Anlagentypen und Verfahren der technischen Algenmassenproduktion. Ber. Deutsch. Bot. Ges. 83, 589-606.

321

Tamiya, H., 1957. Mass culture of algae. Ann. Rev. Plant Physiol. 8, 309-344. Tanticharoen, M., Bunnag, B., Vonshak, A., 1993. Cultivation of Spirulma using secondary treated starch wastewater. Aust. Biotechnol. 3. 223-226. Torzillo, G., 1997. Tubular bioreactors. In: Vonshak, A. (Ed.), Spirulina platensis (Arthrospira): Physiology, cell-biology and biotechnology. Taylor and Francis, London, pp. 101115. Tredici, M.R., Zitelli, G.C., 1997. Cultivation of Spirulina (Arthrospira) platensis in flat plate reactors. In: Vonshak, A. (Ed.), Spirulina platensis (Arthrospira): Physiology, cellbiology and biotechnology. Taylor and Francis, London, pp. 117-130. Tsukada, O., Kawahara, T., Miyachi, S., 1977. Mass culture of Chlorella in Asian countries. In: Mitsui, A., Miyachi, S., San Pietro, A., Tamura, S. (Eds.), Biological Solar Energy Conversion. Academic Press, New York, pp. 363-365. Venkatamaran, L.V., 1986. Blue-green algae as biofertilizer. In: Richmond, A. (Ed.), CRC Handbook of Microalgal Mass Culture. CRC Press, Boca Raton, FL, pp. 455-471. Venkataraman, L.V., Becker, E.W., 1985. Biotechnology and utilization of algae--The Indian experience. Department of Science and Technology, New Delhi. Vonshak, A., 1997. Outdoor mass production of Spirulina: The basic concept. In: Vonshak, A. (Ed.), Spirulina platensis (Arthrospira): Physiology, cell-biology and biotechnology. Taylor and Francis, London, pp. 79-99. Warburg, O., 1919. (0ber die Geschwindigkeit der Kohlens~iurezusammensetzung in lebenden Zellen. Biochemische Zeitschrift 100, 230-270. Watson, A.S., 1979. Aquaculture and Algae Culture. Process and Production. Noyes Data Corporation, NJ. Weissman, J.C., Goebel, R.P., 1987. Design and analysis of microalgal open pond systems for the purpose of producing fuels. Solar Energy Research Institute, Report SERI/ STR-231-2840, 1- 214. Wharton, R.A., Smernoff, D.T., Averner, M.M., 1988. Algae in space. In: Lembi, C.A., Waaland, J.R. (Eds.), Algae and Human Affairs. Cambridge University Press, Cambridge, pp. 485-509.


ii

iiiii

JOURNAL

i

OF



Specific growth rate of Chlamydomonas reinhardtii and Chlorella sorokiniana under medium duration light/dark cycles" 13-87 s Marcel Janssen a,,, Tjibbe Chris Kuijpers a Bram Veldhoen a, Michel Brik Ternbach a, Johannes Tramper a, Luuc R. Mur b, Ren6 H. Wijffels a a Food and Bioprocess Engineering Group, Department of Food Science, Wageningen Agricultural University, PO Box 8129, 6700 EV Wageningen, Netherlands b Amsterdam Research Institute for Substances in Ecosystems, University of Amsterdam, Nieuwe Achtergracht 127, I018 WS Amsterdam, Netherlands


Abstract

The specific growth rate of Chlamydomonas reinhardtii and Chlorella sorokiniana decreased under square-wave light/dark cycles of medium duration, 13-87 s, in comparison to continuous illumination. Three experiments were done in three different turbidostats at saturating and sub-saturating light intensities during the light period, 240-630 ~tmol m - 2 s-1. Within each experiment the light intensity during the light periods of the intermittent light regimes was equal and this intensity was also applied under continuous illumination. The specific growth rate decreased proportional or more than proportional to the fraction of time the algae were exposed to light; this light fraction ranged from 0.32 to 0.88. We conclude that under these light regimes the chlorophyta C. reinhardtii and C. sorokiniana are not able to store light energy in the light period to sustain growth in the dark period at the same rate as under continuous illumination. C. reinhardtii increased its specific light absorbing surface by increasing its chloropyll-a content under light/dark cycles of 13 s duration and a light fraction of 0.67 at 240 ~tmol m - 2 s-1; the chloropyll-a content was twice as high under intermittent illumination in comparison to continuous illumination. The combination of a higher specific light absorption together with a lower specific growth rate led to a decrease of the yield of biomass on light energy under intermittent illumination. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Chlamydomonas reinhardtii; Chlorella sorokiniana" Light/dark cycles; Intermittent illumination; Specific growth rate

* Corresponding author. Tel.: + 31-317-483396; fax: + 31317-482237. E-mail address: [email protected] (M. Janssen)

1. Introduction

The possibility of using photo-autotrophic micro-organisms has been recognized in biotechnol-

0168-1656/99/$ - see front matter 9 1999 Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00084-X

324

M. Janssen et al./Journal of Biotechnology 70 (1999) 323-333

ogy. Microalgae are produced on a commercial scale for the production of single-cell protein, polysaccharides, health-food compounds such as polyunsaturated fatty acids and vitamins. In addition, algae are capable of accumulating heavy metals (Becker, 1994). Other promising applications of photo-autotrophic micro-organisms are the production of hydrogen gas from water and (sun)light (Rao and Hall, 1996; Schulz, 1996) and the fixation of carbon dioxide (Yoshihara et al., 1996; Akimoto et al., 1997; Hu et al., 1998). Large-scale cultivation of photo-autotrophic micro-organisms, however, is still limited by scaleup problems and economical considerations. Development of cheaper and more efficient photobioreactors is a key issue. Light energy often limits reactor productivity and therefore light should be used at the highest possible efficiency. But in high-density cultures only algae close to the illuminated surface, the so-called photic zone, are exposed to light. Algae outside the photic zone receive no light. Moreover, a considerable part of the light energy absorbed in the photic zone will be dissipated as heat because the capacity at which algae fix light energy is limited. This heatdissipated light energy cannot be used anymore by the algae outside the photic zone. It has been shown that faster mixing of highdensity cultures increases the yield of biomass on light energy at high light intensities (Laws et al., 1983; Hu et al., 1996; Hu and Richmond, 1996). Mixing results in movement of algae through the culture and, as such, they are exposed to an intermittent light regime. In other words, algae are alternately exposed to light and no light at all. As a result of faster mixing, algae are exposed to high light intensities more often, but the time of exposure is shorter. These shorter light/dark cycles probably have caused the reported higher biomass yields on light energy because less energy is wasted in the photic zone (Richmond, 1996). The duration of the light/dark cycles in the above studies was in the range of 0.2-1 s. For providing these fast fluctuations reactors were developed with a short light path and a fast mixing rate provided by intense aeration. It would be interesting to investigate the effect of longer light/dark cycles on algal productivity. Light/dark

cycles of medium duration, 10-100 s, can be applied more easily in closed photobioreactors. Reactor types in which the suspension is circulated in a manner resembling plug flow, such as air-lift loop reactors, are suitable for this. Only part of the reactor volume needs to be exposed to (sun)light. The algae are intermittently illuminated as they circulate through the reactor. Such systems can be better temperature controlled, as the dark part of the reactor is not exposed to sunlight and can be cooled more easily. Furthermore, the dark/light cycle can be controlled well by controlling the superficial liquid circulation velocities. Finally, there is a lot of knowledge available on hydrodynamics, mass transfer and scale up of air-lift systems (Chisti, 1989). The influence of medium duration light/dark cycles is not clear. In short term oxygen evolution experiments with non-acclimated Chlorella and Scenedesmus no influence of fluctuating light, of equal time-averaged irradiance, (1-0.0038 Hz, 1263 s cycle duration) was found on specific oxygen production and carbon fixation (Grobbelaar, 1989, 1991; Grobbelaar et al., 1992). However, an increase in specific oxygen production was observed after 3 h of exposure to fluctuating light and apparently physiological changes took place under fluctuating light (Grobbelaar, 1991). Others observed a decrease of the specific growth rate of the chlorophyta Chlamydomonas, Oocystis, Dunaliella, Chlorella and Scenedesmus at frequencies of 0.2-0.1 Hz (5-10 s cycle duration) in comparison to continuous illumination of the same time-averaged irradiance (Walsh and Legendre, 1982; Qu~guiner and Legendre, 1986; Nedbal et al., 1996). On the contrary, Pirt (1986) suggested that Chlorella sorokiniana was able to store enough ATP and reducing power to sustain maximal growth for about 9 s in darkness. In addition, Merchuk et al. (1998) showed that the red microalga Porphyridium sp. could sustain maximal growth in darkness for 6 s during light/dark cycles of 27 s. It is also suggested that photosynthesis in green plants is limited by carbon fixation and that short dark periods (up to 30 s) following a saturating light fleck can result in enhanced carbon fixation (Sharkey et al., 1986a,b; Stitt, 1986).

M. Janssen et al./'Journal of Biotechnology 70 (1999) 323-333

With respect to photobioreactor design it would be very interesting to elucidate the effect of medium duration light/dark cycles on algal productivity. Therefore we measured the specific growth rates of two green algae (chlorophyta) under square-wave light/dark cycles. For comparison the specific growth rate was also determined under continuous illumination. We used Chlamydomonas reinhardtii and Chlorella sorokiniana and they were cultivated in three different systems. The cycle duration ranged from 13 to 87 s and the fraction of time the algae were exposed to light during a cycle (e) ranged from 0.32 to 0.88.

325

2.1. Experiment 1" C. reinhardtii 2.1.1. Organism and medium A C. reinhardtii wild type strain, coded 21 gr, was kindly provided by Dr C. Vilchez from the University of Huelva (Spain). The organism was cultivated in a Sueoka high salt (HS) medium as described by Harris (1989). The medium has the following composition (amounts in g 1-1): NH4C1, 0.5; MgSO4.7H20, 0.02; CaClz.2H20, 0.01; K2HPO4, 1.44; KH2PO4, 0.72 and 5 ml of Htitner's trace elements solution (Harris, 1989). Pure cultures were maintained on agar slants containing Sueoka medium + 0.84 g 1-1 NaHCO3 + 1.75% w/v agar. The cultures were exposed to very mild daylight at room temperature.


Three different cultivation systems were used in three different experiments: (1) C. reinhardtii was cultivated in two air-lift loop reactors under continuous illumination and under a 8.5/4.4 s light/ dark cycle at 240 ~tmol m -2 s - l ; (2) C. sorokiniana was cultivated in a bubble column under continuous illumination at four different light intensities. At the highest and lowest intensity a 30/10 s square-wave light/dark cycle was applied; (3) C. sorokiniana was cultivated in a flat panel reactor under five different square-wave light/dark cycles at 250-290 lamol m - 2 S-1 which are presented in Table 1. Within each experiment the light intensity during the light periods of the intermittent light regimes was equal.

Table 1 Experimental design specific growth rate determination of C. sorokiniana in fiat panel turbidostat under different light/dark cycles Cycle duration, tc (s)

Light fraction, e (-)

25 25 56 87 87

0.32 0.88 0.60 0.32 0.88

2.1.2. Light intensity We used 2-re PAR (Photosynthetic Active Radiation, 400-700 nm) sensors made by I M A G - D L O (Wageningen, The Netherlands) to measure the photon flux density in lamol m - 2 s-1. The photon flux density is used to calculate the average light intensity as is explained later. 2.1.3. Reactor Chlamydomonas was cultivated under non-aseptic conditions in two glass 0.6 1 air-lift loop reactors (Fig. 1). The reactors were equipped with a water jacket connected to a temperature-controlled water bath (25~ The pH was maintained at 7.0 by automatic addition of 1 N sodium-hydroxide solution. The applied gas-flow rate was 30.9 1 h-~: the gas was a mixture of air and carbon dioxide (3% v/v CO2). The downcomer of one reactor was sealed with aluminum foil to create a dark part and the algal suspension circulated between this dark part and the illuminated part. The circulation time (tc) through the loop and the residence time in the downcomer (td) of the reactor liquid were determined by following the movement of a small colored K-carrageenan gel bead (2.6% w/v). tc was 12.9 s and td was 4.4 s leading to a light fraction (E) of 0.66 ( = { t c - td}/tc). The reactors were placed in a closed cabinet with seven fluorescent light tubes placed horizontally at one side (Philips TLD HF, 32W, color

326


~6

ily and it is a compromise between a low level of mutual shading and a cell density high enough for measurements of biomass characteristics and accurate turbidostat control. All these operations were monitored by a datalogger (CR10, Campbell Scientific Ltd., UK). I e

O"

-3

||

o

o

o~

7"---8

Fig. 1. Air-lift loop reactor used for cultivation of C. reinhardtii (front view), (1) gas inlet; (2) water jacket; (3) quantum sensor; (4) pH electrode; (5) suspension outlet; (6) medium inlet; (7) direction of suspension flow; (8) riser; (9) downcomer. 94). Both reactors were placed close to the tubes. One fluorescent tube (Pope FTD, 18W, color 94) was placed vertically at the opposite side of each ALR. The photon flux impinging at a certain height on the surface of the riser of one of the reactors was measured in four directions. The sum of these fluxes was divided by 2 to obtain an estimate of the average flux crossing a flat surface from both sides, called the average light intensity (I); I was 240 ~tmol m - 2 S-1. The reactors were illuminated 24 h a day; no day/night cycle was applied. The reactors were operated as turbidostats. Each reactor was therefore equipped with a quantum sensor positioned behind the riser (Fig. 1) and this sensor was facing the wall with the horizontal fluorescent tubes. When the biomass density increased the photon flux impinging on the sensors decreased. On the moment the photon flux was lower than the set point, 70% of the maximal flux measured when the reactor contained medium without algae, the reactors were automatically diluted until the set point was reached again. This set point was chosen arbitrar-

2.1.4. Specific growth rate The reactors were inoculated with pure C. reinhardtii cultures, maintained on agar slants and they were operated as turbidostats for 66 days. The dilution rates of the reactors were determined by daily measurement of the diluted culture volumes. Since the reactors were operated as turbidostats the biomass density in each reactor was constant and the specific growth rate was equal to the measured dilution rate. 2.1.5. Photosynthetic activity Photosynthetic activity (Po2) was measured in a small reaction vessel with a water jacket connected to a temperature-controlled water bath (25~ The vessel was placed in a closed cabinet. The reaction vessel was equipped with a Clarktype oxygen electrode (YSI 5331, Yellow Springs Instruments Co., USA), connected to a read-out unit (YSI 5300) which in turn was connected to a pen-recorder. The reaction vessel was surrounded by four halogen lamps (Osram 44875 M F L Natura, 12V/ 35-50W/30 ~ which were connected to a power supply equipped with two timers to turn on and off the lamps alternately to apply a light/dark cycle. The light intensity in the vessel could be adjusted by means of neutral density filters (Schott NG-filters, Schott Glaswerke, Germany). The photon flux in the vessel was measured by placing a quantum sensor in the vessel. The photon flux was measured in eight directions. The sum of these fluxes was divided by 4 to obtain an estimate of the average flux crossing a flat surface from both sides, called the average light intensity (I). For each measurement at one light intensity a fresh 6 ml sample was used consisting of 4 ml 75 mM phosphate buffer of pH 6.8 and 2 ml (diluted) culture suspension from the turbidostats; the oxygen activity in the sample was lowered by


purging nitrogen gas; extra carbon dioxide was added as sodium bicarbonate (25 ~tl, 0.56 M). First, the oxygen evolution rate was measured under illumination and after this dark respiration was recorded when it had reached a constant rate. The photosynthetic activity was calculated by adding the measured dark respiration rate to the oxygen evolution rate.

2.1.6. Chlorophyll-a A known volume of the algal suspension (containing 24-98 g of chlorophyll-a) was filtered through a glass-fiber filter (Whatman GF/F) coated with a thin layer of CaCO 3. The filter was put in a glass tube with screw-cap. The extraction was done with 10 ml of 90% v/v ethanol and was facilitated by heating the solution for 3 min (78~ in closed tubes. The tubes were stored in the dark at 4~ to complete extraction overnight. After centrifugation the chlorophyll-a concentration was determined according to Nusch (1980). 2.1.7. Dry weight A known volume (20-50 ml) of the algal suspension was filtered through a pre-dried and preweighed glass-fiber filter (Whatman GF/F). The biomass on top of the filter was washed by filtering demineralised water. The filter with biomass was dried at 105~ and allowed to cool in a desiccator. After this the filter was weighed again. 2.2. Experiments 2 and 3: C. sorok&&na 2.2.1. Organism and medium The chlorophyt C. sorok&iana (CCAP 211/8k) was obtained from the Culture Collection of Algae and Protozoa (Ambleside, UK). This organism was also used by Lee and Pirt (1981) and was called Chlorella vulgar& (Sorokin strain). We used the same A9 medium as described by Lee and Pirt (1981). Since this medium was developed for high density cultures we decided to half the concentration of most components. Only the buffer concentration was maintained at 19.5 mM during the experiments in the bubble column. During the experiments in the flat panel reactor the phos-

327

phate buffer was decreased to 4.85 mM and the urea concentration was also decreased from 33 mM in the bubble column to 10 mM in the flat panel. The medium was not autoclaved but was stored in the dark as four separate solutions. Prior to an experiment the solutions were combined to obtain medium for one week. Growth of contaminant organisms in the medium did not occur in this period.

2.2.2. Reactor C. sorokiniana was cultivated in two types of reactors with different light sources. The first type was a glass bubble column with 380 ml liquid volume with a gas inlet of sintered glass in the bottom. This bubble column was illuminated from two opposite sites by a series of six halogen lamps (Osram Decostar Titan, 12V/50W/60 ~ placed on top of each other to illuminate the whole of the column length. This reactor was illuminated 24 h a day; again no day/night cycle was applied. A 2-rt PAR sensor was continuously placed in the center of the bubble column containing the growing culture suspension The signal of this diode was used to apply a turbidostat control. The photon flux in the center of the column was allowed to drop to 70% of the value measured when the reactor contained medium without algae. The average light intensity in the bubble column was measured in an empty column without liquid. The photon flux density was measured on five different heights and at each height in four directions. The sum of these 20 measurements was divided by 10 to obtain an estimate of the average flux crossing a flat surface from both sides and this was called the average light intensity (I). The second type of reactor was a plexiglass flat panel of 225 ml liquid volume and a depth ( = light path) of 1 cm. The height and width of the rectangular panel was 28 cm and 10 cm, respectively. At the bottom gas was injected via an inlet of sintered glass. The complete reactor surface was illuminated by a screen (9 x 28 cm) of 578 light emitting diodes (LEDs). The emission wavelength of the LEDs (Kingbright, type L 53 SRC E) was 660 nm. Matthijs et al. (1996) concluded

328


there was no need for an additional supply of blue light when using these red LEDs for cultivation of Chlorella pyrenoidosa. The photon flux density leaving the reactor at the opposite side was continuously measured by a 2-~ PAR sensor (LICOR, type LI-190 SA) and this sensor was used to apply turbidostat control. The photon flux leaving the reactor was allowed to drop to 50% of the value measured when the reactor contained medium without algae. The average light intensity in the reactor was measured as the photon flux leaving the reactor, which contained medium without algae. In this flat panel reactor C. sorokiniana was illuminated according to a 14/10 h day/night cycle. Both reactors were equipped with a water jacket connected to a temperature-controlled water bath (35.5-37.5~ Carbon dioxide was added by bubbling the reactors with a carbon dioxide/air mixture, 3 - 5 % v/v CO2. The pH was controlled between 6.5 and 7.2 by automatic addition of 1 N sodium hydroxide or hydrochloric acid. The light sources of both reactors were connected to power supplies each equipped with two timers to turn on and off the lamps alternately to apply a light/dark cycle. 2.2.3. Specific growth rate 50 ml batch cultures were inoculated from pure cultures maintained on agar slants. The batches were cultivated under low light ( < 100 ~tmol m - 2 s-1) and used to inoculate the turbidostat reactors. The turbidostats were operated non-aseptically for one or two weeks under a certain light regime. During this period the specific growth rate was determined over 3 - 5 days by daily measurement of the diluted culture volume. Since the reactors were operated as turbidostats the biomass density in each reactor was constant and the specific growth rate was equal to the measured dilution rate. The growth rate measurements on the different days were used to calculate the 95%confidence interval. After each experimental run at a certain light regime the reactors were emptied, cleaned and inoculated for a new run at another light regime.

3. Results and discussion 3.1. Experiment 1: C. reinhardtii

The air-lift loop reactors were inoculated and 3 days later, denoted as day 0, the biomass density had reached the set point at which the reactors were automatically diluted. From day 0 until day 66 the reactors were operated as turbidostats and the dilution rate was followed. After about a week a considerable amount of algae attached to the reactor surface. The PAR-sensors controlling the dilution rates were placed behind the riser. In the risers attachment of algae and concomitant biofilm growth were negligible due to the rising gas bubbles. As a consequence, the biomass density in the suspension was maintained constant. Attached algae were detached once or twice a day resulting in a temporal increase of biomass density and, consequently, optical density. After detachment the dilution rate increased until the turbidostat set point was reached again. The extra volume diluted in the hour following detachment was estimated and subtracted from the volume diluted daily. Detaxzhment was done by scraping the inner surface with a magnet or with a rubber ring connected to a stick. The magnet was moved around by another magnet outside of the reactor. The specific growth rate of both cultures was measured almost continuously by following the dilution rate. The average specific growth rate of the intermittently illuminated culture, 0.11 h - ~, was 69% of the average growth rate of the continuously illuminated culture, 0.16 h - 1 (Table 2). The specific growth rate of 0.16 h - 1 is higher than the maximal growth rate we measured in another experiment (not shown) with C. reinTable 2 Average specific growth rate (/~) and dry weight concentration (DW) in the continuously and intermittently illuminated cultures during days 0-66. STD represents the estimated standard deviation

Continuous Intermittent

/1 (h- ~)

DW (g 1- ~) average (n)

STD

0.16 0.11

0.22 (6) 0.16 (6)

0.02 0.02

M. Janssen et al. ,'Journal of Biotechnology 70 (1999) 323-333

hardtii, 0.14 h-1. The light intensity of 240 lamol m -2 s -1 obviously is sufficient for maximal growth and this is said to be saturating. The intermittently illuminated culture was exposed to light 66% of the time. Hence the specific growth rate was proportional to the fraction of time the algae were exposed to light. Although the specific growth rate of both cultures is measured, we do not know whether the biomass yield on light energy has changed without information about the amount of light energy that is absorbed by the biomass. The overall biomass yield on light energy (Y~,E) in dry weight per amount of absorbed photons (g mol-1) is calculated according to Eq. (1). x./~ Y~,F-=E. e

[ggmol-1]

329

1.0

0.8

--

-

ratio chl-a content continuous 9intermittent

oo 0 . 6 !

o 0.40.20.0 0

I

I

I

I

I

I

10

20

30

40

50

60

70

Time (days) Fig. 2. Ratio of the chlorophyll-a (chl-a) content of algal biomass from the continuously and intermittently illuminated cultures during the period of operation (e).

(1)

Yx, E represents the ratio of biomass production over energy consumption, including maintenance requirements. The volumetric light absorption rate of the algal suspension (E in gmol 1-~ h-~) is not known. But E will be the same under continuous and intermittent illumination because the optical density of the cultures in both turbidostats was equal. The biomass densities (x) of both cultures were determined several times as dry weight per volume during the 66 days experimental period and are presented in Table 2. The densities of the continuously and intermittently illuminated cultures were 0.22 and 0.16 g 1-~, respectively. The ratio of the yield under continuous and intermittent illumination now can be calculated and was 1"0.76. Yx, E decreased under intermittent illumination in comparison to continuous illumination. The specific light absorption of the intermittently illuminated culture obviously is higher because the biomass density is lower in comparison to continuous illumination. This is caused by a difference in the chlorophyll-a (chl-a) content of both cultures. The chl-a content of the continuously illuminated culture was twice as low as the chl-a content of the intermittently illuminated culture, 0.51 _+ 0.057 (95% confidence interval). This ratio was determined several times during the experimental period and is presented in Fig. 2.

The intermittently illuminated culture acclimated to this light regime by increasing its chl-a content, enabling the algal cells to intercept a larger part of the photon flux. This was also reported in other investigations (Qu6guiner and Legendre, 1986; Shin et al., 1987). The higher chl-a content, however, could not prevent the decrease of the specific growth rate of C. rein~ hardtii under intermittent illumination in comparison to continuous illumination and consequently also y,. E decreased. Increased pigmentation is also found during the process of acclimation to low light intensities, which is called photoacclimation (Falkowski and LaRoche, 1991). Grobbelaar et al. (1996) already concluded that one of the most important factors influencing photosynthetic rates under continuous or intermittent illumination was whether algae were low or high light acclimated. Low light/dark frequencies (20 s period) were perceived as low light conditions and high light/dark frequencies (20 ms period) as high light conditions. Our results are in agreement with these findings since we applied medium frequency light/dark cycles and C. reinhardtii showed low light acclimation. Next to the specific growth rate also the photosynthetic activity (Po2), oxygen production per dry weight, was measured for both C. reinhardtii cultures. This was done at several light intensities to obtain a so-called photosynthesis-irradiance

330


(PI) curve. Po2 of the intermittently illuminated culture was determined under intermittent illumination and Po2 of the continuously illuminated culture was determined under continuous illumination. The results are graphically represented in Fig. 3. The data were fitted with the hyperbolic tangent model (Chalker, 1980), see Eq. (2). PO2

~-- P 0 2

....

"

tan h ( P o~

.I)

[ m g g - l h -1]

9m a x

(2)

In this model, I is the light intensity during the light period in the case of intermittent illumination. At increasing light intensities the slope decreases and the activity increases until the maximal activity (Po2, max) is reached. The fitted constants, the initial slope of the PI curve, and Po 2,m a x are presented in Table 3. These constants were used to calculate the specific oxygen production rates in the turbidostats on a dry weight basis. The light intensity in the turbidostats is 240 ~tmol m - 2 S - - 1 or lower. At this light intensity the Po2 is 248 and 212 mg O2 g-1 h-1 for the continuously and intermittently illuminated cultures under continuous and intermittent illumina-

300 -

..

_-, _-

"7 180 "~

O

j

--'-" 240 -

120 60 0

0

I

1

I

I

1

300

600

900

1200

1500

I (~tmol.m'2.s- 1) Fig. 3. Photosynthetic activity (Po2), oxygen production per dry weight, versus light intensity (I) for the continuously (O) and intermittently ( 9 illuminated cultures under continuous and intermittent illumination (8.5/4.4 s light/dark cycle), respectively. Samples from the turbidostats were analyzed on days 64 and 66, respectively. The light intensity (I) on the x-axis represents the intensity in the light period; the solid lines represent a curve fit with the hyperbolic tangent model.

Table 3 Po2 . . . . and 95% confidence intervals as determined from the data presented in Fig. 3 by a curve fit of the hyperbolic tangent model. The chl-a content on dry weight basis of the continuously and intermittently illuminated cultures was 10.3 and 20.2 mg g - l , respectively

Continuous Intermittent

~(mggh - l (mol m -2 s - l ) - ! )

(mg g-1 h - l )

1.49 + 0.43 1.19 + 0.22

297 + 37 270 _+ 21

P o 2, m a x

tion, respectively. The lower photosynthetic activity per dry weight of the intermittently illuminated culture under intermittent illumination corresponds to the lower specific growth rate of this culture. The decrease of the photosynthetic activity under intermittent illumination, however, is not proportional to the decrease in specific growth rate under intermittent illumination compared to continuous illumination. Under these circumstances specific growth rate and oxygen production do not seem to be linearly correlated. 3.2. Experiments 2 and 3: C. sorokiniana

The specific growth rate of C. sorokiniana was measured under continuous illumination at four different light intensities in the bubble column. The results of these measurements are shown in Fig. 4. As expected the specific growth rate increased with an increasing light intensity. At 630 ~tmol m - 2 s - 1 a specific growth rate of 0.27 h-1 was measured. Lee and Pirt (1981) measured a maximal specific growth rate of 0.235 h-1 for C. sorokiniana and 630 pmol m - 2 s-1 obviously is saturating. At 630 lamol m - 2 S-1 C. sorokiniana was cultivated under a 30/10 s light/dark cycle and the average specific growth rate was 0.18 h-1. This corresponds to a decrease of 33% in comparison to continuous illumination. Lee and Pirt (1981) and Pirt (1986) suggested that under this light/ dark cycle C. sorokiniana would be able to maintain the maximal growth rate. We were not able to confirm their conclusion and we found a reduction of the specific growth rate proportional to the fraction of time the culture was exposed to


light. Unfortunately the variation of the measured growth rates was considerable. This was caused by attachment of algae to the reactor surface and concomitant biofilm growth. C. sorokiniana was also cultivated under intermittent illumination at a low light intensity of 58 ~mol m - 2 s-1. Again a 30/10 s light/dark cycle was applied and the average specific growth rate was 0.060 h-1 in comparison to 0.077 h-1 under continuous illumination. Although the confidence intervals of the averages are big, the specific growth rate decreased by 22% under intermittent illumination in comparison to continuous illumination. Similar experiments were done in the other turbidostat, the flat panel illuminated with the LED screen, at 250-290 ~tmol m -2 s-~. The results of these experiments are shown in Fig. 5. The specific growth rate under a large light fraction of 0.88 was 0.15 h-~ and this is low in comparison to the growth rates measured in the bubble column. The introduction of a 10 h night period caused a reduction in the specific growth rate. Furthermore, the light intensity could be just too low to support the maximal growth rate and maybe is sub-saturating. In Fig. 5 the specific growth rate is plotted as a function of the light fraction (e). The dotted line 0.35

0.28 0.21 "7 =t. 0 . 1 4 continuous illumination intermittent illumination

0.07 -

0.00 0

i

i

I

l

~

100

200

300

400

500

600

700

I (~tmol.m'2.s"1) Fig. 4. Specific growth rate (p) of C. sorokiniana under continuous illumination (O) and intermittent illumination (A), 30/10 s light/dark cycle. Error bars represent the 95% confidence intervals. The light intensity (I) on the x-axis represents the intensity in the light period; the intensity during intermittent illumination was increased by 10 lamol m - 2 s-~ in this graph to allow better comparison of the intervals.

331

25 /

0.20 -

J

0.16 -

56 s / / 1

0.12 -

/// 87s

::s

I

0.08 0.04 -

0.00 0.0

I

I

I

0.2

0.4

0.6

i 0.8

1.0

e(-) Fig. 5. Specific growth rate (it) of C. sorokiniana under intermittent illumination of different cycle duration (tr and light fraction (e) at 250-290 lamol m - 2 s-~, presented as a function of e (O). tr is given in seconds in the graph. The error bars represent the 95% confidence intervals.

represents the specific growth rate when it would be proportional to e. We assumed that the maximal growth rate could be obtained by extrapolating the line from the origin to the point corresponding to the highest growth rate at a light fraction of 0.88 and cycle duration of 25 s. All measured growth rates are on top or below the dotted line. At a low light fraction of 0.32 the growth rate decreased even more than proportional to e. Although large variations occurred in the growth rate measurements all three experiments using C. reinhardtii and C. sorokiniana showed that the specific growth rate decreased proportional or more than proportional to the fraction of time the algae were exposed to light under intermittent illumination in comparison to continuous illumination. We conclude that these algae are not able to store light energy during the light period to sustain growth in the dark period at the same level as under continuous illumination as was suggested by Lee and Pirt (1981) and Pirt (1986). At higher light intensities leading to photoinhibition ( > 600 lamol m - 2 s - 1 ) , algae could react quite differently to intermittent illumination. Under these intensities the dark period could provide time to recover from photoinhibitory damage as was suggested by Merchuk et al. (1998).

332


In this study accurate determination of growth rates in dilute culture suspensions was negatively affected by a t t a c h m e n t of algae to glass or other surfaces in the reactor. This is an i m p o r t a n t aspect, which should be kept in mind when designing experiments. Wall growth itself and also the influence of wall growth on measurements should be minimized. Finally, we found that the overall biomass yield (specific growth rate over specific light absorption rate) can vary significantly under different light regimes. The overall biomass yield of C. reinhardtii under an 8.5/4.4 s light/dark cycle was considerably lower than the yield under continuous illumination. As a result, the dark period will have to be shorter than 4.4 s for optimal light utilization efficiency. Moreover, it is not possible to increase the efficiency of C. reinhardtii by applying light/ dark cycles of m e d i u m duration, 10-100 s, at a saturating light intensity of 240 ~tmol m - 2 S-1.

Acknowledgements We are grateful to S O N for providing a scholarship for this research within the f r a m e w o r k of the p r o g r a m for ' Y o u n g Chemists', no. 4030-08.

Appendix A. Nomenclature PAR

L,E

E P%

Photosynthetic active radiation, all photons between 400 and 700 nm light intensity or [gmol rn -2 s - l ] p h o t o n flux density in P A R range fraction of time [-] algae were illuminated during light/ dark cycle yield of biomass [g lamol-1] on light energy biomass density [g 1-1] specific growth [h-1] rate volumetric light [lamol 1-~ h-~] absorption rate specific photosynthetic activity [mg g-1 h-1]

Po 2, max m a x i m a l specific

photosynthetic activity in PI curve initial slope of PI curve

[mg

g-1

h-l]

[mg g-1 h-1 (~tmol m -2 s-1)-1]

References Akimoto. M., Yamada, H., Ohtaguchi, K., Koide, K., 1997. Photoautotrophic cultivation of the green alga Chlamydomonas reinhardtii as a method for carbon dioxide fixation and alpha-linolenic acid production. J. Am. Oil Chem. Soc. 74, 181-183. Becker. E.W., 1994. Microalgae: Biotechnology and Microbiology. Cambridge University Press, Cambridge. Chalker. B.E., 1980. Modeling light saturation curves for photosynthesis: an exponential function. J. Theor. Biol. 84, 205-215. Chisti, M.Y.. 1989. Airlift Bioreactors. Elsevier, London. Falkowski, P.G., LaRoche, J., 1991. Minireview: acclimation to spectral irradiance in algae. J. Phycol. 27, 8-14. Grobbelaar, J.U., 1989. Do light/dark cycles of medium frequency enhance phytoplankton productivity? J. Appl. Phycol. 1, 333- 340. Grobbelaar, J.U., 1991. The influence of light/dark cycles in mixed algal cultures on their productivity. Bioresour. Technol. 38, 189-194. Grobbelaar, J.U., Kroon, B., Burger-Wiersma, T., Mur, L.R., 1992. Influence of medium frequency light/dark cycles of equal duration on the photosynthesis and respiration of Chlorella pyrenoidosa. Hydrobiologia 238, 53-62. Grobbelaar. J.U., Nedbal, L., Tich~, V., 1996. Influence of high frequency light dark fluctuations on photosynthetic characteristics of microalgae photoacclimated to different light intensities and implications for mass algal cultivation. J. Appl. Phycol. 8, 335-343. Harris, E.H., 1989. The Chlamydomonas Sourcebook: A Comprehensive Guide to Biology and Laboratory Use. Academic Press, San Diego. Hu. Q., Richmond. A., 1996. Productivity and photosynthetic efficiency of Spirulina platensis as affected by light intensity algal density and rate of mixing in a flat plate photobioreactor. J. Appl. Phycol. 8, 139-145. Hu, Q., Guterman, H., Richmond, A., 1996. A flat inclined modular photobioreactor for outdoor mass cultivation of photoautotrophs. Biotechnol. Bioeng. 51, 51-60. Hu, Q.. Kurano, N., Kawachi, M., Iwasaki, I., Miyachi, S., 1998. Ultrahigh-cell-density culture of a marine green alga Chlorococcum litterola in a flat-plate photobioreactor. Appl. Microbiol. Biotechnol. 49, 655-662. Laws, E.A., Terry, K.L., Wickman, J., Chalup, M.S., 1983. A simple algal production system designed to utilize the flashing light effect. Biotechnol. Bioeng. 25, 2319-2335. Lee, Y.K., Pirt, S.J., 1981. Energetics of photosynthetic algalgrowth: influence of intermittent illumination in short (40 s) cycles. J. Gen. Microbiol. 124, 43-52.

M. Janssen et al. ,/Journal of Biotechnology 70 (1999) 323-333 Matthijs, H.C.P., Balke, H., Hes, U.M.V., Kroon, B.M.A., Mur, L.R., Binot, R.A., I996. Application of light-emitting diodes in bioreactors: flashing light effects and energy economy in algal culture (Chlorella pyrenoidosa). Biotechnol. Bioeng. 50, 98-107. Merchuk, J.C., Ronen, M., Giris, S., Arad, S.M., 1998. Light/ dark cycles in the growth of the red microalga Porphyridium sp. Biotechnol. Bioeng. 59, 705-713. Nedbal, L., Tich~, V., Xiong, F., Grobbelaar, J.U., 1996. Microscopic green algae and cyanobacteria in highfrequency intermittent light. J. Appl. Phycol. 8, 325333. Nusch, E.A., 1980. Comparison of different methods for chlorophyll and phaeopigrnent determination. Arch. Hydrobiol. Beiheft 14, 14- 36. Pirt, S.J., 1986. The thermodynamic efficiency (quantum demand) and dynamics of photosynthetic growth. New Phytol. 102, 3-37. Qu6guiner, B., Legendre, L., 1986. Phytoplankton photosynthetic adaptation to high frequency light fluctuations simulating those induced by sea surface waves. Mar. Biol. 90, 483-491. Rao, K.K., Hall, D.O., 1996. Hydrogen production by cyanobacteria: Potential problems and prospects. J. Mar. Biotechnol. 4, 10-15. Richmond, A., 1996. Efficient utilization of high irradiance for production of photoautotrophic cell mass: a survey. J. Appl. Phycol. 8, 381-387.

333

Schulz, R., 1996. Hydrogenases and hydrogen production in eukaryotic organisms and cyanobacteria. J. Mar. Biotechnol. 4, 16-22. Sharkey, D.S., Seemann, J.R., Pearcy, R.W., 1986a. Contribution of metabolites of photosynthesis to postillumination CO2 assimilation in response to lightflecks. Plant Physiol. 82, 1063-1068. Sharkey, D.S., Stitt, M., Heineke, D., Gerhardt, R., Raschke, K., Heldt, H.W., 1986b. Limitation of photosynthesis by carbon metabolism. II. O2-insensitive CO2 uptake results from limitation of triose phosphate utilization. Plant Physiol. 81, 1123-1129. Shin, C.N., Rhee, G.-Y., Chen, J., 1987. Phosphate requirement, photosynthesis and diel cell cycle of Scenedesmus obliquus under fluctuating light. Can. J. Fish. Aquat. Sci. 44, 1753-1758. Stitt, M., 1986. Limitation of photosynthesis by carbon metabolism. I. Evidence for excess electron transport capacity in leaves carrying out photosynthesis in saturating light and CO 2. Plant Physiol. 81, 1115-1122. Walsh, P., Legendre, L., 1982. Effets des fluctuations rapides de la lumi~re sur la photosynth6se du phytoplancton. J. Plankton Res. 4, 313-327. Yoshihara, K.I., Nagase, H., Eguchi, K., Hirata, K., Miyamoto, K., 1996. Biological elimination of nitric oxide and carbon dioxide from flue gas by marine microalga NOA-113 cultivated in a long tubular photobioreactor. J. Ferment. Bioeng. 82, 351-354.


JOU

reNAl.

OF



Modelling of a continuous pilot photobioreactor for microalgae production Daniel Baquerisse b,,, St6phanie Nouals a,b Ars6ne Isambert b, Patrick Ferreira dos Santos a, G6rard Durand b a Thallia Pharmaceuticals S.A, l'Orke d'Ecully, 5 chemin de la ForestiOre, 69130 Ecully, France b Ecole Centrale Paris, Laboratoire de Chimie et Gknie des Prockdks, Grande voie des vignes, 92295 Ch6tenay-Malabry, France


Abstract

In this study, a model of a continuous pilot photobioreactor for microalgae production is proposed. Three aspects have been studied: the modelling of kinetic growth, the gas-liquid transfer and the hydrodynamics in the photobioreactor. The modelling of each aspect has been developed with the dynamic simulation software SpeedUp, after experimental studies, then validated step-by-step. The connection of these three aspects aims to predict and optimise biomass production of the pilot plant. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Porphyridium purpureum; Tubular photobioreactor; Growth model; Gas-transfer modelling; Hydrodynamics modelling

I. Introduction

Modelling of biotechnological processes becomes an important aid. It allows the best understanding of biological phenomena, hydrodynamics, thermodynamics and the interactions of these domains. Given a set of input data, a model is used to predict the output response of the process. A model can be used to solve differ-

* Corresponding author. Fax: + 33-1-41131554. E-mail addresses: [email protected] (D. Baquerisse),

[email protected] (S. Nouals)

ent types of problems like design, economical optimisation, process synthesis, and control. The aim of our work is to describe with a set of equations the behaviour of a microalga in a tubular photobioreactor. It implies the comprehension of the effects of different variables on microalgal growth. Mixing effects in the reactor and gas-liquid mass transfer must also be taken into account. This study includes an experimental part which allows the determination of interactions between different variables and the growth rate. A hydrodynamic study has been done to analyse the flow inside the reactor. At last, a gas-liquid transfer study allows to control variables of interest in growth modelling.


336

D. Baquerisse et al./'Journal of Biotechnology 70 (1999) 335-342

2. Material and methods

2.1. Alga and growth conditions Porphyridium purpureum 1380-1A strain, cultured in the photobioreactor and used for each experimental study, has been obtained from the Sammlung von Algenkulturen Pflanzenphysiologister Institut der Universit/it G6ttingen. The cells have been maintained in liquid culture, on Hemerick medium (Hemerick, 1973). The algal suspensions have been subcultured, by transferring cells to fresh medium every two weeks to maintain cells in exponential growth phase. 2.2. The photobioreactor The reactor we used is a continuous pilot-scale tubular photobioreactor of the Thallia Pharmaceuticals company. It is made of a horizontal snaked tube (length: 84 m; internal diameter: 2.4 cm). This one is connected to an airlift which is composed of an ascending and a descending tube (height: 2.5 m) separated by a 7 1 broth recipient. The total volume of this culture system is 50 1. The horizontal tube is artificially illuminated under white light (halogen lamp) with an intensity of 500 ~tE m - 2 s-1 and a 14/10 photoperiod. In addition, the horizontal tube is immersed in water of which the temperature is controlled around 25~ pH is also controlled by injection of carbon dioxide at the horizontal tube entry.

2.3. Growth kinetics experiments For growth kinetics experiments, we have chosen to work in batch mode with small volume photobioreactor (2.5 1). These thermoregulated double-jacket cylindrical reactors have been illuminated from three sides by fluorescent cool tubes (OSRAM L30W/77 for pink light and OSRAM L30W/72 for white light). An air feed enriched in carbon dioxide has been permitted the supply of carbon and the agitation of the culture. The aim of this experimental work has been to study the effects of different operational conditions on the growth of P. purpureum. In a first time, we studied the effects of: the airflow rate,

the percentage of carbon dioxide in the air and the incident light intensity. Each study on influencing variables has been realised separately, i.e. each factor has been varied on a range centred on its optimal value for the growth of P. purpureum. In the same way, the other operational conditions have been maintained at their optimal value. The optimal values for the growth of P. purpureum and for our little volume reactors are: T = 25~ air-flow = 2.5 --2 V.V.H., %CO2 in the air-flow---2, /in = 120 ~tE m s -1, pH 7 (experimental optimisation done by Thallia Pharmaceuticals Company). Each factor has been varied on an efficient range to obtain a limitation and inhibition of the growth around their optimal value.

2.4. Daily measures A cellular count on Malassez cells and a measure of the Optical Density at 760 nm (which is correlated at the turbidity of our alga) (Dermoun et al., 1992), have been done every day. In the same way, we have measured the incident and outgoing light intensity with a Licor Quantum Photometer and the dissolved carbon dioxide concentration with a probe INGOLD.

2.5. H)'drodynamics experiments The hydrodynamics experiments are implying the experimental determination of the residence time distribution (RTD). For its modelling, the photobioreactor has been separated in two parts, hydrodynamically different (horizontal tubes and airlift). To realise our hydrodynamics study, we have done an impulse of NaC1 to allow normal operation of the photobioreactor during our experiments. Two different gas flow rates have been studied (180 and 540 1 h-1) to obtain different liquid speeds inside the photobioreactor. NaC1 has been measured by use of silvermetry.

2.6. Mass transfer of carbon dioxide in the airlift experiments In our case, gas-liquid mass transfer of C O 2 is

D. Baquerisse et al./ Journal of Biotechnology 70 (1999) 335-342

of prime importance, because the CO2 is the main carbon source. In the aim to study the CO2 mass transfer efficiency, it has been necessary to determine the volumetric mass transfer coefficient KLa (CO2) which permits to characterise the CO2 transfer rate between the gas and liquid phases. According to the literature, volumetric mass transfer coefficients depend on the physical properties of the liquid, the liquid flow and on system and gas injector geometries (Nielsen and Villadsen, 1994a). Our photobioreactor is composed of two parts geometrically and hydrodynamically different, the air-lift and the snaked tube. Thus, the determination of the KLa (CO2) is necessary. In this paper, we present only the study in the air-lift. The calculation of KLa of CO2 has been done from the determination of the K,a of 02: /

KLa(C02) =

.

/ D~

. KLa(02)

(1)

Dco 2

To use this method, we have verified that the absorption of CO2 could be considered as a purely physical process (Talbot et al., 1991; Molina Grima et al., 1993). KLa of 02 has been obtained by the dynamic method with a probe INGOLD (Leveau and Bouix, 1988). A study of the effect of the airflow rate on KLa of CO2 has been done, coupled to a study of the viscosity effect. Three airflows have been studied and repeated twice for different viscosity, i.e. for different microalgal concentrations in the reactor (Table 1).

2. 7. The modelling software The software used for modelling and simulation is the dynamic simulation software SpeedUp, produced by Aspentech.

3. R e s u l t s and discussion

3.1. Growth kinetics Experiments done in batch process lead to growth curves similar to the one presented on Fig. 1. The analysis of exponential phase growth curve has allowed us to obtain the maximal specific growth rate, /z, which is the key variable of the modelling. In the aim to integrate at our modelling reactor internal variables, we have studied the evolution of incident and outgoing light intensity for each experiment. As Fig. 2 shows, the whole incident light intensity is captured by the culture when cell density is greater than 8 • 106 cells ml-~. Thus, we have defined a variable, which allows us to determine the amount of light intensity accessible per cell. This variable, E, has been defined by Krystallidis (1994): E = (/~N -- IOVT) " A

Experiments

Air-flow(l h - l )

Viscosity (cpoises)

1 2 3 4 5 6

540 540 380 380 170 170

1.209 1.264 1.220 1.272 1.206 1.296

(2)

V'X The treatment of all experimental data has allowed us to see that two internal variables, E and the concentration in dissolved CO2, had effects of limitation and inhibition on the growth rate and so an optimal value. So, the equation proposed by Steele (1977) has appeared as an adequate modelling in our case: I

Table 1 Operational conditions for the mass transfer experiments

337

H -- ]'/max /'opt

i 9e ~ l S ~ ]

(3)

Our choice was done on an extension of such an equation in the case of E and dissolved carbon dioxide concentration. Many authors (Frohlich et al., 1983; Engasser, 1988; Cornet et al., 1992; Krystallidis, 1994; Nielsen and Villadsen, 1994b) have established models in the case of a specific growth rate influenced by more than one substrate. For all these models, the specific growth

D. Baquerisse et al./Journal of Biotechnology 70 (1999) 335-342

338 40

1,6

35

1,4

30

1,2

25

1,0

20

0,8

15

0,6

10

0,4

!

~ ,,o ~ * s

~m ~ ~

0,2 -

0

7,

-

,

,,

,,

200

300

.

100

.....

,

0,0

400

Time (h) 1"-*- Cells ~

OD (760 nm) I

Fig. 1. Typical growth of P. purpureum, strain 1380-1A on Hemerick medium. 140 120

!

m,,. - - w

m - . .

.

m - .

.

a .

l

--

m

c

c,

m

a

m

p

0

0

.

r/3

100

~

8o

~'~

40

M

20

i

1o0

"~ec

mm i

cm-

200

cc

9

300

Time (h)

Fig. 2. Evolution of luminous intensity coming in and out the 2.5 1 photobioreactor.

i

400


3.2. Hydrodynamics

rate is obtained by multiplying functions of limitation of the different substrates. Therefore, we have decided to write the complete form of our modelling as follows: fl = flmax

"F,(E)'F2(C02)

The flow has been modelled by a cascade of continuous stirred tank reactors. Indeed such a modelling is able to represent the hydrodynamics of the photobioreactor. The analysis of the experimental curve by calculation of variance and residence time gives the number of reactors in cascade J that allows to represent the flow (Villermaux, 1993). This determination is valid for a symmetrical response of the type of this we obtain for the horizontal part of the photobioreactor. We present in Fig. 4 a response curve of the horizontal part of the photobioreactor to an impulse. It is shown that the deviation between experimental and simulated curves is low and included inside the uncertainty of the chloride concentration measures by silvermetry. When we considered the vertical part of the photobioreactor (airlift), we have had a response that is not symmetrical. The proposed modelling

(4)

We describe only the function of E, the function of C O 2 is similar: E

E . e(1._k_~op)

(5)

F,(E) = ~opt

339

This modelling needs to fit three parameters ]-/max, Eopt, C02opt. We have done this fitting with the batch experiments realised with 2.5 1 photobioreactors. We present here a validation example on a batch experiment, i.e. on an experiment not used to fit our model parameters (Fig. 3). The correlation coefficient of the model with the data used for fitting parameters and validation experiments is RZ = 0.97.

16 14

12 L.., Q

10

..x.. r.~

r,j

9

10

9

20

9

w

30

40

9

50

9

60

Time (h) I

9 Experimental ~,--Simulation]

Fig. 3. Validation curve of the proposed growth model.

70

80

90

D. Baquerisse et al. /Journal of Biotechnology 70 (1999) 335-342

340 0,54 !

0

E r,j o

0,52 0,5 0,48 0,46

FTT~TT~TTT~

eo

o

r,j 0

0,44

9

0,42 0,4 t 200

9

w

220

9

9

240

260

9

280

"'

9" "

300

320

Time (s) Experimental - - - - Simulation

Fig. 4. Validation curve of the hydrodynamics model based on a cascade of CSTR.

in this case is inspired of the airlift geometry. A large volume continuous stirred tank reactor (which corresponds to the broth recipient) is included in the modelling to obtain the good response of the airlift. A good accuracy between the experimental and the simulated curves has been obtained by this way. 3.3. Carbon dioxide mass transfer in the airlift

Four experiments were used to model KLa of CO2. We have used a model found in the literature and adapted on the transfer of CO2 in airlift systems (Lenglace, 1986) which is the following: KLa = ~ " F~" q "/

(7)

The parameters ~, fl and ~ were regressed with a good correlation R 2 - 0.99 (Fig. 5). The modelling obtained has been validated on the other experiments. Fig. 6 shows a good accuracy between the results simulated by the modelling and experimental results R 2 = 0.98.

4. Conclusions

The described model seems to reproduce correctly the experimental results. The modular aspect of the photobioreactor modelling is interesting because it allows us to complete the modelling if necessary. For example, the growth modelling is going on with studying the effect of temperature on the growth rate. We are thinking of studying as well the effect of the photoperiod. Indeed, the industrial reactor of the company for which we are working, will be under natural conditions. It could complicate the growth modelling since the literature shows that according to the photoperiod, cells can split up during the dark period (Dermoun, 1987). A model such the Droop one, which is a structured model and relates the specific growth rate (/1) to the intracellular concentration of the limiting nutrients ('cell quota', q) will have perhaps be considered (Droop, 1973, 1983). The mass transfer study might be also completed by the KLa determination in the horizontal tube of the photobioreactor. So, we will have a


341

1,2E-02

y-x 1,0E-02

.,-,, !

R 2 -

r.~

0,9997

'

-

8,0E-03

~Z 6,0E-03

4,0E-03

2,0E-03

0,0E+00 ~

,

0,0E+00

,

,

4,0E-03

8,0E-03

Experimental

Kla

1,2E-02

(S-1)

Fig. 5. Parameters mass transfer modelling fitting: comparison between experimental and simulated KLa.

1,0E-02

8,0E-03

Y = 1,0974x

,

"-

!

6,0E-03

4,0E-03 ~ v..,.I

2,0E-03

0,0E+00 , 0,0E+00

,

,

2,0E-03

4,0E-03

, ,6,0E-03

'I

8,0E-03

Experimental Kla (S-1) Fig. 6. Mass transfer modelling validation: comparison between experimental and simulated KLa.

1,0E-02

342

D. Baquerisse et al. ,/Journal of Biotechnology 70 (1999) 335-342

c o m p l e t e m a s s transfer m o d e l allowing to simulate the CO2 profile along the p h o t o b i o r e a c t o r .

Appendix A. Nomenclature V.V.H

KLa CO 2 Dco 2

002 02 # I

/opt Iouf A V X E

EOPT

C02 OPT ~, ~, ~,

Fo r/

gas flow rate per liquid culture volu m e a n d per h o u r mass transfer coefficient (s -~) c a r b o n dioxide c a r b o n dioxide diffusion coefficient (m 2 s - ' )

oxygen diffusion coefficient (m2 s-') oxygen specific growth rate (h-') maximal specific growth rate (h-') light intensity (E m -2 s-') optimal light intensity (E m -2 s-') incident light intensity (E m -e s-') outgoing light intensity (E m -2 s-') reactor area (m2) culture v o l u m e (m 3) cell c o n c e n t r a t i o n (106 cells m l - ' ) a m o u n t of light intensity accessible per cell (E s - ' 10 -6 cells) o p t i m a l a m o u n t o f light intensity accessible per (E s-1 10-6 cells) o p t i m a l c a r b o n dioxide c o n c e n t r a t i o n (mol 1-l) c o n s t a n t s o f the KLa m o d e l l i n g air flow rate (1 h - ~ ) viscosity (poiseuille)

References Comet, J.F., Dussap, C.G., Dubertret, G., 1992. A structured model for simulation of cultures of the cyanobacterium Spirulina platensis in photobioreactors: II. Identification of kinetic parameters under light and mineral limitations. Biotechnol. Bioeng. 42, 826-834. Dermoun, D., 1987. Ecophysiologie de Porphyridium cruentum. Validation exprrimentale d'un modrle de croissanceEtude de la production de polysaccharides. PHD thesis of

the Universit6 de Technologie de Compirgne, Compirgne, France. Dermoun, D., Chaumont, D., Thebault, J.M., Dauta, A., 1992. Modelling of growth of Porphyridium cruentum in connection with two interdependent factors: light and temperature. Bioresour. Technol. 42, 113-117. Droop, M.R., 1973. Some thoughts on nutrient limitation in algae. J. Phycol. 9, 264-272. Droop, M.R., 1983. 25 years of algal growth kinetics. Botanica Marina 26, 99-112. Engasser, J.M., 1988. Modelisation des Processus de Fermentation. In: Scriban, R. (Ed.), Biotechnologie. Lavoisier Tec et Doc, Paris, pp. 301-326. Frohlich, B.T., Webster, I.A., Ataai, M.M., Shuler, M.L., 1983. Photobioreactors: models for interaction of light intensity, reactor design and algal physiology. Biotechnol. Bioeng. Symp. 13, 331-350. Hemerick, J., 1973. Culture methods and growth measurements. In: Stein, J.R. (Ed.), Handbook of Physiological Methods. Cambridge University Press, Cambridge, pp. 250-260. Krystallidis, A., 1994. Application du grnie des procrdrs aux biotechnologies marines--Etude de faisabilitr, modrlisation et simulation dynamique d'un procrd6 de culture de micro-algues. PHD thesis of the Ecole Centrale Paris, Chgttenay-Malabry, France. Lenglace, C., 1986. Transferts de matirre dans un biorracteur. PHD thesis of the Institut Polytechnique de Lorraine, Nancy, France. Molina Grima, E., Sanchez Prrez, J.A., Garcia Camacho, F., Robles Medina, A., 1993. Gas-liquid transfer of atmospheric CO2 in microalgal cultures. J. Chem. Tech. Biotechnol. 56, 329-337. Nielsen, J., Villadsen, J., 1994a. Mass transfer. In: Bioreaction and Engineering Principles. Plenum, New York, pp. 296313. Nielsen, J., Villadsen, J., 1994b. Modeling of Reactions Kinetics. In: Bioreaction and Engineering Principles. Plenum, New York, pp. 163-224. Leveau, J.Y., Bouix, M., 1988. Le transfert d'oxygrne. In: Scriban, R. (Ed.), Biotechnologie. Lavoisier Tec et Doc, Paris, pp. 243- 245. Steele, J.H., 1977. Microbial kinetics and dynamics. In: Lapidus, L., Amundson, N.R. (Eds.), Chemical Reactor Theory. Prentice Hall, Englewood Cliffs, NJ, pp. 405-483. Talbot, P., Gortares, M.P., Lencki, R.W., De la Noue, J., 1991. Absorption of CO2 in algal mass culture systems: a different characterization approach. Biotechnol. Bioeng. 37, 834-842. Villermaux, J., 1993. Bilans de population et ditribution de temps de sdjour. Moddlisation de l'~coulement et du mdlange dans les rracteurs rrels. In: Grnie de la rraction chimique--Conception et fonctionnement des rracteurs. Lavoisier Tec et Doc, Paris, pp. 159-233.



Opportunities for marine bioprocess intensification using novel bioreactor design" frequency of barotolerance in microorganisms obtained from surface waters Phillip C. Wright a,* Colin Stevenson b Eileen McEvoy a j. Grant Burgess ~ a Department of Mechanical and Chemical Engineering, Heriot-Watt University, Riccarton, Edinburgh EHI4 4AS, Scotland, UK b CoBoco (UK) Limited, Aberdeen ABI5 6FZ, Scotland, UK Department of Biological Sciences, Heriot-Watt UniversiO', Riccarton, Edinburgh, EH14 4AS, Scotland, UK


Abstract In the context of marine biochemical systems, opportunities exist for the development of novel reactors, with optimization and conversion of current technologies having the potential to yield more efficient units. A limiting factor in the widespread commercial acceptance of a large range of marine metabolites is the efficient production of, for example, sufficient quantities of antibiotics and nutraceuticals to allow for structural analysis and clinical testing. Conventional methods utilised for physical and chemical process intensification require careful analysis of their potential application to shear-sensitive bioprocess systems. Stress induction, for example, provides one route to marine bioprocess intensification due to the expression of metabolites not otherwise possible. Use of high pressure as a stressing agent and/or intensification tool is discussed, and its potential, demonstrated by showing the existence of barotolerant (at 120 MPa) marine microorganisms obtained from shallow surface waters ( < 1.5 m deep), is shown. Microorganisms associated with the surface of, for example, seaweed show a greater likelihood of being barotolerant. 9 1999 Elsevier Science B.V. All rights reserved. Keywords: High hydrostatic pressure; Stress induction; Bioprocess intensification: Bioreactor: Barotolerance

I. Introduction The oceans, due to their vastness and range of biodiversity, offer the potential for innovative bioprocess engineering, particularly in the areas of novel antibiotic/drug development (Fenical, 1997) * Corresponding author. Tel.: + 44-131-4495111; fax: + 44131-4513129. E-mail address: [email protected] (P.C. Wright)

and nutraceuticals (Brower, 1998). Due to the typically low energy density found in bioreactors there is every likelihood that these bioprocesses can probably be intensified. In nonbiological systems, chemical engineers use increased turbulence, temperature and pressure to achieve the desired intensification. M a n y different reactor configurations are used to do this, including bubble columns (airlift), stirred tanks and


344

P.C. Wright et al./Journal of Biotechnology 70 (1999) 343-349

three-phase fluidized beds (Wright and Raper, 1996). However, in biological systems turbulence is not a good option for living materials and much more care must therefore be taken (Chisti and Moo-Young, 1996). With the discovery and increased research into extremophiles, the conventional restrictions of low to moderate bioreactor temperatures and pressures faced by engineers may be circumvented. One particular route to bioprocess intensification (the enhancement of bioprocess productivity) is the harnessing of extremophiles. However, despite the increasing interest and number of these organisms being discovered, the step-change in bioprocess intensity/efficiency that beckons has still not been realised to any great degree. This was highlighted at the recent Extremophiles '98 conference in Japan, where the rate of discovery and characterisation of the mechanisms of extremophiles, rather than process applications, were the main focus of the registrants (Cowan, 1998). It is also obvious from the wider literature that the main thrust of extremophile research rests with (hyper)thermophiles. Screening for thermophiles and hyperthermophiles from the ocean depths is already a widespread practice, with thermostable enzymes such as amylases, proteases and glucosidases being the target (Prieur, 1997). Comparatively little work has been done on barotolerant/barophilic microorganisms.

1.1. High pressure Use of high pressure has some very exciting bioprocessing applications, as new deep-sea microorganisms, particularly members of the Archaea, are being discovered, with the potential for a range of applications in both aerobic and anaerobic environments (DeLong, 1997). One reason for the lack of knowledge on behaviour and novelty of microorganisms obtained from the deep sea is that little enrichment culturing work has been undertaken at very high hydrostatic pressure, despite this being the very environment common to these systems (Prieur, 1997). Both Nelson et al. (1992) and Kato et al. (1998) conducted experiments on marine bacterial strains

taken from large depths. Both found that the growth of these organisms were greater at elevated pressures than at lower pressures. Nelson et al. (1992) reported no growth was achieved at atmospheric pressure for the samples taken from the Mariana Trench. In a significant proportion of conventional bioreactor systems, high pressure or high temperature can have detrimental effects on growth of a wide range of microorganisms. For example, Hauben et al. (1997) reported that pressures in the range of 20 to 130 MPa could inhibit cellular growth, while higher pressures can result in cell death. A significant proportion of the high pressure biotechnology research has been carried out into the food sterilisation and the properties of the resultant treated product (see, for example, Pothakamury et al., 1995; Hill, 1997). Ath6s et al. (1997) carried out experiments on [3-galactosidase derived from Escherichia coli, Aspergillus oryzae and Kluyveromyces lactis, and noted that an important decrease in catalytic activity was observed above 300 MPa, and led to a quite complete inactivation at 500 MPa. However, showing important implications for improved bioprocess intensification, it was found by the same authors (Ath6s et al., 1997) that in the moderately high pressure range of 50 to 250 MPa, the biocatalysed (via [3-galactosidase) lactose hydrolysis reaction could be carried out at a higher temperature, thereby offering enhanced reaction rates. This often synergistic relationship between the important intensification bioprocess variables has also been noted elsewhere (Nelson et al., 1992; Michels and Clark, 1997).

1.2. Aims and scope of present research The scope of this work was to examine a range of options for bioprocess intensification, particularly those arising out of stress induction. A number of high pressure biological systems were examined utilising microorganisms obtained from a range of specific ecological niches. Several different methods have been used to study the effects of pressure on the growth of deep sea microorganisms (Yayanos, 1995; Michels


and Clark, 1997), with the pressure being applied either hydrostatically, or hyperbarically. However, no studies have systematically compared the pressure tolerance of marine bacteria recovered from surface waters as distinct from those recovered from deep sea habitats. In this work, a high pressure batch bioreactor (HPBB) was used throughout to create a maximum hydrostatic pressure of 120 MPa and to survey the pressure tolerance of a number of marine bacterial strains from surface waters (1-5m depth: < 50 Pa). In addition, the comparative barotolerance of strains isolated from open water and algal biofilms was investigated. Deep sea strains were not studied, as the aim was to survey barotolerance of strains from surface waters.


2.1. Escherichia coli EMG-23 A standard strain of E. coli EMG-23, to act as a control, was inoculated into 5 ml of nutrient broth and incubated at 25~ with shaking, for 24 hours. Prior to pressure treatment this inoculated solution was further diluted by adding 2 ml of media to give a final cell concentration of approximately 5 x 108 cell m l - 1. A 300 pl control sample was taken from the cell suspension. A 3 ml aliquot from the cell suspension was then added to the HPBB, which was constructed from stainless steel and consisted of a maximum working volume of 3.9 ml. Hydrostatic pressure was supplied to the bacterially inoculated broth by use of an electrically driven piston. The maximum allowable pressure within the cell was 120 MPa, and this pressure was used throughout all experiments. High pressure cycling was used, as it has been shown that this methodology may lower the ability of the biological species to withstand high pressure (Mozhaev et al., 1994; Hill, 1997). After the E. coli inoculated broth was fed into the HPBB, it was subjected to the maximum pressure of 120 MPa for 5 min at a constant temperature of 25~ The pressure was then reduced to 0 MPa and the broth allowed to rest for 5 min. The time for pressurising and de-pressuris-

345

ing was 1-2 s, and therefore, could be classed as essentially instantaneous. This multiple pulse treatment (high pressure cycling) was carried out until the E. coli was subjected to 80 min of pressure. During the test, 300 pl samples of the broth were taken at time intervals of 5, 10, 20, 40 and 80 min.

2.2. Isolation and culture of mar&e stra&s Marine agar (MA) (Difco, Detroit, USA) and marine broth (MB) (Difco) were prepared in accordance with the manufacturers instructions. The 50% MA and 20% MA contained 18.7 and 7.6 g of Difco Marine Broth 2216 per litre of distilled water, respectively (Boyd et al., 1998; MearnsSpragg et al., 1997, 1998). All media were autoclaved prior to use (121 ~ for 15 rain). Specimens of the seaweed Rhodymenia palmata were collected from sites on the South East Coast of Scotland. Portions of the plant were rinsed with sterile sea-water (2 x 10 ml) and a small area was swabbed with a sterile cotton-tipped swab. The swab was then used to directly inoculate plates. For each specimen a range of media were used for isolation (Boyd et al., 1998). The plates were then incubated at room temperature. Colonies were removed and sub-cultured when the plates showed good growth. Colonies were picked on the basis of colony morphology so as to maximise the diversity of strains isolated. Subculturing provided pure isolates that were stored as stabs in MA. These strains were Rhodymenia epiphytes and designated surface strains since they were associated with surface biofilms. Other strains were isolated from water (1-5 m depth) which were not associated with surfaces. G strains were all recovered from sea-water (G548, G538, G51, G55, and G518), whereas KBRP strains were removed from the seaweed biofilms (KBRP1, KBRP4, KPRP16 and KBRP13). All strains used are currently unidentified, however, each strain number does represent a species obtained from a specific ecological niche. Their natural distribution in the environment has not been investigated. Marine strains were inoculated and grown for 5 days, at 25~ in 5 ml of marine broth, with

346


shaking. The same experimental technique was used for the marine strains as for the E. coli, except that the samples were spread on a marine agar plate. The samples were then incubated for 1 day at 25~ High pressure suspended cultures proceeded in exactly the same manner as the E. coli.

2.3. Population determination The 300 ~tl pressure-treated samples of either the marine strains or E. coli were diluted to give a range of dilutions down to 10-9 cells/ml. Subsequently, 100 lal samples of these dilutions were then spread onto marine agar plates and incubated at 37~ for 24 h. After the incubation period the colonies on the plates were counted. Non-pressure treated samples of identically inoculated solutions were tested in exactly the same manner to provide a standard. A minimum of two tests were performed for each process condition, thus ensuring reproducibility.


3.1. Barotolerance of marine strains The behaviour of liquid supported marine strains subjected to high pressure is presented in Fig. l(a) and (b). (a) Depicts strains obtained from suspended sea-water (strains G55, G518, G538, G548 and G51). (b) Depicts strains obtained from seaweed surfaces (strains KBRP1, KBRP13, KBRP4 and KBRP16). It is interesting to observe that barotolerance is greater in those bacteria obtained from the biofilms on the surface of seaweed, rather than those obtained from open water. This is particularly interesting because these normally surface associated strains were exposed to the pressure cycling whilst suspended in culture, and not attached to a surface. The major exception to this observation is strain G538, which still showed strong survival even after 80 min of exposure (see Fig. la). Strain G55 has been removed from (a) as it showed no survival at the 5 min sample time.

3.2. Barotolerance of Escherichia coli (EMG-23) The results for the liquid-supported E. coli (Fig. 2) show that E. coli appears to be more pressure tolerant than the marine strains. After 20 rain of exposure to 120 MPa, it can be seen that there has been no significant decrease in the number of colonies initially present. This trend continued up to the maximum pressurised time tested, 80 rain. This was not entirely unexpected, as other strains of E. coli (LMM 1010, LMM 1020 and LMM 1030) have shown barotolerant behaviour at pressures over 220 MPa (Hauben et al., 1997). An obvious implication from this demonstration of barotolerance is the use of this variant of E. coli for expression of foreign proteins at high pressure.

3.3. Implications for process intensified bioreactor design The findings in this paper are important from the standpoint of effective bioprocess engineering because so many of the tested strains are barotolerant. In many reaction systems the application of pressure may lead to enhanced yields of target metabolites within the framework of Le Chfttelier's principle, i.e. high pressure will favour an equilibrium state that results in a negative volume change. In addition, the findings presented here form the first stage of an approach to utilising pressure as a stressing agent. This knowledge can then be coupled to reactor design to lead to an improved, and more intensive, bioprocess. As a lead on to a future bioprocess, strain KBRP1, as shown in Fig. l(b), is barotolerant and it has also produced antimicrobial compounds during bench scale operation. For a significant proportion of bioprocessing operations, the use of barotolerant rather than barophilic microorganisms may be preferred, because isolation and enrichment of barophilic cultures is much more difficult in an engineering sense, as they must remain at these significantly high-pressure levels ( > 50 MPa; Kato et al., 1998).


4. Conclusions A number of techniques are being explored as possible routes to bioprocess intensification. To date, most of the techniques fall into the category

347

of alterations to pure transport properties, such as increased turbulence or mass transfer enhancement via bubble size reduction. In this paper, it is discussed that more emphasis could be put onto examination of stress induction as a route to a

100 90 80 70 ~

60

p.

50 40

o~

30 20 10 0

2.5

(a)

I

i

i

,

,

7.5

10

12.5

15

17.5

20

Pressudsed Time (minutes) _s

G518

- B - G548

- x - G51

._,_ G538

lOO[ 9o 8070m to

"~ o~

0

-

50 40 30 20 lOo

i

o

2.5

'

,

'

'

i

5

'

'

'

'

i

7.5

12.5

10

15

17.5

20

Pressurised Time (minutes)

(b)

=

KBRP1

-b

KBRP4

,J

KBRP16

- - o - - KBRP13

Fig. l. Behaviour of marine strains under cyclic high-pressure (120 MPa)" (a) sea-water strains; (b) solid surface associated strains.

348

P.C. Wright et al. /Journal of Biotechnology 70 (1999) 343-349

100

..

~ ' Q , w

9 m l

Lj--_ ~ _ _ . , . . ,

i

9

i ....Q.,,.

--L

=I

"" 4

]

90

80 70 u

. . M

60 50 40

30 20 10 i

L

2.5

5

7.5

~

,.

L

12.5

10

L

,

i

i

i

15

,

~

17.5

i

2O

Pressurised Time (minutes) . ...

Test 1

---o-Test

2

Fig. 2. Behaviour of E. coli (EMG-23) under cyclic high-pressure (120 MPa).

more 'intensive' bioprocess. In addition, in line with predictions from Le Chfitelier's principle, bioprocess improvements may be achievable in reaction systems having a negative volume change. The ultimate aim is enhanced bioproduct generation. Manipulation of barotolerant/barophilic properties was suggested as a strong example of this methodology. As demonstrated by the example microorganisms presented here, a number of barotolerant species may be found at non-extremophile conditions. From our sample group it can be seen that surface strains are more barotolerant, even when grown in suspensions, than strains obtained from open water. As a lead on to a future bioprocess, strain KBRP1 has also produced antimicrobial compounds and is being investigated as a potential model species.

Acknowledgements Phillip Wright would like to thank Pfizer Ltd for provision of an Academic Travel Award, The

British Council (Amsterdam) and the N W O (Netherlands Organisation for Scientific Research) for provision of a fellowship to The University of Groningen, as well as The Royal Society for financial support. Colin Stevenson would like to acknowledge The Department of Mechanical and Chemical Engineering (HWU) for the provision of a Vacation Scholarship. Grant Burgess thanks the Natural Environment Research Council (NERC) and The British Council (Tokyo) for financial support.

References Athes, V., Degraeve, P., Cavaill6-Lefebvre, D., Espeillac, S., Lemay. P., Combes, D., 1997. Increased thermostability of three mesophilic b-galactosidases under high pressure. Biotechnol. Lett. 19 (3), 273-276. Boyd, K.G., Mearns-Spragg, A., Brindley, G., Hatzidimitrou, Rennie, A., Bregu, M., Hubble, M.O., Burgess, J.G., 1998. Antifouling potential of epiphytic marine bacteria from the surfaces of marine algae. In: Le Gal, Y., Mueller-Fuega, A. (Eds.), Marine Microorganisms for Industry. Editions IFREMER (Institut Frangais de Recherche pour l'Ex-

P.C. Wright et al./Journal of Biotechnology 70 (1999) 343-349 ploitation de la Mer). Plouzane, France, ISBN 2-90543494-5, pp. 128-136 Brower, V., (August) 1998. Nutraceuticals: Poised for a healthy slice of the healthcare market? Nature Biotech 16, 728-731. Chisti, Y., Moo-Young, M., 1996. Bioprocess intensification through bioreactor engineering. Trans. IChemE. 74 (A), 575-583. Cowan, D.A., 1998. Hot bugs, cold bugs and sushi. TIBTECH 16 (6), 241-242. DeLong, E.F., 1997. Marine microbial diversity: the tip of the iceberg. TIBTECH 15 (6), 203-207. Fenical, W., 1997. New pharmaceuticals from marine organisms. TIBTECH 15 (9), 339-341. Hauben, K.J.A., Bartlett, D.H., Soontjens, C.C.F., Cornelis, K., Wuytack, E.Y., Michiels, C.W., 1997. Escherichia coli mutants resistant to inactivation by high hydrostatic pressure. Appl. Environ. Microbiol. 63 (3), 945-950. Kato, C., Li, L., Nogi, Y., Nakamura, Y., Tamaoka, J., Horikoshi, K., 1998. Extremely barophilic bacteria isolated from the Mariana Trench, challenger deep, at a depth of 11 000 meters. Appl. Environ. Microbiol. 64 (4), 15101513. Mearns-Sprang, A., Boyd, K.G., Hubble, M.O., Burgess, J.G., 1997. Antibiotics from surface associated marine bacteria. Fourth Underw. Sci. Symp. Society for Underwater Technology, London, ISBN 0 906940 31 1, 147-157.

349

Mearns-Spragg, A., Bregu, M., Boyd, K.G., Burgess, J.G., 1998. Cross-species induction and enhancement of antibiotic production by epiphytic bacteria from marine algae and invertebrates after exposure to terrestrial bacteria. Lett. Appl. Microbiol. 27, 142-146. Michels, P.C., Clark, D.S., 1997. Pressure-enhanced activity and stability of a hyperthermophilic protease from a deepsea methanogen. Appl. Environ. Microbiol. 63 (10), 39853991. Mozhaev, V.V., Heremans, K., Frank, J., Masson, P., Balny, C., 1994. Exploiting the effects of high hydrostatic pressure in biotechnological applications. TIBTECH 12 (12), 493501. Nelson, C.M., Schuppenhauer, R., Clark, D.S., 1992. Highpressure high temperature bioreactor for comparing effects of hyperbaric and hydrostatic pressure on bacterial growth. Appl. Environ. Microbiol. 58 (5), 1789-1793. Hill, S., 12 April 1997. Squeezing the death out of food. New Sci. 1997, 29-32. Pothakamury, U.R., Barbosa-C~inovas, G.V., Swanson, B.G., Meyer, R.S., 1995. The pressure builds for better food processing. Chem. Eng. Prog. March, 45-53. Prieur, D., 1997. Microbiology of deep-sea hydrothermal vents. TIBTECH 15 (7), 242-244. Wright, P.C., Raper, J.A., 1996. A review of some parameters involved in fluidized bed bioreactors. Chem. Eng. Technol. 19 (1), 50-64. Yayanos, A.A., 1995. Microbiology to 10500 meters in the deep sea. Annu. Rev. Microbiol. 49, 777-805.


iii

Jl

| O U I N A L

i

OF



Effect of light-path length in outdoor fiat plate reactors on output rate of cell mass and of EPA in Nannochloropsis sp. Ning Zou, Amos Richmond * Microalgal Biotechnology Laboratory, The Jacob Blaustein Institute for Desert Research, Ben-Gurion University of the Negev, Sede-Boker Campus 84993, Israel Received 12 October 1998; received in revised form 4 December 1998; accepted 22 December 1998

Abstract The effect of light-path length (i.e. reactor width or thickness) of flat plate glass reactors on outdoor production of eicosapentaenoic acid (EPA) and cell mass of Nannochloropsis sp. was tested, using a range of light-paths from 1.3 to 17.0 cm. Volumetric productivity of cell mass and optimal, as well as maximal cell density which represents the highest sustainable cell density under the experimental conditions, decreased with increase in light-path. Daily areal output rate (g dry weight m - 2 d a y - ~) increased with increased light-path, in contrast with results obtained in similar reactors with Spirulina cultures, in which areal output rates increased when the light-path was reduced. Maximal areal productivity of Nannochloropsis sp. (12.8 and 22.4 g ash-free dry weight per day per m 2 of irradiated reactor surfaces, in winter and summer, respectively), reflecting maximal efficiency in light utilization, was obtained with the long light-paths, i.e. 10.4 and 17.0 cm. Increasing the light-path from 1.3 to 17.0 cm resulted in an increase in areal EPA productivity, from 66.7 to 278.2 mg m -2 day -~ in winter and from 232.1 to 515.7 mg m -2 day -~ in summer. This enhancement in areal productivity of EPA stems from increased productivity of cell mass which was associated with the increase in light-path. We concluded that the optimal light-path, which must be defined for each algal species, represents an important parameter which determines optimal culture density (i.e. resulting in the highest output rate of cell mass per irradiated reactor surface), as well as productivity of cell mass and cell products. Under our conditions the optimal light-path for culturing Nannochloropsis in vertical reactors was ca 10 cm. 9 1999 Elsevier Science B.V. All rights reserved.

Keywords: EPA; Light-path; Nannochloropsis sp.; Outdoor cultures; Photobioreactor; Productivity

1. Introduction Light limitation to growth c a n n o t be described solely in terms o f the light flux impinging on the reactor surface, n o r in terms o f the average light * Corresponding author.

flUX available for each cell. Particularly in ultrahigh density cultures, the light regime is the m o s t significant rate-limiting factor of photoautotrophic productivity of cell mass ( H u et al., 1996b, 1998). This s o m e w h a t elusive p a r a m e t e r concerns to overall characteristics involved in the cells' exposure to light, as reflected in light inter-


352

N. Zou, A. Richmond/Journal of Biotechnology 70 (1999) 351-356

mittence due to mutual shading as well as in the intensity and duration of the light flashes, shaping together a dominant factor of the light regime, i.e. the characteristics of the light-dark ( L - D ) cycle to which the cells are exposed, in moving back and forth from the dark to the lit volumes in the reactor (Hu et al., 1998). Together with the population density and the rate of mixing, the length of the light-path (i.e. the width of a flat plate reactor) represents an essential parameter in mass cultures, affecting the light regime through its effect on the L - D cycle. The role of the light-path in optimizing reactor performance was recognized by Hu et al. (1996a), who studied the detailed interactions between optimal cell concentration and light-path length in Spirulina platensis cultures grown in flat plate reactors. In the framework of our efforts to optimize biotechnological aspects involved in mass production of Nannochloropsis, we attempted to delineate the optimal light-path in flat plate reactors for maximal areal output rates of cell mass and of EPA.


.

Culture control: pH was monitored by a microprocessor pH meter (WTW). Dissolved oxygen (DO) was monitored by YSI Model 58. Light intensity was measured with a quantum sensor Li-Cor model Li-185A. Temperature was maintained at 29 + 2~ using tap water spray for evaporative cooling on the reactor front and back panels, pH was maintained between 7.2 and 8.0 by adding 1% CO2 into the compressed air stream used for mixing the culture. Flow rate of the CO2-enriched air was 0.84 1 min-1 1-1 culture.

2.2. EPA analysis The pellet obtained by centrifuging the culture at 2000 rpm for 15 rain was freeze dried (Cohen, 1994). To 25 mg freeze-dried pellet were added: (a) 0.25 mg C17 standard solution (Sigma) and (b) 2 ml H2SO4:methanol solution (2% H2SO4). After scrubbing with Ar gas, the sample was stirred and heated to 80~ for 1 h. Subsequently, 1 ml H20 (to stop the reaction) and 1 ml hexane were added, mixing well with vortex. Centrifuging at 3500 rpm for 5 min resulted in two layers, the upper layer filtered through glass fiber and dried with N2. The sample was redissolved in 100 ~tl hexane and finally chromatographed, using 8590 gas chromatography (GC).

2. I. Growth conditions 1. Organism: Nannochloropsis sp. obtained through the courtesy of Mr Odi Zemora, Oceanographic Inst., Eilat. 2. Reactors: flat plate glass reactors (Hu et al., 1996a), with the following light-paths: 1.3, 2.6, 5.2, 10.4 to 17.0 cm. The total irradiated surfaces (i.e. the front and back panels) of all reactors were identical, 0.52 m 2. 3. Growth medium: unbuffered artificial sea water medium, as follows: 2.7% NaC1, 6.6 g 1MgSO4.7H20, 5.6 g 1-1 MgClz.6H20 , 1.5 g 1-1 CaClz'2H20, 1.45 g 1-1 KNO3, 0.12 g I-1 KHzPO4, 0.04 g 1-1 NaHCO3, 0.01 g 1-1 FeC13.6H20, 0.078 g 1-~ Na2-EDTA, 0.01 mg l-1 CuSO4.5HzO ' 0.022 mg 1-~ ZnSO4"7H20, 0.01 mg 1-~ COC12"6H20, 0.18 mg 1-~ MnC12-4H20, 0.006 mg 1-1 Na2MoOa'2H20.

2.3. Determination of cell mass and cell concentration Cell mass was estimated by measuring: (a) total organic carbon (TOC) in the pellet, obtained by 3500 rpm centrifugation of a culture sample for 6 min, and/or (b) by determining ash free dry weight (AFDW). The latter method was used in winter, the former having been used in summer. The ratio of TOC/AFDW was measured several times in duplicates for different reactors throughout the summer, and it ranged between 0.49 and 0.59 (0.04 standard deviation). TOC was thus multiplied by 1.89 to convert it to AFDW. TOC measurement: samples were diluted according the cell density (to about 1 x l0 s cells m l - ~). One ml diluted sample was centrifuged at ca 3500 rpm for 6 min, adding thereafter HgSO4


353

centrations were maintained. Cultures with cell concentrations which yielded, at steady state, maximal output rates of cell mass were thereby identified as optimal.

(0.2 g), 1 ml DDW, 1 ml H2SO4, 0.5 ml AgzSO4, and 1 ml 0.25 N K2Cr20 7 to the pellet suspension. Another 1 ml H z S O 4 w a s then added to the test tube, which was incubated at 150~ in an oven for 1 h. When cooled, the mixture was poured into flasks for titration. Three drops of Ferroin solution (1/40 mol/1 redox indicator) were added and the sample was finally titrated with 0.1 N Fe(NH4)2(SO4)2"6H20. Dry weight was measured using preweighted Whatman GF/C (0.45 ~tm) 47-ram ~ glass fiber filters. Cells were washed twice with acidified water (pH 4.0) followed with distilled water. The filtered cells were dried at 105~ overnight. Ash content in these samples was ca 6% of the total dry weight (Hu and Richmond, 1994). Cell count was determined using a 0.0025 mm 2 cytometer, with a light microscope ( x 128 magnification). Optimal cell density for each reactor with a given light-path was determined in continuous cultures along which several steady-state cell con-


The length of the light-path exerted a strong effect on the optimal cell density, defined as that cell concentration which results in the highest areal output rate of cell mass (g A F D W m - 2 day-~) (Fig. 1). As expected in a light-limited system, the shorter the light-path, the higher the optimal cell density became. Through the year, however, no striking differences in optimal cell density were observed in a given light-path, except for the 1.3-cm reactor. In the latter the optimal cell density in summer was distinctly the highest of the year (Fig. 1). We suggest this very significant increase in optimal cell density observed for the 1.3-cm reactor only was due to light intensity:

16

~-~

10

u

2O

(cm)

Fig. 2. Areal vs. volumetric productivity of cell mass (in summer) as affected by the light-path. (e) Areal productivity; ( 9 volumetric productivity.

in the very short light-path reactor, it represented, in summer, photoinhibitory radiation for Nannochloropsis sp. This light stress could be eliminated only by increasing areal cell density (reducing thereby the intensity of the light climate for the single cell) associated with longer lightpaths. This effect was elucidated for Spirulina platensis cultures which yielded increased productivity upon exposure to very high radiation levels, provided cell density was increased appropriately, in adjustment to increased radiation (Hu et al., 1997). A similar surge in optimal cell density was not observed in the other reactors with longer light-paths, because the D - L cycle frequency (Hu et al., 1998) in the longer light-paths was obviously lower, whereas areal cell number was higher, compared with the 1.3-cm reactor. Cells in the longer light-path reactors were thus exposed to lesser radiation dose, which was not photoinhibitory. The light-path greatly affected the productivity of cell mass (Fig. 2). Volumetric productivity increased ca 7-fold with a 13-fold decrease in lightpath (from 17.0 to 1.3 cm). In contrast, areal productivity (which relates to the irradiated panels of the reactors measuring 0.52 rn2), sharply increased with the increase in light-path, reaching its peak with the 10.4-cm reactor and becoming slightly lower as the light-path was further increased to 17.0 cm (Fig. 2). Clearly, the high

radiation existing in summer (e.g. ca 1800-2100 ~tE m - 2 S-1 for some 5 h at midday) was best utilized in reactors with the longer light-path. These reactors exhibited on the one hand a larger number of cells per irradiated surface (Fig. 3) and on the other hand affected a decrease in the frequency of the L - D cycle (Zou, 1996; Hu et al., 1998; Tredici and Zittelli, 1998). Thus for the slow growing Nannochloropsis cells, the light regime prevailing in association with the more narrow light-paths (to which Spirulina cultures responded well by enhanced productivity (Hu et al., 1996a; Richmond, 1996)), could not be effectively used by Nannochloropsis. This species reached peak areal productivity, reflecting highest net photosynthetic efficiency, only when light per cell was in effect reduced, as a result of being both distributed to a larger number of cells per given irradiated area (cells m - 2 ) and by being exposed to a lower L - D cycle frequency existing in long light-path reactors, compared with the narrower light-paths. The great increase in volumetric productivity caused by reduction of the light-path (Fig. 2) was due to a surge in the growth rate which resulted from reduction in the extent of light-limitation, an advantage of short light-paths. Nevertheless, the overall lower productivity of the short light-path reactors indicated a less efficient use of radiation, 250 (45.0)

E

200

E3 LL < 150 r

(3.4)

10

,--.T....-

8

1oo

"~ E ..~

50

(27.1) 1 " (13.6)

O

1.3

2.6

5.2 Light

path

10.4

17.0

(cm)

Fig. 3. Effect of light-path length on optimal areal cell density in summer. Numbers in brackets, liters of culture volume in the reactor, all reactors having identical areas of irradiated surface.


355

7.5 a) U

U

-

ca.

o ,m.t"-

~

+ -1.0025X

R= 0.97155

l|

I

5.5

.,.., t" r

._~ 9

= 10.356

LL 6

,

y

6.5

a

131 U.
, presents the unique peculiar property among the diatoms, to produce at its extremities a blue hydrosoluble pigment called >. It is presented the concentration and the desalting of the exocellular pigment by membrane processes (ultrafiltration, nanofiltration, reverse osmosis). Nanofiltration is particularly developed given the potential of this type of application both for the concentration of molecules and for desalting. It is shown the effect of velocity and pressure on performances of nanofiltration membranes. Permeation flux superior to 100 1 h - l m - 2 (at 14.105 Pa) are obtained with the Kiryat Weizmann membrane MP 20 (polyester coated with a polyacrylonitrile layer, cut-off 450 Da). For the desalting of the blue pigment solution, nanofiltration membranes present a few advantages: a low salt rejection (less than 10% at 14.105 Pa) and a high pigment rejection (the nanofiltration membrane MP 20 retains more than 95% of the pigment). This membrane used in diafiltration mode allows an acceptable speed of desalting (700 g of salt eliminated per hour and per m 2 at 25.105 Pa for a concentration of 18 g of salt per litre of solution). 9 1999 Elsevier Science B.V. All rights reserved. Keywords: Concentration; Desalting; Haslea ostrearia; Nanofiltration; Pigment

I. Introduction

* Corresponding author. Tel.: + 33-2-97874531; fax: + 332-97874588. E-mail addresses: [email protected] (L. Vandanjon), [email protected] (P. Jaouen), [email protected] (J.-M. Robert)

The aim of the study is to develop the industrial application of the marine d i a t o m H a s l e a o s t r e a r i a Simonsen, also called >, responsible for the greening of the oysters in France. This microalga presents the unique peculiar prop-


394

L. Vandanjon et al. / Journal of Biotechnology 70 (1999) 393-402

erty among the diatoms, to produce at its extremities a blue hydrosoluble pigment called > (Robert, 1983). Ultrastructurally, the diatom shows marked changes depending on the stage of cell blueing by accumulation of marennine (Nassiri et al., 1998). The interest of this pigment is notably linked to:

9 its colouring properties: the pigment might be used as a new blue natural dye in the food and agriculture fields, 9 its ability for the greening of oysters: the only current industrial application is at the SOPROMA Company in Bouin (Vend6e, Atlantic coast, France), where the production of the diatom is performed in batch cultures, using tanks of 6 m 3 in volume under carefully-controlled incubation conditions, 9 its antiproliferative properties since aqueous extracts of the pigment show inhibitory effects both in vitro and in vivo against solid carcinoma lines (Belt et al., 1997) In this study, it is presented the concentration and the desalting of the exocellular pigment by membrane processes: performances of ultrafiltration, nanofiltration, reverse osmosis membranes are compared. Nanofiltration is particularly developed given the potential of this type of application both for the concentration of molecules and for desalting.

250 ml erlenmeyer flasks filled with 150 ml of ES 1/3 medium (first preculture). The temperature of incubation was 16~ and the light intensity was 3 x 1016 quanta cm -2 s -1 with a 14/10 h light/ dark cycle. The erlenmeyer flask contents were then inoculated into 25 1 flasks containing 20 1 of ES 1//3 medium (second preculture). After incubation of these larger flasks for ca 7 to 10 days in identical temperature and light conditions, their contents were transfered into 500 1 tanks filled with 400 1 underground sea-water from the Bouin district (Rouillard, 1996). In these natural incubation conditions, cell suspensions with concentration ranging between 60 and 100 x 103 cells m l - ~ were obtained after incubation for 5-12 days. The algal mass collected by centrifugation, with a continuous centrifuger CEPA-PATBERG, LE (clarification cylinder type K, inox), was then collected and the blue coloured medium containing the exocellular marennine was recuperated for concentration and desalting by membrane processes.

2.2. Experimental set-up and membranes The experiments are carried out on two pilote plants: Millipore (Prolab Bench Top) and Gamma Filtration (Microlab 80S) built on the same principle (Fig. 1). The both pilote plants are constituted from a centrifugal pump allowing liquid recirculation and


2.1. Biological material The study was performed using an axenic strain of H. ostrearia isolated in the Marine Biology Laboratory from the oyster-pond waters of the Bouin district (Vend6e, France). The clone cells were characterized by an average model length of 65 gm. The algal cultures were maintained by weekly transfer to fresh ES 1/3 medium (Robert, 1983; Lebeau et al., 1999). Algal precultures were precultivated by applying a two steps procedure, before to be used for mass production of the diatom. Cells from the clone pool were first precultured for ca 7 days in

Permeate

[Heatexchanger Throttling valve

Flowmeter

V

Outlet pressure

Filtration module

Feeding tank Recirculating pump

z_._..x

Feedingpump

z_._.a

Inlet pressure Flow control valve

Fig. 1. Schematic representation of the pilote plant.

L. Vandanjon et al. Journal of Biotechnology 70 (1999) 393-402

395

Table 1 Characteristics of membranes used Trade Mark

Manufacturer

Geometry

Technique

Cut-off

Area (m 2)

Material

Iris 3028

Tech-Sep, Or61is Millipore Kiryat Weizmann Kiryat Weizmann

Flat

UF

3 kDa

6.9 x 10- 3

Polyethersulfone

Spiral Tubular

RO NF

reverse osmosis (RO) membrane and two nanofiltration (NF) membranes. These membranes and their characteristics are presented in the Table 1. The UF and N F experiments are carried out by total recycling of the retentate and the permeate. Steady flux and retention rates of the membranes are measured after 120 min. Concentration of the blue pigment solution is carried out with a RO membrane. Working pressure is 41 • 105 Pa and temperature is kept constant at 25~ The concentrated solution is desalted by diafiltration using the same > RO membrane in a continuous mode (Fig. 2).

Qv

"

QR

Qv : Volumetric flow of permeate

The concentration of pigment in solution is determinated by spectrophotometry (Secomam SG 1000). Optical density is measured at 663 nm, which corresponds to the maximum absorbance of the marennine (Robert and Hallet, 1981). Salinity is determinated by measuring resistivity (Tacussel CD 60) of the marennine solution.

3. Resuits~discussion

The pigment > seems to be composed of a mix of macromolecules of different sizes. Size repartition of pigment molecules has been estimated by using six UF membranes (bearing negligible adsorption rate) of different molecular cut-off comprised between 1 and 300 kDa (Vandanjon, 1997). By relying on the model proposed by Ferry (1936), we have shown that a large part of the pigment has a molecular weight comprised between 3 and 7 kDa (Fig. 3). So the ultrafiltration (low cut-off) could be convenient a technique for the concentration and the desalting of marennine solution. High purity molecules can be obtained by using multistage diafiltration (Muller, 1996). However, the existence of a small fraction of pigment whose molecular weight is close to hundreds Dalton involves the use of nanofiltration or reverse osmosis membranes for the recovery of the totality of the marennine.

(= volumetric flow of solvent) QR : Recirculating flow

Fig. 2. Diafiltration in a continuous mode.

3.1. Performance of ultrafiltration membrane (low cut-off) The operating conditions are as following:

396

L. Vandanjon et al. / Journal of Biotechnology 70 (1999) 393-402 %

70 61%

60 50 40 30 17%

20 10

8%

5%

5%

4%

[ 0,85

3,3

7

20

J

35

Molecular weight (kDa) Fig. 3. Molecular weight repartition of marennine molecules. 9

Pressure: 4 x 105 Pa Tangential velocity: 1 m s - ] Results are presented on Fig. 4. The ultrafiltration membrane 3028 3 kDa presents a high rejection percentage for the pigment (90%) and a low rejection percentage for the salts. This membrane could be convenient for the desalting of pigment solution by diafiltration. 9

3.2. Performance of > reverse osmosis membrane

3.2.2. Comments 9 Steady flux of 34 1 m - 2 h-~ is obtained after 26 h. 9 Pigment is highly concentrated while salt concentration is constant. 9 There is a gap between the VRF (VRF = 70) and the CF of pigment (CF = 50). It can be explained by the loose of pigment in the permeate and the fixation of pigment on the membrane separators.

400

3.2.1. Concentration of the pigment solution In order to obtain a high volumic reduction factor (VRF), a volume (Vo) of 350 1 of culture medium containing the hydrosoluble pigment is concentrated. The (culture medium) presents the following characteristics: 9 Optical density (at 663 nm): 0.024 (length of the optical path = 1 cm) 9 Salt concentration: 30 g 1-1 9 pH8 COncentration is achieved with the > reverse osmosis membrane Millipore R45P. Fig. 5 presents the evolution of the permeate flux and the concentration factor (CF) of pigment and salts.

--

100

Pigment rejection eq , cO

3= ._~

200

50

--

.2 u.

.~, O

Salt rejection 0

--"

'

-

J

loo

i

!

~

20O

Time (min) Fig. 4. Evolution of permeate flow and rejection percentage for an ultrafiltration membrane (cut-off 3 kDa).

L. Vandanjon et al.

Journal

o/"

50~

Biotechnology 70 (1999) 3 9 3 - 4 0 2

- -9 50

o~

397

P=41.105Pa T = 25 ~

~0

L__ 40

E

t_

"7 ,.I:7, ~ , 30---

A

/

jzx

~,

20--

,/

1,-, O

/ r--30

Lx

VRF = Vo/Vf o .,.a r

ao~176

CF = Cf/Co

.1=

,~igment

,---- 2 0

V o : Initial volume of blue water

r O

Vf: Final volume of blue water

L)

Co : Concentration at the initial time '---- 1 0 o/

l'-

Salts 0

c

!

c

Cf: Concentration at the end time

I 0

9 q,

I

1--o

10

20

30

Time (h) Fig. 5. Concentration of marennine with a reverse osmosis membrane. 3.2.3. C o n t i n u o u s

diafiltration

Diafiltration with the membrane R45P is achieved at a constant volume (5 1) and a constant pressure (10 x 105 Pa). Fig. 6 presents the evolution of the salt concentration and the rejection rates for salts and pigment in function of time and added pure water volume. It can be noticed that the salts are not totally eliminated after 13 h of diafiltration. Experimental curves are compared to the theoritical model developed in diafiltration (Mulder, 1991). After integrating the overall mass balance sheet:

(m 3 s - 1); TR(%) = 100" [1 -(Cp/C0)], salt rejection rate: and V0, initial volume of blue water (m3). If the salt rejection rate is constant with time, diafiltration experimental curves are close to the theoritical model. In the present case, the gap between theoritical curve and experimental curve is due to a salt rejection rate which is not perfectly constant (evolution of the osmotic pressure during diafiltration, adsorption phenomenom). However, it is possible to estimate the number of diavolumes (DV) necessary for desalting the solution with the relation (Tutunjian, 1985): In

dC -

(1)

Vo " --d~ = Qv " C p

DV =

Co TR

the following equation is obtained: C = C0 9e x p [ -

Qv .(1 - T R / I O 0 ) .

100 t/Vo]

(2)

with: t, filtration time (s); C, salt concentration in the coloured water at the time t (kg m - 3); Co, salt concentration in the coloured water at the initial time (kg m - 3 ) ; Cp, salt concentration in the permeate (kg m - 3 ) ; Qv, volumetric flow of permeate

Cf

(3)

1

with G, salt concentration in the coloured water at the end time (kg m - 3 ) . With the > reverse osmosis membrane (salt rejection rate -~ 60%), it is possible to know the number of diavolumes for desalting from 15 to 1 g l - I :

L. Vandanjon et al. Journal of Biotechnology 70 (1999) 393-402

398

are presented on Fig. 7 (membranes MPT 20 and MPT 31). With the low pressures (AP < 105 Pa), filtrate flux increases linearly and does not depend from the recirculation velocity (Darcy's law). Then flux increases slowly (primary polarization), but it is noticeable that it is strongly influenced by recirculation velocity (Brun, 1989). For these nanofiltration membranes, it can be considered that beyond 20 x 105-25 x 105 Pa, increase of the pressure has no more significant influence on the filtrate flux. However, fluxes superior to 50 1 m -2 h-~ at 30 x 105 Pa can be considered as acceptable performances. The ultrafiltrate fluxes versus tangential velocity are often represented according to the relation (Qu6m6neur and Schlumpf, 1980):

DV = In(1/15)/(60 - 100). 100 = 6.8 On the experimental curve (Fig. 6), it is necessary to add 35 1 of pure water to 5 1 of for desalting the solution, i.e: DV = 35/ 5=7. These results confirm that the model is correct whenever the salt rejection rate is constant.

3.3. Performance of nanofiltration membranes 3.3.1. Permeation flux A few sets of experiments by total recycling of the retentate and the permeate are carried out at different pressures and velocities. For each couple pressure-velocity, the stabilization of the flux is reached after about 2 h. These points allow to draw steady flux versus pressure at different tangential velocities. Results 1 6 -

a

='~

--~_

J = :~ 9u"

. . . . . . .

_~ _:

,_.

(4) :.

--

-

z

~

-

100

Pigment rejection "7

12

o

o

v

o .E

8--

~

0

~

~

-50

O ~ v,,,~ ,4-,,)

o O eO

perimental

_

Theoritical "".........~ model 0

'"

'

I-

'

4

!

'

I 12

8

Time (h) 0 0

~

~

l

7

.

.

.

.

.

14

21 |

28 i

,

35 I

42 l

....

49 :

Added pure water (litres) Fig. 6. D i a f i l t r a t i o n o f m a r e n n i n e s o l u t i o n with a > reverse o s m o s i s m e m b r a n e .

L. Vandanjon et al. , Journal o f Biotechnology 70 (1999) 3 9 3 - 4 0 2 40-I

~'

~

Pure water flux

9 - eo-

- - -o --o

399 9 ....

.

.

.

.

~1~-100

_...Q

Pigment 1 O0 ~

. - - - - ,~.,.~.~

,

I

?'E

,

i

50 ~

_ u = 2 m/s

,/

'

~

,,'"

.'" . ~ J m . .

i

Re = 29,000

o. Membrane MPT20 _.,. Membrane MPT31

u = 1 rrds Re = 14,500

//"'-.~,...__---.z-u

.' ~

,

,--

2o

MembraneMPT 31

....

: _ ~

= 0 5 m/s

-

Re = 7,250

-,.- -- "-.'~=ln~s . . . .

o .,..,

u = 0 5 m/s

--50

20--

j ~ 01..o.

r~

o

I 0

6

lO

20

3o

ot D

~

0

(D

-

Salts

__6.__.o_. . . . . . . . . . .

E

Q....

0

Pressure (10 5 Pa)

0 Fig. 7. Steady flux versus pressure for nanofiltration membranes.

J, ultrafiltrate f l u x ( m 3 m - 2 s - ] ) z(, 1l, constants; and u, fluid velocity (m s-1). In order to verify the validity of the Eq. (4) in nanofiltration, it is drawn on Fig. 8 the logarithmic permeate flux versus the logarithmic tangential velocity under a pressure of 25 x 105 Pa. For each membrane, a straight line is obtained with:

4.B

M

r.-

4.4 ---=

MPT 20 + 4.2--

4.0

.

I -0.8

0"

.

'

.

.

I -0.4

3.8-

.

.

10

20

30

Pressure (105 Pa) Fig. 9. Evolution of salt and pigment rejections with the pressure.

whose equation follows: In J = In z~+ n 9In u

(5)

In turbulent flow, the following values of the slope n can be determined graphically: Membrane MPT 20: n = 0.13 Membrane MPT 31: n = 0.18 These values are inferior to the most common values of the literature: n = 0.69-0.91 (Blatt et al., 1970; Zaitoun, 1979; Aimar, 1992). Generally, the coefficient n depends on hydrodynamic characteristics of the module geometry (flat, tubular, spiral etc.) and of the couple membrane-foulant (Gekas and Hallstr6m, 1987; Nabetani et al., 1990). But the very low concentration of natural pigment and the high salinity of the marennine solution may be responsible for these results.

3.3.2. Selectivio' of nanofiltration membranes

--

~ / / M P M P T

31

.

'

,,,

' 0.0 In U

I 0.4

'

I 0.8

Fig. 8. Experimental validation of the relation J--z~. u" in nanofiltration. Graphic determination of the exponent n.

From a practical or economical point of view, it is interesting to carry out desalting by diafiltration after concentration without any changing of material (pilot-plant and membrane). The > membrane should retain the maximum of pigment and be as permeable as possible for the salts. Fig. 9 describes performance, in term of salt and pigment rejection, of two nanofiltration membranes MPT 20 and MPT 31.

400

L. Vandanjon et al./Journal of Biotechnology 70 (1999) 393-402

The membrane MPT 20 retains the fast totality of the pigment and its permeability for the salts remains high at high pressure. Rejection rate (TR) seems to be independent of the pressure, it would correspond to a mass transfer mechanism intermediate between capillar type mechanism (TR decreases with pressure) and diffusional type mechanism (TR increases with pressure). On the contrary, the nanofiltration membrane MPT 31 presents a rejection rate that increases with the applied pressure. This behaviour is characteristic of a diffusional type mechanism: TR(%) = 100.

1 - A . ( A P - Arc)

(6)

with: A, solvent permeability of the membrane; and B, solute (salts) permeability of the membrane. With this type of nanofiltration membrane, close to reverse osmosis membranes, an increase of the pressure induces an increase of the permeate flux and the salt rejection rate; concurrently, it is observed a light decrease of the pigment rejection. That means that high permeate flows during concentration will induce loss of pigment and will penalize the next step of desalting. So it will be necessary to find a compromise in the aim to optimize the overall process of concentrationdesalting.

The most efficient membrane for desalting is the ultrafiltration membrane 3028 3 kDa. Salts can easily pass through the membrane but the low pigment retention makes the membrane unusable for a diafiltration. At the opposite, the membrane R45P retains perfectly the pigment but its high salt rejection induces a long and difficult desalting. The nanofiltration membrane MPT 20 constitutes the best compromise in term of desalting and pigment rejection and it may be used efficiently in diafiltration mode.

4. Conclusion

Nanofiltration seems to be the best performing technique for desalting the blue coloured solution. The model used in diafiltration with a > reverse osmosis membrane may be used with nanofiltration membranes, whether salt rejection rate is about constant. In these conditions, it is possible to calculate, from laboratory scale experiments, performances of pilot-plant or industrial process. In this study, we have shown the effect of velocity and pressure on performance of nanofiltration membranes. Flows superior to 50 1 m - 2 h-1 (at 30 x 105 Pa) are obtained with the Weizmann membrane MPT 20 (in polyacryloni-

3.4. Comparison of membranes for diafiltration Diafiltration is carried out at a constant volume: Vo = 5 1. Three membranes are tested functionning in diafiltration mode: 9 an ultrafiltration membrane Or61is-Iris 3028 3 kDaatAP-4x 10SPa, 9 a nanofiltration membrane Kyriat Weizmann MPT 20 at A P - 2 5 x 105 Pa, 9 a reverse osmosis membrane Millipore R45P at A P - 10 x 105 Pa. With the aim to evaluate the efficiency of the three membranes (ultrafiltration, nanofiltration and reverse osmosis) to carry out a diafiltration, it has been drawn on Fig. 10 the graphic of desalting speed versus salts concentration in the retentate.

: ~,

1000

-

-

E ,s::

3028 3 kDa (4.105 Pa) MPT20 450 Da (25.105 Pa) R45P (10.10 s Pa) /

"7

UF low c ~ / ~ j

500 - -.=

.

NF

10

15

{/) r

0 9

-

0

5

20

Salt concentration in the retentate (g.1 l ) Fig. 10. Compared efficiency of the membranes for the elimination of the salts.

L. Vandanjon et al. / Journal of Biotechnology 70 (1999) 393-402

trile coated on polyester, cut-off 450 Da). Higher fluxes could be obtained by optimizing hydrodynamic conditions or the couple tangential velocity/pressure. For the desalting of the blue-green pigment solution, nanofiltration membranes present a few advantages over loose reverse osmosis and ultrafiltration: a low salt rejection rate (less than 10% at 14 bars) and a high pigment rejection (the nanofiltration membrane MPT 20 retains the fast totality of the pigment). This membrane used in diafiltration mode allows to obtain an acceptable speed of desalting (for example, 700 g of salt eliminated per hour and per m 2 at 25 x 105 Pa for a concentration of 18 g of salt by litre of solution). So the nanofiltration membrane MPT 20 is interesting in the aim of a simultaneous concentration-desalting (without changing of processunit or membrane) of the marennine solution produced by H. ostrearia. The diafiltration does not allow to eliminate entirely the salts, but it does not constitute an obstacle in the aim of food enhancement of the marennine. Indeed, legislation makes difficult the commercialization of new natural dyestuffs (Labatut and In, 1990). In these conditions, non purified marennine should preferently be enhanced as a food ingredient. However, a high degree of purity is necessary for the use of marennine for therapeutical application in cancerology for example (Riou et al., 1993). The dialysis technique is then well suited for the final elimination of the salts when using small volumes of concentrated pigment solution. One further development of this study is the membrane photobioreactor which combines biological reaction and separation. Biological reaction: the research work consists in optimizing culture conditions of H. ostrearia for the highest pigment production: choice of the clones, study of the shear effects caused by pumps and valves in a recirculating loop (Vandanjon et al., 1999), addition of nutriments. Separation" for the extraction of pigment, two stages of membranes are necessary. Ultrafiltration (cut-off 40 kDa) can be used for the extraction of the exometabolite from the culture in a continuous mode (Rossignol et al., 1999). Then after-

401

wards, the marennine solution is concentrated and desalted by nanofiltration. References Aimar, P., 1992. Limiting flux in membrane separations: a model based on the viscosity dependency of the mass transfer coefficient. Chem. Eng. Sci. 47 (3), 579-586. Belt S.T., Robert J.-M., Roussakis C., Rowland S., 1997. Chemotherapeutic compounds from microalgae. Patent n~ ISOMer (University of Nantes, Nantes, France) and University of Plymouth, Plymouth, UK. Blatt, W.F., Dravid, A., Michaels, A.S., Nelsen, L., 1970. Solute polarization and cake formation in membrane ultrafiltration. In: Membrane Science and Technology, vol. 47. Plenum Press, New York. Brun J.-P., 1989. Proc6d6s de s6paration par membranes. Masson, Paris, 270 p, ISBN 2-225-81573-9. Ferry, J.D., 1936. Ultrafilter membranes and ultrafiltration. Chem. Rev. 18, 373-455. Gekas, V., Hallstr6m, B., 1987. Critical literature review and adaptation of existing Sherwood correlations to membrane operations. J. Membr. Sci. 30, 153-170. Labatut, M.-L., In, T.. 1990. La couleur au naturel. Biofutur November, 37-42. Lebeau, T., Junter, G.-A., Jouenne, T., Robert, J.-M., 1999. Marennine production by agar-entrapped Haslea ostrearia Simonsen. Biores. Technol. 67 (1), 13-17. Mulder, M., 1991. Basic Principles of Membrane Technology. Kluwer, Dordrecht. Muller A., 1996. Proc6d6 d'obtention d'~-lactalbumine de haute puret6: 6tapes 616mentaires du fractionnement des prot6ines de lactoserum et mise en cascade. Ph.D. thesis, ENSAR Rennes, France. Nabetani, H., Nakajima, M., Watanabe, A., Nakao, S.I., Kimura, S., 1990. Effect of osmotic pressure ans adsorption on ultrafiltration of ovalbumine. AIChE J. 36 (6), 907-915. Nassiri, Y., Robert, J.-M., Rinc6, Y., Ginsburger-Vogel, T., 1998. The cytoplasmic fine structure of the diatom Haslea ostrearm (Bacillariophyceae) in relation to marennine production. Phycologia 37 (2), 84-91. Qu~m~neur. F., Schlumpf, J.-P., 1980. Traitement des huiles solubles par ultrafiltration. Entropie 93, 22-29. Riou, D., Roussakis, C., Biard, J.-F., Verbist, J.-F., 1993. Comparative study of the antitumour activity of bistramides A, D and K against a non-small cell broncho-pulmonary carcinoma. Anticancer Res. 13, 2331-2334. Robert J.-M., 1983. Fertilit6 des eaux de claires ostr6icoles et verdissement: utilisation de l'azote par les diatom6es dominantes. Ph.D. thesis, Nantes University, Nantes, France, 281 p. Robert, J.-M., Hallet, J.-N., 1981. Absorption spectrum in vivo of the blue pigment > of the pennate diatom Nat'icula ostrearia Bory. J. Exp. Botany 32 (127), 341-345.

402

L. Vandanjon et al./Journal of Biotechnolog)" 70 (I 999) 393-402

Rossignol N., Vandanjon L., Jaouen P., Qu6meneur F., 1999. Membrane technology for microalgae harvesting: compared performances of cross-flow microfiltration and ultrafiltration. Aquacult. Eng. (submitted for publication). Rouillard I., 1996. Optimisation de la production en masse de Haslea ostrearia Simonsen sur une eau souterraine sal6e: importance de la souche et des conditions de culture: comparaison avec Skeletonema costatum (Grev) Cleve. Ph.D. thesis, Nantes University, Nantes, France, 270 p. Tutunjian, R.S., 1985. Scale-up considerations for membrane processes. Biotechnology 3, 615-626.

Vandanjon k., 1997. Etude d'un proc6d6 de valorisation d'une microalgue marine: concentration et purification par techniques '/i membranes d'un pigment naturel produit par la diatom6e Haslea ostrearia. Ph.D. thesis, Nantes University, Nantes, France, 257 p. Vandanjon, L., Rossignol, N., Jaouen, P., Robert, J.-M., Qu6m6neur, F., 1999. Effects of shear on two microalgae species. Contribution of pumps and valves in tangential flow filtration systems. Biotechnol. Bioeng. 63 (1), 1-9. Zaitoun A., 1979. Osmose inverse et ultrafiltration en milieu organique. Equations de transport. Application ~ l'ultrafiltration des huiles moteur. Ph.D. thesis, INP Lorraine, France.

JOURNAL

OF



Marine bioprocess engineering" the missing link to commercialization Oskar R. Zaborsky Department of Chemistry, College of Natural Sciences, University of Hawaii, 2525 Correa Road--HIG 131, Honolulu, HI 96822, USA

Received 21 November 1998; received in revised form 7 December 1998: accepted 22 December 1998

Abstract

Success of US biotechnology has been and continues to be dependent on new discoveries and their timely transformation into useful products through bioprocess engineering and a systems approach. Bioprocess engineering is an essential element of 'generic applied' or 'precompetitive' research. For marine biotechnology, like biopharmaceutical biotechnology, bioprocess engineering represents the key. The many hundreds of tantalizing bioactive compounds discovered and isolated from varied marine organisms over the past decades have led to only minimal commercialization due to the limited availability of the compounds in question. To address international competitiveness and the revitalization of key US industries, the National Science Foundation launched the Engineering Research Centers Program in the mid 1980s. The essential feature of this program is a partnership among academia, industry and the government to develop next-generation technology through cutting-edge research, relevant education and innovative technology transfer. MarBEC (Marine Bioproducts Engineering Center) is a recently established multi-disciplinary engineering-science cooperative effort of the University of Hawaii and the University of California at Berkeley. Additional partners include three federal laboratories--Argonne National Laboratory, the Edgewood Research, Development and Engineering Center and the Eastern Regional Research Center of the US Department of Agriculture--and the Bishop Museum. MarBEC's research program consists of four major thrusts: Production Systems; Marine Bioproducts and Bioresources; Separation and Conversion; and Bioproduct Formulation. 9 1999 Published by Elsevier Science B.V. All rights reserved. Keywords: Commercialization; Marine bioprocess engineering; Missing link

I. Introduction

Biotechnology is a critical high technology recognized by many countries worldwide including the United States, Japan, Germany, England, France and China. While there are still some skeptics about biotechnology's delivery of prod-

UCtS in relation to its promises, significant success stories are already on hand especially in the biopharmaceutical sector. More than 60 biopharmaceutical products have reached the US market in 16 years and worldwide product sales were $13 billion in 1997 (Thayer, 1998a). Three drugs are above $1 billion in annual sales and another seven

0168-1656/99/$ - see front matter 9 1999 Published by Elsevier Science B.V. All rights reserved. PII: S0168-1656(99)00093-0

404

O.R. Zaborsky /Journal of Biotechnology 70 (1999) 403-408

exceed $250 million. Indeed, without modern biotechnology there would be no life-saving drugs such as erythropoietin, inteferon, colony stimulating factor, glucocerebrosidase, dornase alpha or tissue plasminogen activator. Further, recombinant vaccines and monoclonal antibodies are prevalent among products currently available or in development. In total, more than 350 biotechnology drugs are in development, with the leading targets being cancer, infectious diseases, AIDSHIV related diseases, heart disease and neurologic disorders. Indeed, biotechnology, a 'dream machine' for US entrepreneurs, is now poised to explode into reality not only in the biomedical arena but also agriculture and the chemical industry. In fact, some major chemical firms are becoming life sciences companies through biotechnology (Thayer, 1998b).

2. Marine biotechnology Marine biotechnology, i.e. the application of modern biotechnology to marine organisms and processes, is an area of significant industrial importance whose ramifications will reach almost every major industrial sector including health, environment, energy, food (aquaculture and agriculture), chemicals, advanced materials and national defense (Bioscience, 1996; Oceanography, 1996; Attaway, 1997; Attaway and Zaborsky, 1993; Zaborsky, 1993). In part, the varied nature of areas impacted by marine biotechnology makes it difficult to track and appreciate its full ramifications. Marine biotechnology, like general biotechnology, will no doubt have its most immediate impact in the industrial sectors of drugs, industrial specialty bioproducts and foods/feeds, especially aquaculture. The importance of marine biotechnology has been recognized by a number of organizations including the National Science and Technology Council (NSTC)--Office of Science and Technology Policy (NSTC, 1995), the National Research Council (NRC) of the National Academy of Sciences and National Academy of Engineering (NRC, 1985) and the World Bank (Zilinskas and Lundin, 1993). The NSTC concluded that marine

biotechnology is one of four 'second wave' biotechnology areas that deserves priority attention from federal agencies. In particular, the priorities included: 9 Develop fundamental understanding of genetic, nutritional and environmental factors that control production of primary and secondary metabolites in marine organisms, the basis for new and improved products; 9 Identify bioactive compounds and determine mechanisms of action and function and provide models for new lines of active materials for application in medicine and chemical industry; 9 Develop bioremediation strategies for application in coastal oceans, where multiple uses (including wastewater disposal, recreation, fishing and aquaculture) demand prevention and remediation of pollution; develop bioprocess strategies for sustainable industrial processes; 9 Use tools of modern biotechnology to improve health, reproduction, development, growth and overall well-being of cultivated aquatic organisms; promote interdisciplinary development of environmentally sensitive, sustainable systems that will enable significant commercialization of aquaculture; and 9 Improve understanding of microbial physiology, genetics, biochemistry and ecology to provide model systems for research and production systems for commerce and contribute to understanding and conservation of the seas. Recent activities in marine biotechnology related to drug discovery have been described (Cragg et al., 1997; Fenical, 1997; Wallace, 1997). Activities in Hawaii have focused on aquaculture, specialty bioproducts and bioactive agents effective against cancer and infectious diseases. Most noteworthy developments have included the isolation and synthesis of cyptophycins (a potentially effective class of anti-cancer agents) by Moore and Patterson (Smith et al., 1994), the production of nutritional supplements such as Spirulina pacifica and also astaxanthin by the Cyanotech Corporation and the culturing of varied marine organisms for food applications by firms located at Kona on the Big Island. In terms of research,


many efforts are underway including those of our group in the production of biological hydrogen (Zaborsky, 1998), a collaborative effort with Professor Tredici at Florence employing his specially designed tubular photobioreactors (Szyper et al., 1998) and marine-derived enzymes. Our research group has just established a major culture collection, the Hawaii Culture Collection that houses the former Mitsui-Miami and NREL-Golden collections (http://www.hawaii.edu/hicc). Additionally, the University of Hawaii has exceptional marine and ocean research vessel operations for acquiring even more biological resources from extreme environments such Loihi, an active undersea volcano 20 miles off the Big Island that in time will be the newest island in the Hawaiian archipelago.

3. Bioprocess engineering m the critical link

One of the other three biotechnology areas identified, namely manufacturing/bioprocessing, represents bioprocess engineering as had been also advocated by a NRC study (NRC, 1992). Success of US biotechnology has been and continues to be dependent on new discoveries and their timely transformation into useful products through bioprocess engineering and a systems approach. Without doubt, bioprocess engineering has been the key to success in the commercialization of biotechnology in biopharmaceuticals. Bioprocess engineering, of course, is an essential element of 'generic applied' or 'precompetitive' research (as it is now more commonly referred to), i.e. research that is beyond basic science but that is before development and traditional engineering applications. Generic applied research represents research that addresses generic issues or bottlenecks that go beyond the means of any one company or even any one industry. The production technology for recombinant proteins, the separation of chiral bioproducts or the precise delivery of bioproducts are some examples. While generic applied research or precompetitive research has been supported through a number of programs and special initiatives by the US federal government throughout the past two

405

decades, these efforts have been very vulnerable to the whims of political changes at the presidential or congressional levels. One of the very noted early programs was the Research Applied to National Needs (RANN) program at the National Science Foundation (NSF) in the mid 1970s. This program was dedicated to the application of science and engineering research to national needs such as energy, materials and chemicals, the environment and human safety. Current generic applied research programs addressing key technology areas include the Department of Commerce's Advanced Technology Program (ATP), including its catalysis and biocatalysis thrusts and the now ubiquitous Small Business Research Innovation Research (SBIR) Program that advances technological innovation and commercialization by providing phased funding to small businesses in a number of areas including biotechnology and marine biotechnology. Generic applied research has been recognized for its importance in translating basic research to commercial products but also its vulnerability (Zaborsky, 1984, 1985). To address international competitiveness and the revitalization of key US industries, NSF launched the Engineering Research Centers (ERC) Program in the mid 1980s. The essential feature of this program is a partnership among academia, industry and the government to develop next-generation technology through cuttingedge research, relevant education and innovative technology transfer. A recent analysis of this program has shown it to be quite positive in achieving its objectives (Parker, 1997). For marine biotechnology, like biopharmaceutical biotechnology, bioprocess engineering represents the key from discovery to commercialization. As an example, the many hundreds of tantalizing bioactive compounds discovered and isolated from varied marine organisms by many over the past decades have led to only minimal commercialization due to the limited availability of the compounds in question for clinical trials or further modification by chemical or biological means--a limit to production caused by our limited marine bioprocess engineering skills (Rouhi, 1995).

406

O.R. Zaborsky Journal of Biotechnology 70 (1999) 403-408

4. New opportunities in marine biotechnology: marine bioprocess engineering Several years ago at the Edinburgh, UK and Greifswald, Germany meetings, this author advocated three major areas of attention and opportunity for marine biotechnology. These were: 9 Marine Bioprocess Engineering Project--with the objective of developing reliable bioprocess technologies for marine organisms (microalgae) and marine-derived bioproducts. This also included efforts to establish an international 'marine biosolar' facilities network, with a demonstration facility in Hawaii. 9 Industrial Marine Biodiversity Project--with the objective being to establish an inventory of marine biodiversity (especially the neglected microbes) and crystallize current efforts through an international network of collaboration. 9 Marine Genome Project--with the objective of mapping and sequencing commercially and scientifically significant marine genomes. As with other genome projects, the objective included organizing an international effort. While some activities have been initiated in all three areas, the one of highest priority has been the pursuit of marine bioprocess engineering through a dedicated center that is now named the Marine Bioproducts Engineering Center (MarBEC).

5. MarBEC MarBEC is a multi-disciplinary engineering-science cooperative effort of the University of Hawaii at Manoa, the lead organization and the University of California at Berkeley. Additional partners include three federal laboratories--Argonne National Laboratory, the Edgewood Research, Development and Engineering Center and the Eastern Regional Research Center of the US Department of Agriculture--and the Bishop Museum. MarBEC was officially established on 2 November 1998 and will be a $4-7 million per annum operation once fully functional. NSF funding is $12.4 million for a 5-year period. How-

ever, the full life time of such a center with NSF funding is expected to be 10 years. MarBEC is also expected to become self-sufficient in support. MarBEC's mission is to develop the engineering technology and science base for producing highvalue marine bioproducts essential to the chemical, pharmaceutical, nutraceutical and life sciences industries. Equally important, MarBEC's mission is to produce a new cadre of engineers--i.e. marine biotechnology engineers who will be essential not only to industry but also to government and academia. In essence, MarBEC's mission entails advanced manufacturing of bioproducts that are dependent on marine organisms, solar energy and the technological base offered by modern biotechnology. MarBEC's bioproducts of interest all have multiple uses and multi-million or even billion dollar markets, with significant growth potential. These include carotenoids, polyunsaturated fatty acids, enzymes and bioactive agents. While the importance of marine biotechnology had been recognized, the engineering component had not been. Quite frankly, it is very fragmented throughout academic institutions in the US Equally, education in marine biotechnology engineering is basically non-existent, with no single curriculum dedicated to engineering aspects of marine biotechnology existing at any major US university. Key engineering challenges include: (a) production systems--especially the design, development, modeling and evaluation of sustainable production systems employing phototrophic organisms, development of metabolic engineering and immobilized cell and enzyme bioreactors, environmental aspects of large-scale photobioreactors and near-shore and open-ocean production systems; (b) separation and conversion--more cost-effective separation technologies for labile and chiral bioproducts; and (c) bioproduct formulation, especially in reference to stability and functionality. MarBEC's research program consists of four major thrusts: Production Systems; Marine Bioproducts and Bioresources; Separation and Conversion; and Bioproduct Formulation. These four thrusts are integrated, focusing on engineering yet capturing relevant marine and biological sciences.

O.R. Zaborskv ' Journal o f Biotechnology 70 (1999) 403-408

Each research thrust has near-, mid- and longterm areas of interest to industry and participating faculty. A crucial point, however, is that these four interrelated thrusts are components of a matrix focused on economically significant marine bioproducts. The research conducted will be through multi-disciplinary teams employing the latest advances in science and engineering. Value-added benefits include: a team-work environment, the ability to address cross-cutting engineering issues, a systems approach, the opportunity to tackle long-term engineering challenges and the integration of key disciplines in engineering and marine/biological sciences. Most importantly, MarBEC has a 'critical mass' of expertise that can be called upon to tackle the major barriers or pursue opportunities in a timely fashion. The goal of the education program is to produce the next generation engineer at various levels, especially at the graduate, undergraduate level and post-graduate (industrial) levels. At the undergraduate level, this will be accomplished by introducing marine biotechnology engineering concepts and research project results in appropriate courses at both universities. At the graduate level, the education goal will be accomplished by offering specific courses and specialized short courses dealing with advanced topics. Furthermore, graduate and undergraduate students will be actively engaged in research projects. At the post-graduate level, summer institutes geared to industrial participants will be held. At all levels, research and education will be linked. MarBEC has an industrial collaboration program which aims to establish an effective partnership with industrial firms located in Hawaii, the US mainland and leading technology countries in the Pacific Rim and Europe. The strategy is to have a constant dialogue with industry at every level and opportunity at the research project, research program and center level. MarBEC's outreach program is designed to influence research and education in marine bioproducts engineering in Hawaii, the US mainland and elsewhere where appropriate. As such, MarBEC will work with other organizations to achieve this goal in a bi-directional mode; actions

407

and benefits must flow in both directions. MarBEC's strategy is to work at different levels with organizations within the state, the nation and the world. Important outreach programs are the Marine Options Program and the Sea Grant College Programs of both Hawaii and California. MarBEC is also reaching out to international research organizations. Already in existence are collaborative efforts with organizations in Japan (Tokyo University of Agriculture and Technology) and Europe (University of Florence). As with other ERCs, MarBEC is guided by an Advisory Board consisting of distinguished members of academia, government and industry who can provide counsel on its future direction. MarBEC also has an Industrial Research Advisory Board (IRAB) consisting of experts in the relevant fields of research from member companies. IRAB, in concert with both universities, will establish the formal policies and operating guidelines for industrial participation. A special feature of MarBEC's technology transfer program is its strong linkage to key Hawaii business, financial and economic development organizations as well as to the technology transfer programs of the federal laboratories that will provide a multiplier effect to MarBEC's efforts. In summary, MarBEC builds upon past accomplishments and current research efforts at both universities, yet forges into uncharted areas of immense scientific, engineering and business opportunities of benefit to many.

6. Conclusions

Bioprocess engineering is the critical link that transforms basic research discoveries into commercial reality. Of course, many other aspects come into play when success is achieved such as a market need and timing. As with biopharmaceutical biotechnology, however, bioprocess engineering skills are the critical edge in marine biotechnology. MarBEC represents a new paradigm for research, education and technology transfer in marine biotechnology and a framework for economic development by working with industry in

408


Hawaii, the US mainland, Europe and Asia. MarBEC represents a window to 21st century opportunities of the Pacific--but more importantly a launching pad for real action.

References Attaway, D.A., 1997. A Report on Marine Biotechnology in the National Sea Grant College Program, Washington, DC. Attaway, D.A., Zaborsky O.R. (Eds.), 1993. Marine Biotechnology, Pharmaceutical and Bioactive Natural Products, vol. 1. Plenum, New York. Bioscience, 1996. Marine biotechnology, special issue, 46 (4). Cragg, G.M., Newman, D.J., Weiss, R.B., 1997. Coral reefs, forests and thermal vents: the worldwide exploration of nature for novel antitumor agents. Semin. Oncol. 24, 156163. Fenical, W., 1997. New pharmaceuticals from marine organisms. Trends Biotechnol. 15, 339-341. NRC-1985, 1985. National Research Council--National Academy of Sciences, Marine Biotechnology: Basic Research Relevant to Biomaterials and Biosensors. NRC-1992, 1992. National Research Council--National Academy of Sciences--National Academy of Engineering, Bioprocess Engineering: Putting Biotechnology to Work. NSTC-1995, 1995. National Science and Technology Council-Office of Science and Technology Policy, Biotechnology for the 21st Century: New Horizons. Washington, DC. Oceanography, 1996. Marine Biological Diversity Special Issue, 9 (1).

Parker, L., 1997. The Engineering Research Centers (ERC) Program: An Assessment of Benefits and Outcomes. National Science Foundation, Washington, DC. Rouhi, A.M., 1995. Supply issues complicate trek of chemicals from sea to market. C&EN 11 (20), 42-44. Smith, C.D., Zhang, X., Mooberry, S.L., Patterson, G.M., Moore, R.E., 1994. Cryptophycin: a new antimicrotubule agent active against drug-resistant cells. Cancer Res 54, 3779-3784. Szyper, J.P., Yoza, B.A., Benemann, J.R., Tredici, M.R., Zaborsky, O.R., 1998. Internal gas exchange photobioreactor: development and testing in Hawaii. In: Zaborsky O.R., Benemann J.R., Miyake J., Matsunaga T., San Pietro, A. (Eds.), BioHydrogen. Plenum, New York (in press). Thayer, A.M., 1998a. Great expectations. C&EN 8 (10), 1931. Thayer, A.M., 1998b. Living and loving life sciences. C&EN 11 (23), 17-24. Wallace, R.W., 1997. Drugs from the sea: harvesting the results of aeons of chemical evolution. Mol Med Today 3, 291-295. Zaborsky, O.R., 1984. A critique of government funding of biotechnology: basic and applied research. Chem Econ. Eng. Rev. 16, 9-10. Zaborsky, O.R., 1985. Technology development: the missing link. Visions 2, 6-8. Zaborsky, O.R., 1993. Marine biotechnology. Genet. Eng. News 13, 10. Zaborsky, O.R., Benemann J.R., Miyake J., Matsunaga T., San Pietro A. (Eds.), 1998. BioHydrogen. Plenum, New York. Zilinskas, R.A., Lundin C.G., 1993. Marine Biotechnology and Developing Countries. World Bank Discussion Papers, Washington, DC.

JOURNAL

OF



Author Index Aci6n Fern~ndez, F.G., see Rebolloso Fuentes, M.M. (70) 271 Armstrong, E.D., McKenzie, J. and Goldsworthy, G.T., Aquaculture of sponges on scallops for natural products research and antifouling (70) 163 Asada, Y., see Miyake, J. (70) 89 Asian, K., see Eroglu, i. (70) 103 Baquerisse, D., Nouals, S., Isambert, A., dos Santos, P.F. and Durand, G., Modelling of a continuous pilot photobioreactor for microalgae production (70) 335 Bastianini, A., see Chini Zittelli, G. (70) 299 Battershill, C.N., see Munro, M.H.G. (70) 15 Beelen, T.P.M., see Vrieling, E.G. (70) 39 Bengoa-Ruigomez, M.V., see Kotzabasis, K. (70) 357 Blunt, J.W., see Munro, M.H.G. (70) 15 Borghi, A., see Sponga, F. (70) 65 Borowitzka, M.A., Commercial production of microalgae: ponds, tanks, tubes and fermenters (70) 313 Bowles, R.D., Hunt, A.E., Bremer, G.B., Duchars, M.G. and Eaton, R.A., Long-chain n - 3 polyunsaturated fatty acid production by members of the marine protistan group the thraustochytrids: screening of isolates and optimisation of docosahexaenoic acid production (70) 193 Boyd, K.G., see Burgess, J.G. (70) 27 Bregu, M., see Burgess, J.G. (70) 27 Bremer, G.B., see Bowles, R.D. (70) 193 Burgess, J.G., Jordan, E.M., Bregu, M., Mearns-Spragg, A. and Boyd, K.G., Microbial antagonism: a neglected avenue of natural products research (70) 27 Burgess, J.G., see Wright, P.C. (70) 343 Cabatingan, L.K., see van der Wielen, L.A.M. (70) 363 Camacho P~ez, B., see Robles Medina, A. (70) 379 Cavaletti, L., see Sponga, F. (70) 65 Chini Zittelli, G., Lavista, F., Bastianini, A., Rodolfi, L., Vincenzini, M. and Tredici, M.R., Production of eicosapentaenoic acid by Nannochloropsis sp. cultures in outdoor tubular photobioreactors (70) 299 Chisti, Y., see Molina Grima, E. (70) 231 Chisti, Y., see S~.nchez Mir6n, A. (70) 249 Ciciliato, I., see Sponga, F. (70) 65 Contreras G6mez, A., see S~.nchez Mir6n, A. (70) 249 Cs6g6r, Z., Melgar, D., Schmidt, K. and Posten, C., Production and particle characterization of the frustules of Cyclotella cryptica in comparison with siliceous earth (70) 71 PII: S 0 1 6 8 - 1 6 5 6 ( 9 9 ) 0 0 1 4 5 - 5

de Beukelaer, P.B., see Osinga, R. (70) 155 de Rijk, T.C., see de Swaaf, M.E. (70) 185 de Swaaf, M.E., de Rijk, T.C., Eggink, G. and Sijtsma, L., Optimisation of docosahexaenoic acid production in batch cultivations by Crypthecodinium cohnii (70) 185 Divanach, P., see Kotzabasis, K. (70) 357 dos Santos, P.F., see Baquerisse, D. (70) 335 Duchars, M.G., see Bowles, R.D. (70) 193 Duckworth, A.R., see Munro, M.H.G. (70) 15 Dumdei, E.J., see Munro, M.H.G. (70) 15 Durand, G., see Baquerisse, D. (70) 335 Eaton, R.A., see Bowles, R.D. (70) 193 Eggink, G., see de Swaaf, M.E. (70) 185 Erard-Le Denn, E., see La Barre, S. (70) 207 Eroglu, I., Aslan, K., Gfindfiz, U., Yficel, M. and Tfirker, L., Substrate consumption rates for hydrogen production by Rhodobacter sphaeroides in a column photobioreactor (70) 103 Eroglu, |., see Sediroglu, V. (70) 115 Ero(glu, |., see Yi~it, D.O. (70) 125 Esteban Cerdfin, L., see Robles Medina, A. (70) 379 Etoundi, P., see Helmholz, H. (70) 203 Fern~ndez, F.G.A., see Molina Grima, E. (70) 231 Fern~ndez Sevilla, J.M., see Rebolloso Fuentes, M.M. (70) 271 Gad, G., see Liebezeit, G. (70) 77 Garcia Camacho, F., see Molina Grima, E. (70) 231 Garcia Camacho, F., see Shnchez Mir6n, A. (70) 249 Garcia Sanchez, J.L., see Rebolloso Fuentes, M.M. (70) 271 Garcia-Jimenez, P., Marian, F.D., Rodrigo, M. and Robaina, R.R., Sporulation and sterilization method for axenic culture of Gelidium canariensis (70) 227 German, J.B., see Wood, B.J.B. (70) 175 Gieskes, W.W.C., see Vrieling, E.G. (70) 39 Gim6nez Gim6nez, A., see Robles Medina, A. (70) 379 Goldsworthy, G.T., see Armstrong, E. (70) 163 Grimson, P.H.K., see Wood, B.J.B. (70) 175 Gromek, E., see Turkiewicz, M. (70) 53 Gfindiiz, U., see Eroglu, I. (70) 103 G(indiiz, U., see Sediroglu, V. (70) 115 Gfindfiz, U., see Yi~it, D.O. (70) 125

410

Author Index

Hatziathanasiou, A., see Kotzabasis, K. (70) 357 Helmholz, H., Etoundi, P. and Lindequist, U., Cultivation of the marine basidiomycete Nia vibrissa (Moore & Meyers) (70) 203 Hickford, S.J.H., see Munro, M.H.G. (70) 15 Hirano, S., Nakahira, T., Nakagawa, M. and Kim, S.K.. The preparation and applications of functional fibres from crab shell chitin (70) 373 Hunt, A.E., see Bowles, R.D. (70) 193 Ibfifiez Gonzfilez, M.J., see Robles Medina. A. (70) 379 Isambert, A., see Baquerisse, D. (70) 335 Janssen, M., Kuijpers, T.C., Veldhoen, B., Ternbach. M.B.. Tramper, J., Mur, L.R. and Wijffels, R.H., Specific growth rate of Chlamydomonas reinhardtii and Chlorelhi sorokiniami under medium duration light/dark cycles: 13-87 s (70) 323 Jaouen, P., see Vandanjon, L. (70) 393 Jordan, E.M., see Burgess, J.G. (70) 27 Jozefowicz, M., see La Barre, S. (70) 207 Kalinowska, H., see Turkiewicz, M. (70) 53 Kentouri, M., see Kotzabasis, K. (70) 357 Kim, S.K., see Hirano, S. (70) 373 Kotzabasis, K., Hatziathanasiou, A., Bengoa-Ruigomez. M.V., Kentouri, M. and Divanach, P., Methanol as alternative carbon source for quicker efficient production of the microalgae Chlorella minutissima: Role of the concentration and frequence of administration (70) 357 Koulman, A., Pruijn, L.M.C., Sandstra, T.S.A., Woerdenbag. H.J. and Pras, N., The pharmaceutical exploration of cold water ascidians from the Netherlands: a possible source of new cytotoxic natural products (70) 85 Kreitlow, S., Mundt, S. and Lindequist, U.. Cyanobacteria--a potential source of new biologically active substances (70) 61 Kuijpers, T.C., see Janssen, M. (70) 323 Kiinnemann, T.D., see Liebezeit, G. (70) 77 La Barre, S., Singer, S., Erard-Le Denn, E. and Jozel'oxvicz. M., Controlled cultivation of Alexan~h'ium J~lmututll and [33p] orthophosphate cell labeling towards surface adhesion tests (70) 207 Lavista, F., see Chini Zittelli, G. (70) 299 Lazzarini, A., see Sponga, F. (70) 65 Liebezeit, G., Kiinnemann, T.D. and Gad. G., Biotechnological potential of North Sea salt marsh plants--a review of traditional knowledge (70) 77 Lill, R.E., see Munro, M.H.G. (70) 15 Lindequist, U., see Helmholz, H. (70) 203 Lindequist, U., see Kreitlow, S. (70) 61 Li, S., see Munro, M.H.G. (70) 15 Losi, D., see Sponga, F. (70) 65 Marian, F.D., see Garcia-Jimenez, P. (70) 227 Marinelli, F., see Sponga, F. (70) 65 Matsunaga, T., Takeyama, H., Nakao, T. and Yamazawa, A..

Screening of marine microalgae for bioremediation of cadmium-polluted seawater (70) 33 McEvoy. E., see Wright, P.C. (70) 343 McKenzie, J.D.. see Armstrong, E. (70) 163 Mearns-Spragg, A.. see Burgess, J.G. (70) 27 Meijer, E.M., see Osinga, R. (70) 155 Melgar. D.. see Cs6g6r, Z. (70) 71 Miyake. J.. Miyake. M. and Asada, Y., Biotechnological hydrogen production: research for efficient light energy conversion (70) 89 Miyake. M.. see Miyake, J. (70) 89 Molina Grima. E., Fernfindez, F.G.A., Garcia Camacho, F. and Chisti. Y.. Photobioreactors: light regime, mass transfer. and scaleup (70) 231 Molina Grima. E.. see Rebolloso Fuentes, M.M. (70) 271 Molina Grima. E.. see Robles Medina, A. (70) 379 Molina Grima. E., see Sfinchez Mirdn, A. (70) 249 Mundt. S.. see Kreitlow. S. (70) 61 Munro. M.H.G.. Blunt, J.W., Dumdei, E.J., Hickford, S.J.H., kill. R.E.. ki. S.. Battershill, C.N. and Duckworth, A.R., The discovery and development of marine compounds with pharmaceutical potential (70) 15 Mur. L.R.. see Janssen. M. (70) 323 Nakaga~a. M., see Hirano, S. (70) 373 Nakahira. T.. see Hirano, S. (70) 373 Nakao, T., see Matsunaga, T. (70) 33 Nouals. S.. see Baquerisse, D. (70) 335 Ogbonna. J.C.. Soejima, T. and Tanaka, H., An integrated solar and artificial light system for internal illumination of photobioreactors (70) 289 Ogbonna. J.C.. Tomivama. S. and Tanaka, H., Production of ~.-tocopherol bx sequential heterotrophic-photoautotrophic cultivation of Euelemi gracilis (70) 213 Ogi. T.. see Tsukahara. K. (70) 223 Osinga. R.. de Beukelaer. P.B., Meijer, E.M., Tramper, J. and Wijffeis. R.H.. Grov~th of the sponge Pseudosuberites (aft.) amtrc~.~i in a closed system (70) 155 Pomponi. S.A.. The bioprocess-technological potential of the sea t70) 5 Posten. C.. see Cs6g6r. Z. (70) 71 Pras. N.. see Koulman. A. (70) 85 Pruijn. L.M.C.. see Koulman, A. (70) 85 Quemeneur. F.. see Vandanjon, L. (70) 393 Rebolloso Fuentes. M.M.. Garcia Sfinchez, J.L., Fern/mdez Sevilla. J.M.. Acien Fernfindez, F.G., Sfinchez P6rez, J.A. and Molina Grima. E., Outdoor continuous culture of Pop7~l:.vridiul~ ~ruentum in a tubular photobioreactor: quantitative analysis of the daily cyclic variation of culture parameters (70) 271 Richmond. A.. see Zou. N. (70) 351 Rinkevich. B.. Cell cultures from marine invertebrates: obstacles. new approaches and recent improvements (70) 133 Robaina. R.R.. see Garcia-Jimenez, P. (70) 227

Author hldex Robert, J.-M., see Vandanjon, L. (70) 393 Robles Medina, A., Esteban Cerdfin, L., Gimenez Gim6nez, A., Camacho Pfiez, B., Ibfifiez Gonzfilez, M.J. and Molina Grima, E., Lipase-catalyzed esterification of glycerol and polyunsaturated fatty acids from fish and microalgae oils (70) 379 Rodolfi, L., see Chini Zittelli, G. (70) 299 Rodrigo, M., see Garcia-Jimenez, P. (70) 227 Rossignol, N., see Vandanjon, L. (70) 393 S/mchez Mir6n, A., Contreras G6mez, A., Garcia Camacho, F., Molina Grima, E. and Chisti, Y., Comparative evaluation of compact photobioreactors for large-scale monoculture of microalgae (70) 249 Sfinchez Perez, J.A., see Rebolloso Fuentes, M.M. (70) 271 Sandstra, T.S.A., see Koulman, A. (70) 85 Sawayama, S., see Tsukahara, K. (70) 223 Schmidt, K., see Cs6g6r, Z. (70) 71 Sediroglu, V., Eroglu, |., Yticel, M., Tfirker, L. and Giindfiz, U., The biocatalytic effect of Halobacterium halobium on photoelectrochemical hydrogen production (70) 115 Sijtsrna, L., see de Swaaf, M.E. (70) 185 Singer, S., see La Barre, S. (70) 207 Soejima, T., see Ogbonna, J.C. (70) 289 Sponga, F., Cavaletti, L., Lazzarini, A., Borghi, A., Ciciliato, I., Losi, D. and Marinelli, F., Biodiversity and potentials of marine-derived microorganisms (70) 65 Stevenson, C., see Wright, P.C. (70) 343 Takeyama, H., see Matsunaga, T. (70) 33 Tanaka, H., see Ogbonna, J.C. (70) 289 Tanaka, H., see Ogbonna, J.C. (70) 213 Ternbach, M.B., see Janssen, M. (70) 323 Tomiyama, S., see Ogbonna, J.C. (70) 213 Tramper, J., see Janssen, M. (70) 323 Tramper, J., see Osinga, R. (70) 155 Tredici, M.R., see Chini Zittelli, G. (70) 299 Tsukahara, K., Sawayama, S., Yagishita, T. and Ogi, T., Effect of Ca 2+ channel blockers on glycerol levels in Dunaliella tertiolecta under hypoosmotic stress (70) 223 Ti.irker, L., see Eroglu, |. (70) 103 Tiirker, L., see Sediroglu, V. (70) 115 Tfirker, L., see Yi~it, D.O. (70) 125 Turkiewicz, M., Gromek, E., Kalinowska, H. and Zielifiska, M., Biosynthesis and properties of an extracellular rnetalloo

411

protease from the Antarctic marine bacterium Sphingomonas paucimobilis (70) 53 Turner, M., see Wood, B.J.B. (70) 175 Vandanjon. L.. Jaouen. P., Rossignol, N., Qu6meneur, F. and Robert, J.-M., Concentration and desalting by membrane processes of a natural pigment produced by the marine diatom Haslea ostrearia Simonsen (70) 393 van der Wielen. L.A.M. and Cabatingan, L.K., Fishing products from the sea--rational downstream processing of marine bioproducts (70) 363 ,,an Santen. R.A.. see Vrieling, E.G. (70) 39 Veldhoen, B.. see Janssen, M. (70) 323 Vincenzini. M.. see Chini Zittelli, G. (70) 299 Vrieling, E.G.. Beelen. T.P.M., van Santen, R.A. and Gieskes, W.W.C.. Diatom silicon biomineralization as an inspirational source of new approaches to silica production (70) 39 Wijffels. R.H.. see Janssen, M. (70) 323 Wijffels, R.H.. see Osinga, R. (70) 155 Woerdenbag. H.J.. see Koulman, A. (70) 85 Wood. B.J.B.. Grimson, P.H.K., German, J.B. and Turner, M.. Photoheterotrophy in the production of phytoplankton organisms (70) 175 Wright. P.C.. Stevenson, C.. McEvoy, E. and Burgess, J.G., Opportunities for marine bioprocess intensification using novel bioreactor design: frequency of barotolerance in microorganisms obtained from surface waters (70) 343 Yagishita. T.. see Tsukahara, K. (70) 223 Yamazawa. A.. see Matsunaga, T. (70) 33 Yi~it. D.O.. G/indtiz. U., Tiirker, L., Yficel, M. and Erofglu, [., Identification of by-products in hydrogen producing bacteria: Rhodobacter sphaeroides O.U. 001 grown in the waste water of a sugar refinery (70) 125 Yficel. M., see Eroglu. i. (70) 103 Yficel, M.. see Sediroglu, V. (70) 115 Yticel. M.. see Yi~it, D.O. (70) 125 Zaborsky. O.R.. Marine bioprocess engineering: the missing link to commercialization (70) 403 Zielifiska, M.. see Turkiewicz, M. (70) 53 Zou, N. and Richmond. A., Effect of light-path length in outdoor flat plate reactors on output rate of cell mass and of EPA in Nannochloropsis sp. (70) 351


JOURNAL

OF



Subject Index Acetic acid, (70) 175 Adhesion, (70) 163 Adsorption, (70) 33 Agar diffusion method, (70) 61 Alexandrium minutum, (70) 207 Algae, (70) 175 Antarctic marine bacteria, (70) 53 Antibiotic, (70) 27, 61 Antibiotic activities, (70) 65 Anticancer, (70) 15 Antifungal, (70) 61 Antimitotic, (70) 15 Aplidium glabrum, (70) 85 Aquaculture, (70) 5 Artificial light, (70) 289 Ascidian, (70) 85 Autotrophy, (70) 175 Axenic culture, (70) 227 Bacteriorhodopsin, (70) 115 Barotolerance, (70) 343 Biodiversity, (70) 15, 65 Biogenic silica, (70) 39 Biomineralization, (70) 39 Bioprocess intensification, (70) 343 Bioreactor, (70) 185, 343 Bioremediation, (70) 33 Blue-green algae, (70) 61 Ca 2 + channel blocker, (70) 223 Cadmium, (70) 33 Carbon dioxide consumption rate, (70) 271 Carbon source, (70) 357 Cell culture, (70) 163 Cell lines, (70) 133 Cell numeration, (70) 207 Cellulose acetate membrane, (70) 115 Chitin fibres, (70) 373 Chitin-silk fibroin fibres, (70) 373 Chitosan, (70) 373 Chlamydomonas reinhardtii, (70) 323 Chlamys opercularis, (70) 163 Chlorella, (70) 313 Chlorella minutissima, (70) 357 Chlorella sorokiniana, (70) 323 PII: S0168-1656(99)00146-7

Commercialization, (70) 403 Concentration, (70) 393 Continuous culture, (70) 271 Controlled proliferation, (70) 207 Corn steep liquor, (70) 213 Crvpthecodinium cohnii, (70) 185 Cultivation, (70) 71, 185 Cyanobacteria, (70) 61, 89 Cytotoxicity, (70) 85 Desalting, (70) 393 Diatoms, (70) 71 Didemnum lahillei, (70) 85 Docosahexaenoic acid. (70) 185, 193 Dual substrate, (70) 103 Dunaliella, (70) 313 Dunaliella tertiolecta, (70) 223 Eicosapentaenoic acid, (70) 193, 249, 299 Elemental analysis, (70) 271 Endotoxin, (70) 203 EPA, (70) 351 Ethanol, (70) 213 Euglena gracilis, (70) 213 Extreme environments, (70) 77 Fibroin, (70) 373 Fish farming, (70) 175 Fishing products, (70) 363 Fish oil, (70) 379 Fouling, (70) 163 Frustules, (70) 71 Gas-transfer modelling, (70) 335 Gelidium canariensis, (70) 227 Glucose, (70) 213 Glycerol, (70) 175, 223 Growth, (70) 155 Growth model, (70) 335

Halobacterium halobium, (70) 115 Haslea ostrearia, (70) 393 Heterotrophic culture, (70) 213 High hydrostatic pressure, (70) 343 Hydrodynamics modelling, (70) 335

414 Hydrogen, (70) 89, 125 Hydrogen production, (70) 103, 115 Hydrogen production factor, (70) 103 Hypoosmotic stress, (70) 223 Immobilisation, (70) 115 Indoor culture, (70) 289 Instantaneous fractional yield, (70) 103 Intermittent illumination, (70) 323 Internal illumination, (70) 289 Invertebrate cell culture, (70) 5 In vivo cultivation, (70) 155 Light/dark cycles, (70) 323 Light-path, (70) 351 Light regimen, (70) 231 Lipase-catalyzed esterification, (70) 379 Lipid composition, (70) 175 Marine, (70) 185, 193 Marine bioprocess engineering, (70) 403 Marine bioproducts, (70) 5, 363 Marine Chlorella, (70) 33 Marine epibiotic bacteria, (70) 27 Marine fungi, (70) 203 Marine invertebrates, (70) 133 Marine microalgae, (70) 33 Marine microorganisms, (70) 65 Marine pharmaceuticals, (70) 5 MarinLit, (70) 15 Mass transfer, (70) 231 Medium composition, (70) 185 MeOH, (70) 357 Metalloprotease, (70) 53 Methanol, (70) 357 Microalgae, (70) 231, 249 Microalgae mass cultivation, (70) 299 Microalga oils, (70) 379 Microbial antagonism, (70) 27 Microculture tetrazolium assay, (70) 85 Missing link, (70) 403

Nannochloropsis sp., (70) 299, 351 Nanofiltration, (70) 393 Nia vibrissa, (70) 203 North Sea, (70) 85 N - 3 polyunsaturates, (70) 193

Subject Index Photobioreactor, (70) 71, 103, 289, 351 Photobioreactors. (70) 231, 249, 313 Photoelectrochemical reactor, (70) 115 Photosynthetic bacteria, (70) 89 Photosynthetic culture, (70) 249 Physisorption. (70) 39 Pigment. (70) 393 Platyhelminths, (70) 133 Polyacrylamide gel, (70) 115 Poly-/~'-hydroxy butyric acid, (70) 125 Polymer therapeutics. (70) 15 Polyunsaturated fatty acids, (70) 379 Porphyridium cruentum. (70) 271 Porph.vridium purpureum, (70) 335 Productivity. (70) 351 Protist. (70) 193 Pseudosuberites (aft.) andrewsi, (70) 155 Psychrophilic enzymes. (70) 53 Radioactive labeling. (70) 207 Rational downstream processing, (70) 363 Renewable energy. (70) 89 Rhodobacter sphaeroides. (70) 103, 125 Rhodophyta. (70) 227 Salt marsh plants, (70) 77 Scale-up, (70) 249 Scaleup. (70) 231 Screening. (70) 33 Semiterrestrial, (70) 77 Siliceous earth. (70) 71 Silicon (bio)chemistry, (70) 39 Small angle X-ray scattering (SAXS), (70) 39 Solar energy. (70) 89 Solar light. (70) 289 Specific growth rate, (70) 323 Spirulina. (70) 313 Sponge aquaculture, (70) 15 Sponges, (70) 155 Sporelings. (70) 227 Stress induction, (70) 343 Suberites ficus, (70) 163 Sustainable use, (70) 5

Outdoor cultures, (70) 351 Oxygen generation rate, (70) 271

Textile, (70) 373 Thraustochytrid, (70) 193 z~-Tocopherol, (70) 213 Triglycerides. (70) 379 Tubular photobioreactor, (70) 271, 335 Tubular photobioreactors, (70) 299

Phaeodactylum tricornutum, (70) 249 Photoautotrophic culture, (70) 213

Waste water. (70) 125 Wound dressing. (70) 373

Marine Bioprocess Engineering

Bioprocess Engineering Principles

Solutions Manual Bioprocess Engineering Principles

Putting Biotechnology to Work: Bioprocess Engineering

New Products and New Areas of Bioprocess Engineering

Encyclopedia of Bioprocess Technology

Introduction to Marine Engineering, Second Edition

Dynamics of Marine Vehicles (Ocean engineering)

Introduction to Marine Engineering, Second Edition

Marine

Marine

Marine

Bioanalysis and Biosensors for Bioprocess Monitoring

Bioanalysis and Biosensors for Bioprocess Monitoring

Bioanalysis and Biosensors for Bioprocess Monitoring

Mitzi's Marine

Marine Hydrodynamics

Model Marine

Mitzi's Marine

Marine Biology

Space Marine

Marine Radioactivity

Marine SysOpka

Luna Marine

Lonestar Marine

Space Marine

Marine Macroecology

Marine Biology

Mitzi's Marine

Marine parasitology

Marine Chemistry

Marine Bioprocess Engineering

Bioprocess Engineering Principles

Solutions Manual Bioprocess Engineering Principles

Putting Biotechnology to Work: Bioprocess Engineering

New Products and New Areas of Bioprocess Engineering

Encyclopedia of Bioprocess Technology

Introduction to Marine Engineering, Second Edition

Dynamics of Marine Vehicles (Ocean engineering)

Introduction to Marine Engineering, Second Edition

Marine

Marine

Marine

Bioanalysis and Biosensors for Bioprocess Monitoring

Bioanalysis and Biosensors for Bioprocess Monitoring

Bioanalysis and Biosensors for Bioprocess Monitoring

Mitzi's Marine

Marine Hydrodynamics

Model Marine

Mitzi's Marine

Marine Biology

Space Marine

Marine Radioactivity

Marine SysOpka

Luna Marine

Lonestar Marine

Space Marine

Marine Macroecology

Marine Biology

Mitzi's Marine

Marine parasitology

Marine Chemistry

Recommend Documents