Issuu on Google+


Asymmetric Organic Synthesis with Enzymes

Edited by Vicente Gotor, Ignacio Alfonso, and Eduardo GarcĂ­a-Urdiales


Asymmetric Organic Synthesis with Enzymes

Edited by Vicente Gotor, Ignacio Alfonso, and Eduardo Garc覺織a-Urdiales


Related Titles Eames, J., Todd, M. (eds.)

Cao, L.

Separation of Enantiomers

Carrier-bound Immobilized Enzymes

New Methods and Applications approx. 400 pages Hardcover ISBN: 978-3-527-30502-5

Aehle, W. (ed.)

Enzymes in Industry Production and Applications approx. 520 pages with approx. 120 figures 2007 Hardcover ISBN: 978-3-527-31689-2

Principles, Application and Design 578 pages with 248 figures and 67 tables 2005 Hardcover ISBN: 978-3-527-31232-0

Breslow, R. (ed.)

Artificial Enzymes 193 pages with 81 figures and 8 tables 2005 Hardcover ISBN: 978-3-527-31165-1

Reymond, J.-L. (ed.)

Enzyme Assays High-throughput Screening, Genetic Selection and Fingerprinting 386 pages with 162 figures and 12 tables 2006 Hardcover ISBN: 978-3-527-31095-1

Bisswanger, H.

Practical Enzymology 272 pages with 35 figures and 6 tables 2004 Softcover ISBN: 978-3-527-30444-8

Bommarius, A.S.,Riebel-Bommarius,B.R.

Bornscheuer, U.T., Kazlauskas, R.J.

Biocatalysis

Hydrolases in Organic Synthesis

Fundamentals and Applications

Regio- and Stereoselective Biotransformations

2004 Hardcover ISBN: 978-3-527-30344-1

368 pages with 255 figures and 23 tables 2006 Hardcover ISBN: 978-3-527-31029-6

634 pages with 251 figures and 78 tables


Asymmetric Organic Synthesis with Enzymes

Edited by Vicente Gotor, Ignacio Alfonso, and Eduardo GarcĂ­a-Urdiales


The Editors Prof. Vicente Gotor Dept.de Química Orgánica/Inorg Universidad de Oviedo Avenida Julián Clavería 6 33006 Oviedo Spain Prof. Ignacio Alfonso Dept.de Química Orgánica/Inorg Universidad Jaume I Campus del Riu Sec 12071 Castellón Spain Dr. Eduardo García-Urdiales EMBL - Biocomputing Meyerhofstr. 1 69117 Heidelberg Germany

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: applied for British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Bibliographic information published by the Deutsche Nationalbibliothek Die Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at <http://dnb.d-nb.de>. # 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Typesetting Thomson Digital, Noida, India Printing Strauss GmbH, Mörlenbach Binding Litges & Dopf GmbH, Heppenheim Printed in the Federal Republic of Germany Printed on acid-free paper ISBN: 978-3-527-31825-4


V

Contents Preface X1 List of Contributors

XIII

1

I

Methodology

1

Medium Engineering 3 Giacomo Carrea and Sergio Riva Introduction 3 Modulation of Enzyme Enantioselectivity by Medium Engineering 5 Selectivity Enhancement by Addition of Water-Miscible Organic Cosolvents 5 Selectivity Enhancement in Organic Media with Low Water Activity 7 Organic Solvent Systems 7 Enzyme Properties in Organic Solvents 8 Medium Engineering 9 Rationales 12 Modulation of Enzyme Selectivity: New Trends of Research 14 Ionic Liquids 14 Additives 16 Conclusions and Outlooks 17 References 17

1.1 1.2 1.2.1 1.2.2 1.2.2.1 1.2.2.2 1.2.2.3 1.2.3 1.2.4 1.2.4.1 1.2.4.2 1.3

2

2.1 2.2 2.3

Directed Evolution as a Means to Engineer Enantioselective Enzymes 21 Manfred T. Reetz Introduction 21 Molecular Biological Methods for Mutagenesis 23 High-throughput Screening Methods for Enantioselectivity

27


VI

Contents

2.4 2.4.1 2.4.2 2.4.3 2.4.4 2.4.5 2.4.6 2.4.7 2.4.8 2.4.9 2.4.10 2.4.11 2.4.12 2.4.13 2.5

3 3.1 3.2 3.2.1 3.2.2 3.2.3 3.3 3.3.1 3.3.1.1 3.3.1.2 3.3.1.3 3.3.1.4 3.3.2 3.3.2.1 3.3.2.2 3.3.2.3 3.3.2.4 3.3.2.5 3.3.2.6 3.3.2.7 3.3.2.8 3.4

Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution 28 Lipase from Pseudomonas aeruginosa (PAL) 28 Other Lipases 38 Esterases 38 Hydantoinases 39 Nitrilases 39 Epoxide Hydrolases 41 Phosphotriesterases 45 Aminotransferases 45 Aldolases 46 Cyclohexanone and Cyclopentanone Monooxygenases as Baeyer窶天illigerases and Sulfoxidation Catalysts 48 Monoamine Oxidases 54 Cytochrome P450 Enzymes 55 Other Enzymes 55 Conclusions and Perspectives 56 References 56 The Search for New Enzymes 65 Jean-Louis Reymond and Wolfgang Streit Introduction 65 Mechanism-based Enzyme Design 66 Catalytic Antibodies 66 Rational Design of New Catalysts on Enzyme and Protein Basis Synthetic Enzyme Models 71 Metagenomics 71 Construction of Metagenome-derived DNA Libraries 72 Selection of the Environment 72 Cloning Strategies 73 Screening and Detection Technologies 73 Major Problems that Need to be Addressed 74 The Genomes of Not Yet Cultured Microbes as Resources for Novel Genes 75 Polysaccharide Degrading/Modifying Enzymes 75 Lipolytic Biocatalysts 77 Vitamin Biosynthesis 77 Nitrilases, Nitrile Hydratases, and Amidases 78 Oxidoreductases/Dehydrogenases 79 Proteases 79 Glycerol Hydratases 79 Antibiotics and Pharmaceuticals 79 Conclusion 80 References 80

69


Contents

87

II

Synthetic Applications

4

Dynamic Kinetic Resolutions 89 Belén Martín-Matute and Jan-E. Bäckvall Introduction 89 Synthesis of Enantiomerically Pure Compounds 89 Kinetic Resolution (KR) and Dynamic Kinetic Resolution (DKR) 90 Enzymes in Organic Chemistry 91 Metal-Catalyzed Racemization 92 DKR of Allylic Acetates and Allylic Alcohols 93 DKR of sec-alcohols 94 DKR of Amines 98 Base-Catalyzed Racemization 98 DKR of Thioesters 99 DKR of Activated Esters 100 DKR of Oxazolones 100 DKR of Hydantoins 101 DKR of Acyloins 101 Acid-Catalyzed Racemization 101 Racemization through Continuous Reversible Formation–Cleavage of the Substrate 102 DKR of Cyanohydrins 103 DKR of Hemithioathetals 103 Racemization Catalyzed by Aldehydes 104 Enzyme-Catalyzed Racemization 106 Racemization through SN2 Displacement 106 Other Racemization Methods 107 DKR of 5-Hydroxy-2-(5H)-Furanones 107 DKR of Hemiaminals 107 DKR of 8-Amino-5,6,7,8-tetrahydroquinoline 108 Concluding Remarks 109 References 110

4.1 4.1.1 4.1.2 4.1.3 4.2 4.2.1 4.2.2 4.2.3 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 4.4 4.5 4.5.1 4.5.2 4.6 4.7 4.8 4.9 4.9.1 4.9.2 4.9.3 4.10

5 5.1 5.2 5.2.1 5.2.2 5.2.3 5.2.4

Deracemization and Enantioconvergent Processes 115 Nicholas J.Turner Introduction 115 Deracemization Processes 116 Cyclic Oxidation–Reduction Systems 116 Microbial Deracemization of Secondary Alcohols Using a Single Microorganism 122 Deracemization of Alcohols Using Two Enzyme/Microorganism Systems 124 Epoxides 126

VII


VIII

Contents

5.2.5 5.3 5.3.1 5.3.2 5.4

Carboxylic Acids 126 Enantioconvergent Processes 127 Epoxide Hydrolysis 128 Sulfatases 129 Conclusions and Future Prospects 130 References 130

6

Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides 133 Robert Chênevert, Pierre Morin, and Nicholas Pelchat Introduction and General Aspects 133 Scope 133 Reaction Conditions 133 Kinetic Resolution, Dynamic Kinetic Resolution, and Desymmetrization 134 Enantioconvergent Transformation 137 Enantioselective Biotransformations of Carboxylic Acid Derivatives 137 Ester Hydrolysis 137 Ester Alcoholysis 140 Esterification of Carboxylic Acids 140 Lactones 142 Anhydrides 143 Hydrolysis of Nitriles 144 Hydrolysis of Amides, Lactams, and Hydantoins 146 Enantioselective Enzymatic Transformations of Alcohols 150 Hydrolysis of Epoxides 157 Conclusion 162 References 163

6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6 6.2.7 6.3 6.4 6.5

7 7.1 7.2 7.3 7.4 7.4.1 7.4.2 7.4.3 7.4.4 7.5 7.6 7.7

Aminolysis and Ammonolysis of Carboxylic Acid Derivatives 171 Vicente Gotor-Fernández and Vicente Gotor Introduction 171 Mechanism of Enzymatic Ammonolysis and Aminolysis Reactions 172 Aminolysis and Ammonolysis of Carboxylic Acids 174 Aminolysis and Ammonolysis of Esters 176 Preparation of Nonchiral Amides 176 Resolution of Esters 178 Resolution of Amines, Diamines, and Aminoalcohols 180 Desymmetrization of Diesters 184 Kinetic Resolution of Secondary Amines 185 Toward the Synthesis of b-Aminoacid Derivatives 186 Summary and Outlook 188 References 189


Contents

8 8.1 8.2 8.2.1 8.2.2 8.2.3 8.2.4 8.2.5 8.2.6 8.3 8.3.1 8.3.2 8.3.2.1 8.3.2.2 8.3.2.3 8.3.2.4 8.3.3 8.3.4 8.3.4.1 8.3.4.2 8.3.4.3 8.4 8.4.1 8.4.1.1 8.4.1.2 8.4.2 8.4.3 8.4.4 8.5 8.5.1 8.5.2 8.5.3 8.5.3.1 8.5.3.2 8.6

9 9.1 9.2 9.2.1 9.2.2

Enzymatic Reduction Reaction 193 Kaoru Nakamura and Tomoko Matsuda Introduction 193 Hydrogen Source for Coenzyme Regeneration 193 Alcohol as a Hydrogen Source of Reduction 194 Sugar as a Hydrogen Source of Reduction 194 Formate as a Hydrogen Source of Reduction 194 Molecular Hydrogen as a Hydrogen Source of Reduction 195 Light Energy as a Hydrogen Source of Reduction 196 Electric Power as a Hydrogen Source of Reduction 198 Methodology for Stereochemical Control 199 Screening of Microbes 199 Modification of Biocatalysts by Genetic Methods 201 Modified Yeast 201 Overexpression 202 Coexpression of Genes for Carbonyl Reductase and Cofactor-Regenerating Enzymes 203 Modification of Biocatalysts: Directed Evolution 204 Modification of Substrates 205 Modification of Reaction Conditions 206 Acetone Treatment of the Cell 206 Selective Inhibitor 207 Reaction Temperature 208 Medium Engineering 209 Organic Solvent 209 Soluble Organic Solvent 209 Aqueous-organic Two-phase Reaction 209 Use of Hydrophobic Resin, XAD 211 Supercritical Carbon Dioxide 213 Ionic Liquid 215 Synthetic Applications 216 Reduction of Aldehyde 216 Reduction of Ketone 216 Dynamic Kinetic Resolution and Deracemization 221 Dynamic Kinetic Resolution 221 Deracemization through Oxidation and Reduction 223 Conclusions 224 References 225 Biooxidations in Chiral Synthesis 229 Marko D. Mihovilovic and Dario A. Bianchi Introduction 229 Oxidations of Alcohols and Amines 231 Regioselective Oxidation of Alcohols 231 Desymmetrizations of Diols 233

IX


X

Contents

9.2.3 9.2.4 9.2.5 9.3 9.3.1 9.3.2 9.3.3 9.4 9.5 9.5.1 9.5.2 9.5.3 9.5.4 9.6 9.6.1 9.6.2 9.6.3 9.7 9.7.1 9.7.2 9.8 9.9

10 10.1 10.2 10.3 10.3.1 10.3.2 10.4 10.4.1 10.4.2 10.4.3 10.5 10.6. 10.7 10.8

Kinetic Resolution of Primary and Secondary Alcohols Deracemization of Secondary Alcohols 235 Deracemization of Amines 237 Oxygenation of Nonactivated Carbon Centers 237 Hydroxylations by Wild-Type Whole-Cells 238 Hydroxylations by Recombinant Cytochrome P450 Monooxygenases 239 Hydroxylation via Hydroperoxide Formation 241 Enzymatic Epoxidation 241 Baeyer–Villiger Oxidations 243 Chemoselectivity 245 Desymmetrizations 246 Kinetic Resolutions 248 Regiodivergent Biooxidations 251 Heteroatom Oxidations 253 Sulfide Oxidation 253 Oxidation of Dithio-Compounds 256 Oxidation of other Heteroatoms 256 Aryl Dihydroxylations 257 Dioxygenation in ortho- and meta-Position 257 Dioxygenation in ipso- and ortho-Position 262 Halogenation Reactions 263 Summary and Outlook 265 References 265 Aldolases: Enzymes for Making and Breaking C–C Bonds Wolf-Dieter Fessner Introduction 275 Mechanistic Classification 276 Pyruvate Aldolases 278 N-Acetylneuraminic Acid Aldolase 278 Related Pyruvate Aldolases 281 Dihydroxyacetone Phosphate Aldolases 284 Fructose 1,6-Bisphosphate Aldolase 285 Related DHAP Aldolases 286 Preparative Applications 288 Transketolase and Related Enzymes 302 2-Deoxy-D-Ribose 5-Phosphate Aldolase 305 Glycine Aldolases 308 Summary 311 References 311 Index

319

234

275


XI

Preface The increasing interest for obtaining chiral enantiopure organic molecules has tremendously expanded the development of new technologies and synthetic procedures pursuing this goal. Besides, a need for higher sensibility in developing environmentally friendly and sustainable processes is mandatory for the science and technology of the 21st century. Enzymatic processes combine both aspects with a high degree of success, being able to solve conceptual and practical problems. These are the main reasons for the increasing use of biotransformations both in academia and industry. The recent developments in molecular biology and biochemical tools have expanded the real applications of biotransformations, even on very large scales, and has broad applications in pharmaceutical, alimentary and cosmetic industries, just to cite a few. Enzymes work under very mild reaction conditions with a high degree of efficiency and selectivity. These facts constitute the foundations of an already well documented research field with very solid bases, but continuously expanding thanks to new techniques and methodologies. For this reason, we felt necessary to collect and organize these different bio-catalytic approaches to the synthesis of enantioenriched, ideally enantiopure organic compounds. The book discusses most of the existing biocatalytic solutions to real synthetic problems, which makes it particularly helpful for synthetic, fine-chemicals and pharmaceutical chemists. In all chapters our goal has been to be comprehensive and clear enough in order to also make it a friendly-to-use source of information for undergraduate and PhD students aiming to enter into this fascinating and useful research area. Thus, we believe that this book can be both a tool for studying and a bench-guide for solutions to real problems. The book has been divided in two main parts: one about methodology and a second one about synthetic applications. The first methodological part comprises three chapters focused on the different ways of improving the stereochemical outcome of a given biotransformation. The first chapter describes the different strategies available to improve the stereochemical outcome of a biocatalyst by modifiying the reaction medium. The second chapter summarises the improvement of the enzymatic selectivity by generating new catalysts through directed evolution


XII

Preface

techniques. Finally, a third chapter describing different methods available to find new enzymes with improved properties fulfills the methodology part of the book. The second and larger part has been organized attending to the type of reaction catalyzed by the enzyme or the conceptually special approach. Thus, the fourth chapter deals with dynamic kinetic resolutions, an elegant and attractive way of overcoming the 50% maximum theoretical yield of classical resolutions. Two other conceptually interesting and useful approaches, such as deracemization and enantioconvergent processes, are reviewed in chapter five. The sixth chapter presents the most recent uses of enzymes in the transesterification and hydrolysis of carboxylic acid derivatives, alcohols and epoxides, which are among the most widely used applications of enzymes in organic synthesis. In chapter seven, we have tried to compile the most recent advances in using enzymes for the resolution of amines and amide synthesis. Chapters eight and nine deal with reduction and oxidation reactions, respectively, specifically with the goal of obtaining enantioenriched interesting chiral compounds. Additionally, some intelligent solutions for the regeneration of the cofactors needed for the processes are specially highlighted. Finally, the formation of carbon-carbon bonds using enzymes is reviewed in the last chapter of the book. We would also like to thank very warmly all the chapter authors who have felt the importance of producing a book with these characteristics. They clearly understood the philosophy of the project from the beginning, and they have contributed with exceptionally well-written pieces of work in all senses. We really would like to thank them for their highly enthusiastic dedication. We must say that it has been a real pleasure to collaborate with such an excellent group of scientist from all over the world. And last but not least, we would also like to acknowledge the patience of our families and co-workers who, even without participating in this book, have suffered from the time and dedication devoted to this goal.


XIII

List of Contributors Jan-E. Bäckvall Stockholm University SE-106 Department of Organic Chemistry Arrhenius Laboratory 91 Stockholm Sweden

Wolf-Dieter Fessner Technische Universität Darmstadt Institut für Chemie und Biochemie Petersenstraße 22 64287 Darmstadt Germany

Dario A. Bianchi Vienna University of Technology Institute of Applied Synthetic Chemistry Getreidemarkt 9/163-OC 1060 Vienna Austria

Vicente Gotor Universidad de Oviedo Facultad de Química Departamento de Química Orgánica e Inorgánica Julian Clavería 6 33006 Oviedo Spain

Giacomo Carrea Istituto di Chimica del Riconoscimento Molecolare, CNR Via Mario Bianco 9 20131 Milano Italy Robert Chênevert Université Laval Faculté des Sciences et de Génie Departement de Chemie.CREFSIP. Québec, G1K 7P4 Canada

Vicente Gotor-Fernández Universidad de Oviedo Facultad de Química Departamento de Química Orgánica e Inorgánica Julian Clavería 6 33006 Oviedo Spain Romas J. Kazlauskas University of Minnesota 1479 Gortner Avenue Saint Paul, MN 55108 USA


XIV

List of Contributors

Tomoko Matsuda Tokyo Institute of Technology Department of Bioengineering 4259 Nagatsuta, Midori-ku Yokohama 226-8501 Japan

Nicholas Pelchat Université Laval Faculté des Sciences et de Génie Departement de Chemie.CREFSIP. Québec, G1K 7P4 Canada

Belén Martin-Matute Departamento de Química Orgánica Universidad Autónoma de Madrid Cantoblanco 28049 Madrid Spain

Manfred T. Reetz Max-Planck-Institut für Kohlenforschung Kaiser-Wilhelm-Platz 1 45470 Mülheim an der Ruhr Germany

Stockholm University SE-106 Department of Organic Chemistry Arrhenius Laboratory 91 Stockholm Sweden

Jean-Louis Reymond University of Bern Department of Chemistry & Biochemistry Freiestraße 3 3012 Bern Switzerland

Marko D. Mihovilovic Vienna University of Technology Institute of Applied Synthetic Chemistry Getreidemarkt 9/163-OC 1060 Vienna Austria Pierre Morin Université Laval Faculté des Sciences et de Génie Departement de Chemie.CREFSIP. Québec, G1K 7P4 Canada Kaoru Nakamura Kyoto University Uji Institute for Chemical Research Kyoto 611-0011 Japan

Sergio Riva Istituto di Chimica del Riconoscimento Molecolare, CNR Via Mario Bianco 9 20131 Milano Italy Wolfgang Streit University of Bern Department of Chemistry & Biochemistry Freiestraße 3 3012 Bern Switzerland Nicholas J. Turner The University of Edinburgh Department of Chemistry King’s Buildings West Mains Road Edinburgh EH9 3JJ United Kingdom


j1

I Methodology


j3

1 Medium Engineering Giacomo Carrea and Sergio Riva

1.1 Introduction

‘Efficiency’ and ‘selectivity’ are the two keywords that better outline the outstanding performances of enzymes. However, in some cases unsatisfactory stereoselectivity of enzymes can be found and, in these cases, the enantiomeric excesses of products are too low for synthetic purposes. In order to overcome this limitation, a number of techniques have been proposed to enhance the selectivity of a given biocatalyst. The net effect pursued by all these protocols is the increase of the difference in activation energy (DDG6¼) of the two competing diastereomeric enzyme–substrate transition state complexes (Figure 1.1). The enantioselectivity of biocatalytic reactions is normally expressed as the enantiomeric ratio or the E value [1a], a biochemical constant intrinsic to each enzyme that, contrary to enantiomeric excess, is independent of the extent of conversion. In an enzymatic resolution of a racemic substrate, the E value can be considered equal to the ratio of the rates of reaction for the two enantiomers, when the conversion is close to zero. More precisely, the E value is defined as the ratio between the specificity constants (kcat/KM) for the two enantiomers and can be obtained by determination of the kcat and KM of a given enzyme for the two individual enantiomers. However, considering practical limitations, that is, the availability of optically pure enantiomers, E values are more commonly determined on racemates by evaluating the enantiomeric excess values as a function of the extent of conversion in batch reactions. For irreversible reactions, the E value can be calculated from Equation 1 (when the enantiomeric excess of the product is known) or from Equation 2 (when the enantiomeric excess of the substrate is known) [1a]. For reversible reactions, which may be the case in enzymatic resolution carried out in organic solvents (especially at extents of conversion higher than 40%), Equations 3 or 4, in which the reaction equilibrium constant has been introduced, should be used [1b].


j 1 Medium Engineering

4

[EA]

E+P

[EB]

E+Q

+A E +B

[EA]≠

ΔG

ΔΔG≠

[EB]

E + A or B

E+P

E+Q

Figure 1.1 Energy diagram for an enzyme-catalyzed enantioselective reaction. E ¼ enzyme; A and B ¼ enantiomeric substrates; P and Q ¼ enantiomeric products; [EA] and [EB] ¼ enzyme–substrate complexes; DDG6¼ ¼ difference in free energy; 6¼ denotes a transition state.

For obtaining both the product and the remaining substrate in high enantiomeric excess inone reaction step, the E-value needs to be high, usually around or more than100. E¼

ln ½1cð1 þ eep Þ ln ½1cð1eep Þ

ð1Þ

ln ½ð1cÞð1ees Þ ln ½ð1cÞð1 þ ees Þ

ð2Þ

ln ½1ð1 þ KÞcð1 þ eep Þ ln ½1ð1 þ KÞcð1eep Þ

ð3Þ

ln f1ð1 þ KÞ½c þ ees ð1cÞg ln f1ð1 þ KÞ½cees ð1cÞg

ð4Þ

Where c is conversion of substrate, eep and ees are enantiomeric excesses of product (P) and remaining substrate (S), K ¼ (1  ceq)/ceq, and ceq ¼ conversion at equilibrium.


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

Owing to the logarithmic relationship between E and DDG6¼, a small increase in DDG6¼ produces a dramatic change in E. For instance, when DDG6¼ is increased by only 1 kcal mol1 approximately, the enantiomeric excess of the product is enhanced from 80 to 95% [2]. DDG„ ¼ RTlnE

ð5Þ

The stereochemical outcome of a biocatalyzed reaction is predetermined by the enzyme, the reactant substrates, and also by the reaction conditions (temperature, pH, solvent, additives, etc.). Accordingly, each of them offers a possibility to modulate the enzymatic performances. In this chapter we are going to focus on the latter. With this regard, significant literature examples have shown how a careful tuning of the chemical structure of the substrates as well as of the reaction conditions (temperature and pH) might help to reach the synthetic goals in terms of the enantiomeric excesses of the products (see, for instance, Ref. [3,4]). However, the parameter that has been mostly investigated is the chemical composition of the reaction medium, to the point that the so-called ‘medium engineering’ has emerged in the last years as an important area of research and biotechnological development for industrial applications [ for previous reviews see 5]. In this chapter, experimental evidences of the influence of this parameter on enzymatic enantioselectivity are discussed. As will be shown in the following text, most of the literature data are related to the exploitation of hydrolases (lipases, esterases, proteases); however, some significant examples related to other classes of enzymes have also been reported.

1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

The term ‘medium engineering’, that is the possibility to affect enzyme selectivity simply by changing the solvent in which the reaction is carried out, was coined by Klibanov, who indicated it as an alternative or an integration to protein engineering [5a]. Indeed, several authors have confirmed that the enantio-, prochiral-, and even regioselectivity of enzymes can be influenced, sometimes very remarkably, by the nature of the organic solvent used. In the following text, examples of solvent effects on enzyme selectivity, referred either to systems based (i) on water-miscible organic cosolvents added to aqueous buffers or (ii) on organic media with low water activity, are discussed. 1.2.1 Selectivity Enhancement by Addition of Water-Miscible Organic Cosolvents

Initially, biocatalysis was being conducted in neat aqueous solutions because of the general notion that this environment is optimal for maintaining the enzyme conformation most suitable for binding and catalysis. However, because of the limited water-solubility of many organic substrates, it has been suggested to add varying proportions of water-miscible organic cosolvents to achieve an enhanced concentration

j5


j 1 Medium Engineering

6

R

Pig liver esterase , H2O (pH 7) MeOOC

Organic cosolvent

COOMe

HOOC

1

COOMe 2

Scheme 1.1 Table 1.1 Influence of cosolvents on the asymmetric hydrolysis of

the prochiral diester (1) catalyzed by pig liver esterase. Organic cosolvent, v/v

Temperature ( C)

Enantiomeric excesses of (2)

None Acetonitrile, 5% Acetone, 5% Dimethylsulfoxide, 5% Methanol, 5% Methanol, 20%

20 20 20 20 20 10

79 70 72 81 88 97

of the substrate in the reaction medium [6]. Quite soon, however, it became clear that these cosolvents can also influence the stereoselectivity of the biocatalysts. This area of research was pioneered by Bryan Jones and coworkers in the late 1970s and 1980s [7]. For instance, in the 1980s they showed that the pig liver esterase (PLE)catalyzed hydrolysis of the model prochiral compound 3-methyl-glutarate dimethyl ester (1) to give the (R)-monoester (2) was significantly effected by the added cosolvent (Scheme 1.1 and Table 1.1 )[7b], the higher enantiomeric excess value of (2) being obtained in the presence of 20% v/v MeOH and at low temperature. In another significant example published in the same period, Guanti and coworkers described the desymmetrization of the meso cyclohexene diester (3) to give the hemiester (4), whose enantiomeric excess increased from 55 to 96% when moving from plain buffer to the same buffer containing 10% v/v t-BuOH (Scheme 1.2 and Table 1.2 [8]). Finally, as an ‘old’ example of kinetic resolution of racemic mixtures, mention must be made on the report of Kise and Tomiuchi on the significant effect of acetonitrile on the enantioselectivity of different proteases toward the kinetic resolution of aromatic amino acid ethyl esters (5–8). For instance, (L)-DOPA (8) was obtained with 99% ee in the presence of 90% v/v acetonitrile [9]. COOR¢¢

R

NHR¢

HO

OAc OAc 3 Scheme 1.2 .

5 6 7 8

R = R¢ = H ; R¢¢ = Et R = H ; R¢ = Ac ; R¢¢ = Et R = OH ; R¢ = H ; R¢¢ = Et R = OH ; R¢ = H ; R¢¢ = H

Pig liver esterase , H2O (pH 7)

OH OAc

Organic cosolvent 4


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering Table 1.2 Influence of cosolvents on the asymmetric hydrolysis of

the meso-diester (3) catalyzed by pig liver esterase. Organic cosolvent, v/v

Relative rate of hydrolysis

Enantiomeric excesses of (4)

None Dimethylsulfoxide, 20% Dimethylsulfoxide, 40% Dimethylformamide, 20% tert-Butanol, 5% tert-Butanol, 10%

100 70 28 35 70 44

55 59 72 84 94 96

In a very recent report related to a different group of enzymes, it has been shown that cosolvents, especially methanol, can improve to a great extent the enantioselectivity of three Baeyer–Villiger monooxygenases (phenylacetone monooxygenase, 4-hydroxyacetophenone monooxygenase, and ethionamide monooxygenase) when using organic sulfides as the substrates; depending on the enzyme and on the nature of the solvent and the substrate used, reversal of enantiopreference was also observed [10]. Several other examples have been reported in the literature and most of them have been already reviewed (see, for instance, [11]). However, it must be mentioned that, of course, the selectivity enhancement via addition of water-miscible organic cosolvents may not be taken for granted, as sometimes this approach may be unsuccessful [11a]. The discussion on possible rationales of this phenomenon has been reported at the end of this chapter. 1.2.2 Selectivity Enhancement in Organic Media with Low Water Activity 1.2.2.1 Organic Solvent Systems In the third example of the previous paragraph [9], the reaction conditions described are similar to those used for the enzymatic transformations in ‘bulk’ organic media, an area that was pioneered by Klibanov and coworkers in the 1980s [12] and later investigated and synthetically exploited worldwide [13]. In these systems, solid enzyme preparations (e.g. lyophilized or immobilized on a support) are suspended in an organic solvent in the presence of enough aqueous buffers to ensure catalytic activity. Although the amount of water added to the solvent (as a rule of thumb <5% v/v) may exceed its solubility in that solvent, a visible discrete aqueous phase is not apparent because part of it is adsorbed by the enzyme. Therefore, the two phases involved in an organic solvent system are a liquid (bulk organic solvent and reagents dissolved in it) and a solid (hydrated enzyme particles). The interest and success of the enzyme-catalyzed reactions in this kind of media is due to several advantages such as (i) solubilization of hydrophobic substrates; (ii) ease of recovery of some products; (iii) catalysis of reactions that are unfavorable in water (e.g. reversal of hydrolysis reactions in favor of synthesis); (iv) ease of recovery of insoluble biocatalysts; (v) increased biocatalyst thermostability; (vi) suppression of water-induced side reactions. Furthermore, as already said, enzyme selectivity can be markedly influenced, and even reversed, by the solvent.

j7


j 1 Medium Engineering

8

However, in most cases enzymes show lower activity in organic media than in water. This behavior has been ascribed to different causes such as diffusional limitations, high saturating substrate concentrations, restricted protein flexibility, low stabilization of the enzyme–substrate intermediate, partial enzyme denaturation by lyophilization that becomes irreversible in anhydrous organic media, and, last but not least, nonoptimal hydration of the biocatalyst [12d]. Numerous methods have been developed to activate enzymes for optimal use in organic media [13]. Before discussing the ‘medium engineering’ phenomenon and its synthetic relevance in details, it is useful to offer a brief overview of the ‘fundamentals’ of biocatalysis in organic media. 1.2.2.2 Enzyme Properties in Organic Solvents Enzyme activity in organic solvents depends on parameters such as water activity, pH control, substrate–product solvation, enzyme form, and nature of the solvent. Water Activity It became apparent quite soon that enzyme activity was higher in hydrophobic solvents than in hydrophilic ones and, indeed, Laane et al. [14] found that there was a direct correlation between activity and solvent hydrophobicity, expressed as log P (where P isthepartitioncoefficient of a givensolventbetweenn-octanol and water). Soon after, Zaks and Klibanov [12c] demonstrated, with three unrelated oxidoreductases, that enzyme activity was more or less coincident in the various solvents, provided that the amount of water bound to the protein was the same. The maximal enzymatic activity for the three examined enzymes was attained at about 1000 molecules of water/ enzyme molecule, that is, roughly a monolayer of water on the surface. A more rigorous way to correlate enzyme activity with the water present in the reaction medium was proposed by Halling [15] and Goderis [16], who expressed the water in the medium no longer in terms of content but in terms of thermodynamic water activity (aw). Water activity is correlated to mole fraction of water (ww) and water activity coefficient (gw) by the equation aw ¼ wwgw . Since, in first approximation, the water activity coefficient (gw) increases as a function of solvent hydrophobicity, it follows that a given value of aw will be obtained at a lower water concentration in a hydrophobic than in a hydrophilic medium. Therefore, to attain the same level of enzyme hydration, or of aw, in the enzyme, less water is necessary in hydrophobic than in hydrophilic solvents. Because of the fundamental role played by water, it is clear that the control of aw in the enzymatic reactions carried out in organic solvents is of paramount importance and, as a consequence, different methods have been developed to this end, all of them based on the knowledge that some salts have a fixed hydration state, which gives a well-defined aw [17]. pH and pH Control It is well known that the protonation state of the various groups of an enzyme is important for enzyme activity. However, while in water protonation is simply controlled by pH adjustments, this is not the case in organic solvents, where even the pH concept itself is challenged. A simple way to control the initial state of enzyme protonation was developed by Zaks and Klibanov [12a] who showed that enzyme activity in organic solvents was markedly dependent on the pH of the solution from which the enzyme was recovered (by lyophilization or precipitation).


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

This methodology, the so-called ‘pH adjustment’, is now widely applied and is perfectly suitable when there is no alteration of acid–base concentration in the course of the reaction. If this is not the case, as, for instance, in the synthesis, hydrolysis, or aminolysis of esters, enzyme protonation and activity might be deeply influenced. To overcome this drawback, the use of buffering systems involving highly hydrophobic acids and their sodium salts, and highly hydrophobic bases and their hydrochlorides, has been proposed [18]. The main disadvantage of these buffer systems is that they are appreciably soluble only in solvents with a relatively high polarity. Enzyme Form Proteins are practically insoluble in most organic solvents; therefore, in the absence of any special treatment, they are usually present as a solid suspension. This simplifies catalyst–product separation and enzyme reutilization. The simplest way to prepare a biocatalyst for use in organic solvents and, at the same time, to adjust key parameters, such as pH, is its lyophilization or precipitation from aqueous solutions. These preparations, however, can undergo substrate diffusion limitations or prevent enzyme–substrate interaction because of protein–protein stacking. Enzyme lyophilization in the presence of lyoprotectants (polyethylene glycol, various sugars), ligands, and salts have often yielded preparations that are markedly more active than those obtained in the absence of additives [19]. Besides that, the addition of these ligands can also affect enzyme selectivity as follows. Adsorption on solid matrices, which improves (at optimal protein/support ratios) enzyme dispersion, reduces diffusion limitations and favors substrate access to individual enzyme molecules. Immobilized lipases with excellent activity and stability were obtained by entrapping the enzymes in hydrophobic sol-gel materials [20]. Finally, in order to minimize substrate diffusion limitations and maximize enzyme dispersion, various approaches have been attempted to solubilize the biocatalysts in organic solvents. The most widespread method is the one based on the covalent linking of the amphiphilic polymer polyethylene glycol (PEG) to enzyme molecules [21]. Enzyme Kinetics and Stability Kinetic studies, carried out mostly with hydrolases, have shown that enzymes in organic solvents follow conventional models [12a, 22]. Several reports have indicated that enzymes are more thermostable in organic solvents than in water. The high thermal stability of enzymes in organic solvents, especially in hydrophobic ones and at low water content, was attributed to increased conformational rigidity and to the absence of nearly all the covalent reactions causing irreversible thermoinactivation in water [23]. 1.2.2.3 Medium Engineering In the following text, a few examples of solvent effects on enzyme selectivity are discussed, starting from the ‘classical’ works published by Klibanov’s group. In a first report [24], the enantioselectivities of various proteases were evaluated by comparing the biocatalyzedhydrolysisof 2-chloroethyl esters of N-acetyl-L- and D-amino acids in water and their transesterification with n-propanol in butyl ether. By comparing the ratio of the kcat/KM values for the L- and D-enantiomers in the two reactions, a remarkable relation of the proteases enantioselectivity was observed: apparently, in this case, the organic solvents ‘destroyed’ the selectivity of the tested enzymes. This finding

j9


j 1 Medium Engineering

10

Table 1.3 Influence of the organic solvent on the enantioselectivity

of the protease subtilisin in the kinetic resolution of the racemic amine (9) (expressed as the ratio of the initial rate of acylation of the pure enatiomers, vS/vR). Organic solvent

Selectivity (vS/vR)

Toluene Octane Acetonitrile Carbon tetrachloride Ethyl acetate Butyl ether Pyridine Dimethylformamide Tetrahydrofuran tert-Amyl alcohol 3-Methyl-3-pentanol

0.95 1.3 1.4 1.5 1.6 1.8 2.5 2.9 3.5 4.1 7.7

was synthetically exploited for the preparation of peptides containing D-amino acids, compounds that, in water, could not be substrates for these enzymes [25]. Later on the crucial role played by the solvent was enlightened in the proteasecatalyzed resolution of racemic amines [26]. As shown in Table 1.3, the ratio of the initial rates of acylation of the (S)- and the (R)-enantiomers or racemic a-methylbenzylamine (9) varied from nearly 1 in toluene to 7.7 in 3-methyl-3-pentanol. Similarly, the same authors found a signiďŹ cant solvent effect for the subtilisincatalyzed transesteriďŹ cation of racemic 1-phenylethanol (10) using vinyl butyrate as acyl donor (Table 1.4 [27]). NH2

9

OH

10

Table 1.4 Influence of the organic solvent on the enantioselectivity

of the protease subtilisin in the kinetic resolution of the racemic alcohol (10) (expressed as the enatiomeric ratio E, that is the ratio of the specificity constants of the two enatiomers, (Kcat/KM)S/ (Kcat/KM)R). Organic solvent

(Kcat/KM)S

(Kcat/KM)R

E

Dioxane Benzene Triethylamine Tetrahydrofuran Pyridine Dimethylformamide Nitromethane Acetonitrile

170 13 330 230 43 1.4 16 48

2.8 0.24 6.9 5.8 1.4 0.16 3.3 16

61 54 48 40 31 9 5 3


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

Of course, the influence of organic solvents on enzyme enantioselectivity is not limited to proteases but it is a general phenomenon. Quite soon, different research groups described the results obtained with lipases [28]. For instance, the resolution of the mucolytic drug ()–trans-sobrerol (11) was achieved by transesterification with vinyl acetate catalyzed by the lipase from Pseudomonas cepacia adsorbed on celite in various solvents. As depicted in Scheme 1.3 and Table 1.5, it was found that t-amyl alcohol was the solvent of choice: in this medium, the selectivity was so high (E > 500) that the reaction stopped spontaneously at 50% conversion giving both (þ)-transsobrerol and ()-trans-sobrerol monoacetate in 100% optical purity [29]. O

HO Lipase PS ,

AcO

HO

O

+

Organic solvent OH

OH

OH

(±)-trans -11

(1S, 5R)-12

(1R, 5S)-11

Scheme 1.3 . Table 1.5 Influence of the organic solvent on the enantioselectivity

of the lipase PS (from Pseudomonas species) in the kinetic resolution of racemic trans-sobrerol (10). Organic solvent

E

Tetrahydrofuran Acetone Dioxane 3-Pentanone tert-Amyl alcohol

69 142 178 212 518

In all the reported examples, the enzyme selectivity was affected by the solvent used, but the stereochemical preference remained the same. However, in some specific cases it was found that it was also possible to invert the hydrolases enantioselectivity. The first report was again from Klibanov’s group, which described the transesterification of the model compound (13) with n-propanol. As shown in Table 1.6, the enantiopreference of an Aspergillus oryzae protease shifted from the (L)- to the (D)-enantiomer by moving from acetonitrile to CCl4 [30]. Similar observations on the inversion of enantioselectivity by switching from one solvent to another were later reported by other authors [31]. O Cl

O NHAc

13

When discussing the role of reaction medium on enzyme enantioselectivity, the potential effects of (i) water activity [5b,13f,32], (ii) enzyme form, and (iii) pH, should

j11


j 1 Medium Engineering

12

Table 1.6 Influence of the organic solvent on the enantioselectivity

of the protease from A. oryzae subtilisin in the kinetic resolution of the racemic amino acid (12) (expressed as the ratio of the initial rate of acylation of the pure enatiomers, vS/vR). Organic solvent

Selectivity (vS/vR)

Acetonitrile Dimethylformamide Pyridine Tert-Butanol Acetone Tetrahydrofuran Cycloexanone Methylene chloride 3-Octanone Nitrobenzene tert-Butyl acetate Methyl tert-butyl ether Cyclohexane Toluene Octane Carbon tetrachloride

7.1 5.7 4.3 1.7 1.3 1.3 1.1 0.88 0.73 0.60 0.44 0.34 0.27 0.26 0.24 0.19

not be neglected. As for the first point, the best way to avoid water effects is to carry out experiments at controlled water activity, which indeed has been done in numerous cases [ for a review see [5b,13a]]. When the effects of water activity (or water content) have been specifically investigated, the results were rather contradictory since increases, decreases, or, most often, no variation in the enantioselectivity as a function of the water present in the reaction medium were observed [5b,13e,33,34]. Contradictory results were also obtained for enzyme form, since no influence [35] or influence [36] on enzyme enantioselectivity was reported. In the latter case, the large differences in stereoselectivity as a function of Candida antarctica lipase B form were ascribed to mass transport limitations, with lyophilized preparations more prone to this problem than immobilized preparations [36]. Considering pH, it has been demonstrated that alcohol dehydrogenase enantioselectivity is affected by pH variations in aqueous buffers [37]. However, a study carried out in organic solvents showed that the hydrolytic enzyme cutinase was not influenced by pH changes [34]. 1.2.3 Rationales

The experimental evidences that ‘medium engineering’ might represent an efficient method to modify or improve enzyme selectivity (alternative to protein engineering and to the time-consuming search for new catalysts) were immediately matched by the search for a sound rationale of this phenomenon. The different hypotheses formulated to try to rationalize the effects of the solvent on enzymatic enantioselectivity can be grouped into three different classes. The first hypothesis suggests that


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

the solvent, depending on its polarity, could modify the biocatalyst conformation and, thus, influence the selectivity by altering the molecular recognition process between substrate and enzyme [27]. The second group of studies tries to explain the solvent effects on enantioselectivity by means of the contribution of substrate solvation to the energetics of the reaction [38]. For instance, a theoretical model based on the thermodynamics of substrate solvation was developed [39]. However, this model, based on the determination of the desolvated portion of the substrate transition state by molecular modeling and on the calculation of the activity coefficient by UNIFAC, gave contradictory results. In fact, it was successful in predicting solvent effects on the enantio- and prochiral selectivity of g-chymotrypsin with racemic 3-hydroxy-2-phenylpropionate and 2-substituted 1,3propanediols [39], whereas it failed in the case of subtilisin and racemic sec-phenetyl alcohol and trans-sobrerol [40]. That substrate solvation by the solvent can contribute to enzyme enantioselectivity was also claimed in the case of subtilisin-catalyzed resolution of secondary alcohols [41]. In a third model it has been proposed that solvent molecules could bind within the active site and, depending on their structure, interfere with the association or transformation of one enantiomer more than the other one [28a,42,43]. For instance, a correlation between the enantiomeric ratio E and the van der Waals volume of the solvent molecules was observed in the resolution of 3-methyl-2-butanol catalyzed by C. antarctica lipase B; the van der Waals volume was suggested as one of the parameters that govern solvents effects on enzyme selectivity [44]. Additionally, significant evidence at a molecular level came from a series of crystallographic data reported by Klibanov [45]. In these works, whose main goal was to demonstrate that the tertiary structure of enzymes was the same once the proteins had been dissolved in water or suspended in organic solvents, it was shown that molecules of the organic solvents were able to penetrate into the active site of the protease subtilisin, displacing some of the water molecules present. Quite significantly, molecules of different solvents were found to be located in different regions of the active site and, additionally, this happened not only in pure organic solvents but also in mixtures of organic solvents and water (i.e. dioxane–water 40 : 60 v/v) [45b]. On the basis of these results, it is not surprising that the search for an exhaustive rationale for the ‘medium engineering’ effect is still the object of scientific debate. Furthermore, it should be pointed out that despite the aforementioned experimental results supporting the second and third type of hypotheses, both of them lack fully reliable predictive value and are not sufficient by themselves to explain every case. In fact, the second hypothesis appears to be valid only for the specific enzyme and substrate investigated each time, or when the formation of solvent–enzyme complexes is proposed (third hypothesis), no generalization is possible because of the large number of possible solvent–enzyme complexes and because each complex might behave differently depending on the nature of the substrate. However, whatever the mechanism of action is, the effect of solvents on enzyme selectivity is sometimes really dramatic. For example, Hirose et al. [42] reported that in the Pseudomonas species lipase-catalyzed desymmetrization of prochiral

j13


j 1 Medium Engineering

14

NO 2 Cyclohexane HOOC

COOR

NO 2 (R)-15

Pseudomonas sp. lipase ROOC

N H NO 2

COOR N H

14 , R = t BuC(O)OCH 2 or R = EtC(O)OCH 2

ROOC

COOH

i-Propyl ether

(S)-15

N H

Scheme 1.4 Asymmetric hydrolysis of dihydropyridine diesters: influence of solvent on lipase stereochemical preference.

dihydropyridine dicarboxylates (i.e. 14), the (S)-monoesters as high as 99% ee were obtained in water-saturated di-isopropyl ether, whereas the (R)-isomers were formed preferentially (88–91% ee) in water-saturated cyclohexane (Scheme 1.4). 1.2.4 Modulation of Enzyme Selectivity: New Trends of Research

The aspects of medium engineering summarized so far were a hot topic in biocatalysis research during the 1980s and 1990s [5]. Nowadays, all of them constitute a well-established methodology that is successfully employed by chemists in synthetic applications, both in academia and industry. In turn, the main research interests of medium engineering have moved toward the use of ionic liquids as reaction media and the employment of additives. 1.2.4.1 Ionic Liquids Ionic liquids, which can be defined as salts that do not crystallize at room temperature [46], have been intensively investigated as environmentally friendly solvents because they have no vapor pressure and, in principle, can be reused more efficiently than conventional solvents. Ionic liquids have found wide application in organometallic catalysis as they facilitate the separation between the charged catalysts and the products. In recent years ionic liquids have also been employed as media for reactions catalyzed both by isolated enzymes and by whole cells, and excellent reviews on this topic are already available [47]. Biocatalysis has been mainly conducted in those roomtemperature ionic liquids that are composed of a 1,3-dialkylimidazolium or Nalkylpyridinium cation and a noncoordinating anion [47a]. The polarity of common ionic liquids is in the range of the lower alcohols or formamide, and their miscibility with water varies widely and unpredictably and is


1.2 Modulation of Enzyme Enantioselectivity by Medium Engineering

not strictly correlated with their polarity [47a]. Even though the miscibility of ionic liquids and organic solvents is not yet well documented, generally they mix with lower alcohols and ketones, dichloromethane, and THF, whereas they do not mix with alkanes and ethers [48]. The possibility of using solvents with high polarity, such as that of many ionic liquids, increases the solubility of polar substrates and extends the range of applications of biocatalysis. Hydrolases such as lipases (from C. antarctica and from P. cepacia), proteases (thermolysin, a-chymotrypsin), esterases, and glycosidases have been used in plain ionic liquids, whereas oxidoreductases such as formate dehydrogenase, peroxidases, laccases, and Baker yeast have been employed in waterionic liquid mixtures or in water-ionic liquid biphasic systems [47]. Both increased enzyme stability and activity have been reported as compared to conventional organic solvents media. For example, a-chymotrypsin, lipase from C. antarctica and esterase from Bacillus stearothermophilus were found to be 17, 3, and 30 times more stable in ionic liquids than in organic solvents [47b]. Several reports deal also with the effects of ionic liquids on enzyme enantioselectivity, which is the subject of this chapter. Although in several cases there was no change or even a decrease in enantioselectivity compared to organic solvents [47], in other cases improved enantioselectivity was observed [47,49–56]. In the following text, the latter cases will be examined in some detail. Lipases from C. antarctica and P. cepacia showed higher enantioselectivity in the two ionic liquids 1-ethyl-3-methylimidazolium tetrafluoroborate and 1-butyl-3methylimidazolium hexafluoroborate than in THF and toluene, in the kinetic resolution of several secondary alcohols [49]. Similarly, with lipases from Pseudomonas species and Alcaligenes species, increased enantioselectivity was observed in the resolution of 1-phenylethanol in several ionic liquids as compared to methyl tert-butyl ether [50]. Another study has demonstrated that lipase from Candida rugosa is at least 100% more selective in 1-butyl-3-methylimidazolium hexafluoroborate and 1-octyl-3nonylimidazolium hexafluorophosphate than in n-hexane, in the resolution of racemic 2-chloro-propanoic acid [51]. The examples described so far refer to experiments carried out in plain ionic liquids. However, it has been proved that ionic liquids are also capable of positively influencing enzyme enantioselectivity when employed in mixtures with aqueous buffers or conventional organic solvents. For example, N-ethyl-pyridinium trifluoroacetate improved the selectivity of subtilisin Carlsberg in the resolution of several amino acid esters in comparison with acetonitrile; both the ionic liquid and acetonitrile were present in 15% concentration in an aqueous buffer [52]. A subsequent study demonstrated that subtilisin enantioselectivity was related to the kosmotropicity of individual cations and anions of ionic liquids and that it was higher at higher values of the overall ionic liquid kosmotropicity [53]. Ionic liquids–tertbutanol cosolvent systems markedly boosted the activity, stability, and enantioselectivity of C. antarctica lipase B in the resolution of racemic p-hydroxy-phenylglycine methyl ester [54] and, similarly, cosolvent systems consisting of ionic liquids and chloroform or tert-butanol increased the selectivity of the same enzyme in the resolution of racemic methyl mandelate [55]. Finally, the three-component system

j15


j 1 Medium Engineering

16

made of ionic liquid–isopropanol–water gave the best results in the enantioselective hydrolysis of prochiral diester malonates [56]. 1.2.4.2 Additives Basically, there are three ways to tune enzyme enantioselectivity by means of ‘additives’: (i) the additives are placed in the reaction medium together with the organic solvent, the enzyme, and the reagents; (ii) the additives are co-lyophilized with the biocatalyst before use in the organic solvent; (iii) the additives are complexed with the substrates before their transformation in the organic medium. One example that refers to the first case is the resolution of racemic 2-phenyl-4benzyloxazol-5(4H)-one (12) by alcoholysis with n-butanol catalyzed by C. antarctica lipase B or by Mucor miehei lipase [57]. The enantioselectivity of both enzymes was improved by addition to the organic medium of an organic base such as triethylamine or of a solid-state buffer such as 3-(cyclohexylamino)-2-hydroxy-1-propanesulfonic acid and its sodium salt. The positive effect of the additives on the enantioselectivity (and activity) of the two enzymes was attributed to the control of the protonation state of the biocatalysts [54]. Similarly, the addition of a small amount of an aqueous solution containing metal ions (LiCl or MgCl2) to an organic solvent medium, remarkably enhanced (up to 200 times) the enantioselectivity of C. rugosa lipase in the resolution of racemic 2-(4-substituted phenoxy) propionic acids [58]. EPR experiments suggested that the increase in enzyme selectivity brought about by the metal ions was due to an enhancement of the reaction rate for the R-enantiomer combined with a decrease of the rate for the S-enantiomer [58]. In the second way by which additives can influence enzyme enantioselectivity, it has been shown that including excess salt (e.g. KCl) during lyophilization can enhance not only the activity of subtilisin Carlsberg by more than 20 000 fold, but also its enantioselectivity toward N-acetyl-phenylalanine methyl ester [59]. The changes in selectivity reflected the activity for the (L)-enantiomer, whereas the activity for the (D)-enantiomer was mostly unaffected, which suggests that the favored reaction is more sensitive to the structural integrity of the salt-treated enzyme [59]. Enhanced enantioselectivity (and activity) in organic solvents was also observed for lipase from P. cepacia co-lyophilized with crown ethers and cyclodextrins [60]. Colyophilization with cyclodextrins improved the enantioselectivity of subtilisin Carlsberg and C. rugosa lipase as well [61]. The cyclodextrin effect was ascribed to the preservation of the enzyme active site and not to the formation of additive–substrate or –product complexes [61]. Finally, co-lyophilizing horseradish peroxidase with numerous amino acids influenced enzyme’s subsequent stereoselectivity in the sulfoxidation of methyl phenyl sulfide in 2-propanol. The greatest effect was observed with D-proline, which increased enzyme selectivity from a level that was synthetically meaningless to a useful one [62]. An example that refers to the third method additives can be employed is described below. Markedly enhanced enantioselectivity was reported for P. cepacia lipase and subtilisin Carlsberg with chiral substrates converted to salts by treatment with numerous Bronsted–Lowry acids or bases [63]. This effect was observed in various organic solvents but not in water, where the salts apparently dissociate to regenerate


References

the free substrates. It should be emphasized that in some instances the enzymes exhibited virtually no enantiopreference toward a free substrate but a profound one toward its salt [63].

1.3 Conclusions and Outlooks

Nowadays biocatalysis is a well-assessed methodology that has moved from the original status of academic curiosity to become a widely exploited technique for preparative-scale reactions, up to the point that the so-called ‘industrial biotechnology’ (to which biocatalysis contributes to the most extent) is one of the three pillars of the modern sustainable chemistry. As shown in this chapter, by focusing on the modulation of enzyme selectivity by medium engineering, quite simple modifications of the solvent composition can really have significant effects on the performances of the biocatalysts. The main drawback remains the lack of reliable predictive models. Despite the significant research efforts (particularly in the last decade), it is likely that a reasonable foresight of the enantioselective outcome of an enzymatic transformation will continue to be based solely on a careful analysis of the increasingly numerous literature reports.

References 1 (a) Chen, C.S., Fujimoto, Y., Girdaukas, G. and Sih, C.J. (1982) Journal of the American Chemical Society, 104, 7294–7298. (b) Chen, C.-S. and Sih, C.J. (1989) Angewandte Chemie International Edition in English, 28, 695–707. 2 Jones, J.B. (1990) Pure and Applied Chemistry, 62, 1445–1448. 3 Phillips, R.S. (1992) Enzyme and Microbial Technology, 14, 417–419. 4 Barton, M.J., Hamman, J.P., Fitcher, K.C. and Calton, G.J. (1990) Enzyme and Microbial Technology, 12, 577–583. 5 (a) Wescott, C.R. and Klibanov, A.M. (1994) Biochimica et Biophysica Acta, 1206, 1–9. (b) Carrea, G., Ottolina, G. and Riva, S. (1995) Trends in Biotechnology, 13, 63–70. 6 Butler, L.G. (1979) Enzyme and Microbial Technology, 1, 253–259. 7 (a) Jones, J.B. and Mehes, M.M. (1979) Canadian Journal of Chemistry, 57, 2245–2248. (b) Lam, L.K.P., Hui, R.A. and

8

9 10

11

12

Jones, J.B. (1986) The Journal of Organic Chemistry, 51, 2047–2050. Guanti, G., Banfi, L., Narisano, E., Riva, R. and Thea, S. (1986) Tetrahedron Letters, 27, 4639–4642. Kise, H., and Tomiuchi, Y. (1991) Biotechnology Letters, 13, 317–322. de Gonzalo, G., Ottolina, G., Zambianchi, F., Fraaije, M.W. and Carrea, G. (2006) Journal of Molecular Catalysis B: Enzymatic, 39, 91–97. (a) Faber, K., Ottolina, G. and Riva, S. (1993) Biocatalysis, 8, 91–132. (b) Santaniello, E., Ferraboschi, P., Grisenti, P. and Manzocchi, A (1992) Chemical Reviews, 92, 1071–1140. (a) Zaks, A. and Klibanov, A.M. (1985) Proceedings of the National Academy of Sciences of the United States of America, 82, 3192–3196. (b) Klibanov, A.M. (1986) Chemtech, 16, 354–359. (c) Zaks, A. and Klibanov, A.M. (1988) The Journal of

j17


j 1 Medium Engineering

18

13

14

15

16

17

18

Biological Chemistry, 263, 8017–8021. (d) For a recent review see Klibanov, A.M. (2001) Nature, 409, 241–246. (a) Carrea, G. and Riva, S. (2000) Angewandte Chemie International Edition in English, 39, 2226–2254. (b) Koskinen, A.M.P. and Klibanov, A.M. (eds) (1996) Enzymatic Reactions in Organic Media, Blackie Academic & Professional, London. (c) Hudson, E.P., Eppler, R.K. and Clark, D.S. (2005) Current Opinion in Biotechnology, 16, 637–643. (d) Gupta, M.N. and Roy, I. (2004) European Journal of Biochemistry/FEBS, 271, 2575–2583. (e) Klibanov, A.M. (2003) Current Opinion in Biotechnology, 14, 427–431. (f) Halling, P.J. (2000) Current Opinion in Chemical Biology, 4, 74–80. (g) Wolff, A., Straathof, A.J.J., Jongejan, J.A. and Heijnen, J.J. (1997) Biocatalysis and Biotransformation, 15, 175–184. Laane, C., Boeren, S., Vos, K. and Veeger, C. (1987) Biotechnology and Bioengineering, 30, 81–87. (a) Halling, P.J. (1994) Enzyme and Microbial Technology, 16, 178–206. (b) Halling, P.J. (1989) Trends in Biotechnology, 7, 50–52. (c) Halling, P.J. (1984) Enzyme and Microbial Technology, 6, 513–514. Goderis, H.L., Ampe, G., Feyten, M.P., Fouwe, B.L., Guffens, W.M., Van Cauwenbergh, S.M. and Tobback, P.P. (1987) Biotechnology and Bioengineering, 30, 258–266. (a) Wehtje, E., Jasmedh, K., Adlercreutz, P., Chand, S. and Mattiasson, B. (1997) Enzyme and Microbial Technology, 21, 502–510. (b) Dudal, Y. and Lortie, R. (1995) Biotechnology and Bioengineering, 45, 129–134. (c) Ljunger, G., Adlercreutz, P. and Mattiasson, B. (1994) Enzyme and Microbial Technology, 16, 751–755. (d) Yang, Z. and Robb, D. (1993) Biotechnology Techniques, 7, 37–42. (e) Halling, P.J. (1992) Biotechnology Techniques, 6, 271–276. (a) Xu, K. and Klibanov, A.M. (1996) Journal of the American Chemical Society, 118, 9815–9819. (b) Blackwood, A.D., Curran, L.J., Moore, B.D. and Halling, P.J.

19

20

21

22

23

(1994) Biochimica et Biophysica Acta, 1206, 161–165. (a) Mingarro, I., Gonzales-Navarro, H. and Braco, L. (1996) Biochemistry, 35, 9935–9944. (b) Triantafyllou, A.O., Wehtje, E., Adlercreutz, P. and Mattiasson, B. (1995) Biotechnology and Bioengineering, 45, 406–414. (c) Khmelnitsky, Y.L., Welch, S.H., Clark, D.S. and Dordick, J.S. (1994) Journal of the American Chemical Society, 116, 2647–2648. (d) Dabulis, K. and Klibanov, A.M. (1993) Biotechnology and Bioengineering, 41, 566–571. (e) Yamane, T., Ichiryu, T., Nagata, M., Ueno, A. and Shimizu, S. (1990) Biotechnology and Bioengineering, 36, 1063–1069. (f) Russell, A.J. and Klibanov, A.M. (1988) The Journal of Biological Chemistry, 263, 11624–11626. Reetz, M.T. and Zonta, A. (1995) Angewandte Chemie International Edition in English, 34, 301–303. (a) Bovara, R., Carrea, G., Gioacchini, A.M., Riva, S. and Secundo, F. (1997) Biotechnology and Bioengineering, 54, 50–57. (b) Matsushima, A., Kodera, Y., Hiroto, M., Nishimura, H. and Inada, Y. (1996) Journal of Molecular Catalysis B: Enzymatic, 2, 1–17 and references therein. (a) Martinelle, M. and Hult, K. (1995) Biochimica et Biophysica Acta, 1251, 191–197. (b) Chatterjee, S. and Russell, A.J. (1993) Enzyme and Microbial Technology, 15, 1022–1029. (c) Rizzi, M., Stylos, P., Riek, A. and Reuss, M. (1992) Enzyme and Microbial Technology, 14, 709–714. (a) Volkin, D.B., Staubli, A., Langer, R. and Klibanov, A.M. (1991) Biotechnology and Bioengineering, 37, 843–853. (b) GarzaRamos, G., Darszon, A., Tuena de GomezPuyou, M. and Gomez-Puyou, A. (1989) Biochemistry, 28, 3177–3182. (c) Ayala, G., Tuenade Gomez-Puyou, M., GomezPuyou, A. and Darszon, A. (1986) FEBS Letters, 203, 41–43. (d) Wheeler, C.J. and Croteau, R. (1986) Archives of Biochemistry and Biophysics, 248, 429–434. (e) Reslow, M., Adlercreutz, P. and Mattiasson, B.


References

24

25

26

27

28

29

30

31

32

33

34

(1987) Applied Microbiology and Biotechnology, 26, 1–8. (f) Zaks, A. and Klibanov, A.M. (1984) Science, 224, 1249–1251. Sakurai, T., Margolin, A.L., Russell, A.J. and Klibanov, A.M. (1988) Journal of the American Chemical Society, 110, 7236–7237. Margolin, A.L., Tai, D.F. and Klibanov, A.M. (1987) Journal of the American Chemical Society, 109, 7885–7887. Kitaguci, H., Fitzpatrick, P.A., Huber, J.E. and Klibanov, A.M. (1989) Journal of the American Chemical Society, 111, 3094–3095. Fitzpatrick, P.A. and Klibanov, A.M. (1991) Journal of the American Chemical Society, 113, 3166–3171. See, for instance: (a) Secundo, F., Riva, S. and Carrea, G. (1992) Tetrahedron: Asymmetry, 3, 267–280. (b) Gutman, A.L. and Shapira, M. (1991) Journal of the Chemical Society, Chemical Communications, 1467–1468. (c) Parida, S. and Dordick, J.S. (1991) Journal of the American Chemical Society, 113, 2253–2259. (d) Garcia, M.J., Brieva, R., Rebolledo, F. and Gotor, V. (1991) Biotechnology Letters, 13, 867–870. Bovara, R., Carrea, G., Ferrara, L. and Riva, S. (1991) Tetrahedron: Asymmetry, 2, 931–938. Tawaki, S. and Klibanov, A.M. (1992) Journal of the American Chemical Society, 1143, 1882–1884. (a) Baldoli, C., Maiorana, S., Carrea, G. and Riva, S. (1993) Tetrahedron: Asymmetry, 4, 767–772. (b) Ueji, S., Fujino, R., Okubo, N., Miyazawa, T., Kurita, S., Kitadani, M. and Muromatsu, A. (1992) Biotechnology Letters, 14, 163–168. Ducret, A., Trani, M. and Lortie, R. (1998) Enzyme and Microbial Technology, 22, 212–216. Jacobsen, E.E. and Anthonsen, T. (2002) Canadian Journal of Chemistry, 80, 577–581. Vidinha, P., Harper, N., Micaelo, N.M., Lourengo, N.M.T., da Silva, M.D.R.G.,

35

36 37

38

39

40

41

42

43

44

45

46

Cabral, J.M.S., Afonso, C.A.M., Soares, C.M. and Barreiros, S. (2004) Biotechnology and Bioengineering, 85, 442–449. Ottolina, G., Carrea, G., Riva, S., Sartore, L. and Veronese, F.M. (1992) Biotechnology Letters, 14, 947–952. Rotticci, D., Norin, T. and Hult, K. (2000) Organic Letters, 2, 1373–1376. Secundo, F. and Phillips, R.S. (1996) Enzyme and Microbial Technology, 19, 487–492. Terradas, F., Teston-Henry, M., Fitzpatrick, P.A. and Klibanov, A.M. (1993) Journal of the American Chemical Society, 115, 390–396. (a) Ke, T., Wescott, C.R. and Klibanov, A.M. (1996) Journal of the American Chemical Society, 118, 3366–3374. (b) Wescott, C.R., Noritomi, H. and Klibanov, A.M. (1996) Journal of the American Chemical Society, 118, 10365–10370. Colombo, G., Ottolina, G., Carrea, G., Bernardi, A. and Scolastico, C. (1998) Tetrahedron: Asymmetry, 9, 1205–1214. Saville, C.K. and Kazlauskas, R.J. (2005) Journal of the American Chemical Society, 127, 12228–12229. Hirose, Y., Kariya, K., Sasaki, J., Kurono, Y., Ebike, H. and Achiwa, K. (1992) Tetrahedron Letters, 33, 7157–7160. Nakamura, K., Takebe, J., Kitayama, T. and Ohno, A. (1991) Tetrahedron Letters, 32, 4941–4944. Ottosson, J., Fransson, L., King, J.W. and Hult, K. (2002) Biochimica et Biophysica Acta, 1594, 325–334. (a) Schmitke, J.L., Stern, L.J. and Klibanov, A.M. (1998) Proceedings of the National Academy of Sciences of the United States of America, 95, 12918–12923. (b) Schmitke, J.L., Stern, L.J. and Klibanov, A.M. (1998) Biochemical and Biophysical Research Communications, 248, 273–277. (c) Schmitke, J.L., Stern, L.J. and Klibanov, A.M. (1997) Proceedings of the National Academy of Sciences of the United States of America, 94, 4250–4255. Welton, T. (1999) Chemical Reviews, 99, 2071–2083.

j19


j 1 Medium Engineering

20

47 (a) van Rantwijk, F., Lau, R.M. and Sheldon, R.A. (2003) Trends in Biotechnology, 21, 131–138. (b) Park, S. and Kazlauskas, R.J. (2003) Current Opinion in Biotechnology, 14, 432–437. 48 Bonhote, P., Dias, A.P., Papageorgiu, N., Kalyanasundaram, K. and Gratzel, M. (1996) Inorganic Chemistry, 35, 1168–1178. 49 Kim, K.-W., Song, B., Choi, M.-Y. and Kim, M.-J. (2001) Organic Letters, 3, 1507–1509. 50 Schofer, S.H., Kafzik, N., Wasserscheid, P. and Kragl, U. (2001) Chemical Communications, 425–426. 51 Ulbert, O., Frater, T., Belafi-Bako, K. and Gubicza, L. (2004) Journal of Molecular Catalysis B: Enzymatic, 31, 39–45. 52 Zhao, H. and Malhotra, S.V. (2002) Biotechnology Letters, 24, 1257–1260. 53 Zhao, H., Campbell, S.M., Jackson, L., Song, Z.Y. and Olubajo, O. (2006) Tetrahedron: Asymmetry, 17, 377–383. 54 Lou, W.Y., Zong, M.H., Wu, H., Xu, R. and Wang, J.F. (2005) Green Chemistry, 7, 500–506. 55 Pifissao, C. and Nascimento, M.P. (2006) Tetrahedron: Asymmetry, 17, 428–433.

56 Wallert, S., Drauz, K., Grayson, I., Groger, H., de Maria, P.D. and Bolm, C. (2005) Green Chemistry, 7, 602–605. 57 Quiros, M., Parker, M.-C. and Turner, N.J. (2001) The Journal of Organic Chemistry, 66, 5074–5079. 58 Okamoto, T., Yasuhito, E. and Ueji, S.-I. (2006) Organic and Biomolecular Chemistry, 4, 1147–1153. 59 Hsu, W.-T. and Clark, D.S. (2001) Biotechnology and Bioengineering, 73, 231–237. 60 Mine, Y., Fukunaga, K., Itoh, K., Yoshimoto, M., Nakao, K. and Sugimura, Y. (2003) Journal of Bioscience and Bioengineering, 95, 441–447. 61 Fasoli, E., Castillo, B., Santos, A., Silva, E., Ferrer, A., Rosario, E., Griebenow, K., Secundo, F. and Barletta, G.L. (2006) Journal of Molecular Catalysis B: Enzymatic, 42, 20–26. 62 Yu, J.-H. and Klibanov, A.M. (2006) Biotechnology Letters, 28, 555–558. 63 Ke, T. and Klibanov, A.M. (1999) Journal of the American Chemical Society, 121, 3334–3340.


j21

2 Directed Evolution as a Means to Engineer Enantioselective Enzymes Manfred T. Reetz

2.1 Introduction

As delineated in other chapters of this book, the world market for chiral organic compounds such as pharmaceuticals, plant protecting agents, and fragrances continues to expand, requiring efficient methods for their synthesis [1,2]. Asymmetric catalysis constitutes the most efficient and environmentally benign strategy to meet these synthetic goals. When choosing this approach, organic chemists have several options: transition metal catalysts [1,2], organocatalysts [3], and enzymes [4] or catalytic antibodies [5]. The decision as to which option is the best depends on a number of factors including cost of catalysts, catalyst stability, activity, and enantioselectivity, as well as type of solvent and ease of workup. A fundamentally new approach to the development of enantioselective catalysts for use in organic synthesis was introduced in the 1990s [6]. It makes use of directed evolution, a process that had previously been used to improve the stability and/or activity of enzymes [7]. The basis of this concept is the combination of proper molecular biological methods for random gene mutagenesis and expression coupled with high-throughput screening systems for the rapid identification of enantioselective mutant enzymes. Whenever the natural (wild-type, WT) enzyme shows an unacceptably low enantioselectivity (ee) or selectivity factor (E value) for a given transformation of interest, A!B, a library of mutants is created from which the most highly enantioselective variant is identified. Then the process is repeated as often as necessary using, in each case, the gene (DNA segment) of an improved mutant for the next round of mutagenesis [6,8]. Since the inferior mutants are discarded, an evolutionary character of the overall process is simulated (Figure 2.1). In each cycle, the library of mutated genes is first inserted in a standard bacterial host such as Escherichia coli or Bacillus subtilis. Subsequently, bacterial colonies are plated out on agar plates and harvested individually by a colony picker. Each colony is placed in a separate well of a microtiter plate containing nutrient broth, so that the bacteria grow and produce the protein of interest. Because each colony originates


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

22

Gene (DNA)

Wild-type enzyme

Random mutagenesis

1

2

3

4

5

...

5

... etc.

etc.

Library of mutated genes Repeat

Expression

1

2

3

4

Library of mutated enzymes Screening (or selection) for enantioselectivity 2

2

Positive mutants Figure 2.1 Strategy for directed evolution of an enantioselective enzyme [6,8].

from a single cell, mixtures of mutant enzymes are avoided. A portion of each mutant occurring in the harvested bacterial colony is then robotically placed on a different microtiter plate, where the enantioselective reaction of interest is carried out. Since the enzyme variants and the corresponding mutant genes are spatially addressable, the genotype/phenotype relation is maintained. The best mutant (positive hit) is then the starting point for the next cycle of gene mutagenesis, expression, and screening (Figure 2.2). Knowledge of the structure or mechanism of the enzyme is not required [6,8,9]. This approach is quite different from rational design based on site-directed mutagenesis. However, it is also logical owing to the evolutionary character of the process. As shown later in this chapter, the combination of rational design and randomization procedures constitutes an important option in directed evolution.

Mutagenesis

Repeat

Expression

Library of mutant genes in a test tube

.. . . . . .. . Colony . .. .. . .. .. . . . . . . picking . .. .

Bacterial colonies on agar plate

Screening for enantioselectivity Bacteria producing mutant enzymes grown in nutrient broth

Visualization of positive mutants

Figure 2.2 The experimental stages of directed evolution of enantioselective enzymes [6,8].


2.2 Molecular Biological Methods for Mutagenesis

Following a number of early contributions, the 1980s witnessed significant progress in the development of efficient methods for gene mutagenesis and expression, and libraries of mutant enzymes were constructed and screened for enhanced stability or activity [10]. However, this process in itself is not yet directed evolution, because only one cycle of mutagenesis/screening is involved. Rare cases of at least two such cycles began to appear in the literature, which constitute the first cases of true directed evolution of enzymes [10c]. However, it was not until the period of 1993–1997 that the general idea of directed evolution was fully appreciated [11]. Using known and new methods of gene mutagenesis in combination with screening or selection systems, enzyme properties such as activity, thermostability, and stability against hostile solvents were improved. These developments set the stage for the concept of directed evolution of enantioselective enzymes early on (Figures 2.1 and 2.2) [6,8]. The challenges in this new field of asymmetric catalysis include the elaboration of optimal strategies for scanning protein sequence space for enantioselectivity and the development of efficient high-throughput screening systems for evaluating thousands of potentially enantioselective mutant enzymes.

2.2 Molecular Biological Methods for Mutagenesis

During the last two decades, numerous gene mutagenesis methods have become available for application in directed evolution [7,11e,12]. Some of the most important gene mutagenesis methods are described briefly here. For complete coverage, the reader is referred to recent reviews [11e,12,13]. The use of mutator strains or the treatment of the isolated DNA with UV light or chemicals, as reported in the older literature and sometimes still used today, provides a means to target the whole gene for more or less random point mutagenesis, but it is not trivial to control such processes [7a]. Modern gene mutagenesis methods can be roughly assigned to two different categories: those based on point mutations and those making use of recombination [7,11e,12]. The most frequently used mutagenesis method in directed evolution is error-prone polymerase chain reaction (epPCR), described by Leung [14] and improved by Cadwell and Joyce [15]. It is based on the classical DNA amplification method, polymerase chain reaction (PCR). By varying parameters such as the concentration of MgCl2 (or MnCl2) and/or nucleotide concentration, errors in DNA amplification are induced. The mutation rate can be adjusted empirically so that on average one, two, or more amino acid exchange events occur in the expressed protein. Although epPCR is sometimes considered to induce point mutations randomly over the entire gene (and protein), this is actually not the case. Owing to the degeneracy of the genetic code, inter alia, some amino acid exchanges are favored in a given enzyme whereas others are less so [7,16]. Nevertheless, most directed evolution projects begin with epPCR because it is very easy to perform and provides a convenient platform from which the actual evolutionary optimization can begin, very often by additional cycles of epPCR.

j23


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

24

The major challenge in putting directed evolution into practice has to do with the problem of probing protein sequence space efficiently, so that the screening effort can be minimized. The number of enzyme variants N at the theoretically maximum degree of diversity is given by the following algorithm: N¼

19M  X ! ðX  MÞ!M!

ð2:1Þ

where M is the total number of amino acid exchanges per enzyme molecule and X is the number of amino acids per enzyme molecule [7]. For the purpose of illustration, an enzyme composed of 300 amino acids theoretically has 5700 mutants if one amino acid is exchanged, but 16 million in the case of two simultaneous exchanges, and 30.5 billion if three amino acids are exchanged per enzyme molecule simultaneously [7]. Such numbers are often cited to illustrate the problem of protein sequence space, which of course points to the difficulty in screening such large numbers of enzyme mutants. The formation of focused mutant libraries also provides the experimenter with a useful tool in directed evolution [7]. To apply such methods, information regarding the structure of the enzyme (X-ray data or homology model), may be necessary. An intelligent decision is then made regarding the identification of a ‘hot spot’, a position in the amino acid sequence of the enzyme that is believed to be important with respect to a given catalytic property, for example, near the binding pocket. Following this choice, the position is randomized by saturation mutagenesis leading to an enzyme library containing the 20 theoretically different mutants corresponding to the 20 proteinogenic amino acids at a defined position [7,17]. To ensure >95% coverage (i.e. to make certain that all 20 mutants have been evaluated in an appropriate screen), oversampling is necessary [18]. (For a computer program (CASTER) useful in designing saturation mutagenesis libraries, see homepage of the author: www.mpi-muelheim.mpg.de/ reetz.html). It usually suffices to screen about 100–200 clones for a single amino acid exchange(i.e.topickand evaluate thisnumber of bacterialcolonies ontheagarplates).In a different approach, knowledge of the 3D structure of the enzyme is not required because every single position is randomized separately in a systematic manner [19,20]. In still another strategy, ‘hot spots’ resulting from epPCR are each randomized [21]. In an important extension of the above approaches, cassette mutagenesis was developed, which is a form of saturation mutagenesis [11e,17,22]. The user identifies (or predicts) a ‘hot region’ in the enzyme, either on the basis of structural and mechanistic data or as a result of applying epPCR. One usually focuses on a portion of the enzyme near the binding site and designs a cassette, which is a DNA fragment consisting of a defined number of nucleotides encoding the desired amino acids. If two amino acid positions are chosen for simultaneous randomization, 400 theoretically different mutants are possible, and the necessary oversampling for 95% coverage of protein sequence space can be handled using current screening methods (3000 clones). However, when the number of amino acid positions for randomization is increased, screening problems arise if complete coverage is strived for. For example, if four positions are randomized simultaneously, catalyst diversity is high (204 ¼ 160 000 different mutants), which may be desirable, but oversampling necessary for 95% coverage requires >3 million clones to be screened. Of course, one can


2.2 Molecular Biological Methods for Mutagenesis

j25

settle for much lower coverage, and the greatly reduced catalyst diversity in the screened mutants can still lead to the discovery of hits displaying improved catalytic profiles, as has been demonstrated, for example, in the directed evolution of enantioselective lipases [22]. Recently, a method called iterative saturation mutagenesis (ISM) has been proposed and experimentally implemented, which appears to be exceptionally efficient [23,24]. ISM reduces the screening effort drastically, which means that it can be considered to constitute ‘accelerated directed evolution’. It is based on iterative cycles of saturation mutagenesis at predetermined sites in an enzyme, a given site being composed of one or more amino acid positions. The basis for choosing these sites is crucial and depends upon the nature of the catalytic property to be improved. In the case of substrate acceptance and/or enantioselectivity, the choice is made using the combinatorial active-site saturation test (CAST) [23,25–27]. Accordingly, an X-ray structure or a homology model is used to identify appropriate sites around the complete binding pocket. In the case of thermostability, sites showing highest B-factors, likewise available from X-ray data, are chosen [28]. The most general scheme of ISM is given in Figure 2.3, illustrating the case involving four sites. In a simplified version of ISM, each site is considered only once (Figure 2.4). This means that, in the given case of four sites, 65 libraries of mutants are possible, the overall evolutionary process being convergent. Of course, not all branches have to be explored. Indeed, this embodiment of ISM has been shown to be highly successful in the rapid evolution of enantioselectivity [23,26,27] and thermostability [28]. The reason(s) for the success of ISM has to do with the fact that sequence space has been confined to defined locations in the protein that are most likely to respond positively in an additive or cooperative manner. In sharp contrast, when performing several rounds of epPCR (4–8 are typical!), the whole protein is addressed repeatedly although only a fraction of the amino acid positions are important. Owing to statistical reasons, a given improved mutant (hit) evolved by ISM is not etc.

A CD

B

ABD

C

ABC

BCD

D

A BD

A

A

C

A BC

D

BCD

B

A

B

A BC

A CD

A

D

C

WT Figure 2.3 Schematic illustration of Iterative Saturation Mutagenesis (ISM) involving (as an example) four randomization sites A, B, C and D [23,24].

ACD

B CD

D

B

ABD

C


C

D

D

B B

C C

B C

D

A

D

A

A

C D

C

D

D

A

B

A

C C

A B

D

A

C

D A

B D

B

D

D

A A

B B

D

A B

C

A

B

C

C

A C

A

B C

WT Figure 2.4 ISM employing four sites A, B, C and D, each site in a given upward pathway being visited only once.

B

D

C

B

C D

D

A

B

B

A

26

j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes


2.3 High-throughput Screening Methods for Enantioselectivity

Figure 2.5 Scheme for DNA shuffling illustrated for the case in which the parent genes originate from the WT by some form of mutagenesis [7e].

likely to be found by applying the traditional strategy based on repeating cycles of epPCR. Rather than point mutations, recombination can be considered, which in a genetic sense means the breaking and rejoining of DNA in new combinations. The most prominent version, pioneered by Stemmer in 1994, is DNA shuffling [7e,11c,11d]. In general, one or more genes are first digested with a DNase to yield double-stranded oligonucleotide fragments of 10–50 base pairs, which are then amplified in a PCRlike process. Repeating cycles of strand separation and reannealing in the presence of a DNA polymerase followed by a final PCR-amplification step results in the reassembly of full-length mutant genes. DNA shuffling can be performed with one gene, with two or more natural genes, or with mutant genes (Figure 2.5). A limitation is the requirement of relatively high homology. If homologous genes from different species are chosen, the process is called family shuffling, which appears to be particularly efficient because it provides high catalyst diversity [7e,29]. A certain degree of self-hybridization of parental genes occurs, which lowers the quality of the mutant libraries (some colonies contain WT). Other recombinant methods have also been developed, each having advantages and (some) disadvantages [7,12]

2.3 High-throughput Screening Methods for Enantioselectivity

A number of high-throughput enantiomeric excess assays have been developed, yet none are completely general. This crucial aspect of directed evolution of

j27


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

28

enantioselective enzymes is not treated here and the reader is referred to several reviews [30]. Suffice it to say that many of the screens allow typically 1000–5000 samples to be evaluated per day. If this is not possible in a given case, mediumthroughput GC [31] or HPLC [32] should be used, meaning typically 300–800 enantiomeric excess determinations per day. A prescreen for activity, which is easier than enantiomeric excess determination, is advisable because this eliminates nonactive clones.

2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

Inspite of the availability of numerous mutagenesis methods and enantiomeric excess assays, the best strategy for performing directed evolution of enantioselectivity of a given enzyme for an asymmetric reaction has not yet been established. Following the first report of directed evolution of an enantioselective enzyme [6], which focused on a lipase, numerous individual studies regarding further successful examples have followed [33]. However, only one system, namely, the original lipase-catalyzed reaction, has been studied systematically with respect to illuminating the relative merits of applying the various mutagenesis methods. Therefore, the results and conclusions of these ongoing studies are presented here in detail. 2.4.1 Lipase from Pseudomonas aeruginosa (PAL)

The first case study of the directed evolution of an enantioselective enzyme involving repeating cycles of gene mutagenesis, expression, and screening (Figures 2.1 and 2.2) was carried out with Pseudomonas aeruginosa lipase (PAL) [6]. Lipases (triacylglycerol ester hydrolases, EC 3.1.1.3) [4,34] are the most widely used enzymes in synthetic organic chemistry, catalyzing the chemo-, regio- and/or stereoselective hydrolysis of carboxylic acid esters or the reverse reaction in organic solvents. In the case of enantioselectivity, either desymmetrization of prochiral substrates or kinetic resolution of racemates is relevant, and numerous academic and industrial examples are known [34]. The active site of lipases is characterized by a triad composed of serine, histidine, and aspartate, the oxyanion being the crucial intermediate [34] (Figure 2.6). The kinetic resolution of rac-1 was chosen as a model reaction using the WT lipase from PAL as the catalyst [6]. The WT shows a very low selectivity factor: E ¼ 1.1 in slight favor of (S)-2 (Scheme 2.1). PAL is composed of 285 amino acids, the active site being Ser82 [35]. Initially (in 1995–1996), epPCR at a low mutation rate was chosen to induce an average of only one amino acid exchange per enzyme molecule [6]. This choice was in line with the general opinion of other researchers in the area of directed evolution at that time, although their goals were different (enhanced stability, etc.). Typically, 2000–3000 enzyme variants per generation were screened using a UV–vis–based screening


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

H N

N

H O

Catalytic triad Ser

O O

O

O

OR2

O

ON H

O Substrate

Ser

R1

OR2

R1

O-

O

Ser

His

Asp

j29

+

R2OH

R1

H H Acyl enzyme intermediate

‘Oxyanion’

Aroduct alcohol

Ser + O H

R1CO2H Product acid

Figure 2.6 Mechanism of lipase-catalyzed hydrolysis of esters [34].

O

O

R O

NO2

CH3 rac-1 (R = n-C8H17)

H2O Lipase-variants

R

O OH +

CH3 (S)-2

R

OH + O

NO2

CH3 (R)-2

3

Scheme 2.1

system. The (R) substrate and (S) substrate were studied separately pairwise to assess enantioselectivity. As a result of four cycles of mutagenesis/screening, a variant (I) having an E value of 11.3 in favor of (S)-2 was identified (Figure 2.7). Although this constitutes proof-of-principle, a factor of E ¼ 11.3 is far from practical. A fifth round of epPCR was shown to increase the E value slightly, but such a strategy is far from optimal. Indeed, it became clear that it is better to apply only two or three rounds of epPCR and then to switch to another mutagenesis method [8,36]. One of these methods is saturation mutagenesis. It was assumed that the observed sites of amino acid exchange produced by epPCR are ‘hot spots’ that are important for increasing enantioselectivity, but that the specific amino acids identified by sequencing are not necessarily optimal [21c]. Therefore, several of these positions were chosen for saturation experiments [11e,21e,37]. The same strategy was developed independently in a study concerning the enhancement of the thermostability of a protease [21e,38], and it is now used routinely in directed evolution studies. However, not all of the original hot spots constitute sites that lead to improvements upon applying saturation mutagenesis. One of the successful positions turned out to be 155 (Figure 2.7), which was randomized using the appropriate oligonucleotides. This strategy led to variant II displaying an E value of 20 [21c]. The gene encoding enzyme


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

30

E = 11.3 E = 9.4

E = 4.4 E

S155L S149G

V476 S155L S149G

Enzyme variant I F259L V476 S155L S149G

E = 2.1 S149G

E = 1.1 0

WT 1 2 3 Mutant generations

4

Figure 2.7 Enhanced E values of the PAL-catalyzed hydrolysis of rac-1 by cumulative mutations caused by four rounds of epPCR [6].

variant II was then subjected to another cycle of epPCR, resulting in an even better enzyme variant III having five mutations and a selectivity factor of E ¼ 25 in favor of the (S) substrate. In order to assess the utility of other mutagenesis methods, DNA shuffling was considered, but it was not clear which mutant genes should be shuffled. DNA shuffling of the mutant genes produced by low-mutation-rate epPCR (Figure 2.7) failed to achieve significant improvements, leading to the conclusion that higher gene diversity is necessary [22]. Therefore, epPCR of the WT gene was repeated, this time at a higher mutation rate corresponding to an average of three amino acid exchange events per enzyme molecule. This process led to the identification of several improved enzyme variants, IV and V (E ¼ 3 and E ¼ 6.5) being the best ones [22]. In the original project using a low mutation rate, such selectivity factors were not obtained until the second cycle of epPCR (Figure 2.7) [6]. Thus, it appears that high mutation rates are more successful. The genes encoding the variants III, IV, and V were then subjected to conventional DNA shuffling. This process provided variant VI having the highest enantioselectivity up to that point (E ¼ 32) [22]. However, in a different strategy, a focused library was generated by randomizing amino acids 160–163, a region next to the binding site (Figure 2.8) [22]. This procedure constitutes a fusion of rational design and directed evolution. For this purpose saturation mutagenesis was applied, resulting in simultaneous randomization at all four positions. This provided variant VII with E ¼ 30, which is characterized by four mutations (E160A, S161D, L162G, and N163F) [22]. In this case only 5000 clones were screened, although for 95% coverage of the 160 000 theoretically different mutants an oversampling of about 3 million clones would have been necessary. Clearly, all of these strategies for navigating in protein sequence space are successful, but it is not obvious which one is optimal. Finally, the region of accessible protein sequence space was extended by developing a modified version of Stemmer’s combinatorial multiple-cassette mutagenesis (CMCM)


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

Figure 2.8 Binding pocket of PAL for the acid part of rac-1 showing the geometric position of amino acids 160–163, which were randomized simultaneously by saturation mutagenesis [22].

[22]. The original form of CMCM is a special type of DNA shuffling using the WTgene and cassettes composed of defined sequences to be randomized [39]. In the lipase project, two mutant genes encoding the enzyme variants IV and V and a mutagenic oligocassette allowing simultaneous randomization at the ‘hot spots’ 155 and 162 were subjected to DNA shuffling. This resulted in the most enantioselective variant X, displaying a selectivity factor of E ¼ 51 [22]. It is characterized by six mutations, most of which are remote from the active site (Ser 82) (see the discussion that follows). The results of these and other experiments are summarized in Figure 2.9. A total of 40 000 mutants were screened. It is likely that, upon exploring larger portions of protein sequence space, even better lipase variants can be identified. However, efficient directed evolution is not a matter of generating huge libraries, which then require considerable efforts in screening; the goal is to rather create high-quality libraries while minimizing their size [7,8c,12]. These initial systematic studies regarding the directed evolution of PAL allowed several conclusions to be made. Protein sequence space can be explored successfully by applying the following strategies [8c,33]: (1) generation of mutants by several cycles of epPCR at low or preferably at high mutation rate; (2) identification of hot spots and application of saturation mutagenesis; (3) identification of hot regions empirically or by rational design and subsequent application of saturation mutagenesis; (4) application of epPCR at high mutation rate followed by DNA shuffling of the mutants; (5) extension of the process of CMCM to cover focused positions in the enzyme (hot spots identified by earlier experiments).

j31


4 cycles of epPCR at low mutation rate

Enzyme variant I E = 11

Saturation mutagenesis at hot spots

Enzyme variant II E = 20

Wild-type E = 1.1

1 cycle of epPCR at high mutation rate

Enzyme variants IV with E = 3 V with E = 6.5

epPCR at high mutation rate

No significant improvement

Figure 2.9 Schematic summary of the directed evolution of enantioselective lipase variants originating from the WT PAL used as catalysts in the hydrolytic kinetic resolution of ester rac-1. CMCM Âź Combinatorial multiple-cassette mutagenesis [8c,22].

Further epPCR at low mutation rate

Small improvements

epPCR at low mutation rate

Enzyme variant III E = 25

epPCR at low or high mutation rate

No significant improvement

Modified CMCM with IV, V and oligo-cassette at positions 155/162

Enzyme variant X E = 51

DNA-shuffling with III, IV and V

Enzyme variant VI E = 32

: Generated variant

: Mutagenesis method

Cassette mutagenesis region 160-163

Enzyme variant VII E = 30

Cassette mutagenesis at positions 155/162

Enzyme variants VIII with E = 34 IX with E = 30

32

j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

Efforts were also made to invert the sense of enantioselectivity in the hydrolytic kinetic resolution of ester (1) using PAL with preferential formation of (R)-2 [40,41]. Using epPCR and DNA shuffling, an (R)-selective mutant showing an E value of 30 was evolved by screening about 45 000 clones for the (R) enantiomer. The best mutant is characterized by 11 mutations, which are different from those of the best (S)-selective variant X [41]. Systematic investigations of the type described above provide two kinds of lessons. First, they offer some hints on how to choose the right strategy for probing protein sequence space [8c]. Second, they provide intriguing data for theoretical studies concerning the origin of enantioselectivity [36]. Most of the above results were obtained without prior knowledge of the 3D structure of the enzyme, or of its mechanism; rather, the Darwinian nature of the concept was relied on. Insight into how enzymes function can be gained from directed evolution provided a sound theoretical analysis is performed. This in turn generates information, which can be used to simplify further genetic optimization. The lipase (PAL) used in these studies is a hydrolase having the usual catalytic triad composed of aspartate, histidine, and serine [42] (Figure 2.6). Stereoselectivity is determined in the first step, which involves the formation of the oxyanion. Unfortunately, X-ray structural characterization of the (S)- and (R)-selective mutants are not available. However, consideration of the crystal structure of the WT lipase [42] is in itself illuminating. Surprisingly, it turned out that many of the mutants have amino acid exchanges remote from the active site [8,22,40]. In order to approximate the activity of the best variant X, the supernatants were used to perform kinetics [8c]. The kcat values of variant X turned out to be 9.6 s1 for (S)-1 and 1.2 s1 for (R)-1. In going from the WT to variant X, the kcat/Km value (l mol1 s1) increased significantly: for (S)-1, kcat/Km ¼ 9.0 · 102 (wild type) and 3.7 · 105 (variant X); for (R)-1, kcat/Km ¼ 3.5 · 102 (wild type) and 8.4 · 103 (variant X). Thus, although these are only rough numbers, they show that variant X is much more active than the WT. In order to study the source of enhanced (S) selectivity, variant X and other real and hypothetical mutants were subjected to a detailed molecular mechanics/quantum mechanical (MM/QM) study [36]. The six mutations (amino acid substitutions) of variant X are D20N, S53P, S155M, L162G, T180I, and T234S. With the exception of L162G, they all occur at positions remote from the active center (S82). Remote effects had been observed in earlier investigations of traditional protein engineering [10a], and directed evolution [7,43,44], but these were all cases in which thermostability, chemical stability, or activity was the focus of interest. In contrast, enantioselectivity is a parameter that chemists, biochemists, and enzymologists traditionally associate with the active center and the binding pocket, in the tradition of Emil Fischer’s hypothesis of lock-and-key or Koshland’s model of induced fit [45]. Indeed, all previous attempts to influence the enantioselectivity of an enzyme based on site-specific mutagenesis as suggested by rational design focused on the binding pocket [9]. The initial results of the MM/QM study regarding the source of enhanced enantioselectivity led to several plausible conclusions [36a]. First, only two of the six amino acid substitutions of mutant X influence enantioselectivity substantially.

j33


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

34

Figure 2.10 The oxyanions originating from rac-1 in the WT PAL (left) and mutant X (right) [8c,36].

Second, there is essentially no difference in the binding energetics of the (S) and (R) substrates; rather, stereodifferentiation occurs as the oxyanion is formed. Third, a novel relay mechanism starting at the surface of the enzyme and working inward toward the active site is operating. Accordingly, the H-bond network defined by S161/ S53/H83 in the WT is disrupted by the mutation S53P, setting the sidechain of H83 free to move (Figure 2.10). At the same time, mutation L162G provides additional space for the substrate (larger binding pocket) and also for the imidazole residue of H83 to move toward the active site. In the case of the preferred enantiomer (S)-1, additional stabilization of the oxyanion via H-bonding occurs. This is in contrast to (R)-1, in which case the corresponding oxyanion cannot profit from such stabilization for steric reasons. It is likely that the same effects operate in the transition state of the reaction leading to the oxyanion [36]. These results are fully in line with the conclusions of the observed higher activity of variant X relative to the WT [8b]. Lessons continued to be learnt from directed evolution as a result of careful theoretical analysis in a follow-up theoretical study [36b]. For example, it was predicted that a double mutant characterized by S53P/L162G should also be an excellent catalyst. Therefore, this mutant as well as other mutational combinations resulting from deconvolution of the previously most enantioselective variant having six mutations were prepared and tested. It turned out that the S53P/L162G mutant is even more enantioselective (E ¼ 64) [36b]. In addition to clearing up the reasons for the enhancement of enantioselectivity, this study also raises important questions regarding the efficiency of the various strategies used in probing protein sequence space. Apparently, more than half of the six mutations in the earlier best mutant X (E ¼ 51) are superfluous, which may have several reasons. Systematic experimental and theoretical studies of this kind are helpful in performing in vitro evolution of enantioselectivity. Nevertheless, several questions are not fully answered. Are remote mutations more important than those close to the active site, or is the opposite true? Is it more effective to allow randomization all over the enzyme rather than focusing on the region around the active site (or vice versa)? To be sure, when applying epPCR or any other mutagenesis method that more or less


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

covers the whole gene, there are statistically more remote amino acids for exchange events to occur than amino acids close to the active site. Application of the recently proposed ISM (see Section 2) [23,24,26–28] using the combinatorial active-site saturation test (CAST) [25] may turn out to be the most efficient way to evolve enantioselectivity rapidly. CASTing was initially applied in order to expand the range of substrate acceptance, specifically of PAL. The first step in CASTing is to inspect the X-ray structure or a homology model and to identify sites (A, B, C, D, etc.) at which the amino acid sidechains reside next to the binding pocket. Each site may be composed of one, two or more amino acid positions (usually not more than three). The sites are then subjected to saturation mutagenesis using appropriate nucleotide cassettes with formation of highly focused libraries, which are then screened for activity and/or enantioselectivity. Thus, CASTing means the generation of focused libraries around the complete binding pocket [25]. This distinguishes it from previous studies of focused libraries in which only specific sites were selected [7], as for example, in the simultaneous randomization of amino acids 160–163 of PAL [22]. In the case of CASTing for enhanced substrate acceptance of PAL, inspection of the X-ray structure [42] led to the identification of five sites, A, B, C, D, and E, around the binding pocket harboring the acid moiety of esters [25]. The enzyme was known to catalyze the hydrolysis of fatty acid esters, but not of bulky esters. The initial study focused not only on the original ester (1) but also on the sterically more demanding substrates (4–13). Most of these p-nitrophenyl esters are not accepted by the WT (some background reaction is observed owing to the fact that they are activated esters). OR OR

O

O OR

O 1

4

5

O

OR O

OR 6

OR O

7

8

O

O 9

O

O

H3CO

10

11

O R = p-nitrophenyl

OR OR 12

OR

OR

OR

13

j35


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

36

Figure 2.11 CASTing of the lipase from Pseudomonas aeruginosa (PAL) leading to the construction of five libraries of mutants (A–E) produced by simultaneous randomization at sites composed of two amino acids. (For illustrative purposes, the binding of substrate (1) is shown) [25].

The following amino acid pairs were defined: M16/L17, L118/I121, L131/V135, L159/L162 and L231/V232. Five libraries A, B, C, D, and E, respectively, were then created separately by simultaneous randomization at each pair (Figure 2.11) [25]. All the five libraries of mutants were then tested in the hydrolysis of substrates (1) and (4–13). It turned out that most of the hits originate from libraries A or D. Figure 2.12 summarizes the relative rates of the WT- and mutant-catalyzed hydrolyses, the eight best mutants being featured. The data in Figure 2.12 are the results of initial mutagenesis experiments, which does not yet constitute directed evolution. An evolutionary process was subsequently induced by combining the mutations of two improved mutants of the first round [46]. Thereby new mutants were obtained, which show an increase of activity relative to the WT by more than 2 orders of magnitude. Although enantioselectivity was not the


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

Figure 2.12 Substrate profiles of lipase variants produced by CASTing [25].

j37


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

38

focus of this study, some of the mutants were examined for this property. For example, one of the evolved mutants showed a selectivity factor of E ¼ 49 in the hydrolytic kinetic resolution of the bulky substrate (8) [46]. ISM needs to be tested for comparison. 2.4.2 Other Lipases

The Bacillus subtilis lipase A (BSLA) was the subject of two short directed evolution studies [19,47]. In one case systematic saturation mutagenesis at all of the 181positions of BSLA was performed [19]. Using meso-1,4-diacetoxy-2-cyclopentene as the substrate, reversed enantioselectivity of up to 83% ee was observed. In another study synthetic shuffling (Assembly of Designed Oligonucleotides) was tested using BSLA [47]. In yet another investigation, inversion of enantioselectivity was accomplished by a different approach [48]. The focus of this study was the Burkholderia cepacia KWI-56 lipase as a catalyst in the hydrolytic kinetic resolution of rac-14 (Scheme 2.2). The WT shows (S) selectivity with an E factor of 33. Using X-ray structural data of the lipase, a decision was made concerning the choice of four amino acid residues near the binding product. This site was then randomized using seven hydrophobic amino acids building blocks. By use of a novel in vitro technique for the construction and screening of a protein library by single-molecule DNA amplification via PCR followed by in vitro coupled transcription/translation, a mutant lipase was identified showing reversed enantioselectivity (E ¼ 38) [48]. It would be of interest to test iterative CASTing and to compare the results. Ph

Ph CO2Et

rac-14

Ph CO2H

(R)-15

+

CO2H (S)-15

Scheme 2.2

2.4.3 Esterases

Esterases have a catalytic function and mechanism similar to those of lipases, but some structural aspects and the nature of substrates differ [4]. One can expect that the lessons learned from the directed evolution of lipases also apply to esterases. However, few efforts have been made in the directed evolution of enantioselective esterases, although previous work by Arnold had shown that the activity of esterases as catalysts in the hydrolysis of achiral esters can be enhanced [49]. An example regarding enantioselectivity involves the hydrolytic kinetic resolution of racemic esters catalyzed by Pseudomonas fluorescens esterase (PFE) [50]. Using a mutator strain and by screening very small libraries, low improvement in enantioselectivity was


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

observed [50]. Moderate effects were observed upon applying saturation mutagenesis at hot spots [51]. It was concluded that distal mutations increase enantioselectivity only moderately and that mutations near the active site exert stronger effects [51a]. However, in a later study using the same enzyme PFE and applying epPCR, it was postulated that remote mutations are more important [52]. In another study a hyperthermophilic esterase from Aeropyrum pernix K1 (APE1547) was used as a catalyst in the hydrolytic kinetic resolution of rac-3-octanol acetate [53]. Following a single round of epPCR, a mutant displaying a 2.6-fold increase in enantioselectivity was identified having five amino acid substitutions, which were shown to be spatially distal to the catalytic center. 2.4.4 Hydantoinases

Hydantoinases belong to the E.C.3.5.2 group of cyclic amidases, which catalyze the hydrolysis of hydantoins [4,54]. As synthetic hydantoins are readily accessible by a variety of chemical syntheses, including Strecker reactions, enantioselective hydantoinase-catalyzed hydrolysis offers an attractive and general route to chiral amino acid derivatives. Moreover, hydantoins are easily racemized chemically or enzymatically by appropriate racemases, so that dynamic kinetic resolution with potential 100% conversion and complete enantioselectivity is theoretically possible. Indeed, a number of such cases using WT hydantoinases have been reported [54]. However,ifasymmetricinductionispoororifinversionofenantioselectivityisdesired, directed evolution can come to the rescue. Such a case has been reported, specifically in the production of L-methionine in a whole-cell system (E. coli) (Figure 2.13) [55]. Following epPCR and saturation mutagenesis at hot spots, the D-selective hydantoinase from Arthrobacter sp. DSM 9771 was converted into L-selective variants. The best L-selective mutant showed a value of 20% ee at about 30% conversion, compared to the WT displaying ee ¼ 40% in favor of the D-methionine-derivative. With the help of an appropriate L-carbamoylase, L-methionine itself was produced. This academic/ industrial effort provided several selective hydantoinases of industrial interest (O May, (Degussa-H€ uls), personal communication, 2005). 2.4.5 Nitrilases

Nitrilases catalyze the synthetically important hydrolysis of nitriles with formation of the corresponding carboxylic acids [4]. Scientists at Diversa expanded the collection of nitrilases by metagenome panning [56]. Nevertheless, in numerous cases the usual limitations of enzyme catalysis become visible, including poor or only moderate enantioselectivity, limited activity (substrate acceptance), and/or product inhibition. Diversa also reported the first example of the directed evolution of an enantioselective nitrilase [20]. An additional limitation had to be overcome, which is sometimes ignored, when enzymes are used as catalysts in synthetic organic chemistry: product inhibition and/or decreased enantioselectivity at high substrate concentrations [20].

j39


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

40

L-enantiomer

RH 5-Monosubstituted hydantoin

HN

D-enantiomer

O

Racemase

NH

or pH > 8

HR HN

Hydantoinase

+ H2O O

+ H2O

H R

O OH NH2

OH NH2

N-carbamoyl-amino acid

O

O

RH H2N

X

L-carbamoylase

+ H 2O

Îą-amino acid

NH O

O

R H

O

O

+ CO2

OH

+ NH3

Figure 2.13 Reactions and enzymes involved in the production of L-amino acids from racemic hydantoins by the three-enzyme hydantoinase process [55].

The study concerns the desymmetrization of the prochiral dinitrile (16) with preferential formation of the (R)-17, which was known to be a chiral intermediate in the synthesis of the cholesterol-lowering therapeutic drug (18) (Lipitor, Sortis, Torvast, etc.) as shown in Scheme 2.3. Upon screening genomic libraries obtained from environmental samples, more than 200 new nitrilases that allow mild and selective hydrolysis of nitriles were discovered [56]. One of them catalyzes the (R)-selective hydrolysis of (16) with a value O-

HO

O OH OH NC

CN Nitrilase 16

OH

H2O NC

Several CO2H

F

N

steps

CH3

(R)-17

O HN 18

Scheme 2.3

CH3


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

of 94.5% ee at a substrate concentration of 100 mM [20]. However, when experiments were done at a more practical concentration of 2.25 M, activity and enantioselectivity suffered (only 87.8% ee). Therefore, directed evolution was applied to solve these problems. Saturation mutagenesis was performed systematically at all positions of the 330-amino acid enzyme [20]. The library contained 10 528 genetic variants and was screened using a mass spectrometry-based high-throughput enantiomeric excess assay. Of the 31 584 clones that were considered, 17 led to enhanced enantioselectivity over the WT enzyme. Each of these variants was then tested at 2.25 M substrate concentration. Many performed poorly at this high concentration, but several mutants having newly introduced serine, histidine, or threonine at position 190 (hot spot!) showed enhanced enantioselectivity. The A190H mutant turned out to be the most enantioselective and active catalyst. Within 15 hours, complete conversion of (16) was observed, affording (R)-17 with 98% ee. This dramatic improvement is contrasted with the performance of the less active WT (88% ee). The system works even at 3 M substrate concentration with 90% conversion, 98.5% ee, and essentially no substrate inhibition. Volumetric productivity, which is of great importance in any industrial application, was enhanced by the process of directed evolution [20]. 2.4.6 Epoxide Hydrolases

Several reports regarding the directed evolution of enantioselective epoxide hydrolases (EHs) have appeared [23,57–59]. These enzymes constitute important catalysts in synthetic organic chemistry [4,60]. The first two reported studies concern the Aspergillus niger epoxide hydrolase (ANEH) [57,58]. Initial attempts were made to enhance the enantioselectivity of the ANEH-catalyzed hydrolytic kinetic resolution of glycidyl phenyl ether (rac-19). The WT leads to an E value of only 4.6 in favor of (S)-20 (see Scheme 2.4) [58]. Several libraries of mutant ANEHs were prepared by applying epPCR at various mutation rates and transforming into E. coli BL21 (DE3). About 20 000 clones were screened, the most selective ANEH variant showing a selectivity factor of only E ¼ 10.8 in the kinetic resolution of rac-19 [58]. Thus, this enzyme appeared to be ‘difficult’ to evolve. The ANEH-mutant displaying enhanced enantioselectivity (E ¼ 10.8) was sequenced and shown to be characterized by three mutations, A217V near the active site and K332E and A390E both at remote positions [58]. The X-ray crystal structure of the WT ANEH had been analyzed earlier [61], revealing a dimer comprising identical

O PhO rac-19 Scheme 2.4

HO

H2O EH from A. niger

HO

OH + PhO

PhO (S)-20

(R)-20

OH

j41


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

42

subunits. Each monomeric unit contains tyrosines, which bind and activate the epoxide and the catalytic triad (two aspartates and a histidine), typical of most EHs. Aspartate 192 attacks the epoxide nucleophilically with formation of a glycolmonoester intermediate that is hydrolyzed in a second step. The binding pocket is composed of a very narrow tunnel, which may explain the difficulty in evolving enhanced enantioselectivity. Since the traditional approach using epPCR had not provided mutants displaying notable degrees of improved enantioselectivity, a different strategy was sought. At about the same time, the concept of CASTing for broadening the substrate scope of enzymes was being developed [25]. It thus appeared logical to extend it to iterative CASTing, which is an embodiment of ISM. The CAST analysis of ANEH on the basis of its X-ray structure suggested six sites for saturation mutagenesis, namely, A, B, C, D, E, and F (Figure 2.14). These sites are defined by the following amino acid positions: 193/195/296(A), 215/217/219(B), 329/330(C), 349/ 350(D), 317/318(E), and 244/245/249(F). Thus, the general ISM scheme in Figure 2.3 has been extended by two sites to a total of six. The choice of the particular upward pathway in the kinetic resolution of rac-19, that is, the specific order of choosing the sites in ISM, appeared arbitrary. Indeed, the pathway B C D F E, without utilizing A, was the first one that was chosen, and it led to a spectacular increase in enantioselectivity (Figure 2.15). The final mutant, characterized by nine mutations, displays a selectivity factor of E ¼ 115 in the model reaction [23]. This result is all the more remarkable in that only 20 000 clones were screened, which means that no attempt was made to fully cover the defined protein sequence space. Indeed, relatively small libraries were screened. The results indicate the efficiency of iterative CASTing and its superiority over other strategies such as repeating cycles of epPCR. It will be of interest to see whether other pathways (order of choosing the sites in ISM) also provide highly enantioselective mutants, and whether they are different. Moreover, the source of enantioselectivity needs to be illuminated in structural, kinetic, and theoretical studies. Along synthetic lines, it was observed early on that trans-disubstituted epoxides are not accepted by the WT ANEH. However, preliminary results of applying ‘restricted’ CASTing on the basis of NDT instead of NNK degeneracy are highly promising N: adenine/cytosine/guanine/thymine; K: guanine/thymine; D: adenine/guanine/thymine; T: thymine [62]. The idea behind the use of NDT degeneracy has to do with the fact that only 12 amino acids are employed as building blocks, and that the screening effort for 95% coverage of the focused protein sequence space is greatly reduced (Section 2.2). Using this approach, highly active and enantioselectivity mutants were evolved [62]. This is an important synthetic result because the chiral salen-based Jacobsen catalysts fail in the case of trans-disubstituted epoxides [63]. In another study that appeared prior to the advent of CASTing, the traditional combination of epPCR and DNA shuffling was used to enhance the enantioselectivity of the hydrolytic kinetic resolution of p-nitrophenyl glycidyl ether and other epoxides catalyzed by the EH from Agrobacterium radiobacter [59]. Several mutants were obtained with up to 13-fold improved enantioselectivity. The amino acid exchanges took place around the active site.


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

Figure 2.14 CASTing of the epoxide hydrolase from A. niger (ANEH) based on the X-ray structure of the WT [61]. (a) Defined randomization sites Aâ&#x20AC;&#x201C;E; (b) top view of tunnel-like binding pocket showing sites Aâ&#x20AC;&#x201C;E (blue) and the catalytically active D192 (red) [23].

j43


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

44

120

E = 115 (9 Mutations)

110

100

90

Selectivity factor (E)

80

E (Thr317Trp/Thr318Val) 70

60

50

40

E = 35 30

(Leu249Tyr; 244/245 stay) F

E = 24 E = 21

20

D (Cys350Val; 349 stays)

E = 14 C (Met329Pro/Leu330Tyr) B (Leu215Phe/Ala217Asp/Arg219Ser)

10

WT-ANEH Figure 2.15 Iterative CASTing in the evolution of enantioselective epoxide hydrolases as catalysts in the hydrolytic kinetic resolution of rac-19 [23].


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

2.4.7 Phosphotriesterases

Many chiral phosphates and phosphonic acid esters are of value as agricultural insecticides, while others pose a threat as potential chemical warfare agents [64]. In most cases, the biological effects arise exclusively or mainly from one enantiomeric form, with the mirror image compound showing no biological activity or a reduced one [64,65]. One synthetic approach is to perform hydrolytic kinetic resolution using phosphotriesterases as catalysts, a typical example being the production of chiral organophosphates (R)- or (S)-21 (see Scheme 2.5) [65,66]. O R1O P OR2 O

H2O Phosphotriesterase

NO2 rac-21

O R1O P OR2 O

+

O R1O P OR2 OH

NO2 (R)- or (S)-21

22

Scheme 2.5

The phosphotriesterase from Pseudomonas diminuta was shown to catalyze the enantioselective hydrolysis of several racemic phosphates (21), the SP isomer reacting faster than the RP compound [65,66]. Further improvements using directed evolution were achieved by ďŹ rst carrying out a restricted alanine-scan [67] (i.e. at predetermined amino acid positions alanine was introduced). Whenever an effect on activity/ enantioselectivity was observed, the position was deďŹ ned as a hot spot. Subsequently, randomization at several hot spots was performed, which led to the identiďŹ cation of several highly (S)- or (R)-selective mutants [66]. A similar procedure was applied to the generation of mutant phosphotriesterases as catalysts in the kinetic resolution of racemic phosphonates [68]. 2.4.8 Aminotransferases

In nature, aminotransferases participate in a number of metabolic pathways [4]. They catalyze the transfer of an amino group originating from an amino acid donor to a 2-ketoacid acceptor by a simple mechanism. First, an amino group from the donor is transferred to the cofactor pyridoxal phosphate with formation of a 2-keto acid and an enzyme-bound pyridoxamine phosphate intermediate. Second, this intermediate transfers the amino group to the 2-keto acid acceptor. The reaction is reversible, shows ping-pong kinetics, and has been used industrially in the production of amino acids [69]. It can be driven in one direction by the appropriate choice of conditions (e.g. substrate concentration). Some of the aminotransferases accept simple amines instead of amino acids as amine donors, and highly enantioselective cases have been reported [70].

j45


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

46

O

NH2 23

NH2

(S)-transaminase O (S)-24

Scheme 2.6

However, this is obviously not always the case, a counter example being the transformation of b-tetralone (23) to (S)–aminotetralin (24), with 2-aminobutane serving as the amine donor. The WT of an (S)-selective aminotransaminase leads to a value of only 65% ee. In an important early study, a single round of epPCR was performed, and about 10 000cloneswerescreened(nodetails given)[70].Thisledtoseveralactivemutantswith improved enantioselectivity in the range of 84 –94% ee. Some had only one amino acid exchange, while others had two (e.g. M245V/R405G leading to 94% ee). This combinatorial approach does not constitute an evolutionary process, but one can expect that a second round of epPCR, DNA shuffling, or CASTing should result in higher enantioselectivities. In a follow-up study using a single round of epPCR, (S)-1-methoxypropan-2amine was prepared with >99% ee and 97% conversion – a nice case in which evolutionary optimization is not necessary [71]. A more recent study focused on the directed evolution of the o-transaminase from Vibrio fluvialis JS17, specifically with the aim to eliminate product inhibition by aliphatic ketones while maintaining high enantioselectivity. This was achieved by screening 85 000 clones produced by epPCR [72]. 2.4.9 Aldolases

Although a host of asymmetric transition metal-catalyzed [2] and organocatalyzed [3] aldol reactions are known, many problems persist. Aldolases are often complementary, allowing the enantio- and diastereoselective synthesis of many complex carbohydrates in one or two steps [4,73]. Of course, as in other WTenzyme-catalyzed processes, broad substrate acceptance and stereoselectivity cannot be expected to be general. Thus, directed evolution offers intriguing prospects. The first such study addressed the limited substrate acceptance of D-2-keto-3-deoxy-6-phospho-gluconate (KDPG) aldolase [74]. This enzyme catalyzes the (reversible) addition of pyruvate (25) to certain chiral aldehydes such as (26), with formation of aldol products such as (27). It was known that this aldolase is highly specific for chiral phosphorylated aldehydes with the D configuration at the C2 position leading stereoselectively to a precursor of the corresponding D sugar such as (28) (see Scheme 2.7) [75]: Unfortunately, the phosphorylated form of the starting aldehyde is expensive, and dephosphorylation by a phosphatase requires an additional step. Therefore, the challenge was to obtain a mutant aldolase that not only accepts nonphosphorylated substrates but also turns over the enantiomeric aldehyde (29) stereoselectively with formation of (30), which is a precursor of carbohydrate (31) (see Scheme 2.8) [74]:


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

-O

2C

+

OHC

O

O

KDPG aldolase

OPO32-

-O

OH OPO32-

2C

OH

25

j47

OH

26

Phosphatase

27 O

HO OH

CO2-

OH

28 Scheme 2.7

It was indeed possible to turn the KDPG aldolase into an efficient mutant that accepts aldehyde (29), while maintaining perfect control over the creation of the new stereocenter. A few rounds of epPCR and DNA shuffling led to several highly active and stereoselective mutants. One of the best mutants turned out to have four mutations (T84A, I92F, V118A, E138V), most of which occur at positions remote from the active site [74]. A detailed explanation has not been proposed, but this investigation once more shows that directed evolution may lead to effects not predicted by current rational design. It would be interesting to see how iterative CASTing focusing on the binding pocket would perform. This work was extended by subjecting the Neu5Ac aldolase from E. coli to directed evolution in order to expand its catalytic activity for enantiomeric forms of the usual substrates, including N-acetyl-L-mannosamine and L-arabinose with formation of the synthetically important products L-sialic acid and L-3-deoxymanno-2-octulosonic acid (L-KDO) [76]. The evolved Neu5Ac aldolases were characterized by sequence analysis, kinetics, stereoselectivity, and in one case even by an X-ray structure analysis. Again, remote mutations were identified.  Surprisingly, the X-ray structure of one of the triple mutants at 2.3 A resolution showed no significant difference in folding, relative to that of the wild type. The Wong group has extended their studies to include the directed evolution of D-sialic acid aldolase to L-3-deoxy-manno-2-octulosonic acid (L-KDO) aldolase, which likewise underscores the significance of this approach [77a]. In a similar approach, the industrial synthesis of (3R,5S)-6-chloro-2,4,6-trideoxyhexapyranoside has been accomplished by using a genetically altered aldolase [77b]. In another intriguing directed evolution study, the stereochemical course of aldol additions was significantly altered in a different sense [78]; rather than evolving aldolase mutants that selectively accept stereoisomers of substrates, the —O

2C

O

OHC

OH

+ O 25

Scheme 2.8

OH 29

—O

OH

OH

2C

OH OH 30

O

HO HO

31

CO2-


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

48

OH Wild type O

2—O

OPO32—

3PO

OH

OPO32—

HO

O

OH

33

32 Directed evolution OHC

OPO32—

OH

OH 26

Mutant

2—O

O OPO32—

3PO

OH

OH

34 Scheme 2.9

configuration of the chiral aldehyde as substrate was kept constant and mutants were evolved, which showed the opposite diastereoselectivity (see Scheme 2.9). This goal was reached for the first time in a directed evolution study involving the so-called tagatose-1,6-bisphosphate aldolase [78]. The WT catalyzes the aldol addition of (32) to (26) with selective formation of aldol adduct (33). Upon applying three rounds of DNA shuffling and screening 8000 mutants, an aldolase was obtained showing an 80-fold improvement of kcat/Km of the reaction 26 þ 32 ! 34. Thus, a ‘tagatose aldolase’ was turned into a ‘fructose aldolase’. Investigations of this kind are of great value to synthetic organic chemists interested in the selective synthesis of complex stereoisomeric products such as carbohydrates. 2.4.10 Cyclohexanone and Cyclopentanone Monooxygenases as Baeyer–Villigerases and Sulfoxidation Catalysts

Chemo-, regio-, and stereoselective partial oxidation is an area of intense interest in synthetic organic chemistry [1,2,79,80]. Synthetic transition metal catalysts and enzymes have already contributed successfully, but major challenges remain. The Baeyer–Villiger (BV) reaction of ketones with hydroperoxides affording the corresponding esters or lactones constitutes a synthetically important partial oxidation [81–83]. The reaction is accelerated by acids, bases, or metal complexes, with enantioselective catalysis also being possible in some cases. Alternatively, flavindependent enzymes of the type cyclohexanone monooxygenases (CHMO) can be used as biocatalysts in asymmetric BV reactions [84–87]. In this enzymatic process, oxygen (in air) reacts with the enzyme-bound flavin (FAD) to form an intermediate hydroperoxide, which in the deprotonated form initiates the BV reaction by transferring one oxygen atom originating from O2 to the substrate (ketone), the other being ultimately reduced to water [88]. The flavin cofactor needs to be recycled by reduction, the second cofactor NADPH taking over this function (Figure 2.16).


R

N

N N

NADPH

H

H

O

O2

H2 O

N

H+

N O H - O O

N

R

N H

O R1

O Figure 2.16 Mechanism of CHMO-catalyzed Baeyer窶天illiger reaction [85,88].

N

N

O

N H

O

NADP+

N-

N

R

O

N N OH O

N

R

R2

N

N

R2

N H

O

H

O

R1

O

OCriegee-intermediate

R1

N O H O O

N

R

O R2

2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

j49


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

50

Although NADPH recycling systems are well known, it is currently more convenient to use whole cells containing CHMO. CHMO is known to catalyze a number of enantioselective BV reactions, including the kinetic resolution of certain racemic ketones and desymmetrization of prochiral substrates [84–87]. An example is the desymmetrization of 4-methylcyclohexanone, which affords the (S)-configurated seven-membered lactone with 98% ee [84,87]. Of course, many ketones fail to react with acceptable levels of enantioselectivity, or are not even accepted by the enzyme. The initial results of an early directed evolution study are all the more significant, because no X-ray data or homology models were available then to serve as a possible guide [89]. In a model study using whole E. coli cells containing the CHMO from Acinetobacter sp. NCIMB9871, 4-hydroxy-cyclohexanone(35) was usedas thesubstrate. The WT leads to the preferential formation of the primary product (R)-36, which spontaneously rearranges to the thermodynamically more stable lactone (R)-37. The enantiomeric excess of this desymmetrization is only 9%, and the sense of enantioselectivity (R) is opposite to the usually observed (S)-preference displayed by simple 4-alkyl-substituted cyclohexanone derivatives (see Scheme 2.10) [84–87]. In the directed evolution study, epPCR at various mutation rates was applied (10 000 clones). Some of the hits were (R) selective, and others were (S) selective. Eight of them were sequenced [89]. Of particular interest is mutant 1-K2-F5, characterized by a single mutation F432S, because it leads to reversal of enantioselectivity (79% ee in favor of (S)-37). (R) selectivity was optimized by using the genes of some of the best (R)-selective mutants as starting points for a second round of epPCR [89]. In each case only about 1600 mutants were screened. Although a systematic search was not a goal of this initial investigation, the strategy turned out to be successful. In the case of mutant 1-F1-F5 (40% ee), characterized by a single mutation L143F, a second round of epPCR led to a markedly improved mutant 2-D19-E6, showing 90% ee in favor of (R)-37. Sequencing revealed that four new amino acid exchange events had occurred (E292G, T433I, L435Q, T464A) in addition to the already existing L143F mutation.

O

O

H O

O O2

OH

WT CHMO or mutants

OH

O

35

O OH (S)-36

Scheme 2.10

OH

(R)-37

HO (R)-36 O

O

O (S)-37

H


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

O O2

O

38

O

O

+

CHMOmutant OCH3

O

OCH3 (R)-39

OCH3 (S)-39

Scheme 2.11

As the WT CHMO was known to react (S) selectively with simple four-substituted cyclohexanone derivatives [84–87], it was logical to test mutant 1-K2-F5 as a catalyst in the BV reaction of other ketones. For example, when 4-methoxycyclohexanone (38) was subjected to the BV reaction catalyzed by mutant 1-K2-F5, almost complete enantioselectivity was observed in favor of the (S)-lactone (39) (98.5% ee), in contrast to the WT, which is considerably less selective (78% ee) (see Scheme 2.11) [89]. When using mutant 1-K2-F5, other four-substituted cyclohexanone derivatives such as the methyl, ethyl, chloro, bromo, and iodo compounds were found to react with enantioselectivities of 95–99% ee [89]. These enantiomeric excess values are similar to those of the WT, which shows that the mutational change F432S does not simply shift, but rather expands the substrate range in which >95% ee is achieved. Upon testing a fairly wide range of other substrates including cyclopentanone derivatives, bicyclic ketones, functionalized cyclohexanone, and cyclobutanone derivatives, it was discovered that mutant 1-K2-F5 is astonishingly effective [90] (Table 2.1). Enantioselectivities of 90–99% were generally observed. No synthetic catalyst known to date can compete [82,83]. Shortly after the publication of the early work [89], a paper appeared describing the first X-ray structure of a Baeyer–Villigerase, namely, phenylacetone monooxygenase (PAMO) [91]. It was originally found in the thermophilic bacterium Thermobifida fusca and is the first thermostable Baeyer–Villigerase [92]. Unfortunately, it is characterized by narrow range of substrate acceptance. The X-ray data of PAMO allowed the construction of a homology model for CHMO and the possibility of interpreting the previous mutational results on a molecular level [93]. Although the theoretical analysis has not been completed, a cursory inspection suggests that the newly introduced serine at position 432 of mutant 1-K2-F5 forms a hydrogen bond with arginine at position 337, the amino acid which was postulated to stabilize the Criegee intermediate. Since PAMO is unusually thermostable [91,92], it seemed attractive to engineer a mutant showing broader substrate acceptance [93]. Thus, PAMO was turned into a phenylcyclohexanone monooxygenase (PCHMO) by rational design [93]. The best mutant accepts 2-phenylcyclohexanone, inter alia, and leads to an E value of 100 in kinetic resolution, while retaining high thermal stability. The WT CHMO also shows high enantioselectivity in this reaction, but it is not very stable thermally. This investigation constitutes an important step toward developing robust (and thus industrially viable) Baeyer–Villigerases. Moreover, several parameters in a large scale in vitro reaction were optimized, including buffer composition, solvent and cofactor

j51


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

52

Table 2.1 Desymmetrization of prochiral ketones by the BV

reaction using O2 as the oxidant and the CHMO mutant 1-K2-F5 as the catalyst [90]. Desymmetrization

ee (%) 94

O O

O O Cl

O

Cl

O 91

O O

99

O

O O

97

O 78

O

O O O

O

O

O

O

O

O

O

OH

96

>99

O

>99

O

O

>99

O

OH

regeneration system [94]. A light-driven BV reaction showing >95% ee was also devised [95]. Another FAD/NADPH-dependent enzyme is cyclopentanone monooxygenase (CPMO), which appears to have a somewhat different catalytic profile [96]. For this reason a directed evolution project was initiated based on CASTing, although only very small libraries were screened in the early phase of this research (‘mini’ directed evolution) [26]. Codon usage was restricted to NDTdegeneracy, thus utilizing only 12 amino acids. The added challenge in applying CASTing in this case has to do with the dynamics of this class of enzymes. Large conformational changes occur upon


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

H3C

H 3C

O2 CH3

+

O

CHMOmutants

H2C

40

S

O

..

H2C

S

CH3

(R)-41

H2C

S

..

H3 C

CH3

(S)-41

Scheme 2.12

substrate binding. Nevertheless, on the basis of the X-ray structure of PAMO and sequence alignments, two sites were identified in CPMO for CAST libraries, namely, F156/G157 (library A) and G449/F450 (library B). Saturation mutagenesis at these CAST sites using NDT degeneracy and screening only 150 clones (65% coverage) in each case led to several mutants, which show high enantioselectivity in several BV reactions. For example, in the case of 4-methylcyclohexanone, the WT CPMO shows a value of only 46% ee (R), while a mutant boosted enantioselectivity to 92% ee (R). Iterative CASTing [23] can be expected to deliver even better results. It had been known for a long time that CHMO can also be used as catalysts in the enantioselective air-oxidation of some but not all prochiral thioethers [97]. Therefore, directed evolution of the CHMO from Acinetobacter sp. NCIMB 9871 was applied in those cases in which the WT fails [32]. An example is substrate (40), which reacts with an enantiomeric excess of only 14% in slight favor of (R)-41 (see Scheme 2.12). The original library of 10 000 clones used in the Baeyer–Villiger reaction [89] was screened for performance as potential catalysts in the sulfoxidation [32]. This led to the discovery of at least 20 mutants having enantiomeric excess values in the range of 85–99%, some being (R) selective and others being (S) selective. Five mutants resulting in enantiomeric excess values of >95% were sequenced (Table 2.2) [32]. In general, the mutants are different from those that were previously identified as hits in the BV reaction of prochiral cyclohexanone derivatives, which is not surprising. In contrast, mutant 1-K15-C1, which leads to 98.7% ee in favor of (R)-41, is characterized by amino acid exchange F432S. It is therefore identical to mutant Table 2.2 Selected mutant CHMOs from Acinetobacter sp. NCIMB

9871 for the enantioselective air-oxidation of thio-ether 40 (24 hours; 23–25  C) using whole cells [32].

Mutant

Amino acid exchanges

WT 1-D10-F6 1-K15-C1 1-C5-H3 1-H8-A1

— D384H F432S K229I, L248P Y132C, F246I, V361A, T415A F16L, F277S

1-J8-C5

Yield of 41 (%)

Configuration

ee (%)

Sulfone as side product (%)

75 75 55 77 52

(R) (R) (R) (S) (S)

14.0 98.9 98.7 98.1 99.7

<1 7.9 20.0 5.6 26.6

84

(S)

95.2

5.6

j53


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

54

1-K2-F5 previously identified as a highly enantioselective biocatalyst in the asymmetric BV reaction of a wide range of four-substituted cyclohexanone derivatives (as shown above). Thus, one and the same single mutational change at position 432, namely, the exchange of phenylalanine by serine, leads to a surprisingly versatile biocatalyst [32]. Although a detailed mechanistic study still needs to be performed, S432 may well form a H-bond to the crucial amino acid R337. In the sulfoxidation, appreciable over-oxidation with formation of undesired sulfone was observed [32]. Directed evolution was then applied successfully to eliminate undesired sulfone formation, specifically by going through a second cycle of epPCR [32]. This is significant because it shows for the first time that an undesired side reaction can be eliminated by directed evolution. 2.4.11 Monoamine Oxidases

Monoamine oxidases are enzymes that catalyze the racemization of -amino acids [98]. Both L- and D-selective monoamine oxidases are known (i.e. they catalyze the racemization of either (S) or (R) enantiomers of amino acids). This property has been exploited to obtain enantiomerically pure (R) and (S) amino acids by using an appropriate achiral reducing agent such as NaBH4, NaB(CN)H3 or H3NBH3 in combination with an L- or D-selective monoamine oxidase [99]. This synthetically interesting procedure was extended to the production of chiral amines such as a-methylbenzylamine (42) [100]. Unfortunately, the enzymes display somewhat low activity toward substrates of this kind. For example, the monoamine oxidase from A.niger (MAO-N) racemizes (42), with kcat being only 0.17 min1, although enantioselectivity is not bad ((S) versus (R) selectivity amounts to 17: 1) [100]. Thus, directed evolution was applied (see Scheme 2.13). Following several cycles of mutagenesis using the E. coli XL1-Red mutator strain and transformation of the plasmid library into E. coli, a total of about 150 000 bacterial colonies were assayed for activity using a colorimetric prescreen [100]. The best mutant Asn336Ser showed a 47-fold increase in activity and a 5.8-fold enhancement

NH2 Ph

CH3

(S)-selective MAO

(S)-42 NH H3N . BH3 NH2 Ph

CH3

(R)-42 Scheme 2.13

Ph

CH3 43


2.4 Examples of Enhancing the Enantioselectivity of Enzymes by Directed Evolution

of (S) enantioselectivity [100]. Later it was shown that this mutant displays a fairly broad range of primary and secondary amines [101]. More recently, the chemoenzymatic approach has been extended to include the ‘difficult’ class of tertiary amines [102]. These results illustrate the synthetic power of this approach, which is currently being used industrially. 2.4.12 Cytochrome P450 Enzymes

Cytochrome P450 enzymes occur widely in nature, catalyzing the O2-driven oxidation of an immense variety of organic substrates [4,103]. The mechanism of these Fe-heme-containing enzymes has been investigated extensively, and many have been used as whole-cell catalysts in organic chemistry, as in the CH-activating hydroxylation of steroids [104]. Directed evolution has been applied to the bacterial cytochrome P450 BM-3 from Bacillus megaterium with the aim of turning it into practical mutants showing high activity and regioselectivity toward simple substrates including linear alkanes [105]. For example, using n-octane, previous work had shown that a mixture of 2-, 3-, and 4-octanols is obtained, whereas a mutant was shown to lead to approximately 80% regioselectivity in favor of 2-octanol [105c]. The enantiopurity turned out to be only 40% ee (S), but enantioselectivity was not the goal of the investigation. In a later study, some of the previous mutants were screened as catalysts for the hydroxylation of oxazolines, which led to enantioselectivities in the range of 84–88% ee [106]. A few of the original mutants were also tested in the hydroxylation of aryl-acetic acid esters (44), which led to enantiomeric excesses of 56–95% (see Scheme 2.14) [107]. In general, it would be interesting to exploit iterative CASTing or some other form of directed evolution in the quest to control enantioselectivity of P450 enzymes. Thus far, P450 BM-3 was subjected to directed evolution in the successful quest to improve the enantioselectivity of the epoxidation of 1-hexene (up to 83% ee) [108]. Moreover, a triple mutant of P450 BM-3, originally evolved to turn it into an indolehydroxylating catalyst, was found to catalyze the regioselective hydroxylation of highly branched fatty acids [108]. 2.4.13 Other Enzymes

Several other enzymes have been investigated by directed evolution in order to control substrate specificity, although most of the studies are at an early stage. These

ArCH2CO2R

44 Scheme 2.14

P450 BM-3

Mutant

OH ArCHCO2R 45

j55


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

56

include the engineering of the important class of glycosidases/glycosyl transferases [109] and reductases [110].

2.5 Conclusions and Perspectives

Directed evolution of enantioselective enzymes has emerged as a fundamentally new approach to asymmetric catalysis in synthetic organic chemistry, which involves the proper combination of gene mutagenesis, expression, and high-throughput analysis of enantiomeric purity. Several industrial examples of directed evolution of enantioselective and robust enzymes have been described, others are not directly accessible in the open literature. A range of mutagenesis methods are available, and strategies have been developed for applying them when exploring protein sequences. Most directed evolution projects begin with epPCR or DNA shuffling, but the systematic generation of focused libraries has attracted attention recently. Accordingly, iterative CASTing, which is an example of ISM, offers interesting prospects with respect to the enlargement of substrate acceptance and the enhancement of enantioselectivity. ISM has also been applied in the quest to enhance the thermostability of enzymes substantially. Future work will focus on further generalization to include enzymes that are not yet investigated. Enzyme promiscuity is also an emerging field [111], and directed evolution thereof is another challenge, especially when it includes enantioselectivity. Theoretical and computational approaches, including genetic algorithms or neural networks, may prove to be helpful in specific aspects of directed evolution [112]. Finally, the idea of directed evolution of enantioselective hybrid catalysts comprising a synthetic transition metal center anchored covalently or noncovalently to a protein host also offers intriguing prospects, because this allows the tuning of synthetic catalysts by the methods of molecular biology in a Darwinian sense [8b,8c]. Proof-ofprinciple using iterative CASTing was recently presented for the first time, specifically in the case of Rh-catalyzed olefin-hydrogenation [27]. The enantiomeric excess was shown to increase stepwise from 23–65% ee. This work provides incentive for designing and implementing improved systems.

References 1 (a) Collins, A.N., Sheldrake, G.N. and Crosby, J. (eds)(1997) Chirality in Industry II: Developments in the Commercial Manufacture and Applications of Optically Active Compounds, John Wiley & Sons, Ltd, Chichester. (b) Rouhi, A.M. (2003) Chemical and Engineering News, 81, 45–55. 2 Jacobsen, E.N., Pfaltz, A. and Yamamoto, H. (eds)(1999) Comprehensive Asymmetric Catalysis, Springer, Berlin, Vol. I–III.

3 Berkessel, A. and Gr€oger, H. (eds)(2004) Asymmetric Organocatalysis, Wiley-VCH Verlag GmbH, Weinheim. 4 (a) Drauz, K. and Waldmann, H. (eds) (2002) Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook, 2nd edn., Wiley-VCH Verlag GmbH, Weinheim. Vol. I–III. (b) Faber, K. (eds) (2000) Biotransformations in Organic Chemistry, 4th edn., Springer, Berlin.


References (c) Liese, A., Seelbach, K. and Wandrey, C. (eds)(2006) Industrial Biotransformations, Wiley-VCH Verlag GmbH, Weinheim. (d) Schmid, A., Dordick, J.S., Hauer, B., Kiener, A., Wubbolts, M. and Witholt, B. (2001) Nature (London U.K.), 409, 258–268. (e) Koeller, K.M. and Wong, C.-H. (2001) Nature (London U.K.), 409, 232–240. 5 Schultz, P.G., Yin, J. and Lerner, R.A. (2002) Angewandte Chemie, 114, 4607–4618. (2002) Angewandte Chemie International Edition, 41, 4427–4437. 6 (a) Reetz, M.T., Zonta, A., Schimossek, K., Liebeton, K. and Jaeger, K.-E. (1997) Angewandte Chemie, 109, 2961–2963. (1997) Angewandte Chemie International Edition in English, 36, 2830–2832. (b) Reetz M.T., Zonta A., Schimossek K., Liebeton K. and Jaeger K.-E. (1997) patent application DE-A 19731990.4 (A Process for the Preparation and Identification of Novel Hydrolases Having Improved Properties). 7 (a) Arnold, F.H. and Georgiou, G. (eds) (2003) Directed Enzyme Evolution: Screening and Selection Methods, Humana Press, Totowa. Vol. 230. (b) Brakmann, S. and Johnsson, K. (eds)(2002) Directed Molecular Evolution of Proteins (or How to Improve Enzymes for Biocatalysis), WileyVCH Verlag GmbH, Weinheim. (c) Brakmann, S. and Schwienhorst, A. (eds) (2004) Evolutionary Methods in Biotechnology, (Clever Tricks for Directed Evolution), Wiley-VCH Verlag GmbH, Weinheim. (d) Taylor, S.V., Kast, P. and Hilvert, D. (2001) Angewandte Chemie, 113, 3408–3436. (2001) Angewandte Chemie International Edition, 40, 3310–3335. (e) Powell, K.A., Ramer, S.W., del Cardayre, S.B., Stemmer, W.P.C., Tobin, M.B., Longchamp, P.F. and Huisman, G.W. (2001) Angewandte Chemie, 113, 4068–4080; (2001) Angewandte Chemie International Edition, 40, 3948–3959. (f) Rubin-Pitel, S.B. and Zhao, H. (2006) Combinatorial Chemistry and High Throughput Screening, 9, 247–257. (g) Kaur, J. and Sharma, R.

8

9

10

11

(2006) Critical Reviews in Biotechnology, 26, 165–199. (a) Reetz, M.T. (1999) Pure and Applied Chemistry, 71, 1503–1509. (b) Reetz, M.T. (2002) Tetrahedron, 58, 6595–6602. (c) Reetz, M.T. (2004) Proceedings of the National Academy of Sciences of the United States of America, 101, 5716–5722. (d) Reetz, M.T. (2004) Methods in Enzymology, (eds D.E. Robertson and J.P. Noel), Elsevier Academic Press, San Diego, Vol. 388, pp. 238–256. (a) Imanaka, T. and Atomi, H. (2002) In Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook, 2nd edn (eds K. Drauz and H. Waldmann), Wiley-VCH Verlag GmbH, Weinheim, Vol. 1, pp. 67–95. (b) Cedrone, F., Menez, A. and Quemeneur, E. (2000) Current Opinion in Structural Biology, 10, 405–410. (c) Harris, J.L. and Craik, C.S. (1998) Current Opinion in Chemical Biology, 2, 127–132. (d) Kazlauskas, R.J. (2000) Current Opinion in Chemical Biology, 4, 81–88. (e) Li, Q.-S., Schwaneberg, U., Fischer, M., Schmitt, J., Pleiss, J., Lutz-Wahl, S. and Schmid, R.D. (2001) Biochimica et Biophysica Acta, 1545, 114–121. See for example: (a) Cunningham, B.C. and Wells, J.A. (1987) Protein Engineering, 1, 319–325. (b) Lehtovaara, P.M., Koivula, A.K., Bamford, J. and Knowles, J.K.C. (1988) Protein Engineering, 2, 63–68. (c) Liao, H., McKenzie, T. and Hageman, R. (1986) Proceedings of the National Academy of Sciences of the United States of America, 83, 576–580. (a) Chen, K. and Arnold, F.H. (1993) Proceedings of the National Academy of Sciences of the United States of America, 90, 5618–5622. (b) Arnold, F.H. (1998) Accounts of Chemical Research, 31, 125–131. (c) Stemmer, W.P.C. (1994) Nature (London), 370, 389–391. (d) Stemmer, W.P.C. (1994) Proceedings of the National Academy of Sciences of the United States of America, 91, 10747–10751. (e) Reetz, M.T. and Jaeger, K.-E. (1999) Topics in Current Chemistry, 200, 31–57.

j57


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

58

12 Lutz, S. and Patrick, W.M. (2004) Current Opinion in Biotechnology, 15, 291–297. 13 (a) For an illuminating comparison of directed evolution approaches using the glucuronidase model system, see: Rowe, L.A., Geddie, M.L., Alexander, O.B. and Matsumura, I. (2003) Journal of Molecular Biology, 332, 851–860. (b) Study of saturation mutagenesis versus DNA shuffling:Parikh, M.R. and Matsumura, I. (2005) Journal of Molecular Biology, 352, 621–628. (c) Statistical analysis of random mutagenesis methods: Wong, T.S., Roccatano, D., Zacharias, M. and Schwaneberg, U. (2006) Journal of Molecular Biology, 355, 858–871. 14 Leung, D.W., Chen, E. and Goeddel, D.V. (1989) Technique (Philadelphia), 1, 11–15. 15 Cadwell, R.C. and Joyce, G.F. (1992) PCRMethods and Applications, 2, 28–33. 16 (a) Hughes, M.D., Nagel, D.A., Santos, A.F., Sutherland, A.J. and Hine, A.V. (2003) Journal of Molecular Biology, 331, 973–979. (b) Cirino, P.C., Mayer, K.M. and Umeno, D. (2003) Directed Evolution Library Creation: Methods and Protocols (eds F.H. Arnold and G. Georgiou), Humana Press, Totowa, Vol. 231, pp. 3–10. (c) Eggert, T., Reetz, M.T. and Jaeger, K.-E. (2004) Enzyme Functionality - Design, Engineering and Screening (ed. A. Svendsen), Marcel Dekker, New York, pp. 375–390. (d) Wong, T.S., Tee, K.L., Hauer, B. and Schwaneberg, U. (2004) Nucleic Acids Research, 32, e26. 17 (a) Georgescu, R., Bandara, G. and Sun, L. (2003) Directed Evolution Library Creation (eds F.H. Arnold and G. Georgiou), Humana Press, Totowa, pp. 75–84. (b) Wells, J.A., Vasser, M. and Powers, D.B. (1985) Gene, 34, 315–323. (c) Barettino, D., Feigenbutz, M., Valcarcel, R. and Stunnenberg, H.G. (1994) Nucleic Acids Research, 22, 541–542. 18 (a) Bosley, A.D. and Ostermeier, M. (2005) Biomolecular Engineering, 22, 57–61. (b) Patrick, W.M. and Firth, A.E. (2005) Biomolecular Engineering, 22, 105–112. 19 (a) Funke, S.A., Eipper, A., Reetz, M.T., Otte, N., Thiel, W., van Pouderoyen, G.,

20

21

22

23

24 25

26

27

28

Dijkstra, B.W., Jaeger, K.-E. and Eggert, T. (2003) Biocatalysis and Biotransformation, 21, 67–73.(b) Krumm, H. (2004) Dissertation, Ruhr-Universit€at Bochum, Germany. DeSantis, G., Wong, K., Farwell, B., Chatman, K., Zhu, Z., Tomlinson, G., Huang, H., Tan, X., Bibbs, L., Chen, P., Kretz, K. and Burk, M.J. (2003) Journal of the American Chemical Society, 125, 11476–11477. (a) Miyazaki, K. and Arnold, F.H. (1999) Journal of Molecular Evolution, 49, 716–720. (b) Sneeden, J.L. and Loeb, L.A. (2003) Directed Evolution Library Creation: Methods and Protocols (eds F.H. Arnold and G. Georgiou), Humana Press, Totowa, Vol. 231 pp. 65. literature cited therein; (c) Liebeton, K., Zonta, A., Schimossek, K., Nardini, M., Lang, D., Dijkstra, B.W., Reetz, M.T. and Jaeger, K.-E. (2000) Chemistry and Biology, 7, 709–718. Reetz, M.T., Wilensek, S., Zha, D. and Jaeger, K.-E. (2001) Angewandte Chemie, 113, 3701–3703; (2001) Angewandte Chemie International Edition, 40, 3589–3591. Reetz, M.T. and Wang, L.-W. (2006) Bocola in part M. Angewandte Chemie, 118, 1258–1263. Erratum, 2556 (2006) Angewandte Chemie International Edition, 45, 1236–1241. Erratum, 2494. Reetz, M.T. and Carballeira, J.D. Nature Protocols, submitted. Reetz, M.T., Bocola, M., Carballeira, J.D., Zha, D. and Vogel, A. (2005) Angewandte Chemie, 117, 4264–84268 (2005) Angewandte Chemie International Edition, 44, 4192–4196. Clouthier, C.M., Kayser, M.M. and Reetz, M.T. (2006) Journal of Organic Chemistry, 71, 8431–8437. Reetz, M.T., Peyralans, J.J.-P., Maichele, A., Fu, Y. and Maywald, M. (2006) Chemical Communications (Cambridge, U.K.), 4318–4320. Reetz, M.T., Carballeira, J.D. and Vogel, A. (2006) Angewandte Chemie, 118, 7909–7915. (2006) Angewandte Chemie


References

29

30

31

32

33

34

International Edition, 45, 7745–7751. Crameri, A., Raillard, S.-A., Bermudez, E. and Stemmer, W.P.C. (1998) Nature (London U.K.), 391, 288–291. (a) Reetz, M.T. (2001) Angewandte Chemie, 113, 292–320. (2001) Angewandte Chemie International Edition, 40, 284–310.(b) Reetz, M.T. (2003) Methods in Molecular Biology (eds F.H. Arnold and G. Georgiou), Humana Press, Totowa, Vol. 230, pp. 259–282. (c) Reetz, M.T. (2003) Methods in Molecular Biology (eds F.H. Arnold and G. Georgiou), Humana Press, Totowa, Vol. 230, pp. 283–290. (d) Reetz, M.T. (2004) Evolutionary Methods in Biotechnology (eds S. Brakmann and A. Schwienhorst), Wiley-VCH Verlag GmbH, Weinheim, pp. 113–141. (e) Wahler, D. and Reymond, J.-L. (2001) Current Opinion in Biotechnology, 12, 535–544. (f) Schmidt, M. and Bornscheuer, U.T. (2005) Biomolecular Engineering, 22, 51–56. (g) Reymond, J-.L. (ed.)(2005) Enzyme Assays - Highthroughput Screening, Genetic Selection and Fingerprinting, Wiley-VCH Verlag GmbH, Weinheim. Reetz, M.T., K€ uhling, K.M., Wilensek, S., Husmann, H., H€ausig, U.W. and Hermes, M. (2001) Catalysis Today, 67, 389–396. Reetz, M.T., Daligault, F., Brunner, B., Hinrichs, H. and Deege, A. (2004) Angewandte Chemie, 116, 4170–4173 (2004) Angewandte Chemie International Edition, 43, 4078–4081. Review: Reetz, M.T. (2006) Advances in Catalysis (eds B.C. Gates and H. Kn€ ozinger), Elsevier, San Diego, Vol. 49, pp. 1–69. (a) Schmid, R.D. and Verger, R. (1998) Angewandte Chemie, 110, 1694–1720; (1998) Angewandte Chemie International Edition, 37, 1608–1633. (b) Krishna, S.H. and Karanth, N.G. (2002) Catalysis Reviews, 44, 499–591. (c) Berglund, P. and Hult, K. (2000) Stereoselective Biocatalysis (eds R.N. Patel and N. Ramesh), Marcel

35

36

37 38

39 40 41

42

43

44

Dekker, New York, Chapter 21, pp. 633–657. (d) QM-Studies of serine hydrolases: Hu, C.-H., Brinck, T. and Hult, K. (1997) International Journal of Quantum Chemistry, 89–103. (e) Mechanism of serine proteases: Hedstrom, L. (2002) Chemical Reviews, 102, 4501–4523. Jaeger, K.-E., Ransac, S., Dijkstra, B.W., Colson, C., van Heuvel, M. and Misset, O. (1994) FEMS Microbiology Review, 15, 29–63. (a) Bocola, M., Otte, N., Jaeger, K.-E., Reetz, M.T. and Thiel, W. (2004) Chembiochem: A European Journal of Chemical Biology, 5, 214–223. (b) Reetz, M.T., Puls, M., Carballeira, J.D., Vogel, A., Jaeger, K.-E., Eggert, T., Thiel, W., Bocola, M. and Otte, N. (2007) Chembiochem: A European Journal of Chemical Biology, 8, 106–112. Reetz, M.T. (2000) Pure and Applied Chemistry, 72, 1615–1622. Arnold, F.H., Wintrode, P.L., Miyazaki, K. and Gershenson, A. (2001) Trends in Biochemical Sciences, 26, 100–106. Crameri, A. and Stemmer, W.P.C. (1995) BioTechniques, 18, 194–196. Reetz, M.T. and Jaeger, K.-E. (2000) Chemistry-A European Journal, 6, 407–412. Zha, D., Wilensek, S., Hermes, M., Jaeger, K.-E. and Reetz, M.T. (2001) Chemical Communications (Cambridge, U.K.), 2664–2665. Nardini, M., Lang, D.A., Liebeton, K., Jaeger, K.-E. and Dijkstra, B.W. (2000) Journal of Biological Chemistry, 275, 31219–31225. Iffland, A., Tafelmeyer, P., Saudan, C. and Johnsson, K. (2000) Biochemistry, 39, 10790–10798. (a) Kumar, M., Kannan, K.K., Hosur, M.V., Bhavesh, N.S., Chatterjee, A., Mittal, R. and Hosur, R.V. (2002) Biochemical and Biophysical Research Communications, 294, 395–401. (b) Agarwal, P.K., Billeter, S.R., Rajagopalan, P.T.R., Benkovic, S.J. and Hammes-Schiffer, S. (2002) Proceedings of the National Academy of

j59


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

60

45

46

47

48

49 50

51

52

53

54

55 56

Sciences of the United States of America, 99, 2794–2799. Fersht, A. (1999) Structure and Mechanism in Protein Science, W.H. Freeman, New York. Reetz, M.T., Carballeira, J.D., Peyralans, J.J.-P., H€obenreich, H., Maichele, A. and Vogel, A. (2006) Chemistry-A European Journal, 12, 6031–6038. Zha, D., Eipper, A. and Reetz, M.T. (2003) Chembiochem: A European Journal of Chemical Biology, 4, 34–39. Koga, Y., Kato, K., Nakano, H. and Yamane, T. (2003) Journal of Molecular Biology, 331, 585–592. Moore, J.C. and Arnold, F.H. (1996) Nature Biotechnology, 14, 458–467. (a) Bornscheuer, U.T., Altenbuchner, J. and Meyer, H.H. (1998) Biotechnology and Bioengineering, 58, 554–559. (b) Henke, E. and Bornscheuer, U.T. (1999) Biological Chemistry, 380, 1029–1033. (a) Horsman, G.P., Liu, A.M.F., Henke, E., Bornscheuer, U.T. and Kazlauskas, R.J. (2003) Chemistry-A European Journal, 9, 1933–1939. (b) Park, S., Morley, K.L., Horsman, G.P., Holmquist, M., Hult, K. and Kazlauskas, R.J. (2005) Chemistry and Biology, 12, 45–54. Schmidt, M., Hasenpusch, D., K€ahler, M., Kirchner, U., Wiggenhorn, K., Langel, W. and Bornscheuer, U.T. (2006) Chembiochem: A European Journal of Chemical Biology, 7, 805–809. Zhang, G., Gao, R., Zheng, L., Zhang, A., Wang, Y., Wang, Q., Feng, Y. and Cao, S. (2006) Journal of Molecular Catalysis B: Enzymatic, 38, 148–153. Syldatk, C., May, O., Altenbuchner, J., Mattes, R. and Siemann, M. (1999) Applied Microbiology and Biotechnology, 51, 293–309. May, O., Nguyen, P.T. and Arnold, F.H. (2000) Nature Biotechnology, 18, 317–320. DeSantis, G., Zhu, Z., Greenberg, W.A., Wong, K., Chaplin, J., Hanson, S.R., Farwell, B., Nicholson, L.W., Rand, C.L., Weiner, D.P., Robertson, D.E. and Burk,

57

58

59

60

61

62 63 64

65

66

67

68

M.J. (2002) Journal of the American Chemical Society, 124, 9024–9025. Cedrone, F., Niel, S., Roca, S., Bhatnagar, T., Ait-Abdelkader, N., Torre, C., Krumm, H., Maichele, A., Reetz, M.T. and Baratti, J.C. (2003) Biocatalysis and Biotransformation, 21, 357–364. Reetz, M.T., Torre, C., Eipper, A., Lohmer, R., Hermes, M., Brunner, B., Maichele, A., Bocola, M., Arand, M., Cronin, A., Genzel, Y., Archelas, A. and Furstoss, R. (2004) Organic Letters, 6, 177–180. van Loo, B., Spelberg, J.H.L., Kingma, J., Sonke, T., Wubbolts, M.G. and Janssen, D.B. (2004) Chemistry and Biology, 11, 981–990. Archelas, A. and Furstoss, R. (2001) Current Opinion in Chemical Biology, 5, 112–119. Zou, J.Y., Hallberg, B.M., Bergfors, T., Oesch, F., Arand, M., Mowbray, S.L. and Jones, T.A. (2000) Structure (London), 8, 111–122. Reetz, M.T., Kahakeaw, D. and Lohmer, R. unpublished. Jacobsen, E.N. (2000) Accounts of Chemical Research, 33, 421–431. Benschop, H.P. and de Jong, L.P.A. (1988) Accounts of Chemical Research, 21, 368–374. Wu, F., Li, W.-S., Chen-Goodspeed, M., Sogorb, M.A. and Raushel, F.M. (2000) Journal of the American Chemical Society, 122, 10206–10207. (a) Chen-Goodspeed, M., Sogorb, M.A., Wu, F., Hong, S.-B. and Raushel, F.M. (2001) Biochemistry, 40, 1325–1331. (b) Chen-Goodspeed, M., Sogorb, M.A., Wu, F. and Raushel, F.M. (2001) Biochemistry, 40, 1332–1339. (a) Cunningham, B.C. and Wells, J.A. (1989) Science, 244, 1081–1085. (b) Morrison, K.L. and Weiss, G.A. (2001) Current Opinion in Chemical Biology, 5, 302–307. Hill, C.M., Li, W.-S., Thoden, J.B., Holden, H.M. and Raushel, F.M. (2003) Journal of the American Chemical Society, 125, 8990–8991.


References 69 (a) Rozzell, J.D. and Bommarius, A.S. (2002) In Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook2nd edn. (eds K. Drauz and H. Waldmann), Wiley-VCH Verlag GmbH, Weinheim, Vol. II, pp. 873–893. (b) Li, T., Kootstra, A.B. and Fotheringham, I.G. (2002) Organic Process Research and Development, 6, 533–538. 70 Matcham, G.W. and Bowen, A.R.S. (1996) Chimica Oggi-Chemistry Today, 14 (6), 20–24. 71 Matcham, G., Bhatia, M., Lang, W., Lewis, C., Nelson, R., Wang, A. and Wu, W. (1999) Chimia, 53, 584–589. 72 Yun, H., Hwang, B.-Y., Lee, J.-H. and Kim, B.-G. (2005) Applied and Environment Microbiology, 4220–4224. 73 (a) Fessner, W.-D. and Walter, C. (1996) Topics in Current Chemistry, 184, 97–194. (b) Silvestri, M.G., Desantis, G., Mitchell, M. and Wong, C.-H. (2003) Topics in Stereochemistry (ed. S.E. Denmark), John Wiley & Sons, Ltd, Weinheim, Vol. 23, pp. 267–342. 74 Fong, S., Machajewski, T.D., Mak, C.C. and Wong, C.-H. (2000) Chemistry and Biology, 7, 873–883. 75 Shelton, M.C., Cotterill, I.C., Novak, S.T.A., Poonawala, R.M., Sudarshan, S. and Toone, E.J. (1996) Journal of the American Chemical Society, 118, 2117–2125. 76 Wada, M., Hsu, C.-C., Franke, D., Mitchell, M., Heine, A., Wilson, I. and Wong, C.-H. (2003) Bioorganic and Medicinal Chemistry, 11, 2091–2098. 77 (a) Hsu, C.-C., Hong, Z., Wasa, M., Franke, D. and Wong, C.-H. (2005) Proceedings of the National Academy of Sciences of the United States of America, 102, 9122–9126. (b) Jennewein, S., Sch€ urmann, M., Wolberg, M., Hilker, I., Luiten, R., Wubbolts, M. and Mink, D. (2006) Biotechnology Journal, 1, 537–548. 78 Williams, G.J., Domann, S., Nelson, A. and Berry, A. (2003) Proceedings of the

79

80

81 82

83

84

85

86

87

88

National Academy of Sciences of the United States of America, 100, 3143–3148. Cornils, B. and Herrmann, W.A. (eds) (1996) Applied Homogeneous Catalysis with Organometallic Compounds, Wiley-VCH Verlag GmbH, Vol. 1–2. Holland, H.L. and Weber, H.K. (2000) Current Opinion in Biotechnology, 11, 547–553. Krow, G.R. (1993) Organic Reaction (New York), 43, 251–798. (a) Bolm, C., Schlingloff, G. and Weickhardt, K. (1994) Angewandte Chemie, 106, 1944–1946; (1994) Angewandte Chemie International Edition in English, 33, 1848–1849. (b) Bolm, C., Frison, J.C., Zhang, Y. and Wulff, W.D. (2004) Synlett, 1619–1621. (c) Bolm, C., Palazzi, C. and Beckmann, O. (2004) In Transition Metals for Organic Chemistry: Building Blocks and Fine Chemicals, 2nd edn. (eds M. Beller and C. Bolm), WileyVCH Verlag GmbH, Weinheim, Vol. 2, pp. 267–274. Strukul, G. (1998) Angewandte Chemie, 110, 1256–1267. (1998) Angewandte Chemie International Edition, 37, 1198–1209. Taschner, M.J. and Black, D.J. (1988) Journal of the American Chemical Society, 110, 6892–6893. Flitsch, S. and Grogan, G. (2002) Enzyme Catalysis in Organic Synthesis (eds K. Drauz and H Waldmann), Wiley-VCH Verlag GmbH, Weinheim, Vol. 2, pp. 1202–1245. Stewart, J.D., Reed, K.W., Zhu, J., Chen, G. and Kayser, M.M. (1996) Journal of Organic Chemistry, 61, 7652–7653. (a) Mihovilovic, M.D., Chen, G., Wang, S., Kyte, B., Rochon, F., Kayser, M.M. and Stewart, J.D. (2001) Journal of Organic Chemistry, 66, 733–738. (b) Mihovilovic, M.D., M€ uller, B., Schulze, A., Stanetty, P. and Kayser, M.M. (2003) European Journal of Organic Chemistry, 2243–2249. (a) Donoghue, N.A., Norris, D.B. and Trudgill, P.W. (1976) European Journal of Biochemistry, 63, 175–192. (b) Walsh, C.T.

j61


j 2 Directed Evolution as a Means to Engineer Enantioselective Enzymes

62

89

90

91

92

93

94

95 96

97

98

and Chen, Y.-C.J. (1988) Angewandte Chemie, 100, 342–352 (1988) Angewandte Chemie International Edition in English, 27, 333–343. (c) Sheng, D., Ballou, D.P. and Massey, V. (2001) Biochemistry, 40, 11156–11167. Reetz, M.T., Brunner, B., Schneider, T., Schulz, F., Clouthier, C.M. and Kayser, M.M. (2004) Angewandte Chemie, 116, 4167–4170 (2004) Angewandte Chemie International Edition, 43, 4075–4078. Mihovilovic, M.D., Rudroff, F., Winninger, A., Schneider, T., Schulz, F. and Reetz, M.T. (2006) Organic Letters, 8, 1221–1224. Malito, E., Alfieri, A., Fraaije, M.W. and Mattevi, A. (2004) Proceedings of the National Academy of Sciences of the United States of America, 101, 13157–13162. Fraaije, M.W., Wu, J., Heuts, D.P.H.M., van Hellemond, E.W., Spelberg, J.H.L. and Janssen, D.B. (2005) Applied Microbiology and Biotechnology, 66, 393–400. Bocola, M., Schulz, F., Leca, F., Vogel, A., Fraaije, M.W. and Reetz, M.T. (2005) Advanced Synthesis and Catalysis, 347, 979–986. Schulz, F., Leca, F., Hollmann, F. and Reetz, M.T. (2005) Beilstein Journal of Organic Chemistry, 1, 10. Hollmann, F., Taglieber, A., Schulz, F. and Reetz, M.T. in preparation. (a) Griffin, M. and Trudgill, P.W. (1972) Biochemical Journal, 129, 595–603. (b) Adger, B., Bes, M.T., Grogan, G., McCague, R., Pedragosa-Moreau, S., Roberts, S.M., Villa, R., Wan, P.W.H. and Willetts, A.J. (1997) Bioorganic and Medicinal Chemistry, 5, 253–261. (c) Mihovilovic, M.D., M€ uller, B., Kayser, M.M., Stewart, J.D. and Stanetty, P. (2002) Synlett, 703–706. Alphand, V., Gaggero, N., Colonna, S., Pasta, P. and Furstoss, R. (1997) Tetrahedron, 53, 9695–9706. Dostert, P.L., Benedetti, M.S. and Tipton, K.F. (1989) Medicinal Chemistry Research, 9, 45–89.

99 Beard, T.M. and Turner, N.J. (2002) Chemical Communications (Cambridge U.K.), 246–247. 100 Alexeeva, M., Enright, A., Dawson, M.J. and Mahmoudian, M., Turner, N.J. (2002) Angewandte Chemie, 114, 3309–3312; (2002) Angewandte Chemie International Edition, 41, 3177–3180. 101 Carr, R., Alexeeva, M., Enright, A., Eve, T.S.C., Dawson, M.J. and Turner, N.J. (2003) Angewandte Chemie, 115, 4955–4958 (2003) Angewandte Chemie International Edition, 42, 4807–4810. 102 Dunsmore, C.J., Carr, R., Fleming, T. and Turner, N.J. (2006) Journal of the American Chemical Society, 128, 2224–2225. 103 (a) Sono, M., Roach, M.P., Coulter, E.D. and Dawson, J.H. (1996) Chemical Reviews, 96, 2841–2888. (b) Urlacher, V. and Schmid, R.D. (2002) Current Opinion in Biotechnology, 13, 557–564. (c) Meunier, B., de Visser, S.P. and Shaik, S. (2004) Chemical Reviews, 104, 3947–3980. 104 Holland, H.L. (1998) Biotechnology, 2nd edn. (eds H.-J. Rehm, G Reed, A. P€ uhler and P. Stadler), Wiley-VCH Verlag GmbH, Weinheim, Vol. 8a, pp. 475–533. 105 (a) Farinas, E.T., Schwaneberg, U., Glieder, A. and Arnold, F.H. (2001) Advanced Synthesis and Catalysis, 343, 601–606. (b) Glieder, A., Farinas, E.T. and Arnold, F.H. (2002) Nature Biotechnology, 20, 1135–1139. (c) Peters, M.W., Meinhold, P., Glieder, A. and Arnold, F.H. (2003) Journal of the American Chemical Society, 125, 13442–13450. 106 M€ unzer, D.F., Meinhold, P., Peters, M.W., Feichtenhofer, S., Griengl, H., Arnold, F.H., Glieder, A. and de Raadt, A. (2005) Chemical Communications (Cambridge U.K.), 2597–2599. 107 Landwehr, M., Hochrein, L., Otey, C.R., Kasrayan, A., B€ackvall, J.-E. and Arnold, F.H. (2006) Journal of the American Chemical Society, 128, 6058–6059. 108 (a) Kubo, T., Peters, M.W., Meinhold, P. and Arnold, F.H. (2006) Chemistry-A European Journal, 12, 1216–1220. (b) Budde, M., Morr, M., Schmid, R.D. and


References Urlacher, V.B. (2006) Chembiochem: A European Journal of Chemical Biology, 7, 789–794. 109 (a) Hancock, S.M., Vaughan, M.D. and Withers, S.G. (2006) Current Opinion in Chemical Biology, 10, 509–519. (b) Dion, M., Nisole, A., Spangenberg, P., Andre, C., Glottin-Fleury, A., Mattes, R., Tellier, C. and Rabiller, C. (2001) Glycoconjugate Journal, 18, 215–223. 110 (a) Vedha-Peters, K., Gunawardana, M., Rozzell, J.D. and Novick, S.J. (2006) Journal of the American Chemical Society, 128, 10923–10929. (b) Baik, S.-H., Ide, T., Yoshida, H., Kagami, O. and Harayama, S. (2003) Applied Microbiology and Biotechnology, 61, 329–335. (c) Morikawa, S., Nakai, T., Yasohara, Y., Nanba, H., Kizaki, N. and Hasegawa, J. (2005) Bioscience Biotechnology and Biochemistry, 69, 544–552. (d) Bali, S., O’Hare, H.M. and Weissman, K.J. (2006) Chembiochem: A European Journal of Chemical Biology, 7, 478–484. 111 (a) James, L.C. and Tawfik, D.S. (2003) Trends in Biochemical Sciences, 28, 361–368. (b) Bornscheuer, U.T. and Kazlauskas, R.J. (2004) Angewandte Chemie, 116, 6156–6165; (2004) Angewandte Chemie International Edition, 43, 6032–6040. (c) Aharoni, A., Gaidukov, L., Khersonsky, O., McQ Gould, S., Roodveldt, C. and Tawfik, D.S. (2005) Nature Genetics, 37, 73–76.

112 See for example: (a)Funke, S.A., Otte, N., Eggert, T., Bocola, M., Jaeger, K.-E. and Thiel, W. (2005) Protein Engineering, Design and Selection, 18, 509–514. (b) Kato, R., Nakano, H., Konishi, H., Kato, K., Koga, Y., Yamane, T., Kobayashi, T. and Honda, H. (2005) Journal of Molecular Biology, 351, 683–692. (c) Moore, G.L. and Maranas, C.D. (2004) AIChE Journal, 50, 262–272. (d) Weinreich, D.M., Delaney, N.F., DePristo, M.A. and Hartl, D.L. (2006) Science, 312, 111–114. (e) Fox, R., Roy, A., Govindarajan, S., Minshull, J., Gustafsson, C., Jones, J.T. and Emig, R. (2003) Protein Engineering, 16, 589–597. (f) Nakahara, T., Hindsgaul, O., Palcic, M.M. and Nishimura, S.-I. (2006) Protein Engineering, Design and Selection, 19, 571–578. (g) DiTursi, M.K., Kwon, S.-J., Reeder, P.J. and Dordick, J.S. (2006) Protein Engineering, Design and Selection, 19, 517–524. (h) Bogarad, L.D. and Deem, M.W. (1999) Proceedings of the National Academy of Sciences of the United States of America, 96, 2591–2595 (i) Mayer, M.M., Hochrein, L. and Arnold, F.H. (2006) Protein Engineering, Design and Selection, 19, 563–570. 113 (a) Reetz, M.T., Rentzsch, M., Pletsch, A. and Maywald, M. (2002) Chimia, 56, 721–723. (b) Reetz, M.T. (2001) patent application DE-A 101 29 187.6 (Optimization of Synthetic Catalysts by Means of Directed Evolution).

j63


j65

3 The Search for New Enzymes Jean-Louis Reymond and Wolfgang Streit

3.1 Introduction

Other than the opportunity offered by optimizing a biocatalytic process by engineering and modifying existing enzymes by directed evolution methods [1], one can look for an entirely new enzyme to solve the problem at hand. This option might open unforeseen advantages and allow a total redesign of a process if a new reaction type is found. It might also be necessary if an intellectual property limitation is encountered in a given process. In principle, there are two options to search for a new enzyme: (i) design a new enzyme starting with the knowledge of a reaction mechanism; and (ii) the search for a yet undiscovered enzyme in biodiversity, in particular, through mining the metagenome. Overall, these two approaches represent radically different points of view to the same challenge, yet follow similar workflows via recombinant or synthetic libraries and high-throughput screening for function (Figure 3.1) [2]. In the enzyme design approach, as discussed in the first part of this chapter, one attempts to utilize the mechanistic understanding of chemical reactions and enzyme structure to create a new catalyst. This approach represents a largely academic research field aiming at fundamental understanding of biocatalysis. Indeed, the invention of functional artificial enzymes can be considered to be the ultimate test for any theory on enzyme mechanisms. Most artificial enzymes, to date, do not fulfill the conditions of catalytic efficiency and price per unit necessary for industrial applications. The biodiversity mining approach focuses on expressing possible enzymes found in microbial collections and metagenomic expression libraries [3]. In contrast to the rather academic setting of the enzyme design approach, the screening of microbial collections and the metagenome is integrated in both academia and industry, because it represents a very practical method to solve a catalysis problem whenever a new enzyme is necessary. The approach is particularly successful when new enzymes are searched to carry out a known reaction type on a new type of substrate, or with a different stereoselectivity compared to already known enzymes. In such cases,


j 3 The Search for New Enzymes

66

Synthetic libraries Reaction mechanisms Recombinant libraries

HTS assays

New enzymes

Biodiversity Figure 3.1 The search for new enzymes.

screening enzyme collections such as those offered by a range of enzyme companies is a practical ďŹ rst step. In this chapter, we focus particularly on the metagenome because it represents a generalization of biodiversity mining by extending its boundary to all life forms, including in particular, uncultivable microorganisms.

3.2 Mechanism-based Enzyme Design

The mechanism of any given reaction provides information about the catalytic groups and nature of the transition state involved [4]. This information can be implemented into a structural framework such as a protein or peptide, which is either expressed in recombinant format or prepared by chemical synthesis. The approach includes (i) transition state mimicry-based designs such as catalytic antibodies [5] and the related computational redesign of binding pockets around a transition state based on de novo protein design algorithms [6,7]; (ii) the rational design of new catalysts on protein and enzyme basis, either by design [8], searching for promiscuous activities in existing enzymes by substrate and/or active site modiďŹ cations [9], by cofactor engineering [8,10], or by installing a catalytic metal complex into a noncatalytic protein [11]; and (iii) the search for catalytic activity in synthetic nonprotein macromolecules such as synthetic peptides [12] and dendrimers [13]. 3.2.1 Catalytic Antibodies

Catalytic antibodies were ďŹ rst reported by Lerner and Schultz in the 1980s [5]. Catalytic antibodies are obtained by immunizing a laboratory animal using a stable transition state analog (TSA) of a chemical reaction as hapten. Following the basic principle that chemical catalysis is equivalent to binding to the transition state of a chemical reaction, one can expect that an antibody (Ab) forming a complex with the TSA for a reaction from S to P, with a dissociation constant Ki, might form a similarly tight complex with the real transition state of the reaction (TS ), with a dissociation constant KTS, and therefore provide rate acceleration for the reaction (G #) (Figure 3.2) [14]. TSA design considers the charge distribution and shape of the transition state


3.2 Mechanism-based Enzyme Design

TS* ∆∆G AbTS* S

P

K TS

Ki Ab

AbTS*

AbTSA TSA

TS*

Figure 3.2 Transition state analog principle for the design of catalytic antibodies.

along the same lines used to design TSA enzyme inhibitors. Covalent immunization using a reactive hapten to trap a catalytic side chain on the antibody, which borrows from the concept of suicide inhibitors of enzymes, has also been used with success [15]. Catalytic antibodies are particularly interesting, because they allow one to tackle reaction types not encountered in natural enzymes. One such example is the enantioselective protonation of enol ether (1) to form a single enantiomer of ketone (2) (Figure 3.3). This reaction was found to be catalyzed efficiently and with complete enantioselectivity in favor of the (S) enantiomer by antibody 14D9, a monoclonal antibody produced by a hybridoma cell line resulting from an immunization using hapten (3) as antigen [16]. The hapten design and reaction mechanism of antibody 14D9 is representative of the catalytic antibody approach to new enzymes. Hapten (1) combines a strongly immunogenic aromatic group that can induce substrate recognition with a quaternary ammonium group to induce a hydrophobic binding pocket equipped with a carboxylic acid to be used as catalytic functional Ab 14D9 pH 6, 20 °C MeO

O H

kcat = 0.4 s –1; kcat /kuncat = 104

Ar

Ar

1

(S)-2 Carrier protein Me

N

> 98 % ee

= Ar H N

OH

O 3 Figure 3.3 Antibody 14D9 catalyzes the enantioselective protonation of enol ethers.

j67


j 3 The Search for New Enzymes

68

group for the reaction. The enantioselective protonation of enol ether (1) by antibody 14D9 tolerates a variety of substrates [17], and is sufficiently high to allow preparative scale reactions on the gram scale [18]. The molecular mechanism of the enantioselective protonation reaction by antibody 14D9 was revealed by a crystal structure analysis [19]. A catalytic carboxyl group AspH101 was found at the bottom of the catalytic pocket and found to be necessary for catalysis by mutagenesis to Asn or Ala. The mechanism or protonation involves an overall syn addition of water to the enol ether in a chiral binding pocket ensuring complete enantioselectivity (Figure 3.4). Like many other antibodies, the activity of antibody 14D9 is sufficient for preparative application, yet it remains modest when compared to that of enzymes. The protein is relatively difficult to produce, although a recombinant format as a fusion with the NusA protein was found to provide the antibody in soluble form with good activity [20]. It should be mentioned that aldolase catalytic antibodies operating by an enamine mechanism, obtained by the principle of reactive immunization mentioned above [15], represent another example of enantioselective antibodies, which have proven to be preparatively useful in organic synthesis [21]. One such aldolase antibody, antibody 38C2, is commercially available and provides a useful alternative to natural aldolases to prepare a variety of enantiomerically pure aldol products, which are otherwise difficult to prepare, allowing applications in natural product synthesis [22]. Ar

Ar

H O H O

TyrH35

H O O H H O H O

O H O

TyrH35

AspH101

Ar H

TyrH35

AspH101 TyrL36

TyrL36

O H O

O H O H O H O

H O

O O H

H O

AspH101 TyrL36

Figure 3.4 Protonation mechanism for 14D9.


3.2 Mechanism-based Enzyme Design

3.2.2 Rational Design of New Catalysts on Enzyme and Protein Basis

One of the most direct questions to ask in the perspective of enzyme design is whether an already existing protein with a binding pocket might be turned into a new catalyst by introducing catalytic residues directly, rather than by the elaborated TSA mimicry approach used for catalytic antibodies, hoping to create a new biocatalyst that could harness both the activity and the selectivity, in particular stereoselectivity, that is possible with enzymes. In principle, numerous reports have detailed the possibility to modify an enzyme to carry out a different type of reaction than that of its attributed function, and the possibility to modify the cofactor of the enzyme has been well explored [8,10]. Recently, the possibility to directly observe reactions, normally not catalyzed by an enzyme when choosing a modiďŹ ed substrate, has been reported under the concept of catalytic promiscuity [9], a phenomenon that is believed to be involved in the appearance of new enzyme functions during the course of evolution [23]. A recent example of catalytic promiscuity of possible interest for novel biotransformations concerns the discovery that mutation of the nucleophilic serine residue in the active site of Candida antarctica lipase B produces a mutant (Ser105Ala) capable of efďŹ ciently catalyzing the Michael addition of acetyl acetone to methyl vinyl ketone [24]. The oxyanion hole is believed to be complex and activate the carbonyl group of the electrophile, while the histidine nucleophile takes care of generating the acetyl acetonate anion by deprotonation of the carbon (Figure 3.5).

O

O

Gln106 O O N H H N H O

CAL-B Ser105Ala

O

Thr40

+ O

O

O O H N

Asp187

N H

O

His224

O

O

O Figure 3.5 Michael addition catalyzed by the Ser105Ala C. antarctica lipase B mutant.

j69


j 3 The Search for New Enzymes

70

O

BSA, Cu(I)-phthalocyanin

N

H H

+ Formate buffer pH 4

O N 77% yield 92% de exo 93% ee

Figure 3.6 Enantioselective Diels–Alder reaction catalyzed by BSA and a Cu-phthalocyanin [26].

It has also been observed that noncatalytic proteins can be brought to exhibit unusual catalytic properties. Serum albumins such as bovine serium albumin (BSA), for instance, promote a range of reactions such as the base-catalyzed decomposition of benzisoxazole, which has been studied in great detail [25]. BSA possesses a binding pocket for large hydrophobic aromatic compounds and provides a chiral environment suitable for other types of catalysis as well, in particular, for porphyrin-type metal complexes for a range of metal-catalyzed processes such as reductions, oxidations, and Diels–Alder cycloadditions, as recently reviewed [11b]. One such recent example is the combination of copper phthalocyanine with BSA to produce a new biocatalyst catalyzing a highly enantioselective Diels–Alder cycloaddition reaction [26]. Interestingly, a similar highly enantioselective system has also been reported combining the copper catalyst with DNA as a source of chirality [27]. These examples are part of a broader design scheme to combine catalytic metal complexes with a protein as chiral scaffold to obtain a hybrid catalyst combining the catalytic potential of the metal complex with the enantioselectivity and evolvability of the protein host [11]. One of the first examples of such systems combined a biotinylated rhodium complex with avidin to obtain an enantioselective hydrogenation catalyst [28]. Most significantly, it has been shown that mutation-based improvements of enantioselectivity are possible in these hybrid catalysts as for enzymes (Figure 3.7) [29]. Streptavidin 5 atm H 2

CO2H

CO2H

NHAc

NHAc Quantity yield 94% ee (R)

O HN

P

NH S

N

RhH2 P

O

Figure 3.7 The biotin–streptavidin system for enantioselective hydrogenation.


3.3 Metagenomics

3.2.3 Synthetic Enzyme Models

The field of synthetic enzyme models encompasses attempts to prepare enzymelike functional macromolecules by chemical synthesis [30]. One particularly relevant approach to such enzyme mimics concerns dendrimers, which are treelike synthetic macromolecules with a globular shape similar to a folded protein, and useful in a range of applications including catalysis [31]. Peptide dendrimers, which, like proteins, are composed of amino acids, are particularly well suited as mimics for proteins and enzymes [32]. These dendrimers can be prepared using combinatorial chemistry methods on solid support [33], similar to those used in the context of catalyst and ligand discovery programs in chemistry [34]. Peptide dendrimers used multivalency effects at the dendrimer surface to trigger cooperativity between amino acids, as has been observed in various esterase enzyme models [35]. An interesting case in the perspective of artificial enzymes for enantioselective synthesis is the recently described peptide dendrimer aldolases [36]. These dendrimers utilize the enamine type I aldolase mechanism, which is found in natural aldolases [37] and antibodies [21]. These aldolase dendrimers, for example, L2D1, have multiple N-terminal proline residues as found in catalytic aldolase peptides [38], and display catalytic activity in aqueous medium under conditions where the small molecule catalysts are inactive (Figure 3.8). As most enzyme models, these dendrimers remain very far from natural enzymes in terms of both activity and selectivity, and at present should only be considered in the perspective of fundamental studies.

3.3 Metagenomics

Metagenomics has dramatically increased the speed of discovery of novel biocatalysts by enabling scientists to tap directly into the entire diversity of enzymes held within natural microbial populations. The genomes of as yet uncultured microbes represent a shear unlimited source for novel genes and chemical compounds, to be used by the biotechnology industries for the development of novel products [3]. The dilemma of being unable to cultivate the vast majority of the microorganisms present in most microbial niches has forced scientists to develop methods to exploit the genomes of these bacteria for biotechnology without prior cultivation. The relatively young field of metagenomics developed from advances made in DNA extraction and cloning from environmental samples [39,40]. The key technologies have been outlined in a landmark publication in 1991. This development coincided with the observation that prokaryotic diversity had been greatly underestimated and it became apparent that there remained a great deal of biocatalytic potential locked within the unculturable majority of prokaryotes [z3,41a]. Current computed estimates of soil diversity are in the range of 1 million species per 1 g of soil [42]. Metagenomics now enables access to the biocatalytic potential of unculturable bacteria [43]. Metagenomics pursues two lines of activities: first, it uncovers novel enzymes and molecules for biotechnological

j71


j 3 The Search for New Enzymes

72

O

OH O

Dendrimer L2D1 (1 mol%) H

O2 N

O2N 61% ee HN

+

NH NH

O O

H 3N O NH

O

O H N

N

+

H N

HN

NH

O

O

NH

H 3N

H N

NH 3+ +

NH

NH

H N

O NH

NH O

NH 2 O

O

NH HN

H 3N

N H

HN

H N

O

O +

O

OH O N

O

HN

H N

O

OHN

NH 3 +

NH

O

O

N H

O

+

O

N

N H

NH

NH3 + N H

O O

O O NH

H3N

HN O

O

O

HN

HN NH3 +

NH 3+

NH N H

NH 3+

O

OH

O O

O

N HN

H3N HN

O HN

H N

H N

O

O O

NH 3+

NH L2D1

Figure 3.8 Aldol reaction and peptide dendrimer catalyst.

and pharmaceutical applications; and secondly, it generates new knowledge on the microbial ecology of the studied niches. In this review, we discuss only the biotechnological aspects of metagenomics that are related to the identiďŹ cation of speciďŹ c enzymes and genes. 3.3.1 Construction of Metagenome-derived DNA Libraries 3.3.1.1 Selection of the Environment A range of novel enzyme detection strategies have been developed, which target different stages in the process of metagenomic library construction and screening. The success of a metagenome-based search for enzymes depends, in general, on the


3.3 Metagenomics

selection of the environment. In general, the published studies can be grouped into three categories. The first group selects environments naturally enriched for the target biocatalyst such as the search for xylanases in insect gut [44]. The second group selects highly genomically diverse environments such as soil and either directly extracts and clones the DNA or subjects the community first to enrichment and then proceeds with extraction and cloning. The third group targets extreme environments and searches for biocatalysts that are stable under the conditions experienced by the microorganisms in the environment [45]. 3.3.1.2 Cloning Strategies Other than the careful selection of the environment, the cloning strategy is also of considerable importance. The overall cloning strategy of metagenome-derived DNA fragments is outlined in Figure 3.9. In general, the metagenomic DNA is cloned into either large insert or small insert libraries. Large insert libraries can be generated in cosmids [46,47], bacterial artificial chromosomes (BACs)[48], or fosmids [45,49]. Small insert libraries are maintained in plasmid vectors including pBluescript SKþ [50], pUC19 [51], pZero-2 [52], and ZAP phagemid vector [45]. In the case of pZero-2, a lethal gene is inactivated by the incorporation of environmental DNA, which keeps the level of clones containing a self-ligated vector to a minimum [52]. Small insert libraries contain on average ten times more clones than large insert libraries covering the same amount of environmental DNA. Although the reduced clone number makes large insert libraries easier to screen, the larger insert size and lower copy number makes detection of weakly expressed foreign genes more difficult [42b]. 3.3.1.3 Screening and Detection Technologies With respect to the technologies used for screening, two approaches are used: functional searches and sequence-based searches. Sequence-based searches are always more conservative, but will certainly benefit from the vast amount of data produced by the environmental sequencing projects [53]. However, functional screening strategies aim to detect the biocatalytic reaction first and then characterize the gene responsible. In this way completely novel enzymes can be discovered. This approach requires that the metagenomic DNA be expressed in a heterologous host, usually Escherichia coli. As can be expected with an expression-based system, it is limited when factors required for transcription and translation of the metagenomic DNA are not present in the host. A survey of 32 prokaryotic genomes found that only 40% of genes could be expressed using E. coli [45]. Also, even when a gene is expressed, the level of expression can be so low that it cannot be detected using functional screening. Recently, a transposon that enables inducible expression over T7 promoters in both directions was developed [54]. This transposon could be used to enhance gene expression and increase the likelihood of finding metagenomic enzymes through functional screening. The heterologous hosts used in metagenomics are E. coli, Pseudomonas putida, Streptomyces lividans, and Rhizobium leguminosarum [55,56]. In each of these studies, it was clearly demonstrated that the expression of metagenomic protein is limited owing to a limited number of hosts in most laboratories. Finally, it is noteworthy that the detection systems that are used

j73


j 3 The Search for New Enzymes

74

Figure 3.9 Cloning and DNA–isolation strategies commonly used in metagenomics.

are also of importance for the frequency of gene detection. A very promising system is the SIGEX system. In the SIGEX system, the environmental DNA is cloned into a vector containing a gfp reporter [57]. Clones carrying the catabolic genes are activated in the presence of the target substrate. When the catabolite genes are expressed, they in turn activate the gfp reporter causing the cell to fluoresce. This, in combination with fluorescence activating cell sorting (FACS), greatly increases the number of recombinant clones that can be screened [57]. 3.3.1.4 Major Problems that Need to be Addressed The rapidly increasing number of novel metagenome-derived genes also led to the identification of major bottlenecks. These are mainly linked to the efficiency of heterologous gene expression and the amount of protein produced, in particular,


3.3 Metagenomics

when subsequent biotechnological applications are envisaged. Therefore, a need exists to develop novel experimental systems for high-throughput expression of metagenomic genes and purification of the respective proteins. Additional novel expression host strains must be developed, which should include both prokaryotic and eukaryotic organisms. In any case, the enormous potential of metagenomics for modern biotechnology is obvious, but further and quick advances of technologies are necessary to fully exploit the genetic information of nonculturable microbes as a virtually unlimited resource for biotechnology. Among these technologies, it is mostly the screening that limits the identification of novel biocatalysts. The urgent need to develop screening technologies for the reactions that have been unusual and novel will hopefully help metagenomics to become an even more important key technology in modern biotechnology. 3.3.2 The Genomes of Not Yet Cultured Microbes as Resources for Novel Genes

Metagenomic libraries have been screened for a wide range of biocatalysts [3]. However, many of the published and metagenome-derived enzymes/genes have not been characterized to such an extent that would allow their immediate use in biotechnological applications. In the following section, we highlight examples of the successful isolation enzymes and other valuable biomolecules from metagenomes. 3.3.2.1 Polysaccharide Degrading/Modifying Enzymes Agarases are the enzymes that can liquefy agar, and they can be divided into and agarases, depending on whether they cleave the -L-(1,3) or the -D-(1,4) linkages of the polymer. Although agarases are normally only found in marine microbes, the screening of a soil metagenomic library identified a total of four agarolytic clones containing 12 agarase genes [58]. This highlights the potential of using metagenomics to investigate environments for novel biocatalysts, which are normally detected in isolates from unrelated environments. Amylases have also been in the focus of a number of metagenome studies. At least five articles reporting the detection and characterization of novel amylolytic enzymes from metagenomic DNA libraries have been published [58–61]. Of the 14 amylolytic clones only four have been purified and characterized. One of the characterized amylases, from a soil metagenome, displayed interesting characteristics in that it is stable and active under alkaline conditions, with a pH optimum at pH 9, a characteristic required of amylases in detergents [61,62]. The other three amylases were active under acidic pH and high temperature conditions, and these were then subjected to gene reassembly in order to obtain one enzyme displaying optimal properties [60]. Functional screening of a soil metagenomic library for cellulases revealed a total of eight cellulolytic clones, one of which was purified and characterized [58]. Despite the fact that this library had been generated from a soil sample collected from a

j75


j 3 The Search for New Enzymes

76

Figure 3.10 Typical screening plate for cellulase and amylase positive clones from enrichment cultures. Clear halos indicate the presence of positive clones and the halos diameters give a first idea about the overall activities of the respective enzymes.

nonextreme environment, the cellulase displayed a high level of stability over a broad pH range, up to pH 9, it was stable at 40  C for up to 11 hours, and was highly halotolerant, being active and stable in 3 M NaCl [58]. Metagenomic screening of extreme environments, soda lakes in Africa and Egypt, detected more than a dozen cellulases, some of which displayed habitat-related halotolerant characteristics [63]. One of the earliest articles presenting metagenome-derived biocatalysts reported the detection of cellulases from a thermophilic, anaerobic digester fueled by lignocellulose [64]. While most metagenomic surveys for novel cellulases concentrate on extreme environments, there is sufďŹ cient evidence that nonextreme, and therefore highly genetically diverse, environments also contain a range of cellulases that are highly stable and suitable for industrial applications [65]. Chitinases are responsible for the breakdown and recycling of chitin. With several gigatons of chitin being produced annually, it is the second most abundant polymer


3.3 Metagenomics

in nature after cellulose [66]. Chitinases have several biotechnological applications including increasing plant resistance to fungal disease [67]. To date, there has been only a few metagenomic studies targeting chitinases from marine environments [68]. The identified enzymes need to be characterized on a more detailed level before application in biotechnological processes. Complete hydrolysis of xylan to xylose involves the activity of xylosidases. There have been three studies detecting xylanases/xylosidases in the metagenomes of diverse environments: an insect gut, a thermophilic, anaerobic digester, and waste lagoon of a dairy farm [44,64]. Screening of the lagoon resulted in the detection and characterization of one xylanase, which displayed habitat-related properties in that it was most active at lower temperatures [69]. The four xylosidases detected in the digester were not investigated further, but the cellulases found in the same study were thermophilic [64]. The metagenomic survey of the insect gut discovered four xylanases, which were phylogenetically distant from all other known xylanases, suggesting that they had evolved independently [44]. These xylanases are novel and produce unique hydrolysis products. 3.3.2.2 Lipolytic Biocatalysts The development of metagenomics has greatly accelerated the discovery of novel biocatalysts, many displaying unusual properties. To date, at least 76 esterase or lipase positive clones derived from metagenomes have been reported (Table 3.1). The level of characterization of these novel lipolytic genes ranges from DNA restriction and sequencing analysis to determine clone diversity [48,58] to detailed biochemical analysis of the purified enzyme [45,75]. Of the 76 lipolytic positive clones listed, only 11 have been overexpressed, purified, and subjected to detailed biochemical characterization. It is because of the high number of novel enzymes, which can be detected by the screening of a single metagenomic library, and the amount of time and effort required to fully characterize one enzyme, which causes this bottleneck. Among the characterized metagenomic esterases found so far are two of the largest esterases known: a 325-kDa esterase from a deep-sea hypersaline anoxic basin, and 336-kDa octameric esterase from a drinking water biofilm [45,74]. The esterases from the deep-sea hypersaline basin display habitat-related properties in that they were most active at alkaline pH and displayed higher activities under high-pressure conditions [75]. The esterases from soil and a drinking water biofilm displayed unusual properties, which could not be related to their environment [75]. They were highly stable at alkaline pH and displayed unique substrate spectra with EstA3 being able to hydrolyze substrates such as 7-[3-octylcarboxy-(3-hydroxy-3-methyl-butyloxy)]coumarin, a normally unreactive secondary ester [3]. 3.3.2.3 Vitamin Biosynthesis Metagenomics has been applied to the search for novel genes encoding the synthesis of vitamins such as biotin and vitamin C [47,76]. Seven cosmids were detected in metagenomic libraries obtained after avidin enrichment of environmental samples. The highest levels of biotin production in this study were detected in a cosmid

j77


j 3 The Search for New Enzymes

78

Table 3.1 Lipolytic genes identified in metagenome screens.

Source

Vector

Total size of DNA screened

Number of positive clones

Number of clones screened

Soil

Plasmid Plasmid Plasmid BAC Cosmid Phagemid Phagemid Fosmid Plasmid Fosmid Fosmid Phagemid

474.5 Mb 1170 Mb 637 Mb 100 Mb 49.8 Mb 360 Mb 600 Mb 1179.5 Mb 114 Mb 40 Mb 408.6 Mb 2000 Gb

1 1 2 2 1 2# 1 8 11 4 10 5

730 000 180 000 49 000 1824 1532 30 000 100 000 4213 2727 200 1021 —

[72] [51] [73] [74] [45]

Phagemid Fosmid

77 Mb 210 Mb

11 5

1273 1200

[45] [49]

Cosmid Cosmid Fosmid

56 Mb 87.5 Mb 13.5 Gb

6 1 6

267 2500 64 333

[75]

Soil Soil Kenyan soda lakes Topsoil Pondwater Hot spring Hot spring Deep-sea hypersaline basin Cow gut Mud flats, beach, and forest Drinking water Soil Korean tidal flat

Reference [70] [48] [66] [71]

[69]

obtained from forest soil [47]. In the search for enzymes involved in vitamin C production from glucose, two novel 2,5-diketo-D-gluconic acid reductases were PCR amplified from the environment [76]. 3.3.2.4 Nitrilases, Nitrile Hydratases, and Amidases Chemical hydrolysis of nitriles requires the presence of strong acids or bases at high temperatures and normally results in low yields. Therefore, there is much interest in using biocatalysts to carry out this reaction instead [77]. Nitrile hydratases are used for the production of acrylamide and the vitamin nicotinamide [78]. In one metagenomic study, involving the screening of 3-Gb DNA, 12 novel nitrile dehydratases were detected [79]. In another general screening of a soil metagenomic library for biocatalysts, one amidase positive clone was detected [58]. Amidases are used in the biosynthesis of ß-lactam antibiotics, and in a study targeting amidases of the soil metagenome using enrichment seven amidase positive clones were detected, one of which encoded a novel penicillin acylase [52,80]. Nitrilases are quite rare in bacterial genomes and less than 20 were reported prior to the application of metagenomics for their detection in environmental DNA [81]. Two studies targeting environmental genomes report the detection of more than 337 novel nitrilases. This has dramatically increased the amount of information about nitrilases, and the newly discovered diversity can be applied for the enantioselective production of hydroxy carboxylic acid derivatives [81].


3.3 Metagenomics

3.3.2.5 Oxidoreductases/Dehydrogenases A metagenomic study searching for the diversity of bacteria in the environment capable of utilizing 4-hydroxybutyrate found five clones displaying novel 4-hydroxybutyrate dehydrogenase activity [82]. In a recent metagenomic study, the genes involved in metabolism of poly-3-hydroxybutyrate, a compound being considered as a substitute for fossil fuel-derived polymers, were screened for environmental libraries [83]. They found novel short chain dehydrogenases/reductases, which had <35% similarity to known enzymes and thus could not have been detected using hybridization-based techniques such as PCR. Alcohol oxidoreductases capable of oxidizing short chain polyols are useful biocatalysts in industrial production of chiral hydroxy esters, hydroxy acids, amino acids, and alcohols [83]. In a metagenomic study without enrichment, a total of 24 positive clones were obtained and tested for their substrate specificity. To improve the detection frequency, enrichment was performed using glycerol or 1,2-propanediol and further 24 positive clones were detected in this study. 3.3.2.6 Proteases Proteases are hydrolytic enzymes with important application in industries, in particular, in detergent and in the food industry. A metagenomic study in which 100 000 plasmid clones were screened for proteolytic activity found one positive clone, which was determined to be novel by sequencing analysis [84]. 3.3.2.7 Glycerol Hydratases In industrial processes, 1,3-propanediol is used for the production of polyester fibers, polyurethanes and cyclic compounds [85]. 1,3-Propanediol can be produced from glucose with the limiting step catalyzed by glycerol dehydratase. A metagenomic survey for glycerol hydratases from the environment resulted in seven positive clones, one of which displayed a level of catalytic efficiency and stability making it ideal for application in the production of 1,3-propanediol from glucose. 3.3.2.8 Antibiotics and Pharmaceuticals There have been at least two studies reporting the successful screening of soil metagenomic libraries for indirubin, a microbial product used in the treatment of diseases including leukemia [86]. A range of novel antibiotics have been detected in metagenomic libraries [87]. Palmitoylputrescine was detected in metagenomic DNA from a Bromeliad tank [87] and an isocyanide containing antibiotic was found in a soil metagenomic library [87]. Screening of seven different soil metagenomic libraries revealed 11 clones producing long-chain N-acyltyrosine antibiotics and analysis of their syntheses indicated that 10 of them were novel [87]. All the metagenomederived novel antibiotics detected to date have been found in libraries maintained in E. coli hosts except the terragines, which were in a streptomycete host [56]. In addition to these studies, plenty of information has been gained about the diversity of natural antibiotic resistance mechanisms [88]. The search for novel antimicrobial genes and compounds is probably the field that is advancing most rapidly in metagenomics.

j79


j 3 The Search for New Enzymes

80

3.4 Conclusion

The search for new enzymes is clearly being driven into two parallel and complementary directions. On the one hand, mechanism-based enzyme design is being pursued to push our fundamental understanding of enzyme catalysis. This field provides a driving force in the area of synthetic bioorganic and macromolecular chemistry. As commented in the introduction, most examples of ‘designed’ artificial enzymes fall short of natural enzymes in terms of catalytic efficiency. One must acknowledge that enzyme design remains a largely unsolved problem, which makes it all the more interesting as a challenge capable of yielding significant scientific advances. These might arise by advancing theoretical understanding of structure–function relationship and catalysis itself, or from developing more efficient experimental protocols for mimicking natural selection/evolution in the laboratory. In parallel, at present, metagenomics embodies a complementary side in the search for new enzymes, where the scientific quest is not about refined understanding but rather in exploring what nature has to offer. In truth, the revelation of the enormous genetic diversity in the biosphere beyond all that was imagined justifies the expectation that a very large number of discoveries await to be made in this field. A significant number of metagenome-derived genes have been identified until now. Future work will have to demonstrate whether these genes or their products can be implemented into viable biotechnological processes.

References 1 (a) Reetz, M.T. (2001) Angewandte Chemie International Edition, 40, 284–310. (b) Bloom, J.D., Meyer, M.M., Meinhold, P., Otey, C.R., MacMillan, D. and Arnold, F.H. (2005) Current Opinion in Structural Biology, 15, 447–452. 2 Reymond, J.-L. (ed.) (2006) Enzyme Assays: High-Throughput Screening, Genetic Selection and Fingerprinting, Wiley-VCH, Weinheim, Germany. 3 (a) Streit, W.R. and Schmitz, R.A. (2004) Current Opinion in Microbiology, 7, 492–498. (b) Streit, W.R., Daniel, R. and Jaeger, K.-E. (2004) Current Opinion in Biotechnology, 15, 285–290. 4 (a) Mader, M.M. and Bartlett, P.A. (1997) Chemical Reviews, 97, 1281–1302. (b) Mulholland, A.J. (2005) Drug Discovery Today, 10, 1393–1402.

5 (a) Lerner, R.A., Benkovic, S.J. and Schultz, P.G. (1991) Science, 252, 659–667. (b) Schultz, P.G. and Lerner, R.A. (1993) Accounts of Chemical Research, 26, 391–395. (c) Schultz, P.G. and Lerner, R.A. (1995) Science, 269, 1835–1842. (d) MacBeath, G. and Hilvert, D. (1996) Chemistry and Biology, 3, 433–445. (e) Reymond, J.-L. (1999) Topics in Current Chemistry, 200, 59. (f) Thomas, N.R. (1996) Natural Product Reports, 13, 479–511. (g) Stevenson, J.D. and Thomas, N.R. (2000) Natural Product Reports, 17, 535–537. 6 Dwyer, M.A., Looger, L.L. and Hellinga, H.W. (2004) Science, 304, 1967–1971. 7 (a) Bolon, D.N., Voigt, C.A. and Mayo, S.L. (2002) Current Opinion in Chemical Biology, 6, 125–129. (b) Kuhlman, B. and Baker, D. (2004) Current Opinion in Structural


References

8

9 10

11

12

13

14

15 16

17

18

19

Biology, 14, 89–95. (c) Dwyer, M.A. and Hellinga, H.W. (2004) Current Opinion in Structural Biology, 14, 495–504. (d) Moffet, D.A. and Hecht, M.H. (2001) Chemical Reviews, 101, 3191–3203. (a) Qi, D., Tann, C.-M., Haring, D. and Distefano, M.D. (2001) Chemical Reviews, 101, 3081–3112. (b) Andersson, L.K., Caspersson, M. and Baltzer, L. (2002) Chembiochem: A European Journal of Chemical Biology, 8, 3687–3689. Hult, K. and Berglund, P. (2003) Current Opinion in Biotechnology, 14, 395–400. (a) van de Velde, F., Arends, I.W. and Sheldon, R.A. (2000) Journal of Inorganic Biochemistry, 80, 81–89. (b) Wahler, D. and Reymond, J.-L. (2002) Canadian Journal of Chemistry, 80, 665–670. (a) Thomas, C.M. and Ward, T.R. (2005) Chemical Society Reviews, 34, 337–346. (b) Letondor, C. and Ward, T.R. (2006) Chembiochem: A European Journal of Chemical Biology, 7, 1845–1852. (a) Bradley, L.H., Thumfort, P.P. and Hecht, M.H. (2006) Methods in Molecular Biology, 340, 53–69. (b) Berkessel, A. (2003) Current Opinion in Chemical Biology, 7, 409–419. Kofoed, J. and Reymond, J.-L. (2005) Current Opinion in Chemical Biology, 9, 656–664. (a) Stewart, J.D. and Benkovic, S.J. (1995) Nature, 375, 388–391. (b) Reymond, J.-L. and Chen, Y. (1996) Israel Journal of Chemistry, 36, 199. Tanaka, F. and Barbas, C.F. (2002) Journal of Immunological Methods, 269, 67–79. Reymond, J.-L., Janda, K.D. and Lerner, R.A. (1992) Journal of the American Chemical Society, 114, 2257. Reymond, J.-L., Jahanghiri, G.K. and Lerner, R.A. (1993) Journal of the American Chemical Society, 115, 3909. Reymond, J.-L., Reber, J.-L. and Lerner, R.A. (1994) Angewandte Chemie International Edition in English, 33, 475. Zheng, L., Baumann, U. and Reymond, J.-L. (2004) Proceedings of the National

20

21

22

23

24

25

26

27

28

29

30

Academy of Sciences of the United States of America, 101, 3387–3392. Zheng, L., Baumann, U. and Reymond, J.-L. (2003) Journal of Applied Biochemistry, 5, 577–581. (a) Wagner, J., Lerner, R.A. and Barbas, C.F. III (1995) Science, 270, 1797. (b) Zhong, G., Lerner, R.A. and Barbas, C.F. III (1999) Angewandte Chemie International Edition, 38, 3738. (c) Tanaka, F., Lerner, R.A. and Barbas, C.F. III (2000) Journal of the American Chemical Society, 122, 4835. Sinha, S.C., Sun, J., Miller, G.P., Wartmann, M. and Lerner, R.A. (2001) Chemistry, 7, 1691–1702. Khersonsky, O., Roodveldt, C. and Tawfik, D.S. (2006) Current Opinion in Chemical Biology, 10, 498–508. Svedendahl, M., Hult, K. and Berglund, P. (2005) Journal of the American Chemical Society, 127, 17988–17999. Hollfelder, F., Kirby, A.J., Tawfik, D.S., Kikuchi, K. and Hilvert, D. (2000) Journal of the American Chemical Society, 122, 1022–1029. Reetz, M.T. and Jiao, N. (2006) Angewandte Chemie International Edition, 45, 2416–2419. Roelfes, G. and Feringa, B.L. (2005) Angewandte Chemie International Edition, 44, 3230. Wilson, M.E. and Whitesides, G.M. (1978) Journal of the American Chemical Society, 100, 306. (a) Collot, J., Gradinaru, J., Humbert, N., Skander, M., Zocchi, A. and Ward, T.R. (2003) Journal of the American Chemical Society, 125, 9030. (b) Klein, G., Humbert, N., Gradinaru, J., Ivanova, A., Gilardoni, F., Rusbandi, U.E. and Ward, T.R. (2005) Angewandte Chemie International Edition, 44, 7764. (a) Kirby, A.J. (1996) Angewandte Chemie International Edition, 35, 706–724. (b) Woggon, W.-D. (2005) Accounts of Chemical Research, 38, 127–136.

j81


j 3 The Search for New Enzymes

82

31 (a) Newkome, G.R., Moorefield, C.N. and V€ ogtle, F. (2001) Dendrimers and Dendrons. Concepts. Synthesis and Perspectives, WileyVCH, New York. (b) Tomalia, D.A. and Dvornic, P.R. (1994) Nature, 372, 617–618. (c) Liang, C.O. and Frechet, J.M.J. (2005) Progress in Polymer Science, 30, 385–402. 32 Darbre, T. and Reymond, J.-L. (2006) Accounts of Chemical Research, 39, 925–934. 33 (a) Clouet, A., Darbre, T. and Reymond, J.-L. (2004) Angewandte Chemie International Edition, 43, 4612–4615. (b) Clouet, A., Darbre, T. and Reymond, J.L. (2006) Biopolymers, 84, 114–123. 34 (a) Fonseca, M.H. and List, B. (2004) Current Opinion in Chemical Biology, 8, 319–326. (b) Stambuli, J.P. and Hartwig, J.F. (2003) Current Opinion in Chemical Biology, 7, 420–426. 35 (a) Douat-Casassus, C., Darbre, T. and Reymond, J.-L. (2004) Journal of the American Chemical Society, 126, 7817–7826. (b) Delort, E., Darbre, T. and Reymond, J.-L. (2004) Journal of the American Chemical Society, 126, 15642–15643. (c) Delort, E., NguyenTrung, N.-Q., Darbre, T. and Reymond, J.-L. (2006) Journal of Organic Chemistry, 71, 4468–4480. 36 Kofoed, J., Darbre, T. and Reymond, J.-L. (2006) Organic and Biomolecular Chemistry, 3268–3281. 37 (a) Barbas, C.F., Wang, Y.F. and Wong, C.H. (1990) Journal of the American Chemical Society, 112, 2013–2014. (b) Heine, A., Luz, J.G., Wong, C.H. and Wilson, I.A. (2004) Journal of Molecular Biology, 343, 1019–1034. 38 (a) Kofoed, J., Nielsen, J. and Reymond, J.-L. (2003) Bioorganic and Medicinal Chemistry Letters, 13, 2445–2447. (b) Martin, H.J. and List, B. (2003) Synlett, 1901–1902. (c) Krattiger, P., Kovasy, R., Revell, J.D., Ivan, S. and Wennemers, H. (2005) Organic Letters, 7, 1101–1103.

39 Schmidt, T.M., DeLong, E.F. and Pace, N.R. (1991) Journal of Bacteriology, 173, 4371–4378. 40 Tsai, Y.L. and Olson, B.H. (1992) Applied and Environmental Microbiology, 58, 2292–2295. 41 (a) Amann, R.I., Ludwig, W. and Schleifer, K.H. (1995) Microbiological Reviews, 59, 143–169. (b) Handelsman, J., Rondon, M.R., Brady, S.F., Clardy, J. and Goodman, R.M. (1998) Chemistry and Biology, 5, R245–R249 (c) Hugenholtz, P., Goebel, B.M. and Pace, N.R. (1998) Journal of Bacteriology, 180, 4765–4774. 42 (a) Curtis, T.P. and Sloan, W.T. (2005) Science, 309, 1331–1333. (b) Daniel, R. (2005) Nature Reviews Microbiology, 3, 470–478. (c) Gans, J., Wolinsky, M. and Dunbar, J. (2005) Science, 309, 1387–1390. 43 (a) Lorenz, P. and Eck, J. (2005) Nature Reviews Microbiology, 3, 510–516. (b) Steele, H.L. and Streit, W.R. FEMS Microbiology Letters, 247, 105–111. 44 Brennan, Y., Callen, W.N., Christoffersen, L., Dupree, P., Goubet, F., Healey, S., Hernandez, M., Keller, M., Li, K., Palackal, N., Sittenfeld, A., Tamayo, G., Wells, S., Hazlewood, G.P., Mathur, E.J., Short, J.M., Robertson, D.E. and Steer, B.A. (2004) Applied and Environmental Microbiology, 70, 3609–3617. 45 Ferrer, M., Golyshina, O.V., Chernikova, T.N., Khachane, A.N., Martins dos Santos, V.A.P., Yakimov, M.M., Timmis, K.N. and Golyshin, P.N. (2005) Chemistry and Biology, 12, 895–904. 46 Courtois, S., Cappellano, C.M., Ball, M., Francou, F.-X., Normand, P., Helynck, G., Martinez, A., Kolvek, S.J., Hopke, J., Osburne, M.S., August, P.R., Nalin, R., Guerineau, M., Jeannin, P., Simonet, P. and Pernodet, J.-L. (2003) Applied and Environmental Microbiology, 69, 49–55. 47 Entcheva, P., Liebl, W., Johann, A., Hartsch, T. and Streit, W.R. (2001) Applied and Environmental Microbiology, 67, 89–99. 48 Rondon, M.R., August, P.R., Bettermann, A.D., Brady, S.F., Grossman, T.H., Liles,


References

49

50

51

52

53

54

55

56

M.R., Loiacono, K.A., Lynch, B.A., MacNeil, I.A., Minor, C., Tiong, C.L., Gilman, M., Osburne, M.S., Clardy, J., Handelsman, J. and Goodman, R.M. (2000) Applied and Environmental Microbiology, 66, 2541–2547. Kim, Y.-J., Choi, G.-S., Kim, S.-B., Yoon, G.-S., Kim, Y.-S. and Ryu, Y.-W. (2006) Protein Expression and Purification, 45, 315–323. Henne, A., Daniel, R., Schmitz, R.A. and Gottschalk, G. (1999) Applied and Environmental Microbiology, 65, 3901–3907. Ranjan, R., Grover, A., Kapardar, R.K. and Sharma, R. (2005) Biochemical and Biophysical Research Communications, 335, 57–65. Gabor, E.M., de Vries, E.J. and Janssen, D.B. (2004) Environmental Microbiology, 6, 948–958. Venter, J.C., Remington, K., Heidelberg, J.F., Halpern, A.L., Rusch, D., Eisen, J.A., Wu, D., Paulsen, I., Nelson, K.A., 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. and Smith, H.O. (2004) Science, 304, 66–74. Leggewie, C., Henning, H., Schmeisser, C., Streit, W.R. and Jaeger, K.-E. (2006) Journal of Biotechnology, 123, 281–287. (a) Li, Y., Wexler, M., Richardson, D.J., Bond, P.L. and Johnston, A.W.B. (2005) Environmental Microbiology, 7, 1927–1936. (b) Martinez, A., Kolvek, S.J., Yip, C.L.T., Hopke, J., Brown, K.A., MacNeil, I.A. and Osburne, M.S. (2004) Applied and Environmental Microbiology, 70, 2452–2463. (c) Wexler, M., Bond, P.L., Richardson, D.J. and Johnston, A.W. (2005) Environmental Microbiology, 7, 1917–1926. Wang, G.-Y.-S., Graziani, E., Waters, B., Pan, W., Li, X., McDermott, J., Meurer, G., Saxena, G., Andersen, R.J. and Davies, J. (2000) Organic Letters, 2, 2401–2404.

57 Uchiyama, T., Abe, T., Ikemura, T. and Watanabe, K. (2005) Nature Biotechnology, 23, 88–93. 58 Voget, S., Leggewie, C., Uesbeck, A., Raasch, C., Jaeger, K.-E. and Streit, W.R. (2003) Applied and Environmental Microbiology, 69, 6235–6242. 59 Ferrer, M., Golyshina, O.V., Chernikova, T.N., Khachane, A.N., Reyes-Duarte, D., Martins dos Santos, V.A.P., Strompl, C., Elborough, K., Jarvis, G., Neef, A., Yakimov, M.M., Timmis, K.N. and Golyshin, P.N. (2005) Environmental Microbiology, 7, 1996–2010. 60 Richardson, T.H., Tan, X., Frey, G., Callen, W., Cabell, M., Lam, D., Macomber, J., Short, J.M., Robertson, D.E. and Miller, C. (2002) Journal of Biological Chemistry, 277, 26501–26507. 61 Yun, J., Kang, S., Park, S., Yoon, H., Kim, M.-J., Heu, S. and Ryu, S. (2004) Applied and Environmental Microbiology, 70, 7229–9235. 62 Rothschild, L.J. and Mancinelli, R.L. (2001) Nature, 409, 1092–1101. 63 Grant, S., Sorokin, D.Y., Grant, W.D., Jones, B.E. and Heaphy, S. (2004) Extremophiles, 8, 421–429. 64 Healy, F.G., Ray, R.M., Aldrich, H.C., Wilkie, A.C., Ingram, L.O. and Shanmugam, K.T. (1995) Applied Microbiology and Biotechnology, 43, 667–674. 65 Voget, S., Steele, H.L. and Streit, W.R. (2006) Journal of Biotechnology, 126, 26–36. 66 Howard, M.B., Ekborg, N.A., Weiner, R.M. and Hutcheson, S.W. (2003) Journal of Industrial Microbiology and Biotechnology, 30, 627–635. 67 Carstens, M., Vivier, M.A., van Rensburg, P. and Pretorius, I.S. (2003) Annals of Microbiology, 53, 15–28. 68 (a) Cottrell, M.T., Moore, J.A. and Kirchman, D.L. (1999) Applied and Environmental Microbiology, 65, 2553–2557. (b) Cottrell, M.T., Wood, D.N., Yu, L. and Kirchman, D.L. (2000) Applied and Environmental Microbiology, 66, 1195–1201.

j83


j 3 The Search for New Enzymes

84

69 Lee, C.C., Kibblewhite-Accinelli, R.E., Wagschal, K., Robertson, G.H. and Wong, D.W.S. (2006) Extremophiles, 10, 295–300. 70 Henne, A., Schmitz, R.A., Bomeke, M., Gottschalk, G. and Daniel, R. (2000) Applied and Environmental Microbiology, 66, 3113–3116. 71 Rees, H.C., Grant, S., Jones, B., Grant, W.D. and Heaphy, S. (2003) Extremophiles, 7, 415–421. 72 Lee, S.W., Won, K., Lim, H.K., Kim, J.C., Choi, G.J. and Cho, K.Y. (2004) Applied Microbiology and Biotechnology, 65, 720–726. 73 Rhee, J.-K., Ahn, D.-G., Kim, Y.-G. and Oh, J.-W. (2005) Applied and Environmental Microbiology, 71, 817–825. 74 Kim, J.N., Seo, M.J., Cho, E.A., Lee, S.J., Kim, S.B., Cheigh, C.I. and Pyun, Y.R. (2005) Journal of Microbiology and Biotechnology, 15, 1067–1072. 75 Elend, C., Schmeisser, C., Leggewie, C., Babiak, P., Carballeira, J.D., Steele, H.L., Reymond, J.L., Jaeger, K.E. and Streit, W.R. (2006) Applied and Environmental Microbiology, 72, 3637–3645. 76 Eschenfeldt, W.H., Stols, L., Rosenbaum, H., Khambatta, Z.S., Quaite-Randall, E., Wu, S., Kilgore, D.C., Trent, J.D. and Donnelly, M.I. (2001) Applied and Environmental Microbiology, 67, 4206–4214. 77 Cantarella, M., Cantarella, L., Gallifuoco, A. and Spera, A. (2006) Enzyme and Microbial Technology, 38, 126–134. 78 Brady, D., Beeton, A., Zeevaart, J., Kgaje, C., van Rantwijk, F. and Sheldon, R.A. (2004) Applied Microbiology and Biotechnology, 64, 76–85. 79 Liebeton, K. and Eck, J. (2004) Engineering in Life Sciences, 4, 557–562. 80 Gabor, E.M., Alkema, W.B.L. and Janssen, D.B. (2004) Environmental Microbiology, 6, 879–886. 81 (a) DeSantis, G., Zhu, Z., Greenberg, W.A., Wong, K., Chaplin, J., Hanson, S.R., Farwell, B., Nicholson, L.W., Rand, C.L., Weiner, D.P., Robertson, D.E. and Burk,

82

83

84

85

86

87

M.J. (2002) Journal of the American Chemical Society, 124, 9024–9025. (b) Podar, M., Eads, J.R. and Richardson, T.H. (2005) BMC Evolutionary Biology, 5, 42. (c) Robertson, D.E., Chaplin, J.A., DeSantis, G., Podar, M., Madden, M., Chi, E., Richardson, T., Milan, A., Miller, M., Weiner, D.P., Wong, K., McQuaid, J., Farwell, B., Preston, L.A., Tan, X., Snead, M.A., Keller, M., Mathur, E., Kretz, P.L., Burk, M.J. and Short, J.M. (2004) Applied and Environmental Microbiology, 70, 2429–2436. Wang, C., Meek, D.J., Panchal, P., Boruvka, N., Archibald, F.S., Driscoll, B.T. and Charles, T.C. (2006) Applied and Environmental Microbiology, 72, 384–391. (a) Knietsch, A., Waschkowitz, T., Bowien, S., Henne, A. and Daniel, R. (2003) Applied and Environmental Microbiology, 69, 1408–1416. (b) Knietsch, A., Waschkowitz, T., Bowien, S., Henne, A. and Daniel, R. (2003) Journal of Molecular Microbiology and Biotechnology, 5, 46–56. Gupta, R., Beg, Q.K. and Lorenz, P. (2002) Applied Microbiology and Biotechnology, 59, 15–22. Knietsch, A., Bowein, S., Whited, G., Gottschalk, G. and Daniel, R. (2003) Applied and Environmental Microbiology, 69, 3048–3060. (a) Lim, H.K., Chung, E.J., Kim, J.-C., Choi, G.J., Jang, K.S., Chung, Y.R., Cho, K.Y. and Lee, S.-W. (2005) Applied and Environmental Microbiology, 71, 7768–7777. (b) MacNeil, I.A., Tiong, C.L., Minor, C., August, P.R., Grossman, T.H., Loiacono, K.A., Lynch, B.A., Phillips, T., Narula, S., Sundaramoorthi, R., Tyler, A., Aldredge, T., Long, H., Gilman, M., Holt, D. and Osburne, M.S. (2001) Journal of Molecular Microbiology and Biotechnology, 3, 301–308. (a) Brady, S.F. and Clardy, J. (2004) Journal of Natural Products, 67, 1283–1286. (b) Brady, S.F., Chao, C.J. and Clardy, J. (2004) Applied and Environmental Microbiology, 70, 6865–6870. (c) Brady, S.F. and Clardy, J. (2005) Angewandte Chemie


References International Edition, 44, 7063–7065. (d) Gillespie, D.E., Brady, S.F., Bettermann, A.D., Cianciotto, N.P., Liles, M.R., Rondon, M.R., Clardy, J., Goodman, R.M. and Handelsman, J. (2002) Applied and Environmental Microbiology, 68, 4301–4306.

88 (a) Diaz-Torres, M.L., McNab, R., Spratt, D.A., Villedieu, A., Hunt, N., Wilson, M. and Mullany, P. (2003) Antimicrobial Agents and Chemotherapy, 47, 1430–1432. (b) Riesenfeld, C.S., Goodman, R.M. and Handelsman, J. (2004) Environmental Microbiology, 6, 981–989.

j85


j87

II Synthetic Applications


j89

4 Dynamic Kinetic Resolutions Belen Martın-Matute and Jan-E. B€ackvall

4.1 Introduction 4.1.1 Synthesis of Enantiomerically Pure Compounds

There are three strategies for preparing enantiomerically pure compounds [1]. One way is to use naturally occurring starting materials of defined absolute configuration, provided by nature’s chiral pool. The second approach is to perform a resolution of a racemate. Thirdly, an asymmetric synthesis can be achieved in which one or more chiral centers are created in an achiral starting material. The first two approaches have some important drawbacks: in the first approach, an adequate starting material must be available that can be efficiently transformed to the desired product in a few steps, and in the second approach, the maximum theoretical yield is only 50%. For these reasons, asymmetric synthesis accomplished by the use of a chiral catalyst has been considered as the most refined strategy. In this approach, one enantiopure molecule (catalyst) is able to transform many prochiral molecules in enantiopure products (chiral multiplication). The chiral ligand can be an enzyme, a simple organic molecule, or a metal complex bearing chiral ligands. There are some examples of metal-catalyzed asymmetric transformations that have been developed to produce chiral compounds on a big scale. For example, in the Monsanto process developed by Knowles, chiral diphosphines are employed as chiral ligands in an Rh-catalyzed hydrogenation [2]. Another example is the asymmetric hydrogenation developed by Noyori in which Ru complexes bearing chiral phosphines are used [3]. Despite the still growing number of available methods for the preparation of enantiopure compounds by the use of asymmetric catalysis, kinetic resolution (KR) is still the most employed method in the industry [4], and in most cases biocatalysts (enzymes) are used.


j 4 Dynamic Kinetic Resolutions

90

cat* s Fast

p 50%

ent- s

cat* ent- p Slow

Figure 4.1 Kinetic resolution (s is substrate, p is product, cat* is homochiral catalyst).

4.1.2 Kinetic Resolution (KR) and Dynamic Kinetic Resolution (DKR)

KR is the total or partial separation of two enantiomers from a racemic mixture [5]. KR is based on the different reaction rates of the enantiomers with a chiral molecule (a reagent, a catalyst, etc). In the ideal case, the difference in reactivity is large, and one of the enantiomers reacts very fast to give the product, whereas the other does not react at all (Figure 4.1). Many enzyme-catalyzed KRs do not show this ideal behavior. Usually the reaction does not stop at 50% conversion, but it only slows down. The concentration of each enantiomer is not constant during the reaction, and therefore, the reaction rate of both enantiomers changes with the conversion. To obtain the product with high enantiomeric excess, it is necessary to stop the reaction before it reaches 50% conversion. In a KR the enantiomeric excess of the substrate and product change with the conversion [6]. A direct comparison of the enantiomeric excess of two KRs is relevant only if they have the same conversion. Owing to the intrinsic difficulty that this implies, Sih developed a series of equations to calculate the enantioselectivity (E ) of an enzyme-catalyzed KR [7]. The enantioselectivity known as ‘enantiomeric ratio’ (E ¼ ks/kents) (for biocatalyzed reactions, the ‘binding’ of the substrate enantiomers (which can be neglected with chemical catalysts) usually plays an important role in the chiral selection process and E values of enzyme-catalyzed reactions are therefore defined through Michaelis–Menten kinetics: E ¼ (kcat/KM)R/ (kcat/KM)S), measures the ability of an enzyme to distinguish between two enantiomers [8]. E remains constant throughout the reaction, and allows easy comparison between two KRs. It can be calculated very easily knowing two of the three following parameters: conversion (c), enantiomeric excess of the product (eep), and enantiomeric excess of the substrate (ees) (Faber has developed a simple program (Selectivity) that allows to calculate the E value and to draw plots of the variation of ees and eep as a function of time for an irreversible kinetic resolution. The program can be obtained on: http://borgc185.kfunigraz.ac.at/). As a rule of thumb KRs with E values smaller than 15 do not have applications in synthesis. Values of E of 15–30 are considered as moderate or good, and values >30 are regarded as excellent with useful applications in organic synthesis [1f ]. Enzymatic KRs, as all resolutions, are limited to a maximum theoretical yield of 50%. Strategies to increase the yield are therefore of great importance. The opposite of a resolution, that is, the racemization of a chiral compound, can sometimes be highly desirable and applicable in enantioselective synthesis. By combining a


4.1 Introduction

s Racemization krac ent-s

cat* ks Fast cat*

p 100 %

ent-p

kent-s Slow Figure 4.2 Dynamic kinetic resolution.

racemization with an enzyme-catalyzed resolution, a highly efficient asymmetric transformation to only one enantiomer can be obtained. Such dynamic kinetic resolutions (DKRs) with a theoretical yield of 100% are a powerful approach to prepare enantiomerically pure molecules (Figure 4.2). Racemization can be performed by a chemocatalyst or a biocatalyst, or it can occur spontaneously. For an efficient enzymatic DKR the following requirements must be fulfilled: (i) the KR must be very selective (E > 20); (ii) the racemization must be fast (at least 10 times faster than the enzyme-catalyzed transformation of the slow reacting enantiomer, krac >10 kents); (iii) the racemization catalyst must not react with the product of the reaction; (iv) the KR and the racemization must be compatible under the same reaction conditions. In an ideal DKR, where the substrate stays racemic throughout the reaction process, the optical purity depends only on the enantiomeric ratio (E ) (ee ¼ (E  1)/ (E þ 1)), and is independent of the extent of conversion. The enantiomeric excess of the product formed under racemizing conditions is equal to the initial enantiomeric excess under nonracemizing conditions. 4.1.3 Enzymes in Organic Chemistry

Biocatalysts usually require mild reaction conditions for an optimal activity (physiologic temperature and pH) and, in general, they show high activity, chemo- and enantioselectivity. Furthermore, when using enzymes, many functional group protections and/or activations can be avoided, allowing shorter synthetic transformations. The use of enzymes is therefore very attractive from an environmental and economic point of view. Lipases and proteases belong to the group of hydrolases [1e,9]. Hydrolases are excellent catalysts and many of them are commercially available. They do not require a cofactor and they catalyze the hydrolysis of nonnatural esters very efficiently. However, their use in organic synthesis dates from only two decades ago. Since then, the number of reports on enzyme-catalyzed synthetic transformations has grown exponentially. Before that period, hydrolases were only employed in aqueous media, which is a drawback in synthesis owing to the insolubility of organic compounds in water. During the 1980s it was discovered that enzymes can be employed in organic media since they keep their native structures even in

j91


j 4 Dynamic Kinetic Resolutions

92

anhydrous organic solvents [10]. This stimulated the development of many new enzyme-catalyzed transformations such as transesterifications, aminolysis, or thiotransesterifications. One of the most important developments, the combination of enzymatic resolution with racemization (to give a DKR), could also be realized because of the possibility of running enzymatic reactions in organic solvents. In this chapter, DKRs will be categorized according to the racemization method employed, as being catalyzed by (i) a metal, (ii) a base, (iii) an acid, (iv) an aldehyde, or (v) an enzyme. Also racemizations that take place through continuous cleavage/ formation of the substrate, or through SN2 displacement, among other methods, will be discussed. In most cases, the racemization method of choice depends on the structure of the substrate. In all cases, the KR is catalyzed by an enzyme.

4.2 Metal-Catalyzed Racemization

A range of metal-catalyzed racemization reactions are known in the literature, but it is not always straightforward to combine these with an enzymatic resolution. For example, several complexes of rhodium, iridium, and ruthenium, among others, are known to catalyze rapid racemization of a variety of alcohol substrates [11], but only few have proved compatible with an enzymatic resolution. Most often, the compatibility of the two catalysts under the same reaction conditions is the major problem: the metal may interfere with the enzyme to give poor resolution or the enzyme may slow down or inhibit the racemization by the metal catalyst. Furthermore, the byproducts produced in the metal- and/or enzyme-catalyzed reaction or the surfactants employed to support the enzyme can also interfere with the racemization or with the enzymatic resolution. Another problem that has to be overcome is to find the appropriate solvent for optimal activity of both catalysts, the enzyme and the metal complex: lipases show excellent activity and selectivity in water, but most organometallic complexes do not work in water. Lipases also work very well in apolar organic solvents, such as hexane. However, most organometallic complexes are not soluble in very apolar solvents, resulting in a low rate of the racemization reaction. Also, racemization is usually faster at high temperatures, at which most enzymes lose their selectivity and catalytic activity. Despite this, a deep understanding concerning the compatibility of the two catalysts has been achieved, and excellent results have been obtained. In 1997, St€ urmer already highlighted the importance of the combination of enzymes and transition metals in one pot [12]. Since then, this concept has attracted considerable interest within the scientific community. In all the DKRs presented in this section, the enzyme catalyzes a transesterification process. Thus, enzyme- and metal-catalyzed DKRs will be categorized according to the nature of the substrate as being allylic substrates, secondary alcohols, or primary amines. In the first case, racemization occurs through 1,3-shift of an acetate or a hydroxyl group, and in the last two cases it occurs through hydrogen transfer processes.


4.2 Metal-Catalyzed Racemization

OAc

OAc Ph

i

PrOAc

o

(R)

THF, 25 C, 1.5 d

–Pd(0)

+Pd(0)

OH

CALB, iPrOH

Ph

j93

Ph

71% yield, 98% ee

Ph

Pd L

AcO

Figure 4.3 DKR of allylic acetates catalyzed by a lipase and Pd(0).

4.2.1 DKR of Allylic Acetates and Allylic Alcohols

The first example of chemoenzymatic DKR of allylic alcohol derivatives was reported by Allen and Williams [13]. Cyclic allylic acetates were deracemized by combining a lipasecatalyzed hydrolysis with a racemization via transposition of the acetate group catalyzed by a Pd(II) complex. Despite the limitation of the process with very long reaction times (19 days), this work was a significant step forward in the combination of enzymes and metals in one pot. Some years later, Kim et al. considerably improved the DKR of allylic acetates using a Pd(0) complex for the racemization, which occurs through p-(allylpalladium) intermediates. The transesterification is catalyzed by a lipase (Candida antarctica lipase B (CALB)) using isopropanol as acyl acceptor (Figure 4.3) [14]. A novel approach was developed very recently by Kita et al. [15]. DKR of allylic alcohols was performed by combining a lipase-catalyzed acylation with a racemization through the formation of allyl vanadate intermediates. Excellent yields and enantioselectivities were obtained. An example is shown in Figure 4.4. A limitation with this approach for the substrates shown in Figure 4.4 is that the allylic alcohol must be equally disubstituted in the allylic position (R1 ¼ R2) since CC single bond rotation is required in the tertiary alkoxy intermediate. Alternatively, R1 or R2 can be H if the two allylic alcohols formed by migration of the hydroxyl group are enantiomers (e.g. cyclic allylic acetates). DKR of allylic alcohols can be also performed using ruthenium complexes for the racemization that occurs through hydrogen transfer reactions (vide infra) [16]. EtO OH

OAc

CALB,

OH

VO(OSiPh 3)3 (10 mol %) +[VO(OR)3]

–[VO(OR)3]

OAc (R)

Acetone, 25 oC, 4.5 d 91% yield, 99% ee

OR RO V O O R2

OR V O O

RO

R1

R2 1

R

OR RO V O O R2 R1

Figure 4.4 DKR of allylic alcohols via vanadate intermediates.


j 4 Dynamic Kinetic Resolutions

94

[M–H] + H

O

O

H R

R'

or

R

M

R'

H

O R

H

O R

M

H

H H

R'

R' R or S

rac

Figure 4.5 Simplified mechanism of the racemization of secalcohols catalyzed by transition metal complexes.

4.2.2 DKR of sec-alcohols

The racemization mechanism of sec-alcohols has been widely studied [16,17]. Metal complexes of the main groups of the periodic table react through a direct transfer of hydrogen (concerted process), such as aluminum complexes in Meerwein–Ponndorf–Verley–Oppenauer reaction. However, racemization catalyzed by transition metal complexes occurs via hydrogen transfer processes through metal hydrides or metal dihydrides intermediates (Figure 4.5) [18]. The research groups of Williams [19] and B€ackvall [20] were the first to report the combination of enzymes and transition metals for DKR of sec-alcohols. Williams employed complexes of Al, Rh, or Ir in combination with Pseudomonas fluorescens lipase (PFL) for the DKR of 1-phenylethanol. The best results were obtained using Rh2(OAc)4 as the catalyst for the racemization, and 60% conversion of the alcohol to give 1-phenylethyl acetate in 98% ee was obtained (Figure 4.6) [19]. At higher conversion, the enantiomeric excess dropped and 76% conversion gave 80% ee. The research group of B€ackvall employed the Shvo’s ruthenium complex (1) [21] for the racemization. This complex is activated by heat. For the KR they used p-chlorophenyl acetate as the acyl donor in combination with thermostable enzymes, such as CALB [20] (Figure 4.7). This was the first practical chemoenzymatic DKR affording acetylated sec-alcohols in high yields and excellent enantioselectivities. In the best case 100% conversion (92% isolated yield) with 99% ee was obtained. This method was subsequently applied to a variety of different substrates and it is employed (with a different ruthenium complex) by the Dutch company DSM for the large-scale production of (R)-phenylethanol [22]. Kim and Park subsequently reported that ruthenium precatalyst (2) racemizes alcohols within 30 minutes at room temperature [23]. However, when combined OH Ph

OAc Rh2(OAc)4 (2 mol %)

OAc Ph (R)

PFL, o-phenantroline (6 mol %), PhCOMe (1 equiv) 60% conversion, 98% ee Cyclohexane, 20 oC, 72 h

Figure 4.6 DKR of sec-alcohols catalyzed by PFL and Rh complexes.


4.2 Metal-Catalyzed Racemization Ph

O

j95 Ph

O H

OH

OAc

CALB, p-Cl-C6H4-OAc (3 equiv)

R

R (R)

[Ru] (1) (2 mol %) o

78–92% yield, >99% ee

Toluene, 70 C

R = Alkyl, aryl

Ph Ph

Ph Ph

Ru OC CO

H

Ru CO CO

1

24–72 h

Figure 4.7 DKR of sec-alcohols using CALB and Shvo’s complex (1).

with an enzyme (lipase) in DKR at room temperature, very long reaction times (1.3–7 days) were required, in spite of the fact that the enzymatic KR takes only a few hours (Figure 4.8). Despite these compatibility problems, their results constitute an important improvement, since chemoenzymatic DKR could now be performed at ambient temperature to give high yields, which enables the use of nonthermostable enzymes. More recently, we accomplished a highly efficient metal- and enzyme-catalyzed DKR of alcohols at room temperature (Figure 4.8) [16,24]. This is the fastest DKR of alcohols hitherto reported by the combination of transition metal and enzyme catalysts. Racemization was effected by a very potent hydrogen transfer catalyst (3). Catalysts (2) and (3) have been employed in the DKR of a broad range of secondary alcohols. We have studied the racemization and found that it proceeds via a new mechanism where the intermediate ketone formed in the process stays coordinated to ruthenium [16]. Recently, the dynamic kinetic asymmetric transformation (DYKAT) of diols was improved significantly by using the highly efficient catalyst (3) to achieve very fast epimerization [25,26]. In the DYKATof 2,5-hexanediol catalyzed by (3) [25], the faster epimerization and the low reaction temperature reduce the concentration of intermediates that can evolve by anomalous enzyme-catalyzed (S)-acylations, such as alcohols bearing a carbonyl group or an (R)-acetoxy group at the d-position that give rise to the unwanted formation of meso-diacetates. In the DYKAT of 2,4-pentanediol catalyzed by (3), the fast epimerization outruns unwanted acyl migrations that yield meso-diacetates. Thus, by using complex (3) (R,R)-2,5-hexanediol diacetate and (R,R)-2,4-pentanediol diacetate can now be obtained in quantitative yield with

OH

CALB,

Ph

OAc

OAc Ph

[Ru] (4 mol %), KOt Bu (5 mol %), Na 2CO3 o

Ph

[Ru] = 2, 31 h [Ru] = 3, 3 h

H N

Ph

Ph Ph Ru Cl OC CO

2

(R)

95–99% yield, >99% ee

Toluene, 23 C

Ph

Ph

Ph

Ph

Ph Ph

Ph Ru Cl OC CO

3

Figure 4.8 DKR of sec-alcohols catalyzed by a lipase and Ru complexes at ambient temperature.


j 4 Dynamic Kinetic Resolutions

96

excellent enantiomeric excess (>99%) and high (R,R)/meso ratio (94:6 and 97:3, respectively). Kim and Park et al. have reported a polymer-supported derivative (4), which, together with the enzyme, can be recycled and reused [27].

O O Ph Ph

O Ph

Ph Ru Cl OC CO

4 Sheldon et al. have combined a KR catalyzed by CALB with a racemization catalyzed by a Ru(II) complex in combination with TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl free radical) [28]. They proposed that racemization involved initial ruthenium-catalyzed oxidation of the alcohol to the corresponding ketone, with TEMPO acting as a stoichiometric oxidant. The ketone was then reduced to racemic alcohol by ruthenium hydrides, which were proposed to be formed under the reaction conditions. Under these conditions, they obtained 76% yield of enantiopure 1-phenylethanol acetate at 70 after 48 hours. Naturally occurring lipases are (R)-selective for alcohols according to Kazlauskas’ rule [29] ((R)- and (S)-selective are used for typical sec-alcohols where the large group has a higher priority in the sequential rule for determining the configuration (according to the Cahn–Ingold–Prelog system). Thus DKR of alcohols employing lipases can only be used to transform a racemic alcohol into the (R)-acetate. Serine proteases, a subclass of hydrolases, are known to catalyze transesterifications similar to those catalyzed by lipases, but interestingly, they often show the reversed enantioselectivity to give the (S)-product. Proteases are less thermostable enzymes, and for this reason metal complexes that racemize secondary alcohols at ambient temperature are desirable in DKR of sec-alcohols with these enzymes. Ruthenium complexes (2) and (3) have been combined with subtilisin Carlsberg, affording a method for the synthesis of acetylated (S)-sec-alcohols [30]. With complex (2) the DKR reaction took 3–4 days [30a], whereas with catalyst (3) DKR was obtained with reaction time down to 18 hours [30b], in both cases at room temperature. Very recently the Meerwein–Ponndorf–Verley–Oppenauer (MPVO) reaction has been exploited for the racemization of alcohols using inexpensive aluminum-based catalysts. Combination of these complexes with a lipase (CALB) results in an efficient DKR of sec-alcohols at ambient temperature. To increase the reactivity of the aluminum complexes, a bidentate ligand, such as binol, is required. Also, specific acyl donors need to be used for each substrate [31] (Figure 4.9). Kita et al. have made use of the structure of the acyl donor to develop a domino transformation, a DKR followed by an intramolecular Diels–Alder reaction [32]. They


4.2 Metal-Catalyzed Racemization

OH

OAc

CALB, AlMe3 (10 mol %),

Ph

j97

Ph

OAc

(R)

96% yield, 96% ee (1.2 equiv) Binol (10 mol %) Toluene, rt, 3 h

Figure 4.9 DKR of sec-alcohols using a lipase and inexpensive Al complexes.

O OH

EtO

O

CO2Et

CO2Et

O O

O

O

CO2Et

CALB, Et3N, MS

Cl Ru Cl

81% yield, 97% ee 2

(10 mol %) MeCN, 35 oC, 3 d

Figure 4.10 Tandem DKR-intramolecular Dielsâ&#x20AC;&#x201C;Alder reaction.

have found reaction conditions for three different transformations to effectively take place in one pot: a KR, a racemization, and a Dielsâ&#x20AC;&#x201C;Alder reaction (Figure 4.10). Meijer and Heise and coworkers have developed iterative tandem catalysis with the aim of generating chiral oligomers [33]. The polymerization of 6-methyl-e-caprolactone was achieved by combining enzymatic ring opening with ruthenium-catalyzed racemization in one pot yielding enantioenriched oligomers (Figure 4.11). More recently, Heise and coworkers have shown that DKR can be combined with enzymatic polymerization for the synthesis of chiral polyesters from racemic secondary diols in one pot [34] (Figure 4.12). O OH

O

CALB

O O

O O

OH O n

CALB

O

Ru HN O

n = n +1

O

O O

OH O n

Figure 4.11 Synthesis of chiral oligomers by DKR.

NH


j 4 Dynamic Kinetic Resolutions

98

HO

O

OH

OMe

MeO O

Ru HN

CALB

NH O

O O O O

n

Figure 4.12 Synthesis of chiral polyesters by DYKAT.

4.2.3 DKR of Amines

Racemization of amines is difficult to achieve and usually requires harsh reaction conditions. Reetz et al. developed the first example of DKR of amines using palladium on carbon for the racemization and CALB for the enzymatic resolution [35]. This combination required long reaction times (8 days) to obtain 64% yield in the DKR of 1-phenylethylamine. More recently, B€ackvall et al. synthesized a novel Shvo-type ruthenium complex (5) that in combination with CALB made it possible to perform DKR of a variety of primary amines with excellent yields and enantioselectivities (Figure 4.13) [36]. Jacobs et al. have found that the efficiency of the Pd-catalyzed racemization of amines can be improved by using Pd immobilized on supports such as BaSO4, CaCO3, or BaCO3. The racemization was combined with a KR catalyzed by CALB affording enantiopure acetylated benzylamines in high yields [37]. Very recently Page and coworkers have reported the DKR of sec-amines using a low catalyst loading of an Ir complex for the racemization, and Candida rugosa lipase for the enzymatic resolution [38].

4.3 Base-Catalyzed Racemization

In this chapter, DKR of substrates having a proton with low pKa will be discussed. Racemization occurs by performing the DKR in the presence of a weak base. Also in


4.3 Base-Catalyzed Racemization

NH 2

OAc

CALB,

R

j99

NHAc R

[Ru] (5) (4 mol %), Na 2CO3 Toluene, 90 oC 3d

R = Ph, p-MeOC 6H 4

R

O

R = Ph, 90% yield, 98% ee R = p-MeOC6H4 , 95% yield, 99% ee R

O H

R

R R

Ru

R

R R

H

Ru OC CO CO CO 5 (R = p-MeO-C6 H4 ) Figure 4.13 Ru- and lipase-catalyzed DKR of primary amines.

this section, enzyme- and base-catayzed DKRs will be categorized according to the nature of the substrates as being thioesters, a-activated esters, oxazolones, hydantoins, or acyloins. 4.3.1 DKR of Thioesters

In contrast to oxoesters, the a-protons of thioesters are sufďŹ ciently acidic to permit continuous racemization of the substrate by base-catalyzed deprotonation at the acarbon. Drueckhammer et al. ďŹ rst demonstrated the feasibility of this approach by performing DKR of a propionate thioester bearing a phenylthiogroup, which also contributes to the acidity of the a-proton (Figure 4.14) [39a]. The enzymatic hydrolysis of the thioester was coupled with a racemization catalyzed by trioctylamine. Owing to the insolubility of the substrate and base in water, they employed a biphasic system (toluene/H2O). Using P. cepacia (Amano PS-30) as the enzyme and a catalytic amount of trioctylamine, they obtained a quantitative yield of the corresponding

O PhS

O PhS

SEt

O

Pseudomonas cepacia

PhS

SEt

OH

Oct3N (0.5 equiv) Toluene / H2O, 25 oC 65 h

OPhS

SEt

Figure 4.14 DKR of thioesters using a base for racemization.

>99% yield, 96.3% ee

EtSH


j 4 Dynamic Kinetic Resolutions

100

Cl

Cl O

OH

CF3 Candida cylindr acea lipase O

Oct 3N, o Isooctane/H 2O, 30 C

O

+ CF3CH 2OH

94% yield, 99.5% ee

4d

Figure 4.15 DKR of activated esters using a base for racemization.

carboxylic acid in 96.3% ee. In this DKR, a maximum enantiomeric excess of 96.8% is expected, since the E value for the enzymatic resolution is 62.4. Drueckhammer et al. also reported DKR of thioesters not having activating groups at the a-position [39]. 4.3.2 DKR of Activated Esters

DKR of esters bearing an electron-withdrawing group at the a-carbon can be performed easily under mild reaction conditions due to the low pKa of the a-proton. Tsai et al. have reported an efficient DKR of rac-2,2,2-trifluoroethyl a-chorophenyl acetate in water-saturated isooctane [40]. They used lipase MY from C. rugosa for the KR and trioctylamine as the base for racemization. (R)–chlorophenylacetic acid was obtained in 93% yield and 89.5% ee (Figure 4.15). 4.3.3 DKR of Oxazolones

A very attractive and efficient method for the synthesis of L-aminoacids via DKR has been reported by Turner et al. [41a,b]. They employed enzyme-catalyzed ring opening of 5(4H)-oxazolones in combination with a catalytic amount of Et3N. The relatively low pKa of the C-4 proton (8.9) of oxazolones facilitates racemization. Hydrolysis of the ester obtained through DKR, followed by debenzoylation, yields L-aminoacids in excellent enantiomeric excess (99.5%) (Figure 4.16). In their initial studies, they employed Rhizomucor miehei lipase (Lipozyme) as the biocatalyst [41]. More recently, they have obtained excellent results employing CALB [41b]. This method has also been employed by Bevinakatti [41c,d] and Sih [41e,f ].

O N Ph

O

H HN Et3N (25 mol %), BuOH (2 equiv) o

Toluene, 30 C 5d

H O

Rhizomucor miehei lipase

Ph

OBu O

94% yield, 99.5% ee

Figure 4.16 DKR of oxazolones using a base for racemizing.

O H2N

OH


4.4 Acid-Catalyzed Racemization

NH2

O Ph

HN

HN

D-hydantoinase

NH

Ph Borate buffer o pH = 9, 40 C 2d

O

O CO2H

73% yield, >99% ee

Figure 4.17 DKR of hydantoins under weak basic conditions.

4.3.4 DKR of Hydantoins

Another approach for the synthesis of enantiopure amino acids or amino alcohols is the enantioselective enzyme-catalyzed hydrolysis of hydantoins. As discussed above, hydantoins are very easily racemized in weak alkaline solutions via keto enol tautomerism. Sugai et al. have reported the DKR of the hydantoin prepared from DL-phenylalanine. DKR took place smoothly by the use of D-hydantoinase at a pH of 9 employing a borate buffer (Figure 4.17) [42]. 4.3.5 DKR of Acyloins

Ogasawara et al. took advantage of the easy racemization of acyloins in the presence of a weak base for the DKR of endo-3-hydroxytricyco[4.2.1.02,5]non-7-en-4-one (Figure 4.18) [43]. Acylation of the hydroxyl group was catalyzed by a lipase, and racemization took place via a transient meso-enediol.

4.4 Acid-Catalyzed Racemization

In contrast to enzyme- and base-catalyzed DKRs, there are only a few reports of enzyme- and acid-catalyzed DKRs. A plausible explanation is that deactivation of the enzyme can occur under acidic conditions. Also, decomposition of the substrate has been observed.

O

OH H

+

H HO

O

OH H

Lipase, Et 3N Vinyl acetate

O 75% yield, 97% ee

OH OH Figure 4.18 DKR of acyloins through meso-enediol intermediates.

j101


j 4 Dynamic Kinetic Resolutions

102

O OH

OH

CALB,

O

H-beta zeolite CP814E-22

C7H15

Vinyl octanoate (16 equiv) Octane/H2O, 60 oC, 8 h

–H +, +H2O

+H+, –H2O

90% yield, >99% ee

Figure 4.19 DKR of sec-alcohols catalyzed by acid zeolites and a lipase.

Jacobs et al. employed an acidic zeolite catalyst for the racemization of sec-alcohols, which occurs through the formation of carbocations [44] (Figure 4.19). The KR is catalyzed by CALB in the presence of vinyl octanoate as acyl donor. DKR takes place successfully in a biphasic system (octane/H2O, 1 : 1) at 60  C. Another example of enzyme- and acid-catalyzed DKR has been reported by Bornscheuer [45]. Acyloins were racemized by using an acidic resin through the formation of enol intermediates. The enzymatic resolution was catalyzed by CALB. Since deactivation of this enzyme occurred in the presence of the acidic resin, they designed a simple reactor setup with two glass vials connected via a pump to achieve a spatial separation between the acidic resin and the enzyme (Figure 4.20).

4.5 Racemization through Continuous Reversible Formation–Cleavage of the Substrate

Racemization of some substrates can take place through reversible formation of the substrate via an addition/elimination process. The racemization can be acid or base catalyzed. In this section we will discuss DKR of cyanohydrins and hemithioacetals.

Two-compartment DKR

O

O

CALB, amberlist 15

Ph

O Ph

Ph

OH

OH

Vinyl butyrate Hexane, 40–50 oC

O O

>90% conversion, >91% ee

OH Ph OH Figure 4.20 DKR of sec-alcohols catalyzed by an acidic resin and a lipase.


4.5 Racemization through Continuous Reversible Formation–Cleavage of the Substrate

j103

4.5.1 DKR of Cyanohydrins

Several reports on DKR of cyanohydrins have been developed using this methodology. The unstable nature of cyanohydrins allows continuous racemization through reversible elimination/addition of HCN under basic conditions. The lipase-catalyzed KR in the presence of an acyl donor yields cyanohydrin acetates, which are not racemized under the reaction conditions. In 1992, Oda et al. reported a one-pot synthesis of optically active cyanohydrin acetates from aldehydes, which were converted to the corresponding racemic cyanohydrins through transhydrocyanation with acetone cyanohydrin, catalyzed by a a strongly basic anion-exchange resin [46]. The racemic cyanohydrins were acetylated by a lipase from P. cepacia (Amano) with isopropenyl acetate as the acyl donor. The reversible nature of the base-catalyzed transhydrocyanation enabled continuous racemization of the unreacted cyanohydrins, thereby effecting a total conversion (Figure 4.21). Kanerva et al. have also reported DKR of cyanohydrins [47]. In particular, they obtained very good results with C. antartica lipase A (CAL-A) as the catalyst for the KR of a variety of substrates for which other enzymes such as CALB or PS-C do not give good results (Figure 4.22) [47a]. Burk and coworkers have used a variety of nitrilases for the DKR of cyanohydrins [48]. Nitrilases catalyze the hydrolytic conversion of cyanohydrins directly to the corresponding carboxylic acids. Racemization was performed under basic conditions (phosphate buffer, pH 8) through reversible loss of HCN. (R)-Mandelic acid was obtained in high yield (86% yield) and high enantioselectivity (98% ee) after 3 hours (Figure 4.23).

4.5.2 DKR of Hemithioathetals

As in the case discussed above, hemithioacetals can be racemized by elimination/ addition of a small molecule (a thiol in this case) under weak acidic conditions. Rayner et al. reported the first example of DKR of this kind of substrates yielding homochiral a-acetoxysulfides (Figure 4.24) [49].

NC

O

OH

H

CN Anion-exchange resin

OAc

OH Pseudomonas cepacia lipase

OAc

(OH- form) (10 mol %) (3 equiv)

Figure 4.21 DKR of cyanohydrins.

CN

>96% yield, 84% ee 6.5 d


j 4 Dynamic Kinetic Resolutions

104

NC

OH

N

N

S

O

CHO

S

Amberilite IRA 904 (OH– form)(1 x 10 –5 mol %)

CN

O HO

CAL-A. Acetonitrile

AcO (2equiv)

N S

CN

O AcO

91% yield, 96% ee 48 h

Figure 4.22 Lipase-catalyzed DKR of cyanohydrins.

OH

OH

Nitrilase, MeOH

CN

CO2H

Phosphate buffer pH = 8, 37 oC 3h

86% yield, 98% ee

Figure 4.23 Nitrilase-catalyzed DKR of cyanohydrins.

O O O

H

OH

C8H17SH SiO2

O

SC8H17

O

OAc

Pseudomonas fluorescens lipase

AcO t-Butyl methyl ether, 30 oC

O

SC8H17

O 85% yield, >95% ee

6d

Figure 4.24 DKR of hemithioacetals.

4.6 Racemization Catalyzed by Aldehydes

In 1983, Yamada et al. developed an efficient method for the racemization of amino acids using a catalytic amount of an aliphatic or an aromatic aldehyde [50]. This method has been used in the DKR of amino acids. Figure 4.25 shows the mechanism of the racemization of a carboxylic acid derivative catalyzed by pyridoxal. Racemization takes place through the formation of Schiff-base intermediates. In 1994, Wang et al. reported the DKR of amino acid derivatives by using pyridoxal 5-phosphate as the racemization catalyst [51]. The enzyme employed to catalyze the


4.6 Racemization Catalyzed by Aldehydes

N HOH2C NH2

H2O

OH N

N X

X

R O

O N N

CH3

HOH2C

OH

H2O

CH3 OH

O HOH2C

NH2 X

R

N

O

X

R

CH3

HOH2C

OH

R

O

N

CH3

HOH2C

OH O

X

R

N

CH3

O Figure 4.25 Racemization of amino acids through formation of Schiff-base intermediate.

KR was an alcalase subtilisin Carlsberg as the major enzyme component). To avoid racemization of the ďŹ nal product, they employed a mixture of 2-methyl-2-propanol/ H2O (19 : 1). Under these conditions, the product precipitated during the course of hydrolysis (Figure 4.26). A very similar DKR process was reported two years later by Parmar [52]. Sheldon et al. have reported the DKR of phenylglycine esters via lipase-catalyzed ammonolysis [53]. Racemization was carried out by an aldehyde, such as salicyaldehyde or pyridoxal, under basic conditions. The major problem they found was the racemization by these aldehydes of the ďŹ nal products. However, when performing the DKR at low temperatures (20  C) the substrate was racemized much faster than the product, and DKR was feasible yielding the product in good yield and high enantiomeric excess (Figure 4.27). A very elegant approach has been developed by Kanerva et al. DKR of N-hetrocyclic a-amino esters is achieved using CAL-A [54]. Racemization occurs when acetaldehyde is released in situ from the acyl donor. In this case aldehyde-catalyzed racemization of the product cannot occur (Figure 4.28). This is one of the few examples reported for DKR of secondary amines (For a recent example see the above text and Ref. [38]). O O NH2

O

Pyridoxal 5-phosphate (20 mol %), alcalase

OH NH2

2-Methyl-2-propanol/ H2O (19 : 1) pH = 8.5, 40 oC, 4 h

92% yield, 98% ee

Figure 4.26 DKR of amino acids catalyzed by pyridoxal 5-phosphate.

j105


j 4 Dynamic Kinetic Resolutions

106

O

O O NH2

CH3

Pyridoxal, CALB, NH 3

NH2

t-Butyl alcohol/ t-Butyl methyl ether (70 : 30 v/v) o

NH2 85% yield, 88% ee

–20 C , 66 h

Figure 4.27 DKR of amino acid derivatives.

CAL-A, Et3N

CO2Me

N H

CO2Me

N

O Pr

O t-Butyl methyl ether, 25 oC

O

86% yield, 97% ee

1h Figure 4.28 DKR of -amino esters catalyzed by CAL-A and acetaldehyde released in situ.

4.7 Enzyme-Catalyzed Racemization

Asano et al. have developed an approach for the synthesis of D-amino acids through DKR using a two-enzyme system [55]. They had previously reported the discovery of new D-stereospecific hydrolases that can be applied to KR of racemic amino acid amides to yield D-amino acids. Combination of a D-stereospecific hydrolase with an amino acid amide racemase allows performing DKR of L-amino acid amides yielding enantiomerically pure D-amino acids in excellent yields (Figure 4.29). O H2N

O

D-aminopeptidase

Me NH2

α-Amino-ε-caprolactam racemase Potassium phosphate buffer, 30 oC

HO

Me NH2

7h 99.7% yield, >99% ee

Figure 4.29 DKR using a two-enzyme system.

4.8 Racemization through SN2 Displacement

Williams and coworkers have reported a DKR of a-bromo [56a] and a-chloro esters [56b]. In the latter case, the KR is catalyzed by commercially available cross-linked enzyme crystals derived from Candida cylindracea lipase. The racemization takes place through halide SN2 displacement. The DKR is possible because the racemization of the substrate is faster than that of the product (carboxylate). For the ester, the empty p*(C¼O) orbital is able to stabilize the SN2 transition state by accepting


4.9 Other Racemization Methods

O

O

Candida cylindracea lipase

OMe Cl

OH

ResinPPh 3Cl

Cl

phosphate buffer, pH = 7

90% conversion, 90% ee

Figure 4.30 Enzyme- and chloride-catalyzed DKR.

electron density. However, the carboxylate is more electron-rich and therefore less able to facilitate an SN2 reaction. Racemization is performed by the use of a chloride source. The best results was obtained with a resin-bound phosphonium chloride (Figure 4.30). Faber and coworkers have reported a DKR of mandelic acid by using a lipasecatalyzed O-acylation followed by a racemization catalyzed by mandelate racemase. However, these two transformations do not take place simultaneously in the same pot. When the sequence was repeated four times, (S)-O-acetylmandelic acid was obtained in 80% isolated yield and >98% ee [57].

4.9 Other Racemization Methods

In this section we will discuss some DKRs in which racemization occurs spontaneously during the enzymatic resolution, and without further addition of any reagent. 4.9.1 DKR of 5-Hydroxy-2-(5H)-Furanones

During the KR of this kind of substrates, racemization occurs spontaneously as a consequence of the labile stereogenic center at C-5 [58]. The interconversion occurs by mutarotation, allowing a DKR. The KR is catalyzed by a lipase in the presence of vinyl acetate (Figure 4.31). 4.9.2 DKR of Hemiaminals

Spontaneous racemization also occurs during the enzyme-catalyzed acetylation of hemiaminals [59]. It is thought that the racemization occurs owing to ring opening at

HO

O

O

Lipase PS

AcO

O

O

OAc Hexane 100% conversion, 83% ee

Figure 4.31 DKR in which racemization occurs spontaneously.

j107


j 4 Dynamic Kinetic Resolutions

108

OH

OAc

OH Lipase QL

NR

NR

NR

O

O

O

OAc 100% conversion, >99% ee

Hexane, 60 o C 1d O

NHR O Figure 4.32 DKR in which racemization occurs spontaneously.

the C-3 position under the reaction conditions (Figure 4.32). The KR is catalyzed by lipase QL (Alcaligenes sp.) in the presence of isopropenyl acetate.

4.9.3 DKR of 8-Amino-5,6,7,8-tetrahydroquinoline

Very recently Crawford has reported an interesting enzymatic transformation catalyzed by CALB in which the yield was >60% and the product was obtained in very high optical purity (Figure 4.33) [60]. Furthermore, the remaining starting material was racemic. A mechanistic study performed by the authors showed that CALB and EtOAc are required for racemization to take place. They claimed that enzyme-catalyzed oxidation of the amine substrate takes place, and a ketone intermediate is produced. The ketone reacts with the substrate to form an enamine. This process results in racemization of the substrate (Figure 4.34). It is not clear how the enzyme and EtOAc could dehydrogenate the amine. To conďŹ rm this racemization mechanism, Crawford et al. added 5 mol % of the ketone to the reaction mixture and obtained the product in 78% yield in >98% ee. This DKR is therefore catalyzed by a carbonyl compound, and can be compared to those shown in Section 4.6.

CALB, EtOAc o

Toluene, 50 C

N NH2

2d

N NHAc > 60% yield, >95% ee

Figure 4.33 DKR in which racemization occurs spontaneously.


4.10 Concluding Remarks CALB, EtOAc o

N

Toluene, 50 C

N

N

N

N NH2

N NH

NH2

O

O

N HN

N NH2 rac

Figure 4.34 Mechanism of the racemization suggested in ref. 60.

4.10 Concluding Remarks

During the past few years, great achievements in enzyme-catalyzed DKR have been obtained. The efficiency and scope have improved dramatically. In this process, an enzyme-catalyzed resolution is combined with an in situ racemization allowing a 100% theoretical yield of a variety of enantiopure products to be obtained. The use of enzymes in organic solvents has contributed toward development of DKRs in which racemization is catalyzed by metal complexes. The DKR of alcohols can now be accomplished by the use of a variety of different metals under very mild reaction conditions. It can be applied to the preparation of (R)- or (S)-configured substrates by the use of lipases or proteases, respectively. Also, the acyl group introduced during the enzymatic resolution has been employed in further transformations, increasing the scope of DKR and the development of highly atom-economical processes. DKR has also been used for the synthesis of chiral polymers, which is impossible to accomplish without the presence of a racemization catalyst. Furthermore, DKR of amines can now be performed very efficiently by the combination of enzymes and metal catalysts. Many other racemization methods can be employed in combination with enzymes, such as base- or acid-catalyzed racemizations, aldehyde- or enzymecatalyzed racemizations, or even spontaneous racemizations that take place under the reaction conditions. In all cases, excellent yields and enantioselectivities are obtained, making enzyme-catalyzed DKRs a useful and efficient strategy for the synthesis of enantiopure compounds. The next future will also witness many new contributions to this area. New racemization catalysts, which are more efficient, environmentally friendly, and cheaper are being developed. Also, use of other enzymes will broaden the scope of this transformation. In particular, the new methods available today such as directed evolution, coupled with advances in high-throughput screening technologies, will provide new enzymes with the desired properties to perform new transformations.

j109


j 4 Dynamic Kinetic Resolutions

110

References 1 (a) Halpern, J. and Trost, B.M. (2004) Proceedings of the National Academy of Sciences of the United States of America, 101, 5347. (b) Ojima, I. (ed.) (2000) Catalytic Asymmetric Synthesis, 2nd edn, Wiley-VCH Verlag GmbH, New York. (c) Jacobsen, E.N., Pfaltz, A. and Yamamoto, H. (eds) (1999) Comprehensive Asymmetric Catalysis, Springer, Berlin. (d) Noyori, R. (1994) Asymmetric Catalysis in Organic Synthesis, John Wiley & Sons, Ltd, New York. (e) Drauz, K. and Waldmann, H. (eds) (2002) Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook, 2nd edn, Wiley-VCH Verlag GmbH, Weinheim, Vols. I–III. (f) Faber, K. (2000) Biotransformations in Organic Chemistry, 4th edn, Springer, Berlin. (g) Dalko, P.I. and Moisan, L. (2004) Angewandte Chemie International Edition, 43, 5138–5175. 2 Knowles, W.S., Sabacky, M.J., Vineyard, B.D. and Weinkauff, D.J. (1975) Journal of the American Chemical Society, 97, 2567–2568. 3 Ohkuma, T. Kitamaura, M. and Noyori, R. (2000) Asymmetric hydrogenations, in Catalytic Asymmetric Synthesis (ed. I. Ojima), 2nd edn, Wiley-VCH Verlag GmbH, New York. 4 Breuer, M., Ditrich, K., Habicher, T., Hauer, B., Keeler, M., St€ urmer, R. and Zelinski, T. (2004) Angewandte Chemie International Edition, 43, 788–824. 5 Eliel, E.L., Wilen, S.H. and Mander, L.N. (1994) Stereochemistry of Organic Compounds, John Wiley & Sons, Ltd, New York. 6 Kagan, H.B. and Fiaud, J.C. (1988) Topics in Stereochemistry, 18, 249–330. 7 (a) Sih, C.J. and Wu, S.-H. (1989) Topics in Stereochemistry, 19, 63–125. (b) Chen, C.-S., Wu, S.-H., Girdaukas, G. and Sih, C.J. (1987) Journal of the American Chemical Society, 109, 2812–2817. (c) Chen, C.-S., Fujimoto, Y., Girdaukas, G. and Sih, C.J. (1982) Journal of the American Chemical Society, 104, 7294–7299

8 The selectivity factor used in nonenzymatic kinetic resolutions is known as s, see: (a) reference [6]. (b) Martin, V.S., Woodard, S.S., Katsuki, T., Yamada, Y., Ikeda, M. and Sharpless, K.B. (1981) Journal of the American Chemical Society, 103, 6237–6240 9 Bornscheuer, U.T. and Kazlauskas, R.J. (1999) Hydrolases in Organic Synthesis, Wiley-VCH Verlag GmbH, Weinheim. 10 (a) Klibanov, A.M. (2001) Nature, 409, 241–246. (b) Halling, P.J. (2000) Current Opinion in Chemical Biology, 4, 74–80. (c) Zaks, A. and Klibanov, A.M. (1985) Proceedings of the National Academy of Sciences of the United States of America, 82, 3192–3196. 11 Koh, J.H., Jeong, H.M. and Park, J. (1998) Tetrahedron Letters, 39, 5545–5548. 12 St€ urmer, R. (1997) Angewandte Chemie International Edition, 36, 1173–1174. 13 Allen, J.V. and Williams, J.M.J. (1996) Tetrahedron Letters, 37, 1859–1862. 14 Choi, K.L., Suh, J.H., Lee, D., Lim, I.T., Jung, J.Y. and Kim, M.-J. (1999) Journal of Organic Chemistry, 64, 8423–8424. 15 Akai, S., Tanimoto, K., Kanao, Y., Egi, M., Yamamoto, T. and Kita, Y. (2006) Angewandte Chemie International Edition, 45, 2592–2595. 16 Martın-Matute, B., Edin, M., Bogar, K., Kaynak, F.B. and B€ackvall, J.-E. (2005) Journal of the American Chemical Society, 127, 8817–8825. 17 (a) Clapham, S.E., Hadzovic, A. and Morris, R.H. (2004) Coordination Chemistry Reviews, 248, 2201–2237 (b) Gladiali, S. and Alberico, E. (2004) In Transition Metals for Organic Synthesis (eds M. Beller and C. Bolm), 2nd edn, WileyVCH Verlag GmbH, Weinheim, Vol. 2, pp. 145–166 (c) B€ackvall, J.-E. (2002) Journal of Organometallic Chemistry, 652, 105–111. (d) Huerta, F.F., Minidis, A.B.E. and B€ackvall, J.E. (2001) Chemical Society Reviews, 30, 321–331. (e) Wills, M., Palmer, M., Smith, A., Kenny, J. and Walsgrove, T.


References

18

19

20

21 22

23

24

25

26

(2000) Molecules, 5, 4–18. (f) Palmer, M. and Wills, M. (1999) Tetrahedron: Asymmetry, 10, 2045–2061. (g) MartınMatute, B., Åberg, J.B., Edin, M. and B€ackvall, J.E. Racemization of Secondary Alcohols Catalyzed by Cyclopentadienyl ruthenium Complexes. Insight into the Reaction Mechanism. Chem. Eur. J. 2007, 13, 6063–6072. (a) Pamies, O. and B€ackvall, J.-E. (2001) Chemistry-A European Journal, 7, 5052–5058. (b) Samec, J.S.M., B€ackvall, J.-E., Andersson, P.G. and Brandt, P. (2006) Chemical Society Reviews, 35, 237–248. Dinh, P.M., Howarth, J.A., Hudnott, A.R., Williams, J.M.J. and Harris, W. (1996) Tetrahedron Letters, 37, 7623–7626. (a) Larsson, A.L.E., Persson, B.A. and B€ackvall, J.-E. (1997) Angewandte Chemie International Edition in English, 36, 1211–1212. (b) Persson, B.A., Larsson, A.L.E., Le Ray, M. and B€ackvall, J.-E. (1999) Journal of the American Chemical Society, 121, 1645–1650. Menashe, N. and Shvo, Y. (1991) Organometallics, 10, 3885–3891. Verzijl, G.K.M., De Vries, J.G. and Broxterman, Q.B. (2003) PCT International Application WO 0190396 Al 20011129. “Process for preparation of enantiomerically enriched esters and alcohols.” (a) Choi, J.E., Kim, Y.H., Nam, S.H., Shin, S.T., Kim, M.J. and Park, J. (2002) Angewandte Chemie International Edition, 41, 2373–2376. (b) Choi, J.H., Choi, Y.K., Kim, Y.H., Park, E.S., Kim, E.J., Kim, M.-J. and Park, J. (2004) Journal of Organic Chemistry, 69, 1972–1977. Martın-Matute, B., Edin, M., Bogar, K. and B€ackvall, J.-E. (2004) Angewandte Chemie International Edition, 43, 6535–6539 DYKAT of 1,4-symmetrical diols: Martın-Matute, B., Edin, M. and B€ackvall, J.-E. (2006) Chemistry-A European Journal, 12, 6053–6061. (a) DYKAT of 1,2-diols: Edin, M., Martın-Matute, B. and B€ackvall, J.E. (2006)

27

28

29

30

31

32

33

34

35 36

37

38 39

Tetrahedron: Asymmetry, 17, 708–715. (b) DYKAT of 3,5-piperidinediols: Olofsson, B., Bogar, K., Fransson, A.B.L. and B€ackvall, J.E. (2006) Journal of Organic Chemistry, 71, 8256–8260. Kim, N., Ko, S.-B., Kwon, M.S., Kim, M.-J. and Park, J. (2005) Organic Letters, 7, 4523–4526. Dijksman, A., Elzinga, J.M., Li, Y.-X., Arends, I.W.C.E. and Sheldon, R.A. (2002) Tetrahedron: Asymmetry, 13, 879–884. Kazlauskas, R.J., Weissfloch, A.N.E., Rappaport, A.T. and Cuccia, L.A. (1991) Journal of Organic Chemistry, 56, 2656–2665. S-selective DKR: (a) Kim, M.J., Chung, Y., Choi, Y., Lee, H., Kim, D. and Park, J. (2003) Journal of the American Chemical Society, 125, 11494–11495. (b) Boren, L., Martın-Matute, B., Xu, Y., Córdova, A. and B€ackvall, J.-E. (2006) Chemistry-A European Journal, 12, 225–232. Berkessel, A., Sebastian-Ibarz, M.L. and M€ uller, T.N. (2006) Angewandte Chemie International Edition, 45, 6567–6570. Akai, S., Tanimoto, K. and Kita, Y. (2004) Angewandte Chemie International Edition, 43, 1407–1410. van As, B.A.C., van Buijtenen, J., Heise, A., Broxterman, Q.B., Verzjil, G.K.M., Palmans, A.R.A. and Meijer, E.W. (2005) Journal of the American Chemical Society, 127, 9964–9965. Hilker, I., Rabani, G., Verzijl, G.K.M., Palmans, A.R.A. and Heise, A. (2006) Angewandte Chemie International Edition, 45, 2130–2132. Reetz, M.T. and Schimossek, K. (1996) Chimia, 50, 668–669. Paetzold, J. and B€ackvall, J.-E. (2005) Journal of the American Chemical Society, 127, 17620–17621. Parvulesc, A., De Vos, D. and Jacobs, P. (2005) Chemical Communications, 5307–5309. Stirling, M., Blacker, J. and Page, M.I. (2007) Tetrahedron Letters, 48, 1247–1250. (a) Tan, D.S., G€ unter, M.M. and Drueckhammer, D.G. (1995) Journal of the

j111


j 4 Dynamic Kinetic Resolutions

112

40

41

42

43

44

45

46

47

American Chemical Society, 117, 9093–9094. (b) Um, P.-J. and Drueckhammer, D.G. (1998) Journal of the American Chemical Society, 120, 5605–5610. Wen, W.-Y., Ng, I.-S. and Tsai, S.-W. (2006) Journal of Chemical Technology and Biotechnology, 81, 1715–1721. (a) Turner, N.J., Winterman, J.R., McCague, R., Parrat, J.S. and Taylor, S.J.C. (1995) Tetrahedron Letters, 36, 1113–1116. (b) Brown, S.A., Parker, M.-C. and Turner, N.J. (2000) Tetrahedron: Asymmetry, 11, 1687–1690. See also: (c) Bevinakatti, H.S. Newadkar, R.V. and Banerji, A.A. (1990) Journal of the Chemical Society, Chemical Communications, 1091–1092. (d) Bevinakatti, H.S., Banerji, A.A., Newadkar, R.V. and Mokashi, A. (1992) Tetrahedron: Asymmetry, 3, 1505–1508. (e) Gu, R.-L., Lee, I.S. and Sih, C.J. (1992) Tetrahedron Letters, 33, 1953–1956. (f) Crich, J., Brieva, R., Marquart, P., Gu, R.-L., Flemming, S. and Sih, C.J. (1993) Journal of Organic Chemistry, 58, 3252–3258. (a) Suzuki, M., Yamazaki, T., Ohta, H., Shima, K., Ohi, K., Nishiyama, S. and Sugai, T. (2000) Synlett, 189–192. (b) See also: Lo, H.-H., Kao, C.-H., Lee, D.-S., Yang, T.-K. and Hsu, W.-H. (2003) Chirality, 15, 699–702. Taniguchi, T. and Ogasawara, K. (1997) Chemical Communications, 1399–1400. (a) Wuyts, S., De Temmerman, D., De Vos, D. and Jacobs, P.A. (2003) Chemical Communications, 1928–1929. (b) Wuyts, S., De Temmerman, K., De Vos, D.E. and Jacobs, P.A. (2005) Chemistry-A European Journal, 11, 386–397. Ödman, P., Wessjohann, L.A. and Bornscheuer, U.T. (2005) Journal of Organic Chemistry, 70, 9551–9555. Inagaki, M., Hiratake, J., Nishioka, T. and Oda, J. (1992) Journal of Organic Chemistry, 57, 5643–5649. (a) Paizs, C., Toxa, M., Majdik, C., T€ahinen, P., Irimie, F.-D. and Kanerva, L.T. (2003) Tetrahedron: Asymmetry, 14, 619–627. (b)

48

49 50

51

52

53

54 55

56

Paizs, C., T€ahinen, P., Lundell, K., Poppe, L., Irimie, F.-D. and Kanerva, L.T. (2003) Tetrahedron: Asymmetry, 14, 1895–1904. (c) Paizs, C., T€ahinen, P., Toxa, M., Majdik, C., Irimie, F.-D. and Kanerva, L.T. (2004) Tetrahedron, 60, 10533–10540. (d) Veum, L., Kanerva, L.T., Halling, P.J., Maschmeyer, T. and Hanefeld, U. (2005) Advanced Synthesis and Catalysis, 347, 1015–1021. Desantis, G., Zhu, Z., Greenberg, W.A., Wong, K., Chaplin, J., Hanson, S.R., Farwell, B., Nicholson, L.W., Rand, C.L., Weiner, D.P., Robertson, D.E. and Burk, M.J. (2002) Journal of the American Chemical Society, 124, 9024–9025. Brand, S., Jones, M.F. and Rayner, C.M. (1995) Tetrahedron Letters, 36, 8493–8496. Yamada, S., Hongo, C., Yoshioka, R. and Chibata, I. (1982) Journal of Organic Chemistry, 48, 843–846. Chen, S.-T., Huang, W.-H. and Wang, D.-T. (1994) Journal of Organic Chemistry, 59, 7580–7581. Parmar, V.S., Singh, A., Bisht, K.S., Kumar, N., Belokon, Y.N., Kochetkov, K.A., Iknooikov, N.S., Orlova, S.A., Tararov, V.I. and Saveleva, T.F. (1996) Journal of Organic Chemistry, 61, 1223–1227. (a) Hacking, M.A.P.J., Wegman, M.A., Rops, J., van Rantwijk, F. and Sheldon, R.A. (1998) Journal of Molecular Catalysis B: Enzymatic, 5, 155–157. (b) Wegman, M.A., Hacking, M.A.P.J., Rops, J., Pereira, P., Van Rantwijk, F. and Sheldon, R.A. (1999) Tetrahedron: Asymmetry, 10, 1739–1750. Liljeblad, A., Kiviniemi, A. and Kanerva, L.T. (2004) Tetrahedron, 60, 671–677. (a) Asano, Y. and Yamaguchi, S. (2005) Journal of the American Chemical Society, 127, 7696–7697. (b) See also: May, O. Verseck, S. Bommarius, A. and Drauz, D. (2002) Organic Process Research and Development, 6, 452–457. (a) Jones, M.M. and Williams, J.M.J. (1998) Chemical Communications, 2519–2520. (b) Haughton, L. and Williams, J.M.J. (2001) Synthesis, 943–946.


References 57 Strauss, U.T. and Faber, K. (1999) Tetrahedron: Asymmetry, 10, 4079–4081. 58 Thuring, J.W.J.F., Klunder, A.J.H., Nefkens, G.H.L., Wegman, M.A. and Zwanenburg, B. (1996) Tetrahedron Letters, 37, 4759–4760.

59 Sharfuddin, M., Narumi, A., Iwai, Y., Miyazawa, K., Yamada, S., Kakuchi, t. and Kaga, H. (2003) Tetrahedron: Asymmetry, 14, 1581–1885. 60 Crawford, J.B., Skerlj, R.T. and Bidger, G.J. (2007) Journal of Organic Chemistry, 72, 669–671.

j113


j115

5 Deracemization and Enantioconvergent Processes Nicholas J.Turner

5.1 Introduction

A dominant theme in modern asymmetric catalysis is the development of processes that involve the total conversion, rather than resolution, of both enantiomers of a racemate into a single enantiomer of the product. Such transformations of racemates become equivalent to asymmetric transformations on prochiral substrates and hence are highly desirable, not least because a racemate is often a convenient starting point for the synthesis of a molecule possessing one or more stereogenic centers. Typically the key functional group that is attached to the stereogenic center will be an alcohol, carboxylic acid/amide, amine, nitrile, epoxide, or thiol and hence biocatalytic methods are continually being sought and developed to manipulate these groups in novel ways [1]. Three distinct methods have been developed for achieving the total conversion of a racemate in high yield and enantiomeric excess (Figure 5.1). The first and by far the most popular approach is dynamic kinetic resolution (DKR), which is dealt with in Chapter 2.1 by Jan-Erling Backvall. In a typical DKR, an enantioselective enzyme is used to transform a racemic mixture in which the two enantiomers of the substrate undergo rapid interconversion via racemization. The present chapter deals with two alternative strategies, the first of which is called deracemization. Deracemization is defined here and elsewhere [2] as a reaction whereby two enantiomers are interconverted by a stereoinversion mechanism, allowing the racemate to be transformed into a single enantiomer without any net change in the constitution of the molecule, that is, there is no ‘product’, but simply an optical enrichment of the substrate. The second approach is based upon enantioconvergent transformations in which a necessary condition is that one enantiomer is transformed with inversion of configuration at the stereogenic center. A number of previous reviews [2–6] have dealt with both deracemization and enantioconvergent processes and hence this chapter will focus primarily on the recent literature.


j 5 Deracemization and Enantioconvergent Processes

116

A

P

A [X]

Racemization + (a)

B

Q

A

B

(b)

Retention

+ (c)

B

P +

Inversion

Q

Figure 5.1 Schematic illustration of (a) dynamic kinetic resolution, (b) deracemization, and (c) enantioconvergent processes.

5.2 Deracemization Processes 5.2.1 Cyclic Oxidation–Reduction Systems

One general type of deracemization process depends upon establishing a catalytic cycle involving the combination of (i) an enzyme-catalyzed enantioselective transformation (oxidation) of a racemic mixture into a stable prochiral intermediate and (ii) the reduction of this intermediate back to the starting material in a separate step (Figure 5.2). Provided that the enzyme is enantioselective, the faster reacting enantiomer will be rapidly depleted, and following a number of cycles, the slower reacting enantiomer will accumulate in the reaction. The final enantiomeric excess will be determined directly by the enantioselectivity (kR/kS) of the enzyme-catalyzed reaction. Assuming that the enzymatic reaction is highly enantioselective, then even after only four cycles the enantiomeric excess will have reached 93.4% whereas after seven catalytic cycles the enantiomeric excess is >99% (Figure 5.3). This type of deracemization is really a stereoinversion process in that the reactive enantiomer undergoes stereoinversion during the process. One of the challenges of developing this type of process is to find conditions under which the enzyme catalyst and chemical reactant can coexist, particularly in the case of redox chemistry in which the coexistence of an oxidant and reductant in the same reaction vessel is difficult to achieve. For this

SR

k red

SR, SS = Substrate enantiomers

kR I

I = Achiral/prochiral intermediate

kS SS

k R , k S, k red = Rate constants

k red k R»k S

Figure 5.2 Cyclic deracemization process involving sequential enzyme-catalyzed oxidation and nonenzymatic reduction.


% Enantiomer

5.2 Deracemization Processes

100 90 80 70 60 50 40 30 20 10 0 0

1

2

3

L-Amino

acid

D-Amino

acid

j117

4

Number of cycles Figure 5.3 Progression of an ideal cyclic deracemization process.

reason, perhaps it is not surprising that the only known examples of this type of cyclic oxidation–reduction system that have been reported involve an enzyme as the catalyst for the oxidation step. The first experimental demonstration of this type of deracemization was provided by Hafner and Wellner [7] who used D-amino acid oxidase in combination with sodium borohydride to generate a small quantity (about 5% conversion) of L-alanine from D-alanine (Figure 5.4). In this process, the D-alanine (1: R ¼ CH3) is initially oxidized to the corresponding imine (2) by the D-amino acid oxidase followed by concomitant reduction of the imine nonselectively, which generates a small quantity of L-enantiomer. Interestingly, Hafner et al. estimated that the reduction of the imine intermediate was about 35· faster than hydrolysis to the corresponding a-keto acid (3). Subsequently, Soda et al. [8,9] recognized the possibility of applying this concept

NH 3+ R

Oxidation acid oxidase

D -Amino

CO2 – (D)-1

NH Reduction R [H]

CO2H 2

NH3 + R

O

+ H 2O

CO 2– (L)-1

Figure 5.4 Deracemization of a-amino acids using D-amino acid oxidase in combination with sodium borohydride.

+ R

CO2H 3

NH4+


j 5 Deracemization and Enantioconvergent Processes

118

to preparative deracemization reactions and starting with both D/L-proline and D/Lpipecolic acid were able to achieve high conversions (>95%) and high enantiomeric excesses (>99%) to the corresponding L-enantiomers using the same D-amino acid oxidase/sodium borohydride combination. The work of Soda highlighted two key questions. First, cyclic amino acids are more suitable for this type of deracemization process in that the oxidized imine is cyclic and hence more stable than the corresponding imine derived from an acyclic amino acid, which is prone to hydrolysis to the a-keto acid. Secondly, excesses of sodium borohydride were required to achieve complete deracemization as a result of the instability of sodium borohydride in water. Subsequently, Turner et al. demonstrated that acyclic, as well as cyclic, amino acids could be deracemized in this manner although the yields were slightly lower (about 80%) owing to competing hydrolysis. More importantly, they discovered that water-stable sodium cyanoborohydride could be used in place of sodium borohydride permitting a large reduction in the number of equivalents (about five) of reducing agent [10]. Thereafter, in collaboration with scientists at Great Lakes in the United States, both amine–borane complexes and catalytic transfer hydrogenation using palladium and ammonium formate were found to be effective methods for in situ reduction of the imine [11]. In addition, for the first time D-amino acids were prepared in high yield and enantiomeric excess by use of the L-amino acid oxidase from Proteus myxofaciens. The catalytic transfer hydrogenation approach has been extensively developed by the company Ingenza (www.ingenza.com), which has now devised cost-effective, large-scale processes for the deracemization of specific amino acids using this approach [12–15]. As outlined above, the oxidation–reduction deracemization process is in essence a stereoinversion and hence could in principle be applied to substrates possessing more than one stereogenic center. Turner et al. demonstrated the interconversion of diasteroisomers of amino acids, for example, the conversion of L-isoleucine (4) to Dallo-isoleucine (5) in high yield and enantiomeric excess/diastereoisomeric excess (Figure 5.5) [16]. In a further illustration of this approach, they collaborated with Dowpharma (Chirotech), Cambridge, UK to combine catalytic asymmetric hydrogenation of dehydroamino acids, using chiral rhodium–DuPhos catalysts, with the oxidase stereoinversion technology to prepare a wide range of b-substituted phenylalanines (6) in high diastereoisomeric excess and enantiomeric excess (Figure 5.6) [17].

NH2 CO2 H Me L-isoleucine

NH 2

Proteus myxofaciens L-AAO

4

CO 2 H

[H]

Me D-allo-isoleucine

NH3 -BH 3; 87% yield; 99% d.e. Pd/C-HCO 2NH4 ; 61% yield; 96% d.e. Figure 5.5 Stereoinversion of L-isoleucine to D-allo-isoleucine.

5


5.2 Deracemization Processes

NHAc Ar

CO2 Me

ii. HCl

Me

NH 2

i. Rh(R,R)-EtDuPhos H2 Ar

D-AAO NH 3:BH 3

j119

NH2 Ar

CO2 H

CO2H

Me

Me 6 e.e.'s >98%

Figure 5.6 Synthesis of b-substituted phenylalanines.

Recently Turner and coworkers have sought to extend the deracemization method beyond a-amino acids to encompass chiral amines. Chiral amines are increasingly important building blocks for pharmaceutical compounds that are either in clinical development or currently licensed for use as drugs (Figure 5.7). At the outset of this work, it was known that type II monoamine oxidases were able to catalyze the oxidation of simple amines to imines in an analogous fashion to amino acid oxidases. However, monoamine oxidases generally possess narrow substrate speciďŹ city and moreover have been only documented to catalyze the oxidation of simple, nonchiral

Me

Me O

Me2 N

N

CO 2 H

O Et

N H

O

OEt

Me2 N

O

N

(S)-rivastigmine

(S)-repaglinide

(S)-dapoxetine(phaseII)

O Me

N HO 2 C

N

O

Cl

O

N N

Et Et

O

(R)-levocetirizine

N

N H O

(S)-DMP 777 (phase II) F2 HC O HN

N

N CO 2H

Ph

O O

N

O (R)-garenoxacin (phase III)

Figure 5.7 Drugs containing chiral amines.

(S)-solifenacin (phase III)

O O


j 5 Deracemization and Enantioconvergent Processes

120

NH 2

NH3 :BH 3

CH3 Fast

NH

(S)-α-methylbenzylamine 7 MAO-N (Asn336Ser) Slow

Prochiral imine intermediate 8

NH 2 CH3 NH3 :BH 3 (R)-α-methylbenzylamine 7 77% yield; 93% e.e.

Figure 5.8 Deracemization of a-methylbenzylamine by using a variant monoamine oxidase in combination with NH3: BH3.

amines. The major objective of this work therefore was to apply directed evolution methods to create amine oxidase variants possessing broad substrate specificity and high enantioselectivity. The initial goal was to identify variant enzymes that are able to carry out the deracemization of the model substrate a-methylbenzylamine (7) (Figure 5.8). Monoamine oxidase N (MAO-N) from Aspergillus niger was expressed in Escherichia coli and a library of variants was generated using the E. coli XL-1 Red mutator strain. The variants were screened for activity against a-methylbenzylamine as substrate on solid phase using a high-throughput colorimetric screen. Using this approach, Alexeeva et al. successfully identified a variant (Asn336Ser) that was able to enantioselectively oxidize (S)-a-methylbenzylamine (7) to the corresponding imine (8) [18]. By contrast, the parent wild-type enzyme had very low activity toward (S)-(7), although there was some evidence that it was enantioselective. Introduction of the point mutation resulted in an improvement of catalytic activity of about 47-fold and in combination with ammonia borane, this variant, amine oxidase, was used to deracemize amethylbenzylamine to give the (R) enantiomer in 77% yield and 93% ee. Subsequently Turner and coworkers were able to show that the Asn336Ser variant possessed broad substrate specificity, with the ability to oxidize a wide range of chiral amines of interest [19]. They also discovered a second mutation, Ile246Met, which conferred enhanced activity toward chiral secondary amines as exemplified by the deracemization of racemic 1-methyltetrahydroisoquinoline (MTQ) (9) (Figure 5.9)[20]. By carrying out iterative rounds of mutagenesis and screening, it has been possible to identify further variants that have high activity toward a broad range of chiral amines including tertiary amines (10–13) (Figure 5.10). Interestingly, in this case it is now possible to enter the catalytic cycle from the precursor acyclic amino ketone rather than simply starting with the racemic tertiary amine. This approach is exemplified by the conversion of ketone (14), which under aqueous conditions at


j121

5.2 Deracemization Processes

Me

Me (S)-amine oxidase variant

NH

NH NH 3:BH 3

(R/S)- 9

(R)-9 95% yield; 99% e.e.

Figure 5.9 Deracemization of MTQ 9.

pH 7.0 forms the iminium ion (15). In situ reduction of (15) with ammonia borane generates racemic nicotine (10), which is then subjected to the usual deracemization process producing the unnatural (R) enantiomer of nicotine in >95% yield and >95% ee. Recently, one of the evolved MAO-N variants has been shown to catalyze the enantioselective oxidation of O-methyl-N-hydroxylamines (e.g. 16) yielding the corresponding E-oxime (17) and recovered (R)-hydroxylamine (16) in enantiomerically pure form (Figure 5.11) [21]. Finally in this section on deracemization via cyclic oxidation/reduction methods, there has been some limited work carried out on the deracemization of secondary alcohols. Soda et al. [22] employed lactate oxidase in combination with sodium borohydride to deracemize D/L-lactate (18) via the intermediate pyruvate (19) (Figure 5.12).

H N

N Me

N 10

N

Me

H

H

MeO

Me

N

N

Me

Me

12

11

13 H N Me

N

(S)-10 NHMe

pH 7

N+

O

Me

N

N 14

(S)-amine oxidase variant NH3 :BH 3 H

15

N N

Me (R)- 10

Figure 5.10 Deracemization of tertiary amines.


j 5 Deracemization and Enantioconvergent Processes

122

OMe

HN

HN

MAO-N pH 7.4, 25 oC

Me

OMe

Me

N

OMe

Me

+

18h (R/ S)-16

(E )-17; 46%

(R)-16; 44%; 99% e.e.

Figure 5.11 Enantioselective oxidation of hydroxylamines.

OH H 3C

NaBH4 CO 2H

L-Lactate

L-Lactateoxidase

18

O H 3C

OH H 3C

CO 2 H

Pyruvate 19 NaBH4

CO 2H

D-Lactate

18

Figure 5.12 Deracemization of D/L-lactate (18).

In a related approach, Adam et al. used glycolate oxidase with D-lactate dehydrogenase for the deracemization of a wide range of racemic a-hydroxy acids (20) (Figure 5.13) [23]. 5.2.2 Microbial Deracemization of Secondary Alcohols Using a Single Microorganism

It is well known that certain microorganisms are able to effect the deracemization of racemic secondary alcohols with a high yield of enantiomerically enriched compounds. These deracemization processes often involve two different alcohol dehydrogenases with complementary enantiospeciďŹ city. In this context Porto et al. [24] have shown that various fungi, including Aspergillus terreus CCT 3320 and A. terreus CCT 4083, are able to deracemize ortho- and meta-ďŹ&#x201A;uorophenyl-1-ethanol in good OH R

O

Glycolate oxidase CO 2H

OH

+ R

CO 2H

R

a-Hydroxyacid 20

D-Lactate Dehydrogenase/NADH

Figure 5.13 Deracemization of a-hydroxyacids (20).

CO 2 H


5.2 Deracemization Processes

O

OH Rhizopus oryzae pH = 7.5–8.5

O

(R/S)- 21

Rhizopus oryzae pH = 4.0–5.0

O

Rhizopus oryzae pH = 7.5–8.5 OH

OH (S)-21

(R)-21 Figure 5.14 Deracemization of benzoin (21).

yields and high enantiomeric excesses. Likewise Demir et al. [25] have reported the deracemization of racemic benzoin (21) using Rhizopus oryzae ATCC 9363 (Figure 5.14). Interestingly, they were able to use the pH of the medium to control the absolute configuration of the enantiomer produced. Thus, at pH 7.5–8.5, the (R) enantiomer was obtained in 73–76% yield and 97% ee whereas at pH 4–5 the (S) enantiomer was produced (71% yield; 85% ee). Chadha et al., have published a series of papers on the deracemization of bhydroxyesters using whole cells of Candida parapsilosis. For example, deracemization of racemic ethyl 2-hydroxy-4-phenylbutanoic acid (22: R ¼ H) yielded the (S) enantiomer in 85–90% yield and >99% ee (Figure 5.15) [26]. Analogous deracemizations of racemic ethyl/methyl esters of mandelic acid similarly yielded the (S) products in high enantiomeric excess [27]. A number of para-substituted analogs were also investigated using immobilized cells of C. parapsilosis ATCC 7330 affording the corresponding (S) enantiomers in moderate yield and excellent optical purity [28]. Several substrates with different substitution patterns on the aromatic ring underwent the reaction, but with generally lower enantioselectivity. Studies using model substrates suggested that the sequential activities of an (R)-specific oxidase and an (S)-specific reductase were responsible for the biotransformation. Likewise, C. parapsilosis was investigated for substrate tolerance in the deracemization reactions with aryl a-hydroxy esters (23) (Figure 5.16) [29]. A range of OH

O

OH O

R rac - 22 R = H, Me, Et, OMe, Cl, NO 2

Candida parapsilosis Aqueous 25°C, 6 hr

O O

R

Figure 5.15 Deracemization of b-hydroxyesters using Candida parapsilosis.

(S)- 22

j123


j 5 Deracemization and Enantioconvergent Processes

124

OH

OH O

R'

O Candida parapsilosis

O

R

R'

O

R (S)-23

rac-23 Figure 5.16 Deracemization of a-hydroxyesters (23).

substrates was studied to examine the electronic and steric effects of substituents on the aryl ring. When compared with nonsubstituted, para-Cl substitution resulted in a yield decrease from 75 to 65%, and only a small change in enantiomeric excess, from 99 to 98%. Yields of all substrates tested ranged from 52 to 93%, and enantiomeric excesses from 3 to 100%. Carnell et al. discovered that whole cells of Cunninghamella echinulata NRRL 1384 were able to deracemize racemic N-(1-hydroxy-1-phenylethyl)benzamide (24) to produce the (R) enantiomer (Figure 5.17) [30]. The deracemization involves fast, highly (S)-selective oxidation, followed by slower, partially (R)-selective reduction of the ketone (25). Optimization by removing competing extracellular amidase/protease activity resulted in 82% yield and 92% ee. 5.2.3 Deracemization of Alcohols Using Two Enzyme/Microorganism Systems

An alternative approach to the microbial deracemization of secondary alcohols is to use two different microorganisms with complementary stereoselectivity. Fantin et al. studied the stereoinversion of several secondary alcohols using the culture supernatants of two microorganisms, namely Bacillus stearothermophilus and Yarrowia lipolytica (Figure 5.18) [31]. The authors tested three main systems for deracemization. First, they used the supernatant from cultures of B. stearothermophilus, to which they added Y. lipolytica cells and the racemic alcohols. Secondly, they used the culture supernatant of Y. lipolytica and added B. stearothermophilus cells and the racemic alcohols. Finally, they resuspended the cells of both organisms in phosphate buffer and added the racemic alcohols. The best results were obtained in the ďŹ rst system with 6-penten-2-ol (26) (100% ee and 100% yield). The phosphate buffer system gave OH

Cunninghamella echinulata NRRL 1384

Ph HN

O

O Ph

+

HN

Ph (Âą)-24

OH

O

Ph HN

Ph 25

O Ph

(R)-24 82%, 92% e.e.

Figure 5.17 Deracemization of alcohol (24) using Cunninghamella echinulata.


5.2 Deracemization Processes

OH Yarrowia lipolytica (S)-26

O

OH

Prochiral ketone

Bacillus stearothermophilus (R)-26

Figure 5.18 Stereoinversion of 6-hexen-2-ol (26) using two microorganism preparations.

100% ee and a respectable 91% yield, whereas the Y. lipolytica supernatant gave poor results (40% ee and 85% yield). Faber et al. have reported a novel process for the overall deracemization of racemic mandelic acid derivatives using a combination of an enantioselective lipase and a mandelate racemase activity from Lactobacillus paracasei (Figure 5.19) [32]. Using this approach, racemates of (27) were enantiomerically enriched using a lipase in organic solvent, followed by racemization of the unreacted enantiomer in buffer. Acylated derivatives (S)-(28) were obtained in yields >50% and >99% ee. Lipases with the opposite enantioselectivity produced (R)-28 in >99% ee. Subsequent chemical deacylation of (28) yielded enantiomerically enriched (27). Other racemization systems that may be amenable to conversion to deracemization processes in future have recently been reported by Faber and coworkers [33]. Resting cells of L. paracasei have been used for biocatalytic racemization of open-chain and cyclic dialkyl-, alkyl-aryl-, and diaryl-substituted acyloins (29/30) (Figure 5.20). Both Pseudomonas sp. lipase

OH R

CO2 H rac-27

Vinyl acetate, i-Pr 2 O, 25 ยบC

OH R

OAc CO2H

+

R

(R)-27

CO2 H (S)-28

Lactobacillus paracasei pH 6.5, 42 ยบC Figure 5.19 Deracemization of mandelic acid derivatives (27).

O

Lactobacillus paracasei DSM20207 OH

()n (S)- 29/30

Buffer pH 7 29: n = 1 30: n = 2

Figure 5.20 Racemization of cyclic acyloins.

O OH ()n (R)- 29/30

j125


j 5 Deracemization and Enantioconvergent Processes

126

O

OBn r ac-31

Rhodococcus sp. CBS 717.73 Buffer pH 8 iso-octane (5:1) 30 °C, 24 h

HO

O

OBn +

HO

(R)-31

OBn (R)-32

O

OBn (S)-31

72% yield; >97% e.e. Figure 5.21 Deracemization of epoxides using Rhodococcus sp.

cyclic acyloins (S)-(29) and (S)-(30) were racemized in a clean fashion with no side products detected. 5.2.4 Epoxides

An interesting system has been reported for the deracemization of racemic epoxides (31) (Figure 5.21) [34]. Rhodococcus sp. R312 was used for the enantiospecific hydrolysis of racemic epoxide (31) to the corresponding diol (R)-(32). The unreacted (R)-(31) in the product mixture was converted to (R)-(32) by acidcatalyzed epoxide ring opening, and subsequent chemical cyclization gave the epoxide (S)-(31) in high yield and enantiomeric excess. The enantioselectivity and conversion were also studied for several other whole cell systems and isolated enzymes under various conditions. 5.2.5 Carboxylic Acids

Ohta et al. have reported some elegant studies into the microbial deracemization of 2-aryl- (33) and 2-aryloxypropanoic acids using growing cells of Nocardia diaphanozonaria JCM 3208 (Figure 5.22) [35]. In both cases, the (R) enantiomer was preferentially obtained from the racemate. Although the reaction was found to be very sensitive to the structure of the substrate, in optimal cases, for example, for 2-(4-chlorophenoxy) propanoic acid, yields of up to 95% were obtained with an associated enantiomeric excess of 97%. Recently this group has reported a series of studies to elucidate the underlying mechanism of the deracemization reaction. Labeling studies, together with the use of specific inhibitors, suggest the participation of three enzymes, namely, an acyl-CoA synthetase, an arylpropionyl-CoA epimerase, and a hydrolase [36]. In a subsequent study using a resting cell system of N. diaphanozonaria JCM3208, deracemization of 4-substituted-2-thiopropanoic acids (34) were


5.3 Enantioconvergent Processes

CH 3

CH 3 acyl-CoA synthetase CO2 H

X

COSCoA

X Hydrolase

(R)-33 Net reaction deracemization

2-arylpropionylCoA epimerase

CH 3

CH3 X

CO2 H

X

Hydrolase

COSCoA

X

(S)-33 Figure 5.22 Deracemization of 2-aryloxypropanoic acids (33).

X

CH 3 S

Nocardia diaphanozonaria resting cells

CO2 H

X

48 h

CH 3 S

CO2 H

34 X = H, Cl Figure 5.23 Deracemization of 4-substituted-2-thiopropanoic acids.

examined (Figure 5.23) [37]. It was found that side oxidation reactions occurred when deracemizations were performed under aerobic conditions. Operation under an atmosphere of inert argon was found to promote the deracemization reactions giving good-to-excellent enantiomeric excesses/yields. Evidence was presented that deracemization is a competitive reaction against the b-oxidation pathway of fatty acids.

5.3 Enantioconvergent Processes

As outlined above, enantioconvergent processes require two separate reaction pathways in order to transform a racemic substrate into a single product enantiomer. This is accomplished by employing a catalyst, which transforms one of the substrate enantiomers to the product with retention of conďŹ guration. Concurrently, another catalyst, with opposite enantioselectivity and opposite regioselectivity, transforms the other substrate enantiomer with inversion of conďŹ guration (Figure 5.24).

j127


j 5 Deracemization and Enantioconvergent Processes

128

kR

SR

PR

SR, SS = Substrate enantiomers PR = Nonracemic product

kS SS

kR, kS = Rate constants

Figure 5.24 Principle of an enantioconvergent process.

5.3.1 Epoxide Hydrolysis

Xu et al. employed crude mung bean (Phaseolus radiatus L.) powder for the enantioconvergent hydrolysis of styrene oxides, particularly p-nitrostyrene oxide (35; R ¼ NO2) (Figure 5.25) [38]. The corresponding diol products (36) are important intermediates for the synthesis of dopamine receptor antagonists. The crude extract employed in this process contains two epoxide hydrolases, mung bean epoxide hydrolase A and B (mbEH A and B), neither of which is enantioselective, but both of which show identical regioselectivities with coefficients greater than 90%. That is, both enzymes convert the (R) epoxide with retention of configuration while converting the (S) epoxide with inversion of configuration. Using this crude mung bean preparation, the production of (R)-p-nitrostyrene oxide was achieved in 82% ee and 83.5% yield. After recrystallization, greater than 99% ee was achieved. Furstoss et al. have reported their studies on the use of an epoxide hydrolase with four styrene oxide derivatives (Figure 5.26) [39]. The (R)-diol (43) was obtained in 91% ee at 100% conversion from racemic (42), demonstrating an enantioconvergent HO

O

OH

Mung bean extract pH 6.5, 40 oC

X

X (R)-36

35 X = H, NO2, Cl

O

77.4 – 88.4% 60.7 – 82.4% e.e. O

O

O O O HO2C

CO2H 40

37

38

39

O 41

Figure 5.25 Enantioconvergent hydrolysis of epoxides (35) to the corresponding diols (36) using mung bean epoxide hydrolase.


5.3 Enantioconvergent Processes

OH O

OH

Solanum tuberosum epoxide hydrolase H2O/1 % DMSO

Cl

Cl (R)-43

(rac)-42

Figure 5.26 Enantioconvergent hydrolysis of racemic epoxide (42).

process. A preparative scale reaction, using 3 g of 42 resulted in an enantiomeric excess of 88% of (R)-(43). 5.3.2 Sulfatases

Faber and coworkers have recently investigated the potential for using enantioselective alkyl sulfatases for the enantioconvergent hydrolysis of racemic sulfates to the corresponding alcohols [40]. A subset of known sulfatases are able to enantioselectively hydrolyze racemic alkyl sulfates to the corresponding alcohols with inversion of configuration, thereby fulfilling a necessary condition of an enantioconvergent process. For example, they studied the alkylsulfatase activity of Rhodococcus ruber DSM 44541. Biotransformation of the racemic sulfate ester (44) gave enantiomerically pure (S)-configured alcohol (45) and unreacted sulfate ester (44), also of (S) configuration (Figure 5.27) [41]. In a subsequent report [42], they used an inverting sulfatase from Rhodococcus sp. RS2 for initial resolution followed by treatment of the sulfate/alcohol mixture with aqueous tert-butyl methyl ether and dioxane in the presence of p-toluenesulfonic acid as a catalyst to effect hydrolysis of the remaining unreacted sulfate ester with retention of configuration. The overall process yielded a single stereoisomeric alcohol in an enantioconvergent process. Finally Faber et al. also reported the screening of a range of hyperthermophilic sulfur-metabolizers for the ability to convert racemic sulfate esters (46a) and (46b) to the corresponding alcohols (47a) and (47b) (Figure 5.28). Sulfolobus acidocaldarius DSM 639 gave the most promising results and, under optimized conditions, hydrolysis of (R)-(46a) proceeded with inversion of configuration to give (S)-(47a) in moderate-to-high yield and excellent enantiomeric excess. Several other substrates were also tested.

OSO3 C5 H11

C2 H5 44

Rhodococcus r uber DSM 44541 Tris-buffer pH 7.5

H

OSO3

C 5H 11

C 2H 5 (S)-44

H C5 H11

OH 2

C2 H5

(S)-45 e.e. = 99%

Figure 5.27 Enantioselective hydrolysis of sulfate esters.

SO4

j129


j 5 Deracemization and Enantioconvergent Processes

130

OSO 3- Na + R

Me

Sulfolobus acidocaldarius DSM 639 Culture medium

rac-46a: R = (CH 2) 5CH 3 rac-46b: R = CH2 Ph

OSO 3-Na+

OH R

Me

(S)-47a 32%, >99% e.e. (S)-47b 42%, >99% e.e.

R

Me (S)-46a (S)-46b

Figure 5.28 Enantioselective hydrolysis of sulfate esters using S. acidocaldarius.

5.4 Conclusions and Future Prospects

Enzyme-catalyzed deracemization and enantioconvergent processes have recently emerged as attractive approaches for the efďŹ cient preparation of chiral building blocks in enantiomerically enriched form. Together with enzyme-mediated DKRs, an increasing toolbox of biocatalytic technologies is now available to complement existing methods based upon kinetic resolution of a racemate or asymmetric transformations of a prochiral substrate. Current developments in enzyme discovery/enzyme evolution together with advances in biocatalyst formulation and rapid process development will signiďŹ cantly impact this area in a short-to-medium term to provide an even wider portfolio of biocatalysts for practical application in academic and industrial laboratories.

References 1 Patel, R.N. (2006) Current Opinion in Drug Discovery and Development, 9, 741. 2 Gadler, P., Glueck, S.M., Kroutil, W., Nestl, B.M., Larissegger-schnell, B., Ueberbacher, B.T., Wallner, S.R. and Faber, K. (2006) Biochemical Society Transactions, 34, 296. 3 Gruber, C.C., Lavandera, I., Faber, K. and Kroutil, W. (2006) Advanced Synthesis and Catalysis, 348, 1789. 4 Turner, N.J. (2004) Current Opinion in Chemical Biology, 8, 114. 5 Kroutil, W. and Faber, K. (1998) Tetrahedron: Asymmetry, 9, 2901. 6 Stecher, H. and Faber, K. (1997) Synthesis, 1, 1. 7 Hafner, E.W. and Wellner, D. (1971) Proceedings National Academy of Sciences, 68, 987.

8 Huh, J.W., Yokoigawa, K., Esaki, N. and Soda, K. (1992) Journal of Fermentation and Bioengineering, 74, 189. 9 Huh, J.W., Yokoigawa, K., Esaki, N. and Soda, K. (1992) Bioscience Biotechnology and Biochemistry, 56, 2081. 10 Beard, T. and Turner, N.J. (2002) Chemical Communications, 246. 11 Alexandre, F.-R., Pantaleone, D.P., Taylor, P.P., Fotheringham, I.G., Ager, D.J. and Turner, N.J. (2002) Tetrahedron Letters, 43, 707. 12 Turner, N.J., Fotheringham, I. and Speight, R. (2004) Innovations in Pharmaceutical Technology, 4, 114. 13 Fotheringham, I. (2005) Speciality Chemicals Magazine, 25, 32.


Further Reading 14 Archer, I., Carr, R., Fotheringham, I., Speight, R. and Turner, N.J. (2005) Chimica E L Industria, 87, 100. 15 Fotheringham, I., Archer, I., Carr, R., Speight, R. and Turner, N.J. (2006) Biochemical Society Transactions, 34, 287. 16 Enright, A., Alexandre, F.-R., Roff, G., Fotheringham, I.G., Dawson, M.J. and Turner, N.J. (2003) Chemical Communications, 2636. 17 Roff, G.J., Lloyd, R.C. and Turner, N.J. (2004) Journal of the American Chemical Society, 126, 4098. 18 Alexeeva, M., Enright, A., Dawson, M.J., Mahmoudian, M. and Turner, N.J. (2002) Angewandte Chemie International Edition, 41, 3177. 19 Carr, R., Alexeeva, M., Enright, A., Eve, T.S.C., Dawson, M.J. and Turner, N.J. (2003) Angewandte Chemie International Edition, 42, 4807. 20 Carr, R., Alexeeva, M., Dawson, M.J., Gotor-Fernandez, V., Humphrey, C.E. and Turner, N.J. (2005) ChemBioChem, 6, 637. 21 Eve, T.S.C., Wells, A.S. and Turner, N.J. (2007) Chemical Communications, in press. 22 Oikawa, T., Mukoyama, S. and Soda, K. (2001) Biotechnology and Bioengineering, 73, 80. 23 Adam, W., Lazarus, M., Saha-M€oller, C.R. and Schreier, Peter (1998) Tetrahedron: Asymmetry, 9, 351. 24 Comasseto, J.V., Omori, Á.T., Andrade, L.H. and Porto, A.L.M. (2003) Tetrahedron: Asymmetry, 14, 711. 25 Demir, A.S., Hamamci, H., Sesenoglu, O., Neslihanoglu, R., Asikoglu, B. and Capanoglu, D. (2002) Tetrahedron Letters, 43, 6447. 26 Chadha, A. and Baskar, B. (2002) Tetrahedron: Asymmetry, 13, 1461.

27 Padhi, S.K., Titu, D., Pandian, N.G. and Chadha, A. (2006) Tetrahedron, 62, 5133. 28 Baskar, B., Pandian, N.G., Priya, K. and Chadha, A. (2005) Tetrahedron, 61, 12296. 29 Padhi, S.K. and Chadha, A. (2005) Tetrahedron: Asymmetry, 16, 2790. 30 Cardus, G.J., Carnell, A.J., Trauthwein, H. and Riermeier, T. (2004) Tetrahedron: Asymmetry, 15, 239. 31 Fantin, G., Fogagnolo, M., Giovannini, P.P., Medici, A. and Pedrini, P. (1995) Tetrahedron: Asymmetry, 6, 3047. 32 Larissegger-Schnell, B., Glueck, S.M., Kroutil, W. and Faber, K. (2006) Tetrahedron, 62, 2912. 33 Nestl, B.M., Kroutil, W. and Faber, K. (2006) Advanced Synthesis and Catalysis, 348, 873. 34 Simeó, Y. and Faber, K. (2006) Tetrahedron: Asymmetry, 17, 402. 35 Kato, D., Mitsuda, S. and Ohta, H. (2002) Organic Letters, 4, 371. 36 Kato, D., Mitsuda, S. and Ohta, H. (2003) The Journal of Organic Chemistry, 68, 7234. 37 Kato, D., Miyamoto, K. and Ohta, H. (2004) Tetrahedron: Asymmetry, 15, 2965. 38 Xu, W., Xu, J.H., Pan, J., Gu, Q. and Wu, X.Y. (2006) Organic Letters, 8, 1737. 39 Monterde, M.I., Lombard, M., Archelas, A., Cronin, A., Arand, M. and Furstoss, R. (2004) Tetrahedron: Asymmetry, 15, 2801. 40 Gadler, P. and Faber, K. (2007) TIBTECH, 25, 83. 41 Pogorevc, M., Kroutil, W., Wallner, S.R. and Faber, K. (2002) Angewandte Chemie International Edition, 41, 4052. 42 Wallner, S.R., Pogorevc, M., Trauthwein, H. and Faber, K. (2004) Engineering Life Sciences, 4, 512.

Further Reading Dunsmore, C.J., Carr, R., Fleming, T. and Turner, N.J. (2006) Journal of the American Chemical Society, 128, 2224.

Wallner, S.R. Nestl, B.M. and Faber, K. (2005) Organic and Biomolecular Chemistry, 3, 2652.

j131


j133

6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides Robert Ch^enevert, Pierre Morin, and Nicholas Pelchat

6.1 Introduction and General Aspects 6.1.1 Scope

This chapter presents a review of recent progress in hydrolase-catalyzed enantioselective transformations of carboxylic acid derivatives, alcohols, and epoxides. In light of the large number of such reactions in the literature, this review cannot be fully comprehensive; instead, the aim is to highlight the range of biotransformations that are possible. Emphasis has been placed on methodological advancements and recent applications to the synthesis of natural products and biologically active compounds. This chapter concentrates mainly on advances made during the period 2000 to mid2006. Unless indicated, only examples where the enantiomeric excess (ee) of the product is higher than 90% are included. This introduction briefly presents general concepts that are essential for an understanding of the subject. For more complete coverage of these concepts and earlier literature, the reader is referred to a series of monographs [1–7]. 6.1.2 Reaction Conditions

Hydrolysis of substrates is performed in water, buffered aqueous solutions or biphasic mixtures of water and an organic solvent. Hydrolases tolerate low levels of polar organic solvents such as DMSO, DMF, and acetone in aqueous media. These cosolvents help to dissolve hydrophobic substrates. Although most hydrolases require soluble substrates, lipases display weak activity on soluble compounds in aqueous solutions. Their activity markedly increases when the substrate reaches the critical micellar concentration where it forms a second phase. This interfacial activation at the lipid–water interface has been explained by the presence of a


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

134

helical oligopeptide (lid) that shields the active site. Upon interaction with a hydrophobic interface, the lid opens and the enzyme is rearranged into its active conformation [8]. Aqueous solutions are not suitable solvents for esterifications and transesterifications, and these reactions are carried out in organic solvents of low polarity [9–12]. However, enzymes are surrounded by a hydration shell or bound water that is required for the retention of structure and catalytic activity [13]. Polar hydrophilic solvents such as DMF, DMSO, acetone, and alcohols (log P < 0, where P is the partition coefficient between octanol and water) are incompatible and lead to rapid denaturation. Common solvents for esterifications and transesterifications include alkanes (hexane/log P ¼ 3.5), aromatics (toluene/2.5, benzene/2), haloalkanes (CHCl3/2, CH2Cl2/1.4), and ethers (diisopropyl ether/1.9, tert-butylmethyl ether/ 0.94, diethyl ether/0.85). Exceptionally stable enzymes such as Candida antarctica lipase B (CAL-B) have been used in more polar solvents (tetrahydrofuran/0.49, acetonitrile/0.33). Room-temperature ionic liquids [14–17] and supercritical fluids [18] are also good media for a wide range of biotransformations. Hydrolase-catalyzed reactions are usually performed over a narrow temperature range near room temperature, but recent studies have shown that some reactions proceed at extreme temperatures as low as 60 C or as high as 120 C [19–21]. Microwave irradiation is an alternative to conventional heating [22]. Hydrolase activity can also be improved through the use of additives such as surfactants, inorganic salts, amines, cyclodextrines, and crown ethers [23]. For large-scale or industrial applications, enzymes are immobilized or confined to membrane-restricted compartments [24–27]. Immobilization methods can be grouped into several categories such as noncovalent binding by adsorption on inert supports, covalent immobilization on a solid carrier, and enzyme cross-linking with glutaraldehyde (CLEC). 6.1.3 Kinetic Resolution, Dynamic Kinetic Resolution, and Desymmetrization

A kinetic resolution depends on the fact that the two enantiomers of a racemic substrate react at different rates with the enzyme. The process is outlined in Figure 6.1, assuming that the (S) substrate is the fast-reacting enantiomer (kS > kR) and krac ¼ 0. In ideal cases, only one enantiomer is consumed and the reaction ceases at 50% conversion. In most cases, both enantiomers are transformed and the enantiomeric composition of the product and the remaining starting material varies with the extent

(S)-Substrate

kS

(S)-Product

Fast k rac (R)-Substrate

k S >>kR kR Slow

(R)-Product

Figure 6.1 Kinetic resolution (krac ¼ 0) and dynamic kinetic resolution (krac  kS).


6.1 Introduction and General Aspects

O (a)

OH R

Hydrolase

R'

(r ac)

O O

O

OH R

R'

C 8F17

+

Organic layer

O

(b)

O–

O R

R'

O

Basic aqueous layer O

O

CAL-B

C 8 F17

O

OH +

(rac )

(S) Fluorous phase

(R) Organic phase

Figure 6.2 Separation of products by (a) cyclic anhydrides as acyl donors and (b) fluorous phase technique.

of conversion. The efficacy of a kinetic resolution is measured by the enantiomeric ratio E ¼ kS/kR [28–30]. Resolutions are useful for synthesis when E > 20. Despite its widespread application [31,32], the kinetic resolution has two major drawbacks: (i) the maximum theoretical yield is 50% owing to the consumption of only one enantiomer, (ii) the separation of the product and the remaining starting material may be laborious. The separation is usually carried out by chromatography, which is inefficient on a large scale, and several alternative methods have been developed (Figure 6.2). For example, when a cyclic anhydride is the acyl donor in an esterification reaction, the water-soluble monoester monoacid is separable by extraction with an aqueous alkaline solution [33,34]. Also, fluorous phase separation techniques have been combined with enzymatic kinetic resolutions [35]. To overcome the 50% yield limitation, one of the enantiomers may, in some cases, be racemized and resubmitted to the resolution procedure. Dynamic kinetic resolution (DKR), a more efficient procedure, combines a kinetic resolution with an in situ racemization of a configurationally labile substrate (Figure 6.1, where krac  kS) [36–38]. DKR has three main requirements: (i) the two enantiomers of the starting material must react at very different rates (kS  kR), (ii) the starting material enantiomers must be in equilibrium (krac  kS), and (iii) the product must be inert to racemization. Until recently, DKR was restricted to specific cases involving racemization through exchange of acidic hydrogen (hydantoins [39], oxazolones [40], thioesters [41]), halogen exchange (a-haloesters [42]), and addition– elimination reactions (cyanohydrins [43], reversible Michael addition [44]). The majority of chiral compounds resolved by hydrolase-catalyzed reactions have been secondary alcohols. Recently, the combination of hydrolases and transition-metal catalysts for DKR of secondary alcohols has attracted considerable attention [45–48]. Rutheniumbased organometallic catalysts are commonly used for the racemization of chiral secondary alcohols via hydrogen transfer (oxidation–reduction) as seen in Figure 6.3. Enantioselective enzymatic desymmetrization is the transformation of a substrate that results in the loss of a symmetry element that precludes chirality (plane of

j135


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

136

Ph H

OH Catalyst

R

R'

Ph

O R

Catalyst

HO

H

R

R'

R'

Catalyst =

Ph

Ph Ph OC Ru Cl CO

Figure 6.3 Racemization of a secondary alcohol in the presence of a ruthenium hydrogen-transfer catalyst.

(a)

A

B

Hydrolase

A

B

(b)

BA

A

B

X

X

Prochiral (plane of symmetry)

B

X

C

Y D

meso, achiral (plane of symmetry)

A

(c)

X D

X Y Chiral

X

BA

A

Hydrolase

C

Chiral

A X

Y Hydrolase

X

X

A Achiral (center of symmetry)

A Chiral

Figure 6.4 Hydrolase-catalyzed desymmetrization of a prochiral (a), a meso (b), or a centrosymmetric (c) substrate.

symmetry, center of symmetry) (Figure 6.4). In contrast to the classic kinetic resolutions of racemic mixtures, the theoretical maximum yield of these transformations is 100%. Enzymatic desymmetrization of prochiral or meso compounds to yield enantiomerically enriched products has proved to be a valuable tool in asymmetric synthesis [49,50]. The desymmetrizations of centrosymmetric substrates are very rare. In an asymmetric synthesis, the enantiomeric composition of the product remains constant as the reaction proceeds. In practice, however, many enzymatic desymmetrizations undergo a subsequent kinetic resolution as illustrated in Figure 6.5. For instance, hydrolysis of a prochiral diacetate ďŹ rst gives the chiral monoalcohol monoester, but this product is also a substrate for the hydrolase, resulting in the production of R

R

(S)

(R)

OH

AcO Fast

R AcO

OAc (S)

Slow

Major

R Hydrolase

(R)

Achiral

R

R

(S)

(R)

HO

Slow

Fast

R

R

(S)

(R)

HO

OAc Minor

Figure 6.5 A desymmetrization coupled to a kinetic resolution.

OH Achiral


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

(S)-Substrate +

(S)-Product

(R)-Substrate Figure 6.6 Enantioconvergent transformation of a racemate.

the achiral diol. This second hydrolysis lowers the yield of the monoester but usually it increases its enantiomeric excess by kinetic resolution. The diacetate and the monoester are similar substrates and the enzymes usually have a preference for the same stereogenic center shown by arrows on Figure 6.5, thus removing the minor enantiomer. The second step is usually slower and the chiral product can be isolated in good yield if the reaction is monitored. When the chiral product is a monoacid monoester (HO2C-R-CO2R0 ), the process terminates after the ďŹ rst step because polar carboxylic acids are hydrated in aqueous medium and are not substrates for hydrolases. 6.1.4 Enantioconvergent Transformation

An enantioconvergent transformation leads to a single enantiomeric product from a racemate [51]. Each enantiomer is transformed via independent pathways by the same catalyst or by two different catalysts (Figure 6.6). For example, the hydrolysis of epoxides may proceed with high regio- and stereoselectivity with inversion or retention of conďŹ guration. Several enantioconvergent transformations of epoxides are reported in the last section of this chapter.

6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives 6.2.1 Ester Hydrolysis

Esterases, proteases, and some lipases are used in stereoselective hydrolysis of esters bearing a chiral or a prochiral acyl moiety. The substrates are racemic esters and prochiral or meso-diesters. Pig liver esterase (PLE) is the most useful enzyme for this type of reaction, especially for the desymmetrization of prochiral or meso substrates. The protected E-ring moiety of (S)-camptothecin has been prepared in enantiomerically enriched form by the enzymatic resolution of a triester with PLE in a pH 7 phosphate buffer-acetonitrile (5 : 1) solution (Figure 6.7). The alkaloid camptothecin is an inhibitor of the enzyme topoisomerase and some of its derivatives are anticancer drugs [52]. The resolution of racemic esters via selective hydrolysis catalyzed by hydrolases is a practical way to prepare optically active pharmaceutical intermediates as shown in Figure 6.8 [53,54].

j137


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

138

O

O

O

O

O

PLE O

EtO

CO 2Me OBn

N

O CO 2Me

EtO Phosphate buffer CH 3CN

N

OBn

HO

(R) + Mixture of acids

(r ac)

Camptothecin

O O

Figure 6.7 Chemoenzymatic synthesis of the E-ring portion of camptothecin.

O

H

CO2Me

CO2 Et

NH

(R) Asper gillus melleus protease buffer

S

t -Bu

O N

(R) K lebsiella oxyt oca hydrolase buf fer

(R) Bacillus brevis protease toluene buffer

CO 2(CH 2) 2OMe

Figure 6.8 Resolution of pharmaceutical intermediates (hydrolyzed enantiomer shown).

Several a-amino acids have been prepared from two-substituted malonates through an enzymatic desymmetrization Curtius rearrangement sequence (Figure 6.9). Hydrolysis of a trifluoroethyl cyclopropanedicarboxylate with PLE in pH 7 phosphate buffer provided the (R)-monoacid, which was rearranged to a cyclopropane a-amino acid via a Curtius-type reaction with diphenylphosphoryl azide [55]. The same strategy was applied to the synthesis of orthogonally protected (R)- and (S)-2methylcysteine. Desymmetrization by selective hydrolysis of a malonate with PLE in pH 7.2 phosphate buffer was followed by a Curtius rearrangement to selectively form the carbon–nitrogen bond. Interconversion of the ester and acid functionalities before the rearrangement allows the preparation of the other enantiomer [56]. The antiviral agent virantmycin is an unusual chlorinated tetrahydroquinoline isolated from a strain of Streptomyces (Figure 6.10). Hydrolysis of a prochiral 2,2disubstituted dimethyl malonate with PLE in DMSO-pH 8 phosphate buffer (1 : 4) was a key step in a stereodivergent synthesis of this natural product [57]. A [2 þ 2] photoaddition–cycloreversion was applied to the enantioselective synthesis of the natural product byssochlamic acid (Figure 6.11). Desymmetrization of a meso-cyclopentene dimethyl ester with PLE in pH 7 buffer-acetone (5 : 1) provided a monoacid, one of the photopartners. It is noteworthy that both enantiomers of this natural product were synthesized from the same monoacid [58]. CO 2CH2 CF3 CO 2CH2 CF3 t-BuS MeO2 C

PLE

CO 2H CO 2CH 2CF3

Buffer

PLE CO 2Me Buf fer

NH2

t-BuS

t-BuS HO2C

CO2 H

CO2 Me

H 2N

CO2 Me

Figure 6.9 Chemoenzymatic synthesis of amino acids via desymmetrization of malonates.


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

O MeO

OMe O

O

PLE

HO2C

DMSO buffer MeO

O

Cl

OH O

j139 O

N H

O (–)-Virantmycin

(S) Figure 6.10 Chemoenzymatic synthesis of virantmycin.

MeO 2C

PLE

MeO 2C

O

O

HO2 C O

O

H 2O–acetone MeO2 C Byssochlamic acid

O O

Figure 6.11 Chemoenzymatic synthesis of byssochlamic acid.

A divergent synthesis of tropane alkaloid ferruginine was reported by Node and coworkers [59]. The b-ketoester intermediate was prepared by a novel PLE-catalyzed asymmetric dealkoxycarbonylation (hydrolysis followed by a decarboxylation) of a symmetric tropinone-type diester (Figure 6.12). Dimethyl sulfoxide was added to the phosphate buffer pH 8 (1 : 9) to reduce the activity of PLE and prevent over-dealkoxycarbonylation leading to tropinone. Several lipases were more efficient than PLE and subtilisin Carlsberg for the desymmetrization of an N-t-butoxycarbonyl (Boc) meso-piperidine diester (Figure 6.13). The (3R)-monoester was converted into optically pure isogalactofagomine, a potent galactosidase inhibitor [60].

Bn

Bn

N

PLE

CO 2Et

Me

N

N

CO2 Et

DMSO buf fer EtO2 C O

O

O (+)-Ferruginine

Figure 6.12 Chemoenzymatic synthesis of (þ) or ()-ferruginine.

OH

OH CO2Me

MeO2C N Boc

Lipases H 2O

MeO 2C

OH CO 2H

3

OH

HO

5

N Boc (3R)

Figure 6.13 Chemoenzymatic synthesis of isogalactofagomine.

N H Isogalactof agomine


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

140

i-Bu

O

CAL-B O

i-Bu

i-Bu

O O n-Hexyl

n-Hexanol toluene

(r ac)

O

+ O

(R)

(S)

Figure 6.14 Alcoholysis of ibuprofen vinyl ester.

6.2.2 Ester Alcoholysis

Ester alcoholysis (transesterification) in organic media is an equilibrium reaction and must be shifted in the desired direction. For example, Bornscheuer and coworkers [61] reported the resolution of ibuprofen vinyl ester by transesterification with n-hexanol in the presence of CAL-B. The vinyl alcohol generated during the reaction tautomerizes to acetaldehyde, thus making the reaction irreversible, as illustrated in Figure 6.14. In some cases, the use of a large excess of alcohol is an option to drive the reaction to completion. Alcoholysis of glutamic acid dimethyl ester derivatives with acylase I was regio- and enantioselective (Figure 6.15). An excess of butanol was used as nucleophile and solvent [62]. Lipase-catalyzed methanolysis of racemic N-benzyloxycarbonyl (Cbz) amino acid trifluoroethyl esters carrying aliphatic side chains afforded the L-methyl esters and the D-trifluoromethyl esters (Figure 6.16). The released alcohol (CF3CH2OH) is a weak nucleophile that cannot attack the ester product. The nucleophilicity of the leaving group is depleted by the presence of an electron-withdrawing group [63]. 6.2.3 Esterification of Carboxylic Acids

The direct biocatalytic esterification of a chiral acid with a simple achiral alcohol in organic media is a reversible process and, in order to bias the equilibrium to the

CO 2Me Acylase I R

CO2 Me (r ac)

BuOH

CO2 Me

CO2 Me +

R

R = PrCONH, CbzNH

CO 2Me (R)

R

CO 2Bu (S)

Figure 6.15 Alcoholysis of glutamic acid derivatives.

R

CO 2CH2CF 3 MeOH (4 equiv) R NHCbz (DL)

Carica papaya lipase

CO 2 CH3 NHCbz (L)

R

CO 2CH2CF 3

+

+ CF 3CH2 OH

NHCbz (D)

Figure 6.16 Enantioselective transesterification of N-Cbz-amino acid 2,2,2-trifluoroethyl esters.


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

O

O Burk holderia cepacia OH lipase

O OC 4H 9

1-Butanol hexane Na 2SO4

O (r ac)

j141 OH

+

O (S)

O (R)

Figure 6.17 Biocatalytic esterification in the presence of sodium sulfate as a drying agent.

product side, one of the reactant (alcohol) must be used in excess or one of the products (water) must be removed constantly during the reaction. (R)-3-Phenoxybutanoic acid and the corresponding butyl (S)-ester were obtained by Burkholderia cepacia lipase–catalyzed enantioselective esterification of the racemic acid with 1-butanol in hexane containing anhydrous sodium sulfate to remove the water produced during the reaction (Figure 6.17) [64]. Orthoformates have been used in the lipase-catalyzed esterification aimed at the kinetic resolution of racemic acids such as flurbiprofen, a nonsteroidal anti-inflammatory drug (Figure 6.18). Orthoformates trap the water as it is formed through hydrolysis, and therefore prevent the reverse reaction, and, at the same time, provide the alcohol for the esterification [65]. Several racemic methyl decanoic acids were resolved by esterification with 1-hexadecanol in cyclohexane using immobilized Candida rugosa lipase (CRL) as the catalyst (Figure 6.19). The enantioselectivity was high even when the methyl group was as remotely located as in 8-methyldecanoic acid. Furthermore, for the substrates bearing the methyl group on an even-numbered carbon (i.e. for 2-, 4-, 6-, and 8-methyl decanoic acids), the enzyme showed (S) enantioselectivity and for

CO2H

+

MeCN, 45°C HC(OPr) 3

F (rac)

HC(OPr) 3 + H 2 O

CO2 H

CO2Pr

CAL-B F

F

HCO2 Pr + 2PrOH

Figure 6.18 Irreversible esterification in the presence of an orthoformate.

CO2H n

m

(r ac) Methyldecanoic acid

CRL C 16 H 33 OH Cyclohexane

(S)-2-,-4-,-6-,-8-Methyldecanoic ester + (R)-2-,-4-,-6-,-8-Methyldecanoic acid (R)-3-,-5-,-7-Methyldecanoic ester + (S)-3-,-5-,-7-Methyldecanoic acid

Figure 6.19 Resolution of 2- to 8-methyldecanoic acids.


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

142

OH

PSL O

O (r ac)

R

H2 O, buff er

O +

R

OH

O

(R)

R

O (S)

R = C5 H11, C6 H 13 Figure 6.20 Enantioselective hydrolysis of lactones.

odd-methyldecanoic acids (i.e. for 3-, 5-, and 7-methyl decanoic acids), the selectivity was reversed for the (R) substrate [66]. 6.2.4 Lactones

Enzyme-catalyzed stereoselective hydrolysis allows the preparation of enantiomerically enriched lactones. For instance, Pseudomonas sp. lipase (PSL) was found to be a suitable catalyst for the resolution of d-undecalactone and d-dodecalactone (Figure 6.20). Relactonization of the hydroxy acid represents an efficient method for the preparation of both enantiomers of a lactone [67]. Asymmetric alcoholyses catalyzed by lipases have been employed for the resolution of lactones with high enantioselectivity. The racemic b-lactone (oxetan-2-one) illustrated in Figure 6.21 was resolved by a lipase-catalyzed alcoholysis to give the corresponding (2S,3S)-hydroxy benzyl ester and the remaining (3R,4R)-lactone [68]. Tropic acid lactone was resolved by a similar procedure [69]. These reactions are promoted by releasing the strain in the four-membered ring. Lipases also catalyze the intramolecular transesterification (lactonization) of hydroxy esters. Macrolactonization of a racemic hydroxy ester in the presence of PSL provided the corresponding (R)-lactone (Figure 6.22). This compound is the naturally occurring enantiomer of the pheromone produced by the merchant grain beetle [70]. Chemical macrolactonizations require high dilution to minimize

C 3H 7

Lipase PS O

O (r ac)– Tr ans

C 3H 7 +

O

BnOH

C 3 H7

OBn OH

O (3R,4R)

O

(2S,3S)

Figure 6.21 Enzymatic alcoholysis of a b-lactone.

OH PSL 5

(rac)

CO2 Me

O Isooctane 60°C

Figure 6.22 Biocatalytic macrolactonization.

O (R)-Pheromone

+ (S)-Hydroxy ester


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

OH PPL BnO

OBn Et2O O

CO2Bn O

O

O

H

O

O

(S)-γ -Jasmolactone

Figure 6.23 Desymmetrization via a lipase-catalyzed lactonization.

Ph

Ph

H N

H N

O O

Ph

Ph

O O

O

CAL-B Ph

CH 3OH CH 3CN

Ph

N CO2 CH 3 H 88% (98% ee)

Figure 6.24 DKR of 2-phenyl-4-benzyl-5(4H)-oxazolone.

the competing oligomerizations. Biocatalytic macrolactonizations are temperaturesensitive and at lower temperatures (<45 C), oligomeric esters are the dominant products but better yields of lactones are obtained when the temperature is raised to 60–70 C, which indicate the higher energy of activation for lactonization. The first asymmetric synthesis of ()-g-jasmolactone, a fruit flavor constituent, was achieved via the enantioselective lactonization (desymmetrization) of a prochiral hydroxy diester promoted by porcine pancreas lipase (PPL) (Figure 6.23) [71]. Oxazolones (azlactones) are a form of activated lactones, so they are included in this section. CAL-B is an effective catalyst for the DKR of various racemic foursubstituted-5(4H)-oxazolones, in the presence of an alcohol, yielding optically active N-benzoyl amino acid esters as illustrated in Figure 6.24 [40]. Enantioselective biotransformations of lactides [72,73] and thiolactones [74] have also been reported. 6.2.5 Anhydrides

Enantioselective alcoholysis of racemic, prochiral, or meso cyclic anhydrides can be catalyzed by hydrolases, yielding the corresponding monoesters (Figure 6.25). In most cases, the enantioselectivity was moderate [75–77]. Organometallic catalysts or organocatalysts such as cinchona alkaloids are often more efficient than enzymes for the stereoselective ring opening of cyclic anhydrides.

R2 R1

O

R1 Lipase R3 OH

R2

O O

O

Solvent R3 O

Figure 6.25 Enantioselective ring opening of anhydrides.

O *

OH

j143


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

144

Nitrile hydratase (NHase) R C N H2O

O R C NH 2

O R C OH

Amidase

+ NH 3

H2O

Nitrilase 2H2O Figure 6.26 Two pathways for biocatalytic hydrolysis of nitriles.

6.2.6 Hydrolysis of Nitriles

The hydrolysis of nitriles to carboxylic acids, which proceeds by way of the intermediate amide, requires harsh conditions such as strong bases or acids at a high temperature. Under appropriate conditions, the intermediate amide may be isolated as the major product. The transformation of nitriles by biocatalytic methods offers many advantages, including mild reaction conditions, chemoselectivity, regioselectivity, stereoselectivity, and the absence of by-products [78–81]. The biocatalytic hydrolysis of nitriles may follow two different pathways as illustrated in Figure 6.26. Nitrile hydratases (NHases, EC 4.2.1.84, classified in the lyase group) catalyzes the hydration of a nitrile to the corresponding amide, which may be converted by amidases (EC 3.5.1.4) to the carboxylic acid and ammonia. Nitrilases (EC 3.5.5.1) transform nitriles directly into acids. The lack of commercially available purified enzymes has prompted the use of crude enzyme preparations or whole microbial cells. One particularly interesting application of nitrile biohydrolysis is the largescale industrial production of acrylamide from acrylonitrile [82]. In the nitrile hydratase/amidase system, the enantiodiscrimination predominantly occurs during the hydrolysis of the intermediate amide by the amidase. Nitrilases and some nitrile hydratases can also be enantioselective. A wide range of a-alkyl arylacetonitriles were resolved by nitrile hydrolyzing enzymes [78]. 2-Arylpropionitriles were hydrolyzed to the corresponding arylpropionic acids related to nonsteroidal anti-inflammatory drugs such as ibuprofen. The biotransformation of 2-aryl-4pentenenitriles by Rhodococcus sp. microbial cells provided enantiopure 2-aryl-4pentenoic acids and their amide derivatives [83]. The potential applications of this reaction have been demonstrated by the preparation of (R)-()-baclofen, a specific g-aminobutyric acid receptor agonist and a potent antispastic agent (Figure 6.27). Enantioselective transformations of several cyclopropane or oxirane-containing nitriles were studied using nitrile-transforming enzymes [78]. Microbial Rhodococcus sp. whole cells containing a nitrile hydratase/amidase system hydrolyzed a number CN Rhodococcus sp.

Buffer

Cl ( rac )

H2 N

CONH 2 (S)-Acid

+ Cl

Cl (R )

Figure 6.27 Chemoenzymatic synthesis of (R)-()-baclofen.

CO2 H (R )-Baclofen


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

O Ar

O

Rhodococcus sp. CN

Buffer

(r ac)-tr ans

O CONH2 +

Ar

(2R,3S)

Ar

Ar

CO 2H

CO 2

(2S,3R)

of racemic trans-2,3-epoxy-3-arylpropionitriles to afford (2R,3S)-2-arylglycinamides (Figure 6.28). The arylglycidic acids underwent spontaneous decomposition to arylacetaldehydes under reaction conditions. The overall enantioselectivity originated from the combined effects of a dominantly high (2S)-selective amidase and low (2S)-selective nitrile hydratase [84]. Both cis- and trans-chrysanthemic nitriles and amides were resolved into highly enantiopure amides and acids by Rhodococcus sp. whole cells [85]. The overall enantioselectivity of reactions of nitriles originated from the combined effects of a higher (1R)-selective amidase and a (1R)-selective nitrile hydratase (Figure 6.29). Chrysanthemic acids are related to constituents of pyrethrum flowers and insecticides. The addition of HCN to aldehydes or ketones produces cyanohydrins (a-hydroxy nitriles). Cyanohydrins racemize under basic conditions through reversible loss of HCN as illustrated in Figure 6.30. Enantiopure a-hydroxy acids can be obtained via the DKR of racemic cyanohydrins in the presence of an enantioselective nitriletransforming enzyme [86–88]. Many nitrile hydratases are metalloenzymes sensitive to cyanide and a nitrilase is usually used in this biotransformation. The DKR of mandelonitrile has been extended to an industrial process for the manufacture of (R)mandelic acid [89]. The synthesis of a-amino acids by reaction of aldehydes or ketones with ammonia and hydrogen cyanide followed by hydrolysis of the resulting a-aminonitrile is called the Strecker synthesis. Enzymatic hydrolysis has been applied to the kinetic resolution of intermediate a-aminonitriles [90,91]. The hydrolysis of (rac)-phenylglycine nitrile

R'

Rhodococcus sp.

CONH2

R

+

HO 2C

R

Buffer R

(r ac)–tr ans

R = CH 3 or Cl R' = CN, CONH 2

(1S)

R

(1R)

Figure 6.29 Biotransformation of chrysanthemic nitriles and amides.

OH R

CN (S)

O R

H + HCN

OH R

CN (R)

OH

Nitrilase H2 O

R

Figure 6.30 Nitrilase-catalyzed dynamic kinetic resolution of cyanohydrins.

CHO +

Figure 6.28 Enantioselective hydrolysis of racemic trans-2,3-epoxy-3-arylpropanenitriles.

R

j145

CO2 H

R


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

146

NH2

NH2

Rhodococcus sp.

Ph CN (rac)

Nitrile hydratase

C Ph (r ac) O

NH 2

Amidase

NH 2

Ph

C

NH 2 NH2 +

Ph

C

OH

(L) O

(D) O

Figure 6.31 Kinetic resolution of phenylglycine nitrile.

OH NC

CN

Nitrilase

OH NC

OH

EtOH CO 2H

(R)

H+

NC

CO2 Et

Figure 6.32 Desymmetrization of a meso-dinitrile by a nitrilase.

catalyzed by a Rhodococcus strain containing a nonstereoselective nitrile hydratase and a highly selective amidase gave enantiomerically pure D-phenylglycine amide and L-phenylglycine [91]. This example is illustrated in Figure 6.31. The biocatalytic differentiation of enantiotopic nitrile groups in prochiral or meso substrates has been studied by several research groups. For instance, the nitrilasecatalyzed desymmetrization of 3-hydroxyglutaronitrile [92,93] followed by an esteriďŹ cation provided ethyl-(R)-4-cyano-3-hydroxybutyrate, a useful intermediate in the synthesis of cholesterol-lowering drug statins (Figure 6.32) [94,95]. The hydrolysis of prochiral a,a-disubstituted malononitriles by a Rhodococcus strain expressing nitrile hydratase/amidase activity resulted in the formation of (R)-a,a-disubstituted malonamic acids (Figure 6.33) [96]. 6.2.7 Hydrolysis of Amides, Lactams, and Hydantoins

The main application of the enzymatic hydrolysis of the amide bond is the enantioselective synthesis of amino acids [4,97]. Acylases (EC 3.5.1.n) catalyze the hydrolysis of the N-acyl groups of a broad range of amino acid derivatives. They accept several acyl groups (acetyl, chloroacetyl, formyl, and carbamoyl) but they require a free a-carboxyl group. In general, acylases are selective for L-amino acids, but Dselective acylase have been reported. The kinetic resolution of amino acids by acylasecatalyzed hydrolysis is a well-established process [4]. The in situ racemization of the substrate in the presence of a racemase converts the process into a DKR. Alternatively, the remaining enantiomer of the N-acyl amino acid can be isolated and racemized via the formation of an oxazolone, as shown in Figure 6.34. A recent example of the acylase method is shown in Figure 6.35. N-Benzyloxycarbonylproline is resolved by a proline acylase in a strain of Arthrobacter sp. The

n

Ar NC (r ac)

Rhodococcus sp. CN Nitrile hydratase n = 1, 2

n

Ar H 2 NOC

Amidase CONH2

(r ac)

Figure 6.33 Desymmetrization of prochiral malononitriles.

n

Ar HO2C (R)

CONH2


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives

Racemase (in situ)

COOH

COOH

Acylase

NHAc

H2 N

Buffer

R (r ac )

COOH NHAc R ( D)

+ R

( L)

Ac2 O R

O

R

H 2O

O

Base N

O

N

O

Figure 6.34 Resolution of N-acyl amino acids by acylases.

CO2 H

N O

(DL)

O

Acylase

CO2 H

N

Ar throbact er sp. O

O

(D)

Ph

N H

+

CO 2H

(L)

+ HO

Ph

Ph

+

CO2

Figure 6.35 Acylase-catalyzed resolution of Cbz-(D,L)-proline.

carbamic acid formed during the reaction spontaneously decomposes into CO2 and benzyl alcohol [98,99]. Penicillin G acylase (PGA, EC 3.5.1.11, penicillin G amidase) catalyzes the hydrolysis of the phenylacetyl side chain of penicillin to give 6-aminopenicillanic acid. PGA accepts only phenylacetyl and structurally similar groups (phenoxyacetyl, 4-pyridylacetyl) in the acyl moiety of the substrates, whereas a wide range of structures are tolerated in the amine part [100]. A representative selection of amide substrates, which have been hydrolyzed in a highly selective fashion, is depicted in Figure 6.36. The natural role of peptide deformylases (EC 3.5.1.31) is the hydrolysis of the N-terminal formyl group from nascent peptides. This enzyme combines a good

O O

NH O

HO

R

O Ph

HN

O R'

Ph (S, S) R = Me, Et, CH 2CH=CH2 , Bn

(R) R' = CH 2OH, COOH

Figure 6.36 Examples of amides resolved by penicillin G acylaseâ&#x20AC;&#x201C; catalyzed hydrolysis (the fast-reacting enantiomer is shown).

Ph

OEt N H

(R) R'' = H, TMS

R''

j147


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

148

R

CO2 H

HN (r ac)

H

R

Peptide deformylase

CO2 H

HN

H 2O

H

(D )

O

R

CO 2H

+

NH2 (L)

O

Figure 6.37 Peptide deformylase-catalyzed hydrolysis of N-formyl amino acids.

activity with a very high enantioselectivity in the resolution of N-formylated a- and b-amino acids, a-amino acid amides and a-aminonitriles (Figure 6.37) [101]. Racemic a-amino amides and a-hydroxy amides have been hydrolyzed enantioselectively by amidases. Both L-selective and D-selective amidases are known. For example, a purified L-selective amidase from Ochrobactrum anthropi combines a very broad substrate specificity with a high enantioselectivity on a-hydrogen and a,a-disubstituted a-amino acid amides, a-hydroxyacid amides, and a-N-hydroxyamino acid amides [102]. A racemase (a-amino-e-caprolactam racemase, EC 5.1.1.15) converts the D-aminopeptidase-catalyzed hydrolysis of a-amino acid amides into a DKR (Figure 6.38) [103]. The hydrolysis with a (R)-specific amidase from Klebsiella oxytoca resolved 2-hydroxy- and 2-amino-3,3,3-trifluoro-2-methylpropionamide, giving the (R)-acid and the remaining (S)-amide (Figure 6.39) [104,105]. b-Lactamases (EC 3.5.2.6) produced by bacteria cleave the b-lactam ring and are responsible for their resistance to b-lactam antibiotics. Lactamases are useful catalysts for the enantioselective hydrolysis of b-lactams and other cyclic amides. b-lactams shown in Figure 6.40 were resolved by whole-cell systems containing an amidase [106]. 2-Azabicyclo[2.2.1]hept-5-en-3-one, a bicyclic g-lactam, is an intermediate for the synthesis of antiviral carbocyclic nucleosides (Figure 6.41). This compound was resolved using a cloned lactamase in an industrial-scale process [106,107]. Lipases are generally inactive on the amide bond. One notable exception is the hydrolysis and alcoholysis of b-lactams by CAL-B [108–112]. For example, CAL-B O

NH 2

H2 N

O

NH 2

Racemase

H

H

R (L )

O

NH 2

H

R (D )

X

Amidase NH 2 O

(r ac)

H 2O X = OH, NH2

NH2 R (D )

Figure 6.38 Dynamic kinetic resolution of amino acid amides.

F 3C

OH

D -Aminopeptidase

X

X HO

CF3

+

F3C

NH 2

O

O

(R)

(S)

Figure 6.39 Resolution of 2-hydroxy- and 2-amino-3,3,3-trifluoro2-methylpropionamide by an amidase.


6.2 Enantioselective Biotransformations of Carboxylic Acid Derivatives R

S

NH

S

O

R

O

j149

S

NH

O

R

N H

Figure 6.40 b-Lactams resolved by lactamase-catalyzed hydrolysis (the reactive enantiomer is shown).

N H (rac)

O Lactamase

HN

NH

NH2 +

HO2C

N N

O

(+)

N

(–)

HO

N

NH 2

Abacavir

Figure 6.41 A bicyclic g-lactam resolved by a lactamase.

CO2 H

CAL-B O N H (r ac)

NH 2

H 2O/i-Pr2O 60°C

O

+

HN

(1R,2S)-Cispentacin

(1S,5R)

Figure 6.42 Lipase-catalyzed kinetic resolution of a b-lactam.

catalyzed the enantioselective ring opening of several b-lactams in water-saturated organic media at 60 C as exemplified in Figure 6.42 [110]. Racemic hydantoins result from the reaction of carbonyl compounds with potassium cyanide and ammonium carbonate or the reaction of the corresponding cyanohydrins with ammonium carbonate (Bucherer–Bergs reaction). Hydantoins racemize readily under basic conditions or in the presence of hydantoin racemase, thus allowing DKR (Figure 6.43). Hydantoinases (EC 3.5.2.2), either isolated enzymes or whole microorganisms, catalyze the hydrolysis of five-substituted O

H N

Spontaneous racemization (OH – )

O

H N

O R

N H ( L)

O or hydantoin racemase

R

N H (D )

HO

O

R

N H

D-selective

Hydantoinase

O

N-Carbamoyl-L-Amino

acid CO 2H NH 3 + CO 2 +

NH2

R NH 2 D-Amino acid

HNO2/HCl or carbamoylase

Figure 6.43 Dynamic kinetic resolution of (rac)-hydantoins by a D-hydantoinase.


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

150

hydantoins to N-carbamoyl a-amino acids [39,113–115]. Most hydantoinases are D-enantioselective and convert hydantoins into nonnatural D-amino acid derivatives. The N-carbamoyl group is removed chemically (HNO2/HCl) or biochemically with a carbamoylase (EC 3.5.1.d, carbamoyl amino acid amidohydrolase). D-Phenylglycine and D-4-hydroxyphenyl glycine for the semisynthetic penicillins are produced commercially by the hydantoinase method.

6.3 Enantioselective Enzymatic Transformations of Alcohols

The principal methods for the hydrolase-promoted synthesis of enantiomerically pure alcohols are depicted in Figure 6.44. Biocatalytic acylation and alcoholysis have been reviewed recently [116,117]. Lipases, esterases, and proteases catalyze these reactions, but CAL-B [118–120], CRL [121,122], and diverse lipase preparations from Pseudomonas species are common place. In hydrolytic reactions, water (solvent and nucleophile) is in large excess and the equilibrium is shifted toward the products. In contrast, alcoholysis of esters and acylation of alcohols are reversible transesterifications. The use of an excess of the nucleophile (alcohol) or the removal of the ester product are options to drive alcoholysis to completion. For acylation reactions, the use of activated esters shifts the equilibrium in favor of the products [123,124]. In the case of quasi-irreversible acyl donors such as 2-haloethyl, cyanomethyl, and oxime esters, the released alcohol is a weak nucleophile that cannot attack the ester product. The nucleophilicity of the leaving group is depleted by the presence of electron-withdrawing groups. The most common activated acyl donors are enol esters such as vinyl acetate (VA), isopropenyl acetate (IPA), and 1-ethoxyvinyl esters. The product alcohol is an enol that tautomerizes to a carbonyl compound (acetaldehyde or acetone) or ethyl acetate. VA is more reactive but the liberated acetaldehyde slowly inactivates some lipases (e.g. CRL), probably by formation of imines with side chains of lysines. 1-Ethoxyvinyl esters are as reactive as vinyl esters and the leaving group, ethyl acetate, is inert toward enzymes [125]. Anhydrides (linear, cyclic, mixed carboxylic-carbonic) are also useful acyl donors for irreversible acylations (Figure 6.45). O R*

O

Hydrolysis

+

H 2O

+

R''-OH

R O

R*

O

+

R-CO2 H

R*-OH

+

R''

O

Alcoholysis R

Organic solvent O

R*-OH

R*-OH

+ R'

O

R* Organic solvent

R

O

Acylation R

O

+ O

Figure 6.44 Types of enzyme-catalyzed reactions leading to enantiopure alcohols. Stereogenic centers are included in the alcohol moiety, as indicated by an asterisk.).

R

R'-OH


6.3 Enantioselective Enzymatic Transformations of Alcohols

O R*-OH

+

R

R'

O O

R*

R'

+ R

O

O

HO

R'

R' = H, CH 3 , O Alkyl O R*-OH

+

R

O

O O

R* R'

+ O

R

R'-CO2 H

Figure 6.45 Activated acyl donors for irreversible acylation of alcohols.

Acyl migrations prevailing under aqueous or protic conditions are slower in aprotic organic media. Accordingly, acylation usually gives better results than hydrolysis. Aromatic acyl donors are used when the product is sensitive to acyl migration. Aromatic acyl groups migrate less than aliphatic ones, since the formation of the tetrahedral intermediate involves a greater loss in resonance energy. When both the alcohol and the corresponding ester are substrates for an enzyme, acylation and deacylation (hydrolysis or alcoholysis) reactions are usually complementary and give the opposite enantiomers. Although hydrolysis/alcoholysis and acylation represent reactions in opposite directions, the enzyme favors the same enantiomer (or the same prochiral group) in both cases. This empirical rule applies to kinetic resolutions and desymmetrizations but a few exceptions have been reported [126]. For example, the acylation of meso-tetrahydropyrans (Figure 6.46) by VA in the presence of CAL-B in organic media and the hydrolysis of the corresponding diacetates catalyzed by the same enzyme provided the opposite enantiomers of the monoesters [127]. Both enantiomers of isolevoglucosenone have been prepared by complementary hydrolysis and acylation reactions (Figure 6.47). ()-Levoglucosenone,

OR

OR

OR

OR

CAL-B

OH

CAL-B R

VA

O OH

R

S

S

OAc

Buffer

O

O OH

OH

OAc

Acylation

O OAc

OAc

Hydrolysis

Figure 6.46 Desymmetrization: acylation and hydrolysis yield opposite enantiomers.

O

OH

Lipase AK

O

VA

O

OAc O

(r ac)

O

(R,R,R) Lipase AK Buffer acetone

MnO2

O

+

O

(rac) O

OAc

O

+ O OAc

(S,S,S)

O O

OH

(S,S,S) O

j151

O

(S,S) OH

MnO 2

O

O

(R,R,R)

Figure 6.47 Kinetic resolution: acylation and hydrolysis yield opposite enantiomers.

O

(R,R)

O


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

152

OH Bu 3Sn

OH Ph

Lipozyme VA, i-Pr2 O

OH

Se

Te

Ph

CAL-B VA, hexane

CAL-B VA, hexane

Figure 6.48 Favored enantiomer in lipase-catalyzed acylations of racemic alcohols containing an organometallic substituent.

a regioisomer of isolevoglucosenone, is obtained in low yield by an acid-catalyzed pyrolysis of cellulose. Isolevoglucosenone can be converted into levoglucosenone and both compounds are versatile chiral building blocks [128]. The wide substrate tolerance of lipases is demonstrated by the resolution of organometallic substrates [129–131]. The presence of tin, selenium, or tellurium in the structure of secondary alcohols does not inhibit the lipase activity and enantiopure organometallic alcohols were obtained by acylation in organic media (Figure 6.48). Sulfatases catalyze the hydrolysis of sulfate esters as shown in Figure 6.49. Arylsulfatases generally act through retention of configuration by cleavage of the S–O bond, whereas some alkyl sulfatases break the C–O bond of sulfate esters leading to inversion of configuration. The enantioselective hydrolysis of sec-alkyl sulfate esters by alkyl sulfatases contained in whole cells of Sulfolobus species proceeds with inversion of configuration and furnishes a homochiral product mixture [132]. Biocatalysis has emerged as an important tool for the enantioselective synthesis of chiral pharmaceutical intermediates and several review articles have been published in recent years [133–137]. For example, quinuclidinol is a common pharmacophore of neuromodulators acting on muscarinic receptors (Figure 6.50). (R)-Quinuclidin-3ol was prepared via Aspergillus melleus protease-mediated enantioselective hydrolysis of the racemic butyrate [54,138]. Calcium hydroxide served as a scavenger of butyric acid to prevent enzyme inhibition and the unwanted (R) enantiomer was racemized over Raney Co under hydrogen for recycling.

OSO 3 –Na+ R

+

Sulf olobus sp. whole cells

R'

OSO 3– Na+ R

OH

(Sulfatase) buff er

R'

R

+

R' -

+

OSO3 Na R

R = CH 3, R' = n-C 6 H13 H 3O+ R = CH , R' = CH -Ph 3 2

R'

Figure 6.49 Deracemization of sec-alkyl sulfate esters.

Asper gillus melleus protease

n-PrCO 2 N

buffer

HO

n-PrCO2 + N (S)

Figure 6.50 Protease-catalyzed resolution of quinuclidin-3-ol.

N (R)


6.3 Enantioselective Enzymatic Transformations of Alcohols

HO

Lipase, VA AcO

NHCbz

CH2 Cl2

(r ac)

O

+ (–)-Alcohol

CO2 H HN

Lipase AK

OH

NHCbz

O

OAc

(–)-Kainic acid

VA

(rac)

CO2 H

Figure 6.51 Chemoenzymatic synthesis of kainic acid.

Ogasawara and coworkers reported a concise route to ()-kainic acid (Figure 6.51), an excitatory neurotransmitter of marine origin, via a lipase-mediated kinetic resolution of an N-Cbz aminocyclopentenol [139]. More recently, ()-kainic acid was synthesized from an optically active butenolide prepared by enzymatic DKR [140]. Conduritols and inositols are cyclic polyalcohols with significant biological activity. The presence of four stereogenic centers in the structure of conduritols allows the existence of 10 stereoisomers. Enzymatic methods have been reported for the resolution of racemic mixtures or the desymmetrization of meso-conduritols. For example, Mucor miehei lipase (MML) showed enantiomeric discrimination between all-(R) and all-(S) stereoisomers of conduritol E tetraacetate (Figure 6.52). Alcoholysis resulted in the removal of the four acetyl groups of the all-(R) enantiomer whereas the all-(S) enantiomer was recovered [141]. Gabosines are secondary metabolites isolated from various Streptomyces strains. A rational approach to the synthesis of gabosines and related carbosugars includes the enantioselective acylation of a hydroxy ketal (masked p-benzoquinone) (Figure 6.53). The strategy has been applied to the synthesis of both enantiomers of gabosine N and O and established the absolute configuration of natural gabosine O [142].

OAc

OH

OAc OAc

OAc +

OAc

OAc

OAc

OAc

All-(R)

All-(S)

OAc OH

MML BuOH

OH

t-BuOMe OH All-(R)

OH CAL-B

O

O (r ac)

SPh

VA i-Pr2 O

OH

(R)

O

HO

O

SPh

+(S)-Acetate

Figure 6.53 Chemoenzymatic synthesis of gabosine O.

OAc OAc

Figure 6.52 Enzymatic alcoholysis of racemic conduritol E tetraacetate.

OH

OAc +

HO O (–)-Gabosine O

All-(S)

j153


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

154

O

+ O N (r ac) + O

EtO

OH

O

CAL-B CH 3CN

O

CO 2Et

OH O

N+ O-

O

H

N

(R)

CO2 Et

OH O

H

N CO2 Et

OH (–)-Rosmarinecine

Figure 6.54 A domino DKR-intramolecular 1,3-dipolar cycloaddition reaction.

The simple acyl moiety introduced during the enzymatic acylation of chiral alcohol is usually removed or replaced by other more complex groups during subsequent reactions. Kita and coworkers [143,144] developed sequential transformations in which the acyl moiety bearing functional groups is utilized as part of the constituent structure in a subsequent intramolecular reaction. Following this concept, they reported a concise asymmetric total synthesis of ()-rosmarinecine [145]. The domino reaction was a DKR of a hydroxy nitrone followed by an intramolecular 1,3-dipolar cycloaddition (Figure 6.54). Combination of lipase-catalyzed transesterification with unsaturated vinyl esters as acyl donors and ring-closing metatheses (RCMs) have also been reported [146–148]. Two groups applied this strategy for the synthesis of goniothalamin from cinnamaldehyde [147,148]. The key steps were a transesterification using vinyl acrylate as acyl donor, followed by an RCM, as depicted in Figure 6.55. Enzymatic desymmetrization of prochiral or meso-alcohols to yield enantiopure building blocks is a powerful tool in the synthesis of natural products. For example, a synthesis of conagenin, an immunomodulator isolated from a Streptomyces, involved two enzymatic desymmetrizations [149]. The syn–syn triad of the acid moiety was prepared via a stereoselective acylation of a meso-diol, whereas the amine fragment was obtained by the PLE-catalyzed hydrolysis of a prochiral malonate (Figure 6.56).

O O

OH

O O

O RCM

CAL-B Hexane

(r ac)

O

(R) + (S)-Alcohol

Goniothalamin

Figure 6.55 Combination of lipase-catalyzed transesterification and RCM.

Cbz-HN

CO2 Et

PLE

Cbz-HN

CO2 Et Buff er, CH3 CN Lipase HO

OH

CO2 Et CO2 H

AcO

OH

VA, THF Figure 6.56 Convergent chemoenzymatic synthesis of (þ)-conagenin.

OH

OH

H N

CO2 H

O (+)-Conagenin

OH


j155

6.3 Enantioselective Enzymatic Transformations of Alcohols

OAc

OAc

PLE H 2O

AcO

(+)

HO

Figure 6.57 Desymmetrization of a centrosymmetric diester.

Few examples of chemical or enzymatic desymmetrizations of centrosymmetric molecules (point group S2 ¼ Ci) have been described [150]. The PLE-catalyzed hydrolysis of a centrosymmetric cyclohexanediacetate gave rise to an enantiomerically pure (99.5% ee) cyclohexanediol monoacetate in high yield [151] (Figure 6.57). Polypropionate chains with alternating methyl and hydroxy substituents are structural elements of many natural products with a broad spectrum of biological activities (e.g. antibiotic, antitumor). The anti–anti stereotriad is symmetric but is the most elusive one. Harada and Oku described the synthesis and the chemical desymmetrization of meso-polypropionates [152]. More recently, the problem of enantiotopic group differentiation was solved by enzymatic transesterification. The synthesis of the acid moiety of the marine polypropionate dolabriferol (Figure 6.58a) and the elaboration of the C(19)–C(27) segment of the antibiotic rifamycin S (Figure 6.58b) involved desymmetrization of meso-polypropionates [153,154]. The desymmetrization principle was also exploited in the synthesis of (þ)FR900482, an antitumor antibiotic isolated from Streptomyces sandaensis [155]. A prochiral propanediol was enzymatically desymmetrized using PSL to give the corresponding (S)-monoester as illustrated in Figure 6.59. The stereoselective acylation of a prochiral chromanedimethanol derivative by VA in the presence of CAL-B gave the corresponding (S)-monoester in high enantiomeric

(a)

CRL HO

OH OTBS

HO

OAc

VA Hexane

OTBS

O

CO2 H OTBS

Lipases

(b) OH

OH

OTBS OH

VA

OH

OAc OH

OTBS OH

OH

Figure 6.58 Enzymatic desymmetrization of meso-polypropionates.

OBn

OH OH

OBn

PSL

OH

OH

OH

OAc

VA NO2 OBn

OCONH2

NO 2 OBn

(S)

OHC

O N (+)-FR900482

Figure 6.59 Application of enzymatic desymmetrization to the synthesis of (þ)-FR900482.

NH


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

156

HO

CAL-B OH

O

OH

HO O

VA Et 2O

(R)- α-Tocotrienol

OAc OH

HO

HO O

OH

O (R)

(S)-α-Tocotrienol

Figure 6.60 Chemoenzymatic synthesis of a-tocotrienol.

O O

O CH 2OAc

Phosphate buffer DMSO (10%)

O O O

PPL

CH 2 OH

O O O

OH

CH 2OAc

O

CH 2 OAc

MeO

OMe OMe

(–)-Podophyllotoxin

Figure 6.61 Chemoenzymatic synthesis of ()-podophyllotoxin.

purity (Figure 6.60). In contrast, enzymatic hydrolysis of prochiral diesters failed to give the (R)-monoester in good yield and high enantiomeric excess. Thus, both enantiomers of a-tocotrienol, one of the main vitamin E isoforms, were synthesized from the (S)-monoester [156]. Podophyllotoxin, a plant lignan, is a potent antimitotic agent (Figure 6.61). An enantioselective synthesis of ()-podophyllotoxin was achieved via the enzymatic desymmetrization of an advanced meso-diacetate, through PPL-mediated diester hydrolysis [157]. Stereoselective enzymatic transformation of substrates generated from Diels– Alder adducts is a versatile method for the preparation of complex chiral building blocks. The 1,4-benzoquinone-cyclopentadiene adduct has been the starting material for the chemoenzymatic synthesis of a variety of natural products as illustrated in Figure 6.62. Desymmetrization of an epoxyquinonediol (Figure 6.62a) was a key step in the synthesis of several polyketide-derived natural products such as cycloepoxydon [158]. Reduction of the adduct and desymmetrization of the corresponding meso-ene1,4-diol (Figure 6.62b) provided a chiral intermediate for the synthesis of the aminocyclitol core of hygromycin A [159,160]. Ogasawara and coworkers optimized the ring contraction of the adduct epoxide (Figure 6.62c) and synthesized tanikolide, a marine antifungal compound, via a stereoselective acylation and a retro Diels–Alder reaction [161]. A similar methodology was applied to the enantioselective total synthesis of epoxyquinone natural products such as epoxydon [162] (Figure 6.62d).


6.4 Hydrolysis of Epoxides

O

O

(a) HO

Lipase PS

AcO

VA, THF 0 °C

HO

O

HO

j157

meso O

O O

O O

O OH

OH

Cycloepoxydon (b)

H

O

H

OAc

O

H

2. Enzymatic desymmetrization

O

H

H 2N

H

OH

Hygromycin A subunit

O

OH

H

H

O

meso

TBSO

HO

OAc

Lipase LIP

O H

OH

OH

p-Benzoquinonecyclopentadiene adduct

(c)

O

HO

1. Reduction

O

O C11H 23

VA, Et2 O OH

AcO +

(r ac)

(+)-Tanikolide

Primary (+)-monoacetate

O

H

OH

OAc

Lipase PS

(d)

O

O

O

VA H

OH

TBSO

O

(r ac)

OH

+ (–)-Alcohol

Figure 6.62 Chemoenzymatic synthesis of natural products from the p-benzoquinone-cyclopentadiene adduct.

6.4 Hydrolysis of Epoxides

Chiral epoxides and their corresponding vicinal diols are very important intermediates in asymmetric synthesis [163]. Chiral nonracemic epoxides can be obtained through asymmetric epoxidation using either chemical catalysts [164] or enzymes [165–167]. Biocatalytic epoxidations require sophisticated techniques and have thus far found limited application. An alternative approach is the asymmetric hydrolysis of racemic or meso-epoxides using transition-metal catalysts [168] or biocatalysts [169–174]. Epoxide hydrolases (EHs) (EC 3.3.2.3) catalyze the conversion of epoxides to their corresponding vicinal diols. EHs are cofactor-independent enzymes that are almost ubiquitous in nature. They are usually employed as whole cells or crude

O

Epoxydon


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

158

a b

O R

EH, H 2O

a Retention b

kS >> kR

Inversion

(r ac)

(R )

O

+

R

(S)

R (R)

O

HO HO

R

+

HO HO

(R)

R

Figure 6.63 Stereo- and regioselectivity of epoxide hydrolysis.

cell-free extracts, but a few (e.g. enzymes from Aspergillus niger and Rhodococcus rhodochrous) have recently become commercially available. Parallel to the discovery of new natural EHs [175,176], directed evolution techniques combined with highthroughput screening [177–179] can generate novel enzymes with improved properties. Epoxide biohydrolyses are usually performed in aqueous media; recent research has, however, demonstrated that A. niger EH is more stable and retains some activity in organic media [180]. The efficiency of EHs is sometimes limited by the low solubility of hydrophobic epoxides in aqueous media or the poor stability of the substrate resulting in spontaneous nonenzymatic hydrolysis. The structure and function of several different EHs have been investigated [169,181,182]. The simplified mechanism involves a trans antiperiplanar addition of water to the oxirane ring leading to the trans 1,2-diols. In general, the attack on a stereogenic center leads to an inversion of configuration. Since the ring opening can occur at two different carbon atoms, the regioselectivity of the enzymatic attack is also an important factor. The stereochemical pathways of hydrolysis of monosubstituted epoxides are illustrated in Figure 6.63. The ring opening may proceed via an attack on the less substituted carbon (attack a) leading to retention of configuration or at the stereogenic center (attack b) with inversion of configuration. Thus the absolute configurations of both the diol and the remaining substrate have to be determined separately. In styrene oxide–type epoxides, the attack at the benzylic position is electronically facilitated by resonance stabilization of the carbocation in the transition state. Activity and selectivity vary with the enzyme source and the substrate structure (substitution pattern, steric crowding, and electronic factor) and one cannot make broad generalizations. Pyridyl oxiranes (Figure 6.64) were stereoselectively hydrolyzed by EH from A. niger to give the (S)-epoxide and the (R)-diol [183]. The reactions proceed with high regio- and enantioselectivity on the least hindered carbon atom of the

O

R H

(S) +

H R

O

Asper gillus niger EH

R H

O (S)

+ H R

OH OH

R=

N

(R) N

(R)

Figure 6.64 Kinetic resolution of pyridyl oxiranes.

N


j159

6.4 Hydrolysis of Epoxides

i-Bu

i-Bu

i-Bu i-Bu

A. niger

O

+ EH

O

OH

1.HBr/AcOH

(r ac)

(R)

OH

O

OH

(S)

(S)-Ibuprofen

2.KOH/MeOH Figure 6.65 Chemoenzymatic synthesis of (S)-ibuprofen.

(R)-epoxide. It is noteworthy that this stereoselective reaction cannot be accomplished with transition-metal catalysts. The hydrolysis of various para-substituted a-methylstyrene oxides was studied using 10 EHs [184]. The hydrolysis of the isobutyl compound with the enzyme from A. niger was the key step in the synthesis of (S)-ibuprofen (Figure 6.65). The (R)-diol was recycled via chemical racemization. Enzymatic resolution of epoxides, like other resolution methods, affords a maximum of 50% yield of optically pure product based on racemic starting material. This limitation can be overcome by enantioconvergent processes, in which each of the substrate enantiomers is converted into the same product enantiomer via two independent pathways. Enantioconvergence has been achieved using (i) two biocatalysts, (ii) a single biocatalyst, or (iii) a combination of a bio- and a chemocatalyst. Both A. niger and B. sulfurescens EHs catalyze the enantioselective hydrolysis of rac-styrene oxide [185] (Figure 6.66). These two microorganisms present opposite enantioselectivities toward the epoxide and the absolute configuration of the product diol is R in both cases. Thus, the enantioconvergent biohydrolysis with a mixture of the two fungi in the same reactor provides the (R)-diol in high yield and high enantiomeric excess. Enantioconvergent hydrolysis of para-chlorostyrene oxide (Figure 6.67) using a one-pot sequential bienzymatic strategy provided the corresponding (R)-diol in high yield (96% ee, yield ¼ 93%) [186]. The second enzyme was added after about 50% conversion because the first one was sensitive to inhibition by the (R)-diol.

A.niger

O a +

b O

a Retention B. sulf ur escens b Inversion

(r ac) Both microorganisms Single reactor

OH O (S) (23%, 96% ee)

OH (R) (54%, 51% ee) OH

O OH (R) (19%, 98% ee)

(R) (47%, 83% ee)

OH (R ) (92%, 89% ee) OH

Figure 6.66 Enantioconvergent hydrolysis of styrene oxides using two biocatalysts.


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

160

O

O

OH

HO

Solanum t uber osum (R)

EH (rac)

Cl

Cl

+

(R) (93%, 96% ee) Cl

A. niger EH

Figure 6.67 Convergent hydrolysis using two enzymes in sequence.

OH O

Inversion

OH (R)

(S ) Cl O (R)

Solanum tuber osum EH n tio ten e R

Cl

β-Adrenergic receptor agonists Cl (r ac) Figure 6.68 Enantioconvergent hydrolysis of m-chlorostyrene oxide using a single biocatalyst.

The enantioconvergent biohydrolysis of m-chlorostyrene oxide (Figure 6.68) in the presence of a recombinant S. tuberosum EH afforded the corresponding (R)-diol in a nearly quantitative yield [187]. The (S)-epoxide was attacked at the benzylic (more substituted) carbon whereas the (R)-epoxide was attacked at the terminal (less substituted) carbon. The racemic cis-epoxide shown in Figure 6.69 was hydrolyzed in an enantioconvergent fashion – that is, both enantiomers were converted with opposite regioselectivity (see arrows on Figure 6.69) to give the single (2S,3R)-diol, which underwent spontaneous ring closure to yield an epoxy alcohol [188]. This enzyme-triggered enantioconvergent cascade reaction was applied to the enantioselective synthesis of two natural products, (þ)-pestalotin and a Jamaican rum constituent. Cl

O

O O

O n-C 4H 9 Cl

My cobact er ium par af f inicum

HO HO HO

O

Cl

n-C 4 H9

– HCl

Jamaican rum constituent

OMe

n-C 4H 9

(+)-Pestalotin

n-C 4H 9 O

(r ac) OH

Figure 6.69 An enantioconvergent enzyme-triggered cascade reaction.

O


6.4 Hydrolysis of Epoxides

j161

O O

OH Rhodococcus r uber EH

O

HO

O

(R)

Beer aroma constituent

(r ac) Figure 6.70 Enantioconvergent hydrolysis of a trisubstituted epoxide using a single enzyme.

The enantioconvergent biohydrolysis of sterically demanding trisubstituted oxiranes has also been reported [189,190]. For instance, the enantioconvergent hydrolysis of a trisubstituted rac-epoxide (Figure 6.70) was a key step in the chemoenzymatic synthesis of a volatile constituent of the beer aroma [190]. The enantioconvergent hydrolysis of several 2,2-disubstituted epoxides was accomplished by combined bio- and chemocatalysis [191–194]. For example, the kinetic resolution of the epoxide shown in Figure 6.71 using a Nocardia sp. EH proceeded via an attack at the unsubstituted carbon atom with retention of configuration at the adjacent stereogenic center [194]. In the second step, the remaining epoxide was hydrolyzed under acid-catalysis with inversion of configuration. The sole ring-opening product (S)-diol was further elaborated to the natural product (R)-mevalonolactone, a biosynthetic precursor of terpenes and steroids. Figure 6.72 shows an enantioconvergent multistep process leading to an enantiopure epoxide. The racemic epoxide was resolved by A. niger EH leading to the (R)-diol and the residual (S)-epoxide with excellent optical purity [195]. The chemical O

O

H 2 SO4

OH

N ocar dia sp. EH Ph pH 7.5

O

HO OH

Ph + OH

Ph

Ph

O O (R) - Mevalonolactone

(S) (S) OH

(r ac)

Ph

Figure 6.71 Enantioconvergent hydrolysis of a 2,2-disubstituted epoxide by combined bio- and chemocatalysts. 1. NaH / THF 2. PPh 3 / CCl4 F

F O

F

A. niger EH

( r ac )

F Cl

OH

( S)

H 2O / DMSO (5%) F

+

O

Cl

Figure 6.72 An enantioconvergent process leading to an enantiopure epoxide.

(R) F

Cl OH


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

162

O

R

EH

R

OH

O

O R meso

O

O Cbz

R

N

OH ( R ,R )

O

O

Figure 6.73 EH-catalyzed desymmetrization of cyclic meso-epoxides.

O R

Halohydrin dehalogenase Nu–

OH

O R

+ (S)

R

Nu

Nu– = Cl – , CN– , NO 2– , N3 –

(R)

Figure 6.74 Ring-opening reactions catalyzed by halohydrin dehalogenase.

cyclization of the (R)-diol gave the corresponding epoxy alcohol, which was transformed into the (S)-epoxide via treatment with CCl4/PPh3. Both these reactions were shown to occur without loss of stereochemical integrity. The final product is a key building block for the synthesis of enantiopure azole antifungal agents. meso-Epoxides are also good substrates and have the potential to yield an optically pure diol in 100% yield. The desymmetrization of a wide range of meso-epoxides was studied using high-throughput screening of environmental libraries [196]. In most cases, the (S)-stereogenic center was inverted to yield the (R,R)-diol (Figure 6.73). Moreover, the first EHs providing access to complementary (S,S)-diols were also found. Ring opening of epoxides with nucleophiles other than water (Cl, Br, I, N3, NO2, CN) can also be catalyzed by halohydrin dehalogenase enzymes (EC 3.8.1.5, also named haloalkane dehalogenase or haloalcohol dehalogenase) (Figure 6.74) [197].

6.5 Conclusion

Organic synthesis has been revolutionized over the past two decades by the advent of enantioselective catalysis. Chemocatalysis and biocatalysis were developed in parallel and, undoubtedly, the design of many low-molecular-weight chiral catalysts was inspired by enzyme catalysis. Hydrolase-catalyzed transesterification or hydrolysis of carboxylic acid derivatives, alcohols, and epoxides represent a versatile methodology for the preparation of optically active compounds. Hydrolases are the most frequently used biocatalysts in organic synthesis. They account for about two-thirds of all synthetic biotransformations. Several characteristics make hydrolases useful catalysts in asymmetric synthesis. They do not require sensitive cofactors and many are commercially available. They often show high stereoselectivity on various synthetic nonnatural compounds (broad substrate specificity). They are stable and active in neat organic solvents of low polarity. Unlike heavy metal catalysts, hydrolases are environmentally benign reagents (green chemistry). Recombinant protein technology


References

allows for the production of synthetically useful quantities of enzymes. Hydrolases with novel properties may be obtained by screening environmental samples. Alternatively, genetic engineering techniques such as directed evolution may lead to new biocatalysts.

Acknowledgments

The authors gratefully acknowledge the financial support from the Natural Sciences and Engineering Research Council of Canada (NSERC).

References 1 Bornscheuer, U.T. and Kazlauskas, R.J. (2006) Hydrolases in Organic Synthesis, 2nd edn Wiley-VCH Verlag GmbH, Weinheim. 2 Buchholz, K., Kasche, V. and Bornscheuer, U.T. (2005) Biocatalysts and Enzyme Technology, Wiley-VCH Verlag GmbH, Weinheim. 3 Bommarius, A.S. and Riebel, B.R. (2004) Biocatalysis, Wiley-VCH Verlag GmbH, Weinheim. 4 Faber, K. (2004) Biotransformations in Organic Chemistry, 5th edn Springer, Berlin. 5 Drauz, K. and Waldmann, H. (eds) (2002) Enzyme Catalysis in Organic Synthesis, 2nd edn Wiley-VCH Verlag GmbH, Weinheim, Vol. 1–3. 6 Patel, R.N. (ed.) (2000) Stereoselective Biocatalysis, Marcel Dekker, New York. 7 Roberts, S.M. (ed.),(1999) Biocatalysts for Fine Chemical Synthesis, Wiley-VCH Verlag GmbH, Weinheim. 8 Secundo, F., Carrea, G., Tarabiono, C., Gatti-Lafranconi, P., Brocca, S., Lotti, M., Jaeger, K.E., Puls, M. and Eggert, T. (2006) Journal of Molecular Catalysis B: Enzymatic, 39, 166–170. 9 Hudson, E.P., Eppler, R.K. and Clark, D.S. (2005) Current Opinion in Biotechnology, 16, 637–643. 10 Secundo, F. and Carrea, G. (2003) Chemistry-A European Journal, 9, 3194–3199.

11 Reetz, M.T. (2002) Current Opinion in Chemical Biology, 6, 145–150. 12 Carrea, G. and Riva, S. (2000) Angewandte Chemie International Edition, 39, 2226–2254. 13 Klibanov, A.M. (2001) Nature, 409, 241–246. 14 Park, S. and Kazlauskas, R.J. (2003) Current Opinion in Biotechnology, 14, 432–437. 15 van Rantwijk, F., Lau, R.M. and Sheldon, R.A. (2003) Trends in Biotechnology, 21, 131–138. 16 Kragl, U., Eckstein, M. and Kaftzik, N. (2002) Current Opinion in Biotechnology, 13, 565–571. 17 Sheldon, R.A., Lau, R.M., Sorgedrager, M.J., van Rantwijk, F. and Seddon, K.R. (2002) Green Chemistry, 4, 147–151. 18 Matsuda, T., Harada, T. and Nakamura, K. (2005) Current Organic Chemistry, 9, 299–315. 19 Sakai, T., Kawabata, I., Kishimoto, T., Ema, T. and Utaka, M. (1997) Journal of Organic Chemistry, 62, 4906–4907. 20 Ema, T., Kageyama, M., Korenaga, T. and Sakai, T. (2003) Tetrahedron: Asymmetry, 14, 3943–3947. 21 Cainelli, G., Galletti, P., Giacomini, D., Gualandi, A. and Quintavalla, A. (2003) HeIvetica Chimica Acta, 86, 3548–3559. 22 Rejasse, B., Besson, T., Legoy, M.D. and Lamare, S. (2006) Organic and Biomolecular Chemistry, 4, 3703–3707. 23 Theil, F. (2000) Tetrahedron, 56, 2905–2919.

j163


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

164

24 Cao, L. (2005) Current Opinion in Chemical Biology, 9, 217–226. 25 Roy, J.J. and Abraham, T.E. (2004) Chemical Reviews, 104, 3705–3721. 26 Bornscheuer, U.T. (2003) Angewandte Chemie International Edition, 42, 3336–3337. 27 Reetz, M.T., Tielmann, P., Wiesenh€ofer, W., K€onen, W. and Zonta, A. (2003) Advanced Synthesis and Catalysis, 345, 717–728. 28 Kroutil, W., Kleewein, A. and Faber, K. (1997) Tetrahedron: Asymmetry, 8, 3251–3261. 29 For a web site, see:http://www.orgc. tugraz.at/ (by Faber, K., H€onig, H., Kleewein, A. 1994, updated 2006). 30 Sih, C.J. and Wu, S.H. (1989) Resolution of enantiomers via biocatalysis, in Topics in Stereochemistry, (eds E.L. Eliel, S.H. Wilen and N.L. Allinger), John Wiley & Sons, Ltd, New York, vol. 19, pp. 63–125. 31 Ghanem, A. and Aboul-Enein, H.Y. (2004) Tetrahedron: Asymmetry, 15, 3331–3351. 32 Robinson, D.E.J.E. and Bull, S.D. (2003) Tetrahedron: Asymmetry, 14, 1407–1446. 33 Bouzemi, N., Debbeche, H., AribiZouioueche, L. and Fiaud, J.C. (2004) Tetrahedron Letters, 45, 627–630. 34 Terao, Y., Tsuji, K., Murata, M., Achiwa, K., Nishio, T., Watanabe, N. and Seto, K. (1989) Chemical and Pharmaceutical Bulletin, 37, 1653–1655. 35 Luo, Z., Swaleh, S.M., Theil, F. and Curran, D.P. (2002) Organic Letters, 4, 2585–2587. 36 Turner, N.J. (2004) Current Opinion in Chemical Biology, 8, 114–119. 37 Pellissier, H. (2003) Tetrahedron, 59, 8291–8327. 38 Schnell, B., Faber, K. and Kroutil, W. (2003) Advanced Synthesis and Catalysis, 345, 653–666. 39 May, O., Verseck, S., Bommarius, A. and Drauz, K. (2002) Organic Process Research and Development, 6, 452–457. 40 Brown, S.A., Parker, M.C. and Turner, N.J. (2000) Tetrahedron: Asymmetry, 11, 1687–1690.

41 Um, P.J. and Drueckhammer, D.G. (1998) Journal of the American Chemical Society, 120, 5605–5610. 42 Haughton, L. and Williams, J.M.J. (2001) Synthesis, 943–946. 43 Veum, L., Kanerva, L.T., Halling, P.J., Maschmeyer, T. and Hanefeld, U. (2005) Advanced Synthesis and Catalysis, 347, 1015–1021. 44 Pesti, J.A., Yin, J., Zhang, L. and Anzalone, L. (2001) Journal of the American Chemical Society, 123, 11075–11076. 45 Boren, L., Martın-Matute, B., Xu, Y., Córdova, A. and B€ackvall, J.E. (2006) Chemistry-A European Journal, 12, 225–232. 46 Choi, J.H., Choi, Y.K., Kim, Y.H., Park, E.S., Kim, E.J., Kim, M.J. and Park, J. (2004) Journal of Organic Chemistry, 69, 1972–1977. 47 Pamies, O. and B€ackvall, J.E. (2003) Chemical Reviews, 103, 3247–3261. 48 Kim, M.J., Ahn, Y. and Park, J. (2002) Current Opinion in Biotechnology, 13, 578–587. 49 Garcia-Urdiales, E., Alfonso, I. and Gotor, V. (2005) Chemical Reviews, 105, 313–354. 50 Schoffers, E., Golebiowski, A. and Johnson, C.R. (1996) Tetrahedron, 52, 3769–3826. 51 Faber, K. (2001) Chemistry-A European Journal, 7, 5004–5010. 52 Leue, S., Miao, W., Kanazawa, A., Genisson, Y., Garçon, S. and Greene, A.E. (2001) Journal of the Chemical Society, Perkin Transactions 1, 2903–2905. 53 Truppo, M.D., Journet, M., Shafiee, A. and Moore, J.C. (2006) Organic Process Research and Development, 10, 592–598. 54 Chikusa, Y., Hirayama, Y., Ikunaka, M., Inoue, T., Kamiyama, S., Moriwaki, M., Nishimoto, Y., Nomoto, F., Ogawa, K., Ohno, T., Otsuka, K., Sakota, A.K., Shirasaka, N., Uzura, A. and Uzura, K. (2003) Organic Process Research and Development, 7, 289–296. 55 Salgado, A., Huybrechts, T., Eeckhaut, A., Van der Eycken, J., Szakonyi, Z., F€ ul€op, F.,


References

56 57

58

59

60

61

62

63

64

65

66

67

68

69

70 71

Tkachev, A. and De Kimpe, N. (2001) Tetrahedron, 57, 2781–2786. Kedrowski, B.L. (2003) Journal of Organic Chemistry, 68, 5403–5406. Back, T.G. and Wulff, J.E. (2004) Angewandte Chemie International Edition, 43, 6493–6496. White, J.D., Kim, J. and Drapela, N.E. (2000) Journal of the American Chemical Society, 122, 8665–8671. Katoh, T., Kakiya, K., Nakai, T., Nakamura, S., Nishide, K. and Node, M. (2002) Tetrahedron: Asymmetry, 13, 2351–2358. Liang, X., Lohse, A. and Bols, M. (2000) Journal of Organic Chemistry, 65, 7432–7437. Henke, E., Schuster, S., Yang, H. and Bornscheuer, U.T. (2000) Monatshefte fur Chemie, 131, 633–638. Liljeblad, A., Lindborg, J. and Kanerva, L.T. (2000) Tetrahedron: Asymmetry, 11, 3957–3966. Miyazawa, T., Onishi, K., Murashima, T., Yamada, T. and Tsai, S.W. (2005) Tetrahedron: Asymmetry, 16, 2569–2573. Kawasaki, M., Kakuda, H., Goto, M., Kawabata, S. and Kometani, T. (2003) Tetrahedron: Asymmetry, 14, 1529–1534. Morrone, R., Piattelli, M. and Nicolosi, G. (2001) European Journal of Organic Chemistry, 1441–1443. Hedenstr€ om, E., Nguyen, B.V. and Silks, L.A. (2002) Tetrahedron: Asymmetry, 13, 835–844. Enzelberger, M.M., Bornscheuer, U.T., Gatfield, I. and Schmid, R.D. (1997) Journal of Biotechnology, 56, 129–133. Sakai, N., Ageishi, S., Isobe, H., Hayashi, Y. and Yamamoto, Y. (2000) Journal of the Chemical Society, Perkin Transactions, 1, 71–77. Klomp, D., Peters, J.A. and Hanefeld, U. (2005) Tetrahedron: Asymmetry, 16, 3892–3896. Mori, K. (1995) Synlett, 1097–1109. Missio, L.J. and Comasseto, J.V. (2000) Tetrahedron: Asymmetry, 11, 4609–4615.

72 Jeon, N.Y., Ko, S.J., Won, K., Kang, H.Y., Kim, B.T., Lee, Y.S. and Lee, H. (2006) Tetrahedron Letters, 47, 6517–6520. 73 Kumar, A., Gross, R.A. and Jendrossek, D. (2000) Journal of Organic Chemistry, 65, 7800–7806. 74 Hwang, B.Y., Lee, H.B., Kim, Y.G. and Kim, B.G. (2006) Journal of Molecular Catalysis B: Enzymatic, 41, 125–129. 75 Ozegowski, R., Kunath, A. and Schick, H. (1995) Tetrahedron: Asymmetry, 6, 1191–1194. 76 Fryszkowska, A., Komar, M., Koszelewski, D. and Ostaszewski, R. (2006) Tetrahedron: Asymmetry, 17, 961–966. 77 Fryszkowska, A., Frelek, J. and Ostaszewski, R. (2005) Tetrahedron, 61, 6064–6072. 78 Wang, M.X. (2005) Topics in Catalysis, 35, 117–130. 79 Martinkova, L. and Kren, V. (2002) Biocatalysis and Biotransformation, 20, 73–93. 80 Banerjee, A., Sharma, R. and Banerjee, U.C. (2002) Applied Microbiology and Biotechnology, 60, 33–44. 81 Kobayashi, M. and Shimizu, S. (2000) Current Opinion in Chemical Biology, 4, 95–102. 82 Nagasawa, T. and Yamada, H. (1995) Pure and Applied Chemistry, 67, 1241–1256. 83 Wang, M.X. and Zhao, S.M. (2002) Tetrahedron Letters, 43, 6617–6620. 84 Wang, M.X., Lin, S.J., Liu, C.S., Zheng, Q.Y. and Li, J.S. (2003) Journal of Organic Chemistry, 68, 4570–4573. 85 Wang, M.X. and Feng, G.Q. (2003) Journal of Organic Chemistry, 68, 621–624. 86 Mateo, C., Chmura, A., Rustler, S., van Rantwijk, F., Stolz, A. and Sheldon, R.A. (2006) Tetrahedron: Asymmetry, 17, 320–323. 87 DeSantis, G., Zhu, Z., Greenberg, W.A., Wong, K., Chaplin, J., Hanson, S.R., Farwell, B., Nicholson, L.W., Rand, C.L., Weiner, D.P., Robertson, D.E. and Burk, M.J. (2002) Journal of the American Chemical Society, 124, 9024–9025.

j165


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

166

88 Gr€oger, H. (2001) Advanced Synthesis and Catalysis, 343, 547–558. 89 Yamamoto, K., Oishi, K., Fujimatsu, I. and Komatsu, K.I. (1991) Applied and Environment Microbiology, 57, 3028–3032. 90 Wang, M.X., Lin, S.J., Liu, J. and Zheng, Q.Y. (2004) Advanced Synthesis and Catalysis, 346, 439–445. 91 Wegman, M.A., Heinemann, U., van Rantwijk, F., Stolz, A. and Sheldon, R.A. (2001) Journal of Molecular Catalysis B: Enzymatic, 11, 249–253. 92 DeSantis, G., Wong, K., Farwell, B., Chatman, K., Zhu, Z., Tomlinson, G., Huang, H., Tan, X., Bibbs, L., Chen, P., Kretz, K. and Burk, M.J. (2003) Journal of the American Chemical Society, 125, 11476–11477. 93 Kakeya, H., Sakai, N., Sano, A., Yokoyama, M., Sugai, T. and Ohta, H. (1991) Chemistry Letters, 20, 1823–1824. 94 Bergeron, S., Chaplin, D.A., Edwards, J.H., Ellis, B.S.W., Hill, C.L., Holt-Tiffin, K., Knight, J.R., Mahoney, T., Osborne, A.P. and Ruecroft, G. (2006) Organic Process Research and Development, 10, 661–665. 95 M€ uller, M. (2005) Angewandte Chemie International Edition, 44, 362–365. 96 Wu, Z.L. and Li, Z.Y. (2003) Chemical Communications, 386–387. 97 Liljeblad, A. and Kanerva, L.T. (2006) Tetrahedron, 62, 5831–5854. 98 Shaw, N.M., Robins, K.T. and Kiener, A. (2003) Advanced Synthesis and Catalysis, 345, 425–435. 99 Peterson, M. and Sauter, M. (1999) Chimia, 53, 608–612. 100 Braiuca, P., Boscarol, L., Ebert, C., Linda, P. and Gardossi, L. (2006) Advanced Synthesis and Catalysis, 348, 773–780. 101 Sonke, T., Kaptein, B., Wagner, A.F.V., Quaedflieg, P.J.L.M., Schultz, S., Ernste, S., Schepers, A., Mommers, J.H.M. and Broxterman, Q.B. (2004) Journal of Molecular Catalysis B: Enzymatic, 29, 265–277. 102 Sonke, T., Ernste, S., Tandler, R.T., Kaptein, B., Peeters, W.P.H., van Assema,

103

104 105

106

107

108 109

110 111

112

113 114

115

116

117

F.B.J., Wubbolts, M.G. and Schoemaker, H.E. (2005) Applied and Environment Microbiology, 71, 7961–7973. Asano, Y. and Yamaguchi, S. (2005) Journal of the American Chemical Society, 127, 7696–7697. Shaw, N.M. and Naughton, A.B. (2004) Tetrahedron, 60, 747–752. Shaw, N.M., Naughton, A., Robins, K., Tinschert, A., Schmid, E., Hischier, M.L., Venetz, V., Werlen, J., Zimmermann, T., Brieden, W., de Riedmatten, P., Roduit, J.P., Zimmermann, B. and Neum€ uller, R. (2002) Organic Process Research and Development, 6, 497–504. Lloyd, R.C., Lloyd, M.C., Smith, M.E.B., Holt, K.E., Swift, J.P., Keene, P.A., Taylor, S.J.C. and McCague, R. (2004) Tetrahedron, 60, 717–728. Taylor, S.J.C., Brown, R.C., Keene, P.A. and Taylor, I.N. (1999) Bioorganic and Medicinal Chemistry, 7, 2163–2168. Forró, E. and F€ ul€op, F. (2006) Chemistry-A European Journal, 12, 2587–2592. Forró E., Paal, T., Tasnadi, G. and F€ ul€op, F. (2006) Advanced Synthesis and Catalysis, 348, 917–923. Forró, E. and F€ ul€op, F. (2003) Organic Letters, 5, 1209–1212. Park, S., Forró E., Grewal, H., F€ ul€op, F. and Kazlauskas, R.J. (2003) Advanced Synthesis and Catalysis, 345, 986–995. Adam, W., Groer, P., Humpf, H.U. and Saha-M€oller, C.R. (2000) Journal of Organic Chemistry, 65, 4912–4922. Nam, S.H., Park, H.S. and Kim, H.S. (2005) Chemical Record, 5, 298–307. Burton, S.G. and Dovington, R.A. (2004) Tetrahedron: Asymmetry, 15, 2737–2741. Altenbuchner, J., Siemann-Herzberg, M. and Syldatk, C. (2001) Current Opinion in Biotechnology, 12, 559–563. Ch^enevert, R., Pelchat, N. and Jacques, F. (2006) Current Organic Chemistry, 10, 1067–1094. Santaniello, E., Casati, S. and Ciuffreda, P. (2006) Current Organic Chemistry, 10, 1095–1123.


References 118 Gotor-Fernandez, V., Busto, E. and Gotor, V. (2006) Advanced Synthesis and Catalysis, 348, 797–812. 119 Lutz, S. (2004) Tetrahedron: Asymmetry, 15, 2743–2748. 120 Kirk, O. and Christensen, M.W. (2002) Organic Process Research and Development, 6, 446–451. 121 Dominguez de Maria, P., SanchezMontero, J.M., Sinisterra, J.V. and Alcantara, A.R. (2006) Biotechnology Advances, 24, 180–196. 122 Dominguez de Maria, P., Alcantara, A.R., Carballeira, J.D., de la Casa, R.M., Garcia-Burgos, C.A., Hernaiz, M.J., Sanchez-Montero, J.M. and Sinisterra, J.V. (2006) Current Organic Chemistry, 10, 1053–1066. 123 Hanefeld, U. (2003) Organic and Biomolecular Chemistry, 1, 2405–2415. 124 Otera, J. (2003) Esterification Methods, Reactions and Applications, Wiley-VCH Verlag GmbH, Weinheim. 125 Kita, Y. and Akai, S. (2004) The Chemical Record, 4, 363–372. 126 Izquierdo, I., Plaza, M.T., Rodriguez, M. and Tamayo, J. (1999) Tetrahedron: Asymmetry, 10, 449–455. 127 Ch^enevert, R., Goupil, D., Rose, Y.S. and Bedard, E. (1998) Tetrahedron: Asymmetry, 9, 4285–4288. 128 Kadota, K., ElAzab, A.S., Taniguchi, T. and Ogasawara, K. (2000) Synthesis, 10, 1372–1374. 129 Lee, T. and Kim, S. (2003) Tetrahedron: Asymmetry, 14, 1951–1954. 130 Costa, C.E., Cososki, G.C., Barchesi, H.B., Zanotto, S.P., Nascimento, M.G. and Comasseto, J.V. (2004) Tetrahedron: Asymmetry, 15, 3945–3954. 131 Dos Santos, A.A., Da Costa, C.E., Princival, J.L. and Comasseto, J.V. (2006) Tetrahedron: Asymmetry, 17, 2252–2259. 132 Wallner, S.R., Nestl, B.M. and Faber, K. (2005) Organic and Biomolecular Chemistry, 3, 2652–2656. 133 Gotor-Fernandez, V., Brieva, R. and Gotor, V. (2006) Journal of Molecular Catalysis B: Enzymatic, 40, 111–120.

134 Breuer, M., Ditrich, K., Habicher, T., Hauer, B., Keßeler, M., St€ urmer, R. and Zelinski, T. (2004) Angewandte Chemie International Edition, 43, 788–824. 135 Garcia-Junceda, E., Garcia-Garcia, J.F., Bastida, A. and Fernandez-Mayoralas, A. (2004) Bioorganic and Medicinal Chemistry, 12, 1817–1834. 136 Straathof, A.J.J., Panke, S. and Schmid, A. (2002) Current Opinion in Biotechnology, 13, 548–556. 137 Zaks, A. and Dodds, D.R. (1997) Drug Discovery Today, 2, 513–531. 138 Ikunaka, M. (2004) Catalysis Today, 96, 93–102. 139 Nakagawa, H., Sugahara, T. and Ogasawara, K. (2000) Organic Letters, 2, 3181–3183. 140 Morita, Y. Tokuyama, H. and Fukuyama, T. (2005) Organic Letters, 7, 4337–4340. 141 Lambusta, D., Nicolosi, G., Patti, A. and Sanfilippo, C. (2003) Journal of Molecular Catalysis B: Enzymatic, 22, 271–277. 142 Alibes, R., Bayón, P., de March, P., Figueredo, M., Font, J. and Marjanet, G. (2006) Organic Letters, 8, 1617–1620. 143 Akai, S. and Tanimoto, K. and Kita, Y. (2004) Angewandte Chemie International Edition, 43, 1407–1410. 144 Akai, S., Naka, T., Omura, S., Tanimoto, K., Imanishi, M., Takebe, Y., Matsugi, M. and Kita, Y. (2002) Chemistry-A European Journal, 8, 4255–4264. 145 Akai, S., Tanimoto, K., Kanao, Y., Omura, S. and Kita, Y. (2005) Chemical Communications, 2369–2371. 146 Fujii, M., Fukumura, M., Hori, Y., Hirai, Y., Akita, H., Nakamura, K., Toriizuka, K. and Ida, Y. (2006) Tetrahedron: Asymmetry, 17, 2292–2298. 147 Gruttadauria, M., Lo Meo, P. and Noto, R. (2004) Tetrahedron Letters, 45, 83–85. 148 Sundby, E., Perk, L., Anthonsen, T., Aasen, A.J. and Hansen, T.V. (2004) Tetrahedron, 60, 521–524. 149 Sano, S., Miwa, T., Hayashi, K., Nozaki, K., Ozaki, Y. and Nagao, Y. (2001) Tetrahedron Letters, 42, 4029–4031.

j167


j 6 Transesterification and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides

168

150 Anstiss, M., Holland, J.M., Nelson, A. and Titchmarsh, J.R. (2003) Synlett, 1213–1220. 151 B€ ohm, C., Austin, W.F. and Trauner, D. (2003) Tetrahedron: Asymmetry, 14, 71–74. 152 Harada, T. and Oku, A. (1994) Synlett, 95–104. 153 Ch^enevert, R., Courchesne, G. and Caron, D. (2003) Tetrahedron: Asymmetry, 14, 2567–2571. 154 Ch^enevert, R. and Rose, Y.S. (2000) Journal of Organic Chemistry, 65, 1707–1709. 155 Fellows, I.M., Kaelin, D.E. and Martin, S.F. (2000) Journal of the American Chemical Society, 122, 10781–10787. 156 Ch^enevert, R., Courchesne, G. and Pelchat, N. (2006) Bioorganic and Medicinal Chemistry, 14, 5389–5396. 157 Berkowitz, D.B., Choi, S. and Maeng, J.H. (2000) Journal of Organic Chemistry, 65, 847–860. 158 Mehta, G. and Islam, K. (2004) Organic Letters, 6, 807–810. 159 Konno, H. and Ogasawara, K. (1999) Synthesis, 1135–1140. 160 Donohoe, T.J., Johnson, P.D., Pye, R.J. and Keenan, M. (2005) Organic Letters, 7, 1275–1277. 161 Tanaka, H., Kozuki, Y. and Ogasawara, K. (2002) Tetrahedron Letters, 43, 4175–4178. 162 Mehta, G. and Islam, K. (2004) Tetrahedron Letters, 45, 7683–7687. 163 Yudin A.K. (ed.) (2006) Aziridines and Epoxides in Organic Synthesis, Wiley-VCH Verlag GmbH, Weinheim. 164 Xia, Q.H., Ge, H.Q., Ye, C.P., Liu, Z.M. and Su, K.X. (2005) Chemical Reviews, 105, 1603–1662. 165 Bernhardt, R. (2006) Journal of Biotechnology, 124, 128–145. 166 Meunier, B. de Visser, S.P. and Shaik, S. (2004) Chemical Reviews, 104, 3947–3980. 167 Dembitsky, V.M. (2003) Tetrahedron, 59, 4701–4720. 168 Jacobsen, E.N. (2000) Accounts of Chemical Research, 33, 421–431. 169 Smit, M.S. and Labuschagne, M. (2006) Current Organic Chemistry, 10, 1145–1161.

170 van Loo, B., Kingma, J., Arand, M., Wubbolts, M.G. and Janssen, D.B. (2006) Applied and Environment Microbiology, 72, 2905–2917. 171 Smit, M.S. (2004) Trends in Biotechnology, 22, 123–129. 172 de Vries, E.J. and Janssen, D.B. (2003) Current Opinion in Biotechnology, 14, 414–420. 173 Archelas, A. and Furstoss, R. (2001) Current Opinion in Chemical Biology, 5, 112–119. 174 Steinreiber, A. and Faber, K. (2001) Current Opinion in Biotechnology, 12, 552–558. 175 Kim, H.S., Lee, O.K., Lee, S.J., Hwang, S., Kim, S.J., Yang, S.H., Park, S. and Lee, E.Y. (2006) Journal of Molecular Catalysis B: Enzymatic, 41, 130–135. 176 Xu, W., Xu, J.H., Pan, J., Gu, Q. and Wu, X.Y. (2006) Organic Letters, 8, 1737–1740. 177 Belder, D., Ludwig, M., Wang, L.W. and Reetz, M.T. (2006) Angewandte Chemie International Edition, 45, 2463–2466. 178 Reetz, M.T., Wang, L.W. and Bocola, M. (2006) Angewandte Chemie International Edition, 45, 1236–1241. 179 Cedrone, F., Bhatnagar, T. and Baratti, J.C. (2005) Biotechnology Letters, 27, 1921–1927. 180 Karboune, S., Archelas, A. and Baratti, J. (2006) Enzyme and Microbial Technology, 39, 318–324. 181 Hopmann, K.H. and Himo, F. (2006) Chemistry-A European Journal, 12, 6898–6909. 182 Hopmann, K.H., Hallberg, B.M. and Himo, F. (2005) Journal of the American Chemical Society, 127, 14339–14347. 183 Genzel, Y., Archelas, A., Broxterman, Q.B., Schulze, B. and Furstoss, R. (2001) Journal of Organic Chemistry, 66, 538–543. 184 Cleij, M., Archelas, A. and Furstoss, R. (1999) Journal of Organic Chemistry, 64, 5029–5035. 185 Pedragosa-Moreau, S., Archelas, A. and Furstoss, R. (1993) Journal of Organic Chemistry, 58, 5533–5536. 186 Manaj, K.M., Archelas, A., Baratti, J. and Furstoss, R. (2001) Tetrahedron, 57, 695–701.


Further Reading 187 Monterde, M.I., Lombard, M., Archelas, A., Cronin, A., Arand, M. and Furstoss, R. (2004) Tetrahedron: Asymmetry, 15, 2801–2805. 188 Mayer, S.F., Steinreiber, A., Goriup, M., Saf, R. and Faber, K. (2002) Tetrahedron: Asymmetry, 13, 523–528. 189 Steinreiber, A., Mayer, S.F., Saf, R. and Faber, K. (2001) Tetrahedron: Asymmetry, 12, 1519–1528. 190 Steinreiber, A., Mayer, S.F. and Faber, K. (2001) Synthesis, 13, 2035–2039. 191 Simeo, Y. and Faber, K. (2006) Tetrahedron: Asymmetry, 17, 402–409. 192 Steinreiber, A., Hellstr€om, M., Mayer, S.F., Orru, R.V.A. and Faber, K. (2001) Synlett, 111–113. 193 Orru, R.V.A., Mayer, S.F., Kroutil, W. and Faber, K. (1998) Tetrahedron, 54, 859–874. 194 Orru, R.V.A., Osprian, I., Kroutil, W. and Faber, K. (1998) Synthesis, 1259–1263. 195 Monfort, N., Archelas, A. and Furstoss, R. (2004) Tetrahedron, 60, 601–605.

196 Zhao, L., Han, B., Huang, Z., Miller, M., Huang, H., Malashock, D.S., Zhu, Z., Milan, A., Robertson, D.E., Weiner, D.P. and Burk, M.J. (2004) Journal of the American Chemical Society, 126, 11156–11157. 197 Elenkov, M.M., Tang, L., Hauer, B. and Janssen, D.B. (2006) Organic Letters, 8, 4229–4257.

Further Reading Cardillo, G., Gentilucci, L., Tolomelli, A. and Tomasini, C. (1998) Journal of Organic Chemistry, 63, 2351–2353. Cardillo, G. and Tolomelli, A. and Tomasini, C. (1996) Journal of Organic Chemistry, 61, 8651–8654. Fadnavis, N.W., Radhika, K.R. and Devi, A.V. (2006) Tetrahedron: Asymmetry, 17, 240–244. Topgi, R.S., Ng, J.S., Landis, B., Wang, P. and Behling, J.R. (1999) Bioorganic and Medicinal Chemistry, 7, 2221–2229.

j169


j171

7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives Vicente Gotor-Fernandez and Vicente Gotor

7.1 Introduction

The use of enzymes for organic synthesis has become an interesting area for organic and bioorganic chemists, since many enzymes have demonstrated their activity with nonnatural substrates in organic media, allowing the development of many different synthetic transformations. Among all of them, hydrolases are the most used enzymes owing to their high stability and activity with a broad spectrum of substrates. Besides, many of them are commercially available, do not need cofactor, and work under mild reaction conditions in contrast with drastic reaction conditions required for chemical processes. From the hydrolases, lipases (EC 3.1.1.3) are the most popular biocatalysts and nowadays the ones more used in asymmetric synthesis. Applications of lipases include kinetic resolution (KR) of racemic alcohols, carboxylic acids, esters, or amines [1], as well as the desymmetrization of prochiral compounds [2]. Moreover, lipases are ideal biocatalysts for ammonolysis and aminolysis of esters because they present very low amidase activity. Recently, it has been reported that amidase activity of different benzylamides depends on the acyl residue and also of the substituent groups in the aromatic ring [3]. Over the past years, interest in the preparation of chiral amines and amides by enzymatic ammonolysis or aminolysis reactions [4] has greatly increased for academic and industrial sectors. The role that the enzymatic acylation of amines or ammonia plays for the preparation of some pharmaceuticals is noteworthy [5]. Although the aminolysis of esters to amides is a useful synthetic operation, usually it presents some disadvantages in terms of drastic reaction conditions, long reaction times or strong alkali metal as catalyst, which are usually not compatible with other functional groups in the molecule [6]. For this reason, enzymatic aminolysis of carboxylic acid derivatives offers a clean and ecological way for the preparation of different kind of amines and amides in a regio-, chemo-, and enantioselective manner. In the last decade, biocatalysis in nonaqueous media, using hydrolases, has been widely used for organic chemists. The possibilities that these biocatalysts offer for the preparation of different types of organic compounds, depending upon the nucleophile


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

172

O 1

R

OR

O 1

O 2

R OR Transesterification

O H2O2

R

R OH Hydrolysis

H2O

NHR

Aminolysis O

NH3

Enz

Acyl–enzyme complex O

1

1

2

R

R NH2

2

R OH

O R1 OOH Perhydrolysis

1

2

1

R NH2 Ammonolysis

2

R NHNH2

O

1

2 R NHNHR Hydrazynolysis

Scheme 7.1 Biotransformations of carboxylic acid derivatives using lipases.

used in the process, is shown in Scheme 7.1. The use of Candida antarctica lipase B (CALB) in enzymatic ammonolysis and aminolysis reactions [7] has shown a great utility for the synthesis of a great variety of chiral nitrogenated organic compounds. In this chapter, the desire of the authors is to show a range of examples that cover the different possibilities that hydrolytic enzymes, specially lipases, can offer in the preparation of nitrogenated organic compounds through enzymatic ammonolysis and aminolysis of carboxylic acid derivatives, and thus to present an easily accessible bibliographic comprehension of this subject. Although, transesterification processes have received major attention, in the last few years, the enzymatic acylation of amines for synthetic purposes is being employed as a conventional tool for the synthesis of chiral amines and amides; however, the preparation of other nitrogenated compounds, such as hydrazides, by enzymatic hydrazinolysis reaction, is a less practical process and it has been much less investigated [8].

7.2 Mechanism of Enzymatic Ammonolysis and Aminolysis Reactions

The mechanism for the lipase-catalyzed reaction of an acid derivative with a nucleophile (alcohol, amine, or thiol) is known as a serine hydrolase mechanism (Scheme 7.2). The active site of the enzyme is constituted by a catalytic triad (serine, aspartic, and histidine residues). The serine residue accepts the acyl group of the ester, leading to an acyl–enzyme activated intermediate. This acyl–enzyme intermediate reacts with the nucleophile, an amine or ammonia in this case, to yield the final amide product and leading to the free biocatalyst, which can enter again into the catalytic cycle. A histidine residue, activated by an aspartate side chain, is responsible for the proton transference necessary for the catalysis. Another important factor is that the oxyanion hole, formed by different residues, is able to stabilize the negatively charged oxygen present in both the transition state and the tetrahedral intermediate.


Scheme 7.2 Mechanism of serine hydrolases.

7.2 Mechanism of Enzymatic Ammonolysis and Aminolysis Reactions

j173


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

174

The complex structure of the enzyme can show a very large substrate–enzyme interaction specificity, which can be traduced to a high degree of chemo-, regio-, or stereoselectivity. For this reason, nowadays, the versatility of biotransformations for synthetic proposals is an excellent tool for organic chemists [9].

7.3 Aminolysis and Ammonolysis of Carboxylic Acids

Although, the enzymatic reaction of esters with amines or ammonia have been well documented, the corresponding aminolysis with carboxylic acids are rarer, because of the tendency of the reactants to form unreactive salts. For this reason some different strategies have been used to avoid this problem. Normally, this reaction has been used for the preparation of amides of industrial interest, for instance, one of the most important amides used in the polymer industry like oleamide has been produced by enzymatic amidation of oleic acid with ammonia and CALB in different organic solvents [10]. To carry out the enzymatic amidation of carboxylic acids, normally two strategies are considered: the use of ionic liquids or the removal of water from the reaction media at high temperature or reduced pressure. For instance, one of the first examples of the use of ionic liquids in biocatalysis has been the preparation of octanamide from octanoic acid as starting material and ammonia in the presence of CALB (Scheme 7.3) [11]. Irimescu and Kato have recently described an interesting example of enzymatic KR in ionic liquids instead of organic solvents (Scheme 7.4) [12]. The resolution with CALB is based on the fact that the reaction equilibrium was shifted toward the amide synthesis by the removal of water under reduced pressure. Nonsolvent systems have been also employed in this enantioselective amidation processes, reacting racemic amines with aliphatic acids. The best reaction conditions for the conversion of acids to amides was observed using CALB at 90  C under vacuum. Meanwhile, no O OH

NH3 CALB [C4mim][BF4] 40 ºC, 4 days

O NH2 Quantitative conversion

Scheme 7.3 Example of an ammonolysis process using an ionic liquid-like solvent.

Me Ph

NH2 or

+

Me Ph

Me

Me

NH2

COOH

+

Vacuum CALB Ionic liquid

Ph

NH2

Ph

O

N H

or Me

Me Ph

NH2

Scheme 7.4 Enzymatic acylation of amines with carboxylic acids.

+ Ph

H N O


7.3 Aminolysis and Ammonolysis of Carboxylic Acids O

Ph

O NH2

+ HO R

Ph

j175

NHX 1) Thermolisin KPi buffer pH 7.5 2) TFA: H2O (95:5)

O HO O

N H

NHX R

X= Fmoc, Cbz R= H, CH2CH(CH3)2, CH2Ph, CH2CH2CONH2, + – CH2OH, CH2(imidazole)H , CH2COO , CH2CH2CH2CH3

Scheme 7.5 Enzymatic synthesis of dipeptides on solid phase.

amidation occurred in the absence of enzyme or in the presence of organic solvents, and this methodology has been employed to catalyze the enantioselective amidation of aliphatic acids by racemic 2-ethylhexyl amine [13]. Although, there are not many examples using lipases with solid phase substrates, some processes have been described. The direct enzymatic synthesis of peptides from aminoacids is an interesting alternative to chemical synthesis. The preparation of some dipeptidesonsolidsupporthasresultedinaveryusefulprocess;inthiscasethermolysin, a protease, is more efficient than lipases (Scheme 7.5) [14]. The amine is supported on PEGA [poly-(ethylene glycol)-acrylamide] and owing to the positive charge of the resin, which avoids the ionization of the amine, it is possible to achieve good yields. Alkanolamides from fatty acids are environmentally benign surfactants useful in a wide range of applications. It was found that most lipases catalyze both amidation and the esterification of alkanolamides; however, normally the predominant final products are the corresponding amides, via amidation, and also by esterification and subsequent migration [15]. Recently, an interesting example for the production of novel hydroxylated fatty amides in organic solvents has been carried out by Kuo et al. [16]. Also the impact of various reaction parameters on enzymatic synthesis of amide surfactants from ethanolamine and diethanolamine has been studied, although the possibilities of acyl migration are not investigated. However, it was found that the selectivity of the reaction depended on the solubility of the product in the solvent used, and that the choice of solvent was critical to obtain an efficient process [17]. Recently, an environmentally benign and volume efficient process for enzymatic production of alkanolamides has been described where CALB catalyzes the amidation of lauric acid and ethanolamine in the absence of solvent, at 90  C, to keep the reactants in a liquid state and to remove the water [18]. The enzyme was both very active and stable under the reaction conditions, with about half of the activity remaining after two weeks, obtaining the final amide with a 95% yield (Scheme 7.6). O

O OH

Lipase O

NH2

+ H2N

OH

Lipase O N H

Scheme 7.6 Solvent-free enzymatic synthesis of fatty alkanolamines.

OH


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

176

7.4 Aminolysis and Ammonolysis of Esters

The aminolysis of esters has been much more investigated than the corresponding aminolysis of carboxylic acids. This process is of great utility for the synthesis of nitrogenated organic compounds, especially for the resolution of amines and diamines; moreover, in some cases this process is useful for resolution of esters and for the preparation of amides of industrial interest in benign conditions. Recently, the possibilities of this reaction for the preparation of highly enantiopure b-aminoacids have also been reviewed [19]. In this section, we briefly comment on some representative examples of the enzymatic aminolysis and ammonolysis of esters for the preparation of nitrogenated compounds that are difficult to obtain by conventional chemical methods.

7.4.1 Preparation of Nonchiral Amides

The lipase-catalyzed aminolysis of nonracemic esters has been an excellent methodology for the preparation of amides, which are difficult to obtain by chemical procedures. In this section, the preparation of fatty amides and compounds used in a wide variety of industrial applications is particularly relevant. One of the first processes to obtain oleamide has been carried out by enzymatic ammonolysis reaction of triolein and ammonia. Again, CALB is the most efficient catalyst; this procedure has the advantage that purification steps are not necessary [20]. A practical and chemoselective example of enzymatic aminolysis reaction is the preparation of N-(4-methyl-4-azapentyl) acrylamide (Scheme 7.7), an important building block in the preparation of polymers. A comprehensive study of all possible parameters affecting the enzymatic aminolysis of methyl acrylate and N,N-dimethyl-1, 3-propanediamine and CALB showed that a good selectivity was achieved when the addition of the amine was done by continuous flow addition during long reaction times. Under these conditions, the formation of the Michael adduct is minimized [21]. An efficient biocatalytic method for the production of amides in multigram scale has been developed for the synthesis of a pyrrole-amide, which is an intermediate for the synthesis of the dipeptidyl peptidase IV that regulates plasma levels of the insulinotropic proglucagon. CALB catalyzes the ammonolysis of the ester with ammonium carbamate as source of ammonia (Scheme 7.8) [22]. The use of ascarite and calcium chloride as adsorbents for carbon dioxide and ethanol by-products,

OMe O

+ H2N

NMe2

CAL-B 1,4-Dioxane

H N O

Scheme 7.7 Chemoselective synthesis of an acrylamide.

NMe2

+

H N

MeO O

NMe2


7.4 Aminolysis and Ammonolysis of Esters O N OEt O

O

CALB NH2CO2NH4 CaCl2, Ascarite Toluene 50 ÂşC, 3 days

O

j177

N NH2 O

O

81% isolated yield

Scheme 7.8 Enzymatic ammonolysis using ammonium carbamate as source of ammonia.

respectively, increases the yield to 98% overcoming traditional chemical routes, which yield just 64% of the desired product. Surprisingly, the p-system geometry in a substrate has a notable inďŹ&#x201A;uence in the enzymatic aminolysis of esters. The reaction of diethyl fumarate with different amines or ammonia in the presence of CALB led to the corresponding transamidoesters with good isolated yields, but in the absence of enzyme, a high percentage of the corresponding Michael adduct is obtained (Scheme 7.9). Enzymatic aminolysis of diethyl maleate led to the recovery of the same a,b-unsaturated amidoester, diethyl fumarate, and diethyl maleate. The explanation of these results can be rationalized via a previous Michael/retro-Michael type isomerization of diethyl maleate to fumarate, before the enzymatic reaction takes place. In conclusion, diethylmaleate is not an adequate substrate for this enzymatic aminolysis reaction [23]. Recently Lin and coworkers have developed a selective synthesis of N-acyl and O-acyl propanolol vinyl derivatives by enzyme-catalyzed acylation of propanolol using divinyl dicarboxylates with different carbon chain lengths (Scheme 7.10) [24]. Lipase AY30 in diisopropyl ether demonstrated high chemoselectivity toward the amino

O O OEt EtO

O

NHR OEt

+ EtO O

OEt EtO

O

+ RNH2

O

O

CALB CALB

O

NHR EtO

+

N H

EtO

O OEt

O

O ( )n

N H

OEt O

O

+ RNH2

OEt

O

O

O

O

OEt EtO

OEt

+

OEt O

Scheme 7.9 Enzymatic aminolysis of a,b-unsaturated diesters using CALB.

O

NHR OEt

+ EtO O


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

178

O Lipase AY30 i Pr2O

N OH O

( )n O O

O OH

N H

+

COOCHCH2 ( )n COOCHCH2

MJL 1,4-Dioxane

n = 2, 4, 8 O O O

N H ( )n O O

Scheme 7.10 Controllable selective synthesis of N-acyl or O-accylpropanolol vinyl esters catalyzed by lipases.

group, whereas lipase M from Mucor javanicus in 1,4-dioxane selectively acylated the hydroxyl function. 7.4.2 Resolution of Esters

The enzymatic resolution of esters via aminolysis or ammonolysis processes represents an efficient alternative to the resolution of substrates by transesterification and hydrolysis processes. Of special relevance is the resolution of bifunctional compounds – for instance, an interesting process is the enantiopure preparation of b-hydroxyesters and b-hydroxyamides by enzymatic resolution of 3-hydroxyesters [25]. In this process, CALB was more effective in that Pseudomonas cepacia lipase (PSL) showed opposite selectivity. Azetidine ring is an important structure because it is present in many compounds of pharmaceutical interest; however, its manipulation must be done very carefully owing to the reactivity of these heterocycles of small size. An interesting application of the use of biocatalytic processes is the resolution of azetidine esters (Scheme 7.11). The procedure to choose for the resolution of these compounds is the enzymatic ammonolysis of the corresponding N-substituted azetidines [26]. A derivative of the anti-inflammatory ibuprofen, the corresponding (S)-2-chloroethyl ester, has been obtained with 96% ee by CALB-catalyzed ammonolysis of the

O

O OMe

OMe

O

NH2

CALB

NR

NH3 sat. tBuOH 35 ºC

R = Benzyl, p-methoxybenzyl, allyl

+ NR (R)-Ester > 99% ee

Scheme 7.11 Enzymatic resolution of azetidinoesters.

NR (S)-Amide


7.4 Aminolysis and Ammonolysis of Esters Cl

O

Cl O

CALB NH3

O

NH2 +

O

O

t

BuOH 48 h, rt 56% conversion

(R )-Amide (S)-Ester 96% ee

Scheme 7.12 Resolution of an ibuprofen ester derivative by enzymatic ammonolysis.

corresponding racemic ester (Scheme 7.12) [27]. Reaction was performed in tert-butyl alcohol as solvent and bubbling ammonia through the solution. The resolution of racemic ethyl 2-chloropropionate with aliphatic and aromatic amines using Candida cylindracea lipase (CCL) [28] was one of the first examples that showed the possibilities of this kind of processes for the resolution of racemic esters or the preparation of chiral amides in benign conditions. Normally, in these enzymatic aminolysis reactions the enzyme is selective toward the (S)-isomer of the ester. Recently, the resolution of this ester has been carried out through a dynamic kinetic resolution (DKR) via aminolysis catalyzed by encapsulated CCL in the presence of triphenylphosphonium chloride immobilized on Merrifield resin (Scheme 7.13). This process has allowed the preparation of (S)-amides with high isolated yields and good enantiomeric excesses [29]. The strategy described here explains the different possibilities of enzymatic ammonolysis and aminolysis reaction for resolution of esters or preparation of enantiomerically pure amides, which are important synthons in organic chemistry. This methodology has been also applied for the synthesis of pyrrolidinol derivatives that can be prepared via enzymatic ammonolysis of a polyfunctional ester, such as ethyl ()-4-chloro-3-hydroxybutanoate [30]. In addition, it is possible in the resolution of chiral axe instead of a stereogenic carbon atom. An interesting enzymatic aminolysis of this class of reaction has been recently reported by Aoyagi et al. [31]. The side chain of binaphthyl moiety plays an important role in the enantiodiscrimination of the process (Scheme 7.14). Here, we have selected a few representative examples of the enzymatic resolution of esters by aminolysis or ammonolysis reactions. On the other hand, the enzymatic acylation of racemic amines is also of great utility for the preparation of optically pure O

O OEt Cl

CCL RNH2

+

PPh3 Cl

NHR Cl R=p-OMe-Ph 92% yield 86% ee

O OEt Cl

Scheme 7.13 Dynamic kinetic resolution of 2-Chloropropionate by enzymatic aminolysis.

j179


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

180

O

CN

+ H2N OEt

Lipase

O

TBME

N H

O

+

CN

OEt

(R)-Amide

(S)-Ester

Scheme 7.14 Resolution of a binaphthyl ester by an ammonolysis reaction.

amines and amides. Next, we have summarized some examples related to the utility of these enzymatic processes. 7.4.3 Resolution of Amines, Diamines, and Aminoalcohols

The enzymatic KR between racemic amines and nonactivated esters using a lipase as biocatalyst is shown in Scheme 7.15. In the same manner as in the transesterification of secondary alcohols, this process fits Kazlauskas’ rule [32], where normally if the large group (L) has larger priority than medium group (M), the (R)-amide is obtained. In general, major size differences between both groups result in better enantioselectivities (E). Scheme 7.16 is an example of a general strategy to obtain enantiopure amines, which can be used for the preparation of a great variety of chiral amines and amides. In many cases, for the isolation of the amine, it is convenient to carry out the synthesis through formation of the corresponding salt, or by protection of the amino group, depending on the stability of the starting amine. Normally ethyl acetate is used for the acylation of primary amines, in many cases, as acyl donor and solvent. Other acylating agents such as alkyl methoxy acetates are O

NH2 L

M

O +

1

Lipase

H HN L M

2

R

OR

R1

H2N H +

L

M

Scheme 7.15 Schematic representation of Kazlauskas’ rule in the enzymatic KR of an amine. O NH2 1

R

O 2

+

R

3

R

NH2

Lipase 4

OR

1

R

2

R

(S)-Amine

+

HN

R3

R1 R2 (R)-Amide Acid liquid extraction

(S)-Amine

Basic liquid extraction

(S)-Amine-H+

Scheme 7.16 Enzymatic resolution of amines.

(R)-Amide


7.4 Aminolysis and Ammonolysis of Esters Ph ( )n

Ph

Ph

NH2

NH2

( )n NH2

n = 1, 2

n = 1, 2

Scheme 7.17 Kinetic resolution of cis and trans-2-phenylcycloalkanamines.

also of utility; however, vinyl esters, the best reagents for the resolution of alcohols, are not adequate reagents for the resolution of primary amines owing to their high reactivity. In some cases, the enantiomeric excess can be increased by using ethyl esters of long chain fatty acids. For instance, it has been described that ethyl decanoate is a better acyl agent than other esters of shorter chain for the resolution of sec-butylamine [33]. In recent years, a great variety of primary chiral amines have been obtained in enantiomerically pure form through this methodology. A representative example is the KR of some 2-phenylcycloalkanamines that has been performed by means of aminolysis reactions catalyzed by lipases (Scheme 7.17) [34]. Kazlauskas’ rule has been followed in all cases. The size of the cycle and the stereochemistry of the chiral centers of the amines had a strong influence on both the enantiomeric ratio and the reaction rate of these aminolysis processes. CALB showed excellent enantioselectivities toward trans-2-phenylcyclohexanamine in a variety of reaction conditions (E > 150), but the reaction was markedly slower and occurred with very poor enantioselectivity with the cis-isomer, whereas Candida antarctica lipase A (CALA) was the best catalyst for the acylation of cis-2-phenylcyclohexanamine (E ¼ 34) and trans-2-phenylcyclopropanamine (E ¼ 7). Resolution of both cis- and trans-2-phenylcyclopentanamine was efficiently catalyzed by CALB obtaining all stereoisomers with high enantiomeric excess. The enantiodiscrimination of the biocatalysts is not only restricted to carbon stereogenic centers, but also other chirality elements such as chirality axe or atropoisomerism have been efficiently resolved through lipase-catalyzed acylation of the amino group. Optically active 1,10 -binaphthylamine derivatives are very useful for chiral ligands in asymmetric synthesis. In Scheme 7.18 the KR of three binaphthylamines via lipase-catalyzed amidation is shown. In this case, two immobilized lipases from Pseudomonas aeruginosa were most efficient than CALB [35]. The enantioselectivity depends on the alkyl chain length between the binaphthyl ring and the amino group.

Acyl donor Lipase ( )n

NH2

n = 0,1, 2

i

Pr2O

O ( )n

N H

+ R

n = 0, (S)-enantiomer n = 1, (R)-enantiomer n = 2, (R)-enantiomer

Scheme 7.18 Enzymatic resolution of binapthylamines.

( )n

NH2

n = 0, (R)-enantiomer n = 1, (S)-enantiomer n = 2, (S)-enantiomer

j181


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

182

NH2 1

R

Me

O 2

R

+ Me

HN

Na2CO3 Toluene 90 ºC

Me

O

R

O

CALB catalyst (4 mol%)

R1

Me

R

2

R

ee = 93–99% 78–95% yield

O

H

O

R

R

R

R

H

Ru R R Ru OC CO CO CO Ruthenium catalyst (R= p-MeO-C6H4)

Scheme 7.19 Dynamic kinetic resolution of primary amines.

A KR is a useful method to obtain enantiomerically pure compounds, but suffers from the drawback that the maximum yield is only 50% of the starting material. This limitation can be overcome via a DKR process. This procedure can theoretically lead to a single product enantiomer with 100% yield. This strategy has been widely used for the resolution of alcohols by enzymatic hydrolysis or transesterification processes, the acylation of alcohols being the most explored process [36]. This methodology has also been used to obtain enantiomerically pure amines. Recently, it has been reported a versatile DKR of several primary amines using a ruthenium catalyst for the racemization process, CALB, isopropyl acetate, and sodium carbonate (Scheme 7.19) [37]. An efficient DKR process takes place obtaining the corresponding amides with high yields and enantiomeric excesses; however, still a high temperature (90  C) is needed. The power of biocatalysts for the production of chiral compounds can be elevated to a second stage when proper use of the process leads to a larger number of different enantioenriched products from the same reaction. Some efforts have been made in this direction because the enzyme can be selective to either nucleophile or acyl donor. The elegancy of the reaction is increased when the process is carried out for the resolution of both. Scheme 7.20 depicts this possibility [38]. For this reaction, CALB catalyzes the amidation between a racemic b-hydroxyester and racemic amines, leading to the corresponding amide with very high enantiomeric and diastereomeric excesses. Besides, the remaining ester and amine are recovered from the reaction media, also showing good enantiomeric excesses. By this method, three enantioenriched interesting compounds are obtained from an easy one-step reaction. Another philosophy would be the one-pot resolution of two different nucleophiles, alcohol and amine [39]. An acylated racemic alcohol reacts with a racemic amine in Me

OH O +

Cl

OEt

H2N

R

CALB

OH O Cl

Me N H

+ Cl R

R = Phenyl, furyl, pentyl

Me

OH O + OEt

H2N

R

Scheme 7.20 Double resolution of both racemic esters and amines.


7.4 Aminolysis and Ammonolysis of Esters O Me

O 1

R

O +

2

R

R

Me OH O CALB + + 1 2 2 1,4-Dioxane R1 R R R

NH2 3

4

R

O O

Me

NH2

O HN 3

NH2 Me

Me

Me +

R

NH2 3

R

R4

O O

Me Me

NH2

4

R

O Me

O

j183

NH2

Me OMe

NH2

Me

F3C

Scheme 7.21 One-pot resolution of alcohols and amines.

the presence of CALB to yield four separable enantioenriched compounds. The (R)-alcohol acts as the leaving group in the acylation of the (R)-amine. The remaining (S)-ester and (S)-amine are also recovered from the reaction media (Scheme 7.21). Apart from the synthetic utility of the reaction, the authors use it for studying the effect of the leaving group in the resolution of the nucleophile. In the case of bifunctional compounds, two KRs over the two functional groups are very practical to achieve enantiopure compounds even if in the two steps the enantioselectivity obtained is moderated. A sequential biocatalytic resolution by one-pot double aminolysis has been reported for the resolution of cyclic 1,2-diamines obtaining the final diacylated products from trans-cyclohexane-1,2-diamine and transcyclopentane-1,2-diamine in enantiopure forms [40]. The reaction is being carried out with dimethyl malonate using CALB as biocatalyst. The formation of the (R,R)bisamidoester involves two biocatalytic steps and the enzyme shows the same stereochemical preference toward the (R,R)-enantiomer of the substrate in both steps. A detailed study of the enantiomeric ratio for each step allowed us to suggest an interesting structural effect. The enantioselectivity of the second step is always higher than the first one, in good agreement with Kazlauskas’ rule, because the monoacylated compound has a bigger steric difference in the substituents of the stereocenter than the free diamine. The mechanism for this sequential KR is shown in Scheme 7.22 for the case of trans-cyclohexane-1,2-diamine. The chemoenzymatic synthesis of the analgesic U-()-50,488 [41] and new C2-symmetric bisaminoamide ligands derived from N,N-disubstituted transcyclohexane,1,2-diamine [41] has been possible by a CALB-catalyzed resolution using ethyl acetate as solvent and acyl donor [42]. Aminoalcohols are an important class of compounds in medicinal chemistry because many drugs contain this structure. For their resolution, there are two possibilities: acylation of amino function or an enzymatic transesterification with vinyl esters through the hydroxyl group. However, the amino or hydroxyl group must be protected, because if the starting material is the free aminoalcohol, the O- and N-acylation can take place, and in addition, there are migrations obtaining

NH2


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

184

Ser O NH2 Ser O NH2 NH2

O

OMe

Ser OH

O

OMe

O

O

O Ser OH

O NH OMe NH OMe

NH OMe (E2)

O O (R,R)-trans

O O (R,R)-trans

+

(E1)

NH2

(±)-trans NH2 (S,S)-trans

Scheme 7.22 Mechanism of the sequential kinetic resolution of trans-cyclohexane-1,2-diamine.

low enantioselectivities. For this reason, to achieve a good bioresolution with aminoalcohols, it is necessary to protect the amino or hydroxyl group. Normally, it is easier for the chemical acylation or alkoxycarbonylation of the amino group, and several examples have appeared recently in the literature about the enzymatic resolution of 1,2 and 1,3-aminoalcohols through an enzymatic transesterification of the corresponding amides or carbamates as starting material derivatives [43]. Although the enzymatic resolution of 1,2-aminoalcohols has been exhaustively investigated, the bioresolution of 1,3-aminoalcohols has been scarcely reported, especially through enzymatic aminolysis processes. Recently, the enzymatic resolution of 3-amino-3-phenylpropan-1-ol derivatives has been studied by an enzymatic acylation process (Scheme 7.23). CALA has been identified as the best biocatalyst for the N-acylation reaction of 3-amino-3-phenyl-1-tert-butyldimethylsilyloxy-propan1-ol using ethyl methoxyacetate as acylating agent and tert-butyl methyl ether (TBME) as solvent [44]. This is not a surprising result as CALA has been identified as an ideal biocatalyst for the resolution of sterically hindered compounds [45]. The previous enzymatic study has allowed us to obtain a valuable intermediate for the production of (S)-Dapoxetine that has been synthesized in good overall yield and high optical purity. 7.4.4 Desymmetrization of Diesters

This methodology avoids the inherent 50% yield limit of KR and the difficult separations often encountered in the resolution of racemates. The potential of enzymes, especially lipases, to catalyze the aminolysis and ammonolysis of prochiral O NH2

Me O Si tBu + MeO Me

O

CALA OEt TBME, 30 ºC

HN

OMe Me + O Si tBu Me

Scheme 7.23 Enzymatic resolution of an O-protected 1,3-aminoalcohol.

NH2

Me O Si tBu Me


7.5 Kinetic Resolution of Secondary Amines R1

MeO2C

R2NH2 CALB 1,4-Dioxane 30 ºC

CO2Me

1

1

R

2

R = OH, NHBn, OMe, OAc, Ph, p-F-C6H4

2

MeO2C

R = H, Bu, Bn

CONHR

76–99% ee

Scheme 7.24 Desymmetrization of prochiral glutarates.

substrates has been scarcely studied and only the desymmetrization of prochiral glutarates has been reported [46]. CALB is again the only biocatalyst that yields good results and reacts with the pro-(R)-ester group to yield the monoamidoester of (S) configuration with very high yields and enantiomeric excesses (Scheme 7.24). In some cases, these products are excellent starting material for the synthesis of aminoacids of physiological interest [47].

7.5 Kinetic Resolution of Secondary Amines

The structure of the amine has a great influence in the lipase-catalyzed aminolysis reaction. Although primary amines can be efficiently resolved, few examples have been reported about the acylation or resolution of secondary amines. On the other hand, acyclic secondary amines are not good substrates for this enzymatic reaction. However, cyclic secondary amines are structurally better than acyclic compounds for their resolution, and several examples of resolution of pyrrolidine and piperidine derivatives have been described. Kanerva et al. reported the acylation of pipecolic acid derivatives catalyzed by CALA [48]. As explained above, this enzyme seems to have a larger pocket than CALB in the active site and accepts these bulkier substrates [45]. This reaction has also been applied to pyrrolidine ring and in both cases the enzyme was very efficient, catalyzing the acylation of the secondary amino group with very high enantioselectivity (Scheme 7.25). Enzymatic alkoxycarbonylation gives better results by enzymatic resolution of 1-methyl tetrahydroisoquinoline than the corresponding acylation [49]. The best results are obtained with a 3-methoxyphenyl and allyl mixed carbonate as acylating reagent and Candida rugosa lipase (CRL) as biocatalyst. The best solvent is toluene and the (S)-amine and the (R)-carbamate were obtained with high yields and enantiomeric excesses (Scheme 7.26). Recently, Page and coworkers have reported an efficient process for the DKR of this secondary amine using a novel iridium-based amine racemization catalyst that has allowed the preparation of the corresponding

O N H

OMe + O

R

O

CF3

CALA Solvent

OMe

N O

R

Scheme 7.25 Enzymatic resolution of pipecolic methyl esters.

O

+

N H

OMe O

j185


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

186

OMe CRL Toluene (0.05% H2O) 30 ºC, 8 h

O

NH + O

O

N

O

+

O (R)-Carbamate 98% ee 47% isolated yield

NH

(S)-Amine 99% ee 46% isolated yield

OMe CRL, [IrCp*I2]2

O

NH + O

O

N

Toluene 40 ºC, 23 h 90% conversion

O

O 96% ee 82% isolated yield

Scheme 7.26 Kinetic resolution and DKR of 1-methyl tetrahydroisoquinoline using Candida rugosa lipase.

R2

3

R

* * N

2

R3

R

O

1

R + N H

R4O

CALA or CALB O

TBME

2

+ O

R

3

R O

*

* N H

1

R

Scheme 7.27 Enzymatic preparation of enantiomerically pure indolines.

carbamate in 82% isolated yield and 96% ee using 3-methoxyphenyl propyl carbonate as alkoxycarbonylating agent [50]. An efficient chemoenzymatic route for the synthesis of optically active substituted indolines has been recently developed (Scheme 7.27), and also the alkoxycarbonylation process is more efficient than the acylation reaction. Different lipases have been tested in the alkoxycarbonylation of these secondary amines, CALA being found to be the best biocatalyst for 2-substituted-indolines, and CALB for 3-methylindoline. The combination of lipases with a variety of allyl carbonates and TBME as solvent has allowed the isolation of the carbamate and amine derivatives in a high level of enantiopurity [51].

7.6 Toward the Synthesis of b-Aminoacid Derivatives

Preparation of optically active b-aminoesters, b-aminonitriles, and b-aminocarboxamides are of special relevance for the synthesis of enantiomerically pure b-aminoacids compounds of special relevance in several areas of medicinal chemistry. The resolution of b-aminoesters can be carried out by acylation of the amino groups or by other biocatalytic reactions of the ester groups, such as hydrolysis, transesterification, or aminolysis. The resolution of ethyl ()-3-aminobutyrate


7.6 Toward the Synthesis of b-Aminoacid Derivatives

j187

O NH2 O Me

O

CALB Solvent E = 74–88

+ Me

OEt

OEt

Me

NH O

NH2 O

+ Me

OEt

OEt CbzCl Na2CO3 H2O O CbzHN Me

NH O

O

+ Me

OEt

OEt NH2 O Me

RNH2

+ OEt

R = H, Bn

CbzHN

1) CALB, solvent 2) CbzCl, Na2CO3, H2O E = 4–16

O

CbzHN

Me

NHR

Me

Scheme 7.28 Enzymatic resolution of b-aminoesters.

is more efficient by acylation of the amino group than by the aminolysis of the ester (Scheme 7.28) [52]. Other important derivatives for the preparation of b-aminoacids are the corresponding b-aminonitriles. Lipase-catalyzed N-acylations of racemic cis-2aminocyclopentane and cyclohexane carbonitriles with 2,2,2-trifluoroethyl butanoate have been successfully carried out in organic solvents and ionic liquids [53], PSL yielding better results than CALB (Scheme 7.29). Other derivatives to obtain b-aminoacids are the corresponding carboxamides. Thus, for the preparation of all enantiomers of cis and trans–2-aminocyclopentaneand cyclohexanecarboxamides, the best results obtained are using this acyl donor and CALB [54]. An unexpected change in enantiopreference accompanied by low enantioselectivity was observed when PSL (cis-cyclohexane substrate) or CALA (cis-cyclopentane and cyclohexane substrates) replaced CALB (Scheme 7.30). Recently, a very interesting preparation of b-peptides has been carried out by Kanerva and coworkers through a two-step lipase-catalyzed reactions in which N-acetylated b-amino esters were first activated as 2,2,2-trifluoroethyl esters with CALB [55]. The activated esters were then used to react with a b-aminoester in the presence of CALA in dry diethylether or diisopropylether (Scheme 7.31). In this peptide synthesis, the aminoester was used in excess and the unreacted counterpart was easily separated and later recycled.

CN ( )n

O

Lipase

+ NH2

Pr

OR

Solvent

( )n

CN O N H

CN + ( )n Pr

(±)-cis, n = 1, 2

Scheme 7.29 Enzymatic resolution of b-aminonitriles.

O

+

NH2

OEt


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

188

O

O NH2

( )n

O

Lipase

+ Pr

O

CF3

O NH2

( )n

+

Solvent

HN

NH2

( )n

Pr

NH2 NH2

O

(±)-trans, n = 1, 2

(1R,2R) n = 1 (1R,2R) n = 2

(1S,2S) n = 1 (1S,2S) n = 2

Scheme 7.30 Enzymatic resolution of b-aminocarboxamides.

1

O Me

R N H

O

O OEt

+

1

O

R

O

CALB

Pr

O

CF3

Me 1

R

O

N H

2

R

CF3

R2

O CALA

3

H2N

OR R2 1

O Me

N H

3

O

R

2

R

O

R N H

OR5 4

R

Scheme 7.31 Lipases in b-dipeptide synthesis in organic solvents.

7.7 Summary and Outlook

Hydrolytic enzymes are the most used biocatalysts, and the application of the biocatalysis in organic synthesis is currently a well-established methodology for the chemo- and regioselective modification of nonchiral compounds, resolution of racemates, and desymmetrization of prochiral substrates. In this chapter, we have shown how hydrolases, especially lipases, catalyze enzymatic aminolysis and ammonolysis of carboxylic acids and esters in different solvent systems, providing a potential application for the preparation of amines and amides, compounds which play an important role in the production of fine chemicals and pharmaceuticals. Presently, it is clear that for the resolution of primary amines CALB is the best biocatalyst; however, CALA results are more appropriate for the resolution of secondary amines. In addition, the combination of genetic engineering and molecular modeling techniques are playing a major role in the development of enzymes that will show better results in the future than those currently achieved, and will offer new pathways and possibilities for the synthesis and resolution of interesting manufactures for the industrial sector.


References

References 1 Ghanem, A. and Aboul-Enein, H.Y. (2004) Tetrahedron: Asymmetry, 15, 3331–3351. 2 Garcıa-Urdiales, E., Alfonso, I. and Gotor, V. (2005) Chemical Reviews, 105, 313–354. 3 Torres-Gavilan, A., Castillo, E. and LópezMunguıa, A. (2006) Journal of Molecular Catalysis B: Enzymatic, 41, 136–140. 4 (a) Gotor, V. (1999) Bioorganic and Medicinal Chemistry, 7, 2189–2197. (b) van Rantwijk, F. and Sheldon, R.A. (2004) Tetrahedron, 60, 501–519. (c) Alfonso, I. and Gotor, V. (2004) Chemical Society Reviews, 33, 201–209. (d) GotorFernandez, V. and Gotor, V. (2006) Current Organic Chemistry, 10, 1125–1143. 5 Gotor-Fernandez, V., Rebolledo, F. and Gotor, V. (2007) Preparation of chiral pharmaceuticals through enzymatic acylation of alcohols and amines, in Biocatalysis in the Pharmaceutical and Biotechnology Industry, (ed. R.M. Patel), Dekker, Taylor and Francis, New York, Chapter 7, pp. 203–248. 6 Hayes, K.S. (2001) Applied Catalysis A-General, 221, 187–195. 7 Gotor Fernandez, V., Busto, E. and Gotor, V. (2006) Advanced Synthesis and Catalysis, 348, 797–812. 8 (a) Astorga, C., Rebolledo, F. and Gotor, V. (1993) Synthesis, 287–289. (b) Hacking, M.A.P.J., Akkus, H., van Rantwijk, F. and Sheldon, R.A. (2000) Biotechnology and Bioengineering, 68, 84–91. 9 (a) Bornscheuer, U.T. and Kazlauskas, R.J. (2006) Hydrolases in Organic Synthesis, 2nd edn, Wiley-VCH Verlag GmbH. (b) Faber, K. (2004) Biotransformations in Organic Chemistry, 5th edn, Springer-Verlag, Heildeberg. (c) Bommarius, A.S. and Riebel, B.R. (2004) Biocatalysis, Wiley-VCH Verlag GmbH, Weinheim. 10 Slotema, W.F., Sandoval, G., Guieysse, D., Straathof, A.J.L. and Marty, A. (2003)

11

12

13

14

15

16

17

18

19 20

21

22

Biotechnology and Bioengineering, 82, 664–669. Madeira Rau, R., Rantwijk, F.V., Seddon, K.R. and Sheldon, R.A. (2000) Organic Letters, 2, 4189–4191. (a) Irimescu, R. and Kato, K. (2004) Journal of Molecular Catalysis B: Enzymatic, 30, 189–194. (b) Irimescu, R. and Kato, K. (2004) Tetrahedron Letters, 45, 523–525. Prasad, A.K., Husain, M., Singh, B.K., Gupta, R.K., Manchanda, V.K., Olsen, C.E. and Parmar, V.S. (2005) Tetrahedron Letters, 46, 4511–4514. Ulijn, R.U., Baragaña, B., Halling, P.J. and Flitsch, S.L. (2002) Journal of the American Chemical Society, 124, 10988–10989. Furutani, T., Furui, M., Ooshima, H. and Kato, J. (1996) Enzyme and Microbial Technology, 19, 578–584. Levinson, W.E., Kuo, T.M. and Kurtzman, C.P. (2005) Enzyme and Microbial Technology, 37, 126–130. (a) Fernandez-Perez, M. and Otero, C. (2001) Enzyme and Microbial Technology, 28, 527–536. (b) Fernandez-Perez, M. and Otero, C. (2003) Enzyme and Microbial Technology, 33, 650–660. Tufvesson, P., Annerling, A., Hatti-Kaul, R. and Adlecreutz, D. (2007) Biotechnology and Bioengineering, 97, 447–453. Liljeblad, A. and Kanerva, L.T. (2006) Tetrahedron, 62, 5831–5854. De Zoote, M.C., Kock van-Dalen, A.C., van Rantwijk, F. and Sheldon, R.A. (1996) Journal of Molecular Catalysis B: Enzymatic, 2, 141–145. Torre, O., Gotor-Fernandez, V., Alfonso, I., Garcıa-Alles, L.F. and Gotor, V. (2005) Advanced Synthesis and Catalysis, 347, 1007–1014. Gill, I. and Patel, R. (2006) Bioorganic and Medicinal Chemistry Letters, 16, 705–709.

j189


j 7 Aminolysis and Ammonolysis of Carboxylic Acid Derivatives

190

23 Quirós, M., Astorga, C., Rebolledo, F. and Gotor, V. (1995) Tetrahedron, 51, 7715–7720. 24 Quan, J., Wang, N., Cai, X.-Q., Wu, Q. and Lin, X.-F. (2007) Journal of Molecular Catalysis B: Enzymatic, 44, 1–7. 25 Garcıa, M.J., Rebolledo, F. and Gotor, V. (1993) Tetrahedron: Asymmetry, 4, 2199–2210. 26 Starmans, W.A.J., Doppen, R.G., Thijs, L. and Zwanemburg, B. (1998) Tetrahedron: Asymmetry, 9, 429–435. 27 de Zoete, M.C., Kock-van Dalen, A.C., van Rantwijk, F. and Sheldon, R.A. (1993) Journal of the Chemical Society, Chemical Communications, 1831–1832. 28 Gotor, V., Rebolledo, F. and Brieva, R. (1988) Tetrahedron Letters, 29, 6973–6974. 29 BadjicJ.D., Kadnikova, E.N. and Kostic, N.M. (2001) Organic Letters, 3, 2025–2028. 30 Garcıa-Urdiales, E., Rebolledo, F. and Gotor, V. (1999) Tetrahedron: Asymmetry, 10, 721–726. 31 Aoyagi, N., Kawauchi, S. and Izumi, T. (2004) Tetrahedron Letters, 45, 5189–5192. 32 Kazlauskas, R.J., Weissfloch, A.N.E., Rappaport, A.T. and Cuccia, L.A. (1991) Journal of Organic Chemistry, 56, 2656–2665. 33 Goswami, A., Guo, Z., Parker, W.L. and Patel, R.N. (2005) Tetrahedron: Asymmetry, 16, 1715–1719. 34 (a) Gonzalez-Sabin, J., Gotor, V. and Rebolledo, F. (2005) Tetrahedron: Asymmetry, 15, 481–488. (b) GonzalezSabin, J., Gotor, V. and Rebolledo, F. (2005) Tetrahedron: Asymmetry, 16, 3070–3076. 35 Aoyagi, N. and Izumi, T. (2002) Tetrahedron Letters, 43, 5529–5531. 36 Pamies, O. and B€ackvall, J.E. (2003) Chemical Reviews, 103, 3247–3262. 37 Paetzold, J. and B€ackvall, J.E. (2005) Journal of the American Chemical Society, 127, 17620–17621. 38 Sanchez, V.M., Rebolledo, F. and Gotor, V. (1999) Journal of Organic Chemistry, 64, 1464–1470.

39 Garcıa-Urdiales, E., Rebolledo, F. and Gotor, V. (2000) Tetrahedron: Asymmetry, 11, 1459–1463. 40 (a) Alfonso, I., Astorga, C., Rebolledo, F. and Gotor, V. (1996) Chemical Communications, 2471–2472. (b) Luna, A., Alfonso, I. and Gotor, V. (2002) Organic Letters, 4, 3627–3629. 41 Gonzalez-Sabin, J., Gotor, V. and Rebolledo, F. (2004) Chemistry - A European Journal, 10, 5788–5794. 42 Gonzalez-Sabin, J., Gotor, V. and Rebolledo, F. (2006) Tetrahedron: Asymmetry, 17, 449–454. 43 (a) Maestro, A., Astorga, C. and Gotor, V. (1997) Tetrahedron: Asymmetry, 8, 3153–3159 (b) López-Garcıa, M., Alfonso, I. and Gotor, V. (2004) Chemistry- A European Journal, 10, 3006–3014. 44 Torre, O., Gotor-Fernandez, V. and Gotor, V. (2006) Tetrahedron: Asymmetry, 16, 860–866. 45 Domınguez de Marıa, P., CarboniOerlemans, C., Tuin, B., Bargeman, G., Van de Meer, A. and Van Gemert, R. (2005) Journal of Molecular Catalysis B: Enzymatic, 37, 36–46. 46 (a) Puertas, S., Rebolledo, F. and Gotor, V. (1996) Journal of Organic Chemistry, 61, 6024–6027. (b) Jacobsen, E.E., Hoff, B.H., Moen, A.R. and Anthonsen, T. (2003) Journal of Molecular Catalysis B: Enzymatic, 21, 55–58. (c) López-Garcıa, M., Alfonso, I. and Gotor, V. (2003) Tetrahedron: Asymmetry, 14, 603–609. 47 López-Garcıa, M., Alfonso, I. and Gotor, V. (2003) Journal of Organic Chemistry, 68, 648–665. 48 Liljeblad, A., Lindborg, J., Kanerva, A., Katajisto, J. and Kanerva, L.T. (2002) Tetrahedron Letters, 43, 2471–2474. 49 Breen, G.F. (2004) Tetrahedron: Asymmetry, 15, 1427–1430. 50 Stirling, M., Blacker, J. and Page, M.I. (2007) Tetrahedron Letters, 48, 1247–1250. 51 Gotor-Fernandez, V., Fernandez-Torres, P. and Gotor, V. (2006) Tetrahedron: Asymmetry, 17, 2558–2564.


References 52 Sanchez, V.M., Rebolledo, F. and Gotor, V. (1997) Tetrahedron: Asymmetry, 8, 37–40. 53 Fitz, M., Lundell, K., Lindros, M., F€ ul€op, F. and Kanerva, L.T. (2005) Tetrahedron: Asymmetry, 16, 3690–3697.

54 Fitz, M., Lundell, K., Lindros, M., F€ ul€op, F. and Kanerva, L.T. (2006) Tetrahedron: Asymmetry, 17, 1129–1134. 55 Li, X.-G. and Kanerva, L.T. (2006) Organic Letters, 8, 5593–5596.

j191


j193

8 Enzymatic Reduction Reaction Kaoru Nakamura and Tomoko Matsuda

8.1 Introduction

Reduction is one of the most important, fundamental, and practical reactions â&#x20AC;&#x201C; for example, reduction of ketones affords chiral compounds, which can be transformed into suitable forms with various functionalities to synthesize industrially important chemicals such as pharmaceuticals, agrochemicals, and natural products. The catalysts for the asymmetric reduction can be classiďŹ ed into two categories: chemical and biological. Both have their own peculiarities, and development of both to enable the appropriate selection of the catalysts for particular purposes is necessary to promote green chemistry. For example, biocatalysts have the advantages of being natural, having high chemo-, regio-, and enantioselectivity, and being active under mild reaction conditions. Here, new methodology for improving reactivity and selectivity of enzymatic reduction will be given. Synthetic applications of enzymatic reductions are also discussed.

8.2 Hydrogen Source for Coenzyme Regeneration

Enzymes that perform reduction of carbonyl groups usually require a coenzyme from which a hydride is transferred to the carbonyl carbon. Since reduction of the substrate is concomitant with oxidation of the coenzyme, and the coenzyme is too expensive as a throwaway reagent, it is necessary to recycle and reuse the oxidized form of the coenzyme [1]. The oxidized form of the coenzymes has to be transformed to the reduced form for the next cycle of reduction of substrate. Hydrogen sources are necessary to perform this reduction reaction. For biocatalytic reduction, alcohols such as ethanol and 2-propanol, sugars such as glucose, glucose-6-phosphate, and glucose-6-sulfate, formic acid, and dihydrogen can be used. Some examples are shown in this section.


j 8 Enzymatic Reduction Reaction

194

Geotrichum candidum O

OH Ph

Ph NADH

Cofactor recycling

NAD+

O

OH

Figure 8.1 NADH recycling using alcohol as a hydrogen source for reduction [2].

8.2.1 Alcohol as a Hydrogen Source of Reduction

Alcohols such as ethanol, 2-propanol, and so on, have been widely used to recycle the coenzyme for the reduction catalyzed by alcohol dehydrogenase since the enzyme catalyzes both reduction and oxidation. Usually, an excess amount of the hydrogen source is used to push the equilibrium toward formation of product alcohols. There is an interesting example of the use of secondary alcohols in which the enantioselectivity and reaction yield was improved by recycling the coenzyme using secondary alcohols for the reduction of ketones with Geotrichum candidum Figure 8.1 [2]. Although the enantioselectivity of the reduction of acetophenone with resting cell of the microbe was low, the use of the dried cells as a biocatalyst and addition of a catalytic amount of NAD(P)Ăž and excess amount of secondary alcohols such as 2-propanol and cyclopentanol increased enantioselectivity and also chemical yield of the reduction; the (S)-alcohol was obtained in >99% ee with high yield. However, the addition of stoichiometric amount of NAD(P)H in the dried-cell system gave low enantioselectivity. 8.2.2 Sugar as a Hydrogen Source of Reduction

Glucose and glucose-6-phosphate have been widely used as a reducing energy. Recent example shows that a thermostable glucose-6-phosphate dehydrogenase (G6PDH) from Bacillus stearothermophillus was used for recycling NADPH at high temperature (55  C) for the reduction of 2-butanone by a thermostable alcohol dehydrogenase from Thermoanaerobacter brockii. Glucose-6-sulfate was used instead of glucose-6phosphate and the sulfate is three times more effective than the phosphate, the natural substrate for G6PDH (Figure 8.2) [3]. 8.2.3 Formate as a Hydrogen Source of Reduction

Formate is one of the most representative hydrogen sources for the biocatalytic reduction because CO2 formed by the oxidation of formate is released easily from the reaction system [4]. For example, for the reduction of aromatic ketones by the


8.2 Hydrogen Source for Coenzyme Regeneration

6-Sulfo-gluconolactone

NADPH

j195

2-Butanone

ADH G6PDH (Bacillus stearothermophillus) (Thermoanaerobacter brockii) NADP+

Glucose-6-sulfate

2-Butanol

Figure 8.2 Reduction of ketone with alcohol dehydrogenase from Thermoanaerobacter brockii using glucose-6-sulfate as a hydrogen source [3].

recombinant (S)-alcohol dehydrogenase from Rhodococcus erythropolis, formate dehydrogenase (FDH) from Candida boidinii was used for regeneration of NADH. By using hexane as an organic solvent in biphasic reaction medium, substrate concentrations of 10â&#x20AC;&#x201C;200 mM can be used, and the resulting (S)-alcohols were formed with moderate to good conversion rates, and with up to >99% ee (Figure 8.3) [4]. Formate-FDH system was also applied in the reduction of 6-bromotetralone to (S)-6-bromotetralol, a potential pharmaceutical precursor, with the NADH-dependent ketone reductase from Trichosporon capitatum [4b]. A resin (XAD L-323) was used to bind the product (Figure 8.4). 8.2.4 Molecular Hydrogen as a Hydrogen Source of Reduction

Molecular hydrogen has been used for the recycling of coenzymes [5]. For example, the soluble hydrogenase I (H2 : NADPĂž oxidoreductase, EC 1.18.99.1) from the marine hyperthermophilic strain of the archaeon Pyrococcus furiosus (PF H2ase I) has been used as a biocatalyst in the enzymatic production and regeneration of NADPH utilizing molecular hydrogen. Utilizing the thermophilic NADPH-dependent alcohol dehydrogenase from Thermoanaerobium species (ADH M) coupled to the PF H2ase I in situ NADPH-regenerating system, two prochiral model substrates, acetophenone and (2S)-hydroxy-1-phenyl-propanone, were quantitatively reduced

O

O H3C

H3C CO2

+

NADPH + H

Formate dehydrogenase (Candida boidini)

ADH (Rhodococcus erythropolis )

Hexane layer OH

OH

NADP+

HCO2H

Cl

Cl

H3C

H3C Aqueous layer 95% conversion ee > 99% Figure 8.3 Reduction of ketone with alcohol dehydrogenase from Rhodococcus erythropolis using formate as a hydrogen source [4a].

Cl

Cl


j 8 Enzymatic Reduction Reaction

196

O

XAD L-323 reductase (Trichosporon capitatum)

Br

OH (S) Br

NADH

NAD+

CO2

HCO2H FDH (Candida boidinii)

Figure 8.4 Reduction of 6-bromotetralone with reductase from Trichosporon capitatum using formate as a hydrogen source [4b].

to the corresponding (S)-alcohol and (1R,2S)-diol in which >99.5% ee and >98% de, respectively, with total turnover numbers (ttn: mol product/mol consumed cofactor NADPþ) of 100 and 160 could be reached (Figure 8.5) [5a]. Another example for the use of hydrogen as reductant is observed in the reduction of imine [5b]. New imine reductase activity has been discovered in the anaerobic bacterium Acetobacterium woodii by screening a dynamic combinatorial library of virtual imine substrates, using a biphasic water–tetradecane solvent system. A ruthenium complex [RuCl2(TPPTS)2]2 was used for regeneration of NADPþ to NADPH with hydrogen. Thus, 2-heptanone was reduced with alcohol dehydrogenase from Thermoanaerobacter brockii in the presence of the ruthenium complex, NADPþ, and hydrogen at 60 C to (S)-2-heptanol in 40 % ee. Turnover number was reported to be 18 (Figure 8.6) [5c]. 8.2.5 Light Energy as a Hydrogen Source of Reduction

Photochemical methods [6] have been developed to provide an environmentally friendly system that employs light energy to regenerate NAD(P)H, for example, by the use of a cyanobacterium, a photosynthetic biocatalyst. Using the biocatalysts, the O

OH

H3C

99.5% ee

H3C ADH (Thermoanaerobacter sp.) O

OH

H3C

H3C NADPH

OH

NADP+

H2 Hydrogenase (Pyrococcus furiosus ) Figure 8.5 Reduction of ketone with alcohol dehydrogenase from Thermoanaerobacter species using molecular hydrogen as a hydrogen source [5a].

OH

> 99.5% ee


8.2 Hydrogen Source for Coenzyme Regeneration

O NADPH + H + [RuCl2(TPPTS) 2]2 H2

TBADH

OH

(S) 40% ee

NADP +

Figure 8.6 Reduction of ketone with ruthenium complex and alcohol dehydrogenase using molecular hydrogen as a hydrogen source [5c].

reduction of acetophenone derivatives occurred more effectively under illumination than in the dark (Figure 8.7) [6b,c]. The light energy harvested by the cyanobacterium is converted into chemical energy in the form of NADPH through an electron transfer system, and, consequently, the chemical energy (NADPH) is used to reduce the substrate to chiral alcohol (96 to >99% ee). The light energy, which is usually utilized to reduce CO2 to synthesize organic compounds in the natural environment, was used to reduce the substrate in this case. When a photosynthetic organism is omitted, the addition of a photosensitizer is necessary. The methods use light energy to promote the transfer of an electron from a photosensitizer to NAD(P)þ via an electron transport reagent [6g]. Recently, carbon dioxide was reduced to formic acid with FDH from Saccharomyces cerevisiae in the presence of methylviologen (MV2þ) as a mediator, zinc tetrakis(4-methylpyridyl) porphyrin (ZnTMPyP) as a photosensitizer, and triethanolamine (TEOA) as a hydrogen source (Figure 8.8) [6h].

Light

Light-harvesting system

H2O O

Electron transfer system

NADPH

Cofactor+ recycling NADP

O2 ee 96 – > 99% H OH

X

X (S)

CO2

Calvin cycle

Organic compounds

Cyanobacteria Figure 8.7 Reduction of ketone with photosynthetic biocatalyst using light energy [6b,c].

j197


j 8 Enzymatic Reduction Reaction

198

Formate DH CO2

HCO2H

(HOCH2CH2)3N Light, ZnTMPyP MV2+

Figure 8.8 Reduction of carbon dioxide with formate dehydrogenase and porphyrin complex using light energy [6h].

8.2.6 Electric Power as a Hydrogen Source of Reduction

Electrochemical regeneration of NAD(P)H has long been recognized as a potentially powerful technology from the viewpoint of green chemistry, as it would not require a second enzyme and cosubstrate [1b]. However, the method is not effective because of the necessity of high overpotentials with direct reduction of the coenzyme, electrode fouling, dimerization of the coenzyme, and the fact that only the enzyme in the immediate vicinity of the electrode will be productive. Viologen-diaphorase (lipoamide dehydrogenase) were used for reduction of NAD(P)þ, where viologens were used as a mediator of an electron from electrode to diaphorase (Figure 8.9) [7a]. Organometallics such as rhodium complex were also used for electrochemical regeneration of NAD(P)H from electrode (Figure 8.10) [7b]. Surface-modified electrodes were used for prevention of high overpotentials with direct oxidation or reduction of the cofactor, electrode fouling, and dimerization of the cofactor [7c]. Membrane electrochemical reactors were designed. The regeneration of the cofactor NADH was ensured electrochemically, using a rhodium complex as electrochemical mediator. A semipermeable membrane (dialysis or ultrafiltration) was integrated in the filter-press electrochemical reactor to confine MV+ e–

MV2+

DI'

DIred

NAD+

MV+

DIox

DIox

NADH

Figure 8.9 Regeneration of NADH using diaphoras and electric power. DI: diaphorase [7a].

2+ N 2e–

N Rh

Cp

NADH

OH2 +

H+ N

N

NAD+

Rh Cp

H

Figure 8.10 Regeneration of NADH with rhodium complex using electric power [7b].


8.3 Methodology for Stereochemical Control

[Cp(Me)5Rh(bipy)Cl]+

Cyclohexanone

NADH

ADH

2e– [Cp(Me)5Rh(bipy)H]+

NAD+

Electrode

Cyclohexanol Membrane

Figure 8.11 Reduction of cyclohexanone with alcohol dehydrogenase and rhodium complex using electric power [7b].

the enzyme(s) as close as possible to the electrode surface. Cyclohexanone was reduced effectively to cyclohexanone using horse liver alcohol dehydrogenase (HLADH) (Figure 8.11) [7d,e]. ToincreasetheefficiencyofelectrochemicalreductionofNADPþ,apolyaminoanilinemodified electrode was prepared, and the reduction of NADPþ by ferredoxine NADPþ reductase was investigated. Although the immobilization of only the copolymerized viologen showed no activity, immobilization of both viologen and NADPþ on the electrode showedthe reduction wave at1.2 V versus SCE (saturated calomel electrode), corresponding to that of free NADPþ on the electrode [7].

8.3 Methodology for Stereochemical Control

Since stereoselectivities of biocatalytic reductions are not always satisfactory, modification of biocatalysis are necessary for practical use. This section explains how to find, prepare, and modify the suitable biocatalysts, how to recycle the coenzyme, and how to improve productivity and enantioselectivity of the reactions. 8.3.1 Screening of Microbes

Screening for a novel enzyme is a classical method and still one of the most powerful tools for finding the biocatalytic reduction system [8]. It is possible to discover a suitable biocatalyst by applying the latest screening and selection technologies, allowing rapid identification of enzyme activities from diverse sources [8a]. Enzyme sources used for screening for asymmetric reductions in organic synthesis can be soil samples, commercial enzymes, culture sources, or a clone bank, and so on. Their origin can be microorganisms, animals, or plants. From these sources, enzymes that are regularly expressed and those that are not expressed in the original host can be tested to determine whether they are suitable for the transformation of certain substrates [8a]. For example, a screening of 416 strains (71 bacterial strains, 45 actinomycetes, 59 yeast, 60 basidiomycetes, 33 marine fungi, and 148 filamentous fungi) has been performed to look for microorganisms that display reductase activity in the absence of oxidase activity [8b]. A new microorganism, Diplogelasinospora grovesii IMI

j199


j 8 Enzymatic Reduction Reaction

200

Suitable microbe

Substrate

Product

(a)

Screening

OH

O Diplogelasinospora grovesii

> 98% ee

Reference

71 bacterial strains, 45 actinomycetes, 59 yeast, 60 basidiomycetes, 33 marine fungi , 148 filamentous fungi

[8b]

(S)

(b) O

OH

Candida magnoliae

CO2Et

Cl

Cl

CO2Et

â&#x20AC;&#x201C;1

90 gl , ee 96.6% (ee 99% after heat treatment) (c)

[8c]

450 bacteria

[8d]

OH

Klebsiella pneumoniae IFO 3319

O

400 yeast

(S)

CO2Et

CO2Et Yield 99%, de 99% , ee >99% (2 kg in 200-l fermentor )

(2R, 3S)

(d) O Cl

O

OMe

80 microorganisms (yeast, fungi, bacteria)

[8e]

N

Microbacterium campoquemadoensis (MB5614) OH Cl

O

OMe

(S) 500 mglâ&#x20AC;&#x201C; ee > 95%

N

(e)

KRED 101

O

OH

CO2Et CO2Et NADPH

CO2Et

Keto-reductase-screening kit

Glucose dehydrogenase

[8f]

CO2Et

(3S,4R)

Figure 8.12 Screening for a biocatalyst [8].

171018, was isolated and showed very high activity and stereoselectivity in the reduction of cyclic ketones. For example, the bicycloketone were reduced in >98% ee (S) (Figure 8.12a). In another example, 400 yeasts were screened for the reduction of ethyl 4-chloro-3oxobutanoate, and Candida magnoliae was found to be the best one, as shown in Figure 8.12b [8c]. For the reduction of ethyl 2-methyl-3-oxobutanoate, Klebsiella pneumoniae IFO 3319 out of 450 bacterial strains was found to give the corresponding (2R,3S)-hydroxy ester with 99% de and >99% ee (Figure 8.12c) [8d]. Screening techniques have also been applied to drug synthesis. For example, a key intermediate in the synthesis of the antiasthma drug, Montelukast, was prepared from the


8.3 Methodology for Stereochemical Control

corresponding ketone by microbial transformation as shown in Figure 8.12d. The biotransforming organism Microbacterium campoquemadoensis (MB5614) was discovered as a result of an extensive screening program [8e]. Keto-reductase-screening kit is now available (10 enzyme for 2295 from Biocatalytics Inc.,) and this kit was used for screening of reduction of keto diester (Figure 8.12e). KRED101 afforded (3S,4R)hydroxydiester in 90% de with 100% ee [8f ]. There is also an example for the systematic screening of microbes against aromatic ketones [8g]. A collection of about 300 microbes was surveyed for the ability to generate chiral secondary alcohols by enantioselective reduction of a series of alkyl aryl ketones. Microbial cultures demonstrating utility in reducing model ketones were arrayed in multiwell plates and used to rapidly identify specific organisms capable of producing chiral alcohols used as intermediates in the synthesis of several drug candidates. Approximately 60 cultures were shown to selectively reduce various ketones providing both the R and S enantiomers of the corresponding alcohols in 92–99% ee with yields up to 95% at 1–4 g l1. An alternative approach to chiral alcohols based on selective microbial oxidation of racemic alcohols is also reported. This study provides a useful reference for generating chiral alcohols by selective microbial bioconversion.

8.3.2 Modification of Biocatalysts by Genetic Methods 8.3.2.1 Modified Yeast Recently, various genetic methods for screening, as well as classical methods, have been reported in the search for effective biocatalytic reduction. One of the most interesting examples is the use of glutathione S-transferase (GST)-fusion proteins to allow rapid identification of synthetically useful biocatalysts [9]. A set of fusion proteins consisting of GST linked to the N-terminus of putative dehydrogenases produced by baker’s yeast (S. cerevisiae) was screened for the reduction of various substrates. For example, ethyl 2-oxo-4-phenylbutyrate was reduced rapidly in the presence of NADH and NADPH by two dehydrogenases, Ypr1p and Gre2p, providing the (R)- and (S)-alcohol, respectively, with high stereoselectivities (Figure 8.13a) [9a]. The same enzymes were overexpressed in their native forms in Escherichia coli and growing cells of the engineered strains could also be used to carry out the reductions without the need for exogenous cofactor. A representative set of a- and b-keto esters was also tested as substrates (total 11) for each purified fusion protein (Figure 8.13b,c) [9b]. The stereoselectivities of -keto ester reductions depended both on the identity of the enzyme and the substrate structure, and some reductases yielded both L- and D-alcohols with high stereoselectivities. While a-keto esters were generally reduced with lower enantioselectivities, it was possible to identify pairs of yeast reductases that delivered both alcohol antipodes in optically pure form. These results demonstrate the power of genomic fusion protein libraries to identify appropriate biocatalysts rapidly and expedite process development.

j201


j 8 Enzymatic Reduction Reaction

202

OH (a) CO2Et

[9a]

(R)

Ypr1p ee 97% (GST-Ypr1p, NADP+, G6P, G6PDH)

O

ee 87% (E. coli cell overexpressing Ypr1p) CO2Et

OH Gre2p

CO2Et

Dehydrogenase from Saccharomyces cerevisiae

(S)

ee 90% (GST-Gre2p, NADP+, G6P, G6PDH) ee 91% ( E.coli cell overexpressing Gre2p) OH CO2Et

R1 (b)

R2

OH

L-anti

CO2Et

R1

O

R2

CO2Et

R1

[9b]

OH

R2 Dehydrogenase from Saccharomyces cerevisiae

L-syn

CO2Et

R1

OH

R2

CO2Et

R1 R2

D-syn

D-anti

OH (c)

O R

R

CO2Et

L

[9b]

CO2Et

Dehydrogenase from Saccharomyces cerevisiae

OH R

CO2Et

D

Figure 8.13 Identification of appropriate biocatalysts from fusion protein libraries from baker’s yeast [9].

8.3.2.2 Overexpression Biocatalysts, discovered by screening, can be prepared in a large quantity by overexpression of the enzymes in transformed E. coli [10,11]. For example, the gene encoding (6R)-2,2,6-trimethyl-1-4-cyclohexanedione (levodione) reductase was cloned from the genomic DNA of the soil-isolated bacterium Corynebacterium aquaticum M-13 (Figure 8.14a) [10a]. The enzyme was sufficiently produced in recombinant E. coli cells and the enzyme purified from E. coli catalyzed stereo- and regioselective reduction of levodione. The enzyme was strongly activated by monovalent cations, such as Kþ, Naþ, and NH4. The chiral intermediate (S)-1-(20 -bromo-40 -fluorophenyl) ethanol was prepared by the enantioselective microbial reduction of 2-bromo-4-fluoroacetophenone [10b]. Organisms from genus Candida, Hansenula, Pichia, Rhodotorula , Saccharomyces, Sphingomonas, and baker’s yeast reduced the ketone to the corresponding alcohol in


8.3 Methodology for Stereochemical Control a)

O

Levodione reductase from C. aquaticum (Overexpressed in E. coli)

O (R)

O

[10a]

HO O

b)

j203

ADH from Pichia methanolica (Overexpressed in E. coli)

H 3C X

F

OH H3 C

Glucose, Glucose DH

(S) X

F

[10b]

2 kg scale

X = Br, (CH2)3CO2R c)

O Cl

CO2Et

ADH from Candida parapsilosis (Overexpressed in E. coli) 2-Propanol

OH Cl

CO2Et

(R) [10c] ee > 99% –1 36.6 gl

Figure 8.14 Overexpressed reductase in E. coli [10].

>90% yield and 99% ee. In an alternative approach, the enantioselective microbial reductions of methyl, ethyl, and tert-butyl 4-(20 -acetyl-50 -fluorophenyl) butanoates were demonstrated using strains of Candida and Pichia. Reaction yields of 40–53% and enantiomeric excesses of 90–99% were obtained for the corresponding (S)-hydroxyesters. The reductase, which catalyzed the enantioselective reduction of keto esters, was purified to homogeneity from cell extracts of Pichia methanolica SC 13825. It was cloned and expressed in E. coli with recombinant cultures used for the enantioselective reduction of keto methyl ester to the corresponding (S)-hydroxymethyl ester. On a preparative scale, a reaction yield of 98% and an enantiomeric excess of 99% were obtained (Figure 8.14b). The synthesis of ethyl (R)-4-chloro-3-hydroxy-butanoate ((R)-ECHB) from ethyl 4-chloroacetoacetate (ECAA) was studied using whole recombinant cells of E. coli expressing a secondary alcohol dehydrogenase of Candida parapsilosis [10c]. Using 2-propanol as an energy source to regenerate NADH, the yield of (R)-ECHB reached 36.6 g l1 (more than 99% ee, 95.2% conversion yield) without addition of NADH to the reaction mixture (Figure 8.14c). On the other hand, a novel carbonyl reductase (KLCR1) that reduced ECAA to (S)-ECHB was purified from Kluyveromyces lactis [10d]. KLCR1 catalyzed the NADPH-dependent reduction of ECAA enantioselectively, but not the oxidation of (S)-ECHB. 8.3.2.3 Coexpression of Genes for Carbonyl Reductase and Cofactor-Regenerating Enzymes A new bioreduction system for the production of chiral alcohol using an E. coli transformant coexpressing genes for carbonyl reductase and cofactor-regenerating enzymes has been investigated [11]. An NADPH-dependent carbonyl reductase (S1) isolated from C. magnoliae, which catalyzed the reduction of ethyl 4-chloro-3oxobutanoate (COBE) to ethyl (S)-4-chloro-3-hydroxybutanoate (CHBE), with a 100% ee was used to construct the system. The S1 gene comprises 849 bp and encodes a polypeptide of 30 420 Da, and the deduced amino acid sequence showed a high degree of similarity to those of the other members of the short-chain


j 8 Enzymatic Reduction Reaction

204

alcohol dehydrogenase superfamily [11b]. The E. coli cells expressing both the carbonyl reductase gene and the glucose dehydrogenase (GDH) gene from Bacillus megaterium were used as the catalyst for reduction of ethyl 4-chloro-3-oxobutanoate to ethyl (S)-4-chloro-3-hydroxybutanoate [11c]. In an organic-aqueous two-phase system, 450 g l1 of the product was obtained in 89% yield and 100% ee. The system was also applied for the reduction of analogous substrates affording D-alcohols in high enantiomeric excess (Figure 8.15) [11d]. 8.3.2.4 Modification of Biocatalysts: Directed Evolution Directed evolution of enzymes has been used to improve the reducing function of the enzymes. For example, this method was used to eliminate the cofactor requirement of B. stearothermophillus lactate dehydrogenase, which is activated in the presence of fructose 1,6-bisphosphate [12]. The activator is expensive and representative of the sort of cofactor complications that are undesirable in industrial processes. Three rounds of random mutagenesis and screening produced a mutant that is almost fully

Carbonyl reductase (S1) gene

S1 Plasmid

Gene of Candida magnoliae

S1

More effective !!! Glucose dehydrogenase (GDH) gene for cofactor-recycling S1 Plasmid

Overexpression

GDH E. coli Transformant cells for biocatalysis

Gene of Bacillus megaterium GDH O Organic phase (butyl acetate)

OH

X

CO2 Et

X

CO2 Et

(S)

X = Cl Yield 89% 2.70 M (450 glâ&#x20AC;&#x201C;1) ee 100%

S1

Aqueous phase E. coli

HO

NADPH

GDH

O

HO O HO OH D-Gluconolactone

X = Cl, Br, I, N3, OH, H ee> 99%

NADP+

HO HO HO

O OH D-Glucose

Figure 8.15 Coexpression of genes for carbonyl reductase and cofactor-regenerating enzymes [11c,d].

OH


8.3 Methodology for Stereochemical Control

j205

activated in the absence of fructose 1,6-bisphosphate. The Km of the enzyme without the cofactor was improved from 5 to 0.07 mM for the pyruvate (Figure 8.16). 8.3.3 Modification of Substrates

The enantioselectivity of a biocatalytic reduction can be controlled by modifying the substrate because the enantioselectivity of the reduction reaction is profoundly affected by the structure of the substrate[13]. For example, in the reduction of 4-chloro3-oxobutanoate by bakerÂ&#x2019;s yeast, the length of ester moiety controlled the stereochemical course of the reduction [13aâ&#x20AC;&#x201C;c]. When the ester moiety was smaller than a butyl group, (S)-alcohols were obtained, and when it was larger than a pentyl group, (R)alcohols were obtained, as shown in Figure 8.17a. After reduction, the ester moiety can be exchanged easily without racemization, so both enantiomers of an equivalent synthetic building block could be obtained using the same reaction system. Another example is the introduction of sulfur functionalities into the keto esters at the a- or a0 positions, which can be eliminated after the reduction. Methylthio1

O

OH OR

bsLDH

O

OR O

70-fold activation !!! Evoluted enzyme

Evolution Evolution Native enzyme (bsLDH)

Km pyr = 5 mM Extraction of gene

Step1 Mutation of gene

Point mutation !

STEP4 Collection of Mutant Gene

(R118C, Q203L, After three rounds N307S)

Evolution cycle

Step2 Expression of mutant enzymes

Figure 8.16 Elimination of the cofactor requirement by directed evolution [12].

Km pyr = 0.07 mM

Step3 Selection of mutant enzyme


j 8 Enzymatic Reduction Reaction

206

R

(a)

n = 5 –12 OH

Cl

CO2(CH2)nH

Cl

Baker’s yeast O CO2(CH2)nH

ee (R)

Baker’s yeast ee (S)

OH Cl

CO2(CH2)nH S

n

n=1–4 OH

(b)

CO2Me OH

SR CO2Me

Baker’s yeast

S

ee > 96%

O

Baker’s yeast ee 87%

CO2Me

O

SR CO2Me O PhO2S

CO2Me Baker’s yeast

OH

ee 98% CO2Me

R

OH PhO2S

CO2Me

Figure 8.17 Modification of the substrate to control the enantioselectivities: (a) effect of the length of the ester moiety1 and (b) effect of introducing sulfur functionalities [13d,e].

[12] and phenylsulfonyl1 [13e] groups were used to improve the enantioselectivities, as shown in Figure 8.17d,e. 8.3.4 Modification of Reaction Conditions 8.3.4.1 Acetone Treatment of the Cell A dried cell mass is often used as a biocatalyst for a reduction since it can be stored for a long time and used whenever needed, without cultivation. One convenient method of drying the cell mass is acetone dehydration. For example, dried cells of G.


8.3 Methodology for Stereochemical Control OH

Untreated whole cell

Ph

O

Dried cell of G. candidum (APG4) Ph

ee 28% (R) Catalyst Untreated whole cell Acetone dried cell (APG4) Acetone dried cell (APG4) Acetone dried cell (APG4) Acetone dried cell (APG4) Acetone dried cell (APG4) Acetone dried cell (APG4)

Coenzyme None None NAD + None NAD + NAD + + NADP

NAD+ or NADP+ 2-Propanol or cyclopentanol Additive

Yield (%)

None None None 2-Propanol 2-Propanol Cyclopentanol Cyclopentanol

52 0 1 8 89 97 86

j207

OH Ph

ee > 99% (S) ee (%) 28 (R) 71 (S) 98 (S) > 99 (S) > 99 (S) > 99 (S)

Figure 8.18 Acetone treatment of Geotrichum candidum for the improvement of enantioselectivity [14].

candidum IFO 4597 can be obtained easily by mixing the cells with cold acetone (20 C) followed by filtering and drying under reduced pressure. Cell drying not only aids the preservation of the cell but also contributes to the stereochemical control as shown in Figure 8.18. The reduction of acetophenone catalyzed by untreated, wet whole cell of G. candidum IFO 4597 resulted in poor enantioselectivity (28% ee(R)). When the form of the catalyst was changed from wet whole-cell to dried powdered-cell (APG4), no reduction was observed, which would indicate the loss of the necessary coenzyme(s) and/or coenzyme regeneration system(s) during the treatment of the cells with acetone. The addition of coenzyme NADþ did not have a significant effect on the yield. Addition of 2-propanol resulted in only a small increase in the yield, but a significant improvement in the enantioselectivity was observed. Surprisingly, the addition of both NADþ and 2-propanol profoundly enhanced both chemical yield and enantiomeric excess. Furthermore, the addition of NADH, NADPþ, or NADPHinstead of NADþ and the addition of cyclopentanol instead of 2-propanol also gave enantiomerically pure alcohol in high yield. The improvement in the enantioselectivity from 28% (R) to >99% (S) was due to the suppression of every enzyme that reduces the substrate, followed by the stimulation of an S-directing enzyme by the addition of the coenzyme and an excess amount of 2propanol, agents which push the equilibrium toward the reduction of the substrate. It was confirmed, by separating the enzymes in the powder, that many S- and R-directing enzymes do indeed exist in the dried cells. The addition of a coenzyme and cyclopentanol stimulates only an S enzyme because the specific S enzyme can oxidize cyclopentanol (concomitantly reducing NAD(P)þ), while other S or R enzymes cannot use cyclopentanol as effectively [14c]. This presents a very interesting case, where the experimental conditions of reduction with a cell having both Sand R-directing enzymes was modified and resulted in excellent S enantioselectivity. 8.3.4.2 Selective Inhibitor In case poor overall enantioselectivity is observed owing to the presence of two competing enzymes with different enantioselectivities, one of the most straightfor-


j 8 Enzymatic Reduction Reaction

208

S R

OH

O CO2R

Additive none

O

OH CO2R

CO2R Baker's yeast

Yield 61% ee 4% R = Et Yield 56% ee 96% R = Me

Additive O Cl

OEt Mg2+

Yield 63% ee 94% R = Et Yield 65% ee 99% R = Et

Figure 8.19 Improvement of the enantioselectivity by using an inhibitor of undesired enzymes [15a–c].

ward methods for improving the enantioselectivity is to use an inhibitor of the unnecessary enzyme(s). Ethyl chloroacetate, methyl vinyl ketone, allyl alcohol, allyl bromide, sulfur compounds, Mg2þ, Ca2þ, and so on, have been reported as inhibitors of enzymes in baker’s yeast [15]. For example, the low enantioselectivity in the yeast reduction of b-keto ester was improved by addition of methyl vinyl ketone, Mg2þ, or ethyl chloroacetate as described in Figure 8.19 [15a–c]. By enzymatic studies using purified enzymes from baker's yeast, the enzymes inhibited were identified, and the inhibition mechanism was reported to be noncompetitive [15k]. 8.3.4.3 Reaction Temperature Reaction temperature is one of the parameters affecting the enantioselectivity of a reaction [16]. For the oxidation of an alcohol, the values of kcat/Km were determined for the (R)- and (S)-stereodefining enantiomers; E is the ratio between them. From the transition state theory, the free energy difference at the transition state between (R) and (S) enantiomers can be calculated from E (Equation 2), and G is in turn the function of temperature (Equation 3). The racemic temperature (Tr) can be calculated as shown in (Equation 4). Using these equations, Tr for 2-butanol and 2-pentanol of the Thermoanaerobacter ethanolicus alcohol dehydrogenase were determined to be 26 and 77  C, respectively.

E ¼ ðkcat =K m ÞR =ðkcat =K m ÞS

ð1Þ

From transition state theory, RT lnðEÞ ¼ DDGz

ð2Þ

DDGz ¼ DDHz TDDSz

ð3Þ

When DDGz ¼ 0;

ð4Þ

T r ¼ DDHz =DDSz

Since the transition state for alcohol oxidation and ketone reduction must be identical, the product distribution (under kinetic control) for reducing 2-butanone and 2-pentanone is also predictable. Thus, one would expect to isolate (R)-2-butanol if the temperature of the reaction was above 26  C. On the contrary, if the temperature is less than 26  C, (S)-2-butanol should result; in fact, the reduction of


8.4 Medium Engineering

j209

2-butanone and 2-pentanone at 37  C resulted in 28% ee (R)- and 44% ee (S) alcohol, respectively, as expected [16a]. 8.4 Medium Engineering

Biocatalytic reduction has been performed in nonaqueous solvents to improve the efficiency of the reaction. This section explains the use of organic solvent, supercritical fluids, and ionic liquid. 8.4.1 Organic Solvent 8.4.1.1 Soluble Organic Solvent Soluble organic solvents have often been used as cosolvents to solubilize miscible organic substrates. Since organic compounds including solvents are possibly incorporated inside of the enzyme, they may affect the stereoselectivity of enzymatic reactions. For example, dimethyl sulfoxide (DMSO) (10%) enhance not only chemical yield but also enantioselectivity of yeast reduction. Thus, the poor yield of 23% with 80% ee was increased to 65% yield with >99% ee (Figure 8.20) [17]. 8.4.1.2 Aqueous-organic Two-phase Reaction Aqueous-organic two-phase reaction has been widely performed [18]. One of the purposes of using two-phase reaction system is to control the substrate concentration in aqueous phase where the biocatalysts exist. Hydrophobic substrate and products dissolve easily in the organic phase, so that the concentration in the aqueous phase decreases. The merits of controlling and decreasing the substrate concentration in the aqueous phase are as follows:

1. Substrate and product inhibition can be prevented. 2. When whole cell containing plural enzymes with opposite selectivities and different Km (Michaelis–Menten constant) values are used, problems of low selectivities occur. If the substrate concentration is decreased, one of the enzymes with low Km value catalyzes the reaction so that the selectivity can be improved. 3. The decomposition of unstable substrate/product by aqueous buffer can be prevented by dissolving the substrate and product in the organic phase. Therefore, organic solvents have been widely used for biocatalytic reductions. An interesting example for stereochemical control by using organic solvents for OH

O Cl

Baker's yeast

Figure 8.20 Stereochemicalcontrolbysolubleorganicsolvent[17].

Cl

Yield

ee

Without DMSO

23%

80%

+ 10% DMSO

65% > 99%


j 8 Enzymatic Reduction Reaction

210

the reduction is as follows. G. candidum IFO 4597 catalyzes the reduction of ketones although the reduction of acetophenone, by the resting cell in water, afforded only (R)-alcohol with 28% ee in a 52% yield [18a]. The cell was immobilized on a water-absorbing polymer to spread it on the large surface of the polymer to use it in hexane (Figure 8.21) [18a]. The reduction in hexane by the immobilized cell

Hexane layer

Hexane layer

OH

O

(S)

Reaction in cell

O

OH

NAD(P)H

NAD(P)+

O

OH C4H9

C4H9 Excess amount

Cell

Aqueous layer inside the polymer

Solubility in aqueous layer 1-phenylethanol > acetophenone

Water absorbing polymer O

Geotrichum candidum R

Substrate

Hexane

OH R Yield (%)

52a Acetophenone 2b Acetophenone 73 Acetophenone 81 2-Acetylfuran 99 o-Chloroacetophenone 88 m-Chloroacetophenone 41 p-Chloroacetophenone 59 o-Methylacetophenone 56 m-Methylacetophenone 40 p-Methylacetophenone 97 1-Phenyl-2-propanone 88 Benzylacetone a Reaction in water without immobilization and without 2-hexanol. b Reaction without 2-hexanol.

Figure 8.21 Reduction of acetophenone in hexane by the immobilizedrestingcellofGeotrichumcandidumand2-hexanol[18a].

ee (%) 28 (R)a â&#x20AC;&#x201D; >99 (S) 99 (S) 99 (S) >99 (S) 92 (S) >99 (S) >99 (S) >99 (S) >99 (S) >99 (S)


8.4 Medium Engineering

barely proceeded because of the unfavorable partition of ketone and alcohol between the organic and aqueous layers for the reduction (Figure 4.21). However, when an alcohol such as 2-hexanol was added to the reaction system, the reduction proceeded smoothly to afford (S)-alcohol in excellent enantiomeric excess. 2-Hexanol as well as other 2-alkanols from 2-propanol to 2-octanol worked when they were used in large excess and reached to saturation on the yield at 10â&#x20AC;&#x201C;15 mol equiv. The mechanism of enantioselectivity improvement by this method can be explained as follows: (i) enantiomeric excess is poor for reduction in water because both the S enzyme(s) and R enzyme(s) catalyze the reduction; (ii) when hexane is used, both enzymes are inhibited; and (iii) when 2-alkanol was added, only an S enzyme was reactivated and catalyzed the reduction (2-alkanols are selective to the S enzyme). Various aryl methyl ketones were also reduced smoothly by the same system using 12 equivalents of 2-hexanol to produce the corresponding (S)-alcohols in excellent enantiomeric excess (Figure 8.21). The reduction of ortho-substituted acetophenones gives a relatively better yield than that of para-substituted acetophenones, probably because the reverse reaction, oxidation, does not proceed for the ortho-substituted substrates. The better a substrate is for oxidation, the worse it is for reduction of the substituted acetophenone derivatives. In another example, methyl 7-ketolithocholate (Me-7KLCA) was reduced with Eubacterium aerofaciens JCM 7790 in anaerobic interface bioreactor and dihexyl ether was used as the solvent [18b,c]. The microbe system reduced 12 g l1 of Me-7KLCA to methylursodeoxycholate (Me-UDCA) in more than 50% yield. The product is the precursor of ursodeoxycholic acid, which is used as a cholesterol gallstone-dissolving agent (Figure 8.22). The enantioselective reduction of alkyl 3-oxobutanoates by carbonyl reductase (Sl) from C. magnoliae was also performed in organic-aqueous two-phase reaction system (Figure 8.15) [11c,d]. 8.4.2 Use of Hydrophobic Resin, XAD

Instead of using organic solvents, hydrophobic resin, AmberliteTM XAD, has been used to control the substrate concentration [19]. When XAD is added to the reaction mixture, substrate and products are adsorbed to the hydrophobic resin, XAD, since Me-7KLCA

Me-UDCA CO2Me

HO

O

CO2Me

HO

OH

Organic phase

Microbial film

Glucose Mannitol

Nutrients Water

Hydrophilic carrier

Figure 8.22 Reduction of ketone with anaerobic interface bioreactor [18b].

j211


j 8 Enzymatic Reduction Reaction

212

G. candidum

O R Substrate

1

R

2

Hydrophobic polymer XAD-7

Without XAD-7 Yield/% ee/%

O C6H13

36

OH

With XAD-7

R

82

> 99 (S) O

86 (R)

82

58

18 (R)

76

C3H7

> 99 (S) Ph

O

O

Yield/% ee/%

O

O 19

R2

Substrate Without XAD-7

Yield/% ee/%

60 (S)

1

97 (S)

OPh

With XAD-7 Yield/%

ee/%

14

33 (R)

25

86 (S)

35

94 (S)

32

> 99 (S)

99

3.6 (S)

86

99 (S)

Figure 8.23 Reduction of ketones with Geotrichum candidum in the presence of hydrophobic polymer XAD [19a].

the substrate and product are usually hydrophobic. Therefore, the effective concentration of the substrate around the enzyme is decreased. Recently, XAD was used as material to control the stereochemical course of microbial reductions [19]. In the presence of XAD, simple aliphatic and aromatic ketones were reduced to the corresponding (S)-alcohols in excellent enantioselectivity while low enantioselectivities were observed in the absence of the polymer (Figure 8.23). In the reduction of benzoyloxypropanone, the hydrophobic polymer XAD-7 was used to prevent product inhibition and increase substrate concentration [19b]. Thus, the reduction proceeded in 70 g l1 substrate concentration and afforded 87% (12.4 g) of (S)-1-benzoyloxy-2-propanol in >99% ee (Figure 8.24). The butanone derivative could be reduced with the same method and afforded (S)-alcohol in 72% yield and >99% ee, but the pentanone derivative could not be reduced. An acyclic enone, 2-ethyl-1-phenylprop-2-en-1-one, was reduced with the yeast Pichia stipitis CCT 2617 [19c]. The reduction proceeded chemo- and enantioselectively to afford (S)-2-ethyl-1-phenylprop-2-en-1-ol (65% yield, >99% ee). XAD-7 was used to decrease and control the concentration of both the substrate and the product (Figure 8.25). O

Baker's yeast OBz

XAD-7, ethanol, O 2

OH OBz

(S) Yield 87% ee > 99%

Figure 8.24 Reduction of ketone with bakerÂ&#x2019;s yeast in the presence of hydrophobic polymer XAD [16b].

O

OH Pichia stipitis XAD-7

Figure 8.25 Reduction of ketone with Pichia stipitis in the presence of hydrophobic polymer XAD [19c].

(S) Yield 65% ee > 99%


8.4 Medium Engineering

8.4.3 Supercritical Carbon Dioxide

Reduction using alcohol dehydrogenases is usually conducted in aqueous media. The difficulties encountered in such reactions are the extraction of products that dissolve in aqueous media at low concentration, and an organic solvent is usually 0 used. However, by using scCO2 this becomes unnecessary because CO2 transforms into a gas as the pressure decreases. Therefore, reduction of carbonyl group using alcohol dehydrogenases was conducted in scCO2. For example, immobilized cells of G. candidum were used for the reduction in scCO2 [20a,b]. Since the whole resting cells were used, the addition of expensive coenzymes was avoided. Moreover, the solubility of the coenzymes in scCO2 was not required to be considered. At first, the reduction of o-fluoroacetophenone in scCO2 at 10 MPa was conducted using 2-propanol as a reductant (hydrogen donor) (Figure 8.26), which afforded (S)-1(o-fluorophenyl)ethanol in 81% yield (determined by gas chromatography) after 12 hours [20a]. The time course of the reaction shows that the yield increased with the reaction time, which proved that the alcohol dehydrogenase catalyzed the reduction in the supercritical condition. The substrate specificity was investigated, and as listed in Figure 8.26b, the enzymatic reduction in scCO2 proceeded for various ketones. Acetophenone, acetophenone derivatives, benzyl acetone, and cyclohexanone were used as substrates, and it was found that all of them were reduced by the alcohol dehydrogenase in scCO2 with 2-propanol. The effects of fluorine substitution at the ortho, para, and a-positions of acetophenone were obvious. Compared with the unsubstituted analog, substitution at the ortho position increased the yield, whereas substitution at the para position decreased the yield. Regarding enantioselectivity, very high values (>99% ee) were obtained for the reduction with the majority of the substrates tested, while slightly lower enantioselectivities (96, 97% ee) were observed for a few of them. The enantioselectivities obtained in this system are superior or at least equal to those for most other biocatalytic and chemical systems. The immobilized resting cell of G. candidum was also used as a catalyst for the reduction of o-fluoroacetophenone and cyclohexanone in a semicontinuous flow process using scCO2 [20b]. The apparatus is shown in Figure 8.26c. With flow reactors, the addition of a substrate to the column with a catalyst yielded the product and CO2, which is a gas at ambient pressure, whereas, with the batch reactor, separation of the product from the biocatalyst was necessary after depressurization. Therefore, the flow type was superior to the batch type for achieving virtually no solvent reaction. Moreover, the size of the reactors using the flow process to generate an amount of product comparable with the corresponding batch reactors is smaller, which is particularly attractive for a supercritical fluid system. This reaction using a semicontinuous flow process also resulted in a higher space-time yield than that of the corresponding batch process. An isolated enzyme is also used for the reduction in scCO2 [20c]. HLADH was used for reduction of butyraldehyde as shown in Figure 8.27. In this case, addition of

j213


j 8 Enzymatic Reduction Reaction

214

Figure 8.26 Asymmetric reduction of ketones in CO2 by Geotrichum candidum immobilized whole cell [20]. (a) Time course for the reduction of o-fluoroacetophenone; (b) substrate specificity; (c) apparatus for G. candidum窶田atalyzed reduction with semiflow process using scCO2.


8.4 Medium Engineering

(a)

Horse liver alcohol dehydrogenase (HLADH)

O

OH H

H Ethanol, CO2 (17.9 MPa, 21 oC), 45 h

250 mM (b)

NH 2

O NH2

N O

N O O O P O P O OH O-

O N NH(CH2 )6 NH N

N

CF O CF 2 CF F CF3 CF3 n

O Fluorinated polymer attached

OH

OH

Yield (mM)

(c)

OH

OH

50 40

Coenzymes 1: No added coenzyme 2: NAD+ 3: Fluorinated NAD+

30 20 10 0

4: HLADH lyophilized with NAD+ 1

2 3 Coenzymes

4

Figure 8.27 Reduction of aldehyde in scCO2 by an isolated enzyme, horse liver alcohol dehydrogenase (HLADH) [20c] (a) Reaction scheme; (b) fluorinated coenzyme soluble in CO2; and (c) effect of coenzyme on the reaction.

coenzyme is necessary, and the sample with a soluble coenzyme, fluorinated coenzyme (Figure 8.27b), demonstrated the highest activity (Figure 8.27c). The sample with enzyme alone and the sample with added NADþ showed similar rate of butanol production. NADþ added separately from the enzyme did not change native HLADH activity because the enzyme and NADþ remained separate during the reaction. The HLADH lyophilized with NADþ produced very little butanol owing to mass transfer limitations. 8.4.4 Ionic Liquid

Ionic liquid [bmim]PF6 can be used as a solvent in yeast reduction [21]. The reduction of ketones with immobilized baker’s yeast (alginate) in a 100 : 10 : 2 [bmim]PF6 ionic liquid: water: MeOH mix affords chiral alcohols (Figure 8.28).

j215


j 8 Enzymatic Reduction Reaction

216

OH (S) Yield 40%, ee 79% O R1

OH

Baker's yeast (immobilized with alginate) R2 Me

Bu

(S) Yield 22%, ee 95%

OH R1

N + N

O

R2

PF6–

OH OH

CO2Et (S) Yield 70%, ee 95% CO2Et

[bmim]PF6 : H2O: MeOH = 100 : 10 : 2 OH CO2Et

(S) Yield 75%, ee 84% (S) Yield 60%, ee 76%

Figure 8.28 Use of ionic liquid in yeast reduction [21].

8.5 Synthetic Applications 8.5.1 Reduction of Aldehyde

Many aldehyde reductases transform both aldehydes and ketones. For example, phenylacetaldehyde reductase (PAR) from a styrene-assimilating Corynebacterium strain, ST-10, reduces hexyl aldehyde and phenylacetaldehyde [22a]. Other aldehyde reductases such as one from Sporobolomyces salmonicolor also reduce aldehydes as well as ketones [22b]. Organometallic aldehydes can be reduced enantioselectively with dehydrogenases. For example, optically active organometallic compounds having planar chiralities were obtained by biocatalytic reduction of racemic aldehydes with yeast [22c,d] or HLADH [22e] as shown in Figure 8.29. 8.5.2 Reduction of Ketone

Simple aliphatic ketones as well as aromatic ketones can be reduced with very high enantioselectivity by using biocatalysts [14]. For example, aliphatic ketones such as 2-pentanone, 2-butanone, 3-hexanone, and so on, were reduced with excellent enantioselectivity to the corresponding (S)-alcohols by using the dried cells of G. candidum (Figure 8.30) [14e]. The dried-cell G. candidum system can distinguish between two alkyl groups with a difference of a single methylene unit. The dried-cell G. candidum system is also applied for the reduction of aromatic ketones. Trifluoromethyl ketones were also reduced by the same system and afford (S)-alcohols in excellent enantiomeric excess [14h]. The substrate specificity of the system is shown (Figure 8.30) [14].


8.5 Synthetic Applications

O

O

Yeast

CH2OH Fe(CO) 3

H

Fe(CO) 3

j217

+ Fe(CO) 3

Yield 53% ee 78% (S)

Yield 32% ee > 99% (R)

CH2OH

CHO

H

CHO

HLADH +

NAD + EtOH Cr(CO) 3

Cr(CO) 3

Cr(CO) 3 Yield 33% ee 91%(S)

Yield 51% ee 81%(R)

Figure 8.29 Reduction of organometallic aldehydes to produce alcohols with planar chiralities [22c,e].

Resting cell of G. candidum, as well as dried cell, has been shown to be an effective catalyst for the asymmetric reduction. Both enantiomers of secondary alcohols were prepared by reduction of the corresponding ketones with a single microbe [23]. Reduction of aromatic ketones with G. candidum IFO 5767 afforded the corresponding (S)-alcohols in an excellent enantioselectivity when amberlite XAD-7, a hydrophobic polymer, was added to the reaction system, and the reduction with the same microbe afforded (R)-alcohols, also in an excellent enantioselectivity, when the reaction was conducted under aerobic conditions (Figure 8.31). O

OH R

H3 C

CF3

X3C R Dried cell of G. candidum (APG4) 2-propanol or cyclopentanol

F3 C

R

Product OH

OH

OH

OH

OH

OH Cl

94% ee

>99% ee

OH

98% ee

OH Ph

> 99% ee

OH

OH F3C > 99% ee

> 99% ee

OH F3C

Ph

98% ee

OH CO2Et

> 99% ee

OH

OH F3C Cl

Ph

> 99% ee

> 99% ee

F3C

Figure 8.30 Reduction of ketones by the dried cell of G. candidum, NAD(P)Ăž, and secondary alcohol [14].

OH MeO

Ar

OH

> 99% ee

>99% ee

99% ee

OH CO2Et

Ph > 99% ee

99% ee

S

>99% ee

F3C 98% ee

Ph


j 8 Enzymatic Reduction Reaction

218

OH

O

XAD-7

R

OH

Aerobic conditions R

R G. candidum IFO5767

(S)

(R)

92–> 99% ee

85–> 99% ee

96–> 99% yield

61–> 99% yield

O

O

O

O

O

O

O

N N

Br

Figure 8.31 Synthesis of both enantiomers of secondary alcohols with one kind of microbe [23].

Baker’s yeast has been widely used for the reduction of ketones. The substrate specificity and enantioselectivity of the carbonyl reductase from baker’s yeast, which is known to catalyze the reduction of b-keto ester to L-hydroxyester (L2-enzyme) [15], was investigated, and the enzyme was found to reduce chloro-, acetoxy ketones with high enantioselectivity (Figure 8.32) [24a]. The reduction of hydroxy or acetoxy ketones by baker’s yeast shows an interesting stereoselectivity. For the reduction of acetylbenzofuran derivatives with baker’s yeast, the methyl ketones afforded (S)-alcohol in 20–68% ee. The hydroxyl derivatives afforded (S)-alcohol in 87–93% ee, and the acetoxy derivatives gave (R)-alcohols in 84–91% ee (Figure 8.33) [24b]. Reductase from baker's yeast O R

OH Cl

(R), ee > 99%

1

(R), ee 88%

OH

OH

OH

Ph Cl

R2

R

glucose-6P glucose-6PDH

OH Cl

OH

NADP+ 2

R1

OAc

Ph (R), ee >99%

(R), ee 98%

Ph

OAc

(R), ee 96%

Figure 8.32 Reduction of ketones with reductase from baker’s yeast [24].

OH X

O

HO

O O

Baker's yeast

R

R X = OH or OAc R = H, 5-Br, 5-NO2, 7-MeO

(S) X = OH: ee 87–93% OH AcO

OH O

HO R

Figure 8.33 Reduction of hydroxy and acetoxy ketones [24b].

O

(R) X = OAc: ee 84–91%

R


8.5 Synthetic Applications

OH

O

OH

R1

R1 R2

R1 R2

(S) NADH CPCR

R2

(R)

NAD+ R1 = CH3, R2 = Ar : ee > 99% (S)

CO2 FDH

HCO2H

R1 = CH3, R2 = SiMe3 : ee > 99% (S) R1 = CH3, R2 = H : ee > 60% (R) R1 = C3H7, R2 = H : ee > 99% (S)

Figure 8.34 Reduction of acetylenic ketones [25].

For reduction of acetylenic ketones, two oxidoreductases were used [25]. Lactobacillus brevis alcohol dehydrogenase (LBADH) gave the (R)-alcohols and Candida parapsilosis carbonyl reductase (CPCR) afforded the (S)-isomer, both in good yield and excellent enantioselectivity. By changing the steric demand of the substituents, the enantiomeric excess values can be adjusted and even the configurations of the products can be altered (Figure 8.34). Process with PAR produced by styrene-assimilating Corynebacterium strain ST-10 is one of the best asymmetric reductions ever reported to synthesize chiral alcohols [26]. This enzyme with a broad substrate range reduced various prochiral aromatic ketones and b-keto esters to yield optically active secondary alcohols with an enantiomeric purity of more than 98% ee. The E. coli recombinant cells, which expressed the PAR gene, could efficiently produce important pharmaceutical intermediates: (R)-2-chloro-1-(3-chlorophenyl)ethanol (28 mg ml1) from m-chlorophenacyl chloride, ethyl (R)-4-chloro-3-hydroxy butanoate) (28 mg ml1) from ethyl 4-chloro-3oxobutanoate, and (S)-N-tert-butoxycarbonyl(Boc)-3-pyrrolidinol from N-Boc-3pyrrolidinone (51 mg ml1), with more than 86% yields. The high yields were due to the fact that PAR could concomitantly reproduce NADH in the presence of 3–7% (v/ v) 2-propanol in the reaction mixture (Figure 8.35). Phenylacetaldehyde reductase from Corynebacterium strain ST-10 (Escherichia coli recombinant cells)

O R1

R2

OH R1

2-Propanol

OH

R2 OH

OH

Cl

Cl

OMe

(R), ee > 99%

OH

OH

OH CO2Et

(S), ee > 99%

Cl

OH CO2Et

(S), ee > 99%

Cl (S), ee > 99%

OMe (S), ee > 99%

(S), ee > 99%

Cl

OMe

Br

HO CO2Et

(S), ee 98.4%

Figure 8.35 Reduction of ketones with phenylacetaldehyde reductase from Corynebacterium strain ST-10 [26].

N

CO2But

(S), ee > 99%

j219


j 8 Enzymatic Reduction Reaction

220

(a)

Rhodococcus ruber

O

OH

2-Propanol

R

(S) > 99% ee

R

[27a]

R = alkyl, ph O OH

Rhodococcus ruber

(b)

(S) > 99% ee

2-Propanol

OH

OH

Rapsbery ketone

[27b]

Figure 8.36 Reduction of ketones by Rhodococcus ruber [27].

The biocatalytic process using Rhodococcus ruber DSM 44541 is fully developed. Reduction of ketones with the lyophilized cells using 2-propanol as hydrogen donor afforded (S)-alcohol in high enantiomeric excess (Figure 8.36a) [27a]. Raspberry ketone was also reduced with the biocatalyst and gave (S)-rhododendrol in 87% yield with >99% ee (Figure 8.36b) [27b]. Vegetables could be used as biocatalysts for the reduction of ketones as shown in Figure 8.37 [28]. The reaction of racemic-2-methylcyclohexanone with fresh carrot root [28a] and various vegetables [28b] gave a mixture of (1S,2R)- and (1S,2S)-2-methylcyclohexanol in high enantiomeric excess (Figure 8.37a). The reaction of 2-hydroxycyclohexanone afforded a 1: 2 mixture of (1S,2R)- and (1S,2S)-1,2-cyclohexanediol with >95% ee [28]. For the reduction of 4-phenyl-2-oxo-but-3-enoates, the corresponding (R)-alcohols were obtained in 92–99% ee (Figure 8.37b) [28c]. Prochiral ketones such as aromatic ketones [28d–f] and b-keto esters [28d] were reduced to the corresponding optically active alcohols by carrot and other vegetables (Figure 8.37c) [28d]. O

OH

(a)

(b)

R

Carrot root

O CO2R

X

OH R

R

R = CH3, OH ee > 95%

OH

Plant cell culture of carrot

CO2R

X (R)

(c)

OH

X =H, o-Cl, p-Cl, p-Me R = Me, Et [28c] ee = 92–99%

OH CO2Et

O R1

[28a,b]

R2

OH

Vegetable R1

[28d]

[28d]

R2

OH OH N3

CO2Et

[28d] Figure 8.37 Reduction of ketones by vegetables [28].

X [28d-f]


8.5 Synthetic Applications

8.5.3 Dynamic Kinetic Resolution and Deracemization 8.5.3.1 Dynamic Kinetic Resolution Dynamic kinetic resolution of racemic ketones proceeds through asymmetric reduction when the substrate does racemize and the product does not under the applied experimental conditions. Dynamic kinetic resolution of a-alkyl b-keto ester has been performed through enzymatic reduction. One isomer, out of the four possible products for the unselective reduction (Figure 8.38), can be selectively synthesized using biocatalyst, and by changing the biocatalyst or conditions, all of the isomers can be selectively synthesized [29]. Dynamic kinetic resolution of a-alkyl-b-keto ester was conducted successfully using biocatalysts. For example, baker's yeast gave selectively syn(2R, 3S)-product [29a] and the selectivity was enhanced by using selective inhibitor [29b] or heat treatment of the yeast [29c]. Organic solvent was used for stereochemical control of G. candidum [29d]. Plant cell cultures were used for reduction of 2-methyl-3-oxobutanoate and afforded antialcohol with Marchantia [29e,f ] and syn-isomer with Glycine max [29f ]. Extensive screening methodology was used to find the suitable microorganism. As a result, K. pneumoniae IFO 3319 out of 450 bacterial strains was found to give the corresponding (2R, 3S)-hydroxy ester with 99% de and >99% ee in kilogram scale quantitatively [29g]. A series of 2-(4-chlorophenoxy)-3-oxoalkanoate were reduced by baker’s yeast, and Kluyveromyces marxianus. Yeast reduction of ethyl 2-(4-chlorophenoxy)-3-oxo-3phenylpropanoate (R ¼ Ph) afforded enantiomerically pure ethyl (2R, 3S)-2(4-chlorophenoxy)-3-hydroxy-3-phenylpropanoate out of the four possible stereoisomers in >99% de [29h]. Although baker’s yeast reduction of butanoate (R ¼ CH3) was not selective (92% de), the use of K. marxianus afforded (2R, 3S)-isomer selectively [29i]. The products are intermediates for potential peroxisome proliferator-activated receptor isoform a-agonists (Figure 8.39a). b-Diketones can also be reduced diastereoselectively. Thus, a recombinant alcohol dehydrogenase from L. brevis (recLBADH) overexpressed in E. coli was used for

O

O

OH

O

OEt Fast in aqueous buffer O

O

O

OEt Anti (2R, 3R) OH

OEt

OH

O

OEt Syn (2R, 3S) OH

OEt Syn (2S, 3R)

O OEt

Anti (2S, 3S)

Figure 8.38 Possible products for the reduction of a-methyl b-keto ester.

j221


j 8 Enzymatic Reduction Reaction

222

(a) OH

O CO2Et

Kluyvromyces marxianus

O

OH CO2Et

R O

O

R = Ph

R = CH3 (2R,3S) de > 99%

CO2Et

Ph

Baker's yeast

[29h,i].

Cl

(2R,3S) de > 99%

Cl

(b)

Cl

[29j] O

O CO2But

OH

recLBADH NADP+

O

Yiled 66% CO2But ee 99.2% de 94% (syn)

2-Propanol

OMe

OMe

(c)

S O (S)-ketone

N H

[29k]

S Baker's yeast

(2S, 3S)-Diltiazem

OH N H O Yield 80%, ee > 99% (2S, 3S)

O

OMe S O N (R)-ketone H

O

[29l] Ar

d) O Ar

OH

Curvularia lunata CN

CN

Ar R (R, R)

R

C2H5 Ph C3H7 Ph C4H9 Ph Me2CHCH2 Ph C2H5 m-CH3Ph C2H5 p-CH3Ph C2H5 2-Thienyl C2H5 2-Furyl

Yield de ee (%) (%) (%) 69 38 13 14 42 64 63 59

96 98 86 94 78 92 94 72

98 98 86 97 70 83 93 87

Figure 8.39 Dynamic kinetic resolution.

the reduction of tert-butyl 4-methyl-3,5-dioxohexanoate. The syn-hydroxyester was obtained in 66% yield with 94% de and 99.2% ee (Figure 8.39b) [29j]. Another example of dynamic kinetic resolution is the reduction of a sulfursubstituted ketone. Thus, yeast reduction of (R/S)-2-(4-methoxyphenyl)-1, 5-benzothiazepin-3,4(2H, 5H)-dione gave only (2S, 3S)-alcohol as a product out of four possible isomers as shown in Figure 8.39c [29k]. Only (S)-ketone was recognized by the enzyme as a substrate and reduction of the ketone proceeded


8.5 Synthetic Applications

j223

enantioselectively. The resulting product was used for the synthesis of (2S, 3S)Diltiazem, a coronary vasodilator. White-rot fungus has been used as a biocatalyst for reduction and alkylation. The reaction of aromatic b-keto nitriles with the white-rot fungus Curvularia lunata CECT 2130 in the presence of alcohols afforded alkylation–reduction reaction [29l]. Alcohols such as ethanol, propanol, butanol, and isobutanol could be used (Figure 8.39d). The dynamic resolution of an aldehyde is shown in Figure 8.40. The racemization of starting aldehyde and enantioselective reduction of carbonyl group by baker’s yeast resulted in the formation of chiral carbon centers. The enantiomeric excess value of the product was improved from 19 to 90% by changing the ester moiety from the isopropyl group to the neopentyl group [30a]. Other biocatalysts were also used to perform the dynamic kinetic resolution through reduction. For example, Thermoanaerobium brockii reduced the aldehyde with a moderate enantioselectivity [30b,c], and Candida humicola was found, as a result of screening from 107 microorganisms, to give the (R)-alcohol with 98.2% ee when ester group was methyl [30d]. 8.5.3.2 Deracemization through Oxidation and Reduction Deracemization reaction, which converts the racemic compounds into chiral form in one step in one pot without changing their chemical structures can be performed using microorganism containing several different stereochemical enzymes (Figure 8.41) [31]. For example, for deracemization of 1,2-pentandiol (Figure 8.41a), the (R)-specific NADH-enzyme in C. parapsilosis was reversible and (S)-specific NADPH-enzyme in the same microorganisms was irreversible [31a]. Therefore, whole-cell reaction of racemic 1, 2-pentandiol gave (S)-diol in 93% yield and 100% ee. For the deracemization of phenylethanol derivatives using G. candidum under aerobic conditions (Figure 8.41b), the (S)-specific enzyme was reversible and (R) enzyme was irreversible, so (R)-alcohol accumulated when the cell and racemic alcohols were mixed [31b,c]. Para-substituted phenylethanol derivatives gave better results than meta-substituted derivatives. Sphingomonas was used for

(e)

O CO2R

H

(R)-aldehyde

CO2R

(R)-alcohol

Biocatalysts Baker's yeast Baker's yeast Baker's yeast Baker's yeast

R

Yield (%)

70–80 –CH2CH3 49 –CH(CH 3)2 –CH 2CH(CH3)2 84 78 –CH 2C(CH3)3

Thermoanaerobium –CH 2CH3 brockii

O H

HO

ee (%) Reference

60–65 19 64 90

[30b] [30a] [30a] [30a]

50–80

72

[30c]

— —

98.2 73.6

[30d] [30d]

CO2R Candida humicola Candida humicola

(S)-aldehyde

Figure 8.40 Dynamic kinetic resolution of aldehydes.

–CH 3 –CH 2CH3


j 8 Enzymatic Reduction Reaction

224 (a)

NADH dependent R-specific alcohol dehydrogenase

[31a]

NADPH dependent S-specific reductase

OH

OH

OH (R)

Yield 93% ee 100% (S)

OH

O

OH (S)

Candida parapsilosis

[31b,c] (b) OH

X

X

(S)

X

OH

O

H F Cl Br Me MeO

X (R)

Geotrichum candidum Aerobic conditions

ee (%) Configuration

Yield of alcohol (%) 96 96 93 99 99 65

99 100 98 100 96 100

(c) OH

Sphingomonas paucimobilis

S N

[31c]

OH Yield 85% ee 97% (R)

S N OH

OH

(d)

R R R R R R

[31d] Yield 93% ee 100%

* N OH

N Catharanthus roseus plant cell culture OH Yield 100% ee 87% (R)

N N

Figure 8.41 Deracemization via oxidation–reduction.

deracemization of a thiazole derivative (Figure 8.41c) [31d]. Plant cell culture could be used for deracemization of aromatic alcohols (Figure 8.41d) [31e]. 8.6 Conclusions

Asymmetric reduction of carbonyl groups by biocatalysts has been a useful method for preparation of valuable compounds as shown in this review. Increasingly, bioengineering technology has been applied to alcohol dehydrogenases and reductases as has been applied to hydrolytic enzymes to improve enzyme stability, reactivity and enantioselectivity and to extend substrate specificity. Novel enzymes for reductions created by this technique will be available in large quantities and varieties within a next few years. In the near future, a lot of useful biocatalysts for reduction will be on the market, and expanding number of chemists can use enzymes for reduction more freely than at present owing to the simplification of experimental procedures. Furthermore, the biocatalysts will be even more important with the shift of the raw materials from oil to biomass. Since biomass is a mixture of various multifunctional compounds, chemo-, regio-, and enantioselective catalysts will be


References

indispensable, and biocatalysts will play an important role. Moreover, biocatalysts will perform better using natural substrates from biomass than using manmade substrates.

References 1 (a) Nakamura, K., Yamanaka, R., Matsuda, T. and Harada, T. (2003) Tetrahedron: Asymmetry, 14, 2659–2681. (b) van der Donk, W.A. and Zhao, H. (2003) Current Opinion in Biotechnology, 14, 421–426. (c) Nakamura, K. and Matsuda, T. (2002) in Enzyme Catalysis in Organic Synthesis A comprehensive Handbook (eds K. Drauz and H. Waldmann), Wiley-VCH Verlag GmbH, Weinheim, pp. 991–1047. (d) Santaniello, E., Ferraboschi, P. and Manzocchi, A. ,(2000) in Enzymes in Action (eds B. Zwanenburg, P. Mikolajczyk and M. Kielbasinski), Kluwer Academic Publishers, Dordrecht, pp. 95–115. 2 (a) Matsuda, T., Harada, T., Nakajima, N. and Nakamura, K. (2000) Tetrahedron Letters, 41, 4135–4138. (b) Matsuda, T., Harada, T., Nakajima, N., Itoh, T. and Nakamura, K. (2000) Journal of Organic Chemistry, 65, 157–163. (c) Nakamura, K. and Matsuda, T. (1998) Journal of Organic Chemistry, 63, 8957–8964. 3 Iyer, R.B. and Bachas, L.G. (2004) Journal of Molecular Catalysis B: Enzymatic, 28, 1–5. 4 (a) Gr€ oger, H., Hummel, W., Rollmann, C., Chamouleau, F., H€ usken, H., Werner, H., Wunderlich, C., Abokitse, K., Drauz, K. and Buchholz, S. (2004) Tetrahedron, 60, 633–640. (b) Shorrock, V.J., Chartrain, M. and Woodley, J.M. (2004) Tetrahedron, 60, 781–788. 5 (a) Mertens, R., Greiner, L., van den Ban, E.C.D., Haaker, H.B.C.M. and Liese, A. (2003) Journal of Molecular Catalysis B: Enzymatic, 24–25, 39–52. (b) Williams, P., Micklefield, J., Gardiner, J.M. and Stephens, G. (2004) Tetrahedron, 60, 753–758. (c) Wagenknecht, P.S., Penney, J.M. and Hembre, R.T. (2003) Organometallics, 22, 1180–1182.

6 (a) Nakamura, K., Yamanaka, R., Tohi, K. and Hamada, H. (2000) Tetrahedron Letters, 41, 6799–6802. (b) Nakamura, K. and Yamanaka, R. (2002) Chemical Communications, 1782–1783. (c) Nakamura, K. and Yamanaka, R. (2002) Tetrahedron: Asymmetry, 13, 2529–2533. (d) Utsukihara, T., Chai, W., Kato, N., Nakamura, K. and Horiuchi, C.A. (2004) Journal of Molecular Catalysis B: Enzymatic, 31, 19–24. (e) Utsukihara, T., Misumi, O., Kato, N., Kuroiwa, T. and Horiuchi, C.A. (2006) Tetrahedron: Asymmetry, 17, 1179–1185. (f) Itoh, K., Sakamaki, H., Nakamura, K. and Horiuchi, C.A. (2005) Tetrahedron: Asymmetry, 16, 1403–1408. (g) Parkinson, B.A. and Weaver, P.F. (1984) Nature, 309, 148–149. (h) Miyatani, R. and Amao, Y. (2004) Journal of Molecular Catalysis B: Enzymatic, 27, 121–125. 7 (a) Kim, S., Yun, S.-E. and Kang, C. (1999) Journal of Electroanalytical Chemistry, 465, 153–159. (b) Hollmann, F., Schmid, A. and Steckhan, E. (2001) Angewandte Chemie International Edition, 40, 169–171. (c) Delecouls-Servat, K., Bergel, A. and Basseguy, R. (2001) Journal of Applied Electrochemistry, 31, 1095–1101. (d) Delecouls-Servat, K., Bergel, A. and Basseguy, R. (2004) Bioprocess Biosystems Engineering, 26, 205–215. (e) DelecoulsServat, K., Basseguy, R. and Bergel, A. (2002) Chemical Engineering Science, 57, 4633–4642. (f) Suye, S., Aramoto, Y., Nakamura, M., Tabata, I. and Sakakibara, M. (2002) Enzyme and Microbial Technology, 30, 139–144. 8 (a) Demirjian, D.C., Shah, P.C. and Moris-Varas, F. (2000) Biocatalysis from Discovery to Application (ed. W.-D., Fessner), Springer-Verlag, Berlin,

j225


j 8 Enzymatic Reduction Reaction

226

pp. 1–29. (b) Carballeira, J.D., Álvarez, E., Campillo, M., Pardo, L. and Sinisterra, J.V. (2004) Tetrahedron: Asymmetry, 15, 951–962. (c) Yasohara, Y., Kizaki, N., Hasegawa, J., Takahashi, S., Wada, M., Kataoka, M. and Shimizu, S. (1999) Applied Microbiology and Biotechnology, 51, 847–851. (d) Miya, H., Kawada, M. and Sugiyama, Y. (1996) Bioscience Biotechnology and Biochemistry, 60, 95–98. (e) Shafiee, A., Motamedi, H. and King, A. (1998) Applied Microbiology and Biotechnology, 49, 709–717. (f) Kambourakis, S. and Rozzell, J.D. (2004) Tetrahedron, 60, 663–669. (g) Homann, M.J., Vail, R.B., Previte, E., Tamarez, M., Morgan, B., Dodds, D.R. and Zaks, A. (2004) Tetrahedron, 60, 789–797. 9 (a) Kaluzna, I., Andrew, A.A., Bonilla, M., Martzen, M.R. and Stewart, J.D. (2002) Journal of Molecular Catalysis B: Enzymatic, 17, 101–105. (b) Kaluzna, I.A., Matsuda, T., Sewell, A.K. and Stewart, J.D. (2004) Journal of the American Chemical Society, 126, 12827–12832. 10 (a) Yoshizumi, A., Wada, M., Takagi, H., Shimizu, S. and Nakamori, S. (2001) Bioscience Biotechnology and Biochemistry, 65, 830–836. (b) Patel, R.N., Goswami, A., Chu, L., Donovan, M.J., Nanduri, V., Goldberg, S., Johnston, R., Siva, P.J., Nielsen, B., Fan, J., He, W.X., Shi, Z., Wang, K.Y., Eiring, R., Cazzulino, D., Singh, A., Mueller, R. (2004) Tetrahedron: Asymmetry, 15, 1247–1258. (c) Yamamoto, H., Matsuyama, A. and Kobayashi, Y. (2002) Bioscience Biotechnology and Biochemistry, 66, 481–483. (d) Yamamoto, H., Kimoto, N., Matsuyama, A. and Kobayashi, Y. (2002) Bioscience Biotechnology and Biochemistry, 66, 1775–1778. 11 (a) Kataoka, M., Kita, K., Wada, M., Yasohara, Y., Hasegawa, J. and Shimizu, S. (2003) Applied Microbiology and Biotechnology, 62, 437–445. (b) Yasohara, Y., Kizaki, N., Hasegawa, J., Wada, M., Kataoka, M. and Shimizu, S. (2000) Bioscience Biotechnology and Biochemistry,

64, 1430–1436. (c) Kizaki, N., Yasohara, Y., Hasegawa, J., Wada, M., Kataoka, M. and Shimizu, S. (2001) Applied Microbiology and Biotechnology, 55, 590–595. (d) Yasohara, Y., Kizaki, N., Hasegawa, J., Wada, M., Kataoka, M. and Shimizu, S. (2001) Tetrahedron: Asymmetry, 12, 1713–1718. 12 Allen, S.J. and Holbrook, J.J. (2000) Protein Engineering, 13, 5–7. 13 (a) Zhou, B.-N., Gopalan, A.S., VanMiddlesworth, F., Shieh, W.-R. and Sih, C.J. (1983) Journal of the American Chemical Society, 105, 5925–5926. (b) Chen, C.-S., Zhou, B.-N., Girdaukas, G., Shieh, W.-R., VanMiddlesworth, F., Gopalan, A.S. and Sih, C.J. (1984) Bioorganic Chemistry, 12, 98–117. (c) Shieh, W.-R., Gopalan, A.S. and Sih, C.J. (1985) Journal of the American Chemical Society, 107, 2993–2994. (d) Fujisawa, T., Itoh, T. and Sato, T. (1984) Tetrahedron Letters, 25, 5083–5086. (e) Nakamura, K., Ushio, K., Oka, S., Ohno, A. and Yasui, S. (1984) Tetrahedron Letters, 25, 3979–3982. 14 (a) Nakamura, K., Kitano, K., Matsuda, T. and Ohno, A. (1996) Tetrahedron Letters, 37, 1629–1632. (b) Nakamura, K. and Matsuda, T. (1998) Journal of Organic Chemistry, 63, 8957–8964. (c) Matsuda, T., Harada, T., Nakajima, N. and Nakamura, K. (2000) Tetrahedron Letters, 41, 4135–4138. (d) Hamada, H., Miura, T., Kumobayashi, H., Matsuda, T., Harada, T. and Nakamura, K. (2001) Biotechnology Letters, 23, 1603–1606. (e) Matsuda, T., Nakajima, Y., Harada, T. and Nakamura, K. (2002) Tetrahedron: Asymmetry, 13, 971–974. (f) Nakamura, K., Matsuda, T., Itoh, T. and Ohno, A. (1996) Tetrahedron Letters, 37, 5727–5730. (g) Nakamura, K., Matsuda, T., Shimizu, M. and Fujisawa, T. (1998) Tetrahedron, 54, 8393–8402. (h) Matsuda, T., Harada, T., Nakajima, N., Itoh, T. and Nakamura, K. (2000) Journal of Organic Chemistry, 65, 157–163. (i) Nakamura, K., Matsuda, T. and Harada, T. (2002) Chirality, 14, 703–708.


References 15 (a) Nakamura, K., Kawai, Y. and Ohno, A. (1990) Tetrahedron Letters, 31, 267–270. (b) Nakamura, K., Inoue, K., Ushio, K., Oka, S. and Ohno, A. (1987) Chemistry Letters, 16, 679–682. (c) Nakamura, K., Kawai, Y., Oka, S. and Ohno, A. (1989) Tetrahedron Letters, 30, 2245–2246. (d) Dahl, A.C., Fjeldberg, M. and Madsen, J. (1999) Tetrahedron: Asymmetry, 10, 551–559. (e) Nakamura, K., Kawai, Y., Oka, S. and Ohno, A. (1989) Bulletin of the Chemical Society of Japan, 62, 875–879. (f) Forni, A., Moretti, I., Prati, F. and Torre, G. (1994) Tetrahedron, 50, 11995–12000. (g) Kim, J.-H. and Oh, W.-T. (1992) Bulletin of the Korean Chemical Society, 13, 2–3. (h) Hayakawa, R., Nozawa, K., Shimizu, M. and Fujisawa, T. (1998) Tetrahedron Letters, 39, 67–70. (i) Ushio, K., Hada, J., Tanaka, Y. and Ebara, K. (1993) Enzyme and Microbial Technology, 15, 222–228. (j) Hayakawa, R., Shimizu, M. and Fujisawa, T. (1997) Tetrahedron: Asymmetry, 8, 3201–3204. (k) Nakamura, K., Kawai, Y., Nakajima, N. and Ohno, A. (1991) Journal of Organic Chemistry, 56, 4778–4783. 16 (a) Pham, V.T., Phillips, R.S. and Ljungdahl, L.G. (1989) Journal of the American Chemical Society, 111, 1935–1936. (b) Zheng, C., Pham, V.T. and Phillips, R.S. (1994) Catalysis Today, 22, 607–620. (c) Tripp, A.E., Burdette, D.S., Zeikus, J.G. and Phillips, R.S. (1998) Journal of the American Chemical Society, 120, 5137–5141. (d) Heiss, C., Laivenieks, M., Zeikus, J.G. and Phillips, R.S. (2001) Journal of the American Chemical Society, 123, 345–346. 17 Li, F., Cui, J., Qian, X., Ren, W. and Wang, X. (2006) Chemical Communications, 865–867. 18 (a) Nakamura, K., Inoue, Y., Matsuda, T. and Misawa, I. (1999) Journal of the Chemical Society, Perkin Transactions 1, 2397–2402. (b) Oda, S., Sugai, T. and Ohta, H. (2001) Journal of Bioscience and Bioengineering, 91, 178–183. (c) Oda, S., Sugai, T. and Ohta, H. (2001) Journal of Bioscience and Bioengineering, 91, 202–207.

19 (a) Nakamura, K., Fujii, M. and Ida, Y. (2000) Journal of the Chemical Society, Perkin Transactions 1, 3205–3211. (b) Kometani, T., Toide, H., Daikaiji, Y. and Goto, M. (2001) Journal of Bioscience and Bioengineering, 91, 525–527. (c) Andrade Concei~ao, G.J., Moran, P.J.S. and Rodrigues, J.A.R. (2003) Tetrahedron: Asymmetry, 14, 43–45. 20 (a) Matsuda, T., Harada, T. and Nakamura, K. (2000) Chemical Communications, 1367–1368. (b) Matsuda, T., Watanabe, K., Kamitanaka, T., Harada, T. and Nakamura, K. (2003) Chemical Communications, 1198–1199. (c) Panza, J.L., Russell, A.J. and Beckman, E.J. (2002) Tetrahedron, 58, 4091–4104. 21 Howarth, J., James, P. and Dai, J.F. (2001) Tetrahedron Letters, 42, 7517–7519. 22 (a) Itoh, N., Morihama, R., Wang, J., Okada, K. and Mizuguchi, N. (1997) Applied and Environment Microbiology, 63, 3783–3788. (b) Kita, K., Fukura, T., Nakase, K., Okamoto, K., Yanase, H., Kataoka, M. and Shimizu, S. (1999) Applied and Environment Microbiology, 65, 5207–5211. (c) Howell, J.A.S., Palin, M.G., Jaouen, G., Top, S., Hafa, H.E. and Cense, J.M. (1993) Tetrahedron: Asymmetry, 4, 1241–1252. (d) Howell, J.A.S., Palin, M.G., Hafa, H.E., Top, S. and Jaouen, G. (1992) Tetrahedron: Asymmetry, 3, 1355–1356. (e) Baldoli, C., Buttero, P.D., Maiorana, S., Ottolina, G. and Riva, S. (1998) Tetrahedron: Asymmetry, 9, 1497–1504. 23 Nakamura, K., Takenaka, K., Fuji, M. and Ida, Y. (2002) Tetrahedron Letters, 43, 3629–3631. 24 (a) Ema, T., Moriya, H., Kofukuda, T., Ishida, T., Maehara, K., Utaka, M. and Sakai, T. (2001) Journal of Organic Chemistry, 66, 8682–8684. (b) Paizsa, C., Tosa, M., Majdik, C., Moldovan, P., Novak, L., Kolonits, P., Marcovici, A., Irimie, F.-D. and Poppe, L. (2003) Tetrahedron: Asymmetry, 14, 1495–1501. 25 Schubert, T., Hummel, W., Kula, M.R. and M€ uller, M. (2001) European Journal of Organic Chemistry, 4181–4187.

j227


j 8 Enzymatic Reduction Reaction

228

26 Itoh, N., Matsuda, M., Mabuchi, M., Dairi, T. and Wang, J. (2002) European Journal of Biochemistry, 269, 2394–2402. 27 (a) Stampfer, W., Kosjek, B., Faber, K. and Kroutil, W. (2003) Journal of Organic Chemistry, 68, 402–406. (b) Kosjek, B., Stampfer, W., van Deursen, R., Faber, K. and Kroutil, W. (2003) Tetrahedron, 59, 9517–9521. 28 (a) Baldassarre, F., Bertoni, G., Chiappe, C. and Marion, F. (2000) Journal of Molecular Catalysis B: Enzymatic, 11, 55–58. (b) Utsukihara, T., Watanabe, S., Tomiyama, A., Chai, W. and Horiuchi, C.A. (2006) Journal of Molecular Catalysis B: Enzymatic, 41, 103–109. (c) Baskar, B., Ganesh, S., Lokeswari, T.S. and Chadha, A. (2004) Journal of Molecular Catalysis B: Enzymatic, 27, 13–17. (d) Yadav, J.S., Nanda, S., Reddy, P.T. and Rao, A.B. (2002) Journal of Organic Chemistry, 67, 3900–3903. (e) Mczka, W.K. and Mironowicz, A. (2004) Tetrahedron: Asymmetry, 15, 1965–1967. (f) Andrade, L.H., Utsunomiya, R.S., Omori, A.T., Porto, A.L.M. and Comasseto, J.V. (2006) Journal of Molecular Catalysis B: Enzymatic, 38, 84–90. 29 (a) Nakamura, K., Kawai, Y., Miyai, T. and Ohno, A. (1990) Tetrahedron Letters, 31, 3631–3632. (b) Nakamura, K., Kawai, Y., Nakajima, N., Miyai, T., Honda, T. and Ohno, T. (1991) Bulletin of the Chemical Society of Japan, 64, 1467–1470. (c) Nakamura, K., Kawai, Y. and Ohno, A. (1991) Tetrahedron Letters, 32, 2927–2928. (d) Nakamura, K., Takano, S. and Ohno, A. (1993) Tetrahedron Letters, 34, 6087–6090. (e) Speicher, A., Roeser, H. and Heisel, R. (2003) Journal of Molecular Catalysis B: Enzymatic, 22, 71–77. (f) Nakamura, K., Miyoshi, H., Sugiyama, T. and Hamada, H. (1995) Phytochemistry, 40, 1419–1420. (g) Miya, H., Kawada, M. and Sugiyama, Y.

(1996) Bioscience Biotechnology and Biochemistry, 60, 95–98. (h) Perrone, M.G., Santandrea, E., Scilimati, T., Tortorella, V., Capitelli, F. and Bertolasi, V. (2004) Tetrahedron: Asymmetry, 15, 3501–3510. (i) Perrone, M.G., Santandrea, E., Scilimati, A., Syldatk, C., Tortorella, V., Capitelli, F. and Bertolasi, V. (2004) Tetrahedron: Asymmetry, 15, 3511–3517. (j) Ji, A., Wolberg, M., Hummel, W., Wandrey, C. and M€ uller, M. (2001) Chemical Communications, 57–58. (k) Kometani, T., Sakai, Y., Matsumae, H., Shibatani, T. and Matsuno, R. (1997) Journal of Fermentation and Bioengineering, 84, 195–199. (l) Dehli, J.R. and Gotor, V. (2001) Tetrahedron: Asymmetry, 12, 1485–1492. 30 (a) Nakamura, K., Miyai, T., Ushio, K., Oka, S. and Ohno, A. (1988) Bulletin of the Chemical Society of Japan, 61, 2089–2093. (b) Z€ uger, M.F., Giovannini, F. and Seebach, D. (1983) Angewandte ChemieInternational Edition, 22, 1012. (c) Seebach, D., Z€ uger, M.F., Giovannini, F., Sonnleitner, B. and Fiechter, A. (1984) Angewandte Chemie-International Edition, 23, 151. (d) Matzinger, P.K. and Leuenberger, H.G.W. (1985) Applied Microbiology and Biotechnology, 22, 208–210. 31 (a) Hasegawa, J., Ogura, M., Tsuda, S., Maemoto, S., Kutsuki, H. and Ohashi, T. (1990) Agricultural and Biological Chemistry, 54, 1819–1827. (b) Nakamura, K., Inoue, Y., Matsuda, T. and Ohno, A. (1995) Tetrahedron Letters, 36, 6263–6266. (c) Nakamura, K., Fujii, M. and Ida, Y. (2001) Tetrahedron: Asymmetry, 12, 3147–3153. (d) Allan, G.R. and Carnell, A.J. (2001) Journal of Organic Chemistry, 66, 6495–6497. (e) Takemoto, M. and Achiwa, K. (1998) Phytochemistry, 49, 1627–1629.


j229

9 Biooxidations in Chiral Synthesis Marko D. Mihovilovic and Dario A. Bianchi

9.1 Introduction

Enzyme-mediated oxidation reactions offer highly diverse options for the modification of existing functional groups as well as for the introduction of novel function in chiral catalysis. Biooxidations often enable us to obtain complementary solutions to metal-assisted transformations and organocatalysis and are considered one of the important strategies of ‘green chemistry’. The biooxidations particularly discussed within this chapter focus on enantioselective transformations yielding chiral building blocks for subsequent elaboration. The exceptional chemoselectivity in the majority of biological oxidation reactions enables highly atom efficient synthetic routes by limiting the utilization of protecting groups to a minimum, and hence significantly shortening the number of steps in an elaborate sequence toward complex target compounds. Combined with the usually high regioselectivity of enzyme-mediated reactions and the frequently encountered considerable promiscuity of enzymes vis-a-vis nonnatural substrates, such biooxidations are often highly predictable and allow for their utilization in the development of synthetic strategies. In particular, the utilization of molecular oxygen by oxygenases as primary oxidant in combination with mild reaction conditions makes them very appealing catalytic entities also for industrial applications fulfilling high safety standards. The application of biooxidation reactions both in synthetic laboratories and in fine chemical industry is hampered to a certain degree by the cofactor dependence of these enzymes as a general challenge in redox biocatalysis. Careful process optimization is required to be implemented in up-scaling efforts [1]. Biotransformations utilizing isolated proteins usually require nicotinamide cofactors (reduced nicotinamide adenine dinucleotide (NADH), reduced nicotinamide adenine dinucleotide phosphate (NADPH), etc.), which have to be recycled in order to enable an economically reasonable process. Hence, an auxiliary substrate has to be added in order to regenerate the cofactor in the required oxidation state ultimately closing the catalytic cycle. In general, this complicates the application in synthetic chemistry by increasing


j 9 Biooxidations in Chiral Synthesis

230

the preparative efforts; a more complicated experimental setup is required and product isolation is more demanding. This problem is often solved by creating a closed-loop system with an additional enzyme for cofactor regeneration [2]. Suitable systems are available for recycling of the most abundant cofactors NADH and NADPH as well as their oxidized forms. Recent progress in this field focused on the utilization of cheap auxiliary substrates like formate, glucose, or phosphite in combination with promiscuous systems for the recycling of phosphorylated and nonphosphorylated nicotinamide systems. Such systems bear the advantage of being ‘orthogonal’ to the particular conversion of substrate to product, thus minimizing the danger of unwanted side reactions in the context of cofactor regeneration. Several strategies for nonenzymatic coenzyme recycling have been proposed [3]. Attractive concepts for industrial applications include, in particular, electrochemical cofactor recycling [4] as well as the application of metal complexes (especially Rh systems) to take over the role of the dehydrogenase in closed-loop systems [5]. Another option for recovering cofactors in sufficient amounts is represented by the application of living whole-cells [6,7]. While the microorganism remains intact and its metabolism is operational, cofactor recycling is taken care of by the cell. However, wild-type strains contain a multitude of additional enzymes that may interfere with the required biotransformation, either resulting in a decrease in yield or stereoselectivity. Such obstacles can be overcome by genetically engineering cells to overproduce the required biocatalyst. As a consequence, the desired catalytic entity becomes the dominant fraction in the cell’s proteome and side reactions of competing enzymes become negligible. Eventually, the production of a competing protein can be suppressed by gene knockout to completely delete unwanted biotransformations. Both strategies can be combined if required. Hence, recombinant microorganisms represent efficient biocatalytic entities to produce the required enzyme and simultaneously take care of any cofactor recycling. Redox biocatalysts are often composed of multifunctional domains or represent protein complexes or membrane bound/attached enzymes. Consequently, their purification process may become elaborate and their stability is limited. The latter obstacle can be overcome by recent progress in molecular biology to enable knowledge-based tailoring of enzymes for improving their thermal stability without compromising catalytic efficiency [8]. Application of whole-cells again may be an appealing option as it avoids the protein purification process, the biocatalysts remain in their native environment, and no disruption of the cellular compartments is required. In biooxidation reactions, a majority of transformations is still performed utilizing living whole-cells or a variation thereof (lyophilized cells, crude cell extracts, etc.) to minimize purification efforts by concomitant facile cofactor recycling and maximum biocatalyst activity. The literature describing the characterization and exploitation of a large diversity of biooxidation catalysts has become very rich in recent years and is comprehensively covered by books and reviews in journals [9–15]. Consequently, the current chapter sets out to highlight a few representative and recent examples covering major developments in the field since the beginning of this millennium. The major focus


9.2 Oxidations of Alcohols and Amines

is on contributions of particular relevance for synthetic elaboration. With respect to reaction types, the area of stereoselective biooxidations can be roughly subdivided into four major topics: oxidation reactions (oxidation of alcohols and amines) monooxygenation reactions (hydroxylations, epoxidations, Baeyer–Villiger oxidations, and heteroatom oxidations) dioxygenation reactions (aryl dihydroxylation, peroxidation, and alkene cleavage) halogenation reactions. Several reaction types and functional group transformations will be outlined in the following sections with a major emphasis on those biocatalytic processes of major impact on enantioselective synthesis and chiral product preparation.

9.2 Oxidations of Alcohols and Amines

The enzymatic oxidation of alcohols and amines is catalyzed by various oxidoreductases of the dehydrogenase, oxidase, or peroxidase type [16–19]. This transformation may result in the destruction of chirality (at least at the center of reaction), as is the case in the comprehensive conversion of secondary alcohols to carbonyl species. The unspecific biotransformation can be utilized for cofactor regeneration either in a coupled enzyme or in a coupled substrate approach; in the latter case, the oxidizing enzyme also acts as a stereoselective dehydrogenase. The regioselective oxidation of a particular hydroxyl group may be exploited in selected cases of multifunctional substrates. Certainly, kinetic resolutions or desymmetrizations of bifunctional precursors can be utilized for the production of optically active compounds. The incorporation of a biooxidation step into a redox tandem sequence may be of particular interest, as such a strategy results in a dynamic resolution process, ultimately converting a racemic substrate into a homochiral product. Such transformations are particularly valuable as they overcome the 50% limitation of classical resolutions. Both isolated enzymes and whole-cell systems have been successfully applied in this context. Table 9.1 provides an overview of particularly relevant biocatalytic systems for the biooxidation of alcohols and amines. In the subsequent sections representative examples will be outlined, applying oxidation reactions for the conversion of CO and CN groups. 9.2.1 Regioselective Oxidation of Alcohols

An impressive indication of the high regioselectivity of hydroxysteroid dehydrogenases (HSHDs) was reported for the oxidation of various hydroxyl groups at the steroid core of bile acids [26] (Scheme 9.1). The hydroxy-substituents at positions 3, 7, and 12 could be selectively addressed depending on the hydroxysteroid

j231


j 9 Biooxidations in Chiral Synthesis

232

Table 9.1 Representative isolated enzymes for the oxidation of alcohols.

Representative substrate profile

Biocatalyst/origin Horse liver alcohol dehydrogenase (HLADH) Yeast alcohol dehydrogenase (YADH) Saccharomyces cerevisiae Alcohol dehydrogenase (TBADH) from Thermoanaerobium brockii Glycerol dehydrogenase (GDH) Schizosaccahromyces pombe, Cellulomonas Hydroxysteroid dehydrogenases (HSDH) Arthrobacter BP2 ADH from Lactobacillus kefir ADH from Rhodococcus erythropolis Amino acid oxidase (L-AAO) from Proteus myxofaciens Amino acid oxidase (D-AAO) from porcine kidney Monoamine oxidase (MAO-N) from Aspergillus niger and mutants

[20] [21] [20] [22] [9,19] [9,19] [9,19] [23] [24] [25]

dehydrogenases (HSDH) utilized [27]. In addition to regioselectivity, the absolute stereochemistry was also differentiated by suitable biocatalysts. As the direction of biotransformations could be influenced upon addition of auxiliary substrates (formate versus pyruvate), stereoinversion of particular hydroxyl groups was realized upon sequential biooxidation with a- and b-HSHDs, and vice versa [28]. This technology platform was utilized in the preparation of Gd-labeled cholanoic acids for magnetic resonance imaging [29]. A similar platform of enzymes is available for the regioselective oxidation of carbohydrates (Scheme 9.1) [30]. Carbohydrate dehydrogenases like glucose dehydrogenase (GDH) are capable of specifically attacking the sugar at the anomeric center (position 1) leading to gluconolactone, which is ultimately hydrolyzed to gluconic acid; this reaction is often utilized in cofactor recycling systems [31]. Pyranose oxidases (P2Os) of usually fungal origin oxidize the 2-position in various pyranose derivatives (mono- and disaccharides) [32], and can be utilized in a chemoenzymatic

12-HSDH GAOX OH

GDH

R HO

O

HO HO 3-HSDH

OH OH

Scheme 9.1 Regioselective alcohol oxidations.

OH OH

7-HSDH

OH

P2O


9.2 Oxidations of Alcohols and Amines

j233

sequence (‘Cetus process’) to convert D-glucose to D-fructose [33,34]. Galactose oxidase (GAOX) can be utilized for the specific conversion of the 6-hydroxyl to aldehydes or carboxylic acids [35,36]. Applications of such regioselective biooxidations in the modification of sugar derivatives are particularly important to access food additives. For example, biotransformation of D-galactitol with whole-cells of Enterobacter agglomerans gives the low-calorie sweetener tagatose in 86% chemical yield [37]. In addition, such siteselective biotransformations were utilized in the laccase-mediated [38] derivatization of bioactive compounds like asiaticoside with site-selective oxidation of a particular 6-hydroxyl in the glycon moiety of the target [39].

9.2.2 Desymmetrizations of Diols

The desymmetrization of meso-diols to chiral lactones undoubtedly represents the most prominent reaction by horse liver alcohol dehydrogenase (HLADH) in asymmetric synthesis [40,41]. This enzyme is particularly promiscuous to accept a large range of structurally diverse diols, and even strained ring systems [42] are readily converted to lactones in high chemical yields and optical purity (Scheme 9.2). Multiple chiral centers can be established in a single operation on preparative scale with excellent stereoselectivity (usually >99% ee) and the enzyme is highly selective for the biotransformation of S-alcohols with other oxidizable functional groups (e.g. olefins, sulfur) remaining intact. The overall conversion consists of two biocatalytic steps. Within the first biooxidation, a chiral hydroxy-aldehyde is formed, which is in equilibrium with a lactol species. This compound is then a substrate for the second oxidation. A very simple and efficient cofactor recycling strategy can be set up by adding FMN for the regeneration of NADþ and a facile protocol was established to access multigram quantities. Various biocatalytic options have been presented for the desymmetrization of meso-diols to chiral hydroxyl-ketones. A particularly facile system is represented by

OH

OH HLADH OH

O

HLADH O

CHO OH

NAD+

NADH

NAD+

FMN

FMNH2

Scheme 9.2 Desymmetrization of meso-diols by HLADH.

O NADH


j 9 Biooxidations in Chiral Synthesis

234

Biooxidation

OH

O

R.ruber

( )n meso

R

S

OH

OH

( )n OH O

Bioreduction

O

R.ruber

( )n

( )n

O

OH

O

OH R

( )n