Future Directions in Biocatalysis
This page intentionally left blank
Future Directions in Biocatalysis
Edited by
TOMOKO MATSUDA Tokyo Institute of Technology, Yokohama, Japan
Amsterdam • Boston • Heidelberg • London • New York • Oxford
Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo
Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands Linacre House, Jordan Hill, Oxford OX2 8DP, UK First edition 2007 Copyright © 2007 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email:
[email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-444-53059-2 For information on all Elsevier publications visit our website at books.elsevier.com Printed and bound in The Netherlands 07
08
09
10
11
10
9
8
7
6
5
4
3
2
Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org
1
v
Contents
Preface
xi
Part 1 Novel reaction conditions for biotransformation
1
CHAPTER 1
Biotransformation in ionic liquid
3
Toshiyuki Itoh 1. Introduction 2. Ionic Liquids as a Reaction Medium for Biotransformation 3. Lipase-Catalyzed Reaction in an Ionic Liquid Solvent System 4. Activation of Lipase by an Ionic Liquid 5. Various Biotransformations in an Ionic Liquid Solvent System 6. Concluding Remarks References
3
3
7
10
15
18
18
CHAPTER 2
Temperature control of the enantioselectivity in the lipase-catalyzed
resolutions
21
Takashi Sakai 1. 2. 3. 4.
Introduction Finding of the “Low-Temperature Method” in the Lipase-Catalyzed Kinetic Resolution Theory of Temperature Effect on the Enantioselectivity General Applicability of the “Low-Temperature Method” Examined 4.1. Application to solketal and other primary and secondary alcohols 4.2. Resolution of (±)-2-hydroxy-2-(pentafluorophenyl)acetonitrile 4.3. Immobilization of lipase on porous ceramic support (Toyonite) for acceleration 4.4. Structural optimization of organic bridges on Toyonite 4.5. Practical resolution of azirine 1 by the “low-temperature method” combined
with Toyonite-immobilized lipase and optimized acylating agent 4.6. Resolution of (2R∗ 3S ∗ )- and (2R∗ 3R∗ )-3-methyl-3-phenyl-2-aziridinemethanols 4.7. Resolution of 5-(hydroxymethyl)-3-phenyl-2-isoxazoline 4.8. Application of temperature control to asymmetric protonation 4.9. Lipase-catalyzed resolutions at high temperatures up to 120 C 5. Low-Temperature Reactions in Literatures 6. Lipase-Catalyzed Resolution of Primary Alcohols: Promising Candidates for the
“Low-Temperature Method” 7. Conclusion References
21
22
23
28
28
30
31
32
33
34
36
37
37 37
40
45
45
vi
Contents
CHAPTER 3
Future directions in photosynthetic organisms-catalyzed reactions
51
Kaoru Nakamura 1. Introduction 2. Reduction Reaction 3. Oxidation and Hydroxylation 4. Removal of Organic and Inorganic Substances in Wastewater 5. Conclusion References
51
51
55
56
57
57
CHAPTER 4
Catalysis by enzyme–metal combinations
59
Mahn-Joo Kim, Jaiwook Park, Yangsoo Ahn, Palakodety R. Krishna 1. Introduction 2. Dynamic Kinetic Resolutions by Enzyme–Metal Combinations 2.1. DKR of secondary alcohols 3. Asymmetric Transformations by Enzyme–Metal Combinations 3.1. Asymmetric transformation of ketone 3.2. Asymmetric transformation of enol ester 3.3. Asymmetric transformation of ketoxime 4. Conclusion Acknowledgements References
59
60
60
73
73
75
76
78
78
79
Part 2 Uncomon kind of biocatalytic reaction
81
CHAPTER 5
Biological Kolbe–Schmitt carboxylation
83
Possible use of enzymes for the direct carboxylation of organic substrates
Toyokazu Yoshida, Toru Nagasawa 1. Introduction 2. Enzymes Catalyzing the Carboxylation of Phenolic Compounds 2.1. 4-Hydroxybenzoate decarboxylase (EC 4.1.1.61) 2.2. 3,4-Dihydroxybenzoate decarboxylase (EC 4.1.1.63) 2.3. Phenolphosphate carboxylase (EC 4.1.1.-) in Thauera aromatica 2.4. 2,6-Dihydroxybenzoate decarboxylase 2.5. 2,3-Dihydroxybenzoate decarboxylase 3. Enzymes Catalyzing the Direct Carboxylation of Heterocyclic Compounds 3.1. Pyrrole-2-carboxylate decarboxylase 3.2. Indole-3-carboxylate decarboxylase 4. Structure Analysis of Decarboxylases Catalyzing CO2 Fixation 4.1. Class I decarboxylases 4.2. Class II decarboxylases 4.3. Phenylphosphate carboxylase 5. Conclusion References
83
84
85
87
88
91
95
95
96
99
101
102
103
103
103
104
Contents
vii
CHAPTER 6
Discovery, redesign and applications of Baeyer–Villiger monooxygenases 107
Daniel E. Torres Pazmin˜ o, Marco W. Fraaije 1. Introduction 2. Biocatalytic Properties of Recombinant Available BVMOs 2.1. Discovery of novel BVMOs 2.2. Exploring sequenced (meta)genomes for novel BVMOs 2.3. Screening the metagenome for novel BVMOs 2.4. Redesign of BVMOs 3. Conclusions: Future Directions References
CHAPTER 7
Enzymes in aldoxime–nitrile pathway: versatile tools in biocatalysis
107
110
112
114
118
119
122
125
129
Yasuhisa Asano 1. Introduction 2. Screening for New Microbial Enzymes by Enrichment and Acclimation
Culture Techniques 3. Development of Nitrile-Degrading Enzymes 4. Screening for Heat-Stable NHase 5. Screening for NHase with PCR 6. Nitrile Synthesis Using a New Enzyme, Aldoxime Dehydratase 6.1. Aldoxime-converting enzymes 6.2. Isolation of microorganisms having aldoxime dehydratase activity 6.3. Purification, characterization and primary structure determination of aldoxime
dehydratase 6.4. Synthesis of nitriles from aldoxime with aldoxime dehydratase 6.5. Distribution of aldoxime dehydratase 6.6. Molecular screening for “aldoxime–nitrile pathway” 7. Conclusions Acknowledgments References
CHAPTER 8
Addition of hydrocyanic acid to carbonyl compounds
129
129
131
131
132
133
133
134
134
135
136
136
137
137
137
141
Franz Effenberger, Anja Bohrer, Siegfried Fo¨ rster 1. Introduction 2. Optimized Reaction Conditions for the HNL-Catalyzed Formation of Chiral Cyanohydrins 3. Synthetic Potential of Chiral Cyanohydrins in Stereoselective Synthesis 3.1. Chiral 2-hydroxy carboxylic acids 3.2. Optically active 1,2-amino alcohols 3.3. Stereoselective substitution of the hydroxyl group in chiral cyanohydrins 3.4. Stereoselective synthesis of substituted cyclohexanone cyanohydrins
141
143
145
145
147
148
149
viii
Contents
4. Crystal Structures of Hydroxynitrile Lyases and Mechanism of Cyanogenesis 4.1. Crystal structures of HNLs 4.2. Reaction mechanism of cyanogenesis 4.3. Changing substrate specificity and stereoselectivity applying Trp128 mutants of
wt-MeHNL 5. Conclusions References
149
151
151
152
153
154
Part 3 Novel compounds synthesized by biotransformations
157
CHAPTER 9
Chiral heteroatom-containing compounds
159
Piotr Kiełbasinski, Marian Mikołajczyk ´ 1. Introduction 2. Organosulfur Compounds 2.1. C-chiral hydroxy sulfides and derivatives 2.2. C-chiral hydroxyalkyl sulfones 2.3. C-chiral alkyl sulfates 2.4. Other C-chiral organosulfur compounds 2.5. S-chiral sulfinylcarboxylates 2.6. S-chiral hydroxy sulfoxides 2.7. S-chiral sulfinamides 2.8. S-chiral sulfoximines 3. Organophosphorus Compounds 3.1. C-chiral hydroxy phosphorus derivatives 3.2. C-chiral amino phosphorus compounds 3.3. P-chiral phosphoro-acetates 3.4. P-chiral hydroxy phosphoryl compounds 3.5. P-chiral hydroxy phosphorus P-boranes 3.6. Stereocontrolled transformations of organophosphorus acid esters 4. Organosilanes 5. Organogermanes 6. Future Perspectives References
CHAPTER 10
Enzymatic polymerization
159
160
160
163
165
166
166
168
169
171
172
172
180
183
186
191
192
196
197
197
199
205
Hiroshi Uyama 1. Introduction 2. Enzymatic Synthesis of Polyesters 2.1. Ring-opening polymerization to polyesters 2.2. Polycondensation of dicarboxylic acid derivatives and glycols to polyesters 2.3. Enzymatic synthesis of functional polyesters 3. Enzymatic Synthesis of Phenolic Polymers 3.1. Enzymatic oxidative polymerization of phenols 3.2. Enzymatic synthesis of functional phenolic polymers
205
206
207
212
219
228
228
233
Contents 3.3. Artificial urushi 3.4. Enzymatic synthesis and biological properties of flavonoid polymers 4. Concluding Remarks References
CHAPTER 11
Synthesis of naturally occurring �-D-glucopyranosides based on
enzymatic �-glucosidation using �-glucosidase from almond
ix 238
240
244
245
253
Hiroyuki Akita 1. Introduction 2. Synthesis of -d-Glucopyranoside Under Kinetically Controlled Condition 2.1. Synthesis of naturally occurring -d-glucopyranoside 3. Synthesis of -d-Glucopyranoside Under Equilibrium-Controlled Condition 3.1. Immobilization of -d-glucosidase using prepolymer 3.2. Enzymatic transglucosidation 3.3. Synthesis of naturally occurring benzyl -d-glucopyranoside 3.4. Synthesis of phenethyl -d-glycopyranoside 3.5. Synthesis of (3Z)-hexenyl -d-glycopyranoside 3.6. Synthesis of geranyl -d-glycopyranoside 3.7. Synthesis of Sacranosides A (89) and B (90) 3.8. Synthesis of naturally occurring n-octyl -d-glucopyranosides 3.9. Synthesis of naturally occurring hexyl -d-glucopyranosides 3.10. Synthesis of naturally occurring phenylpropenoid -d-glucopyranoside 4. Future Aspect 5. Conclusion References
253
255
259
262
263
263
267
270
272
275
277
278
280
282
287
289
289
Part 4 Use of molecular biology technique to find novel biocatalyst
291
CHAPTER 12
Future directions in alcohol dehydrogenase-catalyzed reactions
293
Jon D. Stewart 1. Introduction 2. Future Progress in the Discovery Phase of Dehydrogenases 2.1. Accurately predicting dehydrogenase structures 2.2. Predicting dehydrogenase substrate acceptance and stereoselectivities. 2.3. Rapid screening of novel dehydrogenases 2.4. Dehydrogenases for large substrates 2.5. Dehydrogenase modules within larger assemblies as monofunctional catalysts 2.6. Dehydrogenase catalysis of other 1,2-carbonyl additions 3. Future Progress in Dehydrogenase Process Development 3.1. Improving the kinetic properties of dehydrogenases 3.2. Reductions of highly hydrophobic substrates 3.3. Cofactorless dehydrogenases? 4. Conclusions Acknowledgments References
293
295
295
296
296
299
299
300
300
301
301
302
302
303
303
x
Contents
CHAPTER 13
Enzymatic decarboxylation of synthetic compounds
305
Kenji Miyamoto, Hiromichi Ohta 1. Introduction 2. Arylmalonate Decarboxylase 2.1. Discovery of arylmalonate decarboxylase and its substrate specificity 2.2. Purification of the enzyme and cloning of the gene 2.3. Reaction mechanism 2.4. Inversion of enantioselectivity based on the reaction mechanism and homology 2.5. Addition of racemase activity 3. Transketolase-Catalyzed Reaction 3.1. Substrate specificity and stereochemical source of TKase-catalyzed reaction 3.2. Application of TKase-catalyzed reaction in organic syntheses 3.3. Tertiary structure and mutagenesis studies 4. Future Trends of this Area 4.1. Application of decarboxylation reaction to dialkylmalonates 4.2. Decarboxylation of various carboxylic acids 4.3. Oxidative decarboxylation of -hydroxycarboxylic acids 4.4. Carboxylation 4.5. Development of biotransformation via enolate 4.6. Utilization of database and informatics 5. Conclusion References
Index
305
309
310
311
312
317
319
321
322
322
329
331
331
332
333
336
337
339
339
340
345
xi
Preface
Important topics in biocatalysis for organic synthesis are described in this book, for experts and non-experts. Especially, the book focuses on those reactions that are under development now and will be more significant in the future. Therefore, each chapter describing a specific theme summarizes not only the present state but also the direction of the research. The prospects and dreams that will become possible, using biocatalysis, in the future to construct a sustainable society are also included. The book consists of four sections: the enzymatic reaction under unusual conditions, unique biocatalytic reactions, valuable compounds synthesized using biocatalysis and latest molecular biology methods to make useful biocatalysts. The first section dealing with unusual reaction conditions for biocatalysis begins with the use of ionic liquid as a solvent to develop green chemistry. Then, the reaction under extreme temperatures, use of light energy as a driving force to proceed biocatalysis and catalysis by enzyme-metal combinations are described. The second section covers very unique reactions such as carboxylation using decarboxylases and carbon dioxide, Baeyer-Villiger reaction using monooxygenases, reactions in aldoxime-nitrile pathway including dehydration in aqueous solvent and addition of nitrile to carbonyl compounds to synthesize chiral cyanohydrin. The third section highlights novel compounds synthesized using biocatalysts. Chiral heteroatomcontaining compounds such as chiral phosphorus compounds, polymer materials and sugars are selected to demonstrate the usefulness of biocatalysts. The last section describes the use of molecular biology technique to find novel biocatalysts. Alcohol dehydrogenase and decarboxylase are chosen as an example for the use of techniques since detailed and interesting researches have been conducted using these enzymes. Through this book, I hope to introduce the novelty and future directions in biocatalysis. Importantly, I wish to contribute to conserve and beautify the natural environment by editing the book, showing the power of enzymes, treasure from Mother Nature, to catalyze the necessary reaction for mankind in an environmentalfriendly manner. Tomoko Matsuda
This page intentionally left blank
Part One Novel reaction conditions for biotransformation
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
3
Chapter 1
Biotransformation in ionic liquid Toshiyuki Itoh Department of Materials Science, Faculty of Engineering, Tottori University, Tottori, Japan
Abstract The use of ionic liquids (ILs) to replace organic or aqueous solvents in biocatalysis processes has recently gained much attention and great progress has been accomplished in this area; lipase-catalyzed reactions in an IL solvent system have now been established and several examples of biotransformation in this novel reaction medium have also been reported. Recent developments in the application of ILs as solvents in enzymatic reactions are reviewed.
1. INTRODUCTION Ionic liquids (ILs) have very good properties as reaction medium in chemical reactions: they are non-volatile, non-flammable, have low toxicity and good solu bility for many organic and inorganic materials.1 It has long been recognized that an enzymatic reaction proceeds in an aqueous buffer solution under appropriate pH conditions and an enzyme quickly loses its activity in a highly concentrated aqueous salt solution.2 Therefore it seems a foolish notion that enzymatic reaction occurs in a salt medium from the standpoint of biology. However, the use of ILs to replace traditional organic solvents in chemical reactions has recently gained much attention, and even as a novel reaction medium for biotransformation. Lipasecatalyzed reactions in an IL solvent system have now been established,1b−e and several types of non-lipase enzymatic reactions have also been reported recently. I wish to review recent progress in the area of “enzymatic reactions in an IL solvent system” in this chapter.
2. IONIC LIQUIDS AS A REACTION MEDIUM FOR BIOTRANSFORMATION Cull and co-authors3a reported a microbe-mediated transformation of benza mide from benzonitrile in a mixed solvent of IL, 1-butyl-3-methylimidazolium
4
Toshiyuki Itoh
Figure 1: The first enzymatic reaction conducted in a pure ionic liquid solvent system.
hexafluorophosphate ([bmim][PF6 ]), with water (1:4) in July 2000. Then Russell and co-authors3b reported that thermolysin-catalyzed amidation of CBz-asparagine with l-phenylalanine methyl ester proceeded in a mixed solvent of [bmim][PF6 ] with aqueous buffer solution. These examples showed that the IL had no inhibitory action against the enzymes because [bmim][PF6 ] was insoluble in water and enzy matic reactions took place in the water layer. The first example of enzymatic reac tion in a pure IL solvent system was reported by the Sheldon group in December 2000.4 The authors successfully demonstrated two types of Candida antarctica lipase (CAL-B) catalyzed reaction in a pure IL: CAL-B catalyzed amidation of octanoic acid with ammonia and also the formation of octanoic peracid by the reaction of octanoic acid with hydrogen peroxide (Fig. 1). However, the reactions were not enantioselective ones, though the most important aspect of the biocatalysis reaction should be in the enantioselective reaction. We5a and Kragl6 independently reported the first enantioselective lipasecatalyzed reaction in February–March 2001. Since lipase was anchored by the IL solvent and remained in it after the extraction work-up of the product, we succeeded in demonstrating that recyclable use of the lipase in the [bmim][PF6 ] solvent system was possible (Fig. 2).5a Typically the reaction was carried out as follows: to a mixture of lipase in the IL were added this racemic alcohol and vinyl acetate as the acyl donor. The resulting mixture was stirred at 35 C and the reaction course was monitored by GC analysis. After the reaction, ether was added to the reaction mixture to form a biphasic layer, and product acetate and unreacted alcohol were extracted with ether quantitatively. The enzyme remained in the IL phase as expected (Fig. 2). Two months later, Kim and co-workers7a reported similar results and Lozano and Ibora7bd reported other examples of lipase-catalyzed reaction in June. Further Park and Kazlauskas7c reported full details of lipase-catalyzed reaction in an IL solvent
Biotransformation in ionic liquid
5
Figure 2: Lipase-catalyzed reaction system anchored to the solvent. system in August 2001. Studies on the enzymatic reaction in an IL solvent system were thus launched in 2000–2001. We initially tested Candida antarctica lipase using imidazolium salt as sol vent because CAL was found to be the best enzyme to resolve our model substrate 5-phenyl-1-penten-3-ol (1a); the acylation rate was strongly dependent on the anionic part of the solvents. The best results were recorded when [bmim][BF4 ] was employed as the solvent, and the reaction rate was nearly equal to that of the refer ence reaction in diisopropyl ether. The second choice of solvent was [bmim][PF6 ]. On the contrary, a significant drop in the reaction rate was obtained when the reaction was carried out in TFA salt or OTf salt. From these results, we concluded that BF4 salt and PF6 salt were suitable solvents for the present lipase-catalyzed reaction.5a Acylation of 1a was accomplished by these four enzymes: Candida antarctica lipase, lipase QL from Alcaligenes, Lipase PS from Burkholderia cepa cia and Candida rugosa lipase. In contrast, no reaction took place when PPL or PLE was used as catalyst in this solvent system. These results were established in March 2000 but we encountered a serious problem in that the results were significantly dependent on the lot of the ILs that we prepared ourselves. The problem was very serious because sometimes the reaction did not proceed at all. So we attempted to purify the ILs and established a very successful procedure (Fig. 3): the salt was first washed with a mixed solvent of hexane and ethyl acetate (2:1 or 4:1), treated with activated charcoal and passed into activated alumina neutral type I as an acetone solution. It was evaporated and dried under reduced
6
Toshiyuki Itoh
Figure 3: Purification protocol of imidazolium ionic liquid.
pressure at 50 C for 24 h to obtain very clean imidazolium salt. Using the ILs, we succeeded in obtaining more reproducible results; I recommend this as also being very useful to recycle the IL. In fact, we always recycle our ILs after the reaction and have not wasted any in the past. We are still using ILs that have a 7-year history. After establishing the reproducibility of our results of lipasecatalyzed reaction, we submitted our first paper in December 2000 and the paper was accepted on January 5, 2001. Although we lost the chance to be the first to publish in the field for this reason, we learned many things about ILs during that time and these are now important bases of our research group. Very pure ILs are commercially available now and we can use them, but I imagine that all research groups encountered the same problem in the early days of this field, because very clean ILs are required for a biocatalysis system compared to chemical reactions. This story highlights a very important point; we should pay attention to the qual ity of the IL when we evaluate the appropriate one for our desired biocatalyst reaction.
Biotransformation in ionic liquid
7
3. LIPASE-CATALYZED REACTION IN AN IONIC LIQUID SOLVENT SYSTEM We succeeded in showing that recycling of the enzyme was indeed possible in our IL solvent system, though the reaction rate gradually dropped with repetition of the reaction process.5a Since vinyl acetate was used as acyl donor, acetaldehyde was produced by the lipase-catalyzed transesterification. It is well known that acetaldehyde acts as an inhibitor of enzymes because it forms a Schiff base with amino residue in the enzyme. However, due to the very volatile nature of acetaldehyde, it easily escapes from the reaction mixture and therefore has no inhibitory action on the lipase. However, this drop in reactivity was assumed to be caused by the inhibitory action of acetaldehyde oligomer which had accumulated in the [bmim][PF6 ] solvent system. In fact, it was confirmed that the reaction was inhibited by addition of acetaldehyde trimer.5c One of the most important characteristics of IL is its wide temperature range for the liquid phase with no vapor pressure, so next we tested the lipase-catalyzed reaction under reduced pressure. It is known that usual methyl esters are not suit able for lipase-catalyzed transesterification as acyl donors because reverse reaction with produced methanol takes place. However, we can avoid such difficulty when the reaction is carried out under reduced pressure even if methyl esters are used as the acyl donor, because the produced methanol is removed immediately from the reaction mixture and thus the reaction equilibrium goes through to produce the desired product.8 To realize this idea, proper choice of the acyl donor ester was very important. The desired reaction was accomplished using methyl phenylth ioacetate as acyl donor. Various methyl esters can also be used as acyl donor for these reactions; methyl nonanoate was also recommended and efficient optical resolution was accomplished. Using our system, we demonstrated the completely recyclable use of lipase. The transesterification took place smoothly under reduced pressure at 10 Torr at 40 C when 0.5 equivalent of methyl phenylthioacetate was used as acyl donor, and we were able to obtain this compound in optically pure form. Five repetitions of this process showed no drop in the reaction rate (Fig. 4).5bc Recently Kato reported nice additional examples of lipase-catalyzed reaction based on the same idea that CAL-B-catalyzed esterification or amidation of carboxylic acid was accomplished under reduced pressure conditions.9 However, it still remained a problem that the system could not be applied to volatile substrates. We hypothesized that oligomerization of acetaldehyde may be caused by the proton derived from the water molecule trapped by hydrogen bonding at 2-position of the imidazolium ring, because it was suggested that the acidity of the 2-position of imidazolium cation is very high.10 Hence we prepared 1-butyl-2,3-dimethylimidazolium (bdmim) salt and used it as solvent. As we expected, no accumulation of an acetaldehyde oligomer was observed in this solvent system by1 H NMR analysis. The reaction proceeded very smoothly and we were able to use the enzyme repeatedly 10 times while still maintaining perfect
8
Toshiyuki Itoh
Figure 4: Lipase-catalyzed reaction under reduced pressure conditions.
Novozyme-435 Vinyl acetate
OAc
Ph Me
N
N Me
Relative rate (%conv./h)
OH + Ph
Ph Bu BF4
40
300
30
200
100
20
0 Run 1
E value
OH
1 2
3
4
5
6
7
8
9
10
Figure 5: Recyclable use of lipase in [bdmim][BF4 ] solvent system. enantioselectivity and high reactivity (Fig. 5).5d It has now been confirmed that we can use the enzyme more than 20 times for several months using this system. Imidazolium PF6 or BF4 salts were frequently used as solvent for the present lipase-catalyzed reaction. However, these salts are very expensive, and we should develop cheaper ILs. Imidazolium alkyl sulfates might be good candidates because various types of alkyl sulfates can be easily prepared. The imidazolium alkyl sulfate was prepared starting from the corresponding ammonium alkyl sulfate as follows: ammonium alkyl sulfates ([NH4 ][RSO4 ]) are easily prepared by the reaction of
Biotransformation in ionic liquid
9
an alcohol with sulfuric amide, and subsequent anion exchange reaction with [bmim][Cl] gave the corresponding [bmim][RSO4 ].5e The lipase-catalyzed transesterification proceeded in these solvent systems and optically pure acetate was obtained with excellent enantioselectivity. Ethoxyethyl sulfate and phenoxyethyl sulfate gave excellent results and the acetates were obtained with > 99% ee.5e However, interestingly, no reaction took place when [bmim][EtSO4 ] was used as solvent. We recently prepared various types of differently fluorinated alkyl sulfate ILs and discovered that the hydrophobicity was dependent on the content ratio of the fluorine on the alkyl sulfate anion and 2,2,3,3,4,4,5,5-octafluoropentyl sulfate salts showed hydrophobic properties. Melting point and viscosity were also depen dent on the fluorine contents of the anionic part, while conductivity was deter mined by the cationic part and not influenced by the fluorine contents. Efficient lipase-catalyzed transesterification was demonstrated using hydrophobic 1-butyl 3-methylimidazolium 2,2,3,3,4,4,5,5-octafluoropentyl sulfate ([bmim][C5F8]) as solvent (Fig. 6).11 Kim and co-workers12 recently reported an excellent example of dynamic kinetic resolution (DKR) using lipase-ruthenium combo catalyst in an IL solvent system (Fig. 7). Applied to this protocol, the authors succeeded in preparing (R) ester or (S)-ester using lipase PS or subtilisin, respectively. An IL solvent system is truly appropriate for DKR because racemization takes place easily in a highly polar solvent. Lipase-catalyzed reaction is useful for polyester synthesis and IL was employed successfully as solvent. Uyama and Kobayashi13 demonstrated an effi cient polyester synthesis; lipase-catalyzed esterification of agipic acid with butan 1,4-diol proceeded smoothly in [bmim][BF4 ] solvent, particularly under reduced pressure conditions (Fig. 8). Further Russel14 and Nara15 independently reported efficient examples of the lipase-catalyzed polyester synthesis in an IL solvent system.
Figure 6: Lipase PS-catalyzed acylation in hydrophobic alkylsulfate IL.
10
Toshiyuki Itoh
Figure 7: Efficient DKR reaction in an IL solvent system.
Figure 8: Lipase-catalyzed polyester synthesis using IL solvent system. 4. ACTIVATION OF LIPASE BY AN IONIC LIQUID Lozano and co-workers7b reported an interesting stabilization effect of IL for lipase-catalyzed reaction; the authors discovered that the presence of an appropriate substrate was essential for stabilization of enzyme in an IL solvent. The half lifetime of native CAL was only 3.2 h in [emim][PF6 ] solvent, while it lengthened remarkably to 7500 h in the presence of the substrate. The authors succeeded in demonstrating an efficient lipase-recyclable use system based on scCO2 solvent (Fig. 9).1617 In the reaction, it was essential to use an IL as a co-solvent. Lozano, Iborra and co-workers recently reported an interesting stabilizing effect of two types of water-immiscible ILs ([emim][TFSI] and [BuMe3 N][TFSI]) for CAL B-catalyzed transesterification of vinyl butyrate.18 The synthetic activity and the stability of the enzyme in these IL solvent systems were markedly enhanced as compared to those in hexane. CAL-B maintained its activity higher than 75% after 4 days of incubation in [emim][TFSI] solvent, while it showed an activ ity of only 25% when incubated in both water and hexane media at 50 C. Comparison of the ratio of -helix and -sheet by CD spectra showed the activity was closely related with -helix content which reduced to 31% imme diately after lipase was added to hexane and had reached only 2% after 4 days in hexane. On the contrary, no significant reduction of -helix content was
Biotransformation in ionic liquid
11
°
Figure 9: Lipase recyclable use reactor using a mixed solvent of scCO2 with IL.
obtained in [emim][TFSI] solvent. Based on these results, the authors concluded that -helix content might play an important role in maintaining the enzymatic activity. Polyethyleneglycol (PEG) treatment is known to cause stabilization of an enzyme. Goto and co-workers prepared PEG-coated lipase and demonstrated that transesterification of vinyl cinnamate with butanol proceeded smoothly using PEGlipase PS in [omim][PF6 ] solvent system.19 The Russell group investigated the details of PEG treatment on the lipase activity in various types of ILs and reported that this activity was mainly dependent on the anionic part of the ILs (Fig. 10)14 ; high activity was obtained for imidazolium PF6 salt, while no activity was observed for [NO3 ], [OAc], [CH3 SO3 ], [OTf] or [TFA] salt. The authors proposed that nitric anion or acetate anion might have a strong interaction with some parts of the enzyme protein due to the highly nucleophilic nature of these anions and caused deactivation of the enzyme activity.14 Sheldon’s group also investigated the relationship of activity and IL species; CAL-B showed no catalytic activity in some ILs, such as [emim][EtSO4 ], [bmim][lactate], [EtNH3 ][NO3 ] or [bmim][NO3 ], though the enzyme was soluble in these solvents. On the other hand, high activity was obtained in [bmim][PF6 ] or [bmim][BF4 ] although the enzyme was insoluble in them. Interestingly deactivation was recorded for [emim][EtSO4 ], while the enzyme worked in [emim][MeSO4 ]. The authors speculated that this deactivation by the IL might be caused by con formational change due to the interaction of the anionic part of the IL with lipase protein based on the results of FT-IR analysis of the protein.20
12
Toshiyuki Itoh
Figure 10: Lipase-catalyzed acylation: the reaction depended on the counter anionic part of IL. A mixed solvent system of an IL with organic solvent sometimes gave very nice results: Lundell21 reported that enhanced enantioselectivity was obtained when lipase-catalyzed acylation was carried out in a mixed solvent system of [emim][TFSI] with t-BuOMe (1:1), while poor enantioselectivity was recorded for that in the pure [emim][TFSI] solvent (Fig. 11). Ganske and co-workers22ab reported that lipase-catalyzed acylation of a glucose derivative proceeded smoothly in a mixed solvent of [bmim][BF4 ] with t-BuOH, while no reaction took place in [bmim][BF4 ] (Fig. 12). These results taught us that a mixed solvent system of IL with organic solvent may be a good solution if the desired reaction did not take place in a pure IL solvent. Modified enantioselectivity was reported if the reaction was carried out in an IL solvent instead of traditional organic solvent.23 For instance, Kazlauskas and Kim independently reported that the regioselectivity of lipase-catalyzed reaction was enhanced if the reaction was carried out in an IL solvent system (Fig. 13).7c24 Recently Wu and co-workers reported another chip of lipase-catalyzed reaction: a
Figure 11: Lipase-catalyzed reaction in a mixed solvent system using microwave
reaction conditions at 80 C.
Biotransformation in ionic liquid
13
Figure 12: Lipase-catalyzed regioselective esterification of glucose in a mixed
solvent of IL with t-BuOH.
Figure 13: Enhanced regioselectivity of lipase-catalyzed acylation in IL solvent system.
significant enhanced enantioselectivity was obtained for Candida rugosa lipasecatalyzed acylation of ibuprofen in [bmim][BF4 ] compared to the isooctane solvent system (Fig. 14).25 Interestingly no enantioselectivity was obtained for [bmim][MeSO4 ], [bmim][OctylSO4 ], [(n-C6 H13 3 C14 H29 PN3 ], [Bu3 MeP][OTs] or [mmim][MeSO4 ] solvent. These results clearly showed that appropriate choice of the IL is the key point of the reaction.
Figure 14: Great diversity in enantioselectivity among IL solvent systems.
14
Toshiyuki Itoh
We investigated lipase-catalyzed acylation of 1-phenylethanol in the pres ence of various additives, in particular an IL additive using diisopropyl ether as solvent. Enhanced enantioselectivity was obtained when a BEG-based novel IL, i.e., imidazolium polyoxyethylene(10) cetyl sulfate, was added at 3–10 mol% vs. substrate in the Burkholderia cepacia lipase (lipase PS-C) catalyzed transesterification using vinyl acetate in diisopropyl ether or a hexane solvent system.5g Recently Kim reported another very interesting salt-mediated activation of a lipase: lipase PS was mixed with the IL and resulted in “ionic liquid-coated lipase PS” which showed more enhanced enantioselectivity than commercial lipase PS-C in toluene, though no significant modification of the reaction rate was obtained.26 We discovered that lyophilization was essential to activate the lipase effectively by the IL. The novel IL [bdmim][cetyl-PEG10-sulfate] (IL1) worked as an excellent activator of lipase PS-catalyzed acetylation of various types of secondary alcohols using vinyl acetate as acyl donor in i-Pr2 O solvent, while maintaining excellent enantioselectivity. More than 500- to 1000-fold acceler ation was accomplished for some substrates (Fig. 15 and Table 1).27 On the contrary, previously established activation protocol such as PEG coating of a lipase could accelerate the reaction but caused no significant enhanced enantios electivity; also using toluene as solvent sometimes provided enhanced enantios electivity for some substrates, but other times caused a significant reduction in the reaction rate. Since similar activation was observed for CRL, we expect that the activation effect of IL1 might be general for various lipases. We believe this work not only represents a significant advance in the manner of prepara tion of optically active compounds using an enzymatic reaction but also pro vides a new aspect in the application of an IL for such a reaction. Since the present activation was significantly dependent on the substrate, we are also hoping that further optimization of the cationic part of the ILs may make it possible to apply the present protocol of lipase activation to various broader types of substrates.
Figure 15: Results of activation effect by IL1-coating on the lipase PS-catalyzed
enantioselective acylation.
Biotransformation in ionic liquid
15
Table 1 Typical examples of results of activation effect of IL1 coating on the lipase PS-catalyzed enantioselective acylation of (±-1 Substrate
OH
Ratea (above) and E value (below)
Specific activityb of IL1-PS vs. PS-C
Lipase PS
IL1-PS
Rate: 65 E = 16
Rate: 1000 E > 200
15
Rate: 100 E = 39
Rate: 9800 E = 40
98
Rate: 13 E > 200
Rate: 14 E > 200
1
Rate: 10 × 10−2 E = 199
Rate: 11 E > 200
1100
Rate: 12 × 10−2 E > 200
Rate: 6 E > 200
500
(±)-1a
OH NC (±)-1b OH (±)-1c
OH (±)-1d HO
(±)-1e a b
Rate: mM h−1 mg (enzyme)−1 . Rate was determined by GC analysis. Specific activity was calculated by the rate of IL1-PS divided by that of PS-C.
5. VARIOUS BIOTRANSFORMATIONS IN AN IONIC LIQUID SOLVENT SYSTEM An IL solvent system is applicable to not only lipase but also other enzymes, though examples are still limited for lipase-catalyzed reaction in a pure IL solvent. But several types of enzymatic reaction or microbe-mediated reaction have been reported in a mixed solvent of IL with water. Howarth28 reported Baker’s yeast reduction of a ketone in a mixed solvent of [bmim][PF6 ] with water (10:1) (Fig. 16). Enhanced enantioselectivity was obtained compared to the reaction in a buffer solution, while the chemical yield dropped. Antibody-catalyzed aldol condensation was demonstrated in a [bmim][PF6 ] solvent system by Kitazume and co-workers29 (Fig. 17). They tested recyclable use of antibody catalyst in the solvent system and, very interestingly, found that the chemical yield was increased for the second cycle (89%) over the initial run (21%).
16
Toshiyuki Itoh
Figure 16: Baker’s yeast reduction of ketones in a mixed solvent of IL with
water.
Figure 17: Antibody-mediated aldol reaction in IL solvent system.
Okrasa and co-workers reported an interesting combination reaction of glucose oxidase and peroxidase in a mixed solvent of [bmim][PF6 ] with water (Fig. 18). Asymmetric oxidation of sulfide was accomplished successfully in the reaction system.30 Griengl reported the first example of hydroxynitrile lyase-catalyzed cyanohydrin formation in a mixed solvent system of [bmim][BF4 ] and buffer (pH 3.7) (1:1) (Fig. 19).31a In the reaction, a mixed solvent system was essential, but excellent results were obtained. Matsuda32 recently reported an elegant solution for using a delicate enzyme system in an IL solvent system (Fig. 20). The authors prepared Geotrichum candidum IFO5767 dried cell on water-absorbing polymer BL-100 and used
Figure 18: Glucose oxidase-mediated asymmetric oxidation of sulfide in IL
solvent system.
Biotransformation in ionic liquid
17
Figure 19: Enantioselective cyanohydrin formation using hydroxynitrile lyase in
IL solvent system.
Figure 20: Use of IL as reaction medium for asymmetric reduction by
Geotrichum candidum.
it for the asymmetric reduction of methyl ketones. The system made it pos sible to realize easy work-up and provided the corresponding alcohol with excellent enantioselectivity. Of interest was that the reactivity of the enzyme significantly depended on the counter anion of IL. The reaction proceeded smoothly in imidazolium PF6 , BF4 or TFSI salt, while no reaction took place in imidazolium OTf, EtSO4 , MeSO4 or NO3 salt as solvent. As reported in the reactions of lipases,1420 nitric anion or sulfate anion might have a strong interaction with some parts of the enzyme protein, due to the highly nucleophilic nature of these anions, and caused deactivation of the enzyme activity. Chiappe and co-workers33 reported chloroperoxide (CPO)-catalyzed oxi dation in hydrophilic ILs as co-solvents (Fig. 21). The authors investigated the hydrophilic ILs on the activity of CPO and found that CPO showed a higher toler ance toward IL than organic solvent; good activity was obtained when the reaction was carried out in a mixed solvent of [mmim][Me2 PO4 ] and buffer (pH 5.0) (1:1) rather than buffer solution.
18
Toshiyuki Itoh
Figure 21: CPO-catalyzed chlorination in IL solvent system.
6. CONCLUDING REMARKS There might be at least four advantages of the enzymatic reaction using an IL solvent system: (1) the purification process is very easy; (2) the reaction does not waste any water discarded with the organic solvent; (3) it is possible to use the enzyme repeatedly in this system; and (4) it is possible to activate the reaction using an IL-type supporting material. It is now completely established that ILs might be good candidates as the solvent of lipase-catalyzed reaction. However, we should consider the risk of using novel material. For example, Rogers warned that PF6 may be decomposed by moisture at higher temperature and produce toxic hydrogen fluoride.34a The authors also reported that cellulase was inhibited by the addition of [bmim][Cl]; chloride anion caused change of the folding pattern of the enzyme protein, which led to deactivation of the enzyme.34b However, imidazolium cation was reported to be easily degraded in the metabolism system.35 I believe that IL is superior to the conventional organic solvent if we try to create an enzymatic reaction under anhydrous conditions. To meet the challenge in chemistry of developing practical processes, the proper choice of a reaction medium is very important. Breakthroughs have sometimes come along with innovation of a reaction medium in chemical reactions and this is true even in enzymatic reactions. I hope this chapter may provide a hint for the reader’s research.
REFERENCES 1. Reviews see: (a) Wasserscheid, P.; Welton, T. (Eds). Ionic Liquids in Synthesis; Wiley-VCH: Weinheim, 2003. (b) Rogers, R. D.; Seddon, K. R. (Eds). Ionic Liquids as Green Solvents, ACS Symposium Series 856; American Chemical Society: Washington, DC, 2002. (c) Park, S.; Kazlauskas, R. J. Curr. Opin. Biotechnol. 2003, 14, 432. (d) Kragl, U.; Eckstein, M.; Kaftzik, N.
Biotransformation in ionic liquid
2.
3.
4. 5.
6. 7.
8.
9. 10. 11. 12. 13. 14. 15. 16. 17.
18. 19. 20. 21. 22. 23.
19
Curr. Opin. Biotechnol. 2003, 14, 565. (e) Jain, N.; Kumar, A.; Chauhan, S.; Chauhan, S. M. S. Tetrahedron 2005, 61, 1015. (f) Zhao, H. J. Mol. Catal. B Enzym. 2005, 37, 16. Reviews see: (a) Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic Chemistry; Pergamon: Oxford, 1994. (b) Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis: Regio- and Stereoselective Biotransformations; Wiley: Chichester, 1999. (a) Cull, S. G..; Holbrey, J. D.; Vargas-Mora, V.; Seddon, K. R.; Lye, G. J. Biotechnol. Bioeng. 2000, 69, 227. (b) Erbeldinger, M.; Mesiano, A. J.; Russell, A. J. Biotechnol. Prog. 2000, 16, 1131. Lau, R. M.; van Rantwijk, F.; Seddon, K. R.; Sheldon, R. A. Org. Lett. 2000, 2, 4189. (a) Itoh, T.; Akasaki, E.; Kudo, K.; Shirakami, S. Chem. Lett. 2001, 262. (b) Itoh, T.; Akasaki, E.; Nishimura, Y. Chem. Lett. 2002, 154. (c) Itoh, T.; Nishimura, Y.; Kashiwagi, M.; Onaka, M. In Ionic Liquids as Green Solvents: Progress and Prospects, ACS Symposium Series 856; Rogers, R. D.; Seddon, K. R. (Eds); American Chemical Society: Washington, DC; Chapter 21, pp. 251–261, 2003. (d) Itoh, T.; Ouchi, N.; Hayase, S.; Nishimura, Y. Chem. Lett. 2003, 32, 654. (e) Itoh, T.; Nishimura, Y.; Ouchi, N.; Hayase, S. J. Mol. Catal. B Enzym. 2003, 26, 41. (f) Itoh, T.; Ouchi, N.; Nishimura, Y.; Han, S.-H.; Katada, N.; Niwa, M.; Onaka, M. Green Chem. 2003, 5, 494. (g) Itoh, T.; Han, S.-H.; Matsushita, Y.; Hayase, S. Green Chem. 2004, 6, 437. Schöfer, S. H.; Kaftzik, N.; Wasserscheid, P.; Kragl, U. Chem. Commun. 2001, 425. (a) Kim, K.-W.; Song, B.; Choi, M.-Y.; Kim, M.-J. Org. Lett. 2001, 3, 1507. (b) Lozano, P.; De Diego, T.; Guegan, J. P.; Vaultier, M.; Iborra, J. L. Biotechnol. Bioeng. 2001, 75, 563. (c) Park, S.; Kazlauskas, R. J. J. Org. Chem. 2001, 66, 8395. (d) Lozano, P.; De Diego, T.; Carrie, D.; Vaultier, M.; Iborra, J. L. Biotech. Lett. 2001, 23, 1529. (a) Haraldsson, G. G.; Gudmundsson, B. O.; Almarsson, O. Tetrahedron Lett. 1993, 34, 5791. (b) Haraldsson, G. G.; Thorarensen, A. Tetrahedron Lett. 1994, 35, 7681. (c) Sugai, T.; Takizawa, M.; Bakke, M.; Ohtsuka, Y.; Ohta, H. Biosci. Biotech. Biochem. 1996, 60, 2059. (d) Cordova, A.; Janda, K. D. J. Org. Chem. 2001, 66, 1906. Irimescu, I.; Kato, K. Tetrahedron Lett. 2004, 45, 523. Tsuzuki, S.; Tokuda, H.; Hayamizu, K.; Watanabe, M. J. Phys. Chem. B 2005, 109, 16474. Tsukada, Y.; Iwamoto, K.; Furutani, H.; Matsushita, Y.; Abe, Y.; Matsumoto, K.; Monda, K.; Hayase, S.; Kawatsura, M.; Itoh, T. Tetrahedron Lett. 2006, 48, 1801. Kim, M.-J.; Kim, H. M.; Kim, D.; Ahn, Y.; Park, J. Green Chem. 2004, 6, 471. Uyama, H.; Takamoto, T.; Kobayashi, S. Polym. J. 2002, 34, 94. Kaar, J. L.; Jesionowski, A. M.; Berberich, J. A.; Moulton, R.; Russell, A. J. J. Am. Chem. Soc. 2003, 125, 4125. Nara, S. J.; Harjani, J. R.; Salunkhe, M. M.; Mane, A. T.; Wadgaonkar, P. P. Tetrahedron Lett. 2003, 44, 1371. Lozano, P.; De Diego, T.; Carrié, D.; Vaultier, M.; Iborra, J. L. Chem. Commun. 2002, 692. Recent other examples of stabilization of enzyme using IL as co-solvent in scCO2 solvent system: Reetz, M. T.; Wiesenhöfer, W.; Franció, G.; Leitner, W. Adv. Synth. Catal. 2003, 345, 1221. De Diego, T.; Lozano, P.; Gmouh, S.; Vaultier, M.; Iborra, J. L. Biomacromolecules 2005, 6, 1457. (a) Maruyama, T.; Nagasawa, S.; Goto, M. Biotechnol. Lett. 2002, 24, 1341. (b) Maruyama, T.; Yamamura, H.; Kotani, T.; Kamiya, N.; Goto, M. Org. Biomol. Chem. 2004, 2, 1239. Madeira, L. R.; Sorgedrager, M. J.; Carrea, G.; Van Rantwijk, F.; Secundo, F.; Sheldon, R. A. Green Chem. 2004, 6, 483. Lundell, K.; Kurki, T.; Lindroos, M.; Kanerva, L. T. Adv. Synth. Catal. 2005, 347, 1110. (a) Ganske, F.; Bornscheuer, U. T. J. Mol. Catal. B Enzym. 2005, 36, 40. (b) Ganske, F.; Bornscheuer, U. T. Org. Lett. 2005, 7, 3097. (a) Lou, W.-Y.; Zong, M.-H.; Wu, H.; Xu, R.; Wang, J.-F. Green Chem. 2005, 7, 500. (b) Xin, J.-Y.; Zhao, Y.-J.; Shi, Y.-G.; Xia, C.-G.; Li, S.-B. World J. Microbiol. Biotechnol.
20
24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
Toshiyuki Itoh 2005, 21, 193. (c) Nara, S. J.; Mohile, S. S.; Harjani, J. R.; Naik, P. U.; Salunkhe, M. M. J. Mol. Catal. B Enzym. 2004, 28, 39. (d) Noel, M.; Lozano, P.; Vaultier, M.; Iborra, J. L. Biotechnol. Lett. 2004, 26, 301. (e) Rasalkar, M. S.; Potdar, M. K.; Salunkhe, M. M. J. Mol. Catal. B Enzym. 2004, 27, 267. (f) Mohile, S. S.; Potdar, M. K.; Harjani, J. R.; Nara, S. J.; Salunkhe, M. M. J. Mol. Catal. B Enzym. 2004, 30, 185. (g) Ulbert, O.; Frater, T.; Belafi-Bako, K.; Gubicza, L. J. Mol. Catal. B Enzym. 2004, 31, 39. (h) Roberts, N. J.; Seago, A.; Carey, J. S.; Freer, R.; Preston, C.; Lye, G. J. Green Chem. 2004, 6, 475. (i) Kamal, A.; Chouhan, G. Tetrahedron Lett. 2004, 45, 8801. (j) Eckstein, M.; Wasserscheid, P.; Kragl, U. Biotechnol. Lett. 2002, 24, 763. Kim, M.-J.; Choi, M. Y.; Lee, J. K.; Ahn, Y. J. Mol. Catal. B Enzym. 2003, 26, 115. Yu, H.; Wu, J.; Bun, C. C. Chirality 2004, 17, 16. Lee, J. K.; Kim, M.-J. J. Org. Chem. 2002, 67, 6845. Itoh, T.; Matsushita, Y.; Abe, Y.; Han, S.-H.; Wada, S.; Hayase, S.; Kawatsura, M.; Takai, M.; Morimoto, M.; Hirose, Y. Chem. Eur. J. 2006, 12, 9228. Howarth, J.; James, P.; Dai, J. Tetrahedron Lett. 2001, 42, 7517. Kitazume, T.; Jiang, Z.; Kasai, K.; Mihara, Y.; Suzuki, M. J. Fluorine Chem. 2003, 121, 205. Okrasa, K.; Guibé-Jampel, E.; Therisod, M. Tetrahedron Asymm. 2003, 14, 2487. (a) Gaisberger, R. P.; Fechter, M. H.; Griengl, H. Tetrahedron Asymm. 2004, 15, 2959. (b) Matsumoto, M.; Mochizuki, K.; Fukunishi, K.; Kondo, K. Sep. Purif. Technol. 2004, 40, 97. Matsuda, T.; Yamagishi, Y.; Koguchi, S.; Iwai, N.; Kitazume, T. Tetrahedron Lett. 2006, 47, 4619. Chiappe, C.; Neri, L.; Pieraccini, D. Tetrahedron Lett. 2006, 47, 5089. (a) Swatloski, R.P.; Hobrey, J. D.; Rogers, R. D. Green Chem. 2003, 5, 361. (b) Turner, M. B.; Spear, S. K.; Huddlestone, J. G.; Holbrey, J. D.; Rogers, R. D. Green Chem. 2003, 5, 443. Jastorff, B.; Störmann, R.; Ranke, J.; Mölter, K.; Stock, F.; Oberheitmann, B.; Hoffmann, W.; Hoffmann, J.; Nüchter, M.; Ondruschke, B.; Filser, J. Green Chem. 2003, 5, 136.
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
21
Chapter 2
Temperature control of the enantioselectivity in the lipase-catalyzed resolutions Takashi Sakai Division of Chemistry and Biochemistry, Graduate School of Natural Science and Technology, Okayama University, Okayama, Japan
Abstract In the lipase-catalyzed resolution, temperature control of enantioselectivity has been generally accepted for its simplicity and theoretical reliability. Lowering the reaction temperature usually enhances the enantioselectivity. Here, the historical and theoretical backgrounds of the temperature control of enantioselectivity and its applicability to the method are described. Recent literatures for the lipase-catalyzed resolutions to which the “low-temperature method” seems to be promising to enhance the enantioselectivity are also summarized.
1. INTRODUCTION Recently, a variety of methods for increasing the enantioselectivity in lipasecatalyzed reaction1ab have been developed, involving modulation of organic media; additive, acyl donor; and immobilization support. Ionic liquid1cd and supercritical CO2 1e−g recently emerged as efficient reaction media and/or supporting material in enzymatic reactions. Beside these methods, temperature control in the lipasecatalyzed resolution of alcohols is now accepted as a theoretically reliable and gener ally applicable method, since we demonstrated that a lipase exerts its function even at very low temperatures down to −80� C in an organic solvent.2af In our first attempt for resolution of 3-phenyl-2H-azirine-2-methanol (1), the enantioselectivity was found to be maximal at −40� C in ether as reported in 1997. We recently demonstrated that temperature modulation is generally effective in the lipase-catalyzed resolution of primary and secondary alcohols.2a−h Enzymes have long been believed to be tem perature labile and have been used at around ambient temperatures; however, our finding opened the way to the wide range of temperature control of enantioselectivity in enzymatic reactions like chemical asymmetric reaction. Although unusual temperature-induced rate control of enzymatic reaction was well investigated before 1970,3 the temperature control of enantioselectivity in enzymatic reaction (PLE, −10� C) was initially reported by Jones and co-workers
22
Takashi Sakai
in 1986.4a Since then, many attempts4b−i were reported, although the reactions were mostly carried out above 0� C. The rational explanations for the temperature effect were then proposed by Overbeeke and Hult5 with lipases and by Phillips6a−f and co-workers with secondary alcohol dehydrogenase and PLE. In 1997, we reported that the temperature effect is observable with regulation below to very low tem perature, and practically applicable to the lipase-catalyzed resolution of alcohols.2a Moreover, an immobilized lipase on porous ceramic particles (Toyonite), now com mercially available as Amano lipase PS-C II, was found to be highly effective for enhancement of the reaction rate even at low temperatures.2c−h The immobilized lipase was also found to be effective for high-temperature reaction up to 120� C, in which bulky alcohol can be resolved without loss of enantioselectivity,7 maintain ing the transition-state structure.8 On the other hand, the low-temperature method can also be applicable to hydrolytic asymmetric protonation of enol acetate.8 Here, the features and current applications of the temperature modulation of enantiose lectivity are summarized. In addition, reported examples for the lipase-catalyzed resolution of primary alcohols are listed.
2. FINDING OF THE “LOW-TEMPERATURE METHOD” IN THE LIPASE-CATALYZED KINETIC RESOLUTION We first examined the lipase-catalyzed resolution of azirine-2-methanol 1,2a which we expected to have a versatile synthetic utility.2h9 As expected for primary alcohols,710 the enantioselectivity obtained in the transesterification with lipase PS in ether was low (E 11 = 17 at best) at room temperature despite considerable efforts such as screening of lipases, solvents, additives, and acylating agents.12−17 As the final choice, we examined the reaction at low temperatures because of its high reaction rate and found that lowering the temperature to −40� C increased the E value of 17 (at 30� C) to a practically acceptable level of 99; however, further lowering the temperature rather decreased the enantioselectivity (Scheme 1).
N OH
Ph (±)-1
N Ph
N
CH2OH H
(S)-1
Lipase PS Vinyl acetate, Et2O 30°C: E = 17 –40°C: E = 99
+
Ph
H CH2OAc
(R)-2
Scheme 1: “Low-temperature method” in the lipase-catalyzed resolution of
3-phenyl-2H-azirine-2-methanol (1) for enhancement of the enantioselectivity.
Temperature control of enantioselectivity
23
3. THEORY OF TEMPERATURE EFFECT ON THE ENANTIOSELECTIVITY Lowering the temperature in the lipase-catalyzed resolution usually enhances the enantioselectivity. The phenomenon does not come from the temperature-induced conformational change of lipase, but it is understandable on the basis of the theory of physical organic chemistry as explained below.56 In general, the enantioselectivity (E value11 ) in a kinetic-control reaction is determined by the ratio of the rates of two enantiomers and defined by Equation 1: E=
kA ‡ = e−G /RT kB
(1)
where kA and kB are the rate constants for a faster-reacting enantiomer (A) and a slower-reacting enantiomer (B). G‡ is the difference in the transition-state energies (G‡ = G‡fast −G‡slow for each reaction process, and R is gas constant. The lipase-catalyzed kinetic resolution is represented in the same way. Equation 1 is changed to Equation 2: G‡ = −RT ln E
(2)
Figure 1 shows the plots for the relation between G‡ and E at three different temperatures: 50, 0, and −50� C. At the lower temperature (−50� C), the curve flattens quickly, and E value of 100 requires G‡ = 2047 cal mol−1 , less than those of 2506 (at 0� C) and 2965 cal mol−1 (at 50� C), and thus a small difference in transition-state energy (G‡ ) between the enantiomers gives a large effect on the enantioselectivity. Thus, lowering the temperature increases the
ΔΔG‡ = –RT In E
–ΔΔG‡ (cal mol–1)
3500 3000
2965
2500
2506
2000
2047 cal mol–1
1500
1000 500 0
0
20
40 60 E (= kA/kB)
80
100
Figure 1: Relation between G and E value.
24
Takashi Sakai
enantioselectivity if the temperature-induced change of G‡ were small between the temperatures. From the transition-state theory the reaction rate is represented by Equa tion 3, where NA is Avogadro number and h is Plank’s constant: � � RT −G‡ k= exp (3) NA h RT Equation 3 is transformed to Equation 5 by combination with Equation 4: G‡ = H ‡ − TS ‡ H ‡ ln k = R
(4)
� � 1 RT S ‡ + ln + T NA h R
(5)
The relation between ln k and 1/T for the reaction in Scheme 1 is shown in Fig. 2, where H ‡ and S ‡ for faster (A)- and slower (B)-enantiomers are tentatively estimated to be H ‡ = 4 kcal mol−1 , S ‡ = 47 cal deg−1 mol−1 and H ‡ = 7 kcal mol−1 , S ‡ = 9 cal deg−1 mol−1 , respectively, on the basis of the observed data of H ‡ (−3 kcal mol−1 ) and S ‡ (−43 cal deg−1 mol−1 ) which are calculated below (Figs 4 and 5). Figure 2 shows that lowering the temperature (1/T ) decreases both rates (ln kA and ln kB ) by following nearly straight lines, respectively, because the contribution of the second term, lnRT/NA h, is negli gible. Therefore, lowering the temperature increases the difference in the reaction
35
Tr = 425°C
30
Ln k
25 A 20 B 15
10 0.001
0.002
0.003
0.004
0.005
0.006
1/T (K –1)
Figure 2: Relation between ln k and 1/T for faster-reacting and slower-reacting
enantiomers.
Temperature control of enantioselectivity
25
rates for both enantiomers. The temperature effect on the enantioselectivity is thus primarily governed by the difference of H ‡ in the first term of Equation 5. On the other hand, Fig. 2 shows that raising the temperature decreases the enantios electivity, giving racemic product at 425� C. The temperature is called racemic temperature (Tr ).6f Handling of the experimental data for the temperature effect and their theo retical consideration are described as follows. Temperature effect cannot be accu rately discussed by use of Equation 2 because G‡ itself is temperature dependent ‡ ‡ ‡ ‡ − Hslow and S ‡ = Sfast − Sslow : (Equation 6), where H ‡ = Hfast G‡ = H ‡ − TS ‡
(6)
Combination of Equations 2 and 6 gives the Eyring equation 7: H ‡ ln E = R
� � 1 S ‡ + T R
(7)
Figure 3 shows the plot between temperature (T ) and observed enantioselec tivity (E value) for the reaction in Scheme 1. Lowering the temperature suddenly increases the E value until −40� C. The temperature effect can also be represented by the relation between ln E and 1/T as shown in Fig. 4, where ln E is increased linearly from 30 to −40� C to reach the best E value of 99. In contrast, further lowering the temperature rather decreases the E value by following another line. Therefore, the plot consists of two lines intersecting at −40� C, and this temperature is called “inversion temperature” (Tinv ).18−20 The linear correlation suggests that the conformation of lipase is maintained in a course of the observed temperature range. At Tinv (−40� C), the transition-state structure may be changed for some rea sons: a temperature-induced structural change of enzyme and/or a solute–solvent 120 100
E
80 60 40 20 0 3.0
3.5
4.0 1/T (× 10–3, K –1)
4.5
5.0
Figure 3: Relation between observed E value and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).
26
Takashi Sakai
5
30
12
–23
°C
–40
Ln E
99 4
54
3
20
2
7.4
E
1 0 3.0
3.5
4.0 1/T (× 10–3, K –1)
4.5
5.0
Figure 4: Relation between ln E and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).
cluster change,18−20 and so on. The phenomena of Tinv are interesting and further study is required to reveal the reason. The temperature effect is theoretically broadened until 1/T = 0 as shown in Fig. 5, and the line is extrapolated to give the value of S ‡ (ln E = S ‡ /R). From the slope of the line, the value of H ‡ is calculated (H ‡ = −30 kcal mol−1 , S ‡ = −43 cal deg−1 mol−1 ). By this line, raising the tempera ture decreases the E value to reach the racemic temperature (Tr ); however, further raising the temperature begins to increase the E value again toward opposite R/S
Tr = 426
6.0
–40 –70°C
30
5.0 4.0
Ln E
(R)
3.0 2.0
(a)
1.0
(b)
0.0
E=1
–1.0
(S)
–2.0
–3.0 –4.0 0.0
1.0
2.0
3.0
4.0
5.0
1/T (× 10–3, K –1)
Figure 5: Broad range of relation between ln E and 1/T for the lipase-catalyzed
reaction of (1) (Scheme 1).
Temperature control of enantioselectivity
27
selectivity. Tr can be calculated to be 425 � C by Equation 8, which is derived from Equation 6 (G‡ = 0). Usually the observable temperature effect is enthalpydriven like this. On the other hand, in an entropy-driven reaction having a large S ‡ (line a) or a small H ‡ (line b), Tr is observable in a range of lower temperatures. If the Tr is observed at 30 � C, E value is increased by both lowering and raising the temperature from 30 � C, but to opposite R/S enantiomers. H ‡ Tr = S ‡
(8)
Example of the entropy-driven reaction is rare for enzymatic reaction. One example in which Tr appeared at around room temperature in the enantioselective oxidation of secondary alcohols with secondary alcohol dehydrogenase (SADH) from a thermophilic bacterium, Thermoanaerobactor ethanolicus is shown in Fig. 6.6f The reaction of 2-butanol gave an oxidation product together with a remaining optically active alcohol, showing a linear temperature effect log E 6f − 1/T having Tr at 26 � C. Thus, raising or lowering the temperature from 26 � C increases the R or S selectivity of the remaining alcohol, respectively. However, OH R R
Thermoanaerobactor ethanolicus
OH
R = C2H5
O R
R = C3H7 R = C4H9
S R Tr (70°C)
Tr (26°C)
1.0 R 0.5
Log E
0.0 –0.5 –1.0 S
Tr (260°C)
–1.5 –2.0 –2.5 2.8
3.0
3.2 1/T (× 10–3, K –1)
3.4
3.6
Figure 6: Enantioselective oxidation of secondary alcohols with secondary alcohol dehydrogenase (SADH), from a thermophilic bacterium (log E is used in the original paper).
28
Takashi Sakai
a similar reaction with 2-pentanol raised the Tr to 70 � C, and that with 2-hexanol further raised the Tr to 260 � C. These results suggest that the chain length changes the structure of the rate-determining step.
4. GENERAL APPLICABILITY OF THE “LOW-TEMPERATURE METHOD” EXAMINED 4.1. Application to solketal and other primary and secondary alcohols The “low-temperature method” has been applied to some primary and sec ondary alcohols (Fig. 7).2b For example, solketal, 2,2-dimethyl-1,3-dioxolane-4 methanol21 (3) had been known to show low enantioselectivity in the lipasecatalyzed resolution (lipase AK, Pseudomonas fluorescens; E = 16 at 23� C, 27 at 0� C);21a however, the E value was successfully raised up to 55 by lowering the temperature to −40� C (Table 1). Further lowering the temperature rather decreased the E value and the rate was markedly retarded. Interestingly, the loss of the enan tioselectivity below −40� C is not caused by the irreversible structural damage of lipase because the lipase once cooled below −40� C could be reused by allowing it to warm higher than −40� C, showing that the lipase does not lose conformational flexibility at such low temperatures. Similar temperature effect using other racemic alcohols such as 2-hydroxy methyl-1,4-benzodioxane (4), 2-phenylpropanol (5), and 1-cyclohexylethanol (6) was also observed as shown in Fig. 8, obeying Equation 7. These results suggest that the temperature effect is widely applicable regardless of primary or secondary alcohols and an origin of lipase. The thermodynamic parameters (H ‡ and S ‡ ) and the racemic tem peratures Tr (H ‡ /S ‡ 56 estimated for these alcohols are shown in Table 2. The results indicate that the enantioselectivity in all of these reactions is gov erned by the activation-enthalpy differences (H ‡ ), which originates from the difference in the steric and electronic interactions operating between the fasterand slower-reacting enantiomers in the transition state. The higher enantioselec tivity for secondary alcohol 6 comes from the larger negative H ‡ , and the low selectivity for 5 depends on the smaller ones. The negative values of H ‡ are compensated by the negative values of S ‡ in all cases.
OH OH
Ph 1
O
O
N
O 3
OH
O 4
OH
OH
Ph 5
6
Figure 7: Substrates for the lipase-catalyzed resolutions at low temperatures.
Temperature control of enantioselectivity
29
Table 1 Temperature effect in the lipase-catalyzed kinetic resolution of solketal
Entry Temp. (� C) Lipase (mg) Time (h) Ester (% ee) Alcohol (% ee) Conv.b
E
1 2 3 4 5 6 7 8
9 20 26 31 39 55 27 44
30 0 −10 −20 −30 −40 −50 −60
20 20 60 60 100 200c 200c 200c
3 6 8 11 16 24 48 48
63 88 84 92 93 93 74 93
69 25 77 32 39 63 97 51
052 022 048 026 030 041 057 035
MW =ca. 33 000. Lipase (ca. 1% w/w) is absorbed on Celite.
Calculated from ee(s).
c TTN is ca. 3000 at 50% conversion.
a
b
13
8
–23
–40
–51
–73 (°C)
(E ) 6
403
1
Ln E
3 4 4
55 5 6 7.4
2
0
3
4 1/T (× 10–3, K –1)
5
Figure 8: Correlation between ln E and 1/T for the lipase-catalyzed resolution of 1, 3–6.
30
Takashi Sakai
Table 2
Thermodynamic parameters for the lipase-catalyzed resolutions (i-Pr 2 O, vinyl acetate)
Compound
Lipase
1 3 4 5 5 6
PSa AKb AK PS AK PS
a b
H ‡ (kcal mol−1 )
S ‡ (cal deg−1 mol−1 )
Tr (� C)
−301 ± 013 −353 ± 020 −224 ± 012 −113 ± 005 −074 ± 005 −373 ± 025
−431 ± 050 −718 ± 079 −135 ± 040 −177 ± 096 −151 ± 019 −302 ± 092
425 219 1386 365 217 962
With vinyl butyrate. In Et2 O.
4.2. Resolution of (±)-2-hydroxy-2-(pentafluorophenyl)acetonitrile The “low-temperature method” was then applied to the resolution of (±) 2-hydroxy-2-(pentafluorophenyl)acetonitrile (7) (Fig. 9),2c which is usable for the syntheses of a variety of ethane diols, amino alcohols containing C6 F5 groups as novel chiral ligands.22 After screening lipases such as Amano PS and AK, lipase LIP (Pseudomonas aeruginosa lipase immobilized on Hyflo Super-Cel, Toyobo,
F
CN
F
F OAc
OH
F F F
Lipase
F
AcOCH=CH2 i-Pr2O
F
(±)-7
F
CN
CN
F
F
OH
F F
F
F
(S)-8
(R)-7
7 (E = 1097)
Ln E
6 (403) 5 (148) 4 (55) 3 (20)
In E = –8.35 + 3.94 × 10–3 (1/T ) 3.2 (40)
3.4 (21)
3.6 (5)
3.8 (–10°C)
1/T ( × 10–3, K–1)
Figure 9: Correlation between ln E and 1/T in the lipase LIP-catalyzed transesterification of (±)-7.
Temperature control of enantioselectivity
31
Co., Ltd, Japan) was found to be the best choice (E = 113, 1.5 h, 44% conv. at 30� C). The E value was increased regularly by lowering the temperature to 0� C to reach > 427 (4.0 h, 44% conv.), which is a satisfactory and practical level. 4.3. Immobilization of lipase on porous ceramic support (Toyonite) for acceleration2d As described above, the temperature effect is useful for enhancing the enantiose lectivity; however, one problem is the decrease in the reaction rate. For example, although in a lipase AK-catalyzed resolution of solketal, the E value (9 at 30� C, Table 1, entry 1) is increased up to 55 by lowering the temperature to −40� C, 10 times the amount of lipase and 8-fold the reaction time are required as com pared with those at 30� C.2b Thus, the rate of acceleration is an important subject especially to make the low-temperature reaction practical. Acceleration of the reaction rate was attained by using an immobilized lipase on the porous ceramic support (Toyonite).23 The immobilized lipase PS is commercially available as lipase PS-C “Amano” II, which has (methacryloy loxy)propylsilanetrioxyl bridges on the ceramic surface, and lipases are immobi lized on the bridges.23 Immobilization of an enzyme is known to affect the enzyme conformation, rigidity, and aggregation state to alter the enantioselectivity and reactivity.24 As shown in Table 3, the use of the Toyonite-immobilized lipase AK accelerates the reaction from 11 000 (TTN/h, total turnover number per hour) for Celite-immobilized one to 53 000 at 30� C. Observed maximal acceleration is 80 times at −40� C in the case of entry 4 compared with that of entry 3. The great acceleration ability is significant for practical use,25−27 especially at low tempera tures. In an organic solvent, lipase molecules usually form aggregation structures, even on Celite, which reduce the activity, while on Toyonite lipases may be highly dispersed by immobilization with organic bridges on the porous ceramic to exert the inherent high activity.
Table 3
Toyonite-immobilized lipase-catalyzed resolution of solketal (±)-3 (vinyl butylate)
Entry
1 2 3 4
Lipase
AK AK PS PS
Organic bridge
Celite Toyonite Celite Toyonite
E
TTN/h
30� C
−40� C
30� C
−40� C
90 32 68 68
55 21 15 15
11 000 53 000 6400 110 000
110 1400 22 1700
32
Takashi Sakai
4.4. Structural optimization of organic bridges on Toyonite2d Structure of organic bridges attached on the support is also crucial for the conformational engineering of enzymes. The ability of organic bridges in the low-temperature reaction was then examined by using the substrate of 2-hydroxymethyl-1,4-benzodioxane 4 as shown in Table 4 and Fig. 10.2d Figure 10 shows that choice of organic bridges is crucially important both to accelerate the reaction and to enhance the enantioselectivity. As far as examined for 9a–i, termi nal olefin is not necessary, and carbonyl function is requisite. A small structural change of the bridges remarkably affects the temperature effect probably by the difference in a manner of hydrogen bonding between the bridge and functions of Table 4 Effect of organic bridges on the Toyonite-immobilized lipase-catalyzed resolution of (±)-4
Entry
1 2 3 4 5 6 7 8 9 10 11 a
Organic bridge
Nonea Noneb 9a 9b 9c 9d 9e 9f 9g 9h 9ic
E
TTN/h
30� C
−30� C
30� C
−30� C
18 19 8.6 10 6.6 7.9 4.8 11 6.8 5.9 –
33 31 18 17 9.3 .28 21 28 12 6.8 –
3100 5300 7300 14 000 14 000 8700 27 000 12 000 57 000 7200 –
26 52 130 280 230 280 840 100 940 61 –
Commercially available lipase PS immobilized on Celite.
Toyonite without organic bridges.
c Lipase could not be immobilized.
b
Temperature control of enantioselectivity 30
0
–20
3.5
–30
°C
9f 9d 9e 9a 9b
3.0
Ln E
33
2.5
9g 9c
2.0
9h
1.5 3.5
4.0 1/T (× 10–3, K–1)
Figure 10: Temperature effect on the Toyonite-immobilized lipase-catalyzed resolution of (±)-4 by varying the organic bridges.
lipase molecules. The bridge 9d was the best choice from the points of reaction efficiency and availability. 4.5. Practical resolution of azirine 1 by the “low-temperature method” combined with Toyonite-immobilized lipase and optimized acylating agent2e In the first attempt of the low-temperature method for azirine-2-methanol 1 using the Celite-immobilized lipase (Scheme 1), the E value was increased from 17 (30� C) up to 99 (−40� C); however, the TTN/h was decreased from 4700 to 210. On the other hand, use of Toyonite D-M-immobilized lipase28 in the reaction remarkably raised the TTN/h to 4200 at −40� C. In contrast, it lowered the E value to 33. Therefore, adjustment of the two conflicting features, the reaction rate and the enantioselectivity, is essential for the practical use of the low-temperature method. For this purpose, the acylating agents29 were screened and found to be highly effective on the E values. The results of the reactions carried out at −40� C are shown in Table 5.2e Elongation of the acyl chain to vinyl butanoate dramatically increased the E value up to 96, a comparable value in the Celite-immobilized lipase, together with a sufficient TTN/h of 3200. Further improvement was attained by using Toyonite 200M-immobilized lipase, which gave the highest E value (130) and TTN/h (7800) with vinyl butanoate at −40� C. These results indicate that the low-temperature method increases the enantio selectivity, at least above inversion temperature, and the enantioselectivity and reaction rate can be optimized by the use of Toyonite-immobilized lipase and a suitable acylating agent.
34
Takashi Sakai
Table 5 Toyonite-immoblized lipase-catalyzed resolution of ±-1 at −40� C
Entry
Acylating agents
Lipasea
E
TTN/h
A B B B B B B B B B C C D
99 34 96 54 102 35 40 44 71 7 130 77 68
210 4200 3200 1600 2000 1500 1500 1800 3600 5300 7800 8600 1500
R 1 2 3 4 5 6 7 8 9 10 11 12 13
CH3 CH3 CH3 CH2 2 CH3 CH2 4 CH3 CH2 4 CH3 CH2 6 CH3 CH2 8 CH3 CH2 10 PhCH2 CH2 CH2 Cl CH3 CH2 2 CH3 CH2 4 CH3 CH2 2
3a (1)b 3a (1) 3b (1) 3c (1) 3c (2) 3d (2) 3e (2) 3f (2) 3g (2) 3h (1) 3b (1) 3c (2) 3b (2)
a
A: Lipase PS on Celite; B: lipase PS immobilized on Toyonite D-M with 3 (methacryloyloxy)propylsilanetrioxy bridges; C: lipase PS immobilized on Toyonite 200M with 3-(methacryloyloxy)propylsilanetrioxy bridges; D: lipase PS immobilized on Toyonite 200 with 3-(propanoyloxy)hexylsilanetrioxy bridges. b Equivalent of acylation agent.
4.6. Resolution of (2R∗ 3S∗ )- and (2R∗ 3R∗ )-3-methyl-3-phenyl-2-aziridinemethanols2h The lipase-catalyzed resolution of (2R∗ ,3S ∗ )-3-methyl-3-phenyl-2-aziridine methanol (±)-11 by using the “low-temperature method” gave synthetically useful (2R,3S)-11 and its acetate (2S,3R)-11a with (2S)-selectivity (E = 55 at −40� C), while a similar reaction of (2R∗ ,3R∗ )-3-methyl-3-phenyl-2-aziridinemethanol (±) 12 gave (2S,3S)-12 and its acetate (2R,3R)-12a with (2R)-selectivity (E = 73 at −20� C) (Scheme 2). Compound (±)-11 was prepared conveniently via diastereos elective addition of MeMgBr to t-butyl 3-phenyl-2H-azirine-2-carboxylate, which was successfully prepared by the Neber reaction of oxime tosylate of t-butyl
Temperature control of enantioselectivity
1) CH3MgBr 2) LiAlH4
N Ph
OH (2S)selective (±)-11 (2R*,3S*)
(±)-10a
CO2t-Bu
CH3
Lipase PS-C ll
H
CH3
CO2t-Bu
N
H N
Ph
1) PhMgBr 2) LiAlH4
H N
CH3
OH
Ph
(±)-10b
Lipase PS-C ll
H
(2S)selective
(±)-12 (2R*,3S*)
35
H N
CH3
S
Ph
OAc + H
(2S,3R)-11a
H N
CH3 Ph
H
Ph CH3
H N
H
R
OH
(2R,3S)-11
OAc +
R
Ph CH3
H N OH S
H
(2S,3S)-12
(2R,3R)-12a
Scheme 2: Resolution of (2R∗ ,3S ∗ )- and (2R ,3R∗ )-3-methyl-3-phenyl-2-aziridinemethanols. ∗
benzoylacetate. As far as we know, a few examples of the lipase-catalyzed reac tion for such 2-aziridinemethanols having two stereogenic centers at - and carbons are known,30 and none of the reaction with aziridine derivatives without N -protection have been reported. As shown in Fig. 11, the reaction of 11 in ace tone gave higher enantioselectivity, and the best result (E = 55) was obtained at −40� C. However, further lowering the temperature to −50� C did not increase the E value. In contrast, the temperature modulation in THF increased the E value continuously to −60� C. The temperature effect for compound 12 in acetone is shown in Fig. 12. The inversion temperatures (Tinv ) for 11 and 12 are different in acetone.
25
0
–20 –30
–40
–50
–60 °C
E
Ln E
4
50 30
3
20 10
2
3.7
4.1 1/T (× 10–3, K–1)
4.5
Figure 11: Temperature effect in the lipase-catalyzed resolution of ±-11 (�: in acetone; •: in THF; �: in Et2 O).
36
Takashi Sakai 25
0
–20
–40
–50
°C
Ln E
4.5
E 73 67 61 54
4.0
3.5
30
3.6
4.0 1/T (× 10–3, K–1)
4.4
Figure 12: Temperature effect in the lipase-catalyzed resolution of (±)-12 in acetone.
4.7. Resolution of 5-(hydroxymethyl)-3-phenyl-2-isoxazoline2g Optically active 5-(hydroxymethyl)-3-phenyl-2-isoxazoline 13 is a versatile key intermediate for the syntheses of -hydroxy ketones,31 -amino alcohols,32 and -amino acids.33 However, the lipase-catalyzed kinetic resolution of isoxazoline (±)-13 has not been reported so far probably because of the low enantioselectivity expected for primary alcohols (Scheme 3). The enantioselectivity was found to be very low (E value = 4–5 in i-Pr 2 O) at room temperature; however, it could be markedly improved up to an E value of 249 at −60� C by using lipase PS-C II in acetone, which was the best solvent among those tested (THF, i-Pr 2 O) (Fig. 13). The temperature effect involving Tinv was markedly influenced by the solvent, acylating agent, and support. Tinv was not observed in the case with vinyl acetate below to −60� C. Lipase PS-C ll Organic solvent
N O OH
Ph
OCOR
N O Ph
(±)-13 30°C: E = 5 (i-Pr2O) –60°C: E = 249 (acetone)
H
N O O
R
+ Ph
H
OH
O (S)-13 (R)-14a: R = CH3 (R)-14b: R = (CH2)2CH3 (R)-14c: R = (CH2)2Ph
Scheme 3: Resolution of (±)-5-(hydroxymethyl)-3-phenyl-2-isoxazoline
((±)-13).
Temperature control of enantioselectivity
Ln E
0
–20
–40
–60
°C E
6.0
37
5.0
148
4.0
54
3.0
20
2.0
7.4
1.0 3.5
4.0
4.5 1/T (× 10–3, K–1)
5.0
Figure 13: Temperature effect in lipase PS-C II-catalyzed resolution of (±)-13 with vinyl acetate (•), vinyl butyrate (�), and vinyl 3-phenylpropanoate (�) in acetone. 4.8. Application of temperature control to asymmetric protonation8 The low-temperature method is effective not only in the kinetic resolution of alcohols but also in the enantioface-selective asymmetric protonation of enol acetate of 2-methylcyclohexanone (15) giving (R)-2-methylcyclohexanone (16).8 The reaction in H2 O at 30� C gave 28% ee (98% conv.), which was improved up to 77% ee (82% conv.) by the reaction using lipase PS-C II in i-Pr 2 O and ethanol at 0� C. Acceleration of the reaction with lipase PS-C II made this reaction possible because this reaction required a long reaction time. The temperature effect is shown in Fig. 14. The regular temperature effect was not observed. The protons may be supplied from H2 O, methanol, or ethanol, whose bulkiness is important. 4.9. Lipase-catalyzed resolutions at high temperatures up to 120� C Lipase Amano PS-C II was also found to be useful for a high-temperature reaction in the resolution of a bulky substrate, 1,1-diphenyl-2-propanol, which showed no reactivity under usual conditions with lipase PS.7 An enantiopure product was obtained at 40–120� C, and the highest conversion (39%) was obtained at 80–90� C. It is very interesting that a single enzyme is usable in the reaction at a very wide range of temperatures from −80� C to 120� C. 5. LOW-TEMPERATURE REACTIONS IN LITERATURES Recently, temperature control of enantioselectivity is actively studied,34−62 and many examples for the low-temperature reactions have been published. The typical examples for the synthetic purpose are summarized in Fig. 15. The E values of these are generally increased by lowering the temperature. Compound 29 is used
38
Takashi Sakai O
O
Lipase PS-C ll
O
ROH (10 equiv.), i-Pr2O R = H, Me, Et (R)-16
15 60
Ln E
2.5
0 –10 °C
30
%ee
2.0
76
1.5
64
1.0
46
0.5
24
0.0 2.5
3.5 3 1/T (× 10–3, K –1)
EtoH MeOH
4
Figure 14: Lipase-catalyzed enantioface-selective asymmetric protonation. in a media of supercritical CO2 . Compounds 36 are used for the kinetic resolution in the hydrolysis of nitrile and amide, and compound 37 is for the reaction in the hydrolytic ring opening of epoxide. Interestingly, some examples show rather negative temperature effect as shown in Fig. 16. The latter examples may be due to the entropy-driven reactions,6 and are also interesting. F BnO OH
H
H3C C
OTBS
H3C 1835
RT: E = 106 –5°C: E = 307
N 30°C: E = 88 CH3 15°C: E = 127
OH
2138
NH2 39
(In E–1/T graph)
22 40°C: E = 13 –20°C: E = 84
Ph
O
ONO2
1936
2037
40°C: E = 2 4°C: E = 100
22°C: E = 57 –18°C: E = 96
CO2Me O
OH CH3
H
OH
1734
OH
C C
OH O
OCOC3H7
N3 OH 40
23 70°C: E = 18 4°C: E = 72
2441 15°C: E = 18 –13°C: E = 26
Figure 15: Examples of the “low-temperature method.”
Temperature control of enantioselectivity
OH
OH
39
CH3 N O
OH
OH
OH
CO16H33O
N
2542
2643
2744
30°C: E = 40 0°C: E > 750
22°C: E = 17 4°C: E = 32
30°C: E = 18 10°C: E = 31
2845 40°C: E = 6.8 –15°C: E = 33.6
O
OH
Cl 2946 (In E–1/T graph in supercritical CO2)
H3C
OH
HO
OH
F3C
H3C Si H3C Ph
OH 3047
3148
RT: 35% ee 0°C: 49% ee
RT: 93% ee 4°C: >99% ee
OH O
CO2i-Bu
H3C
CO2H O
NH2
3350 25°C: E = 49 0°C: E = 66
3249
25°C: E = 118 0°C: E = 240
H3C
H3C
OH O
C2H5
3451 25°C: E = 14 0°C: E = 78
with LiCl
50°C: E = 12
10°C: E = 1300
3552
O
3653
X
X = CN, CONH2 (Rhodococcus sp. AJ270)
Cl
3754
27°C: E < 40
0°C: E < 100
(A. niger )
30°C: E = 23 20°C: E = 32
Figure 15: Continued.
OAc
CH2OH NHCO2tBu
H N
Bn
Ph O O
3855 60°C: E = 26 22°C: E = 2
3956 RT: E = 127 0°C: E = 13
4057 55°C: E = 130 28°C: E = 24
Ph
O O 4158 30°C: E = 107 4°C: E = 5.8
Figure 16: Temperature effect where lowering the temperature decreases the
enantioselectivity.
40
Takashi Sakai
6. LIPASE-CATALYZED RESOLUTION OF PRIMARY ALCOHOLS: PROMISING CANDIDATES FOR THE “LOW-TEMPERATURE METHOD” Examples of the lipase-catalyzed resolution of primary alcohols are listed in Fig. 17.63−126 They usually give low enantioselectivity because of mechanistic reasons, and no effective method for improving the enantioselectivity is avail able. One of the purposes of this book is to create new ideas and possibilities in this field. The “low-temperature method” is a promising one to improve the enantioselectivity of these alcohols.
Figure 17: Example of lipase-catalyzed resolution of primary alcohols [lit.].
Temperature control of enantioselectivity
Figure 17: Continued.
41
42
Takashi Sakai
Figure 17: Continued.
Temperature control of enantioselectivity
Figure 17: Continued.
43
44
Takashi Sakai
Figure 17: Continued.
Temperature control of enantioselectivity
45
Figure 17: Continued.
7. CONCLUSION Here, the temperature effect is discussed from the aspect of synthetic utility. Temperature control of enantioselectivity, i.e., the “low-temperature method”, is simple and now practically acceptable method. The phenomenon is based on the theory of physical organic chemistry and will be studied further for understanding the enzymatic reaction as organic reaction.
REFERENCES 1. (a) Drauz, K.; Waldmann, H. (Eds) Enzyme Catalysis in Organic Synthesis; Wiley-VCH: Weinheim, 2002. (b) Borncheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis; Wiley-VCH: Weinheim, 2000 (1st edn) and 2006 (2nd edn). (c) Itoh, T.; Han, S.; Matsushita, Y.; Hayase, S. Green Chem. 2004, 6, 437–439. (d) Tsukada, Y.; Iwamoto, K.; Furutani, H.; Matsushita, Y.; Abe, Y.; Matsumoto, K.; Monda, K.; Hayase, S.; Kawatsura, M.; Itoh, T. Tetrahedron Lett. 2006, 47, 1801–1804 and the literatures cited therein. (e) Matsuda, T.; Harada, T.; Nakamura, K. Green Chem. 2004, 6, 440–444. (f) Matsuda, T.; Harada, T.; Nakamura, K. Curr. Org. Chem. 2005, 9, 299–315. (g) Matsuda, T.; Harada, T.; Nakamura, K.; Ikariya, T. Tetrahedron Asymm. 2005, 16, 909–915. Effect of temperature on hydrolytic enzyme in supercritical CO2 is also discussed. 2. (a) Sakai, T.; Kawabata, I.; Kishimoto, T.; Ema, T.; Utaka, M. J. Org. Chem. 1997, 62, 4906– 4907. (b) Sakai, T.; Kishimoto, I.; Ema, T.; Utaka, M. Tetrahedron Lett. 1998, 39, 7881–7884. (c) Sakai, T.; Miki, Y.; Tsuboi, M.; Takeuchi, H.; Ema, T.; Uneyama, K.; Utaka, M. J. Org. Chem. 2000, 65, 2740–2747. (d) Sakai, T.; Hayashi, K.; Yano, F.; Takami, M.; Ino, M.; Korenaga, T.; Ema, T. Bull. Chem. Soc. Jpn 2003, 76, 1441–1446. (e) Sakai, T.; Matsuda, A.; Korenaga, T.; Ema, T. Bull. Chem. Soc. Jpn 2003, 76, 1819–1821. (f) Mini Review: Sakai, T. Tetrahedron Asymm. 2004, 15, 2749–2756. (g) Sakai, T.; Mitsutomi, H.; Korenaga, T.; Ema, T. Tetrahedron Asymm. 2005, 16, 1535–1539. (h) Sakai, T.; Liu, Y.; Ohta, H.; Korenaga, T.; Ema, T. J. Org. Chem. 2005, 70, 1369–1375. 3. Talsky, G. Angew. Chem. Int. Ed. 1971, 10, 548. 4. (a) Lam, L. K. P.; Hui, R. A. H.; Jones, J. B. J. Org. Chem. 1986, 51, 2047–2050. (b) Keinan, E.; Hafeli, E. K.; Seth, K. K.; Lamed, R. J. Am. Chem. Soc. 1986, 108, 162–169. (c) Willaert, J. J.; Lemière, G. L.; Joris, L. A.; Lepoivre, J. A.; Alderweireldt, F. C. Bioorg. Chem. 1988, 16,
46
Takashi Sakai
5. 6.
7.
8.
9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21.
22.
223–231. (d) Parmar, V. S.; Prasad, A. K.; Singh, P. K.; Gupta, S. Tetrahedron Asymm. 1992, 3, 1395–1398. (e) Secundo, F.; Riva, S. R.; Carrea, G. Tetrahedron Asymm. 1992, 3, 267–280. (f) Yasufuku, Y.; Ueji, S. Biotechnol. Lett. 1995, 17, 1311–1316. (g) Kasche, V.; Galunsky, B.; Nurk, E.; Piotraschke, E.; Rieks, A. Biotech. Lett. 1996, 18, 455–460. (h) Galunsky, B.; Ignatova, S.; Kasche, V. Biochem. Biophys. Acta 1997, 1343, 130–138. (i) Miyazawa, T.; Minowa, H.; Miyamoto, T.; Imagawa, K.; Yanagihara, R.; Yamada, T. Tetrahedron Asymm. 1997, 8, 367–370. Holmberg, E.; Hult, K. Biotech. Lett. 1991, 13, 323–326. (a) Pham, V. T.; Phillips, R. S.; Ljungdahl, L. G. J. Am. Chem. Soc. 1989, 111, 1935–1936. (b) Pham, V. T.; Philips, R. S. J. Am. Chem. Soc. 1990, 112, 3629–3632. (c) Pig liver esterase: Andrade, M. A. C.; Andrade, F. A. C.; Phillip, R. S. Bioorg. Med. Chem. Lett. 1991, 1, 373– 376. (d) Phillip, R. S. Enzyme Microb. Technol. 1992, 14, 417–419. (e) Zheng, C.; Pham, V. T.; Phillips, R. S. Catal. Today 1994,22, 607–620. (f) Phillips, R. S. TIBTECH 1996, 14, 13–16. Lipase PS-C II-catalyzed resolution of a bulky substrate at high temperatures up to 120� C in organic solvent: (a) Ema, T.; Kageyama, M.; Korenaga, T.; Sakai, T. Tetrahedron Asymm. 2003, 14, 3943–3947. (b) Ema, T. Tetrahedron Asymm. 2004, 15, 2765–2770. (a) Review: Ema, T. Curr. Org. Chem. 2004, 8, 1009-1-25. (b) Rational design of lipase by site-directed mutagenesis on the basis of stereo-sensing mechanism for secondary alcohol: Ema, T.; Fujii, T.; Ozaki, M.; Korenaga, T.; Sakai, T. Chem. Commun. 2005, 4650–4651. Lipase PS-C II-catalyzed asymmetric protonation: Sakai, T.; Matsuda, A.; Korenaga, T.; Ema, T. Tetrahedron Asymm. 2004, 15, 1929–1932. Review: Padwa, A.; Woolhouse, A. D. In Comprehensive Heterocyclic Chemistry; Lowowski, W. (Ed.); Pergamon Press: Oxford; Vol. 7, Chapter 5, 1984, p. 47. Recent papers cited in Ref. 2a. Weissfloch, A. N. E.; Kazlauskas, R. J. J. Org. Chem. 1995, 60, 6959–6969. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc. 1982,104, 7294–7299. Carrea, G.; Riva, S. Angew. Chem. Int. Ed. 2000, 39, 2226–2254. Overbeeke, P. L. A.; Jongejan, J. A.; Heijnen, J. J. Biotechnol. Bioeng. 2000, 70, 278–290. Theil, F. Tetrahedron 2000, 56, 2905–2919. Loughlin, W. A. Biores. Tech. 2000, 49–62. Sugai, T. Curr. Org. Chem. 1999, 3, 393–406. Cainelli, G.; de Matteis, V.; Galletti, P.; Giacomini, D. Chem. Commun. 2000, 2351–2352. Cainelli, G.; Galletti, P.; Giacomini, D.; Gualandi, A.; Quintavalla, A. Helv. Chim. Acta 2003, 86, 3548–3559. Variable-temperature 13 C NMR experiments show that the solute–solvent clustering effect manages the chemical and enzymatic enantioselectivity. Cainelli, G.; Galletti, P.; Pieraccini, S.; Quintavalla, A.; Giacomini, D.; Spada, G. P. Chirality 2004, 16, 50–56. (a) Vanttinen, E.; Kanerva, L. T. Tetrahedron Asymm. 1997, 6, 923–933. (b) Krief, A.; Provins, L.; Froidbise, A. Tetrahedron Lett. 1998, 39, 1437–1440. (c) Bianchi, D.; Bosetti, A.; Golini, P.; Cesti, P.; Pina, C. Tetrahedron Asymm. 1997, 8, 817–819. (d) Chenault, H. K.; Chafin, L. F.; Liehr, S. J. Org. Chem. 1998, 63, 4039–4045. Synthesis of optically pure 2-amino-1-alcohol,18a–c 1,2-diol,2c and 1,2-diamine2d involving pentafluorophenyl groups: (a) Sakai, T.; Kubo, K.; Kashino, S.; Uneyama, K. Tetrahedron Asymm. 1996, 7, 1883–1886. (b) Sakai, T.; Takayama, T.; Ohkawa, T.; Yoshio, O.; Ema, T.; Utaka, M. Tetrahedron Lett. 1997, 38, 1987–1990. (c) Sakai, T.; Miki, Y.; Nakatani, M.; Ema, T.; Uneyama, K.; Utaka, M. Tetrahedron Lett. 1998, 39, 5233–5236. (d) Korenaga, T.; Tanaka, H.; Ema, T.; Sakai, T. J. Fluorine Chem. 2003, 122, 201–205. (e) Sakai, T.; Korenaga, T.; Washio, N.; Nishio, Y.; Minami, S.; Ema, T. Bull. Chem. Soc. Jpn 2004, 77, 1001–1008. See also Ref. 2c.
Temperature control of enantioselectivity
F
F
F
F
F
F F
FF
F
F
F H2N
F F
F
F
F
F F
FF F
F HO OH (S,S)
F
F
F
F F
F
H2N
OH
OH
(1R,2S)
(1R,2S) chiral 2-amino alcohols F
F
F
F
FF F H 2N
OH
(1S,2R)
F
F
47
F
F F
FF F
F HO OH (R,R)
chiral 1,2-diols
23. Toyonite type 200 without organic bridges on the surface: a new type of spherical porous ceram ics support having 155 ± 5 m average diameter and 60 nm average sized pores. Yamashita, Y.; Kamori, M.; Takenaka, H.; Takahashi, J. Jpn. Kokai Tokkyo Koho, JP 09313179, 1997; US Patent 6004786. Commercially available Toyonite 200M is Toyonite 200 with 3-(2 methylpropenoyloxy)propylsilanetrioxyl groups attached on it. Lipase PS immobilized on Toy onite 200M is also available from Amano Pharmaceutical Co., Ltd, or Wako Pure Chemical Industries, Ltd, as Lipase PS-C “Amano” II. See also Ref.2c. 24. For current examples: (a) Palomo, J. M.; Fernández-Lorente, G.; Mateo, C.; Fuentes, M.; Fernández-Lorente, R.; Guisan, J. M. Tetrahedron Asymm. 2002, 13, 1337–1345. (b) Terreni, M.; Mateo, C.; Bastida, A.; Dalmases, P.; Hugust, L.; Guisan, J. M. Enzyme Microb. Technol. 2001, 28, 389. 25. Kamori, M.; Hori, T.; Yamashita, Y.; Hirose, Y.; Naoshima, Y. J. Mol. Catal. B Enzym. 2000, 9, 269–274. 26. Kato, K.; Gong, Y.; Saito, T.; Kimoto, H. J. Biosci. Bioeng. 2000, 90, 332. 27. Suzuki, M.; Nagasawa, C.; Sugai, T. Tetrahedron 2001, 57, 4841. 28. Toyonite D-M (porous ceramic support, Toyo Denka Kogyo Co. Ltd.) has 120 nm of mean pore size having 3-(2-methylpropenoyloxy)propylsilanetrioxyl bridges on the surface. 29. Recent reports for the enzymatic resolution with various acylating reagents: (a) Kawasaki, M.; Goto, M.; Kawabata,S.; Kodama, T.; Kometani, T. Tetrahedron Lett. 1999, 40, 5223–5226. (b) Hirose, K.; Naka, H.; Yano, M.; Ohashi, S.; Naemura, K.; Tobe, Y. Tetrahedron Asymm. 2000, 11, 1199–1215. (c) Kita, Y.; Takebe, Y.; Murata, K.; Naka, T.; Akai, S. J. Org. Chem. 2000, 65, 83–88. (d) Ottosson, J.; Hult, K. J. Mol. Catal. B Enzym. 2001, 11, 1025. (e) Reaction in ionic solvent: Itoh, T.; Akasaki, E.; Nishimura, Y. Chem. Lett. 2002, 154–155. (f) Akai, S.; Naka, T.; Fujita, T.; Takebe, Y.; Tsujino, T.; Kita, Y. J. Org. Chem. 2002, 67, 411–419. 30. (a) Fuji, K.; Kawabata, T.; Kiryu, Y.; Sugiura, Y. Tetrahedron Lett. 1990, 31, 6663–6666. (b) Davoli, P.; Prati, F. Heterocycles 2000, 53, 2379–2389. (c) Davoli, P.; Caselli, E.; Bucciarelli, M.; Forni, A.; Torre, G.; Prati, F. J. Chem. Soc. Perkin Trans. I 2002, 1948–1953. 31. Lee, J. Y.; Kim, B. H. Tetrahedron Lett. 1992, 33, 2557–2560. 32. Jaëger, V.; Buss, V.; Schwab, W. Tetrahedron Lett. 1978, 19, 3133–3136. 33. Minter, A. R.; Fuller, A. A.; Mapp, A. K. J. Am. Chem. Soc. 2003, 125, 6846–6847. 34. Aoyagi, Y.; Saitoh, Y.; Ueno, T.; Horiguchi, M.; Takeya, K. J. Org. Chem. 2003, 68, 6899–6904.
48
Takashi Sakai
35. de Gonzalo, G.; Brieva, R.; Sánchez, V. M.; Bayod, M.; Gotor, V. J. Org. Chem. 2001, 66, 8947–8953. 36. Cipiciani, A.; Bellezza, F. J. Mol. Catal. B Enzym. 2002, 17, 261–266. 37. Pchelka, B. K.; Loupy, A.; Plenkiewicz, J.; Petit, A.; Blanco, L. Tetrahedron Asymm. 2001, 12, 2109–2119. 38. Miyazawa, T.; Yukawa, T.; Koshiba, T.; Sakamoto, H.; Ueji, S.; Yanagihara, R.; Yamada, T. Tetrahedron Asymm. 2001, 12, 1595–1602. 39. López-Serrano, P.; Wegman, M. A.; van Rantwijk, F.; Sheladon, R. A. Tetrahedron Asymm. 2001, 12, 235–240. 40. Pchelka, B. K.; Loupy, A.; Plenkiewicz, J.; Blanco, L. Tetrahedron Asymm. 2000, 11, 2719–2732. 41. Secundo, F.; Oppizzi, M. L.; Carrea, G.; De Amici, M.; Dallanoce, C. Biocatal. Biotrans. 1999, 17, 241–250. 42. Baczko, K.; Larpent, C. J. Chem. Soc. Perkin Trans. II 2000, 521–526. 43. Haraldsson, G. G.; Thordarson, P.; Halldorsson, A.; Kristinsson, B. Tetrahedron Asymm. 1999, 10, 3671–3674. 44. Kawasaki, M.; Goto, M.; Kawabata, S.; Kodama, T.; Kometani, T. Tetrahedron Lett. 1999, 40, 5223–5226. 45. Avdagi´ c, A.; Lazac, A.; Majer, Z.; Hollòsi, M.; Šunji´c, V. Helv. Chem. Acta 1998, 81, 1567–1583. 46. Matsuda, T.; Kanamaru, R.; Watanabe, K.; Kamitanaka, T.; Harada, T.; Nakamura, K. Tetra hedron Asymm. 2003, 14, 2087–2091. 47. Arai, N.; Chikaraishi, N.; Ikawa, M.; Omura, S.; Kuwajima, I. Tetrahedron Asymm. 2004, 15, 733–741. 48. Ghorpade, S. R.; Kharul, R. K.; Joshi, R. R.; Kalkote, U. R.; Ravindranathan, T. Tetrahedron Asymm. 2004, 15, 733–741. 49. Kawamnami, Y.; Itoh, K. Chem. Lett. 2005, 34, 682–683. 50. Miyazawa, T.; Kaito, E.; Yukawa, T.; Murashima, T.; Yamada, T. J. Mol. Catal. B Enzym. 2005, 37, 63–67. 51. Miyazawa, T.; Imagawa, K.; Minowa, H.; Miyamoto, T.; Yamada, T. Tetrahedron 2005, 61, 10254–10261. 52. Okamoto, T.; Yasuhito, E.; Ueji, S. Org. Biomol. Chem. 2006, 4, 1147–1153. 53. Wang, M.-X.; Feng, G.-Q. J. Mol. Catal. B Enzym. 2002, 18, 267–272. 54. Manoj, K. M.; Archelas, A.; Baratti, J.; Furstoss, R. Tetrahedron 2001, 57, 695–701. 55. Cipiciani, A.; Bellezza, F.; Fringuelli, F.; Silvestrini, M. G. Tetrahedron Asymm. 2001, 12, 2277–2281. 56. Cainelli, G.; Marreis, V. D.; Galletti, P.; Orioli, P. Chem. Commun. 2000, 2351–2352. 57. Kaman, J.; Van der Eycken, J.; Peter, A.; Fulop, F. Tetrahedron Asymm. 2001, 12, 625–631. 58. Goto, M.; Kawasaki, M.; Kometani, T. J. Mol. Catal. B Enzym. 2000, 9, 245–250. 59. Galunsky, B.; Ignatova, S.; Kasche, V. Biochim. Biophys. Acta 1997, 1434, 130–138. 60. Overbeeke, A.; Ottosson, J.; Hult, K.; Jongejan, J. A.; Duine, J. A. Biocatal. Biotrans. 1999, 17, 61–79. 61. Ljubovi´ c, E.; Elenkov, M.; Avdagic, A.; Šunji´c, V. Food Technol. Biotechnol. 1999, 37, 215–224. 62. Persson, M.; Costes, D.; Wehtje, E.; Adlercreutz, P. Enzyme Microb. Technol. 2002, 30, 916–923. 63. Fernández, S.; Brieva, R.; Rebolledo, F.; Gotor, V. J. Chem. Soc. Perkin Trans. I 1992, 2885–2890. 64. Ferraboschi, P.; Grisenti, P.; Manzocchi, A.; Santaniello, E. J. Org. Chem. 1990, 55, 6214–6216. 65. Cambou, B.; Klibanov, A. M. J. Am. Chem. Soc. 1984, 106, 2687–2692.
Temperature control of enantioselectivity
49
66. Francalanci, F.; Cesti, P.; Cabri, W.; Bianchi, D.; Martinengo, T.; Foà, M. J. Org. Chem. 1987, 52, 5079–5082. 67. Terao, Y.; Tsuji, K.; Murata, M.; Achiwa, K.; Nishio, T.; Watanabe, N.; Seto, K. Chem. Pharm. Bull. 1989,37, 1653. 68. Bianchi, D.; Cesti, P.; Battistel, E. J. Org. Chem. 1988, 53, 5531–5534. 69. (a) Bianchi, D.; Cabri, W.; Cesti, P.; Francalanci, F.; Rama, F. Tetrahedron Lett. 1988, 29, 2455–2458. (b) Kanerva, L. T.; Vänttinen, E. Tetrahedron Asymm. 1993, 4, 85–90. 70. (a) Ferraboschi, P.; Brembilla, D.; Grisenti, P.; Santaniello, E. J. Org. Chem. 1991, 56, 5478– 5480. (b) Ferraboschi, P.; Casati, S.; Grisenti, P.; Santaniello, E. Tetrahedron Asymm. 1993, 4, 9–12. 71. Bevinakatti, H. S.; Newadkar, R. V. Tetrahedron Asymm. 1993, 4, 773–776. 72. Egri, G.; Baitz-Gásc, E.; Poppe, L. Tetrahedron Asymm. 1996, 7, 1437–1448. 73. (a) Hof, R. P.; Kellogg, R. M. Tetrahedron Asymm. 1994, 5, 565–568; (b) Hof, R. P.; Kellogg, R. M. J. Chem. Soc. Perkin Trans. I 1996, 2051–2060. 74. Chen, S.-T.; Fang, J.-M. J. Org. Chem. 1997, 62, 4349–4357. 75. Peter, M.; Van der Eycken, J.; Bernáth, G.; Fülöp, F. Tetrahedron Asymm. 1998, 9, 2339–2347. 76. Pallavicini, M.; Valoti, E.; Villa, L.; Piccolo, O. J. Org. Chem. 1994, 59, 1751–1754. 77. (a) Antus, S.; Gottsegen, A.; Kajtár, J.; Kovács, T.; Tóth, T. S.; Wagner, H. Tetrahedron Asymm. 1993, 4, 339–344. (b) Ennis, M. D.; Old, D. W. Tetrahedron Lett. 1992, 33, 6283–6286. 78. Akita, H.; Umezawa, I.; Nozawa, M.; Nagumo, S. Tetrahedron Asymm. 1993, 4, 757–760. 79. Yokomatsu, T.; Nakabayashi, N.; Matsumoto, K.; Shibuya, S. Tetrahedron Asymm. 1995, 6, 3055–3062. 80. Herradón, B.; Cueto, S.; Morcuende, A.; Valverde, S. Tetrahedron Asymm. 1993, 4, 845–864. 81. Weidner, J.; Theil, F.; Kunath, A.; Schick, H. Liebigs Ann. Chem. 1991, 1301. 82. Khlebnikov, V.; Mori, K.; Terashima, K.; Tanaka, Y.; Sato, M. Chem. Pharm. Bull. 1995, 43, 1659–1662. 83. Tanimoto, H.; Oritani, T. Tetrahedron Asymm. 1996, 7, 1695–1704. 84. Johnson, C. R.; Xu, Y.; Nicolao, K. C.; Yang, Z.; Guy, R. K.; Dong, J. G.; Berova, N. Tetrahedron Lett. 1995, 36, 3291–3294. 85. (a) Laval, G.; Audran, G.; Sanchez, S.; Monti, H. Tetrahedron Asymm. 1999, 10, 1927–1933. (b) Laval, G.; Cadillio, G.; Monti, H.; Tolomelli, A.; Audran, G.; Galano, J.-M. Tetrahedron Asymm. 2000, 11, 1289–1294. 86. Hyatt, J. A.; Skelton, C. Tetrahedron Asymm. 1997, 8, 523–526. 87. Mizuguti, E.; Suzuki, T.; Achiwa, K. Synlett 1994, 929–930. 88. Kámán, J.; Forró, E.; Fülöp, F. Tetrahedron Asymm. 2000, 11, 1593–1600. 89. Miyazawa, K.; Yufit, D. S.; Howard, J. A. K.; de Meijere, A. Eur. J. Org. Chem. 2000, 24, 4109–4117. 90. Wakamatsu, H.; Terao, Y. Chem. Pharm. Bull. 1996, 44, 261–263. 91. Nakano, H.; Okuyama, Y.; Iwasa, K.; Hongo, H. Tetrahedron Asymm. 1994, 5, 1155–1156. 92. Ayers, T. A.; Schnettler, R. A.; Marciniak, G.; Stewart, K. T.; Mishra, R. K.; Krysan, D. J.; Bernas, B. R.; Bhardwaj, P.; Fevig, T. L. Tetrahedron Asymm. 1997, 8, 45–55. 93. Mizuno, K.; Sakuda, S.; Nihira, T.; Yamada, Y. Tetrahedron 1994, 50, 10849–10858. 94. (a) Bracher, F.; Papke, T. Tetrahedron Asymm. 1994, 5, 1653–1656. (b) Nordin, O.; Heden ström, E.; Hgberg, H.-E. Tetrahedron Asymm. 1994, 5, 785–788. 95. Schieweck, F.; Altenbach, H.-J. Tetrahedron Asymm. 1998, 9, 403–406. 96. Lemaire, M.; Bolte, J. Tetrahedron Asymm. 1999, 10, 4755–4762. 97. Gais, H.-J.; von der Weiden, I. Tetrahedron Asymm. 1996, 7, 1253–1216. 98. Goj, O.; Burchardt, A.; Haufe, G. Tetrahedron Asymm. 1997, 8, 399–408. 99. Jiménez, O.; Bosch, M. P.; Guerrero, A. J. Org. Chem. 1997, 62, 3496–3499. 100. Kiełbasi´nski, P.; Omela´nczuk, J.; Mikolajczyk, M. Tetrahedron Asymm. 2000, 11, 1313–1321. 101. Forró, E.; Árva, J.; Fülöp, F. Tetrahedron Asymm. 2001, 12, 643–649.
50
Takashi Sakai
102. Nodin, O.; Nguyen, B.-V.; Vörde, C.; Hedenström, E.; Högberg, H.-E. J. Chem. Soc. Perkin Trans. I 2000, 367–376. 103. Hof, R. P.; Kellogg, R. M. J. Org. Chem. 1996, 61, 3423–3427. 104. Rigby, J. H.; Sugathapala, P. Tetrahedron Lett. 1996, 37, 5293–5296. 105. Itoh, T.; Kawai, K.; Hayase, S.; Ohara, H. Tetrahedron Lett. 2003, 44, 4081–4084. 106. Kurokawa, M.; Shindo, T.; Suzuki, M.; Nakajima, N.; Ishihara, K.; Sugai, T. Tetrahedron Asymm. 2003, 14, 1323–1333. 107. Murakami, M.; Kamaya, H.; Kaneko, C.; Sato, M. Tetrahedron Asymm. 2003, 14, 201–215. 108. de Gonzalo, G.; Brieva, R.; Sánchez, V. M.; Bayod, M.; Gotor, V. J. Org. Chem. 2003, 68, 3333–3336. 109. Angoli, M.; Barilli, A.; Lesma, G. D.; Riva, S.; Silvani, A.; Danieli, B. J. Org. Chem. 2003, 68, 9525–9527. 110. Kiełbasi´nski, P.; Albrycht, M.; Luczak, J.; Mikolajczyk, M. Tetrahedron Asymm. 2002, 13, 735–738. 111. Bierstedt, A.; Stölting, J.; Fröhlich, R.; Metz, P. Tetrahedron Asymm. 2001, 12, 3399–3407. 112. de Gonzalo, G.; Brieva, R.; Sánchez, V. M.; Bayod, M.; Gotor, V. J. Org. Chem. 2001, 66, 8947–8953. 113. Forro, E.; Fulop, F. Tetrahedron Asymm. 2001, 12, 2351–2358. 114. Miyazawa, T.; Yukawa, T.; Koshiba, T.; Sakamoto, H.; Ueji, S.; Yanagihara, R.; Yamada, T. Tetrahedron Asymm. 2001, 12, 1595–1602. 115. Ferorelli, S.; Franchini, C.; Loiodice, F.; Perrone, M. G.; Scilimati, A.; Sinicropi, M. S.; Tortorella, P. Tetrahedron Asymm. 2001, 12, 853–862. 116. Olivo, H. F.; Velazquez, F.; Trevisan, H. C. Org. Lett. 2000, 2, 4055–4058. 117. Kawasaki, M.; Goto, M.; Kawabata, S.; Kometani, T. Tetrahedron Asymm. 2001, 12, 585–596. 118. Hamersak, Z.; Ljubovic, E.; Mercep, M.; Mesic, M.; Sunjic, V. Synthesis 2001, 1989–1992. 119. Akita, H.; Nozawa, M.; Mitsuda, A.; Ohsawa, H. Tetrahedron Asymm. 2000, 11, 1375–1388. 120. Ramadas, S.; Krupadanam, G. L. D. Tetrahedron Asymm. 2000, 11, 3375–3393. 121. Kawasaki, M.; Goto, M.; Kawabata, S.; Kodama, T.; Kometani, T. Tetrahedron Lett. 1999, 40, 5223–5226. 122. Majeri´c, M.; Šunji´c, V. Tetrahedron Asymm. 1996, 7, 815–824. 123. Nair, M. S. Anilkumar, A. T. Tetrahedron Asymm. 1996, 7, 511–514. 124. Naemura, K.; Fukuda, R.; Murata, M.; Konishi, M.; Hirose, K.; Tobe, Y. Tetrahedron Asymm. 1995, 6, 2385–2394. 125. Sankaranarayanan, S.; Sharma, A.; Kulkarni, B. A.; Chattopadhyay, S. J. Org. Chem. 1995, 60, 4251–4254. 126. Asensio, G.; Andreu, C.; Marco, J. A. Tetrahedron Lett. 1991, 32, 4197–4198.
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
51
Chapter 3
Future directions in photosynthetic organisms-catalyzed reactions Kaoru Nakamura Institute for Chemical Research, Kyoto University, Uji, Kyoto, Japan
Abstract Present states and future directions of the use of photosynthetic microorganisms, especially microalgae, for biotransformation are discussed.
1. INTRODUCTION Heterotrophs such as fungus, bacteria, and yeasts have been used as biocatalysts for biotransformation of organic compounds to afford useful compounds such as chiral intermediates for medicines. On the contrary, autotrophs such as plant cell and microalgae are rare to be utilized for biotransformations, and investigation is necessary because they are environment-friendly catalysts: they absorb carbon dioxide to generate oxygen using solar energy. Among photosynthetic organisms, microalgae are expected to form a new group of biocatalysts because of the high growth rates compared to plant cells. In this chapter, reduction, oxidation, and hydroxylation using algae are introduced. Environmental remediation using algae is also explained.
2. REDUCTION REACTION The reduction of ketones, aldehydes, and olefins has been extensively explored using chemical and biological methods. As the latter method, reduction by heterotrophic microbes has been widely used for the synthesis of chiral alcohols. On the contrary, the use of autotrophic photosynthetic organisms such as plant cell and algae is relatively rare and has not been explored because the method for culti vation is different from that of heterotrophic microbes. Therefore, the investigation using photosynthetic organisms may lead to novel biotransformations. Indeed, recent research on the use of a cyanobacterium as a biocatalyst has opened up this area: asymmetric reduction of ketones by a cyanobacteria, Syne chococcus elongates PCC 7942, with the aid of light energy proceeded smoothly
52
Kaoru Nakamura
Figure 1: Reduction of ketones by cyanobacteria.
and afforded the corresponding (S-alcohols in excellent enantioselectivities1−3 (Fig. 1). Characteristics of the reduction reaction using cyanobacteria are as follows: (a) Biocatalyst/substrate (b/s) ratios are low: A large amount of biocatalyst is usually required to reduce a considerable amount of substrate (the b/s for baker’s yeast is about 50–350). On the contrary, a low b/s ratio (2.6–0.5) could be achieved using the cyanobacteria. The improvement in the b/s ratio is caused by the fact that the cyanobacterium can utilize the power of light effectively to reduce the substrate. (b) The reaction afforded high selectivity and wide substrate specificity: Ketones used in this report are reduced by the cyanobacterium with excellent enan tioselectivities (> 96% ee). An enzyme exhibiting high enantioselectivity usually shows a relatively strict substrate specificity; hence, there scarcely is a catalyst that reacts with many kinds of substrates and also shows high selectivities. This alga can reduce a wide variety of aryl methyl ketones and afford the corresponding alcohols with high enantioselectivities. (c) Light energy can be used for reduction: Reduction of substrates usually requires a large input of energy, and in microbial reductions, carbohydrates such as sugars have been used as the energy source. These carbohydrates are generated through photosynthesis with sunlight energy. In other words, we have been indirectly using light energy for asymmetric reduction in the heterotrophic organism-catalyzed reduction. However, in the bioreduction using microalgae, the reaction was promoted by the direct use of light energy. For the majority of redox enzymes, nicotinamide adenine dinucleotide [NAD(H)] and its respective phosphate [NADP(H)] are required. These cofactors are prohibitively expensive if used in stoichiometric amounts. Since it is only the oxidation state of the cofactor that changes during the reaction, it may be regen erated in situ by using a second redox reaction to allow it to re-enter the reaction cycle. Usually in the heterotrophic organism-catalyzed reduction, formate, glu cose, and simple alcohols such as ethanol and 2-propanol are used to transform the
Future directions in photosynthetic organisms-catalyzed reactions
53
oxidized form of the coenzyme to the reduced form. These reductants originally stem from bio-products of CO2 with phototrophs with the aid of sunlight. Phototrophs such as algae and plants capture light energy to generate NADPH from NADP+ through photosynthetic electron-transfer reactions. Sub sequently, CO2 is converted into sugar, generally using NADPH. The reducing power of NADPH generated through photosynthesis can also be used in the reduc tion of exogenous substrates such as unnatural ketones to yield useful optically active alcohols. Thus, cofactor-recycling is no problem when photosynthetic living cells are used as biocatalysts for reduction. Accordingly, we can use solar energy directly for bioconversion of artificial substrates.2 (d) The stereochemical course of the reduction is possibly changed by light con ditions: The reduction of ,-difluoroacetophenone by a cyanobacterium proceeded both under light and in the dark, and the poor enantioselectivities (20–30% ee) observed in the dark were improved by irradiation. Thus, the enantioselectivities increased according to the lightness (70% ee under light (1000 lux)). The use of DCMU, a photosynthetic inhibitor, decreased the enantioselectivity of the reduction even under light conditions. The stereochemical course of the reduction is controlled by illumination or by adding DCMU.3 (e) The microalgae is easily manipulated and has high growth rate: Photosynthetic plant cell cultures typically grow quite slowly. However, cyanobac teria grow much faster than plant cell cultures in spite of the fact that the algae are types of phototrophs. Thus, contamination with microbes during the cultivation of algae can be prevented, and sufficient amount for biotransformation can be obtained easily. (f) Ecological system: Algae have the ability to directly utilize sunlight and carbon dioxide for photosyn thesis. Due to this activity, cyanobacteria may help to solve a global environmental problem, the greenhouse effect, which increasingly threatens mankind at the begin ning of the 21st century. Other examples of microalgae-catalyzed reductions of carbonyl groups are summarized below and shown in Fig. 2: • Reduction of carbonyl groups: Terpene4 and aromatic56 aldehydes (100 ppm) were reduced by microalgae. In a series of chlorinated benzaldehyde, m- or p chlorobenzaldehyde reacted faster than the o-derivative.5 Due to toxicity, the substrate concentrations are difficult to increase. Asymmetric reductions of ketones by microalgae were reported. Thus, aliphatic7−9 and aromatic1−31011 ketones were reduced.
Figure 2: Examples of reduction and oxidation reactions.
Future directions in photosynthetic organisms-catalyzed reactions
55
• Reduction of carbon–carbon double bond: Microalgae easily reduce carbon– carbon double bonds in enone.12 Usually, the reduction of carbonyl group and carbon–carbon double bond proceeds concomitantly to afford the mixture of corresponding saturated ketone, saturated alcohol, and unsaturated alcohol because a whole cell of microalgae has two types of reductases to reduce carbonyl and olefinic groups. The use of isolated reductase, which reduces carbon–carbon double bond chemoselectively, can produce saturated ketones selectively. The future direction of reduction of carbonyl compounds is as follows: 1. Development of new reduction systems that reduce sterically hindered compounds: The reported examples of reduction of carbonyl compounds are usually for the substrates that can be easily reduced such as methyl ketones. Since the demand for reduction of various types of com pounds is increasing, investigation of new biocatalytic reductions is required. Photosynthetic organisms are not investigated yet, and they may have new type of enzymes, which can reduce sterically hindered compounds. 2. Development of a new system for biocatalysis by photosynthetic organ isms: Since most of the present reactors of biocatalysis are for heterotrophic microorganisms, their systems are not usable for the reaction by photosyn thetic organisms. Since they require light for growing, it is necessary to develop a new type of photobioreactor to utilize the light energy as reduction power efficiently. 3. Development of new methods for controlling reactions: Although many methods are already known in microbial reactions using heterotroph to control the reaction,13 little is known about the methods using photosynthetic organisms. In usual microbial reactions, three methods, screening of suitable microbes, screening of suitable substrates, and modifying the reaction conditions, have been used for controlling the reaction. These methods have to be modified to use for the reaction by photosynthetic organisms. For example, for the efficient screening of photosynthetic organisms, the formation of large culture stock is necessary.
3. OXIDATION AND HYDROXYLATION Microalgae were used for oxidation and hydroxylation of organic compounds (Fig. 3). For example, hydroxylation of (S)-limonene affords a mixture of cis and trans carveols.14 By hydroxylation and oxidation using Chlorella,15
56
Kaoru Nakamura
Figure 3: Oxidation, hydroxylation and degradation.
valencene, cheaply obtained from Valencia orange, was converted to the product nootkatone, which is the aroma of grapefruit and is used for cos metic and fiber manufactures. Bioconversion of hydrocortisone by Nostoc afforded the degradated ketone16 via oxidation. Oxidation of sulfur atom in thiaoleic acid was done by Chlorella.17 This reaction was catalyzed by oleoyl desaturase. However, in these reactions, over-degradated product is also formed, and it is necessary to develop methods to inhibit the unnecessary reaction to afford a single product in high chemical yield. Stereo- and regioselective reactions are also desired.
4. REMOVAL OF ORGANIC AND INORGANIC SUBSTANCES IN WASTEWATER The prudent disposal of large quantities of agricultural and domestic wastewaters is a major concern in our modern society. Raw wastewaters most often contain significant amounts of bioavailable organic substances such as phenols, endocrine
Future directions in photosynthetic organisms-catalyzed reactions
57
disrupters, and inorganic substances such as ammoniate, nitrate, and phosphate. To remove these substances, algae and microalgae were used.18 For example, inorganic substances such as ammonium,19 phosphorous,18 and arsenate20 compounds and organic substances such as nitro compounds,21 phenols including bisphenol-A,2223 and endocrine-disrupting chemicals24 were removed by using algae or microalgae. The problem of the investigation for removal of haz ardous compounds from wastewater is that the removal and degradation abilities are insufficient for practical use. The easiness of harvest of the biomass is also problematic. Future direction of this field is to develop a novel system for remedia tion: screening of powerful biocatalysts that degrade hazardous compounds rapidly and that can be easily harvested after degradation. The latter is necessary for removal of inorganic substances. Microalgae immobilized in a polymeric matrix or as attached algal communities (biofilm/periphyton) growing in shallow,25 artificial streams or on the surfaces of rotating biological contactors (RBC/biodiscs)26 are expected to be developed.
5. CONCLUSION At present, photosynthetic organisms are not generally used as biocatalysts for bioconversion of organic compounds except for bioremediation of pollutants in the environment, although they are environment-friendly catalysts, and they may contain unusual type of enzymes to establish new reactions. Development of bioreactors specially developed for photosynthetic organism-catalyzed reaction as well as finding effective photosynthetic organisms as a biocatalyst are required in the future.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.
Nakamura, K.; Yamanaka, R.; Tohi, K.; Hamada, H. Tetrahedron Lett. 2000, 41, 6799–6802. Nakamura, K.; Yamanaka, R. J. Chem. Soc. Chem. Commun. 2002, 1782–1783. Nakamura, K.; Yamanaka, R. Tetrahedron Asymm. 2002, 13, 2529–2533. Noma, Y.; Takahashi, H.; Asakawa, Y. Phytochemistry 1991, 30, 1147. Hook, I. L.; Ryan, S.; Sheridan, H. Phytochemistry 1999, 51, 621–627. Noma, Y.; Okajima, Y.; Takahashi, H.; Asakawa, Y. Phytochemistry 1991, 30, 2969–2972. Fiorentino, A.; Pinto, G.; Pollio, A.; Previtera, L. Bioorg. Med. Chem. Lett. 1991, 1, 673–674. Utsukihara ,T.; Chai, W.; Kato, N.; Nakamura, K.; Horiuchi, C. A. J. Mol. Catal. B Enzym. 2004, 31, 19–24. 9. Yoshizako, F.; Kuramoto, T.; Nishimura, A.; Chubachi, M. J. Ferment. Bioeng. 1998, 85, 439–442. 10. Utsukihara, T.; Misumi, O.; Kato, N.; Kuroiwa, T.; Horiuchi, C. A. Tetrahedron Asymm. 2006, 17, 1179–1185. 11. Itoh, K.; Sakamaki, H.; Nakamura, K.; Horiuchi, C. A. Tetrahedron Asymm. 2005, 16, 1403–1408.
58
Kaoru Nakamura
12. Shimoda, K.; Hirata, T. J. Mol. Catal. B Enzym. 2000, 8, 255–264. 13. Nakamura, K.; Yamanaka, R.; Matsuda, T.; Harada, T. Tetrahedron Asymm. 2003, 14(18), 2659–2681. 14. Hamada, H.; Kondo, Y.; Ishihara, K.; Nakajima, N.; Hamada, H.; Kurihara, R.; Hirata, T. J Biosci. Bioeng. 2003, 96, 581–584. 15. Furusawa, M.; Hashimoto, T.; Noma, Y.; Asakawa, Y. Chem. Pharm. Bull. 2005, 53, 1513–1514. 16. Yazdi, M. T.; Arabi, H.; Faramarzi, M. A.; Ghasemi, Y.; Amini, M.; Shokravi, S.; Mohseni, F. A. Phytochemistry 2004, 65, 2205–2209. 17. Nugier-Chauvin, C.; Fauconnot, L.; Noiret, N.; Poulain, S.; Patin, H.; J. Mol. Catal. B Enzym. 1998, 5, 133–135. 18. Hoffmann, J. P. J. Phycol. 1998, 34, 757–763. 19. Talbot, P.; de la Noue, J. Water Res. 1993, 27, 153–159. 20. Murray, L. A.; Raab, A.; Marr, I. L.; Feldmann, J. Appl. Organometal. Chem. 2003, 17, 669–674. 21. Pavlostathis, S. G.; Jackson, G. H. Water Res. 2002, 36, 1699–1706. 22. Yang, S.; Wu, R. S. S.; Kong, R. Y. C. Aquatic Toxicol. 2002, 59, 191–200. 23. Hirooka, T.; Akiyama, Y.; Tsuji, N.; Nakamura, T.; Nagase, H.; Hirata, K.; Miyamoto, K. J. Biosci. Bioeng. 2003, 95, 200–203. 24. Sethunathan, N.; Megharaj, M.; Chen, Z. L.; Williams, B. D.; Lewis, G.; Naidu, R. J. Agric. Food Chem. 2004, 52, 3030–3035. 25. Li, X. Z.; Webb, J. S.; Kjelleberg, S.; Rosche, B. Appl. Environ. Microbiol. 2006, 72, 1639–1644. 26. Kargi, F.; Eker, S. Process Biochem. 2002, 37, 1201–1206.
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Published by Elsevier B.V.
59
Chapter 4
Catalysis by enzyme–metal combinations Mahn-Joo Kim1 , Jaiwook Park1 , Yangsoo Ahn1 , Palakodety R. Krishna2 1
Department of Chemistry, Pohang University of Science and Technology, Pohang, Gyeongbuk, Korea 2 Discovery Laboratory, Organic Chemistry Division III, Indian Institute of Chemical Technology, Hyderabad, India
Abstract Catalytic transformation based on combined enzyme and metal catalysis is described as a new class of methodology for the synthesis of enantiopure compounds. This approach is particularly useful for dynamic kinetic resolution in which enzymatic resolution is coupled with metal-catalyzed racemization for the conversion of a racemic substrate to a single enantiomeric product.
1. INTRODUCTION The search for new and efficient methods for the synthesis of optically pure compounds has been an active area of research.1 The ever-growing market for the “chiral drugs,” especially after the understanding of the causes of Thalidomide Tragedy in 1960s, has put tremendous burden on chemists in discovering new methods and transforming these methods into technologies for the synthesis of single enantiomer drugs by economically viable processes. Traditionally, kinetic resolution (KR) of a racemate is still one of the major methods adopted for the production of the required enantiomer on the industrial front.23 However, this practice is often associated with the problem of either discarding unwanted isomer or recycling it at high costs. Reverting to asymmetric processes in which achiral starting materials are converted to chiral non-racemic products is still an attractive proposition. Consequently, efforts have been directed at discovering such asymmetric reactions/methods or catalytic reagents ideally suited for the production of single enantiomers in good yields and purities. The novel phenomenon of converting racemic substrates into a single enan tiomer of the product by dynamic kinetic resolution (DKR) via racemization of the substrates has been a formidable challenge in asymmetric synthesis.4 Recently, DKR has been receiving increasing attention since it can overcome the limitations
60
Mahn-Joo Kim et al.
of the conventional KR. Particularly, the maximum yield of one stereoisomer of the product in DKR can reach 100% whereas the maximum yield of one stereoiso mer in KR cannot exceed 50%. Accordingly, DKR has been intensively studied as a better alternative for asymmetric synthesis. Attempts are being made to develop efficient catalysts for DKR. This chapter deals with recent developments in this area, in particular DKR by enzyme–metal combinations. Each successful DKR is exemplified with sev eral substrates and novel metal catalyst. Asymmetric transformations of achiral substrates via DKR of racemic intermediates also are described.
2. DYNAMIC KINETIC RESOLUTIONS BY ENZYME–METAL COMBINATIONS DKR requires two catalysts: one for resolution and one for racemization. We and others have developed a novel strategy using enzyme as the resolution catalyst and metal as the racemization catalyst as shown in Scheme 1.5 The R-selective DKR can be achieved by combining a R-selective enzyme with a proper metal catalyst and its counterpart by the combination of the metal catalyst with a S-selective enzyme. This strategy has been demonstrated to be applicable to the DKR of secondary alcohols, allylic esters, and primary amines. Among them, the DKR of secondary alcohols has been the most successful. 2.1. DKR of secondary alcohols The KR of secondary alcohols by some hydrolytic enzymes has been well known. The combinations of these hydrolytic enzymes with racemization catalysts have been explored as the catalysts for the efficient DKR of the secondary alcohols. Up to now, lipase and subtilisin have been employed, respectively, as the R- and S-selective resolution enzymes in combination with metal catalysts (Scheme 2).
X R1
R
EnzR
R2
Y R1
R R2
Metal X S R1
R2
Y
EnzS
S R1
R2
Scheme 1: DKR by enzyme–metal combination.
Catalysis by enzyme–metal combinations OH R1
R
R2
61
OCOR
Lipase RCO2R'
R1
R
R2
Metal OH S R1
R2
OCOR
Subtilisin
S
RCO2R'
R1
R2
Scheme 2: DKR of secondary alcohol. 2.1.1. (R)-selective DKR of secondary alcohols The use of metal catalysts with the hydrolytic enzymes in the DKR of simple secondary alcohols was investigated by several groups including our, Williams, and Bäckvall groups. In 1996, Williams et al. reported for the first time the use of a metal catalyst in the DKR of secondary alcohols.7 The DKR was performed on 1-phenylethanol with the combination of a lipase and a rhodium complex to result in the product albeit in low enantiomeric excess (80% ee). Later, Bäckvall et al. reported a significant improvement of the enantiomeric output by using a diruthenium complex 1 along with an immobilized and thermally stable lipase (Candida antarctica lipase B (CALB); trade name Novozym-435).10 A notable modification in this method was the use of p-chlorophenyl acetate (PCPA), which was found to be more compatible with the racemization catalyst than popular acyl donors such as vinyl and isopropenyl acetate. Thus, the DKR of 1-phenylethanol by this procedure resulted in optically pure (R--phenylethyl acetate in high yield (Table 1).10b However, the procedure has some drawbacks. First, it requires
Table 1 DKR of 1-phenylethanol with diruthenium complex 1 OH Ph
Acyl donor
Cl
CALB 1 Acyl donor
acetophenone
t-BuOH, 70°C
OAc Ph
ee (%)
Yield (%)
OAc
>99
50
OAc
>99
72
>99
100
OAc
62
Mahn-Joo Kim et al.
an elevated temperature (70 C) for the activation of the diruthenium catalyst. The high temperature is unacceptable for thermally less-stable enzymes. Second, a stoichiometric amount of acetophenone should be added as hydrogen mediator for inducing racemization and thus achieving high conversion yield. The absence of acetophenone usually costs 10–15% lowering in yield. Third, unreacted PCPA is often difficult to remove from the acylated products during work-up. To overcome these limitations, more efficient racemization catalyst should be developed, which is highly active at room temperature, compatible with readily available acyl donors such as vinyl and isopropenyl acetate, and does not require a hydrogen mediator for racemization.
Ph Ph
O
O H Ph Ph
Ph
Ph Ph Ru H Ru Ph CO OC CO CO
Cl Cl Ru Ru X Cl
Ru PPh 3 Cl PPh3
1
2
3: X = Cl 4: X = H
Later, in a modification to the above system, we reported the use of an indenylruthenium complex 2 as a racemization catalyst for the DKR of secondary alcohols, which does not require ketones but a weak base like triethylamine and molecular oxygen to be activated.11 The DKR with 2 in combination with immobilized Pseudomonas cepacia lipase (PCL, trade name, Lipase PS-C® ) was carried out at a lower temperature (60 C) and provided good yields and high optical purities (Table 2). This paved the way for the omission of ketones as
Table 2 DKR of secondary alcohols with indenyl ruthenium complex 2 OH R
R′
OAc
PCL 2 PCPA, Et3N, O2 60°C, 43 h
R
R′
R
R
ee (%)
Yield (%)
Ph 4-MeO-Ph 4-Br-Ph 1-Indanyl PhCH2
Me Me Me
96 99 99 82 97
86 82 98 88 60
Me
Catalysis by enzyme–metal combinations
63
Table 3 DKR of secondary alcohols with cymene-ruthenium catalyst 3 OH R
R Ph 4-MeO-Ph 4-Cl-Ph PhCH2
OAc
PCL 3 PCPA, Et3N
CH2Cl2, 40°C
R
ee (%)
Yield (%)
94 99 99 >99
95 93 93 85
hydrogen mediators for the racemization. However, we wanted that ideally the reaction should be carried out at ambient temperature. Towards that goal, soon after, we discovered a remarkable cymene-ruthenium catalyst 3 and its hydride form 4 as effective catalyst systems for a facile DKR of secondary alcohols at 40 C (Table 3). A noticeable feature of this catalyst system was that allylic alco hols12 also underwent facile DKR at room temperature to provide high yields of corresponding chiral acetates with excellent optical purities (Table 4). Addi tionally, even in ionic liquids such as [EMIm]BF4 and [BMIm]PF6 ([EMIm] = 1-ethyl-3-methylimidazolium, [BMIm] = 1-butyl-3-methylimidazolium),13 the
Table 4 DKR of allylic alcohols with cymene-ruthenium catalyst 4 OH R
R Ph 4-Cl-Ph 4-MeO-Ph 2-Furyl c-C6 H11 (CH3 3 C
OAc
PCL 4 PCPA, Et3N CH2Cl2, r.t.
R
ee (%)
Yield (%)
>99 99 99 99 95 >99
84 91 85 92 90 85
64
Mahn-Joo Kim et al.
Table 5 DKR of alcohols with cymene-ruthenium catalyst 3 in [BMIm]PF6 OH R
R′
OAc
PCL 3 R
CH3CO2CH2CF3, Et3N [BMIm]PF6, r.t.
R′
R
R
ee (%)
Yield (%)
Ph PhCHCH2 3-Me-Ph 4-Me-Ph 4-MeO-Ph 4-Cl-Ph 4-Br-Ph 1-Indanyl
Me Me Me Me Me Me Me
85 85 87 87 85 87 92 85
4-[CH3 CH(OH)]Ph
Me
3-[CH3 CH(OH)]Ph
Me
99 99 98 99 99 99 99 99 99 (de 99%) 99 (de 97%)
87 86
same efficacious DKR was retained for benzylic alcohols at room temperature. It was observed that the Ru-catalyzed racemization occurred more rapidly in ionic liquids compared to organic solvents. The added advantage of the use of ionic liquids has been the reusability of both the catalyst and the enzyme by a simple extraction of the products by ether (Table 5).14
Ph Y Ph Ph Ph Ru OC Cl OC
5: Y = NHCH(CH3)2 6: Y = Ph
In an effort directed at developing a racemization catalyst which works uniformly for all the substrates at room temperature, we designed and synthe sized a novel aminocyclopentadienyl ruthenium chloride complex 5.9 The DKR of aromatic as well as aliphatic alcohols could be conducted at room tempera ture. In case of aromatic alcohols, the substituent effects were found insignificant in the DKR; however, aromatic alcohols have comparatively faster conversion rates than their aliphatic counterparts. This is the first ever report of a catalyst
Catalysis by enzyme–metal combinations
65
Table 6 DKR of alcohols with aminocyclopentadienyl ruthenium complex OH R
CALB 5 or 6 Isopropenyl acetate tBuOK, Na2CO3 toluene, 25°C
R′
OAc R
R′
R
R
Catalyst
ee (%)
Yield (%)
Ph
Me
4-Cl-Ph
Me
4-MeO-Ph
Me
1-Indanyl 4-NO2 -Ph 4-CN-Ph c-C6 H11
Me Me Me
CH3 (CH2 4 CH2 =CH PhCH= (Ph)3 COCH2 Ph c-C6 H11
Me Me Me =CH2 CH= =CH2 CH=
5 6 5 6 5 6 5 6 6 5 6 5 5 5 5 5
>99 >99 >99 >99 >99 >99 95 >99 >99 >99 >99 91 98 99 81 >99
95 92 94 91 90 94 89 97 95 86c 98 89c 93 97 62 90
system that works efficiently at room temperature for the successful DKR of secondary alcohols. The resulting acylated products were obtained in high yields and high enantiomeric excess (Table 6). An additional feature is the use of iso propenyl acetate, which is readily available, more active than PCPA, and easily separable from the DKR mixtures.15 Although the mechanism of catalytic racem ization is not yet clear, according to our interpretation, it can be deduced that the amino group in 5 seems to play a crucial role in the racemization, though Bäckvall suggests a different pathway16 for a similar yet modified complex 6 that has no amino functionality. At the same time, Sheldon et al. reported a new catalyst system of [TosN(CH2 2 NH2 ]RuCl(p-cymene) and 2,2,6,6-tetramethyl-1 piperidinyloxy (TEMPO) for the DKR of alcohols17 but resulted in the conversion of 1-phenylethanol to its corresponding acetate with only 76% yield. The DKR of functionalized alcohols such as diols, hydroxy esters, hydroxy aldehydes, azido alcohols and hydroxy nitriles was also taken up as the synthetic utility of the products is very high; besides such a study will bring out the effect of multifunctional substrates under these reaction conditions to broaden the scope of DKR. Initially, the DKR of diols was achieved with diruthenium catalyst 1
66
Mahn-Joo Kim et al.
and CALB in the presence of PCPA to give the corresponding diacetates of (R R-configuration from the mixture of dl- and meso-isomers (Table 7).18 The DKRs of rigid benzylic diols with 1 gave better results in terms of diastereose lectivity compared to those of more flexible aliphatic diols, reflecting that lipase displays higher stereoselectivity towards benzylic diols than aliphatic diols. The DKRs of -, -, - and -hydroxy esters were also accomplished with PCL and 1 at 60–70 C.19 In the DKRs, the enantioselectivities were good in most cases though the yields were moderate. The use of H2 was necessary in the DKR of - and -hydroxy esters to suppress the formation of ketones (Tables 8–10). The DKRs of small functionalized alcohols such as 2-hydroxybutanoic acid, 2-hydroxypropanal, and 1,2-propanediol were carried out after the protection of the terminal groups with a bulky group (Tables 11 and 12) since the bulky pro tecting groups enhanced the enantioselectivity of the enzyme in the DKR.20 In the DKR of hydroxy acids, t-butyl group was found to be the best for carboxyl group protection. The trityl group was a proper choice for the protection of primary alcohols in diols such as 1,2-propanediol, 1,2-butanediol, and 1,3-butanediol.20 1,2-Benzenedimethanol was used for protecting the formyl groups of - and hydroxy aldehydes.20 High enantiomeric excesses (95% and higher) were obtained in the DKRs of the protected diols and hydroxy aldehydes. 2,6-Dimethyl-4 heptanol was used as a hydrogen source to suppress the formation of oxidized side products.
Table 7 DKR of diols with ruthenium complex CALB [Ru]
OH OH X
X CH2 CH2 CH2 CH2 CH2 CH2 =CH CH= 1,3-Ph 1,4-Ph 2,5-Pyr CH2 N(n-Bu)CH2
OAc OAc
OAc OAc +
X (R, R)
Acyl donor toluene
X meso
[Ru]
eea (%)
(R, R)/meso
Yield (%)
1 1 1 1 1 5 1 5 6 1 1
>99 >99 >97 >99 >99 >99 >99 >99 >99 >99 >96
86/14 38/62 90/10 74/26 98/2 99/1 98/2 98/2 99/1 100/0 89/11
63 90 63 43 76 95 77 94 90 78 64
Catalysis by enzyme–metal combinations
67
Table 8 DKR of -hydroxy esters PCL 1
OH R
OAc
PCPA cyclohexane, 60°C
CO2Me
R
CO2Me
R
ee (%)
Yield (%)
94 94 98 98 30 80
80 76 69 80 62 60
Ph 4-MeO-Ph 4-Br-Ph c-C6 H11 PhCH2 CH2 CH3 CH2 CH2 CH2
Table 9 DKR of -hydroxy esters PCL 1
OH R
CO2Me
OAc R
PCPA (CH3)3COCH3 60°C, 6 days
R Ph 4-MeO-Ph c-C6 H11 PhCH2
CO2Me
ee (%)
Yield (%)
95 99 70 96
76 74 82 80
Table 10 DKR of - and -hydroxy esters PCL 1
OH t-BuO2C
n
R
2 3 3
Me Me Et
n
R
OAc t-BuO2C
H2 PCPA, toluene
n
R
Equiv. of PCPA
Temp ( C)
eep (%)
Yield (%)
3.9 3 3
60 70 70
94 98 98
70 89 87
68
Mahn-Joo Kim et al.
Table 11 DKR of hydroxy acids, diols and hydroxy aldehydes O
PCL 1
OH
RO
O RO
PCPA toluene, 70°C
R Ph 4-MeO-Ph 4-Ph-Ph (CH3 3 C
OAc
ee (%)
Yield (%)
86 93 94 >99
88 91 92 88
Table 12 DKR of protected diols and hydroxy aldehydes by lipase-ruthenium combination PCL 1
OH R
n R′
OAc R
PCPA 2,6-dimethyl-4-heptanol toluene, 70°C
n
R′
R
n
R
eep (%)
(Ph)3 CO (Ph)3 CO (Ph)3 CO
1 1 2
Me Et Me
>99 99 95
96 91 97
0
Me
98
95
1
Me
96
90
O
Yield (%)
O O O
The DKRs of -azido alcohols19d and -hydroxy nitriles21a were also accom plished by employing 1 and CALB with PCPA as the acyl donor. The DKRs of -azido alcohols were performed at 60 C while those of -hydroxy nitriles required higher temperature (100 C) primarily to enhance the racemization rate. The optical purities of products were satisfactory in all cases. In the case of -hydroxy nitriles, dehydrogenation lowered the yield.
Catalysis by enzyme–metal combinations
69
2.1.2. (S)-Selective DKR of secondary alcohols The lipase-catalyzed DKRs provide only (R-products; to obtain (S-products, we needed a complementary (S-stereoselective enzyme. A survey of (S-selective enzymes compatible to use in DKR at room temperature revealed that subtilisin is a worthy candidate, but its commercial form was not applicable to DKR due to its low enzyme activity and instability. However, we succeeded in enhancing its activity by treating it with a surfactant before use. At room temperature DKR with subtilisin and ruthenium catalyst 5, trifluoroethyl butanoate was employed as an acylating agent and the (S-products were obtained in good yields and high optical purities (Table 13).22 The (S-selective DKR of alcohols with subtilisin was also possible in ionic liquid at room temperature (Table 14).14 In this case, the cymeneruthenium complex 3 was used as the racemization catalyst. In general, the optical purities of (S-esters were lower than those of (R-esters described in Table 5. 2.1.3. DKR of secondary alcohols by air-stable metal catalyst All the Ru-based racemization catalysts described earlier are air-sensitive and thus difficult to reuse. We found that a modified Ru complex 7 was air-stable and recyclable, in particular, in a polymer-supported form 8. The racemization of secondary alcohols with 7 took place equally well under both oxygen and argon atmospheres. The subsequent DKRs of several alcohols using 7 or 8 under aerobic
Table 13 DKR of secondary alcohols by subtilisin-ruthenium bicatalysisa OH R
R Ph 4-Cl-Ph 4-MeO-Ph c-C6 H11 PhCH2 PhCH2 CH2 CH3 (CH2 4 CH2 =CH PhCH= a
Isolated yields in parentheses.
Subtilisin 5 PrCO2CH2CF3
THF, 25°C
ee (%) 92 99 94 98 92 98 98 95
OCOPr R
Yielda (%) 95 92 93 80 77 80 77 90
(90) (91) (74) (76) (78) (67) (90)
70
Mahn-Joo Kim et al.
Table 14 DKR by subtilisin-ruthenium bicatalysis in [BMIm]PF6 Subtilisin-CLEC 3
OH R
R′
OCOPr R
n-PrCO2CH2CF3, Et3N [BMIm]PF6, r.t.
R′
R
R
ee (%)
Yield (%)
PhCH2 PhCH2 CH2 3-Me-Ph 4-Me-Ph 4-MeO-Ph 4-Cl-Ph 4-Br-Ph 1-Indanyl 4-[CH3 CH(OH)]Ph
Me Me Me Me Me Me Me
89 90 90 90 80 92 91 84 78
3-[CH3 CH(OH)]Ph
Me
97 97 85 85 99 87 82 86 86 (de 52%) 96 (de 63%)
Me
83
conditions at room temperature provided excellent yields with high optical purities (Table 15). Ph Y Ph Ph Ph Ru OC Cl OC
7: Y = OBn 8: Y = OCH2Ph-4-(CO2-Polymer)
2.1.4. DKR of esters The hydrolytic DKR of allyl esters has been studied as a DKR of esters. The first DKR was accomplished through Pd-catalyzed racemization and enzymatic hydrolysis of allylic acetates in a buffer solution.6 However, the DKR under these conditions was limited to cyclohexenyl acetates to give symmetrical palladiumallyl intermediates. Among them, 2-phenyl-2-cyclohexenyl acetate 9 was the only substrate to have been resolved with good results (81% yield, 96% ee). Ph AcO
Ps. fluorescence lipase PdCl2(MeCN)2 Phosphate buffer 37–40°C, 19d
9
Ph HO
Catalysis by enzyme–metal combinations
71
Table 15 DKR of secondary alcohols with air-stable ruthenium catalysts OH R
OAc
CALB 7 or 8
R′
R
Isopropenyl acetate, K3PO4 toluene, 25°C
R′
R
R
Catalyst
ee (%)
Yield (%)
Ph
Me
4-MeO-Ph
Me
4-Cl-Ph
Me
4-MeO-Ph 1-Indanyl 1-Naphthyl
Me
2-Naphthyl
Me
c-C6 H11
Me
CH3 (CH2 4 CH2
Me
7 8 7 8 7 8 7 7 7 8 7 8 7 8 7 8
>99 >99 >99 >99 >99 >99 >99 88 >99 >99 >99 >99 >99 >99 >99 >99
98 98 98 98 98 98 92 91 98 98 98 98 98 95 98 95
Me
We improved the DKR of allylic acetates significantly by replacing the enzymatic hydrolytic reaction with enzymatic transesterification reaction and employing Pd(0) as the racemizing catalyst in organic solvent (Scheme 3).23 Here, Pd-catalyzed racemization is usually accompanied by reductive elimination giving 1,3-dienes as byproducts. The side reaction was effectively suppressed by the addition of a chelating ligand 1,1 -bis(diphenylphosphino)ferrocene (dppf). The OAc R1
R R2
Lipase ROH
OH R1
Pd(0) OAc R1
S R2
Scheme 3: DKR of allylic acetates.
R2
72
Mahn-Joo Kim et al.
DKR reactions were performed with lipase and Pd(PPh3 4 in the presence of dppf and 2-propanol in THF. 2-Propanol was used as an acyl acceptor. Various acyclic allylic acetates were transformed to their corresponding allylic alcohols at room temperature in good yields and excellent optical purities (Table 16). 2.1.5. DKR of amines The DKR of amine is more challenging compared to that of secondary alcohol since no metal catalysts have been known for the efficient racemization of amine. Reetz et al. reported for the first time the DKR of amine, in which 1-phenylethylamine was resolved by the combination of lipase and palladium (Scheme 4). In this procedure, CALB and Pd/C were employed as the combo catalysts.8 However, the DKR required a very long reaction time (8 days) at 50–55 C and provided a poor isolated yield (60%). Recently, an improved procedure using Pd on alkaline earth salts as the racemization catalyst was reported by Jacobs et al.24 The DKR reactions were performed at 70 C for 24–72 h and 75–88% yields were obtained with 99% or greater enantiomeric excess.
Table 16 DKR of allylic acetates by lipase–palladium bicatalysis OAc R
CALB Pd(PPh3)4, dppf i-PrOH THF, r.t.
R
ee (%)
Ph 4-Cl-Ph 4-Me-Ph 2-Furyl 2-Naphthyl
98 97 98 >99 98
NH2 R1
R
R2
OH R
Yield (%) 83 77 81 87 70
NHCOR
Lipase RCO2R'
R1
R
R2
Pd NH2 R1
S
R2
Scheme 4: DKR of amine by lipase–Pd combinations.
Catalysis by enzyme–metal combinations NH2 R1
R
R2
NHCOR
Lipase RCO2R'
73
R1
R
R2
Ru NH2 R1
S
R2
Scheme 5: DKR of amine by lipase–Ru combinations. Recently, Bäckvall et al. reported that a p-methoxy derivative of diruthenium complex 1 racemizes benzyl amines at 90–100 C in toluene and the DKRs using it were achieved successfully (Scheme 5). More than ten benzylic amines were transformed to the corresponding amides with good yields and high ee.25
3. ASYMMETRIC TRANSFORMATIONS BY ENZYME–METAL COMBINATIONS The DKR processes for secondary alcohols and primary amines can be slightly modified for applications in the asymmetric transformations of ketones, enol esters, and ketoximes. The key point here is that racemization catalysts used in the DKR can also catalyze the hydrogenation of ketones, enol esters, and ketoximes. Thus, the DKR procedures need a reducing agent as additional additive to enable asymmetric transformations. 3.1. Asymmetric transformation of ketone The catalytic alcohol racemization with diruthenium catalyst 1 is based on the reversible transfer hydrogenation mechanism. Meanwhile, the problem of ketone formation in the DKR of secondary alcohols with 1 was identified due to the lib eration of molecular hydrogen. Then, we envisioned a novel asymmetric reductive acetylation of ketones to circumvent the problem of ketone formation (Scheme 6). A key factor of this process was the selection of hydrogen donors compatible with the DKR conditions. 2,6-Dimethyl-4-heptanol, which cannot be acylated by lipases, was chosen as a proper hydrogen donor.26 Asymmetric reductive acetylation of ketones was also possible under 1 atm hydrogen in ethyl acetate, which acted as acyl donor and solvent. Ethanol formation from ethyl acetate did not cause critical problem, and various ketones were successfully transformed into the corresponding chiral acetates (Table 17).26b However, reaction time (96 h) was unsatisfactory. Asymmetric reductive acetylation process was also applicable to acetox yaryl ketones.27 For example, m-acetoxyacetophenone 10 was transformed to
74
Mahn-Joo Kim et al.
OH R1 O R1
R2
R2
R
Ru
Lipase
[H]
AcOR
Ru
OAc R1
R2
OH R1
R2
S
Scheme 6: Asymmetric reductive acetylation of ketones. Table 17 Asymmetric reductive acetylation of ketones CALB 1
O R
R′
OAc R′
R
H2, EtOAc 70°C
R
R
ee (%)
Yield (%)
Ph 4-MeO-Ph 4-Cl-Ph 1-Indanyl -Tetralyl c-C6 H11 CH3 (CH2 4 CH2 PhCH2 CH2
Me Me Me
96 99 97 99 99 90 91 72
81 85 72 89 87 87 87 83
Me Me Me
(R-1-(3-hydroxyphenyl)ethyl acetate 11 under 1 atm H2 in 95% yield. The over all reaction seems to involve simple reduction and acyl migration reaction. In fact, however, it is the result of nine catalytic steps: two steps for rutheniumcatalyzed reductions, two steps for ruthenium-catalyzed racemization, two steps for ruthenium-catalyzed deacylations, and three steps for lipase-catalyzed acyla tions (Scheme 7). This process was applicable to a wide range of acyloxyphenyl ketones (Table 18). O OAc
Lipase 1
OAc OH
H2 10
11
Catalysis by enzyme–metal combinations
75
OH OAc
Ru O
OAc
OH OAc
OAc
OAc
H2 Ru
Ac-lipase Lipase
10 Lipase
Lipase
Lipase
Ac-lipase
Ac-lipase
Ac-lipase
O
OAc
OH OH
OH
OH
H2 Ru
Ac-lipase Lipase
11
Ru OH OH
Scheme 7: Multistep acyl transfer of m-acetoxyacetophenone. 3.2. Asymmetric transformation of enol ester After succeeding in the asymmetric reductive acylation of ketones, we ventured to see if enol acetates can be used as acyl donors and precursors of ketones at the same time through deacylation and keto–enol tautomerization (Scheme 8). The overall reaction thus corresponds to the asymmetric reduction of enol acetate. For example, 1-phenylvinyl acetate was transformed to (R-1-phenylethyl acetate by CALB and diruthenium complex 1 in the presence of 2,6-dimethyl-4-heptanol with 89% yield and 98% ee.26a Molecular hydrogen (1 atm) was almost equally effective for the transformation.26b A broad range of enol acetates were prepared from ketones and were successfully transformed into their corresponding (R-acetates under 1 atm H2 (Table 19). From unsymmetrical aliphatic ketones, enol acetates were obtained as the mixtures of regio- and geometrical isomers. Notably, however, the efficiency of the process was little affected by the isomeric composition of the enol acetates.
76
Mahn-Joo Kim et al.
Table 18 Asymmetric transformations of acyloxyphenyl ketones CALB 1
O RCOO
HO
H2 70°C
Substrate
Product
R OCOR
O
HO
RCOO
OCOR
O
ee (%)
Yield (%)
Me Pr
98 98
95 96
Me Pr
93 96
94 93
Me Pr
89 98
88 88
Me Pr
96 98
92 89
OH
RCOO
OCOR
O
RCOO
OH OCOR
O
RCOO
OCOR
HO
3.3. Asymmetric transformation of ketoxime The strategy for the asymmetric reductive acylation of ketones was extended to ketoximes (Scheme 9). The asymmetric reactions of ketoximes were performed with CALB and Pd/C in the presence of hydrogen, diisopropylethylamine, and ethyl acetate in toluene at 60 C for 5 days (Table 20).28 In comparison to the direct DKR of amines, the yields of chiral amides increased significantly. Diisopropy lethylamine was responsible for the increase in yields. However, the major factor would be the slow generation of amines, which maintains the amine concentration low enough to suppress side reactions including the reductive amination. Disap pointingly, this process is limited to benzylic amines. Additionally, low turnover frequencies also need to be overcome.
Catalysis by enzyme–metal combinations
77
OAc
OAc
R
R Lipase
Ac-lipase O
OH
Ru R
[H]
OH
Ru
R
R
Scheme 8: Asymmetric reduction of enol ester. Table 19 Asymmetric hydrogenation of enol acetates CALB 1
OAc R′
R
OAc
H2 or 2,6-dimethyl-4-heptanol
R
R′
R
R
ee (%)
Yield (%)
Ph 4-MeO-Ph 4-Cl-Ph 1-Indanyl c-C6 H11 PhCH2 PhCH2 CH2 CH3 (CH2 4 CH2
Me Me Me
98 98 97 99 99 79 94 91
89 80 91 87 94 90 92 95
Me Me Me Me
NH2 R1 NOH R1
R2
R1
R NHAc Lipase
Pd Pd
[H]
AcOR
NH2 R1
R1
S
Scheme 9: Asymmetric reductive acetylation of ketoxime.
R1
R2
78
Mahn-Joo Kim et al.
Table 20
Asymmetric reductive acetylation of ketoximes
N R
OH
CALB Pd/C
R′
NHAc R
H2, EtOAc (i-Pr)2NEt
toluene, 60°C
R′
R
R
ee (%)
Yield (%)
Ph 3-Me-Ph 4-Me-Ph 4-MeO-Ph Ph 1-Indanyl -Tetralyl 4-Chromanyl
Me Me Me Me Et
98 94 97 96 98 95 97 99
80 81 84 82 76 84 70 89
4. CONCLUSION This chapter dealt with catalysis by enzyme–metal combinations as the novel tool for the conversion of racemic substrates to single enantiomeric products. Despite the inherent complexity of enzyme–metal catalysis for ready access to enantiopure products, rapid progress has been made in the last few years. A wide range of substrates were converted into their respective enantiopure products under DKR and asymmetric transformation conditions. New metal catalyst systems compatible with enzymes were discovered and more importantly the transformations were achieved at ambient temperatures. The focus in future should be directed towards expanding the horizon of substrates and finding applications in the commercial manufacture of specific targets. New strategies for the development of recyclable catalyst systems are required for easy work-up procedures so that DKR becomes not only scalable but also an economically viable option and changes the way of manufacturing chiral molecules in future. It is heartening to note that a recent enantioselective synthesis of lamivudine (3TCTM , a known anti-HIV, was accom plished via DKR method.29
ACKNOWLEDGEMENTS We would like to thank Yoon Kyung Choi for her suggestions and corrections. Generous supports by KOSEF and KRF are also acknowledged.
Catalysis by enzyme–metal combinations
79
REFERENCES 1. Sheldon, R. A. Chirotechnology, Industrial Synthesis of Optically Active Compounds; Marcel Dekker: New York, 1993. 2. (a) Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic Chemistry; Pergamon: Oxford, UK, 1994. (b) Koskinen, A. M. P.; Klibanov, A. M. Enzymatic Reactions in Organic Media; Blackie Academic & Professional: Glasgow, Scotland, 1996. (c) Faber, K. Biotransformations in Organic Chemistry, 3rd edn; Springer: Berlin, 1997. (d) Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis; Wiley-VCH: Weilheim, 1999. (e) Deauz, K.; Waldmann, H. Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook, 2nd edn; Wiley-VCH: Weinheim; Vols I–III, 2002. 3. (a) Kim, M.-J.; Choi, G. –B.; Kim, H.-J. Tetrahedron Lett. 1995, 36, 6253. (b) Kim, M.-J.; Lim, I.-T. Synlett 1996, 2, 138. (c) Kim, M.-J.; Lim, I.-T.; Choi, G.-B.; Whang, S.-Y.; Ku, B.-C.; Choi, J.-Y. Bioorg. Med. Chem. Lett. 1996, 6, 71. (d) Kim, M.-J.; Lim, I.-T.; Kim, H.-J.; Wong, C.-H. Tetrahedron Asymm. 1997, 8, 1507. (e) Lee, D.; Kim, M.-J. Tetrahedron Lett. 1998, 39, 2163. (f) Chung, S.-K.; Chang, Y.-T.; Lee, E. J.; Shin, B.-G.; Kwon, Y.-U.; Kim, K.-C.; Lee, D.; Kim, M.-J. Bioorg. Med. Chem. Lett. 1998, 8, 1503. (g) Lee, D.; Kim, M.-J. Tetrahedron Lett. 1999, 39, 9039. (h) Lee, D.; Kim, M.-J. Org. Lett. 1999, 1, 925. (i) Im, A. S.; Cheong, C. S.; Lee, S. H. Bull. Korean Chem. Soc. 2003, 24, 1269. (j) Kang, H.-Y.; Ji, Y.; Yu, Y.-K.; Yu, J.-Y.; Lee, Y.; Lee, S.-J. Bull. Korean Chem. Soc. 2003, 24, 1819. 4. Ward, R. S. Tetrahedron Asymm. 1995, 6, 1475. 5. (a) Stürmer, R. Angew. Chem. Int. Ed. Engl. 1997, 36, 1173. (b) Azerad, R.; Buisson, D. Curr. Opin. Biotechnol. 2000, 11, 565. (c) Huerta, F. F.; Minidis, A. B. E.; Bäckvall, J.-E.Chem. Soc. Rev. 2001, 30, 321. (d) Kim, M.-J.; Ahn, Y.; Park, J. Curr. Opin. Biotechnol. 2002, 13, 578. (e) Pellissier, H.Tetrahedron 2004, 59, 8291. (f) Pámies, O.; Bäckvall, J.-E.Chem. Soc. Rev. 2003, 103, 3247. (g) Pámies, O.; Bäckvall, J.-E. Curr. Opin. Biotechnol. 2003, 14, 407. (h) Turner, N. J. Curr. Opin. Chem. Biol. 2004, 8, 114. 6. Allen, J. V.; Williams, J. M. J. Tetrahedron Lett. 1996, 37, 1859. 7. Dinh, P. M.; Howarth, J. A.; Hudnott, A. R.; Williams, J. M. J.; Harries, W. Tetrahedron Lett. 1996, 37, 7623. 8. Reetz, M. T.; Schimossek, K. Chimia 1996, 50, 668. 9. Choi, J. H.; Kim, Y. H.; Nam, S. H.; Shin, S. T.; Kim, M.-J.; Park, J. Angew. Chem. Int. Ed. Engl. 2002, 41, 2373. 10. (a) Larsson, A. L. E.; Persson, B. A.; Bäckvall, J.-E. Angew. Chem. Int. Ed. Engl. 1997, 36, 1211. (b) Persson, B. A.; Larsson, A. L. E.; Ray, M. L.; Bäckvall, J.-E. J. Am. Chem. Soc. 1999, 121, 1645. (c) Lee, H. K.; Ahn, Y. Bull. Korean Chem. Soc. 2004, 25, 1471. 11. Koh, J. H.; Jeong, H. M.; Kim, M.-J.; Park, J. Tetrahedron Lett. 1999, 40, 6281. 12. Lee, D.; Huh, E. A.; Kim, M.-J.; Jung, H. M.; Koh, J. H.; Park, J. Org. Lett. 2000, 2, 2377. 13. (a) Kim, K. W.; Song, B.; Choi, M. Y.; Kim, M.-J. Org. Lett. 2001, 3, 1507. (b) Lee, J. K.; Kim, M.-J. J. Org. Chem. 2002, 67, 6845. (c) Kim, M.-J.; Choi, M. Y.; Lee, J. K.; Ahn, Y. J. Mol. Catal. B Enzym. 2003, 26, 115. (d) Erbeldinger, M.; Mesiano, A. J.; Russel, A. Biotechnol. Prog. 2000, 16, 1129. (e) Lau, R. M.; van Rantwijk, F.; Seddon, K. R.; Sheldon, R. A. Org. Lett. 2000, 2, 4189. (f) Itoh, T.; Akasaki, E.; Kudo, K.; Shirakami, S. Chem. Lett. 2001, 262. (g) Schoefer, S. H.; Kraftzik, N.; Wasserscheid, P.; Kragl, U. Chem. Commun. 2001, 425. (h) Park, S.; Kazlauskas, R. J. Org. Chem. 2001, 66, 8395. 14. Kim, M.-J.; Kim, H. M.; Kim, D.; Ahn, Y.; Park, J. Green Chem. 2004, 6, 471. 15. Choi, J. H.; Choi, Y. K.; Kim, Y. H.; Park, E. S.; Kim, E. J.; Kim, M.-J.; Park, J. J. Org. Chem. 2004, 69, 1972. 16. (a) Csjernyik, G.; Bogár, K.; Bäckvall, J.-E. Tetrahedron Lett. 2004, 45, 6799. (b) Martin-Matute, B.; Edin, M.; Bogár, K.; Bäckvall, J.-E. Angew. Chem. Int. Ed. Engl.2004, 43, 6535.
80
Mahn-Joo Kim et al.
17. Dijksman, A.; Elzinga, J. M.; Li, Y. X.; Arends, W. C. E.; Sheldon, R. A. Tetrahedron Asymm. 2002, 13, 879. 18. Persson, B. A.; Huerta, F. F.; Bäckvall, J.-E. J. Org. Chem. 1999, 64, 5237. (b) Edin, M.; Bäckvall, J.-E. J. Org. Chem. 2003, 68, 2216. 19. (a) Huerta, F. F.; Laxmi, S. Y. R.; Bäckvall, J.-E. Org. Lett. 2002, 2, 1037. (b) Runmo, A. B. L.; Pámies, O.; Faber, K.; Bäckvall, J.-E. Tetrahedron Lett. 2002, 43, 2983. (c) Pámies, O.; Bäckvall, J.-E. J. Org. Chem. 2002, 67, 1261. (d) Pámies, O.; Bäckvall, J.-E. J. Org. Chem. 2001, 66, 4022. 20. Kim, M.-J.; Choi, Y. K.; Choi, M. Y.; Kim, M.; Park, J. J. Org. Chem. 2001, 66, 4736. 21. (a) Pámies, O.; Bäckvall, J.-E. Adv. Synth. Catal. 2001, 343, 726. (b) Pámies, O.; Bäckvall, J.-E. J. Org. Chem. 2002, 67, 9006. (c) Pámies, O.; Bäckvall, J.-E. J. Org. Chem. 2003, 68, 4815. 22. Kim, M.-J.; Chung, Y. I.; Choi, Y. K.; Lee, H. K.; Kim, D.; Park, J. J. Am. Chem. Soc. 2003, 125, 11494. 23. Choi, Y. K.; Suh, J. H.; Lee, D.; Lim, I.-T.; Jung, J. Y.; Kim, M.-J. J. Org. Chem. 1999, 64, 8423. 24. Parvulescu, A.; Vos, D. D.; Jacobs, P. Chem. Commun. 2005, 5307. 25. (a) Pámies, O.; Ell, A. H.; Samec, J. S. M.; Hermanns, N.; Bäckvall, J.-E. Tetrahedron Lett. 2002, 43, 4699. (b) Paetzold, J.; Bäckvall, J.-E. J. Am. Chem. Soc. 2005, 127, 17620. 26. (a) Jung, H. M.; Koh, J. H.; Kim, M.-J.; Park, J. Org. Lett. 2000, 2, 409. (b) Jung, H. M.; Koh, J. H.; Kim, M.-J.; Park, J. Org. Lett. 2000, 2, 2487. 27. Kim, M.-J.; Choi, M. Y.; Han, M. Y.; Choi, Y. K.; Lee, J. K.; Park, J. J. Org. Chem. 2002, 67, 9481. 28. Choi, Y. K.; Kim, M.; Ahn, Y.; Kim, M.-J. Org. Lett. 2001, 3, 4099. 29. Goodyear, M. D.; Hill, M. L.; West, J. P.; Whitehead, A. J. Tetrahedron Lett. 2005, 46, 8535.
Part Two Uncomon kind of biocatalytic reaction
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
83
Chapter 5
Biological Kolbe–Schmitt carboxylation Possible use of enzymes for the direct carboxylation of organic substrates Toyokazu Yoshida, Toru Nagasawa Department of Biomolecular Science, Faculty of Engineering, Gifu University, Yanagido, Gifu, Japan
Abstract The occurrence of the enzymes catalyzing the Kolbe-Schmitt carboxylation has been shown, and the characterization of these enzymes has been reviewed. Among them, 2,6-dihydroxybenzoate decarboxylase and pyrrole-2-carboxylate decarboxylase efficiently catalyze the reverse carboxylation reaction and accumulate high concentration of 2,6 dihydroxybenzoate from 1,3-dihydroxybenzene and pyrrole-2-carboxylate from pyrrole, respectively, in the presence of 3 M KHCO3 . The possible application of biological Kolbe-Schmitt carboxylation has been discussed. Primary structure analysis of the decar boxylases suggests the latent distribution of various unidentified enzymes, probably cat alyzing the regioselective carboxylation of aromatic and heterocyclic compounds, in microorganisms.
1. INTRODUCTION Biocatalysis is now one of the most powerful and indispensable tools for organic synthesis due to its environment friendliness and excellent enantio-, regio-, and chemoselectivities. Due to the abundance of CO2 as the major greenhouse gas and C−C bond-forming properties, CO2 -fixing enzymes are of high interest. The quantitatively most important enzyme of this kind is ribulose-1,5-bisphosphate carboxylase catalyzing the central CO2 fixation in all photosynthetic organisms.1 Carboxylation reactions widely occur in nature by the direct use of CO2 or HCO3 − and are mediated by enzymes. Aryl carboxylases are less often described, and in microorganisms they are mainly found in the anaerobic degradation of pheno lic compounds via benzoic acids.2−11 C−C bond-forming enzymes are of inter est for preparative organic chemistry. There are only few applications of such direct enzymatic carboxylation reactions for synthetic purposes at industrial level.
84
Toyokazu Yoshida, Toru Nagasawa
The carboxylation reaction of alkali metal phenoxides with carbon diox ide to hydroxybenzoic acids is known as the Kolbe-Schmitt reaction.1213 The Kolbe-Schmitt reaction proceeds at high temperatures and pressures (100–200 C, 0.1–3 MPa) and has been used for the industrial production of aryl hydrox ycarboxylic acids such as salicylate, 4-hydroxybenzoic acid, 4-aminosalicylic acid, 3- or 6-hydroxy-2-naphthoic acid, etc., for a long time.14 These aryl hydroxycarboxylic acids have been widely used for the production of cosmet ics, pharmaceuticals, food additives, polymer liquid crystals, etc. Despite the fact that Kolbe-Schmitt reaction has been known and exploited for a very long time, the selectivity and yield are not yet completely mastered, and the production of either pure isomer is not easily achieved. In fact, purification process is needed after the synthesis, which is energy-consuming and wasteproducing. Owing to the value of the chemicals produced, it is highly wished to conduct a selective process working under mild conditions. For the devel opment of environment-friendly syntheses, the use of enzymes for the direct carboxylation of organic substrates needs more consideration. Aryl carboxy lases/decarboxylases, which catalyze the synthesis/degradation of aryl carboxylic acids, are characteristic of the region-specificity under mild conditions. In the present review, we have focused on aryl carboxylases/decarboxylases catalyzing CO2 -fixation reaction to explore the possible application of biological KolbeSchmitt carboxylation.
2. ENZYMES CATALYZING THE CARBOXYLATION OF PHENOLIC COMPOUNDS Phenol is an important intermediate in the anaerobic degradation of many complex and simple aromatic compounds. Tschech and Fuchs15 proposed that the carboxy lation of phenol to 4-hydroxybenzoate is the first step in the degradation of phenol under denitrifying conditions. However, 4-hydroxybenzoate is not detected in the cultures or cell extracts of the denitrifying Pseudomonas species in the presence of CO2 and phenol,8 but it is detected if phenol is replaced by phenolphosphate.16 In contrast, 4-hydroxybenzoate is readily detected as an intermediate of phenol degradation in the iron-reducing bacterium GS-15,17 and 4-hydroxybenzoate may prove to be a common intermediate in the anaerobic transformation. Thus, in anaer obic degradation of phenolic compounds, it has been postulated that carboxylation reactions may play important roles. Zhang et al. isolated Clostridium hydroxybenzoicum containing two inducible 4-hydroxybenzoate decarboxylase and 3,4-dihydroxybenzoate decar boxylase that form phenol and catechol (1,2-dihydroxybenzene), respectively.1819 The organism does not further metabolize phenol and catechol produced by these reactions. The carboxylation activities of the two purified decarboxylases are not
Biological Kolbe–Schmitt carboxylation
85
high in vitro, but these decarboxylases possibly play significant physiological roles in the anaerobic metabolism of phenol and catechol.2021 Pure cultures growing anaerobically with catechol and sulfate were iso lated,2223 and the carboxylation of catechol was proposed to be the initial reaction of anaerobic catechol degradation by Desulfobacterium sp. strain Cat2.24 Zhang and Young11 proposed that the initial key reaction for anaerobic degradation of naphthalene and phenanthrene was also carboxylation. Studies on 4-hydroxybenzoate decarboxylase and 3,4-hydroxybenzoate decarboxylase have been restricted to obligate anaerobic bacteria, C. hydroxy benzoicum.2021 Aside from the obligate anaerobic microorganism, C. hydroxy benzoicum, very recently facultative anaerobic bacteria, Enterobacter cloacae strains exhibiting high 4-hydroxybenzoate decarboxylase or 3,4-dihydroxybenzoate decarboxylase activity, have been isolated.2526 In this section, in addition to 4-hydroxybenzoate decarboxylase and 3,4-dihydroxybenzoate decarboxylase, three kinds of enzymes catalyzing the car boxylation of phenolic compounds are focused. 2.1. 4-Hydroxybenzoate decarboxylase (EC 4.1.1.61)
Scheme 1
4-Hydroxybenzoate decarboxylase (EC 4.1.1.61) of anaerobe C. hydroxyben zoicum was purified and characterized for the first time.20 It has an apparent molecular mass of 350 kDa and consists of six identical subunits of 57 kDa. The temperature optimum for the decarboxylation is approximately 50 C, the optimum pH being 5.6–6.2. The activation energy for decarboxylation of 4-hydroxybenzoate is 65 kJ mol−1 (20–37 C). The enzyme also catalyzes the decarboxylation of 3,4-dihydroxybenzoate. The apparent Km and kcat values for 4-hydroxybenzoate are 0.40 mM and 33 × 103 min−1 , and for 3,4-dihydroxybenzoate 1.2 mM and 11 × 103 min−1 , respectively, at pH 6.0 and 25 C. The enzyme activity was not influenced by the addition of biotin or avidin. Secondly, 4-hydroxybenzoate decar boxylase from an anaerobic co-culture, consisting of Clostridium-like strain and an unidentified strain, was purified and characterized.27 The occurrence of 4-hydroxybenzoate decarboxylase was also found in fac ultative anaerobic bacteria, E. cloacae P240, and the enzyme was purified and characterized.25 The activity of the cell-free extract of E. cloacae P240 was deter mined to be 137 mol min−1 (mg protein)−1 at 30 C, which was much higher than
86
Toyokazu Yoshida, Toru Nagasawa
that of C. hydroxybenzoicum, 033 mol min−1 (mg protein)−1 at 25 C.20 The puri fied enzyme of E. cloacae P240 is a homohexamer of identical 60 kDa subunits, which is in good agreement with that of C. hydroxybenzoicum. The properties of 4-hydroxybenzoate decarboxylase from E. cloacae P240 are similar to those of Clostridium strains in optimum temperature and pH, oxygen sensitivity, and substrate specificity. 4-Hydroxybenzoate decarboxylase catalyzes the reverse reactions, that is, the carboxylation of phenol to 4-hydroxybenzoate and of catechol to 3,4 dihydroxybenzoate.2025 Using the E. cloacae P240 purified enzyme, the carboxy lation reaction of phenol was investigated in a tightly sealed reaction vessel to avoid the leakage of CO2 gas. The reaction mixture contained 20 mM phenol, 3 M KHCO3 , 10 mM dithiothreitol, 100 mM potassium phosphate buffer (pH 7.0), and enzyme solution (or whole cells). The reaction was started by the addition of KHCO3 and incubated at 20 C. The decarboxylase catalyzed the reverse car boxylation reaction of 20 mM phenol to form 3.8 mM 4-hydroxybenzoate with a molar conversion yield of 19% (Fig. 1). The Km value for phenol was calcu lated to be 14.8 mM. The conversion ratio of phenol to 4-hydroxybezoate by C. hydroxybenzoicum decarboxylase was 0.5% in the presence of 3.3 mM phenol and 100 mM NaHCO3 by 2 h incubation; 17 M 4-hydroxybenzoate was formed.20 The difference in molar conversion ratios might be caused by the concentration of bicar bonate to the reaction mixture. The reversible conversion of 4-hydroxybenzoate and phenol was also observed by whole-cell suspensions. Assuming that HCO3 − is the co-substrate, the equilibrium constant calculated for the reaction 4 hydroxybenzoate− + H2 O � phenol + HCO−3 was 11.4 (± 0.5). The equilibrium constants explain the low conversion of phenol into 4-hydroxybenzoate.28
Concentration (mM)
20 15 10
5 0
0
50
100 150 Time (min)
200
Figure 1: Time course of carboxylation reaction of phenol. Closed circles,
4-hydroxybenzoate; open squares, phenol.
Biological Kolbe–Schmitt carboxylation
87
2.2. 3,4-Dihydroxybenzoate decarboxylase (EC 4.1.1.63)
Scheme 2 A 3,4-dihydroxybenzoate decarboxylase (EC 4.1.1.63) was purified from C. hydroxybenzoicum and characterized for the first time.21 The estimated molecular mass of the enzyme is 270 kDa. The subunit molecular mass is 57 kDa, sug gesting that the enzyme consists of five identical subunits. The temperature and pH optima are 50 C and pH 7.0, respectively. The Arrhenius energy for decar boxylation of 3,4-dihydroxybenzoate was 325 kJ mol−1 for the temperature range from 22 to 50 C. The Km and kcat for 3,4-dihydroxybenzoate were 0.6 mM and 54 × 103 min−1 , respectively, at pH 7.0 and 25 C. The enzyme catalyzes the reverse reaction, that is, the carboxylation of catechol to 3,4-dihydroxybenzoate, at pH 7.0. The enzyme does not decarboxylate 4-hydroxybenzoate. Although the equilibrium of the reaction is on the side of catechol, it is postulated that C. hydrox ybenzoicum uses the enzyme to convert catechol to 3,4-dihydroxybenzoate.21 The occurrence of 3,4-dihydroxybenzoate decarboxylase was also found widely in facultative anaerobes. Among them, Enterobacter cloacae P241 showed the highest activity of 3,4-hydroxybenzoate decarboxylase,26 and the activity of the cell-free extract of E. cloacae P241 was determined to be 0629 mol min−1 (mg protein)−1 at 30 C, which was more than that of C. hydroxybenzoicum, 011 mol min−1 (mg protein)−1 at 25 C.21 The E. cloacae P241 enzyme has a molecular mass of 334 kDa and consists of six identical 50 kDa subunits. The Km value for 3,4-dihydroxybenzoate was 177 M. The enzyme is also characteristic of its narrow substrate specificity and does not act on 4-hydroxybenzoate and other benzoate derivatives. The properties of E. cloacae P241 3,4-hydroxybenzoate decarboxylase were similar to those of C. hydroxybenzoicum in optimum temper ature and pH, oxygen sensitivity, and substrate specificity. The carboxylation reaction of catechol by E. cloacae P241 was investigated in a tightly sealed vessel to avoid the leakage of CO2 gas.26 The reaction mixture contained 25 mM catechol, 3 M KHCO3 , 10 mM dithiothreitol, 100 mM potassium phosphate buffer (pH 7.0), and enzyme solution. The reaction was started by the addition of KHCO3 , incubated at 20 C. The carboxylation for catechol was also observed using the purified enzyme and the whole-cell suspension of E. cloacae P241 in the presence of 25 mM catechol and 3 M KHCO3 , resulting in the formation of 5.49 mM 3,4-dihydroxybenzoate with a molar conversion ratio of 22% (Fig. 2).
88
Toyokazu Yoshida, Toru Nagasawa
Concentration (mM)
25 20 15 10 5 0
0
60 120 Time (min)
180
Figure 2: Time course of carboxylation reaction of catechol. Closed circles, 3,4-dihydroxybenzoate; open squares, catechol. The reaction product of the reserve carboxylation reaction was isolated and identified to be 3,4-dihydroxybenzoic acid by 1 H NMR and 13 C NMR with the authentic 3,4-dihydroxybenzoic acid as a reference. The carboxylation reaction of catechol to 3,4-dihydroxybenzoate was affected by the concentration of KHCO3 . The carboxylation activity of E. cloacae P241 3,4-dihydroxybenzoate decarboxy lase in the presence of 0.1 M KHCO3 was only 15% of that in the presence of 3 M KHCO3 .26 In the case of C. hydroxybenzoicum 3,4-dihydroxybenzote decarboxy lase, only 0.01 mM 3,4-dihydroxybenzoate was formed from 6 mM catechol in the presence of 50 mM NaHCO3 by 40 min incubation.21 The difference in molar conversion ratios might be caused by the concentration of bicarbonate added to the reaction mixture. 2.3. Phenolphosphate carboxylase (EC 4.1.1.-) in Thauera aromatica COO– CO2 + H2O
ATP + H2O
Pi
AMP + Pi OH
OPO3–
OH
Scheme 3 The anaerobic metabolism of phenol in T. aromatica is initiated by the ATP-
dependent conversion of phenol to phenylphosphate catalyzed by phenylphosphate
Biological Kolbe–Schmitt carboxylation
89
synthase (EC 2.7.9.-). The subsequent para carboxylation of phenylphosphate to 4-hydroxybenzoate is catalyzed by phenolphosphate carboxylase (EC 4.1.1.-).82930 Both enzyme activities are induced in cells grown anoxially on phenol and nitrate and not in cells grown on 4-hydroxybenzoate, the product of this process. Further metabolism of 4-hydroxybenzoate proceeds via benzoyl coenzyme A. Phenylphos phate synthase and phenylphosphate carboxylase were purified and characterized from T. aromatica.831−33 Phenylphosphate synthase consists of three subunits with molecular masses of 70, 40, and 24 kDa. Subunit 1 resembles the central part of classical phospho enolpyruvate synthase which contains a conserved histidine residue. It catalyzes the exchange of free [14 C] phenol and the phenol moiety of phenylphosphate but not the phosphorylation of phenol. Phosphorylation of phenol requires subunit 1, MgATP, and another protein, subunit 2 (40 kDa), which resembles the N-terminal part of phosphoenolpyruvate synthase. Subunit 1 and 2 catalyze the following reaction: phenol + MgATP + H2 O → phenylphosphate + MgAMP + orthophosphate The phosphoryl group in phenylphosphate is derived from the -phosphate group of ATP. The free energy of ATP hydrolysis obviously favors the trapping of phe nol (Km , 0.04 mM), even at a low ambient substrate concentration. The reaction is stimulated several fold by another protein, subunit 3 (24 kDa). The molecular and catalytic features of phenylphosphate synthase resemble those of phospho enolpyruvate synthase, albeit with interesting modifications.33 Phenylphosphate is converted into 4-hydroxybenzoate by phenylphosphate carboxylase. This enzyme consists of four proteins with molecular masses of 54, 53, 18, and 10 kDa. Three of the subunits ( 54, 53, and 10 kDa) were sufficient to catalyze the exchange of 14 CO2 and the carboxyl group of 4-hydroxybenzoate but not phenylphosphate carboxylation. Phenylphosphate carboxylation was restored when the 18 kDa subunit was added. As shown in the reaction model of Fig. 3, the 14 CO2 exchange reaction catalyzed by the three subunits of the core enzyme requires the fully reversible release of CO2 from 4-hydroxybenzoate with formation of a tightly enzyme-bound phenolate intermediate. Carboxylation of phenylphos phate requires the addition of subunit, which is thought to form the same enzyme-bound energized phenolate intermediate from phenylphosphate with vir tually irreversible release of phosphate. Then , , and subunits catalyze the carboxylation of the enzyme-bound energized phenolate intermediate to produce 4-hydroxybenzoate. Phenylphosphate carboxylase acts on phenolic compounds, uses CO2 as a substrate, does not contain biotin or thiamine diphosphate, requires K+ and a divalent metal cation (Mg2+ or Mn2+ ) for activity, and are strongly inhibited by oxygen. The purified E. cloacae P240 4-hydroxybenzoate decarboxylase did not catalyze the carboxylation of phenylphosphate,25 indicating phenolphosphate
90
Toyokazu Yoshida, Toru Nagasawa
COO– Phenol exchange reaction (Subunit 1)
+
Subunit 1
H2O
PO32–
Pi OPO3– Phenylphosphate O– hydrolase
OH Pi
+ H2O
δ -Subunit
ATP H –
Subunit 1
OH
PO3
–
2–
Subunit 1
PO2
AMP Subunit 2 Phosphorylation of subunit 1
H CO2
COO–
CO2 CO2 exchange reaction
α β γ - Subunit Phenylphosphate synthase
Phenylphosphate carboxylase
Figure 3: Possible reaction mechanism of carboxylation of phenol via phenylphosphate.
carboxylase (EC 4.1.1.-) was distinguishable from 4-hydroxybenzoate decarboxy lases (EC 4.1.1.61). Based on the anaerobic metabolism of phenol by T. aromatica, Aresta et al. studied the production of 4-hydroxybenzoic acid from the technological aspect.3435 The crude extract was partially purified by using membranes with a given molecular weight cut-off (MWCO). The enzyme was supported on a solid matrix (agar) to stabilize its activity. The best solid support was revealed to be low-melting agar on which the enzyme had a life of several days to week. In the membrane reactor, as shown in Fig. 4, phenylphosphate was incubated with the enzyme under CO2 atmosphere and the 4-hydroxybenzoic acid accumulated.35 Once all the phosphorylated phenol has been converted, the solution is recovered, and more substrate was added. Membrane was used for the separation of the reaction space where the enzyme is kept, from the space where the reaction product is extracted. The enzyme does not cross the membrane that has a MWCO of 100 kDa. The 4-hydroxybenzoic acid produced in the central reactor, space B, crossed the membrane and was collected in the external part C from where it can be withdrawn with a syringe or downloaded through the stopcock. The lifetime of the enzyme is of several days at room temperature under atmospheric pressure of carbon dioxide that can be used as reaction gas. The process is clean and good results are obtained with a turnover number of approximately 16 000.
Biological Kolbe–Schmitt carboxylation
91
Reagent + cosubstrate
A
B C
Product
Figure 4: Membrane reactor for 4-hydroxybenzoate production using phenolphosphate carboxylase. A membrane (A) separates the reaction space containing the enzyme (B) from water phase where the product is collected (C).
2.4. 2,6-Dihydroxybenzoate decarboxylase OH
OH
COOH
+
OH
CO2
OH
Scheme 4
A novel decarboxylase, 2,6-dihydroxybenzoate decarboxylase, was found in Agrobacterium tumefaciens IAM 12048 at first.36 Thereafter, the same activity was found in Rhizobium species by two groups independently.3738 Furthermore, Pandoraea sp. 12B-2, the most powerful producer of 2,6-dihydroxybenzoate decar boxylase, was isolated.39 These enzymes have been purified and characterized. 2,6-Dihydroxybenzoate decarboxylase activity of these bacteria was induced specifically by 2,6-dihydroxybenzoate. The enzyme activity in a cell-free extract of A. tumefaciens IAM 12048 was stable during storage at 4 C for 7 days in potassium phosphate buffer (pH 7.0) containing 1 mM dithiothreitol. Dif ferent from 4-hydroxybenzoate decarboxylase and 3,4-dihydroxybenzoate decar boxylase, 2,6-dihydroxybenzoate decarboxylase was much less labile and barely
92
Toyokazu Yoshida, Toru Nagasawa
exhibited the sensitivity to oxygen. Therefore, the enzyme purification was per formed with relatively high yield. The purified enzyme of A. tumefaciens IAM 12048 was a homotetramer of identical 38 kDa subunits. The enzyme showed no characteristic absorption maximum other than 280 nm, indicating the absence of a prosthetic group such as pyridoxal 5 -phosphate and thiamine pyrophos phate. The enzyme activity was not diminished after the incubation at 50 C for 30 min. Treatment at 60 and 70 C caused 18 and 100% losses of initial activ ity, respectively. The enzyme was stable on incubation at 30 C for 20 min in the pH range of 5.0–11.0. When the decarboxylation reaction was carried out for 20 min at various temperatures, the activity was maximal at 60 C. The high est activity was observed at pH 8.5 (10 mM Tris–HCl buffer) and the activities under pH 5.0–10.0 were over 80% of that at pH 8.5. 2,6-Dihydroxybenzoate decarboxylases of A. tumefaciens IAM 12048 and Rhizobium species had similar properties. The purified enzyme of A. tumefaciens IAM 12048 catalyzed the regiose lective carboxylation of 1,3-dihydroxybenzene into 2,6-dihydroxybenzoate in the presence of 15 mM 1,3-dihydroxybenzene and 3 M KHCO3 , with a molar conver sion yield of 30%. The Rhizobium enzymes were also confirmed to catalyze the reverse carboxylation. The regioselective carboxylation of 1,3-dihydroxybenzene was optimized using the whole cells of Pandoraea sp. 12B-2. The carboxyla tion reaction was carried out at 30 C in a tightly sealed reaction vessel to avoid the leakage of CO2 gas. The standard reaction mixture contained 12.5 mM 1,3 dihydroxybenzene, 3 M KHCO3 , 100 mM potassium phosphate buffer (pH 7.0), and whole cells (22.0 mg dry weight per 1 ml of reaction mixture). The effect of the KHCO3 concentration on the production of 2,6 dihydroxybenzene was investigated.39 In the presence of 0.1 M KHCO3 , 0.5 mM 2,6-dihydroxybenzoate was formed from 12.5 mM 1,3-dihydroxybenzene. The addition of 1–3 M KHCO3 in the reaction mixture significantly promoted the formation of 2,6-dihydroxybenzoate (Fig. 5a). With 3 M KHCO3 , the molar conversion ratio of 1,3-dihydroxybenzene reached its highest value (48%). The effect of temperature on 2,6-dihydroxybenzoate formation was exam ined (Fig. 5b). When the reaction was carried out for 6 h at 40 or 50 C, the initial velocity of 2,6-dihydroxybenzoate formation was accelerated. How ever, the prolonged incubation over 40 C resulted in the degradation of 2,6 dihydroxybenzoate. To enhance the productivity of 2,6-dihydroxybenzoate, the addition of various kinds of organic solvent was tested. Among them, the addition of 10–20% (v/v) acetone accelerated the initial velocity of 2,6 dihydroxybenzoate formation (Fig. 5c); however, when the equilibrium was achieved, the final molar conversion yield of 1,3-hydroxybenzene reached the same level as in the case without adding acetone. The addition of acetone decreased the viscosity of the whole cells of Pandoraea sp. 12B-2 in the reaction mixture.
2,6-Dihydroxybenzoate (mM)
2,6-Dihydroxybenzoate (mM)
Biological Kolbe–Schmitt carboxylation
6 5 4 3 2 1 0
0
10
20
30
50 40 30 20 10 0
0
20
Time (h)
2,6-Dihydroxybenzoate (mM)
93
40
60
80
Time (h)
100 80 60 40 20 0
0
20
40
60
80
Time (h)
Figure 5: Effect of KHCO3 concentraion (a), temperature (b), and acetone (c) on the carboxylation of 1,3-dihydroxybenzene. (a) The following concentrations of KHCO3 were added in the reaction mixture: open triangles, 0.1 M; closed diamonds, 1 M; open squares, 2 M; closed circles, 3 M. (b) Reactions were carried out using 100 mM 1,3-dihydroxybenzene at 20 C open triangles; 30 C closed circles; 35 C closed diamonds; 40 C open squares. (c) Reactions were carried out using 200 mM 1,3-dihydroxybenzene. Acetone was added at the following concentrations: open squares, 0% (v/v); closed circles 10% (v/v); closed diamonds 20% (v/v).
When the whole-cell reaction was carried out in the presence of 0.1–3 M 1,3-dihydroxybenzene and 3 M KHCO3 , the molar conversion ratio of 1,3-dihydroxybenzene always reached 45–50% (Fig. 6). The high concentration of 1,3-dihydroxybenzene was not inhibitory on 2,6-dihydroxybenzoate decar boxylase. When 3 M 1,3-dihydroxybenzene was incubated with the whole cells of Pandoraea sp. 12B-2 (43.9 mg as dry cell weight) in the presence of 3 M KHCO3 , 220 mg ml−1 of 2,6-dihydroxybenzoate (1.42 M) accumulated after 120 h incubation, with a conversion ratio of 48%. During the carboxylation of 1,3 dihydroxybenzene, no other product except for 2,6-dihydroxybenzoate was formed.
Toyokazu Yoshida, Toru Nagasawa
2,6-Dihydroxybenzoate (mM)
94
1500 1200 900 600 300 0
0
50
100 150 Time (h)
200
250
Figure 6: 2,6-Dihydroxybenzoate production using whole cells of Pandoraea sp. 12B-2. Reactions were carried out in the reaction mixture containing the following concentrations of 1,3-dihydroxybenzne: closed circles, 3 M; open squares, 2 M; closed diamonds, 1 M; open triangles, 0.1 M.
2,3-Dihydroxybenzoate (mM)
The product was isolated and identified by 1 H NMR and 13 C NMR analyses comparing with the authentic 2,6-dihydroxybenzoic acid as a reference. Carboxylation of 20–300 mM 1,2-dihydroxybenzene was carried out using 36.9 mg (as dry cell weight) of whole cells in the presence of 3 M KHCO3 in 1 ml of the reaction mixture. The molar conversion yields were almost the same using 20, 100, and 200 mM 1,2-dihydroxybenzene (approximately 25%) as shown in Fig. 7.
50 40 30 20 10 0
0
6
12 18 Time (min)
24
Figure 7: Time course of carboxylation of 1,2-dihydroxybenzene using whole
cells of Pandoraea sp. 12B-2. Reactions were carried out in the reaction mixture
containing the following concentration of 1,2-dihydroxybenzene: closed circles,
300 mM; open squares, 200 mM; closed diamonds, 100 mM; open triangles,
20 mM.
Biological Kolbe–Schmitt carboxylation
95
2.5. 2,3-Dihydroxybenzoate decarboxylase
HO
OH
HO
COOH
OH
+
CO2
Scheme 5 2,3-Dihydroxybenzoate decarboxylase has been reported to be involved in the metabolism of indole, tryptophan, and anthranilic acid.4041 The primary struc ture of 2,3-dihydroxybenzoate decaroboxylase of Aspergillus niger has been revealed by gene analysis, which exhibits significant homology with those of 2,6-dihydroxybenzoate decarboxylases (unpublished data). This suggests that 2,3 dihydroxybenzoate decarboxylase also catalyzes the regioselective carboxylation of catechol to 2,3-dihydroxybenzoate, although the carboxylation activity has not been examined. 2,3-Dihydroxybenzoate decarboxylase purified from Aspergillus niger has a molecular mass of 120 kDa and consists of four identical 28 kDa sub units. 2,3-Dihydroxybenzoate has a Km value of 340 M and does not act on salicylate, anthranilate, 2,3-dihydroxybenzaldehyde, 2,4-dihydroxybenzoate, 3-hydroxyanthranilate, 3-hydroxybenzoate, 2,3-dihydroxybenzoate, benzoate, or 4-hydroxybenzoate.42 The enzyme of Trichosporon cutaneum is a homodimer of identical 36.5 kDa subunits, which also catalyzes the decarboxylation of 2,3,5-trihydroxybenzoate and 2,3,6-trihydroxybenzoate.41 2,3-Dihdroxybenzoate decarboxylase can be purified, without the addition of sulfhydryl-protecting reagents to the purification buffer, indicating that the decarboxylases are insensi tive to O2 . Recently, the distribution of 2,3-dihydroxybenzoate decarboxylase has been found in a variety of fungal strains (unpublished data), and the carboxylation activ ity for catechol is confirmed by the reaction using resting cells (or cell-free extract) in the presence of 3 M KHCO3 . The detailed comparative studies of enzyme structures and catalytic properties between 2,3-dihydroxybenzoate decarboxylase and 3,4-dihyroxybenzoate decarboxylase might explain how the decarboxylases catalyze the regioselective carboxylation of catechol.
3. ENZYMES CATALYZING THE DIRECT CARBOXYLATION OF HETEROCYCLIC COMPOUNDS Non-oxidative novel aromatic acid decarboxylases, pyrrole-2-carboxylate decar boxylase and indole-3-carboxylate decarboxylase, were found in Bacillus
96
Toyokazu Yoshida, Toru Nagasawa
megaterium PYR291043 and Arthrobacter nicotianae F11612,44 respectively. These enzymes also catalyze the reverse reaction, the regiospecific carboxylation of heterocyclic compound, pyrrole, or indole. 3.1. Pyrrole-2-carboxylate decarboxylase
+ N H
COOH
CO2
N H
Scheme 6
B. megaterium PYR2910, isolated from soil for its ability to grown on pyrrole-2 carboxylate as the sole source of carbon and energy, produces a novel enzyme, pyrrole-2-carboxylate decarboxylase, to catalyze the decarboxylation of pyrrole 2-carboxylate into pyrrole and CO2 .43 Thiophene-2-carboxylate and l-thioproline (each 0.2%, w/v), which are analogous to the substrate pyrrole-2-carboxylate but not converted by the enzyme, were found to be the most effective enzyme inducers, leading to 3-fold higher specific enzyme activity than the substrate inducer pyrrole 2-carboxylate.45 The enzyme was purified from B. megaterium PYR2910. The enzyme has a molecular mass of approximately 98 kDa and consists of two identical subunits. The Vmax and Km values for decarboxylation were determined to be 989 units mg−1 and 24 mM, respectively. The purified enzyme was stable between pH 6 and 9 and at temperature below 50 C. The pH and temperature optima were 6.5 and 45 C, respectively. A unique feature of this enzyme is its requirement of an organic acid, such as acetate, propionate, butyrate, or pimelate, for the activity. The highest activity was found with pimelate (1750 mol min−1 mg−1 , Km 1.8 mM), followed by butyrate (76% relative activity, 45 mM), propionate (74%, 42 mM), and acetate (56%, 43 mM). These acids might reflect the steric dimensions of the enzyme site, where the organic acid affects the catalysis. Throughout the reaction, stoichiometric amounts of pyrrole and CO2 were formed from pyrrole-2-carboxylate. During the reaction, the concentrations of added organic acid, such as acetate, stayed constant.43 Pyrrole-2-carboxylate decarboxylase attains equilibrium in the course of either decarboxylation or carboxylation (Fig. 8). The decarboxylation of 100 mM pyrrole-2-carboxylate was in equilibrium after 1 h, resulting in an equilibrium constant of 0.3 M.45 Due to this “balanced” equilibrium, the enzyme also catalyzed the reverse carboxylation of pyrrole after the addition of HCO3 − , leading to a similar equilibrium constant of 0.4 M and a shift of the [pyrrole]/[pyrrole-2 carboxylate] ratio toward the acid.
Concentration (mM)
Biological Kolbe–Schmitt carboxylation
100
97
100 Pyrrole-2-carboxylate
Pyrrole
80
80 60
60
40
40 Pyrrole-2-carboxylate
20
20 Pyrrole
0
0 0
2
4
6
8
0
2
4
6
8 10 12
Time (min) (a)
(b)
Figure 8: Decarboxylation of pyrrole-2-carboxylate (a) and carboxylation of pyrrole (b) by pyrrole-2-carboxylate decarboxylase.
The reverse CO2 fixation depended on a CO2 source (CO2 or HCO3 − ), which was an additional limiting factor for the reverse reaction. Due to high water solubilities, the best CO2 sources were bicarbonates (HCO3 − with KHCO3 leading to 82 mM pyrrole-2-carboxylate from 100 mM pyrrole, followed by NH4 HCO3 (94% relative activity), NaHCO3 (81%), BaCO3 (17%), and CaCO3 (16%). Other carbonates (CO3 2− ), CO2 gas, or dry ice were with low or without effect due to a low water solubility of CO3 2− and CO2 at neutral pH.46 The reverse reaction showed a substrate saturation dependence, with optimal HCO3 − concentrations above 2.5 M. For the highest carboxylation yield, saturating amounts of 3 M KHCO3 were used, leading to a shift of the reaction equilibrium toward the carboxylate. HCO3 − addition was accompanied by CO2 gas evolution resulting in an increased pressure in the tightly closed reaction vessel of 1.38 atm, which supported the reverse reaction productivity 2.5-fold compared to atmospheric pressure. High pressures are also known to be applied in organic chemical carboxylations.47 As biocatalyst, either concentrated cells with an optical density at 610 nm of 40, previously grown under inducing conditions, or the purified enzyme, both in a concentration corresponding to 100 units enzyme activity ml−1 , were employed. Additionally, acetate as an enzyme cofactor and l-ascorbate as an anti-oxidizing and enzyme-protecting agent were added to the reaction mixture. For maximal CO2 fixation rates, 300 mM pyrrole was optimal. Higher pyrrole concentrations inhib ited the enzyme. In a batch reaction, 255 g l−1 (230 mM) pyrrole-2-carboxylate was formed from 207 g l−1 (300 mM) pyrrole (Fig. 9a). The productivity was increased to 325 mM (361 g l−1 ) by feeding 150 mM pyrrole after 3 h, with initially 250 mM pyrrole (Fig. 9b). The yield after bioconversion was 80%, limited by the equilibrium.
98
Toyokazu Yoshida, Toru Nagasawa
Concentration (mM)
400 Pyrrole-2-carboxylate 300 Pyrrole-2-carboxylate 200 100 Pyrrole
Pyrrole
0 0
2
4
6
8
10
0
2
4
6
8 10 12
Time (min) (a)
(b)
Figure 9: Time course of enzymatic carboxylation of pyrrole. (a) In the batch reaction, initially 300 mM pyrrole was added. (b) The batch-fed reaction was started with 250 mM pyrrole, followed by a second addition of 150 mM after 3 h.
An enzyme mechanism including a cofactor role of organic acid was pro posed. Analogous to the decarboxylation of pyrrole-2-carboxylate by heat, an electrophilic substitution at C2 pyrrole with a C2 protonated intermediate is prob able. The negatively charged organic acid might attack the only positive ring position at N1 attracting its proton. This eases the electron delocalization in the ring, thus increasing the electron density at C2. The C2 proton, therefore, can be substituted by the electrophilic carbon of CO2 (rather than by the less electrophilic carbon of HCO3 − ). The electrophilic substitution also allows the reverse decar boxylation with a protonated intermediate stabilized by the organic acid. Support for the presumed catalytic function of the pyrrole N1-proton might be deduced from the fact that N -methylpyrrole is not converted by the enzyme. The development of CO2 fixation reactions in supercritical CO2 attracts increasing attention due to its gas-like low viscosities and high diffusivities and its liquid-like solubilizing power. Matsuda et al.48 attempted to carry out the conO H
C N
H
O
O N
H
O
C
N
C
O–
H
O–
H O–
O R
O
O R
Scheme 7
O–
O R
Biological Kolbe–Schmitt carboxylation
99
version of pyrrole into pyrrole-2-carboxylate in supercritical CO2 using cells of B. megaterium PYR2910. The reaction was conducted by adding CO2 to 10 MPa to the mixture of pyrrole, the cells, KHCO3 , and NH4 OAc in potassium phos phate buffer at 40 C. The reaction reached an equilibrium position within a few hours and did not proceed further. The yield of the carboxylation reaction in supercritical CO2 (7.6 MPa) was 12 times higher than that under atmospheric pres sure. The effect of pressure on the carboxylation of pyrrole was also investigated, and the maximum yield was between 4 and 7 MPa. This finding suggested the potentiality for biocatalysis in supercritical CO2 in developing synthetic methods utilizing CO2 . Pyrrole-2-carboxylate is employed in the synthesis of various pharmaceu ticals4950 and a potential herbicide.51 A number of organic syntheses have been described52−54 However, they require multiple steps and result in low yields. Furthermore, the chemical carbonation of pyrrole with K2 CO3 depends on high pressure and temperature.55 The one-step enzymatic conversion has advantages with regard to regiospecificity, yield, and mild reaction conditions. 3.2. Indole-3-carboxylate decarboxylase COOH + N H
CO2
N H
Scheme 8 The highest indole-3-carboxylate decarboxylase activity was found in A. nico tianae F11612 isolated from the soil sample through conventional enrichment culture using indole-3-carboxylate as a sole carbon source.44 The occurrence of the enzyme was also found in Fusarium oxysporum IAM5009 and Gibberella fujikuroi IFO6605 from culture collections. The enzyme of A. nicotianae FI1612 was inducibly formed by the addition of 0.05% (w/v) indole-3-carboxylic acid, and 3-cyanoindole and l-tryptophan also induced the enzyme activity slightly. No activity was found without the addition of inducers. In the screening for indole-3-carboxylate decarboxylase, its weak activity was frequently lost during aerobic cultivation. However, after the optimization of culture conditions of A. nicotianae FI1612, increased activity of whole cells was found even under aerobic conditions. The highest total activity of about 500 nmol ml−1 min−1 was observed after 32 h at 28 C. When a cell extract prepared from A. nicotianae FI1612 cells was stored without the addition of sulfhydryl-protecting reagents, 80% of the initial activity was lost after storage at 4 C for 4 days. The enzyme activity was stabilized
100
Toyokazu Yoshida, Toru Nagasawa
by the addition of a mixture of 1 mM dithiothreitol, 50 mM Na2 S2 O3 , and 20% (v/v) glycerol to the purification buffer. However, much loss of activity seemed inevitable during enzyme purification. Due to its lability, the enzyme could not be purified to be homogeneous; however, the molecular mass of the enzyme was estimated to be 258 kDa by gel permeation. The decarboxylation activity was maximal at 50 C and pH 7.0. The A. nicotianae FI1612 enzyme did not require any cofactor for its activity, although the pyrrole-2-carboxylate decarboxylase activity of B. megaterium PYR2910 was completely dependent on an organic acid such as acetic acid.43 The carboxylation of indole into indole-3-carboxylate was observed by the purified indole-3-carboxylate decarboxylase as well as by the whole cells. For the carboxylation reaction, temperatures over 30 C were not appropriate. The activi ties at 10, 20, and 30 C were about the same. The activity was maximal at pH 8.0 (Tris–HCl buffer, 100 mM). As shown in Fig. 10, the resting cells of A. nicotianae FI1612 also catalyzed the carboxylation of indole efficiently in the reaction mix ture containing 20 mM indole, 3 M KHCO3 , 100 mM potassium phosphate buffer (pH 6.0) in a tightly closed reaction vessel. By 6 h, 6.81 mM indole-3-carboxylic acid accumulated in the reaction mixture with a molar conversion yield of 34%. Compared to the carboxylation of pyrrole by pyrrole-2-carboxylate decarboxylase, the lower value compared might derive from the lower solubility of indole in the reaction mixture. Resting cells of A. nicotianae FI1612 also catalyzed the carboxylation of 2-methylindole and quinoxaline. The activities toward 2-methylindole and quinoxaline were 37 and < 1% of the activity toward indole, respectively. The reaction products of the reverse carboxylation, indole-3-carboxylic acid and 2-methylindole-3-carboxylic acid, were isolated and identified through physico chemical analyses with the authentic compounds as reference.
Indole-3-carboxylate (mM)
25 20 15 10 5 0
0
2
4 Time (h)
6
Figure 10: Carboxylation of indole by whole cells of Arthrobacter nicotianae
FI1612.
Biological Kolbe–Schmitt carboxylation
101
4. STRUCTURE ANALYSIS OF DECARBOXYLASES CATALYZING CO2 FIXATION Up to now, primary structures of the following enzymes catalyzing CO2 fixation have been revealed from nucleotide sequencing of their genes: 4-hydroxybenzoate decarboxylase,2556 3,4-dihydroxybenzoate decarboxylase,26 pyrrole-2-carboxylate decarboxylase (unpublished data), indole-3-carboxylate decarboxylase (unpublished data), and 2,6-dihydroxybenzoate decarboxylase.3738 As for 2,3-dihydroxybenzoate decarboxylase, the carboxylation activity for cat echol has not been reported. However, we presume that 2,3-dihydroxybenzoate decarboxylase also catalyzes the reverse carboxylation reaction, since the primary structure is homologous to that of 2,6-dihydroxybenzoate decar boxylase. These decarboxylases can be divided into two enzyme groups as shown in Table 1, according to similarity of their primary structures. Class I contains oxygen-sensitive decarboxylases, their subunit molecular mass being 52–60 kDa. 4-Hydroxybenzoate decarboxylase, 3,4-dihydroxybenzoate
Table 1 Classification of decarboxylases catalyzing CO2 fixation Sub. MWa (kDa)
Enzyme
Source
Class I 4-Hydroxybenzoate
Chyd
57
3-Octaprenyl-4-hydroxybenzoate decarboxylase
Eclo240 Eclo241
60 59
YclC VdcC (vanillate decarboxylase) Unidentified hypothetical proteins
Bmeg
52
Anic
60
Atum Rrad Rsp Anig
38 34 38 28
decarboxylase 3,4-Dihydroxybenzoate decarbnoxylase Pyrrole-2-carboxylate decarboxylase Indole-3-carboxylate decarboxylase Class II 2,6-Dihydroxybenzoate decarboxylase 2,3-Dihydroxybenzoate decarboxylase
Homologous proteins
5-Carboxyvanillate decarboxyolate 2-Amino-3-carboxyymuconate-6 semialdehyde decarboxylase Unidentified hypothetical proteins
Chyd, Clostridium hydroxybenzoicum; Eclo240, Enterobacter cloacae P240; Eclo241, Enterobacter cloacae P241; Bmeg, Bacillus megaterium PYR2910; Anic, Arthrobacter nicotiacae FI1612; Atum, Agrobacterium tumefaciens IAM12048; Rrad, Rhizobium radiobacter WU0108; Rsp, Rhizobium sp. strain MTP-10005.
102
Toyokazu Yoshida, Toru Nagasawa
decarboxylase, pyrrole-2-carboxylate decarboxylase, and indole-3-carboxylate decarboxylase belong to this enzyme group. The decarboxylases in class II are insensitive to oxygen with subunit molecular mass of 28–38 kDa, such as 2,6 dihydroxybenzoate decarboxylase and 2,3-dihydroxybenzoate decarboxylase. The members of class I show no significant homology with class II enzymes. The structural and evolutionary relationships between class I and II enzymes have not been elucidated yet, because no tertiary structure has so far been presented.
4.1. Class I decarboxylases In the aligned primary structures of class I decarboxylases, the conserved amino acid residues are scattered over their primary structures. There have been few reports to identify the amino acid residues essential for catalytic activity or sub strate binding. Huang et al.56 reported the E-X-P motif in the alignment analysis for 4-hydroxybenzoate decarboxylase of C. hydroxybenzoicum and its homologous unidentified proteins. The E-X-P motif is also conserved in pyrrole-2-carboxylate decarboxylase and indole-3-carboxylate decarboxylase (unpublished data). How ever, the corresponding motif sequence is not observed in the primary structures of 3,4-dihydroxybenzoate decarboxylase of E. cloacae P241.26 Recent homology search using 4-hydroxybenzoate decarboxylase of E. cloacae P240 has shown that its primary structure exhibits 53, 72, and 28% identities to 4-hydroxybenzoate decarboxylase of C. hydroxybenzoicum, VdcC protein (vanillate decarboxylase) of Streptomyces sp. strain D7, and 3 octaprenyl-4-hydroxybenzoate decarboxylase of Escherichia coli, respectively.25 On the database of protein primary structure, there are 11 unidentified pro teins showing over 50% identities with 4-hydroxybenzoate decarboxylase of E. cloacae P240. The homologous proteins are representatively named as probable 4-hydroxybenzoate decarboxylase, YclC protein, or 3-octaprenyl (or polyprenyl)-4-hydroxybenzoate decarboxylase. Hypothetical YclC of E. coli, puta tive 3-polyprenyl-4-hydroxybenzoate decarboxylase of Salmonella typhimurium, a hypothetical protein of S. enterica, YclC protein of Kluyvera citrophila, and putative 4-hydroxybenzoate decarboxylase of Shigella flexneri exhibited 96–97% identities to E. cloacae P240 4-hydroxybenzoate decarboxylase. These putative proteins might be 4-hydroxybenzoate decarboxylase, although the enzyme activity has not been reported in such bacteria. Although 4-hydroxybenzoate decarboxylase of E. cloacae P240 shows sim ilar properties to 3,4-dihydroxybenzoate decarboxylase of E. cloacae P241 with the exception of substrate specificity, the identity of primary structures was only 24%.26 Probably, two hydroxyl groups of 3,4-dihydroxybenzoate are strictly rec ognized by the substrate-binding region of 3,4-dihydroxybenzoate decarboxy lase. Pyrrole-2-carboxylate decarboxylase and indole-3-carboxylate decarboxylase exhibit reasonable homology (20–60%) with hypothetical proteins of various
Biological Kolbe–Schmitt carboxylation
103
microorganisms (unpublished data), suggesting the latent distribution of decar boxylases catalyzing the carboxylation of N -heterocyclic compounds. 4.2. Class II decarboxylases With respect to class II enzymes, the primary structure analysis and sitedirected mutagenesis study of 2,6-dihydroxybenzoate decarboxylase have been reported.3738 2,6-Dihydroxybenzoate decarboxylase shows 20–50% identi ties to 2,3-dihydroxybenzoate decarboxylase, 5-carboxyvanillate decarboxylase, 2-amino-3-carboxymuconate-6-semialdehyde decarboxylase, and several hypo thetical proteins. The multiple alignment of 2,6-dihydroxybenzoate decarboxy lase of Rhizobium strains and the orthologous proteins including hypothetical proteins indicated that two histidine residues were conserved. A site-directed mutagenesis study revealed the histidine residues to be essential for the cat alytic activity.38 As for 2,3-dihydroxybenzoate decarboxylase, the presence of histidine, tryptophan, and cystein residues was suggested at the active site of the enzyme from Aspergillus niger.5758 To elucidate the reaction mechanism of decarboxylation/carboxylation reaction without the involvement of cofactor, further site-directed mutagenesis studies and X-ray crystallographic analysis are required. 4.3. Phenylphosphate carboxylase Primary structure analysis of phenylphosphate carboxylase of T. aromatica is performed in detail, to clarify the reaction mechanism involving four kinds of subunits.32 The subunits have molecular masses of 54, 53, 18, and 10 kDa, respectively, which make up the active phenylphosphate carboxylase. The primary structures of and subunits show homology with 3-octaprenyl 4-hydroxybenzoate decarboxylase, 4-hydroxybenzoate decarboxylase, and vanil late decarboxylase, whereas subunit is unique and not characterized. The 18 kDa subunit belongs to a hydratase/phosphatase protein family. Taking 4-hydroxybenzoate decarboxylase into consideration, Schühle and Fuchs postulate that the core enzyme catalyzes the reversible carboxylation.32
5. CONCLUSION The enzymes catalyzing the Kolbe–Schmitt carboxylation seem to occur ubiq uitously. Some of them, such as 2,6-dihydroxybenzoate decarboxylase and pyrrole-2-carboxylate decarboxylase, catalyze efficiently the reverse carboxyla tion reaction and accumulate high concentration of 2,6-dihydroxybenzoate from 1,3-dihydroxybenzene and pyrrole-2-carboxylate from pyrrole, respectively, in the
104
Toyokazu Yoshida, Toru Nagasawa
presence of 3 M KHCO3 . The facts rouse us to explore the possible applica tion of biological Kolbe–Schmitt carboxylation for the production of aromatic acids. The biological Kolbe–Schmitt carboxylation is a promising new tool for the regiospecific introduction of carboxy groups in precursors of pharmaceuti cals and agrochemicals in order to functionalize them and, thus, alter biological activities. Unfortunately, so far, aromatic acid decarboxylases are highly specific for each aromatic substrate, and a number of its analogs are not carboxylated. The gene resources of various aromatic decarboxylases have been gathered and preserved. Since there are various other valuable targets for regiospecific car boxylations, the finding of novel CO2 -fixing enzymes with a wider substrate spectrum is desirable. The screening for other reversible decarboxylases, showing wide substrate specificity and catalyzing regiospecific carboxylation, is signifi cant and potential. Further survey of microbial resources having aromatic decar boxylases will be continued. From now on, for the further application of the reverse carboxylation of decarboxylase, the most important thing is to develop new excellent techniques to shift the reaction equilibrium to the direction of carboxylation.
REFERENCES 1. Hartman, F. C.; Harpel, M. R. Annu. Rev. Biochem. 1994, 63, 197–234. 2. Bisaillon, J.-G.; Lèpine, F.; Beaudet, R.; Sylvestre, M. Appl. Environ. Microbiol. 1991, 57, 2131–2134. 3. Brackmann, R.; Fuchs, G. Eur. J. Biochem. 1993, 213, 563–571. 4. Gallert, C.; Winter, J. Appl. Microbiol. Biotechnol. 1992, 37, 119–124. 5. Gorny, N.; Schink, B. Appl. Environ. Microbiol. 1994, 60, 3396–3400. 6. Knoll, G.; Winter, J. Appl. Microbiol. Biotechnol. 1989, 30, 318–324. 7. Kuever, J.; Kulmer, J.; Jansen, S.; Fischer, U.; Blotevogel, K.-H. Arch. Microbiol. 1993, 159, 282–288. 8. Lack, A.; Fuchs, G. J. Bacteriol. 1992, 174, 3629–3636. 9. Ramanand, K.; Suflita, J. M. Appl. Environ. Microbiol. 1991, 49, 1689–1695. 10. Sharak-Genthner, B. R.; Townsend, G. T.; Chapman, P. J. FEMS Microbiol. Lett. 1991, 78, 265–270. 11. Zhang, X.; Young, L. Y. Appl. Environ. Microbiol. 1997, 63, 4759–4764. 12. Kolbe, H. Liebigs Ann. Chem. 1860, 113, 125–127. 13. Otta, K. Bull. Chem. Soc. Jpn 1974, 47, 2343–2344. 14. Dewar, M. S. J. The Electronic Theory of Organic Chemistry; Oxford University Press: London; 1949, pp. 1–324. 15. Tschech, A.; Fuchs, G. Arch. Microbiol. 1987, 148, 213–217. 16. Glockler, R. A.; Tschech, A.; Fuchs, G. FEBS Lett.1989, 251, 237–240. 17. Lovley, D. R.; Lonergan, D. J . Appl. Environ. Microbiol.1990, 56, 1858–1864. 18. Zhang, X.; Mandelco, L.; Wiegel, J. Int. J. Syst. Bacteriol. 1994, 44, 214–222. 19. Zhang, X.; Wiegel, J. Microb. Ecol. 1990, 20, 103–121. 20. He, Z.; Wiegel, J. Eur. J. Biochem. 1995, 229, 77–82. 21. He, Z.; Wiegel, J. J. Bacteriol. 1996, 178, 3539–3543. 22. Szewzyk, R.; Pfenning, N. Arch. Microbiol. 1987, 147, 163–168.
Biological Kolbe–Schmitt carboxylation 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.
105
Schnell, S.; Bak, F.; Pfennig, N. Arch. Microbiol. 1989, 152, 556–563. Gorny, N.; Schink, B. Appl. Environ. Microbiol. 1994, 60, 3396–3400. Matsui, T.; Yoshida, T.; Hayashi, T.; Nagasawa, T. Arch. Microbiol. 2006, 186, 21–29. Matsui, T. Ph.D. Thesis, Gifu University, 2006. Li, T.; Juteau, P.; Beaudet, R.; Lépine, F.; Villemur, R.; Bisaillon, J. Can. J. Microbiol. 2000, 46, 856–859. Zhang, Z.; Wiegel, J. Appl. Environ. Microbiol, 1994, 60, 4182–4185. Lack, A.; Fuchs, G. Arch. Microbiol. 1994, 161, 132–139. Lack, A.; Fuchs, G. Eur. J. Biochem. 1991, 197, 473–479. Breining, S.; Schiltz, E.; Fuchs, G. J. Bacteriol. 2000, 182, 5849–5863. Schühle, K.; Fuchs, G. J. Bacteriol. 2004, 186, 4556–4567. Schmeling, S.; Narmandakh A.; Schmitt, O.; Gad’on, N.; Schühle, K.; Fuchs, G. J. Bacteriol, 2004, 186, 8044–8057. Aresta, M.; Quaranta, E.; Liberio, R.; Dileo, C.; Tommasi, I. Tetrahedron 1998, 54, 8841–8846. Aresta, M.; Dibenedetto, A. Rev. Mol. Biotechnol. 2002, 90, 113–128. Yoshida, T.; Hayakawa, Y.; Matsui, T.; Nagasawa, T. Arch. Microbiol. 2004, 181, 391–397. Yoshida, M.; Fukuhara, N.; Oikawa, T. J. Bacteriol. 2004, 186, 6855–6863. Ishii, Y.; Narimatsu, Y.; Iwasaki, Y.; Arai, N.; Kino, K.; Kirimura, K. Biochem. Biophys. Res. Commun. 2004, 324, 611–620. Matsui, T.; Yoshida, T.; Yoshimura, T.; Nagasawa, T. Appl. Microbiol. Biotechnol. 2006, 73, 95–102. Kamath, A. D.; Vaidyanathan, C. S. Appl. Environ. Microbiol. 1990, 56, 275–280. Anderson, J. J.; Dagley, S. J. Bacteriol. 1981, 146, 291–297. Kamath, A. D.; Dasgupta, D.; Vaidyanathan, C. S.Biochem. Biophys. Res. Commun. 1987, 145, 586–595. Omura, H.; Wieser, M.; Nagasawa, T. Eur. J. Biochem. 1998, 253, 480–484. Yoshida, T.; Fujita, K.; Nagasawa, T. Biosci. Biotechnol. Biochem. 2002, 66, 2388–2394. Wieser, M.; Fujii, N.; Yoshida, T.; Nagasawa, T. Eur. J. Biochem. 1998, 257, 495–499. Asada, K. In Organic and Bioorganic Chemistry of Carbon dioxide, Biological Carboxylation; Inoue, S.; Yamazaki, N. (Eds); Halsed: New York; 1982, pp. 185–251. Haruki, E. In Organic and Bioorganic Chemistry of Carbon dioxide, Organic Synthesis with Carbon Dioxide; Inoue, S.; Yamazaki, N. (Eds); Halsed: New York; 1982, pp. 5–78. Matsuda, T.; Ohashi, Y.; Harada, T.; Yanagihara, R.; Nagasawa, T.; Nakamura K. Chem. Commun. 2001, 2194–2195. Kerwin, J. F. Jr; Wagenaar, F.; Kopecka, H.; Lin, C. W.; Miller, T.; Witte, D.; Srashko, M.; Nadzan, A. M. J. Med. Chem. 1991, 34, 3550–3359. Kasum, B.; Prager, R. H.; Ksopelas, C. Aust. J. Chem. 1990, 43, 355–365 Peterson, M. L.; Corey, S. D., Fond, J. L.; Walker, M. C.; Sikorsky, J. A. Bioorg. Med. Chem. Lett. 1996, 6, 2853–2858. Doyle, F. P.; Mehta, M. D.; Sach, G. S.; Pearson, J. L. J. Chem. Soc. 1958, 4458–4466. Kreszu, G.; Firl, J. Angew. Chem. 1964, 76, 439. Terry, W. G.; Jackson, A. H.; Kenner, G. W.; Kornis, G. J. Chem. Soc. 1965, 4389–4393. Smissman, E. E.; Gaber, M. B.; Winzler, R. L. J. Am. Pharm. Assoc. 1956, 45, 509. Huang, Z.; He, Z.; Wiegel, J. J. Bacteriol. 1999, 181, 5119–5122. Kamath, A. V.; Rao, N. A.; Vaidyanathan, C. S. Biochem. Biophys. Res. Commun. 1989, 165, 20–26. Santha, R.; Savithri, H. S.; Rao, N. A.; Vaidyanathan, C. S. Eur. J. Biochem. 1995, 230, 104–110.
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
107
Chapter 6
Discovery, redesign and applications of Baeyer–Villiger monooxygenases Daniel E. Torres Pazmiño, Marco W. Fraaije Laboratory of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Nijenborgh, Groningen, The Netherlands
Abstract Baeyer–Villiger monooxygenases (BVMOs) represent valuable oxidative biocatalysts. A special feature of these atypical monooxygenases is that they catalyze not only Baeyer– Villiger oxidations but also sulfoxidations and a number of other oxidation reactions. Except for this promiscuity in reactivity, BVMOs are often very enantio-, regio- and/or chemoselective while accepting a broad range of substrates. This chapter provides an overview of recent developments concerning these special oxidative biocatalysts and sketches some future perspectives.
1. INTRODUCTION The Baeyer–Villiger oxidation reaction was discovered more than 100 years ago by Adolf von Baeyer and Victor Villiger.1 By this reaction, ketones are converted into the corresponding esters. In organic chemistry, peracids are commonly used as catalyst to perform this atypical oxidation reaction that results in oxygen insertion into a carbon−carbon bond (Fig. 1). Already in 1948 it was recognized that enzymes that catalyze Baeyer– Villiger reactions exist.2 This was concluded from the observation that a biological Baeyer–Villiger reaction occurred during the biotransformation of steroids by fungi. It took two decades for the first Baeyer–Villiger monooxygenases (BVMOs, EC 1.14.13.X), sometimes also referred to as Baeyer-Villigerases, to be isolated and characterized.34 From then on, a number of microbial BVMOs have been reported revealing several recurring biochemical characteristics (for a recent review see Ref. 5). All characterized BVMOs contain a flavin cofactor that is crucial for cataly sis while NADH or NADPH is needed as electron donor. An interesting observation is the fact that most reported BVMOs are soluble proteins. This is in contrast to many other monooxygenase systems that often are found to be membrane-bound or membrane-associated. In 1997, Willetts6 concluded from careful inspection of
108
Daniel E. Torres Pazmiño, Marco W. Fraaije O X1
O
Baeyer–Villiger oxidation
X2
X1
O
X2
Figure 1: The Baeyer–Villiger oxidation reaction. all available biochemical data on BVMOs that at least two classes of BVMOs exist. The type I BVMOs consist of only one polypeptide chain, contain FAD as tightly bound cofactor and are dependent on NADPH for activity. They contain two Rossmann sequence motifs, GxGxxG, indicating that these enzymes bind the two cofactors (FAD or NADPH) using separate dinucleotide-binding domains.7 The type II BVMOs use FMN as flavin cofactor and NADH as electron donor and are composed of two different subunits. At the time of the initial classification, the respective N-terminal sequences did not provide any clue concerning the struc ture of these two-component monooxygenases. However, recent sequence data suggest a sequence relationship with the flavin-dependent luciferases.8 Therefore, it is likely that type II BVMO oxygenase subunits will also have a TIM-barrel fold. Recent findings hint to at least two other BVMO classes of which one also represents a group of flavoproteins. It was found that the bacterial flavoprotein monooxygenase MtmOIV is involved in the biosynthetic pathway of the antitumor drug mithramycin.9 Sequence analysis indicates that it is related to flavoprotein monooxygenases that typically perform hydroxylation or epoxidation reactions.8 Crystals of this novel monooxygenase have recently been reported.10 A crystal structure would reveal what structural features separate this BVMO from the sequence-related hydroxylases and epoxidases. Also a heme-containing BVMO has recently been reported belonging to the cytochrome P450 superfamily.11 This plant enzyme was shown to convert a specific plant steroid. Earlier studies already suggested Baeyer–Villiger activity of other eukaryotic P450s.12 Future studies will reveal whether these novel oxidative enzymes can be of use for biocat alytic applications. However, the first results suggest that these newly identified BVMOs are dedicated to convert very specific and complex molecules suggesting a narrow substrate specificity. Nevertheless, the finding of these novel BVMO types indicates that during evolution several different enzymes have evolved into BVMOs. Therefore, more BVMO types may be discovered in the coming years. Most biochemical and biocatalytic studies have been performed with type I BVMOs.5 This is partly because of the fact that they represent relatively uncompli cated monooxygenase systems. These monooxygenases are typically soluble and composed of only one polypeptide chain. Expression systems have been devel oped for a number of type I BVMOs while no recombinant expression has been reported for a type II BVMO. Cyclohexanone monooxygenase (CHMO) from an Acinetobacter sp. NCIMB9871 was the only recombinant available BVMO
Discovery, redesign and applications of BVMOs R
R N
109
N
NADPH-binding flavin reduction
O
– N
N
NH
NH N H
N O
O
NADP+
Reaction with oxygen Formation of peroxyflavin
Release of NADP+ Release of water
R
R N
N
Baeyer–Villiger reaction
O
N
N
O NH
NH N H OH NADP+
O
Substrate-binding product release O
O
N HO NADP+
O–
O
O O
Figure 2: Scheme of the catalytic mechanism of type I BVMOs. Phenylacetone is taken as example substrate.
for a long time as it was cloned and overexpressed already in 1988.13 CHMO has been subjected to several sophisticated kinetic studies which have revealed that in BVMOs also, catalysis is achieved by formation of a peracid catalyst: a peroxyflavin (Fig. 2).1415 Upon reaction with NADPH and molecular oxygen, the flavin cofactor is able to form this peroxygenated flavin intermediate. This reactive intermediate is equivalent to the peracids used in organic chemistry and will react with a ketone to form an ester. In fact, it is the ability of BVMOs to form and stabilize a negatively charged peroxyflavin intermediate that enables these enzymes to perform Baeyer– Villiger reactions and other oxygenation reactions. The formation of the reactive oxygenated enzyme intermediate is not regulated by substrate binding which sets these BVMOs mechanistically apart from other well-studied monooxygenase sys tems. For example, cytochrome P450s and flavin-containing hydroxylases will only form the equivalent reactive enzyme intermediate after binding of a substrate. The peroxyflavin in CHMO is stabilized by active-site residues and the bound NADP+ coenzyme. Structural details concerning this enzyme complex are lacking as no CHMO structure is available. After its formation, the peroxyflavin enzyme interme diate waits until a suitable substrate enters the active site upon which oxygenation will take place. In the case that no suitable substrate is present, the peroxyflavin
110
Daniel E. Torres Pazmiño, Marco W. Fraaije
will decay to form hydrogen peroxide. However, this NADPH oxidase function of BVMOs is very inefficient (< 01 s−1 ) due to the effective stabilization of the peroxy intermediate. This prevents intracellular formation of toxic hydrogen per oxide. A striking feature of the catalytic mechanism of CHMO is the fact that the coenzyme NADPH/NADP+ remains bound to the enzyme throughout the catalytic cycle. Only when the oxygenation reaction and the decay of the hydroxyflavin into oxidized flavin have occurred, NADP+ is released (Fig. 2). This mechanistic feature was recently confirmed for another type I BVMO: 4-hydroxyacetophenone monooxygenase (HAPMO)16 . By kinetic inhibition studies and ESI-MS experi ments, it could be demonstrated that the coenzyme remains bound during the whole catalytic cycle. Also, with the newly identified phenylacetone monooxygenase (PAMO)17 we have found that NADP+ is a competitive inhibitor with respect to NADPH (M. W. Fraaije, personal communication), suggesting that NADP+ release is again the last step in the catalytic cycle. Another indirect proof for binding of NADP+ throughout the catalytic cycle came from a study where an artificial elec tron donor was tested.18 This revealed that NADP+ binding is essential to maintain the enantioselectivity of PAMO, indicating that the bound coenzyme forms parts of the active site determining the positioning of the substrate. It also indicates that, in vivo, BVMOs are virtually always occupied by NADP+ or NADPH. This is in line with the observation that BVMOs are highly stabilized when NADP+ is bound.16
2. BIOCATALYTIC PROPERTIES OF RECOMBINANT AVAILABLE BVMOS The list of heterologously expressed type I BVMOs has grown significantly in recent years (Table 1). Except for BVMOs primarily acting on small cyclic ketones (cyclopentanone and cyclohexanone)1920 , variants specific for larger cyclic ketones (cyclododecanone and cyclopentadecanone) have also been discovered.2122 Addi tionally, BVMOs that readily accept aromatic ketones (4-hydroxyacetophenone and phenylacetone derivatives) have been described.1723 Also, a BVMO acting on steroids has been reported.24 Substrate profiling studies suggest that BVMOs have a rather broad specificity and often display overlapping substrate speci ficities. For example, bicycloheptenone and aromatic sulfides have been shown to be converted by several BVMOs. Illustrative for the broad substrate speci ficity of BVMOs is the fact that in 2002 it was reported that for CHMO from Acinetobacter sp. over 100 different substrates had been reported.25 Since that time, a number of other ketones have been shown to be converted by this monooxygenase extending the impressive list of CHMO substrates even further.25−29 Except for exploring its catalytic potential, CHMO from Acinetobacter has also been used as a model system for upscaling BVMO-mediated biocatalysis.
Table 1 Recombinant available BVMOs Acronym
Cyclopentanone monooxygenase
CPMO
Cyclohexanone monooxygenase
CHMO
Primary substrate
Origin
O
Comamonas sp.
Year of cloning 20022041
Acinetobacter sp.
19881319
Rhodococcus ruber
200121
Pseudomonas sp.
200622
Pseudomonas fluorescens
200123
Thermobifida fusca
200417
O
Cyclododecanone monooxygenase
CDDMO
Cyclopentadecanone monooxygenase
CPDMO
O
O
4-Hydroxyacetophenone monooxygenase
HAPMO
Phenylacetone monooxygenase
PAMO
O HO
O
Ethionamide monooxygenase
EtaA
Steroid monooxygenase
STMO
Physiological substrates unidentified Mycobacterium tuberculosis Rhodococcus rhodochrous
200447
Discovery, redesign and applications of BVMOs
BVMO
199924
O
O
111
List of BVMOs that have been overexpressed in E. coli and of which the isolated enzyme has been characterized to some extend.
112
Daniel E. Torres Pazmiño, Marco W. Fraaije
In recent years, the main focus has been on using whole cells expressing CHMO as biocatalyst. By this, the problem of coenzyme usage by the enzyme can be circumvented, as NADPH will be regenerated by the cellular machinery. During the years, problems related to, e.g., product/substrate inhibition and oxygen sen sitivity have been tackled.3031 An efficient methodology has been developed that is applicable on kilogram scale.32 Using whole cells expressing CHMO in combi nation with resin-based in situ substrate feeding and product removal (SFPR) and a tuned oxygen supply, a highly productive process was developed yielding two nearly enantiopure (ee > 98%) regioisomeric lactones in good yield. This nicely illustrates that BVMOs can be applied on a scale that is relevant for synthesis of fine chemicals. While the substrate acceptance of a specific BVMO is often relaxed, cat alytic efficiencies and regio- and/or enantioselectivities can differ significantly when comparing BVMOs. A number of CHMO and CPMO homologs (>35% sequence identity) have recently been identified and explored concerning their biocatalytic potential. These comparative biocatalytic studies using highly sim ilar enzymes have revealed that, as expected, all studied CHMOs and CPMOs cover a similar substrate range.293334 However, regio- and/or enantioselectivi ties differ significantly. It was observed that CPMOs and CHMOs often dis play opposite enantioselectivities.34 This illustrates the need for a large library of BVMOs to meet the demand for enantio- and/or regioselective oxidative biocatalysts. Although BVMOs often display a broad substrate specificity, each BVMO has a certain preference for a specific type of substrate. While CHMO and CPMO are highly active with a range of (cyclic) aliphatic ketones, HAPMO and PAMO prefer aromatic substrates.172335−37 Figure 3 illustrates the overlapping substrate specificities for several well-studied BVMOs displaying several typical substrates for each BVMO. For several recombinant available BVMOs, no extensive substrate profiling studies have been performed. For steroid monooxygenase, only steroid substrates have been tested while cyclododecanone and cyclopentadecanone monooxyge nases have only been recently identified and await further exploration. Based on the first biocatalytic data obtained with cyclopentadecanone monooxygenase, it appears that this novel BVMO is attractive when relatively large compounds are targeted.22 The enzyme proved to be effective in enantioselective Baeyer–Villiger oxidations of a range of bulky cycloketones. By this, it complements the substrate range that is covered by other known BVMOs. 2.1. Discovery of novel BVMOs While the number of available recombinant BVMOs has grown significantly over the last few years, there is still a demand for other BVMOs to expand the biocat alytic diversity. Most BVMOs that have been described are dedicated to efficiently
Discovery, redesign and applications of BVMOs
CHMO
O
113
O
HAPMO R
R
R
X
O
O
O H
O N O
O
O H N
O
X
R R
H
HO
O
O
O S R
O O
O R R
O
O O
O
R
O
O MeO Me
nPr O
CPMO
PAMO O
Figure 3: Illustration of overlapping substrate specificities of four BVMOs. Cyclohexanone monooxygenase (CHMO), cyclopentanone monooxygenase (CPMO), 4-hydroxyacetophenone monooxygenase (HAPMO) and phenylacetone monooxygenase (PAMO). For each enzyme, several typical substrates are shown.
convert cyclohexanone and related cyclic aliphatic ketones. To cover a broader range of substrate types and enantio- and/or regioselectivities, new BVMOs have to be discovered. This can be done in a number of ways. In the past, it was com mon practice to isolate new microbes that were able to grow on a target substrate after which the respective enzyme/gene was retrieved. This approach has been successful in obtaining most of the presently available BVMOs (Table 1). Only PAMO and ethionamide monooxygenase have been discovered via other meth ods.1436 However, the classical approach of isolating a specific microorganism, enzyme purification and subsequently cloning is laborious, time consuming and often unsuccessful. Frequently, the enzyme responsible for the observed reaction is difficult to purify and hence the respective gene cannot be retrieved. Therefore, it is attractive to exploit other newly developed methods that circumvent these pitfalls.
114
Daniel E. Torres Pazmiño, Marco W. Fraaije
2.2. Exploring sequenced (meta)genomes for novel BVMOs Nowadays, the genomes of a wide variety of organisms have been sequenced and are publicly available, offering a new and efficient way of retrieving BVMO genes. Currently, as genome sequences of over 600 microbes are available (see http://www.ncbi.nlm.nih.gov/sutils/genom_table.cgi), it is attractive to look directly into this pool of genes in silico without growing and isolating any microor ganism. In addition to the database of partially and fully sequenced genomes, it is also informative to survey the database of sequenced environmental genomes. Especially, the ‘Sargasso sea’ metagenome database is rich in new sequences.38 Metagenomic databases contain a large number of unexplored genes (>1 million sequences in the Sargasso database!). However, in an in silico (meta)genome mining approach it is fundamental to have bioinformatic tools to identify, with some certainty, genes that encode BVMOs. By simply searching for sequences that show homology with known BVMOs, novel putative BVMO genes may be found. However, many sequence-related genes may represent flavoprotein monooxygenases that do not catalyze Baeyer–Villiger reactions. A more reliable identification of BVMO genes has become feasible since a type I BVMO-specific motif (FxGxxxHxxxWD /P was identified by comparing sequences of characterized BVMOs.39 This allows an effective survey of all (meta)genome databases concern ing the occurrence of type I BVMOs (Table 2: note that this table only includes completely sequenced genomes). The identification of new BVMO genes will obviously facilitate production of novel biocatalysts. Except for obtaining the new gene sequence of a specific BVMO, a thorough genome analysis may also provide more valuable information. As genes belonging to specific degradation pathways are often clustered on microbial genomes, analyzing the sequence regions flanking a BVMO gene may give hints concerning the physiological role of the enzyme and accordingly provide clues for the corresponding substrate specificity. In fact, we have found that many BVMO genes are flanked by an esterase/hydrolase gene. The corresponding esterase or lactonase activity would hydrolyze the ester or lactone formed by BVMO activity. Such a co-localization of BVMO with an esterase/lactonase has also been observed in sequenced genome fragments containing a BVMO gene, e.g., in the case of the CHMO, cyclopentanone monooxygenase, PAMO, HAPMO and cyclopentadecanone genes.17222340−42 This suggests that BVMOs often play a role in a specific catabolic pathway and is in agreement with the fact that most described BVMOs are part of a degradation pathway. A genomic trawl using the above-mentioned type I BVMO motif as filter currently yields 174 putative BVMO genes when searching for all finished micro bial genomes (Table 2). This indicates that type I BVMOs are not very rare but are frequently utilized by microbes. All identified BVMO genes originate from bacteria or fungi while none could be found in archaebacteria, plants, animals or the human genome. On average, roughly one out of two (174/348) micro bial genomes contain a BVMO gene. This suggests that at present ∼400 novel
Discovery, redesign and applications of BVMOs
115
Table 2 Genomic occurrence of BVMOs Source
Number of screened genomes
Number of putative BVMOsa
Bacteria
293
138
Archaea
Examples of genomes containing multiple BVMOsb Mycobacterium aviumc (12) Novosphingobium aromaticivorans (8) Rhodococcus RHA1 (20) –
27
0
d
Fungi
28
36
Other eukaryotese
22
0
–
Sargasso sea
–
32
–
370
206
–
Total
Aspergillus nidulans FGSC-4A (16) Gibberella zeae PH-1 (10) Magnaporthe grisea 70-15 (7)
a
Genomes were surveyed for the presence of putative type I BVMOs by: (1) searching for sequence homologs of phenylacetone monooxygenase and (2) filtering for sequences that contain the type I BVMO-sequence motif.39 b The number of BVMO genes for each genome is indicated in brackets.
c Mycobacterium avium subsp. paratuberculosis K-10.
d Sixteen yeast and 12 fungal genomes have been screened: 33 BVMOs discovered in fungi and
three in the yeast Candida albicans SC5314. e Other eukaryotes included Apis mellifera, Bombyx mori, Caenorhabditis elegans, Drosophila melanogaster, Homo sapiens, Mus musculus, Tetraodon nigroviridis.
type I BVMO genes are present in the genome sequence database (including the unfinished genomes). Strikingly, BVMO genes are unevenly distributed among microbial genomes with only a few microorganisms containing a large number of BVMOs while the majority of genomes are devoid of putative type I BVMOs (Fig. 4). In fact, the number of genomes containing only one BVMO is more or less equal to the number of genomes containing four or more BVMO genes. It is also worth noting that a relatively large number of bacterial BVMO genes (80) were found in actinomycetes. This may suggest a role of BVMOs in the synthesis of secondary metabolites. Also, the Sargasso metagenome database contains a sig nificant number of BVMO genes indicating that in the sea environment also many microbes employ BVMOs for specific but yet unknown metabolic routes (Table 2). These genes cannot simply be obtained using PCR techniques as the microbes from which the genes originate have not been isolated. However, by gene synthesis it is in principle feasible to explore these newly identified putative biocatalysts.
Daniel E. Torres Pazmiño, Marco W. Fraaije
Frequency (number of genomes screened)
116
260
260
250 Bacteria Fungi
24 20 14
11 6 3
2 0
0 0
1
1
0 2
3
>3
Number of BVMOs/genome
Figure 4: Distribution of putative type I BVMO genes among bacterial and fungal genomes (see Table 2).
In contrast to the type I BVMOs discussed above, type II BVMOs have been explored to a limited extent. In fact, an in silico search for type II BVMOs in the genome sequence database is hampered by the fact that only one type II BVMO sequence (limonene monooxygenase, gi47116765) has been deposited in the database. A BLAST search with the limonene monooxygenase sequence at NCBI (http://www.ncbi.nlm.nih.gov) reveals that only 12 bacterial sequences show a relatively high sequence homology (>40% sequence identity). Except for these sequence homologs, a large number of other sequences also show lim ited sequence homology and appear to belong to the luciferase class of flavindependent monooxygenases. This hints to an evolutionary relationship between type II BVMOs and luciferases. In the Sargasso sea database, only five sequences can be found that display high sequence identity (>40%) with limonene monooxy genase. These findings suggest that type II BVMOs are less widespread when compared with type I BVMOs explaining to some extent why these BVMOs have been reported in the literature less frequently. Recently, several BVMOs have been reported in the literature that had been found by genome mining. The first example concerned the discovery of a ther mostable BVMO. The well-studied BVMO, CHMO, is not a very robust biocata lyst. Often, conversions using this biocatalyst suffer from enzyme inactivation. To circumvent this problem, it was of interest to obtain a more (thermo)stable BVMO. As no BVMO genes have been identified in genomes of archaebacteria, genomes of (semi)thermophilic bacteria were surveyed. Using the above-mentioned BVMO sequence motif, it was found that the genome of Thermobifida fusca contains two
Discovery, redesign and applications of BVMOs
117
type I BVMO genes. This actinomycete typically grows at 55–60 C and there fore should yield thermostable biocatalysts. The two genes have been cloned and the corresponding enzymes have been overexpressed in Escherichia coli. Only one of the two expressed BVMOs could be purified and characterized and was shown to be primarily active on a range of aromatic ketones and sulfides. The highest catalytic efficiency has been obtained with phenylacetone and hence its name phenylacetone monooxygenase. The genes flanking the PAMO gene also suggest a role in the degradation of aromatic compounds like phenylacetone. Except for phenylacetone, a number of other ketones and sulfides are accepted by the enzyme.1736 With several aromatic prochiral ketones and sulfides, excellent enantioselectivity was observed. However, the enzyme is only marginally active with cyclic aliphatic ketones and therefore has only a limited overlap in substrate specificity when compared with CHMO (Fig. 3). The enzyme indeed proves to be of superior stability when compared with other known BVMOs as it is stable for days when stored at moderate temperatures (<40 C). The enzyme is also active in the presence of organic solvents and, interestingly, it was found that organic solvents can tune the enantioselectivity.43 More recently the group of Grogan reported on another genome mining project targeting BVMOs from Mycobacterium tuberculosis.44 The genome of M. tuberculosis harbors six putative type I BVMO genes. Also, other mycobac terial genomes are relatively rich in BVMO genes with M. marinum setting the record at present: 15 BVMO genes. All six type I BVMOs genes that are part of the genome of M. tuberculosis were cloned and four of them could be overexpressed in E. coli. Unfortunately, by using three test substrates, Baeyer– Villiger activity of only three mycobacterial BVMOs could be confirmed. One of them corresponds to ethionamide monooxygenase that has been shown to be involved in the activation of commonly used antitubercular drugs.4546 In a previous study, Baeyer–Villiger activity of this monooxygenase had already been verified with a large number of ketones.47 However, another overexpressed BVMO from M. tuberculosis exhibited exquisite enantioselectivity with racemic bicyclo[3.2.0]hept-2-en-6-one.44 This bicyclic aliphatic ketone is often used to probe the biocatalytic potential of BVMOs. This chiral ketone is of high value for the synthesis of fine chemicals when available in its enantiomerically pure form. Also, the chiral lactones formed from this ketone by Baeyer–Villiger oxidation are highly interesting for synthetic purposes. For this reason, CHMO has been exploited in several biocatalytic studies as mentioned above.3032 Using whole cells containing the novel mycobacterial biocatalyst, it was possible to perform a res olution of racemic bicyclo[3.2.0]hept-2-en-6-one yielding enantiomerically pure (1R,5S)-(+)-bicyclo[3.2.0]hept-2-en-6-one. This was not feasible with CHMO or any other available BVMO and demonstrates again the value of a genome mining approach. By exploring the catalytic properties of novel BVMOs identified only by sequence, new substrate specificities and regio- and/or enantioselectivities may be uncovered.
118
Daniel E. Torres Pazmiño, Marco W. Fraaije
2.3. Screening the metagenome for novel BVMOs Except for exploring sequenced genomes, yet other interesting approaches for discovering novel BVMOs exist. A recent development in the area of biocatalyst discovery is the exploitation of metagenomic DNA libraries. By this methodology, DNA is directly isolated from a certain sample that is expected to contain a variety of microbes (e.g., soil). Subsequently, the obtained DNA is randomly fragmented and cloned into a suitable vector and expression host. The resulting gene library, containing a huge number of randomly cloned genome fragments, can subsequently be screened for any desired biocatalytic activity. In this way, bacterial enzymes exhibiting specific activities can be found without the need of isolating or cultivating a specific bacterium. This culture-independent methodology has been shown to be very successful to uncover novel biocatalysts and has quickly become a standard approach at academic and industrial institutions. However, no BVMO discovered by metagenome screening has yet been reported. As shown above, BVMOs occur frequently in bacterial genomes and therefore metagenomic gene libraries should typically contain a multitude of BVMOs. However, it is not trivial to screen libraries for specific enzyme activities. Often screening methods rely on visualization of enzyme activity by color formation upon substrate conversion or growth that is triggered by the specifically desired enzyme activity. Several assays that can be used for screening cells for BVMO activity have been reported but none of them appears to be suited for generic screening of a massive number of clones.48−50 The most commonly used assay to detect BVMO activity in vitro is measuring the depletion of NADPH which absorbs light at 340 nm. This approach cannot be used when handling whole cells. Recently, several alternative methods have been reported to detect cells exhibiting BVMO activity. In one report, it was shown that the BVMO-associated ester/lactone product formation can be coupled to an esterase/lactonase which will result in a drop of pH. By using pH indicators, this change of pH could be visualized.51 A disadvantage of this approach is that it is limited by the substrate range of the hydrolytic partner enzyme and very sensitive to other activities that influence the pH. Another reported method for detecting BVMO activity is also dependent on hydrolysis of the formed product. It was shown that HAPMO readily catalyzes conversion of 3-acetylindole to indoxyl acetate. Chemical or enzymatic hydrolysis of this product yields indoxyl which spontaneously reacts with molecular oxygen forming the blue-colored compound, indigo (Fig. 5).52 Recently, two fluorogenic assays were also described that can be used for detecting BVMO activity.4849 However, these methods are biased toward specific chromogenic and fluorogenic substrates and will only yield enzymes that are active toward these and similar compounds. An effective and generic method of screening of BVMO activity is indis pensable for screening metagenome libraries as these libraries are often in the range of 100 000–1 000 000 clones. Therefore, it is a challenge to develop a novel way to efficiently detect BVMO activity in gene libraries. This would facilitate
Discovery, redesign and applications of BVMOs
O O BVMO N H
OH
O
3-Acetyl indole
O Oxidation
Esterase/KOH N H
119
N H
H N
N H
O Indigo blue
Figure 5: Chromogenic BVMO assay using 3-acetyl indole resulting in the formation of indigo blue.52
discovery of a new set of BVMOs displaying novel biocatalytic properties and may also disclose new types of BVMOs. 2.4. Redesign of BVMOs For enzyme redesign studies, structural information of a targeted biocatalyst is very valuable. However, at present, no structure of the well-studied BVMO, CHMO, is available. In order to obtain a predictive structural model of the active site of CHMO, a cubic space model has been built based on the substrate acceptance of this biocatalyst.53 While this model can predict to some extent the specificity and selectivity of CHMO, it is of little use for enzyme redesign studies. However, even if no structural information is available, it is still possible to exploit random mutagenesis methods in order to change specific biocatalytic properties of an enzyme. Such an approach of directed evolution has been performed with CHMO by the Reetz group.54 By error-prone PCR, a library of CHMO mutants was created. By using chiral GC analysis, a library of 10 000 mutants was screened for improved enantioselectivity using the prochiral substrate 4-hydroxycyclohexanone. Upon Baeyer–Villiger oxidation by wild-type CHMO, this ketone is converted into both corresponding lactone enantiomers (only 9% ee for the (R)-enantiomer). The same library was also used to find mutants that show an improved enantioselectivity with methyl-p-methylbenzyl thioether. This sulfide is readily oxidized by wildtype CHMO forming the (R-sulfoxide in low enantiomeric excess (ee = 14%). Although only a relatively small library of mutants was screened due to the limited screening efficiency (800 mutants per day), impressive results were obtained. For both reactions, multiple improved mutants were found that could be improved even further in a second round of mutagenesis. For the Baeyer–Villiger reaction, the enantioselectivity could be improved from 9 to 90% ee while mutants with the opposite enantioselectivity could also be retrieved. For the sulfoxidation reaction, several mutants were found that showed an excellent enantioselective behavior (>98% ee). Interestingly, both screens yielded mutants with improved biocatalytic properties in which only one specific residue was replaced: F432. Most other retrieved mutants contained multiple mutations suggesting that multiple mutations (additive effects) are needed to improve enantioselectivity.
120
Daniel E. Torres Pazmiño, Marco W. Fraaije
Except for the directed evolution study on CHMO, only a few other enzyme redesign studies on BVMOs have been reported. By sequence alignment of BVMO sequences, residues have been identified as targets to change the coenzyme speci ficity of CHMO and HAPMO.55 As mentioned above, most BVMOs are only active with the relatively expensive coenzyme NADPH, thereby compromising costeffective biocatalytic applications. It would be advantageous to engineer BVMOs which are active with NADH. By site-directed mutagenesis, it was found that one specific residue (K326 in CHMO) is crucially involved in recognizing the 2 -phosphate of NADPH. This lysine residue is conserved in all BVMO sequences which is in line with the observation that all type I BVMOs are NADPH-specific. However, while it was possible to change the selectivity of both CHMO and HAPMO toward NADH, all prepared mutants showed a modest to low activity with NADH. This indicates that a more thorough enzyme redesign strategy is needed in which the presently available structural information could be exploited (see below). Another site-directed mutagenesis study was performed by Cheesman et al.56 in which all histidine residues in CHMO were replaced to probe their respective function. This revealed that replacement of H59 prevented expression while the H163Q mutant exhibited only 10% activity. This latter observation is in line with the fact that H163 is part of the BVMO-specific motif mentioned above. Replacing the corresponding histidine in HAPMO also affected the activity of the respective enzyme strongly.39 Other sequence–function relationship data concerning a type I BVMO come from studies that focused on elucidating the genetic basis of drug resistance of M. tuberculosis isolates toward thioamides.4546 It was found that a range of point mutations in a specific gene, etaA, were associated with thioamide drug resistance. The resulting drug resistance can be ascribed to the inactivation of the corresponding enzyme: ethionamide monooxygenase. Future studies will reveal whether the reported mutations are fatal for expression or result in inactive enzyme. Ethionamide monooxygenase represents a type I BVMO and is able to convert a range of ketones into the corresponding esters.47 Except for catalyzing Baeyer– Villiger oxidations, the enzyme is also able to oxidize the sulfide moieties of several antitubercular thioamide drugs. The oxidized drugs appear to be highly toxic for mycobacteria. This indicates that ethionamide monooxygenase acts as a prodrug activator. In vitro, ethionamide monooxygenase only displays a very low activity with all tested substrates. This low activity may be due to the fact that all tested substrates are unrelated to the (unknown) physiological substrate of ethionamide monooxygenase. However, it might also indicate that ethionamide monooxygenase needs other components to be fully active. As mentioned above, the first BVMOs were already purified several decades ago. Subsequent biochemical studies have revealed that these enzymes are typi cally soluble and often easy to express at high levels in, e.g., E. coli. These features suggest that BVMOs are perfect candidates for X-ray crystallography studies. However, crystallization of several type I BVMOs has been attempted and proven
Discovery, redesign and applications of BVMOs
121
to be difficult. Several groups have tried to crystallize the prototype BVMO: CHMO from Acinetobacter. However, all attempts have been without success, which is probably caused by the relative instability of this specific monooxy genase. This is in line with the observation that CHMO is sensitive to cysteine oxidation leading to enzyme inactivation.57 Also, ethionamide monooxygenase has been subjected to crystallization trials. Unfortunately, the recombinant enzyme withstands crystallization as it tends to aggregate in its pure form (M. W. Fraaije, personal communication). We have successfully crystallized HAPMO resulting in large well-shaped yellow crystals. However, X-ray analysis of these crystals revealed that the diffracting power was very poor with a resolution limit of 6 Å on a synchrotron source. As flexibility in N-termini has been shown to introduce heterogeneity in protein crystals, truncation could be a solution for the diffraction problem as HAPMO contains an extended N-terminus when compared with other BVMOs.58 To test this, several truncated HAPMO mutants were also crystallized.52 However, no improvement in diffraction properties was found. Several BVMOs have been examined with respect to their stability. This has revealed that CHMO is not a very stable biocatalyst. The enzyme is rapidly degraded intracellularly when expressed in E. coli.59 In its isolated form, the enzyme is also quite unstable: t1/2 = 24 h at 25 C.60 Additives, e.g., kosmotropic salts, have a stabilizing effect but are not desirable for biocatalytic applications. It was also reported that immobilization on a solid carrier is a more effective way to stabilize CHMO. HAPMO was shown to inactivate rapidly at elevated temperatures: t1/2 = 80 min at 36 C.16 For this BVMO, it was shown that the lifetime could be increased 4-fold by adding the FAD cofactor. Addition of a NADPH coenzyme analog resulted in a more drastic effect as the enzyme remained fully active for 120 min at 36 C. While the addition of a coenzyme analog will inhibit efficient catalysis, it can be used as additive when a BVMO has to be stored. In order to obtain a more robust BVMO, we decided to search for a BVMO gene from a thermophilic microorganism. While no BVMO genes could be identified in hyperthermophilic archaea (Table 2), we discovered two genes in the genome of the semi-thermophile T. fusca (see above). One of the identified BVMOs was found to be overexpressed in E. coli as a soluble, fully flavinylated and active enzyme.17 As indicated above, the enzyme was found to be highly active on phenylacetone and is thermostable. Only at temperatures above 50 C, it tends to inactivate (t1/2 = 24 h at 52 C). Possibly due to its robustness, PAMO readily crystallizes yielding crystals with good diffraction properties (<2Å). This resulted in the elucidation of its crystal structure in 2004.61 The availability of the PAMO structure offers new possibilities for redesigning this or other BVMOs. One structure-inspired enzyme redesign study concerning PAMO has already been reported.62 By sequence alignment of CHMO and PAMO, it was deduced that PAMO contains an extended active site loop when compared with CHMO. It was probed whether, by deleting one or two loop residues, the substrate acceptance of PAMO could be altered. The designed mutants indeed brought about significant changes in substrate acceptance.
122
Daniel E. Torres Pazmiño, Marco W. Fraaije
While wild-type PAMO was unable to convert 2-phenylcyclohexanone efficiently, all deletion mutants readily accepted this ketone as substrate. All mutants also displayed a similar thermostability when compared with the parent enzyme. The most active mutant (deletion of S441 and A442) was used for examining its enantioselective properties. It was found that the mutant preferably formed the (R) enantiomer of the corresponding lactone (E = 100). While CHMO also shows a similar enantioselective behavior, this PAMO deletion mutant is a better candidate for future applications due to its superior stability. This clearly demonstrates that PAMO can be used as parent enzyme to design thermostable BVMO variants. It also illustrates that the available crystal structure of PAMO will be of great help for BVMO redesign efforts.5
3. CONCLUSIONS: FUTURE DIRECTIONS As each BVMO is limited in substrate specificity, it is crucial to have a large collection of these oxidative biocatalysts available. Except for expanding the scope of possible reactions, a large toolbox of BVMOs would also increase the chance of being able to perform any wanted specific chemo-, regio- and/or enantioselective reaction. This contrasts with the present situation as only a relatively small number of BVMOs can be exploited for biocatalytic purposes. Therefore, it is still crucial to discover or engineer BVMOs with novel biocatalytic properties. An obvious way to generate new BVMOs is by taking advantage of the genome sequence information: genome mining. Table 2 illustrates that many puta tive BVMO genes can be identified which can be explored for their catalytic potential. Two examples of such a genome mining approach have already been reported yielding two novel BVMOs with interesting biocatalytic properties (see earlier1744 ). However, it also shows the risks of such an approach. Of the six putative BVMO genes from M. tuberculosis that were cloned, only four resulted in significant expression of the corresponding protein.44 Of the four expressed pro teins, only one showed an interesting enzymatic activity. These are complications inherent to genome mining. It is still impossible to predict if a gene can be easily expressed in a certain host nor what specific activities it will exhibit. Nevertheless, as hundreds of unexplored putative BVMO genes can be identified in the genome database, it is attractive to delve into this wealth of sequence information. An effective way to clone novel BVMO genes from sequenced genomes is by using gene synthesis. This technology is becoming relatively inexpensive and allows, e.g., codon optimization for specific hosts. It also circumvents the need to obtain a specific organism or its genomic DNA. By this, obtaining novel BVMOs from sequenced genomes will become more efficient and will yield new BVMOs in the near future. The recently developed methodology for enzyme discovery that is based on “random” DNA isolation and subsequent screening (metagenome mining) is
Discovery, redesign and applications of BVMOs
123
not yet applicable for BVMO discovery. So far, no screening method has been described that is effective enough to allow screening of massive metagenomic gene libraries. As soon as this technological hurdle has been crossed, the number of available biocatalytically relevant BVMOs will grow. Another more sophisticated way of obtaining novel BVMOs is to redesign a specific BVMO in order to tune catalytic properties. Such an approach would ideally yield biocatalysts tailor-made to perform any wanted reaction. Random mutagenesis methods have been very popular in the last decade to obtain enzyme variants with improved biocatalytic properties. However, these directed evolution methods typically involve creation of huge libraries of enzyme mutants that have to be screened for a newly introduced characteristic. As a consequence, efficient screening techniques are required to fully screen these libraries. However, methods enabling ultra high-throughput screening are often unavailable and as a result only relatively small mutant libraries are screened. This limits the extent to which enzyme properties can be changed. One cannot expect that by introducing a few random mutations, enzymatic properties will fundamentally change.63 The recently reported directed evolution study on CHMO nicely exemplifies these limitations. As discussed above, screening of a CHMO mutant library for mutants that exhibit altered enantioselective behavior is currently feasible only at low throughput (800 clones per day). A relatively small library (10 000 clones) was screened for CHMO mutants with altered enantioselectivity when converting a ketone or a sulfide.54 The wild-type enzyme already showed some enantioselectivity on both test compounds and therefore the desired change in enzyme reactivity represent energetically seen a small change. Nevertheless, the approach was very successful as several dozens of mutants were found displaying improved enantioselectivity. Several mutants were sequenced and allowed identification of the replaced residues. The mutants represented single, double and triple mutants, indicating effective error-prone PCR. One residue was found in both screens and suggests that this is a hotspot in tuning enantioselective Baeyer–Villiger oxidations and sulfoxidations. As a structure of a sequence-related BVMO, PAMO (40% sequence identity), is available at present, it is now possible to build a homology model structure of CHMO and locate the observed random mutations. We have recently built such a CHMO model structure (unpublished results). Interestingly, inspection of the model reveals that most of the observed mutations are clustered in a specific part of the structure located at the re side of the flavin cofactor (Fig. 6). When disregarding mutations of surface residues in double or triple mutants, only one mutation was found in the NADPH-binding domain, three mutations were located in the helical domain while all other (10) mutations were part of the interior of the FAD-binding domain. All the observed mutations in the FADbinding domain are in loops and are close to the predicted substrate-binding pocket next to the flavin.61 Mutations in the same structural region in PAMO introduced by Bocola et al.62 confirmed that this is a hotspot for affecting sub strate specificity and enantioselectivity. These results attest to the effectiveness
124
Daniel E. Torres Pazmiño, Marco W. Fraaije
Figure 6: Model structure of CHMO. The FAD cofactor is shown as sticks. Mutations observed in a directed evolution study resulting in altered enantioselectivities are highlighted (spheres).54 Note that a number of mutations were found in double/triple mutants and cannot be linked with certainty to the altered catalytic properties. Mutations in double/triple mutants that are on the surface were omitted.
of directed evolution not only to generate valuable new biocatalysts but also to identify residues or structural regions that are important for enzyme functioning. By combining the information obtained by mutagenesis studies and the available structure of PAMO, more directed enzyme redesign studies can be performed in the near future. Such structure-inspired enzyme redesign efforts should effectively generate BVMO variants performing reactions that are yet inaccessible. As CHMO and CPMO homologs are closely related to PAMO at protein sequence level, it is feasible to build structural models of these biocatalysts. Except for tuning substrate specificity or regio- and/or enantioselectivity, it should also be possible to, e.g., engineer mutant enzymes that accept NADH as coenzyme. By solving the structure of PAMO, the residues interacting with the 2 -phosphate of NADPH have also been identified. This could fuel biocatalytic applications that involve usage of isolated enzymes. Generation of a BVMO that can both use NADH or NADPH might also be beneficial for biocatalysis based on whole cells. Except for extending the toolbox of type I BVMOs, it can also be worthwhile to explore other types of Baeyer–Villiger catalysts. As mentioned above, type II BVMOs have hardly been explored and new types of BVMOs have been discov ered in recent years. Furthermore, Baeyer–Villiger oxidation activity has also been introduced into enzymes that normally catalyze other (hydrolytic) reactions.64 This
Discovery, redesign and applications of BVMOs
125
indicates that the field of BVMO discovery and engineering is still expanding and it is expected that the number of biocatalytic applications based on these oxidative biocatalysts will grow accordingly.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Baeyer, A.; Villiger, V. Ber. Dtsch. Chem. Ges. 1899, 32, 3625–3633. Turfitt, G. E. Biochem. J. 1948, 42, 376–383. Conrad, H. E.; DuBus, R.; Namtvedt, M. J.; Gunsalus, I. C. J. Biol. Chem. 1965, 240, 495–503. Forney, F. W.; Markovetz, A. J. Biochem. Biophys. Res. Commun. 1969, 37, 31–38. Kamerbeek, N. M.; Janssen, D. B.; van Berkel, W. J. H.; Fraaije, M. W. Adv. Synth. Catal. 2003, 345, 667–678. Willetts, A. TIBTECH 1997, 15, 55–62. Wierenga, R. K.; Terpstra, P.; Hol, W. G. J. J. Mol. Biol. 1986, 187, 101–107. van Berkel, W. J. H.; Kamerbeek, N. M.; Fraaije, M. W. J. Biotechnol. 2006, 124, 670–689. Rodríguez, D.; Quirós, L. M.; Braña, A. F.; Salas, J. A. J. Bacteriol. 2003, 185, 3962–3965. Wang, C.; Gibson, M.; Rohr, J.; Oliveira, M. A. Acta Crystallogr. 2005, 61, 1023–1026. Nomura, T.; Kushiro, T.; Yokota, T.; Kamiya, Y.; Bishop, G. J.; Yamaguchi, S. J. Biol. Chem. 2005, 280, 17873–17879. Fischer, R. T.; Trzaskos, J. M.; Magolda, R. L.; Ko, S. S.; Brosz, C. S.; Larsen, B. J. Biol. Chem. 1991, 266, 6124–6132. Chen, Y. C.; Peoples, O. P.; Walsh, C. T. J. Bacteriol. 1988, 170, 781–789. Ryerson, C. C.; Ballou, D. P.; Walsh, C. Biochemistry 1982, 21, 2644–2655. Sheng, D.; Ballou, D. P.; Massey, V. Biochemistry 2001, 40, 11156–11167. van den Heuvel, R. H. H.; Tahallah, N.; Kamerbeek, N. M.; Fraaije, M. W.; van Berkel, W. J. H.; Janssen, D. B.; Heck, A. J. R. J. Biol. Chem. 2005, 280, 32115–32121. Fraaije, M. W.; Wu, J.; Heuts, D. P. H. M.; van Hellemond, E. W.; Lutje Spelberg, J. H.; Janssen, D. B. Appl. Microbiol. Biotechnol. 2005, 66, 393–400. de Gonzalo, G.; Ottolina, G.; Carrea, G.; Fraaije, M. W. Chem. Commun. 2005, 29, 3724–3726. Donoghue, N. A.; Norris, D. B.; Trudgill, P. W. Eur. J. Biochem. 1976, 63, 175–192. Griffin, M.; Trudgill, P. W. Eur. J. Biochem. 1976, 63, 199–209. Kostichka, K.; Thomas, S. M.; Gibson, K. J.; Nagarajan, V.; Cheng, Q. J. Bacteriol. 2001, 183, 6478–6486. Iwaki, H.; Wang, S.; Grosse, S.; Bergeron, H.; Nagahashi, A.; Lertvorachon, J.; Yang, J.; Konishi, Y.; Hasegawa, Y.; Lau, P. C. K. Appl. Environ. Microbiol. 2006, 72, 2707–2720. Kamerbeek, N. M.; Moonen, M. J. H.; van Der Ven, J. G. M.; van Berkel, W. J. H.; Fraaije, M. W.; Janssen, D. B. Eur. J. Biochem. 2001, 268, 2547–2557. Morii, S.; Sawamoto, S.; Yamauchi, Y.; Miyamoto, M.; Iwami, M.; Itagaki, E. J. Biochem. 1999, 126, 624–631. Mihovilovic, M. D.; Müller, B.; Stanetty, P. Eur. J. Org. Chem. 2002, 22, 3711–3730. Wang, S.; Kayser, M. M.; Iwaki, H.; Lau, P. C. K. J. Mol. Catal. B Enzym. 2003, 22, 211–218. Mihovilovic, M. D.; Müller, B.; Schulze, A.; Stanetty, P.; Kayser, M. M. Eur. J. Org. Chem. 2003, 12, 2243–2249. Mihovilovic, M. D.; Rudroff, F.; Grötzl, B.; Stanetty, P. Eur. J. Org. Chem. 2005, 5, 809–816. Mihovilovic, M. D.; Snajdrova, R.; Grötzl, B. J. Mol. Catal. B Enzym. 2006, 39, 135–140. Hilker, I.; Baldwin, C.; Alphand, V.; Furstoss, R.; Woodley, J. M.; Wohlgemuth, R. Biotechnol. Bioeng. 2006, 93, 1138–1144.
126
Daniel E. Torres Pazmiño, Marco W. Fraaije
31. Alphand, V.; Carrea, G.; Wohlgemuth, R.; Furstoss, R.; Woodley, J. M. TIBTECH 2003, 21, 318–323. 32. Hilker, I.; Wohlgemuth, R.; Alphand, V.; Furstoss, R. Biotechnol. Bioeng. 2005, 92, 702–710. 33. Kyte, B. G.; Rouvière, P. E.; Cheng, Q.; Stewart, J. D. J. Org. Chem 2004, 69, 12–17. 34. Mihovilovic, M. D.; Rudroff, F.; Grötzl, B.; Kapitan, P.; Snajdrova, R.; Rydz, J.; Mach, R. Angew. Chem. Int. Ed. 2005, 44, 3609–3613. 35. de Gonzalo, G.; Torres Pazmiño, D. E.; Ottolina, G.; Fraaije, M. W.; Carrea, G. Tetrahedron Asymm. 2006, 17, 130–135. 36. de Gonzalo, G.; Torres Pazmiño, D. E.; Ottolina, G.; Fraaije, M. W.; Carrea, G. Tetrahedron Asymm. 2005, 16, 3077–3083. 37. Mihovilovic, M. D.; Kapitan, P.; Rydz, J.; Rudroff, F.; Ogink, F. H.; Fraaije, M. W. J. Mol. Catal. B Enzym. 2005, 32, 135–140. 38. Venter, J. C.; Remington, K.; Heidelberg, J. F.; Halpern, A. L.; Rusch, D.; Eisen, J. A.; Wu, D. Y.; Paulsen, I.; Nelson, K. E.; Nelson, W.; Fouts, D. E.; Levy, S.; Knap, A. H.; Lomas, M. W.; Nealson, K.; White, O.; Peterson, J.; Hoffman, J.; Parsons, R.; Baden-Tillson, H.; Pfannkoch, C.; Rogers, Y. H.; Smith, H. O. Science 2004, 304, 66–74. 39. Fraaije, M. W.; Kamerbeek, N. M.; van Berkel, W. J. H.; Janssen, D. B. FEBS Lett. 2002, 518, 43–47. 40. van Beilen, J. B.; Mourlane, F.; Seeger, M. A.; Kovac, J.; Li, Z.; Smits, T. H. M.; Fritsche, U.; Witholt, B. Environ. Microbiol. 2003, 5, 174–182. 41. Iwaki, H.; Hasegawa, Y.; Wang, S.; Kayser, M. M.; Lau, P. C. K. Appl. Environ. Microbiol. 2002, 68, 5671–5684. 42. Brzostowicz, P. C.; Blasko, M. S.; Rouvière, P. E. Appl. Microbiol. Biotechnol. 2004, 58, 781–789. 43. de Gonzalo, G.; Ottolina, G.; Zambianchi, F.; Fraaije, M. W.; Carrea, G. J. Mol. Catal. B Enzym. 2006, 39, 91–97. 44. Bonsor, D.; Butz, S. F.; Solomons, J.; Grant, S.; Fairlamb, I. J. S.; Fogg, M. J.; Grogan, G. Org. Biomol. Chem. 2006, 4, 1252–1260. 45. Baulard, A. R.; Betts, J. C.; Engohang-Ndong, J.; Quan, S.; McAdam, R. A.; Brennan, P. J.; Locht, C.; Besra, G. S. J. Biol. Chem. 2000, 275, 28326–28331. 46. DeBarber, A. E.; Mdluli, K.; Bosman, M.; Bekker, L. G.; Barry, C. E. Proc. Natl Acad. Sci. USA 2000, 97, 9677–9682. 47. Fraaije, M. W.; Kamerbeek, N. M.; Heidekamp, A. J.; Fortin, R.; Janssen, D. B. J. Biol. Chem. 2004, 279, 3354–3360. 48. Sicard, R.; Chen, L. S.; Marsaioli, A. J.; Reymond, J.-L. Adv. Synth. Catal. 2005, 347, 1041–1050. 49. Gutiérrez, M. C.; Sleegers, A.; Simpson, H. D.; Alphand, V.; Furstoss, R. Org. Biomol. Chem. 2003, 1, 3500–3506. 50. Bicalho, B.; Chen, L. S.; Grognux, J.; Reymond, J.-L.; Marsaioli, A. J. J. Braz. Chem. Soc. 2004, 15, 911–916. 51. Watts, A. B.; Beecher, J.; Whitcher, C. S.; Littlechild, J. A. Biocatal. Biotransform. 2002, 20, 209–214. 52. Kamerbeek, N. M. Ph.D.Thesis, University of Groningen, 2004. 53. Ottolina, G.; Carrea, G.; Colonna, S.; Rückemann, A. Tetrahedron Asymm. 1996, 7, 1123–1136. 54. Mihovilovic, M. D.; Rudroff, F.; Winninger, A.; Schneider, T.; Schulz, F.; Reetz, M. T. Org. Lett. 2006, 8, 1221–1224. 55. Kamerbeek, N. M.; Fraaije, M. W.; Janssen, D. B. Eur. J. Biochem. 2004, 271, 2107–2116. 56. Cheesman, M. J.; Kneller, M. B.; Rettie, A. E. ChemicoBiol. Interact. 2003, 146, 157–164. 57. Bennett, A. J. Undergr. Res. 2004, 6 (http://www.clas.ufl.edu/jur/200409/papers/paper_bennett. html). 58. Price, S. R.; Nagai, K. Curr. Opin. Biotech. 1995, 6, 425–430.
Discovery, redesign and applications of BVMOs
127
59. Walton, A. Z.; Stewart, J. D. Biotechnol. Prog. 2002, 18, 262–268. 60. Zambianchi, F.; Pasta, P.; Carrea, G.; Colonna, S.; Gaggero, N.; Woodley, J. M. Biotechnol. Bioeng. 2002, 78, 489–496. 61. Malito, E.; Alfieri, A.; Fraaije, M. W.; Mattevi, A. Proc. Natl Acad. Sci. USA 2004, 101, 13157–13162. 62. Bocola, M.; Schulz, F.; Leca, F.; Vogel, A.; Fraaije, M. W.; Reetz, M. T. Adv. Synth. Catal. 2005, 347, 979–986. 63. Morley, K. L.; Kazlauskas, R. J. TIBTECH 2005, 23, 231–237. 64. Carboni-Oerlemans, C.; Domínguez de María, P.; Tuin, B.; Bargeman, G.; Van der Meer, A.; Van Gemert, R. J. Biotech., 2006, 126, 140–151.
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
129
Chapter 7
Enzymes in aldoxime–nitrile pathway: versatile tools in biocatalysis Yasuhisa Asano Biotechnology Research Center, and Department of Biotechnology, Faculty of Engineering, Toyama Prefectural University, Imizu, Toyama, Japan
Abstract The discovery and exploitation of enzymes in aldoxime–nitrile pathway: nitrile hydratase, amidase, nitrilase, aldoxime dehydratase, etc., are shown along with the use of method ologies, such as organic chemistry, microbial screening by enrichment and acclimation culture techniques, enzyme purification, gene cloning, molecular screening by polymerase chain reaction (PCR).
1. INTRODUCTION The microbial and enzymatic production of useful materials has been successfully applied to several chemical processes1 and has become not only one of the most important research areas in synthetic organic chemistry, but also one of the most hot scientific topics in “Green Chemistry” in which materials are expected to be efficiently used, emissions and waste reduced and inherently safer processes sought.2 The recent success in microbial transformation has been based on the screen ing for microbial enzymes catalyzing new reactions or screening known enzymes for an unknown activity with synthetic substrates. Although the advantages of the use of enzymes in organic synthesis have been fully illustrated, not much attention has been paid to the systematic exploitation of microorganisms producing new enzymes.3 The isolation of microorganisms based on wide biological diversity is very important to initiate the study of the microbial transformation of chemicals and exploit new enzymes.
2. SCREENING FOR NEW MICROBIAL ENZYMES BY ENRICHMENT AND ACCLIMATION CULTURE TECHNIQUES Little is known about the substrate specificities of most of the enzymes as compared with common reagents used in organic synthesis. The velocity of the enzymatic
130
Yasuhisa Asano
reactions is much affected by the neighboring substituents: it cannot be predicted whether an unnatural chemical can be a substrate or not with most of the enzymes. Therefore, we screen microorganisms for an enzyme when (i) no information is available for the desirable reaction, (ii) nothing is known about the specific reaction but a homologous reaction is known to occur, (iii) an enzyme is known only in another biological source, (iv) insufficient amount of the enzyme is produced by a certain organism for a practical transformation, (v) increased stability of the enzyme for practical use is required, etc. Enrichment culture is a technique to isolate microorganisms having special growth characteristics.3 Some microorganisms grow faster than others in media with limited nutrients, high temperature, extreme pH values, etc. Microorgan isms that grow faster than other species become dominant after several transfers of the culture. The isolation method utilizing the difference of growth veloc ity gives us only limited group of microorganisms, since only a few kinds of fast-growing microorganisms usually appear on the plate containing a starting material in the successive transfer of microorganisms in the enrichment cul ture. The isolates are classified according to their biochemical and morphological characteristics. At the same time, microorganisms deposited among the culture collections are tested for the activity, dependent on or independent of the iso lation experiments. During such studies, there may be unexpected findings and discovery. An acclimation technique3 is applied when a toxic or unnatural compound is used as a substrate and usually run long term to isolate microorganisms which are not easily isolated by the enrichment culture. An adaptation to a syn thetic medium containing a target compound sometimes results in isolation of microorganisms having a desirable new enzyme. Genetic changes in the microor ganisms may be expected. Since fast-growing microorganisms do not always possess the desired enzyme activity and suppress the growth of other microor ganisms, one may look for microorganisms growing rather slowly on certain compounds by methods such as spreading diluted soil sample directly to the medium allowing an isolation of microorganisms having an ability to convert the compound. If the product is located in the downstream of the starting material in the metabolism, enzymes converting such compounds could be screened. In some cases, a new metabolic pathway of the compound may be discovered during the screening. Many industrial microorganisms have been isolated in such ways. How ever, when one wants to synthesize compounds located at the end of anabolism, such as polysaccharides, whose synthesis requires multistep enzyme reactions and energy, the strategy cannot be applied, since a compound with higher energy could be synthesized by coupling of an energy-yielding reaction. Oxidoreductases uti lizing NAD+ or NADP+ as cofactors are also useful in the oxidation or reduction reactions because of their reversible nature.
Enzymes in aldoxime–nitrile pathway
131
3. DEVELOPMENT OF NITRILE-DEGRADING ENZYMES4�5 Nitrile hydratase (NHase, EC 4.2.1.84) has become one of the most important industrial enzymes, although it was discovered during the studies on the microbial degradation of polyacrylonitrile. The properties of NHase and its nature to be able to accumulate extraordinarily high concentration of amides in a strikingly fast manner were discovered after a number of planning, experiments and assumptions. Although the details could be reported elsewhere, they are summarized as follows: (i) at first, polyacrylonitrile was a target of microbial degradation and there has been no description of an enzyme hydrating nitrile activity, although such reac tions have been well described in organic chemistry; (ii) equivalent transforming thinking was done, searching for degraders of oligomer or monomer rather than polyacrylonitrile degrader; (iii) an acclimation technique to isolate acrylonitrile, one of the most toxic nitriles, degrading microorganisms was used; (iv) purification of nitrile hydratase and its characterization were done for the first time, and a proof of involvement of amidase in nitrile hydrolysis to carboxylic acid was obtained; (v) equivalent transforming thinking was done to search for amide-accumulating microorganism among various nitrile-degrading microorganisms based on the dif ference in the substrate specificities of nitrile hydratase and amidase, since the only acrylonitrile degrader isolated was proved to have a nitrilase; (vi) discovery of amide-producing microorganisms in high amount, etc.4
4. SCREENING FOR HEAT-STABLE NHASE It goes without saying that not all microorganisms have been isolated by human beings. Once there is a report on a new enzyme produced by mesophiles, ther mostable enzymes catalyzing the same reaction could be sought from the environ ment. To seek for thermophilic microorganisms, volcano, hot springs, chimneys at the bottom of the sea, etc., are the sources of isolation of such microorganisms. With the increasing knowledge in genomic information, screening for new bio catalysts based on various aspects of similarities (primary, secondary and tertiary structures, motifs and domain structures of the proteins and their positions in the metabolism, etc.) from the database is also carried out. It is known that NHase used in industry has a lower optimum temperature,45 therefore many reports have been concerned with screening for thermostable NHase.6 Miyanaga et al. has succeeded to analyze the X-ray structure of such a thermostable NHase from Pseudomonas thermophila. Bacillus sp. BR449 pro ducing NHase with optimum temperature of 55 C has been isolated. Similarly, Bacillus sp. RAPc8 has a growth optimum at 65 C. Takashima et al. has isolated Bacillus smithii strain SC-J05-1, with optimum temperature at 40 C, and whose NHase has an optimum temperature and pH of 50 C and 10, respectively. Its crystal structure has also been elucidated. Bacillus pallidus strain Dac521 has
132
Yasuhisa Asano
an optimum growth temperature at 50 C. Pseudonocardia thermophila JCM3095 shows an optimum growth temperature at 60 C and NHase, whose higher structure is also solved, is stable at more than 60 C. It has been possible to use the wildtype strains of Pseudomonas chlororaphis B23 and Rhodococcus rhodochrous J-1, in the selective acrylamide production from acrylonitrile, although these recent isolates with higher optimum growth temperature would also metabolize amide produced by NHase quite fast; therefore, it would be necessary to produce amide by NHase expressed in heterologous hosts to block further metabolism of amide.
5. SCREENING FOR NHASE WITH PCR Recently, many studies have been concerned with the precise number of microor ganisms in the environment.7 The microorganisms isolated from paddy fields in Japan are less than 1% in number than those found under microscopic observa tions. It is reported that the number of culturable microorganisms isolated from the feces of human being has become almost the same in number by a microscopic result by the recent introduction of the anaerobic culture technology of microor ganisms. Previously, it was known that the number of culturable microoganisms from the feces was 0.01–0.1% of microscopic observations. It is reported that various unculturable microorganisms play important role in the removal of organic and inorganic chemicals in activated sludge. Contrary to the above-mentioned microbial screening for enzymes having unique properties, it has become possible to screen useful enzyme by PCR from the environment. It is reported that more than several micrograms of environ mental DNA is obtained from 1 g of soil. Many 16S rRNA sequences have been determined from environmental DNA. There are two possible reasons why the number of colonies is always lesser than that counted under microscopy: one is that microorganisms somehow give smaller number on the plates, the other is that there are many kinds of microorganisms which are unable to be cultivated. Evidences are appearing that the second one reflects the truth for the number of microorganisms in nature.8 Although there is difficulty in classifying the uncultur able microorganisms using the DNA database from the culturable microorganisms, many new gene clusters are discovered and characterized from environmental DNA. Therefore, it is expected that much fundamental and applied research will be carried out using the unculturable microorganisms and environmental DNA as sources of new DNA. Duran et al. recently reported a detection of new NHases using PCR.9 It is well known that there are Fe- and Co-requiring NHases. A PCR protocol was developed for the specific detection of genes coding NHase. The primer design was based on the highly conserved sequences found in the coding region of the -subunit gene of Rhodococcus erythropolis corresponding to the highly conserved sequence of the metal-binding site (326–352 amino acid of NHase gene). When
Enzymes in aldoxime–nitrile pathway
133
the genomes of R. erythropolis JCM3132 and JCM3191 were used as models, they gave PCR products with the primers. The sequences of R. erythropolis JCM3132 and JCM3191 NHases were homologous to that of the -subunit of R. rhodochrous IFO15564 NHase by 91 and 96.6%, respectively. Using the genomic DNA of some soil isolates, two types of NHase sequences were detected. The use of soil DNA (about 25 g per g soil) taken from acrylamide-producing plants, etc., gave Fe-type NHase sequences (450 bp DNA) by PCR. Using the PCR method, both types of Fe- and Co-NHase-encoding genes were detected in DNA from pure cultures and soil samples. Furthermore, consensus primers allowed rapid cloning and expression of novel NHases in Escherichia coli.
6. NITRILE SYNTHESIS USING A NEW ENZYME, ALDOXIME DEHYDRATASE 6.1. Aldoxime-converting enzymes Much studies have been carried out on nitrile-transforming enzymes, since NHase was shown to be highly chemoselective even with the whole microbial cells, and the accumulation of amides was achieved in a high concentration (up to 40%, w/v or more).4 The enzyme has been used in the industrial production of various chemicals. Not only the primary structure of the enzyme, but also the X-ray structure of NHase has been clarified.6ae−g At the beginning of the studies, NHase was considered to be a rare enzyme,4 but the following studies showed that the distribution of the enzyme is rather wide in microoganisms,6 although its physiological function is not yet clear. On the other hand, cyanogenesis is an ability of plants to release hydrogen cyanide in the cells and is considered a mechanism to protect plants from attack by fungi and predators. Cyanogenesis is believed to be distributed in over 2650 species of higher plants, and more than 20 kinds of cyanogenic glycosides are known.10 They are glycosides of aliphatic and aromatic cyanohydrins. The aglycons of the cyanohydrin have been considered to be biosynthesized from hydrophobic l-amino acids such as tyrosine, phenylalanine, valine, isoleucine, via aldoxime compounds, although enzymatic and genetic knowledge are scarce. Aldoximes are prepared from aldehydes and hydroxylamine by condensation reaction, and the dehydration reaction of aldoxime is one of the most important methods of nitrile synthesis in organic chemistry.11 We speculated that it would become one of the most important examples in “Green Chemistry” if the dehy dration reaction could be realized by an enzymatic method, and started studies on a new enzyme, aldoxime dehydratase, and its use in enzymatic nitrile synthe sis. Furthermore, we clarified the relationship between aldoxime dehydratase and nitrile-degrading enzymes in the genome of the microorganisms and the physio logical role of the enzyme.
134
Yasuhisa Asano
6.2. Isolation of microorganisms having aldoxime dehydratase activity12 By using acclimation for 3–4 months of soil microorganisms to a medium contain ing Z-phenylacetaldoxime (PAOx), a degrader Bacillus sp. OxB-1 was isolated, which has an activity of converting PAOx to phenylacetonitrile (PAN) and a strong nitrilase activity. Partial purification of the enzyme activity revealed that the enzyme catalyzed the stoichiometric conversion of PAOx to PAN. On the other hand, we isolated a degrader of E-pyridine-3-aldoxime (PyOx), Rhodococcus sp. YH3-3 from soil after 6 months of acclimation of soil. The strain contained a dehydratase converting PyOx to 3-cyanopyridine (CyPy). Different from Bacillus sp. OxB-1, Rhodococcus sp. YH3-3 possessed nitrile hydratase and amidase rather than nitrilase, active in the metabolism of CyPy. 6.3. Purification, characterization and primary structure determination of aldoxime dehydratase13 A novel dehydratase that catalyzes the stoichiometric dehydration of Z-phenylacetaldoxime to phenylacetonitrile has been purified to homogeneity from a cell-free extract of Bacillus sp. strain OxB-1 isolated from soil. It has a Mr of about 40 000 and is composed of a single polypeptide chain with a loosely bound protoheme IX. The enzyme is inactive unless FMN is added to the assay, but low activity is also observed when sulfite replaces FMN. The enzyme is active with arylalkylaldoximes and to a lesser extent with alkylaldoximes. The enzyme prefers the Z-form of phenylacetaldoxime over its E-isomer. The gene coding for the enzyme was cloned, and a 1053 bp open reading frame that codes for 351 amino acid residues was identified as the oxd gene. A nitrilase, which participates in aldoxime metabolism in the organism, was found to be coded by the region just upstream from the oxd gene. In addition, an open reading frame (orf2), whose gene product is similar to bacterial regulatory (DNA-binding) proteins, was found just upstream from the coding region of the nitrilase. These findings provide genetic evidence for a novel gene cluster that is responsible for aldoxime metabolism in this microorganism. With these results, we named the new enzyme as phenylacetaldoxime dehy dratase (EC 4.99.1.7). It was also suggested that the enzyme utilizes FMN as an electron acceptor, because the Vmax value was increased about five times under anaerobic condition and the sulfite ion could replace FMN, although the enzyme requires oxidized form of FMN. It was revealed that the enzyme is a quite unique enzyme whose apparent function is to catalyze a dehydration reaction. The reaction mechanism is of much interest.14 The gene for the dehydratase was expressed in E. coli under lac promoter, and an expression plasmid pOxD 9OF was constructed. The transformant was cul tivated under optimal condition at 30 C when much of the enzyme was expressed in a soluble form with more than thousand and several hundred times than the wild-type strain per culture (up to more than 50% of the total soluble protein of the
Enzymes in aldoxime–nitrile pathway Aldoxime OH N dehydratase
R H
Aldoxime
H2O
R C N
Nitrile Nitrile hydratase
135
Nitrilase
O R C OH
Carboxylic acid O R C NH2
Amidase
Amide
Figure 1: Microbial aldoxime-nitrile pathway. transformant). On the other hand, when the transformant was cultivated at 37 C, the enzyme was expressed as inclusion body. We also found that the occurrence of aldoxime dehydratase is as wide as that for nitrile-degrading enzymes such as nitrile hydratase, amidase and/or nitri lase. All of the nitrile degraders hitherto isolated contained aldoxime dehydratase activities. The author would like to propose that the pathway in which aldoximes are successively degraded via nitrile could be named as the “aldoxime–nitrile pathway” (Fig. 1). 6.4. Synthesis of nitriles from aldoxime with aldoxime dehydratase15 The dehydration reaction of aldoxime to form nitriles using the resting cells of Rhodococcus sp. YH3-3 was optimized. We found that the enzyme was induced by aldoxime and catalyzed the stoichiometric synthesis of nitriles from aldoximes at pH 7.0 and 30 C. Phenylacetonitrile once synthesized from phenylacetaldoxime was hydrolyzed to phenylacetic acid, since the strain has nitrile degradation enzymes such as nitrile hydratase and amidase. We have been successful in syn thesizing phenylacetonitrile and other nitriles stoichiometrically by a selective inactivation of nitrile hydratase by heating the cells at 40 C for 1 h. Various nitriles were synthesized under optimized conditions from aldoximes in good yields. On the other hand, 3-phenylpropionitrile was synthesized from Z-3-phenyl propionaldoxime (0.75 M) in a quantitative yield (98 g l−1 ) by the use of cells of E. coli JM 109/pOxD-9OF, a transformant harboring a gene for a new enzyme, phenylacetaldoxime dehydratase, from Bacillus sp. strain OxB 1. Other arylalkyl- and alkyl-nitriles were also synthesized in high yields from the corresponding aldoximes. Moreover, 3-phenylpropionitrile was success fully synthesized by the recombinant cells in 70 and 100% yields from 0.1 M unpurified E/Z-3-phenylpropionaldoxime, which is spontaneously formed from 3-phenylpropionaldehyde and hydroxylamine in a butyl acetate/water biphasic system and aqueous phase, respectively. This is the first report of synthesizing nitriles by an enzymatic method, a new technique of green chemistry.
136
Yasuhisa Asano
6.5. Distribution of aldoxime dehydratase16 The distribution of phenylacetaldoxime-degrading and pyridine-3-aldoxime degrading ability was examined with intact cells of 975 microorganisms, including 45 genera of bacteria, 11 genera of actinomyces, 22 genera of yeasts and 37 genera of fungi, by monitoring the decrease of the aldoximes. The abilities were found to be widely distributed in bacteria, actinomyces, fungi and some yeasts: 98 and 107 strains degraded phenylacetaldoxime and pyridine-3-aldoxime, respectively. All of the active strains exhibited not only the aldoxime dehydration activity to form nitrile but also the nitrile-hydrolyzing activity. On the other hand, all of the 19 nitrile-degrading microorganisms (13 species, 7 genera) were found to exhibit aldoxime dehydration activity. It is shown that aldoxime dehydratase and nitrile-hydrolyzing activities are widely distributed among 188 aldoxime and 19 nitrile degraders and that the enzymes were induced by aldoximes or nitriles. The NHase responsible for aldoxime metabolism from the E-pyridine 3-aldoxime-degrading bacterium, Rhodococcus sp. strain YH3-3, was purified and characterized. Addition of cobalt ion was necessary for the formation of enzyme. The native enzyme had a Mr of 130 000 and consisted of two sub units (-subunit, 27 100; -subunit, 34 500). The enzyme contained approximately 2 mol cobalt per mol enzyme. The enzyme had a wide substrate specificity: it acted on aliphatic saturated and unsaturated as well as aromatic nitriles. The N-terminus of the -subunit showed good sequence similarities with those of other NHases. Thus, this NHase is part of the metabolic pathway for aldoximes in microorganisms. 6.6. Molecular screening for “aldoxime–nitrile pathway”17 A molecular screening procedure using Southern hybridization and PCR to iden tify aldoxime dehydratase (Oxd) encoding genes (oxds) among 14 aldoxime- or nitrile-degrading microorganisms was developed. When an oxd gene of R. ery thropolis N-771 was used as a probe, positive hybridization signals were seen with the chromosomal DNA of eight strains, suggesting that these strains have similar oxd genes as R. erythropolis N-771. By analyzing the PCR-amplified fragmentswith degenerate consensus primers, the occurrence of homologous Oxd coexisting with Fe-containing NHase in the active eight strains was demonstrated coinciding with the results of Southern hybridization. The whole length of oxd gene was cloned as an example from one of the positive strains, Pseudomonas sp. K-9, sequenced and expressed in E. coli. Analysis of the primary structure of the protein (OxdK) encoded by the oxd gene of Pseudomonas sp. K-9 led to identify an Oxd having a new primary structure. Thus, the PCR-based analysis of oxd gene is a useful tool to detect and analyze the “aldoxime–nitrile pathway” in nature, since Oxd is the key enzyme for the pathway.
Enzymes in aldoxime–nitrile pathway
137
7. CONCLUSIONS As described above, the procedure and some of our examples of successful isolation of microorganisms having new enzymes were briefly described. Furthermore, historic changes of the way and methods of the research on the degradation of nitrile compounds and the exploitation of some industrially important enzymes including NHase, etc., were introduced. It may be possible to see that the series of research done in three decades could be summarized as follows: (i) basic research on the microbial polyacryloni trile degradation and metabolism with an extension of knowledge from organic chemistry, (ii) an extension to use the newly discovered enzyme NHase to amide synthesis, (iii) studies on the microbial physiology and the properties of the enzymes (being not straightforward with many trials and errors in the steps (i)–(iii)), (iv) discovery of high accumulation of amides from nitriles by NHase, eventually leading to industrial acrylamide production with tremendous efforts involving many researchers of Prof. H. Yamada’s laboratory of Kyoto University and Mitsubishi Rayon Co. Ltd, (v) an extension of the knowledge from organic chemistry and plant physiology to discover “aldoxime dehydratase,” (vi) the real ization of the first enzymatic nitrile synthesis and (vii) the starting of basic studies on aldoxime dehydratase.17 On the other hand, we have been very successful in evolving a new enzyme as an industrial enzyme.18 The technology requires exactly the same strategy as in the screening for new enzymes, although some wider fluctuation of the fidelity of PCR or related molecular screening should be necessary. Recently, the genetic information has been tremendously expanded and the functions of the proteins are assumed and discussed by homology comparison of their primary and higher structures. However, it goes without saying that their functions are not fully understood by these methods, until the properties of the proteins are proven by experiments. It is one of the surest ways to express the gene and prove its functions by actual experiments, if one would really like to use the genetic resources.
ACKNOWLEDGMENTS Sincere thanks are due to Dr H. Yamada, Dr Y. Kato, staff members of my laboratory, postdoctoral fellows and students, whose names appear in References.
REFERENCES 1. Yamada, H.; Shimizu, S. Angew. Chem. Int. Ed. Engl. 1988, 27, 622–642. 2. (a) Raber, L. Chem. Eng. News 1997, 75, 7–8. (b) Wilkinson, S. L. Chem. Eng. News 1997, 75, 35–43.
138
Yasuhisa Asano
3. Asano, Y. J. Biotechnol. 2002, 94, 65–72. 4. (a) Yamada, H.; Asano, Y.; Hino, T.; Tani, Y. J. Ferment. Technol. 1979, 57, 8–14. (b) Asano, Y., Ando, S.; Tani, Y.; Yamada, H. Agric. Biol. Chem. 1980, 44, 2497–2498. (c) Asano, Y.; Tani, Y.; Yamada, H. Agric. Biol. Chem. 1980, 44, 2251–2252. (d) Asano, Y.; Ando, S.; Tani, Y.; Yamada, H. Agric. Biol. Chem. 1981, 45, 57–62. (e) Asano, Y.; Yasuda, T.; Tani, Y.; Yamada, H. Agric. Biol. Chem. 1982, 46, 1183–1189. (f) Asano, Y.; Fujishiro, K.; Tani, Y.; Yamada, H. Agric. Biol. Chem. 1982, 46, 1165–1174. (g) Bandyopadhyay, A. K.; Naga sawa, T.; Asano, Y.; Fujishiro, K.; Tani, Y.; Yamada, H. Appl. Environ. Microbiol. 1986, 51, 302–306. 5. (a) Asano, Y. Nippon Nogei Kagakukaishi 1991, 65, 1617–1626. (b) Yamada, H.; Kobayashi, M. Biosci. Biotechnol. Biochem.1996, 60, 1391–1400. 6. (a) Miyanaga, A.; Fushinobu, S.; Ito, K.; Wakagi, T. Biochem. Biophys. Res. Commun. 2001, 288, 1169–1174. (b) Padmakumar, R.; Oriel, P. Appl. Biochem. Biotechnol. 1999, 77–79, 671–679. (c) Pereira, R. A.; Graham, D.; Rainey, F. A.; Cowan, D. A. Extremophiles 1998, 2, 347–335. (d) Takashima, Y.; Yamaga, Y.; Mitsuda, S. J. Ind. Microbiol. Biotechnol. 1998, 20, 220– 226. (e) Hourai, S.; Miki, M.; Takashima, Y.; Mitsuda, S.; Yanagi, K. Biochem. Biophys. Res. Commun. 2003, 312, 340–345. (f) Cramp, R.; Gilmour, M.; Cowan, D. A. Microbiology 1997, 143, 2313–2320. (g) Yamaki, T.; Oikawa, T.; Ito, K.; Nakamura, T. J. Ferment. Bioeng. 1997, 83, 474–477. (f) Huang, W.; Jia, J.; Cummings, J.; Nelson, M.; Schneider, G.; Lindqvist, Y. Structure 1997, 5, 691–699. (g) Nakasako, M.; Odaka, M.; Yohda, M.; Dohmae, N.; Takio, K.; Kamiya, N.; Endo, I. Biochemistry 1999, 38, 9887–9898. 7. Hattori, M. Biseibutsugaku no Kiso;Gakkai syuppan Center: Tokyo, 1986. 8. (a) Ueda, T.; Suga, Y.; Matsuguchi, T. Eur. J. Soil. Sci. 1995, 46, 415–421. (b) Suzuki, M. T.; Rappe, M. S.; Haimberger, Z. W.; Winfield, H.; Adair, N.; Strobel, J.; Giovannoni, S. J. Appl. Environ. Microbiol. 1997, 63, 983–989. (c) Hugenholtz, P.; Goebel, B. M.; Pace, N. R. J. Bacteriol. 1998, 180, 4765–4774. 9. Precigou, S.; Goulas, P.; Duran, R. FEMS Microbiol. Lett. 2001, 204, 155–161. 10. (a) Asano, Y.; Tamura, K.; Doi, N.; Ueatrongchit, T.; H-Kittikun, A.; Ohmiya, T. Biosci. Biotechnol. Biochem. 2005, 69, 2349–2357. (b) Nanda, S.; Kato, Y.; Asano, Y. Tetrahedron 2005, 61, 10908–10916. (c) Nanda, S.; Kato, Y.; Asano, Y. Tetrahedron Asymm. 2006, 17, 735–741. (d) Vetter, J. Toxicon 2000, 38, 11–36. (e) Lechtenberg, M.; Nahrstedt, A. In Ikan, R. (Ed.) Naturally occuring glycosides; Wiley: New York; 1999, pp. 147–191. 11. Maruoka, K.; Yamamoto, H. In Comprehensive Organic Synthesis; Trost, B. M. (Ed.); Pergamon Press: Oxford; Vol. 6, 1991, pp. 763–793. 12. (a) Asano, Y.; Kato, Y. FEMS Microbiol. Lett. 1998, 158, 185–190. (b) Kato, Y.; Ooi, R.; Asano, Y. Arch. Microbiol. 1998, 170, 85–90. 13. (a) Kato, Y.; Nakamura, K.; Sakiyama, H.; Mayhew, S. G.; Asano, Y. Biochemistry 2000, 39, 800–809. (b) Kato, Y., Asano, A. Protein Expr. Purif. 2003, 28, 131–139. 14. (a) Kobayashi, K.; Yoshioka, S.; Kato, Y.; Asano, Y.; Aono, S. J. Biol. Chem. 2005, 280, 5486– 5490. (b) Kobayashi, K.; Pal, B.; Yoshioka, S.; Kato, Y.; Asano, Y.; Kitagawa, T.; Aono, S. J. Inorganic Biochem. 2006, 100, 1069–1074. 15. (a) Kato, Y.; Ooi, R.; Asano, Y. Appl. Environ. Microbiol. 2000, 66, 2290–2296. (b) Kato, Y.; Tsuda, T.; Asano, Y. Eur. J. Biochem. 1999, 263, 662–670. 16. Kato, Y.; Yoshida, S.; Asano, Y. FEMS Microbiol. Lett. 2005, 246, 243–249. 17. (a) Kato, Y.; Asano, Y. J. Mol. Catal. B Enzym. 1999, 6, 249–256. (b) Xie, S.-X.; Kato, Y.; Asano, Y. Biosci. Biotechnol. Biochem. 2001, 65, 2666–2672. 18. (a) Asano, Y.; Mihara, Y.; Yamada, H. J. Mol. Catal. B Enzym. 1999, 6, 271–277. (b) Asano, Y.; Mihara, Y.; Yamada, H. J. Biosci. Bioeng. 1999, 87, 732–738. (c) Mihara, Y.; Utagawa, T.; Yamada, H.; Asano, Y. Appl. Environ. Microbiol. 2000, 66, 2811–2816. (d) Ishikawa, K.;
Enzymes in aldoxime–nitrile pathway
139
Mihara, Y.; Gondoh, K.; Suzuki, E.; Asano, Y. EMBO J. 2000, 19, 2412–2423. (e) Mihara, Y.; Utagawa, T.; Yamada, H.; Asano, Y. J. Biosci. Bioeng. 2001, 92, 50–54. (f) Ishikawa, K.; Suzuki, E.; Mihara, Y.; Utagawa, T.; Asano, Y. Protein Eng. 2002, 15, 539–543. (g) Mihara, Y.; Ishikawa, K.; Suzuki, E.; Asano, Y. Biosci. Biotechnol. Biochem. 2004, 68, 1046–1050. (h) Komeda, H.; Ishikawa, N.; Asano, Y. J. Mol. Catal. B Enzym. 2003, 21, 283–290.
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
141
Chapter 8
Addition of hydrocyanic acid to carbonyl compounds Franz Effenberger, Anja Bohrer, Siegfried Förster Institut für Organische Chemie der Universität Stuttgart, Stuttgart, Germany
Abstract The hydroxynitrile lyase (HNL)-catalyzed addition of HCN to aldehydes is the most important synthesis of non-racemic cyanohydrins. Since now not only (R)-PaHNL from almonds is available in unlimited amounts, but the recombinant (S)-HNLs from cassava (MeHNL) and rubber tree (HbHNL) are also available in giga units, the large-scale productions of non-racemic cyanohydrins have become possible. The synthetic potential of chiral cyanohydrins for the stereoselective preparation of biologically active compounds has been developed during the last 15 years.
1. INTRODUCTION The release of hydrocyanic acid (HCN) as a defense against herbivores is widely distributed in higher plants (cyanogenesis). The highly toxic HCN is chemically masked in the plants in the form of cyanohydrins, which are stabilized by an O--glycosidic linkage to saccharides (mainly d-glucose).1 During cyanogenesis, a specific -glycosidase cleaves the cyanoglycoside first into the corresponding cyanohydrin and a carbohydrate. In a second step, the cyanohydrin is decomposed enzymatically into HCN and the corresponding carbonyl compound by hydroxynitrile lyases (HNLs) (Scheme 1). It is interesting to note that in the case of R1 = R2 , an optically active cyanohydrin results in all known cases of cyanogenesis. For the enzymatic
Scheme 1: Cyanogenesis in higher plants.
142
Franz Effenberger et al.
Scheme 2: HNL catalyzed enantioselective addition of HCN to aldehydes. cleavage of the resulting cyanohydrins, enantiospecific HNLs (R- or S-HNLs) are required. As any other catalyst, HNLs catalyze the reaction in both directions. An enantioselective addition of HCN to carbonyl compounds (R1 = R2 ) form ing optically active cyanohydrins (Scheme 2) could therefore be deduced from cyanogenesis. The first attempts to prepare optically active cyanohydrins using the HNL from bitter almonds (Prunus amygdalus), (R)-PaHNL [EC.4.1.2.10], as catalyst gave only moderate optical yields.2 Especially with slower-reacting aldehydes, enantioselectivity was very poor because the non-enzymatic catalyzed chemical addition yielding racemates becomes dominant. The decisive breakthrough for synthetic applications of HNLs in the enantioselective synthesis of cyanohydrins was the discovery that the chemical addition can be suppressed considerably by working in organic solvents immiscible with water.3 Table 1 shows a compari son of the optical yields of the (R)-PaHNL-catalyzed cyanohydrin formation in water/ethanol and in an organic solvent, respectively.3 Until 1987, the (R)-PaHNL from almonds was the only HNL used as catalyst in the enantioselective preparation of cyanohydrins. Therefore, it was of great interest to get access to HNLs which catalyze the formation of (S)-cyanohydrins. (S)-SbHNL [EC 4.1.2.11], isolated from Sorghum bicolor, was the first HNL used for the preparation of (S)-cyanohydrins.4 Since the substrate range of SbHNL is limited to aromatic and heteroaromatic aldehydes as substrates, other enzymes with (S)-cyanoglycosides have been investigated as catalysts for the synthesis of (S)-cyanohydrins. The (S)-HNLs from cassava (Manihot esculenta, MeHNL)5 and from Hevea brasiliensis (HbHNL)6 proved to be highly promising candidates for the preparation of (S)-cyanohydrins. Both MeHNL5 and HbHNL7 have been overexpressed successfully in Escherichia coli, Saccharomyces cerevisiae and Pichia pastoris. In Table 2, the properties and characteristics of the five HNLs presently applied in the enzyme-catalyzed large-scale preparation of chiral cyanohydrins are summarized.78 Although considerable progress has been made in metal-catalyzed syntheses of non-racemic cyanohydrins in recent years, the HNL-catalyzed preparation of chiral cyanohydrins is by far the most important method especially for large-scale processes, due to the easy access of (R)- and (S)-HNLs and the high optical and chemical yields of the cyanohydrins obtained.9−12
Addition of hydrocyanic acid to carbonyl compounds
143
Table 1
(R)-PaHNL catalyzed cyanohydrin formation in H2 O/EtOH and in organic solvents,
respectively OH
O C R
Aldehyde R =
+
HCN
(R)-PaHNL
C R
H
H2 O/EtOH
H CN
Organic solvent
Time (h)
Yield (%)
ee (%)
Solv.
Ph
1
99
86
EtOAc i-Pr2 O
3-PhO-C6 H4 2-furyl =CH MeCH=
5 2 15
99 86 68
105 69 76
EtOAc EtOAc EtOAc
Me3 C
25
56
45
i-Pr2 O
MeS(CH2 2
3
87
60
i-Pr2 O
Time (h)
Yield (%)
25 3 192 4 3 45 16
ee (%)
95 96
99 >99
99 88 68
98 99 97
84
83
98
96
Table 2 HNLs currently applied as catalysts in the enantioselective preparation of chiral cyanohydrins78 MeHNL/HbHNL SbHNL
PaHNL
LuHNL
Family
Euphorbiaceae
Gramineae
Rosaceae
Linaceae
FAD Carbohydrate Structure MW of subunits R/S specificity Homologies
No No Homotetramer 30 000 S pir7 protein
No Yes Heterotetramer 33 000, 18 000 S Serine carboxypeptidases
Yes Yes Monomer 60 000 R Flavoproteins
No No Homodimer 42 000 R Alcohol dehydrogenases
2. OPTIMIZED REACTION CONDITIONS FOR THE HNL-CATALYZED FORMATION OF CHIRAL CYANOHYDRINS For practical applications of HNLs as catalysts for the preparation of chiral cyanohydrins, three objectives have been achieved: first, to get high enantioselec tivity it is decisive to suppress the non-enzymatic addition of HCN to the substrate;
144
Franz Effenberger et al.
secondly, especially for industrial applications high reaction rates are desirable for efficient large-scale productions; thirdly, in situ formation of HCN is favorable to avoid free handling of toxic HCN. The highest enantioselectivities can be achieved by using the enzymes on a solid support (mainly cellulose) in an organic sol vent (diisopropylether, methyl-tert-butylether). For large-scale productions using reasonable priced enzymes, a two-phase system (methyl-tert-butylether/water) is the method of choice because of higher reaction rates. Reactions in organic sys tems using HNLs adsorbed on a support, however, require considerably smaller amounts of enzymes, and product recovery is much easier compared to the bipha sic system. The “support-bound” enzyme can be filtered off after completion of the reaction and reused as catalyst several times. A comparison of the advantages and disadvantages of the various reaction conditions is summarized in recent com prehensive reviews.3b13 The avoidance of toxic “free” HCN is achieved by in situ preparation from sodium cyanide and acetic acid14 or by transcyanation with acetone cyanohydrin.15 It is also possible to replace isolated enzymes by whole cells, for example, by almond or apple meal instead of isolated PaHNL or by sorghum shoots instead of isolated SbHNL.16 In Table 3, results of the preparation of chiral cyanohydrins using almond meal as catalyst and the in situ formation of HCN are summarized.14 (R)-HNLs isolated from other plants do not show any improvements as catalysts compared to the enzyme from almond and therefore cannot compete with (R)-PaHNL for practical applications.16 Recently, the use of ionic liquids instead of organic solvents has been published for the biphasic system.17 For PaHNL and SbHNL, the reaction rates are increased in comparison to organic solvents without a change of enantioselectivity.17 Table 3 Synthesis of optically active cyanohydrins using almond meal OH
O
(R)-PaHNL, ethyl acetate
C H
R
Aldehyde R =
Ph 3-PhO-C6 H4 5-Me-fur-2-yl nPr =CH MeCH=
C
NaCN/OAc (pH 5.5)
Reaction
R
H CN
Cyanohydrins
Temp. ( C)
Time (h)
Conversion (%)
ee (%)
4
16
100
99
20 4 4 4
89 17 41 41
47 70 100 100
99 99 89 99
Addition of hydrocyanic acid to carbonyl compounds
145
During the last years, the overexpression of MeHNL and HbHNL has been improved considerably18 so that large quantities of these enzymes are now cheaply available.
3. SYNTHETIC POTENTIAL OF CHIRAL CYANOHYDRINS IN STEREOSELECTIVE SYNTHESIS Stereoselective follow-up reactions of non-racemic cyanohydrins enable the syn thesis of many other classes of important compounds with one or more stereogenic centers, such as 2-hydroxycarboxylic acids, 2-amino acids, etc. (Scheme 3).9−12 3.1. Chiral 2-hydroxy carboxylic acids In contrast to 2-amino acids, only a few non-racemic 2-hydroxy acids are found in nature. Hydrolysis of non-racemic cyanohydrins offers an interesting general route to (R)- as well as (S)-2-hydroxy carboxylic acids. (R)- and (S)-cyanohydrins are, respectively, hydrolyzed in hydrochloric acid to give the corresponding (R)- and (S)-2-hydroxy acids in excellent yields and with complete retention of configuration.3b Under milder reaction conditions (lower temperature, shorter reaction times), the corresponding 2-hydroxy acid amides can be obtained selectively (Scheme 4).19 Aromatic 2-hydroxy carboxylic acids are of special interest for applica 13 tions. Among them, optically active mandelic acids are regarded as most impor tant commercially.13 The synthetic potential of non-racemic 2-hydroxy acids lies in
Scheme 3: Stereoselective follow-up reactions of non-racemic cyanohydrins.9−12
146
Franz Effenberger et al.
Scheme 4: Stereoselective synthesis of (R)-2-hydroxy acid amides and (R)-2-hydroxy acids by hydrolysis of (R)-cyanohydrins. the possibility to activate the hydroxy group, for example, by sulfonylation.20 The O-activated 2-hydroxy acids can be reacted with any kind of nucleophile to give the corresponding substitution product under complete inversion of configuration without any racemization (Scheme 3).20 The industrial production of CLOPIDOGREL (PLAVIX® ) is an interesting example which demonstrates the synthetic potential of non-racemic cyanohydrins (Scheme 5).21 In the first step, (R)-2-chlorobenzaldehyde cyanohydrin is prepared by the almond meal-catalyzed addition of HCN to 2-chlorobenzaldehyde. The cyanohy drin is then transformed into the corresponding 2-hydroxy ester by Pinner reaction. Subsequently, the hydroxy group is activated by sulfonylation and reacted with tetrahydrothieno[3.2-c]pyridine to give, under complete inversion of configuration, the (S − -amino acid derivative CLOPIDOGREL.21 Another important class of pharmaceuticals which is prepared from chi ral 2-hydroxy acids is the angiotensin-converting enzyme (ACE) inhibitors.22 (R)-3-phenylpropionaldehyde cyanohydrin is transformed into the corresponding 2-hydroxy ester which after activation by sulfonylation reacts with dipetides to give, under inversion of configuration, ACE inhibitors known as “prils” (Scheme 6).22
Scheme 5: Stereoselective synthesis of Clopidogrel (Plavix® ) starting from 2-chlorobenzaldehyde.21
Addition of hydrocyanic acid to carbonyl compounds
147
Scheme 6: Stereoselective synthesis of ACE inhibitors from 3-phenylpropionaldehyde cyanohydrin. 3.2. Optically active 1,2-amino alcohols 1,2-Amino alcohols, which have a broad spectrum of biological activities,23 can be categorized as adrenaline-like with one chiral center at C-1 or as ephedrine-like with two chiral centers at C-1 and C-2 (Scheme 7). Although a variety of methods have been developed for the stereoselective preparation of 1,2-amino alcohols,24 in most cases it is easier and more efficient to prepare these important compounds stereoselectively starting from chiral cyanohydrins (Scheme 7).2526
Scheme 7: Synthesis of (1R)- and (1R 2S)-1,2-amino alcohols starting from
(R)-cyanohydrins.
148
Franz Effenberger et al.
Although it is known that only adrenaline-like compounds with (1R) configuration and ephedrine-like compounds with (1R 2S)-configuration are responsible for the desired biological activity, in most cases racemates are still applied as pharmaceuticals. (1R 2S)-ephedrine is produced by a fermentation pro cess, which is limited to very specific substrates. When starting from non-racemic cyanohydrins, however, almost any structural variation for the stereoselective synthesis of 1,2-amino alcohols is possible. One interesting example for this pos sibility of structural variation is the preparation of heteroaromatic analogues of l-ephedrine.27 Follow-up reactions of optically active 1,2-amino alcohols enable the stereoselective preparation of other interesting compounds. 2-Hydroxy-3-amino carboxylic acids, for example, can be prepared stereoselectively from furyl 1,2-amino alcohols by ozonization of the furyl group to give the carboxyl group.28 3.3. Stereoselective substitution of the hydroxyl group in chiral cyanohydrins The synthetic potential of non-racemic cyanohydrins can be extended considerably by converting the hydroxyl group into a good leaving group, which could be exchanged stereoselectively with any kind of nucleophile. Nucleophilic substitutions of O-activated 2-hydroxy carboxylic acids and esters, respectively, are well established,2021 but little is known about the analogous reactions of activated cyanohydrins. Chiral 2-sulfonyloxynitriles, accessible from non-racemic cyanohydrins, have a relatively high configurational stability.29 They react with nucleophiles under very mild conditions under inversion of configuration (Scheme 8).3031
Scheme 8: Stereoselective reactions of non-racemic 2-sulfonyloxy nitriles with
nucleophiles.
Addition of hydrocyanic acid to carbonyl compounds
149
The chemical yields in all substitution reactions are normally higher than 80%. The optical yields in case of 2-sulfonyloxy nitriles derived from aliphatic aldehydes (R = aliphatic) are very high (ee 94–99%), whereas in case of aromatic derivatives (R = Ar) partly racemization occurs (ee ∼ 80%). For these substrates, the Mitsunobu reaction represents an alternative to the O-activation combined with nucleophilic substitution.32 The (S)-configurated products obtained by nucleophilic substitution of (R)-2-sulfonyloxynitriles (Scheme 8) can be transformed stereoselectively into other interesting compounds. (S)-2-azidonitriles can be hydrogenated to give (S)-2 aminonitriles or chiral 1,2-diamines.30b Optically active 2-aminothiols are accessible from the substitution product with thioacetate by hydrogenation of the nitrile func tion.31 Chiral 2-aminothiols could be of interest as complexing agents for metal ions in chiral catalysts or as educts for the synthesis of optically active heterocycles. 3.4. Stereoselective synthesis of substituted cyclohexanone cyanohydrins Substituted 1-hydroxy cyclohexane-1-carboxylic acids, which could be prepared from the corresponding cyanohydrins by acid hydrolysis as described above, are important as pharmaceuticals and plant-protective agents.33 Although the compounds derived from 2- and 3-cyclohexanones have two stereogenic centers, stereoselective syntheses of these interesting products have been published only very recently.34 Completely unexpected are the results of HNL-catalyzed additions to 4-substituted cyclohexanones, which do not possess a prochiral center. The (R)-PaHNL-catalyzed addition affords almost exclusively trans-isomers, whereas with (S)-MeHNL cis-addition is favored (Table 4).35 Since the chemical addition of HCN always results in mixtures of cis/trans isomers, the stereoselective HNL-catalyzed addition is of great advantage in the synthesis of natural products. The syntheses of the natural monoterpenes cis-p menth-8-ene-1,7-diol and cis-p-menthane-1,7,8-triol are interesting examples for the application of this methodology (Scheme 9).36 While the chemical addition of HCN to 4-(2-propenyl)cyclohexanone affords a cis/trans ratio of 13:82, the MeHNL-catalyzed addition gives almost exclusively the cis-isomer (≥ 96%).35 The chemoenzymatic syntheses of rengyol and isorengyol are other examples for applications of HNL-catalyzed additions of HCN to cyclohexanones.37
4. CRYSTAL STRUCTURES OF HYDROXYNITRILE LYASES AND MECHANISM OF CYANOGENESIS The properties and characteristics of the five HNLs used as catalysts in stereose lective syntheses are listed in Table 2. The crystal structures of these HNLs have been determined during the last decade. From the crystal structures and kinetic measurements, the mechanistic pathways of cyanogenesis could be established.
150
Franz Effenberger et al.
Table 4 HNL-catalyzed addition of HCN to 4-substituted cyclohexanones OH
CN O
R
CN H
(S)-MeHNL R iPr2O
+ HCN
(R)-PaHNL
R H
H cis
trans
Aldehyde R =
Me Et nPr iPr tBu =CHCH3 CH2 = Ph
OH
iPr2O
(S)-MeHNL-catalyzed
(R)-PaHNL-catalyzed
Time (h)
Conv. (%)
cis:trans
Time (h)
Conv. (%)
cis:trans
2 5 5 5 3 9 3
80 Quant. 93 Quant. 82 99 95
35:65 74:36 96:4 97:3 99:1 98:2 99:1
15 22 31 22 216 24 264
99 99 83 94 50 99 71
3:97 2:98 2:98 1:99 10:90 1:99 4:96
Scheme 9: Stereoselective synthesis of cis-p-menthane-1,7,8-triol (A) and
cis-p-menth-8-ene-1,7-diol (B).
Addition of hydrocyanic acid to carbonyl compounds
151
4.1. Crystal structures of HNLs Several structurally different types of HNLs occur in nature, which likely origi nated by convergent evolution from different ancestral proteins.38 The enzyme from almond (PaHNL) was first crystallized in 1994 and the structure was solved by multiple wavelength anomalous dispersion of a mercury derivative.39 The first 3D structure analysis of PaHNL was performed in 2001.40 (R)-PaHNL from almond uses FAD as cofactor and is related to oxidoreductases; it exhibits HNL activity only in the oxidized form of FAD.41 The crystal structure of the HNL isolated from S. bicolor (SbHNL) was determined in a complex with the inhibitor benzoic acid.42 The folding pattern of SbHNL is similar to that of wheat serine carboxypeptidase (CP-WII)43a and alcohol dehydrogenase.43b A unique two-amino acid deletion in SbHNL, however, is forcing the putative active site residues away from the hydrolase binding site toward a small hydrophobic cleft, thereby defining a completely different active site architecture where the triad of a carboxypeptidase is missing. The most extensively investigated HNL structures are those from H. brasiliensis (HbHNL)44 and M. esculenta (MeHNL),45 which are highly homolo gous (76% identity). For MeHNL, the crystal structure of the wild-type enzyme complexed with acetone has been reported in 2001 (Fig. 1).45 Both enzymes belong to the family of ,-hydrolases.43 The active site of MeHNL is located inside the protein and connected to the outside through a small channel, which is covered by the bulky amino acid tryptophane 128.45 It was possible to obtain the crystal structure of the complex with the natural substrate acetone cyanohydrin with the mutant Ser80A1a of MeHNL. This complex allowed the determination of the mode of substrate binding in the active site.45 A summary of 3D structures of known HNLs was published recently.4647 4.2. Reaction mechanism of cyanogenesis Mechanisms of cyanogenesis have been published first based on the crystal struc tures of SbHNL from S. bicolor42 and of PaHNL from almonds.47 In this article, the mechanism of cyanogenesis will be discussed on the most proven mechanism of MeHNL from M. esculenta.45 Combining X-ray structure and site-directed mutagenesis results in a mech anism as shown in Scheme 10 in which His236 acts as a general base abstracting a proton from Ser80, thereby allowing proton transfer of the cyanohydrin to Ser80. The His236 imidazolium cation then facilitates leaving of the cyano group by pro ton transfer (Scheme 10).45 It was demonstrated that the active site lysine residue Lys237 in MeHNL is necessary for the catalytic activity.48 The (S)-enantioselectivity of MeHNL can be explained considering Fig. 1. As shown in the figure, the carbonyl oxygen is hydrogen bonded to Ser80 and Thr11. One methyl group of acetone (C1) is held in position by van der Waals contacts to Leu149, Thr11 and Ile12. The side chains of these residues define a
152
Franz Effenberger et al. CYS 81 SER 80 THR 11
ILE 12 3.3
2.9
3.0 PHE 211
4.1
HIS 14
4.0
HIS 236 C3
ASP 208 3.8
4.0
C1 4.3
3.7
ILE 210 LEU 158
LEU 149
TRP 128
Figure 1: Structure of selected active site residues of MeHNL complexed with acetone. small hydrophobic site S1 . The second methyl group (C3) points in the opposite direction toward the active site channel, defining the putative second subsite S2 in the binding cavity. Aldehydes are now fixed in a way that the small substituent hydrogen is situated in S1 and the larger substituent R in S2 . This mode of binding suggests that the incoming cyanide exclusively attacks the Si-face of the carbonyl compound. This implies that HCN is deprotonated by His236, which is located on the Si-face of the substrate in order to be consistent with the (S)-enantioselectivity of the addition.45 4.3. Changing substrate specificity and stereoselectivity applying Trp128 mutants of wt-MeHNL As mentioned earlier, it is known from the crystal structure of MeHNL that the active site of the enzyme is accessible by a narrow channel. The channel entrance is capped by the large amino acid Trp128. Substitution of Trp128 by amino acids with decreasing size gives the corresponding MeHNL mutants. The MeHNL W128A mutant, for example, could be prepared and overexpressed in E. coli.49 Since the entrance to the active site of this mutant is significantly less hindered, the
Addition of hydrocyanic acid to carbonyl compounds
153
Scheme 10: Mechanism of cyanogenesis catalyzed by MeHNL.45 access of sterically demanding substrates is facilitated considerably. A comparison of the reactions of 3-phenoxy benzaldehyde, which is a very bulky substrate, with the wt-enzyme and the W128A mutant, respectively, reveals the superiority of the mutant. The aldehyde was converted quantitatively in a much shorter time and with high enantiomeric excess using the mutant W128A, nearly independent of the amount of enzyme present.49
5. CONCLUSIONS Although considerable progress has been made in metal-catalyzed preparations of non-racemic cyanohydrins, the HNL-catalyzed reaction is still the most impor tant method for the synthesis of chiral cyanohydrins, especially for large-scale reactions. The usefulness of HNLs as catalysts for the stereoselective addition of HCN to carbonyl compounds has increased substantially because (R)-PaHNL
154
Franz Effenberger et al.
from almonds and two recombinant (S)-HNLs, MeHNL from M. esculenta and HbHNL from H. brasiliensis, are fairly stable and nowadays easily available in large amounts. During the last 15 years the 3D structures of the most important HNLs have been determined. Combining the X-ray structure information with the kinetics of certain mutants of the active site made it possible to elucidate unambiguously the mechanism of cyanogenesis for both MeHNL and HbHNL, respectively. It is possible to improve yields and stereoselectivity of HNL-catalyzed formations of chiral cyanohydrins in a straightforward manner through the knowledge of the active site and the possibility of applying more reactive mutants. Most of the stereoselective follow-up reactions of non-racemic cyanohydrins have been investigated only conceptually so far. Starting from (R)-cyanohydrins, (R)-2-hydroxy carboxylic acids, (R)-2-hydroxy aldehydes, (1R)-2-amino alcohols and (1R 2S)-2-amino alcohols are easily available in high optical yields and chemical yields by chemical transformations of the cyano group. Analogous reac tions are possible starting from (S)-cyanohydrins. Sulfonylation of the hydroxy group in non-racemic cyanohydrins subsequently allows nucleophilic substitu tion of the sulfonate group with complete inversion of configuration. By this means, non-racemic 2-azido nitriles, 2-amino nitriles, 1,2-diamines and 2-sulfanyl nitriles can be prepared in high optical yields. The HNL-catalyzed addition of HCN to 4-substituted cyclohexanones unexpectedly exhibits high cis/trans selectivity, which is very useful for the stereoselective synthesis of natural products. Therefore, applications of chiral cyanohydrins for the preparation of biologically active compounds with stereogenic centers, which are applied as pharmaceuticals or plant-protecting agents, will play a major role in future developments.
REFERENCES 1. Conn, E. E. In The Biochemistry of Plants, Secondary Plant Products; Stumpf, P. K.; Conn, E. E. (Eds); Academic Press: New York; Vol. 7, 1981, pp. 479–500. 2. Becker, W.; Freund, H.; Pfeil, E. Angew. Chem. Int. Ed. Engl. 1965, 4, 1079. 3. (a) Effenberger, F.; Ziegler, T.; Förster, S. Angew. Chem. Int. Ed. Engl. 1987, 26, 458–460. (b) Effenberger, F. Angew. Chem. Int. Ed. Engl. 1994, 33, 1555–1564. 4. (a) Effenberger, F.; Hörsch, B.; Förster, S.; Ziegler,T. Tetrahedron Lett. 1990, 31, 1249–1252. (b) Niedermeyer, U.; Kula M. R. Angew. Chem. Int. Ed. Engl. 1990, 29, 386–387. 5. Förster, S.; Roos, J.; Effenberger, F.; Wajant, H.; Sprauer, A. Angew. Chem. Int. Ed. Engl. 1996, 35, 437–439. 6. Griengl, H.; Hickel, A.; Johnson, D. V.; Kratky, C.; Schmidt, M.; Schwab, H. Chem. Commun. 1997, 1, 1933–1940. 7. Hasslacher, M.; Schall, M.; Hayn, M.; Griengl, H.; Kohlwein, S. D.; Schwab, H. J. Biol. Chem. 1996, 271, 5884–5891. 8. (a) Jansen, I.; Woker, R.; Kula, M. R. Biotechnol. Appl. Biochem. 1992, 15, 90–99. (b) Wajant, H.; Mundry, K. W.; Pfizenmaier, K. Plant Mol. Biol. 1994, 26, 735–746.
Addition of hydrocyanic acid to carbonyl compounds
155
9. (a) Effenberger, F. Hydroxynitrile lyases in stereoselective synthesis. In Stereoseletive Biocatal ysis; Patel, R. N. (Ed.); Marcel Dekker: New York; 2000, pp. 321–342. (b) Effenberger, F.; Förster, S.; Wajant, H. Curr. Opin. Biotechnol. 2000, 11, 532–539. 10. North, M. Tetrahedron 2004, 60, 10371–10568. 11. Schmidt, M.; Griengl, H. Top. Curr. Chem. 1999, 200, 193–226. 12. Gregory, R. J. H. Chem. Rev. 1999, 99, 3649–3682. 13. Gröger, H. Adv. Synth. Catal. 2001, 343, 547–558. 14. Zandbergen, P.; van der Linden, J.; Brussee, J.; van der Gen, A. Synth. Commun. 1991, 21, 1386–1391. 15. Ognyanov, V. I.; Datcheva, V. K.; Kyler, K. S. J. Am. Chem. Soc. 1991, 113, 6992–6996. 16. (a) Kiljunen, E.; Kanerva, L. T. Tetrahedron: Asymmetry 1997, 8, 1225–1234. (b) Solis, A.; Luna, H.; Perez, H. I.; Manjarrez, N.; Sanchez, R.; Albores-Velasco, M.; Castillo, R. Biotechnol. Lett. 1998, 20, 1183–1185. (c) Effenberger, F.; Gaupp, S. Tetrahedron: Asymmetry 1999, 10, 1765–1775. 17. Gaisberger, R. P.; Fechter, M. H.; Griengl, H. Tetrahedron: Asymmetry 2004, 15, 2959–2963. 18. Effenberger, F.; Wajant, H.; Förster, S. PCT Int. Appl., WO 200148178 A1, CAN 135: 45300, 2001. 19. Rochowski, A. Ph.D. Thesis, Universität Stuttgart, 1997. 20. (a) Burkard, U.; Effenberger, F. Chem. Ber. 1986, 119, 1594–1612. (b) Effenberger, F.; Burkard, U.; Willfahrt, J. Liebigs Ann. Chem. 1986, 314–333. 21. Bousquet, A.; Musolino, A. PCT Int. Appl. WO 9918110 A1, CAN 130: 296510, 1999. 22. Sheldon, R. A.; Zeegers, H. J. M.; Houbiers, J. P. M.; Hulshof, L. A.; Andeno, B. V. Chim. Oggi 1991, 9, 35–47. 23. Kleemann, A.; Engel, J. Pharmaceutical Substances, 3rd edn, Thieme: Stuttgart, 1999. 24. (a) Corey, E. J.; Link, J. O. Tetrahedron Lett. 1990, 31, 601–604. (b) Enders, D.; Reinhold, U. Liebigs Ann. Chem. 1996, 11–26. (c) Li, G.; Angert, H. H.; Sharpless, K. B. Angew. Chem. Int. Ed. Engl. 1996, 35, 2813–2817. (d) Bakale, R. P. Spec. Chem. 1995, 15, 249–250, 253. 25. (a) Ziegler, T.; Hörsch, B.; Effenberger, F. Synthesis 1990, 575–578. (b) Effenberger, F.; Jäger, J. J. Org. Chem. 1992, 62, 3867–3873. 26. (a) Brussee, J.; Dofferhoff, F.; Kruse, C. G.; van der Gen, A. Tetrahedron 1990, 46, 1653–1658. (b) Effenberger, F.; Gutterer, B.; Jäger, J. Tetrahedron: Asymmetry 1997, 8, 459–467. 27. Effenberger, F.; Eichhorn, J. Tetrahedron: Asymmetry 1997, 8, 469–476. 28. Tromp, R. A.; van der Hoeven, M.; Amore, A.; Brussee, J.; Overhand, M.; van der Marel, G. A.; van der Gen, A. Tetrahedron: Asymmetry 2003, 14, 1645–1652. 29. Effenberger, F.; Stelzer, U. Angew. Chem. Int. Ed. Engl. 1991, 30, 873–874. 30. (a) Effenberger, F.; Stelzer, U. Chem. Ber. 1993, 126, 779–786. (b) Effenberger, F.; Kremser, A.; Stelzer, U. Tetrahedron: Asymmetry 1996, 7, 607–618. 31. (a) Effenberger, F.; Gaupp, S. Tetrahedron: Asymmetry 1999, 10, 1765–1775. (b) Effenberger, F.; Gaupp, S. Tetrahedron: Asymmetry 1999, 10, 1777–1786. 32. (a)Warmerdam, E. G. J. C.; Brussee, J.; Kruse, C. G.; van der Gen, A. Tetrahedron 1993, 49, 1063–1070. (b) Decicco, C. P.; Grover, P. Synlett 1997, 529–530. (c) Hughes, D. L. Org. Prep. Proced. Int. 1996, 28, 127–164. 33. (a) Skinner, W. A.; Fuhrmann, F.; Rutledge, L. C.; Moussa, M. A.; Schreck, C. E. J. Pharm. Sci. 1980, 69, 196–198. (b) Shapiro, S. L.; Rose, I. M.; Roskin, E.; Freedman, L. J. Am. Chem. Soc. 1959, 81, 386–390. (c) Fischer, R.; Bretschneider, T.; Krüger, B. W.; Santel, H.-J.; Dollinger, M.; Wachendorff-Neumann, U.; Erdelen, C. Ger. Offen. DE 4337853 A1, CAN 123:32947, 1995. 34. Kobler, C.; Bohrer, A.; Effenberger, F. Tetrahedron 2004, 60, 10397–10410. 35. Effenberger, F.; Kobler, C.; Roos, J. Angew. Chem. Int. Ed. 2002, 41, 1876–1879. 36. Kobler, C.; Effenberger, F. Chem. Eur. J. 2005, 11, 2783–2787. 37. Kobler, C.; Effenberger, F. Tetrahedron 2006, 62, 4823–4828.
156
Franz Effenberger et al.
38. Wajant, H.; Effenberger, F. Biol. Chem. Hoppe-Seyler 1996, 377, 611–617. 39. Lauble, H.; Müller, K.; Schindelin, H.; Förster, S.; Effenberger, F. Proteins Struct. Funct. Genet. 1994, 19, 343–347. 40. Dreveny, I.; Gruber, K.; Glieder, A.; Thompson, A.; Kratky, C. Structure 2001, 9, 803–815. 41. Baerwald, K. R.; Jaenicke, L. FEBS Lett. 1978, 90, 255–260. 42. Lauble, H.; Miehlich, B.; Förster, S.; Wajant, H.; Effenberger, F. Biochemistry 2002, 41, 12043–12050. 43. (a) Ollis, D. L.; Cheah, E.; Cygler, M.; Dijkstra, B.; Frolow, F.; Franken, S. M.; Harel, M.; Remington, J.; Silman, I.; Schrag, J.; Sussman, J. L.; Verschueren, K. H. G.; Goldman, A. Protein Eng. 1992, 5, 197–211. (b) Cygler, M.; Schrag, J. D.; Sussman, J. L.; Harel, M.; Silman, I.; Gentry, M. K.; Doctor, B. P. Protein Sci. 1993, 2, 366–382. 44. (a) Wagner, U. G.; Hasslacher, M.; Griengl, H.; Schwab, H.; Kratky, C. Structure 1996, 4, 811–822. (b) Gruber, K.; Gugganig, M.; Wagner, U. G.; Kratky, C. Biol. Chem. 1999, 380, 993–1000. 45. (a) Lauble, H.; Förster, S.; Miehlich, B.; Wajant, H.; Effenberger, F. Acta Crystallogr. Sect. D Biol. Crystallogr. 2001, D57, 194–200. (b) Lauble, H.; Miehlich, B.; Förster, S.; Wajant, H.; Effenberger, F. Protein Sci. 2001, 10, 1015–1022. 46. Gruber, K.; Kratky, C. J. Polym. Sci. Part A Polym. Chem. 2004, 42, 479–486. 47. Dreveny, I.; Kratky, C.; Gruber, K. Protein Sci. 2002, 11, 292–300. 48. Gruber, K.; Gartler, G.; Krammer, B; Schwab, H.; Kratky, C. J. Biol. Chem. 2004, 279, 20501–20510. 49. Bühler, H.; Effenberger, F.; Förster, S.; Roos, J.; Wajant, H. ChemBioChem 2003, 4, 211–216.
Part Three Novel compounds synthesized by biotransformations
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
159
Chapter 9
Chiral heteroatom-containing compounds Piotr Kiełbasinski, Marian Mikołajczyk ´ Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza, Łód´z, Poland
Abstract Enzyme-mediated syntheses of chiral non-racemic hetero-organic compounds, with a stereogenic centre located either on a heteroatom or on a carbon atom in a side chain, are comprehensively presented. Particular attention is paid to the use of common hydrolytic and reducing enzymes. On the basis of the results presented, some conclu sions are drawn and proposals presented, which concern possible future directions in the applications of enzymes to the synthesis and transformations of chiral hetero-organic derivatives.
1. INTRODUCTION Chiral hetero-organic compounds, particularly those containing sulfur or phospho rus, continue to receive intense interest of researchers due to their applicability in organic synthesis and interesting biological activities.1 This concerns both the derivatives with a stereogenic centre located on a carbon atom, with a hetero organic moiety being only one of the substituents, and the compounds with a stereogenic centre located on a heteroatom. Their synthesis still remains a subject of interest for organic chemists. Among many synthetic methodologies, those employing enzymes have recently become a valuable supplement. The three major types of selectivities that are displayed by enzymes, i.e. chemoselectivity, regioselectivity and diastereo- or enantioselectivity, caused an explosion of their applications in organic synthesis and particularly in the preparation of chiral non-racemic products. The possibil ity of using racemic mixtures, meso- and prochiral compounds as substrates has made the use of enzymes applicable to a broad variety of transformations. More over, enzymes have been shown to accept and stereoselectively recognize different types of chirality and various stereogenic centres, including those located on het eroatoms.2 This has removed all the obstacles concerning the types of substrates and made enzymes highly versatile catalysts.
160
Piotr Kiełbasi´nski, Marian Mikołajczyk
The present chapter will concentrate on the application of enzymes to the synthesis of chiral non-racemic hetero-organic compounds of all kinds. The pre sentation will focus mainly on the use of common hydrolytic enzymes, whose natural function is a transformation of carboxylic acid derivatives, and on the enzymes (often in the form of whole cells or living organisms) capable of reduc ing a carbonyl moiety. Therefore, in most cases the heteroatom present in the substrate molecule will not be involved in the chemical reaction catalysed by the enzyme, being, however, crucial for its stereochemical outcome. By taking such an approach to the problem, certain particular fields of enzyme-promoted trans formations will be omitted. For example, the use of nucleases – phosphomono – and diesterases will not be discussed, as it does not involve any stereochemical change in the reacting molecule. Also, enzyme-catalysed stereoselective oxida tions of sulfides to sulfoxides will not be presented; they have been exhaustively reviewed elsewhere.3 However, certain hydrolytic enzyme-promoted transforma tions of heteroatom substrates, in which the heteroatom will be involved both as a stereogenic and as a reacting centre, will be discussed as the examples of the enzyme catalytic promiscuity.
2. ORGANOSULFUR COMPOUNDS 2.1. C-chiral hydroxy sulfides and derivatives Searching for a method of synthesis of enantiopure lamivudine 1, the compound having a monothioacetal stereogenic centre, Rayner et al. investigated a lipase catalysed hydrolysis of various racemic -acetoxysulfides 2. They found out that the reaction was both chemoselective (only the acetate group was hydrolysed with no detectable hydrolysis of the other ester moieties) and stereoselective. As a result of the kinetic resolution, enantiomerically enriched unreacted starting compounds were obtained. However, the hydrolysis products 3 were lost due to decomposition.4 In this way, the product yields could not exceed 50% (Equation 1). The product 2 (R = CH2 CH(OEt)2 ) was finally transformed into lamivudine 1 and its 4-epimer.5
Chiral heteroatom-containing compounds
161
When a reverse procedure was applied, i.e. enzymatic acetylation of racemic 3, formed in situ from the appropriate aldehydes and thiols, the reaction proceeded under the conditions of dynamic kinetic resolution and gave enantiomerically enriched acetates 2 with 65–90% yields and with ees up to 95% (Equation 2).6 It must be mentioned that the addition of silica proved crucial, as in its absence no racemization of the initially formed substrates 3 occurred and the reaction stopped at the 50% conversion.
When -diacetoxysulfides 2, R1 = AcOCH2 , were subjected to enzymemediated hydrolysis under kinetic resolution conditions, a completely regioselec tive for the primary acetyl group and highly enantioselective reaction occurred (Equation 3).7
In the compounds discussed above, the sulfur atom was attached directly to the stereogenic carbon atom. However, some structures that have remote sulfur
162
Piotr Kiełbasi´nski, Marian Mikołajczyk
moieties are also worth mentioning. The approach to their synthesis was based on the use of fermenting baker’s yeast as the reducing agent of the keto group present in the substrate (for reviews see Refs. 8 and 9). In this way, a variety of -hydroxyalkyl sulfides were obtained in reasonable to good yields and with good to excellent enantiomeric excesses (e.g. Equation 4), although a low concentration of the substrates had to be applied.10−13
-Keto dithioacetals were similarly transformed to non-racemic –hydroxy derivatives (e.g. Equation 5).14−18 In general, the reaction rate and stereochemical outcome depended on the hydrocarbon chain length and the type of the substituents in the substrate.
C-chiral racemic -hydroxy sulfides were also resolved using PFL under kinetic resolution conditions. The products were transformed into optically active 3-(alkanesulfonyloxy)thiolane salts (Scheme 1).19a Similarly, 1,2-cyclic sulfite glycerol derivatives (cis and trans) were resolved into enantiomers via a Pseu domonas cepacia-catalysed acylation with vinyl butyrate. The E values depended on the solvent used and varied from 2 to 26.19b
Scheme 1
Chiral heteroatom-containing compounds
163
2.2. C-chiral hydroxyalkyl sulfones Chiral hydroxyalkyl sulfones are usually secondary alcohols and as such can be obtained in their non-racemic forms using general methods developed for these classes of compounds. Among the biocatalytic methods, two approaches were applied: microbiological reductions of the corresponding keto sulfones 5 and resolution of racemic hydroxyalkyl sulfones 6. In the former case, the earliest procedures utilized fermenting baker’s yeast as the reducing agent,1011 which gave the products in high yields and with high enantiomeric excess. However, this method allowed to obtain only one enantiomer of a hydroxyalkyl sulfone – for - and -hydroxyalkyl sulfones it is the (S)-enantiomer.10112021 The same result was obtained using Corynebacterium equi22 and the fungus Aspergillus niger.2022 However, the (R)-enantiomers were produced during the reduction carried out with other fungi, namely Geotrichum candidum and, with somewhat inferior results, Mortierella isabellina (Equation 6).20 The ees of the products were strongly depen dent on the length of the alkyl group R – the longer the chain the lower the ee value.102123 The same influence was exerted by the phenyl group, although changing Baker’s yeast for Saccharomyces kyokai 7 gave the opposite enantiomer, i.e. (R)-6, R = Ph.24
Another approach to the synthesis of chiral non-racemic hydroxyalkyl sul fones used enzyme-catalysed kinetic resolution of racemic substrates. In the first attempt, Porcine pancreas lipase was applied to acylate racemic , and hydroxyalkyl sulfones using trichloroethyl butyrate. Although both enantiomers of the products could be obtained, their enantiomeric excesses were only low to moderate.25 Recently, we have found that a stereoselective acetylation of racemic -hydroxyalkyl sulfones can be successfully carried out using several lipases, among which CAL-B and lipase PS (AMANO) proved most efficient.26 Moreover, application of a dynamic kinetic resolution procedure, in which lipase-promoted kinetic resolution was combined with a concomitant ruthenium-catalysed27 racem ization of the substrates, gave the corresponding -acetoxyalkyl sulfones 8 in yields
164
Piotr Kiełbasi´nski, Marian Mikołajczyk
higher than 50% and with ees up to 99%. Also, in this case, the length of the alkyl chain proved crucial – only the compounds with the methyl or ethyl group attached to the stereogenic carbon atom were accepted by the enzymes. The acetates 8 were then quantitatively reduced to the corresponding -hydroxyalkyl sul fones 7 by treatment with the borane/dimethyl sulfide complex (Equation 7, Table 1).26
Table 1 Dynamic kinetic resolution of –hydroxyalkyl sulfones Entry
1 2 3 4 5 6
Ar
Ph Ph p-Tol p-Tol Ph Ph
R
Me Me Me Me Et Et
Lipase
CAL-B PS CAL-B PS CAL-B PS
Acetate 8
Alcohol 7
Yield (%)
[]D (CHCl3
[]D (CHCl3
ee (%)
66 64 79 75 42 34
−1 0 −1 1 −1 1 −0 9 −4 4 −4 4
−12 0 −11 8 −11 6 −10 3 −9 2 −9 2
>99 >99 >99 90 >99 >99
-Hydroxyalkyl sulfones 10 resemble -hydroxyalkyl sulfides 3 (Equa tion 1) in the sense that they are unstable in their unprotected form. Therefore, the enzyme-promoted hydrolysis of the corresponding racemic O-acetyl derivatives 9 led to optically active unreacted substrates, while the resulting hydroxyalkyl sulfones 10 underwent decomposition to sulfinic acids and aldehydes. An attempt to achieve a dynamic kinetic resolution based on the assumed reaction of sulfinic acids with aldehydes and subsequent acetylation of the resulting hydroxyalkyl sulfones 10 (similar to the procedure shown in Equation 2) failed due to the lack of formation of the latter under the conditions applied (Equation 8).28 It should be added that such a formation was reported in an early literature as an easy and efficient process.29
Chiral heteroatom-containing compounds
165
2.3. C-chiral alkyl sulfates C-chiral alkyl sulfates 11, which are sulfate esters of secondary alcohols, are hydrolytically cleaved by the enzymes sulfatases.30 Interestingly, there are two types of sulfatases – the one which cleaves the S–O bond31 and the second which cleaves the C–O bond.32 The former does not alter the configuration at the stereogenic carbon atom (Equation 9), acting in the stereochemical sense like lipases, esterases and proteases; the latter causes inversion of configuration at the carbon chirality centre (Equation 10). In the first case, a normal kinetic resolution of the racemic sulfate can be performed to give unreacted sulfate 11 and the alcohol 12, both having opposite absolute configurations at the stere ogenic carbon atom. In the second case, starting from the racemic substrate a mixture of the unreacted substrate 11 and the alcohol 12 is obtained, each of them having the same absolute configuration. Such a result makes it possible to use this type of sulfatase in a deracemization process, in which a racemic sub strate is converted into a single enantiomer of a product. This approach was used by Faber and coworkers32 for deracemization of secondary alcohols. It should be added that the final transformation of 11 into 12 was stereospecifically per formed by hydrolysis using p-toluenesulfonic acid in aqueous t-butyl methyl ether–dioxane.32e
166
Piotr Kiełbasi´nski, Marian Mikołajczyk
2.4. Other C-chiral organosulfur compounds Desymmetrization of meso-3,4-dihydroxythiolane 13 was accomplished either via its lipase-catalysed acetylation or hydrolysis of the corresponding diacetate 15. Each method led to the opposite enantiomer of the monoacetate 14 (Equa tion 11).33 It is noteworthy that, although the resulting monoacetate 14 exhibited only a low to moderate enantiomeric excess (up to 36%), it turned out to be configurationally relatively stable in contrast to its carba-analogue, cis-1-acetoxy 2-hydroxycyclopentane, which was found to undergo fast racemization due to a 1,2-acetyl shift.
2.5. S-chiral sulfinylcarboxylates Enzymatic resolution of racemic sulfinylcarboxylates or desymmetrization of prochiral sulfinyldicarboxylates resembles the procedures applied for the C-chiral or prochiral analogues. Thus, the stereogenic (like in A) or prostereogenic (like in B) centre of the substrate is recognized and stereoselectively bound in the active site of the enzyme, but is not involved in the chemical transformation, which takes place at a more or less remote reacting site. The latter is here an ester group, while the stereogenic centre may be located on carbon or any heteroatom (Scheme 2).2
Chiral heteroatom-containing compounds
167
Scheme 2 The first attempt at the kinetic resolution of racemic S-chiral sulfinylacetates and sulfinylpropionates rac-16 was made in 1986 when the use of bacteria allowed to obtain the recovered esters 16 in good yields and with ees of 90–97%.34 Then, application of isolated lipase from Pseudomonas sp. K-10 made it possible to obtain both the unreacted esters 16 and the corresponding acids 17 in high yields and with ees exceeding 95% (Equation 12).35 However, the use of pig liver esterase (PLE), known as an enzyme of high potential for ester hydrolysis, gave disappointing results (ees from 21 to 80%).36
Later on, a variety of other S-chiral sulfinylcarboxylates were resolved to give both the non-racemic recovered esters and acids with different ees. Among them, there were 3-arylsulfinylpropenoates (Scheme 2, structure A, Z = S(O), L = ethenyl), resolved by CCL or -chymotrypsin with ees of 65–91%,37 2 (alkanesulfinyl)benzoates (Z = S(O), L = o-C6 H4 , resolved by CRL (E > 100)38 and cyclic sulfinylcarboxylates – derivatives of 3,4-didehydrothiopyran 1-oxides 18, resolved by PLE, -chymotrypsin and PPL with moderate to high ees.39
To outweigh disadvantages of the kinetic resolution presented above, an enzymatic desymmetrization of prochiral sulfinyldiacetates 19 was performed. The use of various enzymes, PLE,40a41 -chymotrypsin (-CT)40a and PPL,41 made it
168
Piotr Kiełbasi´nski, Marian Mikołajczyk
possible to obtain each enantiomer of the monoacetate 20 and to determine their absolute configurations40b (Equation 13).
2.6. S-chiral hydroxy sulfoxides The procedures used in the synthesis of non-racemic hydroxy sulfoxides are in some respects similar to those applied to the preparation of hydroxy sulfones, the difference being the location of a stereogenic centre on the sulfinyl sulfur atom. Thus, a cholesterol esterase (CE)-promoted kinetic resolution of alkyl aryl sulfox ides having pendant acetoxy group in the ortho position of the aryl ring 21 gave both the unreacted acetate and the corresponding hydroxyphenyl product 22 with moderate enantioselectivity (Equation 14).42 In turn, S-racemic -hydroxyalkyl sulfoxides 23 and hydroxyalkylaryl sulfoxides 24, having the reacting primary hydroxy group very remote from the stereogenic centre (four bonds in 23 and six bonds in 24), were successfully resolved via enzymatic acetylation.43
An attempt at the enzymatic hydrolysis of racemic -acetoxymethyl sulfox ides rac-25 gave the recovered esters in yields up to 40% and with ees up to 95%. The hydrolysis products, -hydroxymethyl sulfoxides 26, underwent decom position like in the case of analogous hydroxy sulfides (Equation 1) and hydroxy
Chiral heteroatom-containing compounds
169
sulfones (Equation 8). A possible reverse reaction remained out of the question because of the inaccessibility of the corresponding sulfenic acids 27 (Equation 15).44
Some enzymes were found to exhibit high diastereoselectivity towards hydroxy sulfoxides. It was observed in the acetylation of the epimeric mixture of -hydroxyalkyl sulfoxides 28 and 29, leading to the improvement of the diastere omeric ratio,45a and in the acetylation of (R R)-dihydroxythiolane 1-oxide 30, lead ing exclusively to cis-31 (Equation 16).33 Interestingly, single biotransformations of 1-phenylthio-2-propanone 32 by the fungus Helminthosporium sp. 4671 resulted in both sulfur oxidation to the sulfoxide and carbonyl reduction to the alcohol to give the hydroxy sulfoxide 33 with high diastereomeric excess (Equation 17).45b
2.7. S-chiral sulfinamides Subtilisin E was found to efficiently catalyse enantioselective hydrolysis of cer tain N -acyl arenesulfinamides 34 to arenesulfinamides 35. Noteworthy, among
170
Piotr Kiełbasi´nski, Marian Mikołajczyk
the potential substrates only those having the chloroacetyl, bromoacetyl and dihy drocinnamoyl groups were recognized and hydrolysed by the enzyme in a highly stereoselective manner. Some other N -acyl derivatives gave inferior results and the N -acetyl and N -butanoyl analogues proved totally unreactive (Equation 18, Table 2).46 The molecular modelling that was made for the first time for this kind of compound suggested that N -haloacetyl and N -dihydrocinnamoyl groups mimic the phenylalanine moiety and thus bind the sufinamide to the enzyme active site. The enantioselectivity is believed to stem from a favourable hydropho bic interaction between the aryl group of the fast-reacting (R)-enantiomer of the arenesulfinamide and the S1 leaving group pocket in subtilisin E.46
However, when subtilisin E was replaced by subtilisin Carlsberg, the hydrolysis of the S–N bond in some N -acyl arenesulfinamides 34 unexpectedly became the main hydrolytic process giving under the kinetic resolution condi tions, in addition to the unreacted substrates, the corresponding sulfinic acids and Table 2 Enzymatic hydrolysis of N -acyl arenesulfinamides46 Entry
Ar
R
1 2 3 4 5 6 7 8 9 10 11 12 13
Ph Ph p-Tol p-Tol p-Tol p-ClC6 H4 p-ClC6 H4 p-MeOC6 H4 p-MeOC6 H4 p-Tol p-Tol p-Tol p-Tol
CH2 Cl (CH2 2 Ph CH2 Cl CH2 Br (CH2 2 Ph CH2 Cl (CH2 2 Ph CH2 Cl (CH2 2 Ph CH2 OMe (CH2 2 Pri Me (CH2 3
a
Conversion (%)
ee of (S)-34
ee of (R)-35
14 14 36 36 47 35 9 21 40 28 13 0 0
16 16 56 99 89 52 10 25 65 33 14 0 0
97 99 97 89 99 97 99 94 98 84 91 0 0
Ea >150 >150 >150 >150 >150 >150 >150 >150 >150 26 24 0 0
The substrates also underwent 1–3% spontaneous chemical hydrolysis; the E values shown were corrected for this hydrolysis.
Chiral heteroatom-containing compounds
171
carboxamides (Equation 19).47 Several experiments were carried out to prove that in this case also the enzyme active site was involved.48 The crucial proof was obtained on the basis of the electrospray mass spectroscopy, which clearly indi cated the formation of an O-sulfinyl enzyme (sulfinylated most probably on the active site serine).49 Thus, this reaction constitutes an example of the catalytic promiscuity of enzymes, which was demonstrated by the fact that the enzyme favoured the unnatural functional group in the presence of the normal functional group.
2.8. S-chiral sulfoximines Two types of sulfoximinocarboxylates (analogous to sulfinylcarboxylates 16), namely S-aryl-S-methoxycarbonylmethyl-N -methyl sulfoximine 36 and S-methylS-phenyl-N -ethoxycarbonyl sulfoximine 37, were subjected to hydrolysis in the presence of PLE in a phosphate buffer. As a result of a kinetic resolution, both the enantiomerically enriched recovered substrates and the products of hydrolysis and subsequent decarboxylation 38 and 39, respectively, were obtained with moderate to good ees (Equations 20 and 21). Interestingly, in each case the enantiomers of the substrates, having opposite spatial arrangement of the analogous substituents, were preferentially hydrolysed.50 This was explained in terms of the Jones PLE active site model.51
172
Piotr Kiełbasi´nski, Marian Mikołajczyk
3. ORGANOPHOSPHORUS COMPOUNDS 3.1. C-chiral hydroxy phosphorus derivatives C-chiral hydroxy phosphorus derivatives, which have been described so far in the literature, are secondary alcohols. Thus, the syntheses of non-racemic compounds of this type comprise two main approaches (cf. C-chiral hydroxyalkyl sulfones, Section 2.2): asymmetric reduction of the corresponding keto derivatives and resolution of racemic hydroxyalkanephosphorus substrates. Thus, 2-oxoalkanephosphonates 40 were successfully reduced using baker’s yeast to give the corresponding 2-hydroxyalkanephosphonates 41 in reason able yields and with ees of ca. 97% (Equation 22).52a The same con cerns 3-oxoalkanephosphonates52b (for a review see Ref. 52c). Interestingly, 1-oxoalkanephosphonates could not be used as substrates in this reaction due to their instability in aqueous media.52a However, this obstacle was removed by applying lyophilized cells of baker’s yeast or various types of fungi, immobilized on Celite in anhydrous media. The corresponding 1-hydroxyalkanephosphonates were obtained in yields up to 85% and with ees up to 99%.52d
This methodology was also applied to the enantioselective reduction of 3-halo-2-oxopropanephosphonates 40 (R = CH2 Cl, CH2 Br) to give the corre sponding 3-halo-2-hydroxypropanephosphonates in good yield and with ees up to 81%.5354 The 3-chloro derivative was then used in the synthesis of 2,3 epoxyphosphonates and phosphorus analogues of (R)-GABOB53 and of phos phocarnitine (vide infra).54 A series of 2-oxoalkanephosphonates were also screened for reduction with the fungi Geotrichum candidum and only diethyl
Chiral heteroatom-containing compounds
173
2-oxopropanephosphonate 40 (R = Me) proved to undergo the desired reaction to give (+)-(R)-diethyl 2-hydroxypropanephosphonate 41 (R = Me) in 78% yield and with ee = 98%.55 However, the most common and important method of synthesis of chiral non-racemic hydroxy phosphoryl compounds has been the resolution of racemic substrates via a hydrolytic enzyme-promoted acylation of the hydroxy group or hydrolysis of the O-acyl derivatives, both carried out under kinetic resolution conditions. The first attempts date from the early 1990s5657 and have since been followed by a number of papers describing the use of a variety of enzymes and various types of organophosphorus substrates, differing both by the substituents at phosphorus and by the kind of hydroxy (acetoxy)-containing side chain. Thus, these methodologies were used to resolve 1-hydroxyalkanephospho nates 41 by either their enzyme-mediated acylation (Method A) or hydrolysis of the corresponding acyloxy derivatives 42 (Method B) (Equation 23). Selected examples are collected in Table 3.
In a similar way, non-racemic 2-hydroxyalkanephosphonates (phosphine oxides) 43 and their O-acyl derivatives were obtained (Equation 24, Table 4).
The results presented in Tables 3 and 4 deserve some comments. First, a variety of enzymes, including whole-cell preparations, proved suitable for the res olution of different hydroxyalkanephosphorus compounds, giving both unreacted substrates and the products of the enzymatic transformation in good yields and, in some cases, even with full stereoselectivity. Application of both methodologies, acylation of hydroxy substrates rac-41 and rac-43 or the reverse (hydrolysis of the acylated substrates rac-42 and rac-44), enables one to obtain each desired enantiomer of the product. This turned out to be particularly important in those cases when a chemical transformation OH → OAc or reverse was difficult to perform. As an example, our work is shown in Scheme 3. In this case, chemical hydrolysis of the acetyl derivative 46 proved difficult due to some side reactions and therefore an enzymatic hydrolysis, using the same enzyme as that in the acyla tion reaction, was applied. Not only did this provide access to the desired hydroxy derivative 45 but it also allowed to improve its enantiomeric excess. In this way,
174
Table 3 Kinetic resolution of 1-hydroxy(acetoxy)alkanephosphonates R1
R2
R3
Method
Enzyme
Hydroxy compound 41 Yield (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
BnO MeO i-PrO i-PrO MeO i-PrO i-PrO i-PrO i-PrO EtO i-PrO i-PrO i-PrO i-PrO i-PrO i-PrO EtO
Et Ph Ph Ph E-MeCH= CH Et n-Pr n-C5 H11 n-C7 H15 n-C5 H11 o-Tol 2-Naphth 2-Furyl 3-Pyridyl Et i-Bu n-Pr
Me Me Me Me Me Me Me Me CH2 Cl Me CH2 Cl Me Me Me CH2 Cl CH2 Cl n-Pr
A B B B B B B B B B B B B B B B B
PFL F-AP F-AP AP-6 F-AP AP-6 AP-6 AP-6 AP-6 AP-6 AP-6 AP-6 F-AP AP-6 SP-254 P-2 WCPC
41 31 37 46 39 37 39 43 51 38 38 38 34 40 34 44 29
Acyloxy compound 42
[]D
ee (%) (abs. conf.)
Yield (%)
−51 −46 −28 2 −25 1 −9 8 +12 5 +16 2 +15 5 +11 9 +14 0 −28 2 −30 4 −15 8 −30 2 16.2 −22 5 +15
>99 >99 (S) >99 (S) 95 (S) 89 (S) 77 (S) 83 (S) 87 (S) 73 (S) 69 (S) 55 (S) 90 (S) 93 (S) 77 (S) 95 (S) 76 75
− 35a 46a 29a 45 47a 45a 36a 43a 49a 50a 38 26 34 43 41 26
[]D
(+) +41 2a +22 1a +24 6a +16 5 −17 −22 6 −21 8 −22 7 −14 8 +15 2 +44 7 +48 1 +26 2 −38 +38 5 −27
E
Ref.
ee (%) (abs. conf.) − 90a (R) 86a (R) 91a (R) (R) 51 (R) 68 (R) 80 (R) 84 (R) 48 (R) 46 (R) 78(R) 67 (R) 66 (R) n.r. 99 n.r.
n.r. >100 >100 138 35 n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r. n.r.
58 59 59 59 59 60 60 60 60 60 61 61 61 61 62 62 63
Piotr Kiełbasi´nski, Marian Mikołajczyk
Entry
EtO EtO i-PrO MeO EtO EtO i-PrO EtO EtO EtO EtO EtO
Ph Me Me Et Et CH2 = CH CH2 = CH Me Et CH2 = CH Ph 3- PhOC6 H4 (CH2 3
n-Pr Me Me Me Me Me Me Meb Meb Meb Meb Mec
B A A A A A A A A A A A
WCPFL CAL-B CAL-B CAL-B CAL-B CAL-B CAL-B CAL-B CAL-B PS-C PS-C GCL
20 43 42 50 47 41 43 n.r n.r. n.r. n.r. 62
−29 −5 5 −6 1 −4 3 −4 9 −12 7 −14 4 n.r (−) n.r.(−) n.r.(−) n.r. n.r.(−)
80 >95 (R) >95 (R) 51 (R) 80 (R) >95(R) >95 (R) 99 (R) 90 (R) 60 (R) 3 n.r. (R)
50 44 43 31 41 40 44 50 45 40 2 34
+20 +26 5 +20 1 +44 2 +41 3 +23 8 +26 5 n.r.(+) n.r.(+) n.r.(+) n.r. +7 5
n.r. >95 (S) >95 (S) >95 (S) >95 (S) >95 (S) >95 (S) >99 (S) 98 (S) 84 (S) >95 95
n.r. >100 >100 50 95 >100 >100 >500 300 21 n.r. n.r.
63 64 64 64 64 64 64 65 65 65 65 66
Enzymes: PFL, lipase from Pseudomonas fluorescens; F-AP, lipase from Rhizopus orizae; AP-6, lipase from Aspergillus niger; SP-254, lipase from
Aspergillus oryzae; P-2, Chirazyme®; WCPC, whole cell cultures of Penicillium citrinum; WCPFL, whole cell cultures of Pseudomona fluorescens;
CAL-B, lipase from Candida antarctica B; PS-C, lipase from Pseudomonas cepacia; GCL, lipase from Geotrichum candidum. n.r.: not reported.
a Alcohol obtained by chemical hydrolysis of the corresponding enzymatically resolved acyloxy derivative.
b p-Chlorophenyl acetate used as the acetylating agent.
c Isopropenyl acetate used as the acetylating agent.
Chiral heteroatom-containing compounds
18 19 20 21 22 23 24 25 26 27 28 29
175
176
Table 4 Kinetic resolution of 2-hydroxy(acetoxy)alkanephosphonates or phosphine oxides Entry
R1
R2
R3
Method
Enzyme
Hydroxy compound 43 Yield (%)
Ph Ph Ph EtO EtO MeO EtO EtO MeO EtO EtO EtO MeO MeO EtO EtO EtO
Me Et MeOCH2 Me Me n-Bu 2-Pyridyl ClCH2 Me Me Et Me Me Et Ph 2-CF3 C6 H4 2-Furyl
Me Me Me Me Me Me Me Me Me Me Me Mec Mec Mec n-Pr n-Pr n-Pr
A A A A A A A A A A A A A A B B B
RGL RGL RGL AK PS AK LPL AH-S CAL-B CAL-B CAL-B CAL-B CAL-B CAL-B CRL CRL CRL
n.r. n.r. n.r. 44 45 30 46 40 39 40 43 n.r. n.r. n.r. 41 44 45
ee (%) (abs. conf.)
−7 2 +9.8 +8 5 +6 7 +6 4 −11 6a −18 8 +7 0 +13 6 +15 2 +13 1 n.r. (+) n.r. (+) n.r. (+) n.r. n.r. n.r.
88 47 80 90 (R) 87 (R) 92 (R) 62 89 (R) >95 (Sb >95 (Sb 85 Sb 98 90 13 >95 >95 88.7
Yield (%) n.r. n.r. n.r. 47 46 61 45 45 40 39 35 49 47 12 42 42 40
[]D
ee (%) (abs. conf.)
n.r. n.r. n.r. +7 2 +7 1 +59 +32 −1 3 +5 7 +9 7 +8 2 n.r. (+) n.r. (+) n.r. (+) n.r. n.r. n.r.
69 76 77 93 (S) 89 (S) 49 (S) 72 88 (S) >95 (Rb >95 (Rb >95 (Rb 99 >99 >99 >95 >95 93.8
E
16 12 18 95 50 8 11 n.r. >100 >100 >100 >500 >500 >225 >100 >100 93
Ref.
57 57 57 55 55 55 55 54 64 64 64 65 65 65 68 68 68
n.r.: not reported. Enzymes: RGL: rabbit gastric lipase; AK: lipase from Pseudomonas fluorescens (AMANO); PS: lipase from Pseudomonas cepacia (AMANO); LPL: lipoprotein lipase from Pseudomonas aeruginosa; CAL-B: lipase from Candida antarctica B; AH-S: lipase AH-S (AMANO).
a
Specific rotation of the corresponding acetate.
Absolute configurations were ascribed on the basis of an empirical model by analogy to 1-hydroxyalkanephosphonates. They are opposite to those
obtained from the NMR spectra of the Mosher’s esters of 2-hydroxyalkanephosphonates (entries 4–6) and whose absolute configuration is known.52a67 c p-Chlorophenyl acetate was used as the acetylating agent.
b
Piotr Kiełbasi´nski, Marian Mikołajczyk
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
[]D
Acyloxy compound 44
Chiral heteroatom-containing compounds
177
both enantiopure alcohols 45 were obtained and then transformed into (R)- and (S)-phosphocarnitine 47 (Scheme 3).54
Scheme 3
Interestingly, for the transformation of both the racemic 1 hydroxyalkanephosphonates 41 and 2-hydroxyalkanephosphonates 43 into almost enantiopure acetyl derivatives 42 and 44, respectively, a dynamic kinetic reso lution procedure was applied. Pamies and Backvall65 used the enzymatic kinetic resolution in combination with a ruthenium-catalysed alcohol racemization and obtained the appropriate O-acetyl derivatives in high yields and with almost full stereoselectivity (Equation 25, Table 5). It should be mentioned that lowering
178
Piotr Kiełbasi´nski, Marian Mikołajczyk
Table 5 Dynamic kinetic resolution of hydroxyalkanephosphonates R1
Et Et Me Et Me
R2
Me Me Me Et Me
n
Enzyme
0 0 0 0 1
Acetate Yield (%)
ee (%)
86 87 83 85 62
>99 >99 >99 >99 >99
CAL-B PS-C CAL-B PS-C CAL-B
of the yields of 2-acetoxyphosphonates was due to the formation of byproducts, 2-ketophosphonates, arising from the oxidation of the substrates by the catalyst; to prevent this side reaction, a source of hydrogen was sometimes added to the reaction mixture. There are several other examples of C-chiral hydroxy phosphorus com pounds which were obtained in enantiomerically enriched forms using enzymatic methodology. Thus, cis-1-diethylphosphonomethyl-2-hydroxycyclohexane 48 was resolved into enantiomers by enzymatic acetylation: the highest enantioselectivity was achieved using lipase PS in THF or lipase AK without solvent and vinyl acetate as the acetylating agent (Equation 26).69
Lipase
PS AK
Solvent
THF none
48
E
49
Yield (%)
ee (%)
Yield (%)
ee (%)
60 41
62 >99
35 35
97 93
62 152
Chiral heteroatom-containing compounds
179
Prochiral 2-( -phosphono)alkyl-1,3-propanediols 50 were enantioselec tively acetylated to give the corresponding O-acetyl derivatives 51 in very high yields and with high ees (Equation 27).70
Lipase/solvent
PS/THF PS/i-Pr2 O PS/THF PS/THF PS/i-Pr2 O
X
None None CH2 CF2 CF2
Product 51 Yield (%)
ee (%)
98 92 85 98 98
98 98 93 98 98
Two types of racemic 3-hydroxy phosphonates, in which the phosphono and hydroxy moieties are separated by double bond, were successfully resolved using a common enzyme-catalysed acetylation. Both acyclic 52 (Equation 28)71 and cyclic 54 (Equation 29)72 derivatives underwent easy acetylation under the kinetic resolution conditions to give the products in high yield and with almost full stereoselectivity.
Lipase Lipozyme AK PS PPL
(S)-52 ee (%)
(R-53 ee (%)
E
99 97 81 22
96 96 98 99
>200 >200 >200 130
180
Lipase
AK PS AK PS
Piotr Kiełbasi´nski, Marian Mikołajczyk
n
2 1 3 3
54
E
55
Yield (%)
ee (%)
Yield (%)
ee (%)
48 41 48 56
98 <99 >99 70
50 48 50 42
>99 84 >99 97
>200 57 >200 130
Finally, 2-keto-4-hydroxyalkanephosphonates 56 were resolved either by CAL-B-promoted acetylation or by lipase from Candida rugosa (CRL)-mediated hydrolysis of the corresponding O-butyrates 57 (Scheme 4).73
Scheme 4
3.2. C-chiral amino phosphorus compounds C-Chiral amino phosphorus compounds, particularly 1-amino- and 2 aminoalkanephosphonic acids, have been a subject of interest for organic chemists due to their interesting biological activity.1cd Among many methods of their
Chiral heteroatom-containing compounds
181
stereoselective synthesis, some enzymatic approaches have been reported, which constitute a nice alternative to chemical procedures. Stereoselective hydrolysis of racemic 1-(N -phenylacetylamino) alkanephos phonic acids performed in the presence of penicillin acylase under the kinetic resolution conditions gave both the unreacted substrates and the products – the corresponding 1-aminophosphonic acids in high yields and with full enantioselec tivity. The unreacted N -acyl derivatives were hydrolysed chemically and in this way each enantiomer of the free acid was obtained (Scheme 5).74 A number of 1- and 2-aminophosphonates were resolved by a straight CAL B-promoted acetylation of the amino group in the substrates rac-58. Surprisingly, ethyl acetate had to be used as an acetylating agent, since the commonly applied vinyl acetate reacted with aminoalkanephosphonates even in the absence of an enzyme (Equation 30, Table 6).75
Scheme 5
R
Me Me Ph Ph i-PrCH2 PhCH2
1-Aminophosphonic acid Yield (%)
[]D
Abs. conf.
80 80 73 71 76 77
−17 6 +17 0 +24 0 −23 0 −34 5 −50
(L) (R (D) (S (L) (R (D) (S (L) (R (L) (R
182
Piotr Kiełbasi´nski, Marian Mikołajczyk
Table 6 Kinetic resolution of 1- and 2-aminophosphonates75 Entry
1 2 3 4 5 6 7 8 9
R1
R2
Me Me Me Et CH2 = CH Me Me Me Et
n
Et n−Pr i-Pr Et Et Et n-Pr i-Pr Et
1 1 1 1 1 0 0 0 0
Aminophosphonate 58
Acetylaminophosphonate 59
Yield (%)
[]D
ee (%)
Yield (%)
40 41 40 44 >90 41 42 44 73
+6 6 +10 9 +6 6 +7 3 0 −8 0 −1 2 −1 5 −2 7
99 5 100 100 64 0 99 7 90 96 18
54 53 55 42 0 48 42 43 10
[]D
E
ee (%)
+12 +12 5 +16 4 +23 0 +57 2 +56 4 +55 3 +14 4
78 76 72 79 0 90 98 98 100
64 >80 >70 16 0 >100 >200 >200 >200
The aminoalkanephosphonic acids which bear an additional hydroxy group in the molecule were usually resolved via enzymatic acylation of this hydroxy group. For example, resolution of N -Cbz-phosphoserine dimethyl ester 60 using various lipases gave poor results. However, lipase PS-promoted acetylation of N -Cbz-phosphoisoserine diethyl ester 61 gave both the unreacted substrate 61 and the O-acetylated product 62 with almost 100% enantiomeric excess (E = 1000).76 O O Ph
O
O
P(OMe)2 N H
60
OH
Ph
O
O P(OEt)2
N H HO H (R)-61
O Ph
O
N H
O P(OEt)2 H OAc
(S)-62
In turn, an attempt at the resolution of phosphothreonine diester using both the acetylation and hydrolysis approaches proved that the reactions were chemo and diastereoselective, but completely non-enantioselective.77 Another approach to the synthesis of phosphoisoserine was reported by Hammerschmidt et al.78 In the first step, they resolved diisopropyl 2-azido-1-hydroxyethanephosphonate 64
Chiral heteroatom-containing compounds
183
via enzymatic hydrolysis of the acetoxy derivative 63 and then transformed it into the free acid 65 by reduction (Equation 31).78
O
O N3
P(OPri)2 OAc
Lipase, buffer pH 7.0 N3
conv. 50%
O
P(OPri)2
+ N3
P(OPri)2
OH
OAc
(–)-(S)-64
(–)-(R)-63
cy 48%, ee = 97%
cy 47%
rac-63
(31)
MeOH/Et3N
O H2N
P(OH)2
OH (–)-(R)-65
Pd/H2
N3
O P(OPri)2 OH (+)-(R)-64 ee = 97%
(l)-Phosphinotricin 67, which is the active component of naturally occur ring antibiotic “biolaphos,” was synthesized from the corresponding keto acid 66 via reductive amination catalysed by l-glutamate dehydrogenase (EDH) (Equation 32).79
Each enantiomer of 67 was earlier obtained by an -chymotrypsin-catalysed resolution of its diethyl ester 68 or its cyclic analogue 69, followed by chemical hydrolysis (Scheme 6).80 Patent literature reports on the analogous resolutions of phosphinotricin using, among others, penicillin G-acylase,81 penicillin G-amidase,81 subtilisin81 or microorganisms such as Enterobacter aerogenes, Klebsiella oxytoca, Corynebac terium sp., Rhodococcus rubropertinctus and others.82
3.3. P-chiral phosphoro-acetates Resolution of various racemic P-chiral phosphorylacetates 70 involved the same approach as was shown for sulfinylcarboxylates (Scheme 2). However, unlike the case of sulfinyl compounds, only PLE proved efficient for their P(O) analogues.
184
Piotr Kiełbasi´nski, Marian Mikołajczyk
Scheme 6 Thus, a series of racemic phosphinylacetates,8384 phosphonylacetates8586 and phos phorylacetates8586 were resolved into enantiomers via their PLE-promoted hydrol ysis to give products in high yields and with high ees (Equation 33). The most representative, selected examples are collected in Table 7.
The same methodology was applied to the desymmetrization of prochiral phenylphosphinyldiacetate 72. As a result, the monoester 73 was obtained in good yield but with moderate ee (Equation 34).87
No.
1 2 3 4 5 6 7 8
R1
Ph Ph Ph Ph PhO PhO PhO CF3 CH2
R2
Me Et =CH CH2 = MeO MeO EtO EtO MeO
R3
Me Me Me Me Et Me Et Me
Recovered ester 70
Acid 71
Yield (%)
[]D
ee (%)
Abs. conf.
Yield (%)
50 45 40 40 42 66 46 43
+22 +09 7 −54 −16 −10 −3 0 n.r −6 2
82 >96 ∼100 95 99 20 86 99
R R S S n.d. n.d. n.d. S
42 41 22 44 52 22 53 53
[]D −22 2 −8 2 +54 5 +9 1 n.r. +8 7 n.r. +5 2
ee (%)
Abs. conf.
82 81 n.d. 64 77 52 73 82
S S R R n.d. n.d. n.d. R
E
Ref.
n.r. n.r. n.r. n.r. 39 n.r. 17 52
83,84 83,84 83,84 85 86 85 86 86
Chiral heteroatom-containing compounds
Table 7 Enzymatic hydrolysis of phosphoryl-acetates 70
185
186
Piotr Kiełbasi´nski, Marian Mikołajczyk
3.4. P-chiral hydroxy phosphoryl compounds The first P-chiral hydroxy phosphoryl compounds that were enzymatically resolved into enantiomers were o-hydroxyaryl phosphines and their oxides 75. The resolution was achieved via enzyme-assisted hydrolysis of their O-acetyl deriva tives 74, the most effective enzymes being CE and lipase from C. rugosa (CRL) (Equation 35). The highest enantioselectivity was observed in the case of naphthyl derivatives (Equation 36), having a P=O moiety.88
X
Enzyme
O O Lone pair
CE CRL CE
X
Enzyme
O O Lone pair Lone pair
CE CRL CE CRL
ee of 74A (%) (abs. conf.)
ee of 75A (%) (abs. conf.)
E
49 (S 12 (S 33 (R
53 (R 19 (R 49 (S
4.8 1.6 4.0
ee of 74B (%) (abs. conf.)
ee of 75B (%) (abs. conf.)
E
61 (S 69 (S 44 (R 11 (R
89 (R 95 (R 43 (S 15 (S
32 81 3.8 1.5
Chiral heteroatom-containing compounds
187
Hydroxymethylphosphine oxides, hydroxymethylphosphinates and hydrox ymethylphosphonates 76, the primary alcohols with a phosphorus stereogenic cen tre, have been a subject of our intense investigations for several years. This was due to the fact that some of them proved to exhibit herbicidal activity (as part of larger molecules), which was strongly dependent on their absolute configuration.89 Thus, the resolution of hydroxymethylphosphinates and -phosphonates was achieved using lipase-promoted acylation or hydrolysis of the corresponding O-acyl deriva tives (Equation 37) in various solvents: organic solvents (i-Pr2 O, CH2 Cl2 , tBuOMe),90 ionic liquids (IL)91 eg. BMIM PF6 and supercritical carbon dioxide (scCO2 .92 A similar resolution of hydroxymethylphosphine oxides was carried out by Shioji et al.93 in an excess of the acylating agent. Selected results are presented in Table 8.
Inspection of Table 8 clearly shows that the highest enantioselectivity was observed in the case of phosphine oxides when the acylation was performed with a more bulky acyl donor, serving also as solvent (entries 16, 17 and 19).93 The ionic liquid91 enhanced stereoselectivity in comparison with organic solvents8990 only for the substrates which contained larger organic substituents (entry 9 versus entry 8).91 In turn, supercritical carbon dioxide proved to be the worst solvent among those investigated. The main reason seems to be a relatively high polarity of the sub strates for which scCO2 with its low polarity is not a suitable reaction medium.92 An exceptional reversal of the sense of enantioselectivity for the sole example, when CAL-B was used in the acylation of t-butyl(hydroxymethyl)phenylphosphine oxide (entry 17 versus entries 15 and 16), must be at present left without explanation.93 Kinetic resolution of diastereomerically pure racemic 1 hydroxyalkylphosphinates having two stereogenic centres 78 proceeded in a highly enantioselective manner to give both the unreacted substrates and the acetylated products 79 with ees exceeding 98% (Equation 38).94 However, when each diastereomerically pure 1-acetoxy-H-phosphinate 80 was subjected to enzymatic hydrolysis, only one of them, namely (R∗ ,SP∗ )-80, underwent the desired reaction, the other one, (R∗ ,R∗P )-80, being totally unreactive (Scheme 7).95
188
Table 8 Kinetic resolution of racemic hydroxymethylphosphoryl compounds rac-76 Entry
Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph i-PrO Et Et Et Ph Ph t-Bu Ph Ph Ph
R2
MeO MeO MeO MeO EtO EtO EtO i-PrO i-PrO i-PrO MeO i-PrO i-PrO i-PrO t-Bu t-Bu Ph Me c-C6 H11 Me
R3
Me Me Me Me Me Me Me Me Me Me Me Me Me Me Me n-Pr Me n-Pr n-Pr Me
Method
A A A A A A A A A A B A A A A A A A A A
Solvent
i−Pr2 O CH2 Cl2 IL scCO2 i−Pr2 O IL scCO2 i−Pr2 O IL scCO2 CH2 Cl2 /buffer i−Pr2 O IL scCO2 IL PrC(O)Ovinyl MeC(O)Opropenyl PrC(O)Ovinyl PrC(O)Ovinyl MeC(O)Ovinyl
IL, ionic liquid – BMIM PF6 ; for lipase see Tables 3 and 4.
Lipase
PFL PS AK CAL-B PFL AK CAL-B PFL AK CAL-B PFL PFL AK CAL-B AK AK CAL-B AK AK PS
Alcohol 76
Acetate 77
Yield ee (%) (abs. (%) conf.)
Yield ee (%) (abs. (%) conf.)
44 42 34 81 42 36 6 37 36 52 55 39 32 0 33 30 40 40 45 60
39 44 38 5 53 37 82 46 48 46 45 58 55 100 42 55 n.r. 55 45 30
80 (R) 92 (R) 89 (R) 4 (R) 54 (R) 79 (R) 88 (R) 80 (R) 95 (R) 28 (R) 16 86 95 − 43 (S) >98 (S) >98 (R) >98 (S) 90(S) 51
89 (S) 86 (S) 89 (S) 7 (S) 47 (S) 83 (S) 6 (S) 21(S) 80 (S) 27(S) 34 64 50 − 53 (R) 68 (R) n.r. 86(R) 94 (R) 88
E
45 40 51 1.5 5 26 3 5 32 3 n.r. 12 12 − 5 27 21 65 95 23
Ref.
90 90 91 92 90 91 92 90 91 92 90 89 91 92 91 93 93 93 93 93
Piotr Kiełbasi´nski, Marian Mikołajczyk
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
R1
Chiral heteroatom-containing compounds
189
R
Lipase
Conversion (%)
(RP ,R)-78 ee (%)
(SP ,S)-79 ee (%)
Me Me Et Et i-Pr i-Pr
CAL AK CAL AK CAL AK
50 50 17 50 No reaction 9
>98 >98 28 >98 – 2
>98 >98 >98 >98 – >98
Scheme 7 On the basis of this result, a method was developed which allowed to stereoselectively obtain one diastereomer of the product with a high enantiomeric excess, but in a maximum yield of 25%, starting from a diastereomeric mixture of the substrate (Equation 39).95
190
R
p-Tol Ph PhCH2
Piotr Kiełbasi´nski, Marian Mikołajczyk
Starting 80 d.r.
1.3:1 1.5:1 1.3:1
(R,SP )-81
Recovered 80
Yield (%)
ee (%)
Yield (%)
Diastereomeric ratio
14 17 14
>99 >99 74
39 45 76
2.0:1 2.8:1 1.4:1
Finally, prochiral bis(hydroxymethyl)phenylphosphine oxide 82 was desym metrisized using either a lipase-catalysed acetylation (Method A) or hydrolysis of the corresponding diacetyl derivative 83 (Method B), to give the chiral monoac etate 84. Application of the two reverse procedures made it possible to obtain both enantiomerically enriched forms of 84 (Equation 40).87
Lipase
PFL PFL PS PFL PFL
Solvent
CHCl3 i-Pr2 O/buffer THF BMIM PF6 BMIM PF6 /buffer
Method
A B A A B
Monoacetate 84 []D
Yield (%)
ee (%)
Abs. conf.
+3 9 −3 4 +1 8 +0 6 −3 3
50 50 57 35 28
79 68 37 14 65
(R) (S) (R) (R) (S)
Chiral heteroatom-containing compounds
191
3.5. P-chiral hydroxy phosphorus P-boranes Among the useful chiral organophosphorus compounds, the trivalent derivatives play a special role. They are used both as chiral auxiliaries in stoichiometric reactions and as chiral ligands in transition metal catalysts. However, trivalent phosphorus compounds are prone to oxidation and usually difficult to handle. Therefore, their borane complexes, which are stable and which can be easily converted into the parent compounds without racemization, are becoming more and more popular.9697 Among various attempts at their synthesis, the enzy matic approach proved suitable for the preparation of chiral non-racemic hydrox yalkylphosphine P-boranes. Thus, racemic hydroxyalkylphosphine P-boranes rac 85 were subjected to stereoselective acetylation in the presence of various enzymes. As a result of a kinetic resolution, both the unreacted alcohols 85 and acetylation products 86 were obtained in reasonable yields but with mod erate ees only (Equation 41, Table 9). The ee values were higher for the same substrates when the reaction was performed in a highly non-polar solvent – cyclohexane.9899
Also, desymmetrization of prochiral hydroxyalkylphosphine P-boranes was successfully performed using similar reagents and conditions. In the case of bis(hydroxymethyl)phenylphosphine P-borane 87, both its acetylation and hydrol ysis of the diacetyl derivative 89 gave good results, although in addition to the expected monoacetate 88, the diol 87 and diacetate 89 were always present in the reaction mixture (Equation 42).100
192
Lipase
ANL BCL PPL PFL CAL-B CAL-B
Piotr Kiełbasi´nski, Marian Mikołajczyk
Method
Product distribution 87:88:89
A A A A A B
Monoacetate 88
63:36:1 75:23:2 65:34:1 41:53:6 0:48:52 48:49:3
ee (%)
Abs. conf.
70 21 21 62 92 90
R R R R R S
Lipase: ANL, lipase from Aspergillus niger; BCL, lipase from Burkholderia cepacia (formerly Pseudomonas cepacia); CAL-B, lipase from Candida antarctica B; PFL, lipase from Pseudomonas fluorescens; PPL, pig pancreatic lipase.
In turn, bis(2-hydroxyethyl)phenylphosphine P-borane 90 underwent acety lation only in the presence of the lipase from Pseudomonas fluorescens, but its stereochemical outcome depended on the solvent used (Equation 43). The absolute configuration of 91 was not determined.99
Solvent
CHCl3 i-Pr2 O
Monoacetate 91 Yield (%)
[]D
ee (%)
43 42
−10 +1
90 10
3.6. Stereocontrolled transformations of organophosphorus acid esters A variety of phosphoric acid triesters and their derivatives are used as pesticides. Although there are no natural phosphorotriesters, those artificial ones undergo decomposition in the soil, implying that some microorganisms exist which are capable of hydrolysing them. The first report on a stereoselective enzymatic phos photriester hydrolysis was published in 1973, when Dudman and Zerner succeeded
R
Solvent
Enzyme
Recovered alcohol 85 Yield (%)
1 2 3 4 5 6 7 8
Et n-Bu i-Pr t-Bu EtO i-PrO i-PrO i-PrO
c-C6 H12 c-C6 H12 c-C6 H12 c-C6 H12 i-Pr2 O i-Pr2 O i-Pr2 O c-C6 H12
CAL CAL CAL CAL CAL-B AK CAL-B CAL-B
31 23 32 44 38 51 43 31
Acetate 86
[]D
ee (%)
Abs. conf.
Yield (%)
[]D
ee (%)
Abs. conf.
−23 3 −15 3 −10 1 +2 9 −18 9 −9 1 −7 9 −24 1
92 91 79 53 20 14 12 37
R R R R n.d. S S S
58 62 61 42 38 21 39 49
n.r. n.r. n.r. n.r. +17 +12 +7 +11
34 55 67 n.d. n.d. 31 18 29
S S S S n.d. R R R
E
Ref.
6 10 3 6 n.r. n.r. n.r. n.r.
98 98 98 98 99 99 99 99
Chiral heteroatom-containing compounds
Table 9 Enzymatic acetylation of hydroxymethylphosphine P-boranes
193
194
Piotr Kiełbasi´nski, Marian Mikołajczyk
in the enzymatic preparation of optically active n-butyl methyl p-nitrophenyl phos phate starting from the racemic substrate. The enzyme involved in this reaction was an ingredient of the horse or beef serum.101 Later on, phosphotriesterase, an enzyme found in certain native soil bacteria (Pseudomonas diminuta and Flavobacterium sp.), was shown to degrade organophosphates.102 Raushel et al. used this enzyme for hydrolysis of the phosphonothionate 92 and found out that it hydrolysed exclu sively the (S)-enantiomer of the substrate. The reaction proceeded with inversion of configuration at phosphorus and gave the (S)-enantiomer of the thioacid 93 (Equation 44).103
Phosphotriesterase from P. diminuta (PTE) was found to exhibit high hydrolytic activity towards various types of tetracoordinated phosphorus acid esters. Apart from the phosphonothionate 92, phosphoric acid triesters 94 (Equation 45),104 benzenephosphonic acid diester 95 (Equation 46)105 and methyl phenylphosphinic acid ester 96 (Equation 47)106 were also stereoselectively hydrol ysed under kinetic resolution conditions. Of course, in the case of the latter three kinds of substrates, half of the reacting ester was irreversibly lost due to the formation of achiral phosphorus acids.
In the case of the phosphotriesters 94, the use of engineered mutants of phos photriesterase allowed not only to enhance but even to reverse the stereoselectivity of the native enzyme. The ees of the recovered esters exceeded 95%.104
In turn, for the phosphonic acid diester 95, stereoselectivity of the native phosphodiesterase was enhanced by over three orders of magnitude by alteration of the pKa values of the leaving group phenol. For example, for X = CO2 Me, Y = H the stereoselectivity was 5000 higher than for X = NO2 , Y = 2-F.105
Chiral heteroatom-containing compounds
195
Both types of manipulation were applied to achieve the required enantiomer and the highest enantioselectivity in the hydrolysis of the phosphinic acid diester 96.106 Finally, non-racemic phosphorothioic and phosphonothioic acids 98 were obtained via a PTE-catalysed stereoselective hydrolysis of the prochiral substrates 97 (Equation 48).107 The absolute configurations of the thioacids 98 depended on whether native PTE or its mutants were used.
In addition to phosphotriesterase from P. diminuta (PTE) discussed above, two other types of enzymes were found to exhibit phosphotriesterase activity. Interestingly, both are peptidases – the enzymes which in nature hydrolyse a peptide bond. The first one – organophosphorus acid anhydrolase (OPAA) from Alteromonas sp. JD6.5 – is a proline dipeptidase; its original activity is to cleave a dipeptide bond with a prolyl residue at the carboxy terminus.108109 The second one – aminopeptidase P (AMPP) from Escherichia coli – is a proline-specific peptidase that catalyses hydrolysis of N-terminal peptide bonds containing a proline residue.110111 OPAA was found to exhibit stereoselectivity towards phosphorotriester sub strates 94, with a preference for the (S)-enantiomer. Surprisingly, the selectivity was most apparent for the substrates in which the non-hydrolysing substituents did not differ too much. For example, for 94 (R1 = Me, R2 = Et) the chiral preference was 112-fold, while for 94 (R1 = Me, R2 = i-Pr) it was 100-fold.108 Similarly, p-nitrophenyl analogues of sarin 99 and soman 100 (fluorine replaced by the p-nitrophenoxy group) were stereoselectively hydrolysed by OPAA with a 2- to 4-fold preference for the (R)-enantiomer. In the case of the soman analogue 100, the enzyme also exhibited an additional preference for the configuration of the stereogenic carbon atom, which depended on the configuration at phosphorus.109 In turn, AMPP was found to stereoselectively hydrolyse the phosphonoth ionate 101, exhibiting preference towards the (S)-enantiomer. The hydrolysis
196
Piotr Kiełbasi´nski, Marian Mikołajczyk
proceeded with inversion of configuration at phosphorus, similar to that shown in Equation 44.111 All the enzymes discussed above belong to the class of dimetalloenzymes.112 In this context, it should be mentioned that serine-type hydrolases are irre versibly inhibited by organophosphorus esters, among them highly toxic chemical warfare agents. However, in some cases, for example of human butyrylcholi noesterase, the inhibited enzyme could be reactivated by proper mutations.113 Moreover, such mutations were found to confer phosphotriesterase activity in this enzyme!114
4. ORGANOSILANES Stereoselective acylation of racemic hydroxyalkylsilanes 102 was achieved using a crude papain preparation and, surprisingly, 4-phenylpentanoic acid as the acyl donor. The ees of the products 103 and the recovered substrates varied from 30 to 99% (Equation 49).115
In turn, a lipase-promoted acylation of prochiral substrates 104, using lipases from Candida cylinracea and Chromobacterium viscosum and methyl isobutyrate or acetoxime isobutyrate as the acyl donors, gave the product 105 in yields up to 80% and with ees up to 75% (Equation 50).116
Chiral heteroatom-containing compounds
197
5. ORGANOGERMANES Immobilized PLE was applied to promote stereoselective acetylation of prochiral bis(hydroxymethyl)methyl-phenylgermane 106 (R1 = Me) with vinyl acetate as a solvent and acyl donor.117 Later on, the same group reported that each enan tiomer of hydridogermane monoacetates 107 (R1 = H) was obtained either via acetylation of the bis-hydroxy derivative 106 (R1 = H) or hydrolysis of the cor responding diacetate 108 (R1 = H). In both methods, porcine pancreatic lipase was used and, obviously, each reaction led to a different enantiomer of 107 (Equation 51).118
6. FUTURE PERSPECTIVES The concise, though comprehensive, overview presented above clearly shows that practically all kinds of commonly available enzymes can be applied to the trans formations of hetero-organic compounds. In this way, enantiomerically enriched products can be obtained, having a stereogenic centre located either on a het eroatom or on a carbon atom of an organic substituent. This lack of obstacles and important limitations makes the enzymatic methodology very promising and creates a possibility of opening new ways of its use in the preparation of a variety of heteroatom derivatives: (1) A promising direction may be the use of known enzymes to catalyse new unnatural reactions. This approach is based on the so-called “catalytic promiscuity of enzymes,” i.e. on the ability of a single active site of an enzyme to catalyse more than one chemical transformation. The partic ular achievements of Raushel and co-workers (application of prolidases as phosphotriesterases) and Kazlauskas and co-workers (the use of sub tilisin as a sulfinic ester hydrolysing enzyme) paved the way for new research. For example, such an approach makes it possible to consider the sulfinyl and phosphoryl groups as the reacting sites in an enzyme-catalysed chemical transformation with a concomitant stereorecognition of a remote stereogenic centre of a substrate by the enzyme active site. In other words,
198
Piotr Kiełbasi´nski, Marian Mikołajczyk
the phosphate ester and sulfinate ester groups would serve the same pur pose as the carboxylic ester group for the common hydrolytic processes (see Equation 2), opening a new field for the synthesis of C-chiral and heteroatom-chiral compounds. (2) Determination of molecular interactions between unnatural heteroatom sub strates and the aminoacids of the active site of an enzyme should allow broadening the scope of the performed transformations. In this context, a very useful approach may be an initial use of simplified empirical models, which proved to be applicable to heteroatom-chiral and prochiral substrates and irreplaceable for the enzymes whose crystal structure cannot be deter mined (e.g. PLE). A more sophisticated approach, which is based on the molecular modelling, rests upon an appropriate location of a hetero-organic substrate in a known enzyme active site to establish which enantiomer can be preferably bound by the enzyme. It should be noted that such an approach is an almost completely new field for hetero-organic substrates, the only exception being the work by Kazlauskas et al. on the subtilisin-catalysed hydrolysis of sulfinamides. Finally, the most desirable approach would be the experiments in which heteroatom-containing compounds (recognized as good substrates for a given enzyme) are properly transformed to serve as transition state analogues capable of inhibiting this enzyme. This would undoubtedly reveal the real molecular interactions. Moreover, it would be particularly useful for lipases which are known as “induced fit” enzymes, and therefore the binding mode of the hetero-organic substrates within their active sites may substantially differ from that of simple organic molecules. (3) Deracemization of heteroatom-chiral compounds, unlike their C-chiral ana logues, is so far not known at all. It may be quite difficult to achieve since the stereogenic centres located on heteroatoms are sterically stable and, in general, difficult to racemize. Nevertheless, a combination of appro priate enzymes and racemizing agents remains imaginable, among which racemases can also be considered. (4) A search for new enzymes should be continued. Nature proves to be still very unpredictable and sometimes startling discoveries can be made, which are, for example, connected with finding new enzymes of unexpected activity. Phosphotriesterases are the best examples as they catalyse the transformation which is unknown in living systems. (5) Mutagenesis of known enzyme towards a desired activity will be the fastest developing direction. The use of mutants of simple serine-hydrolases, which exhibit the phosphotriesterase activity (in contrast to the native enzymes, which are irreversibly inhibited under such conditions), clearly shows that practically any kind of substrates can be enzymatically transformed. The
Chiral heteroatom-containing compounds
199
analytical methods, which are developing extremely fast, now enable us to have an insight into the enzyme action at the molecular level. (6) Although the use of ribozymes and DNA-zymes is limited to the field of nucleic acid transformations, with small exceptions only that concern their use as catalysts in the Diels–Alder cycloaddition, a strong interest to use them in the stereoselective synthesis of heteroatom compounds can be expected. (7) In the light of the continuously growing number of industrial applications of enzymatic methods, it would be desirable to develop large-scale procedures enabling the synthesis of enantiomerically pure hetero-organic derivatives. Among the most interesting products of this kind, there are enantiomeri cally pure water-soluble phosphines bearing additional substituents capable of coordinating metal ions, which might be used as chiral ligands for organometallic catalysts.
REFERENCES 1. (a) Mikołajczyk, M.; Drabowicz, J.; Kiełbasi´nski, P. Chiral Sulfur Reagents: Applications in Asymmetric and Stereoselective Synthesis; CRC Press: Boca Raton, 1997. (b) Pietrusiewicz, K. M.; Zabłocka, M. Chem. Rev. 1994, 94, 1375–1411. (c) Kafarski, P.; Lejczak, B. Phosphorus Sulfur Silicon 1991, 63, 193–215. (d) Palacios, F.; Alonso, C.; de los Santos, J. M. Chem. Rev. 2005, 105, 899–931. 2. Kiełbasi´nski, P.; Mikołajczyk, M. In Hydrolytic Enzymes in the Synthesis of Non-racemic Heteroorganic Compounds with a Stereogenic Centre on the Heteroatom, Enzymes in Action: Green Solutions for Chemical Problems; Zwanenburg, B.; Mikołajczyk, M.; Kiełbasi´nski, P. (Eds); Kluwer: Dordrecht; 2000, pp. 161–191. 3. (a) Holland, H. Chem. Rev. 1988, 88, 473–485. (b) Colonna, S.; Gaggero, N.; Richelmi, C.; Pasta, P. In Enantioselective Oxidations Catalyzed by Peroxidases and Monooxygenases, Enzymes in Action: Green Solutions for Chemical Problems; Zwanenburg, B.; Mikołajczyk, M.; Kiełbasi´nski, P. (Eds); Kluwer: Dordrecht; 2000, pp. 133–160. (c) Holland, H. L.; Brown, F. M.; Lozada, D.; Mayne, B.; Szerminski, W. R.; van Vliet, A. Can. J. Chem. 2002, 80, 633–639. (d) Bakker, M.; van Rantwijk, F.; Sheldon, R. A. Can. J. Chem. 2002, 80, 622–625. (e) Hodgson, D.; Buist, P. H. Tetrahedron Asymm. 2003, 14, 641–644. 4. Milton, J.; Brand, S.; Jones, M. F.; Rayner, C. M. Tetrahedron Asymm. 1995, 6, 1903–1906. 5. Milton, J.; Brand, S.; Jones, M. F.; Rayner, C. M. Tetrahedron Lett. 1995, 36, 6961–6964. 6. Brand, S.; Jones, M. F.; Rayner, C. M. Tetrahedron Lett. 1995, 36, 8493–8496. 7. Brand, S.; Jones, M. F.; Rayner, C. M. Tetrahedron Lett. 1997, 38, 3595–3598. 8. Csuk, R.; Glanzer, B. I. Chem. Rev. 1991, 91, 49–97. 9. Servi, S. Synthesis 1990, 1–25. 10. Crumbie, R. L.; Deol, B. S.; Nemorin, J. E.; Ridley, D. D. Aust. J. Chem. 1978, 31, 1965. 11. Iriuchijima, S.; Kojima, N. Agric. Biol. Chem. 1978, 42, 451. 12. Yamazaki, T.; Asai, M.; Ohnogi, T.; Lin, J. T.; Kitazume, T. J. Fluorine Chem. 1987, 35, 537. 13. Itoh, T.; Nakai, M.; Sato, T. Tetrahedron Lett. 1985, 26, 771–774. 14. Fujisawa, T.; Kojima, E.; Itoh, T.; Sato, T. Chem. Lett. 1985, 1751–1754. 15. Han, C. Q.; Di Tullio, D.; Wang, Y. F.; Sih, C. J. J. Org. Chem. 1986, 51, 1253–1258.
200
Piotr Kiełbasi´nski, Marian Mikołajczyk
16. Takaishi, Y.; Yang, Y. L.; Di Tullio, D.; Sih, C. J. Tetrahedron Lett. 1982, 23, 5489–5492. 17. (a) Guanti, G.; Banfi, L.; Guaragna, A.; Narisano, E. J. Chem. Soc., Chem. Commun. 1986, 138–139. (b) Guanti, G.; Banfi, L.; Narisano, E. Tetrahedron Lett. 1986, 27, 3547–3550. 18. Ghiringhelli, D. Tetrahedron Lett. 1983, 24, 287–290. 19. (a) Urban, F. J.; Breitenbach, R.; Vincent, L. A. J. Org. Chem. 1990, 55, 3670–3672. (b) Lemaire, M.; Bolte, J. Tetrahedron Asymm. 1999, 10, 4755–4762. 20. Robin, S.; Huet, F.; Fauve, A.; Veschambre, H. Tetrahedron Asymm. 1993, 4, 239–243. 21. Kozikowski, A. P.; Mugrage, B. B.; Li, C. S.; Felder, L. Tetrahedron Lett. 1986, 27, 4817–4820. 22. Ohta, H.; Kato, Y.; Tsuchihashi, G. I. J. Org. Chem. 1987, 52, 2735–2739. 23. Tanikaga, R.; Hosoya, K.; Kaji, A. J. Chem. Soc. Perkin Trans. I 1987, 1799–1803. 24. Nakamura, K.; Ushio, K.; Oka, S.; Ohno, A.; Yasui, S. Tetrahedron Lett. 1984, 25, 3979–3982. 25. Chinchilla, R.; Najera, C.; Pardo, J.; Yus, M. Tetrahedron Asymm. 1990, 1, 575–578. 26. Kiełbasi´nski, P.; Rachwalski, M.; Mikołajczyk, M.; Moelands, M. A. H.; Zwanenburg, B.; Rutjes, F. P. J. T. Tetrahedron Asymm. 2005, 16, 2157–2160. 27. (a) Menasche, N.; Shvo, Y. Organometallics 1991, 10, 3885–3891. (b) Huerta, F. F.; Minidis, A. B. E.; Backvall, J.-E. Chem. Soc. Rev. 2001, 30, 321–331. 28. Kiełbasi´nski, P.; Rachwalski, M.; Mikołajczyk, M. Unpublished results. 29. (a) von Meyer, E. J. Prakt. Chem. 1901, 63, 167. (b) Field, L.; Settlage, P. H. J. Am. Chem. Soc. 1951,73, 5870–5872. (c) Bredereck, H.; Bader, E. Chem. Ber. 1954, 87, 129. 30. Faber, K.; Kroutil, W. Curr. Opin. Chem. Biol. 2005, 9, 181–187. 31. (a) Boltes, I.; Czapi´nska, H.; Kahnert, A.; von Bulow, R.; Dierks, T.; Schmidt, B.; von Figura, K.; Kertesz, M. A.; Uson, I. Structure 2001, 9, 483–491. (b) Faber, K.; Glueck, S. M.; Nestl, B. M.; Larisseger-Schnell, B.; Wallner, S. R.; Kroutil, W. BIOTRANS 2005, July 3–8, 2005, Delft, NL, Abstract book, p. 1. 32. (a) Pogorevc, M.; Kroutil, W.; Wallner, S. R.; Faber, K. Angew. Chem. Int. Ed. Engl. 2002, 41, 4052–4054. (b) Pogorevc, M.; Faber, K. Tetrahedron Asymm. 2002, 13, 1435–1441. (c) Pogorevc, M.; Strauss, U. T.; Riemeier, T.; Faber, K. Tetrahedron Asymm. 2002, 13, 1443– 1447. (d) Wallner, S. R.; Nestl, B. M.; Faber, K. Org. Lett. 2004, 6, 5009–5010. (e) Wallner, S. R.; Nestl, B. M.; Faber, K. Tetrahedron 2005, 61, 1517–1521. 33. Kiełbasi´nski, P.; Albrycht, M.; Mikołajczyk, M.; Wieczorek, M. W.; Majzner, W. R.; Filipczak, A.; Ciołkiewicz, P. Heteroatom Chem. 2005, 16, 93–103. 34. Ohta, H.; Tsuchihashi, G. Chem. Lett. 1986, 217–218. 35. (a) Burgess, K.; Henderson, I. Tetrahedron Lett. 1989, 30, 3633–3636. (b) Burgess, K.; Henderson, I.; Ho, K.-K. J. Org. Chem. 1992, 57, 1290–1295. 36. Kiełbasi´nski, P. Tetrahedron Asymm. 2000, 11, 911–915. 37. Cardellicchio, C.; Naso, F.; Scilimati, A. Tetrahedron Lett. 1994, 35, 4635–4638. 38. Allenmark, S. G.; Andersson, A. C. Tetrahedron Asymm. 1993, 4, 2371–2376. 39. (a) Kiełbasi´nski, P.; Mikołajczyk, M.; Zwanenburg, B.; de Laet, R. C. Phosphorus Sulfur Silicon 1994, 95–96, 495–496. (b) Kiełbasi´nski, P.; Zwanenburg, B.; Damen, T. J. G.; Wieczorek, M. W.; Majzner, W. R.; Bujacz, G. Eur. J. Org. Chem. 1999, 2573–2578. ˙ 40. (a) Mikołajczyk, M.; Kiełbasi´nski, P.; Zurawi´ nski, R.; Wieczorek, M. W.; Błaszczyk, J. Synlett 1994, 127–129. (b) Wieczorek, M. W.; Błaszczyk, J.; Mikołajczyk, M.; Kiełbasi´nski, P.; ˙ Zurawi´ nski, R. Heteroatom Chem. 1994, 5, 167–173. 41. Tamai, S.; Miyauchi, S.; Morizane, C.; Miyagi, K.; Shimizu, H.; Kume, M.; Sano, S.; Shiro, M.; Nagao, Y. Chem. Lett. 1994, 2381–2384. 42. Serreqi, A. N.; Kazlauskas, R. J. Can. J. Chem. 1995, 73, 1357–1367. 43. (a) Morita, S.; Matsubara, J.; Otsubo, K.; Kitano, K.; Ohtani, T.; Kawano, Y.; Uchida, M. Tetrahedron Asymm. 1997, 8, 3707–3710. (b) Kitano, K.; Matsubara, J.; Ohtani, T.; Otsubo, K.; Kawano, Y.; Morita, S.; Uchida, M. Tetrahedron Lett. 1999, 40, 5235–5238. 44. Kiełbasi´nski, P; Albrycht, M.; Mikołajczyk, M. Unpublished results.
Chiral heteroatom-containing compounds
201
45. (a) Medio-Simon, M.; Gil, J.; Alemon, P.; Varea, T.; Asensio, G. Tetrahedron Asymm. 1999, 10, 561–566. (b) Holland, H. L.; Ihasz, N.; Lounsbery, B. J. Can. J. Chem. 2002, 80, 640–642. 46. Saville, C. K.; Magloire, V. P.; Kazlauskas, R. J. J. Am. Chem. Soc. 2005, 127, 2104–2113. 47. Mugford, P. F.; Magloire, V. P.; Kazlauskas, R. J. J. Am. Chem. Soc. 2005, 127, 6536–6537. 48. Our parallel experiments, in which subtilisin Carlsberg was used to promote hydrolysis of N -acetyl-N -benzyl arenesulfinamides, led to exclusive S–N bond breaking. However, the recovered substrates were racemic. Moreover, blank experiments showed that a spontaneous chemical hydrolysis contributed to the process to a much higher degree than that in the cases shown in Ref. 47. Hence, a conclusion was drawn that in our case the hydrolysis proceeded without involvement of the subtilisin active site: Kiełbasi´nski, P.; Albrycht, M.; Mikołajczyk, M. Unpublished results. 49. Mugford, P. F.; Kazlauskas, R. J. BIOTRANS 2005, July 3–8, 2005, Delft, NL, Abstract book, p. 243. 50. Kiełbasi´nski, P. Pol. J. Chem. 1999, 73, 735–738. 51. Toone, E. J.; Werth, M. J.; Jones, J. B. J. Am. Chem. Soc. 1990, 112, 4946–4952. ˙ 52. (a) Zyma´ nczyk-Duda, E.; Lejczak, B.; Kafarski, P.; Grimaud, J.; Fischer, P. Tetrahedron ˙ 1995, 51, 11809–11814. (b) Zyma´ nczyk-Duda, E.; Kafarski, P.; Lejczak, B. Enzym. Microbiol. Technol. 2000, 26, 265–270. (c) Kafarski, P.; Lejczak, B. J. Mol. Catal. B Enzym. 2004, 29, ˙ 99–104. (d) Brzezi´nska-Rodak, M.; Zyma´ nczyk-Duda, E.; Kafarski, P.; Lejczak, B. Biotechnol. Prog. 2002, 18, 1287–1291. 53. Yuan, C.; Wang, K.; Li, Z. Heteroatom Chem. 2001, 12, 551–556. 54. Mikołajczyk, M.; Łuczak, J.; Kiełbasi´nski, P. J. Org. Chem. 2002, 67, 7872–7875. ˙ 55. Zurawi´ nski, R.; Nakamura, K.; Drabowicz, J.; Kiełbasi´nski, P.; Mikołajczyk, M. Tetrahedron Asymm. 2001, 12, 3139–3145. 56. Ramos Tombo, G. M.; Bellus, D. Angew. Chem. Int. Ed. Engl. 1991, 30, 1193–1215. 57. Kagan, H. B.; Tahar, M.; Fiaud, J.-C. Tetrahedron Lett. 1991, 32, 5959–5962. 58. Khushi, T.; O’Toole, K. J.; Sime, J. T. Tetrahedron Lett. 1993, 34, 2375–2378. 59. Li, Y.-F.; Hammerschmidt, F. Tetrahedron Asymm. 1993, 4, 109–120. 60. Drescher, M.; Hammerschmidt, F.; Kahlig, H. Synthesis 1995, 1267–1272. 61. Eidenhammer, G.; Hammerschmidt, F.; Synthesis 1996, 748–754. 62. Hammerschmidt, F.; Wuggenig, F. Tetrahedron Asymm. 1999, 10, 1709–1721. 63. Skwarczy´nski, M.; Lejczak, B.; Kafarski, P. Chirality 1999, 11, 109–114. 64. Zhang, Y.; Yuan, C.; Li, Z. Tetrahedron 2002, 58, 2973–2978. 65. Pamies, O.; Backvall, J.-E. J. Org. Chem. 2003, 68, 4815–4818. 66. Patel, R. N.; Banerjee, A.; Szarka, L. J. Tetrahedron Asymm. 1997, 8, 1055–1059. 67. (a) Kitamura, M.; Tokunaga, M.; Noyori, R. J. Am. Chem. Soc. 1995, 117, 2931–2932. (b) Aboujaoude, E. E.; Collignon, N.; Savignac, P.; Teulade, M. P. Phosphorus, Sulfur, Silicon 1985, 25, 57–61. 68. Zhang, Y.; Li, Z.; Yuan, C. Tetrahedron Lett. 2002, 43, 3247–3249. 69. Yokomatsu, T.; Nakabayashi, N.; Matsumoto, K.; Shibuya, S. Tetrahedron Asymm. 1995, 6, 3055–3062. 70. Yokomatsu, T.; Sato, M.; Shibuya, S. Tetrahedron Asymm. 1996, 7, 2743–2754. 71. Attolini, M.; Iacazio, G.; Peiffer, G.; Maffei, M. Tetrahedron Lett. 2002, 43, 8547–8549. 72. Attolini, M.; Iacazio, G.; Peiffer, G.; Charmasson, Y; Maffei, M. Tetrahedron Asymm. 2004, 15, 827–830. 73. Zhang, Y.; Xu, C.; Li, J.; Yuan, C. Tetrahedron Asymm. 2003, 14, 63–70. 74. Solodenko, V. A.; Kasheva, T. N.; Kukhar, V. P.; Kozlova, E. V.; Mironenko, D. A.; Svedas, V. S. Tetrahedron 1991, 47, 3989–3998. 75. Yuan, C.; Xu, C.; Zhang, Y. Tetrahedron 2003, 59, 6095–6112. 76. Heisler, A.; Rabiller, C.; Douillard, R.; Goalou, N.; Hagele, G.; Levayer, F. Tetrahedron Asymm. 1993, 4, 959–960.
202
Piotr Kiełbasi´nski, Marian Mikołajczyk
77. Heisler, A.; Rabiller, C.; Hagele, G. Phosphorus, Sulfur, Silicon 1995, 101, 273–280. 78. Hammerschmidt, F.; Lindner, W.; Wuggenig, F.; Zarbl, E. Tetrahedron Asymm. 2000, 11, 2955–2964. 79. Fang, J.-M.; Lin, C.-H.; Bradshaw, C. W.; Wong, C.- H. J. Chem. Soc. Perkin Trans. I 1995, 967–978. 80. Natchev, I. J. Chem. Soc. Perkin Trans. I 1989, 125–131. 81. Willms, L.; Fulling, G.; Keller, R. EP 0382113, 1990. 82. Willms, L.; Bartsch, K. EP 0690133, 1996. ˙ 83. Kiełbasi´nski, P.; Zurawi´ nski, R.; Pietrusiewicz, K. M.; Zabłocka, M.; Mikołajczyk, M. Tetra hedron Lett. 1994, 35, 7081–7084. ˙ 84. Kiełbasi´nski, P.; Zurawi´ nski, R.; Pietrusiewicz, K. M.; Zabłocka, M.; Mikołajczyk, M. Pol. J. Chem. 1998, 72, 564–572. 85. Kiełbasi´nski, P.; Góralczyk, P.; Mikołajczyk, M.; Wieczorek, M. W.; Majzner, W. R. Tetra hedron Asymm. 1998, 9, 2641–2650. 86. Sano, S.; Kujime, E.; Takemoto, Y.; Shiro, M.; Nagao, Y. Chem. Pharm. Bull. 2005, 53, 131–134. ˙ 87. Kiełbasi´nski, P.; Zurawi´ nski, R.; Albrycht, M.; Mikołajczyk, M. Tetrahedron Asymm. 2003, 14, 3379–3384. 88. Serreqi, A. N.; Kazlauskas, R. J. J. Org. Chem. 1994, 59, 7609–7615. 89. Spangler, L. A.; Mikołajczyk, M.; Burdge, E. L.; Kiełbasi´nski, P.; Smith, H. C.; Ły˙zwa, P.; Fisher, J. D.; Omela´nczuk, J. J. Agric. Food Chem. 1999, 47, 318–321. 90. Kiełbasi´nski, P.; Omela´nczuk, J.; Mikołajczyk, M. Tetrahedron Asymm. 1998, 9, 3283–3287. 91. Kiełbasi´nski, P.; Albrycht, M.; Łuczak, J.; Mikołajczyk, M. Tetrahedron Asymm. 2002, 13, 735–738. 92. Albrycht, M.; Kiełbasi´nski, P.; Drabowicz, J.; Mikołajczyk, M.; Matsuda, T.; Harada, T.; Nakamura, K. Tetrahedron Asymm. 2005, 16, 2015–2018. 93. Shioji, K.; Ueno, Y.; Kurauchi, Y.; Okuma, K. Tetrahedron Lett. 2001, 42, 6569–6571. 94. Shioji, K.; Tashiro, A.; Shibata, S.; Okuma, K. Tetrahedron Lett. 2003, 44, 1103–1105. 95. Yamagishi, T.; Miyamae, T.; Yokomatsu, T.; Shibuya, S. Tetrahedron Lett. 2004, 45, 6713–6716. ˙ 96. Kiełbasi´nski, P.; Zurawi´ nski, R.; Bałczewski, P.; Mikołajczyk, M. In Alkylphosphorus Com pounds, Comprehensive Organic Functional Group Transformations; Katritzky, A. R.; Taylor, R. J. K. (Eds); Elsevier: Amsterdam; Vol. 2, 2005, pp 383–461. 97. Imamoto, T. Pure Appl. Chem. 2001, 73, 373. 98. Shioji, K.; Kurauchi, Y.; Okuma, K. Bull. Chem. Soc. Jpn 2003, 76, 833–834. ˙ 99. Kiełbasi´nski, P.; Albrycht, M.; Zurawi´ nski, R.; Mikołajczyk, M. J. Mol. Catal. B Enzym. 2006, 39, 45–49. 100. Wiktelius, D.; Johansson, M. J.; Luthmann, K.; Kann, N. Org. Lett. 2005, 7, 4991–4994. 101. Dudman, N.; Zerner, B. J. Am. Chem. Soc. 1973, 95, 3019–3021. 102. Munnecke, D. M. Appl. Environ. Microbiol. 1976, 32, 7–13. 103. Lewis, V. E.; Donarski, W. J.; Wild, J. R.; Raushel, F. M. Biochemistry 1988, 27, 1591–1597. 104. Wu, F.; Li, W. S.; Chen-Goodspeed, M.; Sogorb, M. A.; Raushel, F. M. J. Am. Chem. Soc. 2000, 122, 10206–10207. 105. Li, Y.; Aubert, S. D.; Raushel, F. M. J. Am. Chem. Soc. 2003, 125, 7526–7527. 106. Li, Y.; Aubert, S. D.; Maes, E. G.; Raushel, F. M. J. Am. Chem. Soc. 2004, 126, 8888–8889. 107. Li, W.-S.; Li, Y.; Hill, C. M.; Lum, K. T.; Raushel, F. M. J. Am. Chem. Soc. 2002, 124, 3498–3499. 108. Hill, C. M.; Wu, F.; Cheng, T.-C.; DeFrank, J. J.; Raushel, F. M. Bioorg. Med. Chem. Lett. 2000, 10, 1285–1288. 109. Hill, C. M.; Li, W.-S.; Cheng, T.-C; DeFrank, J. J.; Raushel, F. M. Bioorg. Chem. 2001, 29, 27–35.
Chiral heteroatom-containing compounds
203
110. Jao, S.-C.; Huang, L.-F.; Tao, Y.-S.; Li, W.-S. J. Mol. Catal. B Enzym. 2004, 27, 7–12. 111. Huang L.-F.; Su, B.; Jao, S. C.; Liu, K.-T.; Li, W.-S. ChemBioChem 2006, 7, 506–514. 112. Vanhooke, J. L.; Benning, M. M.; Raushel, F. M.; Holden, H. M. Biochemistry 1996, 35, 6020–6025. 113. Millard, C. B.; Lockridge, O.; Broomfield, C. A. Biochemistry 1995, 34, 15925–15933. 114. Lockridge, O.; Blong, R. M.; Masson, P.; Froment, M.-T.; Millard, C. B.; Broomfield, C. A. Biochemistry 1997, 36, 786–795. 115. Fukui, T.; Kawamoto, T.; Tanaka, A. Tetrahedron Asymm. 1994, 5, 73–82. 116. Djerourou, A. H.; Blanco, L. Tetrahedron Lett. 1991, 32, 6325–6326. 117. Tacke, R.; Wagner, S. A.; Sperlich, J. Chem. Ber. 1994, 127, 639–642. 118. Tacke, R.; Kosub, U.; Wagner, S. A.; Bertermann, R.; Schwarz, S.; Merget, S.; Gunther, K. Organometallics 1998, 17, 1687–1699.
This page intentionally left blank
Future Directions in Biocatalysis Edited by Tomoko Matsuda © 2007 Elsevier B.V. All rights reserved.
205
Chapter 10
Enzymatic polymerization Hiroshi Uyama Department of Applied Chemistry, Graduate School of Engineering, Osaka University, Suita, Japan
Abstract In this chapter, described are recent advances in enzymatic polymerization, defined as chemical polymer syntheses in vitro (in test tubes) via non-biosynthetic pathways cat alyzed by an isolated enzyme. The major target macromolecules formed via the enzymatic polymerizations in this chapter are polyesters and phenolic polymers. For synthesis of polyesters, hydrolases are used as catalyst; hydrolases, enzymes catalyzing a bond-cleavage reaction by water, induce the reverse reaction of hydrolysis, leading to polymer produc tion by a bond-forming reaction. Specific enzyme catalysis provides a novel synthetic route for useful polyesters, many of which are difficult to be synthesized by conventional methodologies. Peroxidase and laccase act as catalyst for oxidative polymerization of var ious phenol derivatives to produce a new class of phenolic polymers without using toxic formaldehyde under mild reaction conditions. Artificial urushi is developed by laccase catalyzed curing of new urushiol analogs. Flavonoid polymers obtained by an enzymatic oxidative polymerization show amplified biological properties.
1. INTRODUCTION Enzymes are proteins catalyzing all in vivo biological reactions. Enzymatic cata lysis can also be utilized for in vitro reactions of not only natural substrates but some unnatural ones. Typical characteristics of enzyme catalysis are high catalytic activity, large rate acceleration of reactions under mild reaction conditions, high selectivities of substrates and reaction modes, and no formation of byproducts, in comparison with those of chemical catalysts. In the field of organic synthetic chemistry, enzymes have been powerful catalysts for stereo- and regioselective reactions to produce useful intermediates and end-products such as medicines and liquid crystals.12 In the recent decades, an enzyme-catalyzed polymerization (“enzymatic polymerization”) has been of increasing importance as a new trend in macro molecular science.3−7 Enzyme catalysis has provided new synthetic strategy for useful polymers, most of which are difficult to produce by conventional chemical
206
Hiroshi Uyama
Table 1 Classification of enzymes and in vitro production of typical polymers catalyzed by respective enzymes Enzymes
Typical polymers
Oxidoreductases Transferases Hydrolases
Phenolic polymers, polyanilines, vinyl polymers Polysaccharides, cyclic oligosaccharides, polyesters Polysaccharides, polyesters, polycarbonates, poly(amino acid)s, polyphosphates
Lyases Isomerases Ligases
catalysts. In vitro enzymatic syntheses of polymers via non-biosynthetic (non metabolic) pathways are, therefore, recognized as a new area of precision polymer syntheses. The enzymatic polymerization also affords a great opportunity for using nonpetrochemical renewable resources as starting substrates of functional polymeric materials, which much contributes to global sustainability without depletion of scarce resources. In the enzymatic polymerizations, the product polymers can be obtained under mild reaction conditions without using toxic reagents. Therefore, the enzymatic polymerization has a large potential as an environmental-friendly synthetic process of polymeric materials, providing a good example to achieve “green polymer chemistry.” Enzymes are generally classified into six groups. Table 1 shows typical poly mers produced with catalysis by respective enzymes. The target macromolecules for the enzymatic polymerization have been polysaccharides, poly(amino acid)s, polyesters, polycarbonates, phenolic polymers, poly(aniline)s, vinyl polymers, etc. In the standpoint of potential industrial applications, this chapter deals with recent topics on enzymatic synthesis of polyesters and phenolic polymers by using enzymes as catalyst.
2. ENZYMATIC SYNTHESIS OF POLYESTERS Many studies concerning syntheses of aliphatic polyesters by fermentation and chemical processes have been made in viewpoint of biodegradable materials.89 Recently, another approach to synthesis of biodegradable polyesters has been developed by polymerization using lipase as catalyst.10−15 It is generally accepted that an enzymatic reaction is virtually reversible, and hence, the equilibrium can be controlled by appropriately selecting the reaction conditions. Based on this concept, many of the hydrolases, which are enzymes catalyzing a bond-cleavage
Enzymatic polymerization
207
Ring-opening polymerization of lactones O C O
O
Lipase
ORC R
n
Polycondensation of dicarboxylic acids or their derivatives with glycols Lipase XO2CRCO2X + HOR'OH
O O CRC OR'O
–XOH X: H, alkyl, halogenated alkyl, vinyl, etc.
n
Polycondensation of oxyacids or their esters Lipase
O
HORCO2X
ORC n –XOH X: H, alkyl, halogenated alkyl, vinyl, etc.
Scheme 1 reaction by hydrolysis, have been employed as catalyst for the reverse reaction of hydrolysis, leading to polymer production by a bond-forming reaction. Lipase is an enzyme which catalyzes the hydrolysis of fatty acid esters normally in an aqueous environment in living systems. However, lipases are some times stable in organic solvents and can be used as catalyst for esterifications and transesterifications. By utilizing such catalytic specificities of lipase, functional aliphatic polyesters have been synthesized by various polymerization modes. Typ ical reaction types of lipase-catalyzed polymerization leading to polyesters are summarized in Scheme 1. Lipase-catalyzed polymerizations also produced poly carbonates and polyphosphates. 2.1. Ring-opening polymerization to polyesters Various cyclic esters have been subjected to lipase-catalyzed ring-opening poly merization. Lipase catalyzed the ring-opening polymerization of 4- to 17 membered non-substituted lactones.16−18 In 1993, it was first demonstrated that medium-size lactones, -valerolactone (-VL, six-membered) and -caprolactone (-CL, seven-membered), were polymerized by lipases derived from Candida cylindracea, Burkholderia cepacia (lipase BC), Pseudomonas fluorescens (lipase PF), and porcine pancreas (PPL).1920 -Propiolactone (-PL, 4-membered) was polymerized by Pseudomonas family lipases as catalyst in bulk, yielding a mixture of linear and cyclic oligomers with molecular weight of several hundreds,21 whereas poly(-PL) of high molec ular weight (molecular weight > 5 × 104 ) was obtained by using Candida rugosa lipase (lipase CR).22
208
Hiroshi Uyama
Substituted four-membered lactones were polymerized via the lipase catal ysis. The lipase-catalyzed polymerization of -butyrolactone (-BL) produced poly(-hydroxybutyrate) (PHB),23 which is a polyester having similar structure produced in vivo by bacteria for an energy-storage substance. The molecular weight of PHB reached 7300 in the polymerization using PPL as catalyst at 100 C.24 Lipase CR also showed high catalytic activity toward the polymeriza tion at high temperature. The resulting products contained a significant amount of cyclic oligo(-hydroxybutyrate)s, which were formed by the lipase-catalyzed reaction of linear PHB.25 Poly(malic acid) is a biodegradable and bioadsorbable water-soluble polyester having a carboxylic acid in the side chain. The chemoenzymatic syn thesis of poly(malic acid) was achieved by the lipase-catalyzed polymerization of benzyl -malolactonate, followed by the debenzylation.26 The molecular weight of poly(benzyl -malolactonate) increased on copolymerization with a small amount of -PL using lipase CR catalyst.27 Five-membered unsubstituted lactone, -butyrolactone (-BL), is not poly merized by conventional chemical catalysts. However, oligomer formation from -BL was observed by using PPL or Pseudomonas sp. lipase as catalyst.2328 Enzy matic polymerization of six-membered lactones, -VL and 1,4-dioxan-2-one, was reported. -VL was polymerized by various lipases of different origins.1929 The molecular weight of the enzymatically obtained polymer was relatively low (less than 2000). In polyester synthesis via ring-opening polymerizations, metal catalysts are often used. For medical applications of polyesters, however, there has been con cern about harmful effects of the metallic residues. Enzymatic synthesis of a metal-free polyester was demonstrated by the polymerization of 1,4-dioxan-2-one using Candida antarctica lipase (lipase CA).29 Under appropriate reaction con ditions, the high molecular weight polymer (molecular weight = 41 × 104 ) was obtained. -CL (seven-membered lactone) is industrially manufactured and its oligomer having a hydroxy group at both ends is widely used as soft segment of polyurethanes. High molecular weight poly(-CL) is a commercially avail able biodegradable plastic. So far, various commercially available lipases have catalyzed the -CL polymerization. In the case of crude industrial lipases (PPL, lipases CR, BC, and PF), a large amount of catalyst (often more than 40 wt% for -CL) was required for the efficient production of the polymer.3031 On the other hand, lipase CA showed high catalytic activity toward the -CL polymerization; a very small amount of lipase CA (less than 1 wt% for -CL) was enough to induce the polymerization.32 Under appropriate conditions, poly(-CL) with the molec ular weight close to 105 was obtained.3334 During the polymerization of -CL, degradation simultaneously took place.35 The polymerization in bulk produced the linear polymer, whereas the main product obtained in organic solvents was of cyclic structure, suggesting that intramolecular condensation took place during the
Enzymatic polymerization
209
polymerization. Recently, microwave-assisted lipase-catalyzed polymerization of -CL was reported.36 In the polymerization of -CL catalyzed by lipase CA in organic solvents, the polymer was obtained efficiently in solvents having log P (a parameter of hydrophobicity) values from 1.9 to 4.5, whereas solvents with log P from −11 to 0.5 showed low propagation rate.37 Among the solvents examined, toluene was the best solvent to produce high molecular weight poly(-CL) efficiently. Variation in the ratio of toluene to -CL in the reaction at 70 C showed that the monomer conversion and polymer molecular weight were the largest for a ratio about 2:1. Furthermore, lipase CA could be reused for the polymerization. In the range of five cycles, the catalytic activity hardly changed. The kinetics of the -CL bulk polymerization by lipase CA showed linear relationships between the monomer conversion and the molecular weight of the polymer; however, the total number of the polymer chains was not constant during the polymerization.38 The monomer consumption apparently followed a first-order rate law. Ring-opening polymerization of -methyl-substituted medium-size lactones, -methyl--valerolactone and -methyl--caprolactone, proceeded by using lipase CA catalyst in bulk.39 Lipase CA efficiently catalyzed the ring-opening poly merization of 1,5-dioxepan-2-one.40 A linear relationship between the monomer conversion and the molecular weight of the polymer was observed. The monomer consumption followed a first-order rate law, suggesting no termination and chaintransfer reaction. The enzymatic polymerizability of 1,5-dioxepan-2-one was much larger than that of -CL. A nine-membered lactone, 8-octanolide (OL), was enzymatically polymer 41 ized. Lipases BC and CA showed high catalytic activity for the polymer ization. Four unsubstituted macrolides, 11-undecanolide (12-membered, UDL),42 12-dodecanolide (13-membered, DDL),43 15-pentadecanolide (16-membered, PDL),44−46 and 16-hexadecanolide (17-membered),47 were enzymatically polymer ized. Various lipases catalyzed the polymerization of these macrolides. For the polymerization of DDL, the activity order of the catalyst was lipase BC > lipase PF > lipase CR > PPL. The lipase CA-catalyzed polymerization of PDL proceeded fast in toluene to produce a high molecular weight polymer with the molecular weight higher than 8 × 104 . Enzymatic ring-opening polymerization of macrolides (UDL, DDL, and PDL) proceeded even in an aqueous medium. The single crystals of the aliphatic polyesters enzymatically synthesized from these macrolides were prepared and their crystal structure was examined.48−50 Enzymatic synthesis of aliphatic polyesters was also achieved by the ringopening polymerization of cyclic diesters. Lactide was not enzymatically polymer ized under mild reaction conditions; however, poly(lactic acid) with the molecular weight higher than 1 × 104 was formed using lipase BC as catalyst at higher tem peratures (80–130 C).5152 Protease (proteinase K) also induced the polymerization; however, the catalytic activity was relatively low.
210
Hiroshi Uyama
Scheme 2 Lipases CA, BC, and PF catalyzed the polymerization of ethylene dode canoate and ethylene tridecanoate to give the corresponding polyesters.53 The enzyme origin affected the polymerization behaviors; in using lipase BC catalyst, these bislactones polymerized faster than -CL and DDL, whereas the reactivity of these cyclic diesters was in the middle of -CL and DDL in using lipase CA. It is well known that the catalytic site of lipase is a serine residue and lipasecatalyzed reactions proceed via an acyl–enzyme intermediate. The enzymatic poly merization of lactones is explained by considering the following reactions as the principal reaction course (Scheme 2). The key step is the reaction of lactone with lipase involving the ring opening of the lactone to give an acyl–enzyme interme diate (“enzyme-activated monomer,” EM). The initiation is a nucleophilic attack of water, which is contained partly in the enzyme, onto the acyl carbon of the intermediate to produce -hydroxycarboxylic acid (n = 1), the shortest propagat ing species. In the propagation stage, the intermediate is nucleophilically attacked by the terminal hydroxyl group of a propagating polymer to produce a one-unit more elongated polymer chain. The kinetics of the polymerization showed that the rate-determining step of the overall polymerization is the formation of the enzyme-activated monomer. Thus, the polymerization probably proceeds via an “activated-monomer mechanism.” Reactivity of cyclic compounds generally depends on their ring size; small and intermediate ring-size compounds possess higher ring-opening reactivity than macrocyclic lactones (macrolides) due to their larger ring strain. Table 2 summa rizes dipole moment values and reactivities of lactones with different ring size. The dipole moment (indication of ring strain) of the macrolides is lower than that of -VL and -CL and close to that of an acyclic fatty acid ester (butyl caproate). The rate constant of the macrolides in anionic polymerization is much smaller than
Enzymatic polymerization
211
Table 2 Dipole moments and reactivities of unsubstituted lactones Ring size of lactones
6 7 9 12 13 16 17 Butyl caproate a b
Dipole moment (Cm)
4.42 4.45 — 1.86 1.86 1.86 — 1.75
Relative polymerization rate Enzymatica
Chemicalb
— 0.10 — 0.13 0.19 0.74 1.00 —
2500 330 21 09 10 09 10 —
Calculated from Michaelis–Menten constants using lipase catalyst.
Polymerization with zinc octanoate/butyl alcohol initiator system in bulk.
that of -VL and -CL. These data clearly show that the macrolides have much lower ring strain, and hence, show less anionic reactivity and polymerizability than the medium-size lactones. On the other hand, the macrolides showed unusual enzymatic reactivity. Lipase PF-catalyzed polymerization of the macrolides proceeded much faster than that of -CL. The lipase-catalyzed polymerizability of lactones was quantitatively evaluated by Michaelis–Menten kinetics. For all monomers, linearity was observed in the Hanes–Woolf plot, indicating that the polymerization followed Michaelis– Menten kinetics. The Vmax lactone and Vmax lactone /Km lactone values increased with the ring size of lactone, whereas the Km lactone values scarcely changed. These data imply that the enzymatic polymerizability increased as a function of the ring size, and the large enzymatic polymerizability is governed mainly by the reaction rate (Vmax , but not to the binding abilities, i.e., the reaction process of the lipase–lactone complex to the acyl–enzyme intermediate is the key step of the polymerization. Lipase catalyzed the ring-opening copolymerization of cyclic monomers. In 1993, the first example of the enzymatic ring-opening copolymerization of lactones was demonstrated; -VL and -CL were copolymerized by lipase PF catalyst.54 The resulting copolymer was of random structure having both units. Random ester copolymers were also enzymatically synthesized from other combinations: -CL OL, -CL-PDL, and OL-DDL. The formation of the random copolymers in spite of the different enzymatic polymerizabilities of these lactones suggests that the intermolecular transesterifications of the polyesters frequently took place during the copolymerization. By utilizing this specific lipase catalysis, random ester copolymers were synthesized by the lipase-catalyzed polymerization of macrolides in the presence of poly(-CL).55
212
Hiroshi Uyama
2.2. Polycondensation of dicarboxylic acid derivatives and glycols to polyesters So far, various dicarboxylic acid derivatives, dicarboxylic acids, their activated and non-activated esters, cyclic acid anhydrides, and polyanhydrides have been polymerized with glycols through lipase catalysis to give polyesters. Many diacids are commercially available; however, their enzymatic reactiv ity is relatively low. Therefore, the enzymatic polymerization of dicarboxylic acids and glycols was often carried out under vacuum for production of high molecular weight polyesters.5657 In the Mucor miehei lipase (lipase MM)-catalyzed poly merization in hydrophobic solvents of high boiling points such as diphenyl ether and veratrole using a programmed vacuum profile, the molecular weight reached higher than 4 × 104 . The increase of the molecular weight of aromatic polyesters was also observed by the polymerization under vacuum.58 In the polymerization of isophthalic acid and 1,6-hexanediol using lipase CA as catalyst at 70 C, the polyester with molecular weight of 55 × 104 was formed, whereas lipase MM only produced the corresponding oligomer. The effects of substrates and solvents on the formation, molecular weight, and end-group structure of the polymer in the polycondensation using lipase CA as catalyst were systematically investigated.59 Diphenyl ether was found to be the preferred solvent that gave the polyester of the highest molecular weight. Concerning the effect of the monomer structure, the longer chain length diacids (sebacic and adipic acid) and diols (1,8-octanediol and 1,6-hexanediol) gave higher enzymatic reactivity than the shorter chain length of diacids (succinic and glutaric acid) and 1,4-butanediol. Aliphatic polyesters were reported to be synthesized by enzymatic polymer ization of dicarboxylic acids and glycols in a solvent-free system.6061 Lipase CA efficiently catalyzed the polymerization under mild reaction conditions, despite the heterogeneous mixture of the monomers and catalyst. Methylene chain length of the monomers greatly affected the polymer yield and molecular weight. The polymer with molecular weight higher than 1 × 104 was obtained by the reaction under reduced pressure. A small amount of adjuvant was effective for the poly mer production when both monomers were solid at the reaction temperature.62 Scale-up experiment produced the polyester from adipic acid and 1,6-hexanediol in more than 200 kg yield.63 This solvent-free system claimed a large potential as an environmental-friendly synthetic process of polymeric materials owing to the mild reaction conditions and no use of organic solvents and toxic catalysts. The polymerization of adipic acid and 1,8-octanediol in bulk was investi gated by using lipase CA immobilized on different resins as well as lipase CA free of the immobilized resin.64 The immobilized lipase induced the polymerization more efficiently to produce the polyester of higher molecular weight. Under a wide range of reaction conditions, the molecular index of the resulting polymer without fractionation was less than 1.5, suggesting that lipase CA catalyzes the chain growth with chain length selectivity.
Enzymatic polymerization
213
The effects of the feed ratio in the lipase CA-catalyzed polymerization of adipic acid and 1,6-hexanediol were examined by using NMR and MALDI-TOF mass spectroscopies.65 1 H NMR analysis showed that the hydroxyl terminated product was preferentially formed at the early stage of the polymerization in the stoichiometric substrates. As the reaction proceeded, the carboxyl-terminated product was mainly formed. Even in the use of an excess of the dicarboxylic acid monomer, the hydroxy-terminated polymer was predominantly formed at the early reaction stage, which is a specific polymerization behavior due to the unique enzyme catalysis. A dehydration reaction is generally realized in non-aqueous media. Since a product water of the dehydration is in equilibrium with starting materials, the solvent water disfavors the dehydration to proceed in an aqueous medium due to “the law of mass action.” Nevertheless, the present authors have found that lipase catalysis provided a dehydration polymerization of a dicarboxylic acid and glycol in water.6667 The view of dehydration in an aqueous medium is a new aspect in organic chemistry. Lipases BC, CA, CR, MM, and PF were active for the polymerization of sebacic acid and 1,8-octanediol. In the polymerization of dicarboxylic acid and glycol, the polymerization behavior greatly depended on the methylene chain length of the monomers. The polymer was obtained in good yields from 1,10-decanediol, whereas no polymer formation was observed using 1,6-hexanediol, suggesting that the combination of the monomers with appropriate hydrophobicity is favored for efficient polymer formation. Alkyl esters often show low reactivity for lipase-catalyzed transesterifi cations with alcohols. Therefore, it is difficult to obtain high molecular weight polyesters by lipase-catalyzed polycondensation of dialkyl esters with glycols. The molecular weight greatly improved by polymerization under vacuum to remove the formed alcohols, leading to a shift of equilibrium toward the product polymer; the polyester with molecular weight of 2 × 104 was obtained by the lipase MMcatalyzed polymerization of sebacic acid and 1,4-butanediol in diphenyl ether or veratrole under reduced pressure.68 Activated esters of halogenated alcohols, such as 2-chloroethanol, 2,2,2 trifluoroethanol, and 2,2,2-trichloroethanol, have been often used as substrate for enzymatic synthesis of esters, owing to an increase in the electrophilicity (reactivity) of the acyl carbonyl and avoid significant alcoholysis of the products by decreasing the nucleophilicity of the leaving alcohols.1 The enzymatic synthesis of polyesters from activated diesters was achieved under mild reaction conditions. The polymerization of bis(2,2,2-trichloroethyl) glutarate and 1,4-butanediol proceeded in the presence of PPL at room temperature in diethyl ether to produce the polyesters with molecular weight of 82 × 103 .69 Vacuum was applied to shift the equilibrium forward by removal of the activated alcohol formed, leading to the production of high molecular weight polyesters.68 The polycondensation of bis(2,2,2-trifluoroethyl) sebacate and aliphatic diols took place using lipases BC, CR, MM, and PPL as catalyst in diphenyl ether. Under the
214
Hiroshi Uyama
appropriate reaction conditions, the polymer with molecular weight higher than 4 × 104 was obtained and lipase MM showed the highest catalytic activity.5770 In the PPL-catalyzed polymerization of bis(2,2,2-trifluoroethyl) glutarate with 1,4 butanediol in 1,2-dimethoxybenzene, the periodical vacuum method increased the molecular weight to nearly 4 × 104 .71 An irreversible procedure for the lipase-catalyzed acylation using vinyl esters as acylating agent has been developed, where a leaving group of vinyl alcohol tautomerizes to acetaldehyde. In these cases, the reaction with the vinyl esters proceeds much faster to produce the desired compounds in higher yields, in comparison with the alkyl esters. Divinyl esters reported first by us are efficient monomers for polyester pro duction under mild reaction conditions.72 In the lipase PF-catalyzed polymerization of divinyl adipate and 1,4-butanediol in diisopropyl ether at 45 C, a polyester with molecular weight of 67 × 103 was formed, whereas adipic acid and diethyl adipate did not afford the polymeric materials under similar reaction conditions (Scheme 3). Lipase-catalyzed polymerization of divinyl ester and glycol is proposed to proceed as follows (Scheme 4). First, the hydroxy group of the serine residue nucleophilically attacks the acyl carbon of the divinyl ester monomer to produce an acyl–enzyme intermediate (EM) involving elimination of acetaldehyde. The reaction of EM with the glycol produces 1:1 adduct of both monomers. In the propagation stage, the nucleophilic attack of the terminal hydroxy group takes place on the acyl–enzyme intermediate formed from the vinyl ester group of the
Scheme 3
Enzymatic polymerization
215
Scheme 4
monomer and 1:1 adduct, and subsequently the propagation steps keep going similarly. Lipases BC, CA, MM, and PF showed high catalytic activity toward the polymerization of divinyl adipate or divinyl sebacate with -glycols with dif ferent chain lengths.73 A combination of divinyl adipate, 1,4-butanediol, and lipase PC afforded the polymer with molecular weight of 21×104 . The yield of the poly mer from divinyl sebacate was higher than that from divinyl adipate, whereas the opposite tendency was observed in the polymer molecular weight. The polyester formation was observed in various organic solvents, and among them, diisopropyl ether gave the best results. During the lipase-catalyzed polymerization of divinyl esters and glycols, there was a competition between the enzymatic transesterification and the hydro lysis of the vinyl end group, resulting in the limitation of the polymer growth.74 A mathematical model showing the kinetics of the polymerization predicts the product composition (terminal structure).75 A comparison of the experimental data and model predictions suggests that the molecular weight and terminal group func tionality of polyesters can be controlled by selection of biocatalysts. The reaction calorimetry was used to monitor the kinetics of the polymerization.76 The reaction rate increased as a function of the monomer concentration. As the polymerization proceeded, the rate constant for the polyester synthesis was significantly reduced. A batch-stirred reactor was developed to minimize temperature and mass-transfer effects.77 Using this reactor, poly(1,4-butylene adipate) with the molecular weight of 23 × 104 was synthesized in only 1 h at 60 C.
216
Hiroshi Uyama
Aromatic polyesters were efficiently synthesized from aromatic diacid divinyl esters. Lipase CA induced the polymerization of divinyl esters of isoph thalic acid, terephthalic acid, and p-phenylene diacetic acid with glycols to give polyesters containing aromatic moiety in the main chain.78 The highest molecular weight (72 × 103 ) was attained from a combination of divinyl isophthalate and 1,10-decanediol. Enzymatic polymerization of divinyl esters and aromatic diols also afforded the aromatic polyesters.79 A combinatorial approach for biocatalytic production of polyesters was demonstrated.80 A library of polyesters were synthesized in 96 deep-well plates from a combination of divinyl esters and glycols with lipases of different origin. In this screening, lipase CA was confirmed to be the most active biocatalyst for the polyester production. As acyl acceptor, 2,2,2-trifluoroethyl esters and vinyl esters were examined and the former produced the polymer of higher molecular weight. Various monomers such as carbohydrates, nucleic acids, and a natural steroid diol were used as acyl acceptor. Lipase-catalyzed copolymerization of divinyl esters, glycols, and lac tones produced ester copolymers with molecular weight higher than 1 × 104 (Scheme 5).81 Lipases BC and CA showed high catalytic activity for this copoly merization. 13 C NMR analysis showed that the resulting product was not a mix ture of homopolymers, but a copolymer derived from the monomers, indicating that two different modes of polymerization, polycondensation and ring-opening polymerization, simultaneously take place through enzyme catalysis in one pot to produce ester copolymers. Furthermore, this result strongly suggests the fre quent occurrence of transesterification between the resulting polyesters during the polymerization. Acid anhydride derivatives are also good acylating reagents through lipase catalysis. A new type of enzymatic polymerization involving lipase-catalyzed ringopening poly(addition–condensation) of cyclic anhydride with glycols was demon strated (Scheme 6).82 The polymerization of succinic anhydride with 1,8-octanediol using lipase PF catalyst proceeded at room temperature to produce the polyester. Glutaric and diglycolic anhydrides were polymerized with -alkylene glycols
Scheme 5
Enzymatic polymerization
217
Scheme 6 in the presence of lipase CA in toluene to give the polyesters.83 Under appropriate reaction conditions, the molecular weight reached 1 × 104 . Polyester synthesis was carried out by insertion–dehydration of glycols into polyanhydrides using lipase CA as catalyst (Scheme 6).84 The insertion of 1,8-octanediol into poly(azelaic anhydride) took place at 30–60 C to give the corresponding polyester with molecular weight of several thousands. Effects of the reaction parameters on the polymer yield and molecular weight were systematically investigated.83 The dehydration reaction also proceeded in water. The reaction behaviors depended on the monomer structure and reaction media. Lipase-catalyzed synthesis of polyesters from cyclic anhydrides and oxi ranes was reported.8586 The polymerization took place by PPL catalyst and the molecular weight reached 1 × 104 under the selected reaction conditions. During the polymerization, the enzymatically formed acid group from the anhydride may open the oxirane ring to give a glycol, which is then reacted with the anhydride or acid by lipase catalysis, yielding the polyesters. Oxyacids or their esters were enzymatically polymerized to produce the cor responding polyesters. Polyesters of relatively high molecular weight were enzy matically produced from 10-hydroxydecanoic acid87 and 11-hydroxyundecanoic acid88 using a large amount of lipase CR catalyst (10 weight fold for the monomer). In the case of 11-hydroxyundecanoic acid, the corresponding polymer with molec ular weight of 22 × 104 was obtained in the presence of activated molecular sieves. Lipase CA also polymerized hydrophobic oxyacids efficiently.89 The DP value was beyond 100 in the polymerization of l6-hydroxyhexadecanoic acid, 12-hydroxydodecanoic acid, or 10-hydroxydecanoic acid under vacuum at high temperature (90 C) for 24 h, whereas the polyester with lower molecular weight was formed from 6-hydroxyhexanoic acid under similar reaction conditions. This difference may be due to the lipase–substrate interaction. Lipase-catalyzed polymerization of oxyacid esters was reported. PPL cat alyzed the polymerization of methyl 6-hydroxyhexanoate.20 The polymer with DP up to 100 was synthesized by polymerization in hexane at 69 C for more than 50 days. The PPL-catalyzed polymerization of methyl 5-hydroxypentanoate for 60 days produced the polymer with DP of 29. Solvent effects were
218
Hiroshi Uyama
Scheme 7
systematically investigated; hydrophobic solvents such as hydrocarbons and diiso propyl ether were suitable for the enzymatic production of high molecular weight polymer. Ester–thioester copolymers were enzymatically synthesized (Scheme 7).90 The lipase CA-catalyzed copolymerization of -caprolactone with 11 mercaptoundecanoic acid or 3-mercaptopropionic acid under reduced pressure pro duced the polymer with molecular weight higher than 2 × 104 . The thioester unit of the resulting polymer was lower than the feed ratio. The transesterification between poly(-caprolactone) and 11-mercaptoundecanoic acid or 3-mercaptopropionic acid also took place by lipase CA catalyst. Recently, aliphatic polythioesters were synthesized by lipase CA-catalyzed polycondensation of diesters with 1,6 hexanedithiol.91 Enzymatic synthesis of polyesters in green solvents such as supercritical fluids and ionic liquids were demonstrated. Supercritical carbon dioxide (scCO2 ) was employed as solvent for the polycondensation of divinyl adipate and 1,4 butanediol.92 Quantitative consumption of both monomers was achieved to give the polyester with molecular weight of 39 × 103 , indicating that scCO2 was a good medium for the enzymatic polycondensation. The polymerization of bis(2,2,2 trichloroethyl) adipate and 1,4-butanediol using PPL catalyst proceeded in a supercritical fluoroform solvent to give the polymer with molecular weight of several thousands.93 By changing the pressure, the low-dispersity polymer fractions were separated. Room-temperature ionic liquids have received much attention as green designer solvents. We first demonstrated that ionic liquids acted as good medium for lipase-catalyzed production of polyesters. The polycondensation of diethyl adipate and 1,4-butanediol using lipase CA as catalyst efficiently proceeded in 1-butyl-3-methylimidazolinium tetrafluoroborate or hexafluorophosphate under reduced pressure.94 The polymerization of diethyl sebacate and 1,4-butanediol in 1-butyl-3-methylimidazolinium hexafluorophosphate took place even at room temperature in the presence of lipase BC.95
Enzymatic polymerization
219
2.3. Enzymatic synthesis of functional polyesters Lipase catalysis is often used for enantioselective production of chiral compounds. Lipase induced the enantioselective ring-opening polymerization of racemic lac tones. In the lipase-catalyzed polymerization of racemic -BL, the enantioselec tivity was low; an enantioselective polymerization of -BL occurred by using thermophilic lipase to give (R)-enriched PHB with 20–37% enantiomeric excess (ee).96 The enantioselectivity was greatly improved by the copolymerization with 7- or 13-membered non-substituted lactone using lipase CA catalyst (Scheme 8); the ee value reached ca. 70% in the copolymerization of -BL with DDL.97 It is to be noted that in the case of lipase CA catalyst, the (S)-isomer was prefer entially reacted to give the (S)-enriched optically active copolymer. The lipase CA-catalyzed copolymerization of -caprolactone (6-membered) with DDL enan tioselectively proceeded, yielding the (R)-enriched optically active polyester with ee of 76%. Optically active polyesters were synthesized by lipase CA-catalyzed ring-opening polymerization of racemic 4-methyl or ethyl--caprolactone. The (S)-isomer was enantioselectively polymerized to produce the polyester with >95% ee.98 Quantitative reactivity of 4-substituted -caprolactone using lipase CA as catalyst was analyzed.99 The polymerization rate decreased by a factor of 2 upon the introduction of a methyl substitutent at the 4-position. Furthermore, 4-ethyl--caprolactone polymerized five times slower than the 4-methyl--caprolactone. This reactivity difference is strongly related to the enan tioselectivity. Interestingly, lipase CA displayed S-selectivity for 4-methyl or ethyl--caprolactone, and the enantioselectivity was changed to the (R)-enantiomer in the case of 4-propyl--caprolactone. Lipase BC induced the enantioselective polymerization of 3-methyl-4-oxa 6-hexanolide (MOHEL).16 The initial reaction rate of the S-isomer was seven times
Scheme 8
220
Hiroshi Uyama
Scheme 9 larger than that of the R-isomer, indicating that the enantioselective polymerization of MOHEL took place through lipase catalysis. (S)-MOHEL was also polymerized by lipase PF, whereas no polymerization of the R-isomer took place with lipase PF. PPL catalyzed an enantioselective polymerization of bis(2,2,2 trichloroethyl) trans-3,4-epoxyadipate with 1,4-butanediol in diethyl ether to give a highly optically active polyester (Scheme 9).100 The molar ratio of the diester to the diol was adjusted to 2:1 to produce the (–) polymer with enantiomeric purity of >96%. The polymerization of racemic bis(2-chloroethyl) 2,5-dibromoadipate with excess of 1,6-hexanediol using lipase A catalyst produced optically active trimer and pentamer.101 The polycondensation of 1,4-cyclohexanedimethanol with fumarate esters using PPL catalyst afforded moderate diastereoselectivity for the cis/trans monocondensate and markedly increased diastereoselectivity for the dicondensate product.102 An optically active oligoester was enantioselectively prepared from racemic 10-hydroxyundecanoic acid using lipase CR catalyst.103 The resulting oligomer was enriched in the (S-enantiomer to a level of 60% ee and the residual monomer was recovered with a 33% ee favoring the antipode. Lactic acid was converted to the corresponding oligomer with DP up to 9 by lipase CA catalyst.104 HPLC analysis showed that the (R)-enantiomer possessed higher enzymatic reactivity. Optically active oligomers (DP < 6) were also synthesized from racemic -substituted- hydroxy esters using PPL catalyst.105 The enantioselectivity increased as a function of bulkiness of the monomer substituent. The enzymatic copolymerization of the racemic hydroxyacid esters with methyl 6-hydroxyhexanoate produced the optically active polyesters with molecular weight higher than 1 × 103 . Enzymatic enantioselective oligomerization of a symmetrical hydroxy diester, dimethyl hydroxyglutarate, produced a chiral oligomer (dimer or trimer) with 30–37% ee.106 Polyols such as glycerol and sugars were enzymatically regioselectively polymerized with dicarboxylic acid derivatives to form soluble polyesters. The lipase-catalyzed polycondensation of glycerol and divinyl esters produced polyesters having a secondary hydroxy group in the main chain.107 NMR analy sis of the polymer obtained from divinyl sebacate and glycerol using lipase CA
Enzymatic polymerization
221
Scheme 10 catalyst at 60 C in bulk showed that 1,3-diglyceride was a main unit and that a small amount of the branching unit (triglyceride) was contained.108 The regios electivity of the acylation between primary and secondary hydroxy groups was 74:26. In the polymerization at 45 C, the regioselectivity was perfectly controlled to give a linear polymer consisting exclusively of 1,3-glyceride units (Scheme 10). The polymerization of divinyl sebacate with 1,2,4-butanetriol or 1,2,6 hexanetriol at 60 C produced the polymer containing the branched unit. In the polymerization at lower temperature, the regioselectivity was perfectly controlled to give the linear polymer consisting of exclusively -disubstituted unit.109 The lipase origin and feed ratio of monomers greatly affected the microstructure of the polymer. The lipase MM-catalyzed polymerization of divinyl sebacate and glycerol produced a linear polymer consisting of 1,2- and 1,3-disubstituted units, whereas the 1,3-disubstituted and trisubstituted units were observed in the poly mer obtained using lipase PC catalyst. Interestingly, the highly branched polyester with branching factor >07 was formed by the lipase CA-catalyzed reaction of poly(azelaic anhydride) and triols such as glycerol. As a possible application of glycerol-based polyesters, new crosslinkable polyesters were synthesized by lipase CA-catalyzed polymerization of divinyl sebacate and glycerol in the presence of unsaturated higher fatty acids derived from renewable plant oils (Scheme 11).110111 The polymerization under reduced pressure improved the polymer yield and molecular weight. The curing of the polymer obtained using linoleic or linolenic acid proceeded by cobalt naphthenate catalyst or thermal treatment to give a crosslinked transparent film. Biodegrad ability of the obtained film was evaluated by biochemical oxygen demand (BOD) measurement in an activated sludge. The degradation gradually took place and the biodegradability reached 45% after 42 days, indicating the good biodegradability of this crosslinked film. Epoxide-containing polyesters were enzymatically synthesized via two routes using unsaturated fatty acids as starting substrate (Scheme 11).112 Lipase catalysis was used for both polycondensation and epoxidation of unsaturated fatty acid group. One route was synthesis of aliphatic polyesters containing an
222
Hiroshi Uyama
Scheme 11
unsaturated group in the side chain from divinyl sebacate, glycerol, and the unsat urated fatty acids, followed by an epoxidation of the unsaturated fatty acid moiety in the side chain of the resulting polymer. In another route, epoxidized fatty acids were prepared from the unsaturated fatty acids and hydrogen peroxide in the presence of lipase catalyst, and subsequently the epoxidized fatty acids were poly merized with divinyl sebacate and glycerol. The polymer structure was confirmed by NMR and IR, and for both routes, the high epoxidized ratio was achieved. Curing of the resulting polymers proceeded thermally, yielding transparent poly meric films with high gloss surface. Pencil scratch hardness of the present film improved, compared with that of the cured film obtained from the polyester hav ing an unsaturated fatty acid in the side chain. The obtained film showed good biodegradability, evaluated by BOD measurement in an activated sludge. For regioselective acylation of sugars, proteases were often used as catalyst.1 Polyesters having a sugar moiety in the main chain were synthesized via the protease catalysis. In the polycondensation of sucrose with bis(2,2,2-trifluoroethyl) adipate catalyzed by an alkaline protease from Bacillus sp. showing an esterase activity, the regioselective acylation of sucrose at the 6 and 1 -positions was claimed to yield the sucrose-containing polyester (Scheme 12).113 The reaction proceeded slowly; the molecular weight reached larger than 1 × 104 after 7 days.114
Enzymatic polymerization
223
Scheme 12 Two-step synthesis of sugar-containing polyesters by lipase CA catalyst was reported (Scheme 13).115 Lipase CA catalyzed the condensation of sucrose with an excess of divinyl adipate to produce sucrose 6,6 -O-divinyl adipate, which was reacted with , -alkylene diols by the same catalyst, yielding polyesters containing a sucrose unit in the main chain. This method conveniently affords
Scheme 13
224
Hiroshi Uyama
Scheme 14 the sugar-containing polyesters with relatively high molecular weight. Similarly, a trehalose-containing polyester was obtained from trehalose 6,6 -O-divinyl adipate through the catalysis of lipase CA. We first demonstrated that lipase CA produced the reduced sugar-containing polyesters regioselectively from divinyl sebacate and sorbitol, in which sorbitol was exclusively acylated at the 1 and 6-positions (Scheme 14).116 Mannitol and meso-erythritol were also regioselectively polymerized with divinyl sebacate. The enzymatic formation of the high molecular weight sorbitol-containing polyester was confirmed by combinatorial approach.80 The lipase CA-catalyzed polycondensation of adipic acid and sorbitol also took place in bulk.117 In the polymerization at 90 C, the molecular weight reached 1 × 104 ; however, the regioselectivity decreased (85%), probably due to the high temperature and/or the bulk condition. These data suggest that the divinyl ester is a suitable monomer for regioselective synthesis of sugar-containing polymers. The copolymerization of adipic acid, sorbitol, and 1,8-octanediol enhanced the molec ular weight of the sugar-containing polyesters. The melting and crystallization temperatures as well as the melting enthalpy decreased with increasing sorbitol content.118 This is attributed to the polyol units along the polyester chain which disrupt crystallinity. The biocompatibility of the sugar-containing polyester from adipic acid, sorbitol, and 1,8-octanediol was examined by using a mouse fibroblast 3T3 cell line in vitro.119 Polyester-sugar or polyester-polysaccharide conjugates were regioselec tively synthesized via enzyme catalysis. Lipase CA-catalyzed polymerization of -CL in the presence of alkyl glucopyranosides produced polyesters bearing a sugar at the polymer terminal (Scheme 15).120121 In the initiation step, the primary hydroxy group of the glucopyranoside was regioselectively acylated. Polysaccha rides also initiated the lipase-catalyzed polymerization of -CL.122 The enzymatic graft polymerization of -CL on hydroxyethyl cellulose produced cellulose-graft poly(-CL) with degree of substitution from 0.10 to 0.32. Reactive polyesters were enzymatically synthesized. Lipase catalysis chemoselectively induced the ring-opening polymerization of a lactone having exo-methylene group to produce a polyester having the reactive exo-methylene group in the main chain (Scheme 16).123124 This is in contrast to the anionic
Enzymatic polymerization
225
Scheme 15
Scheme 16 polymerization; the vinyl polymerization of this monomer took place by a con ventional anionic initiator or catalyst. Terminal-functionalized polymers such as macromonomers and telechelics are very important as prepolymer for construction of functional materials. Singlestep functionalization of polymer terminal was achieved via lipase catalysis. Alco hols could initiate the ring-opening polymerization of lactones by lipase catalyst. The lipase CA-catalyzed polymerization of DDL in the presence of 2-hydroxyethyl methacrylate gave the methacryl-type polyester macromonomer, in which 2 hydroxyethyl methacrylate acted as initiator to introduce the methacryloyl group quantitatively at the polymer terminal (“initiator method”).125 This methodology was expanded to the synthesis of -alkenyl- and alkynyl-type macromonomers by using 5-hexen-1-ol and 5-hexyn-1-ol as initiator, respectively. A methacryl-type polyester macromonomer was synthesized by lipase PFcatalyzed polymerization of DDL using vinyl methacrylate as terminator (“termina tor method”), in which the vinyl ester terminator was present from the beginning of the reaction (Scheme 17).126 In using divinyl sebacate as terminator, the telechelic polyester having a carboxylic acid group at both ends was obtained.127 Various non-protected thiol compounds were used as initiator or terminator for the thiol end-functionalization of poly(-CL).128 Long-chain unsaturated , -dicarboxylic acid methyl esters and their epox idized derivatives were polymerized with 1,3-propanediol or 1,4-butanediol in the presence of lipase CA catalyst to produce reactive polyesters.129 The molec ular weight of the polymer from 1,4-butanediol was higher than that from
226
Hiroshi Uyama
Scheme 17 1,3-propanediol. All the resulting polymers possessed melting point in the range from 23 to 75 C. Unsaturated ester oligomers were synthesized by lipase-catalyzed polymer ization of diesters of fumaric acid and 1,4-butanediol.130 Isomerization of the dou ble bond did not occur to give all-trans oligomers showing crystallinity, whereas the industrial unsaturated polyester having a mixture of cis and trans double bonds is amorphous.131 The enzymatic polymerization of bis(2-chloroethyl) fumarate with xylylene glycol produced the unsaturated oligoester containing aromaticity in its backbone.132 Chemoenzymatic synthesis of alkyds (oil-based polyester resins) was demonstrated.133 PPL-catalyzed transesterification of triglycerides with an excess of 1,4-cyclohexanedimethanol mainly produced 2-monoglycerides, followed by thermal polymerization with phthalic anhydride to give the alkyd resins with molecular weight of several thousands. The reaction of the enzymatically obtained alcoholysis product with toluene diisocyanate produced the alkyd-urethane. The enzymatic polymerization of 12-hydroxydodecanoic acid in the presence of 11-methacryloylaminoundecanoic acid conveniently produced the methacrylamide-type polyester macromonomer.134135 Lipases CA and CC were active for the macromonomer synthesis. Enzymatic selective monosubstitution of a hydroxy-functional dendrimer was demonstrated.136 Lipase CA-catalyzed poly merization of -CL in the presence of the first generation dendrimer gave the poly(-CL)-monosubstituted dendrimer.
Enzymatic polymerization
227
Poly(ethylene glycol) (PEG)-containing polyesters were synthesized via lipase catalysis.137 The lipase CA-catalyzed polymerization of dimethyl 5 hydroxyisophthalate, , -alkylene glycol, and PEG having a hydroxy group at both ends gave the PEG-containing polyesters. The chemoenzymatic synthesis of amphiphilic polyesters was examined by the lipase CA-catalyzed polymerization of dimethyl 5-hydroxyisophthalate and PEG, followed by the coupling with alkyl bromide.138 The supramolecular organization of the resulting polymeric nanomi celles in an aqueous medium was confirmed by NMR and light scattering. In vivo studies by encapsulating anti-inflammatory agents in the polymeric nanomicelles and by applying topically resulted in significant reduction in inflammation. The reduction ratio in inflammation using the polymeric nanomicelles was about 60%. A novel chemoenzymatic route to polyester polyurethanes was developed without employing highly toxic isocyanate intermediates.139 First, diurethane diols were prepared from cyclic carbonates and primary diamines, which were subse quently polymerized with dicarboxylic acids and glycols by using lipase CA as catalyst, yielding the polyurethanes under mild reaction conditions. Fluorinated polyesters were synthesized by the enzymatic polymerization of divinyl adipate with fluorinated diols. Lipase CA was effective in producing the fluorinated polyesters.140 The highest molecular weight (52×103 ) was achieved by the polymerization using 3,3,4,4,5,5,6,6-octafluorooctan-1,8-diol in bulk. A silicon oligomer was synthesized by the polycondensation of diethoxydimethylsilane using lipid-coated lipase from Rhizopus delemar as catalyst in isooctane containing a small amount of water.141 The polymerization is proposed to be initiated at the OH head group of the coating lipid. Block copolymers were synthesized by a combination of lipase-catalyzed polymerization and atom transfer radical polymerization (ATRP).142143 At first, the polymerization of 10-hydroxydecanoic acid was carried out by using lipase CA as catalyst. The terminal hydroxy group was modified by the reaction with -bromopropionyl bromide, followed by ATRP of styrene using CuCl/2,2 bipyridine as catalyst system to give the polyester–polystyrene block copolymer. Trichloromethyl-terminated poly(-CL), which was synthesized by lipase CAcatalyzed polymerization with 2,2,2-trichloroethanol initiator, was used as initiator for ATRP of styrene. Enzyme catalysts are useful for polymer recycling. We first proposed a new concept of chemical recycling of polymers using lipase catalyst.144 Aliphatic polyesters were subjected to hydrolytic degradation by lipase catalyst in organic solvents. The lipase CA-catalyzed degradation of poly(-CL) with molecular weight 4 × 104 readily took place in toluene at 60 C to give oligomers with molec ular weight less than 500. The degradation behavior catalyzed by lipase was quite different than an acid-catalyzed degradation (random bond cleavage of polymer). After the removal of the solvent from the reaction mixture, the residual oligomer was polymerized in the presence of the same catalyst of lipase. These data pro vide a basic concept that the degradation–polymerization could be controlled by
228
Hiroshi Uyama
Scheme 18 the presence or absence of the solvent, providing a new methodology of plastics recycling (Scheme 18).
3. ENZYMATIC SYNTHESIS OF PHENOLIC POLYMERS Phenol–formaldehyde resins using prepolymers such as novolaks and resols are widely used in industrial fields. These resins show excellent toughness and thermalresistant properties, but the general concern over the toxicity of formaldehyde has resulted in limitations on their preparation and use. Therefore, an alternative process for the synthesis of phenolic polymers avoiding the use of formaldehyde is strongly desired. The enzymatic synthesis of phenolic polymers has been extensively investi gated. In living cells, various oxidoreductases play an important role in maintaining the metabolism of living systems. So far, several oxidoreductases, peroxidase, laccase, polyphenol oxidase (tyrosinase), etc., have been reported to catalyze oxidative polymerization of phenol derivatives, and among them, peroxidase is most often used.145−149 Peroxidase is an enzyme whose catalysis is an oxidation of a donor to an oxidized donor by the action of hydrogen peroxide, liberating two water molecules. Horseradish peroxidase (HRP) is a single-chain -type hemoprotein that catalyzes the decomposition of hydrogen peroxide at the expense of aromatic proton donors. 3.1. Enzymatic oxidative polymerization of phenols Peroxidase-catalyzed oxidative coupling of phenols proceeds fast in aqueous solu tions, giving rise to the formation of oligomeric compounds. However, the resulting oligomers have not been well characterized, since most of them show low solubil ity towards common organic solvents and water. In 1987, enzymatic synthesis of a new class of phenolic polymers was first reported.150 An oxidative polymerization of p-phenylphenol using HRP as catalyst was carried out in a mixture of water and water-miscible solvents such as 1,4-dioxane, acetone, DMF, and methyl for mate. The polymerization proceeded at room temperature and during this process, powdery polymers were precipitated. The reaction medium composition greatly affected the molecular weight, and the polymer with the highest molecular weight (26 × 104 ) was obtained in 85% 1,4-dioxane. Afterwards, peroxidase-catalyzed
Enzymatic polymerization
229
Scheme 19
oxidative polymerization of various phenol derivatives has been examined to pro duce functional and useful phenolic polymers.148149 Phenol, the simplest and industrially more important phenolic compound, is a multifunctional monomer when considered as a substrate for oxidative poly merizations, and hence conventional polymerization catalysts afford insoluble macromolecular products with non-controlled structure.151 Phenol was subjected to oxidative polymerization using HRP or soybean peroxidase (SBP) as catalyst in an aqueous-dioxane mixture, yielding a polymer consisting of phenylene and oxyphenylene units (Scheme 19). The polymer showed low solubility; it was partly soluble in DMF and dimethyl sulfoxide (DMSO) and insoluble in other common organic solvents.152153 When a water-soluble alcohol and a buffer were used as the solvent, the prod uct showed an improved solubility toward DMF and DMSO.154155 For instance, in an equivolume mixture of methanol and phosphate buffer (pH 7), a DMF-soluble polymer was obtained in good yields. The solubility of the resulting polymer strongly depended on the buffer pH and content of the mixed solvent. The molecu lar weight of the polymer was in the range of several thousand Daltons. The result ing phenolic polymer showed relatively high thermal stability and no clear glass transition temperature (Tg ) was observed below 300 C. Molecular weight control of the polymer was achieved by the copolymerization with 2,4-dimethylphenol to give a soluble oligomer with a molecular weight of 500.156 The control of the polymer structure was achieved by solvent engineering. The ratio of phenylene and oxyphenylene units was strongly dependent on the sol vent composition. In the HRP-catalyzed polymerization of phenol in a mixture of methanol and buffer, the oxyphenylene unit increased by increasing the methanol content, while the buffer pH scarcely influenced the polymer structure.154155 The proposed polymerization mechanism is shown in Scheme 20. A phenoxy free radical is first formed, then two molecules of the radical species dimerize via coupling. Since peroxidase often does not recognize larger molecules, a radical transfer reaction between a monomeric phenoxy radical and a phenolic polymer takes place to give the polymeric radical species. In the propagation step, such propagating radicals are subjected to oxidative coupling, producing polymers of higher molecular weight. The polymerization outcome depended on the monomer structure as well as on the enzyme origin. For instance, using HRP and p-n-alkylphenols, the
230
Hiroshi Uyama
Scheme 20
polymer yield increased by increasing the chain length of the alkyl group from one to five in an aqueous 1,4-dioxane as medium, while the amount of poly mer obtained with hexyl- or heptylphenol was almost the same as that obtained with the pentyl derivative.157158 The polymer formation by HRP catalysis was observed with all the cresol isomers investigated.159 A polymer was obtained in high yields using p-i-propylphenol, whereas o- and m-isomers were not poly merized under similar reaction conditions. Poly(p-n-alkylphenol)s prepared in the aqueous 1,4-dioxane showed low solubility toward organic solvents. On the other hand, soluble oligomers with molecular weights lower than 1000 were formed
Enzymatic polymerization
231
from p-ethylphenol using aqueous DMF as a solvent.160 Poly(p-t-butylphenol) enzymatically synthesized in water–dioxane showed Tg and melting point (Tm ) at 182 and 244 C, respectively. When the HRP-catalyzed polymerization of p-substituted phenols was con ducted in an equivolume mixture of organic solvent and pH 7 phosphate buffer, the regioselectivity was influenced by the monomer substituents and the solvent nature.161162 The hydrophobic nature of the monomer substituent and of the organic solvent (evaluated as and log P, respectively) strongly affected the polymer structure. A significative first-order correlation between these parameters and the polymeric structure was observed. The structures of these macromolecules covered a wide range of the unit ratio between a phenylene and an oxyphenylene (from 94/6 to 4/96), indicating that the regioselectivity can be controlled by varying the solvent and substituent nature, yielding poly(phenylene) or poly(oxyphenylene). The peroxidase-catalyzed polymerization of m-alkyl substituted phenols in aqueous methanol produced soluble phenolic polymers.163 The mixed ratio of buffer and methanol greatly affected the yields and the molecular weight of the polymer. The enzyme source greatly affected the polymerization pattern of m-substituted monomers. Using SBP catalyst, the polymer yield increased as a function of the bulkiness of the substituent, whereas the opposite tendency was observed when HRP was the catalyst. In order to better understand the solvent effect on the enzymatic polymeriza tion of phenols, the self-association of m-cresol in water–organic solvent mixtures was examined.164 Clustering of m-cresol in these solvents was observed by UV absorption spectroscopy and mass spectrometry for clusters. The pattern of the clustering formation in the solvents of different composition was significantly related to the results of the enzymatic polymerization in these mixed solvents, suggesting that the clustering of the phenol monomer in water–organic solvent mixtures affords the phenolic polymer more efficiently than that in the absence of the organic solvent. Fluorinated phenols, 3- and 4-fluorophenols, and 2,6-difluorophenol, were subjected to peroxidase-catalyzed polymerization in an aqueous organic solvent, yielding fluorine-containing polymers.165 Elimination of fluorine atom partly took place during the polymerization to give polymers with complicated structures. Various bisphenol derivatives were also polymerized by peroxidase under selected reaction conditions, yielding soluble phenolic polymers. Bisphenol-A was polymerized by peroxidase catalyst to give a polymer soluble in acetone, DMF, DMSO, and methanol.166 The polymer was produced in higher yields using SBP as a catalyst. This polymer showed a molecular weight of 4 × 103 and a Tg at 154 C. The HRP-catalyzed polymerization of 4,4 -biphenol produced a polymer showing high thermal stability.167 Peroxidase also induced the polymerization of an industrial product, bisphenol-F, consisting of 2,2 -, 2,4 -, and 4,4 -dihydroxydiphenylmethanes.168 Under the selected reaction conditions, the quantitative formation of a soluble
232
Hiroshi Uyama
phenolic polymer from bisphenol-F was achieved. Among the isomers, 2,4 - and 4,4 -dihydroxydiphenylmethanes were polymerized to give the corresponding poly mers in high yields, whereas no polymerization of the 2,2 -isomer occurred, sug gesting the frequent occurrence of a radical transfer reaction between a phenoxy radical of the enzymatically polymerizable monomer and a phenol moiety of the enzymatically non-polymerizable monomer. In the case of 4,4 -dihydroxyphenyl monomers, the bridge structure enormously affected the polymerization behaviors and the thermal properties of the resulting polymers. Laccase is a protein containing copper as its active site and uses oxygen as an oxidizing agent. An oxidative polymerization of phenol and its derivatives was performed using laccase as catalyst without hydrogen peroxide in aqueous organic solvents at room temperature under air.169 Laccase derived from Pycnoporus coc cineus (PCL) efficiently induced the polymerization to produce phenolic polymers consisting of a mixture of phenylene and oxyphenylene units. The unit ratio of the polymer could be precisely controlled by selection of the solvent nature and the monomer substituent. A bi-enzymatic system (glucose oxidase + HRP) was also used to catalyze the synthesis of phenolic polymers. The polymerization of phenol, albeit in mod erate yield, was accomplished in the presence of glucose avoiding the addition of hydrogen peroxide (Scheme 21),170 which was formed in situ by the oxidation of glucose catalyzed by glucose oxidase. As described above, the enzymatic polymerization of phenols was often carried out in a mixture of a water-miscible organic solvent and a buffer. By adding 2,6-di-O-methyl--cyclodextrin (DM--CD), the enzymatic polymerization of water-insoluble m-substituted phenols proceeded in buffer.171 The water-soluble complex of the monomer and DM--CD was formed and was polymerized by HRP to give a soluble polymer. In the case of phenol, the polymerization took place in the presence of 2,6-di-O-methyl--cyclodextrin (DM--CD) in a buffer.172 Only a catalytic amount of DM--CD was necessary to induce the polymerization efficiently. Coniferyl alcohol was oxidatively polymerized in the presence of -CD in an aqueous solution.173
Scheme 21
Enzymatic polymerization
233
PEG was found to be a template in the oxidative polymerization of phenol in water.174175 The presence of the PEG template in an aqueous medium greatly improved the regioselectivity of the polymerization, yielding a phenol polymer with the phenylene unit content higher than 90%. During the reaction, the poly mer was produced in high yields as precipitates in complexing with PEG. The molecular weight of PEG strongly affected the polymer yield. The unit molar ratio of polyphone and PEG was ca. 1:1. The FT-IR and DSC analyses exhibited the formation of the miscible complex of the phenolic polymer and PEG by hydrogenbonding interaction. PEG monododecyl ether, a commercially available non-ionic surfactant, was also a good template for the polymerization of phenol in water.176 Advantages for enzymatic synthesis of phenolic polymers are summarized as follows145 : (i) the polymerization of phenols proceeds under mild reaction conditions without use of toxic reagents (environmentally benign process); (ii) phenol monomers having various substituents are polymerized to give a new class of functional polyaromatics; (iii) the structure and solubility of the polymer can be controlled by changing the reaction conditions; (iv) the procedures of the polymerization as well as the polymer isolation are very convenient. 3.2. Enzymatic synthesis of functional phenolic polymers Poly(2,6-dimethyl-1,4-oxyphenylene) (poly(phenylene oxide), PPO) is a mate rial widely used as high-performance engineering plastics, thanks to its excel lent chemical and physical properties, e.g., a high Tg (ca. 210 C) and mechan ically tough property.151 PPO was first prepared from 2,6-dimethylphenol monomer using a copper/amine catalyst system. 2,6-Dimethylphenol was also polymerized via HRP catalysis to give a polymer exclusively consisting of 1,4-oxyphenylene unit,177 while small amounts of Mannich-base and 3,5,3 ,5 tetramethyl-4,4 -diphenoquinone units are always contained in the chemically pre pared PPO. Laccase (PCL) as well as peroxidases (HRP and SBP) induced a new type of oxidative polymerization of the 4-hydroxybenzoic acid derivatives, 3,5-dimethoxy 4-hydroxybenzoic acid (syringic acid) and 3,5-dimethyl-4-hydroxybenzoic acid. The polymerization involved elimination of carbon dioxide and hydrogen from the monomer to give PPO derivatives with molecular weight up to 18 × 104 (Scheme 22).178179 -Hydroxy- -hydroxyoligo(1,4-oxyphenylene)s were formed in the HRPcatalyzed oxidative polymerization of 4,4 -oxybisphenol in an aqueous methanol.180 During the reaction, hydroquinone was formed. Scheme 23 shows the postulated mechanism of the trimer formation; the redistribution and/or rear rangement of the quinone-ketal intermediate takes place, involving the elimination of hydroquinone. New positive-type photoresist systems based on enzymatically synthesized phenolic polymers were developed.181 The polymers from the bisphenol monomers
234
Hiroshi Uyama
Scheme 22
Scheme 23
exhibited high photosensitivity, comparable with a conventional cresol novolak. Furthermore, the present photoresist showed excellent etching resistance. Nano scale polymer patterning was reported to be fabricated by the enzymatic oxidative polymerization with dip-pen nanolithography technique.182
Enzymatic polymerization
235
Scheme 24
HRP catalysis induced a chemoselective polymerization of a phenol deriva tive having a methacryloyl group.183 Only the phenol moiety was polymerized to give a polymer having the methacryloyl group in the side chain. The resulting polymer was readily cured thermally and photochemically (Scheme 24). A phe nol with an acetylenic substituent in the meta position was also chemoselectively polymerized by HRP to give a polymer bearing acetylenic groups (Scheme 25).184 For comparison, the reaction of the monomer using a copper/amine catalyst, a con ventional catalyst for oxidative coupling, was performed, producing a diacetylene derivative exclusively. The resulting polymer was converted to a carbon polymer
Scheme 25
236
Hiroshi Uyama
in much higher yields than enzymatically synthesized poly(m-cresol). Therefore, it might have a large potential as precursor of functional carbon materials. Hydroquinone mono-PEG ether was polymerized by HRP in aqueous 1,4 dioxane.185 High ionic conductivities (4 × 10−5 S cm−1 ) were found in the film con sisting of the lithiated phenolic polymer and PEG. A thiol-containing polymer was synthesized by peroxidase-catalyzed copolymerization of p-hydroxythiophenol and p-ethylphenol in reverse micelles.186 CdS nanoparticles were attached to the copolymer to give polymer-CdS nanocomposites. The reverse micellar system was also effective for the enzymatic synthesis of a poly(2-naphthol) consisting of quinonoid structure.187 It showed a fluorescence characteristic of the naphthol chromophore. A novel photoactive azopolymer, poly(4-phenylazophenol), was synthesized using HRP catalyst.188 A reversible trans to cis photoisomerization of the azobenzene group with long relaxation time was observed. Phenolic copolymers containing fluorophores (fluoroscein and calcein) were synthesized by SBP catalysis and used as array-based metal-ion sensor.189 Selec tivity and sensitivity for metal ions could be controlled by changing the polymer components. Combinatorial approach was made for efficient screening of specific sensing of the metals. A natural phenolglucoside, 4-hydroxyphenyl -d-glucopyranoside (arbutin), was subjected to regioselective oxidative polymerization using a peroxidase catalyst in a buffer solution, yielding the water-soluble polymer consisting of 2,6-phenylene units, in turn converted to poly(hydroquinone) by acidic deglycosy lation (Scheme 26).190 The resulting polymer was used for a glucose sensor exploit ing its good redox properties.191 Another route for the chemoenzymatic synthesis of poly(hydroquinone) was the SBP-catalyzed polymerization of 4-hydroxyphenyl benzoate, followed by alkaline hydrolysis.192 A polynucleoside with an unnatural polymeric backbone was syn thesized by SBP-catalyzed oxidative polymerization of thymidine 5 -p hydroxyphenylacetate.193 Chemoenzymatic synthesis of a new class of poly(amino acid), poly(tyrosine) containing no peptide bonds, was achieved by the peroxidasecatalyzed oxidative polymerization of tyrosine ethyl esters, followed by alkaline hydrolysis.194 Amphiphile higher alkyl ester derivatives were also polymerized in
Scheme 26
Enzymatic polymerization
237
micellar solution to give macromolecules showing surface activity at the air–water interface.195196 The molecular weight of the enzymatically synthesized phenolic polymers was in the range of several thousands. Ultrahigh molecular weight polymers were synthesized by the oxidative coupling of the phenolic polymers using Fe salen as a catalyst, which can be regarded as a mimic of peroxidases.197198 The oxidative coupling of the precursor polymers with molecular weight of several thousands produced DMF-soluble polymers with Mw > 106 quantitatively, where a crosslinking is suppressed by selection of reaction conditions. The intermolecular coupling of the polymer efficiently proceeded at the initial stage of the reaction, leading to the rapid increase of the molecular weight. On the other hand, the molecular weight of the polymer obtained from the monomer was much lower. This result provides a new synthetic method of ultrahigh molecular weight polymeric materials via oxidative couplings. Furthermore, the oxidative coupling behaviors between the precursor polymers and the corresponding phenolic monomers were examined. The structure of the starting monomers had a great influence on the oxidative coupling of the precursor polymers. The polymers from o- and msubstituted monophenols were subjected to efficient oxidative coupling to give high molecular weight polymers quantitatively; on the other hand, the oxidative coupling of the polymers from p-substituted monophenols scarcely proceeded. Polymers from bisphenol and triphenol were also oxidatively polymerized, leading to the quantitative production of the high molecular weight polymers under the appropriate reaction conditions. We prepared a phenol-containing hyaluronan derivative, which was inter molecularly coupled by HRP to yield a crosslinked hydrogel (Scheme 27).199 The sequential injection of this hyaluronan derivative and peroxidase formed
Scheme 27
238
Hiroshi Uyama
biodegradable hydrogels in vivo, offering high potential as a promising biomaterial for drug delivery and tissue engineering. Morphology of the enzymatically synthesized phenolic polymers was con trolled under the selected reaction conditions. Monodisperse polymer particles in the sub-micron range were produced by HRP-catalyzed dispersion polymerization of phenol in 1,4-dioxane-phosphate buffer (3:2 v/v) using poly(vinyl methyl ether) as stabilizer.200−202 The particle size could be controlled by the stabilizer con centration and solvent composition. Thermal treatment of these particles afforded uniform carbon particles. The particles could be obtained from various phenol monomers such as m-cresol and p-phenylphenol. Blends of enzymatically synthesized poly(bisphenol-A) and poly(p-t butylphenol) with poly(-CL) were examined.203 FT-IR analysis showed the expected strong intermolecular hydrogen-bonding interaction between the phe nolic polymer with poly(-CL). A single Tg was observed for the blend, and the value increased as a function of the polymer content, indicating their good miscibility in the amorphous state. In the blend of enzymatically synthesized poly(4,4 -oxybisphenol) with poly(-CL), both polymers were miscible in the amorphous phase also.204 The crystallinity of poly(-CL) decreased by poly(4,4 oxybisphenol). 3.3. Artificial urushi As “japan” implies the meaning of “a lacquer or varnish giving a hard, glossy finish” and/or “objects decorated and lacquered in the Japanese style,”205 urushi wares are regarded as one of the most typical symbols of Japanese art. Oriental lacquer (urushi) of Japan and China is a natural resinous sap of the Rhus vernicifera tree. Urushi coating is hard enough to give a brilliant polish and is highly durable (for more than a thousand years under appropriate conditions) and solvent-resistant in comparison with the synthetic coatings. Majima’s pioneering work in the early days of the 20th century revealed that main components of urushi are “urushiols,” whose structure is a catechol derivative with unsaturated hydrocarbon chains consisting of a mixture of monoenes, dienes, and trienes at 3- or 4-position of catechol.206 Crosslinking of the urushiol is supposed to be accomplished mainly by a laccase-catalyzed oxidative coupling of the phenol moiety of the urushiol and a subsequent autoxidation of unsaturated alkyl chains in air. Urushi can be regarded as the only example of practical natural paints utilizing in vitro enzymatic catalysis for hardening. Urushi is curable under air at room temperature without organic solvents, and hence, urushi seems very desirable for coating system from the environmental standpoint. However, modeling study of urushi has been limited, mainly due to the difficulty in preparation of the urushiol. The reason why synthesis of natural urushiols involves multistep, tedious procedures is that the reactive unsaturated group cannot be directly introduced on the catechol moiety; protection and deprotection of the catechol moiety are
Enzymatic polymerization
239
Scheme 28 required. We designed new urushiol analogues, in which the unsaturated group is connected with the phenolic group through an ester linkage.207−210 The analogs were synthesized by a lipase-catalyzed esterification of phenols having a pri mary alcohol with unsaturated fatty acids of different numbers of double bonds (Scheme 28). The primary aliphatic hydroxy group in the phenolic precursors was regioselectively acylated by lipase BS to produce oily urushiol analogs. In this new approach, the analogs were obtained by one or two reaction steps via facile procedures from commercially available reagents. Furthermore, it is to be noted that the urushiol analogs showed no dermatitis-causing activity. The curing was performed using laccase (PCL) as catalyst in the presence of acetone powder, a third component of the sap acting as emulsifier of oily urushiol and aqueous laccase solution. The urushiol analogs were crosslinked under mild reaction conditions without use of organic solvents to produce the polymeric film with high gloss surface and good elastic properties. Cardanol, a main component obtained by thermal treatment of cashew nut shell liquid (CNSL), is a phenol derivative having mainly the meta substituent of a C15 unsaturated hydrocarbon chain with one to three double bonds as the major. Since CNSL is nearly one-third of the total nut weight, a great amount of CNSL is obtained as byproducts from mechanical processes for the edible use of the cashew kernel. Only a small part of cardanol obtained in the production of cashew kernel is used in industrial fields, though it has various potential industrial utilizations such as resins, friction-lining materials, and surface coatings. Therefore, development of new applications for cardanol is very attractive. A new crosslinkable polymer was synthesized by the SBP-catalyzed poly merization of cardanol.211 When HRP was used as catalyst for the cardanol polymerization, the reaction took place in the presence of a redox mediator (phe nothiazine derivative) to give the polymer.212 Fe-salen efficiently catalyzed the polymerization of cardanol in organic solvents (Scheme 29).213214 The polymeriza tion proceeded in 1,4-dioxane to give the soluble polymer with molecular weight of several thousands in good yields. The curing of the polymer took place in the presence of cobalt naphthenate catalyst at room temperature or thermal treat ment (150 C for 30 min) to form yellowish transparent films (“artificial urushi”
240
Hiroshi Uyama
Scheme 29 in a broad sense). The resulting crosslinked film exhibited good elastic properties comparable with natural urushi. FT-IR monitoring of the curing showed that the crosslinking mechanism is similar to that of the oil autoxidation. 3.4. Enzymatic synthesis and biological properties of flavonoid polymers Bioactive polyphenols are present in a variety of plants and used as important com ponents of human and animal diets.215 Flavonoids are a broad class of low molec ular weight secondary plant polyphenolics, which are benzo--pyrone derivatives consisting of phenolic and pyrane rings. Their biological and pharmacological effects including antioxidant, antimutagenic, anticarcinogenic, antiviral, and anti inflammatory properties have been demonstrated in numerous human, animal, and in vitro studies. These properties are potentially beneficial in preventing diseases and protecting the stability of the genome. Many of these activities have been related to their antioxidant actions. Green tea is derived from Camellia sinensis, an evergreen shrub of the Theaceae family. Most of the polyphenols in green tea are flavanols, commonly known as catechins; the major catechins in green tea are (+)-catechin, (−)-epicatechin, (−)-epigallocatechin, (−)-epicatechin gallate, and (−)-epigallocatechin gallate (EGCG). Numerous biological activities have been reported for green tea and its contents; among them, the preventive effects against cancer are most notable.216 We have designed not only polymerized flavonoids but also flavonoid con jugates of various polyamines, in consideration of extension of the amplification of physiological properties of the flavonoids. Polymeric flavonoids were synthesized by the enzymatic oxidative coupling.217 The HRP-catalyzed polymerization of (+)-catechin was carried out in an equivolume mixture of 1,4-dioxane and buffer (pH 7) to give the polymer with molecular weight of 30 × 103 in 30% yield.218 Using methanol as co-solvent improved the polymer yield and molecular weight.219 In the polymerization of
Enzymatic polymerization
241
catechin by using laccase derived from Myceliophthora (ML) as catalyst, the reaction conditions were examined in detail.220 A mixture of acetone and acetate buffer (pH 5) was suitable for the efficient synthesis of soluble poly(catechin) with high molecular weight. The mixed ratio of acetone greatly affected the yield, molecular weight, and solubility of the polymer. The polymer synthesized in 20% acetone showed low solubility toward DMF, whereas the polymer obtained in an acetone content less than 5% was completely soluble in DMF. In the UV–vis spectrum of poly(catechin) in methanol, a broad peak centered at 370 nm was observed. In alkaline solution, this peak was red-shifted and the peak intensity became larger than that in methanol. In the ESR spectrum of the enzymatically synthesized poly(catechin), a singlet peak at g = 1982 was detected, whereas the catechin monomer possessed no ESR peak. Superoxide anion scavenging activity of the enzymatically synthesized poly(catechin) was evaluated. Poly(catechin), synthesized by HRP catalyst, greatly scavenged superoxide anion in a concentration-dependent manner, and almost completely scavenged at 200 M of a catechin unit concentration.218 The laccase catalyzed synthesized poly(catechin) also showed excellent antioxidant property.219 Catechin showed pro-oxidant property in concentrations lower than 300 M. These results demonstrated that the enzymatically synthesized poly(catechin) possessed much higher potential for superoxide anion scavenging, compared with intact catechin. EGCG is a major ingredient of green tea possessing powerful antioxi dant activity and cancer-chemopreventive activity due to actions of radical scav enging, enzyme inhibition, and metal chelation. The polymer obtained by the laccase-catalyzed oxidative coupling of EGCG showed much higher superoxide anion scavenging activity than the EGCG monomer and enzymatically synthesized poly(catechin).221 Xanthine oxidase (XO) is not only an important biological source of ROS but also the enzyme responsible for the formation of uric acid associated with gout leading to painful inflammation in the joints. The XO inhibition effect by the enzy matically synthesized poly(catechin) increased as an increasing concentration of catechin units, while the monomeric catechin showed almost negligible inhibition effect in the same concentration range.219220 This markedly amplified XO inhibi tion activity of poly(catechin) was considered to be due to effective multivalent interaction between XO and the condensed catechin units in the poly(catechin). PolyEGCG also showed excellent inhibition effect for XO. The XO inhi bition effect of EGCG monomer was quite low with inhibition less than about 5% over a range of tested concentrations. In contrast, polyEGCG showed greatly amplified XO inhibition effect in a concentration-dependent manner.221 Moreover, the inhibition of polyEGCG was higher than that of allopurinol, the frequently used commercial inhibitor for gout treatment. Thus, polyEGCG is expected as one of the leading candidates of therapeutic molecules against various diseases
242
Hiroshi Uyama
induced by free radicals and/or enzymes including gout. Kinetic analysis showed that polyEGCG is an uncompetitive inhibitor of XO. Mutans streptococci are the major pathogenic organisms of dental caries in humans. The pathogenicity is closely related to production of extracellular, water-insoluble glucans from sucrose by glucosyltransferase and acid release from various fermentable sugars. Poly(catechin) obtained by HRP catalyst in a phosphate buffer (pH 6) markedly inhibited glucosyltransferase from Streptococcus sorbrinus 6715,222 whereas the inhibitory effect of catechin for this enzyme was very low. Rutin is one of the most commonly found flavonol glycosides identified as vitamin P with quercetin and hesperidin. An oxidative polymerization of rutin using ML as catalyst was examined in a mixture of methanol and buffer to produce a flavonoid polymer.223 Under selected conditions, the polymer with molecular weight of several thousands was obtained in good yields. The resulting polymer was readily soluble in water, DMF, and DMSO, although rutin monomer showed very low water solubility. UV measurement showed that the polymer had broad transition peaks around 255 and 350 nm in water, which were red-shifted in an alkaline solution. ESR measurement showed the presence of a radical in the polymer. Quercetin and kaempferol were polymerized by laccase and tyrosinase, and the morphology of the resulting aggregate was examined.224 The antioxidant activity of rutin was greatly amplified by the laccase catalyzed oxidative coupling; poly(rutin) also showed much higher scavenging capacity toward superoxide anion than rutin.223 We examined the protection effects of rutin and poly(rutin) against endothelial cell damage caused by 2,2 -azobis(2 amidinopropane)dihydrochloride. Poly(rutin) enhanced cell viability with higher protection effects against the oxidative damage than that of the rutin monomer at low concentration. In particular, in high concentration,