HETEROGENEOUS CATALYSIS IN RACEMIZATION AND KINETICRESOLUTION ALONG A JOURNEY IN PROTEIN ENGINEERING

Tamás Görbe

Heterogeneous catalysis inracemization and kinetic resolutionalong a journey in protein engineering

Tamás Görbe

©Tamás Görbe, Stockholm University 2018 ISBN print 978-91-7797-185-6ISBN PDF 978-91-7797-186-3 Cover picture: Gergely JuhászPrinted in Sweden by Universitetsservice US-AB, Stockholm 2018Distributor: Department of Organic Chemistry, Stockholm University

Sometimes it is the people noone imagines anything of whodo the things that no one canimagine.- Alan Turing

��������� ������� ����������������� � ���

En katalysator är en kemisk förening som möjliggör och accelererar en kemisk reaktion utan att konsumeras själv. Katalys öppnar en alternativ väg och ger mildare reaktionsförhållanden, till exempel lägre temperatur. Katalysatorer används vanligen i organisk kemi och innehåller ofta övergångsmetaller. Naturen har också sina egna katalysatorer, proteiner vid namn enzymer. Dessa makromolekyler är ansvariga för vår ämnesomsättning, DNA-replikation, transkription och translation. Enzymer är miljövänliga, biologiskt nedbrytbara och de har möjlighet att utföra reaktioner under milda rektionsförhållanden.

Avhandlingen fokuserar på utveckling av heterogena katalysatorer för racemicering och kinetisk resolvering. Under de kinetiska resolveringarna applicerades ett enzym som en katalysator för enantioselektiv acylering av alkoholer eller aminer. Nackdelen med den kinetiska resolveringen är att endast 50% av substratet transformeras till produkten. För att kunna nyttja de resterande 50% substrat så kan resolveringar kombineras med racemiseringsreaktioner. Genom dessa reaktioner kan alkoholerna eller aminerna omvandlas till den eftersökta enantiomeren i 100% utbyte. Racemiseringen av tertiära alkoholer har beskrivits i min avhandling och allyliska tertiära alkoholer kan användas som startmaterial i en migrerande dynamisk kinetisk resolvering (DKR). Dessutom beskriver min avhandling utvecklandet av en biohybridkatalysator, vilken utgöres av CalB-enzymet och Pd-nanopartiklar för DKR. Denna katalysator karakteriserades med flera analytiska tekniker.

Den sista delen av min avhandling handlar om proteindesign, som är en metod som möjliggör modifiering av naturliga enzymer. Detta kan göras genom att mutera den genetiska koden, vilken är syntesmallen för enzymet. Efter modellering av enzymstrukturen med målföreningarna valdes några aminosyror som muterades. De producerade enzymvarianterna visade sig vara effektiva för kinetiska resolveringar med tidigare oreaktiva substrat.

���

����

�� ���������������� �

This thesis is based on the following papers, which will be referred to by Ro-man numerals I-VI. Reprints were made with the kind permission from the publishers and the contribution by the author to each publication is clarified in Appendix A.

I.� Heterogeneous Acid-Catalyzed Racemization of Tertiary Alcohols

Tamás Görbe, Richard Lihammar, Jan-E. Bäckvall Chemistry – A European Journal 2018, 24, 77–80

II.� Migratory Dynamic Kinetic Resolution of Carbocyclic Allylic Al-cohols Chicco M. Sapu, Tamás Görbe, Richard Lihammar, Jan-E. Bäckvall, Jan Deska Organic Letters 2014, 16, 5952–5955

III.� Design of a Pd(0)-CalB CLEA Biohybrid Catalyst and Its Appli-

cation in a One-Pot Cascade Reaction Tamás Görbe, Karl P. J. Gustafson, Oscar Verho, Gabriella Kervefors, Haoquan Zheng, Xiaodong Zou, Eric V. Johnston, Jan-E. Bäckvall ACS Catalysis 2017, 7, 1601–1605

IV.� Application and Further Structure Elucidation of Pd(0)-CalB

CLEA Biohybrid Catalyst – Chemoenzymatic Dynamic Kinetic Resolution of Primary Benzylic Amines Karl P. J. Gustafson†, Tamás Görbe†, Gonzalo de Gonzalo-Calvo, Ning Yuan, Cynthia Schreiber, Andrey Shchukarev, Cheuk-Wai Tai, Ingmar Persson, Jan-E. Bäckvall Manuscript † Authors contributed equally to the publication

V.� Improved Enantioselectivity of Subtilisin Carlsberg Towards Sec-

ondary Alcohols by Protein Engineering Robin Dorau, Tamás Görbe, Maria S. Humble ChemBioChem 2018, 19 (4), 338–346

���

VI.� Transesterification of tert-Alcohols by Engineered Candida antarc-tica Lipase A Tamás Görbe†, Johanna Löfgren†, Michael Oschmann, Maria S. Hum-ble, Jan-E. Bäckvall Manuscript † Authors contributed equally to the publication

Publications not included in this thesis: Synthesis of Regioisomeric 17��-N-Phenylpyrazolyl Steroid Derivatives and Their Inhibitory Effect on 17�-Hydroxylase/C17,20-lyase Zoltán Iványi, János Wölfling, Tamás Görbe, Mihály Szécsi, Tibor Witt-mann, Gyula Schneider Steroids 2010, 75, 450–456 Androsztánvázas 17-es Helyzetü 5’-Jód-triazolok Regiospecifikus Szín-tézise Tamás Görbe Magyar Kémikusok Lapja (Journal of Hungarian Chemical Society) 2012, 10, 298–300 Synthesis of Novel 17-(5’-Iodo)triazolyl-3-methoxyestrane Epimers via Cu(I)-catalyzed Azide–Alkyne Cycloaddition, and an Evaluation of Their Cytotoxic Activity in vitro Gyula Schneider, Tamás Görbe, Erzsébet Mernyák, János Wölfling, Tamás Holczbauer, Mátyás Czugler, Pál Sohár, Renáta Minorics, István Zupkó Steroids 2015, 98, 153–165 Synthesis of Novel 17-Triazolyl-androst-5-en-3-ol Epimers via Cu(I)-cat-alyzed Azide–Alkyne Cycloaddition and their Inhibitory Effect on 17�-Hydroxylase/C17,20-lyase Anita Kiss, Bianka Edina Herman, Tamás Görbe, Erzsébet Mernyák, Barnabás Molnár, János Wölfling, Mihály Szécsi, Gyula Schneider Accepted Manuscript (Steroids), 2018

��

������ �

Populärvetenskaplig sammanfattning på svenska i

List of Publications iii Abbreviations vii Amino Acid Abbreviations viii 1. Introduction 1 1.1 Enzymes – Nature’s catalysts 1 1.1.1 Transition-metal catalysis 2 1.1.2 Homogeneous and heterogeneous catalysis 3 1.1.3 Metal nanoparticles as catalysts 3 1.1.4 One-pot cascade reactions 5 1.2 Asymmetric reactions and desymmetrization 7 1.2.1 Enzyme-catalyzed kinetic resolution 8 1.2.2 Racemization techniques 11 1.2.3 Dynamic kinetic resolution 12 1.3 Protein engineering 13 1.3.1 Directed evolution 13 1.3.2 Rational and semi-rational design 14 1.3.3 High-throughput screening of an enzyme library 16 1.3.4 Immobilization of enzymes 17 1.4 Objective of this thesis 19 2. Heterogeneous Acid-Catalyzed Racemization of Tertiary Alcohols (Paper I.) 20

2.1 Introduction 20 2.2 Results and discussion 20 2.3 Conclusion 24

3. Migratory Dynamic Kinetic Resolution of Carbocyclic Allylic Alco-hols (Paper II) 25

3.1 Introduction 25

���

3.2 Results and discussion 26 3.3 Conclusion 30

4. Design of a Pd(0)-CalB CLEA Biohybrid Catalyst and Its Catalytic Applications (Paper III and IV) 31

4.1 Introduction 31 4.2 Synthesis of the Pd(0)-CalB CLEA catalyst 32 4.3 Results and discussion – Cascade reaction 33 4.4 Results and discussion – DKR of benzylic primary amines 37 4.5 Characterization and analysis of Pd(0)-CalB CLEA 44 4.6 Conclusion 48

5. Improved Enantioselectivity of Subtilisin Carlsberg Towards Second-ary Alcohols by Protein Engineering (Paper V) 50

5.1 Introduction 50 5.2 Results and discussion 52 5.3 Conclusion 58

6. Transesterification of Tertiary Alcohols by Engineered Candida antarctica Lipase A (Paper VI) 59

5.1 Introduction 59 5.2 Results and discussion 60 5.3 Conclusion 65

7. Concluding Remarks 66 Appendix A 67 Appendix B 68 Acknowledgements 69 References 72

�

����

���������� �

Abbreviations and acronyms are used in agreement with standards of the subject.[1] Only nonstandard and unconventional ones that appear in the thesis are listed here.

CalA/B Candida antarctica lipase A/B CAST combinatorial active site saturation test conv. conversion CLEA cross-linked enzyme aggregates CLEC cross-linked enzyme crystals DKR dynamic kinetic resolution ee enantiomeric excess epPCR error-prone polymerase chain reaction HAADF STEM high-angle annular dark-field-scanning transmission electron microscopy KR kinetic resolution ICP-OES inductively coupled plasma-optical emission spectroscopy MCF mesocellular foam OE-PCR overlap extension polymerase chain reaction pNP-ester p-nitrophenyl ester rac. racemic SC subtilisin Carlsberg SDM site-directed mutagenesis sec. secondary tert. tertiary TEM transmission electron microscopy wt wild-type XAFS X-ray absorption fine structure XAS X-ray absorption spectroscopy XPS X-ray photoelectron spectroscopy

�����

��������������������� �

Name Three letters Single letter

Alanine Ala A

Arginine Arg R

Asparagine Asn N

Aspartate Asp D

Cysteine Cys C

Glutamine Gln Q

Glutamate Glu E

Glycine Gly G

Histidine His H

Isoleucine Ile I

Leucine Leu L

Lysine Lys K

Methionine Met M

Phenylalanine Phe F

Proline Pro P

Serine Ser S

Threonine Thr T

Tryptophan Trp W

Tyrosine Tyr Y

Valine Val V

��

����������������

������������ �����������������Catalysis constitute a fundamental concept in (organic) chemistry as it offers a way to speed up chemical reactions by enabling new reaction pathways (Fig-ure 1.1). A catalyst is an additive substance, which enables a lower energy barrier pathway during the reaction without affecting the standard Gibbs free energy or being consumed during the process. Instead, the catalyst is regener-ated at the end of the catalytic cycle and is available to partake in another catalytic cycle. Catalysis has a major effect on our everyday life where ~90% of our chemical products are produced using catalysis to some extent.[2] There are several different types of catalysis, but perhaps the two most widely stud-ied ones during recent years are enzyme and transition metal catalysis. In Na-ture, catalytic processes are carried out by amino acid or nucleic acid derived macromolecules called enzymes. Enzymes have been evolved over millions of years of evolution and defines the state of the art of catalysis in living or-ganisms. Nature designed these macromolecular machineries to catalyze a wide range of biochemical reactions involved in metabolism, DNA replica-tion, transcription and translation, under mild conditions with impressive ef-ficiency and selectivity. Because of their biological origin, they are also the ultimate green catalysts and a renewable source. These properties make en-zymes attractive and sustainable catalysts and as a result they are currently being exploited in large-scale fine chemical processes and other industrial ar-eas.[3] A major drawback associated with enzymes have traditionally been their sensitivity and incompatibility with organic solvents. However, recent studies have shown that these issues are not general for all enzymes, and under certain conditions some of them can efficiently work in organic media.[4] This discovery laid the foundation for what would become the field of biocatalysis, and have had a tremendous positive impact on our society.

��

���������������������������������������The field of catalysis can be divided into two major categories: homogeneous and heterogeneous catalysis. In homogeneous catalysis, as the name implies, the catalyst is in the same phase as the substrate(s) and the reactant(s), and for the heterogeneous case the catalyst is in a different phase.[7] Heterogeneous catalysts are easier to separate and recycle, which makes them attractive for industrial applications. Over 80% of the industrial processes used today are performed with heterogeneous catalysts.[8] One of the most widely used appli-cations is the contact process, which is the current method for producing sul-furic acid in high concentration and high purity. In the early version of the process, platinum was used as heterogeneous catalyst. Later vanadium oxide (V2O5) was applied, which was not as prone as platinum to deactivation by the arsenic impurities derived from the sulfur source.[9] Aside from the simpler separation and recycling of heterogeneous catalysts, they also exhibit higher stabilities in many cases, which is beneficial from an economic perspective.[8]

Although heterogeneous catalysts possess several positive properties, they are unfortunately more difficult to characterize and design. The characteriza-tion and mechanistic studies require multidisciplinary knowledge of physical chemistry, surface characteristics, organic chemistry and material science. In contrast, homogeneous catalysts can be precisely characterized via a number of different analytical methods, including spectroscopical techniques such as FT-IR and NMR. Furthermore, the possibility to precisely determine the struc-ture of the homogeneous catalyst makes rational design and mechanistic stud-ies more straightforward. As a result, the development of asymmetric proto-cols are generally straighter for homogeneous catalysts than their heterogene-ous counterparts.[10]

�������������������������������������In 1959, Richard Feynman predicted the possibility of nanosized machines and the fruitful future of nanotechnology during his lecture "There's Plenty of Room at the Bottom". He also speculated that scientists would one day be capable of "arranging the atoms the way we want", and be able to do chemical reactions by mechanical manipulations (Figure 1.2).[11] From the 1980s the area of nanotechnology emerged to become a new scientific field.

��

Figure 1.2. R. Feynman is teaching undergraduate students.

Metal nanoparticles are essentially larger groups of metal atoms. Based on

the size of the clusters, they can contain tens to several thousands of atoms. Nanoparticles have attracted significant attention during the last decade be-cause of their unusual properties, and researchers from several different fields, such as chemistry, medicine, electronics, and energy technology, have envi-sioned a large future demand for these materials.[12] The synthesized nanopar-ticles have opened up new possibilities in the field of heterogeneous cataly-sis.[13] In most of the cases, they have high atom economy, activity, and selec-tivity in combination with recyclability, which is of high priority in green chemistry.

A unique feature of nanoparticles is that their catalytic properties can be modulated through tailoring of their size and shape.[14] The catalytic activity is very dependent on the change of the surface-to-volume ratio, where smaller particles have a higher ratio of surface atoms than larger ones (Figure 1.3). This allows for a more effective atomic arrangement and provides more inter-actions with the reaction environment, thus increasing the number of atoms that are available to partake in the catalytic reaction/process. Also, the large number of coordinatively unsaturated metal atoms often confers the nanopar-ticles with a unique set of catalytic properties that can be very different from those of the corresponding bulk metal.[15]

��

ammonia and the only byproduct is water, which makes it a good green chem-istry alternative towards the synthesis of amines. The amines were obtained in high yields and excellent stereoselectivities (Scheme 1.3).17

Scheme 1.3. Enantioselective amine synthesis by an in vitro enzyme cascade.

One of the latest applications is combining biocatalysis with metal cataly-

sis to be able to carry out an even broader range of transformations. The vari-ety of metals and catalysis approaches are very broad and the presence of an enzyme can enable new and stereoselective transformation. Our group has de-veloped a heterogeneous artificial metalloenzyme based on MCF material, which combined Pd nanoparticles and CalB enzyme. The enzyme catalyzed the resolution of the amine with high sterespecificity, meanwhile the Pd(0) racemized in situ the non-reactive enantiomer of the amine.[18] Another inter-esting example was reported by Okuda and Hayashi,[19] where they studied a biohybrid catalyst for olefin metathesis in water. During their studies, differ-ent types of Hoveyda–Grubbs–catalyst precursors were incorporated into the cavity of the mutated �-barrel protein nitrobindin. With this artificial metal-loenzyme, the ring-opening metathesis polymerization (ROMP)[20] and the ring-closing metathesis (RCM)[21] reactions could be performed in water under relatively mild conditions (pH 6, 25–40 ºC).�

��������������������������������������The word chirality is based on the Greek word for hand, χέρι (kheir), and is probably the most straightforward representation of the phenomenon. Some molecules are non-superimposable on their mirror images, just like our hands. For example, a carbon atom that binds to four different substituents constitutes a stereogenic center, which gives rise to two mirror forms. The two mirror images are called enantiomers and designated as left-handed (S) or right-handed (R). Enantiomers shares the same physical properties, except that they rotate plane-polarized light in opposite directions. The chirality of molecules

R1 R2

OH

R1 R2

NH2

R1 R2

O

NAD+

NADHADH

NH3/NH4+

H2OAmDH

�

is an important concept when it comes to biological applications, such as phar-maceuticals, as a pair of enantiomers can exhibit very different biological ac-tivities. This is because the biological building blocks, such as the amino ac-ids, sugars, and lipids that make up all biological macromolecules are chiral themselves. Therefore, when a small molecule is introduced to this chiral bi-ological environment, it can exhibit a very different effect depending on its stereochemistry, and that is why there is a large interest from the pharmaceu-tical, agricultural, flavor, and fragrance industries to synthesize enantiomeri-cally pure compounds and study them separately.[22] For example, spearmint leaves contain (R)-(–)-carvone and the caraway seeds have the other enantio-mer, (S)-(+)-carvone. Anyone can easily smell the difference between the two molecules caused by the different interaction with the olfactory receptors in our nose.[23]

���������� ���������!����������������Because of the different biological behavior of chiral molecules, there is a demand to synthesize each molecule selectively, especially in the case of po-tential drug candidates. There are three main techniques for the synthesis of enantiomerically pure compounds: (i) chiral pool synthesis, where a simple chiral natural product is applied as a building block in the synthesis of more complex chiral molecules,[24] (ii) asymmetric synthesis using chiral catalysts or reagents,[25] and (iii) resolution of racemic mixtures.[26] The simplicity of the resolution is appealing but it only converts 50% of the starting material (one enantiomer) into the product, whereas the other half remains untouched.

Kinetic resolution (KR) is an efficient method for the separation of alco-hols and amine enantiomers. The process relies on the difference between the reaction rate for the two enantiomers with the resolving agent. A variety of different protocols and resolving agents have been reported based on organo-catalysts, transition metals, or enzymes.[27]

Enzymes are a great choice for the KR process because of their high reac-tivity and high stereoselectivity. The enzyme class referred to as hydrolases have attracted considerable attention from chemists because of their natural behavior to catalyze the cleavage of a chemical bonds using water (hydroly-sis).[28] For example, hydrolases convert lipids to glycerol and fatty acids and proteases cleave the amide bonds of proteins to create smaller peptides or amino acid fragments.[29] The major breakthrough came when scientists real-ized that the enzymes do not necessarily need aqueous environment.[30] It was found that some enzymes, especially hydrolases, need only a layer of water to maintain their activity.[31] This finding made it possible to apply enzymes for organic reactions and opened up for a new avenue of applications. The organic

��

Scheme 1.4. Hydrolase-catalyzed ping-pong bi-bi mechanism.

To be able to quantify the enantioselectivity of an enzymatic reaction, the

enantiomeric ratio (E value) has been developed. This dimensionless parame-ter defines the ratio of the rate constant of the two enantiomers kfast/kslow and it is unaffected by the conversion. A higher E value implies a larger difference in the reaction rates of the two enantiomers. Three equations have been de-rived since both of the enantiomers are difficult to synthesize individually and the rate would be difficult to study. The E value can be calculated using any two of the following data: enantiomeric excess of the substrate (eeS), enantio-meric excess of the product (eeP) or the conversion of the reaction (c).[36] Equations:

� ������ ������

����� ������ Eq. (1)

� ������ ������

����� ������ Eq. (2)

� ������� ������ ���������

������ ������ ��������� Eq. (3)

N N

His

HH

OSer

Asp/Glu

OO

N N

His

H

OSer

Asp/Glu

OO

OO R2

H

R1

N N

His

H

OSer

Asp/Glu

OO

R1O

H

R2O

N N

His

H

OSer

Asp/Glu

OO

OO R2

H

R3

O

R4

O

R2

R4

R3

O

O

R2R1

R3

OH

R4

���

������"���������������#���The variety and efficiency of methods for carrying out resolutions are wide. However, they all have in common that they only convert one of the enantio-mers into the product, whereas the other one remains untouched, thus limiting the maximum yield to 50%. The racemization process can overcome this drawback by simply converting the non-reactive enantiomer into the racemic form (Scheme 1.5).

Scheme 1.5. Racemization of sec-alcohols.

This interconversion can occur through a variety of mechanisms: racemi-

zation via radical and redox reactions, acid or base-catalyzed racemization, enzyme-catalyzed racemization, photochemical racemization, thermal race-mization, racemization by nucleophilic substitution, racemization proceeding via meso-intermediates or Schiff-base intermediates.[37] The combination of KR and racemization is a powerful process for asymmetric synthesis.

In this thesis for instance, examples will be presented where acid and Pd(0)–catalysts are employed for racemization. The acid-catalyzed racemiza-tion of alcohols employed here proceeds via SN1 carbocation formation in which the hydroxy group is protonated by the acid, resulting in exclusion of water to create a planar sp2-carbocation, which is prochiral (Scheme 1.6). Sub-sequent re-addition of water is nonselective and results in a racemic mixture.

Scheme 1.6. Racemization of tertiary alcohols.

On the other hand, the racemization of amines in this thesis proceeds

through a transfer hydrogenation mechanism. In this case, the substrate will undergo a dehydrogenation and after that, since there is no other hydrogen acceptor, the hydride is transferred back to the oxidized substrate from the transition metal catalyst. Here, the oxidized substrate is prochiral and the re-addition of the hydride is not an enantioselective process, thereby leading to a racemic mixture of the amine.

R1

OH

R2

racemization catalyst

R1 R2

OH

R1 R2

R3 OH +H+

-H2OR1 R2

R3

-H+

+H2O

R1 R2

R3 OH

���

������$������!����������������The drawback of the KR process can be circumvented via a racemization pro-tocol and if it occurs in situ, a so called dynamic kinetic resolution (DKR) can be achieved. The racemization generates/establishes a dynamic equilibrium between the two enantiomers, which gives a continuous load of the reactive enantiomer for the resolution. The DKR process can thus increase the conver-sion of the desired product up to 100%, which makes the reaction more atom economical and industrially attractive (Scheme 1.7).

Scheme 1.7. DKR of sec-alcohols.

The efficiency and compatibility between the two combined reactions, rac-

emization and resolution, are the most important factors for a successful DKR. However, the design of such systems is quite difficult to realize. In general, an efficient KR reaction is the starting point and thereafter a suitable racemi-zation catalyst is chosen for the system. When designing this cooperative cat-alytic scheme, one has to be certain that the two catalytic reactions do not inhibit each other.[38] An efficient KR system can be further developed to a DKR if it has a high selectivity (E value > 20)[36] and the rate of the racemiza-tion is at least equal to the rate of the fast reacting enantiomer (krac ≥ kfast). Chemoenzymatic DKR has turned out to be an exciting research field within asymmetric catalysis.[18b,38,39] There are several examples, of transition metals that have been successfully used as racemization catalysts in combination with enzymatic reactions.[38,39,40] In the Bäckvall group, several protocols have been developed for the DKR of sec-alcohols,[41] allylic alcohols,[42] homoallylic al-cohols,[43] diols,[44] N-heterocyclic 1,2-amino alcohols,[45] primary amines,[46] �-amino esters,[47] hydroxy esters[48] and chlorohydrins.[49] For these sub-strates, several different enzymes have been used with different transition metal-based racemization catalyst. DKR protocols have been also applied to the total synthesis of pharmaceutically active molecules, such as bufuralol,[50] duloxetine[51] and salbutamol.[52]

R1

OH

R2

in s

itura

cem

izat

ion

R1

OH

R2

lipase enzymeacyl donor

R1

O

R2

O

R3

R1

O

R2

O

R3

99% (99% ee)

���

However, the portfolio/toolbox of available enzymes is constantly grow-ing, while the number of compatible homogeneous racemizaton catalysts are still rather limited.[41,53] Several enzymes have on the other hand been shown to be compatible with heterogeneous catalysts, such as Raney Ni and Co,[54] Ru(OH)3/Al2O3,[55] SO4

2-/TiO2,[56] VOSO4·5H2O,[57] acidic zeolites,[58] acidic resins[59] and Pd(0) immobilized on supports,[46a,47a,60] in the DKR of sec-alco-hols and amines.

����%���������������The term protein engineering refers to the process of artificially modifying protein macromolecules. According to this definition, protein engineering can be achieved either via synthetic modifications or through molecular biology techniques. The engineering occurs at the specific gene sequence which is translated to the change of the protein’s properties. There exist several differ-ent protein engineering techniques ranging from those that are fully random-ized, which mimics the Darwinian evolution, to those based on more rational approaches where the protein design is based on in silico modeling. The com-mon denominator between each technique is the polymerase chain reaction (PCR), which is used to carry out the modification of the genetic material.[61]

������$�������&��������In 1859, the British biologist/naturalist Charles Darwin published his book “The Origin of Species”, which explains the theory of adapted changes over generations. The basic statement of the book is that genetic variation is based on the sexual recombination between individuals or by mutation of the genetic material. Natural selection ascertains that only the species with the beneficial mutations will survive, a concept also referred to as “survival of the fittest”.[62]

These random mutations can for example originate from exposure to UV-light or emerge through chemically-induced changes in the DNA, or simply be mistakes caused by the DNA polymerase during the natural transcription process.63 Most of the mutations do not trigger a phenotype change and are called silent mutations. They usually consist of only an alteration in a single nucleotide in the three-letter code; however, they have an important role in the evolution until the phenotypic property changes. The mutations that provide advantages for the fitness of the organism will favor its survival and/or repro-duction.

During the directed evolution process the natural evolution is mimicked in vitro, through several rounds of mutagenesis, followed by a screening and se-lection process. The random mutagenesis is performed via the error-prone

���

high-resolution structural information, such as a crystal structure of the de-sired protein. Based on this structure, it is possible to model and simulate to identify interesting target points. The protein of interest can be mutated by SDM using mismatch-containing PCR primers, resulting in changes at a spe-cific site.[68] The enzyme mutants are produced by a host cell system, which will be extracted and screened for the altered phenotypic properties. SDM can for example be used to evaluate the contribution of a specific amino acid res-idue by changing it to an inactive glycine or alanine residue.[69]

The semi-rational design approach combines the advantages of SDM and directed evolution. The selection of the specific site is rationally chosen but the insertion of the mutation incorporates degenerated PCR primers, which incorporates 1–4 different nucleotides at the same specific position, enabling the generation of a combinatorial gene library. In 2005, the Reetz group pub-lished a design method based on directed evolution, which was called combi-natorial active site saturation test (CAST).[70] The targeted enzymatic crystal structure was computationally studied and a set of amino acids were identified. These segments of the gene sequence were simultaneously randomized, usu-ally two or three sites at the time. The power of this technique is that it allows for identification of potential synergetic effects exerted by a certain combina-tion of amino acids, which cannot be achieved from the substitution of one amino acid at time. Such synergistic effects are unlikely to arise in the directed evolution process, since the 2–3 random mutations that introduced usually does not occur in proximity to one another. Iterative CASTing is to apply hits from a library as templates for the next round of CASTing where another site at the enzyme is targeted.

The typical way to generate a saturation mutagenesis library is to apply the NNK degenerated codon (N = any nucleotide, K = guanine or thymine), which constitutes all possible 32 codons, covering all 20 proteinogenic amino acids. If the aim is to reduce the size of the library, it is advantageous to be sure that each amino acid is presented by one codon, which will eliminate the over-representation of certain amino acids. If the desired degenerated codon is not available, it is still possible to mix the different primers in a suitable ratio. The NDT codon (D = adenine/guanine/thymine, T = thymine) used by the Reetz group could decrease the size of the library while maintaining comparable quality. NDT comprises 12 codons covering 12 amino acids (Asn, Asp, Arg, Cys, Gly, Ile, Phe, Leu, Tyr, Ser, His, Val), which is a good representation of all the chemical properties, such as polarity, aromatic, aliphatic, positively or negatively charged. There are existing possibilities for visualizing the process using an evolution tool, such as CASTER.[71] The concept was investigated by Reetz and coworkers when they studied the epoxide hydrolase from Aspergil-lus niger in which it was established that the hit rate was increased using the NDT codon.[72]

� �

���������� ��������������������'����������(�����One of the laborious and time-consuming part of protein engineering is the screening of the enzyme libraries in an efficient and reproducible fashion. The analysis time per sample and the automization of the sampling is crucial for practicality standpoint. However, there are other important factors that should be taken into account as well, like for example, the generality, robustness of assay and the overall cost of the entire process. One of the first published en-zyme screening assay for enantioselectivity was based on using UV-Vis spec-troscopy. The analysis (Quick-E-Test) monitors the cleavage of p-nitrophenyl ester (pNP-ester), a process that results in the release of p-nitrophenol, a com-pound with a specific absorption at 405 nm in its phenolate form (Scheme 1.9).[73]

Scheme 1.9. Enzymatic hydrolysis of p-nitrophenyl ester.

In order to determine the enantioselectivity of the hydrolases, both of the

(R)- and (S)-substrate enantiomers have to be studied individually with subse-quent comparison of their relative hydrolysis rates.[74] Our group has success-fully used the abovementioned method in three separate studies which in-volved the hydrolysis of pNP-esters to screen lipase variants.[75] To be able to extend the procedure, it is possible to look at the other products that are formed. For example, the formed acid can be analyzed by its pH. Addition of a pH indicator to the solution will show a change in color and the intensity can subsequently be measured by UV-vis spectroscopy at the indicator’s absorb-ance, a test commonly called Kazlauskas’ test.[76] However, this method re-quires that there exists a linear correlation between the acid concentration and the absorbance of the indicator. Also, the buffer has to be wisely chosen as it has to have approximately the same pKa value and its strength has to fit the preferred pH change.

It is also worth highlighting that the media of all these methods are aqueous buffers, which is not optimal for most organic reactions. To overcome this problem, Konarzycka-Bessler and Bornscheuer developed a transesterifica-tion process, a method that is suitable for screening the activity of hydrolases in an organic solvent (Scheme 1.10).[77] During the assay, a vinyl ester is em-ployed as the acyl donor and is allowed to react with an alcohol compound in the presence of the enzyme in an non-polar organic solvent, such as n-hexane, iso-octane or petroleum ether. The produced acetaldehyde will then react with

O2N

O

O

Rlipase mutants

bufferO2N

O

O

R HO

O

R

O

NO2

+ +

���

a fluorescent hydrazine to form an adduct that has a distinct, strongly fluores-cent character.

Scheme 1.10. High-throughput method for transesterification of alcohols.

����)�*�(�����������'������During an immobilization process, a catalyst is attached to a solid support. There are several ways to immobilize an enzyme, but it can be separated into three categories: (i) binding to a surface (by adsorption, or cova-lent/ionic/chelative attachment), (ii) encapsulation, (iii) cross-linking of the enzymes to each other (Figure 1.6). The most commonly used materials have a high surface area to be able to incorporate as many active catalytic species as possible. The immobilization of a biocatalyst on a support has several ad-vantages, for example it behaves as a heterogeneous catalyst, which makes the system easier to handle and store. It also simplifies the work-up and purifica-tion procedure of the product. In addition, it provides the possibility to recycle the catalyst.[78] Specifically for lipases, the binding to a solid material can cause positive conformational changes, similarly to the interfacial activation, and can thus increase the catalytic activity.[79]

However, the heterogenization of biocatalysts does not only confer posi-tive effects. For example, it adds an extra step to the preparation of the cata-lyst. In certain cases, activity loss can be observed and also the mass transfer issues can be problematic. The most common analysis of the enzymatic sys-tem is to evaluate the free enzyme and the immobilized versions. The free enzyme is considered to be freeze-dried. The freeze-dried enzyme has usually decreased activity, which relates to the structural perturbations, and aggregate formation especially in an organic solvent. These structural modifications can be overcome by using additives, such as salts, carbohydrates, crown ethers or ligands.[4,79]

O R1

O+ R2 OH

lipase

organic solvent O R1

OR2

+O

H

O

H+

NO

N

NO2

NHNH2

NO

N

NO2

HN N CH CH3

highly fluorescent

-H2O

���

cross-linked, which increases the stability and robustness. The disadvantage of this method is the crystallization process, which can be highly challenging to achieve. To overcome this issue, an alternative process was developed in-volving the creation of enzyme aggregation by addition of water-miscible sol-vents (for example iso-propanol), saturated salt solutions (such as NH4SO4), or non-ionic polymers. These aggregates can then be cross-linked together us-ing a linker such as glutaraldehyde, this process called cross-linked enzyme aggregates (CLEA). The CLEA procedure was used by the immobilization of several different enzymes and enzymatic cascades.[84]

The immobilized metal ion affinity chromatography (IMAC) is another way to bind the enzyme on a solid support. The resin of the IMAC column contains metal ions which can chelate with the histidine tagged (His-tag) pro-teins. The His-tag is genetically encoded in the enzymes and placed at the N- or C-terminal end. This technique is commonly used by the selective separa-tion and purification of the enzymes. Next to the purification, metal ion (usu-ally Co, Ni and Cu) containing 96-well plates are also available which can be used for screening assays or synthetic applications during the protein engi-neering.[85]

��)�+(,���&���'������������The overall objective of this thesis is the development of new enantioselective methods for kinetic resolution and dynamic kinetic resolution. To achieve this goal, we envisioned a racemization protocol for tert-alcohols via heterogene-ous acid catalysis. Another vision was to find a new-way to immobilize and combine enzymes with metal nanoparticles and expend the biohybrid-heter-genenous catalyst era, which would be suitable for several different transfor-mations. Nevertheless, I had a personal interest to modify enzymes with pro-miscuous activity. Based on protein engineering, the aim was to increase ac-tivity and selectivity of two different enzymes from the hydrolases family to-wards more sterically demanding substrates.

��

��� ������� �����!�����"#��$����#���������%�����"�����&�� �'������(�

����*������������A high demand for enantiomerically pure tert-alcohols exists but unfortu-nately only a few examples of chemical or enzymatic resolution have been reported for these compounds. The group of Zhao has studied the chemical resolution of 3-hydroxy-3-substituted oxindoles, in which an N-heterocyclic carbene (NHC) was applied as a resolving catalyst to obtain enantioenriched tert-alcohols.[86] An enzymatic resolution approach was reported by Born-schauer and co-workers within a great observation of a specific amino acid motif (GGGX-motif, G = glycine and X = any amino acids), which has high importance for the activity towards tert-alcohols.[87]

From these few examples, it is clear that there exists a need for the devel-opment of a compatible racemization system for tert-alcohols, which has not been presented in the literature. To be able to racemize the quaternary carbon center, the reaction cannot proceed through the typical transition metal-cata-lyzed mechanism (dehydrogenation/hydrogenation) since there are no hydro-gen atoms that can be shuffled.[88] The only way to racemize tert-alcohols is by cleavage of the C–O bond to give a carbocation. The planar carbocation formed from this process is pro-chiral, and upon trapping with water the ste-reochemical information is lost. The method has also been used successfully for sec-alcohol compounds, even in the context DKRs.[89] For our study, we were able to utilize this concept for the racemization of tert-alcohols.

����"����������$����������As a model substrate, 2-phenylbut-3-yn-2-ol (rac)-1a was chosen from the previous studies of enzymatic kinetic resolutions. The substrate was synthe-sized by a Grignard reaction of the corresponding acetophenone derivative with ethynylmagnesium bromide. The racemic mixture of the substrate (rac)-1a-e was subsequently acylated with acetic anhydride under microwave con-ditions to produce 2-phenylbut-3-yn-2-yl acetate (rac)-2a-e. For the synthesis of enantiomerically-pure 1a-e, enzyme-catalyzed KR was applied on the ra-cemic acetate 2a-e. The wt-CalA enzyme was used for this KR reaction but

���

unfortunately it was too reactive and did not afford (R)-2-phenylbut-3-yn-2-ol (R)-1a-e in high enantioselectivity. However, the untouched acetate (S)-2a-e could be delivered in high ee. The enzyme was used together with its immo-bilized form and Accurel MP-1001 hydrophobic support was employed in pH 7.6 phosphate buffer solution. The corresponding enantiomerically enriched acetate (S)-2a-e was hydrolyzed chemically to (S)-1a-e in a mixture of MeOH and aqueous saturated Na2CO3 (4:1). In this basic media, the selectivity was retained and afforded the desired alcohols for the racemization studies.

Scheme 2.1 Synthesis of the tert-alcohol substrates.

Carboxylic acid or sulfonic acid functionalized styrene–divinylbenzene

(gel) polymer resins were studied which are shown in Table 2.1.

Table 2.1 Screening of acidic resins.[a]

Enrty Resin Acidic group ee [%]

1 Amberlite 120 –COOH 13 2 Amberlyst 15 –SO3H <1

3 Dowex 50WX2 –SO3H 23

4 Dowex 50WX4 –SO3H <1

5 Dowex 50WX8 –SO3H 2 [a] 0.1 mmol substrate was added to 1 mL of water with 40 mg of acidic resin and stirred at room temperature. 0.05 mL of sam-ples were extracted with EtOAc and analyzed by GC-FID equipped with chiral column.

R

O

R

O OH

R

O

R

O

(rac)-1a-e (rac)-2a-e

(S)-2a-e

CalAon Accurel

O

O O

DMAP, TEA

MgBr

THF

buffer

OOH

R

MeOH

Na2CO3

(S)-1a-e

O

R

O

(rac)-2a-e

OHacidic resin

OH

(S)-1a (rac)-1a

H2O, 5 h

���

Based on the mechanism for tert-alcohols (Scheme 1.6, p. 11) it seemed to be an obvious choice to evaluate the acids in water as a solvent. The reactions were conducted on a 0.1 mmol scale in water (1 mL) using 40 mg of hetero-geneous acid. The acid should be strong enough to be able to racemize the substrate (S)-1a, and unfortunately the carboxylic acid functionalized resin, Amberlite 120, was not found to be sufficiently acidic to give full racemiza-tion (Table 2.1, Entry 1). However, Amberlyst 15 continaing sulfonic acid functionalities was found to be sufficient for this process (Table 2.1, entry 2). Dowex type of resins were also tested with various degree of cross-linking (2, 4, and 8%, Table 2.1, Entries 3-5). The cross-linkages had a high influence on the activity, which is related to the fact that the catalytically active substituents are spaced closer together on a volume basis thereby reducing the additional volume of the water. Because of these facts, Dowex 50WX4 and 50WX8 were acidic enough to be able to racemize (S)-1a (Table 2.1, Entries 4 and 5). Dur-ing the evaluation of different acids, the reactions were also analyzed by 1H NMR spectroscopy since it is known that upon carbocation formation there is a possibility for formation of styrene 3 as a side-product through elimination. However, at that point we were not aware of the side-product formation re-lated to propargylic alcohols, that is the Mayer-Schuster rearrangement.[90] In this rearrangement, unsaturated (E)- and (Z)- aldehydes can be produced. In line with this, both (E)-3-phenylbut-2-enal 4 and (Z)-3-phenylbut-2-enal 5 were detected by 1H NMR (Table 2.2). Table 2.2 Byproduct formation during the racemization.[a]

Entry Resin 1a [%][b] 3 [%][b] 4 [%][b] 5 [%][b]

1 Amberlite 120 81 0 10 9 2 Amberlyst 15 53 29 10 8

3 Dowex 50WX2 100 0 0 0

4 Dowex 50WX4 100 0 0 0

5 Dowex 50WX8 100 0 0 0 [a] 0.1 mmol substrate (rac)-1a was added to 1 mL of water with 40 mg of acidic resin and stirred at room temperature. 0.05 mL of samples were extracted with EtOAc, the organic phase was separated and dried. The samples were dissolved in CDCl3 and analyzed by 1H NMR. [b] NMR yield.

OHOH

+Ph

Ph

OPh

O+ +

Ph

(rac)-1a 3 4 5(S)-1a

H2O, 5 h

acidic resin

���

The application of Dowex resins did not lead to any byproduct formation; however, they displayed activity differences. Dowex 50WX8 was chosen for further studies since it was shown to have the highest racemization capability (Table 2.1, Entry 5). Subsequently, different solvents were studied for the rac-emization of (S)-2a using a reaction time of 5 h with the Dowex 50WX8 resin (Figure 2.1).

Figure 2.1. The screening of solvents.

It was already demonstrated that water gave the best results without any

byproduct formation. The iso-octane would be a suitable solvent for the enzy-matic KR and hence showed a good reaction rate but afforded 8% of 3 and 33% of 4 and 5 as byproducts. Toluene and acetonitrile were also evaluated but gave slow racemization. After 5 h, toluene gave 54% ee and acetonitrile 40% ee. An iso-octane/water (1:1) two-phase system was studied as well, based on its possible DKR application, when an organic solvent is necessary, it shows after 5 h, 9% ee of 1a was detected. The optimal conditions turned out to be 0.1 mmol substrate and 40 mg of Dowex 50WX8 in 1 mL water.

After the reaction optimization, differently substituted tert-alcohols were studied in our process (Figure 2.2).

�

��

��

��

��

���

� �� ��� ��� ��� ��� ��

����

�

��������

�� � ������� ���� �

�� �������� ������� ��� �

���

Figure 2.2 Substrate scope of the acid-catalyzed racemization of tert-alcohols.

The applied Dowex 50WX8 resin was active during the racemization of

(S)-1a and (S)-1b within 5 h. The electron-withdrawing group substituted sub-strates, such as 4-fluorophenyl (S)-1c, 4-chlorophenyl (S)-1d, and 4-nitro-phenyl (S)-1e, were also tested and revealed that (S)-1c was racemized within 5 h; however, (S)-1d required 28 h to reach full racemization. Carbocation formation is challenging for the 4-nitrophenyl substrate (S)-1e due to the elec-tron-withdrawing nature and therefore required an additional 120 mg acid, el-evated temperature (70 ºC) and a reaction time of 52 h to be fully racemized. It would also have been interesting to study electron-donating substituents. Unfortunately, acylation of 2-(4-methoxyphenyl)but-3-yn-2-ol failed. The acetylated compound was found to be unstable during the work-up procedure in which hydrolysis occurred to deliver the starting alcohol (S)-2-(4-methox-yphenyl)but-3-yn-2-ol.

����-�����������An easy-to-use and easily modifiable racemization protocol has been devel-oped for tert-alcohols. Five different heterogeneous acids were tested for ac-tivity and Dowex 50WX8 was found to be the optimal acid for racemization. Different types of solvents were evaluated in process, where water was found to give the best results, allowing for full racemization without any byproduct formation under mild reaction conditions. Three differently substituted tert-alcohols was successfully racemized at room temperature, but the 4-nitro-phenyl substituted (S)-1e required elevated temperature and longer reaction time in order to reach full racemization.

OH

OH

F

OH

H3C

OH

OH

O2N(rac)-1a

ee 1% (r.t., 5 h)

(rac)-1cee 1.7% (r.t., 5 h)

(rac)-1bee 1.2% (r.t., 5 h)

(rac)-1dee 1.4% (r.t., 28 h)

(rac)-1eee 3% (70 ºC, 52 h)

OHDowex 50WX8

OH

Cl

H2OR R

���

)��*�������"�+"������,������$ �����������������"��������"��������&�� �'�������(�

����*������������The application of lipases during the kinetic resolution reactions is common in organic synthesis to produce enantiomerically enriched alcohols both pri-mary and secondary.[18] However, tertiary alcohols are still a challenging sub-strate class due to the lack of activity and selectivity of the commonly em-ployed enzymes.[91] There are a few exceptions, for example, the tertiary pro-pargylic carbinols where serine hydrolase can accommodate the terminal al-kyne moiety.[92] The Deska group has been studying the total synthesis of taylofuran A. The envisioned starting material (rac)-6a of taylofuran A with an ethynyl substituent seemed to be a good platform to set the enantioselec-tivity using a lipase-catalyzed kinetic resolution (Scheme 3.1).

Scheme 3.1. Envisioned reaction pathway.

During our collaboration with the Deska group, attempts were made to use our heterogeneous acid-catalyzed racemization technique on substrate (rac)-6a in order to develop a DKR system. It was an unexpected astonishment that Candida antarctica lipase B converted 6a to the product with high selectivity. It was surprising to find that product (R)-7a was formed instead of the desired (R)- or (S)-6a. The formation of the sec-alcohol (R)-7a proceeds via a 1,3-transposition process that occurs during the mildly acidic, acetylation step in an acetic anhydride/pyridine system.

OH OH+

OH

1) Ac2O, Py

2) lipase, aq buffer

(R)- or (S)-6a0% yield

(R)-7a47% yield, 99% ee

(rac)-6a

� �

Scheme 3.2. Migratory DKR of the (rac)-6a.

This type of isomerization has been reported previously by Mulzer and his

group using trifluoroacetic acid.[93] Our system is milder than the previous protocols, and based on this observation we have focused to develop the trans-formation in a productive fashion. These allylic alcohols are attractive analogs for stereoselective transformations based on their dynamic behavior.[94,95] The synthesis of cyclic tertiary allylic carbinols is usually an easier task but the preparation of the secondary isomers typically involves several steps. It was therefore worth to study our system, which is based on the isomerization of tert-alcohols to generate sec-alcohols. The sec-alcohol generated a lipase-cat-alyzed transesterification and racemization to achieve a DKR process with high selectivity and isolated yield of the allylic carbinol derivatives in one pot (Scheme 3.2).

����"����������$����������Both the 1,3-isomerization of the tert-alcohol and the racemization of the sec-alcohol depend on the formation of transient allylic cationic intermediates. 3,4-Dihydro-[1,1’-biphenyl]-1(2H)-ol (rac)-6b and (S)-3,4,5,6-tetrahydro-[1,1’-biphenyl]-3-ol (S)-7b were chosen as model substrates. Both of the re-actions were individually investigated. The compound (rac)-6b was used to test the 1,3-isomerization and the enantioenriched (S)-7b was studied in the racemization. In both cases several heterogeneous acidic resins were evaluated (Table 3.1).

OH

(rac)-6a

heterogeneous acid

lipase enzyme + acyl donor

OO

R

(R)-10a

���

Table 3.1. Screening of the isomerization and racemization.[a]

Entry Alcohol Resin Solvent 7b [%] 8 [%]

1 (rac)-6b Amberlite 120 n-pentane 31 34

2 (S)-7b Amberlite 120 n-pentane 24 (31% ee) 36

3 (rac)-6b Amberlyst 15 n-pentane 39 26

4 (S)-7b Amberlyst 15 n-pentane 9 (5% ee) 35

5 (rac)-6b Dowex 50WX4 n-pentane 26 31

6 (S)-7b Dowex 50WX4 n-pentane 21 (25% ee) 36

7 (rac)-6b Amberlyst 15 iso-octane/H2O 68 5

8 (S)-7b Amberlyst 15 iso-octane/H2O 99 (5% ee) 0

9 (rac)-6b Dowex 50WX4 iso-octane/H2O 74 0

10 (S)-7b Dowex 50WX4 iso-octane/H2O 99 (22% ee) 0

[a] (rac)-6b (10 mg) or (S)-7b (10 mg, 99% ee) solved in 4 mL of n-pentan or iso-octane/water (1:1), the acidic resin was added to the solution and stirred at room temperature. The product ratios were determined by 1H NMR and the enantiose-lectivity was analyzed by HPLC equipped with chiral column.

Based on the enzymatic kinetic resolution, n-pentane was chosen as the

solvent. Amberlite 120, which is the weakest acidic resin, was found to be active in the isomerization of (rac)-6b (Table 3.1, Entry 1) and during the racemization of (S)-7b (Table 3.1, Entry 2), but the analysis showed several side-products being formed where the ether 8 was one of the major ones formed. The sulfonic acid-functionalized acids Amberlyst 15 and Dowex 50WX4 showed a similar trend. The activity during the isomerization was good but ether formation was extensive (Table 3.1, Entry 3–6). As presented before, a two-phase system consisting of water, as a nucleophile, seemed to

OHor

OH

(S)-7b(rac)-6b

OH

+

O

(rac)-7b 8

heterogeneous acid

solvent, 5 h, rt

��

be a possibility in order to trap the cationic species and avoid formation of the dehydration product. An iso-octane/water mixture where the two phases are well separated was chosen. Therefore, the acids were tested with this system and Amberlyst 15 provided good results with only a few percent of the side-product after a reaction time of 5 h. The racemization of (S)-7b also proceeded well with full recovery of the racemic sec-alcohol (rac)-7b. Dowex 50WX4 gave similar results albeit with a slower reaction rate, but without any ether formation. Since both of the desired reactions proceeded well with Dowex 50WX4, further studies were pursued with this resin.

During the acid-catalyzed isomerization and racemization the importance of water as co-solvent was clearly understandable, but the enzymatic kinetic resolution of sec-alcohol has to be performed in pure organic solvents other-wise the reverse hydrolysis reaction of the lipase will take place. In the litera-ture, it is known that this problem can be overcome by keeping the enzymes in a closed compartment in the organic layer.[58] We realized that a simple nylon material can serve for this function rather easily without using any spe-cial equipment. Therefore, a vessel containing the enzyme beads functioned like a tea bag during our reaction, which maintains the enzyme at the organic layer. The regular fine-woven nylon mesh material proved to be good enough for this purpose and worked well in combination with the immobilized Novo-zym 435 (CalB) enzyme.

The key of the migratory DKR procedure is maintaining a good separation of the reaction steps. The water solubility of each substrate has a crucial role and avoids decomposition and side-product formation. The acyl donor of the kinetic resolution has two important roles, first of all it has to be hydrophobic enough to stay in the organic layer and during the reaction the produced ester has to have high lipophilicity to avoid ending up in the acidic water environ-ment. Both of these effects can be relatively easily overcome by the use of an ester with a long carbon chain as the, acyl donor; however, the chain length of the acyl donor has an influence on the kinetic resolution process. During a simple KR this effect has only a marginal effect on the selectivity and the relative reaction rates (k(R)/k(S) = E value) were always greater than 200. How-ever, during the combined process with Dowex 50WX4, the optical purity of the products was highly dependent on the structure of the vinyl ester. The most commonly used, and partially water-miscible, vinyl acetate produced the ace-tyl ester with 81% ee (Table 3.2, Entry 1). As we had predicted, an increase of the lipophilicity of the acyl donors would overcome this challenge. The vinyl caprate 9 donor gave excellent enantioselectivity (99% ee) with a good isolated yield (Table 3.2, Entry 3).

���

Table 3.2. Acyl donor screening within the two phase system.[a]

Entry R– Yield [%] ee [%]

1 CH3 62 81

2 C3H7 85 95

3 C9H19 76 99

Several different tertiary carbinols were subsequently synthesized in order

to study the generality of the developed approach. The synthesis of the starting materials involved addition of organolithium (RLi) or organomagnesium nu-cleophiles (RMgX) to allow for the formation of the desired enones. The pro-duced racemic allylic alcohols were tested by our migratory DKR procedure (Figure 3.1).

Figure 3.1. Substrate scope of the migratory DKR reaction. Reaction conditions: (rac)-6 (0.5 mmol), 9 (2.5 mmol), Dowex 50WX4 (100 wt%), iso-octane (20 mL), water (20 mL), room temperature, 300 rpm. Isolated yields after flash chor-matogarphy; ee determined by HPLC equipped with chiral column.

OH

(rac)-6b

Acyl donorNovozym 435 (CalB)

Dowex 50WX4

i-octane/H2O18 h, rt

O R

O

CH3

(R)-10a 82%99% ee, 22h

O

O

C9H19 OO

O

O O O O

O

C9H19

O

C9H19

C9H19

O

Cl

O

C9H19

O

C9H19

O

C9H19

O

C9H19

O C9H19

O

(R)-10b 72%99% ee, 22h

(R)-10c 85%84% ee, PSL, 22h

(R)-10d 67%99% ee, 120h

(R)-10e 81%99% ee, 96h

(R)-10f 90%99% ee, 120h

(R)-10g 89%99% ee, 120h

(R)-10h 26%99% ee, 96h

(R)-10i 82%99% ee, 22h

OH

(rac)-6a-i

+O C9H19

O

O

O

C9H19

R1

R2

nn

lipaseDowex 50WX4

iso-octane/H2O, rt

(R)-10a-i9

��

The five-membered carbinol (rac)-6a was more sensitive to the acidic treatment than the cyclohexanol-based derivatives and extensive decomposi-tion was observed under our reaction conditions. However, if the amount of the acidic resin was reduced, a sufficient reaction profile was obtained and the enantiomerically pure conjugated enyne (R)-10a was isolated in 82% yield. Similarly, as (rac)-6a, the related five-membered ring (rac)-6c required a re-duced amount of the acid, which allowed (R)-10c to be isolated in high yield and good selectivity. If the ring-size was expanded to seven-membered rings, the cycloheptenol (rac)-6d isomerized significantly slower than (rac)-6a and a longer reaction time had to be applied to achieve a higher yield. Moreover, after a reaction time of 4–5 days, different substituted cyclohexenyl substrates could be successfully applied to provide the corresponding product (R)-10d in good to excellent yields with high enantioselectivity. Substrate (rac)-6h with an aliphatic chain as substituent gave low yield during the DKR due to the slow isomerization rate to the sec-alcohol. Finally, an acyclic allylic car-binol (rac)-6i was also evaluated to give the product (R)-10i in 82% yield with 99% ee.

Due to the fact that both of our catalysts are heterogeneous, recycling of the catalysts seemed to be feasible, especially since product isolation was sim-ple from the developed system. Indeed, it was possible to reuse both the Dowex 50WX4 resin and the Novozym 435 (CalB) for our migratory DKR six times without any noticeable loss of activity or selectivity. After each re-action, the organic layer was separated and the aqueous layer was extracted to remove all remaining product. After this step, a new batch of substrate in iso-octane was added and the teabag containing the CalB catalyst were placed into the organic layer. All along the recycling study, the conversions for (R)-10b were retained between 80% and 90% with excellent ee (>99%) over 22 h re-action time

����-�����������Tertiary allylic carbinols were successfully applied in a migratory DKR sys-tem under mild acidic conditions on a preparative scale to provide enantiopure carbocyclic esters. The polymeric resin-based heterogeneous acids afforded an easy to use approach for the 1,3-isomerization of the substrates and race-mization of the rearranged alcohol. The ad hoc designed tea bag for the No-vozym 435 was compatible with the two-phase solvent system. This approach provides several new and valuable carbocyclic building blocks for future ap-plications.

���

-��+ �����������'.(!���/���0��/��&"����������" �������� ������"��������������� �'��������������1(�

)���*������������The development of novel heterogeneous catalysts has been dramatically in-creased by the advances in material science and nanotechnology during the last decade. Basically, any type of catalytically active species can be effi-ciently immobilized on heterogeneous surfaces, such as organocatalysts[96], enzymes,[97] organometallic complexes[98] and metal nanoparticles.[99] The number of available solid supports is also increasing with tunable morphology and surface functionalization, including carbon-based materials[100], metal-or-ganic frameworks,[101] metal oxides,[102] polymers,[103] and silica-based mate-rials.[104] The ideal support material should have high stability within a wide range of reaction conditions together with a large internal surface area to ac-cept high loadings of the catalytic species. In general, the supports are inert, which ensure that side reactions do not take place during the reaction. The main problem associated with heterogeneous/supported catalysts is the large mass of non-catalytically-active material present by the support material com-pared to the catalytic species. Some of the next generation of supports have been designed to be able to take part in reactions themselves by exhibiting various form of catalytic activities which can be harnessed during synergistic cascade reactions. For example, the acidic or basic supports can be further functionalized with metal nanoparticles, which are subsequently applied in different redox-condensation reaction processes.[105] Unfortunately, the possi-ble reactions are rather limited with these types of acidic/basic supports as their main function is to promote condensation-type reactions. A different ap-proach to design bifunctional catalysts was disclosed by the Zare[106] and Pal-omo[107] groups where pure enzymes were converted to heterogeneous mate-rials upon treatment with metal salts. Surprisingly, the enzymes retained their catalytic activity in these biohybrid materials. The Palomo group was able to apply the biohydrids in several different reactions and was able to show that both the CalB and the Pd-nanoparticles function separately and synergistically during a DKR reaction. The metal salt-based heterogenization of enzymes has been applied by the last literature examples, which concept-wise is not far from the cross-linked enzyme aggregates (CLEA) method, where the enzymes

���

creates a heterogeneous composite based on aggregation and a cross-linker agent. This procedure is well established and has successfully been applied to several different enzymes also on an industrial scale. The major advantage of the CLEA procedure is the reduction of the catalytically inert solid support, which is usually about 90–99 wt% of the heterogeneous catalyst. However, a possible drawback of such a system is the less profound enzyme stability and the diffusion problem associated with the entangled structure.

Our vision was to combine the two approaches mentioned above to be able to improve the applicability and scalability of the biohybrid synthesis. In par-ticular we hypothesized that aggregated enzymes generated through the CLEA method could be a good heterogeneous scaffold for metal nanoparticles. This would constitute an alternative to our earlier work, where a Pd-CalB hybrid was prepared, in which the Pd nanoparticles were attached to the cavities of a siliceous mesocellular foam (MCF) and the enzyme (CalB) was linked to the free sites through glutaraldehyde.[18] Using this new concept, the Pd immobi-lization would be more challenging to control; however, on the other hand by circumventing the need of the MCF material it would be possible to remove most of the non- catalytically productive mass with this biohybrid catalyst (hereon referred to as Pd(0)-CalB CLEA catalyst).

)���.���������'����%�/01 -��2�-3������������The synthetic protocol used for the synthesis of our Pd(0)-CALB CLEA cat-alyst was inspired by a previously reported CLEA protocol developed by the Sheldon group.[108] Here, commercially available CalB solution was combined with glutaraldehyde and Na(CN)BH3 in iso-propanol. Subsequently, the ob-tained solid pellet was washed with neutral phosphate buffer, acetone, and diethyl ether using centrifugation technique. The washing step with buffer was particularly important since it removed the iso-propanol solvent, which could otherwise inhibit the enzyme. The prepared CalB CLEA catalyst, in phosphate buffer (pH 7.2, 100 mM), was then dissolved to an MeCN solution containing Pd(OAc)2. The resulting solution was stirred overnight (16 h) at room temper-ature. Thereafter, the Pd(II)-CalB CLEA was separated by centrifugation, and the non-coordinated Pd(II) species was washed out of the solution using neu-tral phosphate buffer. The centrifuged Pd(II)-CalB-CLEA was dissolved in phosphate buffer and the bound Pd(II) was then reduced to Pd(0) nanoparticles (Scheme 4.1)

���

Scheme 4.2. Cascade reaction catalyzed by Pd(0)-CalB CLEA.

The enantioselective transesterification catalyzed by CalB is a well-estab-lished protocol but lactones have not been employed as acyl donors in KRs previously. The enantiomerically pure 1-phenylethyl 4-oxopentanoate prod-ucts (R)-14a-f are of great synthetic value and can easily be further function-alized to more complex target compounds. Employing a functionalizable acyl donor in a kinetic resolution is known in the literature. Akai and Kita used unsaturated ethoxyvinyl esters in the DKR of 3-vinylcyclohex-a-en-1-ols, which subsequently underwent an intramolecular Diels–Alder reaction.[110]

The Pd(0)-CalB CLEA catalyst showed high activity and selectivity during the cascade transformation, the optimal reaction conditions were optimized (Table 4.1).

O

OH +

OH Pd(0)-CalB CLEATEA O

O

O+

OH

O

OPd

CalB CLEA

toluene (dry)60 ºC

(rac)-13a11

OH

12

(R)-14a (S)-13a

���

Table 4.1. Optimization of reaction conditions for the cascade reaction.

Entry 11 [mmol]

(rac)-13a [mmol]

Catalyst [mg]

TEA [mmol] T [ºC] Conv.

[%] 1[a] 0.2 0.1 5.0 0.1 60 46 2[a] 0.2 0.1 2.5 0.1 60 31 3[b] 0.2 0.1 5.0 0.1 60 50 4[b] 0.2 0.1 2.5 0.1 60 47 5[b] 0.05 0.1 2.5 0.1 60 29 6[b] 0.1 0.1 2.5 0.1 60 49 7[b] 0.1 0.1 2.5 0.075 60 48 8[b] 0.1 0.1 2.5 0.075 40 15 9[b] 0.1 0.1 2.5 0.075 50 41 10[c] 1.0 1.0 25 0.75 60 50

[a] Conditions: Pd(0)-CalB CLEA and substrate 11 were dispersed in 0.5 mL tolu-ene, and then substrate (rac)-13a and TEA were added and the reaction was sealed and stirred at the indicated temperatures. The conversion was determined by 1H NMR with 1,3,5-trimethoxybenzene as internal standard and the ee was 99% in every reaction, as determined by chiral GC. [b] Conditions: every condition param-eters were similar instead of the 0.1 mL toluene. [c] Conditions: Pd(0)-CalB CLEA and substrate 11 were dispersed in 0.75 mL toluene, and then substrate (rac)-13a and TEA were added and the reaction was sealed and stirred at the indicated tem-peratures. The conversion was determined by 1H NMR with 1,3,5-trimethoxyben-zene as internal standard and the ee was 99% in every reaction, as determined by chiral GC.

Based on our previous study, TEA provided the best results during the cy-

cloisomerizaton reaction and therefore only the amount of TEA was screened. Toluene was chosen as solvent since it has shown good compatibility with lipase enzymes previously. The first step was to investigate the loading of the catalyst in dry toluene (0.5 mL) using 0.1 equiv of TEA and 2 equiv of 11, with respect to (rac)-13a at 60 ºC. The initial 5 mg of catalyst loading provided 46% conv., while the use of 2.5 mg catalyst gave only 31% conv. Based on the substrate and reagent’s high solubility in toluene, the concentration could be increased and by utilizing 5 mg of catalyst in 0.1 mL toluene it was possible to achieve full conversion into (R)-14a. Decreasing the loading of Pd(0)-CalB CLEA to 2.5 mg was also evaluated under these reaction conditions and gave almost similar conversion (47%). The next step was to control the amount of

OH

OH

OPd(0)-CalB CLEA, TEA

toluene, 3h

O

OOH

+O

+

11 (rac)-13a (R)-14a (S)-13a

� �

4-pentynoic acid 11. The use of 1 equiv of 11 worked well; however, decreas-ing the loading to 0.5 equiv dramatically decreased the conversion. The amount of TEA could also be reduced, which is important during the recycling of the catalyst since the Pd could potentially coordinate to TEA, and this co-ordination could lead to increased metal leaching. Unfortunately, the reaction had to be conducted at 60 ºC since temperature decrease of the temperature to 40 or 50 ºC resulted in incomplete conversion. To conclude, the best results were to be: 1.0 equiv of 4-pentynoic acid 11, 0.75 equiv of TEA and 2.5 mg of catalyst in dry toluene (1 mL) at 60 ºC.

To demonstrate the scope of the method, a few differently substituted 1-phenyethanol derivatives (rac)-13a-f were tested with different steric and electronic properties (Figure 4.2). The p-Me substituted 1-phenylethanol (rac)-13b gave similar results as the model substrate (rac)-13a and gave 46% isolated yield of the enantiopure ester (R)-14b. The introduction of a strongly electron-donating substituent lowered the yield slightly to 42% of 1-(4-meth-oxyphenyl)ethanol (R)-14c. On the other hand, the more sterically demanding 1-(naphthalene-2-yl)ethanol (rac)-13d worked more efficiently and gave 45% isolated yield of (R)-14d. The substrates that contain electron-withdrawing groups, such as 1-(4-chlorophenyl)ethanol (rac)-13e and 1-(4-nitro-phenyl)ethanol (rac)-13f, were also compatible, providing the products in 43% and 45% yield, respectively. For 1-(4-nitrophenyl)ethanol, the substrate alcohol (S)-13f was also isolated in 44% yield. All of the above-mentioned reactions showed exclusive selectivity towards the (R)-products with >99% ee.

Figure 4.2. Substrate scope of the cascade reaction.

O

O

O

O

O

O

H3C

O

O

O

Cl

O

O

O

H3CO

O

O

O

O

O

O

O2N

OH

OH

OPd(0)-CalB CLEA, TEA

toluene, 3h, 60 ºC

O

OOH

+O

+

RR

R

11 (rac)-13a-f (R)-14a-f (S)-13a-f

(R)-14a47%

(R)-14b46%

(R)-14c42%

(R)-14d45%

(R)-14e43%

(R)-14f45%

���

Subsequently, the recyclability of the Pd(0) CalB-CLEA catalyst was stud-ied using the model substrate (rac)-13a. Here, six consecutive cycles were carried out. After a reaction time of 3 h, the mixture was separated by centrif-ugation and the catalyst was washed two times with dry toluene. The isolated yields of each run were within the range of 42–46% and the optical purity was in all cases >99%. To quantify the degree of metal leaching, the supernatant toluene from the reaction of substrate (rac)-13a was subjected to elemental analysis (ICP-OES) and it was found that the Pd leaching was <1 ppm, demon-strating that the Pd nanoparticles are indeed strongly bound in the enzyme-aggregate structure.

)�)�"���������������������4�$5"��'�(�������������������After applying the Pd(0)-CalB CLEA catalyst in the cascade reaction of sec-alcohols, it was speculated if the catalyst could be used for the DKR of ben-zylic primary amines. The first attempt was to test the racemization of amines by our catalyst. For this purpose, 1-phenylethylamine (rac)-15a was chosen as the model substrate, which was not racemzied by the Na(CN)BH3 reduced Pd(0)-CalB CLEA during the initial tests. However, the application of stronger reducing agent (NaBH4), provided Pd(0)-CalB CLEA, that showed initial racemization in toluene (Table 4.2). The supporter free Pd(0)-CalB CLEA catalyst turned out to be less efficient than the previously reported Pd(0)-Amp-MCF.[111] After 6 h reaction time, the ee of (R)-15a dropped to 31% at 90 ºC under 1 atm of hydrogen, and 29% of byproduct 16 was formed (Table 4.2). The side-product formation was expected, based on the previous reports of DKR on amines catalyzed by heterogeneous catalysts (Scheme 4.3).[111] The racemization of amine proceeds via formation of an imine that has the possibility to undergo condensation with a second molecule of the starting amine, resulting in the release of ammonia. The imine formation can be reduced by dihydrogen and split to an amine and the ethyl benzene 16.

Scheme 4.3. Possible side-product formation during the racemization of amines.

The concentration and temperature were reduced to circumvent the large

amount of side product formation, but these two modifications also reduced the rate of racemization. 1,4-Dioxane and THF solvent was used to evaluate the solvent effects, it turned out that at 90 ºC, 1,4-dioxane gave slower rate of

Ar

NH

Ar

NH2 -H2 Ar

NH2

-NH3

Ar

N

Ar+H2

Ar

HN

Ar

Ar

NH2Ar+

(rac)-15a (rac)-15a 16

��

the reaction, but the side product formation was also lower. The racemization in THF showed significantly slower racemization without any by-product for-mation.

Table 4.2. Optimization of the racemization of (R)-15a.[a]

Entry Solvent T [ºC] t [h] ee [%][b] 16 [%][c]

1 toluene 90 6 31 29 2 1,4-dioxane 90 6 62 9

3 THF 90 6 58 3

4 THF 60 6 94 <1

5 1,4-dioxane 60 6 96 <1 [a] (R)-15a (0.25 mmol, 0.166 M), Na2CO3 (0.5 mmol) and Pd(0)-CalB CLEA (22 mg, 4.4 mol% Pd) was mixed and stirred at 90 ºC under partial pressure of hydro-gen. [b] Determined by chiral-GC/FID. [c] Determined by 1H NMR using 1,3,5-trimethoxybenzene as internal standard.

To develop the DKR procedure, 1-phenylethylamine (rac)-15a was used

in combination with ethyl methoxyacetate 17 as acyl donor in different or-ganic solvents. It was demonstrated that the Pd(0)-CalB CLEA catalyst, re-duced by NaBH4, was able to efficiently and selectively catalyze the DKR of (rac)-15a to give the chiral amide (R)-18a (Table 4.3).

NH2 Pd(0)-CalB CLEANa2CO3, H2

Solvent, T [ºC]

(R)-15a 16

NH2

(rac)-15a

+

���

Table 4.3. Optimization of the DKR of 1-phenylehtylamine (rac)-15a.

During the initial optimization reactions, 2.0 equiv of ethyl methoxyacetate

17, 22 mg of Pd(0)-CalB CLEA and 1.1 equiv of Na2CO3 were applied in different dry solvents under atmospheric hydrogen gas. The reaction was car-ried out in toluene at 90 ºC, delivering 81% NMR yield with 95% ee of (R)-18a. Unfortunately, the remaining mass balance of 19% was found to be the side-product, ethyl benzene 16.

After this observation, different solvents were tested in order to increase the enantioselectivity and reduce the amount of side-product 16. The reaction in THF was carried out at 60 ºC and afforded 60% NMR yield of (R)-18a without any side product formation. To be able to increase the NMR yield of

NH2

+O

OO

HN

OO

+

Pd(0)-CalB CLEANa2CO3, H2

Solvent, T [ºC] / 18 h

(rac)-15a (R)-18a 1617

Ent. Solvent Base Additive T [ºC]

(R)-15a[a] [%]

ee[b] [%]

16[a] [%]

1 Toluene Na2CO3 None 90 81 95 19 2 THF Na2CO3 None 60 60 99 <1 3 THF Na2CO3 None 90 65 99 <1 4 1,4-Dioxane Na2CO3 None 90 82 96 5 5 2-MeTHF Na2CO3 None 90 69 89 <1 6 Bu2O Na2CO3 None 90 80 77 13 7 CPME Na2CO3 None 90 72 93 8 8 Anisole Na2CO3 None 70 78 95 7 9 TBME Na2CO3 None 60 61 97 <1 10 THF K2CO3 None 60 45 99 <1 11 THF NaOAc None 60 53 99 <1 12 THF NEt3 None 60 66 91 <1 13 Toluene None None 90 67 88 25

14[c] Toluene Na2CO3 Fe(OAc)2 90 85 50 13 15[d] Toluene Na2CO3 BPh3 90 72 80 5 16[e] Toluene Na2CO3 NH4HCO2 90 43 66 3 [a] Determined by 1H NMR using 1,3,5-trimethoxybenzene as internal standard. [b] Determined by chiral-GC/FID. [c] 20 mol% of Fe(OAc)2. Reaction in absence of catalyst led to 11% of amide and 6% of side product after 22 h. [d] 10 mol% of BPh3. [e] Reactions using 4.0 equivalents of NH4HCO2.

��

(R)-18a, the reaction temperature was increased to 90 ºC but only a 5% in-crease of the yield of (R)-18a was obtained. 1,4-Dioxane provided good re-sults with 82% NMR yield and with 96% ee along with only 5% of ethyl ben-zene 16 after 18 h. Other solvents, such as 2-methyltetrahydrofuran (2-MeTHF), dibutyl ether (Bu2O), cyclopentyl methyl ether (CPME), anisole and tert-butyl methyl ether (TBME) were also evaluated but resulted in all cases in lower yields and selectivity (Table 4.2, Entries 1–9). Next, different bases were tested. The use of K2CO3 or NaOAc afforded lower conversions than Na2CO3. TEA showed a good yield but gave only 91% ee. The most desired solvent was toluene, since it is not coordinating to Pd, and it was tested without any base at 90 ºC, yielding 67% of (R)-18a with 25% of the side-product 16. The DKR reaction was also run with different Lewis-acidic additives in order to reduce the formation of 16. Unfortunately, the presence of Fe(OAc)2, BPh3 or NH4CO2 decreased the activity and selectivity of the hybrid catalyst.

For the solvents that displayed the best activity (THF, 1,4-dioxane and tol-uene) a further screening of the substrate concentration revealed that increas-ing this to 0.25 M significantly improved the results (Table 4.4). The reactions were finished in a shorter period of time with only a slightly decreased enan-tioselectivity and the amount of side-product 16 was only slightly increased from 1% to 5%. Increasing the equivalents of the acyl donor (ethyl methoxy-acetate) from 2.0 to 5.0 equiv did not have any significant effect on the out-come of the reaction. A further increase of the amine to 0.5 M in THF and 1,4-dioxane decreased the ee of the product and did not improve the yield. Finally, the catalyst loading was also investigated for the DKR of (rac)-15a in 1,4-dioxane at 90 ºC. Here, 11 mg of the catalyst (2.2 mol% of Pd) was capable to convert the substrate, delivering the product in 82% NMR yield with 3% of the side product. The double amount of Pd(0)-CalB CLEA (4.4 mol% Pd) in-creased the NMR yield to 87% and 5% of ethyl benzene 16 was detected. As a last test, 44 mg of catalyst (8.8 mol% of Pd) was applied, providing 83% of (R)-18a. The loading of the catalyst did not have any effect on the enantiose-lectivity.

���

Table 4.4. Optimization of concentration and catalyst loading in the DKR[a]

Entry Solvent (rac)-15a

Pd [mol%]

(R)-18a[a]

ee [%][b] 16 [%][a]

1 1,4-dioxane 0.25 4.4 87 94 5

2 THF 0.25 4.4 82 98 5

3 toluene 0.25 4.4 77 93 18

4 THF 0.5 4.4 83 95 6 5 1,4-dioxane 0.5 4.4 90 92 10

6 1,4-dioxane 0.25 2.2 82 93 3

7 1,4-dioxane 0.25 8.8 83 93 17

[a] Determined by 1H NMR using 1,3,5-trimethoxybenzene as internal standard. [b] Determined by GC/FID equipped with chiral column.

The final conditions for the DKR were chosen to 0.25 M (rac)-15a sub-

strate, 4.4 mol% of the catalyst, Na2CO3 as base on either THF or 1,4-dioxane at 90 ºC. Several different substituted racemic amine derivatives were inves-tigated during the substrate scope (Figure 4.3).

NH2

+O

OO

HN

OO

+

Pd(0)-CalB CLEANa2CO3, H2

Solvent (x M)90 ºC / 18 h

(rac)-15a (R)-18a 1617

���

Figure 4.3. Substrate scope for the DKR of amines (rac)-15a-g. [a] The reaction per-formed in 1,4-dioxane. [b] The reaction performed in THF.

When 1,4-dioxane was used as solvent, the conversions were higher but

side-product formation was also increased. THF as solvent increased the se-lectivity and led to <3% of side-product formation but the NMR yields were generally lower. Great results were obtained in the DKR of (R)-2-methoxy-N-(1-phenylpropyl)acetamide (R)-18g in 1,4-dioxane and (R)-2-methoxy-N-(1-(naphthalene-1-yl)ethyl)acetamide (R)-18e in THF, with 80% NMR yield and 97% enantioselectivity. The p-fluoro substituted compound (rac)-15c gave relatively low activity; however, 1,2,3,4-tetrahydronaphthalen-1-amine (rac)-15f afforded 94% of the corresponding product and 86% ee. In the tandem reaction described above involving generation of a functionalized acyl donor and KR of sec-alcohols, the Pd(0)-CalB-CLEA catalyst was easily and effi-ciently recyclable. It was therefore of interest to investigate if it was also pos-sible to recycle the hybrid catalyst in the DKR reaction (Figure 4.4 and 4.5).

R2

NH2 Pd(0)-CalB CLEA (4.0 mol%)Na2CO3 (1.1 eq), H2

solvent (1mL), 90 ºCOO

O+

R2

HN

OO

R1

R1

HN

OO

H3C

HN

OO

F

HN

OO

H3CO

HN

OO

(R)-18b[a]

yield: 84%ee 94%

(R)-18c[b]

yield: 71%ee 95%

HN

OO

(R)-18a[a]

yield: 87%ee 94%

(R)-18d[b]

yield: 80%ee 91%

(R)-18e[b]

yield: 80%ee 97%

(R)-18g[a]

yield: 80%ee 91%

17, 2.0 eq

HN

OO

HN

OO

(R)-18f[b]

yield: 94%ee 86%

(rac)-15a-g (R)-18a-g

���

Figure 4.4. Recycling of the Pd(0)-CalB CLEA catalyst by the DKR of (rac)-15a in toluene.

The recycling was clearly more challenging than the cascade reaction with

sec-alcohols. The protocol was tested with two solvents, toluene and 1,4-di-oxane. For this purpose, the model substrate (rac)-15a was used at 90 ºC in each solvent. After a catalytic cycle, the heterogeneous mixture was separated by centrifugation and the catalyst was recovered. It was washed using the same procedure as was used in the synthesis of the catalyst (phosphate buffer, acetone and diethyl ether) in order to remove all residual substances. The ac-tivity of the catalyst in toluene slightly decreased from 78% to 68% after the fifth cycle but the enantioselectivity was only maintained until the third cycle, after which it dropped to 78% (Figure 4.4). Although, the recycling in 1,4-dioxane was less successful, the ee dropped already after the first reaction and kept this trend during the cycles. After the fourth reaction, the ee was 68% and the NMR yield was only 63% (Figure 4.5).

���� ���� ��� ���� ����

���� ���� ���� ��� ���

�

��

��

��

��

���

� � � � ���� ������ ��!� �"�� ��� �#�$

���

Figure 4.5. Recycling of the Pd(0)-CalB CLEA catalyst by the DKR of (rac)-15a in 1,4-dioxane.

)�6�-�����������������������������'�%�/01 -��2�-3���Elemental analysis (ICP-OES) was performed on the hybrid Pd(0)-CalB CLEA catalyst to be able to determine the palladium content in the heteroge-neous system. It was found that an average of 4.7 wt% of the catalyst is Pd. After each recycling studies, the supernatants were analyzed to determine the leaching of Pd from the catalyst. During our first applied protocol, the loss of Pd was always <1%, which means that the Pd is confined within the hetero-geneous structure. Different microscopy techniques were applied to analyze the structure and morphology. Using scanning electron microscopy (SEM), the morphological differences were compared between the non-Pd-function-alized and Pd(0)-functionalized CalB CLEA catalyst (Figure 4.6). It has been reported by the Sheldon group that CalB CLEA has a well-defined spherical morphology (Figure 4.6, a), and for our Pd-functionalized catalyst this mor-phology was changed to a more amorphous one (Figure 4.6, b). This result clearly shows that the Pd nanoparticles are incorporated and bound to the ma-trix but the incorporation and morphological changes did not significantly re-duce the enzymatic activity or selectivity.

��� ��������

����

��� ��� ��� ���

�

��

��

��

��

���

� � � ���� ������ ��!� �"�� ��� �#�$

���

cases, the recyclability of the catalyst was examined and showed good results without any leaching of Pd nanoparticles (<1 ppm). The catalyst was charac-terized by several surface techniques including EXAFS and XPS analysis.

��

2���������0������ ������"����3������ ������� ����%�4��� �3������"�����&�� ��"��������0���������'�����1(�

6���*������������The enantioselectivity of the transacylation reaction is controlled by R- or S-selective enzymes, where the lipases show R-selectivity and the serine prote-ases favor the S-enantiomer. For the S-selective KR, subtilisin Carlsberg (SC) and the rationally designed CalB W104A mutant[112], which has been engi-neered to display the reverse selectivity, have been commonly used. SC is naturally secreted from Bacillus licheniformis[113] and used as an additive in various detergent products.[114] From a synthetic chemistry perspective, SC has been successfully applied in the preparation of enantiomerically enriched S-alcohols through DKR using ruthenium-catalyzed racemization.[115]

Both SC and CalB originate from the same enzyme family, called hydro-lases. More specifically, these two are serine hydrolases, which operate through identical reaction mechanism with the Asp-His-Ser catalytic triad and a stabilizing oxyanion hole. However, the enzymes have different folds and their active sites are the mirror images of each other.[116] According to Kazlauskas’ rule, SC reacts faster if the medium-sized group (M) of the sec-alcohol is fitted in the small (S1′) pocket and the larger group (L) is directed towards the solvent. For the SC enzyme, the methyl substituent is pointing toward the S1′ pocket and the phenyl unit being the large group is pointing towards the solvent (Figure 5.1). Unfortunately, in all cases SC shows lower enantioselectivity and a lower tolerance towards organic solvents compared to CalB.

���

mutagenesis.[119] However, there are only a few studies on the protein engi-neering of SC towards the transacylation reaction.[120] This work concerns the organic solvent sensitivity and improving the SC enzyme by protein engineer-ing.

6���"���������������������1-phenylethanol (rac)-13a was chosen as model substrate and during the ini-tial experiments the enantioselectivity of SC (AGN35600.1[121]) was tested in THF (Scheme 4.1). Different enzyme preparations were tried, such as free form, free form but treated with surfactants (Brij56® and octyl-�-D-glycopy-ranoside), immobilized on EziG 2® and treated with surfactants, as well as immobilized on Accurel MP1000 and treated with KCl.

Scheme 5.1. Model reaction.

After the different preparations, the enzymes were lyophilized for 24 h,

and all reactions were performed under an argon atmosphere. The free enzyme in THF solution showed an E value close to 4 during our model reaction setup, which is in accordance with previous literature reports.115 Pre-treatments quickly showed to be essential since a ten-fold increase was detected during the first tests, increasing the E value to about 40.

The free enzymes and the immobilized ones were studied without any sur-factants or KCl treatment, and none of them showed any activity towards 1-phenylethanol. However, all of them were active in a proteolytic activity assay using N-succinyl-Ala-Ala-Pro-Phe p-nitroanilide. Based on the moderate E values obtained through immobilization, it was decided to apply a protein en-gineering approach guided by molecular modelling and literature datasets. For the modelling of SC, the 1YU5[122] (PDB ID) crystal structure was used. The S and R configurations of the tetrahedral intermediates were built into the en-zyme structure and studied. The methyl group of the substrate should be point-ing toward the S1′ pocket, and because of its size the phenyl group should be directed toward the solvent. The opposite position should not be possible since the medium pocket is too small to host the phenyl ring. The size restriction of the S1′ pocket is dependent on a methionine (M221) residue, which corre-sponds to M222 in 1YU6. The reason for the shift is the different way of counting the amino acid residues, in the structure it started from residue 56,

OH

+O

OSC wt or variants

Na2CO3

THF

O

OOH

+

(S)-20a (R)-13a(rac)-13a 19

���

relative to that of the mature protein sequence of SC (AGN35600.1). Earlier studies on subtilisin Bacillus amyloliquefaciens (SBA) demonstrated that the sulfur atom in the methionine residue next to the catalytic serine was prone to oxidation. The produced sulfoxide yields an inactive enzyme because of the destabilization of the transition state intermediate.[123] This is most likely the reason why the reactions required inert atmospheric conditions. The Estell group performed mutations at the M222 position in SBA and created 19 mu-tants where two of these variants (M222A and M222S) presented higher re-sistance towards oxidation.[123a] It was also shown that the mutation of M222F in subtilisin BPN′ resulted in increased peptide ligation activity in organic sol-vents. Since this position seemed to have a vital role for the catalysis and sta-bility, we also chose this site as the first mutation point. During the first round, five variations of the M221 position were chosen and mutated to M221A/C/S/F/W by site-directed mutagenesis. The mutants were tested with skim milk agar plates and a proteolytic activity assay showed positive results. However, the proteolytic activity was low for M221F and M221W. Proteolytic activity is a highly important test where inactivity means that the pro-peptide cannot be cleaved off after secretion and the enzyme will thus remain inac-tive.[124]

Of the different produced variants, two of the them (M221C and M221F) showed higher E values for our desired transacylation reaction than the natural protein (Table 5.1).

Table 5.1. Enantioselectivities of immobilized[a] SC wild type and mutants for the transesterification of (rac)-13a in THF after 48 h.[b]

SC variant eeS [%][c] eeP [%][d] Conv. [%] E (S) wt 3 95 3 39

M221A 0 87 0.2 n.d. M221C 6 99 5 >100 M221F 5 97 5 67 M221S 3 98 3 41 M221W 0 86 0.1 n.d. G165L 8 95 8 46

G165L/M221C 18 98 16 >100 G165L/M221F 25 98 20 >100 G165L/M221S 6 99 6 >100 [a] The enzymes were immobilized on EziG 2 material and co-lyophilized with surfactants. [b] The immobilized enzymes (10 mg, EziG 2), Na2CO3 (53 mg), vinyl butyrate 19 (75 mM), 1-phenylethanol (rac)-13a (50 mM), dodecane (100 mM) were added into 1 mL of dry THF. The reaction was shacked at 25 ºC under Ar

���

atmosphere. The samples were analyzed by GC equipped with chiral column. [c] Enantiomeric excess of substrate. [d] Enantiomeric excess of product.

The variant M221C afforded a synthetically useful E value (> 100), while

M221F had an E value of 67; however, the conversions remained low (5%) after 48 h. The E value of variant M221S was similar to that of the wild type and no E values were determined for M221A and M221W because of the low conversion (< 1%) after 48 h. Due to the low conversions, new mutations that retain a high degree of enantioselectivity were selected from a structure- and literature search. The G165 variant (G166 in 1YU6) was found, which was a part of a previous mutagenesis study in SBA. This residue has an effect on the size of the S1′ pocket and the natural peptide amino acid side-chain prefer-ence.[125] Also, the G165L/I/Y mutants showed improved perhydrolysis activ-ity by increasing the substrate specificity towards methyl esters.[126]

The above-mentioned literature data and molecular modelling guided us to design a variant at position G165. The G165L point-mutated variant was gen-erated by site-directed mutagenesis and within the enzyme preparation, it was tested for our desired transacylation reaction. The results were slightly better than the wild type enzyme, providing an E value of 46 at 8% conversion.

The two point mutations, G165L and M221C/F/S, were combined and fur-ther studied for the model reaction. The double mutant showed good E values (> 100) and the variant G165L/M221F gave the highest conversion (20%) af-ter 48 h reaction time. The S- and R-substrate specificities of the mutants were evaluated and compared to the natural protein (Table 5.2). To increase the E value towards (S)-1-phenylethanol (S)-13a, the substrate specificity (kcat/KM) towards (S)-20a should be improved. The two variants SC M221C and G165L/M221F showed the desired increased specificity towards (S)-13a, and reduced specificity towards (R)-13a. In contrast, SC G165L/M221C and G165L/M221S gave higher specificity towards both of the enantiomers.

���

Table 5.2. Kinetic measurements of the immobilized SC wild type and variants for transacylation of (S)- and (R)-13a.[a]

SC variant (kcat/KM)R [s-1]

(kcat/KM)S [s-1] E (S) (KM)R [M] (KM)S [M]

wt 0.064 2.5 39 0.41 0.25 M221C 0.042 5.9 140 2.1 0.38

G165L/M221C 0.11 15 140 3.4 0.51

G165L/M221F 0.047 7.7 160 >6.8 0.64

G165L/M221S 0.22 13 57 6.8 1.4 [a] The kinetic was performed under pseudo-one-substrate conditions. The concen-tration of vinyl butyrate 19 was kept constant at 200 mM. The immobilized en-zymes (10 mg on EziG 2), Na2CO3 (53 mg), vinyl butyrate 19 (200 mM), 1-phe-nylethanol (rac)-13a (0.1–1.0 M), dodecane (100 mM) were added into 1 mL of dry THF. The reaction was shacked at 25 ºC under Ar atmosphere. The samples were analyzed by GC equipped with chiral column.

Since the SC G165L/M221F variant showed the highest E value and con-

version it was further studied with different sec-alcohols in comparison to the wild-type biocatalyst (Table 5.3).

� �

Table 5.3. Substrate scope of transacylation reaction catalyzed by SC wild type[a] and G165L/M221F mutant.[a,b]

Wild type G165L/M221F

Entry Subst. R1 R2 E conv. [%] E conv.

[%] 1 13a H CH3 44 3 120 36 2 13b CH3 CH3 7 <1 43 31 3 13c Cl CH3 41 6 110 36 4 13d OMe CH3 2 <1 140 26 5 13e H C2H5 0 0 210 6 6 13f H C3H7 0 0 220 8 7 13g H C4H9 0 0 210 6 8 13h H C6H13 0 0 210 7

[a] The enzymes were immobilized on Accurel MP1000 in phosphate buffer (0.1 M, pH 7.8) with 1 M KCl. [b] The immobilized enzymes (10 mg), Na2CO3 (53 mg), vinyl butyrate 19 (750 mM), 1-phenylethanol derivatives (rac)-13a-h (500 mM), dodecane (100 mM) were added into 1 mL of dry THF. The reaction was shacked at 25 ºC under Ar atmosphere. The samples were analyzed by GC equipped with chiral column.

By substituting the para-position, the electronic effects were explored.

Here, 1-(4-methylphenyl)ethan-1-ol (rac)-13b worked great with our mutant but the selectivity was lower (Table 5.3, Entry 2). The electron-donating p-OMe substituted 1-phenylethanol (rac)-13d showed a dramatic increase in se-lectivity, giving an E value of 140, while the wild-type displayed low activity. The electron-withdrawing substrate 1-(4-chlorophenyl)ethan-1-ol (rac)-13c yielded a high E value (110). The aliphatic part of the substrate was also mod-ified since the modelling predicted that the enzyme could potentially host longer carbon chain in the R2-position. Four different products containing dif-ferent alkyl chains were tested, such as 1-phenylpropan-1-ol (rac)-13e, 1-phe-nylbutan-1-ol (rac)-13f, 1-phenylpentan-1-ol (rac)-13g), and 1-phenylheptan-1-ol (rac)-13h). All of them gave high enantioselectivity with E values > 200 and excellent enantiopurity (> 99%); however, the conversions were low.

The (S)-20a and (S)-20f-h products were modelled into the double mutant (G165L/M221F) structure based on the 1YU6 crystal structure. This model-ling study demonstrated that the new variant acquired a new cavity in the S1′

R2

OH

+O

OSC wt or double mutant

Na2CO3

THF R2

O

O

R2

OH

+

R1

R1R1

(rac)-13a-h 19 (S)-20a-h (R)-13a-h

��

the highest E value was obtained at 7 ºC. Meanwhile, the wild-type enzyme reaches an E value of 40 at the same temperature. The E value of the variant was lower at room temperature (23 ºC) than the earlier kinetic measurements using the two alcohol enantiomers in separate reactions.

For the last evaluation of our mutant, we tested the stability of the G165L/M221F enzyme in THF and compared this with the wild-type. The purified enzyme solutions (1 mgmL–1) in phosphate buffer were diluted in THF and the retained activity was measured using a proteolytic activity assay during the storage in THF. After 15 min, the wild-type enzyme lost 30% of its activity and within an hour of incubation the activity was less than 1%. How-ever, the mutant retained 50% active after one hour of incubation. To be able to evaluate the stability behavior of each mutation, the single mutants were also tested under these conditions. All three mutants showed higher stability than the natural biocatalyst in THF and it was also confirmed that the stability was dependent on the M221F mutation.

6���-�����������As explored by others in the literature, the enantioselectivity of SC was highly dependent on the pre-treatment of the enzyme rather than the immobilization technique. However, the immobilization on Accurell MP1000 in phosphate buffer with 1M KCl as additive gave an efficient immobilization after 90 min. The lyophilized enzymes on the support were stable for months if it was stored in aqueous saturated LiCl solutions. A new SC mutant was created that in-cluded two mutation points, G165L/M221F, which showed an E value of >100 towards (S)-1-phenylethanol (S)-13a in THF under an inert atmosphere. The enzyme kinetic study clearly showed an increase in enantioselectivity towards the (S)-alcohol substrate; however, this selectivity was found to be tempera-ture dependent. Here, the highest E value was obtained at 7 ºC. Not only did the M221F mutation point open up the S1′ pocket, but it also increased the stability of the enzyme in THF. The second mutation point (G165L) gave the possibility to convert longer alkyl chain-containing sec-alcohols, which is not possible with the wild-type enzyme.

���

5��%��� �������������������!����&�� ��"�0��������������������������� ���'�����1�(�

7���*������������The enantioselective synthesis of tertiary alcohols is still a remaining chal-lenge in contemporary synthetic organic chemistry.[127] However, these com-pounds and their related ester derivatives are of high synthetic importance as they are found in numerous natural products[128] and building blocks in organic chemistry.[129] The enzyme class hydrolases, such as lipases and proteases, have already several established applications in synthetic chemistry,[29,130] and they are also employed on an industrial scale.[131] However, the chiral com-pounds most commonly resolved are primary alcohols, secondary alcohols, and carboxylic acids. The tert-alcohols are significantly more challenging since only a few hydrolases have been found to accept these substrates. The reason for the low acceptance of tert-alcohols is the steric effects of the qua-ternary carbon center. The Bornscheuer group has disclosed a specific amino acid motif, which needs to be present to be able to catalyze the kinetic resolu-tion of tert-alcohols. The GGGX-motif, where X denotes any amino acid, was found through sequence analysis and alignment, and was further confirmed using several experimental examples.[132] This amino acid residue is located in the oxyanion hole of the lipases. Most of the reported procedures make use of hydrolysis techniques to resolve the tert-alcohol esters; however, there is a single example of transesterification of tert-acohols by the Bornscheuer group. Although the enantioselectivity was shown to be promising, the applied CalA enzyme displayed low activity even with extremely high loadings.[133]

Our group has already developed an efficient racemization protocol for tert-alcohols.[134] These results encouraged us to start a protein engineering project to develop an efficient biocatalyst from the wild-type CalA enzyme. It was imagined that by the end of the project, we would have access to both a racemization system and a biocatalyst, thereby enabling the development of a DKR procedure.

��

After the modelling study, we could identify seven key amino acids, which would be good targets for mutation. Five of these amino acids should not have any electronic interactions in the active site; however, they provide a larger space for the substrate and the generated product. Two amino acids (Y93 and L367) seemed to have an important role in the TI of (S)-substrate, which should maintain a hydrogen bond between the oxygen atom of the alcohol and the catalytic H366. Our hypothesis was that the modification of only these two amino acids would not have a dramatic increase for the activity, which is needed for the development of a DKR procedure. All of the seven amino acids were modified by altering them to smaller sizes, such as glycine or alanine, while maintaining the possibility to have the natural residue as well (Table 6.1). Based on these seven mutations with three amino acid possibilities the size of the library became rather large, having 37 possible variants.

Table 6.1. Suggested amino acid (Aa) residue positions to be included in the CalA library.

Aaposition Aawt Aalibrary

93 Y YAG 183 Y YAG

230 F FAG

233 F FAG

274 L LAG 336 I IAG

367 L LAG

DNA primers were designed to include the mutation points in the selected

sites. With the primers in hand, an overlap extension PCR (OE-PCR) method was used with a high-performance polymerase. By OE-PCR, each primer con-tained a single mutation and on one site three of these primers were propor-tionally mixed together to be able to get equal possibilities of the wt/G/A amino acids. The mixtures were used with our pBGP-CalA plasmid to include the mutations and create small DNA fragments. The seven fragments were created with an overlap region and bound together via PCR. The resulting “megaprimer” was again applied with our plasmid to create the whole plasmid sequence including specific mutations. Furthermore, the plasmid gene se-quence also contained a His-tag sequence, which plays a major role in the purification and immobilization of the enzymes. The mutations of the gene were evaluated by several gene sequencing to be able to identify that it con-tains the mutations and it has good variations. The creation of the library was performed according to a developed protocol, which was applied for the CalA

��

enzyme for another project in the group.[136] Pichia pastoris X33 yeast cell line was used as host for our modified plasmid, which is beneficial since it produces high quantities of the enzyme. However, its growing time is far longer than E. coli cells. The eukaryotic system needs at least 9–10 days to grow in cultures to express the protein. Since one of the aims of the project was to apply the enzyme variants in a DKR procedure, this more robust system was used. The cell culture with the mutated plasmid was grown on tributyrine-agar plates, which functions as a pre-screening process as the individual cell colonies that create active enzyme variants show a shallow circle around the colony (Scheme 6.2). Tributyrin is a natural triglyceride, which can be used on agar plates. In case of extracellular lipase activity, the tributyrin get hydro-lyzed to butyric acid and glycerol, which is the visible hallow around the ac-tive colonies.

Scheme 6.2. Tributyrin hydrolysis catalyzed by extracellular lipases on agar-plate.

After applying this process, the library size was reduced from the original

2187 (= 37) variants to 1392 colonies (16 plates). These colonies were re-grow in deep-well plates. After the growing time, the cultures were centrifuged down and the supernatant, which contains the enzyme, was immobilized on Ni-coated well-plates, since the incorporated His-tag, which was encoded into the gene, chelates to Ni. The procedure has three important roles: it serves as a purification step; the bound enzyme amount is also similar in each well since the number of binding sites are relatively similar and measured by the provider of the plates; it is also possible to switch from the buffer solution to organic solvent, where the transacylation reaction is performed. The immobilization requires about 2 h of shaking the supernatant in the wells. Afterwards, the enzyme immobilized plates are washed three times with iso-octane to remove the leftover buffer solution. In each plate, the last well was kept as a blank and column six contained the wild-type enzyme in order to compare the results more precisely. RP-UPLC-MS was used during the screening due to its high efficiency and quick analysis time. In the reaction mixtures, our model sub-strate (rac)-1a was introduced together with vinyl butyrate 19 as acyl donor in iso-octane (Scheme 6.1).

The substrate concentration was optimized by enzyme kinetic study based on the wild-type CalA. The His-tagged wild-type CalA was produced and im-mobilized by the same procedure as mentioned above. The concentration of

O

OO

O

O

O HOOH

OH

HO

O

+HO

O

HO

O

lipase

��

the substrate (rac)-1a was varied between 5–160 mM meanwhile the acyl do-nor concentration was kept constant at 250 mM in iso-octane. The Michaelis–Menten curve showed that the applied highest (160 mM) concentration is still not saturated due to the wt-CalA enzyme and it did not get deactivated (Figure 6.2). However, the solubility of (rac)-1a is limited in iso-octane. To be able to dissolve up to the 160 mM concentration, the solution has to be heated and stirred for long time. Since this observation, 80 mM concentration was used by the experiments with 250 mM vinyl butyrate 19.

Figure 6.2. Michaelis–Menten saturation curve by the transacylation of (rac)-1a cat-alyzed by wt-CalA.

The reactions were conducted in the immobilized Ni-coated wells on a ta-ble top shaker at room temperature. After the 24 h reaction time, samples were taken, diluted with iso-octane and screened by RP-UPLC-MS. After the screening process, nine possible hits were selected, further screened and ana-lyzed (Figure 6.3).

����%&������%�������%������%�������%�������%�������%�������%�������%�������%���

� �� ��� ��� ���

'

��(����� �������#�)$

��

Figure 6.3. The identified hits based on the RP-UPLC-MS screening.

All hits from the different plates were picked and re-grown on a deep well

plate, to enable analysis of the enantioselectivity and the conversion. The measurements were compared to the wild type enzyme and analyzed by gas chromatograph equipped with chiral column. Due to the small amounts of en-zyme on the plates, the reactions are rather slow and gave low conversions; however, the enantioselectivity was retained >90% in each case.

To be able to identify the mutation points of the better performing en-zymes, the single colonies were inoculated for an overnight culture and the plasmid were removed from the cells. The concentration of the plasmids was low (~10 nguL-1) in all cases. To be able to properly sequence our entire gene, PCR amplification has been used to increase the concentration of the DNA with specifically designed primers, which can bind both ends in our sequence. After the readings, we could identify that P1–H1 and P1–E5 colonies contain the wild-type CalA gene, but both of the best performing P2–C8 and P7–A11 contains a mutation point at the L367 position. In the natural enzyme, this specific point contains a leucine (CTC), which was mutated to the smallest glycine (GGT) residue. The sequencing of the other possible mutants is ongo-ing in our lab at the moment.

The P2–C8, as our best performing candidate so far, was re-grown in large (500 mL) scale to be able to further optimize with our reaction conditions. We were especially curious about the temperature of the reaction and thermosta-bility of the enzyme. Both the wild type CalA and the P2–C8 variant were

���

�����

���

���

������

���

���

�

��

���

���

���

���*�+

,- .��%� .��/� .��/� .��0� .��1� .���� .����� .��2�

��

immobilized on Accurel MP1000 with KCl additive. The reactions were per-formed in iso-octane (1 mL) with 80 mM of substrate (rac)-1a, 250 mM of vinyl butyrate 19 and 25 mg of the immobilized enzyme (Figure 6.4).

Figure 6.4. Scope of temperature by the transacylation of (rac)-1a catalyzed by wt-CalA and P2–C8 CalA variant.

To run the reactions in different temperatures gave a big improvement to

our reaction, after 48 h reaction time the mutant P2–C8 (L367G) enzyme var-iant showed 35% conversion, but unfortunately the enantioselectivity was dropped down to 93%. However, the reaction gave 18% conversion of the product (R)-21 at 40 ºC with 96% ee.

7���-�����������Based on the literature and our studies of the racemization of tert-alcohols, a protein engineering project was performed to be able to increase the activity of the wt-CalA enzyme. Seven amino acid sites were mutated based on mod-eling the tetrahedral intermediates in the active site. The mutations were per-formed by OE-PCR method with a pre-screening on tributyrine agar-plates. The selected colonies were individually grown and the transacylation reac-tions were performed on Ni-coated plates. After the screening, the hits were further evaluated and the most potent P2–C8 (L367G) mutant was tested in synthetic scale. The mutant enzyme showed higher conversion and good ee on elevated temperatures.

�

�

��

��

��

��

�

�

��

� �� �� � �� ��

���3�*�

+

��� �*4+

����50

�����50

�����50

.��0����50

.��0�����50

.��0�����50

�

6�������������$���� �

My doctoral thesis introduces new heterogeneous catalysts for racemization and kinetic resolution processes. Meanwhile, we also took a step away from the synthetic chemistry and got introduced into the world of protein engineer-ing. The first two project describe the application of heterogeneous acidic res-ins in catalysis. During the first project, the racemization of tert-alcohols was studied, where we could develop a system that can essentially racemize tert-alcohols without any byproduct formation. This procedure has a high synthetic value, since before, the unreacted tert-alcohol enantiomer (50%) of the reso-lution was just waste. The second research topic is a good example, how a new project can originate from a previous one. Meanwhile our tert-alcohol substrates could be racemized by acidic resins, the tertiary allylic carbinols were isomerized to sec-alcohols. Later on, it was studied that the combination of CalB enzyme and the acidic resin perform a migratory DKR reaction, where the acidic resin was involved in two disparate processes: the isomerization to sec-alcohol and the racemization of the alcohol in a two-phase solvent system.

Chapter 4 describes the synthesis, applications and characterization of a biohybrid catalyst. At the beginning, the Pd(0)-CalB CLEA catalyst was de-signed to be able to combine the Pd nanoparticles with the CalB enzyme. The first application was a one-pot cascade reaction, where Pd catalyzed the cy-cloisomerization of 4-pentynoic acid to a lactone, which in turn could serve as an acyl donor in enzymatic kinetic resolution of sec-alcohols. During the work on this project, another application, the DKR of amines was already in our mind. It was achieved later with a slightly modified Pd(0)-CalB CLEA cata-lyst. The biohybrid catalyst was also characterized with several different ana-lytical techniques to be able to understand its composition.

In the last two chapters, the modification of two different enzyme was per-formed by two different protein engineering aspects. The first case, specific point mutations were introduced to provide a more rigid subtilisin Carlsberg enzyme. The second project contained seven mutation sites, which were ran-domly mutated based on the natural amino acid/Gly/Ala codon in the CalA enzyme. In both cases, the mutants had higher performance in our organic transformations.

��

������7��8���������������� ��

The author’s contribution to publications I-VII:

I.� Performed the major part of the experimental work and wrote the article.

II.� Contributed to the experimental work and to the writing of the arti-

cle. III.� Performed the major part of the experimental work and contributed

to the writing of the article.

IV.� Devised the synthetic method of the biobybrid catalyst, performed the Pd(0)-CalB CLEA catalyst synthesis and wrote the major part of the article.

V.� Contributed to the experimental work and to the writing of the arti-

cle.

VI.� Performed the major part of the experimental work and wrote the article.

�

������7�/8�$���������� ��� �

Reprint permissions were kindly granted for each publication by the following publishers: I.� Tamás Görbe, Richard Lihammar, Jan-E. Bäckvall

Copyright © 2018 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chemistry – A European Journal 2018, 24, 77–80

II.� Chicco M. Sapu, Tamás Görbe, Richard Lihammar, Jan-E. Bäck-vall, Jan Deska Copyright © 2014 American Chemical Society Organic Letters 2014, 16, 5952–5955

III.� Tamás Görbe, Karl P. J. Gustafson, Oscar Verho, Gabriella Kervefors, Haoquan Zheng, Xiaodong Zou, Eric V. Johnston, Jan-E. Bäckvall Copyright © 2017 American Chemical Society ACS Catalysis 2017, 7, 1601–1605

IV.� Robin Dorau, Tamás Görbe, Maria S. Humble

Copyright © 2018 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemBioChem 2018, 19 (4), 338–346

V.� Picture of Prof. R. P. Feynman was kindly granted by Caltech Copyright © 2018 Institute Archives California Institute of Technology

��

�����4������ �

I want to express my gratitude to: Prof. Jan-Erling Bäckvall for accepting me as a graduate student in 2013.

It has been a long way since then, and I always got great inspiration and en-couragement from you. The freedom of science, what I got from you was scary in the beginning. Afterwards, I enjoyed every moment, when I could come up with an idea, and you always supported them. I also learned from you how to coordinate a research group in nice and efficient way.

Prof. Frances H. Arnold for hosting my four months exchange at Caltech

and giving me the opportunity to do research on two highly interesting pro-jects. I have learned incredible a lot about directed evolution, protein engi-neering and molecular biology. Practically, you opened up a new world for me. Also, you are highly encouraging and a cool leader of your group.

Prof. Gyula Schneider for guiding me during my undergraduate studies. I

know that I was not the easiest student to working with, but you were always very helpful and supportive during my time in your lab. I learned most of my basic, synthetic chemistry skills from you. I am glad to remember all of our great discussions next to the column chromatography of steroids.

Prof. Pher G. Andersson for your interest about my thesis and for your

valuable comments. Dr. Maria Humble for teaching me and helping a lot in the field of molec-

ular biology. I was always glad to work with you during our two projects. Prof. Jan Deska for your friendship and the work together on our migratory

DKR project. Markus, Oscar, Arnar and Bin for proofreading this thesis and giving me

lots of valuable comments.

��

All of my collaborators/colleagues, who I have had a pleasure to working with during the years: Johanna Löfgren, Michael Oschmann, Haoquan Zheng, Xiaodong Zou, Eric Johnston, Robin Dorau, Gonzalo de Gonzalo, Ning Yuan, Andrey Shchukarev, Cheuk-Wai Tai, Ingmar Persson.

Karl for being one of my best friend since the time I arrived to Stockholm.

I think you are the best organic chemist who I met and your passion about it is just incredible. I am very glad that we have spent the last 5 years next to each other, working on projects together and now pretty much defend/party together. I was more than happy to have lunch with you when we could share our life issues and stories.

Richard for your welcome and guidance in the JEB group. I learned a lot

from you about lab skills, KR, DKR, NMR and also how to be happy and funny in every single day.

Arnar for your friendship, jokes, weekend works together and also the

fights in Star Wars and Guitar Hero! Do not forget about chia seeds, latte, scarf, flannel shirt, Macbook and so on, you know the “hipster life”.

Oscar for being a great friend and all of your guidance in the heterogeneous

catalysis field, nevertheless to the introduction of the nanoworld. My exchange graduate student colleagues: Chicco, Cynthia and Chris-

toph. I was very happy to work with you guys and share lots of fun moments with you inside and outside of the lab.

My first diploma-worker, Gabriella, I think we shared quite some suffer

together during the bio-nano catalyst time, when nothing worked and then much more fun when it was finally working. I am very happy to be your friend and to see you now as a PhD student. Magnus, as my next student, we have still a lot to do, but I believe your curiosity about science will lead us to the goal.

All of the past and present members of the JEB group: Tuo, Teresa, Madde,

Bin, Anuja, Youai, Johanna, Daniels, Ramesh, Alex, Can, Ylva, Maria K., Javier, Karim, Nerea, Gilian, Laura, Santosh, Beneesh, Karin, Andreas, Deng, Jie, Man-bou, Michael, Alex, Viola and many more.

All of the past and present people at the Department: Thanks for every-thing guys, it was always fun to work with you!

���

All the past and present members of the Broffice, you guys made the office like home, maybe a bit crazy home sometimes! I was very happy to share my life with you, especially with Tove S., Paz, Tove K. Fredrik, Alexey.

All of the past and present TA-staff: Thank you for keeping the department

up and running, especially thanks to Carin, who helped me and teach me how to get my compound’s HRMS data.

All of my friends at the Frances group and at Caltech (particularly my awesome supervisor, Christopher Prier) you guys made my exchange unfor-gettable and fun. I hope to see you all again soon!

The foundations and agencies that have supported me financially during

my exchange and the many conferences. I want to acknowledge the following: Stiftelsen Olle Engkvist Byggmästare, Kungliga Skogs- och Lantbruksakade-mien, C F Liljevalch:s resestipendier, Kungliga Vetenskapsakademien, Läng-manska kulturfonden, Helge Ax:son Johnsons Stiftelse, Stockholm universitets donationsstipendier, Svenska Kemistsamfundet, Ångpanneföreningens forsk-ningsstiftelse and K & A Wallenberg Foundation. I have not only learned new things during the trips but also explored the world.

The Greens villa crew for the “Thursday pub” nights: Andrey, Tim, Fabian, Erik L., Eric, Markus R., Göran, Martin, Sybrand…

Markus, I have realized, that you are pretty much my older brother. We

have shared so many fun and very sometimes sad moments together. You were always ready to help me out from any scenario in life and science. Your atti-tude always helped me to look at things from the right and bright angle. You are a great chemist as well, and I really enjoyed our scientific discussions to-gether.

My family for your constant support with everything, even if they were a little bit crazy sometimes. You were always with me during the good or bad time and never stop to believe in me. Anyu, Apu és Adri, nagyon köszönöm nektek mindazt a gondoskodást, bíztatást és szeretetet, amit tőletek kaptam az életem során!

Ohh, yeah, just one more thing! All fellow scientists should never forget: “Sometimes science is more art than science!” -Rick Sanchez

���

$���� �

[1] Following ACS Standard Abbrevation/Acronyms, 2018 Guidelines for Authors, Journal of Medicinal Chemistry Home Page. http://pubs.acs.org/page/jmcmar/submission/authors.html

[2] A. In, R&D Magazine 2005, p 29. [3] a) J. Poliana, A. P. McCabe (Eds.), Industrial Enzymes Structure, Func-

tion and Applications, Springer-Verlag: Berlin, 2007; b) K. Faber, Bio-transformations in Organic Chemistry, 4th Ed. Springer-Verlag: Berlin, 2000.

[4] M. N. Gupta, Eur. J. Biochem. 1992, 203, 25–32. [5] a) R. H. Crabtree, The Organic Chemistry of Transition Metals, 5th Ed.

John Wiley & Sons Inc.: New York, USA, 2009; b) P. B. Kettler, Org. Proc. Res. Dev. 2003, 7, 342–354; c) F. R. Hartley (Ed.), Chemistry of the Platinum Group Metals: Recent Developments, Elsevier: Amster-dam, New York, 1991.

[6] R. P. Reddy, H. M. L. Davies, Org. Lett. 2006, 8 (22), 5013–5016. [7] G. Rothenberg, Catalysis: Concepts and Green Applications, Wiley-

VCH: Weinheim, 2008. [8] R. A. Sheldon, H. van Bekkum, Fine Chemicals Through Heterogene-

ous Catalysis, Wiley-VCH: Weinheim, 2001. [9] E. Cook, Nature 1926, 117, 419–421. [10] a) C. E. Garret, K. Prasad, Adv. Synth. Catal. 2004, 346, 889–900; b)

B. Konigsberger, G. –P. Chen, R. R. Wu, M. J. Girgis, K. Prasad, O. Repic, T. J. Blacklock, Org. Process Res. Dev. 2003, 7, 733–742.

[11] R. Feynman, Engineering and Science 1960, 22–36. [12] a) W. Jiang, C. A. von Roemeling, Y. Chen, Y. Qie, X. Liu, J. Chen, B.

Y. S. Kim, Nature Biomedical Engineering 2017, 1, 0029; b) K. Watanabe, D. Menzel, N. Nilius, H. -J. Freund, Chem. Rev. 2006, 106 (10), 4301–4320; b) C. N. R. Rao, A. Müller, A. Cheetham (Eds.), The Chemistry of Nanomaterials: Synthesis, Properties and Applications, Wiley-VCH: Weinheim, 2004; c) G. Schmid (Ed.), Nanoparticles: From Theory to Application, Wiley-VCH: Weinheim, 2004; d) R. Shenhar, V. M. Rotello, Acc. Chem. Res. 2003, 36, 549–561; e) P.

���

Poizot, S. Laurelle, S. Grugeon, L. Dupont, J. –M. Tarascon, Nature 2000, 407, 496–499.

[13] a) L. L. Chng, N. Erathodiyil, J. Y. Ying, Acc. Chem. Res. 2013, 46, 1825–1837; b) M. J. Climent, A. Corma, S. Iborra, Chem. Rev. 2011, 111, 1072–1133; c) G. A. Somorjai, Y. Li, Top. Catal. 2010, 53, 832–847; d) D. Astruc (Ed.), Nanoparticles and Catalysis, Wiley-VCH: Weinheim, 2008.

[14] a) K. An, G. A. Somorjai, ChemCatChem 2012, 4, 1512–1524; b) G. A. Somorjai, J. Y. Park, Angew. Chem., Int. Ed. 2008, 120, 9352–9368; c) Y. –F. Han, D. Kumar, D. W. Goodman, J. Catal. 2005, 230, 353–358; d) R. Narayanan, M. A. El-Sayed, Nano Lett. 2004, 11, 798–802.

[15] a) F. Zaera, Catal. Lett. 2012, 142, 501–516; b) B. R. Cuenya, Thin Solid Films 2010, 518, 3127–3150; c) H. –J. Freund, Chem. Eur. J. 2010, 16, 9384–9397; d) L. C. Grabow, M. Mavrikakis, Angew. Chem., Int. Ed. 2008, 47, 7390–7392.

[16] a) J. Alemán, A. V. Chadwick, J. He, M. Hess, K. Horie, R. G. Jones, P. Kratochvíl, I. Meisel, I. Mita, G. Moad, S. Penczek, R. F. T. Stepto, Pure Appl. Chem. 2007, 79, 1801–1829; b) W. Ostwald, Lehrbuch der Allgemeinen Chemie, 2nd Ed. Engelmann: Leipzig, 1896.

[17] F. G. Mutti, T. Knaus, N. S. Scrutton, M. Breuer, N. J. Turner, Science 2015, 349, 1525−1529.

[18] a) A. K. Engström, E. V. Johnston, O. Verho, K. P. J. Gustafson, M. Shakeri, C. W. Tai, J. -E. Bäckvall, Angew. Chem., Int. Ed. 2013, 52, 14006−14010; b) O. Verho, J. -E. Bäckvall, J. Am. Chem. Soc. 2015, 137, 3996–4009.

[19] D. F. Sauer, T. Himiyama, K. Tachikawa, K. Fukumoto, A. Onoda, E. Mizohata, T. Inoue, M. Bocola, U. Schwaneberg, T. Hayashi, J. Okuda, ACS Catal. 2015, 5, 7519–7522.

[20] S. Sutthasupa, M. Shiotsuki, F. Sanda, Polymer Journal 2010, 42, 905–915.

[21] R. H. Grubbs, S. J. Miller, G. C. Fu, Acc. Chem. Res. 1995, 28 (11), 446–452.

[22] a) V. Farina, J. T. Reeves, C. H. Senanayake, J. J. Song, Chem. Rev. 2006, 106, 2734–2793; b) M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Kesseler, R. Stürmer, T. Zelinski, Angew. Chem., Int. Ed. 2004, 43, 788–824.

[23] C. C. C. R. de Carvalho, M. M. R. da Fonseca, Food Chemistry 2006, 95 (3), 413–422.

[24] J. M. Hawkins, T. J. N. Watson, Angew. Chem., Int. Ed, 2004, 43, 3224–3228

[25] a) A. Moyano, R. Rios, Chem. Rev. 2011, 111, 4703–4832; b) P. I. Dalko, L. Moisan, Angew. Chem., Int. Ed. 2004, 43, 5138–5175; c) B.

���

M. Trost, Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5348–5355; d) C. J. Douglas, L. E. Overman, Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5363–5367; e) I. Ojima (Ed.), Catalytic Asymmetric Synthesis, 2nd Ed. Wiley-VCH: New York, 2000; f) E. N. Jacobsen, A. Pfaltz, H. Yama-moto (Eds.), Comprehensive Asymmetric Catalysis, Springer: Berlin, 1999; g) R. Noyori, Asymmetric Catalysis in Organic Synthesis, John Wiley & Sons Inc.: New York, 1994.

[26] E. Fogassy, M. Nogradi, D. Kozma, G. Egri, E. Palovics, V. Kiss, Org. Biomol. Chem. 2006, 4, 3011–3030.

[27] KR with transition metal, organo cat and enzyme [28] E. García-Urdiales, I. Alfonso, V. Gotor, Chem. Rev. 2011, 111, 110–

180. [29] U. T. Bornscheuer, R. J. Kazlauskas, Hydrolases in Organic Synthesis:

Regio- and Stereoselective Biotransformations; Wiley-VCH: Wein-heim, 1999.

[30] E. A. Sym, Biochem. Z. 1933, 258, 304–324; P. J. Halling, L. Kvittingen, Trends Biotechnol. 1999, 17, 343–344.

[31] a) G. Carrea, S. Riva, Organic Synthesis with Enzymes in Non-aqueous Media; Wiley-VCH: Weinheim, 2008; b) A. Zaks, A. M. Klibanov, Sci-ence 1984, 224, 1249–1251; c) A. Zaks, A. M. Klibanov, Proc. Natl. Acad. Sci, U. S. A. 1985, 82, 3192–3196; d) P. J. Halling, Phil. Trans. R. Soc. Lond. B 2004, 359, 1287–1297.

[32] R. J. Kazlauskas, A. N. E. Weissflock, A. T. Rappaport, L. A. Cuccia, J. Org. Chem. 1991, 56, 2656–2665.

[33] K. Drauz, H. Waldmann, Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook; Wiley-VCH: Weinheim, 1995.

[34] a) J. Kraut, Annu. Rev. Biochem. 1977, 46, 331–358; b) M. -P. Bousquet-Dubouch, M. Graber, N. Sousa, S. Lamare, M. -D. Legoy, Biochimica et Biophysica Acta 2001, 1550, 90–99.

[35] L. Brady, A. M. Brzozowski, Z. S. Derewenda, E. Dodson, G. Dodson, S. Tolley, J. P. Turkenburg, L. Christiansen, B. Huge-Jensen, L. Nor-skov, L. Thim, U. Menge, Nature 1990, 343, 767–770.

[36] C. -S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc. 1982, 104, 7294–7299.

[37] E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink, B. Zwanenburg, Tetrahedron 1997, 53, 9417–9476.

[38] B. Martín Matute, J. -E. Bäckvall, Curr. Opin. Chem. Biol. 2007, 11, 226–232.

[39] a) Y. Ahn, S. -B. Ko, M. -J. Kim, J. Park, Coord. Chem. Rev. 2008, 252, 647–658; b) M. T. El Gihani, J. M. J. Williams, Curr. Opin. Chem. Bio. 1999, 3, 11–15; c) R. Stürmer, Angew. Chem., Int. Ed. 1997, 36, 1173–

���

1174; d) D. S. Tan, M. M. Guenter, D. G. Drueckhammer, J. Am. Chem. Soc. 1995, 117, 9093–9094.

[40] a) J. H. Lee, K. Han, M. -J. Kim, J. Park, Eur. J. Org. Chem. 2010, 999–1015; b) M. -J. Kim, Y. Ahn, J. Park, Korean Chem. Soc. 2005, 26, 515-522; c) O. Pàmies, J. -E. Bäckvall, Trends Biotechnol. 2004, 22, 130–135.

[41] a) B. Martín Matute, M. Edin, K. Bogár, F. B. Kaynak, J. -E. Bäckvall, J. Am. Chem. Soc. 2005, 127, 8817–8825; b) B. Martín Matute, M. Edin, K. Bogár, J. -E. Bäckvall, Angew. Chem., Int. Ed. 2004, 43, 6535–6539.

[42] a) M. C. Warner, A. Nagendiran, K. Bogár, J. -E. Bäckvall, Org. Lett. 2012, 14, 5094–5097; b) J. Norinder, K. Bogár, L. Kanupp, J. -E. Bäck-vall, Org. Lett. 2007, 9, 5095–5098; c) K. Bogár, P. Hoyos Vidal, A. R. Alcántar León, J. -E. Bäckvall, Org. Lett. 2007, 9, 3401–3404.

[43] M. C. Warner, G. A. Schevchenko, S. Jouda, K. Bogár, J. -E. Bäckvall, Chem. Eur. J. 2013, 19, 13859–13864.

[44] K. Leijondahl, L. Borén, R. Braun, J. -E. Bäckvall, Org. Lett. 2008, 10, 2027–2030.

[45] R. Lihammar, R. Millet, J. -E. Bäckvall, Adv. Synth. Catal. 2011, 353, 2321–2327.

[46] a) M. Shakeri, C. -W. Tai, E. Göthelid, S. Oscarsson, J. -E. Bäckvall, Chem. Eur. J. 2011, 17, 13269–13273; b) L. K. Thalén, D. Zhao, J. -B. Sortais, J. Paetzold, C. Hoben, J. -E. Bäckvall, Chem. Eur. J. 2009, 15, 3403–3410; c) J. Paetzold, J. -E. Bäckvall, J. Am. Chem. Soc. 2005, 127, 17620–17621.

[47] a) K. Engström, M. Shakeri, J. -E. Bäckvall, Eur. J. Org. Chem. 2011, 1827–1830; b) M. Shakeri, K. Engström, A. Sandström, J. -E. Bäckvall, ChemCatChem 2010, 2, 534–538.

[48] O. Pámies, J. -E. Bäckvall, J. Org. Chem. 2002, 67, 1261–1265. [49] A. Träff, K. Bogár, M. Warner, J. -E. Bäckvall, Org. Lett. 2008, 10,

4807–4810. [50] E. V. Johnston, K. Bogár, J. -E. Bäckvall, J. Org. Chem. 2010, 75,

4596–4599. [51] A. Träff, R. Lihammar, J. -E. Bäckvall, J. Org. Chem. 2011, 76, 3917–

3921. [52] A. Träff, C. E. Solarte, J. -E. Bäckvall, Collect. Czech. Chem. Commun.

2011, 76, 919–927. [53] a) N. Menashe, Y. Shvo, Organometallics 1991, 10, 3885–3891; b) R.

M. Haak, F. Berthiol, T. Jerphagnon, A. J. A. Gayet, C. Tarabiono, C. P. Postema, V. Ritleng, M. Pfeffer, D. B. Janssen, A. J. Minnaard, B. L. Feringa, J. G. de Vries, J. Am. Chem. Soc. 2008, 130, 13508–13509; c) J. H. Choi, Y. K. Choi, Y. H. Kim, E. S. Park, E. J. Kim, M. -J. Kim,

� �

J. Park, J. Org. Chem. 2004, 69, 1972–1977; d) D. Mavrynsky, M. Pä-iviö, K. Lundell, R. Sillanpää, L. T. Kanerva, R. Leino, Eur. J. Org. Chem. 2009, 1317–1320.

[54] A. N. Parvulescu, P. A. Jacobs, D. E. De Vos, Adv. Synth. Catal. 2008, 350, 113–121.

[55] O. Långvik, T. Saloranta, A. Kirilin, A. Liljeblad, P. Mäki-Arvela, L. T. Kanerva, D. Y. Murzin, R. Leino, ChemCatChem 2010, 2, 1615–1621.

[56] G. Xu, L. Wang, Y. Chen, Y. Chen, J. Wu, L. Yang, Tetrahedron Lett. 2013, 54, 5026–5030.

[57] S. Wuyts, J. Wahlen, P. A. Jacobs, D. E. De Vos, Green Chem. 2007, 9, 1104–1108.

[58] a) S. Wuyts, K. De Temmerman, D. E. De Vos, P. A. Jacobs, Chem. Eur. J. 2005, 11, 386–397; b) S. Wuyts, K. De Temmerman, D. E. De Vos, P. A. Jacobs, Chem. Commun. 2003, 1928–1929.

[59] Y. Cheng, G. Xu, J. Wu, C. Zhang, L. Yang, Tetrahedron Letters 2010, 51, 2366–2369.

[60] a) M. -J. Kim, W. -H. Kim, K. Han, Y. K. Choi, J. Park, Org. Lett. 2007, 9, 1157–1159; b) A. N. Parvulescu, P. A. Jacobs, D. E. De Vos, Chem. Eur. J. 2007, 13, 2034–2043; c) M. S. Kwon, N. Kim, C. M. Park, J. S. Lee, K. Y. Kang, J. Park, Org. Lett. 2005, 7, 1077–1079.

[61] K. Mullis, F. Faloona, S. Scharf, R. Saiki, G. Horn, H. Erlich, Cold Spring Harbor, Symp. Quant. Biol. 1986, LI, 263–273.

[62] C. Darwin, On the Origin of Species by Means of Natural Selection, Or, The Preservation of Favoured Races in the Struggle for Life, J. Murray, 1859.

[63] B. Alberts, A. Johnson, J. Lewis, M. Raff, K. Roberts, P. Walter, Mo-lecular Biology of the Cell, 4th Ed., Garland Science: United States of America, 2002.

[64] R. C. Cadwell, G. F. Joyce, Genome Res. 1992, 2, 28–33. [65] C. K. Prier, R. K. Zhang, A. R. Buller, S. Brinkmann-Chen, F. H. Ar-

nold, Nature Chemistry 2017, 9, 629–634. [66] M. T. Reetz, J. D. Carballeira, A. Vogel, Angew. Chem., Int. Ed. 2006,

45, 7745–7751. [67] K. L. Morley, R. J. Kazlauskas, Trends Biotechnol. 2005, 23, 231–237. [68] C. A. Hutchison, S. Phillips, M. H. Edgell, S. Gillam, P. Jahnke, M.

Smith, J. Biol. Chem. 1978, 253, 6551–6560. [69] K. L. Morrison, G. A. Weiss, Curr. Opin. Chem. Biol. 2001, 5, 302–

307. [70] M. T. Reetz, M. Bocola, J. D. Carballeira, D. Zha, A. Vogel, Angew.

Chem., Int. Ed. 2005, 44, 4192–4196. [71] Evolution Tools CASTER and B-FItter, M. Reetz, Vol. 2015.

���

[72] M. T. Reetz, D. Kahakeaw, R. Lohmer, ChemBioChem 2008, 9, 1797–

1804. [73] M. T. Reetz, A. Zonta, K. Schimossek, K. -E. Jaeger, K. Liebeton, An-

gew. Chem., Int. Ed. 1997, 36, 2830–2832. [74] M. Baumann, R. Stürmer, U. T. Bornscheuer, Angew. Chem., Int. Ed.

2001, 40, 4201–4204. [75] a) A. G. Sandström, Y. Wikmark, K. Engström, J. Nyhlén, J. -E. Bäck-

vall, Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 78–83; b) K. Engström, J. Nyhlén, A. G. Sandström, J. -E. Bäckvall, J. Am. Chem. Soc. 2010, 132, 7038–7042.

[76] L. E. Janes, A. C. Löwendahl, R. J. Kazlauskas, Chem. Eur. J. 1998, 4, 2324–2331.

[77] M. Konarzycka-Bessler, U. T. Bornscheuer, Angew. Chem., Int. Ed. 2003, 42, 1418–1420.

[78] a) U. T. Bornscheuer, Angew. Chem., Int. Ed. 2003, 42, 3336–3337; b) A. Liese, L. Hilterhaus, Chem. Soc. Rev. 2013, 42, 6236–6249; c) G. W. Huisman, S. J. Collier, Curr. Opin. Chem. Biol. 2013, 17, 284–292.

[79] P. Adlercreutz, Chem. Soc. Rev. 2013, 42, 6406–6436. [80] P. Villeneuve, J. M. Muderhwa, J. Graille, M. J. Haas, J. Mol. Catal. B:

Enzym. 2000, 9, 113–148. [81] R. A. Sheldon, Adv. Synth, Catal. 2007, 349, 1289–1307. [82] K. Faber, Biotransformations in Organic Chemistry: A Textbook;

Springer-Verlag, 2004. [83] A. F. A. Quiocho, F. M. Richards, Proc. Natl. Acad. Sci. U. S. A. 1964,

52, 833–839. [84] S. Talekar, A. Joshi, G. Joshi, P. Kamat, R. Haripurkar, S. Kambale,

RSC Adv. 2013, 3, 12485–12511. [85] Y. Wikmark, M. S. Humble, J. -E. Bäckvall, Angew. Chem., Int. Ed.

2015, 54, 4284–4288. [86] S. Lu, S. B. Poh, W. Y. Siau, Y. Zhao, Angew. Chem., Int. Ed. 2013,

53, 1731–1734. [87] J. Rehdorf, G. A. Behrens, G. S. Nguyen, R. Kourist, U. T.

Bornscheuer, Appl. [88] M. C. Warner, J. -E. Bäckvall, Acc. Chem. Res. 2013, 46 (11), 2545–

2555. [89] a) S. Wuyts, K. De Temmerman, D. De Vos, P. A. Jacobs, Chem. Eur.

J. 2005, 11, 386–397; b) P. Ödman, L. A. Wessjohann, U. T. Bornscheuer, J. Org. Chem. 2005, 70, 9551–9555; c) S. Wuyts, K. De Temmerman, D. De Vos, P. A. Jacobs, Chem. Commun. 2003, 15, 1928–1929.

[90] S. Swaminathan, K. V. Narayanan, Chem. Rev. 1971, 71 (5), 429–438.

��

[91] E. Henke, J. Pleiss, U. T. Bornscheuer, Angew. Chem., Int. Ed. 2002,

41, 3211–3213. [92] R. Kourist, P. Domínguez de María, U. T. Bornscheuer, ChemBioChem

2008, 9, 491–498. [93] J. Ramharter, J. Mulzer, Org. Lett. 2011, 13, 5310–5313. [94] a) C. Morrill, R. H. Grubbs, J. Am. Chem. Soc. 2005, 127, 2842–2843;

b) T. Mitsudome, S. Sueoka, S. Ideka, T. Mizugaki, K. Jitsukawa, K. Kaneda, ChemCatChem 2013, 5, 2879–2882; c) J. Li, C. Tan, J. Gong, Z. Yang, Org. Lett. 2014, 16, 5370–5373.

[95] S. Akai, R. Hanada, N. Fujiwara, Y. Kita, M. Egi, Org. Lett. 2010, 12, 4900–4903.

[96] a) R. Munirathinam, J. Huskens, W. Verboom, Adv. Synth. Catal. 2015, 357, 1093−1123; b) T. Tsubogo, T. Ishiwata, S. Kobayashi, Angew. Chem., Int. Ed. 2013, 52, 6590−6604.

[97] a) J. M. Bolivar, I. Eisl, B. Nidetzky, Catal. Today 2016, 259, 66−80; b) Y. Zhang, J. Ge, Z. Liu, ACS Catal. 2015, 5, 4503−4513.

[98] a) M. Peixoto de Almeida, S. A. C. Carabineiro, ChemCatChem 2012, 4, 18−29; b) A. Zsigmond, F. Notheisz, Curr. Org. Chem. 2006, 10, 1655−1680.

[99] a) S. Peiris, J. McMurtrie, H. Y. Zhu, Catal. Sci. Technol. 2016, 6, 320−338; b) L. L. Chng, N. Erathodiyil, J. Y. Ying, Acc. Chem. Res. 2013, 46, 1825−1837; c) K. An, G. A. Somorjai, ChemCatChem 2012, 4, 1512−1524; d) D. Astruc, F. Lu, J. R. Aranzaes, Angew. Chem., Int. Ed. 2005, 44, 7852−7872.

[100] a) P. Zhang, H. Zhu, S. Dai, ChemCatChem 2015, 7, 2788−2805; b) D. S. In Su, Nanomaterials in Catalysis, 1st Ed; P. Serp, K. Phillipot, Eds.; Wiley−VCH: Weinheim, Germany, 2013; pp 331−374; c) L. F. Ma-bena, S. Sinha Ray, S. D. Mhlanga, N. J. Coville, Appl. Nanosci. 2011, 1, 67−77.

[101] a) S. Li, K. Yang, C. Tan, X. Huang, W. Huang, H. Zhang, Chem. Com-mun. 2016, 52, 1555−1562; b) H. L. Jiang, Q. Xu, Chem. Commun. 2011, 47, 3351−3370.

[102] a) M. B. Gawande, P. S. Branco, R. S. Varma, Chem. Soc. Rev. 2013, 42, 3371−3393; b) M. Flytzani-Stephanopoulos, B. C. Gates, Annu. Rev. Chem. Biomol. Eng. 2012, 3, 545−574.

[103] a) Z. Li, J. Li, Curr. Org. Chem. 2013, 17, 1334−1349; b) E. Guibal, Prog. Polym. Sci. 2005, 30, 71−109; c) M. Králik, A. J. Biffis, Mol. Catal. A: Chem. 2001, 177, 113−138.

[104] a) R. Luque, A. M. Balu, M. Campelo, M. D. Gracia, E. Losada, A. Pineda, A. A. Romero, J. C. Serrano-Ruiz, Catalysis 2012, 24, 253−280; b) D. Astruc, Inorg. Chem. 2007, 46, 1884−1894; c) A. Taguchi, F. Schüth, Microporous Mesoporous Mater. 2005, 77, 1−45.

���

[105] a) R. Srivastava, B. Sarmah, B. Satpati, RSC Adv. 2015, 5,

25998−26006; b) Y. Horiuchi, D. Do Van, Y. Yonezawa, M. Saito, M. Dohshi, T. H. Kim, M. Matsuoka, RSC Adv. 2015, 5, 72653−72658; c) M. Li, X. Xu, Y. Gong, Z. Wei, Z. Hou, H. Li, Y. Wang, Green Chem. 2014, 16, 4371−4377.

[106] J. Ge, J. Lei, R. N. Zare, Nat. Nanotechnol. 2012, 7, 428−432. [107] M. Filice, M. Marciello, M. del Puerto Morales, J. M. Palomo, Chem.

Commun. 2013, 49, 6876−6878. [108] R. Schoevaart, M. W. Wolbers, M. Golubovic, M. Ottens, A. P. G. Kie-

boom, F. van Rantwijk, L. A. M. van der Wielen, R. A. Sheldon, Bio-technol. Bioeng. 2004, 87, 754–762; c) P. Adlercreutz, Chem. Soc. Rev. 2013, 42, 6406−6436.

[109] a) A. Nagendiran, O. Verho, C. Haller, E. V. Johnston, J. -E. Bäckvall, J. Org. Chem. 2014, 79, 1399−1405. b) O. Verho, F. Gao, E. V. John-ston, W. Wan, A. Nagendiran, H. Zheng, J. -E. Bäckvall, X. Zou, APL Mater. 2014, 2, 113316.

[110] S. Akai, K. Tanimoto, Y. Kita, Angew. Chem., Int. Ed. 2004, 43, 1407–1410.

[111] K. P. J. Gustafson, R. Lihammar, O. Verho, K. Engström, J. -E. Bäck-vall, J. Org. Chem. 2014, 79 (9), 3747–3751.

[112] a) A. O. Magnusson, M. Takwa, A. Hamberg, K. Hult, Angew. Chem., Int. Ed. 2005, 44, 4582–4585; b) A. O. Magnusson, J. C. Rotticci-Mul-der, A. Santagostino, K. Hult, ChemBioChem 2005, 6, 1051–1056; c) K. Engström, M. Vallin, P. O. Syren, K. Hult, J. -E. Bäckvall, Org. Bi-omol. Chem. 2011, 9, 81–82; d) K. Engström, M. Vallin, K. Hult, J. -E. Bäckvall, Tetrahedron 2012, 68, 7613–7618.

[113] K. Linderstrøm-Lang, M. Ottesen, Nature 1947, 159, 807–808. [114] D. Kumar, X. Savitri, N. Thakur, R. Verma, T. C. Bhalla, Res. Micro-

biol. 2008, 3, 661–672. [115] a) M. -J. Kim, Y. I. Chung, Y. K. Choi, H. K. Lee, D. Kim, J. Park, J.

Am. Chem. Soc. 2003, 125, 11494–11495; b) L. Borén, B. Martín-Matute, Y. Xu, A. Córdova, J. -E. Bäckvall, Chem. Eur. J. 2006, 12, 225–232; c) L. K. Thalén, A. Sumic, K. Bogár, J. Norinder, A. K. Å. Persson, J. -E. Bäckvall, J. Org. Chem. 2010, 75, 6842–6847.

[116] D. L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, F. Frolow, S. M. Franken, M. Harel, S. J. Remington, I. Silman, J. Schrag, J. L. Sussman, K. H. G. Verschueren, A. Goldman, Protein Eng. 1992, 30, 321–331.

[117] a) W. -T. Hsu, D. S. Clark, Biotechnol. Bioeng. 2001, 73, 231–237; b) V. Stepankova, S. Bidmanova, T. Koudelakova, Z. Prokop, R. Chaloupkova, J. Damborsky, ACS Catal. 2013, 3, 2823–2836; c) J. Mukherjee, M. N. Gupta, Tetrahedron Lett. 2015, 56, 4397–4401.

�

[118] Y. L. Khmelnitsky, R. Hilhorst, A. J. Visser, C. Veeger, Eur. J. Bi-

ochem. 1993, 211, 73–77. [119] N. B. Philip, Biochim. Biophys. Acta Protein Struct. Mol. Enzymol.

2000, 1543, 203–221. [120] M. Dickman, R. C. Lloyd, J. B. Jones, Tetrahedron: Asymmetry 1998,

9, 4099–4102. [121] M. Rachinger, S. Volland, F. Meinhardt, R. Daniel, H. Liesegang, Ge-

nome Announc. 2013, 1, e00525–13. [122] J. T. Maynes, M. Cherney, A. Qasim, J. Laskowski, N. G. James, Acta

Crystallogr. Sect. D 2005, 61, 580–588. [123] a) D. A. Estell, T. P. Graycar, J. A. Wells, J. Biol. Chem. 1985, 260,

6518–6521; b) R. Bott, M. Ultsch, A. Kossiakoff, T. Graycar, B. Katz, S. Power, J. Biol. Chem. 1988, 263, 7895–7906; c) H. Gron, L. M. Bench, S. Branner, K. Breddam, Eur. J. Biochem. 1990, 194, 897–901.

[124] S. D. Power, R. M. Adams, J. A. Wells, Proc. Natl. Acad. Sci. U. S. A. 1986, 83, 3096–3100.

[125] J. J. Perona, C. S. Craik, Protein Sci. 1995, 4, 337–360. [126] a) D. Despotovic, L. Vojcic, R. Prodanovic, R. Martinez, K. -H.

Maurer, U. Schwaneberg, J. Biomol. Screening 2012, 17, 796–805; b) D. Despotovic, L. Vojcic, M. Blanusa, K. -H. Maurer, M. Zacharias, M. Bocola, R. Martinez, U. Schwaneberg, J. Biotechnol. 2013, 167, 279–286.

[127] a) E. J. Corey, A. Guzman-Perez, Angew. Chem., Int. Ed. 1998, 37, 388–401; b) J. Christoffers, A. Mann, Angew. Chem. Int. Ed. 2001, 40, 4591–4597; c) J. Christoffers, A. Baro, Quarternary Stereocenters, Wiley-VCH, Weinheim, 2005.

[128] a) J. L. Bicas, A. P. Dionisio, G. M. Pastore, Chem. Rev. 2009, 109, 4518–4531; b) E. Brenna, C. Fuganti, F. G. Gatti, S. Serra, Chem. Rev. 2011, 111, 4036–4072.

[129] D. O’Hagan, N. A. Zaidi, Tetrahedron: Asymmetry 1994, 5, 1111–1118.

[130] U. T. Bornscheuer, C. Bessler, R. Srinivas, S. Hari Krishna, Trends Bi-otechnol. 2002, 20, 433–437.

[131] A. Liese, K. Seelbach, C. Wandrey, Industrial Biotransformations, Wiley-VCH, Weinheim, 2000.

[132] a) E. Henke, J. Pleiss, U. T. Bornscheuer, Angew. Chem., Int. Ed. 2002, 41, 3211–3213; b) B. Heinze, R. Kourist, L. Fransson, K. Hult, U. T. Bornscheuer, Prot. Eng. Des. Sel. 2007, 20, 123–131; c) J. Rehdorf, G. A. Behrens, G-S. Nguyen, R. Kourist, U. T. Bornscheuer, Appl. Micro-biol. Biotechnol. 2012, 93, 1119–1126.

[133] S. H. Krishna, M. Persson, U. T. Bornscheuer, Tetrahedron: Asymmetry 2002, 13, 2693–2696.

��

[134] T. Görbe, R. Lihammar, J. -E. Bäckvall, Chem. Eur. J. 2018, 24, 77–

80. [135] D. J. Ericsson, A. Kasrayan, P. Johansson, T. Bergfors, A. G. Sand-

ström, J. -E. Bäckvall, S. L. Mowbray, J. Mol. Biol. 2008, 376 (1), 109–119.

[136] Y. Wikmark, M. S. Humble, J. -E. Bäckvall, Angew. Chem., Int. Ed. 2015, 54, 4284–4288.