Citation
The Effects of Active Site Mutations on Old Yellow Enzyme 1

Material Information

Title:
The Effects of Active Site Mutations on Old Yellow Enzyme 1
Creator:
Conerly, William C
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (124 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Chemistry
Committee Chair:
Stewart, Jon D
Committee Members:
Richards, Nigel G
Horenstein, Nicole A
Mcelwee-White, Lisa A
Maupin, Julie A
Graduation Date:
5/5/2012

Subjects

Subjects / Keywords:
Active sites ( jstor )
Alcohols ( jstor )
Alkenes ( jstor )
Amino acids ( jstor )
Dehydrogenases ( jstor )
Enantiomers ( jstor )
Enzyme substrates ( jstor )
Enzymes ( jstor )
Esters ( jstor )
Mutagenesis ( jstor )
Chemistry -- Dissertations, Academic -- UF
biocatalysis -- oye1
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Chemistry thesis, Ph.D.

Notes

Abstract:
Biocatalysis is increasing in demand in the industrial setting. Many enzymes, however, are not suited for industrial chemistry without adaptations made to their amino acid sequence. Though there has been some success predicting which amino acid replacements would give the desired outcome, many labs still find error-prone PCR and high-throughput screening to be more practical. Relying on these methods requires them to search through large libraries of enzyme mutants. These libraries often reach thousands of mutations. Due to the lack of equipment and resources, or the high cost involved, it is impractical for many labs to analyze data sets that large. Previous studies have shown that mutations at or near the active site can have the greatest impact on catalytic properties. In an effort to optimize an enzyme using scaled down libraries, mutations were made to the old yellow enzyme 1 (OYE1). OYE1 is an alkene reductase that can reduce the double bond of many compounds but is more closely associated with the reduction of a, Beta unsaturated ketones and aldehydes. One study focused on improving the substrate scope by changing the reductive potential of the FMN cofactor. To do so, we used the substitution T37A, but this did not result in improvement of the enzyme's properties. Another focus was to change the enantioselectivity of the enzyme. To do this, a mutant library was made at position 116. We attempted to find if these mutants would change the enantioselectivity for some 2,5 alkyl substituted cyclohexenone compounds, and evidence was found to suggest that this does take place. These mutants were also screened with some phenyl substituted alkenes and also Baylis-Hillman adduct compounds. While the investigation featuring phenyl substituted alkenes did not yield any results towards altered stereochemical outcomes, the work featuring Baylis-Hillman adduct compounds revealed mutants with both improved conversion percentages compared to OYE1 as well as improved enantioselectivity. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Stewart, Jon D.
Statement of Responsibility:
by William C Conerly.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Conerly, William C. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
864879773 ( OCLC )
Classification:
LD1780 2012 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 THE EFFECTS OF ACTIVE SITE MUTATIONS ON OLD YELLOW ENZYME 1 By WILLIAM COLIN CONERLY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 William Colin Conerly

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank all of the individuals whose advice has led me toward the path would never earn a living as a musician, and my father (a musician) for agreeing wit h her. I will be forever indebted to the professors at Northwestern State who urged me to continue my education through graduate school. I would like to acknowledge all of the Stewart group members who helped me through this experience: Santosh Kumar Padh i, who in a short amount of time taught me sterile technique and how to properly handle proteins; Despina Bougiouku, for her sound advice laced with priceless sarcasm; and to the rest of the Stewart group members, Dimitri Dacier, Brad Sullivan, Neil Stowe, Adam Rothman, and Adam Walton, for their humor and understanding. I would like to thank my committee members. To Nicole Horenstein, Nigel Richards, Lisa McElwee White, and Julie Maupin Furlow, your advice to me during my oral has been sincerely appreciat ed. And to Dr. Stewart, thank you for your guidance and for the advice I eventually learned was always right after many times being reluctant to follow it. I would like to thank my family, especially my wife for her constant support through this process. And I would like to thank my son, who is a joyful reminder that it is time to move forward with my life.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 TECHNIQUES AND MOTIVATIONS FOR ENGINEERED PROTEINS .................. 17 Background and Significance ................................ ................................ ................. 17 Mutagenesis Methods ................................ ................................ ............................. 19 Examples of Carbonyl Reductions ................................ ................................ .......... 20 Changing Cofactor Dependence ................................ ................................ ...... 21 Substrate Scope ................................ ................................ ............................... 22 Solvent Stabilit y ................................ ................................ ................................ 26 Product Tolerance ................................ ................................ ............................ 27 Thermostability ................................ ................................ ................................ 29 Industrially Useful Carbonyl Reductase Variants ................................ .............. 29 Synthetic Enzymes ................................ ................................ ........................... 32 Direct anchoring ................................ ................................ ......................... 32 Indirect anchoring ................................ ................................ ...................... 33 Alkene Reductases ................................ ................................ ................................ 33 Mechanism ................................ ................................ ................................ ....... 35 Substrates for the Oxi dative Half Reaction ................................ ....................... 35 Product Stereochemistry ................................ ................................ .................. 36 Overview ................................ ................................ ................................ ................. 40 2 MEASURING REVERSED SUBSTRATE BINDING OF CARVONE MIMIC COMPOUNDS ................................ ................................ ................................ ........ 52 W116 Mutants ................................ ................................ ................................ ......... 52 Results and Discussion ................................ ................................ ........................... 53 QuikChange site directed mutagenesis ................................ ............................ 53 Reduction of 2,5 disubstituted cyclohexenones ................................ ............... 53 Materials and Methods ................................ ................................ ............................ 57 Subcloning From a Saccharomyces cerevisiae Expression Plasmid ............... 57 QuikChange Site Directed Mutagenesis ................................ .......................... 58 Protein Purificati on Procedure ................................ ................................ .......... 58 Experimental Procedure for Screening ................................ ............................. 59

PAGE 6

6 Synthesis and Characterization ................................ ................................ ........ 60 3 REDUCTIONS OF ACIDS AND ESTERS ................................ .............................. 71 Reduction of Halogenated Acids ................................ ................................ ............. 71 Results and Discussion ................................ ................................ ........................... 72 Esters with OYE1T37A ................................ ................................ ..................... 72 Haloacrylic Acids ................................ ................................ .............................. 73 Materials and Methods ................................ ................................ ............................ 73 Mutagenesis ................................ ................................ ................................ ..... 74 Protein Purification Procedure ................................ ................................ .......... 74 Procedure for Haloacrylate Reductions ................................ ............................ 75 4 ALCOHOL FUNCTIONAL IZED ALKENES ................................ ............................. 83 Baylis Hillman Adducts ................................ ................................ ........................... 83 Results and Discussion ................................ ................................ ........................... 83 Screening of Enzymes ................................ ................................ ..................... 83 Crystallographic Study ................................ ................................ ..................... 85 Materials and Methods ................................ ................................ ............................ 86 Subcloning From a Saccharomyces cerevisiae Expression Plasmid ............... 86 QuikChange Site Directed Mutagenesis ................................ .......................... 87 Protein Purification Procedure ................................ ................................ .......... 87 Experimental Procedure for Screening ................................ ................................ ... 89 Small Scale Reactions ................................ ................................ ..................... 89 Reduction of Alkene 96 by GST OYE 2.6 Crude Extract ................................ .. 89 Reduction of Alkene 97 by GST OYE2.6 Whole Cells ................................ ..... 90 5 BETA PHENYL COMPOUNDS ................................ ................................ .............. 95 Ph enyl Compounds ................................ ................................ ............................. 95 Results and Discussion ................................ ................................ .................... 96 Materials and Methods ................................ ................................ ..................... 97 Experiment al Procedure for Screening ................................ ................................ ... 97 6 FUTURE WORK ................................ ................................ ................................ ... 101 Future Work ................................ ................................ ................................ .......... 101 Concluding Remarks ................................ ................................ ............................. 102 APPENDIX A SCHEME FOR QUIKCHANGE METHOD ................................ .......................... 104 B SCHEME FOR RESTRICTION DIGESTIONS AND LIGATIONS ......................... 105 C SEQUENCING CHROMATOGRAMS ................................ ................................ ... 106

PAGE 7

7 D PR IMERS USED ................................ ................................ ................................ .. 111 E NMR DATA OF PREVIOUSLY UNREPORTED COMPOUNDS ........................... 112 LIST OF REFERENCES ................................ ................................ ............................. 118 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 124

PAGE 8

8 LIST OF TABLES Table page 1 1 Goals set by Codexis set for the production of ( R ) 2 methylpentanol ................. 41 1 2 Targets set by Codexis for a biocatalytic route to tetrahydrothiophene 3 ol ....... 41 2 1 Starting materials and products for the synthesis of the compounds chosen to mimic carvone. ................................ ................................ ............................... 63 2 2 Enzymes that gave 95% conversion or better at least once during the screening ................................ ................................ ................................ ............ 63 3 1 Data corresponding to the comparison of OYE1 and OYE1T37A. ..................... 78 4 1 Results of the Bay lis Hillman ad duct reductions. ................................ ................ 92

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Examples of alcohol dehydrogenase enzymes that are selective for NADP or NAD ................................ ................................ ................................ ................... 42 1 2 Reaction stereochemistry of Sporobolomyces salmonicolor carbonyl reductase. ................................ ................................ ................................ ........... 42 1 3 Synthetically desired reaction performed by Thermoanaerobacter ethanolicus ................................ ................................ ................................ ...... 42 1 4 Resolution strategy using the alcohol dehydrogenase from Thermoanaerobacter ethanolicus ................................ ................................ ...... 43 1 5 Reactio n scheme of galactose oxidase. ................................ ............................. 43 1 6 Secondary aryl alcohol o x idations by engineered galactose oxidase mutants res olving secondary aryl alcohols. ................................ ................................ ...... 43 1 7 Reduction of 2,5 hexanedione by Pyrococcus furiosus alcohol dehydrogenase. ................................ ................................ ................................ .. 43 1 8 Ketone reduction using Rhodococcus erythropolis alcohol dehydrogenase MAK154. ................................ ................................ ................................ ............. 44 1 9 NADPH regene ration glucose dehydrogenase oxidation followed by the spontaneous hydrolysis of the gl uconolactone to gluconic acid. ........................ 44 1 10 pH in dicator assay used for high throughput screening ................................ ..... 44 1 11 Reaction scheme of Penicillium citrinum Beta keto ester reductase with 4 bromo 3 oxobutyrate (BAM) ................................ ................................ ............... 45 1 12 Process patented by BASF for the production of ( R ) 2 methylpent anol. ........... 45 1 13 Process developed by Codexis for the production of ( R ) 2 methylpentanol. ...... 45 1 14 Pfizer process for R tetrahydrothiophene 3 ol. ................................ ................. 46 1 15 Codexis process for ( R ) tetrahydrothiophene using a ketone reductase. ........... 46 1 16 Structure of Montelukast sodium ................................ ................................ ........ 46 1 17 Asymmetric carbonyl reduction as a method for indtoducing the chiral center of montelukast. ................................ ................................ ................................ ... 47 1 18 Hydrogenation/isomerization of cis stilbe ne ................................ ...................... 47

PAGE 10

10 1 19 Biotin bound ligand for anchoring a Rhodium complex into avidin ..................... 47 1 20 BINAP structures and a basic representation of a Noyori Hydrogenation .......... 48 1 21 PyMol image of OYE1 co crystallized with p hydroxybenzaldehyde. ................. 48 1 22 Mechanism of OYE1. ................................ ................................ .......................... 49 1 23 Typ ical stereochemical outcome of both chiral centers. ................................ ..... 49 1 24 Enantiocomplementary reductions of perillaldehyde. ................................ ......... 49 1 25 Nitroalkene reduction by alkene reductases. ................................ ...................... 50 1 26 Complementary stereochemical outcomes using substrate engineering. ........... 50 1 27 Synthesis of Roche ester precursors using alkene reductases. ......................... 50 1 28 Sterecomplementar y reductions of ( S ) carvone by wild t ype OYE1 and the W116I mutant. ................................ ................................ ................................ .... 51 1 29 Altered stereochemistry of alkene reductase Yqjm using i terative saturation mutagenesis. ................................ ................................ ................................ ...... 51 2 1 Pymol stereoimage of the OYE1 active site with p hydroxybenzaldehyde as a bound inhibitor. ................................ ................................ ................................ ... 64 2 2 Structures of R and S carvone ................................ ................................ ............ 64 2 3 Error of QuikChange method. ................................ ................................ .......... 65 2 4 Synthesis of the Carvone mimic compounds. ................................ ..................... 65 2 5 Chiral GC MS result for the L uche reduction. ................................ ..................... 66 2 6 Chiral GC result with 2,5 dimethylcyclohexanon e .. ................................ ............ 66 2 7 All possible outcomes when reducing 2,5 dimet hylcyclohexenone. .................. 67 2 8 Palladium reduced standards of substituted c yclohexenone compounds on nonchiral GC MS. ................................ ................................ .............................. 67 2 9 Data for the reduction of 2,5 dimethylcyclohexenone. ................................ ........ 68 2 10 Data fo r the reduction of 2 methyl 5 ethylcyclohexenone ................................ ... 68 2 11 Data for the reduction of 2 metyl 5 propylcyclohexenone ................................ .. 69 2 12 Data for the reduction of 2 methylcyclohexenone ................................ ............... 69

PAGE 11

11 2 13 General reaction and NADPH regeneration pathway ................................ ......... 70 2 14 Synthesis pathway a nd GC results. ................................ ................................ .... 70 3 1 Amino acids that hydrogen bond to the isoalloxazine ring of flavin. .................... 79 3 2 neighboring electron withdrawing halides ................................ ........................... 79 3 3 Methacrylic acid chromatograms. ................................ ................................ ....... 80 3 4 Chlo roacrylic acid chromatogram. ................................ ................................ ...... 81 3 5 Derivatized product of the OYE1 Bromoacrylic acid r eduction on nonchiral GC MS. ................................ ................................ ................................ .............. 81 3 6 Synthesis of trans 2,2,2 trifluoroethyl but 2 enoate ................................ ............ 82 4 1 Electron density map with the amino ac ids from OYE1W116I modeled in. ........ 93 4 2 Substrate modeled into the electron density map. ................................ .............. 93 4 3 Figure demonstrating the different conformations of I116, H191, and Y196. ...... 94 5 1 methyl trans cinnamaldehyde binding modes In OYE1. ................................ .. 99 5 2 Synthesis pathway for the racemic standard of reduced alpha methyl trans cinnamaldehyde ................................ ................................ ................................ 99 5 3 Enzyme reductions with Methyl trans cinnamaldehyde ................................ 100 5 4 Enzyme reductions with (Z) ethyl 2 fluoro 3 phenylacrylate ............................. 100 C 1 W116G ................................ ................................ ................................ ............. 106 C 2 W116H ................................ ................................ ................................ ............. 106 C 3 W116Q ................................ ................................ ................................ ............. 106 C 4 W116M ................................ ................................ ................................ ............. 107 C 5 W116Y ................................ ................................ ................................ .............. 107 C 6 W116S ................................ ................................ ................................ .............. 107 C 7 W116V ................................ ................................ ................................ .............. 108 C 8 W116T ................................ ................................ ................................ .............. 108

PAGE 12

12 C 9 W116N ................................ ................................ ................................ ............. 108 C 10 W116D ................................ ................................ ................................ ............. 109 C 11 W116E ................................ ................................ ................................ .............. 109 C 12 W116R ................................ ................................ ................................ ............. 109 C 13 W116C ................................ ................................ ................................ ............. 110 C 14 W116K ................................ ................................ ................................ .............. 110 C 15 T37A ................................ ................................ ................................ ................. 110 E 1 1 H NMR of 2,5 dimethyl cyclohexenone ................................ ........................... 112 E 2 13 C NMR of 2,5 dimethylcyclohexenone. ................................ .......................... 113 E 3 1 H NMR of 2 methyl 5 propylcyclohexenone. ................................ ................... 114 E 4 13 C NMR of 2 methyl 5 propylcyclohexenone. ................................ ................. 115 E 5 1 H NMR of trans 2 (1 phenylprop 1 en 2 yl) 1,3 dioxolane. ............................. 116 E 6 13 C NMR of trans 2 (1 phenylprop 1 en 2 yl) 1,3 dioxolane. ............................ 117

PAGE 13

13 LIST OF ABBREVIATION S BINAP bis(diphenylphosphino) binaphthyl bp Base pair CASTing Combinatorial active site saturation test de D iastereomeric excess DMAP 4 Dimethylaminopyridine DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid dPE D Pseudoephedrine EAM (RS) 1 phenyl 1 keto 2 ethylaminopropane ee Enantiomeric excess FMN F lavin mononucleotide GC Gas chromatography GC FID Gas chromatography Flame ionization detection GC MS Gas chromatography Mass Spectrometry GDH Glucose dehydrogenase HPLC High pressure liquid chromatography IPA Isopropylalcohol IPTG I so propylthio D g alactoside (IPTG) ISM Iterative saturation mutagenesis KER Keto ester reductase NAD + Nicotinamide adenine dinucleotide NADP + Nicotinamide adenine dinucleotide phosphate NMR Nuclear magnetic resonance OYE1 Saccharomyces pastorianus Old Yellow Enzyme 1

PAGE 14

14 PCR Polymerase chain reaction PETN Pentaerythritol tetranitrate ProSar Protein sequence activity relationships SDR Short chain dehydrogenase reductase TBDMS Tert butyldimethylsilyl TeSADH Thermoanerobacter ethanolicus dehydrogenase THF Tetrahydrofuran TLC Thin layer chromatography

PAGE 15

15 Abstract of Dissertation Presente d to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EFFECTS OF ACTIVE SITE MUTATIONS ON OLD YELLOW ENZYME 1 By WILLIAM COLIN CONERLY May 2012 Chair: Jon D. Stewart Major: Chemistry Biocatalysis is increasing in demand in the industrial setting. Many enzymes, however, are not suited for industrial chemistry without adaptations made to their amino acid sequence. Though there has been some success pred icting which amino acid replacements would give the desired outcome, many labs still find error prone PCR and high throughput screening to be more practical. Relying on these methods requires them to search through large libraries of enzyme mutants. Thes e libraries often reach thousands of mutations. Due to the lack of equipment and resources, or the high cost involved, it is impractical for many labs to analyze data sets that large. Previous studies have shown that mutations at or near the active site can have the greatest impact on catalytic properties. In an effort to optimize an enzyme using scaled down libraries, mutations were made to the old yellow enzyme 1 (OYE1). OYE1 is an alkene reductase that can reduce the double bond of many compounds bu t is more closely associated with the reduction of unsaturated ketones and aldehydes. One study focused on improving the substrate scope by changing the reductive potential of the FMN cofactor. To do so, we used the substitution T37A, but this did no

PAGE 16

16 properties. Another focus was to change the enantioselectivity of the enzyme. To do this, a mutant library was made at position 116. We attempted to find if these mutants would change the enantioselectivity for som e 2,5 alkyl substituted cyclohexenone compounds, and evidence was found to suggest that this does take place. These mutants were also screened with some phenyl substituted alkenes and also Baylis Hillman adduct compounds. While the investigation featurin g phenyl substituted alkenes did not yield any results towards altered stereochemical outcomes, the work featuring Baylis Hillman adduct compounds revealed mutants with both improved conversion percentages compared to OYE1 as well as improved enantioselect ivity.

PAGE 17

17 CHAPTER 1 TECHNIQUES AND MOTIV ATIONS FOR ENGINEERE D PROTEINS Background and S ignificance Biological systems can respond differently to the different optical isomers of compounds. For example, enantiomeric compounds can smell different ly as in S and R carvone (caraway and spearmint respectively). 1 S tudies have also shown that giving medications as one optically pure isomer increases drug potency and decreases side effects. 2 However, many drugs are still sold as racemates or mixtures of diastere omers because the added cost of producing optically pure compounds is too high. This has stimulated interest in developing new methods for pr oducing chiral compounds. One particularly fruitful strategy to form chiral sp 3 centers is the hydr ogenation of double bonds, particularly carbonyl and alkene moieties. Many metal catalyzed examples of these reactions have been published, more recently, enzymes capable of performing these reactions have also been used as the basis of synthetic methodology. Whole microbial cells have been popular as biocatalysts 3 In these cases, intracellular enzymes reduce ketones an d alkenes to chiral sp 3 centers. The y require only a cheap carbon source such as glucose for energy in order to regenerate the NADPH and NADH reducing equivalents. 4 While such processes can be employed commercially 5 they also have drawbacks. Since cells express many enzymes enantiomeri c purity can be poor due to competing catalytic activities. In order to solve this problem, one can delete all of the competing proteins, though this process is laborious when many proteins are involved. 6 Alternatively, purified enzymes can be used inste ad. In order to simplify purifying enzymes, it is possible to clone enzymes into an E. coli expression vector that is tagged for affinity chromatography. 7 Affinity tagging

PAGE 18

18 allows each protein to be purified by a common protocol without the need to develo p a customized purification procedure for each enzyme. Easy enzyme purification allows one to focus on characterizing the reactions they catalyze, rather than on the mechanics of isolation. In order to be practical bioca talysts need to be robust, cost effective, and environmentally friendly. 8 Inexpensive bulk materials therefore, represent especially challenging targets for biocatalytic methods since the petrochemical industry is extremely cost effective. 9 Conversely, fine chemical manufacturing is a more logical arena for biocatalysis. Th e pharmaceutical industry currently uses more than ten pr ocesses that utilize carbonyl reductases 10 This is usually possible because priority is placed on compound purity above all else However, in some instance s, d ue to their inherent regioselective and stereoselective properties, p rotection and deprotection pathways can be avoided, circumventing synthetic steps thus lowering manufacturing costs 10 It is for these reasons t he popularit y of biocatalysts is incr easing. E nzymes are not usually suited for industrial chemis try in their native form They have evolved over millions of years in support of their host organism, and not for the chemical process for which they may be applied 11 One problem that often ari se s is that only one product stereoisomer can be easily accessed from the native biocatalyst In the case of carbonyl reductases, enzymes that produce both alcohol enantiomers are available. These catalysts are classified as being Prelog ( S ) or anti Prel og ( R ), named 12 The situation is different for alkene reductases, however, and most of these enzymes reduce substrates with the same stereoselectivity. 13 This class of enzyme is

PAGE 19

19 also burdened by expensive cofactor dependence, limite d substrate scope, and poor stability in the presence of organic substances. Our group has focused its attention on overcoming all of these problems. This chapter will review some methods of enzyme engineering as they relate to producing chiral compounds using carbonyl and alkene reductases with an emphasis on synthetic application. The strategies range from replacing catalytic metals of enzymes to more practical and even industrially relevant mutagenesis. The final discussion of this chapter will be the developing field of alkene reductases with a focus on the old yellow enzyme family. Mutagenesis Methods A molecular biologist has many tools that can be utilized for the creation of a better p rotein. Error prone PCR is commonly employed since it is simpl e and free of intellectual property constraints 14 This method involves replicating DNA with a low fidelity process, but only works well when there is a good high throughput screening method available to identify a useful variant. Since many of the resulting mutations will have little effect or will even be d etrimental, researchers routinely examine thousands of samples at a time in order to find the desired result. Another key disadvantage to error prone PCR is that not all codons are accessible by base changes at a single position. For example, there is on ly one codon that codes for tryptophan (TGG). 15 If the starting codon codes leucine using CTT, it is unlikely that all three positions will be changed by methods that create sequence changes at random positions. For this reason, error prone PCR favors am ino acids encoded by multiple codons. The large screening efforts demanded by completely random mutagenesis has motivated the search for better ways to use knowledge of enzyme structure and properties to lessen reliance on blind

PAGE 20

20 screening. This problem h as led to the development of rational mutagenesis methods in an effort to shrink the necessary screening library size. One noteworthy example is provided by Codexis, which has applied the method of protein sequence activity relationships (ProSAR) to impro ve ketoreductases, and another is Iterative Saturation Mutagenesis (ISM), developed by the Reetz group. Reetz and coworkers developed ISM as a way to implement Combinatorial Active Site Saturation Test (CASTing). 16 CASTing rests on the simple philosophy t hat if a scientist wants to influence the way an enzyme binds its substrate(s), then success is more probable if one concentrates on making mutations close to the active site. 17 Such mutations to the active site are made at more than one position at a tim e (in combination). The major drawback to traditional CASTing is that simultaneous saturation mutagenesis at multiple positions leads to very large libraries and a correspondingly large screening effort. While ISM also relies on more than one mutation wi thin the active site, each position is examined individually, and then the best substitutions are combined. This involves the creation of multiple first generation rounds of mutagenesis. 18 Examples of Carbonyl Reductions Carbonyl reductions can introduce a single chiral center into a molecule or more than one if dynamic kinetic resolution is involved. 19 The new functional group can be useful for other downstream applicati ons. 20 The following sections will give examples of some evolved carbonyl reductases categorized by the reasons for their directed evolution.

PAGE 21

21 Changing Cofactor Dependence Carbonyl reductase enzymes may use either NADPH or NADH to supply the hydride requi red for alcohol formation Enzymes are often highly selective for one cofactor or the other ; unfortunately, NADPH is more expensive than NADH by an order of magnitude. The expense can be minimized by using a cofactor recycling system, but industrial sc al es still require large cofactor quantities This has motivated a search to uncover the origin of NADPH selectivity, with the goal of switching to NADH. On the other hand, it is important to preserve the original catalytic efficiency in the engineered var iants 21 These issues have been addressed in the family of carbonyl reductases called short chain dehydrogenase reductases (SDR). Previous X ray crystallography studies have shown that a conserved aspartate side chain forms a hydrogen bond to the adenine for the NADH dependent enzymes, 22 while in NADPH dependent reductases, a threonine at the same position hydrogen bonds to the phosphate through a water molecule. 23 Nakanishi et al. changed the coenzyme preference from NADPH to NADH in mouse lung carbonyl reductase using only a single mutation (T38D). 24 The Kcat value reported for the mutant enzyme with NADH (3.6s 1 ) was an improvement over the wild type value with NADPH (2.4 s 1 ). No changes in stereoselectivity were reported. Xiao applied the results of sequence and structural analysis summarized above to a carbonyl reductase from Candida parapsilosis 25 Using all nine possible combinations of the mutations S67D, H68D, and P69D, they were able to switch the selectivity of the enzyme to NADH dependence Interestingly, the enzyme also switched its enantioselectivity. The wild S

PAGE 22

22 alcohols. Using 2 hydroxyacetophenone as a model substrate, the mutant that produced the best results was S67D/H68D which ga ve 83% yield and 90% ee ( R ). These same concerns were also addressed in the R specific alcohol dehydrogenase from Lactobacillus brevis which was first discovered during a search for synthetically interesting alcohol dehydrogenases. 26 It is an interesting enzyme because its substrate scope includes diaryl ketones and not just substituted acetophenones like a single mutation in its active site G37D (Figure 1 1 ). 27 This accomplishment was heralded as an important advance by the authors; unfortunately, the switch in cofactor specificity came with a rate penalty. This problem was overcome by a different research group who performed computational modeling on the enzyme. 21 model suggested that four possible changes might work, but experimentally, three of them failed. Based on computations, the A38P mutation altered the hydrogen bonding pattern of main chain amide bonds to NADPH, and this mutation gr eatly increased the rate of NADH oxidation with a Vmax of 80.3 U / mg. This system was further probed by generating the double mutant G37D / A38P, but this variant proved to be inferior to the single mutant A38P. Substrate Scope Using docking studies the enantioselectivity of the enzyme S porobolomyces salmonicolor carbonyl reductase (SSCR) was enhanced. It had been previously reported that SSCR reduced a number of ketones to chiral alcohols with good enantiomeric purity (> 95% ee) 28 By contrast, it reduced para substituted acetophenones with poor selectivity (14 59% ee favoring R ). 29 Using the docking program ICM Pro 3.4.9.d, computational binding studies were carried out with 4

PAGE 23

23 methoxyacetophenone as an initial test substrate. These efforts revea led that the binding modes for Prelog and anti Prelog hydrogenation were energetically close. 30 The docking studies also showed that the residues Q245 and M242 lay very near the par a position of the substrate. When the researchers mutated Q245 to all pos sible amino acids, binding of 4 methoxyacetophenone changed, and stereoselectivity was enhanced by three mutations. The H, L, and P replacements afforded higher ee values and also changed the enantiomeric preference from R to S (Figure 1 2). The latter a lteration was unexpected. However, since the wild type enzyme provided only 57% ee for this substrate, any increase in optical purity was still a welcome result. Other para substituted substrates were studied, and the same trend was observed. All had i mproved ee values, many greater than 90% along with higher specific activity in several cases. Unfortunately, the mutant displayed poorer stereoselectivity for substrates reduced by the wild type enzyme with high stereoselectivity. Thermoan a erobacter etha nolicus dehydrogenase (TeSADH) is a secondary alcohol dehydrogenase that is highly thermostable ( half life of 1.7 h at 90C ) and solvent stable. 31 These robust properties make it attractive for industrial processes. Ziegelmann Fjeld et al. used the exis ting crystal structure published under the name Thermoanaerobacter brockii dehydrogenase (seq uence data had confirmed they w ere identical) to choose a site to mutate and broaden the scope of the enzy researchers w ere specifically interest ed in reducing phenylacetone to the S alcohol 1 phenyl 2 propanol (Figure 1 3) since this is an important precursor for the synthesis of amphetamine and derivatives of amphetamine. However, no thermo or solvent tolerant enzyme is currently used for this reduction, and TeSADH cannot naturally perform this

PAGE 24

24 reaction. Though the enzyme can produce many secondary alcohols including ones with long aliphatic chains, such as 2 decanol, it is not reactive toward ketone substrates containing rings with more than three carbons. 32 Within the crystal structure the active site has two binding pockets, designated small and large. The small binding pocket has a higher affinity for alkyl groups than the large pocket. The active site has a tryptophan in it that is sterically bulky and the researchers used site directed mutagenesis to change it to an alanine (W110A). This single p oint mutation was successful in creating an enzyme that could reduce phenylacetone to the S alcohol with 99% conversion and 99% ee. To further explore the selectivity of the enzyme the researchers used the enzyme to oxidize a selection of racemic alcohol s and resolved the R enantiomer s to 99% ee (Figure 1 4) The driving force for the reaction uses solvent as a direct resource for reducing equivalents to force the equilibrium. If oxidizing the alcohol to resolve enantiomers is desired then acetone in w ater (10% v / v) was used as the solvent If reducing the ketone to a chiral alcohol is desired then the solvent consists of 2 propanol in water (30% v / v). 33 This mutation also allowed TeSADH to reduce/oxidize other secondary alcohols that were attach ed to a ring. An extreme example of changing the substrate scope of an enzyme is galactose oxidase. Whereas most researchers choose enzymes with activities that are close to the desired outcome, galactose oxidase has been through enough generations of mut agenesis to include substrates very different than its original. Galactose oxidase is an oxido reductase enzyme that catalyzes the oxidation of the alcohol to an aldehyde at the C 6 position of galactose (Figure 1 5) 34 Though the name is similar to gluco se oxidase the two act on different positions of their respective sugars (glucose oxidase

PAGE 25

25 oxidi zes the anomeric carbon to yield a lactone ). Since hydrogen bonding patterns are often used for enzyme recognition, g alactose oxidase is very specific for gala ctose. Also, s ince sugars are biologically active they are interesting molecules for synthetic study as drug target molecules. Sugars are a great example of the need for regioselective chemical methods, otherwise, protection and deprotection steps must be utilized to avoid oxidizing the wrong hydroxyl group the lack of an available enzyme that would oxidize glucose at the 6 position. Since galactose oxidase acts at the C 6 position they hypothesized that it may be engin eered to accept glucose using mutagenesis methods. These researchers chose to mutate R330 and Q 406 to eliminate the hydrogen bonds that are specific for the hydroxyls on C 4, C 3, and C 2 and W290 which interacts with the carbon backbone. 34 After mutatin g these positions the enzyme no longer had activity for its physiological substrate, galactose, but some mutants were able to oxidize glucose at the C 6 position The enzyme that they found to be best was mutated in the positions R330K, Q406T, and W290F. Starting with the sequence from Turner, and using error prone PCR other researchers were able to change the activity of the enzyme to include a range of secondary alcohols (Figure 1 6) 35 These mutants were quickly identified using a colorimetric screening method on agar plates. Modeled after a previously published method, 36 this method involves treating the agar plate with a peroxidase enzyme and a substrate of that enzyme that will change col or upon oxidation. In this procedure, the plates contained the chromogenic substrate 4 chloronapthol. Peroxidase captured hydrogen peroxide, using it to oxidize 4 chloronapthol so that colonies positive for

PAGE 26

26 oxidase activity appeared dark purple. This al lowed the authors to screen 100,000 mutants per round of mutagenesis and find mutants that were able to resolve 13 racemic alcohols to chiral alcohols and ketones with a Kcat > 200 and 99% ee. Solvent Stability When using purified enzymes tolerance to org anics can be a major concern. The most important force holding an enzyme together is the hydrophobic effect. However, substrates that are required in process chemistry are often only minutely soluble in water. Reaction conditions therefore often include water soluble alcohols or other miscible co solvents to help dissolve the substrate. This solubility problem is often compounded by the need for high concentrations of starting materials or products in the reaction mixture to achieve high volumetric prod uctivities. Typical physiological substrate concentrations are millimolar; however, industrial processes often exceed these levels by 100 times or more This often leads to protein instability. For this reason, h igh substrate concentrations cannot be a c hieved without finding solvent tolerant enzymes 8 There is often a correlation between solvent tolerant and thermotolerant enzymes. 19 Hyperthermotolerant enzymes might therefore seem a logical answer. Unfortunately, many of these enzymes show little or n o catalytic activity at moderate temperatures. One strategy used to find a solvent tolerant enzyme has been to take a hyperthermotolerant enzyme and make it less thermally stable, hoping that the engineered enzyme will maintain a higher degree of solvent tolerance while still catalyzing a reaction at or near room temperature Pyrococcus furiosus is a hyperthermophile that has an optimal growth temperature near 100C. 37 Its alcohol dehydrogenase is capable of reducing diones to the corresponding S,S diols 38 This

PAGE 27

27 enzyme is very robust with reported half life values of 150h at 80C, 22.5h at 90C, and displays its maximum activity at 90C. Chiral diol synthesis is best done at room temperature or near 30C to minimize energy cost. 19 Unfortunately, P. fur iosus ADH has only 5% of its maximal activity at this temperature. Error prone PCR followed by activity screenings were performed to increase the activity at low temperature. The protein with the best activity at 30C contained two mutations, R11L and A1 80V. The substitution R11L is located in the NADPH binding site. This position is not critical for NADPH binding, but it had previously been shown that a hydrophobic residue is preferred there to enhance both cofactor binding and turnover rate. 39 The se cond substitution (A180V) is located in the substrate binding pocket and is also found in many mesophilic alcohol dehydrogenases. 19 It may be important for maintaining the position of the catalytic residue T183. These two mutations increased the specific activity of the enzyme with 2,5 hexanedione at 30C, but maintained enough stability to still allow high concentrations of 2 propanol as both a co solvent and a hydride source. The researchers were interested in producing S,S, 2,5 hexanediol as a chiral building block in downstream applications (Figure 1 7). As may be expected, the doubly mutated enzyme is not as stable as the wild type at 100C. The half life for the mutant enzyme at that temperature is 44 min while that of the wild type is 178 min. T he study failed to report whether the increased rate at 30 C resulted in any penalty in product optical purity. Product Tolerance The amino alcohol dehydrogenase from Rhodococcus erythropolis MAK154 has a substrate scope that includes amino alcohols Fo r example, this enzyme reduced S 1 phenyl 1 keto 2 methylaminopropane to d pseudoephedrine(dPE) (Figure 1 8) 40

PAGE 28

28 Given the commercial interest in this product, this is potentially useful. Unfortunately the enzyme cannot tolerate high product concentrations, which makes it impractical for industrial use. In response, Urano used high th roughput screening and error prone PCR to test new mutants of the enzyme for greater product tolerance. 41 The most interesting aspect of this experiment was the manner in which the reaction was detected. Since the goal of the study was to increase product toler ance, the starting rea ction mixtures contained 80 mg / mL dPE as well as ( R,S ) 1 phenyl 1 keto 2 ethylaminopropane (EAM) which was th e actual substrate for reduction A glucose dehydrogenase NADPH recycling system served as both the hydride source and also a method to monitor the reaction progress. For every molecule of glucose oxidized to gluconolactone, spontaneous hydrolysis of the lactone product (Figure 1 9) yielded a free proton, which caused a p H drop in the reaction mixture. This was used to monitor the activity by u sing phenol red as an indicator (Figure 1 10) Using the colorimetric screen, 5,000 different clones were visually examined in microtiter plates. Approximately 100 clones changed the co lor of the indicator from red to yellow Of those two were selecte d after measuring the bioconversion rate on an HPLC. Each had one single amino acid substitution, either G73S or S214R. To uncover the way in which two mutations were related to stability in the presence of high concentrations of product, the correspondi ng double mutant (G73S, S214R) was prepared. Mutants with the G73S substitution had higher Km values than the wild type, indicating an important function in substrate binding; h owever tests proved the double mutant was the most stable in high concentrati ons of dPE. It also gave the best yield (87%) in reaction conditions.

PAGE 29

29 Thermostability Asako et al. reported Penicillium citrinum keto ester reductase (KER) catalyze the reduction of methyl 4 bromo 3 oxobutyrate (BAM) to ( S ) 4 bromo 3 hydroxybutyrate ( S BHBM) with more than 90% ee (Figure 1 11) 42 This compound is an important intermediate in the synthesis of an inhibitor for the ra te limiting step of cholesterol synthesis. Though the enzyme can reduce the desired substrate it is not useful for industrial synthesis due to its low thermal stability. 43 Using error prone PCR, a library of 4 000 mutants was generated, then screened for reductase activity by measuring consumption of NADPH after heating at 45C for 30 min. From thes e screenings, 4 positions w ere picked as potentially useful for increasing stability. After further site directed mutagenesis to mix and match substitutions at these positions it was decided that the mutant L54Q possessed the most desirable propertie s with respect to stereoselectivity (98% ee S ) and good thermotolerance ( retaining 54% of the initial activity after 2 h at 45C) Industrially Useful Carbonyl Reductase Variants All of the enzymes mentioned previously were mutated in an effort to improve their industrial efficacy. However there is no mention or evidence that these developed enzymes were actually employed commercially Even with the improved properties described above, it is likely that many of them are still not suitable In the follow ing section, engineered carbonyl reductase enzymes that are used in an industrial process will be discussed. One good example was reported by workers at Codexis, who described the development of a carbonyl reductase suitable for producing ( R ) 2 methylpenta nol. 44 This alcohol is a key chiral building block for pharmaceuticals 45 and liquid crystals. 46

PAGE 30

30 BASF patented a process for ( R ) 2 methylpentanol in 2006. Their synthesis started from 2 methylpent 2 enal and involved aldehyde hydrogenation, followed by a lkene hydrogenation using a ruthenium complex and H 2 at 200 bar, then a final lipase mediated kinetic resolution to remove the minor enantiomer ( Fig ure 1 12). 47 route involved selective reduction of racemic 2 methylvaleraldehyde by an alcohol deh ydrogenase (Figure 1 13). The route was expected to meet several goals, summarized in Table 1 1. A library of wild type enzymes was initially screened for the desired conversion, although the extent of the library was not disclosed. Lactobac illus k efir KRED gave the best initial hit, although the wil d type enzyme did not meet the performance targets because it provided only 85% ee. 44 After three rounds of mutagenesis, a final mutant met all of the targeted criteria, achieving 98.2% ee and 46% conversion at a substrate loading of 220 g / L. This mutant had a total of five amino acid substitutions (G82S, S96A, E145S, L153Q, I223V). ( R ) tetrahydrothiopene 3 ol is a useful building block for synthesizing the prodrug sulopenem The original Pfizer process starts with L aspartic acid and requires five steps to reach the final target (Figure 1 14) 48 Though the route yields an optical pur ity of 96% ee, it involves hazardous conditions, high energy intermediates, and a noxious reagent. A simple strategy to reach the target compound would involve the asymmetric reduction of tetrahydrothiophene 3 one (Figure 1 15). Unfortunately, the near s ymmetrical structure of the starting ketone makes this reduction highly challenging. Using CBS borane gives only 23% ee, 48 and catalytic hydrogenation using Ir BINAP gave a maximum of 82% ee. 49

PAGE 31

31 To overcome these problems, Codexis set out to develop an enz yme that could reduce tetrahydrothiophene 3 one since no enzyme was available that possessed adequate performance with respect to enough optical purity. 20 Cost and performance targets were also set for this conversion ( Table 1 2 ). After screening a libra ry of enzymes, Lactobacillus kefir ADH again gave the most promising initial results (63% ee). Several rounds of mutagenesis were employed to improve product optical purity. The most dramatic improvements occurred in the first two rounds when 95% ee was reached. Positive changes were only incremental. After a total of eight rounds of mutagenesis, the best mutant contained ten mutations. Using the final enzyme ( R ) tetrahydrothiop h ene 3 ol was produced in 85 89% yield with > 99% purity and 99.3% ee in ba tches larger than 100 kg. Montelukast sodium is an asthma and allergy medication developed by Merck and sold as Singulair (Figure 1 16) 50 The main drawback of the original synthesis was its requirement for ( ) DIP Cl, which is corrosive, moisture sensiti ve, requires a tedious workup, and places a heavy burden on waste stream (Figure 1 17). 51 The key challenge for a biocatalytic alternative was that the precursor molecule is not soluble in water and suitable enzymes were not stable in the presence of hig h cosolvent levels The starting material is most soluble in DMSO, and in an aqueous system, it is most solvent in a tertiary solvent system consisting of 1:5:3 THF/IPA/Water. The latter mixture is desirable for carbonyl reduction, because the isopropyl alcohol (IPA) can also be used as a hydride source for the reduction. To dissolve higher substrate concentrations, the bioconversion was run at 40C. The patent for this process cites several ketoreductases as being possible candidates, but does not give the final

PAGE 32

32 process enzyme. 52 The publication that supports the patent discloses that the final process was not run in the initial ternary solvent mixture. At 45C, the enzyme was most stable in a 1:5:3 mixture of toluene/IPA/100m M pH 8.0 triethanolamine HCl with 2 mM MgSO 4 Conversions can be run for 24 hours in 125 kg batches. Synthetic Enzymes The most extreme example of protein engineering involves the construction of synthetic enzymes by incorporating a catalytic metal comp lex into the protein active site binding pocket. In general, t he catalytic outcomes of these synthetic enzymes have no relation to the origina l function of the natural proteins Rather, catalysis depends completely on the chemistry of the incorporated metal and the ste reoch emical outcome is dictated by the steric and hydrophobic environment provided by the prote in architecture. Artificial metal cofactors can be placed into the active site of a protein using two different strategies, and two examples have been applied to red ucing alkene and carbonyl groups. Direct a nchoring Direct anchoring occurs when a catalytically active metal is bound directly to the active site of the protein. This was done to human carbonic anhydrase II (HCAII), which normally has a zinc ion in its ac tive site. This zinc ion was removed via dialysis using 2,6 pyridinedicarboxylate, t he n the active site was filled with rhodium using dialysis. 53 Other active site mutations w ere made to encourage the release of zinc by removing histidines from the activ e site for some assays. These modifications completely abolished the original catalytic activity of HCAII ( converting CO 2 to bicarbonate ) and allowed the completely unrelated catalytic activity of alkene reduction to be observed. D i fferent mutants w ere s creened with cis stilbe ne. The goal of this study was to

PAGE 33

33 develop a method for alkene reduction. Unfortunately, this has not been a very p ractical approach since the best result ( 80% conversion ) was compromised by the presence of free rhodium in solution which caused isomerization to trans stilbe ne This isomer was not reduced by the enzyme (Figure 1 18) Indirect a nchoring Indirect anchoring, also called supramolecular anchoring depends on binding the catalytic metal ion to an inhibitor of an enzyme, thereby directing it into the active site of the protein. This approach was applied to avidin by Whitesides ( F igure 1 19) 54 An achiral diphosphine rhodium catalyst conjugated to biotin was mixed with avidin, yielding a catalyst for hydrogenations of acrylates Any asymmetric preference observed for this system was due to the tertiary structure The same principle was then applied to streptavidin a prokaryotic protein that a lso binds biotin with high affinity. 55 Streptavidin has a larger binding pocket that affected stereoselectivity. 56 The initial study that compared a vidin and streptavidin directly concluded that some combinations of phosphine ligands streptavidin and av idin yield enan tiocomplementary products, although s treptavidin generally provided better ee values. This concept was later applied to ruthenium complexes combined with avidin or streptavidin for chiral ketone reductions. Using acetophenone as a test sub strate the best result was 92% conversion along with 94% ee R 57 Many of the results in these three examples depend ed on genetic optimization as well as chemical optimization e.g. chain length between metal and biotin, and the ligand sphere surrounding the coordinated metal. Alkene Reductases The Nobel Prize in chemistry was awarded to Ryoji Noyori in 2001 for his method of cis hydrogenation using r uthenium 2,2' bis(diphenylphosphino) 1,1' binaphthyl

PAGE 34

34 (BINAP) complexes Ru BINAP complexes allow for an in tuitive synthetic approach. O ne only has to choose either R or S BINAP to produce the enantiomer needed. S BINAP as the catalyst would give one product while R BINAP would give the enantiomer ( Figure 1 20 ). 58 This method has had great success in the presence of certain functional groups like alkenes functionalized with a carboxylic acid, but only moderate gains with other aprotic groups, such as alkenes functionalized with esters or ketones. 59 With all of this success the procedure for the Noyori asymmetric hydrogenation also highlights some of the general problems of catalysis for these reactions : high pressure, high heat, limited substrate range and a limited product range. Also the products result from a cis hydrogenation mechanism which leaves out a way to produce isomers that could result from a trans hydrogenation. Biotransformations using alkene reductases offer a n alternative to these metal catalyzed reactions Such conversions use r enewable resources, atmospheric pressure s at or near room temperature and often with such great enantioselectivity and regioselectivity that one can avoid unnecessary synthetic steps. 60 The Old Yellow Enzyme family has proven particularly useful in asymm etric alkene hydrogenation s Old yellow enzyme 1 (OYE1) was the first protein containing a non covalently associated cofactor to be isolated. It was purified from Saccharomyces pastorianus (originally named Saccharomyces carl sbergensis ) in 1933 by Warburg and Christian 61 and its FMN cofactor was discovered by Thorell in 1935. 62 This cofactor which is critical to reactivity, is held in the middle of the active site using steric and hydrogen bonding interactions. The enzyme has an 8 8 b arrel shape, also known as a TIM Barrel ( Fi gure 1 21 ) and the protein exists in the cell as a homodimer. 63 The history of

PAGE 35

35 the enzyme is rich with studies that provide an understanding of the chemistry that makes enzymatic catalysis possible, especially when it comes to t he use of cofactors, both in vivo and in synthetic applications Though there is a wealth of knowledge for this protein, one question that has remained unanswered is the physiological role 64 Mechanism OYE1 follows a ping pong bi bi mechanism. T he firs t stage is the reductive half reaction in which NADPH delivers a hydride to N5 of the active site FMN 65 The oxidative half reaction can be performed by several substrates including O 2 but for synthetic purposes involves reducing the alkene of an aldehyde. While t hese are unlikely to be the physiolo gical substrates for the enzyme, their reductions are efficient and stereoselective. The side chains of His 191 and Asn 194 donate hydrogen bonds to the enone or enal carbony l oxygen, which help s position the substrate and further polarize s the alkene bond to enhance the electrophilicity of its carbon. This facilitates a hydride transfer from the carbon along with protonation of the enone species in a step t hat involves Y196. The net result is a trans addition of H 2 across the alkene (Figure 1 22 ). 66 This mechanism nicely complements the Noyori type methods that provide net cis hydrogenation. Substrates for the Oxidative Half Reaction Many early OYE1 studies used oxygen as the electron acceptor for the oxidat ive half reaction. In 1995 it was discovered that 2 cyclohexenone could also a ct as an electron acceptor for the reduced FMN 67 This stimulated a larger search by Massey to identify additional O YE1 substrates He concluded OYE1 reduced the carbon to carbon bond of many substituted alpha beta un saturated ketones and aldehydes, although

PAGE 36

36 there were some steric limitations. 67 In general, most unsaturated acids and esters were not reduced by OYE 1 Subsequent work by other groups has better defined the substrate range of OYE1. For example, in the case of 2 substituted 2 cyclohexenones, OYE1 reduces the methyl derivative efficiently, but the ethyl analog is reduced much more slowly. 68 The conclus ion is that this site can accommodate small functional groups only. Consistent with this notion, halide or alcohol substitution at this position is well tolerated, and the added electron withdrawing effects further activate the double bond and thereby inc reases the reaction rates 69 The 3 position can vary greatly in size as long as large groups are not on both sides of the alkene bond 69 A few other examples of the enzyme s limitations were also revealed in the substrate library. OYE1 will not catalyze the reduction of an alkene that does not have sufficient electron withdrawing groups near it. For example an alkene activated by an alcohol will not react, nor will it reduce a carbonyl if there is no alkene to reduce. 68 OYE1 also catalyzes a dismutation reaction for some substrates in the absence of NADPH. For example, incubating 2 cyclohexenone with OYE1 yields 0.5 equivalents of cyclohexenone and an equal amount of phenol. In this case, the C5/C6 alkane moiety is the electron source for FMN reduction The resulting cyclohexadienone tautomerizes to yield phenol and the FMNH 2 reduces another substrate molecule to yield the saturated ketone. 64 Product S tereochemistry A limitation shared by all enzyme catalyzed synthetic methodologies concerns the abilit guided organometallic approaches such as the Noyori asymmetric hydrogenation either enantiomer can be produced

PAGE 37

37 simply by selecting the appropriate BINAP enantiomer. 59 Since all natural enzymes are assembl ed from only L amino acids, obtaining the enantiomer of a biocatalyst is not a viable strategy. Fortunately, some enantiocomplementary pairs of enzymes have been identified, 70 but this is not generally proven the case for alkene reductases from the OYE su perfamily (Figure 1 23). 13 Successful examples of using alkene reductases to produce both enantiomeric products in optically pure form are summarized below. The most straightforward approach to identifying an alkene reductase with complementary stereosele ctivity to that of OYE1 is to source naturally occurring enzymes. Alkene reductases outside the OYE1 superfamily are particularly attractive targets One example is rat leukotriene B4 12 hydroxydeh ydrogenase ( Ltb4DH ) Ltb4DH is a non flavoprotein whose only cofactor is NADPH 7 1 A uniq ue feature of this enzyme is its ability to catalyze either trans or cis addition of H 2 depending on whether R or S perillaldehyde is used as the substrate. 72 Ltb4DH reduces both R and S perillaldehyde to the same product cis isomer (Figure 1 24 ) ; however, the mechanism involves net trans addition of H 2 in one case and net cis addition is the other. Unfortunately, this mechanistic divergence was only observed for a single substrate. A rel ated strategy for finding reductases with different stereochemical preferences is to develop a library of OYE 1 homologs. This approach assumes that even homolog o us proteins have enough differences to alter the stereochemical outcome, possibly by changing the substrate binding orientation (Figure1 25). 73 ,7 4 While this approach was very successful for carbonyl reductases, it has shown only limited utility for alkene reductases One notable exception is the reduction of 1 nitro 2 phenylpropene, where the enzyme NCR reductase from Zymomonas mobilis can

PAGE 38

38 reduce this compound to the S configuration and 99% conversion while OYE1 reduces it to the R configuration and 90% conversion. 7 3 Substrate engineering is a nother strategy for changing the stereochemical outcome s of alkene reductases. Faber has reported one example. 7 5 This allows for an optimized pairing between substituents and alkene reductases In one example, 2 methoxy 2 cycl ohexenone is reduced by both OYE1 and NCR reductases with the same stereochemical preference, although the latter provides a lower % ee value. By contrast, the allyl analog is not a substrate for OYE1 ; however, NCR reduces it to the opposite stereoisomer with good % ee (Figure 1 26) Despite its success, this method is not ideal for synthesis since at least one enantiomer would require an additional two steps to account for the protection and deprotection. Faber also investigated a substrate engineering s trategy in the synthesis of Roche ester using alkene reductases (Figure 1 27). 76 Roche ester is an important chiral tocopherol, 7 7 the fragrance component muscone, 7 8 and many antibiotics. 79,80 ,8 1 e the stereochemical outcomes of alkene reductase catalyzed bioconversions was to adjust the steric bulk near the alkene by forming ether derivatives (allyl, benzyl, and tert butyldimethylsil yl (TBDMS)). Unfortunately, the attempt to use altered steric bu lk to flip substrate binding within the active sites failed. While different substrate/enzyme pairs gave a wide range of conversions, only R selectivity was observed, and no combination that yielded the S enantiomer was identified. The best result involved TBDMS ether and xenobiotic reductase A, from Pseudomonas putida which gave the reduced product with > 99% ee and > 99% conversion.

PAGE 39

39 The final strategy for altering the stereoselectivities of alkene redu ctases is the one adopted by this lab. This approach uses protein engineering to influence the orientations of substrate binding. Sites for mutagenesis are selected using information from X ray crystallography. In favorable cases, a single amino acid su bstitution can flip substrate binding (Figure 1 28) 8 2 A method pioneered by Reetz ( iterative saturation mutagenesis ) is more generally effective, however. 83 Instead of making mutations and screening those with many substrates, this method is based on ch oosing a substrate and making generations of mutations to fit that substrate. This coax es the enzyme into binding the substrate in a new conformation. Mutations to build from next are decided using conversion and ee calculations from the previous generat ion of enzymes Experimental results clearly show that these mutations do not change the mechanism of the enzyme, preserving the trans hydrogenation, and thus still act as a potential complement to the Noyori hydrogenation. In a previously published exa mple of this, the substrate 3 methylcyclohexenone was chosen as a substrate, and the OYE family member YqjM was the starting wild type enzyme. Using only two generations of mutagenesis, the researchers were able to reverse the binding of 3 methylcyclohexe none with YqjM (Figure 1 29). A similar approach was taken by Toogood with the enzyme Pentaerythritol tetranitrate (PETN) reductase. PETN reductase is an OYE family member from the organism Enterobacter cloacae 8 4 In a study to reveal the structural and binding impact of H181 and H184, Toogood et al. made site saturated mutants at both positions. 8 5 The researchers first made a library of single mutants at each position and then made double mutants. Paying attention to these positions is interesting because they both

PAGE 40

40 have been implicated as being important for substrate binding by donating hydrogen bonds and polarization of the double bond on the substrate. Changing them could be devastating to catalytic activity. W hen reducing the nitroalkene 2 phenyl 1 nitropropene however, the majority of H181X/H184X mutants showed a dramatic increase in the enantiopurity of the reduced product. The new ee values w ere in excess of 80% whereas the wild type enzyme ee was 54%. T his did come with some unfortunate compromise as many of the enzymes with good conversion also increased the yield of an undesired oxime. One of the mutations H184N was the most promising with conversion of 84% an ee of 87% and 67% yield for the alka ne. Of the product yield, 17% was the undesired oxime byproduct. Overview The following chapters of this dissertation follow an effort to improve catalysis by alkene reductase enzymes focusing on OYE1 The a ttempts to improve the enzyme include mutagenesis to allow reversed substrate binding, investigating the limitations of previously reported reversed substrate binding, and changing the reductive potential to aid in expanding substrate scope. Many of the examples are model compounds utilized as a tool to understand the substrate binding environment with some examples that are utilized for downstream chemistry.

PAGE 41

41 Table 1 1. Goals set by Codexis set for the production of ( R ) 2 methylpentanol Substrate (g/L) 220 Enzyme (g/L) 2 Reaction time ( h) 24 NADP cost contribution ($/kg) 10 Conversion (%) ee (%) 98 Table 1 2. Targets set by Codexis for a biocatalytic route to tetrahydrothiophene 3 ol Substrate loading 100 g / L ( M) KRED loading 1 g / L conversion 98% Reaction time 24 h Optical purity of prod. 99% ee

PAGE 42

42 Figure 1 1. Examples of alcohol dehydrogenase enzymes that are selective for NADP or NAD. a) Wild type Lactobacillus brevis alcohol dehydrogenase bound to NADP. b) Lactobacillus brevis alcohol dehydrogenase G37D bound to NAD. Reprinted with permission from J. Mol. Biol. 2005 349, 801 813 Copyright 2005 Elsivier Figure 1 2. Reaction stereochemistry of Sporobolomyces salmonicolor carbonyl reductase. Org. Lett 2008, Vol. 10, No. 4 525 528 Figure 1 3. Synthetically desired reaction performed by Thermoanaerobacter ethanolicus Prot. Eng. Des. & Selec 2007 vol. 20 no. 2 pp. 47 55

PAGE 43

43 Figure 1 4. Resolution strategy using the alcohol dehydrogenase from Thermoanaerobacter ethanolicus J. Org. Chem 2007 72 30 34 Figure 1 5. Reaction scheme of galactose oxidase. ChemBioChem 2002, 3, 781 783 Figure 1 6. Secondary aryl alcohol oxidations by engineered galactose oxidase mutants resolving secondary aryl alcohols. ChemBioChem 2008 9, 857 860 Figure 1 7. Reduction of 2,5 hexane dione by Pyrococcus furiosus alcohol dehydrogenase. Extremophiles 2008 12, 587 594

PAGE 44

44 Figure 1 8. Ketone reduction using Rhodococcus erythropolis alcohol dehydrogenase MAK154. Lett. Appl. Microbiol 2006 43, 430 435 Figure 1 9. NADPH regeneration glucose dehydrogenase oxidation followed by the spontaneous hydrolysis of the gluconolactone to gluconic acid. J. Biosci Bioeng 2011 11, 3, 266 271 Figure 1 10. pH indicator assay use d for high throughput screening from left to right 1. Positive control, 2. Negative control 3. W1. mutant S214R. 4. W2. Mutant G73S 5. Double mutant G72S, S214R. Reprinted with permission from. J. Biosci Bioeng 2011 11, 3, 266 271 Copyright 2011 Elsi vier.

PAGE 45

45 Figure 1 11. Reaction scheme of Penicillium citrinum Beta keto ester reductase with 4 bromo 3 oxobutyrate (BAM) to ( S ) 4 bromo 3 hydroxybutyrate Appl Environ Microbiol 2005 71, 1101 1104 Figure 1 12. Process patented by BASF for the production of ( R ) 2 methylpentanol. World Patent Application WO 2006/034812, 2006. Figure 1 13. Process developed by Codexis for the product ion of ( R ) 2 methylpentanol. Org. Proc. Res. Dev 2010, 14, 119 126

PAGE 46

46 Figure 1 14. Pfizer process for R tetrahydrothiop h ene 3 ol Tetrahedron Lett 1993 34 785 Figure 1 15. Codexis process for ( R ) tetrahydrothiophene using a ketone reductase. Org. Proc. Res. & Dev 2010 14, 188 192 Figure 1 16. Structure of Montelukast sodium

PAGE 47

47 Figure 1 17. Asymmetric carbonyl reduction as a method for indtoducing the chiral center of montelukast. Org. Proc. Res. Dev 2010, 14, 193 198 Figure 1 18. Hydrogenation/isomerization of cis stilbene. The complex of HCAII and Rhodium cannot catalyze the reduction of trans stilbene. Chem. Eur. J. 2009 15, 1370 1376 Figure 1 19. Biotin bound ligand for anchoring a Rhodium complex into avidin J. Am. Chem. Soc. 1978 100 306

PAGE 48

48 Figure 1 20. BINAP structures and a basic representation of a Noyori Hydrogenation Figure 1 2 1. PyMol image of OYE 1 co crystallized with p hydroxybenzaldehyde. PDB file downloaded from Swiss Protein Data Bank.

PAGE 49

49 Figure 1 22. Mechanism of OYE 1. J Biol. Chem 1998 273, 32763 32770 Figure 1 23. Typical s tereochemical o utcome of b oth c hiral c enters. Biochemistry 1995 34, 4246 4256 Figure 1 24. Enantiocomplementary reductions of perillaldehyde. J. Am. Chem. Soc 2008 130 7655 7658

PAGE 50

50 Figure 1 25. Nitroalkene reduction by alkene reductases. Eur. J. Org. Chem. 2008 1511 1516 F igure 1 26. Complementary stereochemical outcomes using substrate engineering. E ur. J. Org. Chem. 2010 6354 6358 Figure 1 27. Synthesis of Roche ester precursors using alkene reductases. Adv. Synth. Catal 2010 352, 2663 2666

PAGE 51

51 Figure 1 28. Sterecomplementary reductions of ( S ) carvone by wild type OYE1 and the W116I mutant. J. Am. Chem. Soc. 2009 131 3271 3280 Figure 1 29. Altered stereochemistry of alkene reductase Yqjm using iterative saturation mutagenesis. Adv. Synth. Catal 2009 351, 3287 3305

PAGE 52

52 CHAPTER 2 MEASURING REVERSED S UBSTRATE BINDING OF CARVONE MIMIC COMPOUNDS W116 Mutants Tryptophan 116 in OYE1 was originally chosen for replacement as a way to relieve steric hindrance within the active site since this side chain is very bulky (Figure 2 1) 8 2 It was expected that amino acids with smaller substituents would allow OYE1 to accept 2 cyclohexenones with larger substituen ts. In this regard, the mutagenesis study was a failure: none of the mutants had greater activity against sterically encumbered substrates. There was however an unintended but favorable consequence F or some substrates a subset of the mutants yielded the opposite stereochemical outcome compound compared to the wild type. For example, with ( S ) carvone, two changes occurred. First, the poor conversion seen for wild type OYE1 was eliminated; and second, the stereochemical outcome of this reduction was opposite that of wild type ( trans or cis ) While W116 had not been implicated in the catalytic mechanism of OYE1, sequence alignment shows that it is highly conserved among OYE homologs. NMR studies revealed that the W116I mutant followed the same mechanism as the wild type OYE1 (net trans additio n of H 2 ) Thus the reversed stereochemical outcome orientation of the substr ate in the active site. Analogous behavior was also observed for ( R ) peril l aldehyde and to a lesser extent neral as well as the W116F mutant of OY E1 One intriguing result from this study was that no change in substrate binding orientation occurred when OYE1W116I was used to reduce 2 methylcyclohexenone. This led to the conclusion that the methyl group at the 2 position (which is also present in ca rvone) does not dictate binding orientation. The only d ifference between R and S

PAGE 53

53 carvo ne is the configuration of the substituent at the 5 positio n (Figure 2 2). Given the 5 position substituent, we carried out a more systematic set of carvone analogs that differed in this position We also decided to create all possible substitution mutants for W116 since only a subset had been examined previously. Results and Discussion QuikChange site directed mutagenesis We used t he QuikChange method for site directed mutagenesis with optimization for the OYE1 gene The classical QuikChange method uses overlapping primers with T M >78C to allow annealing despite imperfect complementarity. Once thermal cycling is complete, the new ly synthesized DNA lack s methylations on the a denines of GATC regions while the template strands contain these modifications. This allows selective destruction of the template by Dpn1 leaving only amplified DNA in the solution (see scheme in appendix A) This mixture c an be used to transform E. coli cells to recover the desired mutants after in vivo nick repair We originally employed the recommended E. coli strain (XL1 Blue); however, primer concatemers were found in the final plasmids (Figure 2 3). Substituting E. coli JM109 overcame this problem, although poor transformation efficiency was sometimes observed. We also altered the primer design to incorpor encoding all amino acids at position 116 were generated. The proline, leucine, and alanine mutants were prepared by Bradford Sullivan. Reduction of 2,5 disubstituted cyclohexenones The goal of this project was to uncover the structural basis for substrate flipping in 5 substituted 2 cylclohexenones. Substituted cyclohexenone molecules were chosen

PAGE 54

54 (Table 2 1) to cover the structural space between the two extremes examined earlier. The final preparative step was to develop an analytical method that allows the enantio and diastereomeric preferences of the OYE1 mutants to be established. A problem with the synthesis of the 2,5 disubstituted cyclohexenones is that all were obtained as racemic mixture s (Figure 2 4) It was not possible to separate this mixture into pure enantiomers on a preparative scale. All enzyme catalyzed reductions were therefore kinetic resolutions, so we attempted to identify the substrate enantiomers unambiguou sly. Since we were unable to resolve the substrate enantiomers directly by chiral phase GC a derivatization pathway was designed. An initial Luche reduction allow ed the carbonyl to be selectively reduced to an alcohol while preserving the alkene. Becau se of the conformational preference of the starting material and the high axial selectivity of the Luche conditions, attack by NaBH 4 yields only an enantiomeric pair of cis alcohols. 86 This was verified by 1 H NMR and confirmed the previous report by Trach etenberg 87 The products of this d erivatization could be partially separated by chiral phase GC, but baseline resolution was not achieved (Figure 2 5). The enantiomeric alcohols were therefore acetylated, and baseline resolution on chiral phase GC was ac hieved successfully. Unfortunately, while this reduction/acetylation allowed enantiomeric separation of the starting material, it proved unsuitable when applied to time point samples from bioconversions. After enzymatic screening, Luche reduction, and ac etylation, not enough material was left to fully characterize an accurate conversion and ee/de values. Even after reactions had been scaled up by tenfold, the very small quantities made it impossible to measure peak areas reliably.

PAGE 55

5 5 We therefore turned to an alternative strategy in which the starting alkenes were reduced with H 2 /Pd on carbon, followed by chiral phase GC analysis. From a racemic starting material, four possible diastereomers would be expected. Unfortunately, of all separations attempted o n three different chiral phase GC stationary phases (Chirasil Dex CB, Dex 225 and Chirasil Val), only three peaks instead of four could be abilities to separate the en antiomers of one diastereomer but not the other (Figure 2 6). It is still possible to get some information from these reactions. If it is assumed that OYE1 only binds this mixture in the expected manner, then only two diastereomers could be formed (Figur e 2 7). These two diastereomers can be analyzed by a nonchiral GC (Figure 2 8). If any of the starting materials is reduced in the reversed binding orientation, then one of the product peaks will disappear and the other will grow larger. Neither of thes e analyses is ideal or completely conclusive. All one can say is one of the two starting material enantiomers has bound in the opposite orientation compared to the other. One cannot determine which enantiomer did so. Another problem for this analysis is that substrate conversion values must be near 100% in order to be able to speculate whether an enzyme mutant is able to flip a substrate. Lesser extents of conversion skew the de values. This effect is clearly seen in figures 2 9, 2 10, and 2 11 where m any of the enzymes that have poor conversion also have high % de. This is presumed to occur because the mutant is selective for only one of the isomers. All of the enzymes were screened with the substrate 2 methylcyclohexenone, and of the enzymes that gav e significant conversion, all produced the same enantiomer (Figure 2 12). This does not seem intuitive since the carbon that would normally

PAGE 56

56 interact with the 116 position is the methyl at the 2 position of the cyclohexene ring. This illustrates the same point as before. Mutating the W116 amino acid site affects the orientation of other amino acids in the active site, which allows for the 5 position of the cyclic substrate to dictate which direction the molecule binds. When reducing 2,5 dimethylcyclohex enone, wild type OYE1 performed exactly as expected, yielding two diastereomers with equal facility (Figure 2 9). Based on the GC analysis, an insignificant amount of substrate flipping occurred. However, with this substrate, W116I already showed some te ndency for reversed substrate binding. The W116I mutant had the best evidence for substrate flipping when reducing 2 methyl 5 ethylcyclohexenone (Figure 2 10). Both the selectivity and reactivity drop for the substrate with the propyl substituent. This is also true for W116F, W116L, and W116N (Table 2 2). For the W116I mutant, this was expected from the previously reported data because when bond rotation is considered, the ethyl group takes up a similar amount of space as the isopropenyl group from the carvones. The conversion percentages of the enzymes with the propyl substituent tended to be poor. In fact, only two enzymes actually reduced the substrate with the propyl substituent to 100% conversion (W116A and W116L Figure 2 11). It is also intrigui ng that the W116L mutant actually reduced substrates with larger substituents more efficiently than those with smaller substituents. A similar trend can also be seen for the W116S and W116T mutants, which had little to no conversion with 2 methylcyclohexe none and better conversions for the ethyl and propyl analogs. For the W116L mutant, this is due to a larger hydrophobic binding pocket. Entropic effects help the substrate into the active site and away from water.

PAGE 57

57 It was also found throughout this experi ment, when a particular set of diastereomers is favored, they all favor the same set. It is as if the enzymes can only change the binding of one of the starting material enantiomers. This, however, cannot be guaranteed. For example, OYE1W116A has conver sion values ranging from 95% to 100% and for all of those substrates de is low. It is possible this enzyme is changing the binding of both substrates to an equivalent degree. Doing so would cause the peaks corresponding to each set of enantiomers to chan ge very little. This would lead to the appearance of no flipping even if reversed substrate binding were occurring. To move further, this experiment must be done with optically pure starting materials. Materials and Methods Restriction endonucleases and T4 DNA ligase were purchased from New England Biolabs. Primers were ordered from Integrated DNA Technologies. Glucose dehydrogenase was graciously provided by Codexis. Plasmids were purified on large scales using CsCl density gradient ultracentrifugatio n. DNA sequencing was performed by the University of Florida ICBR using the Sanger capillary fluorescence methods. LB medium contained 5 g / L Bacto Yeast extract, 10 g / L Bacto Tryptone and 10 g / L NaCl. Kanamycin and ampicillin were used at concent rations described by Maniatis 40 4C. Crotonaldehyde, 2 hexene al and 2 methylcyclohexanone were purchased from Sigma Aldrich. E thyl propiony l acetate was purchased from Alfa Aesar. G lutathione resin was purchased from Clontech. Subcloning From a Saccharomyces cerevisiae Expression Plasmid I generated the mutants W116M, W116Y, and W116S by digesting separately with restriction endonuclease enzymes Bsa I and Kpn I from the yeast shuttle vector

PAGE 58

58 PSKP3W116X. 82 The resulting 577 bp fragment containing position 116 was isolated and purified by low melting agarose gel electrophoresis. Simultaneously the E coli vector pDJB5, which housed the wild type OYE1 ge ne, was digested with Bsa I and Kpn I and the 7,150 bp fragment was purified using low melting agarose gel electrophoresis. Fragments were joined together using T4DNA ligase and transformed into E coli JM109. Colonies were analyzed for the presence of the insert, first by PCR and then by Sanger DNA sequencing (Appendix B). QuikChange Site Directed Mutagenesis I produced the W116 mutants W116G, W116H, W116Q, W116V, W116T, W116N, W116D, W116E, W116R, W116C, and W116K using an adaptation of the QuikChange met hod. The procedure was 18 cycles of amplification using 95 C (30 s), 55 C (1 min) and 68C (9 min) followed by final extension time of 9 min at 68C. Each reaction had a Pfu turbo buffer, 50 ng pDJB5, 125 ng Pfu turbo DNA polymerase. The resulting DNA mixture was digested with DpnI overnight at 37C to fragment the template DNA (Appendix A). Protein Purification P rocedure I purified the mutated proteins using the overexpression plasmids transformed into E. c oli BL 21 (DE3) using electroporation. Proteins tagged to GST were purified as described previously. 8 2 A 5 mL overnight culture of the appropriate strain was grown using LB growth me d ium and 40 / mL kana mycin, with shaking at 37 C. The overnight culture was diluted 1 : 100 into a 500 mL culture of LB/kanamycin and grown at 37C with shaking. When the optical density O.D. 600 nm of the culture was between 0.5 and 1. 0, protein expre ssion was induced by adding isopropyl thio D g alactoside

PAGE 59

59 (IPTG) to a final concentration of 100 At this stage, the culture was shaken at room tempera ture for 6 hours. The cells were then pelleted by centrifugation, washed twice with cold sterile wa ter and pelleted by centrifugation again. They were resuspended in 30 mL of cold loading buffer ( 50 mM Tris Cl, 4 mM MgCl 2 1 mM DTT ( add ed immediately before use), 1mM PMSF ( also added immediately before use), and 10% glycerol, pH 7.5 ). C ells were then lysed using a French press ure cell (10,000 20,000 psi), and insoluble debris was removed by centrifugation at 15,000 x g for 20 min at 4 C. The supernatant was then passed through a 0 onto a column containing 10 mL of glutathione resin It was then recirculated using a peristaltic pump for 3 5 hrs using a flow rate of ~0.5 mL / min. Unbound proteins were then allowed t o flow though to waste and the column was washed twice with 20 mL portions of cold loading buffer (sam e as ab ove). The bound flavoprotein was then eluted using 40 mL elution buffer (39.6 mL loading buffer, 0 .40 mL 2M NaOH, and 0 .31 g reduced glutathione). The eluant was concentrated by ultrafiltration (Amicon YM 30 membrane) and then dialyzed overnight i n buffer ( 20 mM Tris Cl, 4 mM MgCl 2 55 mM NaCl, 2 mM EDTA, 1 mM DTT, 50% glycerol, pH 7.5) prior to storage at 20 C Protein concentrations were determined by Bradford analysis. 88 The glutathione resin was regenerated using high salt buffers to remove any remaining proteins. Buffer A (0.1M Tris Cl, 0.5M NaCl, pH 8.5) and buffer B (0.1 M Sodium acetate, 0.5 M NaCl, pH 4.5) were each passed through the column in alternating 200 mL amounts until each had been used 3 times. Experimental P rocedure for S cree ning I screened the substrates in 1 mL 0.1 M potassium phosphate buffer pH 7.0 for 24 hours at room temperature. Reactions contained 14 mol glucose, 5 g glucose

PAGE 60

60 dehydrogenase, 0.2 + GC sample 13). Synthesis and Characterization Synthesis of 2,5 di methyl cyclohexenone and 2 methyl 5 propylcyclohexen one. I synthesized two substrates, each as a racemic mixture, starting with 2,5 dimethyl cyclohexenone (alkene 70 )and 2 methyl 5 propylcyclohexenone. The plan was to increase the steric hindrance through ethyl and propyl at the 5 position (Table 2 1). The title compounds were synthesized by a modification of literature methods. 89 Anhydrous m ethanol 25 mL was chilled on an ice bath the n a catalytic amount of sodium metal was added to produce NaOCH 3 as a catalytic nucleophile. Once the sodium had dissolved ethyl propiony l acetate and crotonaldehyde were added sequentially in a one to one molar ratio. The reaction mixture was left on ice for 30 min and was then heate d to reflux over night. The reaction was quenched by slowly adding 1 M HCl. The products were extracted using diethyl ether added in portions. The combined organics were washed with 1 M NaOH followed by brine, then dried with MgSO 4 concentrated on a rotary evaporator. The crude products were purified by flash chromatography on silica gel with 1 : 10 ethyl acetate : hexane. To check compound purity, conversion analys is was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R 9.11 min. 1 H NMR 1.05 (3H d), 1.77 (3H m), 2.12 (3H m), 2.4 (1H dm), 2.50 (1H dt), 6.71 (1H m); 13 C NMR 200.3, 145.0, 135.5, 46.6, 34.5, 31.0, 21.4, 15.9; (EI) 124(M), 109(16), 82(100), 69(22), 54(42); 2 methyl 5 propylcyclohexenone (alkene 74 ) was synthesized using this procedure replaci ng crotonaldehyde with 2 hexen al (Figure 2 11) GC MS 60 C 2 min, 10 C / min

PAGE 61

61 increase, 200 C 5 min, T R 12.35 min. 1 H NMR .903 (3H m) 1.329 (4H m), 1.767(1H m), 1.79(1H m), 2.04(1H m), 2.4 (3H m), 2.53 (2H dd), 6.72(1H m). 13 C NMR 200.5, 145.1, 135.7, 44.8, 38.2, 35.6, 32.8, 19.8, 15.9, 14.3. (EI) 152(M), 109(17), 82(100), 54(20). Luche reduction of 2,5 dimethylcyclohexenone 86 I derivatized 2,5 dimethylcyclohexenone to compound 83 using the Luche reduction. To 10 mL of cold ethanol, 0.41mmol (100 mg) of cerium chloride was added followed by 0.41 mmol (15.5 mg) Na BH 4 After stirring for 30 min, 0.4 mmol (49 mg) 2,5 dimethylcyclohexenone was slowly added to the mixture. The progress of the reaction was monitored using GC (Figure 2 14). NMR verified predominantly cis geometry as previously reported. 87 To check co mpound purity, conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R 8.15 min. Chiral separation was attempted using a GC MS (Hewlett Packard) with a chirasil dex column 60 C 2 min, 10 C / min increase, 180 C 5 min, R T 8.17 min enantiomer 1, 8.22 min enantiomer 2; 1 H NMR 5.50 (1H s), 4.20 (1H m), 2.05 (3H m),1.75 (3H m), 1.67 (2H s), 1.21 (1H m), 0.93 (3H d); 13 C NMR 136.3, 124.5, 71.0, 42.2 34.5, 28.8, 22.1, 19.2; (EI) 126 (M), 111 (100), 97 (41), 84(77), 69(63), 55(43). Acetylation of 2,5 dimethylcyclohex 2 en 1 ol I then derivatized 2,5 dimethylcyclohex 2 en 1 ol using an acetylation. To a flask containing 50 mg of 2,5 dimethylcyclohex 2 en 1 ol, 5 mL ethyl acetate was added, stirred, and 1 mL was removed. To this aliquot a twofold molar excess of acetic anhydride was added along with one single crystal of 4 dimethylaminopyridine and was left to stir over night (Figure

PAGE 62

62 2 14). The reac tion was quenched by washing with water and compound 85 was analyzed by GC MS. Conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R 10.28 min, (EI) 168(M), 108(58.7), 93(100), 91(22.5), 55(25.1). Chiral separation to measure enantiomeric excess was achieved using a GC MS (Hewlett Packard) with a Chirasil Dex column with helium as the carrier gas, 60 C 2 min, 1 C / min, 120 C 1 min, 1 0 C / min, 180 C 5 min. R T 37.1 min enantiomer 1, 41.2 min enantiomer 2. Reduced standards I made three racemic standards for this project. One for each 2,5 dimethylcyclohexenone, 2 methyl 5 ethylcyclohexenone, and 2 methyl 5 propylcyclohexenone. For e ach of these starting materials 20 mg was dissolved in 5 mL ethanol, then 5 mg Pd on carbon was added and the flask was flushed with H 2 and then left over night under H 2 balloon. Reaction conversion was monitored by GC MS and then compared with the result s from enzyme screening, 60 C 2 min, 10 C / min increase, 200 C 5 min, 2,5 dimethylcyclohexanone R T 7.86 min diastereomer 1, 8.24 min diastereomer 2. EI 126(M), 82(84), 69(100), 56(45). 2 methyl 5 ethylcyclohexanone R T 9.8 min diastereomer 1, 9.97 min dia stereomer 2. (EI) 140(M), 111(59), 96(49), 83(100), 55(70). 2 methyl 5 propylcyclohexanone R T 11.21 min diastereomer 1, 11.36 min diastereomer 2, (EI) 154(M), 111(100), 97(87), 55(89)

PAGE 63

63 Table 2 1. Starting materials and products for the synthesis of the compounds chosen to mimic carvone. keto ester Enals Product Name 2,5 dimethyl 2 cyclohexenone 2 methyl 5 ethyl 2 cyclohexenone 2 methyl 5 propyl 2 cyclohexenone Table 2 2 Enzymes that gave 95% conversion or better at least once during the screening % conv % de % conv % de % conv % de OYE 1 Wild Type 100 4.5 82 24.4 89.5 14.3 W116A 95 22.8 100 20 100 13.4 W116F 100 5.9 100 21.3 86.6 25 W116I 100 35 100 73.5 90.8 39.3 W116L 50.6 22.8 100 61.9 100 11.1 W116N 96.3 29 97.8 81.7 78.4 62.6 W116Q 100 14.7 86.9 58 61.6 100 W116Y 100 4.4 77.6 30.2 83.3 24.8

PAGE 64

64 Figure 2 1. Pymol stereoimag e of the OYE1 active site with p hydroxybenzaldehyde as a bound inhibitor. Fi gure 2 2 Structures of R and S carvone

PAGE 65

65 OYE1 CGTTCGTTTGGGTTCAGTTA TGG GTTTTGGGTTGGG............ Seq. CGTTCGTTTGGGTTCAG TTA GGC GTTTT GGGTTGGGTT C AGTTA GGC G Primer CGTTCGTTTGGGTTCAGTTT GGC GTTTTGGGTTGGG Primer CGTTCGTTTGGGTTCAG TTT GGC G ............CTGCTTTCCCAGACAATC TTTTGGGTTGGGCTGCTTTCCCAGACAATC TTTTGGGTTGG G Figure 2 3. Error of Q ui kChange method when following S manual. First line : native OYE1 sequen ce dots represent where the mutated sequence has add itional bases. Second line: Resulting sequence of the attempted W116G mutation. Third line: Forward W116G primer to illustrate where it first occurs. Fourth line: Forward W116G primer to illustrate that it repeats in the resulting DNA sequence causing an undesired insertion. Figure 2 4. Synthesis of the Carvone mimic compounds. Compound 70 is the desired product that was obtained with about 30% yield and compound 75 is the side product obtained with about 20% yield. J. Org. Chem. 1997 62 9323 9325

PAGE 66

66 Figure 2 5. Chiral GC MS result for the Luche reduction. Figure 2 6. Chiral GC result with 2,5 dimethylcyclhexanone. When using a 1 degree per minute method on a chirasil dex column only 3 peaks can be resolved. Integration values indicate that the first peak approximately equals the sum of the following two peaks, indicating it contains two of the four isomers.

PAGE 67

67 Figure 2 7. All possible outcomes when reducing 2,5 dimethylcyclohexenone. When reducing 2,5 dimethylcyclohexenone, OYE1 binds each enantiomer in the expected manner, which places the methyl near the W116 and gives two diastereomers. The two normal product diastereomers are structures 77 and 78. The two that could possibly occur from reversed binding are structures 79 and 80. A. B C. Figure 2 8. Palladium reduced standards of substituted cyclohexenone compounds on nonchiral GC MS. A. 2,5 dimethylcyclohexenone. B. 2 methyl 5 ethylcyclohexenone C. 2 methyl 5 propylcyclohexenone.

PAGE 68

68 Figure 2 9. Data for the reduction of 2,5 dimethylcyclohexenone. Figure 2 10. Data for the reduction of 2 methyl 5 ethylcyclohexenone

PAGE 69

69 Figure 2 11. Data for the reduction of 2 metyl 5 propylcyclohexenone Figure 2 12. Data for the reduction of 2 methylcyclohexenone

PAGE 70

70 Figure 2 13. General reaction and NADPH regeneration pathway Figure 2 14. Synthesis pathway and GC results. GC results are of the acetylated product f rom the development of the strategy to analyze consumption of starting materials 2,5 dimethylcyclohexenone

PAGE 71

71 CHAPTER 3 REDUCTIONS OF ACIDS AND ESTERS Reduction of Halogenated Acids Previously published reports have established that OYE1 can reduce an alkene bond that is conjugated to an aldehyde or a ketone. Subsequent reports later proved that the enzyme can also reduce alkenes with nitro substituents. By contrast, OYE1 cannot reduce alkene bonds of conjugated acids or esters. Some exceptions do exist; h owever, many of these exceptions are diactivated acids or esters. 90 The need for an enzyme to reduce these substrates has led to the class of enzymes being studied called enoate reductases. 69 These enzymes can reduce the alkene of conjugated acids or est ers but must do so under anaerobic conditions due to oxygen sensitivity. It would be better if OYE1 could perform these reactions because it is not sensitive to oxygen. Early kinetic experiments showed that adding additional electron withdrawing groups to the alkene bond increased reaction rates. 66 Increased polarity increases the partial positive charge on the beta carbon making hydride transfer easier from the flavin. In the following experiments, electron withdrawing groups were considered to aid OY E1 in reducing compounds it normally cannot. As an electron withdrawing group, halogens have not been well studied with OYE1. Reduced haloacrylate compounds are interesting because some of them have st suggested that it could not reduce acids unless there is a chlorine atom adjacent to the double bond. 91 Many years later an enzyme called haloacrylate reductase was shown to reduce chloroacry lic acid to ( S ) 2 chloropropionic acid with 99% ee. Even tho ugh the enzyme is called haloacrylate reductase, only results pertaining to the chloro derivative have been reported 92,93 In an

PAGE 72

72 effort to expand the scope of OYE1 catalyzed reactions, haloacrylate derivatives w ere scr eened with OYE1. Another approach tak en to find ways to get OYE1 to reduce an organic acid and esters was modifying the reduction potential of the FMN cofactor. The mutant T37A was made for this project because it had been previously reported that this mutant changed the reduction potential of the flavin. Threonine 37 forms hydrogen bonds with both the flavin N5, which is the hydride delivering nitrogen, and the C4 carbonyl (Figure 3 1). Loss of these interactions slows the reductive half reaction but increases the oxidative half reactions (the portion of the mechanism that involves alkene substrates). 94 It was hypothesized that an altered reduction potential may help the enzyme reduce compounds that it usually cannot. Results and Di scussion Esters with OYE1T37A The first substrates that were tested for this project were acrylate esters. Ethyl 3,3 dimethylacrylate, and the trifluoro ester were screened. Neither esters gave any conversion with either the wild type OYE1 or the T37A mutant and only starting material was observed by GC MS. However, to get a better picture of the catalytic activity of the OYE1T37A mutant, it was also used to reduce known OYE1 substrates 2 methylcyclohexenone and methyl trans cinnamaldehyde, giving poor conversion for both. This suggests that even though th is variant has a greater reduction potential, the net effect of the slower reductive half reaction is an overall poorer enzyme (Table 3 1). This may be due to structural alterations caused by the T37A mutation. FMN is held in the active site using hydrog en bonding and steric interactions alone. T37 not only has the role of regulating reduction potential, but is also one of the hydrogen bonding

PAGE 73

73 partners responsible for anchoring the flavin in the active site. Given these disappointing results, we did not pursue the T37A mutant further. Haloacrylic Acids Because both the starting materials and products are highly water soluble, significant efforts were expended on developing analytical methods to monitor these reductions. Each substrate presented its ow n challenges. Bromoacrylic acid could be converted to its methyl ester using boron trifluoride in methanol and this derivative could be analyzed by GC MS. OYE1 reduced this substrate, although with only 70% conversion. Unfortunately, this strategy was n ot applicable to methacrylic or 2 chloroacrylic acids. These were analyzed using FID detection GC due to poor ionization when using GC MS. At the outset, we predicted that OYE1 catalyzed reduction would be increased by increased electron withdrawing abili ty of the halides (F > Cl > Br) (Figure 3 2). Methacrylic acid was not expected to be a substrate for OYE1. It was included as a roughly isosteric analog for the haloacrylates. To our surprise, methacrylic acid was a substrate for OYE1, although the e xtent of conversion was less than 25% (Figure 3 3). Chloroacrylic acid also appeared to be a relatively poor substrate (Figure 3 4), although accurate conversion percentages were difficult to reproduce due to sample loss from some samples. This may be du e to loss of the volatile saturated acid during the lyophilization step that preceded GC analysis. This was also a problem for 2 fluoroacrylic acid as well. We were unable to monitor reductions of 2 fluoroacrylic acid. Materials and Methods ( S ) 2 bromopropionic acid, bromoacrylic acid, trifluoroethanol, ethyl 3,3 dimethylacrylate, and boron trifluoride in methanol 14% were purchased from Sigma

PAGE 74

74 Aldrich. Chloroacrylic acid was purchased from Alfa Aesar. 2 Fluoropropenoic acid, and 2 fluoropropi onic acid were purchased from Matrix Scientific. Mutagenesis I mutated the OYE1 sequence to include the amino acid replacement T37A using the QuikChange method The procedure was 18 cycles of amplification using 95 C (30 s), 55 C (1 min) and 68C (9 min) followed by final extension time of 9 min at 68C. Pfu turbo buffer, 50 2.5 U Pfu turbo DNA polymerase. Th e resulting DNA mixture was digested with DpnI overnight at 37C to fragment the template DNA (Appendix A). Protein Purification P rocedure I purified the mutated proteins using the overexpression plasmids transformed into E. c oli BL 21 (DE3) using electropo ration. Proteins tagged to GST were purified as described previously. 8 2 A 5 mL overnight culture of the appropriate strain was grown using LB growth me dium and 40 / mL kana mycin, with shaking at 37 C. The overnight culture was diluted 1 : 100 into a 500 mL culture of LB/kanamycin and grown at 37C with shaking. When the optical density O.D. 600 nm of the culture was between 0.5 and 1. 0, protein expression was induced by adding isopropyl thio D g alactoside (IPTG) to a final concentration of 100 At this stage, the culture was shaken at room tempera ture for 6 hours. The cells were then pelleted by centrifugation, washed twice with cold sterile water and pelleted by centrifugation again. They were resuspended in 30 mL of cold loading buffer ( 50 mM Tris Cl, 4 mM MgCl 2 1 mM DTT ( add ed immediately before use), 1 mM PMSF (also added immediately before use), and 10% glycerol, pH 7.5 ). C ells were then lysed using a French press ure cell (10,000 20,000

PAGE 75

75 psi), and insoluble debris was removed by centrifu gation at 15,000 x g for 20 min at 4 C. The supernatant was then passed through a 0 a column containing 10 mL of glutathione resin It was then re circulated using a peristaltic pump for 3 5 hrs using a flow rate of ~0.5 mL / min. Unbound proteins were then allowed t o flow though to waste and the column was washed twice with 20 mL portions of cold loading buffer (same as ab ove). The bound flavoprotein was then eluted using 40 mL elution buffer (39.6 mL loading buffer 0 .40 mL 2M NaOH, and 0 .31 g reduced glutathione). The eluant was concentrated by ultrafiltration (Amicon YM 30 membrane) and then dialyzed overnight in buffer ( 20 mM Tris Cl, 4 mM MgCl 2 55 mM NaCl, 2 mM EDTA, 1 mM DTT, 50% glycerol, pH 7.5) prior to st orage at 20 C Protein concentrations were determined by Bradford analysis. 88 The glutathione resin was regenerated using high salt buffers to remove any remaining proteins. Buffer A (0.1M Tris Cl, 0.5M NaCl, pH 8.5) and buffer B (0.1 M Sodium acetate, 0.5 M NaCl, pH 4.5) were each passed through the column in alternating 200 mL amounts until each had been used 3 times. Procedure for Haloacrylate R eductions I screened the substrates in 1 mL 0.1 M potassium phosphate buffer pH 7.0 for 24 hours at room temperature. Reactions contained 14 mol glucose, 5 g glucose dehydrogenase, 0.2 + Methacrylic aci d, chloroacrylic acid, and fluoroacrylic acid A screening reaction mixture as described previously was lyophilized and carefully monitored (prolonged freeze drying affected yield of reduced acid). To the dried sample, 1 drop of concentrated HCl was adde d and then extracted using 1 mL dichloromethane. Analysis

PAGE 76

76 of these compounds was c arried out on a GC FID detector 60 C 2 min, 10 C / min increase, 200 C 5 min, R T methacrylic acid 17.09 min, chloroacrylic acid 21.17 min. Esterified bromoacrylic acid I u sed a derivatization procedure to analyze the amount of starting material and product left after the screening. The reduction product was first derivatized to the methyl ester before analysis by GC MS. To the lyophilized product 0.5 mL of 14% boron trifl uoride in methanol solution was added then heated to 100 C for 1 hour in a pressure vessel. The reaction was then quenched by adding, 0.5 mL H 2 O and extracted using methylene chloride (Figure 3 5). This procedure was adapted from the supporting informati on from Faber. 90 Conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R 6.02 min (EI) 166(M), 164(M), 135(100), 133(95), 107(68), 105(70), 85(4 2), 59(35). This very same procedure was used with the purchased reduced racemic standard as well. GC MS T R 5.69 min (EI) 168(M), 166(M), 109(78), 107(80), 87(100), 59(77).. Reduced standard of methacrylic acid : For GC FID analysis I made a reduced stan dard of methacrylic acid by putting 0.5 mL of methacrylic acid in 5 mL of methanol. Palladium activated carbon 5% 25 mg was added to the solution, and the flask was purged using H 2 gas from a balloon. This solution was left stirring over night and analyzed by GC the next day. Any activated carbon in the analysis sample was removed by filtering over Celite. Conversion analysis was performed using a GC FID (Hewlett Packard) with a D B 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 180 C 5 min, T R 17.06 min,

PAGE 77

77 Trans 2,2,2 trifluoroethyl but 2 enoate I synthesized the ester trans 2,2,2 trifluoroethyl but 2 enoate (alkene 87 ) using the following procedure. Dr y tetrahydro furan 20 mL was placed in a round bottom flask and chilled to 0C. T riflu oroethanol 1.9 mL and dry triethylamine 4 mL were added. C rotonyl anhydride 6 mL was slowly added last (Figure 3 6) This r eaction was monitored by GC MS. When the rea ction reached c ompletion, it was quen ched by washing with water and saturated NaHCO 3 The solution was then concentrated on a rotary ev aporator and distilled using a K ugelrohr apparatus. NMR data was compared to previously published results. 95 Conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 40 C 5 min, 10 C/min increase, 200 C 5 min, T R 5.49 min 1 H NMR 7.10 (1H m), 5.92 (1H d), 4.53 (2H q), 1.92 (3H d). 13 C NMR 18.3, 77.4, 121, 125 148, 165. (EI) 168(M), 153(23), 83(30), 69(100). 2,2,2 trifluoroethyl buta noate I synthesized a reduced standard of trans 2,2,2 trifluoroethyl but 2 enoate by placing 20 mg into ethyl acetate 2 mL with Pd on carbon 5 mg. This reaction mixture was stirred under an H 2 atmosphere using a balloon. Conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 40 C 5 min, 10 C / min, 200 C 5 min, T R 4.32 min. (EI) 170(M), 155(25), 142(100), 83( 46), 71(80)

PAGE 78

78 Table 3 1. Data corresponding to the comparison of OYE1 and OYE1T37A. Substrate OYE1 OYE1T37A 0% c 0% c 0% c 0% c 99% c 99% ee 10% c 99% ee 93% c 61% ee 34% c 66% ee

PAGE 79

79 Figure 3 1. Amino acids that hydrogen bond to the isoalloxazine ring of flavin. Changing T37 to an alanine has been previously published to change the flavin reduction potential from 230 mV to 263 mV. Proc. Natl. Acad. Sci. 1999 Vol. 96, 3556 3561 Figure 3 neighboring electron withdrawing halides

PAGE 80

80 A. B. C. Figure 3 3. Methacrylic acid chromatograms. A. negative control B. Methacrylic acid with OYE. C. Coinjection of sample B along with the Palladium on carbon reduced acid. 17.0 is the reduced product 17.9 is the methacrylic acid starting material and 19.6 is the methyl benzoate internal standard.

PAGE 81

81 Figure 3 4 Chloroacrylic acid chr omatogram. The peak at 19.985 is the OYE1 reduced product and the peak at 21.175 is the starting material Figure 3 5. Derivatized product of the OYE1 Bromoacrylic acid reduction on nonchiral GC MS. Peak at 5.69 is the OYE1 reduced product and 6.02 is the starting material.

PAGE 82

82 Figure 3 6. Synthesis of trans 2,2,2 trifluoroethyl but 2 enoate

PAGE 83

83 CHAPTER 4 ALCOHOL FUNCTIONALIZ ED ALKENES Baylis Hillman Adducts Previous efforts with OYE1 have focused on simple model compounds that have little downstream use. Chiral centers that are functionalized with an alcohol can be useful as building blocks for downstream chemistry. It was for this reason that three Baylis Hillman adducts whose reduction products are chiral synthons were chosen for study with OYE1. 96 Reduction of methyl 2 (hydroxymethyl)acrylate yields Roche ester, down stream products of which were discussed in Chapter 1. 2 (hydroxymethyl) 2 cyclohexenone when reduced to the S configuration can be later used to make th e spasmolytic drug R,R rociverine. 97 The final compound chosen was 2 (hydroxymethyl) 2 cyclopentenone, which after reduction can be oxidized using a Baeyer Villiger monooxygenase into a compound that is eventually converted into the jasmine lactone. 98 Res ults and Discussion Screening of Enzymes Wild type OYE1, reduced methyl 2 (hydroxymethyl) acrylate with only 19% conversion but 98% ee R (Table 4 1). For the 2 (hydroxymethyl) 2 cyclohexenone, OYE1 gave very little conversion. This fact was not surprising because it has been previously published that OYE1 can only reduce 2 ethyl 2 cyclohexenone with 14% conversion. 68 Lastly 2 hydroxymethyl 2 cyclope nteneone was reduced with 51% conversion and 60% ee R Among the OYE1W116 mutants screened with these substrates, W116Q and W116H performed well, giving % ee values of 98% S 97% S respectively and conversion of 78% and 67% respectively. The two best con verters

PAGE 84

84 were W116L and W116N, both reaching 98% conversion, but each one had poor optical purity 20% ee S and 41% ee S For this substrate, production of pure R was poor. None of the enzymes that favored R gave more than 70% conversion and of the two tha t gave 90% ee R or better conversion was only 19% and 9% (wild type and W116A respectively). This does not solve the problem, but it does suggest that further protein engineering may solve the problem of the production of the R enantiomer. This would be ideal since the previously published example producing R (mentioned in Chapter 1) involved the addition and subsequent removal of TBDMS, and no direct pathway to produce the Roche ester. 7 6 For the substrate 2 (hydroxymethyl) 2 cyclohexenone, many of the W 116 mutants gave conversions of 98% combined with 98% ee S Absolutely no mutants gave the R enantiomer. For the substrate 2 hydroxymethyl 2 cyclopentenone OYE1 wild type actually gave the R enantiomer which would result from the flipped binding conforma tion. This is somewhat interesting because all of the OYE1W116 mutants that had measurable conversion gave the S enantiomer. The best result from this substrate being the reduction with the mutant W116F which gave the conversion of 98% and also a 98% ee S These results are interesting when compared to the enzyme OYE2.6 from Pichia stipitis This enzyme was chosen in the past because in the position that is equivalent to W116 in wild type OYE1, OYE2.6 has an isoleucine. In previous studies OYE2.6 had st ereoselectivity most resembling wild type OYE1. 8 2 Though it has in some instances resulted in complementary stereoselectivity. 99 When OYE2.6 is directly compared to OYE 1, it outperforms OYE1 for all three substrates with respect to both conversion % and optical purity. Binding and enantioselectivity of OYE2.6 do not resemble OYE1

PAGE 85

85 since for all three substrates the product produced is in the S configuration. Comparing OYE2.6 to the mutants for the first two substrates, it outperforms all of the W116 rep lacements. However, for 2 hydroxymethyl 2 cyclopenteneone, seven of the mutants matched the OYE2.6 conversion percentage and outperformed its stereoselectivity. One mutant W116F gave 98% conversion and 98% ee S. Crystallographic Study To be sure that the trans addition was preserved and the different enantioselectivities were due to substrate flipping, the OYE1W116I mutant enzyme was crystallized and soaked with 2 hydroxymethyl 2 cyclopentenone. The substrate bound with 80% occupancy, and when it was modeled in the active site with one binding mode, there were two regions of unaccounted for electron density. The electron density was adequately accounted for after modeling the enone as a one to one mixture of binding modes (Figures 4 1, 4 2). The two binding modes represent what is expected inside of the enzyme with the carbonyl oxygen sitting between the residues N194 and H191. The electron density map also shows two orientations for Y196 and also H191. This probably corresponds to the two binding m odes of the substrate (Figure 4 3). In the binding orientation where the hydroxymethyl group is pointing toward I116 for one of the orientations of Y196, there should be steric clash between the oxygen of the substrate and the phenolic oxygen of Y196. Th e second conformation of Y196 moves this residue away from the substrate but also causes steric clash with H191. It is presumed that this steric clash between tyrosine and histidine is relieved by histidines other conformation. The other binding mode seen in the electron density does not represent a favorable Michaelis complex. It has been observed in past studies that the angle

PAGE 86

86 formed by FMN N10 N5 substrate carbon should be between 96 and 117 for a binding mode to lead to hydride transfer. 100 The a ngle in this binding mode is 78 One last point that helps explain why this binding mode produces little conversion is that Y196 puts its oxygen 4.45 away from the substrates carbon which is too far for proton transfer. This accounts for the low pro duction of the enantiomer that results from this binding mode. Materials and Methods Restriction endonucleases and T4 DNA ligase were purchased from New England Biolabs. Primers were ordered from Integrated DNA Technologies. Glucose dehydrogenase was gra ciously provided by Codexis. Plasmids were purified on large scales using CsCl density gradient ultracentrifugation. DNA sequencing was performed by the University of Florida ICBR using the Sanger capillary fluorescence methods. LB medium contained 5 g / L Bacto Yeast extract, 10 g / L Bacto Tryptone and 10 g / L NaCl. Kanamycin and ampicillin were used at concentrations described by Maniatis 40 4C. ( R ) ter was purchased from Sigma Aldrich. Chiral GC analyses were Dex column with helium as the carrier gas and an FID detection. GC MS analyses were carried out using a DB 17 column and EI ionization at 70 eV. Enzyme screening reactions and GC analysis were performed by Adam Walton, the substrates were synthesized by Dr. Bradford Sullivan, and crystallography trials were performed by Yuri Pompeu. Subcloning From a Saccharomyces cerevisiae Expression Plasmid I generated the mutants W116M, W116Y, and W116S by digesting separately with restriction endonuclease enzymes Bsa I and Kpn I from the yeast shuttle vector

PAGE 87

87 PSKP3W116X. 82 The resulting 577 bp fragment containing position 116 was isolated and purified by low melt agarose gel electrophores is. Simultaneously the E coli vector pDJB5, which housed the wild type OYE1 gene, was digested with Bsa I and Kpn I and the 7150 bp fragment was purified using low melting agarose gel electrophoresis. Fragments were joined together using T4DNA ligase and transformed into E coli JM109. Colonies were analyzed for the presence of the insert first by PCR and then by Sanger DNA sequencing. QuikChange Site Directed Mutagenesis I produced the W116 mutants W116G, W116H, W116Q, W116V, W116T, W116N, W116D, W116E, W116R, W116C, W116K using an adaptation of the QuikChange method. The procedure was 18 cycles of amplification using 95 C (30 s), 55 C (1 min) and 68C (9 min) followed by final extension time of 9 min at 68C. Each reaction had a Pfu turbo buffer, 50 ng pDJB5, 125 ng Pfu turbo DNA polymerase. The resulting DNA mixture was digested with DpnI overnight at 37C to fragment the template DNA. The resulting plasmid was transformed into E coli JM109 using electroporation. Protein Purification P rocedure I purified the mutated proteins using the overexpression plasmids transformed into E. c oli BL 21 (DE3) using electroporation. Proteins tagged to GST were purified as described previously. 8 2 A 5 mL overnight culture of the appropriate strain was grown using LB growth me dium and 40 / mL kana mycin, with shaking at 37 C. The overnight culture was diluted 1 : 100 into a 500 mL culture of LB / kanamycin and grown at 37C with shaking. When the optical density O.D. 600 nm of the culture was between

PAGE 88

88 0.5 and 1. 0, protein expression was induced by adding isopropyl thio D g alactoside (IPTG) to a final concentration of 100 At this stage, the culture was shaken at room tempera ture for 6 hours. The cells were then pelleted by centrifugation, washed twice with cold sterile water an d pelleted by centrifugation again. They were resuspended in 30 mL of cold loading buffer ( 50 mM Tris Cl, 4 mM MgCl 2 1 mM DTT ( add ed immediately before use), 1 mM PMSF (also added immediately before use), and 10% glycerol, pH 7.5 ). C ells were then lysed using a French press ure cell (10,000 20,000 psi), and insoluble debris was removed by centrifugation at 15,000 x g for 20 min at 4 C. The supernatant was then passed through a 0 onto a column containing 10 mL of glutathione resin It was then re circulated using a peristaltic pump for 3 5 hrs using a flow rate of ~0.5 mL / min. Unbound proteins were then allowed t o flow though to waste and the column was washed twice with 20 mL portions of cold loading buffer ( same as ab ove). The bound flavoprotein was then eluted using 40 mL elution buffer (39.6 mL loading buffer, 0 .40 mL 2M NaOH, and 0 .31 g reduced glutathione). The eluant was concentrated by ultrafiltration (Amicon YM 30 membrane) and then dialyzed overnig ht in buffer ( 20 mM Tris Cl, 4 mM MgCl 2 55 mM NaCl, 2 mM EDTA, 1 mM DTT, 50% glycerol, pH 7.5) prior to storage at 20 C Protein concentrations were determined by Bradford analysis. 88 The glutathione resin was regenerated using high salt buffers to rem ove any remaining proteins. Buffer A (0.1 M Tris Cl, 0.5 M NaCl, pH 8.5) and buffer B (0.1 M Sodium acetate, 0.5 M NaCl, pH 4.5) were each passed through the column in alternating 200 mL amounts until each had been used 3 times.

PAGE 89

89 Experimental Procedure for Screening Small Scale Reactions Adam Walton screened the enzymes in 50 mM KPi, pH 7.5 (300 L total volume). Reactions contained 200 mM glucose, 25 U / mL glucose dehydrogenase, 0.3 mM NADP + 10 mM alkene substrate, and 100 g alkene reductase. Reaction s were run with agitation for 24 h, and then extracted using 0.5 mL ethyl acetate. Reaction products were analyzed by GC. Reduction of Alkene 96 by GST OYE 2.6 Crude Extract Adam Walton screened Alkene 96 using an OYE 2.6 crude extract using the following procedure. One colony of E.coli BL21(DE3) with the plasmid pBS2 ( a derivative of pET22b ( ) with the coding sequence of Scheffersomyces stipitis OYE2.6 fused to GST) was picked from an LB ampicillin plat e and used to inoculate 40 mL of LB ampicillin medium. The culture was grown to saturation by growing at 37 C overnight, then added to 4 L of LB ampicillin supplemented with 20% glucose (80 mL), and AF 204 (0.5 mL, Sigma). Cells were grown in a New Bruns wick M19 fermenter at 37 C, 700 rpm, and 1 vvm airflow for 2.5 h until reaching O.D. 600 = 0.8, then the temperature was reduced to 30C and fusion protein overexpression was induced by adding IPTG to 100 M. After 4 h the cells were harvested by centrifu gation and stored at 20C. Cells were then thawed resuspended in 30 mL .1M KPi, 10 M PMSF, pH 7.5 and passed through a French pressure cell twice. Insoluble material was pelleted by centrifugation at 27,000 x g for 45 min at 4C. Large scale reactions were performed in 100 mK Kpi, pH 7.5 (90 mL) with glucose (8.0g), the mixture was then degassed by bubbling with Ar for 1 h. Glucose dehydrogenase 1.0 mg, NADP + 10 mg, and OYE2.6 crude extract 450 U were added and equilibrated for 15 min. Alkene 96 0.63 g was

PAGE 90

90 added and the reaction was monitored using a pH stat at 7.5 with 1 M KOH as the titrant. GC MS analysis confirmed that no substrate remained when base demand dropped. The product was extracted using methylene chloride (first 100 mL and then 2 x 50 mL). The organic layers were combined and then filtered through Celite washed with brine, dried with Na 2 SO 4 and concentrated using a rotary evaporator. Enantiomeric excess was calculated using GC MS and ( S ) configuration was assigned using optical rot ation. 10 1 Reduction of Alkene 97 by GST OYE2.6 Whole Cells Adam Walton screened alkene 97 using the following procedure. A colony of E. coli BL21(DE3) with the plasmid pBS2 was picked from an LB ampicillin plate and used to inoculate 15 mL LB ampicillin medium. The culture was grown to saturation by growing at 37 C overnight with shaking, and then the culture was added to 500 mL LB amicillin with 0.40% glucose. The cul ture was agitated in a baffled flask at 200 rpm until O.D. 600 reached 1.0. At this point protein overexpresion was induced by adding IPTG to a final concentration of 100 M, and shaken at 30 C for 4 h. The cells were harvested by centrifugation and resu spended in 300 mL of 50 mM KPi, pH 7.5 with 0.10 M glucose. Alkene 97 300 mg was added and the mixture was shaken at 30C for 1.5 h. GC MS analysis showed that all of the starting material had been consumed. To extract the product the reaction mixture w as slowly stirred overnight with 500 mL CH 2 Cl 2 at room temperature. The emulsion formed was filtered through Celite to remove insoluble material, washed with brine, dried over Na 2 SO 4 and the solvent was removed using a rotary evaporator. The absolute configuration of the reduction product was determined by Dr. Bradford Sullivan to be S after Baeyer Villier oxidation to the

PAGE 91

91 corresponding lactone. Spectral data were consistent with pr eviously reported data by Lees and Whitesides. 10 2

PAGE 92

92 Table 4 1. Results of the Baylis Hillman adduct reductions. Appears as published in ACS Catalysis 2011, 1, 989 993 % conv % ee % conv % ee % conv % ee S. p astorianus OYE1W116 mutants Trp (wt) 19 >98( R ) 5 N.D. 51 60 ( R ) Ala 9 90( R ) 84 >98 ( S ) > 98 72 ( S ) Val 52 86( R ) 84 >98 ( S ) 97 92 ( S ) Tyr 68 76( R ) >98 >98 ( S ) >98 87 ( S ) Phe 37 70( R ) >98 >98 ( S ) 98 >98 ( S ) Ser 13 46( R ) 84 >98 ( S ) 87 >98 ( S ) Ile 50 9( R ) >98 >98 ( S ) >98 91 ( S ) Arg N.D. 5 N.D. 5 N.D. Pro 5 N.D. 14 >98 ( S ) 16 77 ( S ) Thr 5 N.D. 28 >98 ( S ) 44 >98 ( S ) Cys 5 N.D. 31 >98 ( S ) 47 77 ( S ) Lys 5 N.D. 60 >98 ( S ) 75 76 ( S ) Glu 5 N.D. 93 90 ( S ) 96 88 ( S ) Asp 5 N.D. >98 91 ( S ) 95 77 ( S ) Gly 14 16 ( S ) 98 >98 ( S ) >98 86 ( S ) Leu >98 20 ( S ) >98 >98 ( S ) >98 57 ( S ) Asn >98 41 ( S ) >98 >98 ( S ) >98 89 ( S ) Met 15 64 ( S ) >98 >98 ( S ) >98 86 ( S ) His 67 97 ( S ) >98 >98 ( S ) >98 77 ( S ) Gln 78 >98 ( S ) >98 >98 ( S ) >98 89 ( S ) OYE 2.6 >98 >98 ( S ) >98 >99 ( S ) >98 76 ( S )

PAGE 93

93 Figure 4 1. Electron density map with the amino acids from OYE1W116I modeled in. Reprinted with permission from ACS Catal. 2011, 1 989 993. Copyright 2011 American Chemical Society Figure 4 2. Substrate modeled into the electron density map. A. shows there is unaccounted for electron density when the compound is modeled in one conformation. B shows the fit by two binding modes which explains the electron density. Reprinted with permission f rom ACS Catal. 2011, 1 989 993. Copyright 2011 American Chemical Society.

PAGE 94

94 Figure 4 3. Figure demonstrating the different conformations of I116, H191, and Y196.

PAGE 95

95 CHAPTER 5 BETA PHENYL COMPOUNDS Phenyl Compounds Compounds with a phenyl group attached to the carbon, like m ethyl trans cinnamaldehyde are interesting compounds to study with OYE1. One reason is that some of them represent previously published abnormal binding modes in OYE1. 67 Another reason is that these compounds have different complications associated with synthesizing or purchasing the cis vs. trans alkene isomers. M ethyl trans cinnamaldehyde being the predominant example in this study fits both of these descriptions. As previ carbon, trans to the carbonyl, and larger than a methyl group are typically bad substrates. 68 However, it has also been previously reported that m ethyl trans cinnamaldehyde is a very good substrate for reduction with a turnover number of 110 min 1 67 Upon further review, it has been noted that this substrate is reduced to the S enantiomer which represents a flipped binding mode in the active site (Figure 5 1). 10 3 In addition, this substrate is a natural product that is readily available and sold predominantly in the trans isomer (95%) but is not available in the cis Compounds like this are a good example of why the strategy of using cis vs trans isomers as star ting materials to get a desired reduction product may not be a suitable strategy to follow. For these reasons, two compounds were chosen for study with the enzymes, m ethyl trans cinnamaldehyde and (Z) ethyl 2 fluoro 3 phenylacrylate

PAGE 96

96 Results and Discuss ion We first employed standard conditions to obtain a reduced standard for m ethyl trans cinnamaldehyde (H 2 with palladium on carbon) but obtained no reaction. A derivatization pathway was therefore utilized to break up any resonance stabilization of the alkene caused by the pi bonding system. This involved protection of the aldehyde using an acetalization, reduction using palladium on carbon under a hydrogen atmosphere, then deprotection of the aldehyde (Figure 5 2). None of the mutants screened with m ethyl trans cinnamaldehyde showed any complementary enantioselectivity compared to OYE1 (Figure 5 3). The Tyr and Phe mutants at position 116 did show higher enantioselectivity for the same enantiomer and the Phe mutant even showed nearly identical conv ersion compared to the wild type (93.4% conversion for OYE1 and 92.7% for OYE1W116F). Having no enantiocomplementary products was expected because of how sterically limited the large phenyl ring is (Figure 5 1). Even though this is an expected result, th ere has been some success reported on finding enzyme homologs that can produce the enantiocomplementary product of cinnamaldehyde derivatives. 10 4 Unfortunately for the substrate (Z) ethyl 2 fluoro 3 phenylacrylate none of the mutants had any favorable act ivity. OYE1 displayed only 15% conversion, and no other enzyme reduced the substrate at all (Figure 5 4). OYE1 is usually poor at reducing an alkene conjugated with an ester, so finding a mutant that could better reduce the compound would have also been a desirable result. Attempts at increasing the activity with some substrates in the past have been successful.

PAGE 97

97 Materials and Methods methyl trans cinnamaldehyde and ethylene glycol was purchased from Sigma Aldrich. Palladium on carbon and p toluene s ulfonic acid were purchased from Fisher. (Z) ethyl 2 fluoro 3 phenylacrylate and its reduced standard were gifts from Dr. Neil Stowe. Experimental P rocedure for S creening I screened the substrates in 1 mL 0.1 M potassium phosphate buffer pH 7.0 for 24 hou rs at room temperature. Reactions contained 14 mol glucose, 5 g glucose dehydrogenase, 0.2 + Trans 2 (1 phenylprop 1 en 2 yl) 1,3 dioxolane In order to make a reduced standard, I first derivatized m ethyl trans cinnamaldehyde into an acetal using the following procedure. methyl trans cinnamaldehyde 20 mL and ethylene glycol 20 mL with p toluene sulfonic acid 100 mg was dissolv ed i n toluene 100 mL and heated to reflux in a Dean Stark trap. The resulting acetal was purified using recrystal l ization (Figure 5 2 compound 98). 1 H NMR (CDCl 3 300Mz) 4.05 (ddd, 4H) 5.3 (s 1H) 6.68 (d, 1H) 7.28 (m, 5H) 13 C NMR, 155.4, 136.9, 134.6, 130.2, 129.3, 128.3, 127.2, 107.9, 65.7, 12.1, (EI) 190(M), 175(100), 131(30), 117(50.9), 91(23.6) 73(45.7), MP. 59 60 C. Conversion analysis was performed using a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R =17.5 min. 2 (1 phenylpropan 2 yl) 1,3 dioxolane I then reduced the derivatized product using the following procedure. The product from the acetalization 200 mg was placed in 25 mL etha nol with 10 mg of p a lladium activated carbon under an H 2 atmosphere using

PAGE 98

98 a balloon (Figure 5 2 compound 99 ). A small sample of thi s was removed using a capillary, filtered with Celite and run on GC to verify reaction c onversion Upon completion of the reaction t he activate d carbon was filtered away using Celite and the resulting solution was concentrated on a rotary evaporator. NMR data was compared to previously published results 10 4 1 H NMR .9 (d, 3H), 2.0 (m 1H), 2.4 (dd, 1H), 2.9 (dd, 1H), 3.9 (dm, 4H), 4.8 (d, 1H), 7.2 (m, 5H), 13 C NMR 140.6, 129.4, 128.4, 126, 107, 65.29, 65.26, 39, 37.8, 13.5. (EI) 192(M), 130(2.6), 114(8.45), 91(16.5), 73(100). Conversion analysis was performed usin g a GC MS (Hewlett Packard) with a DB 17 column with helium as the carrier gas, 60 C 2 min, 10 C / min increase, 200 C 5 min, T R =16.0 min. 2 methyl 3 phenylpropanal I then deprotected the acetal back to the aldehyde which was used as a reduced standard o f m ethyl trans cinnamaldehyde 2 (1 phenylpropan 2 yl) 1,3 dioxolane 100 mg was added to 5 mL ethanol. Water 100 L was added with 10 mg paratoluene sulfonic acid to deprotect the aldehyde (Figure 5 2 compound 100) This reac tion mixture was heated ov er a B unsen burner to facilitate the reaction. A 0.25 mL aliquot of this reaction was washed with 1mL water and extracted using 2x0.5mL ethyl acetate and run on nonchiral GC MS for analysis. Conversion analysis was performed using a GC MS (Hewlett Packar d) with a DB 17 column with helium as the carrier gas, 60 C 2 min 10 C / min, 200 C 5 min, T R 12.25 min. (EI) 148(M), 105(22), 91(100), 78(26) 65(20). Chiral separation and ee calculations were achieved using GC FID with helium as the carrier gas and a beta dex column, 60 C 2 min, 1 C / min increase for 70 min, 130 C 1 min, 10 C / min increase, 180 C 5 min, T R = R enantiomer 50.1 min, S enantiomer 50.5 m in.

PAGE 99

99 A. B. Figure 5 methyl trans cinnamaldehyde binding modes In OYE1. The binding mode that is recognized as the normal mode for most substrates is represented in A. However, i t is also known that this is not sterically favorable due to the large phenyl ring trans to the carbonyl. This mode is also known to be the non catalytically active binding mode for this substrate. The catalytically active binding mode is represented in B. which flips the substrate in the active site. Dalton Trans. 2010 39 8472 8476 Figure 5 2 Synthesis pathway for the racemic standard of reduced alpha methyl trans cinnamaldehyde

PAGE 100

100 Figure 5 3. Enzyme reductions with M ethyl trans cinnamaldehyde Figure 5 4. Enzyme reductions with (Z) ethyl 2 fluoro 3 phenylacrylate

PAGE 101

101 CHAPTER 6 FUTURE WORK Future Work With respect to site directed mutagenesis of OYE1, some optimization of the procedure still needs to be performed. This work has been championed by Stewart group members Dr. Bradford Sullivan and doctoral candidate Adam Walton (work performed mostly on OY E2.6). Under their plan, a faster more reliable method has been developed that saturates each amino acid position. The goals of this work are important in a cooperative effort for other projects, such as engineering a protein to give R selectivity in the Baylis Hillman adduct work. If one follows the path of iterative site directed mutagenesis, the most promising starting points for further mutation were either wild type or W116 mutants of OYE1. Catalytic study will not be limited to Baylis Hillman adduc ts in future work. The original purpose of using substrates that can be made into chiral synthons downstream can be applied to other compounds. Enzymatic alkene reductions are well characterized with model compounds which have very little downstream impo rtance. In order for the study of alkene reductions to move forward, the question of industrial viability must be addressed. With respect to the project of reducing halogenated acids, the original plan for this project included analyzing whether OYE1 and the mutant enzymes could reduce the acid and the corresponding methyl ester compounds. Fortunately this very question was answered recently by Parmaggiani. 10 5 According to this study, OYE1, OYE2, and OYE3 can reduce the methyl ester of chloroacrylic and bromoacrylic acid with good conversion (95 100%) and ee values ranging from 94 98% S The fact that only S

PAGE 102

102 enantiomers resulted means some future work may be pursued for finding a way to produce the R enantiomers. This may pose a formidable challenge con sidering the bulkiness of the bromine atom. The 2,5 disubstituted cyclohexenones were chosen as a model to measure the hindrance on substrates caused by the binding pocket of OYE1 and how that changes as groups get bulkier. The results of this project wil l be difficult to interpret without producing enantiomerically pure starting materials. Perhaps derivatization followed by preparative HPLC could solve this. One issue, however, is whether or not this information is worth pursuing. New data suggests sig nificant differences in selectivity can be caused by conservative changes in amino acid identity. A fine example of this is the precursor molecule to the Roche ester with respect to OYE1W116I and OYE1W116L. Without molecular dynamics study, it is difficu lt to predict the outcome of adding steric bulkiness to a substrate and how that will respond to an amino acid change. Concluding Remarks For an enzyme like the old yellow enzyme that has 400 amino acids in its sequence, the total number of possible sequen ces is 20 400 Some perspective can be reached on the scale of this number by comparing it to the current estimate of the total number of atoms in the known universe (10 80 atoms). 10 6 So ev en for studies involving error prone PCR and high throughput screening, one of the first concerns to be addressed is how to limit the number of samples down to a manageable size. In addition, many of the mutations that will be incorporated will have a wide range of effects from absolut ely no impact to completely deleterious. Incorporating the strategy of limiting mutations to areas of the protein that are more likely to have the greatest

PAGE 103

103 effect on binding, such as the active site, may help decrease the costs associated with these studi es and allow smaller laboratories to have an impact on biocatalytic processes. The work investigating whether an enzyme may be made to be suitable in process development will continue, involving more amino acid positions to produce new chiral building blo cks. This lab has proven that it is possible through active site substrates for which the wild type enzyme is a poor performer. This has been done with both model compound s and compounds that are important for the production of other products. Using mutated enzymes from the Old Yellow E nzyme super family may be a viable approach for some niche chemical processes.

PAGE 104

104 APPENDIX A SCHEME FOR QUIKCHANGE METHOD

PAGE 105

105 APPENDIX B SCHEME FOR RESTRICTION DIGESTIONS AND LIGATIONS Bsa I then Kpn I T4 DNA Ligase

PAGE 106

106 APPENDIX C SEQUENCING CHROMATOGRAMS Figure C 1. W116G Figure C 2. W116H Figure C 3. W116Q

PAGE 107

107 Figure C 4. W116M Figure C 5. W116Y Figure C 6. W116S

PAGE 108

108 Figure C 7. W116V Figure C 8. W116T Figure C 9. W116N

PAGE 109

109 Figure C 10. W116D Figure C 11. W116E Figure C 12. W116R

PAGE 110

110 Figure C 13. W116C Figure C 14. W116K Figure C 15. T37A

PAGE 111

111 APPENDIX D PRIMERS USED Primer alignments Forward Primers A: OYE 1 AAG AAA TCG TTC GTT TGG GTT CAG TTA TGG GTT TTG GGT TGG GC T GCT TTC CCA W116C AA TCG TTC GTT TGG GTT CAG TTA TGC GTT TTG GGT TGG G W116D AA TCG TTC GTT TGG GTT CAG TTA GAT GTT TTG GGT TGG G W116E AA TCG TTC GTT TGG GTT CAG TTA GAA GTT TTG GGT TGG G W116G CG TTC GTT TGG GTT CAG TTT GGC GTT TTG GGT TGG G W116H AA TCG TTC GTT TGG GTT CAG TTA CAT GTT TTG GGT TGG G W116K C GTT TGG GTT CAG TTA AAG GTT TTG GGT TGG GCT GCT TTC CC W116N AA TCG TTC GTT TGG GTT CAG TTA AAC GTT TTG GGT TGG G W116Q C GTT TGG GTT CAG TTA CAG GTT TTG GGT TGG GCT GCT TTC CC W116R AA TCG TTC GTT TGG GTT CAG TTA CGC GTT TTG GGT TGG G W116T AA TCG T TC GTT TGG GTT CAG TTA ACC GTT TTG GGT TGG G W116V AA TCG TTC GTT TGG GTT CAG TTA GTG GTT TTG GGT TGG G B: Reverse primers W116C G GAA AGC AGC CCA ACC CAA AAC GCA TAA CTG AA C CC W116D G GAA AGC AGC CCA ACC CAA AAC ATC TAA CTG AA C CC W116E G GAA AGC AGC CCA ACC CAA AAC TTC TAA CTG AA C CC W116G G GAA AGC AGC CCA ACC CAA AAC GCC TAA CTG AA C CC W116H G GAA AGC AGC CCA ACC CAA AAC ATG TAA CTG AA C CC W116K GTC TGG GAA AGC AGC CCA ACC CAA AAC CTT TAA CTG AAC CC W116N G GAA AGC AGC CCA ACC CAA AAC GTT TAA CTG AA C CC W116Q GTC TGG GAA AGC AGC CCA ACC CAA AAC CTG TAA CTG AAC CC W116R G GAA AGC AGC CCA ACC CAA AAC ACG TAA CTG AAC CC W116T G GAA AG C AG C CCA ACC CAA AAC GGT TAA CTG AA C CC W116V G GAA AGC AGC CCA ACC CAA AAC CAC TAA CTG AA C CC

PAGE 112

112 APPENDIX E NMR DATA OF PREVIOUS LY UNREPORTED COMPOU NDS Figure E 1. 1 H NMR of 2,5 dimethyl cyclohexenone

PAGE 113

113 Figure E 2. 13 C NMR of 2,5 dimethylcyclohexenone.

PAGE 114

114 Figure E 3. 1 H NMR of 2 methyl 5 propylcyclohexenone.

PAGE 115

115 Figure E 4. 13 C NMR of 2 methyl 5 propylcyclohexenone.

PAGE 116

116 Figure E 5. 1 H NMR of trans 2 (1 phenylprop 1 en 2 yl) 1,3 dioxolane

PAGE 117

117 Figure E 6. 13 C NMR of trans 2 (1 phenylprop 1 en 2 yl) 1,3 dioxolane

PAGE 118

118 LIST OF REFERENCES 1) Heth, G.; Nevo, E.; Ikan, R.; Weins tein, V.; Ravid, U.; Duncan, U. Experintia 1992 48 897. 2) Filmore, D. Mod. Drug. Discov June, 2003 35 39 3) Liese, A.; Seelbach, K.; Wandrey, C. Industrial Biotransformations Wiley: New York, 2006 4) Stewart, J. D. Curr. Opin. Biotech. 2000 11 363 368. 5) Chartraina M.; Robergea, C.; Chungb, J.; McNamarab, J.; Zhaob, D.; Olewinskia, R.; Hunta, G.; Salmona, P.; Rousha, D.; Yamazakia, S.; Wangc ,T.; Grabowskib, E; Bucklanda, B; Greashama R. Enzyme and Microb. Technol 1999 25 489 496 6) Ste w art, J. D. Org. Lett. 1999 1 1155 1159 7) Martzen, M. R.; McCraith, S. M.; Spinelli, S. L.; Torres, F. M.; Fields, S.; Grayhack, E. J.; Phizicky, E. M. Science 1999 286 1153 1155. 8) Huisman, G. W.; Liang J.; Krebber, A. Curr. Opinn. Chem. Biol ., 2010 14 122 129 9) Wohlgemouth, R. New Biotech 2009 25 4, 204 213 10) Pollard, D. J.; Woodley, J. M. Trends Biotechnol 2007 25 66 73 11) Huisman, G. W.; Lalonde, J. J. Enzyme Evolution for Chemical Process Applications. In Biocatalysis in the Pharmaceutical and Biotechnology Industries ; Patel, R. Ed; CRC Press, USA, 2006 717 742 12) Prelog V. Pure and Applied Chemistry 1964 9 119 130. 13) Bougioukou, D. J. Evaluation of old and new alkene reductases as potential biocatalysts. Ph.D. T hesis, University of Florida, 2006 14) Pritchard, L.; Corne, D.; Kell, D.; Rowland, J.; Winson, M. J Theor Biol 2005 234 497 509 15) Eggert, T.; Reetz M.T. ; Jaeger, K. E. Directed evolution by random mutagenesis: A critical evaluation. In Enzyme Functionality: Design, Engineering, and Screenin g ; Svendsen, A., Ed.; Marcel Dekker, New York, 2004 ; pp 375 390. 16) Reetz, M.; Wang, L. Angew. Chem. Int. Ed. 2006 45 1236 1241

PAGE 119

119 17) Reetz, M. T.; Bocola, M.; Carballeira, J. D.; Zha, D.; Vogel, A. Angew. Chem. Int. Ed 2005 44 4192 4196 18) Reetz M T.; Carballeira J D. Nat Protoc 2007 2 891 903. 19) Machielsen R.; Leferink N.; Hendriks A.; Van der Oost J. Extremophiles 2008 12 587 594 20) Liang, J., Mundorff, E., Voladri, R., Jenne, S., Wong, J., Huisman, G., Truesdell, S., Lalonde, J., Org. Proc. Res. & Dev 2010 14, 188 192 21) Machielsen, R.; Looger, L.; Raedts, J. Eng. Life Sci 2009 9 38 44 22) Varughese, K. I.; Xuong, N. H.; Kiefer, P. M.; Matthews, D. A.; Whiteley J. M. Proc. Natl. Acad. Sci USA 1994 91 5582 5586. 23) Tanaka, N.; Nonaka, T.; Nakanishi, M.; Deyashiki, Y.; Hara, A.; Mitsui Y. Structure 1996 4 33 45. 24) Nakanishi, M.; Matsuura, K.; Kaibe, H.; Tanaka, N.; Nonaka, T.; Mitsui, Y.; Hara, A J. Biol. Chem. 1997 272 2218 2222. 25) Zhang, R.; Xu, Y.; Sun Y.; Zhang, W.; Xiao R. Appl. Environ. Microbiol 2009 75 2176 2183 26) Hummel, W. Trends Biotechnol 1999 17 487 492. 27) Schlieben, N.; Niefind, K. J. Mol. Biol 2005 349 801 813 28) Zhu, D.; Hua, L. J. Org. Chem. 2006 71 9484. 29) Zhu, D.; Yang, Y.; Buynak, J. D.; Hua, L. Org. Biomol. Chem 2006 4 2690. 30) Zhu D. Yang, Y.Majkowicz, S.; Hua, L. Org. Lett 2008 10 525 528. 31) Ziegelmann Fjeld, K.; Musa, M.; Phillips, R. Prot. Eng. Des. & Selec 2007 20 47 55. 32) Keinan, E.; Hafeli, E. K .; Seth, K. K.; Lamed, R.; J. Am. Chem. Soc 1986 108 162 169. 33) Musa, M.; Ziegelmann Fjeld K.; Vieille J.; .Zeikus, G J. Org. Chem 2007 72 30 34 34) Sun, L.; Bulter, T.; Alcalde, M.; Petrounia, I. P.; Arnold, F. H. ChemBioChem 2002 3 781 783.

PAGE 120

120 35) Escalettes, F.; Turner, N. J. ChemBioChem 2008 9 857 860. 36) Alexeeva M.; Enright A.; Dawson M.; Mahmoudian, M.; Turner N., Angew. Chem. Int. Ed 2002 41 No. 17 3177 3180. 37) Fiala, G.; Stetter, K. Arch Microbiol 1986 145 56 61 38) Van der Oost J.; Voorhorst W.; Kengen S. Eur. J. Biochem 2001 268 3062 3068 39) Wierenga R.; Terpstra P.; Hol W. J Mol Biol 1986 187 101 107 40) Kataoka, M.; Nakamura, Y.; Urano, N.; Ishige, T.; Shi, G.; Kita, S.; Sa kamoto, K.; Shimizu, S. Lett. Appl. Microbiol 2006 43 430 435 41) Urano, N.; Fukui, S.; Kumashiro S.; Ishinge T.; Kita S.; Sakamoto K.; Kataoka M.; Shimizu S. J. Biosci Bioeng 2011 1 1 1 266 271 42) Asako H.; Wakita R.; Matsumura K.; Shimizu M.; Sakai J.; Itoh N. Appl Environ Microbiol 2005 71 1101 1104 43) Asako H.; Shimizu M.; Itoh N. Appl Microbiol Biotechnol 2008 80 805 812 44) Gooding, O.; Voladri, R.; Bautista, A.; Hop kins, T.; Huisman, G.; Jenne, S.; Ma, S.; Mundorff, E.; Savile, M. Org. Proc. Res. Dev 2010 14, 119 126 45) Evans, M. C.; Franklin, L. C.; Murtagh, L. M.; Nanninga, T. N.; Pearlman, B. A.; Saenz, J. E.; Willis, J. U.S. Patent Application 2,714,168A1, 2007 46) Morita, K. ; Hachiya, S.; Moriwaki, F.; Endo, H. U.S. Patent 5,281,685, 1994 47) Jakel, C.; Heydrich, G.; Sturmer, R.; Paciello, R. World Patent Application WO 2006/03 4812, 2006. 48) Quallich, G. J.; Woodall, T. R. Tetrahedron Lett. 1993 34 785 49) Zhang, X.; Taketomi, T.; Yoshizumi, T.; Kumobayashi, H.; Akutagawa, S.; Mashima, K.; Takaya, H. J. Am. Chem. Soc. 1993 115 3318 50) King, E. G. Corley, R. K. Anderson, R. D. Larsen, T. R. Verhoeven, P. J. Reider J. Org. Chem 1993 58 3731 3735 51) Liang, J.; Lalonde, J.; Borup, B.; Mitchell, V.; Mu ndorff, E.; Trinh, N.; Kochrekar, D.; Cherat, R.; Pai, G. Org. Proc. Res. Dev 2010 14, 193 198

PAGE 121

121 52) Liang, J.; Borup, B.; Mitchell, V.; Mundorff, E.; Lalonde, J.; Huisman, G. PCT Patent Application WO/2009/042984, 2009, pp 115 125 53) Jing Q.; Okrasa K.; Kazlauskas R. Chem. Eur. J. 2009 15 1370 1376 54) Wilson, M. E.; Whitesides, G. M. J. Am Chem. Soc. 1978 100 306. 55) Reetz, M.; Peyralans, J.; Maichele, A.; Fu, Y., Maywald, M. Chem. Commun 2006 13 4318 4320 56) Collot, J.; Grandinaru, J.; Humbert, N.; Skander, M.; Zocchi, A.; Ward, T. J. Am. Chem. Soc 2003 125 9030 9031 57) Letondor, C.; Humbert, N.; Ward, T. Proc N atl. Acad Sci 2005 102 13, 4683 4687 58) Miyahita, A.; Yasuda, A;Takaya, H.; Touriumi, K.; Ito, T.; Souchi, T.; Noyori, R. J. Am. Chem. Soc 1980 102 7932 7934 59) Tetsuo, O.; Tsutomu, M.; Nobuo, S.; Hidenori, K.; Hidemasa, T. J. Org. Chem. 1995 60 357 363 60) Chaparro Riggers, J. F.; Rogers, T. A.; Vazquez Figueroa, E.; Polizzi, K. M.; Bommarius, A. S. Adv. Synth. Catal 2007 349 1521 1531 61) Warburg, O.; Christian, W. Naturwissenschaften 1932 20 688. 62) Theorell, H. Biochem. Z 1935 278 263 290 63) Fox, K. M.; Karplus, P. A. Structure 1994 2 1089 1105 64) Murthy, Y. V. S. N.; Meah, Y.; Massey, V. J. Am. Chem. Soc 1999 121 5344 5345 65) Massey, V.; Schopfer, L. M. J. Biol. Chem 1986 261 1215 1222 66) Ko hli, R. M.; Massey, V. J. Biol. Chem 1998 273 32763 32770 67) Vaz, A. D. N.; Massey, V. Biochemistry 1995 34 4246 4256 68) Swiderska, M. A.; Stewart, J. D. J. Mol. Catal 2006 42 52 54 69) Stuermer, R.; Bernhard, H.; Hall, M.; Faber, K. Curr. Opin. Chem. Bio 2007 11 203 213

PAGE 122

122 70) Mugford, P. F.; Wagner, U. G.; Jiang, Y.; Faber, K.; Kazlauskas. Angew. Chem. Int. Ed 2008 47 8782 8793 71) Yokomizo, T.; Ogawa, Y.; Uozumi, N.; Kume, K.; Izumi, T.; Shimizu, T. J. Biol. Chem. 1996 271 2844 2850. 72) Bougioukou, D. ; Stewart, J. J. AM. CHEM. SOC. 2008 130 7655 7658 73) Faber, K. Eur. J. Org. Chem. 2008 73 1511 1516 74) Hall, M.; Stu eckler, C.; Faber, K. Adv. Synth. Catal 2008 350 411 418 75) Winkler, C. ; Faber, K. Eur. J. Org. Chem. 2010 33 6354 6358 76) Stueckler, C.; Winkler, C.; Bonnekessel, M.; Faber, K. Adv. Synth. Catal 2010 352 2663 2666 77) Cohen, N.; Eichel, W. F.; Lopersti, R. J.; Neukom, C.; Saucy, G. J. Org. Chem 1976 41 3505 3511 78) Branca, Q.; Fischli, A. Helv. Chim. Acta 1977 60 925 944 79) Evans, D. A.; Sacks, C. E.; Kleschick, W. A. Taber, T. R. J. Am. Chem. Soc 1979 101 6789 6791. 80) Prez, M.; Prez, D. I.; Marteinez, A.; Castrol, A.; Gmez, G.; Fall, Y. Chem. Commun 2009 10 3252 3254. 81) Smith B. III; Adams, C. M.; Lodise S. A., Barbosa,; Degnan, A. P. Proc. Natl. Acad. Sci. USA 2004 101 12042 12047. 82) Padhi, S. ; Bougioukou, D.; Stewart, J. J. Am. Chem. Soc. 2009 131 3271 3280 83) Bougioukou, D. ; Reetz, M. Adv. Synth. Catal 2009 351 3287 3305 84) Binks, P. R.; French, C. E.; Nicklin, S.; Bruce, N. C. Appl. Environ. Microbiol 1996 62 1214 1219. 85) Toogood, H. S.; Fryszkowska, A. H.; Mansell, D.; Stephens, Gill.; Gardine r, J. M.; Scrutton, N. S. ChemBioChem 2011 12 738 749 86) Luche, J. L. J. Am. Chem. Soc 1978 100 2226 222 7 87) Trachtenberg, E.; Carver, J. J. Org. Chem 1970 35 1646 1653 88) Bradford, M. Anal. Biochem 1976 72, 248 254.

PAGE 123

123 89) Chong, B. D.; Ji, Y. I.; Oh, S. S.; Yang, J. D.; Baik, W.; Koo, S. J. Org. Chem. 1997 62 9323 9325 90) Stueckler, C.; Hall, M.; Ehammer, H.; Pointner, E.; Kroutil, K.; Macheroux, P.; Faber, K. Org. Lett ., 2007 9 5409 5411 91) Utaka M. J. Org. Chem. 1989 54 4989 4992 92) Kurata et al. Tet Asym. 2004 15 2837 2839 93) Atsushi K. J. of Bio and Bioeng 2008 105 429 431. 94) Xu, D.; Kohli, R. M.; Massey, V. Proc. Natl. Acad. Sci. 1999 96 3556 3561 95) Kreiselmeier G. Eur. J. Org. Chem. 2005 12 4687 4698 96) Walton, A. Z.; Conerly, W. C.; Pompeu, Y.; Sullivan, B.; Stewart, J. D. ACS Catal. 2011 1 989 993 97) Chen, A. ; Xu, J.; Chiang, W.; Christina L.L. ; Chai, C. L. L. Tetrahedron 2010, 66 1489 1495 98) Mase, N.; Inoue, A.; Nishio, M.; Takabe, K. Bioorg. Med. Chem. Lett 2009 19 3955 3958 99) Bougioukou, D. J.; Walton, A. Z.; Stewart, J. D. Chem. Commun 2010 46 8558 8560. 100) Fraaije, M. W.; Mattevi, A. Trends Biochem. Sci. 2000 25 126 132. 101) Chen, A.; Xu, J.; Chiang, W.; Chai, C. L. L. Tetrahedron 2010 66, 1489 1495. 102) Lees, W. J.; Whitesides, G. M. J. Org. Chem 1993 58, 1887 1894. 103) Stueckler, C.; Mueller, N.; Winkler, C. K. ; Glueck, S.; Gruber, K.; Steinkellner a, G.; Faber K. Dalton Trans. 2010 39 8472 8476 104) Benohoud, M.; Tuokko, S.; Pihko P. Chem. Eur. J. 2011 17 8404 8413 105) Brenna, E.; Gatti, F. G.; Manfredi, A.; Monti, D.; Parmeggiani, F. Org. Proc. Res. Dev Available online 106) Turner, N. Nat. Chem Bio 2009 5 567 573

PAGE 124

124 BIOGRAPHICAL SKETCH W. Colin Conerly was born in DeRidder Louisiana. From there he moved to Pineville La. where he graduated from Pineville High School Colin was recruited by Northwestern State University in Natchitoches La. to perform on scholarship in the university symphony as a bass p layer. He earned his Bachelor of Science in c hemistry in May 2006. From there he traveled to the University of Florida to pursue his Ph. D. in the division of b iochemistry where he joined the laboratory of Dr. Jon Stewart. Shortly before completing his Ph.D., Colin was hired by Evolugate LLC where he hopes to be an impactful scientist, as a part of their team, for years to come.