Advances in Applied Microbiology, Volume 43

ADVANCES IN

Applied Microbiology VOLUME 43

This Page Intentionally Left Blank

ADVANCES IN

Applied Microbiology Edited by

SAUL L. NEIDLEMAN Oakland, California

ALLEN I. LASKIN Somerset, New Jersey

VOLUME 43

Academic Press San Diego New York Boston London Sydney Tokyo Toronto

This book is printed on acid-free paper. @ Copyright 0 1997 by ACADEMIC PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher's consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 01923), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1997 chapters are as shown on the title pages, if no fee code appears on the title page, the copy fee is the same as for current chapters. 0065-2164/97 $25.00

Academic Press

a division of Harcourt Brace b Company 15 East 26'h Street, 15'h floor, New York, New York 10010,USA http:/lwww,apnet.com

Academic Press Limited 24-28 Oval Road, London NW1 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Serial Number: 0065-2164 International Standard Book Number: 0-12-002643-0

PRINTED IN THE UNITED STATES OF AMERICA 97 98 99 00 01 02 BB 9 8 7 6 5

4

3

2

1

CONTENTS

Production of Acetic Acid by Clostridium thermoaceticum

MUMRCHERYAN, S A W PAREKH, MINISHSHAH, AND KUSUMAWITJITRA I. 11. 111. IV. V. VI. VII.

Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acetic Acid Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cfostridiuin fhermoaceticum . . . . . . . . . , . . . , . . . . . . . . . . . . . . . . . . . . . . . . Strain Improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low-Cost Media . . . . . . . . . . . . . . . . . . . . , . , . . . . . . . . . . . . . . . . . . . . . . . . Bioreactors for Improving Productivity . . , . . . . . . . . . . . . . . . . . . . . . . . . . . Downstream Processing of Acetate Fermentation Broths. . . . . . . . * . . . . . . . ...,.*........ References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

.

1 3 5 13 16 24 29 31

Contact Lenses, Disinfectants, and Acanthamoeba Keratitis

DONALDG. AHEARNAND MANALM. GABRIEL ..................

I. Introduction and Taxonomy 11. Biology. ... ... . . . . . . . .. .. . .. .. . . . .

V. VI.

................. ................. . . .. . .... ... , . .. .. .. .. . ................ Disinfection. . . . . . . . . . . . . . . Adherence to Lenses. . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . .

35 3a 40 42 44

47 51

Marine Microorganisms as a Source of New Natural Products

V. S. BERNAN,M. GREENSTEIN, AND W. M. MAIESE I. 11. 111. IV V.

Introduction .................................................... Natural Products from Marine Microorganisms . , . . . . . . . . . . . . . . . . . . . Overview of Wyeth-Ayerst (W-AR) Marine Natural Products Program . . . . Marine Biotechnology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summ......................................................... References .....................................................

.

..

V

.

.

.

57

58 70 86 a7 a7

vi

CONTENTS

Stereoselective Biotransformations in Synthesis of Some Pharmaceutical Intermediates &MESH

N . PATEL

I . Introduction ....................................................

I1. Tax01 Semisynthesis ............................................. EI. Thromboxane A 2 Antagonist ......................................

N . ACE Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

V Anticholesterol Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

VI. Antiinfective Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII . Calcium Channel Blocking Drugs .................................. Vm . Antipsychotic Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX . X. XI. XII.

Antiarrhythmic Agents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Potassium Channel Openers., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Antiinflammatory Drugs .......................................... Antiviral Agents ................................................ XI11 Prostaglandin Synthesis .......................................... References .....................................................

.

91 92 98

101 107 113 120 121 125 127 129 130 132 133

Microbial Xylanolytic Enzyme System: Properties and Applications

P R ABAJPAI ~ I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

II. Structure of Xylan and Its Interaction with Plant Cell Walls. . . . . . . . . . . . . III. Properties of Xylanolytic Enzymes ................................. N . Production of Xylanolytic Enzymes ................................. V. Application of Xylanases .........................................

VI. Conclusions .................................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

141 142 149 163 167 184 185

Oleaginous Microorganisms: An Assessment of the Potential JACEK LEMAN

.

I Introduction .................................................... I1. Microbial Oil ...................................................

Single Cell Oil .................................................. Specialty Fats and Oils ........................................... Valuable Metabolites ............................................. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

195 197 203 206 225 234 235

..............................................................

245

C~NTENTSOFPREVIOUSVOLUMES ..........................................

251

III. N. V. VI.

INDEX

Production of Acetic Acid by Clostridium thermoaceticum MUNIRCHERYAN, AND

S A W PAREKH, MINISHSHAH, KUSUMAWITJITRA

Agricultural Bioprocess Laboratory University of Illinois Urbana, Illinois 61802

I. Historical Background 11. Acetic Acid Production A. Aerobic Process B. Anaerobic Process 111. Clostridium thermoaceticum A. Substrates for Fermentation B. Mineral Requirements IV. Strain Improvement V. Low-Cost Media VI. Bioreactors for Improving Productivity VII. Downstream Processing of Acetate Fermentation Broths References

I. Historical Background

Acetic acid (ethanoic acid, methyl carboxylic acid) has been produced as long as wine making has been practiced and therefore dates back to at least 10,000 BC (Nickol, 1979; Agreda and Zoeller, 1993), although it could be predated by certain fermented foods made from milk (Allgeier et al., 1974).It is assumed that the first vinegar, which is an aqueous solution of acetic acid, resulted from spoiled wine (Ghose and Bhadra, 1985), given that the Latin word acetum means sour or sharp wine. It initially functioned as a medicinal agent and was most likely the first known antibiotic. For most of human history, all acetic acid was derived by the same age-old process of sugar fermentation to ethyl alcohol and subsequent oxidation to acetic acid by rnicroorganisms to produce vinegar. This was the sole source of acetic acid. Late in the nineteenth century, this process was supplemented by the advent of wood distillation, which provided an additional source of acetic acid (Agreda and Zoeller, 1993). In 1995, annual production of acetic acid by the petrochemical route in the United States was 4.68 billion pounds, 1 ADVANCES IN APPLIED MICROBIOLOGY,VOLUME 43 Copyright 0 1997 by Academic Press, Inc. All rights of reproduction in any form reserved. 0065-2164/97$25.00

2

M. CHERYAN et 01.

ranking 35th among all chemicals produced (Anonymous, 1996). Production increased at an annual rate of 18% in 1993-95. Vinyl acetate ranked 41st, averaging 3 billion lb in 1993-95. In 1916, the first dedicated plant for the production of acetic acid by chemical rather than biological means became commercial (LeMonnier, 1965). This method was based on acetylene-derived acetaldehyde, and it marks the advent of inexpensive, industrial-grade acetic acid and the birth of a viable industry based on its use (Agreda and Zoeller, 1993). The advantages to chemical synthetic routes include high acetate concentrations (35-45 YOby weight), high production rates, and acetic acid generated in the free-acid form. The major disadvantages are the need for high temperatures and high pressures, good agitation, the threat of explosion, the high cost of catalysts, and the dependence on nonrenewable uncertain sources of raw materials (crude oil). Fermentation production routes have traditionally been aimed at the food market. Vinegar production usually requires lower capital investment, has shorter start-up times, and can generate different types and flavors of vinegar when different carbohydrate sources are used. Furthermore, the raw material (e.g., corn) is a renewable resource. The cost of acetic acid from chemical synthesis has ranged from 25 to 35 a/lb on a 100% basis, while it is 35-45 allb from aerobic fermentation. Clearly the latter value must be decreased if fermentation production is to supply the demand for nonfood uses. The industrial importance of acetic acid can be understood from Fig. 1. The major outlet today is for vinyl acetate, which is used for vinyl plastics, adhesives, textile finishes, and latex paints. This market has grown rapidly during the past few years due to the demand for synthetic fibers. In 1979 calcium magnesium acetate (CMA) was identified as a noncorrosive environmental-friendly alternative to chloride salts for deicing roads (Marynowski et al., 1985). Road salt use is 10-12 million tons per year in the United States, and CMA in solid form could supply about 5-10% of that market within the next decade (Wise, 1992). Liquid potassium acetate is being used now as a deicer for airport runways and as a heat exchange fluid; in the latter role it could serve as a partial replacement for ethylene glycol (1995 annual production of 5.23 billion lb). In addition, there are reports that CMA or calcium acetate can also be used as an additive to coal-fired combustion units, for example, boilers used by electrical utilities (Levendis, 1991; Manivanan and Wise, 1991; Sharma, 1991). Here calcium acts as a “grabber” for sulfur in the coal, reduces sulfur dioxide emissions, and partially relieves the problem of acid-rain pollution. If these environment-related substitu-

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

3

ACETIC ACID

I

1

Acetic Anhydride

1

r

1

Pharmaceuticals

I Textile Finishes Vinyl Plastics Latex Paints Adhesives

1

Plasticizers

Cellulose Acetate

1

Pharrnaceut.

/Solvents1

Transparent Sheets Textile Fibers Photo Film

Heat Transfer Liquids Meat Preservative Neutralizer Fungicide De-icers

FIG.1. Uses of acetic acid.

tions take place, the demand for acetic acid would increase tremendously. II. Acetic Acid Production

Acetic acid as an industrial chemical is presently produced from fossil fuels and chemicals by three processes: acetaldehyde oxidation, hydrocarbon oxidation, and methanol carbonylation. It can also be produced by biological routes, which forms the basis of this review. A. AEROBICPROCESS Food-grade acetic acid is produced by the two-step vinegar process (Allgeier et a]., 1974; Ebner and Follman, 1983, Crueger and Crueger, 1990; Shreve and Brink, 1977). The first step is the production of ethanol from a carbohydrate source such as glucose. This is carried out at 30°C using the anaerobic yeast Saccharomyces cerevisiae. C6H12O6

+ 2C02 + 2CH3CH20H.

The second step is the oxidation of ethanol to acetic acid. Although a variety of bacteria can produce acetic acid, only members of Acetobacter are used commercially, typically the aerobic bacterium Acetobac-

4

M. CHERYAN et al.

ter aceti at 27-37°C. This fermentation is an incomplete oxidation because the reducing equivalents generated are transferred to oxygen and not to carbon dioxide: 2CH3CHzOH + 0

2

-+ ZCHsCOOH + 2Hz0.

The overall theoretical yield is 0.67 g acetic acid per gram glucose. At the more realistic yield of 76% (of 0.67, i.e., 0.51 g per gram glucose), this process requires 2.0 pounds of sugar or 0.9 pounds of ethyl alcohol per pound of acetic acid produced (Busche et al., 1982). Complete aeration and strict control of the oxygen concentration during fermentation are important to maximize yields and keep the bacteria viable (Muraoka et d.,1980; Osuga et a]., 1984). Submerged fermentation has almost completely replaced surface fermentation methods. The drawand-fill mode of operation can produce acetic acid at concentrations up to 10% wt/wt in continuous culture at pH 4.5 in about 35 hours (Crueger and Crueger, 1990; Ebner and Follman, 1983; Nickol, 1979). B. ANAEROBIC PROCESS

In the 1980%another process for production of acetic acid emerged based on anaerobic fermentation using Clostridia. These organisms can convert glucose, xylose, and some other hexoses and pentoses almost quantitatively to acetate according to the following reactions: CsHlZOe + 3CH3COOH, + 5CH3COOH.

ZCSH1005

Typical acidogenic bacteria are Clostridium aceticum (Braun et al., 1981), Clostridium thermoaceticum (Fontaine et al., 1942; Andreesen et al., 1973), Clostridium formicoaceticum (Andreesen et d., 1970) and

Acetobacterium woodii (Balch et d., 1977). Many can also reduce carbon dioxide and other one-carbon compounds to acetate (Ljungdahl, 1983).

This fermentation route has several advantages. It is anaerobic and thus should have lower fermentation cost. The theoretical yields are higher than the aerobic fermentation: 3 moles of acetic acid are produced per mole of glucose consumed, that is, 1 g acetic acid/g glucose (Brownell and Nakas, 1991; Brumm, 1988; Parekh and Cheryan, 1990a,b, 1991; Schwartz and Keller, 1982b; Wise et al., 1991). Actual yields with C. thermoaceticum have ranged from 0.85 (Fontaine et al., 1942; Ljungdahl et al., 1986; Wang et al., 1978) to 0.90 g acetic acid per gram glucose and greater (Parekh and Cheryan, 1990a, 1994a; Shah and Cheryan, 1995b). Until 1967, C. thermoaceticum was the only acetogen

PRODUCTION OF ACETIC ACID BY C. thermoaceticurn

5

Glucose

t

Pyruvate

CO,

Formate

--+

-7-

Formyl-THF -CH,-THF

2 Acetyl CoA

/--

CoA

Lzz-

Corrinoid enzyme

CoA Acetyl CoA

FIG.2. Embden-Myerhoff pathway for production of acetic acid.

easily available for study (Ljungdahl, 19861. Consequently, the most detailed studies of acetate synthesis have been performed with this organism. III. Clostridium thermoaceticum C. thermoaceticum was isolated from horse manure. It is an obligate anaerobe, Gram-positive, spore-forming, rod-shaped, thermophilic organism with an optimum growth temperature of 55-60°C and an optimum pH of 6.6-6.8 (Ljungdahl et al., 1985). The wild strains produce 2.55 moles of acetic acid per mole of glucose fermented (actual yield) and only 13-20 g/liter acetic acid in batch fermentation (Fontaine et al., 1942; Wang et a)., 1978; Sugaya et al., 1986).The fermentation of sugars to acetate is a complex process. As shown in Fig. 2, one mole of hexose is metabolized by the Embden-Meyerhof pathway to yield 2 moles of pyruvate, which are further metabolized to 2 moles of acetate (formed from carbons 2 and 3 of the pyruvate] and to 2 moles of COz (formed from the carboxyl groups) (Wood, 1952a). The two moles of COz serve as electron acceptors, and one mole COz is finally reduced to methyltetrahydrofolate (CH,THF). The CH3THF

6

M. CHERYAN et al.

then combines with the second C 0 2 and coenzyme-A (CoA), forming acetyl-CoA, and finally the formation of the third mole of acetate (Barker and Kamen, 1945; Wood, 1952b). The overall reaction can be written as follows: C6HI2O6+ 2H20 -+ ZCHBCOOH+ 2 C 0 2 + 8H’ + 8e-, 2C02 + 8H’ + Be- -+ CH3COOH + 2H20.

Enzymes involved in formation of the third mole of acetate are tetrahydrofolate enzymes, carbon monoxide dehydrogenase (CODH),NADPdependent formate dehydrogenase (FDH), and a corrinoid enzyme (Ljungdahl, 1986). These enzymes are metalloproteins, for example, CODH contains nickel, iron and sulfur (Drake et al., 1980; Ragsdale et al., 1983), and FDH contains iron, selenium, tungsten, and a small quantity of molybdenum (Yamamoto et al., 1983), while the corrinoid enzyme (vitamin BIZ compound) contains cobalt (Hu et al., 1984). A. SUBSTRATES FOR FERMENTATION

In most physiological studies conducted on C. thermoaceticum, cells were cultivated in a complex undefined growth medium containing substantial quantities of yeast extract and tryptone. Thus, the true anabolic capabilities of C. thermoaceticum remained unclear for a long time. Lundie and Drake (1984) attempted to establish the basic nutritional requirements of C. thermoaceticum and thus define its anabolic potential and limitations. In developing a minimally defined growth medium, after a series of deletion experiments they found that C. thermoaceticum did not display any specific amino-acid requirement and required nicotinic acid as the sole essential vitamin. Koesnandar et al. (1990) determined the optimum concentration of five metals in the glucose-minimal medium as iron 100 pM, cobalt 400 pM, molybdateGO0 pM, selenite 1pM, and nickel 0.15 pM. C. thermoaceticum is now routinely grown in a rich medium containing glucose, a complex nitrogen source such as yeast extract or corn steep liquor, tryptone, reducing agent, bicarbonate, phosphate, and minerals (Andreesen et al., 1973; Schwartz and Keller, 1982a). This is shown in Table I as the “base-standard” medium. Tryptone was excluded from the culture media by Ljungdahl et al. (1985),although yeast extract was used. However, yeast extract is an expensive ingredient; Ljungdahl et al. (1986)suggested that yeast extract could be replaced by such cheaper materials as yeast autolysates and corn steep liquor. This . has been confirmed by Cheryan and Shah (1996), Witjitra et ~ l(1996), and Shah and Cheryan (1995a), as discussed later.

7

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

TABLE I NUTRIENT MEDIUM FOR CLOSTRIDIUM THERMOACETICUM (g/liter)

Component Glucose Buffering components: KHCO, KZHPO, KHZPO, Yeast extract Salts: (NH,I,SO, MgSO,. 7H2O Fe(NH,),(SO,), . 6H,O COSO,. 7.5HzO Na,WO, 2H,O Na,MoO, 2H,O NiC1, 6H,O ZnSO, . 7H,O Na,SeO, Cysteine . HC1 H,O

Base (standard] medium (1x1

Medium at 2X level

As requireda

As required

9.0

1.8

1.40 1.10

1.40 1.10 10.0

5.00

1.0

2.0

0.25

0.50

0.04 0.03 0.0033 0.0024 0.00024 0.00029

0.08 0.06

0.0066 0.0048

0.000017

0.00048 0.0005 0.000034

0.25

0.25

Vsually 20 g/liter for maintenance and initial growth,

Typical fermentation patterns are shown in Figs. 3 and 4. There is usually a lag period before growth and acetate production commences. This is followed by a sharp drop in cell numbers as indicated by OD measurements, and a simultaneous decrease in the rate of acetate production. Growth and acetate production by C. thermoaceticum is critically dependent on nutrient type and levels. When nutrients are supplied at the “normal” level ( I X concentration shown in Table I), the final acetate concentration and yield are low (22.5 g/liter and 0.77 g acetate/g glucose consumed), as shown in the upper graph of Fig. 3. A large portion of the substrate (glucose) is unutilized. When the nutrient supply was increased to 2X in Experiment 2 (Fig. 3, bottom), the final acetate concentration increased significantly to 35 g/liter. Additional fermentation parameters are shown in Table 11. Increasing the nutrients to 3X and supplying additional glucose during the fermentation resulted in a higher acetate concentration of 47.8 g/liter (Expt. 3; Fig. 4). The concentration of cells also increased with

8

M. CHERYAN et al.

4

Acetate

. 3

2

f X \ I.,

Cells

1

L I

20

40

60

00

100

128

Time (hours) FIG.3. Typical fermentation patterns of C. thermoaceticum using yeast extract and ammonium sulfate as nutrients. Temperature of all fermentations was 6OOC. TOP: Experiment 1 with 1X level of yeast extract (5 g/liter), ammonium sulfate (1g/liter), and salts (see Table I1 for concentrations). BOTTOM: Experiment 2 with 2X concentration of yeast extract, ammonium sulfate and salts (Cheryan and Shah, 1996).

increased supply of nutrients. Maximum cell concentration increased from 1.8 g/liter in Experiment 1to 4.8 g/liter in Experiment 3. Fructose, an intermediate product associated with nutrient depletion or microorganism stress (Witjitra et al., 1996), was also produced when glucose was added the second time in Experiment 3 (Fig. 4) but was gradually metabolized by the end of the experiment. In subsequent experiments (listed in Table 11) the yeast extract concentration was reduced and ammonium sulfate concentration was increased while keeping the concentration of salts fixed at 2X. In Experiment 4 the yeast extract concentration was reduced to 5 g/liter, while other nutrients were supplied in the same concentration as in Experi-

9

PRODUCTION OF ACETIC ACID BY C. thermoaceticum 5

4

i 39

v)

*6 1

0

Time (hours) FIG.4. Fermentation of dextrose by C. thermoaceticum using yeast extract and ammonium sulfate, each at 3X level (see Table 11).Additional glucose was added as shown after 28 hours of fermentation [Cheryan and Shah, 1996).

ment 2. This resulted in acetate concentration and yields of 31.5 g/liter and 0.77 g/g glucose, respectively; both these values were lower than Experiment 2. In addition, fructose was produced as a by-product and some glucose remained unutilized at the end. Increasing the ammonium sulfate concentration to 3 g/liter (3X) caused the acetate concentration to increase significantly to 38 g/liter, improving the yield to 0.86 g/g, with no by-product formation. However, further reduction of yeast extract to 0.5X reduced the acetate concentration to 33 g/liter and the yield to 0.78 g/g; in addition, fructose was produced toward the end of the experiment (Expt. 6, Table 11). Increasing the ammonium sulfate to 3.5X restored performance (Expt. 7: 36 g/liter acetate produced in 71 h, with an acetate yield of 0.86 g/g and no fructose). Thus, it appears that ammonium sulfate could reduce yeast extract requirements by 75% with minimal effects on acetate production and yield. In fact, the productivity was higher, increasing from 0.39 g/liter-h in Experiment 2 to 0.5 g/liter-h in Experiment 7 (Table 11). The fermentation is clearly growth-associated. As shown in Figs. 3 and 4, cell concentration increased exponentially initially and then decreased until the end of the fermentation. Acetate concentration also increased and then leveled off. Acetate productivity was more than 1 g/liter-h initially and declined with the time of fermentation, parallel to the decrease in OD. This emphasizes the necessity for keeping the culture viable and active. A by-product, which appears to be fructose, appeared after the OD started to decrease in almost all cases, especially

TABLE I1 EFFECT OF NUTRIENTLEW ON FERMENTATIONP.m.o.mnsa Experiment no.

Nitrogen source and level*

YE,1x YE,2X YE,3x YE,1x YE,1x YE,0.5X YE,0.5X CSL,1x CSL,2x

SaltsC level 1x 2x 3x 2x 2x

2x 2X

2X 2x

(NHJLQ level

'Erne (hl

1x 2x 3x 2x 3x 3x 3.5x

115 90 125 72 65 100 71 90 95

3x 2X

Acetate (glliter) 22.5 35.0 47.8 31.5 38.0 33.1 36.1 31.1 30.8

Yield of acetate

Yield of fructose

(g/d

(!A4

0.77 0.83

-

0.19

0.01 0.05 -

0.39 0.38 0.43

n

0.58 0.33 0.51 0.35 0.32

z z

0.85 0.77 0.86 0.78 0.84

0.03

0.78 0.82

0.08 0.08

Wtial glucose concentration = 46 goiter. Source:Cheryan and Shah (1996). %E = yeast extract (see Table 1 for IX concentrations). CSL = corn steep liquor (X= 10 g solidsfliter medium). Q o e s not include ammonium sulfate.

-

Productivity (g acetate/ litedh)

5

E

z 4

Y

PRODUCTION OF ACETIC ACID BY C. therrnoaceticurn

11

when the nutrients were inadequate either in quality and/or quantity. High glucose concentration inhibits the initial growth of C. thermoaceticum (Witjitra et al., 1996). However, after adaptation (which takes time and energy, and results in slow initial growth rate and some glucose utilization without acetate production), the fermentation proceeds rapidly. There appears to be a minimum ratio of nutrient concentration to glucose concentration to produce acetic acid. If glucose was still available but not the nutrient, the microorganism produced fructose instead of acetate. Acetate production from glucose by C. thermoaceticum generates five moles of ATP per mole of glucose consumed (Fuchs, 1986). This results in high levels of cell mass per mole of glucose consumed (Andreesen et al., 1973). To maintain productivity, the cells must balance their ATP supply and demand. Since growth consumes more ATP than maintenance, most of the acetic acid produced by C. thermoacetjcum occurs during the growth phase. In all our work, especially those with suboptimal levels of nutrients, a decrease in cell density is observed 24-48 hours into the fermentation, which is followed by a decrease in the rate of production of acetate. Thus, to increase acetic acid production, the cells must be continuously growing. This requires a continuous supply of nutrients, as shown in Experiments 1-3 (Table 11). This can also explain why a large excess of nutrients was necessary to obtain very high (>80 glliter) acetate concentrations (Parekh and Cheryan, 1994a). It also explains the limitations of a continuous cell-recycle bioreactor (see later): merely maintaining higher cell concentrations or using low dilution rates could not produce high acetate concentrations without also proportionally increasing the nutrient concentration. When cells use yeast extract as a source of amino acids, nucleotides, and fatty acids, they will need less ATP than if they have to synthesize these compounds using ammonium ions as the starting material. Thus, assimilation of ammonium ions is important if cells are to recycle the ATP generated during production of acetic acid. Therefore, much smaller amounts of ammonium sulfate could replace yeast extract (e.g., compare levels in Expts. 2 and 7, Table 11). However, since yeast extract contains growth factors in addition to nitrogen, it could not be completely substituted by ammonium sulfate. It appears that acetate production could be sustained only with sufficient ATP-consuming reactions taking place in the cells. If sufficient nutrients are not supplied, fructose was produced (e.g., Expts. 4, 6, 8, and 9). By increasing ammonium sulfate (e.g., Expts. 5 and 7), by-product formation was avoided and acetate yields were good (0.84-0.86).

12

M. CHERYAN et a].

Thus, by substituting the much cheaper ammonium sulfate for yeast extract, the ratio of acetate produced to yeast extract needed can be substantially increased. As will be seen later, the medium cost could be lowered further by substituting corn steep liquor for yeast extract. C. thermoaceticum grows well on pyruvate as the primary substrate (Barker, 1944; Andreesen et a]., 1973). Parent (wild) strains of C. thermoaceticum did not grow on DL-lactate but an adapted strain grew on both D- and L-lactate with a yield of acetate of 0.95-1.00. Both strains grew on and consumed lactate when pyruvate, glucose, fructose, or xylose was used as the second substrate (Brumm, 1988). It is important to maintain a low level of the carbon source in the fermenter at all times to ensure viability of the cells (Shah and Cheryan, 1995b). By using different concentrations of externally added sodium salts to the growth media, the relative growth inhibition caused by the anion was found to be in the order of acetate > chloride > sulfate. Various externally added cations of acetate were also examined, and the relative magnitude of inhibition on the growth rate was found to be ammonium > potassium > sodium (Wang and Wang, 1984). This could be specific to a particular strain (e.g., DSM 521 used by these investigators), since recent work has shown that strain ATCC 47907 could successfully produce ammonium acetate up to 50 glliter acetic acid (Cheryan and Shah, 1996). B . MINERAL REQUIREMENTS

The fact that metal is involved in acetate production explains why metal is needed in the culture media. Andreesen et al. (1973) found positive benefits of metals (ferrous, molybdate, and calcium) on growth yield, enzymes, and synthesis of acetate from CO,. Formate dehydrogenase was the only enzyme of those assayed that was affected by the addition of metals to the growth medium. Addition of selenite and molybdate or tungstate to the growth medium stimulated the formation of formate dehydrogenase during growth of C. thermoaceticum (Andreesen and Ljungdahl, 1973). Later, Ljungdahl and Andreesen (1978) demonstrated that, although their strain of C. thermoaceticum could grow without including selenite, tungstate, or molybdate in the medium, high formate dehydrogenase activity was obtained only when these metals were present in the medium. Shoaf et al. (1974) found that NH: or K+ but not Na+ increased the thermostability of formyltetrahydrofolate synthetase from C. thermoaceticum, but that phosphate ions inhibited the enzyme, and this inhibition was stronger in the presence of NH:. Nickel (Ni2+)has no effect

PRODUCTION OF ACETIC ACID BY C. fhermoaceticum

13

on the growth of C. thermoaceticum, but less carbon monoxide dehydrogenase was formed when Ni2+was omitted from the medium (Diekert and Thauer, 1980). Increasing cobalt by fourfold (to 400 pl4) did not affect cell growth, glucose consumed, acetate produced, or its molar yield, but it did increase corrinoid production (Koesnandar et al., 1990). Cysteine is used in the medium as a source of sulfur (Koesnandar et a]., 1990) and as a reducing agent. Since cysteine is expensive, studies have been conducted with alternate low-cost sulfur-containing reducing agents (Shah et al., 1996). It appears that the recommended dosage of 0.25 g/liter of cysteine . HC1. H,O is probably too much, at least for strain ATCC 47907 with yeast extract as the complex nitrogen source. Decreasing it to 0.05 g/liter actually improved the final acetate concentration in the broth while simultaneously reducing the maximum OD in the fermentation broth by 50%. Thus, the increase in acetate yield may be a result of a decrease in cell mass yield that allowed more carbon to be channeled into acetate production. Cysteine could also be successfully replaced with sodium thioglycolate and sodium sulfide (Na2S). The sulfur source should not be eliminated completely: with no sulfurcontaining reducing agent, acetate and cell concentrations were lower (Shah et al., 1996). IV. Strain Improvement

Better strains of C. thermoaceticum were developed by Schwartz and Keller (1982a,b) that could grow and produce acetic acid at pH 4.5 (available as ATCC 31490). Reed et al. (1987) developed strains tolerant to high acid-low pH conditions to allow recovery of product in the acid or undissociated form rather than the salt form. An improved strain of ATCC 39289, an acid-tolerant strain, produced 29 g/liter acetate in 140 hours at pH 6.65 and could grow in the presence of 70 g/liter sodium acetate (Parekh and Cheryan, 1990a). In a fed-batch mode, this particular improved strain performed well (Parekh and Cheryan, 1990b). Several other strains of C. thermoaceticum have also been screened (Ljungdahl, 1983; Ljungdahl et al., 1985; Wang and Wang, 1984). Ljungdahl et al. (1986) reported on a particular mutant strain that could apparently produce high levels of acetate (up to 100 g/liter acetate or more). However, no further studies with that strain have been reported. Parekh and Cheryan (1991) developed several strains by treatment with chemical mutagenic agents (NTG and nitrous acid) and selective enrichment procedures. Mutagenesis yielded several mutants, some of which exhibited growth at pH 5.6 and above, as well as acetate yields

14

M. CHERYAN et al.

above 0.8 g acetate per gram pyruvate, when 10-20-mM pyruvate was supplied as the sole energy source. Table 111 summarizes the behavior of a few mutants that showed consistent growth when screened in selective media. All strains exhibited growth in media containing up to 5% acetate at pH 6.6 while producing an additional 1.0% acetate with yields (Yp,s) of 0.8 or better. Prior to mutagenesis, the parent strain (C. thermoaceticum ATCC 39289, a spontaneous mutant isolated by Schwartz and Keller (1982b)by repeated culturing at low pH) could be grown and cultivated in media at pH 6.4 and above containing 5% acetate or 10-mM pyruvate or formate. However, media containing 1% acetate at pH below 5.3 inhibited growth and repressed acetate production. In contrast, two mutant cultures, G-10 and 5-40, showed visible signs of growth after 140 h of cultivation in media containing 1% acetate at pH 5.3. However, these cultures could not be subcultured regularly in the same 4% acetate media at pH 5.3 and attain good growth. Mutant strain G-20 was the only one that appeared to grow when cultivated at pH 5.6 in 50-mMpyruvate as the sole energy source (Table 111). Subsequent experiments showed it could grow and attain acetate yields of 0.81 in media containing 80-mMpyruvate controlled at pH 5.5 and incubated for 4 days. The surviving cells could be readily retransferred to an 80-mMpyruvate medium and showed visible growth in 72 h. Cells from this culture could be subcultured regularly in 80-mM pyruvate without difficulty when the pH was controlled at pH 5.6. In addition, culture G-20 could also tolerate dolime concentrations as high as 6% when cultured at pH 6. Wiegel et al. (1991) also reported that their C. thermoaceticum mutants derived after EMS treatment showed excellent tolerance to dolime. A similar procedure with the other mutants did not result in any growth when the media contained over 20-mM pyruvate (Table 111). The rationale of using pyruvate- and formate-containing media was that these compounds play a fundamental role in the metabolism of acetogenic bacteria (Ljungdahl, 1983). Pyruvate is the precursor of acetate in the pyruvate dehydrogenase reaction, and it is also the source of the “CO” unit in the CO-Ni-E complex. This unit also plays an important role in synthesis of acetate from COz. Similarly, formate is the precursor of the methyl group in acetate synthesis via the autotrophic pathway. This approach has yielded several mutants with better acetate producing capabilities (Brumm, 1988; Wiegel et al., 1991). Pyruvate and formate inhibited growth of some mutants, especially at pH 5.5 and below (Parekh and Cheryan, 1991; Wiegel et al., 1991). While some strains could grow in the presence of 10-mM pyruvate at pH 5.6 (Table 111), they lacked tolerance to 50-mM pyruvate and dolime

z0

TABLE 111

aL

SELECTION CONDlTIONS'

Glucose/pyruvate

Pyruvate (pH 6.5)

Strains

pH 5.5

pH 6.5

10 mM

Parent strain

-

+

+

50mM

-

2

Pyruvate (pH 5.61 1OmM

20 mM

+

-

50mM

2 80mM

-

-

$ > n

Mutant strainb G-10 G Z O 640 N-10

N-40 5-40 s-100 0

- - no growth or acetate production; + = growth and acetate production. Source: Parekh and Cheryan (1991).Reprinted with permission

( I -

from Springer-Verlag. bc = NTG treatment;N = nitrous acid treatment; S = EMS treatment. Numbers refer to the time (minutes) of exposure of cells to t h e reagent.

a

16

M. CHERYAN et

01.

and did not show any improvement in acetate production (Parekh and Cheryan, 1991). Only G-20 could grow in 80-mM pyruvate and 50-mM formate media when used as the sole energy source (Table111). Attempts to grow this mutant in higher-pyruvate media and at low pH (below 5.4) were unsuccessful. Growth resumed in media with pH above 5.6. In addition, even though the parent strain was unable to grow in 10-mM pyruvate media at pH 5.3, it metabolized pyruvate in the presence of glucose when cultivated above pH 6.5. Some C. thermoaceticum strains utilize mixed substrates (glucose and lactate or formate in a 1:l ratio) above pH 6.5 independently of the starting substrate (Brumm, 1988). The ability of this mutant to grow and produce acetate below pH 6 makes it useful in production of industrial acetates such as CMA, since it would assist in the soluhilization of dolime and in the recovery of soluble CMA. When evaluated in a fed-batch fermentation system, this mutant strain performed better than the parent strain (Fig. 5): 18% higher concentration of acetate and 25% faster rate compared to the parent strain (Parekh and Cheryan, 1991). Experiments with this mutant strain have been repeated and carried out for several years in laboratory and pilot scale (up to 40 liters) in a variety of bioreactors (batch, fed-batch, multiple-stages, cell-recycle, and membrane) operated up to 54 days continuously and producing several salts of acetate (ammonium, calcium, calcium-magnesium, potassium, and sodium). No alterations in stability or performance were observed. This strain has produced as much as 10% CMA in a fed-batch reactor in 140 hours, with 93% substrate utilization and acetate yields of 0.80 g/g glucose (Parekh and Cheryan, 1994a). Such robustness is important for industrial production of bulk chemicals. This mutant culture is available as ATCC 49707 and DSM 6867. V. Low-Cost Media

The superior performance of mutant strains of C. thermoaceticum comes at a price: substantial quantities of high-quality yeast extract are needed and fermentation times are long, resulting in low productivity. This leads to high fermentation cost, For example, with a yield of 0.85 g/g and a glucose cost of $0.16 per kilogram, the cost of CMA would be $2.49 per kilogram if even the lowest-cost commercial yeast extract ($8.50 per kilogram) is used. This far exceeds today’s price of petroleum-derived CMA of $0.70-0.80 per kilogram. Thus, the cost of nutrients must be reduced, while maintaining a reasonably high concentration of acetate and good productivity.

17

PRODUCTION OF ACETIC ACID BY C. thermoaceticum 70

i

Calcium-mepesium acetate

2 20

70 h

20

;

10

y

0

c

z a

10

(I)

0

0

0

50

100

150

0

50

100

150

200

Fermentation time (hours) FIG.5. Fed-batch production of acetate by improved and mutant strain G-20 of C. therrnonceficurn.pH was maintained at 6.3 f 0.5 by the addition of 20% high-magnesium lime, LEFT: Calcium/magnesium acetate produced (expressed as acetic acid concentration). RIGHT: Glucose consumed during the fed-batch fermentation (Parekh and Cheryan, 1991). Reprinted with permission from Springer-Verlag.

Table IV is a list of the various nutrients studied by Witjitra et al. (1966). Bacto yeast extract from Difco is universally recognized as an

excellent stimulator of bacterial growth. With C. thermoaceticurn fermentation, it results in excellent growth and acetogenesis, especially at concentrations of 15-30 g/liter or more (Parekh and Cheryan, 1991, 1994a). It is not clear which components of yeast extract are important. Lundie and Drake (1984) suggested that nicotinic acid is the sole essential vitamin for C. thermoaceticum: yeast extract is rich in water-soluble vitamins, including nicotinic acid (Bridson and Brecker, 1970; Difco Laboratories, 1984). Although hydrolyzed cottonseed is an excellent nutrient source for Lactobacillus amylovorus (Cheng et al., 1991), C. thermoaceticum showed no growth or acetate production even after 100 h of incubation. It might lack the component[s) required for the growth of C, therrnoaceticum, or it might have contained a component from the cottonseed that inhibited its growth (e.g., gossypol). The fermentation pattern with hydrolyzed corn gluten meal was similar to that observed with yeast extract except that much less acetate ( 4 5 g/liter) was produced. It is possible that corn gluten meal contains a growth inhibitor for C. thermoaceticum that might explain the lag phase. For example, phytic acid, a component of corn, binds strongly with minerals, making them biologically unavailable as nutrients (Cheryan, 1980).

TABLE TV

C o ~ p o m OF o ~NUTRIEN~ SOURCE(% wtlwt)"

Hydrolyzed pratein nutrients Heavy

Corn gluten

N-&Soy

stillagd

3.4

3 .O

4.4

2.9

2.6

2.4

8.7

10.1

13.7

93.0 na 0.39

na

na na

na

0.28

na

0.14

CottonsMd

Moistum a-amino nitrogen Total nitrogen Fat Fiber Lactic acid

Ash

na

-

-

corn

Uefatted

Difco

steep liquo$

SOY

yeast

flow

exkdct

50.0

7.0

5.5

na

7.0

8.3 na

na 1.84 1.0 na

1.47

11.3

-

9.5 na nn na

7.3

6.0

10.0

0.11

0.01

na

4.3

6.1

6.3

3.4 0.36

5.7

Sodium

8.1. 0.69

0.04

0.56 0.03

PH

6.3

6.0

6.8

4.3

"Source:Witjitra (1994). "Received in a concenhted liquid form. All others were in solid powdered form. na = data no1 available.

3.0

g

Fi 5z a

n

!-

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

19

The performance of the corn refining by-products was also good, although some pretreatment may have to be given to them to optimize their utilization. “Stillage” is the liquid residue after the ethanol has been stripped away in the beer well of a yeast-based ethanol fermentation plant. The solids consisted mainly of dead yeast cells, traces of ethanol, low-molecular-weight sugars, organic acids (lactic, succinic), glycerol, and amino acids. With untreated stillage instead of yeast extract at 10 g soliddliter (Fig. 6), there was a lag period at the beginning of the fermentation and fructose was produced instead of acetate. When growth began, fructose levels dropped and acetate was produced. When cell concentration started to decrease, fructose started accumulating again. Fermentation was better when the stillage was filtered before being added to the media (Fig. 6). Once again a clear correlation between decreased cell concentration, decreased acetate production rate, and appearance of fructose can be observed in the data. Filtration appeared to eliminate a growth inhibitor in stillage. When the filtered stillage was used in the medium without dilution, that is, at a level of 64 g solids/liter, the fermentation profile was also very good (Fig. 6). There was slow growth initially but no fructose was produced during this period. After adapting itself, the culture continued to grow and produce acetate. When the cells started lysing and decreasing in numbers (after 160 h), the same phenomenon occurred, that is, subsequent production of fructose. The high concentration of nutrients in the undiluted stillage was obviously sufficient for extended cell growth and production of acetate, but perhaps too concentrated to be used in the initial stages, since there appeared to be some inhibition initially. A better method of utilizing stillage would be in a fed-batch mode, where the stillage is added with the glucose as needed to sustain viability and productivity. The yield with the undiluted filtered stillage was 1.07 g acetate per gram glucose (Table V). This is because there was lactic acid and sugars in the stillage that were probably utilized by the organism to produce acetate (Brumm, 1988). Corn steep liquor (CSL) contains soluble components leeched out from the corn during soaking. It is rich in organic nitrogen (4446% protein on a dry matter basis), with about half the nitrogen as free amino acids; the balance is small peptides with very little intact protein (Christianson et al., 1965). It contains relatively high levels of several important vitamins, trace elements, and lactic acid (10-30%, dry basis); the latter can be utilized by C. fhermoaceficum, thus increasing the yield of acetate. It has been industrially used as a nutrient in ethanol fermentation and for production of pharmaceuticals. In acetate fermen-

20

M. CHERYAN et al. 40

thewnoaceticum

35

.

Stillage, g/L Unfiltered.1 C Filtered, 10

b

._ € 3 1

U '

0

'

'

'

50

100

"

'

,

150

200

'

.

250

Time (hours) FIG.6. Fermentation of glucose by C. thermoaceticum using corn-refining by-products as nutrients. Open points = with corn steep liquor (CSL)at 10 g soliddliter. Closed points = with stillage at concentrations shown on dry solids basis. (Adapted from Witjitra et al., 1996.)

tation, yield of acetate was similar to fermentation using yeast extract, but less acetate was produced with corn steep liquor, probably due to lower concentration of essential nutrients. Fructose began to accumulate as soon as acetate production stopped (Fig. 6). Increasing the concentration of CSL improves the fermentation, as shown in Fig. 7. Pretreatment of the CSL with dolime and filtering out the precipitate also is beneficial, possibly by removing the phytic acid (Shah and Cheryan, 1995a). Figure 7 shows the effect of different levels of pretreated and filtered CSL on the fermentation pattern. The 1 X experiment was conducted with corn steep liquor at 5 g/liter (dry basis), salts at the concentration shown in Table I, and added vitamins. The final acetate concentration was only 19 g/liter, and cell concentration

TABLE V FERMENTATION s-P FOR C. THERMOACETICUMWITH VARIOUS NUTRIENTS AT A CONCENTRATION OF 10 g Soms/LITER OF F~RMENTATION BROTH,UNLESS OTHERWISE STATED'

Glucose utilizationb Nubient Yeast extract, Difco Hydrolyzed corn gluten N-Z-Soy@BL Hydrolyzed soy flour Corn steep liquor Stillage, unfiltered Stillage, filtered Stillage, filtered (64g/liter)

(%I 91 46 55 90

aa 57 63 74

Acetate produced (g/liter) 34.6 14.5 14.1 33.9 31.1 17.6 19.9 36.0

Fructose produced (g/liter)

0 U

5

g

Yield (S/d

Productivity (g/liter-h)

-

0.80 0.72

0.47 0.15

0.7

0.57

0.29

>

0.79 0.83 0.68 0.71 1.07

0.39 0.32 0.15 0.21 0.15

.c 9

-

1.5

4.5 4.9 2.7

2

> R

=! n

8m

"Source: Witjitra (1994). bGlucose in media initially was 50 glliter.

N Y

22

M. CHERYAN et al.

Time (hours)

FIG.7. Effect of CSL concentration on fermentation of glucose (initial glucose = 46 g/liter). Open points = dolime-treated and filtered. 1 X is CSL at 5 g dry solids/liter. Closed points = No treatment (2x1. Medium was supplemented with thiamine (0.15 glliter), riboflavin (0.35 glliter), pyridoxine (0.175 g/liter), nicotinic acid (3 g/liter), and pantothenic acid (1g/liter). (Adapted from Shah and Cheryan, 1995a).

reached only 0.9 g/liter before decreasing. Doubling the CSL and salts doubled the maximum cell density and increased acetate level to 3 1 g/liter. At 2.7X concentration of CSL and salts, the final acetate level was 38.5 g/liter and maximum cell density was also three times the cell density at 1X. The importance of proper pretreatment of the CSL can be seen in Fig. 7. At the 2X level, untreated (i.e., no dolime treatment and no prefiltration) CSL resulted in only 26 g/liter of acetate and a much lower productivity than with pretreated CSL.

PRODUCTION OF ACETIC ACID BY C. thermooceticum

23

Table I1 shows the interactions when CSL and ammonium sulfate in the medium were varied. Experiments 2 and 9 are similar except that CSL was substituted for the yeast extract. Acetate concentration was lower, but yields were similar. Experiments 4 and 8 were also similar in design and results. Fructose was produced in all CSL experiments. However, as will be seen later, adding additional CSL during fermentation (e.g., by fed-batch operation) caused the fructose to disappear and be replaced with more acetate. Low-hydrolysis soy flour is also a promising low-cost nutrient. Soybean meal is the residue from soybeans after the extraction of the oil. It is a complex mixture of protein (50% dry basis), carbohydrates (oligosaccharides and fiber 30%), fat (l%), and lecithin (1.8%).It is a good source of nutrients for industrial fermentation, especially for antibiotics (Crueger and Crueger, 1990). Enzymatically hydrolyzing the protein, even to a limited extent, has a beneficial effect on fermentation by making the nitrogedprotein in soy meal more available, and perhaps releasing other bound nutrients. It is possible that the residual lipids and/or lecithin in the soy flour is the key nutrient. Lipids are essential components of the membranes of Clostridia. Baumann et al. (1965) reported that about 4.5% of the dry weight of log phase cells of C. butyricum is lipid; out of this, 70% is phospholipid. The lipid material in the soy meal could contribute to the stability of the cell wall during fermentation, reducing the rate of cell lysis and allowing better fermentation. It is worth noting that no fructose was detected with low-hydrolysis soy flour as the nutrient. On the other hand, a high degree of hydrolysis [with the resultant high levels of free amino acids) yielded less product than the less extensively hydrolyzed soy flour. Similar results were observed with L. amylovorus (Cheng et al., 1991) and L. bulgaricus (Leh and Charles, 1989). Apparently, certain microorganism may be stimulated by peptides of a certain length or of a certain amino-acid composition. Since hydrolyzed soy flour, corn steep liquor, and ethanol stillage cost less than $0.50/kg each, their use will significantly reduce the overall nutrient cost and the cost of fermentation-derived acetate. The economic significance of using stillage as a nutrient source goes beyond the acetate industry. At present, the economic value of stillage, after evaporation and drying, is marginal at best (moisture removal of stillage requires substantial amounts of energy). If, on the other hand, stillage is used as the nutrient source for acetate fermentation, there will be little or no stillage handling costs. This will improve ethanol economics while simultaneously improving acetate economics by making available a good nutrient source at almost no cost, except for the filtration (which

24

M. CHERYAN et al. 60

1

2

3

4

5

4

67

-

C

.-4-0 m

-2 -

-1

b C

0 0 C

u"

Time (hours) FIG.8. Fed-batch fermentation with 3X levels of CSL and salts. Arrows indicate addition of nutrient solutions in the following amounts: 1,2, and 3 = 50 ml of solution A (CSL and salts 6.6X, glucose 260 g/liter); 4 = 100 ml of solution B (CSL and salts 10x1; 5 and 6 = 50 ml of solution B; 7 = 20 ml of solution C (glucose 500 g/liter). Initial volume = 1.0 liter, final volume = 1.47 liter. From Shah and Cheryan (1995a). Reprinted with permission from The Macmillan Press Ltd.

removes the growth inhibitors and suspended solids and simultaneously sterilizes the stillage). VI. Bioreactors for Improving Productivity

Much of the work done with this fermentation has been done with batch fermenters, where all the carbohydrate and nutrients are added at the start of fermentation. With fermentations that are substrate-inhibited, a better method is to use the fed-batch mode of operation. This significantly improves the performance of C. thermoaceticum fermentation (Parekh and Cheryan, 1990b, 1994a). Figure 8 shows a typical fed-batch fermentation using CSL as the complex nitrogen source. Since it is substrate-inhibited, the initial glucose concentration was low (20 g/liter), while CSL and salts were at the normal 1 X concentration shown in Table I. Subsequently, about 24 hours into the fermentation, when the OD decreased, indicating a decrease in viable cell numbers, more nutrients and glucose were added (in concentrated form to minimize volume changes). Eventually, after 70 hours, cell numbers went up again, As before, fructose was produced during the death phase of the cells. But when cell viability improved the fructose disappeared, apparently being converted into acetic acid. At the end of the fermentation, there was no fructose in the broth. This pattern has been repeatedly observed in the fed-batch mode (Shah and Cheryan, 1995a).

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

25

Alkali PH controller

1-1

1 cr”----i - -*El&

Prod&

(Permme)

FIG.9. Schematic of a continuous bioreactor using membrane separations for cell recycle.

Continuous fermentation (Sugaya et al., 1986) and immobilized whole cells have been used to increase the productivity of homoacetogenic fermentations (Reed and Bogdan, 1985; Wang and Wang, 1983). However, with Ca-alginate-immobilized cells of C. thermoaceticum, steady-state performance was not achieved when cultured at high acetate concentrations (Wang and Wang, 1983). On the other hand, cell-recycle bioreactors using a membrane module as the separation device have been shown to vastly increase the productivity of several anaerobic fermentations, such as ethanol and lactic acid (Cheryan and Mehaia, 1986) and may have some advantages over immobilized cells, such as higher concentration of free cells, no diffusion limitation, excellent mixing in the bioreactor, and a cell-free product stream. The biggest advantage is that cell concentrations far in excess of normal levels can be used with no danger of cell washout. For example, yeast concentrations of 100 g/liter can be used for production of ethanol from glucose (equivalent to 1011-1012 cells/ml). This is far greater than the 10-20 g/liter ( lo8 celldml) normally used in industrial fermentations. These higher cell concentrations result in much faster fermentations, thus vastly improving productivity (Cheryan, 1986). Figure 9 is a schematic of a continuous cell-recyle membrane bioreactor. The reaction vessel of the fermentation system is coupled in a semi-closed loop configuration to an ultrafiltration or microfiltration module of the appropriate chemical nature and physical configuration (Cheryan, 1986). The reaction vessel is initially charged with the cells

26

M. CHERYAN et al. 1 .o

2.0

=:

\ 1.5 (r

v

.-> .+d

c

0.9

0.8

2\

0.7

1.0

0

4-l

0.6

0

: a

-2

0.5

.E >-

0.5 0.4

0.0

1 301

--. -A

5 20

0

GI u c 0 s e,*A

,

l 0 L 4 0 0.005 0.010 0.015 0.020 0.025 0.030

Dilution r a t e (h-')

FIG.10. Fermentation of glucose by C. thermonceficum in a membrane recycle bioreactor. Cell concentration (X) = 17 g/liter, glucose concentration (S) = 58 g/liter. Yeast extract and salts were at the 2X level. Data taken after at least 5 volume changes. From Parekh and Cheryan (1994b). Reprinted with permission from Elsevier Science Inc.

and adjusted to the required cell concentration. Feed is then pumped into the reaction vessel and product is removed as permeate at the same rate, thus keeping the volume in the bioreactor constant. For growth-associated fermentation, some provision has to be made to bleed excess cells to avoid pumping problems. Figure 10 shows typical results obtained with such a bioreactor with C. thermoaceticum at a cell concentration of 1 7 g/liter (normal cell concentrations in batch and fed-batch fermentations are 2-4 g/liter). An increase in dilution rate (i.e., a decrease in residence time) resulted in a decrease in acetate concentration, a decrease in acetate yield, and an increase in unutilized glucose (Parekh and Cheryan, 1994b). High dilution rates also resulted in the appearance of a by-product (not shown in Fig. 10; the authors suggested the by-product was lactate, but it could have been fructose). The long-term stability of this bioreactor was very good with this strain of C. thermoaceticum. Interestingly, the usual bell shape of the cell-density curve observed in batch operations was not present: no loss

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

27

Cells 0

50

100

150

200

260

300

350

400

Time (hours)

FIG.11. Performance of a two-stage continuous-membrane bioreactor. The membrane was attached only to the second stage. The feed was pumped into the first stage reaction vessel. Broth from the first stage was pumped continuously into the second vessel at a rate to keep fermentation volumes constant. Fermentation broth from second stage was recycled through the membrane module. The cell-free permeate was removed at the same rate as the feed flow rate to the first stage. The retentate containing the cells was recycled back to the first stage at 25% of the feed flow rate to the first stage, and the remaining was recycled to the second stage. Dilution rate is based on the total volume used in the two stages. From Shah and Cheryan (1995h). Reprinted with permission from Humana Press Inc.

of viability or vitality was noted during the 54 days of continuous operation. Microscopy examination did not indicate any dramatic change in the external morphology (cell size or shape). The recycled cells were isolated and preserved in 50% glycerol broth at -20°C. On retrieving, the culture did not demonstrate any change in the rate or efficiency of acetate/CMA production, suggesting that the mutant culture is stable (Parekh and Cheryan, 1994b). Similar microbial stability was shown with a 2-stage membrane bioreactor in which the membrane was attached only to the second stage and a part of the retentate containing the cells was recycled to the first stage. As shown in Fig. 11, with a cell concentration of 4.5 g/liter in the first stage and 12.5 g/liter in the second stage, a dilution rate of 0.02 h-l resulted in acetate concentrations of 25 and 37.5 g/liter in the first and second stages, respectively. This outlet concentration was slightly better than a one-stage membrane reactor (Shah and Cheryan, 1995b). When the dilution rate was increased to 0.033 h-l, the two-stage bioreactor resulted in lower acetate concentration, but productivity was higher. In addition, the cell concentration increased to 15.5 g/liter. These data

28

M. CHERYAN et a].

Ok

56

160

160

260

’

260

300

’

361

Time (hours) FIG.12. Performance of a draw-and-fill bioreactor. At the end of the fermentation, 80% of the fermentation volume was withdrawn and clarified with a cross-flow microfiltration membrane. The retentate containing the cells was recycled to the fermenter, which was then recharged with fresh substrate and fermentation allowed to continue. “Acetate” is expressed in terms of acetic acid. pH was maintained at 6.2 with ION NaOH. From Shah and Cheryan (1995b). Reprinted with permission from Humana Press Inc.

confirmed previous reports that high productivity and high product concentration are mutually exclusive in such high-rate fermenters (Cheryan, 1986; Parekh and Cheryan, 1994b). The yield of acetate was 0.85-0.9 g/g glucose consumed. A “draw-and-fill” bioreactor in combination with a membrane appeared to be the optimum design (Shah and Cheryan, 1995b). In this design, the reaction vessel is operated as a batch fermenter. At the end of the fermentation, a portion of the fermentation broth is withdrawn through the membrane module. The cells are recycled and the reaction vessel charged with fresh substrate. Figure 1 2 shows a series of such draw-and-fill operations with 80% removal of the volume in each cycle. With feed glucose concentrations of 50 g/liter, fermentation times were usually about 34 hours, acetate concentration (expressed as acetic acid) was 38 g/liter, and overall productivity was 0.93 g/liter. There was a lag in acetic acid production at the beginning of each cycle. Also, in each cycle of fed-batch operation, the cells passed through a growth phase, followed by a significant decrease in OD, probably due to cell lysis, caused perhaps by nutrient depletion. These characteristics were similar to those observed in conventional batch fermentations. However, the average cell concentration increased in later cycles, which allowed the fermentation to be completed faster (every 24 hours).

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

I I

Sugars

'

1

29

Nutrients ,

Nutrient recycle

Water recyle

f

Concentrated Organic Acid

Nutrient r e c y c l e d Alkali recycle

FIG.13. Possible downstream processing of acetate fermentation broths by membrane technology. MF is cross-flow microfiltration, NF is nanofiltration, ED is electrodialysis, LR is low-rejection membranes, and HR is high-rejection membranes.

In batch fermentation without cell recycle, acetic acid production is proportional to the amounts of yeast extract and trace salts supplied in the medium. For all types of bioreactors studied, increasing dilution rate increases volumetric productivity but decreases specific productivity (g acetate produced per gram cells). Thus, in cell-recycle bioreactors the nutrient supply should be increased in proportion to cell concentration to realize the full potential of the microorganism. VII. Downstream Processing of Acetate Fermentation Broths

Downstream processing refers to the series of unit operations used to isolate, purify, and concentrate the product. Downstream processing often determines the economic feasibility of the process. One possible downstream processing scheme for acetate utilizes membrane separations technology, as shown in Fig. 13. The first operation is cell separation, which can be done by cross-flow microfiltration. Cell harvesting by membranes is rapidly replacing conventional filtration and centrihgation techniques. When combined in a semiclosed loop configuration to the bioreactor or fermenter, it becomes a powerful tool to dramatically improve the productivity of the fermentation, while simultaneously

30

M. CHERYAN st al.

providing a cell-free broth for subsequent downstream processing (Cheryan, 1986). Depending on the physical and chemical nature of the fermentation products, the cell-free broth is subjected to chromatography, electrophoresis, crystallization, precipitation, extraction, distillation, and/or membranes. Solvent extraction with azeotropic distillation is the preferred method for chemically derived acetic acid, while freeze concentration is used for vinegar. Both require substantial amounts of energy since a change in phase of the solvent is required. Furthermore, if the acetate is required in the free-acid form, there will be additional cost to convert the salt form produced in the anaerobic fermentation to the free-acid form. Among the membrane techniques, electrodialysis and nanofiltration are particularly useful for separating and partially concentrating acetates. A relatively new membrane technology, nanofiltration (NF) can separate charged compounds from noncharged ones and from each other, depending on the relative sizes of the ions and the degree of dissociation (Raman et d., 1994). In Fig. 13 the cell-free broth is first passed through a low-rejection (LR) NF membrane, which separates the acetate from most of the rest of the broth components, including the sugars and many of the nutrients. The substantially purified, but dilute, acetate solution then has to be concentrated. If the acetate is in the salt form, it can be either evaporated or possibly processed through high-rejection (HR) membranes to partially concentrate it prior to evaporation and/or drying. Several NF membranes have recently been screened for the separation of acetic acid (Han and Cheryan, 1995), and preliminary economic calculations suggest it is an attractive technique for this purpose (Han and Cheryan, 1996). Electrodialysis (ED) is a membrane-separation process that separates and concentrates charged compounds from liquid feed solutions by transport under the application of electrical energy and through anionor cation-selective membranes. Ultimately, the ED unit could be coupled to a high-performance membrane bioreactor that would integrate fermentation and separation steps for continuous production of acetates, as shown in Fig. 13. The first stage of the ED could be done with conventional anion- and cation-exchange membranes as a prepurification step, followed by a second stage with bipolar membranes that would generate the alkali and the free acid form of acetate. This would substantially reduce the cost of alkali needed for fermentation as well as reducing waste treatment costs. ED has been found to be particularly useful in concentrating vinegar, for example, to save on transportation costs (Chukwu and Cheryan, 1996).

PRODUCTION OF ACETIC ACID BY C. thermoaceticum

31

In summary, the industrial production of acetic acid by fermentation using Clostridium thermoaceticum appears to be feasible. The mutant strain ATCC 47907 is especially promising. It is a robust organism that can be adapted to a variety of nutrient sources. Considerable research has been done to lower costs by adapting the culture to corn steep liquor and ammonium sulfate as nitrogen sources instead of yeast extract, by reducing the level of reducing agent (cysteine) or by substituting it with a cheaper source such as sodium sulfide, and by using fed-batch fermentation systems in combination with cell recycle by cross-flow microfiltration. Of the material costs, the sugar (e.g., dextrose) will cost about $150/ton of acetic acid and the nutrients should account for only $20/ton. Considering that acetic acid from petroleum sources sells for $550-660/ton in 1996, it indicates good potential for the fermentation process to provide a significant share of the market. ACKNOWLEDGMENTS

Research on acetate production by Clostridium thermoaceticum in the first author’s laboratory has been supported by the Illinois Corn Marketing Board, the Minnesota Corn Promotion and Research Council, the U S . Department of Agriculture through the NRICGP program, and the Illinois Agricultural Experiment Station at Urbana-Champaign. REFERENCES Agreda, V. H., and Zoeller, J. R. (1993). “Acetic Acid and Its Derivatives.” Dekker, New York. Allgeier, R. J., Nickol, G. B., and Connor, H. A. (1974). Food Prod. Dev. 8(6),50-56. Andreesen, J. R., and Ljungdahl, L. G. (1973). f. Bacteriol. 116, 867-873. Andreesen, J. R., Gottschalk, G., and Schlegel, H. G. (1970). Arch. Mikrobiol. 72,154-174. Andreesen, J. R., Schaupp, A,, Neurauter, C., Brown, A., and Ljungdahl, L. G. (1973). J. Bacteriol. 114, 743-751. Anonymous. 1996. Chem. Eng. News 74(15), 15-19. Balch, W. E., Schoberth, S., Tanner, R. S., and Wolfe, R. S. (1977). Int. f . Syst. Bacteriol. 27, 355-361.

Barker, H. A. (1944). Proc. Natl. Acad. Sci. U.S.A. 30,88-90. Barker, H. A., and Kamen, M. D. (1945). Proc. Nut. Acad. Sci. U.S.A. 31,219-225. Baumann, N. A., Hagen, P.-O., and Goldfine, H. (1965). f . Biol. Chem. 240, 1559-1567. Braun, M.,Mayer, F., and Gottschalk, G. (1981). Arch. Microbiol. 128, 288-293. Bridson, E.Y., and Brecker, A. (1970). In “Methods in Microbiology” (J. R. Norris and D. W. Ribbons, eds.), Vol. 3A, pp. 229-304. Academic Press, New York. Brownell, J. E., and Nakas, J. P. (1991). J. Ind. Microbiol. 7, 1-6. Brumm, P. J. (1988). Biotechnol. Bioeng. 32, 444-450. Busche, R. M., Shimshick, E. J., and Yates, R. A. (1982). Biotechnol. Bioeng. Syrnp. Ser. 12, 249-262.

32

M. CIIERYAN et al.

Cheng, P., Mueller, R. E., Jaeger, S., Bajpai, R., and Iannotti, E.L. (1991). 1.Ind. Microbiol. 7, 27-34. Cheryan, M. (1980). CRC Crit. Rev. Food Sci. Nutr. 13, 297-335. Cheryan, M. (1986). “Ultrafiltration Handbook.” Technomic, Lancaster, PA. Cheryan, M., and Mehia, M. A. (1986). Chemtech 16(11), 676-681. Cheryan, M., and Shah, M. M. (1996). Unpublished data. Christianson, D. D., Cavins, J. F., and Wall, J. S. (1965). J. Agric. Food Chem. 13, 277-280. Chukwu, U. N., and Cheryan, M. (1996). Concentration of vinegar by electrodialysis. J. Food Sci. 61. In press. Crueger, W., and Crueger, A. (1990). In “Biotechnology: A Textbook of Industrial Microbiology,” 2nd ed., pp. 143-147. Sinauer Associates, Inc., Sunderland, MA. Diekert, G., and Thauer, R. K. (1980). FEMS Microbiol. Lett. 7, 187-189. Difco Laboratories (1984). Difco Manual: Dehydrated Culture Media a n d Reagents for Microbiology.Difco Laboratories Inc., Detroit. Drake, H. L., Hu, S. I., and Wood, H. G. (1980).1.Bid. Chem. 255, 7174-7180. Ebner, H., and Follmann, H. (1983). In “Biotechnology” (H. J. Rehm and G. Reed, eds.), pp. 387-407. Verlag Chemie, Weinheim. Fontaine, F. E., Peterson, W. H., McCoy, E., Johnson, M. J., and Ritter, G. J. (1942). I. Bacteriology 43, 701-715. Fucbs, G. (198fi).FEMS Microbiol. Rev. 39, 181-213. Ghose, T. K., and Bhadra, A. (1985). In “Comprehensive Biotechnology” (M. Moo-Young, ed.), Vol. 3 , pp. 701-729. Pergamon, New York. Han, I. S., and Cheryan, M. (1995). J. Membrane Science 107, 107-113. Han, I. S., and Cheryan, M. (1996). Appl. Biochem. Biotechnol. 57/58, 19-28. Hu, S. I., Pezacka, E., and Wood, H. G. (1984). J. Biol. Chem. 259, 8892-8897. Koesnandar, Nishio, N., and Nagai, S. (1990). 1. Ferment. Bioeng. 71, 181-185. Leh, M. B., and Charles, M. (1989).J. Ind. Microbiol. 4,77-80. LeMonnier, E. (1965). In “Kirk-Othmer Encyclopedia of Chemical Technology,” Vol. 8, pp. 386-404. Wiley-Interscience, New York. Levendis, Y. A. (1991). In “Calcium Magnesium Acetate (CMA)” (D. Wise, Y. Levendis, and M. Metghalchi, eds.), pp. 211-223. Elsevier, Amsterdam. Ljungdahl, L. G. (1983). In “Organic Chemicals from Biomass” (D. L. Wise, ed.), pp. 219-248. Benjamin/Cummings Publishing Co, Inc., Menlo Park, CA. Ljungdahl, L. G. (1986). Ann. Rev. Microbrol.40, 415-450. Ljungdahl, L. G., and Andreesen, J. R. (1978). Meth. Enzymol. 53, 360-372. Ljungdahl, L. G., Carreira, L. H., Garrison, R. J., Rahek, N. E., Gunter, L. F., and Wiegel, J, (1985). Biotecbnol. Bioeng. Symp. 15, 207-223. T.jungdah1, L. G., Carreira, L. H., Garrison, R. J., Rahek, N. E., Gunter, L. F.,and Wiegel, J. (1986). CMA Manufacture (11):Improved Bacterial Strain for Acetate Production. U.S. Department of Transportation Report No. FHWA/RD-86/117. Lundie, L. L., and Drake, H. L. (1984). J. Bacteriol. 159, 700-703. Manivanan, S., and Wise, D. (1991). In “Calcium Magnesium Acetate (CMA)” (D. Wise, Y. Levendis, and M. Metghalchi, eds.), pp. 257-272. Elsevier, Amsterdam. Marynowksi, C. W., Jones, J. L., Tuse, D., and Boughton, R. L. (1985). Ind. Eng. Chem. Prod. Res. Dev. 24(3), 457-465. Muraoka, H., Watabe, Y., and Ogasawara, N. (1980). 1. Ferment. Techno!. 60, 171-180. Nickol, G. R. (1979). In “Microbial Technology” (H. J. Peppler and D. Perlnian, eds.), Vol. 2, pp, 155-172. Academic Press, New York, NY. Osuga, J., Mori, A,, and Kato, J. (1984). 1.Ferment. Techno]. 62, 139-149. Parekh, S. R., and Cheryan, M. (199Oa). Process Biochem. 25(4), 117-121.

PRODUCTION OF ACETIC ACID BY C. thermooceticum

33

Parekh, S . R., and Cheryan, M. (199Ob). Biotechnol. Lett. 12, 861-864. Parekh, S . R., and Cheryan, M. (1991). Appl. Microbiol. Biotechnol. 36,384-387. Parekh, S.R., and Cheryan, M. (1994a). Biotechnol. Lett. 16,139-142. Parekh, S . R., and Cheryan, M. (1994b). Enzyme Microb. Technol. 16, 104-109. Ragsdale, S. W., Clark, J. E., Ljungdahl, L. G., Lundie, L. L., and Drake, H. L. (1983). J. Biol. Chem. 258, 2364-2369. Raman, L. P., Cheryan, M., and Rajagopalan, N.(1994). Chem. Eng. Progr. 90(3), 68-74. Reed, W. M., and Bogdan, M. E. (1985). Biotechnol. Bioeng. Symp. Ser. 15, 641-647. Reed, W. M., Keller, F. A., Kite, F. E., Bogdam, M. E., Ganoung, J. S . (1987). Enzyme Microb. Technol. 9, 117-120. Schwartz, R. D., and Keller, F. A. (1982a). Appl. Environ. Microbiol. 43,1385-1392. Schwartz, R. D., and Keller, F. A. (1982b). Appl. Environ. Microbial. 43,117-123. Shah, M. M., and Cheryan, M. (1995a). J. Ind. Microbiol. 15, 424-428. Shah, M. M., and Cheryan, M. (1995h). AppI. Biochem. Biotechnol. 51/52, 413-422. Shah, M. M., Akanbi, F., and Cheryan, M. (1996). Appl. Biochem. Biotechnol. In press. Sharma, P. K. (1991). In “Calcium Magnesium Acetate (CMA)” (D. Wise, Y. Levendis, and M. Metghalchi, eds.), pp. 273-298. Elsevier, Amsterdam. Shoaf, W. T., Neece, S. H., and Ljungdahl, L. G. (1974). Biochim. Biophys. Acta. 334, 448-458. Shreve, R. N., and Brink, J. A,, Jr. (1977). “Chemical Process Industries.” McGraw-Hill, New York. Sugaya, K., Tuse, D., and Jones, J. L. (1986). Biotechnol. Bioeng. 28, 678-683. Wang, D. 1. C., Fleishchaker, R. J., and Wang, G. Y. (1978). AIChESymp. Ser, 182,105-110. Wang, G., and Wang, D. I. C. (1983). Appl. Biochem. Biotechnol. 8, 491-503. Wang, G., and Wang, D. I. C. (1984). Appl. Environ. Microbiol. 47,294-298. Wiegel, J., Carreira, L. H., Garrison, R. J., Robek, N. E., and Ljungdahl, L. G . (1991). In “Calcium Magnesium Acetate (CMA)” (D. Wise, Y. Levendis, and M. Metghalchi, eds.), pp. 359-416. Elsevier, Amsterdam. Wise, D. L. (1992). In ”Biochemical Engineering for 2001’’ (S. Furusaki, I. Endo, and R. Matsuno, eds.), pp. 723-726. Springer-Verlag, Tokyo. Wise, D. L., Levendis, Y. A., and Metghalchi, M., eds. (1991). “Calcium-Magnesium Acetate (CMA).” Elsevier Science Publishers, New York. Witjitra, K. (1994). M.S. Thesis, University of Illinois, Urbana. Witjitra, K., Shah, M. M., and Cheryan, M. (1996). Enzyme Microb. Technol. 19(7), 322-327. Wood, H. G. (1952a). J. Bid. Chem. 199, 579-583. Wood, H. G. (1952b). J. Biol. Chem. 199, 905-931. Yamamoto, I., Saiki, T., Liu, S. M., and Ljungdahl, L. G. (1983). J. Bid. Chem. 258, 1826-1832.

This Page Intentionally Left Blank

Contact Lenses, Disinfectants, and Acantharnoeba Keratitis DONALDG. AHEARNAND MANALM. GABRIEL Department of Biology Georgia State University Atlanta, Georgia 30302-4010

I. Introduction and Taxonomy 11. Biology 111. Infections

IV. Ecology V. Disinfection VI. Adherence to Lenses References

I. Introduction and Taxonomy

Small filose, free-living amoebae with a cyst stage with ostioles are classified in the genus Acanthamoeba, family Acanthamoebidae, order Amoebida, phylum Rhizopoda. Acanthamoeba are characterized by typically uninucleate trophozoites with fine protoplasmic projections (acanthopodia) arising anteriorly or laterally from the clear hyaline ectoplasm and by a prominent cyst stage (see Page, 1967; Sawyer and Griffin, 1975). The trophozoite divides by binary fission with the early disappearance of the nucleolus and nuclear membrane. In axenic broth cultures the trophozoites are somewhat globose to slightly irregular and of variable size (12 to 25 pm in diameter). In the presence of bacteria and particularly when traversing a surface, the trophozoites are more irregular in form and usually larger (15 to over 60 pm for some species) in one dimension. A large central or posterior vacuole is present. The trophozoite may become coated with bacteria or fungal cells that ultimately are ensnared in capsular material; a pincer-like pseudopod may be present in some species. In broth cultures with low densities of bacteria, trophozoites may agglutinate in clumps of 10 to 30 globose cells. A cellulosic cyst stage (endocyst wall reacts most strongly for cellulose) that progresses from a round precyst to a double layered wall form is prominent. The cysts contain a variable number of ostioles with closed apertures. Under scanning electron microscopy, the round to 35 ADVANCES IN APPLIED MICROBIOLOGY, VOLUME 4 3 Copyright 0 1997 by Academic Press, Inc All rights of reproduction in any form reserved. 0065-2164197 1625.00

36

D. G. AHEARN AND M. M. GABRIEL

FK:.1. Cysts and trophozoites of Acanfhnmoeba castellonii on surfaces in contact lens cases contaminated with bacteria and fungi. Note that bacteria adhered to surface of cyst (top right) and variable (round to stellate-like) cysts (top left); characteristic acanthapodia of trophozoite (bottom left]. Trophozoites “capped” with ensriared cells of Pseudoinonas aeruginnsa in PBS (bottom right).

stellate cysts appear polyhedral with ridges defining the facets. Cysts may have bacteria or fungi adhering to their surfaces (Fig. 1). Pussard and Pons (1977) divided the genus into three subgroups on the basis of cyst morphology: Group I-deeply scalloped stellate endocysts with a rounded, slightly rippled or smooth wall, ostioles at tips of the rays, ectocyst and endocyst walls usually separated; Group 11-cysts generally smaller than those in Group I, globular, ovoid, polygonal or triangular endocysts with ectocyst wrinkled or rippled, ostioles usually at the angles of the rays; Group 111-cysts round to slightly angular to irregular with ectocyst wall thin and delicate, single or obscure ostioles. A listing of species as to their probable group on the basis of cyst morphology is presented in Table I. These groupings are helpful in distinguishing some species, but variabilities in cyst morphology between strains and with culture conditions make some species assignments difficult or arbitrary. Species in Group I are generally larger than those in the other two groups. Supplemental information such as isoen-

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

37

TABLE I PROVISIONAL GROWINCS OF ACANTHAMOEBA SPECIES ON THE BASISOF CYST MORPHOLOGY" Group I Mostly smooth ectocyst, stellate scalloped endocyst

A. astronyxis (Ray and Hayes, 1954) Page, 1967

Group I1 Mostly wrinkled ectocyst, variable endocyst

Group I11 Ectocyst thinly rippled to smooth, round to slightly angular endocyst

A. castellanii (Douglas, 1930) Page, 1967' (9-14, 11)

A. culbertsoni Singh and Das, 1970' (13-28, 18)

A. hatchetti Sawyer, Visvesvara, and Harke,

A. jacobsi Sawyer, Thomas, Nerad, and Visvesvara, 1992 cysts spherical (12.5-1 7.5, 15)

(16.0-28.0, 19)h

A. comandoni Pussard, 1964 (21-30, 22)

1977' (11.5-21, 16)

A. echinulata Pussard and Pons, 1977 (10-32, 25)

A. polyphaga (Puschkarew, 1917) Page, 19fi7" (9-16, 13)

A. palestinensis (Reich, 1933) Page, Page, 1967 (13-22, 18)

A. griffini Sawyer, 1971r (14-25, 20)

A. rhysodes (Singh, 1652) 1979' (10-26, 18)

A. lenticulata Molet and Ermolieff-Braun, 1976 (n/a)

A. pearcei Nerad, Sawyer, Lewis, and McLaughlin,

A. stevensoni Sawyer, Nerad, Lewis, and McLaughlin, 1993

A. royreba Willaert, Stevens, and Tyndall, 1978 (n/a)

1995 (17.5-25, 21)

Singh and Hanumaiah,

(10-23, 17)

A. tubiashi Lewis and Sawyer, 1979 (18-30, 23) "Based on Pussard and Pons (1977) and Sawyer rt al. (1992: 1993). bRanges and mean diameters of cysts in @m in parenthesis ( 1 are approximated from literature reports and observations of representative cultures: nla = data not available. "Species associated with human eye infections.

zyme profiles (propionyl esterase, leucine aminopeptidase, and acid phosphatase) and temperature tolerance for growth are necessary for distinguishing species (Stratford and Griffiths, 1978; Sawyer et al., 1993; Nerad et al., 1995). Molecular studies have further demonstrated the classification complexities within Acantharnoeba. Restriction-fragment-length polymorphism (RFLP) analyses of mitochondria1 DNA (mtDNA) of isolates of A. castellanii and A. polyphaga indicate that the two species may represent a single species complex (Byers et al., 1983; Yagita and Endo, 1990). Digestion with BgIII, EcoRI, and Hind111 of whole-cell DNA of 33 clinical isolates ostensibly representing these two species produced RFLPs that differentiated the strains into seven multiple-strain and three sin-

38

D. G. AHEARN AND M.M. GABRIEL

gle-strain groups (Kilvington et al., 1991). The study included four strains of A. polyphaga (identified on the basis of cyst morphology) that were placed among three RFLP groups. The authors indicate that the number and positions of the EcoRI RFLPs were identical to those obtained by Byers et al. (1983) in their study of mtDNA from A. castellanii. This fact and the similarity in sizes (about 40 kb) of the RFLPs indicated that their RFLPs originated from mtDNA. Ledee et al. (1996) studied PCR-amplified nuclear 18s rDNA genes from three isolates of A. griffini. These genes, approximately 2800 bp in length in A. griffini and A. Zenticulatu, contain a Group I Intron (Gast et al., 1994). Comparison of the sequences of the RFLPs from isolates of A. griffini, one from an eye, one from tap water, and one from a contact lens associated with the same individual, showed that the isolates were identical and differed by less than 1%in bp sequences from the type culture (Ledee et al., 1996). The possible presence of endosymbionts in Acanthamoeba spp. presents an added difficulty in comparisons of mtDNA and in PCR-based studies. Byers (1986) in his extensive review expresses the need for sequence data on more conservative regions of DNA because of high interstrain variability in the mtDNA. Additional data on strains of varied environmental and clinical origin showed no overall correlation between mtDNA fingerprint groups and environmental source and further demonstrated the diversity of mtDNA fingerprint groups (Gautom et a]., 1994). 11. Biology

Acanthamoeba species are voracious predators of various Gram-negative bacteria, cyanobacteria, and fungi (Nero et al., 1964; Wright et al., 1981; Schuster et a]., 1993). Species are cannibalistic or pathogenic and may be grown in axenic enrichment culture and on defined media. An enrichment broth of peptone (2.0%), glucose (l.80/0),and yeast extract (0.2%)in a basic salts mixture is frequently employed for axenic culture of Acanthamoeba, but peptose-glucose broths fortified with salts, or skim milk broth, tissue culture fluids, etc., will support growth of many strains (Neff, 1957). Acanthamoeba spp. may be cultivated also in defined media composed of basic salts, various mixtures of amino acids, biotin, and thiamine and fortified with acetate or glucose (Adam, 1959, 1964). The generation times for Acanthamoeba species vary with the strain and culture conditions but in axenic culture frequently range from 13 to 18 h at 25°C.

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

39

Neff (1957) demonstrated that live or dead Aerobacter aerogenes and Escherichia coli and dead cells of Saccharomyces cerevisiae served as excellent food sources for the growth of an isolate of Acanthamoeba. Live yeasts, including Candida famata, Cryptococcus neoformans, and Rhodotorula rubra, are known to support the growth of various strains of Acanthamoeba (Castellani, 1930; Nero et al., 1964; Bunting et al., 1979). Schuster et al. (1993) found that A. castellanii migrated in a wateragar medium to both Gram-negative and Gram-positive bacteria but preferably ingested Enterobacter cloacae, Klebsiella pneumoniae, Shigella boydii, and Bacillus cereus. A weak feeding response was observed to Serratia marcescens and the response to Staphylococcus aureus was moderate; Pseudomonas aeruginosa was not ingested. Stenotrophom o n m (Xanthonomonas) maltophilia cocultivated in saline with A. castellanii or A. polyhaga for 24 h supported more luxuriant growth of the amoebae than Escherichia coli (Bottone et a]., 1994). Pseudomonas aeruginosa, a species that may be lethal for Acanthamoeba species (Qureshi et al., 1993), E. coli, Staphylococcus aureus, and S. epidermidis were only sparsely internalized. Bottone et al. (1994) also noted that a “capping” phenomenon occurred with Stenotrophomonas maltophilia and F! aeruginosa where the bacterial cells seemed to be ensnared in mass in a sticky-surface substance secreted by the amoebae. Capping was not observed with E. coli nor the nonmotile S. epidermidis. Most Acanthamoeba species will grow on live or dead E. coli (the use of E. coli for the isolation and enumeration of Acanthamoeba is a standard practice), but growth on staphylococci and particularly pigmented Gram-negative bacilli is frequently reported as negative. The inhibition of Acanthamoeba species by bacteria may be associated with toxic pigments or other metabolic by-products (Singh, 1945). Nevertheless, inhibitory bacteria such as r! aeruginosa and S. marcescens have been found to support the growth of A. castellanii in saline (Wang and Ahearn, in press). Various investigators have indicated that Acanthamoeba migrate to bacteria that are not internalized. The food preferences demonstrated by strains of Acanthamoeba in laboratory studies are most likely influenced by nutrient carryover with the bacterial inocula, the metabolic state of the bacteria, and the density ratios and strains of amoebae to bacteria. Certain bacteria and yeasts ingested by strains of Acanthamoeba may be maintained as endosymbionts for varied time periods (Drozanski and Chmielwski, 1979; Fritsche et al., 1993). Bacillus spp. and Saccharomyces spp. may be maintained in amoebae cysts, whereas Legionella spp. may proliferate in the trophozoites (Moffat and Tompkins, 1992).

40

D. G. AHEARN AND M. M. GABRIEL

The cysts of Acanthamoeba species are produced under varied conditions that include low temperature and low nutrients, age in culture, and the presence of inhibitors. A. castellanii and A. palestinensis may produce over 70% cysts within 48 h when shifted from enrichment media to 0.5 to 0.8M MgC12, or to NaCl or KC1 in the case of the latter species (Griffiths and Hughes, 1968; Lasman and Shafran, 1978). An encystment factor produced by transforming cells induces or enhances cyst formation by both these species (Akins et al., 1985; Lasman, 1987).

I l l . Infections Acanthamoeba species are the cause of rare human infections. Visvesvara (1995) includes A. culbertsoni, A. castellanii, A. polyphaga, A. astronyxis, A. healy, and A. divionensis as agents of chronic granulomatous amoebic encephalitis. This disease of the central nervous system occurs primarily among the immunosuppressed or chronically ill, particularly AIDS patients. Only about 100 cases have been reported worldwide (Visvesvara, 1995). The first reported case of Acanthamoeba keratitis developed in 1973 in a Texan farmer who had rinsed an eye with water from a horse trough following trauma to the eye from a shaft of straw (Jones et a)., 1975). Both trophozoites and cysts of A. polyphaga were isolated repeatedly from corneal scrapings. Shortly afterwards, two eye infections caused by Acantharnoeba were recognized in the United Kingdom. One occurred in a teacher of 32 years who had mild unilateral keratoconjunctivitis and uveitis that did not respond to treatment. Within 6 months, the eye deteriorated, with corneal ulceration, pain, and loss of vision. A. polyphaga was isolated frequently from the affected eye. A corneal graft was eventually rejected. The second case involved a farmer of 59 years with very similar eye conditions (unilateral central or paracentral corneal infiltrate) who eventually lost his eye. Acanthamoeba was isolated from his eye tissues (Nagington et al., 1974). During the next few years, a series of reports associated amoebic keratitis with contact lens wear (CDC, 1986; Kyle and Noblet, 1987; Auran et al., 1987; Stehr-Green et al., 1989). In contrast to the mostly fatal granulomatous encephalitis, Acanthamoeba keratitis (also a rare disease) occurs most commonly among healthy normal individuals and seems associated with the minor trauma to the eye that accompanies contact lens wear. The species diagnosed from eye infections, in order of their decreasing frequency, are A. castellanii, A . polyphaga, A. rhysodes, A. culbertsoni, A. hatchetti, and A. griffini.

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

41

The symptoms of Acanthamoeba keratitis include a foreign body sensation followed by severe pain, tearing, photophobia, and blepharospasm. The syndrome emerges over a period ranging from only a few days to months. Initial corneal changes may range from opaque streaks to epithelial opacities, or corneal epithelial stippling and microcystic edema. The syndrome worsens to dense central infiltrates with thinning, circumferential peripheral guttering, and peripheral vascularization. In most cases, the recurrent cycle of healing and breakdown in epithelium overlying the stromal infiltrate is apparent (Cohen et al., 1987; Lindquist eta]., 1988).The disease responds poorly to chemotherapy and has frequently resulted in loss of vision. Initially, amoebic keratitis associated with contact lens wear was frequently misdiagnosed; the disorder and its lesions were often confused with fungal, bacterial, or herpetic keratitis. It is not clear whether cysts or trophozoites or both are the infectious entities in Acanthamoeba keratitis, nor is there information available on an infectious dose. It is clear, however, that Acanthamoeba spp. are common in the environment and that Acanthamoeba keratitis is of rare occurrence, much less frequent than contact-lens-related Gram-negative bacterial keratitis. In turn, contamination of contact lens cases with Gram-negative bacteria is not uncommon, and Acanthamoeba may occur in storage cases of wearers with no symptoms of infection. Failure to comply with proper hygienic procedures for cleaning and disinfecting lenses appears to be the major factor in most contact-lens-related eye infections, although some disinfectant systems are less forgiving of deviations from recommended regimen. Bacterial and fungal colonization of the lens case appears to be the primary step in most Acanthamoeba keratitis related to contact lens wear. Proliferation of the amoebae in the case is followed by transport of the amoeba to the eye via the lens. Trophozoites of A. castellanii and A. polyphaga will bind to the cornea and seem to share a mannose-type receptor with l? aeruginosa (Morton et a]., 1991). Predisposition of the cornea of the eye from the minor trauma of contact lens wear or by a coinfecting agent follows binding of the trophozoite. Cysts do not adhere to submerged surfaces, but they do become entrapped in biofilms and the cyst may potentially be embedded in the lens. The cyst itself could traumatize the cornea. Recently, Mathers et al. (1996) suggested that Acanthamoeba keratitis is more prevalent in the United States than suspected. In a tandem-scanning-confocal-microscopy screening of 2 1 7 keratitis patients where 5 1 patients were suspected to be infected with Acanthamoeba, 36 of these demonstrated highly refractile ovoid bodies typical of Acanthamoeba.

42

D. G. AHEARN AND M. M. GABRIEL

None of the patients were culture-positive for Acanthamoeba; although epithelial smears demonstrated trophozoite- and cyst-like cells. After antiamoebic therapy was commenced (polyhexamethylene or, preferably, chlorhexidine biguanide, each at 0.02% and in combination with propamidine), 43 patients had fewer symptoms, clinical signs, and refractile bodies indicated by tandem-scanning-confocal microscopy. Only 1 7 of the 43 patients were wearing contact lenses when symptoms developed. The investigators suggest that their apparent mini-epidemic of amoebic keratitis could indicate a much more common infection, possibly caused by a new species of amoebae. Detection of Acanthamoeba keratitis by confocal microscopy, a rapid, noninvasive in vivo procedure, is of value for early diagnosis of the disease. The diagnosis is accurate even when culture results are negative (Pfister et al., 1996). Various chemotherapeutic agents have been used in vitro: propamidine, dibromopropamidine (Wright et al., 1985), hydroxystilbamidine, paromomycin, 5-fluorocytosine (Jones et a]., 1975), and clotrimazole (Borochovitz et al., 1981; Stevens and Willaert, 1980) for the treatment of Acanthamoeba keratitis. Kilvington and White (1994) suggested that propamidine is the drug of choice in the treatment of Acanthamoeba. At the present time, propamidine combined with polyhexamethylene biguanide (PHMB) or propamidine combined with chlorhexidine digluconate have been employed successfully (Hay et al., 1994; Seal et al., 1995b; Ledee et al., 1996). The biguanides chlorhexidine and PHMB are used topically at concentrations of 0.02%. They have broad-spectrum antimicrobial activity against bacteria and fungi as well as amoebae and are included as disinfectants or preservatives in various products. Their broad antimicrobial activity may be important in treatment of Acanthamoeba keratitis because there is a possibility that at least with some strains of amoeba virulence is related to the presence of coinfecting bacteria or endosymbionts (see Badenoch, 1991). IV. Ecology

Species of Acanthamoeba dwell in damp soils and upper zones of mud in rivers and lakes and in anthropogenic water reservoirs (Daggett et al., 1982; Sawyer et al., 1987; Bhattacharya et al., 1987; Martinez, 1985; Visvesvara, 1986). Cysts and trophozoites are carried by humans and distributed by wind, water, and animals. Extensive studies by Sawyer and his colleagues have demonstrated that distributions of Acanthamoeba in soils, fresh waters, and marine coastal habitats may coincide with sewage contamination (see Sawyer, 1989; Sawyer et al., 1993; Sawyer et al., 1996). Even species initially described from true

CONTACT LENSES AND ACANTHAMOEBA KEFUTITIS

43

marine habitats probably are of terrestrial or estuarine origin (Bhattacharya et al., 1987; Nerad et al., 1995; Sawyer et al., 1996). Species of Acanthamoeba readily colonize surface biofilms in swimming pools, hot tubs, heating-, ventilation-, and air-conditioning systems, and domestic water taps and eye-wash stations (Martinez, 1985; Bier and Sawyer, 1990; Paszko-Kolva et al., 1991; Mergeryan, 1991). The association of eye infection with contact lens wear in the United States coincided with the use of home-prepared saline solutions (CDC, 1986, 1987). The amoebae grow on microbial contaminants in nonpreserved saline or inadequately preserved storage solutions and are transferred on the contact lens to the eye (Ubelaker, 1991). As many as 50% of normal contact lens wearers have contaminating microorganisms in their contact lens cases at some time during their use of lenses (Wilson et al., 1990). Donzis et al. (1989) reported that Acanthamoeba spp. contaminated lens cases and solutions only when bacteria or fungi were present. No Acanthamoeba spp. were found in the parent bottles of disinfectant solutions. Of ten individuals with Acanthamoeba spp. in their lens care systems, Gram-negative bacteria, mainly Pseudomonas spp., were also present; in six instances in conjunction with Bacillus spp. More than 70% of the cornea tissues and lens cases of patients with Acanthamoeba keratitis were found to be cocontaminated with bacteria and amoebae (Bacon et d.,1993). Giovannini et al. (1994) suggested that the early onset of bacterial conjunctivitis may enrich the natural conjunctival microbiota, as well as the contact lens case, with a n abundant bacterial supply, thus favoring the growth and the development of virulent Acanthamoeba spp. in the lens case and eye. In vivo experiments with rats have shown that the virulence of Acanthamoebae spp. in keratitis is enhanced by coinoculation with bacteria (Lawin-Brussel et al., 1993). Bacterial numbers in a system might decrease from phagocytosis by amoebae or bacterial numbers may increase because of their higher growth rate (Bamforth, 1985; Griffiths, 1990). Larkin et al. (1990) proposed that bacteria in the external eye or in the lens system could stimulate cysts of Acanthamoeba spp. to excyst to the infectious vegetative stage (trophozoites). Trophozoites have not been observed to grow in unadulterated contact lens solutions, and, though encystment may result, the cysts show poor viability. Cysts, however, particularly those from cultures grown on Stenotrophomonas and Escherichia, will persist for at least up to several weeks in otherwise microbe-free disinfectant solutions. In the United States, the withdrawal of salt tablets for home-prepared saline from the market in the mid-1980s resulted in a decreased incidence of Acanthamoeba keratitis. In England and Asia during the

44

D. G. AHEARN AND M. M. GABRIEL

199Os, reports of Acanthamoeba keratitis seem to be increasing (see Kilvington and White, 1994). Rinsing lenses in tap water (particularly if the water supply is from a roof-storage facility deficient in disinfectant) may also be implicated in Acanthamoeba keratitis (Talamo and Larkin, 1993; Ledee et al., 1996). In the report by Ledee et al. (1996), the same strain of A. griffini was isolated from the cornea, the contact lens storage case, and the domestic water supply. Contact-lens-associated Acanthamoeba keratitis in Hong Kong and England in the mid-1990s may be related to disinfectant deficiencies in reservoir-type water systems. In general, primary contamination of the lens care system, particularly with Gram-negative bacteria, including Pseudomonas spp., seems to be a prerequisite for the establishment of infectious populations of Acanthamoeba. Studies that indicate that Acanthamoeba spp. and P aeruginosa are mutually exclusive in lens cases may not recognize that survival or presence of both bacterium and amoeba in the lens case varies with time and population dynamics.

V. Disinfection

Various reports have indicated that cysts of Acanthamoeba species may resist desiccation from months to years, tolerate 2.0% concentrated HC1 (0.25 M) and resist exposure to chlorine, chlorides, ozone, and peroxides (Sawyer, 1970; Cursons et al., 1980; Bryant et a]., 1982). Penley et al. (1989) found that trophozoites and cysts of two corneal isolates of Acanthamoeba survived in most contact lens solutions beyond the manufacturer’s recommended disinfectant times. Only an isopropyl alcohol cleaning system (now approved for disinfection), and an “02 conditioner” killed both (2 x lo3) cysts and trophozoites of A. polyphaga and A. castellanii within 30 min of exposure. Solutions with chlorhexidine (0.005 and 0.006%) and chlorhexidine and alkyl triethanol ammonium chloride (0.013%) plus thimerosal (0.002%) appeared to be the next most effective against A. polyphaga; no viable amoebae were recovered at 18 h. An isolate of A. castellanii was more resistant, surviving in all tested solutions (except the O2 conditioner and isopropyl alcohol). A similar pattern of relative efficacies for contact lens solutions was reported by Connor et al. (1989) for cysts of A. culbertsoni; the isopropyl alcohol solution and solutions containing thimerosal appeared more effective than hydrogen peroxide or solutions with PHMB.

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

45

Ludwig et al. (1986) indicated that heat disinfection (80°C for 10 min) in a preserved saline solution (NaCl 0.7% with sodium borate, boric acid and thimerosal O.OOl%, and EDTA 0.1%) killed pellet inocula of 5 x lo2 cysts and 5 x lo3 trophozoites of A . castellanii and A. polyphaga suspended in 2 ml of disinfectant solution. A chlorhexidine (0.005°/~)/thimerosal(O.OO1°/o) solution after 4 h at room temperature killed the trophozoites and cysts of A. castellanii but not of A. polyphaga. Both species of Acanthamoeba survived 4-hour exposure to alkyl triethanol ammonium chloride (0.013%)/thimerosal 0.002% and chlorhexidine (0.005%), and 10 and 30 min exposure to hydrogen peroxide (3%). Davies et al. (1990) found that strains of A. polyphaga and A. castellanii (lo3 to lo4 cysts per milliliter) were killed by the triethanol ammonium chloride-thimerosal solution within 4-6 h. Hydrogen peroxide (3%) was effective at 4 h, whereas solutions with 0.001% polyquaternium and 0.00005% polyaminopropyl biguanide (equals polyhexamethylene biguanide or PHMB) yielded viable amoebae after 24-h exposure. Zanetti et al. (1995) studied disinfectant solutions diluted in a peptone-yeast extract-glucose medium. Hydrogen peroxide (1.5%) killed 100% of the trophozoites (1.5 x 104/ml) of a clinical strain of A. castellanii within 30 min. This solution was more effective than a diluted chlorhexidine-thimerosal solution (100% kill at 3 h). A mixed cyst-trophozoite suspension (over 90% trophozoites) yielded viable amoeba after 6 h but not after 9 h of exposure to these same preparations. Other diluted commercial disinfection solutions containing benzalkonium chloride, PHMB, or polyquaternium yielded viable amoeba after 9 h exposure to the disinfectants. Connor et al. (1989) compared the effectiveness of chlorhexidine digluconate at concentrations from 0.0001% to O . O l % , a commercial solution with PHMB (0.0005%), hydrogen peroxide YO), thimerosal (O.OOZ%), and heat (80°C for 10 min) against cysts (1.5 x lo4 per ml) of A. culbertsoni. Chlorhexidine at 0.005% appeared equivalent to the PHMB solution; 42% survival of the cyst inocula was reported after 24-h exposure. Hydrogen peroxide for 60 min and thimerosal after 24 h appeared least effective, whereas heat was most effective, but still 6% of the cyst inocula was reported to survive 80°C for 10 min. Viability of cysts after exposure to the disinfectants was based on hemocytometer counts of calcofluor white-stained cysts that showed a decrease in cyst number. The surviving cysts were incubated in thioglycolate broth, and cells were enumerated 1 week later to determine if any growth had occurred. Growth occurred in all systems. When the concentration of chlorhexidine was increased (0.09%),the cysts were destroyed.

46

D. G. AHEARN AND M.M.GABRIEL

The biguanides chlorhexidine and PHMB have demonstrated clinical efficacy at concentrations of 0.02% (Larkin et al., 1992). Burger et al. (1994) found that PHMB at 0.009% (90 ppm) in a borate buffer killed over 98% of cysts (1O5-1O6/ml) of A. castellanii and A. polyphaga within 30 s. Hay et al. (1994), in an extensive comparison and pairing of 1 2 antimicrobials against 20 isolates of Acanthamoeba, found that susceptibilities varied with the strain, but chlorhexidine and PHMB were the most active single compounds after 48-h exposure. Concentrations below 0.005% were active against 2 x lo4 cysts and trophozoites of most strains. The two biguanides each showed slight in vitro synergy when in combination with pentamidine. In a peptone-yeast extractglucose medium, concentrations of 0.005 to 0.01% of either chlorhexidine or PHMB were biocidal at 24 h for 105 trophozoites or cysts per ml of four Acanthamoeba spp. (Tirado-Angel et al., 1996). The biocidal effect was dependent upon exposure time, concentration of inocula and inhibitor, and strain. Some synergism between the two biguanides was noted. Nearly all the above investigations noted that trophozoites were more susceptible than cysts to inhibitors. Most indicated that biguanides, particularly chlorhexidine, were the more active antiamoebal compounds. The antimicrobial activities usually showed a concentration (both of inocula and inhibitor) and time dependence. The various discrepancies in the data on the relative efficacies of commercial systems have been attributed to strain differences and procedural differences, particularly the quantitation of surviving cells. Trophozoites and some precysts lyse in the presence of preservatives, particularly biguanides. Dependent upon the strain of amoeba and other test conditions, some inhibitors (particularly at sublethal concentrations) may induce encystment. Some of the differences in the results also may be attributed to failure to quantitate the viability of “surviving” cysts and failure to use neutralizers in the recovery process (or the use of an inappropriate neutralizer). Also, cysts of some strains may have relatively poor viability or demonstrate highly variable degrees of recalcitrance (see Schuster and Jacob, 1992). We have made similar observations, particularly with cysts that have been produced in axenically maintained cultures. Such cultures may need to be grown on selected bacteria to regain the more typical resistant cysts. In particular, bacterial contamination may be of some importance in determining the resistance of cysts. Outgrowth of bacteria or yeasts (endosymbionts or exogenous contaminants) in the test system can stimulate trophozoites to feed and cysts to excyst, resulting in increased susceptibility of the amoeba to the disinfectant.

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

47

VI. Adherence to Lenses

Various reports have indicated that trophozoites adhere preferentially over cysts to contact lenses and that various species and strains differentially adhere to the four FDA hydrogel lens groups (Table 11). In a comparison of two nonionic lens types, a polymacon 38.6%-water-content lens and a lidofilcon 70%-water-content lens, Kilvington and Larkin (1990) found that both cysts and trophozoites showed greater adherence to the higher-water-content lenses. In contrast, John et al. (1991) reported lower adherence of A. castellanii to ionic, 58% watercontent etafilcon A lenses than to nonionic, 38% water-content polymacon lenses and nonionic, 70Y0 water-content lidofilcon lenses. Kilvington (1993) found different degrees of adherence to the four FDA contact lens groups for two strains of Acanthamoeba. Trophozoites of both strains showed greater adherence than cysts. Inocula of 1 x 106 washed cells suspended in normal saline were placed into wells of a microtiter plate, each well containing a single lens. After 4 h of incubation at room temperature, each lens was removed and gently submerged in two separate 10-ml volumes of saline. The lens was then placed in 1 ml of ice-cold saline for 15 min. The tube was vortexed for 10 s and the cells in suspension were counted. Trophozoites of one strain had greater adherence to a Group I lens (38% water content), with no significant differences in adherence to the other lenses (water contents ranging from 45 to 70%). The other strain showed equivalent adherence to Groups I, 111, and IV lenses and reduced adherence to the Group I1 lens (70% water content). Cleansing and rinsing procedures of all four lens types with commercial systems as per the manufacturer’s instructions removed all amoeba from the lenses. Kelley and Xu (1995) examined the effect of increasing concentrations of trophozoites of A. polyphaga on adherence to lens fragments for four different lenses: polymacon, nonionic, 38% water content; bufilcon A, ionic, 45% water content; etafilcon A, ionic, 58% water content; and lidofilcon A, nonionic, 70% water content. After 2 h of incubation of the lens materials in amoeba suspensions with agitation at room temperature, followed by three washes in phosphate-buffered saline (PBS) (volumes not given), the higher concentrations of trophozoites ( 105/lens fragment) adhered more to the nonionic 70%-watercontent lens than to the nonionic 45%-water-content lens. Adherence to the nonionic lenses was significantly greater than to the ionic lenses. With inocula concentrations per lens fragment of about 250-500 trophozoites, no significant difference in adherence to lidofilcon and bufilcon materials was detected. The etafilcon A lens at all inocula concentra-

TABLE I1 FDA GROUPINGS OF HYDROGEL CONTACTLENSES

Group 1 low water (50% H,O) nonionic polymers

Group 3 low water (50% H,O) ionic polymers

Tefilcon (38%) Tetrafilcon A (43%) Crofilcon (38%) Nefilcon A&B (45%) Isofikon (36%) Mafilcon (33%) Polymacon (38%)

Lidofilcon B (79%) Surfilcon (74%) Lidofilcon A (70%) Netrafilcon A (65%)

Bufilcon A (45%) Deltafilcon A (43%) Droxifilcon A (47%) Phemfilcon A (38%) Ocufilcon (44%)

Bufilcon A (55%) Perfikon (71%) Etafilcon A (58%) Ocufilcon B (53%) Ocufilcon C (55%) Phemfilcon A (55%) Methafilcon (55%) Vifilcon A (55%)

CONTACT LENSES AND ACANTHAMOEBA KERATITIS

49

tions appeared to have the lowest levels of adherence. Later, Kelly et al. (1995) submerged segments of polymacon, nonionic, 38%-water-content, etafilcon A, ionic, 58%-water-content, siloxane acrylate, and polymethylmethacrylate (PMMA) lenses in cell suspensions of A. CQStellanii (105/ml, trophoz0ite:cyst ratio, 90:lO) in a peptone-yeast extract-glucose medium. After incubation with agitation for varying periods u p to 2 h, the lens segments were submerged for 3 s in PBS and than examined under a microscope. The number of trophozoites adherent to the PMMA and RGP segments increased with time of incubation. This phenomenon was not observed for cysts. More trophozoites adhered to these lenses (mean adherence: 50 to 75 per mmz) than to the hydrogel lenses, which retained no more than 25 cells per mm2 after 2 h. No statistically significant differences in adherence existed between the two types of rigid lenses or between the two hydrogel lenses. In this latter study, where lens segments were incubated for 2 h with cells in a peptone-yeast extract-glucose medium, the mean number of adherent trophozoites, about 10 to 15 mm2, was considerably higher than those adhering from PBS (65% and optical purity of 99% were obtained for S-(-)-2J using each enzyme. S - ( - ) - B was obtained in 90% reaction yield and 99.8% optical purity using C. viscosum lipase under similar conditions. The stereoselective hydrolysis of dimethylesters of symmetrical dicarboxylic acids including meso-diacids such as cis-l,2-cycloalkane dicarboxylic acids and diacids with prochiral centers has been demonstrated by Mohr et al. (1983) using pig liver esterase (PLE). The product of these stereoselective hydrolyses, chiral monoacetate of dicarboxylic acids, were obtained with an enantiomeric excess (e.e.) from 10 to 90% depending upon the substrate. Enantioselective hydrolysis of cis-1,2-

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

107 NHZ

I

coa

HCOOH

Formnte Dehydrogsnass CBZ-L-lysine

33

L-z-Oxylyslne 32

Keto d d 34

i"; 0

C.nnoprll3l FIG.12. Synthesis of chiral synthon for ceranopril. Conversion of CBZ-L-lysine 33 to L-Z-oxylysine 32 (Pate1 et al., '1992b). Reprinted with permission from Springer-Verlag.

diacetoxycycloalkane and 2-substituted 1,3-propanediol diacetate have been demonstrated using PPL by Laumen and Schneider (1985) and Tombo et al. (1986), respectively. Ceranopril 31 is another ACE inhibitor (Karenewsky et al., 1988) that requires chiral intermediate carbobenzoxy (CBZ)-L-oxylysine 32 (Fig. 12). A biotransformation process was developed to prepare CBZ-Loxylysine (Hanson et al., 1992). NE-CBZ-L-lysine 33 was first converted to the corresponding keto acid 34 by oxidative deamination using cells of Providencia alculifaciens SC-9036 that contained L-amino acid oxidase and catalase. The keto acid 34 was subsequently converted to CBZ-L-oxylysine 32 using L-2-hydroxyisocaproate dehydrogenase from Lactobacillus confisus. The NADH required for this reaction was regenerated using formate dehydrogenase from Candidu boidinii (Fig. 12). The reaction yield of 95% along with 98.5% optical purity was obtained in overall process. V. Anticholesterol Drugs

Chiral P-hydroxy esters are versatile synthons in organic synthesis specifically in the preparation of natural products (Mori and Tanida,

108

R. N. PATEL

1984; Hirama and Uei, 1982; Gopalan and Sih, 1984). The asymmetric reduction of carbonyl compounds using Baker’s yeast has been demonstrated and reviewed (Ward and Young, 1990; Csuk and Glanzer, 1991; Sih et a]., 1984). In the stereoselective reduction of P-ketoester of 4chloro and 4-bromo-3-oxobutanoic acid, specifically 4-chloro-3-oxobutanoic acid methylester, Sih and Chen (1984b) demonstrated that the sterteoselectivity of yeast-catalyzed reductions may be altered by manipulating the size of the ester group using y-chloroacetoacetate as substrate. They also indicated that the e.e. of the alcohol produced depended upon the concentration of the substrate used. Nakamura et al. (1989) demonstrated the reduction of P-keto ester with Baker’s yeast and controlled stereoselectivity by the addition of a$-unsaturated carbony1 compounds. The additive tended to shift the stereoselectivity of the reduction reaction toward the production of R-hydroxy ester. The shift in stereoselectivity was accounted for based on inhibition of the competitive enzyme that produced S-hydroxy ester. They also used immobilized Baker’s yeast to improve the stereoselectivity of the reduction reactions (Nakamura et a]., 1990). The enantiomeric excess of alcohols produced was improved to 90% by using immobilized cells compared to 31% obtained with free cells when methyl 4-chloroacetoacetate was use as the substrate. Pate1 et al. (1992d) have described the reduction of 4-chloro-3-oxobutanoic acid methyl ester 35 to S-(-)-4-chloro-3-hydroxybutanoicacid methyl ester 36 (Fig. 13) by cell suspensions of Geotrichurn candidurn SC-5469. S-(-)-E is a key chiral intermediate in the total chemical synthesis of 37, a cholesterol antagonist that acts by inhibiting hydroxymethyl glutaryl CoA (HMG-CoA)reductase. In the biotransformation process, a reaction yield of 95% and optical purity of 96% were obtained for &6 by glucose-, acetate-, or glycerol-grown cells (10% wthol) of G. candidurn SC-5469. Substrate was used at a concentration of 10 g/liter. The optical purity of S was increased to 99% by heat treatment of cell suspensions (55% for 30 min) prior to conducting bioreduction of 35. Glucose-grown cells of G. candidurn SC-5469 have also catalyzed the stereoselective reduction of ethyl-, isopropyl-, and tertiary-butyl esters of 4-chloro-3-oxobutanoic acid and methyl and ethyl esters of 4-bromo3-oxobutanoic acid. A reaction yield of ~ 8 5 %and an optical purity of >94% were obtained. NAD+-dependent oxidoreductase responsible for the stereoselective reduction of P-keto esters of 4-chloro- and 4-bromo3-oxobutanoic acid was purified 100-fold. The molecular weight of purified enzyme was 950,000 Da. The purified oxidoreductase was

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

109

Q F

HMGCoA Reductme InhibitorX FIG.13. Synthesis of chiral synthon for anticholesterol drug. Stereoselective reduction of 4-chloro-3-oxobutanoic acid methyl ester 35 (Pate1 et al., 1992a). Reprinted with permission from Elsevier Science Inc.

immobilized on a Eupergit C and used to catalyze the reduction of 35 to S. The cofactor NADH required for the reduction reaction was regenerated by glucose dehydrogenase. Nakamura e f al. (1991) have isolated four different oxidoreductases from Baker's yeast that catalyzed the reduction of P-keto esters to P-hydroxy esters. Two oxidoreductases, namely, D-enzyme-1 (molecular weight 25,000) and D-enzyme 2 (MW 1,600,000 Da), catalyzed the reduction of P-keto ester stereoselectivity to D-P-hydroxy ester. In contrast, one other enzyme, L-enzyme (MW 321,000 Da), reduced substrates to L-J3-hydroxy esters. The NADP-oxidoreductases, designated the Sand R-enzymes, have been purified and characterized from cell extracts of Saccharomyces cerevisiae, and catalyze the enantioselective reductions of ~ - o x o - 4-oxo-, , and 5-0x0-esters (Heidlas et al., 1988). The S-enzyme had a molecular weight of 48,000 and reduced 3-oxo-esters, ~ - o x o -and , 5-0x0-acids and esters enantioselectively to S-hydroxy compounds in the presence of NADPH. This enzyme may be located in the mitochondria1 fraction. The R-enzyme, which had a molecular weight of 800,000 and contained subunits having molecular weights of 200,000 and 210,000, specifically reduced 3-0x0-esters to R-hydroxy esters using NADPH as coenzyme. The R-enzyme, which occurs in the cytosol, was considered to be identical to a subunit of the fatty acid synthetase complex.

110

R. N. PATEL

Most microorganisms and enzymes derived therefrom have been used in reduction P-keto or a-keto compounds involved reduction of singleketo groups (Jones and Beck, 1986; Keinan et al., 1986; Patel et d.,1981; Bradshaw et al., 1992a; Christen and Crout, 1988). Patel et al. (1993b) have demonstrated the stereoselective reduction of a diketone 3,5-dioxo-6-(benzyloxy) hexanoic acid ethyl ester 38 to (3S,5R)-dihydroxy-6(benzyloxy) hexanoic acid ethyl ester (Fig. 14). The compound is a key chiral intermediate required for the chemical synthesis of [4-[4a,6~(E)ll-6-[4,4-bis-[4-fluorophenyl)-3-(l-methyl-lH-tetrazol-5-yl) -1,3-butadienyl]-tetrahydro-4-hydroxy-2H-pyren-2-one. Compound Ra new anticholesterol drug, acts by inhibition of HMG-CoA reductase (Sit et al., 1990). Among various microbial cultures evaluated for the stereoselective reduction of diketone 38, cell suspensions of Acinetobacter calcoaceticus SC-13876 reduced 38 to -. The reaction yield of 85% and optical purity of 97% were obtained using glycerolgrown cells. The substrate was used at 2 g/liter and cells at 20% (wt/vol, wet cells) concentration. Cell extracts of A. calcoaceticus SC-13876 in the presence of NAD+, glucose, and glucose dehydrogenase reduced 38 to the corresponding monohydroxy compounds 41 and 42 (3-hydroxy-5-oxo-6-benzyloxy hexanoic acid ethyl ester 41, and 5-hydroxy-3-oxo-6-benzyloxy hexanoic acid ethyl ester 4 2 ) . Both 41 and 112 were further reduced to (3S,4R)-dihydroxy compound by cell extracts (Fig. 14). A reaction yield of 9 2 % and an optical purity of 99Y0 were obtained when the reaction was carried out in a 1-liter batch using cell extracts. The substrate was used at 10 g/liter. Product 39a was isolated from the reaction mixture in 72% overall yield. The GC and HPLC area purity of the isolated product was 99% and the optical purity was 99.5%. Reductase that converted 38 to was purified about 200-fold from cell extracts of A. calcoaceticus SC-13876. The purified enzyme gave a single protein band on SDS-PAGE corresponding to 33,000 Da. Using a resolution process, chiral alcohol R-(+)-@ was also prepared by the lipase-catalyzed stereoselective acetylation of racemic 40 in organic solvent (Patel et al., 1992e). They evaluated various lipases, among which lipase PS-30 (Amano International Enzyme Co.) and BMS lipase efficiently catalyzed acetylation of the undesired enantiomer of racemic 40 to yield S-(-)-acetylated product 3 and unreacted desired R-(+)-* (Fig. 15). A reaction yield of 49 MY" (theoretical maximum 50 MYv) and an optical purity of 98.5% were obtained for R - ( + ) - a when the reaction was conducted in toluene as solvent in the presence of isopropenyl acetate as acyl donor. Substrate was used at a concentration

-

(+)-a,

-

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

0

111

FIG.14. Synthesis of chiral synthon for anticholesterol drug. Stereoselective reduction of 3,5-dioxo-6-~enzyloxy)hexanoic acid ethyl ester 38 (Pate1 et al., 1993b). Reprinted with permission from Elsevier Science Inc.

+ +

N

F F

+ ____)

F F

FIG.15. Stereoselective acetylation of racemic 40 to S-(-)-acetate 43 and R-(+)-m, an anticholesterol drug (Pate1 et al., 199Ze). Reprinted with permission from Springer-Verlag.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

113

of 4 g/liter. In methyl ethyl ketone at a 50-g/liter substrate concentration, a reaction yield of 46 M% and optical purity of 96.4% were obtained for R-(+)-@. Lipase PS-30 was immobilized on Accurel PP and the immobilized enzyme reused five times without any loss of activity or productivity in the resolution process to prepare R-(+)-@. The enzymatic process was scaled up to a 640-liter preparative batch using immobilized lipase PS-30 at 4 g/liter and racemic substrate 40 in toluene as solvent. From the reaction mixture R - ( + ) - a was isolated in 35 M*h overall yield with 98.5% optical purity and 99.5% chemical purity. The undesired S-(-)-acetate 43 produced by this process was enzymatically hydrolyzed by lipase PS-30 in a biphasic system to prepare the corresponding S-(-)-alcohol @. Thus, both enantiomers of alcohol 40 were produced by the enzymatic process. Pravastatin 44 and Mevastatin 45 are anticholesterol drugs that act by competitively inhibiting HMG-CoA reductase (Endo et al., 1976a). Pravastatin sodium is produced by two fermentation steps. The first step is production of compound ML-236B 46 by Penicillium citrinum (Endo et al., 1976a,b; Hosobuchi et al., 1993a,b). Purified 46 was converted to its sodium salt with sodium hydroxide and in the second step was hydroxylated to Pravastatin sodium 44 (Fig. 16) by Streptomyces carbophilus (Serizawa et al., 1983). A cytochrome P,,,-containing enzyme system has been demonstrated from S. carbophilus that catalyzed the hydroxylation reaction (Matsuoka et al., 1989). The chiral intermediate 2,4-didoexyhexose derivative required for HMG-CoA reductase inhibitors has also been prepared using 2-deoxyribose-5-phosphate aldolase (DERA).This enzyme accepts a wide variety of acceptor substrates and has been useful in organic synthesis (Barbas et a]., 1990; Chen et al., 1992). As shown in Fig. 1 7 , the reactions start with stereospecific addition of acetaldehyde 47 to a substituted acetaldehyde to form a 3-hydroxy-4-substituted butyraldehyde 48, which subsequently reacts with another acetaldehyde to form a 2,4-dideoxyhexose derivative 49 (Gijsen and Wong, 1994). DERA has been overexpressed in Escherichia coli.

VI. Antiinfective Drugs

During the past several years, synthesis of a-amino acids has been pursued intensely (Williams, 1989; O’Donnell, 1988; O’Donnell et al., 1989, 1970; Evans et al., 1987; Schmidt et al., 1989; Imada et a]., 1981) because of their importance as building blocks of compounds of medicinal interest, particularly antiinfective drugs. The asymmetric synthesis

114 R. N.PATEL

0

c

FIG. 16. Stereoselective hydroxylation of ML-236B to Pravastatin 44.

0

"i a' B

a

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

'

L 2 - J

a

FIG.17. Preparation of 2,4-dideoxyhexosederivatives 49 by aldolase (Gijsen and Wong, 1994). Reprinted with permission from the American Chemical Society.

115

R. N.PATEL

116

NADH

NAD*

\

H 2 N A o H

I

NH4*

Coda

a-Keto-&hydroxylsovalerate 52

L-&hydroxyvallneB

\COa

[(CH3)3N*CH2CH20H]a

Tigemonam 51 FIG.18. Synthesis of chiral synthon for tigemonam. Stereoselective reductive amination of a-keto-P-hydroxyisovalerate52 (Hanson et ol., 1990).

of P-hydroxy-a-amino acids by various methods has been demonstrated (Bold et al., 1989;Evans et al., 1987;Ito et al., 1988;Saito et al., 1985; Gordon et al., 1982)because of their utility as starting materials for the total synthesis of monobactam antiobiotics. L-P-hydroxyvaline So (Sykes e f a]., 1981) is a key chiral intermediate required for the total synthesis of an orally active monobactam, Tigemonam 51 (Fig. 18).The resolution of CBZ-0-hydroxyvaline by chemical methods has been demonstrated (Godfiey et al., 1986;Shanzer et al., 1979;Berse and Bessette, 1971). Leucine dehydrogenase from strains of Bacillus (Schutte et al., 1985;Monot et al., 1987)has been used for the synthesis of branchedchain amino acids. Hanson et al. (1990)have described the synthesis of ~-P-hydroxyvaline50 from a-keto-P-hydroxyisovalerate52 by reductive amination using leucine dehydrogenase from B. sphaericus ATCC-4525 (Fig. 18). NADH required for this reaction was regenerated by either formate dehydrogenase from Candida boidinii or glucose dehydrogenase from B. megaferium. Such immobilized cofactors as polyethylene glycol-NADH and dextrans-NAD were effective in the biocatalytic proc-

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

O HO Amoxicillin 99

'COpH

H

0

117

CHS

@-CON NH2 Cefadroxyl

a

Cephalexin51

FIG.19. Structure of antiinfective agents amoxicillin, cefadroxil, and cephalexin.

ess. The required substrate 52 was generated either from a-keto-P-bromoisovalerate or its ethyl esters by hydrolysis with sodium hydroxide in situ. In an alternate approach, the substrate 52 was also generated from methyl-2-chloro-3,3-dimethyloxiran-carboxylate and the corresponding isopropyl and 1,I-dimethyl ethyl ester. These glycidic esters are converted to substrate 52 by treatment with sodium bicarbonate and sodium hydroxide. In this process, an overall reaction yield of 98% and an optical purity of 99.8% were obtained for the L-P-hydroxyvaline So. D-phenylglycine is required for the semisynthetic antibiotic ampicillin, and D-hydroxyphenylglycine (Fig. 19) is used in the production of amoxicillin and cefadroxyl (Aida et al., 1986; Syldatk et al., 1987). The use of D-p-hydroxy phenylglycine will significantly increase because such new drugs as aspoxicillin, cefbuperazine, and cepyramide are expected to be marketed. Currently, D-amino acids are commercially produced by a chemoenzymatic route using D-hydantoinase. In this process, chemically synthesized DL-5-substituted hydantoin 53 is hydrolyzed to N-carbamoyl-D-amino acid 54 by microbially derived D-hydantoinase (Fig. 20). The latter compound undergoes rapid and spontaneous racemization under the reaction conditions; therefore, theoretically a 100% yield of 54 can be obtained. The compound 54 is further chemically converted to the corresponding D-amino acid 55 (Moller et al., 1988; Morin et al., 1986; Takahashi et al., 1979; Yamada et al., 1978; Kamphuis et a]., 1990b). Microbial N-carbamoylases have

118

R. N. PATEL

DL-5-8ubstltutd Hydantdn

Dsmlno acid I

1

D-HydantolnsM,

PN-carbmoyl acld 54

FIG.20. Synthesis of D-amino acids by D-hydantoinase and carbamoylase.

been demonstrated that catalyzed the conversion of N-carbamoyl-Damino acid to the corresponding D-amino acid. Some organisms contained both D-hydantoinase and N-carbamoylase activity (Olivieri et al., 1979; Runser et a]., 1990; Yokozeki et al., 1987; Kim and Kim, 1995). L-hydantoinase has also been described from a microbial source that catalyzes the conversion of DL-5-substitutedhydantoin to N-carbamoylL-amino acid. This process has been used in the production of L-amino acids (Yokozeki et al., 1987; Syldatk et al., 1987; Nishida et al., 1987; Tsugawa et al., 1966). D-amino acids and L-amino acids have also been prepared by D-Specific or L-specific acylases derived from microbial sources. In this process, DL-N-acetyl amino acid (Fig. 21) is resolved by hydrolytic reaction to yield the D- or L-amino acid 57 depending upon D- or L-selective acylase used in the reaction (Sugi and Suzuki, 1980; Chibata et al., 1988; Chenault et a]., 1989; Sakai et al., 1991). Sandoz commercialized the production of cyclosporin A, which is used in the treatment of transplant rejection and autoimmune disease. L-valine is an essential component of cyclosporin A (Ice and Agathos, 1989; Kobe1 and Traber, 1982). L-a-amino acids have been prepared by the resolution of racemic a-amino acid amide by the L-specific aminopeptidase from Pseudomonas putida ATCC-12633 (Kamphuis et a]., 1990a; Vriesema et al., 1986).Enzyme from Pseudomonas putida ATCC12633 cannot be used to resolve a-alkyl-substituted amino acid amides. Amino amidase from Mycobacterium neoaurum ATCC-25795 has been

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

)r'

R-C-COOH

I

NH

I

o=c

-

I

D-Acylsse

R

I

H,

D&-N-pcetyl Pmlao edd

COOH Hn I

HN &OOH

I o=c

4ch

HZN

I

CHI

119

CH3

L-N-pcetyl -no

d d

D-smino add 9

FIG.21. Synthesis of D-amino acids by o-acylase.

FIG.22. Synthesis of L-a-alkylamino acids by amidase.

used in the preparation of L-acid 58 and D-amide of a-alkyl-substituted amino acids 59 (Fig. 22) by an enzymatic resolution process using racemic a-alkylamino acid amide as a substrate (Kamphuis et al., 1990b, 1992).Amidase from Ochrobactrum anthropi catalyzed the resolution of a,a-disubsituted amino acids, N-hydroxyamino acids, and a-hydroxy acid amides. The resolution process could lead to the production of chiral amino acids or amides in 50% yield. Amino-acid racemases have been used to get 100% yield of chiral amino acids (Kamphuis et al., 1992). Aminotransferases have been used in the production of chiral amino acids (Primrose, 1984; Rozzell, 1987, 1989; Carlton et al., 1986). Aminotransferases catalyze the transfer of an amino group, a proton, and a pair of electrons from a primary amine substrate to the carbonyl group of an acceptor molecule such as oxaloacetate, a-ketoglutarate, or pyruvate (Fig. 23). o-aminotransferases from Pseudomonas sp. F-126 have also been used to produce homochiral amine products (Burnett et al., 1979). The stereoselective removal of either the pro-(S) or the pro-(R) proton has been demonstrated (Bouclier et al., 1979; Tanizawa et al., 1982). w-aminotransferase specific for the secondary amines has also been demonstrated from Bacillus megaterium and Pseudomonas aerugiRosa (Stirling, 1992).

120

R. N. PATEL

HOOCHsC

L-Aspartic acid

SKeto a d d Amlnotransferase COOH

0 Oxaloacatate FIG. 23.

L-Amino acid

Synthesis of L-amino acids by aminotransferase.

The bioconversion of nitriles and primary amides have been used in the production of such optically active a-hydroxy or a-amino acids as L-phenylalanine, L-lactic acid, and L-phenylglycine, which are a chiral synthon in many pharmaceutical syntheses. Nitrilases have been isolated from organisms belonging to genus Brevibacterium, Rhodococcus, and Pseudornonas (Asano et al., 1982; Bui et al., 1982; Kobayashi et al., 1988; Nagasawa et al., 1986, 1987). After the discovery of the antibiotic thienamycin, compounds that contain the carbapenem and penem ring systems have attracted much attention. The importance of the stereochemistry of the hydroxyethyl group is demonstrated by the fact that this group must be in the R configuration for antimicrobial activity. Previously, synthesis of carbapenem and penem compounds have often utilized the optically active p-lactam intermediates [Fujimoto et al., 1986; Alpegiani et al., 1985; Shibata et al., 1985). ~-(-)-3-hydroxybutyric acid prepared by the microbial hydroxylation of butyric acid have been used in p-lactam synthesis (Iimori and Shibazaki, 1985; Ohashi and Hasegawa, 1992). VII. Calcium Channel Blocking Drugs

a,

Dilthiazem a benzothiazepinone calcium channel blocking agent that inhibits influx of extracellular calcium through L-type voltage-operated calcium channels, has been widely used clinically in the treatment of hypertension and angina [Chaffman and Brogden, 1985). Since dilthiazem has a relatively short duration of action (Kawai et al., 1981), an 8-chloroderivative has been introduced in the clinic as a more potent

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

121

analogue of dilthiazem (Isshiki et al., 1988). Lack of extended duration of action and little information on the structure-activity relationship in this class of compounds led Floyd et al. (1990) and Das et al. (1992a,b) to prepare isosteric 1-benzazepin-2-ones and led to the identification of the 6-trifluoro-methyl-2-benzazepin-Zone derivative as a longer-lasting and more potent antihypertensive agent. A key chiral intermediate 52 ((3R-cis)-l,3,4,5-tetrahydro-3-hydroxy-4-(4-methoxyphenyl)-6-(trifluoromethyl)-2H-l-benzazepin-Z-one) was required for the total chemical synthesis of the new calcium channel blocking agent @ ((cis)-3-(acetoxy)~-[2-(dimethylamino)ethyl]-~,3,4,5-tetrahydro-4-(4-methoxyphenyl)-6-triflutomethy1)-2H-1-benzazepin-2-one). A stereoselective microbial process (Fig. 24) was developed for the reduction of 4,5-dihydro-4-(4methoxyphenyl)-6-(trifluoromethyl)-1H-l-benzazepin-2,3-dione 64 to chiral g (Pate1 et al., 1991b). Compound % exits predominantly in the achiral enol form, which is in rapid equilibrium with the 2-keto form enantiomers. Reduction of @ could give rise to formation of four possible alcohol stereoisomers. Remarkably, conditions were found under which only the single-alcohol isomer 52 was obtained by microbial reduction. Among various cultures evaluated, microorganisms from the genera Nocardia, Rhodococcus, Corynebacterium, and Arthobacter reduced compound @ to compound g with a 60-70% conversion yield at 1g/liter substrate concentration. The most effective culture, Nocardia salmonicolor SC-6310, catalyzed the bioconversion of to s2 in a 96% reaction yield with 99.9% optical purity at a Z-g/liter substrate concentration. Product g was isolated and identified by NMR and MS. A preparative-scale fermentation process for growth of N.salmonicolor and a bioreduction process using cell suspensions of the organism were demonstrated. VIII. Antipsychotic Agents

Much effort has been directed towards the understanding of the Sigma receptor system in the brain and endocrine tissue. This effort has been motivated by the hope that the Sigma site may be a target for a new class of antipsychotic drugs (Ferris et al., 1991; Junien and Leonard, 1989; Walker et al., 1990). Characterization of the Sigma system helped clarify the biochemical properties of the distinct haloperidol-sensitive Sigma binding site, the pharmacological effects of Sigma drugs in several assay systems, and the transmitter properties of a putative endogenous ligand for the Sigma site (Steinfels et a]., 1989; Massamiri and Duckles, 1990; Martinez and Bueno, 1991; Taylor et al., 1991). R-(+) compound @ (BMY-14802)is a Sigma ligand and has a high affinity for

sc 6310 &-

Dilth&uem_BL

Calcium Channel B l o c k P

I SO 32191 P

FIG.24. Synthesis of chiral synthon for calcium channel blocker. Stereoselective reduction of 4,5-dihydro methoxyphen~~l)-6-~trifluoromethyl~-lH-benzazepin-2,3-dione @ (Patel et al., 1991b). Reprinted with permission from Els Science Inc.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

Mlcrobydtlon

Ketone p

123

~

Antlpsychotlc agent R-(+)-BMY 14002 @

FIG.25. Preparation of R-(+)-BMY-14802, an antipsychotic agent. Stereoselective reduction of ~-(~-fluorophenyl)-~-[~-(5-fluoro-~-pyrimidinyl)-l-piperazinyl]-l-butanone 66 (Patel et al., 1993c). Reprinted with permission from Portland Press Ltd.

Sigma binding sites and antipsychotic efficacy. The stereoselective microbial reduction of keto compound l-(4-fluorophenyl)-4-[4-(5-fluoro-2pyrimidiny1)-1-piperazinyll-1-butanone @ to yield the corresponding hydroxy compound R-(+)-BMY-14802@ (Fig. 25) has been developed by Patel et al. (1993~). Among various microorganisms evaluated for the reduction of ketone 66, Mortierella ramanniana ATCC-38191 predominantly reduced compound &6 to R-(+)-BMY-14802and Pullularia pullulans ATCC-16623 reduced compound @ to S-(-)-BMY-14802. An optical purity of >98% was obtained in each reaction. In a two-stage process for reduction of compound @, cells of M . ramanniana ATCC-38191 were grown in a 380-liter fermenter containing 250 liter of medium and harvested after 31 hours of growth. Cells harvested from a 380-liter fermenter were used for the reduction of ketone 66 in a 15-liter fermenter using 10 liter of cell suspensions (20% wt/vol, wet cells). Ketone @ was used at a 2-g/liter concentration and glucose was supplemented at a 20-g/liter concentration during the biotransformation process to generate NADH required for reduction. After a 24-h biotransformation period, about a 90% yield (99.0% optical purity) of R-(+)-BMY-14802was obtained. The R-(+)-BMY-14802 was isolated from the 10-liter fermentation broth in overall 70 MY0 yield and 99% GC and HPLC purity. Isolated R-(+)-BMY-14802 gave an optical purity of 99.5”/0as analyzed by chiral HPLC. A single-stage fermentation-biotransformation process was demonstrated for reduction of ketone 66 to R-(+)-BMY-14802by cells of M . ramanniana ATCC-38191. Cells were grown in a 20-liter fermenter containing 15 liter of medium. After 40 h of growth in a fermenter, when the residual glucose was depleted (0.1”/0)and the pH of the medium

124

R. N. PATEL

dropped to 4.5, the biotransformation process was initiated by addition of 30 g of ketone 54 and 300 g of glucose. The biotransformation process was completed within a 24-h period, with a reaction yield of 100% and an optical purity of 98.9% for R-(+)-BMY-14802.At the end of the biotransformation process, cells were removed by filtration and product was recovered from the filtrate in overall 80% recovery. The isolated product had 99% HPLC purity and 98.8% optical purity. A reductase with a molecular weight of 29,000 Da has been purified to homogeneity which catalyzed the conversion of ketone 66 to R-(+)-BMY-14802, For the production of optically active alcohols, reduction of the inexpensive prochiral ketones is a promising method. Commercially available alcohol dehydrogenases derived from horse liver and Thermoanaerobium brockii (Jones, 1986; Yevich et al., 1986) have been used in the preparation of chiral alcohols. Alcohol dehydrogenase from T brockii is heat stable and has broad substrate specificity toward aliphatic ketones. Substrates with bulky side chains (such as acetophenone) are poor substrates. Alcohol dehydrogenase from yeast, horse liver, and T brockii transfer the pro-R hydride to the reface of the carbonyl to give (S) alcohols, a process described by Prelog's rule (Jones and Beck, 1986; Keinan et al., 1986). Alcohol dehydrogenases from Pseudomonas sp. strain PED and Lactobacillus kefir and Mortierella isabellina have been shown to catalyze the enantioselective reduction of aromatic, cyclic, and aliphatic ketones to the corresponding chiral alcohols (Bradshaw et al., 1992a,b). Both enzymes exhibit anti-Prelog specificity, transferring the pro-R hydride to form (R) alcohols. Most oxidoreductases used for the preparation of optically active alcohols involve the use of NADH as cofactor. Simon (1990) demonstrated the use of reductases from anaerobic Clostridium strains that catalyzed the reduction of a variety of compounds to optically active alcohols using methyl or benzyl viologen as an electron donor. 2-enoate reductase and 2-oxocarboxylate reductase have been used in the stereoselective reduction of carbon-carbon- and carbon-oxygen-containing compounds R-(+)-BMY-14802@ has also been prepared by lipase-catalyzed resolution of racemic BMY-14802 acetate ester fl (Hanson et al., 1994b). Lipase from Geotrichum candidum (GC-20 from Amano Enzyme Co.) catalyzed the hydrolysis of acetate @ t o R-(+)-BMY-14802(Fig. 26) in a biphasic solvent system in 48% reaction yield and 98% optical purity. The rate and enantioselectivity of the hydrolytic reaction was dependent upon the organic solvent used. The enantioselectivity ( E values) ranged from 1in the absence of solvent to >loo in dichloromethane and

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

1

GC-20 Toluene /H20

125

t

R-(+)-BMY-14802 45

FIG.26. Preparation of R-(+)-BMY-14802,an antipsychotic agent. Stereoselective hydrolysis of BMY-14802 acetate 67 (Patel et al., 1995). Reprinted with permission from Elsevier Science t t d .

toluene. S-(-)-BMY-l4802 was also prepared by chemical hydrolysis of undesired BMY-14802 acetate obtained during the enzymatic resolution process. IX. Antiarrhythmic Agents Larsen and Lish (1964) reported the biological activity of a series of phenethanolamine-bearing alkyl sulfonamido groups on the benzene ring. Within this series, some compounds possessed adrenergic and antiadrenergic actions. D-(+)-SOtalOlis a beta-blocker (Uloth et al., 1966) that, unlike other beta-blockers, has antiarrhythmic properties and has no other peripheral actions (Lish et al., 1965). The P-adrenergic-blocking drugs such as propranolol and sotalol have been separated chemically into the dextro- and levorotatory optical isomers, and it has been demonstrated that the activity of the lev0 isomer is 50 times that of the corresponding dextro isomer (Somani and Bachand, 1969). Chiral alcohol X fJ is a key intermediate for the chemical synthesis of D-(+)-sotalol m. The stereoselective microbial reduction of N-(4-(2-chloroacety1)phenyl)methanesulfonamide 70 to the corresponding (+)-alcohol@ (Fig. 2 7 ) has been demonstrated (Patel el al., 1993d). Among numbers of microorganisms screened for the transformation of compound 70 to compound @, Rhodococcus spp. ATCC-29675, ATCC-21243, Nocardia salmonicolor SC-6310, and Hansenula pofymorpha ATCC-26012 gave the desired chiral alcohol @ in >90% optical purity. H. polymorpha ATCC-26012 catalyzed the efficient conversion of compound 70 to

126

R. N. PATEL

___T

H Ketone LQ

H. Potymorpha ATCC 26012

CHaOY\N 0 1 H Chlral Alcohol

FIG.27. Synthesis of chiral synthon for ~-(+)-sotalol.Stereoselective reduction of N-(4-(2-chloroacetyllphenyl)methanesulfonamide 70 (Pate1et al., 1 ~ 3 d )Reprinted . with permission from Springer-Verlag.

compound @ in 95"/0 reaction yield and >99% optical purity. Growth of 13. polymorpha ATCC-26012 culture was carried out in a 380-liter fermenter, and cells harvested from the fermenter were used to conduct transformation in a 3-liter preparative batch. Cell suspensions (20% wet cells in 3 liter of 10 mM potassium phosphate buffer, pH 7.0) were supplemented with 1 2 g of compound 70 and 2 2 5 g of glucose, and the reduction reaction was carried out at 25"C, 200 rpm, pH 7. Complete conversion of compound 211 to chiral alcohol SS was obtained in a 20-h reaction period. Using preparative HPLC, 8.2 g of compound @ were isolated from the reaction mixture in overall 68% yield with >99% optical purity. Both enantiomers of solketal (2,2-dimethyl-l,3-dioxolane-4-methanol) and their corresponding aldehydes are attractive building blocks for the preparation of enantiomerically pure and biologically active compounds (Hirth and Kindler, 1982; Peters et a]., 1987), specifically S-P-blocking agents. Since solketal is relatively inexpensive and commercially available, enantioselective oxidation of its alcohol function has provided an economically feasible process for the production of R-(-)-solketal by preferentially oxidizing the S-enantiomer to its corresponding acid (Bertola et al., 1987). Quino-haemoprotein ethanol dehy-

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

127

drogenase from Comamonas testosteroni has been used in the preparation of S-(+)-solketal (Geerlof et al., 1994). X. Potassium Channel Openers

The study of potassium (K) channel biochemistry, physiology, and medicinal chemistry has flourished, and numerous papers and reviews have been published (Edwards and Weston, 1990; Robertson and Steinberg, 1989,1990).It has long been known that K channels play a major role in neuronal excitability, and it is now clear that K channels play a critical role in the basic electrical and mechanical function of a wide variety of tissues, including smooth muscle, cardiac muscle, and glands (Hamilton and Weston, 1989). A new class of highly specific pharmacological compounds has been developed that either open or block K channels (Robertson and Steinberg, 1990; Ashwood et al., 1986). K channel openers are powerful smooth muscle relaxants with in vivo hypotensive and bronchodilator activity (Hamilton and Weston, 1989). The synthesis and antihypertensive activity of a series of novel K-channel openers (Bergmann et al., 1990; Jacobsen et al., 1991; Evans et al., 1983; Atwal et al., 1991) based on monosubstituted trans-4-amino-3,4dihydro-2,2-dirnethyl-2H-l-benzopyran-3-o1 71 have been demonstrated. Chiral epoxide 22 and diol 73 are potential intermediates for the synthesis of K channel activators, important as an antihypertensive and bronchodilator agents. The stereoselective microbial oxygenation of 2,2-dimethyl-2H-l-benzopyran-6-carbonitrile 74 to the corresponding chiral epoxide 2 and chiral d i o l B (Fig. 281 has been demonstrated (Pate1 et al., 1994b). Among microbial cultures evaluated, the best culture, Mortierella ramanniana SC-13840, gave reaction yields of 67.5 MY0 and optical purities of 96% for (+)-trans diol 73. A single-stage process (fermentation-epoxidation) for the biotransformation of 74 was developed using Mortierella ramanniana SC-13840. In a 25-liter fermenter, the (+)-transd i o l B was obtained in the reaction yield of 60.7 MY0 and optical purity of 92.5%. In the two-stage process using a %liter cell suspension (10% wt/vol, wet cells) of Mortierella ramanniana SC-13840, the (+)-trans diol 73 was obtained in 76 MY0 yield with an optical purity of 96%. The reaction was carried out in a 5-liter Bioflo fermenter with a 2 g/liter substrate and a 10 g/liter glucose concentration. Glucose was supplied to regenerate NADH required for this reaction. From the reaction mixture, (+)-transd i o l 3 was isolated in 65 M% 14.6 g) overall yield. An optical purity of 97% and a chemical purity of 98% were obtained for the isolated (+)-trans diol 3.

CH:,

Potassium Channel Opener 7r

FIG.28. Oxygenation of 2,2-dimethyl-ZH-l-benzopyran-6-carbonitrile 74 to the corresponding chiral epoxide 22 and (+ d i o l B by Morfierella ramanniana SC-13840 (Patel et al., 1994b). Reprinted with permission from Elsevier Science Ltd.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

129

In an enzymatic resolution approach, chiral diol 73 was prepared by the stereoselective acetylation of racemic diol with lipases from Candida cylindraceae and Pseudomonas cepacia. Both enzymes catalyzed the acetylation of the undesired enantiomer of racemic diol to yield monoacetylated product and unreacted desired (+)-trans diol 73. A reaction yield of >40% and an optical purity >go%were obtained using each lipase (Pate1 et al., 1995).

XI. Antiinflammatory Drugs Naproxen, (S)-2-(6-methoxy-2-naphthyl)propanoic acid 75, is a nonsteroidal antiinflammatory and analgesic agent first developed by Syntex (Fried and Harrison, 1967; Harrison et al., 1970). Biologically active desired S-naproxen has been prepared by enantioselective hydrolysis of the methyl ester of naproxen by esterase derived from Bacillus subtilis Thai 1-8 (Giordano et al., 1992). The esterase was subsequently cloned in Escherichia coli with over an 800-fold increase in enzyme activity. The resolution of racemic naproxen amide and ketoprofen amides have been demonstrated by amidases from Rhodococcus erythropolis MP50 and Rhodococcus sp. C311 (Yamamoto et al., 1990; Gu et al., 1986b; Battistel et al., 1991; Layh et al., 1994). S-naproxen 75 and S-ketoprofen 76 (Fig. 29) were obtained in 40% yields (theoretical maximum 50%) and 97% e.e. The enantioselective esterification of naproxen has been demonstrated using lipase from Candida cylindraceae in isooctane as solvent and trimethylsilyl as alcohol. The undesired isomer of naproxen was esterified, leaving desired S-isomer unreacted (Tsai and Wei, 1994) Ibuprofen 22 is another well-known analgesic antiinflammatory drug, and it is believed that it will be marketed as a single-isomer drug. The kinetic enzymatic resolution of racemic ibuprofen has been reported (Trani et al., 1995). The reaction for resolution has been scaled up to make gram quantities of S-ibuprofen. This was accomplished by two enantioselective reactions each catalyzed by Novozyme 435. In the first reaction, 300 g of racemic ibuprofen were esterified with 1-dodecanol to yield the R-ester and S-ibuprofen to produce 89 g of S-ibuprofen in 85% enantiomeric excess. In the second reaction, 75 g of the 85% e.e. material were used to prepare 39 g of S-ibuprofen with a 97.5% e.e. Another approach for the enzymatic preparation of S-ibuprofen has been demonstrated by de Zoete et al. (1994). The enantioselective ammonolysis of ibuprofen-2-chloroethyl ester by Candida antarctica lipase (lipase SP435) gave the remaining ester S-(+)-enantiomerin 44% yield and 96% e.e.

130

R. N.PATEL

0

H

S-(-)-Ketoprofen26 FIG. 29. Structure of antiinflanlatury drugs naproxen 25, ketoprofcn

X ,and ibuprofen 22.

S-2-chloropropionic acid used as a chiral synthon €or pharmaceutical synthesis of various nonsteroidal antiinflammatory drugs has also been prepared by stereoselective dehalogenase reactions. Pseudomonas p u tida contained two dehalogenases, one a low-molecular-weight enzyme showing 100% specificity for S-Z-chloropropionate and the other enzyme with a higher molecular weight with specificity towards R-2-chloropropionate (Hardman and Slater, 1981; Hardman, 1991; Barth et al., 1992; Fetzner and Lingens, 1994). Future use of‘ R- and S-specific dehalogenases in enzymatic resolution processes will be very promising. XII. Antiviral Agents

Purine nucleoside analogues have been used as antiviral agents (Mansuri and Martin, 1987). Lamivudine, Zidovudine, and Didanosine are effective antiviral agents. Lamivudine 28, a highly promising drug candidate for HIV2 and HIV3 infection, provides a challenge to the synthetic chemist due to the presence of two acetal chiral centers, both sharing the same oxygen atom. The use of cytidine deaminase from Escherichia coli (Mahmoudian et a]., 1993) has been demonstrated to deaminate 2’-deoxy-3’-thiacytidine enantioselectively to prepare optically pure (2’-R-cis)-2’-deoxy-3’-thiacytidine (3TC, Lamivudine, Fig. 30).

131

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

-

Cytidine Deaminase

Racsmlc 24

Ha(>

(4-B 3TC 0

I

Didanosim (Vldex)

Zldovudins (Retrovir)

FIG.30. Preparation of (-)-lamivudine 28, an antiviral agent, by cytidine deaminase (Mahmoudian et al., 1993). Reprinted with permission from Elsevier Science Inc.

A novel enzymatic resolution process has been developed for the preparation of chiral intermediate for lamivudine synthesis. An enzymatic resolution of a-acetoxysulfides by Pseudomonas fluorescens lipase has been demonstrated to give a chiral intermediate in >45% yield and 97% optical purity (Milton et al., 1995). Toluene cis-dihydrodiols has been used as a synthetic building block for the preparation of a lactam that was converted into a new antiviral drug. cis-dihydrodiols are synthetic building blocks for pharmaceutical synthesis (Sheldrake, 1992). The enzymatic cis-dihydroxylation of aromatic compounds to give cis-dihydrodiols offers potential as raw materials for the preparation of chiral synthon needed for medicinal usage. Gibson et al. (1975) first demonstrated this reaction using benzene as a substrate. Subsequently, cis-dihydroxylation of numerous aromatic compounds was demonstrated using a bacterial system. In contrast to the bacterial system, a mammalian enzyme system produced trans-dihydrodiols as major products (Gibson and Subramanian, 1984). Imperial Chemical Industries has developed industrial processes to prepare kilogram quantities of arene-cis-dihydrodiols (Taylor, 1987).

132

R. N. PATEL

HO'

FIG. 31. Synthesis of chiral synthons for prostaglandins and leukotrienss. Stereoselective reduction of bicyclo[3.2.0lhept-2-en-6-one 81 and 7,7-dimethylbicyclo[3.2.O]l~ept-2-ene-6-one @ (Roberts, 1985).Reprinted with permission from Kluwer Academic Publishers.

XI1I. Prostaglandin Synthesis

Optically active epoxides are useful chiral synthons in the pharmaceutical synthesis of prostaglandins. Microbial epoxidation of olefinic compounds was first demonstrated by van der Linden (1963). Subsequently, May et al. (1975) demonstrated the epoxidation of alkenes in addition to hydroxylation of alkanes by an w-hydroxylase system. Oxidation of alk-1-enes in the range C6-Cl2, a,w-dienes from C6-C12, alkyl benzene, and ally1 ethers was demonstrated using an w-hydroxylase enzyme system from Pseudomonas oleovorans. R-epoxy compounds in greater than 75% e x . were produced by epoxidation reactions using the w-hydroxylase system (Abbott and Hou, 1973; May et al., 1975, de Smet et a]., 19831. The epoxidation system from Nocardia corallina is very versatile, has broad substrate specificity, and reacts with unfunctionalized aliphatics as well as aromatic olefins to produce R-epoxides (Takagi et al., 1990; Furuhashi, 1992).

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

133

Chiral bicyclo[3.2.0]heptanone has been recognized as a chiral precursor for (+)-prostaglandinA2 79 and (+)-prostaglandin-F2a 80 synthesis. The reduction of racemic bicyclo[3.2.0]hept-2-en-6-one 81 in high optical purity by cells of Mortierella ramanniana has been demonstrated (Fig. 31). The same organism has been used for the stereoselective reduction of 7,7-dimethylbicyclo[3.2.0]hept-2-ene-6-one 82 to prepare the chiral synthon for (+)-leukotriene-B4 synthesis 83 (Roberts, 1985). Cycloalkonone oxygenase from Pseudomonas putida AS1 and Acinetobacter sp. NCIMB9871 has been used to catalyze the regio- and stereoselective Baeyer-Villiger type oxidation of [3.2.01hept-Z-en-6-one. The enantiomerically pure lactones prepared by this enzymatic reaction are chiral synthons for prostaglandin syntheses (Shipston et al., 1992; Lenn and Knowles, 1994). Asymmetric hydrolysis of diethyl-3-hydroxy-3-methylglutyrate to its corresponding monoester in high optical purity by pig liver esterase has been demonstrated (Haung et al., 1975; Chen et ~ l . 1981; , Gais and Lukas, 1984). Chiral monoesters are synthons for chemoenzymatic preparation of prostaglandins. REFERENCES

Abbott, B. J., and Hou, C. T. (1973). A p p . Microbiol. 26,86-90. Aida, K., Chibata, I., Nakayama, K., Takinami, K., and Yamada, H., eds. (1986). Biotechnology of Amino Acid Production, Prog. Ind. Microbiol., Vol. 24. Elsevier, Amsterdam. Alpegiani, M., Bedeschi, A., Giudichi, F., Perrone, E., and Franceschi, G. (1985). J. Am. Chem. Soc. 107, 6398-6400. Asano, Y., Yasuda, T., Tani, Y., and Yamada, H. (1982). Agric. Biol. Chem. 46,1183-1189. Ashwood, V. A,, Buckingham, R. E., Cassidy, F., Evans, J. M., Hamilton, T. C., Nash D. J., Stempo, G., and Willcocks, K. J. (1986). J. Med. Chem. 29,2194-2201. Atwal, K., Grover, G. J., and Kim, K. S. (1991). U.S. Pat. 5,140,031. Barbas, C. F., Wang, Y.-F., and Wong, C.-H. (1990). J. A m . Chem. Soc. 112,2013-2016. Barth, P. T., Bolton, L., and Thonison, J. C. (1992). J. Bocteriol. 174,2612-2619. Battistel, E.,Bianchi, D., Cesti, P., and Pina, C. (1991). Biotechnol. Bioeng. 38,659-664. Bergmann, R.,Eiermann, V., and Gericke, R. J. (1990). J. Med. Chem. 33, 2759-2761. Berse, C., and Bessette, P. (1971). Can. J. Chem. 49,2610-2611. Bertola, M.A., Koper, H. S., Phillips G. T., Merx, A. F., and Claussen, V. P. (1987). Eur. Pat. Appl. No. 0244912Al. Bold, G., Duthaler, R. O., and Riediker, M. (1989). Angew. Chem. Int. Ed. Engl. 28, 497-498. Bouclier, M., Jung, M. J., and Lippert, B. (1979). Eur. J. Biochem. 98,363-367. Bradshaw, C. W., Fu, H., Shen, G. J., and Wong, C.-H. (1992a). J. Org. Chem. 57,15261531. Bradshaw, C. W., Hummel, W., and Wong, C.-H. (1992b). J. Org. Chem. 57, 1532-1536.

134

R. N. PATEL

Brieva, R., Crich, J. Z., and Sih, C. J. (1993). J. Org. Chem. 58, 1068-1075. h i , K., Arnaud, A., and Galzy, P. (1982). Enzyme Microb. Technol. 4, 95-197. Burnett, G., Walsh, C., Yonaha, K., Toyama, S., and Soda, K. (1979). J. Chem. SOC.Chem. Commun. 25, 826-835. Carlton, G. J., Wool, I,. I,.. Updike, M. H., Lanty, L., and Hamman. J. P. (1986). Bio/Technology 5, 317-325. Chaffman, M., and Brogden, R. N. (1985). Drugs 29, 387-390. Chen, L., Dumas, D. P., and Wong, C.-H. (1992). J. Am. Chem SOC.114, 41-43. Chen, C.-S., Fnjimoto, Y., and Sih, C. J. (1981). J. Am. Chem Soc. 103, 3580-3585. Chenault, H. K., Dahiner, J., and Whitesides, G. M. (1989). J. Am. Chem. SOC.11, 63546364. Chibata, I., Tosa, T., Sato, T., and Mori, T. (1988). In “Methods in Enzymology” (K. Mosbach, ed.), Vol. 44, pp. 746-759. Academic Press, New York. Christen, M., and Crout, D. H. G. (1988). J. Chem. Soc. Chem. Commun. 45, 264-266. Christen, A. A , , Gibson, D. M., and Bland, J. (1991). U.S. Pat. 5,019,504. Cole, D. C. (1994). Tetrahedron 32, 9517-9582. Crosby, J. (1991). Tetrahedron 47,4789-4846. Crout, D. H. G., Davies, S., Heath, R. J., Miles, C. O., Rathbone, D. R., and Swohoda, B. E. P. (1994). Biocatdysis 9, 1-30. Csuk, R.. and Glanxer, B. I. (1991). Chem. Rev. 96, 556-566. Cushman, D. W., and Ondetti, M. A. (1980). Biochem. Pharrn. 29,1871-1875. Cushman, D. W., Cheung, M. S., Saho, E. F., and Ondetti, M. A. (1977). Biochem. 16, 5484-5491. Das. J., Haslanger, M. F., Gougoutas, J. Z., and Malley, M. F. (1987). Synthesis 12, 1100-1112. Das, J., Floyd, D. M., Kimball, S. M., Patel, R. N., and Thottathil, J. K. (199Za). Ind. J . Chem. 31B, 817-820. Das, J., Floyd, D. M., Kimball, S. M., Duff, K. J., Vu, T., Lago, M. W., Moquin, R. V., Lee, V. G., Gougoutas, J. Z., Malley, M. F., Moreland, S., Brittain, R. J., Hedberg, S. A., and Cucinotta, G. (1992b). J. Med. Chem. 35, 773-778. Dnvies, H. G., Green, R. H., Kelly, D. R., and Roberts, S. M. (1990). Biotechnol. 10, 129-1 52. Delaney, N. G.. Gordon, E. N., DeForrest, 1. M., and Cushman, D. W. (1988). Eur. Pat. EP361365 Denis, J.-N., Greene, A. E., Aarao Serre, A., and Luche, M.-J. (1986). J. Org. Chem. 51, 46-50. de Smet, M.-J., Kingma, J., Wynberg, H., and Witholt, B. (1983). Enzyme Microb. Technol. 5, 352-356. de Zoete, M. C., Kock-van Dalen, J., van Rantwijk, F., and Sheldon, R. A. (1994). Biocatalysis 10, 307-316. Edwards, G., and Weston, A. H. (1990). TIPS 11, 417-422. Endo, A., Kuroda, M., and Tsujita, Y. (1976a). 1. Antibiot. 29, 1346-1348. Endo, A,, Kuroda, M., and Tanzawa, K. (1976h). FEBS Lett. 72, 323-326. Evans, D. A., Ellman, J. A., and Dorow, R. L. (1987). Tetrahedron Lett. 28, 1123-1126. Evans, J, M., Fake, C. S., Hamilton, T. C., Poyser, R. H., and Watts, E. A. (1983). J. Med. Chem. 26, 1582-1586. Feng, J.-M., Lin, C.-H., Bradshaw, C. W., and Wong, C.-H. (1995). J. Chem. Soc., Perkin Trans. 1 , 967-978. Ferris, C. D., Hirsch, D. J., Brooks, B. P., and Snyder, S. H. (1991). J. Neurochem. 5 7 , 729-737.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

135

Fetzner, S., and Lingens, F. (1994). Microbiol. Rev. 58, 641-685. Floyd, D. M., Moquin, R. V., Atwal K. S., Ahmed, S. Z . , Spergel, S.H., Gougoutas, J. Z., and Malley, M. F. (1990). I. Org. Chem. 55,572-5575. Ford-Hutchinson, A. W. (1991). CJin. Exper. Allergy 21, 272-276. Fried, J. H., and Harrison, I. T. (1967). Br. Pat. 1211134. Fujimoto, K., Iwano, Y., Hirai, K., and Sugawara, S. (1986). Chem. Param. Bull. 34, 999-1004. Furuhashi, K. (1992). In “Chirality in Industry” (A. N. Collins, G. N. Sheldrake, and J. Crosby, J., eds.), pp. 167-188. Wiley, New York. Gais, H.-J., and Lukas, K. L. (1984). Angew Chem. Int. Ed. EngJ. 23, 142-148. Geerlof, A., Stoorvogel, J., Jongejan, J. A., Leenen, E. J. T. M., van Dooren, T. J. G. M., van den Twell, W. J. J., Duine, J. A. (1994). Appl. Microbiol. Eiotechnol. 42, 8-15. Gibson, D. T., and Subramanian, V. (1984). In “Microbial Degradation of Aromatic Compounds” (D. T. Gibson, ed.). pp. 181-252. Dekker, New York. Gibson, D. T., Mahadevan, V., Jerina, D. M., Yagi, H., and Yeh, H. J. C. (1975). Science 189,295-298. Gijsen, H. J., and Wong, C.-H. (1994). 1.A m Chem. SOC. 116,8422-8423. Giordano, C., Villa, M., and Panossian, S. (1992). In “Chirality in Industry” (A. N. Collins, G. N. Sheldrake, and J. Croshy, eds.), pp. 303-312. Wiley, New York. Godfrey, J. D., Mueller, R. H., and Van Langen, D. J. (1986). Tetrahedron Lett. 27,27932796. Goodhue, C. T., and Schaeffer, J. R. (1971). Biotechnol. Bioeng. 13,203-214. Gopalan, A. S., and Sih, C. T. (1984). Tetrahedron Lett. 25, 5235-5238. Gordon, E. M., Ondetti, M. A., Pluscec, J., Cimarusti, C. M., Bonner, D. P., and Sykes, R. B. (1982). 1.Am. Chem. Soc. 104,6053-6060. Gou, D.-M., Liu, Y.-C., and Chen C . 3 . (1993). J. Org. Chem. 58,1287-1289. Gu, Q. M., Reddy D. R., and Sih, C. J. (1986a). Tetrahedron Lett. 27,5203-5206. Gu,Q.M., Chen, C. S., and Sih, C. J. (1986b). Tetrahedron Lett. 27, 1763-1766. Hamaka, N., Seko, T., Miyazaki, T., and Kawasaki., A. (1990). Advs. in Prostaglandin, Thromboxane and Leukotriene Research 21, 359-362. Hamilton, T. C., and Weston, A. H. (1989). Gen. Pharm. 20,1-9. Hanson, R.L., Singh, J., Kissick, T. P., Patel, R. N., Szarka, L. J., and Mueller, R. H. (1990). Bioorg. Chem. 18, 116-130. Hanson, R. L., Bembenek, K. S., Patel, R. N., and Szarka, L. J. (1992). Appl. Microbiol. Biotechnof. 37,599-603. Hanson, R. L., Wasylyk, J. M., Nanduri, V. B., Cazzulino, D. L., Patel, R. N., and Szarka, L. J. (1994a).J. Eiol. Chem. 269,22145-22149. Hanson, R. L., Banerjee, A., Comezoglu, F. T., Mirfakhrae, D., Patel, R. N., and Szarka, L. J. (1994b). Tetrahedron Asymm. 5,1925-1934. Hardman, D. J. (1991). Crit. Rev. Eiotechnol. 11,1-40. Hardman, D. J., and Slater, J. (1981). J. Gen. Microbiol. 123,117-128. Harrison, I. T., Lewis, B., Nelson, P., Rooks, W., Roszwski, A., Tomolonis, A., and Fried, J. H. (1970). 1.Med. Chem 13,203-208. Hasegawa, J., Ogura, M., Kanema, H., Noda, N., Kawaharada, H., and Watanabe, K. J. (1982). J. Ferment. Techno]. 60,501-508. Haung, F.-C., Lee, L. F. H., Mittal, R. S. D., Ravikumar, P. R., Chan, J. A,, Sih, C. I., Caspi, E., and Eck, C. R. (1975). J. Am. Chem. Soc. 97,4144-4145. Heidlas, J., Engel, K.-H., and Tressl, R. (1988). Eur. J. Biochem. 172,633-639. Hirama, M., and Uei, M. (1982). 1.A m . Chem. Soc. 104,4251-4256. Hirth, G., and Barner, R. (1982). Helv. Chim. Acta 65,1059-1084.

136

R. N. PATEL

Holton, R. A., Juo, R. R., Kim, H. B., Williams, A. D., Harusawa, S., Lowenthal, R. E., and Yogai, S. (1988). J. Am. Chem. Soc. 110,6558-6560. Hosobuchi, M., Kurosawa, K., and Yoshikawa, H. (1993a). Biotechnol. Bioeng. 42, 815820.

Hosobuchi, M., Shioiri, T., Ohyama, J., Arai, M., Iwado, S., and Yoshikawa, H. (1993b). Biosci. Biotech. Biochem. 57,1414-1419. Ice, J., and Agathos, S. N. (1989). Biotechnol. Lett. 11, 77-82. Iimori, T., and Shibazaki, M. (1985). Tetrahedron Lett. 26, 1523-1526. Imada, A,, Kitano, K., Kintaka, K., Muori, M., and Asai, M. (1981).Nature 289, 590-591. Isshiki, T., Pegram, B., and Frohlich, E. (1988). Cardiovasc. Drug Ther. 2,539-544. Ito, Y., Sawamura, M., Shirakawa, E., Hayashizaki, K., and Hayashi, T. (1988). Tetrahedron 44,5253-5262. Jacobsen, E. N., Zhang, W., Muci, A. R., Ecker, J, R., and Deng, L. (1991). J. Am.Chem. SOC. 113, 7063-7067.

Jones, I. B. (1986). Tetraliedron 42,3351-3403. Jones, J. B., and Beck, J. F. (1986). 117 “Application of Biochemical Systems in Organic Synthesis” IJ. B. James, C. J. Sih, and D. E. Perlman, eds.), pp. 248-376. Wiley, New York. Jones, J. B., and Francis, C. (1984). Can. J. G e m . 62,2578-2584. Junien, J. L., and Leonard, B. E. (1989). Clin. Neuropharmacol. 12,353-374. Kamphuis, J., Boesteri, W. H. J., Broxterman, Q. B., Hermes, H. F. M., van Balken, J. A. M., Meijer, E. M., and Schoemaker, H. E. (199Oa). Adv. Biochem. Eng. Biotech. 42, Js

133-1 8fi.

Kamphuis, J. Hermes, H. F. M., van Balken, J. A. M., Schoemaker, H. E., Boesten, W. H. J., and Meijer, M. E. (199Ob). In “Amino Acids, Chemistry, Biology and Medicine” [G. Lubec and G. A. Rosenthal, eds.), pp. 119-125. ESCOM Science Puhlishers, Vienna. Kamphuis, J., Meijer, E. M., Boesten, W. H. J., Broxterman, Q. B., Kaptein, B., Hermes, H. F. M., and Schoemaker, H. E. (1992). In “Production of Amino Acids and Derivatives” (D. Rozzell and F. Wagner, eds.), pp. 117-200. Hanser Publishers, Munich. Karenewsky, D. S., Badia, M. C., Cushman, D. W., DeForrest, J. M., Dejneka, T., Loots, M. J., Perri, M. G., Petrillo, E. W., and Powell, J. R. (1988). J. Med. Chern. 31, 204-212. Kawai, C., Konishi, T., Matsuyama, E., and Okazaki, H. (1981). Circulation 63, 1035-1038. Keinan, E., Hafeli, E. K., Seth, K. K., and Lamed, R. R. (1986). J. Am. Chem. Soc. 108, 162-168.

Kim, G.-J., and Kim, H . 3 . (1995). Enzyme Microb. Technol. 17,63-67. Kingston, D.G. I. (1991). Pharm. Ther. 52, 1-34. Kohayashi, M., Nagasawa, T., and Yamada, H. (1988). A p p l . Microbiol. Biotechnol. 29, 2 3 1-2 3 3 .

Kohel, H., and Traber R. (19821. Eur. 1.Appl. Microbiol. Biotechnol. 14,237-242. Larsen, A. A . , and Lish, I? M. (1964). Nature 203, 1283-1284. Laumen, K.,and Schneider, M. (1985). Tetrahedron Lett. 26,2073-2076. Layh, N.,Stolz, A., Bohme, J., Effenberger, F., and Knackmuss, H.-J. (1994). I. Biotechnol. 33, 175-182.

Lenn, M. 7.. and Knowles, C. J. (1994). Enzyme Microb. Technol. 16, 964-969. Lish, P. A,, Weikel, J. H., Dungan, K. W. (1965). J. Pharmacol. Exp. Ther. 149, 161-173. Lok, K.P., Jakovac, T. J., and Jones. J. B. (1985). Am. Chem. Soc. 107, 2521-2526. Mahmoudian, M., Baines, B. S., Drake, C. S., Hale, R. S., Jones, P., Piercey, J, E., Montgomery, D. S., Purvis, I. 7.. Storer, R., Dawson, M. J., and Lawrence, G. C. (1993). Enzyme Microb. Technol. 15, 749-754.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

137

Mansuri, M., and Martin, J. C. (1987). Ann. Report Med. Chem. Vol. 22, Chapter 15. Margolin, A. L. (1993). Enzyme Microb. Technol. 15, 266-280. Martinez, J. A., and Bueno, L. (1991). Eur. J. Pharmacol. 202, 379-383. Massamiri, T., and Duckles, S. P. (1990). J. Pharmacol. Exp. Ther. 253, 124-129. Matsuoka, T., Miyakoshi, S., Tanzawa, K., Nakahara, K., Hosobuchi, M., and Serizawa, N. (1989). Eur. J. Biochem. 184, 707-713. May, S. W., Schwartz, R. B., Abbott, B. J., and Zaborsky, 0. R. (1975). Biochim. Biophys. Acta. 403, 245-250. Milton, R. A., Brand, S., Jones, M. F., Rayner, C. M. (1995). Tetrahedron Lett. 36, 69616964. Mohr, P., Waespe-Sarcevic, N., and Tamm, C. (1983). Helv. Chim. Acta. 66, 2501-2511. Moller, A., Syldatk, C., Schulze, M., and Wagner, F. (1988). Enzyme Microb. Technol. 25, 91-96. Moniot, J. L. (1988). U.S. Pat. Appl. CN88-100862. Monot, F., Bemoit, Y., Lemal, J., Honorat, A., and Ballerini, D. (1987). In “Proc. Eur. Cong. Biotechnol., 4th” (0.M. Neijssel, R. R. ven der Meer, and K. C. A. M. Luyben, eds.), Vol. 2, 42-45. Elsevier, Amsterdam. Mori, K. (1995). Synlett., November, pp. 1097-1109. Mori, K., and Tanida, K. (1984). Tetrahedron Lett. 40, 3471-3476. Morin, A., Hummel, W., and Kula, M. (1986). Appl. Microbiol. Biotechnol. 25, 91-96. Nagasawa, T., Ryuno, K., and Yamada, H. (1986). Biochem. Biophys. Res. Commun. 139, 1305-1312. Nagasawa, T., Nanaba, H., Ryuno, K. Takeuchi, K., and Yamada, H. (1987). Eur. J. Biochem. 162, 691-698. Nakamura, K., Kawai, Y. Oka, S., and Shue, A. (1989). Tetrahedron Lett. 30, 2245-2246. Nakamura, K., Kawai, Y., Miyai, T., and Ohno, A. (1990). Tetrahedron Lett. 31, 3631-3632. Nakamura, K., Kawai, Y., Nakajima, N., and Ohno, A. (1991). 1.Org. Chem. 56,4778-4783. Nakane, M. (1987). U.S. Pat. 4,663,336. Nanduri, V. B., Hanson, R. L., LaPorte, T. L., KO,R. Y., Patel, R. N., and Szarka, L. J. (1995). Biotech. Bioeng. 48, 547-550. Nishida, Y., Nakamichi, K., Nabe, K., and Tosa, T. (1987). Enzyme Microb. Technol. 9, 72 1-72 5. O’Donnell, M. J. (1988). Tetrahedron 44, 5253-5614. O’Donnell, M. J., Bennett, D. W., and Wu, S. (1989). J. Am. Chem. SOC.111, 2353-2355. Ohashi, T., and Hasegawa, J. (1992). In “Chirality in Industry” (A. N. Collins, G. N. Sheldrake, and J. Crosby, eds.), pp. 269-278. Wiley, New York. Olivieri, R., Fascetti, E., Angelini, L., and Degen, L. (1979). Enzyme Microb. Technol. 1, 201-204. Ondetti, M. A,, and Cushman, D. W. (1981). J. Med. Chem. 24, 355-361. Ondetti, M. A,, Rubin, B., and Cushman, D. W. (1977). Science 196, 441-444. Ondetti, M. A., Miguel, A., and Krapcho, J. (1982). U.S. Pat. 4,316,906. Patel, R. N., Hou, C. T., Laskin, A. I., and Derelanko, P. (1981). J. Appl. Biochem. 3, 218-23 2. Patel, R. N., Robison, R. S., and Szarka L. J. (1990). Appl. Microbiol. Biotechnol. 34, 10-14. Patel, R. N., Howell, J. M., Banerjee, A., Fortney, K. E , and Szarka, L. J. (199la). Appl. Microbiol. Biotechnol. 36, 29-34. Patel, R. N., Robison, R. S., Szarka, L. J., Kloss, J., Thottathil, J. K., Mueller, R. H. (199lb). Enzyme Microb. Technol. 13, 906-912.

138

R. N. PATEL

Palel, R. N., Liu, M., Banerjee, A., Thottathil, J. K., Kloss, J., and Szarka, L. J. (1992a). Enzyme Microb. Technol. 14, 778-784. Patel, R. N., Liu, M., Banerjee, A., and Szarka, L. J. (1992b). Appl. Microbiol. Biotechnol. 37,180-183. Patel, R. N., Howell, J. M., McNarnee, C. G., Fortney, K. F., and Szarka, I,. J. (1 9 9 2 ~ ). Biotechnol. Appl. Biochem. 16, 34-47. Patel, R. N., McNarnee, C. G., Banerjee, A., Howell, J. M., Robison, R. S . , and Szarka, L. J. (1992d). Enzyme Microb. Technol. 14, 731-738. Patel, R. N., McNarnee, C. G., and Szarka, L. J. (1992e). Appl. Microbiol. Biotechnol. 38, 56-60. Patel, R. N., Banerjee, A., Howell, J. M., McNamee, G. M., Brzozowski, D., Mirfakhrae, D., Thottathil, J. K., and Szarka, L. J. (1993a). Tetrahedron A s y m m . 4, 2069-2084. Patel, R. N., Banerjee, A., McNamee, C. G., Brzozowski, D., Hanson, R. L., and Szarka, L. J. (1993b). Enzyme Microb. Technol. 15, 1014-1021. Patel, R. N., Banerjee, A., Liu, M., Hanson, R. L., KO, R. Y., Howell, J. M., and Szarka, L. J. ( 1 9 9 3 ~ )Biotechnol. . Appl. Biochem. 17, 139-153. Patel, R. N., Banerjee, A., McNamee, C. J., and Szarka, L. J. (1993d). Appl. Microbiol. Brotechnol. 40,241-245. Patel, R. N., Banerjee, A,, KO, R. Y., Howell, J. M., Li, W.-S., Comezoglu, F. T., Partyka, R. A., and Szarka, I.. J. (1994a). Biotechnol. Appl. Biochem. 20, 23-33. Patel, R. N., Banerjee, A., Davis, B., Howell, J. M., McNarnee, C. J., Brzozowski, D., North, J., Kronenthal, D., and Szarka, L. J. (1994b). Bioorg. Med. Chem. 6, 535-542. Patel, R. N., Banerjee, A., McNamee, C. G., and Szarka, L. J. (1995). Tetrahedron A s y m m . 6, 123-130. Peters, U., Bankova, W., and Welzel, P. (1987). Tetrahedron 43, 3803-3816. Primrose, S. B. (1984). Eur. Pat. Appl. 84100421.8 Roberts, S. M. (1985). In “Enzymes as Catalysts in Organic Synthesis” (M. P. Schneider, ed.], Vol. 178, 55-75. Kluwer, Dordrecht. Robertson, D. W., and Steinberg, M. I. (19891. Ann. Med. Chem. 24, 91-100. Kohertson, D. W., and Steinberg, M. I. (1990). J. Med. Chem. 33, 1529-1533. Rozzell, J. D. (1989). U.S. Pat. 4,876,766. Runser, S., Chinski, N., and Ohleyer, E. (1990). Appl. Microbiol.Biotechnol. 33, 382-388. Saito, S . , Bunya, N., Inaha, M., Moriwake, T., and Torii, S. (1985). Tetrahedron Lett. 26, 5 309-53 1 2 , Sakai, K., Oshima, K., and Moriguchi, M. (1991). A p p l . Environ. Microbiol. 57, 25402543. Sakimas, A., Yuri, K., Ryozo. N., and Hisao, 0. (1986). Eur. Pat. 0172614. Santaneillo, E., Ferraboschi, P., Grisenti, P., and Manzocchi, A. (1992). Chem. Rev. 92, 1071-1140. Schiff, I? B., Fant, J., and Horowitz, S. B. (1979). Nature 277, 665-667. Schimazaki, M., Hasegawa, J., Kan, K., Nemura, K., Nose, Y.,Kondo, H., Seymour, A. A., Swerdel, T. N., and Abboa-Offei, B. (1991). 1. Cardiovasc. Pharmacol. 17, 456-465. Schmidt, U., Respondek, M., Lieberknecht, A., Werner, J., and Fischer, P. (19891. Synthesis 28, 256-261. Schutte, H., Hummel, W., Tsai, H., and Kula, M. (1985). Appl. Microbiol. Biotechnol. 22, 306-31 7. Serizawa, N., Serizawa, S., Nakagawa, K., Furuya, K., Okazaki, T., and Tarahara, A. (1983). J. Antibiot. 36, 887-891.

SYNTHESIS OF SOME PHARMACEUTICAL INTERMEDIATES

139

Seymour, A. A., Swerdel, T. N., Fennell, S. A,, Druckman, S.P., Shadid, B., van der Plas, H. C., Boesten, W. H. J., Kamphuis, J., Meijer, E. M., and Schoemaker, H. E. (1991). Tetrahedron 46, 913-917. Shanzer, A., Somekh, L., and Butina, D. (1979). J. Org. Chem. 44, 3976-3969. Sheldrake, G. N. (1992). In “Chirality in Industry” (A. N. Collins, G. N. Sheldrake, and J. Crosby, eds.), pp. 127-166. Wiley, New York. Shibata, T., Iino, K., Tanaka, T., Hashimoto, T., Kameyama, Y., and Sugimura, Y. (1985). Tetrahedron Left. 26, 4739-4743. Shipston, N. F., Lenn, M. J., and Knowles, C. J. (1992). J. Microsc. Meth. 15, 41-52. Sih, C. T. (1987). Eur. Pat. 87264125. Derwent International Publication #W087105328. Sih, C. J., and Chen, C. S.(1984a). Angew. Chem. 96, 556-566. Sih, C. J.. and Chen, C.-S. (1984b). Angew. Chem. Inf. Engl. 23, 570-578. Sih, C. J., Zhan, B. N., Gropalan, A. S.,Chen, C. S., Girdaukais, G., and van Middlesworth, F. (1984). Ann. N.Y Acad. Sci. 434, 186-193. Sih, C. J., Gu, Q.-M., Holdgrun, X., and Harris, K. (1992). Chiralify4, 91-97. Simon, H. (1990). In “Biocatalysis” (D. A. Abramowicz, ed.), pp. 218-242. Van NostrandReinhold, New York. Sit, S. Y., Parker, R. A., Motoe, I., Balsubramanian, H. W., Cott, C. D., Brown, P. J., Harte, W. E., Thompson, M. D., and Wright, J. (1990). J. Med. Chem. 33, 2982-2999. Somani, P., and Bachand, T. (1969). Eur. J. Pharmacol. 7,239-247. Steinfels, G. F., Tam, S.W., and Cook, L. (1989). Neuropsychopharmcol. 2, 201-207. Stinson, S . C. (1992). Chem. Eng. News, 2 8 September, pp. 46-79. Stirling, D. I. (1992). In “Chirality in Industry” (A. N. Collins, G. N. Sheldrake, and J. Crosby, eds.), pp. 209-222. Wiley, New York. Sugie, M., and Suzuki, H. (1989). Agric. Biol. Cliem. 44, 1089-1095. Sykes, R. B., Cimarusti, C. M., Bonner, D. P., Bush, K., Floyd, D. M., Georgopadakou, N. H., Koster, W. H., Liu, W. C., Parker, W. L., Principle, P. A., Rathnim, M. L., Slusarchyk, W. A., Trejo, W. H., and Wells, J. S. (1981). Nature 291,489-491. Syldatk, C., Cotoras, D., Dombach, G., GroB, C., Kallwafi, H., and Wagner, F. (1987). Bioteclinol. Lett. 9, 25-30. Takagi, M., Uemura, N., and Furuhashi, K. (1990). Ann. N.Y Acad. Sci. 613, 697-702. Takahashi, S., Kii, Y., Kumagai, H., and Yamada, H. (1979). J. Ferment. Technol. 56, 492-498. Tanizawa, K., Yoshimara, T., Asada, Y., Sawoda, S., Misono, H., and Soda, K. (1982). Biochem. 21, 1104-1109. Taylor, D. P., Eison, M. S.,Moon, S . L., and Yocca, F. D. (1991). Adv. Neuropsychol. Psychopharm. 1, 307-315. Taylor, S . C . (1987). Eur. Pat. Appl. EP179603. Tombo, G. M. R., Schar, H. P., Busquets, X. F., and Ghisalba, 0. (1986). Tetrahedron Left. 21,5707-5710. Trani, M., Ducret, A., Pepin, P., and Lortie, R. (1995). Biotechnol. Left. 17, 1095-1098. Tsai, S.-W., and Wei, H.-J. (19941. Enzyme Microb. T~chnol.16, 328-333. Tsugawa, R., Okumura, S., Ito, T., and Katsuya, N. (1966). Agric. Bid. Chem. 30, 27-34. Uloth, R. H., Kirk, J. R., Gould, W. A., and Larsen, A. A. (1966). J. Med. Chem. 9, 88-96. van der Linden, A. C. (1963). Biochim.Biophys. Acta. 77, 157-162. Vriesema, B. K., ten Hoeve, W., Wijnberg, H., Kellogg, R. M., Boesten, W. H. J., Meijer, E. M., and Schoemaker, H. E. (1986). Tetrahedron Lett. 26, 2045-2048. Walker, J. M., Bower, W. D., Walker, F. O., Matsumoto, R. R., Costa, B. D., and Rice, K. C. (1990). Phorm. Rev. 42, 355-402.

140

R. N. PATEL

Wani, M. C., Taylor, H. L., Wall, M. E., Coggon, P., and McPhail, A. T. (1971). J. Am. Chem. SOC. 93, 2325-2327.

Ward, 0. P., and Young, C. S. (1990). Enzyme Microb. Techno]. 12, 482-493. Williams, R. M. (1989). In “Synthesis of Optically Active a-Amino Acids” (J. E. Baldwin and P. D. Magnus, eds.), Vol. 7, pp. 130-150. Pergamon, Oxford, New York. Wong, C.-H., and Whitesides, G. M. (1994). “Enzymes in Synthetic Organic Chemistry,” Tetrahedron Organic Chemistry Series, Vol. 12. Pergamon, Elsevier Science, New York. Yamada, H., and Shimizu, S.(1988). Angew Chem. h t . Ed. Engl. 27, 622-642. Yamada, H., Takahashi, S., Kii, Y., and Kumagai, H. (1978). J. Ferment. Techno]. 55, 484-491,

Yamamoto, K . , Uneo, Y., Otsubo, K., Kawakami, K., and Komatsu, K. (1990). Appl. Environ. Microhiol. 56, 3125-3129. Yevich, J. P., New, J. S.,and Smith, D. W. (1986). J. Med. Chem. 29, 359-369. Yokozeki, K., Nakamori, S., Yamanaka, S., Eguchi, C., Mitsugi, K., and Yoshinaga, F. (1987). Agric. B i d . Chem. 51, 715-719.

Microbial Xylanolytic Enzyme System: Properties and Applications PRATIMA BAJPAI Chemical Engineering Division Thapar Corporate Research and Development Centre Patiala 147 001, India

I. Introduction 11. Structure of Xylan and Its Interaction with Plant Cell Walls 111. Properties of Xylanolytic Enzymes A. Xylanases B. 0-Xylosidases C. a-Arabinosidases D. a-Glucuronidases E. Esterases IV. Production of Xylanolytic Enzymes V. Application of Xylanases A. In Pulp and Paper Making 1. Prebleaching of Kraft Pulps 2. Enzymatic Debarking 3. Fiber Modification 4. Production of Dissolving Pulp 5. Removal of Shives 6. Retting of Flax Fibers B. Other Applications VI. Conclusions References

I. Introduction

Xylan is a major component of plant hemicellulose. After cellulose, it is the most abundant renewable polysaccharide in nature. Xylan and cellulose are the predominant hemicellulosic polysaccharides found in the cell walls of land plants, in which they may constitute more than 30% of the dry weight. Aside from terrestrial plants, in which xylans are based on a P-1,4-linked D-xylosyl backbone, marine algae also synthesize xylans of different chemical structure based on a P-1,3-linked D-xylosyl backbone. In some species of Chlorophyceae and the Rhodophyceae where cellulose is absent, xylans form a highly crystalline fibrillar material. 141 ADVANCES IN APPLIED MICROBIOLOGY. VOLUME 4 3 Copyright 0 1997 by Academic Preas, Inc. All rights of reproduction i n any form reserved. 0065-2164197 $25.00

142

P. BAJPAI

Xylanolytic enzymes of microorganisms have received a great deal of attention in the last 10 years. These enzyme systems are of interest for several reasons. On the one hand, they are clearly involved in providing sources of carbon and energy for the organisms that produce them. They act in the same fashion for hosts harboring xylanase-producing organisms, and they are involved in the growth, maturation, and ripening of cereals and fruits. Moreover, xylanases appear to be involved in the invasion of plants and fruits by pathogens. On the other hand, xylandegrading enzyme systems have considerable potential in several biotechnological applications. The present article reviews the properties and application of microbial xylanolytic enzyme systems. Attention is focused on recent advances, and particularly on several aspects that were not covered in earlier reviews.

I I . Structure of Xylan and Its Interaction with Plant Cell Walls Xylan structure is variable, ranging from linear 1,4-P-linked polyxylose chains to highly branched heteropolysaccharides. The prefix “hetero” denotes the presence of sugars other than D-xylose. Some major structural features are summarized in Fig. 1. The main chain of xylan is analogous to that of cellulose but composed of D-xylose instead of D-glucose. Branches consist of L-arabinofuranose linked to the 0-3 positions of D-xylose residues and of D-glucuronic acid or 4-0-methyl-D-ghcuronic acid linked to the 0-2 position. Both side-chain sugars are linked a-glycosidically. The degree of branching varies depending on the source. Xylans of several wood species, particularly of hardwoods, are acetylated: for example, birch xylan contain >1mol of acetic acid per 2 mol of D-xylose. Acetylation occurs more frequently at the 0-3 than the 0-2 position, and double acetylation of a D-XyloSe unit has also been reported. There is a relationship between the chemical structure of xylans and their botanical origin and also in their cytological localization. This results in a certain degree of complexity of xylan-containing materials that may possess several different xylan polymers of related structures but differ by more or less important features. Xylans of terrestrial plants are present in various proportions in the cell wall of all lignified tissues, but many may also be found in plant species as diverse as mosses and ferns (Aspinall, 1959; Joseleau, 1980). They usually are constituents of the secondary walls of tissues having structural functions, but they are also present to some extent in the primary walls of growing cells (Joseleau and Barnoud, 1974; McNeil et a]., 1979), as well as in the

143

MICROBIAL XYLANOLYTIC ENZYME SYSTEM

H

H

H OH

H

H

H

H

H OH

COOH

A

OAC

0

n a y Hc@H

"C H3O @H

Araf

a

-

***

lo

Ac I* 3 3 4xylgl-4xyl~l-1xyl~1-~xylpI- Lxylbl-4xylfll-4xyl~1-4xyl~l- lxylgl-4Xylpl-

2

7 0 a 1

IC

Ac

Me G l c A

f10 U

Me GicA

endo-1,bp-xylanase (EC 3.2.1.8) P-xylosidase (EC 3.2.1.37) acetyl esterase (EC 3.1.1.6) or acetyl xylan esterase

0 a-glucuronidase (EC 3.2.1) c) a-L-arabinofuranosidase (EC 3.2.1.55) FIG.1. Structure of hardwood xylan and the site of attack by xylanolytic enzymes.

primary walls of seeds and bulbs (Shaw and Stephen, 1966a,b) of certain plant species in which they have reserve functions. In all the various ultrastructural localizations that they have in plant walls, xylans interact with other structural components, in particular with cellulose microfibrils, with other noncellulosic polymers, and in most cases with lignin. Noncovalent interactions of xylans with other polysaccharides essentially involve hydrogen bonding whereas covalent bonds interconnect xylans, lignins, and some phenolic acids. Although relatively little is known about the conformation of xylan chains within the cell wall lattice, conformational analysis of xylans has been carried out by crystallographic studies after purification of the isolated hemicelluloses. The development of techniques that permit the observation and analysis of polymers in their natural location in the cell wall-like

144

P. BAJPAI

electron microscopy, solid-state NMR, and Fourier transform infrared spectroscopy-are now being employed. Like most polysaccharides of plant origin, xylans display a large polydiversity and polymolecularity (Joseleau et al., 1992). This corresponds to their being present in a variety of plant species and to their distribution in several types of tissues and cells. All land plant xylans are characterized by a p-1,4-linked D-xylopyranosyl main chain that carries a variable number of neutral or uronic monosaccharide substituents or short oligosaccharide side chains. Very few unbranched linear xylan homopolysaccharides from land plants have been isolated. The best known is the xylan from esparto grass (Chanda et al., 1950), which, because of the peculiarity of its structure, served as a model for chemical and physical studies (Marchessault et a]., 1961). Other unsuhstituted linear xylans have been isolated from tobacco stalks (Eda et al., 1976) and guar seed husks (Montgomery et al., 1956). From the simple p-1,4-D-xylopyranosyl chain, the structural complexity of xylan rises with the number of substituting mono- and oligosaccharides attached to the P-1,4-linked xylosyl main chain. Timell reviewed the chemistry of hemicelluloses from angiosperms (Timell, 1964) and gymnosperms (Timell, 1965), and Wilkie described the structural characteristics of the different xylans from monocotyledonous species (Wilkie, 1979). The general features of all of these xylans from hardwoods, softwood, and grasses are given in Fig. 2. Much has been known for a rather long time about the chemical structure of alkali-soluble xylans. Most of the chemical structures reported in the literature were acquired by chemical, enzymatic, or spectroscopic analytical methods used separately or in combination. This led to description of the most characteristic features of polysaccharides belonging to the xylan family and corresponded to averaged structures with no or only little information about minor structural elements that were considered insignificant or due to impurities. The structural features of complex heteroxylans were obtained with more powerful investigative tools principally involving the use of the numerous glycohydrolases, esterases, and glycanases, of HPLC, and of 2D NMR spectroscopy. However, a lot less is known about their true primary structure, that is, if the arabinosyl, uronic acid, or acetyl substituents are attached to the xylosyl backbone randomly or as regular repeating sequences. The main structural elements commonly found in land plant cell wall xylans are shown in Table I. On the basis of the nature of the substituents, four main families can be considered where the complexity increases from linear to highly

( O IA c )

-

~

o

~

-

o

~

o

~

-

HOo

w0 o

~

L

~

Bo Ei

E F

OR

OH

X

5z

R: a-~-GIcpA(1+2)Xyl ... 4-OMe-a-~-GlcpA(1+2)Xyl ... a-L-Araf(l+3)Xyl ... a-L-Araf(l+2)Xyl ... P-D-Galp(1+S)a-L-Araf(l+3)Xyl.. p-D-xylp(l-+2)a-~-Araf(l+3)XyI.. a-L-Araf(l-3-2, 1+3 and 1+2,3 Araf),(l-+3)Xyl Feruloyl

0

s2

n m

5

v)

...

p. coumaroyl Lignin FIG.2. Principal side-chain substitution on xylan backbone.

3 4 z

~

146

P. BAJPAI TABLE I PRINCIPAL STRUCTU TYPES ~ L FOUND IN THE XYLANFAMILY

Structural tY Pe

Nature of side chains

Source

Reference'

Linear homoxylan

= none

Esparto grass Tobacco stalk

Chanda et al., 1950 Eda et a]., 1976

Arahinoxy Ian Low branching

Terminal a-L-Araf

Common harhery monocots Primary walls Flours Gramineae pericarp

Henderson & Hay, 1972

High branching Complex side chain

a-L-arabinan oligomers

McNeil et al., 1979 Ewold & Perlin, 1959 Brillouet & Joseleau, 1987

Glucuronoxylan

CY-D-G~C~A a-4-0-Me-o-GlcpA

Soybean hull Hardwood Gramineae Legumes

Aspinall et a]., 1966 Timell, 1965 Wilkie, 1979 Reicher eta]., 1989

Glucuronoarabinoxylan

Terminal a - ~- Glc pA a-4-Me-o-GlcpA a-L-Araf &D-GalO

Softwoods Gramineae Dicot primary wall

Xmell, 1965 Wilkie, 1979 McNeil e t a ] . . 1979

T h e references cited here illustrate examples of typical xylans.

substituted xylans. A broad distinction may thus be made between the arabinoxylans having only side chains of single terminal units of a-Larabinofuranosyl substituents, the true glucuronoxylans in which a-Dglucuronic acid and/or its 4-0-methyl ether derivative represent the only substituent, and the more complex glucuronoarabinoxylan in which a-L-arabinose and a-D-glucuronic acid and 4-0-methyl-a-D-glucuronic acid are present at the same time. In addition to these three main families, one may distinguish arabinoxylans having a high degree of substitution by more or less short side chains of 2,3,5- and 2,3-linked arabinofuranosyl oligosaccharides attached to the 0-3 position of the xylosyl main chain, and galactoglucurono-arabinoxylans,characterized by the presence of terminal P-D-galactopyranosyl residues on complex oligosaccharide side chains of xylans from several perennial plants (Buchala and Meirer, 1972; Wilkie, 1979). Table I1 gives some of the most commonly encountered side chains that substitute heteroxylans of vari-

147

MICROBIAL XYLANOLYTIC ENZYME SYSTEM TABLE I1 VARIABILITY IN SIDE-CHAIN SUBSTITUTION I N THE HETEROXYLAN FAMILY ~

__

Side chains

~

~

Usual position of attachment to main chain

Source

References

Terminal single unit U-D-GICPA a-4-0-Me-D-GlcpA a-L-Araf

Angiosperms Gymnosperms

Timell, 1964 Timell, 1965

Corn-cob Barnhoo leaves

Dekker & Richard, 1975 Wilkie & Woo, 1977 Buchala & Meirer, 1972

-3

Wheat kernel

Brillouet

+2.+3,+2,3

Angiosperms

Karacsonyi et a]., 1983 Saavedra et ol.,1988 Reicher eta]., 1989 Kato & Nevins. 1985

-13

Complex oligosaccharidcs P-D-xylp(lj2)a-L-Araf P-D-Galp(l+5]a-L-Araf P-D-Galp(l-141-D-Xylp (+Z)-a-L-Araf 4-Me-a-D-GlcpA(1+4)-DXylp(1 +4)-D-Galp

-13 +3 +3

Arahinan side chains +Z)-L-Araf -+3)-L-Araf +5)-L-Araf +2,3)-~-Araf

1

&

Joseleau, 1987

Nonsaccharide side chains Acetyl

L-Araf Feruloyl

+BL-Araf

Monocots

ous organics. Substitution by nonsaccharide groups or molecules that are covalently attached to the xylan backbone or side chains are also listed in Table 11. Aside from the typical structures reported in Tables I and 11, several polymers having unusual characteristics have been reported. For example, an acidic xylan substituted only with a-D-glucuronosyl residues with the exclusion of the 4-0-methyl ether derivative was purified from the husk of a legume (Swamy and Salimaih, 1990). A neutral arabinoxylan containing an average structural unit corresponding to a xylosyl residue doubly substituted with a single arabinofuranosyl residue at the 0-3 position and two (1-3)linked L-arabinofuranosy1 residues at 0-2 was isolated from the bark of Litsea glutinosa

148

P. BAJPAI

(Herath et al., 1990). This polysaccharide had the very unusual feature of having both a - ~and - P-1,-arabinofuranosylresidues, as demonstrated by lH and 13CNMR and by a homonuclear lHlH 2D-COSY experiment. Similarly, a 2D heteronuclear correlated spectrum was the determinant for assigning signals of C-1 and H-1 of arabinosyl substituents attached to the same xylose residue of the backbone rather than to different xylose residues of the main chain (Ebrigerova et al., 1990). All of the structural data describing the ratio of substituents and their various positions on the main chain are insufficient for a complete description of the primary structure of xylan, since only little information is provided about the sequence of the distribution of the side chains along with the P-1,4-linked xylan backbone. Yet this information is needed for an understanding of the conformation of xylan in solution or in the cell walls and to understand their interaction with other cell wall polymers. Both the relative distribution of the side chains and the physical conformation of the polysaccharide are of great importance for the action of enzymes and may influence their mode of attack and hydrolysis yield. Only a few techniques and approaches provide information about the distribution of the side chains and substituents. In this area, specific degradation involving chemical reagents can provide useful results (Aspinall, 1982), but certainly the best tools are enzymes. There is a great diversity of available xylan-degrading enzymes, glycohydrolases, and glycanhydrolases that when used in conjunction with modern techniques for analysis of oligosaccharides (like HPLC, NMR and FAB mass spectroscopy) provide the best results for a detailed description of the true primary structures of xylans (Comtat and Joseleau, 1981; Debeirre e f al., 1990; Kovac et al., 1982). After exhaustive hydrolysis of the highly acidic arabinoxylan from Sequoia sempervirens (Dutton and Joseleau, 1977) by a purified fraction of endoxylanase, the analysis of the released oligosaccharides and of the undegraded polymer residue suggested that the uronic acid substituents were irregularly distributed on the xylosyl backbone (Comtat and Joseleau, 1981). The results also constituted evidence for the existence on this xylan chain of open hydrolyzable regions of unsubstituted or less substituted portions where endoxylanases could readily attack, and of nonhydrolyzable blocks, immune to endoxylanase attack because of the higher density of substituents. Similar results had been found for the distribution of uronic acid in the xylan from another softwood (Shimizu et al., 1978), whereas specific alkaline-catalyzed p-elimination demonstrated that glucuronic acid groups were randomly dispersed on the backbone of hardwood xylan (Rose11 and Svensson, 1975).

MICROBIAL XYLANOLYTIC ENZYME SYSTEM

149

Ill. Properties of Xylanolytic Enzyme Systems

Due to the complex structure of xylans, several different enzymes are needed for their enzymatic degradation and modification. The two main glycanases that depolymerize the hemicellulose backbone are endo-l,4P-xylanase (EC 3.2.1.8) and endo-l&P-mannanase (EC 3.2.1.78). Small oligosaccharides are further hydrolyzed by 1,4-p-D-xylosidase, 1,4-p-Dmannosidase, and 1,4-P-D-glucosidase. The side groups are split off by a-L-arabinosidase, a-D-glucuronidase and a-D-galactosidase. Esterified side groups are liberated by acetyl xylan esterase and acetyl galactoglucomannan esterases. A. XYLANASES

Nonspecific xylanases from Trichoderma spp. can attack cellulose (Beldman et al., 1985), carboxymethylcellulose, p-nitrophenyl-p-glucoside (Beldman et al., 1985; Shikata and Nisizawa, 1975), cello-oligomers (Biely et al., 1991; Claeyssens et al., 1990; Shikata and Nisizawa, 1975), cellobiose (Beldman et al., 1985), laminarin (Lappalainen, 1986), and p-nitrophenyl-0-cellobioside (Shikata and Nisizawa, 1975). Carboxymethyl-cellulose, p-nitrophenyl-glucoside, and xylan apparently compete for the same active site on an enzyme from T viride that does not attack insoluble cellulose (Shikata and Nisizawa, 1975). A number of nonspecific glycanases have been characterized in Trichoderma spp. With one exception, they have relatively larger MWs (32-76 kDa) and more acidic PI values (3.5-5.3). This PI range excludes all of the apparently specific xylanases except for the 21-kDa xylanase from 'I: lignorum. These observations corroborate the hypothesis that there is a cluster of nonspecific glycanases with acidic PIS (Hrmova et al., 1986). The observation that these glycanases were induced by sophorose but not xylobiose suggests that they could be considered cellulases rather than xylanases. A multiplicity of xylanases has been documented in numerous organisms (Wong et al., 1988), with evidence for the occurrence of three to five xylanases in bacteria and fungi (Berenger et al., 1985; Fournier et al., 1985; Fredrick et al., 1981, 1985;John et al., 1979; Marui et al., 1985; Mitsuishi et al., 1987; Okhoshi et al., 1985; Shei et a]., 1985; Sreenath and Joseph, 1982; Takenishi and Tsujisaka, 1973, 1975; Tsujibo et al., 1990; Yoshioka et al., 1981). Recent analyses at the molecular genetic level have verified the occurrence of multiple xylanases in bacterial species (Flint et al., 1989; Gilbert et al., 1988; Mondou et al., 1986;

150

P. BAJPAI

Sakka et al., 1990; Vats-Mehta et al., 1990; Yang et al., 1989). Five xylanases have been detected in Trichoderma spp. (Dekker, 1983; Wong et al., 1986a) and three have been purified and characterized in T harzianum E58. The functional and genetic basis of these multiple enzymes has not been completely elucidated. Electrophoretically distinct xylanases may arise from posttranslational modifications of the same gene product, such as differential glycosylation or proteolysis. Trichoderma xylanases have been reported to be glycosylated in some cases (Lappalainen, 1986; Toda et al., 1971) but not in others (John and Schmidt, 1988; Ujiie et al., 1991; Wood and McCrae, 1986). The latter group of enzymes includes one pair of xylanases isolated from IT: koningii and another from 'I:lignorum. It therefore appears that differential glycosylation cannot explain the occurrence of multiple xylanases in these cases. Furthermore, a comparison of aminoacid compositions suggests that the three xylanases purified from IT: harzianum are distinct gene products (Wong et al., 1986a). The amino-acid composition of seven different Trichoderma xylanases has been reported. These data suggest high similarity between xylanase A from T harzianum and low-MW xylanases from ?: koningii and 7: viride. The other two xylanases from IT: harzianum appear to be distinct because of their relatively high alanine content. Furthermore, xylanase B is distinguished by a lack of tryptophan and a high cysteine content. These observations would suggest that there is a class of similar low-MW xylanases that occurs among Trichoderma spp. Amino-acid sequence comparisons have already suggested that xylanase A from T harzianum and a low-MW xylanase from T viride have over 90% homology (Roy et al., 1991; Yaguchi et al., 1992). Their similarity to corresponding enzymes from Bacillus pumilus, B. subtilis, and Schizophyllum commune is over 40% (Roy et al., 1991; Yaguchi et al., 1992). Preliminary X-ray diffraction analyses have been reported for xylanase A (Rose et al., 1987) and a 22.5-kDa xylanase from B. pumilus (Katsube et al., 1990; Moriyama et al., 1987). Completion of these studies would provide a means of directly comparing the structure of related xylanases from different sources. Amino-acid sequence comparisons of the catalytic domains of numerous glycanases have suggested that two classes of xylanase occur in microorganisms: the low-MW xylanases containing 182-234 aminoacid residues and the high-MW xylanases containing 269-809 residues (Gilkes et al., 1991). The low-MW xylanase class includes enzymes from B. pumilus, B. subtilis, and Clostridium acetobutylicum. Species of these two bacterial genera have most clearly illustrated the consistent occurrence of two forms of xylanases in one organism (Wong et al.,

MICROBIAL XYLANOLYTIC ENZYME SYSTEM

151

1988). It is not known whether these two genetic units correspond to genetic units in Trichoderma spp. Historically, the functional division of xylanases has been related to the ability of certain xylanases to release arabinosyl substituents from arabinoxylan (Dekker, 1985). The work of Takenishi and Tsujisaka (1975) had shown a functional importance for the occurrence of an arabinose-releasing xylanase in Aspergillus niger. This enzyme was apparently responsible for the removal of arabinosyl substituents while another was responsible for the hydrolysis of xylotriose; the two together cooperated to increase the hydrolysis yield from arabinoxylan. However, such cooperation was not observed between two xylanases isolated from T. koningii: IM173022 (Wood and McCrae, 1986) or among xylanases from ?: harzianum E58 (Wong et al., 1986b),probably because their arabinose-releasing and xylotriose-cleaving activities were not complimentary. Nevertheless, the three cellulase-free xylanases from T. harzianum cooperated in the hydrolysis of aspen xylans. The degree of cooperativity apparently increased with increasing complexity of the substrate, from the form of the deacetylated polymer to that of the acetylated polymer and, most significantly, to the form of aspen holocellulose. Furthermore, the three xylanases isolated from Talaromyces byssochlarnydoides YH-50 also cooperated to maximize the hydrolysis of hardwood xylan (Yoshioka et al., 1981). The functional contributions of the individual xylanases have not yet been elucidated. Other characteristics proposed for classifying xylanases include the distinction between xylan and xylodextrin-hydrolyzing enzymes (Bailey and Gaillard, 1965), between enzymes with endo- and exo-type hydrolytic mechanisms (Reilly, 1981), and between enzymes with and without preference for the substitution sites on the xylan backbone (Fredrick et al., 1985; Nishitani and Nevins, 1991). The characterization of a 43-kDa xylanase from B. subtilis suggested that its hydrolysis sites are specifically oriented to the glucuronosyl rather than the arabinosyl substituents (Nishitani and Nevins, 1991). The optimal conditions for activity of Trichoderma xylanases range from 45 to 65OC and from pH 3.5 to 6.5. As might be expected, the xylanases with higher temperature optima are relatively more thermally stable (hie et al., 1990; Tan et al., 1985a; Wood and McCrae, 1986) than those with lower temperature optima (Hashimoto et a]., 1971; John and Schmidt, 1988; Tan et al., 1985a; Wood and McCrae, 1986). Two xylanases have been reported to be stable at 50°C for 1 h (hie et al., 1990; Tan et al., 1985a) and one at 60°C for 20 min (Wood and McCrae, 1986). These properties are relatively moderate when compared to xylanases isolated from thermophilic microorganisms. For example, a xylanase

152

P. BAJPAI

isolated from Thermostoga sp. strain FjSS3-B.1 has a temperature optimum of 105°C at pH 5.5 and a half-life of 90 min at 95°C (Simpson et al., 1991). Furthermore, alkaline-tolerant xylanases have been isolated from Bacillus spp. that have a broad range of pH optima and stabilities, ranging up to pH 10 (Honda et al., 1985b; Horikoshi and Atsukawa, 1973; Okazaki et al., 1985). Mercury ions have been found to be inhibitory to the activity of Trichoderma xylanase at concentrations ranging from 0.1 to 10 mM (Hashimoto et al., 1971; Huang et al., 1991: John and Schmidt, 1988). One exception is a partially purified xylanase from 7: viride that was not inhibited by 1 mM Hgz+(Gibson and McCleary, 1987). Of the other ions tested (Gibson and McCleary, 1987; Hashimoto et al., 1971; Huang et al., 1991; John and Schmidt, 1988; Tan et al., 1985b), 1mM Ca2+was found to be inhibitory in one case (Huang et al., 1991) and 1 mM Cu2+ in another (John and Schmidt, 1988). Trichoderma xylanases have been found to be active on xylans from different sources, usually producing xylooligomers (xylobiose and xylose). Xylose is not usually the major product, and it is typically produced after an accumulation of xylooligomers. Of the xylanases characterized, one isolated from T pseudokoningii (Baker et al., 1977) and two isolated from T. viride (Dean et al., 1991) were reported to be unable to produce xylose. Two nonspecific glycanases from T viride were also found to be unable to produce xylose during xylan hydrolysis (Beldman et a]., 1988; Shikata and Nisizawa, 1975). One of these glycanases produced xylobiose as an initial product, indicating that it acts like an exoxylanase (Shikata and Nisizawa, 1975). The hydrolysis patterns of Trichoderma xylanases, however, have suggested that most are endoxylanases. Xylans are not completely hydrolyzed by crude culture filtrates (Poutanen et al., 1987, 1990a) or purified xylanases from Trichoderma spp. (Poutanen and Puls, 1989; Wong et al., 1986b; Wood and McCrae, 1986). However, hydrolysis yields from certain xylans could be improved by using mixtures of three different xylanases purified from T harzianum (Wong et al., 1986b). Xylose yields obtained using a purified xylanase from T reesei were increased when a purified 0-xylosidase was added (Poutanen and Puls, 1989). They were further increased where the relevant debranching enzymes were added to the hydrolysis reaction. When acetylated xylooligomers were partially deactivated by freezedrying over ammonia, they became more accessible to hydrolysis by xylanase. All these observations suggest that the substituents on xylans can restrict their hydrolysis by xylanases.

MICROBIAL XYLANOLYTIC ENZYME SYSTEM

153

The occurrence of debranching activities in xylanases appears to be variable. The four xylanases isolated from ?: reesei (Dekker, 1985) were reported to be arabinose-releasing enzymes as were the 18-kDa xylanases from 7: koningii and T viride. However, six other characterized Trichoderma xylanases lack this activity for removing the arabinosyl substituents from arabinoxylans. On the other hand, the removal of glucuronosyl residues is not an activity that has been associated with xylanases. Although this activity has seldom been examined in Trichoderma xylanases, it has been reported to be absent from the 20-kDa xylanase from T reesei (Poutanen and Puls, 1989),from the 18-kDa xylanase from T viride (Sinner and Dietrichs, 19761, and from a crude preparation of 20-kDa xylanase from T harzianum (Clark et al., 1990, 1991). This debranching activity has generally been attributed to other enzymes, namely, a-glucuronidases. There is, however, some evidence for the release of some acetic acid during the hydrolysis of an acetylxylan by a purified xylanase from ?: reesei (Poutanen and Puls, 1989). Xylotriose is the smallest oligomer hydrolyzed by most of the characterized xylanases. Of the three reported exceptions, the 20-kDa xylanase from ?: harzianum hydrolyzed xylotetrose using a pathway initiated by transglycosylation (Tan et al., 1985b). Transferase activity was also reported in the 20-kDa xylanase from 'I: lignorum (John and Schmidt, 1988)and in two nonspecific glycanases (Beldman et a]., 1988; Biely et al., 1991). This type of reaction may enable a nonspecific glycanase to split xylobiose after it has been incorporated into cello-oligomers (Biely et al., 1991).Transglycosylation also appears to be an important reaction in the hydrolysis of xylobiose by a P-xylosidase from T viride (Matsuo and Yasui, 1984). The activity of this xylobiase apparently decreases with increasing chain length of xylooligomers. However, other xylobiases have higher activity on xylooligomers than on xylobiose (Beldman et al., 1988; John and Schmidt, 1988) and can have substantial hydrolytic activity on xylans (Beldman et al., 1988). Trichoderma xylanases are known to solubilize carbohydrates from cell wall preparations, holocellulose substrates, and kraft pulps. Only 20% of the xylosyl residues from corn shoot cell walls and 10% of those from bean shoot cell walls were solubilized by a xylanase purified from T pseudokoningii (Baker et al., 1977). Much lower levels of other sugar residues were solubilized, suggesting the high degree of selectivity observed in certain Trichoderma xylanases (Clark et al., 1990, 1991; Senior et a!., 1988; Sinner et al., 1976, 1979; Tan et a].,1985b; Viikari et al., 1990; Wong et al., 198fia). Although xylose was not detected in

154

P. BAJPAI

the cell wall hydrolysates obtained using the 'T: pseudokoningii xylanase, it was found in holocellulose and pulp hydrolysates obtained using xylanases from 7: harzianum, 'T: reesei, and 'T: viride. These latter enzymes could solubilize 11-71% of the xylan in hardwood and softwood holocellulose (Sinner et a]., 1976, 1979), 9-25% of that in kraft pulps (Clark et a]., 1990; Senior et al., 1988; Viikari et a]., 1990), and 54% of that in a bleached hardwood h a f t pulp (Senior et al., 1988). There was a decrease in the degree of polymerization (DPJ of the xylan remaining in h a f t pulps after xylanase treatment (Miller et al., 1991). In beechwood holocellulose, the percentage of xylan solubilized by a xylanase from 'T: viride increased with decreasing particle size of the substrate (Sinner eta]., 1976). At the smallest particle size tested (0.070.1 mm), xylan accessibility could not be increased in beechwood holocellulose using a cellulase treatment (Sinner et al., 1976) or in spruce holocellulose using a mannanase treatment (Sinner et al., 1979). These observations suggest that xylan accessibility in solid substrates is not dependent on the other carbohydrates. In intact h a f t pulp fibers, however, xylan accessibility to a xylanase from Aspergillus foetidus could be increased using a cellulase (Puls et a]., 1990) and that to a xylanase from T harzianum using an extraction with sodium hydroxide (Clark et al., 1991). Xylan accessibility in solid substrates, therefore, appears to be dependent on the nature of these substrates. This is also illustrated by the unexpected observation that xylan accessibility increases in a series of radiata pine kraft pulps containing increasing levels of residual lignin, even though the percentage of xylan content decreases (Clark et al., 1990). In addition, the nature of solid substrates could have an inhibitory effect on xylanases from I: harzianum (Senior et a]., 1990, 1991). Furthermore, the hydrolytic properties of the enzymes is important because different xylanases from this fungus provided different hydrolysis yields from aspen holocellulose, with the maximum achieved using a mixture of all three enzymes (Wong et a]., 1986b).

B. P-XYLOSIDASES Exo-1,4-P-D-xylosidases (EC 3.2.1.37) hydrolyze xylooligosaccharides and xylobiose to xylose by removing successive D-xylose residues from the nonreducing termini. P-xylosidase is part of most microbial xylanolytic systems, but the highest extracellular production levels have been reported for fungi. P-xylosidases are rather large enzymes with molecular weights exceeding 100 kDa and are often reported to consist of two or more

MICROBIAL XYLANOLYTIC ENZYME SYSTEM

155

subunits (Kersters-Hilderson et al., 1982; Matsuo and Yasui, 1984; Poutanen and Puls, 1988; Rodionova et al., 1983; Ujiie et al., 1985). Most purified P-xylosidases show highest activity toward xylobiose and no activity toward xylan. The activity toward xylooligosaccharides generally decreases rapidly with increasing chain length (Rodionova et al., 1983; Van Doorslaer et al., 1985). In addition to formation of xylose, many 0-xylosidases produce transfer products with higher molecular weights than that of the substrate (Conrad and Nothen, 1984; Rodionova et al., 1983). Some P-xylosidases have also been reported to possess P-glucosidase activity (Rodionova et a]., 1983; Ujiie et al., 1985). An important characteristic of P-xylosidases is their susceptibility to inhibition by xylose, which may significantly affect the yield under process conditions (Dekker, 1983; Poutanen and Puls, 1988; Rodionova et a]., 1983). P-xylosidase is the key enzyme for production of monomeric xylose from soluhilized xylan fragments, such as those obtained from a steaming process (Poutanen and Puls, 1988). They have been shown to act in synergy with suhstitutent-cleaving enzymes in the hydrolysis of suhstituted xylooligosaccharides (Poutanen, 1988a; Puls et al., 1987). The P-xylosidase of Trichoderma reesei was not able to hydrolyze xylobiose bearing an acetyl substituent at the nonreducing end without the presence of acetyl xylan esterase (Poutanen et d., 1990h). C. ~-ARABINOSIDASES

a-L-arabinofuranosidases(EC 3.2.1.55) hydrolyze nonreducing a-Larahinofuranosyl groups of arabinans, arabinoxylans, and arabinogalactans, as reviewed by Kaji (1984). The production of arabinosidases in microorganisms is often associated with the production of pectinolytic or hemicellulolytic enzymes, for example, in Corticiuni rolfsii (Kaji and Yoshihara, 1970), Sclerotina fructigena (Feilding and Byrde, 1969), T reesei (Poutanen et al., 1987; Poutanen, 1988b), and different Streptomyces species (Johnson et al., 1988a,h; Kaji et al., 1981). Some reported molecular characteristics of a-arabinosidases are presented in Table 111. The purified a-arabinosidase of Aspergillus niger (Tagava and Kaji, 1969), as well as that partially purified from a commercial pectinase preparation (Neukom et al., 1967), was able to release L-arahinose from wheat L-arabino-D-xylan. As the reaction proceeded, an amorphous precipitate consisting mainly of D-xylan with only traces of arahinose was formed. Adrewartha et 01. (1979) prepared a series of arabinoxylans

156

P. BAJPAI TABLE 111 CX-ARABINOSIDASES'

Molecular weight (kDa)

Microorganism

Aspergillus niger Trichoderma reesei Streptomyces spp. S trep tomyces p urp urescens Ruminococcus nlhus

53h

PI 3.6

53"

7.5

92'

4.4

495,h 62'

3.9

305,b 75'

6.8

"Sources: Poutanen et a].,1991;Kaji et al., 1969;Komae et al., 1982:Tagava and Kaji, 1988. 'Gel chromatography. 'SDS-PAGE.

from purified wheat-flour arabinoxylan by partial removal of arabinosyl side branches using an a-L-arabinosidase. They suggested that the solublizing effect of the arabinosyl substituents was not a result of increased hydration, but was due to their ability to prevent intermolecular aggregation of unsubstituted xylose residues. Cereal endospermic arabinoxylans especially are known to form viscous solutions and gels. It is obvious that suitable a-arabinosidases could be used to control the degree of substitution and hence the water-binding capacity of these pentosans. In a similar way, a-galactosidases have been used in adjusting the degree of a-galactosyl substitution and hence the gelling properties of galactomannans (Fujita and Nakamura, 1986; Overbeeke et al., 1987).

D. a-GLUCURONIDASES a-glucuronidases are required for hydrolysis of the a-l,2-glycosidic linkage between xylose and D-glucuronic acid or its 4-0-methyl ether. The presence of acidic oligosaccharides in xylan hydrolysates produced by hemicellulolytic enzyme preparations indicates the absence or inadequacy of this enzyme (Poutanen et al., 1987; Sinner and Dietrichs, 1976). 4-0-methylglucuronic acid was first detected in the enzymatic hydrolysates of glucuronoxylan by Sinner et al. (1972). The presence of a uronic acid-liberating enzyme, together with P-xylosidase, was

157

MICROBIAL XYLANOLYTIC ENZYME SYSTEM TABLE IV CX-GLIJCIJRONUIASES AND ITS SUBSTRATES’

Organisms

Substrate

A. bisporus T reesei

4-O-MeGlcAXz

S. olivochromogenes

Wheat bran

M.paranaguensis

Larchwoodxylan, wheat bran xylanase, p-nitorphenyl-uglucuronoside

S . olivochromogenes D. dendroides T palustris Trichoderma spp. l? versicolor L. sulphureus A. bisporus l? ostreatus Streptomyces spp. T reesei T aurantiacus

+ xylanase t

(4-0-MeGfcAJ-xylitol

Larchwoodxylan

+ xylanase

4-O-MeGlc AXz 4-0-MeGlcAX3

“Sources: Puls, 1992; Puls et al., 1986; Mackenzie et a]., 1987; Ishihara and Shimizu, 1988: Fontana ef ol., 1988.

claimed to increase the xylose yield in the enzymatic hydrolysis of hardwood xylan (Puls et al., 1976). The presence of a-glucuronidase in the hemicellulolytic system of T reesei was demonstrated in 1983 by Dekker. The production of a-glucuronidase by many fungi and bacteria (Table IV) has been reported (Puls, 1992). Only a few a-glucuronidases have been totally or even partially purified and characterized (Table V). The a-glucuronidase isolated from a culture filtrate of Agaricus bisporus by gel chromatography is a very large protein (450 kDa) (Puls et al., 1987). The enzyme had a very low isoelectric point and a pH optimum of about 3.3. A series of 4-0-methylglucuronosubstituted xylooligosaccharides with a DP up to 5 tested as substrates showed highest activity against 4-0-methylglucuronoxylobiose (Korte, 1980). The a-glucuronidase of A . bisporus had no activity toward polymeric xylan. The a-glucuronidase of ?: reesei also had an acidic isoelectric point (Poutanen, 1988a). It had a molecular weight of about 70 kDa as estimated by gel chromatography and a pH optimum at 6 with 4-0-methylgucurono-xylobiose as substrate. The a-glucuronidase of the thermo-

158

P. BAJPAI TABLE V PURIFIED CL-GLLJCURONIDASES" Molecular

PH

weight

Organism

Trichoderma reesei Rut C-30 Trichoderma reesei Thermoascus aurantiacus Agaricus bisporus

(kDa)

PI

optimum

>lo0 100