INFORMATION to USERS the Most Advanced Technology Has Been

INFORMATION TO USERS

The most advanced technology has been used to photo graph and reproduce this manuscript from the microfilm master. UMI films the original text directly from the copy submitted. Thus, some dissertation copies are in typewriter face, while others may be from a computer printer.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyrighted material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are re produced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each oversize page is available as one exposure on a standard 35 mm slide or as a 17" x 23" black and white photographic print for an additional charge.

Photographs included in the original manuscript have been reproduced xerographically in this copy. 35 mm slides or 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Accessing the World'sUMI Information since 1938

300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA Order Number 8820378

Stereochemical studies in anaerobic metabolism

Zydowsky, Lynne Douthit, Ph.D.

The Ohio State University, 1988

UMI 300 N. Zeeb Rd. Ann Aibor, M I 48106 PLEASE NOTE:

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark V .

1. Glossy photographs or pages_____

2. Colored illustrations, paper or print______

3. Photographs with dark background_____

4. Illustrations are poor copy______

5. Pages with black marks, not original copy ______

6. Print shows through as there is text on both sides of page ______

7. Indistinct, broken or small print on several pages.

8. Print exceeds margin requirements ______

9. Tightly bound copy with print lost in spine ______

10. Computer printout pages with indistinct print ______

11. Page(s)______lacking when material received, and not available from school or author.

12. Page(s) seem to be missing in numbering only as text follows.

13. Two pages numbered . Text follows.

14. Curling and wrinkled pages

15. Dissertation contains pages with print at a slant, filmed as received

16. Other

UMI STEREOCHEMICAL STUDIES IN ANAEROBIC METABOLISM

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Lynne Douthit Zydowsky, B.S.

The Ohio State University

1988

Dissertation Committee: Approved By

Heinz G. Floss

Michael H. Klapper

Robert M. Mayer Adviser

Ming-Daw Tsai Department of Chemistry To Tom, Who is both my husband and best friend

ii ACKNOWLEDGMENTS

I would like to express my sincere appreciation to my research adviser, Dr. Heinz G. Floss, for his support and guidance during my stay in his research group. He gave me the opportunity to join his group even though he knew that my chemical backgound and research skills needed substantial development. I will always be greatly indebted to him for taking this chance with me.

I would like thank Dr. John N Reeve for opening his laboratory to me and providing research assistance along the way. I would also like to thank my dissertation committee members; Dr. Michael H.

Klapper, Dr. Robert M. Mayer, and Dr. Ming-Daw Tsai for their patience during the final stages of my graduate career.

I am indebted to my coworkers in the Floss group. They offered me the scientific assistance and friendship which made my stay at

The Ohio State University a rewarding and enjoyable experience. A special thanks goes to Jim Brown and Elizabeth Haas. They offered me not only their sceintific knowledge, but also their home and their hearts at a time when I needed both.

Dr. Leroy Salerni, my friend and former professor from Butler

University, did not let me settle for second best. He started my graduate career in the right direction by suggesting that I go to Purdue University and work for Dr. Heinz G. Floss. I will always be thankful for his advice and confidence in me.

To Tom Zydowsky, my husband and best friend, I owe a very special thanks. His selfless and loving support, both scientific and personal, was invaluable to me. He offered suggestions and comments which helped me scientifically, and gave me the support and encouragement to continue in pursuit of my career. For this I will always love him.

iv VITA

May 22,1958 ...... Born - Olney, Illinois

May, 1 9 8 1...... B.S., Butler University, Indianapolis, Indiana

August, 1981 - December, 1982 ...... Graduate Student PurdueUniversity, West Lafayette, Indiana

January, 1983 - January 1987 Graduate Research November 1987 - present Associate in the Department of Chemistry, The Ohio State University

Feburary 1987 - November 1987 ...... Proctor and Gamble Predoctoral Fellowship

PUBLICATIONS

1. L.D. Zydowsky, T.M. Zydowsky, E.S. Haas, J.W. Brown, J.N. Reeve and H.G. Floss. "Stereochemical Course of Methyl Transfer from Methanol to Methyl Coenzyme M in Cell-Free Extracts of Methanosarcina barkeri." J. Am. Chem. Soc,. 1987. 109. 7922.

2. S. Raybuck, N.R. Bastian, L.D. Zydowsky, K.Kobayashi, H.G. Floss, W.H. Orme-Johnson and C.T. Walsh. "Nickel-Containing CO Dehydrogenase Catalyses Reversible Decarbonylation of Acetyl CoA with Retention of Stereochemistry at the Methyl Group." J. Am. Chem. Soc. 1987. 109. 3171.

v 3. H. Lebertz, H. Simon, L.F. Courtney, S.J. Benkovic, L.D. Zydowsky, K. Lee and H.G. Floss. "Stereochemistry of Acetic Acid Formation from 5-Methyl Tetrahydrofolate by Clostridium thermoaceticum." J. Am. Chem. Soc. 1987. 109. 3173.

FIELDS OF STUDY

Bioorganic Chemistry and Mechanistic Enzymology. TABLE OF CONTENTS

PAGE

DEDICATION i i

ACKNOWLEDGMENTS...... i i i

VITA...... v

LIST OF FIGURES...... x i

LIST OF ABBREVIATIONS...... x iv

PART I INTRODUCTION...... 1

RESULTS AND DISCUSSION...... 10 Preparation of (R)- and (S)-[ 2H i, 3H]Acetyl Coenzyme A ...... 1 0 The Carbonylation Reaction with Purified Carbon Monoxide Dehydrogenase (CODH) from Clostridium thermoaceticum ...... 11 Stereochemical Analysis of Acetic Acid Formation from (Methyl-R)- and (Methyl-S)-[ 2 H i, 3 H]methyltetrahydrofolate by a Cell-Free Extract of Clostridium thermoaceticum. 1 3

CONCLUSION...... 16

EXPERIMENTAL...... 20 Materials and Methods ...... 2 0 Synthesis of (R)- and (S)-[ 2H-| 3H]Acetyl Coenzyme A 21 Carbonylation Exchange with Purified CO Dehydro genase (CODH) from Clostridium thermoaceticum with (R)- and (S)-[ 2H-| 3H]Acetyl Coenzyme A ...... 22 Conversion of (Methyl-R)- and (Methyl-S)- [2H i ,3 H]methyltetrahydrofolate with the Cell-Free Extract From Clostridium thermoaceticum ...... 23 PART II INTRODUCTION...... 24 Methanogenesis from CO 2 and H 2...... 27 Methanogenesis from Methanol, Methylamine and Acetate ...... 31

RESULTS AND DISCUSSION...... 3 6 Cultivation of Methanosarcina barkeri on Methanol Methylamine, and Acetate ...... 3 6 Preparation of an Active Cell-Free Extract from Methanosarcina barkeri ...... 3 9 Conversion of CD 3OH to CD3H with the Cell-Free Extract of Methanosarcina barkeri ...... 39 Synthesis of (R)- and (SJ-pH-j , 3H]Methanol ...... 40 Conversion of (R)- and (S)-[ 2H-| ,3H]Methanol (lk) and (la) to Acetate for Configurational Analysis ...... 52 Conversion of 14C-Methanol into Methyl Coenzyme M using the Cell-Free Extract of Methanosarcina barkeri and Isolation of Methyl Coenzyme M ...... 5 9 Conversion of (R)- and (S)-[ 2H i, 3 H]Methanol into (R)- and (S)-[2H i, 3H]Methyl Coenzyme M with the Cell-Free Extract of Methanosarcina barkeri ...... 60 Degradation of [ 2H-|,3H]Methyl Coenzyme M to Acetic Ac id ...... 61 Results from the Stereochemical Analysis of (2H i ,3H]Methyl Coenzyme M Dervied from (R)- and (S)- pH-i 3H]Methanol ...... 63

CONCLUSION...... 65

EXPERIMENTAL...... 6 8 Material and Methods ...... 6 8 Cultivation and Fermentation ...... 7 0 Harvesting and Preparation of the Cell-Free Extract ...... 74 Conversion of CD 3OH to CD3H with the Cell-Free Extract of Methanosarcina barkeri ...... 76 Formation of Trideuterated Methane by Methanosarcina barkeri Grown of Trideuterated Subtrates ...... 77 Synthesis of (S)-[2Hi , 3H]Methanol (la ) ...... 78 Conversion of (S)-[ 2H-| ,3H]Methanol (la ) to Acetate...... 79

viii Enzymatic Conversion of (S)-[ 2H i, 3 H]Methanol (la) to (S)-[2H i, 3H]Methyl Coenzyme M (Za) using a Cell-Free Extract of Methanosarcina barkeri ...... 81 Enzymatic Conversion of Trideuterated Methanol to Trideuterated Methyl Coenzyme M using the Cell-Free Extract of Methanosarcina barkeri ...... 82 Conversion of (S)-[ 2H i, 3H]Methyl Coenzyme M (Za) to Acetate ...... 8 3

PART III INTRODUCTION...... 8 5

RESULTS AND DISCUSSION...... 9 5 Preparation of the Cell-Free Extract of Methano sarcina barkeri ...... 95 Conversion of 2-Propenyl Coenzyme M to Propene with the Cell-Free Extract of Methanosarcina barkeri 9 5 Proton Exchange in the Methyl Coenzyme M Reduc tase reaction ...... 98 Synthesis of (R)- and (S)-2-[1- 2H i, 3 H ]P ropenyl Coenzyme M ...... 9 9 Stereospecific Reduction of 9,10-Ethanoanthracene- 11-[132Hi]methan-13-al (ID to (13S)- and (13R)- 9.10-Ethanoanthracene-11 -[13-2Hi]methan-13-ol (13a) and 03b) ...... 1 2 2

CONCLUSIONS...... 129

EXPERIMENTAL...... 132 Materials.and Methods ...... 1 32 2-Propenyl Coenzyme M (H ) ...... 1 33 Enzymatic Conversion of 2-Propenyl Coenzyme M ( 1 1 ) to Propene using the Cell-Free Extract of Methanosarcina barkeri ...... 134 Ethyl 9,10-Ethanoanthracene-11-carboxy!ate (12 ) ...... 135 9.10-Ethanoanthracene-11-[13-2H2]methan-13-ol (13).... 135 9.10-Ethanoanthracene-11-[13-2Hi]methan-13-al (ID---- 1 3 6 (13S)-9,10-Ethanoanthracene-11-[13-2Hi]methan- -13-01 (1 3 a )...... 1 37 (S )-2 -[1 -2 Hi]Propen-1-ol (15a) and (R)-2- [1 -2Hi]Propen-1-ol (1 5 b )...... 139 (S )-2 -[1 -2 Hi]Propenyl (S)-(+)-0-Acetylmandelate M6a1... 139 (R )-2 -[1 -2H i]P ro p e n yl (S )-(+)-0-A cetylm andelate (ISJa.)... 140 2-[1-2H2]Propen-1-olQ5) ...... 140 2-[1-2H2]Propenyl p-toluenesulfonate ( 1 2 ) ...... 141 2-[1 -2H2]Propenyl Coenzyme M (IS) ...... 141 9,10-Ethanoanthracene-11-[13-2H2]methyl-13-p- toluenesulfonate (IS) ...... 142 3-(Benzylthio)-9,10-Ethanoanthracene-11- [13-2H2]methane(2Q) ...... 142

APPENDIX...... 145 Configurational Analysis of Chiral Methyl Groups ...... 145

LIST OF REFERENCES...... 149

x LIST OF FIGURES

FIGURES PAGE

1 Scheme proposed in 1966 for the synthesis of acetate 3 in Clostridium thermoaceticum

2 Stereochemical predictions of acetate synthesis in 8 Clostridium thermoaceticum from chiral 5-methyltetrahydrofolate

3 Enzymatic synthesis of (R)- and (S)-[ 2H it3 H]acetyl 10 coenzyme A.

4 Degradation of chiral 5-methyltetrahydrofolate to 14 acetic acid for configurational analysis

5 Stereochemical outcome from the enzymatic conversion 1 5 of chiral 5-methyltetrahydrofolate to acetate with Clostridium thermoaceticum

6 Postulated mechanism for carbonylation reaction of 1 7 acetyl coenzyme A with CODH from Clostridium thermoaceticum.

7 Carbonylation reaction with organoiron species 19

8 Substrates utilized by methanogenic bacteria 24

9 Scheme proposed by Barker to account for 2 5 methanogenesis from methanol and CO 2

10 Strucuture of coenzyme M 26

11 Structures of methanofuran and tetrahydromethano 28 p te rin

1 2 Methanogenesis from CO 2 3 0

13 Proposed scheme for methanogenesis from methanol in 3 2 Methanosarcina barkeri.

xi 14 Proposed scheme for methanogenesis from acetate

15 Deprotonation of the diazonium ion to give diazomethane

1 6 Synthesis of methionine carrying a chiral methyl group

17 Strategy of the synthesis of chiral methanol

1 8 Synthesis and configurational analysis of chiral m ethanol

19 Proposed mechanism for racemization of chiral methanol by a transesterification reaction.

20 Proposed mechanism for racemization of chiral methanol by solvent participation with HMPA

21 Proposed mechanism for racemization of chiral methanol by a mixed hydrolysis reaction with LiOH.

22 Synthesis of methyltosylate carrying a chiral methyl group

23 Steric course of the enzymatic synthesis of methyl coenzyme M from methanol, and configurational analysis of methyl coenzyme M

24 Structure of F 430

25 Proposed mechanism of methyl coenzyme M reductase

2 6 Postulated role of HS-HTP (component B)

2 7 Substrates for methyl coenzyme M reductase

28 Conversion 2-propenyl coenzyme M to propene

29 Mass spectral fragmentation of propene

30 Original strategy for the synthesis of 2-propenyl coenzyme M

31 Diels-Alder chemistry for the synthesis of 2 -[ 1-2H 2]propenyl coenzyme M

xii 32 Diels-Alder chemistry for the synthesis of 106 (R)- and (S)-2-[1- 2H-|]propenyl coenzyme M

33 1H NMR spectrum of dideuterated allyl alcohol in 1 08 benzene-d 6

34 1H NMR spectrum of dideuterated 1 1 0 2-propenyl coenzyme M

35 2H NMR spectrum of dideuterated 111 2-propenyl coenzyme M

36 Possible reaction mechanisms for the synthesis of 112 2-propenyl coenzyme M

37 Regiochemical outcome of 1,1- and 3,3-dideuteroally 113 bromide with sulfur anions.

38 Structure of the allyl sulfonium salt 117

39 Structure of 2-propenyl coenzyme M sulfonic acid 117 propenyl ester

40 New strategy for the synthesis of 2 -[1-2 H 2]p ro p e n y l 119 coenzyme M using Diels-Alder chemistry.

41 Stereospecific synthesis of 1 23 2-[1-2H-|]propenyl coenzyme M

4 2 Strucuture of (S)-(+)-0-acetylmandelate ester 125 of allyl alcohol

43 1H NMR of the (S)-(+)-0-acetylmandelate ester of (R)- 127 and (S)-2-[1- 2Hi]propen-ol

4 4 2H NMR of the (S)-(+)-0-acetylmandelate ester of 128 (R)- and (S)-2-[1- 2 H-|]propen-1-ol

45 Mass spectral fragmentation of propene 130

46 Configurational analysis of chiral methyl groups 147

XIII LIST OF ABBREVIATIONS

Common Abbreviations

Alpine-BoraneR pinanyl-9-borabicyclo[3.3.1]nonane ATP adenosine 5'-triphosphate BrES bromoethanesulfonic acid Coenzyme M 2-mercaptoethanesulfonic acid DMF N,N-Dimethylformamide HMPA hexamethylphosphoramide HPLC high performance liquid chromatography J coupling constant in hertz Hz h e rtz PCC pridinium chlorochromate NMR nuclear magnetic resonance S.A. specific acitivity in uCi/umol TES N-tris-( hydroxy methyl)-methyl-2- aminoethanesulfonic acid THF tetrahydrofuran TLC thin layer chromatography uCi m ic ro c u rie

xiv PART i INTRODUCTION

Clostridium thermoaceticum is an anaerobic thermophile that can utilize CO 2 as its sole carbon source. In 1942 Fontaine reported that this microorganism ferments hexoses stoichiometrically to give 3 moles of acetic acid per one mole of hexose1. This observation raised an interesting mechanistic question. Was the organism catalyzing a new type of hexose cleavage resulting in three 2-carbon compounds or did the hexose cleavage produce in normal fashion two moles of acetate and two moles of CO 2 and the latter were converted into another mole of acetate? This question was first addressed in studies by Barker and Kamen who found that fermentation of glucose in the presence of 14C 0 2 resulted in 14C label in both the methyl and the carboxyl carbon of acetate2. In 1952 Wood showed that fermentations conducted in the presence of 13CC>2 produced acetate which according to mass spectral analysis contained molecules of mass 62, i.e., two mass units greater than normal acetate3.

Therefore, 13C-label was incorporated into both carbons within the same molecule. From the data it appeared that in the fermentation of hexoses by this organism, one mole of acetate could be formed by the reduction of two moles of CO 2 .

1 2

Further investigations pointed to the involvement of corrinoids, specifically a methyl cobalamin intermediate (CH 3-B 12), in the synthesis of acetate from CO 24-5. The following scheme was proposed in 1966 to explain acetate synthesis from CO 2 (Figure 1).

At the time it was known that in methionine biosynthesis, methyltetrahydrofolate acted as a methyl carrier by transferring its methyl group to a corrinoid enzyme which then transferred the methyl group to homocysteine. It was proposed that a similar type of transformation was taking place in acetate biosynthesis in

C. thermoaceticum. In this scheme, one mole of CO 2 is reduced to methyltetrahydrofolate and the methyl group is transferred to the cobalt of a corrinoid enzyme. Another mole of CO 2 then combines with the methyl group to yield acetate. In support of this hypothesis, Wood and coworkers demonstrated in 1971 by 14C 0 2 labeling studies that the methyl group of methyltetrahydrofolate was indeed labeled with 14C 6. They also demonstrated that 14CH 3- tetrahydrofolate was a precursor of the methyl group in acetate7.

In more recent work the conversion of one mole of CO 2 to methyltetrahydrofolate was substantiated. CO 2 is reduced to formate which then combines with tetrahydrofolate to form 5- formyltetrahydrofolate and then 5,10-methenyltetrahydrofolate. A series of reductions leads to 5,10-methylenetetrahydrofolate and finally to 5-methyltetrahydrofolate. This species is the actual precursor of the methyl group in acetate biosynthesis8-9. The 3

2 CH3 - COCOOH ------— 2 CH3 - COOH

HCOOH ------2 C 0 2

THF

Formyl-THF

CH ECo

ECo

Figure 1. Scheme proposed in 1966 for the synthesis of acetate in Clostridium thermoaceticum. 4 enzymes responsible for this transformation have been purified and characterized8*9.

The other mole of CO 2 is reduced to CO and then combined with the methyl group in the presence of coenzyme A (CoASH) to yield acetyl coenzyme A10. The enzyme responsible for this reduction is

CO dehydrogenase (CODH), which was first discovered by Thauer and coworkers11. This enzyme was purified by Ragsdale £ la l, and found to contain nickel12*13. The role of this enzyme was originally thought to be limited to the reduction of CO 2 to CO, since it was assumed that the acetate was formed by a direct carbonylation of the CH 3-B 12 species.

Recent work by Wood and coworkers, however, suggests that

CODH not only reduces CO 2 to CO, but is also responsible for the assembly of acetate. The evidence that supports this hypothesis comes from two labeling studies: The first study showed that purified CODH catalyzes the exchange of 3H-CoASH with the CoASH in acetyl coenzyme A14. The second study demonstrated that the purified enzyme also catalyzes the exchange of [ 1- 14C ]ace tyl coenzyme A with CO15. These exchange reactions suggest that the enzyme binds acetyl coenzyme A and can reversibly cleave it into the methyl, carbonyl and CoASH fragments.

EPR studies of purified CODH in the presence of CO points to involvement of a Ni 3+-iron-carbon species. Furthermore, when the enzyme was reacted with acetyl coenzyme A or CoASH, the EPR spectrum showed a signal which was representative of a carbon- 5 metal species14-16. These data are suggestive of a carbonylation

(carbonyl insertion reaction or migratory insertion) reaction occurring on a nickel or iron site. The involment of nickel as the site of carbonylation in the enzymatic reaction has precedent in some organonickel chemistry reported by Kohara s ia i17- They exposed a methyl-nickel compound to one mole of CO and were able to crystallize the corresponding acetyl-nickel compound. The carbonylation product was then reacted with p-cyanophenol resulting in acetyl transfer to afford the activated acetate ester.

This type of reaction (carbonylation reaction with metals) has ample precedent in organometallic chemistry, however, it has never been demonstrated in biological systems.

Wood's proposed mechanism of acetate synthesis involves two sequential transfers of the methyl group, the first from 5- methyltetrahydrofolate to the cobalt in B 12, and the second from

CH 3-B 12 to CODH. The methyl-metal intermediate undergoes a carbonylation reaction and the resulting acetyl-metal species is subject to a displacement reaction with coenzyme A to yield acetyl coenzyme A. The formation of the activated acetate ester in the work by Kohara supports this mechanism of acetyl coenzyme A formation proposed by Wood.

The stereochemical analysis of chemical transformations offers valuable information about the reaction mechanism taking place.

With this information it is possible to distinguish a reaction mechanism which is stereospecific, such as an Sn 2 mechanism from 6 one which leads to racemization, an Sn1 mechanism. Once the steric course of particular types of reactions or reaction steps has been established, the determination of the overall stereochemical outcome of a reaction sequence can serve to reveal the number and nature of the individual reaction steps in the sequence. The stereochemical course of methyl group transfer reactions in both biological and chemical reactions can be investigated using the chiral methyl group methodology. This method involves the use of a methyl group which is chiral by virtue of isotopic substitution with tritium and deuterium. The configurational analysis of chiral methyl groups, subsequently referred to as the chirality analysis, involves analyzing [ 2H-| ,3 H]acetic acid by the method of Arigoni and coworkers and Cornforth e la i.18-19-. Their method consists of conversion of acetic acid into acetyl coenzyme A, followed by a condensation with glyoxylate which is catalyzed by malate synthase.

The malate is isolated and then equilibrated with fumarase. Based on the primary kinetic deuterium isotope effect of the malate synthase reaction, the percent tritium retention in the fumarase reaction, termed the F value, indicates the configuration and optical purity of the methyl group in acetic acid. An optically pure (R)- methyl group has an F value of 79, and an optically pure (S)-methyl has an F value of 21. For a more complete discussion of the chirality analysis see the Appendix.

Single displacements of methyl groups in biological methyl transfer reactions have been shown to proceed with inversion of 7 configuration20-21. Chemical (non-enzymatic) carbonylation reactions proceed with retention of the alkyl group configuration.

This has been demonstrated with organoiron and organopalladium compounds22-27.

In Wood's mechanism of acetate synthesis, the two consecutive methyl transfers and the subsequent carbonylation reaction should lead to an overall retention of configuration, assuming that the methyl transfer and carbonylation reactions proceed with a stereochemistry as previously discussed (Figure 2).

On the other hand, a process which involves a single methyl transfer to B i2 followed by a carbonylation reaction on the CH 3-B 12 species, would result in net inversion of configuration (Figure 2). The stereochemical analysis of the methyltetrahydrofolate to acetate conversion should allow us to distinguish between the two proposed mechanisms.

In the present study, the following reaction mechanisms were investigated: 1. The stereospecificty of the carbonylation reaction of CODH in C. thermoaceticum was studied using (R)- and (S)-

[2H i 3 H]acetyl coenzyme A. 2. The steric course of acetate biosynthesis was examined using (methyl-R)- and (methyl-S)-

[2H 1,3 H]methyltetrahydrofolate and a cell-free extract from C. thermoaceticum. Both of these reactions afford acetyl-coenzyme A or acetate which can be analyzed directly in the chirality analysis.

The first study was carried out in collaboration with Professor

Walsh and coworkers. The second study involved a collaboration 8

5-CH,-H„folate

inversion

▼ \ / CH, V C B1S enzyme

retention inversion

\ / CH,-CO^fo^ CHj-X B u enzyme CO-dehydrogenase

retention 1 CH,-CO-X CO-dehydrogenase * CH.-COOH CH.-COOH

net inversion net retention of configuration of configuration

Figure 2. Stereochemical predictions of acetate synthesis in Clostridium thermoaceticum from chiral 5-methyltetrahydrofolate 9 with Professor Benkovic and coworkers and Professor Simon and cow orkers. RESULTS AND DISCUSSION

Preparation of (R)- and (S)-[2H i,3H]Acetyl Coenzyme A

(R)- and (S)-[2Hi , 3H]acetyl coenzyme A were prepared enzymatically from (R)- and (S)-[ 2H i, 3H]acetate, ATP, and Co ASH using acetate kinase and phosphotransacetylase. This procedure was adapted from the procedure developed by Cornforth and coworkers to convert acetate into acetyl coenzyme A for the chirality analysis 1 9

(Figure 3).

CH3CO2H + ATP — » CH3CO2PO3H2 CH3CO2PO3H2 + CoASH ■" '» CH3COSC0A + H3PO4 Figure 3

The synthesis was conducted by coupling the two enzymatic reactions together and allowing the conversion to proceed at room temperature in 100 mM carbonate buffer (pH 7.4). The formation of acetyl coenzyme A was monitored by HPLC (high performance liquid chromatography) and the crude product was purified by HPLC using a

Waters C-18 reverse phase semipreparative column (7.8 mm x 300 mm), with 12.5% methanol in 50 mM potassium phosphate buffer pH

10 11

6 as the mobile phase. At a flow rate of 4 to 5 mL/min, the retention times were approximately 8 to 10 minutes for coenzyme A and 20-25 minutes for acetyl coenzyme A. Unreacted acetate eluted with the void volume of the column.

Monitoring the reaction at several different pH values showed that a pH of 7.4 was optimal for the production of acetyl coenzyme A and minimized the chemical hydrolysis of acetyl coenzyme A to acetate. A reaction time of 80 minutes was found to be optimal for the production of acetyl coenzyme A. Longer reaction times resulted in the hydrolysis of acetyl coenzyme A back to acetate. The best yield that could be obtained starting from (R)-t 2H i(3 H]acetate, was a radiochemical yield of 17% for (R)-[ 2H i, 3H]acetyl coenzyme A, and the radiochemical yield of (S)-[ 2H-| ,3 H]acetyl coenzyme A was 10% starting from (S)-[2H i, 3 H ]a ce ta te .

The optical purity of the acetyl coenzyme A could be analyzed directly by the chirality analysis. The starting (R)- and (S)-

[2H i 3 H]acetate had been previously synthesized in our group and found to be 89% ee R (F = 76.1) and 93% ee S (F = 23.0)23. The resulting (R)- and (S)-[ 2H-| 3 H]acetyl coenzyme A analyzed as 84% ee

R (F = 74.5) and 79% ee S (F = 27.2).

The Carbonylation Reaction with Purified Carbon Monoxide Dehydrogenase (CODH) from Clostridium thermoaceticum.

The stereospecificity of the carbonylation reaction was investigated by incubating purified CODH from C. thermoaceticum with (R)- and (S)-[ 2H-| 3 H]acetyl coenzyme A24. The enzyme purification and the subsequent incubations were performed in

Professor Walsh's laboratories at M.I.T. To insure that the exchange reaction was indeed taking place, [1- 14C]acetyl coenzyme A was mixed with the chiral acetyl coenzyme A and the 3H/14C ratios were monitored. The samples incubated with the pure enzyme for 45 minutes in the presence of unlabeled CO showed that approximately

70-75% of the 14C-label was exchanged. Therefore, 70-75% of the chiral acetyl coenzyme A molecules had undergone reversible cleavage between the methyl and carbonyl carbons.

Using the chirality analysis, the chiral purity of the methyl group of acetyl coenzyme A after the exchange reaction analyzed. The (R)-

[2 H 1,3H]acetyl coenzyme A that was reisolated and purified from the reaction mixture from two separate experiments gave F values of

73.4 and of 73.9 corresponding to 81% ee and 82% ee R after an exchange of 70% and 72.5% of the 14C-label. The (S)-[ 2H-| 3 H]acetyl coenzyme A reisolated from two experiments gave F values of 26.7 and 27.4 corresponding to 80% and 78% ee S after 75.5% and 77.9% exchange respectively24. These data demonstrate that the carbonylation reaction of CODH in C. thermoaceticum occurs with no racemization. 1 3

Stereochemical Analysis of Acetic Acid Formation from

(Methyl-R)- and (Methyl-S)-[2 H i (3 H]methyltetrahydrofolate by a Cell-Free Extract of Clostridium thermoaceticum.

To investigate the mechanism proposed by Wood and coworkers for the biosynthesis of acetate in C. thermoaceticum. (methyl-R)- and (methyl-S)-[ 2 H i, 3 H]methyltetrahydrofolate which had been

previously synthesized by Benkovic and coworkers 25 were incubated anaerobically with a cell-free extract of C. thermoaceticum. The preparation of the cell-free extract and the incubations were carried out in the laboratories of Professor Simon.

The chiral methyltetrahydrofolate had been previously degraded by diazotization and KMnC >4 oxidation to form methylamine which, by standard procedures, was converted stereospecifically to acetic acid for the chirality analysis (Figure 4). The chiral methyltetrahydrofolate samples were found to be 44% ee R and 37% ee S, respectively25.

The acetic acid obtained from two incubations with (R)- methyltetrahydrofolate gave an F value of 59.5 and 58.3 which corresponds to 33% ee R and 29% ee R configuration of the methyl group respectively. The sample of acetic acid from (S)- methyltetrahydrofolate gave an F value of 37.2 which corresponds to 44% ee S configuration of the methyl group26. Therefore, the synthesis of acetate from the 5-methyltetrahydrofolate occurs with overall retention of configuration (Figure 5). H H H © NaN0 2 I ® NaOH/TsCl J. C. © 0 ► . 0 . J s d 7 snh,ci ©NoH/TsCI d*Vnn( WR ® ^ Ts

KCN/HMPT

H H I © H 20 2 /NaOH I VCvCO^H © NaNC^/F^SC^ ‘C = N D

Figure 4. Degradation of chiral 5-methyltetrahydrofolate to acetic acid for configurational analysis. H I I " y C — IH4 f o l a t e ] inversi0-n Hm" c [B12ENZYME] , I D 11inversion 44 % e.e. R

37 % e.e. S

...... f r [CODEHYDROGENASE] H - C0 _ [CODEHYDROGENASE] t '

H 1 » y c — COOH

31 % e.e. R

44 % e.e. S

Figure 5. Stereochemical outcome from the enzymatic conversion of chiral 5 -methyltetrahydrofolate to acetate in Clostridium thermoaceticum CONCLUSION

The methyl group of methyltetrahydrofolate is transformed into the methyl group of acetate by C. thermoaceticum with overall retention of configuration. The results argue against a mechanism of acetate biosynthesis in which acetyl coenzyme A is formed directly on the B 12 enzyme and are consistent with the mechanism proposed by Wood and collaborators. This mechanism involves a transfer of the methyl group from methyltetrahydrofolate to the cobalt on the B 12 enzyme, followed by a transfer to CODH where a carbonylation reaction takes place to afford acetyl coenzyme A26.

Overall retention of configuration, will be obtained if two methyl transfers occur, each with inversion of configuration, followed by a carbonylation reaction with retention of configuration.

The results of the CODH exchange study with (R)- and (S)-

[2H i ,3 H]acetyl coenzyme A indicate that the carbonylation reaction is stereospecific and proceeds with little or no racemization24. The following mechanism shown in Figure 6 is proposed for the carboylation reaction. Since there is literature precedent for nickel being capable of undergoing such a carbonylation reaction, nickel is postulated as the site where the carbonylation occurs. The cleavage reaction involves the formation of an acetyl-nickel species formed by a nucleophilic substitution reaction on acetyl coenzyme A. The

1 6 0 K ^ C ^ i^ S C o A ,SCoA ..SCO A * C0 SCoA Enz O Enz* — —» + v < r En< Enz-N i ' N i ^ C H . N i-C H , N i-C H S I d ♦CO m

Enz-Ni .SCoA .SCoA + Enz* 0 Enz 0 N N i ^ C H . kNi-CH- H 3 c ^ SCoA i CO

Figure 6 . Postulated mechanism for the carbonylation reaction of acetyl coenzyme A with CODH from Clostridium thermoaceticum. carbonyl group could then dissociate to form a methyl-nickel intermediate which is then capable of undergoing a recarbonylation reaction with another molecule of carbon monoxide. This constitutes the exchange reaction between the carbonyl carbon of acetyl coenzyme A and carbon monoxide catalyzed by CODH. The insertion of the carbon monoxide back into the nickel-methyl bond represents the normal carbonylation reaction resulting in the synthesis of acetyl coenzyme A. The observed net retention of the methyl group configuration of acetyl coenzyme A in the carbonyl exchange reaction does not reveal the steric course of the carbonylation insertion reaction, but it does show that decarbonylation and the carbonylation reaction proceed with the same stereochemistry, i.e., both with retention or both with inversion. This obviously suggests that they both proceed by the same mechanism, i.e., the exchange reaction is reversible rather than a cyclic process. Based on the chemical precedent it seems likely that both the decarbonylation and the carbonylation reaction proceed with retention of the methyl group configuration. The results also show that the carbonylation reaction proceeds completely stereospecifically, which again had precedent in chemical models.

Whitesides and coworkers have shown that carbonyl insertion reactions in organo-iron compounds proceed with greater than 95% retention of configuration at the carbon. In their study the p- 1 9 cyclopentadienyldicarbonyl-iron erythro- and threo-3,3- dimethylbutyl- 1 ,2 - 2 H 2 was treated with triphenylphosphine to afford the corresponding threo and erythro carbonyl insertion products (Figure 7)27. The stereochemistry of the starting compounds and the carbonyl insertion products after isolation were confirmed by conformational analysis as determined by 1H NMR. This technique allowed for direct observation of the iron-acyl intermediate and also lends support to the existence of an acyl- metal intermediate in the enzymatic reaction. The carbonylation reaction is novel in biological systems. Further studies are required to determine if a nickel-methyl or a iron-methyl species is involved.

c ( C i u C(CH3);

PPh

Fe(CO)aCp < T Fe(PPhj)Cp erythro-2 CO

Figure 7 EXPERIMENTAL

Materials and Methods

Materials: All chemicals and solvents used were of the highest grade commercially available. The organic chemicals were purchased from Aldrich Chemical Co. and the biochemicals from

Sigma Chemical Co.

Analytical Methods: Radioactive samples were counted using

Purification using high performance liquid chromatography (HPLC) was done with a Waters model 590 programmable solvent delivery system, a Hitachi L-3000 multi-channel photodiode array detector, and a Hitachi D-2000 chromato-integrator, using an Waters C-18 reverse phase column (7.8 mm x 300 mm, flow rate 4-5 mL/min, isocratic system 12.5% methanol in 50 mM phosphate buffer, pH 6.0).

2 0 21

Synthesis of (R)- and (S)-[2 H i, 3 H]Acetyl Coenzyme A.

The following reagents were placed in a 25 ml_ flask at 4 °C :

100 mM carbonate buffer (pH 7.4), 8.3 mM ATP, 1 mM EDTA, 5 mM

M gC2 l , 6 umol (R)-[2Hi , 3 H]acetate (41 uCi, S.A. 11.36 uCi/umol, 89% ee R)23, nonlabeled carrier acetate to give a total of 10 umol, 185 III of phosphotransacetylase, and 35 IU of acetate kinase. The mixture was incubated at room temperature and the reaction was initiated by the addition of .75 mM coenzyme A, giving a total reaction volume of 10 mL19. After 80 min at room temperature the reaction was stopped by the addition of .450 mL of 2N HCI. The mixture was lyophilized and the residue was dissolved in 1 mL of 50 mM potassium phosphate buffer (pH 6.0). The crude product was purified by HPLC using a C-18 reverse phase semi-preparative column

(mobile phase 12.5% methanol in 50 mM potassium phosphate buffer, pH 6.0). At a flow rate of 4-5 mL/min, the the retention times were

8-10 minutes for coenzyme A and 20-25 minutes for acetyl coenzyme A. Unreacted acetate eluted with the void volume and could be repurified by steam distillation at an acidic and basic pH.

The fractions corresponding to the acetyl coenzyme A were pooled and lyophilized. This afforded (R)-[ 2 H if3 H]acetyl coenzyme A in 17% radiochemical yields (7 uCi, S.A. 4.3 uCi/umol).

In an identical manner, 1.9 umol of (S)-[ 2 H i, 3 H]acetate (56 uCi,

S.A. 6.42 uCi/umol, 93 ee S ) 23 was converted into (S)-[ 2 H-| ,3 H ]acetyl coenzyme A in a radiochemical yield of 10% (5.6 uCi, S.A. 5.3 22 uCi/umol). The acetate samples were then further prepared by a series of basic and acidic distillations for the chirality analysis.

Carbonylation Exchange with Purified CO Dehydrogenase

(CODH) from Clostridium thermoaceticum with (R)- and (S)-

£2Hi ,3 H]Acetyl Coenzyme A.

The purification of the CODH from C. thermoaceticum and the enzyme incubations with (R)- and (S)-acetyl coenzyme A were performed in the laboratories of Professor Walsh at M.l.T in the following manner: The (R)-[ 2 H i, 3 H]acetyl coenzyme A was mixed with [1-1 4 C]acetyl coenzyme A to give a total of 200 nmol of acetate with a 3 H/14C ratio of approximately 10. This material was incubated anaerobically at 55 °C with .4 mg of purified CODH in the presence of 150mM potassium phosphate buffer (pH of 6.0) and .2 mM methyl viologen in a total reaction volume of 1 mL14. Aliquots (.100 mL) of the reaction mixture were taken and counted to establish the

3h/14C ratio. The reaction was stopped when 75% of the 1 4 C -label had exchanged out of the acetyl coenzyme A, which was after approximately 40 min. The acetyl coenzyme A was then purified from the reaction mixture by HPLC using a C-18 reverse phase column (mobile phase, 12.5% methanol in 50 mM sodium phosphate buffer, pH 5.5)24. 23

Conversion of (Methyl-R)- and (Methyl-S)-

[ 2 H i, 3 H]methyltetrahydrofolate with the Cell-Free Extract from Clostridium thermoaceticum.

The preparation of the cell-free extract and enzyme incubations were performed in the laboratories of Professor Simon. (Methyl-R)- and (methyl-S)-[2Hi ( 3 H]methyltetrahydrofolate (1.0 x 10 6 dpm and

1.15 x 105 dpm respectively in less than .1 umol) which had been previously synthesized in Professfor Benkovic laboratories25, were incubated anaerobically with the cell-free extract of C . thermoaceticum (16.6 mg protein/mL) under an CO atmosphere for

30 min at 55 °C. The following reagents were added; .1 mM CoASH, 4 mM ATP, 1.4 mM Fe2+, 16 mM dithiothreitol, and 90 mM potassium phosphate buffer (pH 6 ) in a total reaction volume of .550 mL 28.

The reaction was terminated by the addition of .5 mL of 2.2 N HCICU, followed by .25 mL of 1 N NaOH. Following the denaturation, 10 umol of non-labeled carrier acetic acid was added, and the reaction mixture was passed through a cation exchange column (Dowex 50 W,

50-100 mesh, H+, .8 x 8 cm, water), and further purified by steam distillations. This resulted in 17-24% conversion of the chiral methyltetrahydrofolate to acetate26. PART II INTRODUCTION

Methanogenic bacteria are a diverse group of anaerobic organisms which can metabolize various one carbon compounds to produce methane. They have been found to contain a unique set of cofactors.

The synthesis of methane has been reported to be coupled to energy formation. Most methanogenic bacteria grow on CO 2 and H 2 , although some can utilize formate, carbon monoxide, acetate, methylamine, and methanol to produce methane (Figure 8 ).

C 0 2 + 4 H2 ------> CH 4 + 2 H 2 0 4 HCOOH ------> CH4 + 2 H20 + 3 C 02

4 CO + 2 H20 ------> CH4 + 3 C 0 2

CH 3 COOH ------> CH4 + c o 2

4 CH3 OH ------> 3 CH4 + C 0 2 + 2 H20

4 CH3 NH 2 + 2 H20 ------> 3 CH4 + C 0 2 + 4 NH3

2 (CH3)NH + 2 H20 ------> 3 CH4 + 3 C 0 2 + NH3

4 (CH3)3N + 6 H20 ------> 9 CH4 + 3 C 0 2 + 4 NH3

Figure 8

24 25

Early work by Barker and coworkers showed that not only was

CO 2 reduced to methane, but the methyl group of acetate and methanol were also converted to methane. By using trideuterated acetate or methanol, they showed that the methyl group was transferred intact to methane. Barker postulated that the conversion to methane involved the stepwise reduction of a C 1 unit and in 1956, he proposed the following scheme to account for methanogenesis from methanol, acetate, and H 2/CO 2 (Figure 9). In this scheme, an unknown methyl carrier X is proposed as the direct precursor of methane during methanogenesis. The compound X-CH 3 would be the common intermediate in the reduction of CO 2 , acetate, and methanol to methane29.

CH3OH +XH

XH + CO, h2o

+2H +2H +2H +2H XCOOH XCHO XCH2OH XCH, CH4 + XH -h2o -H 20

-2H -CO,

CH3COOH + XH

Figure 9 26

In 1963, Blaylock and Stadtman reported that methylcobalamin

(CH 3 - B i 2 )was a substrate for the formation of methane by a cell- free extract of Methanosarcina barkeri 3 Q>31. They were the first to suggest that methylcobalamin was the direct precursor of methane.

They also reported that 1 4 CH 3 - B i 2 could be isolated from cell-free extracts which had been incubated with 14CH 3 0 H. From these observations, it was proposed that the conversion of methanol to methane proceeds by a transfer of the methyl group from methanol to the cobalt of vitamin B 12, followed by a direct reduction to methane. They also reported the isolation of an enzyme system responsible for this enzymatic conversion which could be separated into four components; a corrinoid protein, an unidentified protein, a ferredoxin compound, and a heat stable acidic cofactor32.

The notion that CH 3 -B 12 was the intermediate (X-CH 3 ) w as questioned when, in 1971, Wolfe and McBribe discovered a heat stable acidic compound which could also act as a methyl carrier.

They called this new methyl carrier coenzyme M 33 and found it to be present in all methanogenic bacteria. In 1974, Taylor and Wolfe determined the structure of this cofactor, which is unique to methanogens, to be 2-mercaptoethanesulfonic acid (Coenzyme M )34

(Figure 10).

HSCH2 CH2 SO3 H Figure 10 27

Using 14CH 3 -B -|2 as substrate, they found that coenzyme M was methylated and then reduced to form 14C-methane. Therefore, the methylated form of this cofactor, 2 -(methylthio)ethanesulfonic acid

(methyl coenzyme M), is the direct precursor of methane. To date no other compound has been found which acts as the direct precursor to methane.

Methanogenesis from CO2 and H2

Until recently, the steps involved in the reduction of CO 2 to methane and the identity and role of the C 1 carriers were not well understood. This pathway involves the participation of three cofactors, all of which are unique to methanogens; a carbon dioxide reduction factor (CDR, later renamed methanofuran )3 5 *36 methanopterin ,3 7 -38 and coenzyme M. The structure of methanofuran was elucidated by Leigh and Wolfe and was found to be that of a 2,4- disubstituted furan 39 (Figure 11).

The aminomethyl group at C -2 of methanofuran is thought to participate in the reduction of CO 2 to methane. The CO 2 is bound to the amino group and then reduced to the formaldehyde oxidation level. This N-formyl-methanofuran is the first intermediate in the reduction of CO 2 to methane40.

The active form of methanopterin is tetrahydromethanopterin

This form was found to be a carrier of C-1 moieties at the methenyl, methylene, and methyl oxidation levels. The chemistry of COOH O COOH O COOH O __ CHzNHj HOC)CC^2CHCHCH2aH2CNHCT(>l2(>^l!N I^(>l2(> l2^ H ( > i2CH2- ^ OCH2 q

COOH

METHANOFURAN

NH OH

H i HO i-ui COOH HO

OH OH COjH

TETRAHYDROMETHANOPTERIN

Figure 11. Strucutre of methanofuran and tetrahydromethanopterin 29 tetrahydromethanopterin is similar to that seen with tetrahydrofolate 3 8 -41 (Figure 11).

The N-formyl group of methanofuran is transferred to tetrahydromethanopterin where it is further reduced to the methyl group oxidation level. The methyl group is then transferred to coenzyme M. To complete the sequence, methyl coenzyme M is reduced to methane by the methyl coenzyme M reductase system.

The reduction of CO 2 to methane does not require ATP, however, it does require the reduction of methyl coenzyme M. The stimulation of methane production from CO 2 by methyl coenzyme M was first reported by R. P. Gunsalus and hence it is referred to as the "RPG effect." He found that if methyl coenzyme M was added to a cell- free extract in the presence of hydrogen and CO 2 , then the rate of methane formation increased. In fact, each mole of methyl coenzyme M resulted in 11 mols of CO 2 being activated and reduced to methane. From these results it can be deduced that the terminal reaction in methane formation is coupled to the activation of CO 2 . It is thought that perhaps the energy produced by methanogenesis is conserved and used for the activation of CO 2 (reaction step 7 to 1), suggesting, therefore, that methanogenesis from H 2/CO 2 is a cyclic process (Figure 1 2 ). 30

HCO-MFR

h>mpt h c o - h4 m p t HS-CoM <3> IH* 0 mo-(=CH-)H4MPT

® / z * CH?=HaMPT ® > 2e 3 ' ^ M P T

Figure 1 2 Methanogenesis from CO 2 . 3 1

Methanogenesis from Methanol, Methyiamine, and Acetate

The organism, M. barkeri. is one of the most versatile species of methanogens; it is capable of utilizing methanol, methyiamine, and acetate as well as H 2/CO 2 to produce methane and cellular carbon compounds. Isotopic labeling studies with trideuterated methanol, methyiamine, and acetate have shown incorporation of the deuterium into methane. Early evidence indicated that vitamin B 12 w a s involved in the conversion of methanol to methane, however, with the discovery of coenzyme M as the direct precursor to methane, the role of vitamin B 12 in methanogenesis became uncertain. Blaylock had originally reported an enzyme system that was responsible for the conversion of 1 4 CH 3 - B i2 to methane. Taylor and Wolfe also reported the purification of a CH 3 - B i 2 :methyl coenzyme M methyltransferase, however, they did not establish its significance34.

Working with M. barkeri. Vogels and coworkers demonstrated that there are two methyltransferases involved in the conversion of methanol; a methanol:5-hydroxybenzimidazolecobalamide methyltransferase (MT-|), and methyl cobalamin:coenzyme M methyltransferase (MT 2 )42. MT1 is an oxygen sensitive corrinoid enzyme (122,000 mol. wt.) and this is probably the enzyme Blaylock originally isolated (Blaylock, 1968)32. This enzyme requires preactivation with ATP, Mg2+, and H 2 to reduce the cobalt in the enzyme from Co2+ to Co 1 + , In whole cells, this process is performed 32

by a hydrogenase, F 420, and a ferredoxin. The reduced cobalt species

(Co1+) can then be alkylated by the methyl group from methanol.

MT2 (43,000 mol wt) is not oxygen sensitive and is the enzyme responsible for the transfer of the methyl group from free or bound methylated corrinoidsto coenzyme M. This was the enzyme isolated by Wolfe and coworkers34. This enzyme has been found in several species of methanogens and its biological role is probably not restricted to this particular function. In fact, it has been speculated that this enzyme can also act to transfer methyl groups from vitamin B 12 to various intermediates involved in acetate synthesis. The acetate is then used in the synthesis of cellular carbon compounds.

Based on these findings, the conversion of methanol to methyl coenzyme M is expected to involve two sequential transfers of the methyl group; first to a cobalt (Co1+) in a cobalamin and the second to the sulfur in coenzyme M (Figure 13).

CH,

N N Enz--A Co \ CH,OH c h 3s c h 2c h 2s o 3 N —j------N

HBI

Figure 13 Despite a vast amount of research on the methanol to methane conversion, the mechanism of methanol activation is still unanswered. It is likely that methanol is converted to an activated species or protonated prior to transfer to the cobalamin, since the direct displacement on methanol would necessitate having hydroxide as the leaving group.

One would expect to see the same transformations occurring when methyiamine is the substrate in the enzymatic reactions.

When trimethylamine was used as the substrate, a methylated coenzyme M intermediate was seen43. However, there is no other direct evidence to confirm the involvement of a cobalamin in trimethylamine metabolism.

Coenzyme M has been reported as an intermediate in methanogenesis from acetate44-45. The methyl coenzyme M reductase system has also been reported to be involved in the conversion of the methyl group of acetate to methane. The participation of corrinoids has been postulated since corrinoid inhibitors have been shown to inhibit methane production from acetate, but not from H 2/CO 2 . The cleavage of acetate in methanogenesis is thought to involve the enzyme carbon monoxide dehydrogenase (CODH), which is also the enzyme responsible for the assembly of the methyl and the carbonyl carbon in acetate synthesis.

It has been reported that M. barkeri grown on acetate has levels of

CODH which are five times higher than those in cells grown on 34

H 2/CO 2 . Removal of this enzyme prevented acetate cleavage in the cell free extracts. CODH has also been shown to catalyze an exchange reaction between the carbonyl carbon of acetate and CO 246.

It has been postulated by Thauer that there are, in fact, two corrinoid enzymes involved in methanogenesis from acetate47.

Inhibition studies with 1-iodopropane (a known inhibitor of corrinoids) have shown two distinct levels of inhibition, one at low

1-iodopropane concentration (< 10 uM), and one at higher 1 - iodopropane concentrations (100 uM).

In their proposed pathway acetyl phosphate and acetyl coenzyme

A are the actual substrates for CODH (Figure 14). Their proposal is further based on the finding that acetate kinase and phosphotransacetylase were induced when the organism was grown on acetate. They found that acetate formation, methanogenesis, and the CO 2 exchange reaction were all inhibited by the high concentration of 1 -iodopropane whereas the low concentration of 1 - iodopropane inhibited only methanogenesis. Therefore, they speculated that the same corrinoid containing enzyme is involved in acetate synthesis, the exchange reaction, and methanogenisis and this enzyme transfers the methyl group from CODH to the cobalt on

B 1 2 . Next the methyl group (CH 3 -B-i2 )is then transferred by another vitamin B 12 containing enzyme to a cobalt on another B 1 2 . Finally, the methyl is transferred to the sulfur on coenzyme M to give methyl coenzyme M. Based on this proposal, the conversion of acetate to methyl coenzyme M would be expected to proceed by three sequential methyl group transfers.

n ADP ------^— ► n ATP

CH3 X —♦C H 3 [Co| E —♦ CH3Cd M CH4

CH^COCr^^C^CO-©^—►CH3CO-CoA*-/ >.— 2[H]------

1ATP 1ADP CO ~ Y «------^

m AOP ------► m ATP Figure 14

In the present study, the stereochemical course of the methyl group transfers from methanol, methylamine, and acetate to coenzyme M were explored using chiral methyl group methodology

(see Appendix).

In order to study the stereochemical course of the methyltransferase reaction, the following tasks needed to be completed: 1. Cultivation of M barkeri on methanol, methylamine, and acetate. 2. Preparation of an active cell-free extract which will convert these substrates into methyl coenzyme M. 3. Synthesis of chiral substrates and their subsequent degradation to acetate for determination of optical purity. 4. Enzymatic conversion of the chiral substrates into methyl coenzyme M. 5. Degradation of methyl coenzyme M to acetate and chirality analysis. RESULTS AND DISCUSSION

Cultivation of Methanosarcina barkeri on Methanol, Methylamine, and Acetate.

Methanogens are strict anaerobes, therefore, their cultivation requires the use of special equipment and techniques. Similar techniques are also required to complete the enzymology portion of this work. For this part of the study, the work was carried out in the laboratories of Professor John Reeve at The Ohio State

University. Professor Reeve and coworkers instructed me in the special anaerobic techniques required to perform the following experiments. The stringent requirement for anaerobic conditions must be adhered to in order to work with methanogens. It has been reported that M. barkeri can withstand exposure to oxygen for up to

30 hours, however, all of the activity in the cell-free extract is immediately destroyed upon exposure to oxygen. A great deal of time was spent learning the techniques necessary to grow cells and to prepare an active cell-free extract.

M. barkeri strains 227 and MS were both available for use in this study from the culture collection of Professor John Reeve and were transferred anaerobically from a frozen glycerol stock solution to

ER medium (H 2/CO 2 , 60:40, 40 psi)48. Both strains of M. barkeri were cultivated initially to determine which strain would show

36 37 better growth under the final fermentation conditions. The cultures were then incubated at 37 °C for 7 days with gentle shaking. When the culture reached turbidity, several transfers (1 mL) were made into new medium until good growth was established. At this time, the growing times were 2.5 days and the organism was now ready for adaptation to growth on methanol, methylamine, and acetate.

An aliquot of the cell suspensions (1 mL) was transferred into ER medium which had been prepared under a N 2/CO 2 atmosphere (60:40,

40 psi) and which contained a 100 mM concentration of either methanol, methylamine. or acetate49. The onset of turbidity in the methanol and methylamine cultures was approximately 4-5 days, however, for the acetate grown cultures this took substantially longer, occurring around 30 days. Other groups had also experienced similar problems in establishing growth on acetate, so this was not unexpected. To insure that the organism was producing methane, it was grown in ER medium containing 100 mM concentrations of the trideuterated substrates. Aliquots were removed from the head space with a gas tight syringe and the presence of methane was detected using GC/MS. This experiment showed the production of

CD 3 H from all three carbon sources44-50.

After good growth was established on the desired carbon sources, the production was scaled up by innoculating a 20 mL aliquot of the cell suspension into a sealed one liter bottle with 300 mL of ER medium under an N 2/CO 2 atmosphere (60:40, 15 psi). These bottles 38 were used as inoculum for the 24 liter fermentations. It was observed at this point that the Methanosarcina barkeri strain 227 showed better growth on the alternate carbon sources, therefore, the study was completed with that strain. The methanol adapted organism showed the best growth of the three; therefore, it was decided to scale up production of that organism for the cell-free stu d ie s.

The fermentations of the methanol adapted organism were conducted in a 28 liter New Brunswick Microferm fermentor which had been modified for anaerobic cultivation. The medium contained

100 mM methanol and the fermentations were carried out under a positive pressure of N 2 /CO 2 (60:40) at 37 °C with agitation. The culture was allowed to grow until it became yellow and turbid, which usually required 4 days. The cells were then harvested anaerobically to yield 1 gm of cells (wet weight) per liter of medium. The cells were then resuspended in 50 mM TES buffer (pH

7.2) containing 15 mM MgCl 2 and 2 mM dithiothreitol. The resulting cell paste could be stored for up to one year under an atmosphere of

H 2 (40 psi) at - 70 0C. 39

Preparation of an Active Cell-Free Extract from Methanosarcina barkeri.

The cell-free extract was prepared anaerobically by breaking the

cells with a French Pressure cell followed by centrifugation to

remove the cell debris. The cell-free extract contained 12-24

mg/mL of protein as determined by the amido-black assay using

bovine serum albumin as the standard51. The crude enzyme extract

could be stored under an H 2 atmosphere (40 psi) at -70 °C for up to 6

months without loss of activity.

Conversion of CD 3 OH to CD3 H with the Cell-Free Extract of Methanosarcina barkeri

The activity of the cell-free extract was tested by monitoring

the enzymatic conversion of trideuterated methanol to trideuterated

methane. In order to insure the conversion of methanol to methane,

all of the substrate preparations and enzymatic reactions were

performed under strict anaerobic conditions in a sealed Wheaton

bottle. The cell-free extract was activated by preincubation with

7.5 mM ATP and 15 mM MgCl 2 in TES buffer (pH = 7.2) for 15 minutes at 37 °C, followed by addition of 2 mM dithiothreitol, 18 mM coenzyme M and, finally, 10 mM CD 3 OH to initiate the reaction.

These conditions had been previously shown by Vogels and coworkers to result in maximal enzymatic activity42. 40

The reaction mixture was incubated at 37 °C overnight with gentle shaking and an aliquot from the head space was removed for determination of methane by GC/MS

Synthesis of (R)- and (S)-[2H i,3H]Methanol

In order to probe the steric course of methyl coenzyme M formation from methanol, samples of (R)- and (S)-[ 2 H i, 3 H]methanol needed to be synthesized. The originally envisioned strategy for the synthesis of chiral methanol was diazotization of chiral methylamine. The chiral methylamine can be easily prepared from chiral acetate using the Schmidt reaction, which affords methylamine with predominant retention of configuration. The diazotization of methylamine was a literature procedure which had been shown to afford methanol in yields of 6-25%52. If the yields of methanol could be optimized, then this route would allow for a direct synthesis of chiral methanol starting from a readily available su b stra te .

In the diazotization reaction, methylamine is converted into the methyldiazonium ion, which can be trapped by solvent or added nucleophiles by an Sn 2 mechanism, leading to methyl group transfer.

A side reaction proceeds through the intermediacy of diazomethane, which is formed by deprotonation of the methyldiazonium ion. This can now undergo proton exchange with the solvent protons. This proton exchange would lead to racemization in the synthesis of chiral methanol. This exchange reaction, however, was shown to be both pH and buffer dependent, with the least amount of exchange occurring at an acidic pH 53 (Figure 15).

CHON?

CH3 N2+ ch2 dn2 + chd2 n2 +

Jd20 | o20

CH3OD CH2 D00 chd2od CD30D

Figure 15

The diazotization reaction was investigated by reacting chiral methylamine with NaN02 and HCI and it was found that the reaction proceeds with a high degree of inversion of configuration54.

However, as observed previously, the yields were very low (yields were less than 1%), and despite many attempts, the yields could not be improved or made reproducible. For this reason, this route to chiral methanol was abandoned.

The alkylation of water with methylnitrosourea (MNU) carrying a chiral methyl group was investigated as an alternative synthesis of methanol, since it is known that MNU decomposes to the methyldiazonium ion53. MNU was heated in 0.1 M Tris HCI buffer (pH

7.4) for 3 hours at 34 °C. The methanol was lyophilyzed from the reaction mixture and its conversion into acetate followed by chirality analysis showed that the methyl group was approximately

40% racemized55. The most likely explanation for these data is that at pH 7.4, proton abstraction from the methyldiazonium ion is competitive with nucleophilic displacement by water or hydroxide ion, resulting in a partially racemized product. The alkylation of water by dimethylnitrosamine (DMN) containing a chiral methyl group in a metabolic activation system of rat liver microsomes was also investigated previously in our laboratory56. This reaction also resulted in low yields of methanol (1-3.5%) which was approximately 40% racemized. Based on these results, it was apparent that an alternative synthesis which would provide chiral methanol in better yields and with useful optical purity was needed.

As mentioned earlier, the overall result of the diazotization reaction is that the amino group of methylamine is converted into a good leaving group which can be displaced by solvent or by added nucleophiles, resulting in the formation of the critical carbon- nucleophile bond. The same general strategy has been used in our laboratories to convert methylamine into a methyl transfer reagent.

This can be accomplished by replacing the amino hydrogens of methylamine with p-toluenesulfonyl groups, making the amino group a sufficently good leaving group for nucleophilic displacement reactions57-58. This approach had previously been employed with

(R)- and (S)-N-[ 2 H i, 3 H]methyl-N,N-di-p-toluenesulfonimide as the 43

electrophile for the alkylation of the homocysteine anion in HMPA.

This reaction affords methionine in 37-45% yield and was shown to

proceed with inversion of configuration (Figure 16)59.

If a similar displacement could be achieved by an oxygen

nucleophile, this would result in the formation of the critical

carbon-oxygen bond required for the synthesis of chiral methanol.

The displacment of the N-methyl-N.N-di-p-toluenesulfonimide with

an alkoxide anion, for example, would produce the corresponding

ether, but the hydrolysis of this would be expected to be difficult.

The reaction of N-methyl-N,N-di-p-toluenesulfonimide with a carboxylate anion should proceed to form a carboxylic acid ester, with the expected inversion of configuration at the methyl group.

Hydrolysis of this ester should afford the desired chiral methanol.

This approach, using N-methyl-N.N-di-p-nitrobenzenesulfonimide

and potassium 3,5-dinitrobenzoate in DMF, had been shown to afford the desired ester in 50-70% yields60.

Based on this strategy, the alternative synthesis of chiral

methanol shown in Figure 17 was developed. The starting materials for the synthesis are (R)- and (S)- [ 2 H i 3 H]acetic acid, which had

been previously prepared by modifications of a literature p ro c e d u re 23 with high specific activity and chiral purities of 99% ee

R and 95% ee S 61. The chiral acetate was converted into methylamine by a Schmidt reaction (H 3 P 0 4 , NaNa), which occurs with almost exclusive retention of configuration. The two step D H n H n H % I NoN3,H2SO^ X | s TsCI.NoOH \ \

/ C\ > / ° \ * / C \ T COONa T NH2 T NHTs

S-benzyl-L_- COOH TsCI,NoH '''✓£$ homocysteine, DMF / \ / Ts No, HMPA T N I Ts

Figure 16. Synthesis of Methionine carrying a chiral methyl group. 45

H \ H N 3 \

/ S ------/ < \ COOH ' T NH3CI ' T

1) TsCI, NaOH H . D ArCOOK, HMPA

yC- ' V, inversion 2) TsCI, NaH N(Ts)2 * T

Ar = p - METHOXYPHENYL

u T O 1 M LiOH, RT f

€ OH /% Ar

Figure 17. Strategy of the synthesis of chiral methanol 46

conversion of methylamine into (R)- and (S)-N-[ 2 H i, 3 H]methyl-N,N- di-p-toluenesulfonimide ( 2 b and 2 a, respectivly) proceeds in approximately 60% yield. In order to establish the chiral purity of the methyl group after the Schmidt reaction, aliquots of 2a and 2 b were reacted with KCN in HMPA at 90 °C to afford acetonitrile with the expected inversion of configuration at the chiral methyl group.

The acetonitrile was then is converted into acetamide by an alkaline hydrogen peroxide oxidation followed by diazotization with HNO 2 to afford acetic acid. This two step reaction sequence has been shown to minimize proton exchange and racemization62. The acetic acid was purified by a series of basic and acidic distillations prior to the chirality analysis 18-19 (see Appendix).

The displacement reaction with KCN results in an inversion of configuration, therefore, the methyl group of the acetate will have the opposite configuration as the methyl group in the N-methyl-N,N,- di-p-toluenesulfonimide. This analysis showed that the methyl group in 2b was 94% ee R and in 2a it was 82% ee S.

The first attempts at methyl ester formation were made using potassium benzoate in HMPA as the nucleophile and N-methyl-N,N- di-p-toluenesulfonimide as the electrophile. Due to the insolubility of potassium benzoate in most organic solvents, the reaction was carried out in HMPA, with the added benefit that an increase in nucleophilicity is seen in going to this polar, aprotic solvent. A related synthesis of carboxylic acid esters from the corresponding acid and alkyl halides has been reported to proceed in quantitative yields using HMPA as the solvent63. After 6 days at 85 °C, the starting material had completely disappeared, presumably with the accompanying formation of the desired methyl benzoate. Workup of the reaction involves an extensive series of extractions which serve to separate the HMPA from the product, and it was soon discovered that any methyl benzoate formed in the displacement reaction was lost in the workup, due to its volatility. Therefore, a carboxylate anion would have to be chosen which would result in the formation of a less volatile methyl ester.

Methyl 3,5-dinitrobenzoate is an ester which could be easily isolated and purified on the small scale that was needed for the reaction with [ 2 H i, 3 H]N-methyl-N,N-di-p-toluenesulfonimide. As previously mentioned, the formation of this ester had also been demonstrated using the N-methyl-N,N-di-p-toluenesulfonimides in

DMF60. In the hope of avoiding the aforementioned problems associated with the separation of HMPA from the desired product, another reaction solvent was chosen. Based on literature precedent, freshly distilled acetonitrile was used as the solvent and in order to help solubilize the potassium carboxylate anion by formation of the

"naked" anion64-65, 18-crown-6 was added to the reaction. After 24 hours at 85 °C, only a trace of methyl 3,5-dinitrobenzoate was observed and after several weeks at that temperature, the desired product was formed in less than 1% yield as determined by GC/MS. 48

Potassium p-methoxybenzoate was the next carboxylate anion examined. Since it contains an electron donating group para to the carboxyl group, it was expected to be more nucleophilic than either of the previously used carboxylate anions. The displacement reaction was first tried in refluxing acetonitrile with added 18- crow n- 6 , however, after several attempts the yield of product was to low too be useful. The reaction was then attempted using 3 equivalents of potassium p-methoxybenzoate to 1 equivalent of N- methyl-N,N-di-p-toluenesulfonimide in HMPA. At 85 °C, the reaction mixture remained heterogeneous and no product was formed. Upon increasing the reaction temperature to 110 °C, the reaction mixture became homogeneous and after 2.5 days, the GC/MS showed that no N-methyl-N,N-di-p-toluenesulfonimide remained. It was subsequently found that the reaction time could be decreased to

3 hours if the reaction temperature was increased to 160 °C, however, at this temperature decomposition of product was also a serious side reaction. Optimized yields of 3a (52%) and 3b (56%) were obtained by running the reaction at 110 °C for 2.5 days

(Figure 18).

It was very critical to the success of the reaction that dry, freshly distilled HMPA was used and that the reaction was performed under an argon atmosphere in order to keep the HMPA as dry as possible. At lower reaction temperatures, which resulted in longer reaction times, the reaction mixture apparently n DTsCI/NoOH D ^ / - \ ____ u - MeO—(OV-COOK H ^ R “ ‘■ V H .P Q ,. h « 4r 2)Tsa/W°M r hJ r / \jOOH J \ H, T/ V T’ HMRft T 2 T (inverskx T* R=F«78.7 R:94%e.e. " * « • • • S:82%e.«. S:F*22.5 (based on degradation 95%e.e. to ocetic add16)

* 4 ' t T N.N-dimethylanUine |

LiO H _ 3,5-dinitrobenzoyt __ V ^ O H*JUC — - ——— c / \ chloride D OH

S’35% e.e. OMe R:44-58%e.e.

D D D KCN/HMft^ R NoOH/HgOg R HNO ^ R (inversion) J \ J \ J \ __ j CN r CONH2 T COOH

Ts * p - toluenesulfonyl R • F* 60.2 * 35%e.e. S:F»37.2t 36 2, 33.2«44-58%e.i

Figure 18. Synthesis and configurational analysis of chiral methanol 50 absorbed water from the atmosphere and hydrolytic formation of N- methyl-p-toluenesulfonimide was seen. This compound is a poor electrophile and its formation results in a decrease of the overall yield. It has been previously shown that the reaction of N,N- diarylsulfonimides with hydroxide anion results in a sulfur-nitrogen bond cleavage rather than in a carbon-nitrogen bond cleavage60.

This explains the formation of the N-methyl-p-toluenesulfonimide which was seen when using lower reaction temperatures or undistilled HMPA.

The separation of the desired ester from the HMPA required the development of an extensive purification regimen. The reaction mixture was diluted with water and then extracted with ether (5 x

50mL). The ether layer was concentrated and then redissolved in water which was again extracted with ether/hexane (3:1) (5 x 50 mL). Finally, the combined organic extracts were dried and concentrated in vacuo to afford the crude product which was further purified by HPLC.

An HPLC system was first developed using a Waters C-18 reverse phase semi-preparative column (7.8 mm x 300 mm), with

CH 3 CN/H 2O/CH 3 OH, (1:9:10) as solvent. A flow rate of 3 mL/min, gave a retention time of 14.5 minutes for the desired ester and 45 minutes for the N-methyl-N,N,-di-p-toluenesulfonimide. Upon concentration of the eluent, however, no product remained, probably as a result of steam distillation of the desired ester from the aqueous solution. This is a common occurrence with volatile esters.

Therefore, to prevent its loss, the product would have to be extracted out of the aqueous mobile phase with organic solvents, or alternatively, the HPLC purification of the product would have to involve non-aqueous solvent systems. The latter was the better solution. The alternate HPLC system that was developed used an IBM silica semi-preparative column (10 mm x 250 mm), with the mobile phase being hexane/methylene chloride (35:65). At a flow rate of 8 mL/min. the retention times were approximately 8 minutes for the desired ester and 11 to 12 minutes for the N-methyl-N,N,-di-p- toluenesulfonimide. This afforded 3a in 51% radiochemical yield and 3b in 15% radiochemical yield (the lower yield is due to loss in the purification), which were pure as determined by GC/MS.

The hydrolysis of 3a and 3b was performed using 1 M LiOH (1 mL) with vigorous stirring at room temperature for 60 hours. The ester was not soluble in the aqeous medium, however, as the hydrolysis proceeded, the reaction mixture became homogenous. After 60 hours, the methanol was purified from the reaction mixture by lyophilization into a cooled receiving flask. This gave an aqueous solution of (S)- or (R)-[ 2 H it3 H]methanol (1a or 1b, respectively) in greater than 90% radiochemical yield based on the ester. 52

Conversion of (R)- and (SHPHi^HJMethanol (1 b and la ) to Acetate for Configurational Analysis.

The conversion of 1a and 1b into acetate was carried out as shown in Figure 18 . An aqueous solution of the methanol was diluted with nonlabeled carrier material to give a total of .250 mmol of methanol. To this was added 3,5-dinitrobenzoyl chloride and N,N- dimethylaniline in benzene and the reaction was allowed to proceed with stirring at room temperature for 2 days. After purification,this reaction afforded methyl 3,5-dinitrobenzoate which now contained a chiral methyl group. The methyl group was then displaced from the ester with KCN in HMPA to give acetonitrile with inversion of configuration. It had previously been demonstrated that in HMPA, the ester carbon-oxygen bond can be cleaved by a direct displacement of the methyl group with KCN56>66.

The acetonitrile was then converted into acetamide by an alkaline hydrogen peroxide oxidation followed by diazotization with HNO 2 to afford acetic acid. This degradation scheme produced acetic acid in which the methyl group had the opposite configuration as the methyl group in methanol. Chirality analysis of the acetate showed that the methanol 1a was 35% ee S and 1b was 44-58% ee R. Therefore, the synthesis proceeded with a substantial amount of racemization, i.e., the conversion of 82% ee 2a and 94% ee 2b was accompained by 38% and 62% racemization respectively. The racemization could have occurred by one of the following mechanisms: 1 . Sn 2 attack at the ester methyl carbon of methyl p- methoxybenzoate by potassium p-methoxy benzoate (Figure 19).

o c h 3 o c h 3 o c h 3

Figure 19

2. Solvent participation of HMPA during sulfonimide displacement by potassium p-methoxybenzoate (Figure 20).

CH 3 N ( S 0 2 CF 3 ) 2 + OP[N(CH3)2]3— - HMPA

CH 3 0 P [ N f C H 3 ) 2 ] 3 N ( S 0 2 CF 3 ) 2 t N u c Figure 20

3. A mixed hydrolysis mechanism which involves attack of the hydroxide anion on the methyl group as well as the carbonyl carbon of the ester (Figure 21). 54

o o

O^'OH ° r 0 ~ " ’ OCH 3 OCH 3

Figure 21

The multiple displacement mechanism occurring as a side reaction (Figure 19), would result in a transesterification reaction, leading to racemization of the chiral methyl group. This type of reaction is rarely observed, however, it has been demonstrated to occur in aprotic solvents such as HMPA, DMF, and pyridine. For example, the reaction of methyl esters with sodium cyanide in HMPA results in S n 2 displacement of the methyl group, affording the carboxylate anion and acetonitrile66-67. Eschenmoser and coworkers have also reported examples in which the attack on the methyl carbon atom predominates in solvents such as DMF and pyridine68.

The O-alkylation of phenols by esters has also been demonstrated69. Kito and collegues suggested that an electron donating group in the para position would increase the electron density on both the oxygen of the carbonyl group and the methyl carbon atom of the ester. This increased electron density would retard the attack of the phenoxide anion on the methyl carbon if this was the rate determining step. Their results clearly show that attack at the methyl carbon was indeed rate determining. This oxygen to oxygen alkyl transfer might support the hypothesis that 55 the racemization is due to a side reaction in which the methyl group of 3a and 3b is displaced by the carboxylate anion. No examples of ester to ester alkyl transfers have been reported.

The stoichiometry of the reaction (3 equivalents of the carboxylate anion to 1 equivalent of N-methyl-N,N-di-p- toluenesulfonimide) certainly increased the probability of this side reaction occurring. An experiment in which the stoichiometry is set at 1:1 might establish if indeed excess nucleophile was the factor responsible for the extensive racemization.

The possibility of solvent participation could also be an explanation for the observed racemization (Figure 20). It has been reported that the reaction of bis trifluoromethanesulfonimides with

NaCN in HMPA results in formation of the HMPA salt. The HMPA salt in turn could react with nucleophiles to afford the desired product with racemization of the methyl group70-71. This reaction has been studied further by Townsend & Theis and was shown to occur with methyltriflylimide as well72. They found that solvent participation could be advoided when mixed sulfonimides were used. These methyl transfer reagents can be prepared from the corresponding tosylimides and trifluoromethanesulfonic anhydride. Using NMR, they demonstrated that there is no solvent (HMPA) intermediate in reactions with the mixed sulfonimides when performed at room temperature. They also looked at the reaction of N-methyl-N,N-di- p-toluenesulfonimide in HMPA at room temperature and did not 56 detected a solvent intermediate. Two pieces of evidence argue against solvent participation in the ester formation reaction: 1) The methyl transfer from N-[ 2 H if3 H]methyl-N,N-di-p-toluenesulfonimide to the sulfur of homocysteine during methionine synthesis was performed in HMPA at 80 9C 59 (Figure 16). This displacement resulted in a high degree of inversion of configuration, therefore, the absence of a solvent intermediate with HMPA can be implied. 2)

As previously mentioned, the direct displacement of the chiral methyl group on 2a and 2b with KCN in HMPA at 90 °C also proceeded with clean inversion of configuration. From these results it seems unlikely that solvent participation is the cause of the racemization seen in the synthesis of 1 a and 1 b.

There is ample literature precedent for a tetrahedral intermediate in the hydrolysis of methyl p-methoxybenzoate with

LiOH, however, the possibility of a mixed hydrolysis mechanism can not be neglected. A mixed hydrolysis mechanism would involve attack by the hydroxide ion on the methyl carbon as well as on the carbonyl carbon, leading to racemization of the methyl group (Figure

21). Hydrolysis of methyl esters by displacement at the methyl carbon is rarely observed. Two cases in which hydrolysis is known to proceed by this mechanism are: 1. Hydrolysis of p- propiolactones73, in which attack on the p-carbon allows for the release of ring strain. 2 . Hydrolysis of methyl 2,6- 57 dibutylbenzoate74, where attack at the carbonyl carbon is blocked by the bulky ortho substituents.

A key experiment that might determine which step of the synthesis resulted in the racemization would be to perform a direct displacement on methyl p-methoxybenzoate with KCN in HMPA. This would result in the formation of acetonitrile which could be converted in the usual manner to acetic acid for the chirality analysis. This experiment would reveal the chiral purity of the methyl group prior to ester hydrolysis, which would be an indication of the reaction mechanism taking place in the ester formation step.

An additional experiment which would probe for a mixed mechanism for the hydrolysis reaction would be to synthesize the ester from CH 3 18OH then perform the hydrolysis under the same reaction conditions as before and look at the 180 content of the resulting methanol following derivatization to methyl 3,5- dinitrobenzoate. If the hydrolysis were only occurring by attack at the carboxylate carbon, this would result in the benzoate anion and

CH 3 18OH, whereas hydroxide attack at the methyl carbon would result in the formation of CH 3OH and 180-labeled benzoate anion.

Despite its low optical purity, the chiral methanol derived from methyl p-methoxybenzoate was used to determine the stereochemical outcome of methyl coenzyme M formation in M . barkeri. A synthesis of chiral methanol with higher optical purity is 58 still needed to confirm the results obtained from the enzyme catalyzed reaction.

A possible alternative route involves the hydrolysis of (R)- and

(S)-[2 H i, 3 H]methyl p-toluenesulfonate, which can be synthesized in a three step procedure starting from chiral methylamine. Arigoni and coworkers developed a synthesis which afforded the chiral methyl p-toluenesulfonate with an optical purity of 66-70% ee

(Figure 22)75. In several attempts, however,the optical purity in our hands was consistently lower (54-59% ee). This procedure would afford chiral methanol of somewhat higher optical purity than the material now at hand, however, neither synthesis is completely satisfactory.

T T

N a N j, H3PO, NaOH, TsCI COOH D

T T

HNO.

N -S 0 2C7 H7 N-SOjCrHy

D CHLOROBENZENE, 95°C

inversion j f 0S 02C7H7

Figure 22 59

Conversion of 14C-Methanol into Methyl Coenzyme M using the Cell-Free Extract of Methanosarcina barkeri and Isolation of Methyl Coenzyme M.

With the following exceptions, conditions used to convert the

14C-labeled substrates to methyl coenzyme M were the same as those used in the conversion of the trideuterated substrates to methane : (1). The 14C-labeled substrates were used in place of the trideuterated substrates. (2). Bromoethanesulfonic acid was added to inhibit the production of methane. Bromoethanesulfonic acid

(BrES) is a known inhibitor of the methyl reductase-catalyzed conversion of methyl coenzyme M to methane. It does not, however, inhibit the methyltransferase reaction. The use of BrES minimizes the loss of label due to methane formation and results in the accumulation of methyl coenzyme M. Vogels and coworkers demonstrated that 250 uM concentrations of BrES completely inhibited methane production 42 and that concentration was used in this study as well.

The enzymatic reaction was stopped after 5-6 hours by venting the H 2 and adding an equal volume of cold ethanol. A initial workup consisted of centrifugation and passage of the supernatant through a cation exchange column [AG 50 W-X 8 (H+)] and subsequent TLC on cellulose. After development of the TLC plate (methanol/1,3- dioxolane/water/ammonium hydroxide, 3:6:1:1)76, two radioactive bands (Rf = .72, Rf = . 6 ) were isolated. In order to determine which 60

band contained the methyl coenzyme M, an additional experiment was

perform ed.

The enzymatic reaction was repeated using tracer amounts of

14CH 3 0 H with 5 mM CD3OH as carrier. The reaction was worked up

as before and the two bands were isolated and separately applied to

a cation exchange column which served to convert the methyl

coenzyme M sodium salt into the methyl coenzyme M sulfonic acid

form. The eluent was concentrated and the residue from each band

was treated with diazomethane. This step converted the sulfonic

acid portion of methyl coenzyme M into the methyl ester so that it could now be analyzed by GC/MS44. GC/MS analysis determined that

only the higher Rf (Rf = .72) band contained the trideuterated methyl coenzyme M. The methyl group of methyl coenzyme M was fully deuterated, therefore, there is no apparent proton exchange in the

methyltransferase reaction. The 14C-methyl coenzyme M was

isolated in radiochemical yields of 30% to 40%. This high level of incorporation permitted the use of the chiral substrate for the final experiment.

Conversion of (R)- and (SH 2 Hi,3H]Methanol into (R)- and

(S)-[ 2 H ii3 H]Methyl Coenzyme M with the Cell-Free Extract of Methanosarcina barkeri.

The conditions used for the enzymatic conversion of chiral methanol and for the isolation of methyl coenzyme M were the same as those previously described for the 14C labeled substrates. The samples of chiral methanol (R: 9.32 x 106 and 1.15 x 10 6 dpm; S: 2.1 x 107 dpm) were used without added carrier. This resulted in a final methanol concentration in the enzymatic reaction of .5-1 mM. As previuosly described, labeled methyl coenzyme M was isolated from the reaction mixture in 30-40% radiochemical yield. The chiral samples were then mixed with nonlabeled carrier and degraded to acetic acid for the chirality analysis 77 (Figure 23).

Degradation of [2 H i, 3 H]Methyl Coenzyme M to Acetic Acid.

Methyl coenzyme M was converted into acetic acid by a modification of a procedure developed by Arigoni and coworkers for the degradation of methionine78. This sequence started by converting methyl coenzyme M into the dimethylsulfonium salt using trimethyloxonium tetrafluoroborate (Figure 23). The sulfonium salt was then treated with potassium p-nitrobenzoate in refluxing cyclohexane to afford the thioester with inversion of configuration of the methyl group. Arigoni and coworkers found that the direct displacement reaction of a related sulfonium salt with KCN resulted in extensive proton exchange. This exchange, probably due to ylide formation, would result in significant racemization of the chiral methyl group. The less basic p-nitrothiobenzoate anion did not lead to proton exchange. This step results in the loss of 50% of the n O to

— s o ^ 2 2 ch ch \ i R A S :42%S e.e. R s33—38%R ae. KCN/HMRA (inversion) Methyl Coenzyme M Coenzyme Methyl 2 no V=o 161 ¥ HJs A

M. borkeri M.

ooh . ENZ ^ \ : 37.9 = 4 2 = % 37.9 e.e. 4 f t = = 3 3 -3 8 % e.e. \ R: F* 60.9; 59.6 S:F* VT ^HRI AT\ U - 7 — N (inversion) 3 3 2 - s o OH/H^)2 2 o CH 1) N 1) 2) HNO \ | R T' \ © T ^s— ch M. M. borkeri

n \: '" iR '" 'C R * 1 Figure 23. Steric course of the enzymatic synthesis of methyl coenzyme M coenzyme from M. methanol, and configurational analysis of methyl (CH R :4 4e.e. —58% S 35%e.e. : HBI =5—hydroxybenzimidozole HBI

Figure 23 63 radioactivity, since there is an equal chance of displacing either methyl group on the sulfonium salt.

After purification, the methyl p-nitrothiobenzoate was treated with KCN in HMPA to afford acetonitrile. This Sn 2 displacem ent reaction proceeds with inversion of configuration and the acetonitrile was converted into acetic acid as previously described.

The overall radiochemical yield for the conversion of methyl coenzyme M to acetic acid was 1-2%. This is consistent with the yields seen by Arigoni and coworkers for the degradation of methionine to acetic acid78.

Results from the Stereochemical Analysis of [2H-|,3H]Methyl Coenzyme M Derived from (R)- and (S)- [2H it3H]Methanol.

The procedure for the degradation of [ 2 H i, 3 H]methyl coenzyme M to acetate involves two Sn 2 displacements of the methyl group.

Each Sn2 displacement proceeds with inversion of configuration, resulting in overall net retention of configuration. The degradation thus affords [ 2 H i, 3 H]acetic acid whose methyl group has the same configuration as the starting pHi^Hjmethyl coenzyme M.

Methanol of 44-58% ee R configuration after conversion to methyl coenzyme M and subsequent degradation gave acetic acid of

33-38% ee R configuration (F = 60.9; 59.6). The methyl coenzyme M from methanol of 35% ee S gave acetic acid of 42% ee S 64 configuration (F = 37.9). -These results demonstrate that the methyl group of methanol is transformed into methyl coenzyme M with net retention of configuration 77 (Figure 23). CONCLUSION

This study has determined that the generation of methyl coenzyme M from methanol in M. barkeri proceeds with overall retention of methyl group configuration77. This result supports the double displacement mechanism proposed by Vogels and coworkers in which the methyl group of methanol is transferred first to the cobalt in B 12 and then to the sulfur of coenzyme M42.

The B 12-dependent methionine synthase from E.coli. catalyzes a similar transformation as that seen in the methyl transfer from methanol to methyl coenzyme M in M. barkeri. The stereochemical analysis of this reaction showed that the methyl group of 5- methyltetrahydrofolate is also transferred to homocysteine with net retention of configuration25. This result is consistent with a proposed mechanism which involves the transfer of the methyl group from 5-methyltetrahydrofolate to the cobalt of vitamin B 1 2 , followed by a second transfer to the sulfur of homocysteine to afford methionine.

Both enzymatic reactions pose the same question: How is the carbon-oxygen bond of methanol or the carbon-nitrogen bond of 5- methyltetrahydrofolate activated prior to the transfer of the methyl group to vitamin B 12. In the present case with methanol, one might postulate that the direct nucleophilic displacement would not be

6 5 66 facile since hydroxide is a rather poor leaving group. One can postulate that methanol is converted into an activated intermediate prior to transfer of the methyl group to the cobalamin of B 1 2 .

Phosphorylation of methanol, for example, would produce methylphosphate, which should be a suitable substrate for nucleophilic displacement reactions. A similar activation strategy is seen in the shikimate pathway and in terpene biosynthesis. To date there is no evidence to support any mechanism of methanol activation in methanogens.

A cell-free system from the methylamine adapted M. barkeri has also been developed for the enzymatic conversion of 14C- methylamine into 14C-methyl coenzyme M in 30-40 % yields. Chiral methylamine will be converted into methyl coenzyme M with this cell-free extract. Before the experiment with the chiral material could be carried out, however, a diminished cell growth was observed and for this reason the experiment has been delayed.

Acetate metabolism in methanogens is quite controversial and poorly understood. A stereochemical analysis of the conversion of acetate into methyl coenzyme M would help to establish a reaction mechanism and might answer the question about the involvement of one or two corrinoid enzymes in acetate metabolism to methane. If the theory of Thauer and coworkers is correct, then methyl coenzyme M formation from acetate would involve three sequential transfers of the methyl group and the stereochemical outcome would be net inversion of configuration. 67

The organism shows very slow growth on acetate, however, a gift of cell paste from Dr. Joe Krzycki has allowed the preparation of an active cell-free extract. Initial studies with this extract and chiral acetate have resulted in very low incorporation of radioactivity into methyl coenzyme M. The majority of the radioactivity is contained in an lower moving band (Rf=. 6 ). This band was also seen in the work with methanol as the substrate, but it was a minor product in that reaction. The structure of this unknown compound was not determined, although it is possible that it might be the Ci carrier

(N-methyltetrahydromethanopterin ?). If this were indeed the case, its degradation by diazotization and oxidation with KMn 0 4 should afford methylamine75'79.

The formation of radioactive methylamine by such a degradation would be direct evidence that this band did indeed contain a N- methylated compound. The experiment with chiral acetate could then be used not only to determine the stereochemical outcome of the conversion into methyl coenzyme M, but also of the methyl transfer to nitrogen

The sucessful completion of these experiments will help to establish the reaction mechanism of the conversion of methylamine and acetate into methyl coenzyme M. EXPERIMENTAL

Materials and Methods Materials: All chemicals and solvents used were of the highest grade commercially available and most were used without further purification. Hexamethylphosphoramide (HMPA) was vacuum distilled from calcium hydride. The organic chemicals were purchased from Aldrich Chemical Co. and the biochemicals from

Sigma Chemical Co. Trimethyloxonium tetrafluoroborate was purchased from Alfa Chemical Co. The following radioactive substrates were obtained from ICN: 14CH 3 0 H (S.A. 50 uCi/umol),

14CH 3 NH 3 CI (S.A. 50 uCi/umol). The (R)- and (S)-[2Hi 3H]acetate 61 and (R)- and (S)-[ 2 H if3 H]methyl-N,N-di-p-toluenesulfonimide which were used as the starting materials in several syntheses were prepared by Dr. Thomas Spratt and Dr. Thomas M. Zydowsky80. The following deuterated compounds were purchased from Aldrich

Chemical Co. (atom % 2 H): Methyl-d 3 -amine hydrochloride (98), m e th y l-d 3 alcohol-d (99.5), and acetic-d 3 acid, sodium salt (99).

Thin layer chromatography (TLC) separations were carried out using pre-coated cellulose plates (.5mm thickness, Brinkmann). Gases used in the cultivation and fermentation of M. barkeri and as analytical standards were purchased from Matheson.

6 8 69

Analytical Methods: Radioactive samples were counted using

Aquasol-2 (New England Nuclear) or 963 (New England Nuclear) scintillation cocktail in a Packard Minaxi-p Tri-Carb 4000 liquid scintillation counter. The counting efficiencies were determined using [ 14C]- and [ 3 H]toluene as internal standards. Radioactivity on chromatograms was located by scanning the developed TLC plates with a Berthold radiochromatogram scanner. The configuration and chiral purity of the acetate samples were determined by the method of Arigoni and co-workers 18 and Cornforth et al .19 by a procedure used routinely in this laboratory (see Appendix). The gas chromatography-mass spectrometry (GC/MS) for the determination of methane and ethane was carried out on a Finnigan GC/MS at The

Ohio State Chemical Instrument Center using an Alltech carbosieve

S-ll column. Routine GC/MS identification of synthetic intermediates was performed on a Hewlett-Packard 5790A mass selective detector and 5790 gas chromatograph with a 9000-216 data system using a Supelco SPB-5 capillary column. Mass spectral data are reported in the following order: m/z (relative intensities).

Purification using high performance liquid chromatography (HPLC) was done with a Waters model 590 programmable solvent delivery system, a Hitachi L-3000 multi-channel photodiode array detector, and a Hitachi D-2000 chromato-integrator, using an IBM silica gel column (5 urn, 10 mm x 250 mm, flow rate 8 mL/min, isocratic system hexane/methylene chloride, 35:65). 70

Cultivation and Fermentation

M. barkeri strain 227 was transferred anaerobically from a frozen glycerol stock solution into a 125 mL Wheaton bottle containing anaerobically prepared ER medium (20 mL). The culture was incubated at 37 °C for 7 days with constant shaking under an atmosphere of H 2/CO 2 (60:40, 40 psi). Once the culture became turbid, 1 mL of the cell suspension was anaerobically inoculated into fresh medium which was then incubated under the same conditions until turbid. Consecutive transfers were subsequently performed until good growth had been established under these conditions. At this point, optimal cultivation time was approximately 2.5 days. An aliquot (1 mL) of the cell suspension was then transferred into ER medium (20 mL) which had been prepared under an atmosphere of

N 2/CO 2 (60:40, 40 psi) and which contained a 100 mM concentration of either methanol, methylamine, or acetate49. Another consecutive transfer regimen was conducted with the alternate carbon substrate.

Linder these conditions, several days were required before turbity indicated that the growth was sufficient for inoculation into the fresh medium. After induction of the growth on the alternate substrates, the onset of turbity for the methanol and methylamine cultures was approximately 4-5 days, while the growing time for the acetate grown cultures was considerably longer.

A 20 mL volume of the culture in ER medium was inoculated via a double needle into a one liter serum bottle containing 300 mL of ER medium under an atmosphere of N 2/CO 2 (60:40, 15 psi) and incubated under the same conditions described previously. These cultures were used as the inoculum for the 24 liter fermentations.

The fermentations for the methanol-adapted M. barkeri were conducted in ER medium which contained 100 mM methanol in a 28 liter New Brunswick Microferm fermentor which had been modified for anaerobic cultivation. The fermentor was equipped for a continous sparging of gas into the medium and the fermentation was carried out under a positive pressure of N 2/CO 2 (60:40) at 37 °C with agitation at 300 rpm. Twice daily, sterile 15% Na 2 S solution (20 mL) was injected into the fermentor through the inoculation port.

The culture was allowed to grow until it became yellow and turbid.

This stage was usually reached after about 4 days. Cell growth was monitored by periodically withdrawing a small sample of the medium through the sample port for visual inspection.

The following medium (ER medium) was prepared anaerobically and used for this study: Double distilled water (880 mL), NH 4 CI (2.7 g), NaHC03 (5.0 g ), Na2C 0 3 (1.25 g), CH3COONa (2.5 g), .5% FeS04 (2 mL, freshly prepared), mineral solution I (50 mL), mineral solution II

(50 mL), trace mineral solution (10 mL), vitamin solution (10 mL), and resazurin solution (4 mL). Mineral solution I consisted of

K 2 HPO 4 (6 g) in double distilled water (1 L). Mineral solution solution II contained the following: KH 2 PO 4 ( 6 g), (NH4 )2 S 0 4 (6 g),

NaCI (1 2 g), M gS 04 (1 .2 g), CaCl2 (.16 g) in double distilled water (1 72

L). Trace mineral solution consisted of Na 2 SeC >3 (24 mg), C0 CI2 (1 0 0 mg), Na2 MoC>4 2 H2 O (24 mg), NiCl2 -6 H2 0 (24 mg), and concentrated

HCI (10 mL) in double distilled water (1 L). The vitamin solution consisting of biotin (2 mg), folic acid (2 mg), pyridoxine HCI (10 mg), thiamine HCI (5 mg), riboflavin (5 mg), nicotinic acid (5 mg),

DL-Ca pantothenate (5 mg), vitamin B 12 (.1 mg), p-aminobenzoic acid

(5 mg) and lipoic acid (5 mg) in double distilled water (1 L) was filter-sterilized and stored at 0 to 4 °C. The reducing agent was prepared in the following way: Double distilled water (400 mL) was boiled and cooled to room temperature after which cysteine.HCI (5 g) was added. The pH was adjusted to 12 with NaOH pellets and

Na 2 S.9 H2 0 (5 g) was added. The solution was stored in 10 portions and dispensed in 125 mL serum bottles. These bottles were purged with N2 , stoppered, crimped, and autoclaved for 15 min. The reducing agent was kept at 0 to 4 °C. The resazurin solution was prepared by dissolving two pellets of resazurin in double distilled water (8 8 mL)48.

ER medium was prepared in a 2 liter flask and heated to boiling over a Bunsen burner. While the media was being heated, four serum bottles (one liter), were being purged with H 2/CO 2 (60:40). Media

(250 mL) was dispensed into each serum bottle while it was still hot and gas purging was continued for 5 min. The bottles were sealed with black butyl rubber stoppers, crimped, and placed into the anaerobic chamber. Inside the chamber, ER media was dispensed 20 73 mL each into 125 mL serum bottles which were then sealed, crimped, and removed from the chamber. Using a gas/vacuum manifold, these bottles were evacuated and regassed (H 2/CO 2 60:40 three times), and then pressurized to 10 psi. The bottles were placed in a covered pan of water and autoclaved for 25 min at 120

°C. After the media had cooled to room temperature, reducing agent

(.25 mL per 20 mL of media) was added with a sterile syringe to each bottle and the bottles were pressurized to 40 psi with H 2/CO 2

(60:40). After 12-15 h the media became colorless and ready for inoculation. A pink color indicated that either the media contained oxygen or that the pH was incorrect and those bottles were discarded.

For preparation of one liter serum bottles, the hot media was dispensed directly into the bottles and autoclaved. Reducing agent

(1 ml per 1 0 0 ml of media) was added directly to those bottles after they had cooled to room temperature and each was pressurized to a final pressure of 15 psi with N 2 . The same media was used throughout this study, however; the gas atmosphere was changed as in d ica te d .

Methanol was autoclaved in a sealed serum bottle and added via a sterile syringe after the medium had cooled to room temperature.

Methylamine and sodium acetate stock solutions were dissolved in sterile anaerobic water, autoclaved and added to the medium via a sterile syringe. 74

The same medium was used for the 24 liter fermentations (20-

fold scale up). The medium was prepared in the fermentor and

purged while stirring at 200 rpm for 15 minutes under the desired

gas atmosphere. The exhaust port was then closed, the gas sparger

was shut off, and the medium was allowed to autoclave for 2 0 min.

After the temperature had dropped below 100 °C, the gas sparger

and exhaust were opened and agitation was increased to 300 rpm.

When the temperature reached 10 °C above the fermentation temperature (37 °C), reducing agent (200 mL) was added through the

inoculation port with a sterile double needle. After 8 h, 15% Na2 S

(20 mL) and methanol (80 mL) were added to the fermentor and then

the inoculum (1 L) of the organism was added.

Harvesting and Preparation of the Cell-Free Extract

A collection carboy (25 liter) was fitted with a black butyl

rubber stopper containing 3 ports for the following connections; a gas inlet port, a gas outlet port, and a harvesting port. The gas inlet from the carboy was connected to a N 2 cylinder and purged for 15-20

minutes at low pressure with the other two ports open. Before the carboy was connected to the fermentor for harvesting, 15% Na 2 S

(40 mL) was injected into both the empty carboy and into the fermentor. In preparation for harvesting, the carboy was connected to the fermentor harvest port while maintaining the positive N 2 flow through the system. The gas exhaust valve on the fermentor was then closed and the sparger was left on. This allowed pressure to build up inside the fermentor, thereby forcing the culture broth into the carboy. When the carboy was full, the harvest port from the fermentor was closed, and the exhaust port on the fermentor was opened. The remaining ports on the carboy were closed while simultaneously shutting off the N 2 gas. The carboy was then connected to the anaerobic chamber while maintaining an anaerobic atmosphere in the system. The harvesting tube from the carboy was connected to the inlet port on the anaerobic chamber and the carboy once again connected to a N 2 gas cylinder.The exhaust tube was clamped off allowing the gas to flow in, and the increased interior pressure forced the contents of the carboy into the anaerobic chamber. The cells were collected by vacuum filtration and resuspended in a minimal amount of anaerobically prepared 50 mM

TES buffer (pH 7.2) containing 15 mM MgCl 2 and 2 mM dithiothreitol.

The cell paste was divided into 30 mL aliquots, sealed, and stored in

125 mL serum bottles at -70 °C under 40 psi H 2 . The yield of the fermentations was 1 g of cells per liter (wet weight). The cell paste could be stored for as long as one year with no apparent loss of activity.

To prepare the cell free extract, the cell slurry (30 mL) was thawed under an atmosphere of H 2 and loaded into a French pressure cell in the anaerobic chamber. The outlet of the French pressure cell was connected via tygon tubing to an Erlenmeyer flask (125 mL) fitted with a stopper having three ports clamped off. This arrangement allowed for removal of the filled cell from the chamber while maintaining an anaerobic atmosphere. After placing the

French pressure cell in the hydraulic press, the flask was connected to a N2 tank and continuously sparged throughout the procedure. The tube connecting the cell to the flask was opened and a pressure of

20,000 psi was applied. When breakage of the cells was completed, the tubes to the flask were clamped shut and the flask was transferred back into the anaerobic chamber. This procedure was repeated one additional time. The extract was dispensed into centrifuge tubes equipped with O rings, sealed in the anaerobic chamber, transferred out, and centrifuged (20,000 g, 30 min) to remove cell debris. The pellet was discarded and the crude enzyme extract was divided into aliquots (20 mL) in the anaerobic chamber and placed into sealed 65 mL serum bottles which were then pressurized to 40 psi with H 2 gas. Cell-free extracts stored at - 70

°C for 6 months showed full enzymatic activity. The crude enzyme preparation contained 12-24 mg/mL of protein as determined by the amido-black assay with bovine serum albumin as the standard51.

Conversion of CD3OH to CD3H With the Cell-Free Extract of Methanosarcina barkeri

Trideuterated methanol was converted to methane with a cell free extract of M. barkeri strain 227 which had been grown on 77 methanol. The extract and all substrates were prepared anaerobically and the reaction was performed in a Wheaton serum bottle (65 mL) under a 40 psi H 2 atmosphere. The enzyme extract

(28 mg protein, 12-24 mg/mL) was activated by preincubation with

7.5 mM ATP and 15 mM MgCl 2 in 10 mM TES buffer (pH of 7.2) for 15 minutes at 37 0C. The following were added anaerobically: 2 mM dithiothreitol, 18 mM coenzyme M, and finally 10 mM CD 3 OH to initiate the reaction. The reaction mixture, which had a final volume of 12 mL, was incubated at 37 °C with gentle shaking overnight. Aliquots were removed from the head space of the serum bottle with a gas tight syringe and analyzed for methane production via GC/MS; MS m/z_19 (M+, 100), 18 (46), 17 (56), 16 (7.5).

This experiment was also carried out using CD 3 NH 3 CI as the substrate and a cell free extract from M. barkeri grown on methylamine. The mass spectral data were identical to those obtained from the experiment done with methanol as the substrate.

Formation of Trideuterated Methane by Methanosarcina barkeri Grown on Trideuterated Substrates.

M. barkeri strains 227 and MS were grown in ER medium containing 100 mM CD 3 OH under 40 psi N 2/CO 2 (60:40) atmosphere.

The cells were grown in Wheaton bottles (65 ml) at 37 °C with gentle shaking until turbid. An aliquot from the headspace was 78 withdrawn with a gas tight syringe and analyzed for methane production by GC/MS.

The mass spectral data from this experiment were identical to those of the cell free experiment. This experiment was also performed using trideuterated methylamine and trideuterated sodium acetate in final concentrations of 100 mM44-50.

Synthesis of (S)-[2 H i,3 H]Methanol (1 a)

Under an argon atmosphere, (R)-N-[ 2 H i, 3 H]methyl-N,N-di-p- toluenesulfonimide (2b) [95 uCi, S.A. 6 .2 uCi/umol, 94% e.e.R]80, diluted with nonlabeled carrier to give a total of 59 umol, and potassium p-methoxybenzoate (33 mg, 177 umol) were dissolved in freshly distilled HMPA (5 mL) and stirred at 110 °C for 2.5 days. The progress of the reaction was monitored by GC/MS. The reaction mixture was cooled to room temperature, poured into water (50 mL) and extracted with ether (5 x 50 mL). The combined ether layers were back-extracted with water (5 x 50 mL), thoroughly dried over anhydrous MgS 0 4 , and concentrated in vacuo. In order to remove the last traces of HMPA, the crude product was dissolved in an ether/hexane (3:1, 100 mL) mixture and using this mixed solvent system, the extraction sequence described above was repeated. The combined organic extracts were dried over anhydrous MgS 0 4 and concentrated in vacuo. The crude product was dissolved in methylene chloride and purified by HPLC (silica, mobile phase 79

hexane/methylene chloride, 35:65). This gave (S)-[ 2 H it3 H]methyl p- methoxybenzoate (3a) [48 uCi, S.A. 1.61 uCi/mmol, 51% radiochemical yield] which was pure as determined by GC/MS and analytical HPLC.

Hydrolysis of 3a was conducted in 1 M LiOH (1 mL) for 62 h at room temperature. Lyophilization of the water/methanol reaction mixture into a cooled receiving flask yielded an aqueous solution of

1 a in greater than 90% yield. The aqueous solution of 1a was then converted to acetate for chirality analysis by the method described below .

In an identical manner, (S)-N-[ 2 H-| 3 H]methyl-N,N-di-p- toluenesulfonimide (2a) [91 uCi, S.A. 47 uCi/umol, 82% e.e.S]80, afforded (R)-[ 2 H i, 3 H]methyl p-methoxybenzoate (3b) [13.46 uCi, S.A.

1.54 uCi/umol, 15% radiochemical yield]. Hydrolysis of 3b also gave a greater than 90% yield of 1b.

Conversion of (S)-[ 2 H i, 3 H]Methanol (1 a) to Acetate

An aqueous solution (1.5 mL) of 1a was diluted with nonlabeled carrier methanol to give a total of .250 mmol. To this solution was added 3,5-dinitrobenzoyl chloride (461 mg, 2 mmol), N,N- dimethylaniline (242 mg, 2 mmol, .253 mL), benzene (30 mL) and the reaction was vigorously stirred for 2 days at room temperature. The mixture was washed with 1 N NaOH (2 x 30 mL) and the aqueous layer was back-extracted with chloroform (2 x 30 mL). The 80

combined organic extracts were dried over anhydrous Na 2 S 0 4 and concentrated in vacuo to afford a dark oil. The crude product was dissolved in a minimal amount of acetone and purified by TLC

(alumina, 1mm preparative plates, hexane/benzene, 1:1). The band corresponding to (S)-[ 2 H-| 3 H]methyl 3,5-dinitrobenzoate (4a) was scraped off and the product was eluted with chloroform.

To a solution of 4a in HMPA (3 mL), KCN (65 mg, 1 mmol) was added and the mixture was frozen in a 25 mL round bottom flask containing a magnetic stir bar. This flask was connected, via a lyophilization bridge, to another flask containing water (1 mL). Both flasks were frozen in dry ice/isopropanol and the system was evacuated with a vacuum pump. The reaction flask (4a + KCN +

HMPA) was heated to 90 °C while the receiving flask was cooled to -

78 °C. The reaction was continued for one week, after which time the vacuum was released and the frozen flask [now containing water and (R)-[ 2 H i, 3 H]acetonitrile (5b)] was allowed to thaw.

The aqueous solution of 5b was treated with 30% hydrogen peroxide (3 mL) and 6 N NaOH (.100 mL). The solution was heated with stirring for 6 h at 50-55 °C. During the early stages of the reaction vigorous bubbling occurred. After about 5 h, the solution was cooled to room temperature and concentrated in vacuo at 35 °C to yield (R)-[2Hi , 3 H]acetamide (6 b).

Amide 6 b was dissolved in water (1 mL) and cooled to 0 °C. Cold

5 N H2 SO 4 (2 mL) was added dropwise with stirring, followed by the 8 1

dropwise addition of aqueous NaN 0 2 (400 mg in 1 mL of H2 O). At this point brown fumes appeared and the mixture was allowed to stir at

0 °C for 1 h. The reaction was warmed to room temperature and stirred for an additional 5 h. The acetic acid was isolated by a steam distillation and then further purified by a series of basic and acidic distillations in preparation for the chirality analysis56-59.

In an identical manner, 1b was converted to acetate for chirality an alysis.

Enzymatic Conversion of (S)-[ 2 H i, 3 H]Methanol (1 a) to (S)-

[2 H 1 ,3 H]Methyl Coenzyme M (7a) using a Cell-Free Extract of Methanosarcina barkeri.

An aqueous solution of 1a was converted enzymatically to 7 a using the cell free extract of M. barkeri strain 227 which had been adapted to grow on methanol. The cell free extract and the substrates had been prepared anaerobically and the reaction was performed in a sealed Wheaton bottle (65 mL). The crude extract (17 mg of protein, 12 mg/mL) was activated by preincubation for 15 min at 37 °C under 40 psi H 2 atmosphere with 7.5 mM ATP and 15 mM

M g2 CI in 10mM TES buffer adjusted to a pH of 7.2. The following were added anaerobically: 2 mM dithiothreitol, 18 mM coenzyme M,

.25 mM bromoethanesulfonic acid (to inhibit the methyl reductase) and an aqueous solution of 1a (0.5-1 mM) to initiate the reaction.

The reaction mixture, which had a final volume of 6 mL, was 82 s incubated at 37 °C with gentle shaking for 5-6 hours under 40 psi

H242. The reactions were stopped by venting the H2, and then adding an equal volume of cold ethanol. After centrifugation, the supernatant was concentrated in vacuo, diluted with water (2 mL), and passed through a cation exchange column [AG 50 W-X 8 (H+), 1.5 cm x 8 cm, water] and further purified by TLC (cellulose, .5mm preparative plate, methanol/ 1 ,3 -dioxolane/water/ammonium hydroxide,

3 :6 :1 :1 ) 76 The fastest moving band (Rf=.72) was eluted with 50% aqueous methanol to afford 7a in 30-40% yields77. An aqueous solution of 1b was converted enzymatically to 3b in an analogous manner.

Enzymatic Conversion of Trideuterated Methanol to Trideuterated Methyl Coenzyme M using a Cell-Free Extract of Methanosarcina barkeri.

In an identical manner an experiment using tracer amounts of

1 4 CH 3 0 H with 5 mM CD3 OH as the substrate to initiate the reaction provided methyl coenzyme M. Upon working up the reaction as before, two radioactive bands (Rf=.75, Rf=. 6 ) were isolated from the cellulose TLC plate. Each band was applied to a cation exchange column [AG 50 W-X 8 (H+), .5cm x 6 cm, water] which served to convert the sodium salt of methyl coenzyme M to the corresponding sulfonic acid. After concentration in vacuo, the residue was 83

dissolved in a minimal amount of methanol (.2 mL) and excess

diazomethane was added to derivatize the free acid into the sulfonic

acid methyl ester. Analysis by GC/MS determined that only the band

with the higher Rf contained the trideuterated methyl coenzyme M .44

MS m/i_173 (M+, 7.6), 77 (100), 64 ( 1 1 ), 43 (24).

Conversion of (S)-[ 2 H i, 3 H]Methyl Coenzyme M (7a) to Acetate.

Under an argon atmosphere, 7a was diluted with nonlabeled

carrier methyl coenzyme M to give a total of .506 mmol followed by

the addition of 98% formic acid (.300 mL) and methylene chloride (5

mL). The solution was cooled to 0 °C after which trimethyloxonium tetrafluoroborate (96 mg, .646 mmol) was added and the two phase

mixture was vigorously stirred for 12-15 h at room temperature.

The mixture was then concentrated in vacuo and the product was dried by azeotropic distillation with toluene (3x5 mL). Finally, the

product was dried overnight on a vacuum pump to afford the methyl sulfonium salt of methyl coenzyme M ( 8 a) [51% radiochemical yield].

p-Nitrothiobenzoic acid (200 mg, 1 mmol) was suspended in water (5 mL), made alkaline with 30% NaOH, and then neutralized with 30% H3 PO 4 . Half of this solution was added to a solution of 8 a

(42.5 mg, .253 mmol) in cyclohexane (30 mL) and the mixture was heated to 80 °C with stirring for 18 h. The remaining portion of the thioacid was then added to the reaction and heating was continued 84 for an additional 24 h. The cyclohexane layer (top layer) was removed and replaced with fresh cyclohexane (30 mL) every 8-10 h.

The pH of the aqueous layer was monitored at these times and it readjusted to neutrality as needed. The combined cyclohexane extracts were dried over anhydrous MgSC >4 and concentrated in vacuo.

The crude product was purified by TLC (silica gel, 2 mm preparative plates, benzene, Rf=.5 ) 78 This yielded the pure (R)-[ 2 H it3 H ]m ethyl p-nitrothiobenzoate (9b) [34% radiochemical yield].

To a solution of 9b in dry HMPA (3 mL), KCN (30 mg) was added and the mixture was frozen in a flask equipped with a magnetic stir bar. This flask was then connected via a lyophilization bridge to another flask containing water (1 mL).. Both flasks were frozen in dry ice/isopropanol, the system was evacuated, and the temperature of the reaction flask (9b + KCN + HMPA) was gradually increased to

160 °C over 12 h while the receiving flask was cooled to -78 °C.

The reaction was allowed to proceeded at 160 °C for an additional

1 2 h.

The aqueous (S)-[ 2 Hi3H]acetonitrile (10a) solution was converted to acetic acid following the procedure described earlier for the conversion of methanol to acetate ( 6 .6 % radiochemical based on 9b, 1% overall yield for the entire sequence).

The degradation of 7b to acetate proceeded with yields nearly identical to those for 7a also giving a 1% overall yield for the entire sequence. PART III INTRODUCTION

The methyl coenzyme M reductase system catalyzes the reduction of methyl coenzyme M to methane in methanogenic bacteria. This two electron reduction is the last step in the overall conversion of

H 2 and CO 2 and other substrates, such as methanol, methylamine, and acetate, to methane. Although the components of the methyl reductase system have not yet been fully characterized, it seems to consist of a hydrogenase (component A 1)81, an ATP binding protein

(component A 2 )82, a protein containing the tightly bound cofactor

F4 3 0 (component C)83-84, the cofactor (7- mercaptoheptanoyl)threonine phosphate (component B or HS-HTP)85-

87, and a oxygen sensitive component (component A 3)88, The reaction also needs catalytic amounts of ATP, FAD, and Mg.

Component C has been proposed to be the site where the reduction actually takes place and is thought to be the methyl coenzyme M reductase. However, to achieve the reduction reaction in vitro, it is necessary to use an anaerobically prepared cell-free extract or a mixture of the partially purified components. All of the preparations reported to date result show than 1% of the activity found in whole cells.

85 86

Methyl coenzyme M reductase purified from Methanobacterium thermoautotrophicum has a molecular weight of 300,000 and consists of the following subunits, 012 (68,000), P 2 (48,000), 72

(38,000). The purified enzyme contains two molecules of tightly bound F 4 3 0 , two covalently bound molecules of coenzyme M (present in a heterodisulfide form), and an unknown amount of component B

89-92. Recently, Ankel-Fuchs and Thauer have developed an in vitro system consisting of component B, vitamin B 1 2 , dithiothreitol, and pure methyl reductase, which catalyzes the reduction of methyl coenzyme M to methane. This reaction was found to be strictly dependent on component B and to be stimulated 20 fold by vitamin

B 1293-

Coenzyme F 430 is a unique nickel tetrapyrrole discovered by Wolfe and coworkers whose structure and physical properties have been determined in Thauer's and Eschenmoser's laboratories94-95. These workers established that the structure contains characteristics of both porphyrins and corrins and have, therefore, called F 430 a

"tetrahydrocorrin" (Figure 24).

HjCOOC

H.NOC W t.-CH,H

HmC

COOCH H.COOC

COOCH, Figure 24 87

Various physical studies of bound F 430 (Raman spectroscopy, EXAFS, and X-ray) showed the nickel to be in a 6 coordinate form with axial ligation and long nickel nitrogen distances of approximately 2.1 A; this could represent a species which is axially reactive in substrate binding or catalysis96' 100.

In work with both whole cells and purified enzyme, Albracht gj a],, showed that an EPR signal is present during methanogenesis101.

This signal could be attributed to the nickel in F 4 3 0 in either the Ni 1 or Ni3 state. This finding demonstrates that a redox change is taking place during catalysis.

To date very little mechanistic information is available on this enzyme system. Recent work has given some insight into the possible role played by coenzyme F 4 3 0 . Juan and Pfaltz 102 have investigated the redox properties of F 4 3 0 using the pentamethyl ester (F 4 3 0 M) analog. They have demonstrated that the Ni 2 F4 3 oM can be reduced to Ni 1 F4 3 oM without affecting the porphinoid ligand, and this species could be considered as a potential intermediate in the reduction reaction. Recently they investigated the reactivity of the

N i1 form of F4 3 0 M with a methyl group bound to a halogen, oxygen, or a sulfur103. They found that in DMF, methyl iodide is much more reactive than either methyltosylate or trimethyl- and cyclic sulfonium salts. The sulfonium salts caused a lag period before the production of methane began, however, this lag period was not observed when the reaction was conducted in the presence of excess n-propanethiol. They speculated that a species is slowly formed 88 which serves to accelerate the reaction and has a similar effect as the added thiol. Possible candidates would be hydroxide ion or dimethylamine formed from the decomposition of DMF. In fact, with the addition of LiOH and diethylamine to the reaction, the lag period was again eliminated. They also demonstrated that upon using DMF- d 6 , CH3 D was not formed, whereas added CH 3 CH 2CH 2 SD or

(CH 3 CH 3 )3 NDCI did afford CH 3 D. These results suggest that the reduction reaction involves proton transfer to a methyl-nickel intermediate to yield methane and the Ni 2 F4 3 oM species. There is literature precedent for protonolysis of methyl-nickel species in aqueous solution to yield methane104.

The reaction of F 4 3 0 M with methyl coenzyme M, however, resulted in no methane production. Therefore, in nature methyl coenzyme M might be converted into an actived form or perhaps there exists a species (thiol?) which accelerates the reaction. The formation of an alkyl-NiF 4 3 o species would be similar to the alkyI-

C 0 -B 12 intemediate seen in methionine synthase105. This type of a mechanism has been recently postulated by Walsh and Orme-

Johnson 106 (Figure 25).

N -j N

f CH4 N N Figure 25 89

All of these findings suggest that the reduction of methyl coenzyme

M to methane may involve the nickel site in F 430; perhaps the enzyme acts as a biological equivalent of a Raney nickel catalyst.

The role of HS-HTP was postulated in the past to be that of a potential methyl group carrier. Ellermann alai-. synthesized the S- methylated HS-THP and investigated its effect on the reduction reaction using purified methyl reductase, vitamin B 12 and dithiothreitol107. They demonstrated that the S-methylated HS-HTP cannot be substituted for methyl coenzyme M and in fact is a competitive inhibitor with HS-HTP. Methane production was found to be dependent on the presence of HS-HTP, which indicates that the free thiol group is necessary. Since the purified methylreductase is always found to contain a heterodisulfide with a molecular weight between 400 to 600, they suggested that the heterodisulfide is the mixed disulfide formed between coenzyme M and HS-HTP82.108.

They have proposed the following reaction mechanism (Figure 26):

CH3 SC0 M + HS-HTP CH4 + CoMSS-HTP

CoMSS-HTP + DTTred HSCoM + HS-HTP + DTT0X

Figure 26 90

Reaction (1) is mediated by methyl coenzyme M reductase using F 430 as the catalyst with HSHTP being the electron donor. Reaction (2) would be nonenzymatic and catalyzed by vitamin B 12. One might speculate that the cofactor HS-HTP is the thiol species that accelerates the reaction or enhances the reactivity of Ni 1 F 4 3 o in methyl reductase reaction with methyl coenzyme M.

A recent paper by Walsh and coworkers reported the synthesis of various substrate analogs made by modifying the heteroatom and alkyl region of methyl coenzyme M. The analogs were used to probe the reaction mechanism of methyl coenzyme M reductase109. They reasoned that replacement of sulfur (covalent radus 1.02 A) with selenium, (covalent radus 1.16 A) would minimally alter the size of the heteroatom, but markedly alter bond strength (carbon-selenium bond dissociation energy is 28 kcal/mol weaker) and softness, thereby, affecting the nickel-heteroatom interaction upon substrate binding. This could in turn alter the dissociation constant since the heteroatom could be interacting with the enzyme's active site, affecting the Km term. The bond strength should have an effect on the KCat. especially if the bond cleavage is rate determining. A threefold increase in Km and Kcat was reported, therefore the catalytic efficiency for this analog (Kcat/KM) remained the same.

Replacement of sulfur with oxygen would result in a carbon-oxygen bond dissociation energy that is about 21 kcal/mol greater than that of the sulfur analog. Oxygen is a harder atom, therefore, the 9 1 interaction with the nickel should be decreased; and in fact, they saw that this analog was not turned over by the enzyme.

The substitution of electron withdrawing groups such as fluorine for hydrogen on the methyl group of methyl coenzyme M, would have a stabilizing effect if an alkyl-nickel species were involved110. A radiacal intermediate (CF 2 H), resulting from a homolytic cleavage of the carbon-sulfur bond, would also be stabilized as would a carbanion which could be an intermediate before protonolysis to afford methane. Walsh and coworkers found that difluoromethyl coenzyme M did turn over in the enzymatic reaction and the rate was

1.8 fold higher than that seen with methyl coenzyme M. However, trifluoromethyl coenzyme M was not a substrate for the enzyme.

One possible explanation is that three electron withdrawing fluorine atoms could greatly decrease the electron density on the heteroatom. This is turn could affect the nickel-heteroatom interaction in a similar manner as the substrate which had oxygen in place of sulfur as the heteroatom . It should be noted that the trifluoromethylseleno-coenzyme M analog was also synthesized in order to rule out the possibility that the increased CF 3 -S bond energy was the determining factor. The selenium derivative was also not a substrate for the enzymatic reaction. All of these findings suggest that there is a nickel-heteroatom interaction in the enzyme which is important for catalysis and that carbon-sulfur bond cleavage might be the rate determining step in the reduction rea ction . 92

A stereochemical analysis of the methyl reductase reaction might provide important information concerning the mechanism of this unique reaction. However there is an obvious problem: Methane has four hydrogens, and hence, four different isotopes of hydrogen would be needed to produce an isotopically chiral version of this molecule. However, only three hydrogen isotopes are known.

Therefore, it is necessary for this stereochemical analysis to replace one hydrogen by a different ligand. For example, the replacement of one hydrogen on the S-methyl group of methyl coenzyme M would afford ethyl coenzyme M as the substrate and the enzyme catalyzed reduction would produce ethane. For the purpose of the stereochemical study, ethyl coenzyme M containing one tritium and one deuterium at the enantiotopic positions on C-1 could be synthesized. The enzymatic reduction would now afford ethane with concommitant formation of a chiral methyl group from C -1

(Figure 27).

CH3 3 Com Methyl Reductase CH

.oT Methyl Reductase ,.*T CH3-C -D ► CH3-C - D \ H SCoM Figure 27 H 93

Chirality analysis requires the methyl group to be in the form of acetic acid (see Appendix), therefore, one would have to convert ethane to acetate for analysis.

Ethyl coenzyme M was found by Gunsalas to be reduced to ethane at a rate of about 20% that for methyl coenzyme M111.

Therefore, this approach would be feasible. However, the conversion of ethane to acetate by reactions of known stereochemical outcome was considered problematic, requiring oxidation at an unfunctionalized carbon atom. In addition, half of the label would be lost. A product that could be more easily converted to acetate for chirality analysis is propene. If 2-propenyl coenzyme M were found to be a substrate for methyl coenzyme M reductase, then the resulting propene could be easily converted to acetate for chirality analysis by a Lemieux oxidation. The 2 -propenyl coenzyme M would, therefore, have to be stereospecifically labeled with tritium and deuterium at C-1. Reduction of this substrate would then generate a chiral methyl group at C-3 of propene (Figure 28).

T 'T Methyl Reductase CH 2 SCH -C -D CH2 = C H -C -D ► -u \ H SCoM

Figure 28 94

Thus in order to study the steric course of the methyl coenzyme M reductase reaction, the following tasks needed to be accomplished:

Preparation of an active cell-free extract from M. barkeri. 2.

Conversion of 2 -propenyl coenzyme M to propene with the cell free extract. 3. A stereospecific synthesis of (R)- and (S)- 2 -[1 -

2 Hi3H]propenyl coenzyme M. 4. Conversion of the chiral propene into acetate for the chirality analysis.

A great deal of time was spent in learning the anaerobic techniques so that the organism could be sucessfully grown and an active extract could be prepared. These tasks were discussed in detail in Part II. However, the task of synthesizing the chiral substrate also proved to be very difficult and time consuming. This project is still underway and will hopefully be completed soon. RESULTS AND DISCUSSION

Preparation of the Cell-Free Extract of M ethanosarcina barkeri

The successful preparation of the cell-free extracts from M . barkeri was previously discussed in the section on the methyl transferase reaction. For the present study, the cell-free extract was prepared using the methanol adapted M. barkeri. The methanol adapted organism was chosen because the growth on this carbon source was the fastest and gave the best yield of cells per liter of medium. Therefore, it was easy to scale up to 20 liters and prepare

2 0 grams of cell paste from one fermentation.

Conversion of 2-Propenyl Coenzyme M to Propene with the Cell-Free Extract of Methanosarcina barkeri.

The synthesis of 2-propenyl coenzyme M (11) was performed using the method developed by Wolfe and coworkers111. This procedure involves the alkylation of 2 -mercaptoethanesulfonic acid (coenzyme

M) with allyl iodide in concentrated NH 4 OH to give 2-propenyl coenzyme M. Conditions for the enzymatic reduction reaction were the same as those used in the cell free system for conversion of trideuterated methanol to methane, except that 2 -propenyl coenzyme M was substituted for coenzyme M and the trideuterated

95 96 methanol was eliminated. To insure that the cell-free system was viable, trideuterated methyl coenzyme M (prepared from trideuterated methyl iodide) was used in place of the 2 -propenyl coenzyme M and the production of trideutrated methane was confirmed by GC/MS.

After several unsuccessful attempts, formation of a small amount of propene in the enzymatic reaction with 2 -propenyl coenzyme M was detected. Selective ion monitoring with GC/MS was used to confirm the presence of propene in the reaction headspace.

The spectrum of the mass spectral fragmentation is shown in Figure

29. The yields of propene from the enzymatic reaction, however, were to low to quantitate. It was reported by Walsh and coworkers that 2 -propenyl coenzyme M can also be an inhibitor of the reduction reaction with an I 50 of 23 uM109. In our enzyme incubation, the concentration of 2-propenyl coenzyme M was 18 mM. This dictates that the stereospecifically labeled 2-[1- 2 H i(3 H]propenyl coenzyme M be synthesized with a high specific radioactivity. This would offer two advantages: First, it would allow the use of a low concentration of the substrate in the enzymatic reaction which might result in less inhibition, therefore, increasing the yield of propene. Second, the high specific activity should insure that in the end, degradation of the resulting propene to acetate produces enough radioactivity for the chirality analysis. Therefore, encouraged by the results of the tuss srecnuH BATAi ZKHT2 178 BASE It/E i *4/23/87 13:54.88 ♦ 1:18 CALI. CAL42387 83 U C i SAIinJE: SAIffLE 86H. 2M. EMUNCEB (S 15* 2N 8T> 41

58. 8 *

*38

...... wr~ T" 25 36

Figure 29. Mass spectral fragmentation of propene. 98 enzymatic reaction, the synthesis of stereospecifically labeled 2- propenyl coenzyme M was started.

Proton Exchange in the Methyl Coenzyme M Reductase Reaction

In the conversion of trideuterated methanol to methyl coenzyme

M using the cell free extract of M. barkeri as described in the section on the methyl transferase reaction, the methyl coenzyme M that was isolated and analyzed by GC/MS was found to be exclusively trideuterated in the methyl group. The diagnostic base peak of rn/z

77 and the molecular ion peak of m/£_173 were used to confirm this result. In the conversion of trideuterated methanol to methane using the same cell free extract, the percentage of deuterium in trideuterated methane was less than that in trideuteromethyl coenzyme M. However, the percentage of deuterium was identical to that seen in the conversion of trideuterated methyl coenzyme M to methane using the cell free system. The distribution of deuterium in the methane is in accord with the results obtained by Gottschalk and coworkers from M. barkeri grown on trideuterated methanol50. Using a fragmentation pattern of CD 3 H and CD 2 H 2 reported in the literature, they came up with the following isotopic analysis; CD 3 H

(83%), CD2 H 2 (14%), CH4 (3%). The data seem to indicate a slight proton exchange during the conversion of methanol (and also methylamine and acetate) to methane, and our results imply that the 99 exchange occurs specifically in the methyl coenzyme M reductase re a ctio n .

Synthesis of (R)- and (S)-2-[1-2H i,3H]Propenyl Coenzyme M

Our original strategy for the synthesis of (R)- and (S)- 2 -[1 -

2 H it3 H]propenyl coenzyme M is outlined in Figure 30. Based on literature precedent, the reduction of ethyl acrylate to 2 -[1 -

2 H 2 ]propen- 1 -ol and its subsequent oxidation to acrolein was expected to proceed smoothly. The stereoselective reduction of aldehydes with a-pinanyl-9-borabicyclo[3.3.1]nonane (tx-pinanyl-9-

BBN, Alpine-BoraneR) is well known, and labeled primary alcohols with high optical purity are available by this method. Moreover, both enantiomers of optically pure a-pinene (the starting material for a- pinanyl-9-BBN) are commercially available and the preparation of the tritium-labeled reagent was recently developed in our laboratory112. Finally, the alkylation of coenzyme M with various alkyl halides is a literature procedure for the preparation of alkyl coenzyme M derivatives111. The same procedure should be applicable using an alkyl tosylate in place of the alkyl halide.

Rather than proceeding directly to the synthesis of the chiral, tritium- and deuterium-labeled 2-propenyl coenzyme M, a decision was made to evaluate the synthetic route using dideuterated compounds. This strategy offers the advantage that the regiochemical outcome of the individual steps in the synthesis could be evaluated. Of particular concern was the possibility that the * ° LiAID4, Et20 _ Ag,CO,, CHLOROBENZENE CH2*C H -C ------CH2*C H -C D 2OH ■ M2 » - OCH2CH3

Ethyl Acrylate 2-[1-2H 2]P ro p e n -1 -o l

♦° (♦) - PINANYL-9-BBN-T „.D CH2'CH-C CH2*CH-C -T

n oh Acrolein (R)-2-{1-2H 1t2H ]P ro p e n -1 -o l

1) NaH, THF, 50 °C,15 hr CoMi THF* NH40H ..*D inversion ► CH2« C H -C ^ -T ------c h 2»c h - c - d 2) TsCI, -20 to 0 C OTs 0 - 4 °C, 12 hr NSCH2CH2S03NH4

(R )-2 -[1*2H 1,3H]Propenyl (S )-^1-2H 1,3H ]P ro p e n y l 1-p-Toluenesulfonate Coenzyme M

CoM * HSCH2CH2S 0 3NH4

Figure 30. Original strategy for the synthesis of 2 -propenyl coenzyme M 101 final displacement reaction might proceed, at least in part, by an

S N 1 or by an Sn 2 ' rather than a pure Sn 2 mechanism, leading to scrambling of the isotopic label.

The reduction of ethyl acrylate was carried out using lithium aluminum deuteride in tetrahydrofuran at 25 °C. Following the procedure described by Amundsen and Nelson, excess reducing agent was destroyed with a calculated amount of water followed by the addition of aqueous base113. This procedure served to convert the lithium and aluminum salts to lithium aluminate (UAIO 2 ). The mixture was filtered through glass wool to remove the UAIO 2 and the tetrahydrofuran was removed by careful distillation. The remaining solution was then analyzed by 1H NMR and GC/MS and shown to be a complex mixture containing 2 -[1-2 H 2]pro pe n- 1 -ol (bp

98 °C), [1,1,2,3-2 Hi]propanol (bp 97 °C), ethyl acrylate (bp 99 °C), and ally! acrylate. Allyl acrylate was formed during the removal of the tetrahydrofuran by distillation, presumably by a transesterification reaction of the unreacted ethyl acrylate with the allyl alcohol.

It was apparent that the reaction conditions would have to be altered to control the unwanted reduction of the carbon-carbon double bond and also to minimize the presence of unreacted starting material. It had been reported in the literature that with other a,p- unsaturated carbonyl compounds the reduction of the carbon-carbon double bond might be controlled by switching to a solvent of weaker

Lewis basicity such as diethyl ether. Other changes, such as lower 102 reaction temperature, shorter reaction times and inverse order of addition had also been shown to minimize the reduction of the carbon-carbon double bond. For example, Hochstein and Brown reported that the reduction of cinnamaldehyde using excess UAIH 4 in diethyl ether at 25 °C resulted in an 87% yield of the saturated alcohol. The inverse addition of 1.1 equivalents of UAIH 4 to a solution of cinnamaldehyde in diethyl ether at -10 °C, however, resulted in a 90% yield of the unsaturated alcohol114.

In our case, however, modification of the reaction conditions did not result in a cleaner product. The reduction of ethyl acrylate when carried out with UAID 4 in diethyl ether at -10 °C, still gave a mixture of [ 1 ,1,2,3-2 H]propanol, 2 -[1 -2 H 2]propen- 1 -ol, and starting material, all of which have similar boiling points and could not be easily separated. It began to appear that the reduction of ethyl acrylate was not a practical route to labelled 2 -[1-2 H 2]p ro p e n - 1 -ol.

Concurrently, work was started on the second step of the proposed synthesis, namely the oxidation of 2 -propen- 1 -ol to acrolein. Following a literature procedure, silver carbonate on celite was the first oxidizing agent tried115. This reagent has the advantage that the oxidation is done at 80 °C in the high boiling solvent, chlorobenzene (bp 132 °C), so that the acrolein (bp 59 °C) could be distilled from the reacton mixture as it was formed. The product was condensed in a dry ice/isopropanol trap attached to the reflux condenser. Contrary to the high yields reported by Golfier and 103

Prange, this oxidation procedure gave very low yields of acrolein, primarily due to incomplete oxidation of 2 -propen- 1-ol.

Manganese dioxide (Mn 0 2 ) in toluene, a reagent known to oxidize allylic and benzylic alcohols to the corresponding aldehydes, was then tried. This procedure required large amounts of Mn 0 2 and solvent, therefore, the recovery of acrolein from the reaction mixture necessitated the distillation of a small amount of acrolein out of large quantity of toluene. The acrolein rapidly polymerized upon attempted distillation from the reaction mixture, therefore, this procedure was abandoned. The attempted oxidation of 2 - propen- 1-ol with other oxidizing agents such as pyridinium chlorochromate (PCC) or pyridinium dichromate (PDC) were equally unsuccessful. Having run into serious difficulties early on in this synthetic route, alternate strategies were then considered.

One possible alternative to the reduction/reoxidation sequence would be to reduce the ethyl acrylate directly to acrolein and then continue the synthesis as previously outlined. This would completely eliminate the troublesome oxidation of 2 -propen- 1-ol to acrolein. A reagent that permits the direct reduction of a,p- unsaturated esters to a,p-unsaturated aldehydes is diisobutylaluminum hydride (DIBAL-H). Moreover, this reagent can be prepared with a deuterium label116, so that the direct synthesis of deuterated acrolein might be possible. However, two potential problems dissuaded us from attempting this reaction: The first problem is the rapid polymerization of acrolein which was seen 104 during the attempted oxidations of allyl alcohol, since in this synthesis the acrolein would also have to be removed from the reaction mixture by distillation. The second problem is the tedious preparation of the DIBAL-D which would be required for this route.

The problems associated with the polymerization of the acrolein and the volatility of the intermediates prompted the reevaluation of the original synthetic strategy.

The aforementioned paper by Golfier and Prange 115 also described the preparation of C-1 deuterated acrolein and 2 -[1 -

2 H 2]propen- 1 -ol utilizing some Diels-Alder chemistry reported earlier by Bartlett and Tate117. Based on this paper, a new strategy for the synthesis of stereospecfically labeled and dideuterated 2 - propenyl coenzyme M was developed (Figures 31 & 32). This approach offers the advantages that the carbon-carbon double bond is protected during most of the synthesis and that the easily purified, crystalline intermediates can be synthesized on a large scale due to the ready availability and low cost of the starting m a te ria ls .

The Diels-Alder reaction of ethyl acrylate with anthracene in refluxing nitrobenzene provided ethyl 9,10-ethanoanthracene-11- carboxylate (12) in 50% yield. Reduction of 12 with LiAI2 H 4 afforded the 9,10-ethanoanthracene-11-[13-2H2]methan-13-ol (13) in 50% yield117. The retro Diels-Alder reaction of alcohol 13 w a s carried out by heating 13 in a COOCH2CH3 / H.^C00CH2CH CH NITROBENZENE Li AID. CHj REFLUX ( 4 hr THF , REFLUX 10 hr

1 2

CD2OH

350 °C 1) NaH, THF, 50 °C, 15 hr CH2* C H - C D 2OH 2) TsCI, -20 to 0 °C 1 5

Co M = HSCH2CH2S 03NH4

CoM, THF/NH4OH (1:5) CH2*CH-CD2OTs inversion " CH2*C H -C D 2 0 to 4 °C, 12 hr s c h 2c h 2s o 3n h 4 1 7 1 8

Figure 31. Diels-Alder chemistry for the synthesis of 2-[1- 2H2]propenyl coenzyme M. COOCH2CH3 H.^COOCH,CH CH NITROBENZENE LIAID, + c h 2 REFLUX , 4 hr THF , REFLUX 10 hr

CD2OH p c c / a l 2o 3 (+) - PINANYL-9-BBN-H

BENZENE , Rm Temp

CDHOH ,.D 350 °C 1) NaH, THF, 50 °C, 15 hr , c h 2»c h - c - h \ OH 2) TsCI, -20 to 0 °C 15a

Co M, THF/NH4OH (1:5) c h 2* c h - c - h c h 2«c h - c - d s in v e rs io n 17a OTs 0 to 4 °C, 12 hr 18b n s c h 2c h 2s o 3n h 4

Figure 32. Diels-Alder chemistry for the synthesis of (R)- and (S)- 2-[1-2 Hi]propenyl coenzyme M. 107

pear shaped flask which was connected to a short path distillation

head with a 10 cm trap attached. The trap was cooled in a salt/ice water bath while the reaction flask was heated to 350 °C with a

Woods metal bath. The heating was maintained for 40 minutes during which time the 2 -[ 1-2 H 2]p ro p e n - 1 -ol (15) distilled over115.

This procedure afforded 15 in 82% yield. Comparison of the 1H NMR spectrum of 15 with the 1H NMR spectrum of the unlabeled alcohol showed complete disappearence of the peak at 3.85 ppm (Figure 33).

Thus it can be concluded that the 2 -[1 -2 H 2]pro pe n- 1 -ol from the

retro Diels-Alder reaction is completely deuterated at C-1. No scrambling of the label by allylic rearrangement had occurred during the retro Diels-Alder reaction.

To continue with the synthesis of 2 -[1-2 H 2]propenyl coenzyme M, dideuterated alcohol 15 was converted into 2 -[1-2 H 2]propenyl 1-p- toluenesulfonate (17). Normal workup followed by preparative TLC afforded the unstable tosylate 17 in 30% yield.

The final step in the proposed synthesis of 2 -[1-2 H 2]p ro p e n y l coenzyme M (18) is based on the preparation of alkyl coenzyme M derivatives described earlier. This procedure involved the addition of a tetrahydrofuran solution of the dideuterated tosylate 17 to a solution of 2-mercaptoethanesulfonic acid (coenzyme M) in concentrated, oxygen-free NH 4OH, followed by stirring overnight at

0 to 4 °C. After a minimal workup, the crude reaction mixture, of which 18 was the major component, was dissolved in D 2O and its 1H

NMR spectrum recorded for comparision with that of the unlabeled 1 -Dideuterated Allyl Alcohol in Benzene-d 6

I J U l L

6.00 5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 PPM

Allyl Alcohol in Benzene-d 6

T " 1 'i'' 6.00 5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 PPM 108 Figure 33. 1H NMR spectrum of dideuterated allyl alcohol in b e n z e n e -d 6 from the retro Diels-Alder reaction. 109

reference sample 11 which had been previously prepared from allyl

iodide.

The 1H NMR spectrum of crude 18 clearly showed that the

protons at C-1 of 18 were aol fully deuterated (Figure 34).

Moreover, the 2H NMR spectrum of crude 18 showed that deuterium

was present both at C-1 (3.26 ppm) and C-3 (5.29 ppm, 5.25ppm)

positions and that the ratio of deuterium in the two positions is

about 1:1 (Figure 35). The nucleophilic displacement of dideuterated

tosylate 17 by the sulfur anion, therefore, did not proceed by the

desired Sn 2 mechanism.

The data are suggestive of a Sn 1 mechanism which would involve the formation of an allylic carbocation followed by attack of the

nucleophile at either terminus. This mechanism would result not

only in the scrambling of deuterium between the two methylene groups, but would also yield racemic 2-propenyl coenzyme M when carried out with the chirally labeled tosylate (Figure 36).

Allylic substrates undergo rapid nucleophilic substitution reactions and both the Sn 1 and Sn 2 processes are accelerated compared to saturated substrates. It has been shown that when allylic substrates are reacted with nucleophiles in a polar reaction medium, the reaction proceeds by a Sn1 mechanism. In most reactions, the Sn 1 rates will increase in solvents of increasing polarity and decrease in nonpolar, aprotic solvents. This is due to the ability of the polar solvent to stabilize the carbocation which is formed in the Sn1 reactions. A given reaction, therefore, can Dideuterated 2-Propenyl Coenzyme M

fCJ 6.20 6.00 .60 3.40

Standard 2 -Propenyl Coenzyme M 0 1 1 Figure 34. 1H NMR spectrum of dideuterated 2-propenyl coenzyme M INTEGRAL iue 5 2 NR pcrm f ietrtd -rpnl onye M coenzyme 2-propenyl dideuterated of spectrum NMR 2H 35. Figure PPM R-1.56 SR P / M .162PPM/CM 7.464 HZ/CN F2 DC I5.909P FI IS I H X21.00 CX V13.00 2 CV NC 6B LB 5300.000 20HOP CPD 02 N1300 FN N500 12 ON OR DE NS N8.2 no PN TE RG AO O0.0 VO HZ/PT SN2 N1000.000 SN TO 16384 SI 01 sr sr F 2300.130 SF02 F 46.072 SF2 F46.072 SFO SF TINE DATE A L L V D C O M .2 H * 16.0 4096 1000.000 20-11-87 12:51 627.0 303 400 180.000 40 46.072 3 2.507P £.048 -.400 1.000 0.0 .05 .500 .122 CH2*C H -C D 2SCbM + CD2 = CH-CH2SCoM

c h 2= c h - c d 2 CH, CH-CD2SCoM /OTS SCoM

CD, = CH-CH,SCoM

SCoM

Figure 36. Possible reaction mechanism for the synthesis of 2 - propenyl coenzyme M. 113

proceed by a Sn 1 mechanism in one solvent and by a Sn 2 mechanism

in another.

Several recent literature examples serve to illustrate the

solvent dependence of nucleophilic displacement reactions with allylic substrates. The preparation of dideuteroallyl phenyl sulfide from a 3.3:1 mixture of 3,3- and 1,1-dideuteroallyl bromide with sodium thiophenoxide in methanol afforded the desired product as a

55:45 mixture of 1,1- and 3,3-dideuteroallyl phenyl sulfide118. The same substrate, 1 ,1 -dideuteroallyl bromide, was reacted with excess sodium methiolate in refluxing pentane to afford only 1 ,1 - dideuteroallyl methyl sulfide (Figure 37)119.

H2C-CHCD2Br PhSNa* H2C=CHCD2SPh

+ CH3OH + D2C=CHCH2Br D2C=CHCH2SPh

3.3:1.1 55:45

NaSCH H *C -CHCD2Br H2C-CHCD2SCH3

Figure 37

The regiochemical outcome of these two reactions indicates that the choice of solvent has a pronounced effect on the reaction mechanism. Methanol, which was the solvent that resulted in a scrambling of the deuterium by a Sn 1 process in the first reaction, 114

is several orders of magnitude less polar then NH 4 OH, the solvent which was used in the preparation of 18. The use of pentane, which

resulted in complete retention of the deuterium at C- 1 , allowed the

reaction to proceed by a Sn 2 mechanism.

A second explanation for the results and One which should not be overlooked involves a competing S n 2' reaction. This mechanism would also lead to a scrambling of deuterium into the terminal olefinic protons. To produce the observed equal distribution between C-1 and C-3, the Sn 2 and the SN2'process would fortuitously have to proceed with the same rates. The S n 2' reaction usually occurs when there is a large, bulky substituent at the site of nucleophilic displacement. This type of reaction mechanism can usually be minimized by the use of a smaller leaving group such as methanesulfonate in place of the bulkier p-toluenesulfonate, however, this approach was not investigated (Figure 36).

In order for the proposed synthesis to afford 18 fully deuterated at C-1, the mechanism of the displacement reaction needs to be pure

Sn2. A direct bimolecular displacement mechanism would insure the stereochemical integrity of the chirally labeled product. Based on the results described above, pentane would be the solvent of choice for maximizing Sn2 character in the displacement reaction. The major problem, though, is the insolubility of the 2 - mercaptoethanesulfonic acid (coenzyme M) and 18 in organic solvents. Coenzyme M is only sparingly soluble in methanol and completely insoluble in other organic solvents. Therefore, in order 115 to change the reaction solvent to pentane, two changes would have to be made: The first change would be to use dideuterated allyl bromide, since the ailyl tosylate 17 is not expected to be soluble in pentane. This should pose no synthetic problem, since there is a literature procedure for the synthesis starting from the 2 -[ 1 -

2 H 2 ]propen- 1-ol which is known to afford the allyl bromide with complete deuterium retention at C- 1 120. This reaction proceeds by a single displacement mechanism, thereby preventing racemization at the chiral center due to multiple displacements.

The second change would be to convert coenzyme M to a derivative which would be soluble in nonpolar solvents and capable of undergoing an S-alkylation reaction. This constitutes a major problem due to the nucleophilicity of the sulfur as the free SH in coenzyme M. The disulfide form of coenzyme M could be converted into a derivative which is soluble in organic solvents and reduced before use in the synthesis. However, this would still present problems. It was previously found that in the disulfide form of coenzyme M, the sulfur is still very nucleophilic and the sulfur as well as the oxygen anion were alkylated by diazomethane, resulting in formation of the methyl ester of methyl coenzyme M. Therefore, the sulfur would have to be protected in order to alkylate the sulfonic acid and then deprotected under conditions which the methyl ester would survive. Because of these practical difficulties, the synthesis of 18 by this route was not attempted. 116

The polar aprotic solvent hexamethylphosphoramide (HMPA) is known to minimize carbocation formation and their related rearrangements. Coenzyme M and its derivatives should be soluble in

HMPA, therefore, it might be the solvent of choice for the preparation of 18 without racemization. A major problem with the use of HMPA, however, is its separation from the desired product.

This step usually involves the extraction of HMPA into the aqueous phase, however, 2-propenyl coenzyme M is water soluble and its separation from HMPA would be troublesome.

The 2H NMR spectrum of crude 18 also showed a deuterium peak at 3.60 ppm (Figure 35). Examination of the1H NMR spectrum of crude 18 indicates the presence of a minor product which contains allylic protons [peaks at 5.46 ppm and 5.97 ppm] (Figure 34) which are shifted downfield from those in 2-propenyl coenzyme M. The unknown product also shows two multiplets corresponding to the methylene protons in coenzyme M derivatives which are also shifted downfield ( 3.08 ppm and 3.32 ppm). Wolfe and coworkers have shown that the dimethyl sulfonium salt of methyl coenzyme M affords a 1H NMR spectrum in which the methyl signal and the methylene protons in the coenzyme M portion are shifted downfield from the corresponding signals in methyl coenzyme M34. Therefore, an allyl sulfonium salt could be the minor product which is seen in the 1H NMR spectrum of crude 18 (Figure 38). 117 + H2C= c h c h 2 s c h 2c h 2s o 3n h 4

c h 2c h - c h 2

Figure 38

One would expect to see, however, the required deuterium peak

resulting from deuterium in the olefinic region since formation of the sulfonium salt must proceed through the intermediacy of 2 -

propenyl coenzyme M. This peak is not present in the 2H NMR spectrum, therefore, this explanation is questionable.

The additional signals in the 1H and 2H NMR spectra of crude 18

might be due to 2-propenyl coenzyme M sulfonic acid propenyl ester

(Figure 39), which could arise from the displacement of allyl tosylate 17 by the sulfonate anion

h 2c =c h c h 2 s c h 2c h 2s o 3c h 2c h = c h 2

Figure 39

This explanation would require that the 1H NMR signals for the S- propenyl portion of the minor componet would overlap with those of

2-propenyl coenzyme M. The shifted NMR signals would result from 118

the O-propenyl group and two methylene protons. This explanation would also imply that the sulfonate anion displaces tosylate 17 by a

Sfsj2 mechanism since there is no corresponding deuterium signal in the olefinic region. It is known, however, that sulfonic acid esters are prone to undergo hydrolysis with aqueous bases, therefore, the ester would not be expected to be stable in concentrated NH 4 OH. At this time no further explanation of the reaction mechanism or the possible identification of the minor product can be offered. It became clear, though, that since the reaction mechanism could not be controlled and this might lead to randomization of the label, an alternative synthesis of 18 was necessary.

One way to eliminate the possibility of an allylic rearrangement by either Sn 1 or competing Sn 2 ' reaction mechanisms would be to make the portion containing the allyl group the nucleophilic component in the Sn2 reaction. The current strategy for the synthesis of 2-propenyl coenzyme M, based on this concept, is shown in Figure 40. This route offers the additional advantage that it utilizes several steps from the previous strategy. The synthesis of

9,10-ethanoanthracene-11-[13-2H2]methyl 13-p-toluenesulfonate

(19) from alcohol 13 proceeded in 51% yield. Tosylate 19 is considerably more stable than allyl tosylate 17 and requires no special handling. A suspension of NaH in tetrahydrofuran was cooled to 0 °C and benzyl mercaptan was added dropwise under an argon atmosphere. After 30 minutes, a solution of 19 in tetrahydrofuran was added and the reaction was slowly warmed to room H_C 0 0 CH2CH NITROBENZENE CH LiAID4 + ~ CH REFLUX , 4 hr THF , REFLUX 10 hr

1) NaH, THF, 50 °C

2) TsCI, -20 to 4 °C, 12 hr

1 3 1 9

H.^CD2SCH2C8H HSCH2C6H5, NaH 380 °C inversion CH2*CH-CD2 THF, 50 °C 57 hrs 20 I sch2c 6 h 5 21

1) NH3, Na CH2«CH-CD2 2) BrEs | 2 2 SCH2CH2S03Na 119

Figure 40. New strategy for the synthesis of 2 -[1 -2 H 2]p ro p e n y l coenzyme M utilizing Diels-Alder chemistry. 120

temperature. If the reaction was allowed to proceed for 19 hours at

room temperature, the yield of 13-(benzylthio)-9,10-

ethanoanthracene-11-[13-2H2]methane (20) was only 8.7%. It was found that heating the reaction mixture increased the yields, even though this also increased the risk of a competing elimination

reaction of the tosylate 19. After several trials, it was found that the yield of the desired product 2 0 could be increased significantly

by heating the reaction at 50 °C. For example, heating the reaction for 14.5 hours gave a 34% yield of 20 while heating for 57 hours afforded 20 in a 67% yield.

The retro Diels-Alder reaction of 20 is expected to afford allyl benzyl sulfide 21 [lit.122 bp 79-80 °C at 6 Torr]. Reactions of this type have been used to synthesize not only allyl alcohol, but also acrolein, 1 -m ethoxy- 2 -propene, labeled allyl amines and allylic nitro compounds from the corresponding Diels-Alder adducts. Since it has been determined previously that the retro Diels-Alder reaction of dideuterated alcohol 13 afforded the dideuterated allyl alcohol 15 without allylic rearrangement, the retro Diels-Alder reaction of 2 0 should also proceed without allylic rearrangement.

Due to the high boiling point of 21, the retro Diels-Alder reaction must be done at reduced pressure in order to distill 21 from the reaction mixture. In trial experiments, the formation of 21 in the retro Diels-Alder reaction of 20 has been demonstrated by analysis of the distillate by GC/MS (the mass spectrum gave a molecular ion 121

of m /2. 166 corresponding to 2 1 ); however, the reaction still requires considerable optimization.

The last step of the proposed synthesis involves the cleavage of allyl benzyl sufide 21 to yield the 1-th io -2 -propenyl anion which can participate in a S n 2 displacement reaction with brom oethanesulfonic aGid to afford 2 -[1-2 H 2 ]propenyl coenzyme M

(22). This reaction is analogous to the last step in the synthesis of chirally labeled methionine59. In that reaction,- the sulfur anion formed by the Na/NH 3 reduction of S-benzyl-homocysteine is alkylated with chiral methyl labeled N-methyl-N,N-di-p- toluenesulfonimde to afford the chiral methyl labeled methionine.

It should be noted that there also exists the possibility of an allylic cleavage taking place to afford the thiobenzylate anion.

Literature precedent suggests that the benzylic cleavage is thermodynamically favored over the allylic cleavage but a mixture of products will probably be formed. However, the allylic cleavage would give propene and unlabeled S-benzyl coenzyme M, which should be easily removed by purification. Thus, allylic cleavage might reduce somwhat the yield, but it should not lead to erroneous results. To date, the reductive alkylation reaction of 21 has not been attempted. 122 *»

Stereospecific Reduction of 9,10-Ethanoanthracene-11- [13-2H i]methan-13-al (14.) to (13S)- and (13R)-9,10- Ethanoanthracene-11-[13-2Hi]methan-13-ol ( H a ) and ( H k ) .

The synthesis of (R)- and (S)-2-[1- 2 H-|]- and [1-2Hi 2 H]propenyl coenzyme M (Figure 41) requires the stereospecfic reduction of

9,10-ethanoanthracene-11-[13-2Hi]methan-13-al (14) by Midland’s reagent (Alpine-BoraneR) to yield the (13S)- and (13R)-9,10- ethanoanthracene-11-[13-2Hi]- or [13-2Hi3H]methan-13-ol (13a and 13b respectively). Before proceeding on to the stereospecfic reduction, however, a practical synthesis of aldehyde 14 was needed.

The oxidation of dideuterated alcohol 13 to aldehyde 14 had been reported in the literature 115 using silver carbonate on celite, however, this oxidizing agent gave very low yields of the desired aldehyde and is a very expensive reagent to use on a large scale.

Freshly prepared PCC on aluminum oxide, however, gave complete oxidation of 13 to afford 14 in 75% yield of purified material.

The reduction of 14 with Alpine-BoraneR, however, had not been previously demonstrated and there was some concern over the ability of this reagent to produce the chiral alcohols with high optical purity. The rates of Alpine-BoraneR mediated reductions are substrate dependent. Most aldehydes are rapidly reduced to afford alcohols with high enantiomeric excess121. Substrates which react COOCH2CH3 h _ c o o c h 2c h : C 'H NITROBENZENE LiAID, CH, REFLUX , 4 hr THF , REFLUX 10 hr

PC C /A I20 3 (+) - PINANYL-9-BBN

BENZENE, Rm Temp

1 3

1) NaH, THF, 50 °C HSCH2C6H5, NaH in v e rs io n 2) TsCI, -20 to 4 °C, 12 hr THF, 50 °C 57 hrs

1 3 a 1 9 a

H ..C D H S C H 2C6H5

...H 1) NH3, Na 380 °C c h 2*c h - c - d CH2«CH-C -D 2) BrEs ~ SCH2CH2S 0 3 20b 21b SSCH2C6 Hs 22

Figure 41. Stereospecific synthesis of 2 -[1 -2 H 2]p ro p e n y l 123 coenzyme M. 124

more slowly, due to either electronic or steric effects, provide

alcohols of lower optical purity.

It appears that in slower reactions, the normal cyclic pathway is

replaced by a two-step dehydroboration pathway. This pathway

produces the achiral 9-BBN which actually carries out the reduction

step, resulting in products of lower enantiomeric excess. Aldehyde

14 is rather bulky, therefore, its reduction with Alpine-BoraneR

may result in alcohols with lower chiral purity.

A second concern was the fact that 14 itself is chiral and

present as the racemate. The reagent may discriminate between the

enantiomers and reduce one of them with high selectivity and the

other with lower selectivity. Following the retro Diels-Alder

reaction the result would be alcohols with lower than 1 0 0 %

enantiomeric purity. The magnitude of this effect is not easily

p re d icte d .

To establish if the Alpine-BoraneR mediated reduction is

stereospecific, deuterated aldehyde 14 was reduced with (R)- and

(S)-Alpine-BoraneR to give the corresponding (13S) and (13R)-

alcohols (13a and 13b respectively) in 60% yield. As mentioned

earlier, 14 is a racemic mixture, therefore, reduction of 14 will

result in the formation of a mixture of diasteromers. If the

reduction was 1 0 0 % stereoselective, only two diastereomers would

be formed. The absolute configuration at C-13 (the isotopically chiral center) would be the same in both diastereomers, however, the center at C-1 2 would be racemic. On the other hand, if the 125 reduction was not 100% stereospecific then a mixture of 4 stereoisomers would result.

In order to assign the absolute configuration and chiral purity of

13a and 13b, the retro Diels-Alder reaction was performed on 13a and 13b affording the desired (S)- and (R)-2-[1- 2 Hi]-propen-1-ol

15a and 15b in 8 8 % and 77% yield, respectively. Again, if reduction of aldehyde 14 was 100% stereospecific, then only one enantiomer would result after the retro Diels-Alder reaction and if it was not stereospecific, then the result would be a mixture of the two enantiomers of the allyl alcohol. Using the chirally pure derivatizing agent (S)-(+)-0-acetylmandelic acid, allyl alcohols 15a and 15b were converted into the corresponding (S)-(+)-0- acetylmandelate esters 16a and 16b122(Figure 42).

CH = CH2

Figure 42

This now converted the mixture of the enantiomers (if the reduction was not stereospecific) into a mixture of two diastereomers, the ratio of which will be equal to the ratio of the enantiomers present in the original alcohol, providing no racemization occurred in the 126

derivatization step. This ratio can be determined by integrating the

signals for the diastereotopic methylene protons in the 1H or 2H NMR

spectrum of the two samples. These signals were clearly separated

in the 1H NMR spectrum of the (S)-(+)-0-acetylmandelate ester of

the unlabeled alcohol. Parker has found that in a series of (S)-(+)-0-

acetylmandelate esters of primary alcohols, the NMR signal for the

pro-R hydrogen is shifted upfield from the signal for the pro-S

hydrogen122. Based on Midland's work, it was predicted that the alcohol derived from R-(+)-pinanyl-9-BBN (13a) should have the (S) configuration and that derived from S-(+)-pinanyl-9-BBN (13b) should have the (R) configuration. These predictions were confirmed

using Parker's method. The 1H NMR spectrum of the (S)-(+)-0- acetylmandelate ester of allyl alcohol 15a showed a significant decrease in the signal intensity for the pro-S hydrogen and the complimentary result was obtained for 15b (Figure 43 & 44). From the integration of the 1H and 2H NMR signals it is judged that the reduction proceeded with 8 8 % stereospecificity with both R- and S-

Alpine-BoraneR. It can be concluded from this analysis, that alcohols

15a and 15b can be prepared in 84% ee by this procedure. CHIRAL NANOELATE FROM (R)-ALPINE BORANE OEUTERIUN OECOUPLEO J.N.B.

ill - . ■M j — I— —i— >— i— ■— i— i— i— ■— i— ■— i— ■— i— '— i— ■— i— ■— i— ■— i— ■— i— 1-- 1-- 9.60 9.90 9.40 9.30 9.20 9.10 9.00 4.90 4.BO 4.70 4.60 4.90 4.40 4.30 4 .20 4.10 PPN CHIRAL NANOELATE FRON (SI-ALPINE BORANE OEUTERIUN OECOUPLEO J.N.B.

J llu lL 1-- — I— I-- 1-- 9.60 9.50 9.40 9.30 9.20 9.10 9.00 4.90 4.80 4.70 4.60 4.90 4.40 4.30 4.20 4.10 PPN Figure 43. 1H NMR of the (S)-(+)-0-acetylmandelate ester of (S)- and (R)-2-[1- 2 H-|]propen-1-ol. 128

MAN0SALP.H2 MANORALP.2H OATE 9 -2 -8 8 OATE 9 -2 -8 8 TIME 18:37 TIME 10: 16 SF SF 46 .072 46.072 SFO 48 .072 SFO 46.072 SF02 300.130 SF02 300.130 SF2 4 6 .0 7 2 SF2 46.072 SY 46 0 SY 46 .0 01 67 .2 3 6 01 92.871 SI 4096 SI 4096 TO 4096 _ TO 4096 SM 368.732 SM 322.789 SM2 368.732 SM2 322.789 HZ/PT .180 HZ/PT .158 VO 0.0 VO 0 .0 PM PM 6.0 2 0 .0 RD 0.0 RO 0 .0 AO 9.SS4 AO 6 .3 4 9 RB 400 RG 400 N9 380 NS 639 TE 308 TE 308 OE 1697.0 OE 1938.0 OR 12 OR 12 ON 1398 ON 1949 FM 900 FM 500 02 4800.000 02 4800.000 DP 20H CPO OP 20H CPD LB -.7 0 0 LB -1 .0 0 0 OB .200 OB .150 NC 4 NC 1 CX 3 .0 0 CX 3 .0 0 CY 12.00 CY 12.00 FI 4.424P FI 4.424P F2 4.217P F2 4.216P MI .01 MI .03 DC 1.000 OC 1.000 HZ/CM 3.1 81 HZ/CM 3 .2 0 9 PPM/CM .069 PPM/CM .070 IS 0 IS 0 SR -1 6 8 .1 0 SR -1 1 9 .0 7

4 .4 0 4.30 4 .4 0 4 .3 0 PPM PPM

Figure 44. 2h NMR of the S-(+)-0-acetylmandelate ester of (R)- and (S )-2 0 [1 -2 H-|]propen-1-ol. CONCLUSIONS

In conclusion this work has established a viable approach for determining the steric course of methyl coenzyme M reductase. It has demonstrated that 2-propenyl coenzyme M is a substrate for methyl coenzyme M reductase, producing propene that can be easily converted into acetate for the chirality analysis.

A feasible route was also worked out for the synthesis of chirally labeled 2 -propenyl coenzyme M (Figure 41). The reaction sequence was developed using the dideuterated compounds in order to probe the regiochemical outcome of each step. Having optimized the reactions up to and including the synthesis of 13-(benzylthio)-

9,10-ethanoanthracene-11-[13-2H2]methane (20), the stage is now set to complete the synthesis of 2 -[1 -2 H 2 ]propenyl coenzyme M.

A key step in the synthesis of the chirally labeled compound, the reduction of 9,10-ethanoanthracene-11-[13-2Hi]methan-13-al (14) with (R)- and (S)-Alpine-BoraneR to give (13R)- and (13S)-9,10- ethanoanthracene-11-[13-2Hi]methan-13-ol (13b and 13a) has been shown to be highly stereoselective. For the synthesis of 2-[1-

2 H i(3 H]propenyl coenzyme M, the required reduction will be performed using tritium-labeled (R)- and (S)-Alpine-BoraneR, the preparation of which has previously been worked out in this laboratory112. Using this reagent, the reduction of 14 should afford

129 130

the corresponding labeled alcohols which now contain a deuterium

and tritium at C-13. These alcohols will then be carried on to the

final product as previously outlined.

The dideuterated 2-propenyl coenzyme M will also be used as a

substrate for methyl coenzyme M reductase to obtain additional

information about the mechanism of the reaction. The distribution

of the deuterium label in the C-1 and C-3 positions of propene would

be diagnostic of a particular reaction mechanism, i.e., Sn 1 , S n 2 , or

S n2\

The presence of propene will be determined by GC/MS using

selective ion monitoring to confirm the results. The mass spectral

fragmentation of propene results in two diagnostic peaks; the

molecular ion peak m/Z. 42, and a peak at rn/Z. 27 which represents

the olefinic portion of propene (Figure 45).

CH 2 sC H - C H 3 CH 2 = CH

m /2 . 42 m /2 . 27 Figure 45

By making a direct comparison of these two peaks, the deuterium enrichment of C-1 and C-3 can be determined. This will allow 131 evaluation of the enzymatic reduction reaction prior to the synthesis of (R)- and (S)-2-[1- 2 H-|f3 H]propenyl coenzyme M.

The enzymatic conversion of the chiral 2-propenyl coenzyme M using the cell-free extract of M. barkeri will be performed as previously mentioned. The propene generated from the methyl coenzyme M reductase catalyzed reduction of the chiral (R)- and (S)-

2 -[1 -2H i ,3 H]propenyl coenzyme M will be converted into acetate by a

Lemieux oxidation. This procedure oxidizes the carbon-carbon double bond in propene, affording a mixture of [ 2 H-|,3 H]acetic acid, formic acid, and CO 2 . The mild conditions (Nal 0 4 , catalytic amounts of KMn 0 4 , 25 °C) used in this oxidation should preserve the stereochemical integrity of the methyl group.

Formic acid is a known inhibitor of malate synthase, one of the enzymes used in the chirality analysis of acetic acid. It will be necessary, therefore, to remove any formic acid produced by the oxidation. This can be easily accomplished by further treatment with H2 S 0 4 /H g S0 4 which specifically oxidizes formic acid to CO 2 .

The acetic acid will then be purified by a series of acidic and basic distillations prior to the chirality analysis.

The sucessful completion of these goals will provide for the first time stereochemical information on methyl coenzyme M reductase.

This study could also shed further light on the catalytic role of nickel in cofactor F 4 3 0 . EXPERIMENTAL

Materials and Methods

Materials: All chemicals and solvents used were of the highest grade commercially available and most were used without further purification. Tetrahydrofuran (THF) was distilled from sodium and benzophenone. The organic chemicals were purchased from Aldrich

Chemical Co. and the biochemicals from Sigma Chemical Co. The following deuterated compounds were obtained from Aldrich

Chemical Co. (atom % 2 H): Methyl-d 3 -amine hydrochloride (98), m e th y l-d 3 alcohol-d (99.5), acetic-d 3 acid, sodium salt (99), lithium aluminum deuteride [LiAID 4 ] (98), benzene-d 6 (99.5), and chloroform-d (99.8). Thin layer chromatography (TLC) separations were carried out on pre-coated silica gel plates (.25mm or 2 mm thickness, Brinkmann) with fluorescent indicator or pre-coated cellulose plates (.5mm thickness, Brinkmann). Column chromatography was carried out on silica gel (230-400 mesh, 60A,

Merck) which was used both for flash and gravity columns. Gases used in the cultivation and fermentation of M. barkeri and as analytical standards were purchased from Matheson.

Analytical Methods: Proton and carbon-13 nuclear magnetic resonance (1H NMR and 13C NMR) spectra were recorded on an IBM

300 FT-NMR spectrometer operating on a field strength of 7.1 T. The

132 133

1H NMR data are reported in the following sequence: chemical shift

in ppm (multiplicity, coupling constant in Hertz, number of protons).

The gas chromatography-mass spectrometry (GC/MS) of methane, ethane, and propene was carried out on a Finnigan GC/MS at The Ohio

State Chemical Instrument Center using an Alltech carbosieve S-ll column for the determination of methane and ethane gas,and for propene. Routine GC/MS identification of synthetic intermediates was performed on a Hewlett-Packard 5790A mass selective detector and 5790 gas chromatograph with a 9000-216 data system using

Supelco SPB-5 capillary column. Mass spectral data are reported in the following order: m/z (relative intensities).

2-Propenyl Coenzyme M (11)

A round bottom flask containing 2-mercaptoethanesulfonic acid

(coenzyme M) [1.25 g, 7.6 mmol] and a magnetic stir bar was purged with argon for 20 minutes. Concentrated NH 4 OH (35 mL) was purged with argon for 2 0 minutes and transferred via a double needle to the flask containing coenzyme M. This solution was purged with argon for an additional 5 min and then allyl iodide (1.68 g, 10 mmol) was added via a syringe. The flask was covered with foil and the reaction was vigorously stirred at 0 to 4 °C overnight. The reaction mixture was concentrated in vacuo and dried overnight on a vacuum pump. Mp 226-230 °C; ^H NMR (300 MHz, D 2 0) 5 5.94 (m, 1 H), 5.33- 134

5.25 (m, 2 H), 3.33 (d, J=7.24 Hz, 2 H), 3.22 (m, 2 H), 2.92 (m, 2 H);

13C NMR (75.4 MHz, D 2 0 ) 8 132.9, 116.9, 49.9, 32.9, 23.4.

The same procedure was used for the synthesis of methyl and ethyl coenzyme M, substituting methyl iodide or ethyl iodide for the allyl iodide 1 11

Enzymatic Conversion of 2-Propenyl Coenzyme M (11) to

Propene using a Cell Free Extract of Methanosarcina barkeri

An aqueous solution of 11 was converted enzymatically to propene using a cell free extract of M. barkeri strain 227 which had been grown on methanol. The cell free extract and all substrates were prepared anaerobically and the reaction was performed in a

Wheaton serum bottle (65 mL). The entire crude enzyme extract

(21.6 mg protein, 12 mg/mL) was activated by preincubation with

7.5 mM ATP and 15 mM MgCI 2 in 10 mM TES buffer (adjusted to a pH of 7.2) for 15 minutes at 37 °C under 40 psi H2. The following, 2 mM dithiothreitol and 18 mM 2-propenyl coenzyme M (11) were added anaerobically to give a final reaction volume of 6 mL. The reaction . was allowed to proceed overnight at 37 °C with gentle shaking.

Aliquots (2 mL) from the head space of the serum bottle were removed with a gas tight syringe and the presence of propene was determined by GC/MS using selective ion monitoring; MS mJz. 42 (M+,

70), 41 (100), 39 (75), 38 (18), 37 (16), 27 (32). 135

Ethyl 9,10-Ethanoanthracene-11-carboxylate (12.)

A solution of anthracene (25 g, 140 mmol) and ethyl acrylate

(43.5 g, 435 mmol) in nitrobenzene (400 mL) was vigorously refluxed for 4 h. The solution was concentrated in vacuo at 100 °C until approximately 40 mL remained and the solution was then poured into diethyl ether (40 mL) and placed in a freezer overnight. The resulting solid was filtered and washed with diethyl ether to remove colored impurities. Recrystallization from diethyl ether afforded colorless crystals (19.5 g, 50% yield ) 1 1 7 mp 98 °C (lit.1 1 7 mp 117 0C); 1H NMR (300 MHz, CDCI3) 5 7.33-7.21 (m, 4 H), 7.13-7.04

(m, 4 H), 4.65 (d, J=2.59 Hz, 1 H), 4.33 (t, J=2.70 Hz, 1 H), 4.10-3.94

(m, 2 H), 2.85 (ddd, J=10.2, 4.60, 2.60 Hz, 1 H), 2.19 (ddd, J=12.5,

4.60, 2.70 Hz, 1 H), 1.98 (ddd, J=12.5, 10.3, 2.70 Hz, 1 H), 1.18 (t,

J=7.10 Hz, 3 H); 13C NMR (75.4 MHz, CDCI 3 ) 8 173.4, 144.02, 143.67,

142.52, 140.02, 126.18, 126.10, 125.71, 125.63, 124.78, 123.63,

123.40, 123.23, 60.6, 47.0, 44.1, 43.9, 30.7, 14.25; MS m/z. 278 (M+,

1.9), 178 (100).

9,10-Eth an o anthracene-11-[13-2H2]methan-13-ol (1 31

Linder an argon atmosphere, lithium aluminum deuteride (1 g, 24 mmol) was added to freshly distilled tetrahydrofuran (600 mL) and the suspension was heated to reflux. The Diel-Alder adduct 12 (9.29 g, 33.4 mmol) was added in portions over 1 h to the vigorously stirred, refluxing LiAID 4 suspension and the mixture was refluxed 136

for an additional 12-15 h. After cooling to room temperature, the

reaction flask was further cooled to 0 °C and the mixture was

hydrolyzed with 10% sulfuric acid (100 mL). Tetrahydrofuran was

removed under reduced pressure and the aqueous residue was

extracted with diethyl ether (4 x 150 mL). The combined ether

extracts were washed with sodium bicarbonate (3 x 30 mL), dried

over anhydrous MgS 0 4 and concentrated in vacuo. The crude product was recrystallized from ether/petroleum ether to give a colorless

powder (4.28 g, 50% yield); (TLC on silica gel, methylene chloride/hexane, 9:1, Rf .125)117 mp 105-107 °C (lit .117 mp 1 1 0

0C); 1H NMR (300 MHz, CD3 OD) 5 7.23-6.98 (m, 8 H), 4.31 (d, J=2.30

Hz, 1 H), 4.16 (t, J=2.60 Hz, 1 H), 2.0-1.96 (br m, 1 H), 1.83 (ddd,

J=12.3, 10.5, 2.60 Hz, 1 H), 0.98 (ddd, J=12.3, 4.33, 2.60 Hz, 1 H); 13C

NMR (75.4 MHz, CD 3 OD) 6 145.52, 145.40, 145.31, 141.89, 126.90,

126.62, 126.6, 126.44, 126.40, 124.30, 124.21, 123.87, 65.66, 46.84,

45.42, 42.30, 32.46; MS m/Z_ 238 (M+, .9), 178.(100).

9,10-Eth an o anthracene-11-[13-2Hi]methan-13-al (14)

Pyridinium chlorochromate (PCC) on aluminum oxide was prepared by the following procedure: Chromium trioxide (25 g, 25 mmol) was added to 6 M hydrochloric acid (45 mL). After stirring the solution for 10 min at 40 °C, pyridine (19.8 g, 20.2 mL, 250 mmol) was added to the mixture (not in the opposite order!). The mixture was cooled to 0-4 °C at which time a yellow precipitate 137

formed. Upon rewarming the mixture to 40-55 °C, the precipitate

dissolved and 1 0 0 g of neutral aluminum oxide was added with

vigorous stirring. The water was removed in vacuo at 100 °C for 2 h

and the remaining orange solid was dried overnight on a vacuum

p u m p 124-

In a 500 mL round bottom flask, alcohol 13 (3 g, 12.6 mmol) was

dissolved in benzene (240-300 mL). PCC on aluminum oxide (63 g)

was added and the reaction was stirred for 5 h at room temperature.

The reaction slurry was filtered through a 3" cake of Florisil in a

sintered glass funnel and the Florisil was washed with methylene

chloride (300 mL). The filtrate was concentrated in vacuo and

purified by column chromatography (300 g silica gel, methylene

chloride/hexane, 9:1). Evaporation of solvent gave a colorless oil

which yielded a white solid after drying with a vacuum pump (2.3 g,

75% yield); (TLC on silica gel, methylene chloride/hexane, 9:1,

Rf=.55); mp 92-94 °C (lit.115 mp 97.5 0C); 1H NMR (300 MHz, CDCI3),S

7.33-7.23 (m, 4 H), 7.15-7.05 (m, 4H), 4.67 (d, J=2.52 Hz, 1H), 4.38

(t, J=2.68 Hz, 1H), 2.75 (m, 1H), 2.10 (m, 1H), 1.97 (m, 1H); MS m/z

235 (M+, 4.4), 178 (100).

(13S)-9,10-Ethan o anthracene-11-[13-2H1]m ethan-13-ol (13a) A solution of 14 (2.3 g, 9.8 mmol) in methylene chloride was transferred to a dry 50 mL 3 neck flask closed with a septum. The 138

solution was concentrated and the resulting oil was dried with a

vacuum pump for 2-3 h. Under an argon atmosphere, R-(+)-pinanyl-

9-borabicyclo[3.3.1]nonane (R-Alpine-BoraneR) [25 mL of .5M

solution in THF, 12.5 mmol] was added and the reaction was stirred

for 12-15 h at room temperature and then refluxed for 1 h. After cooling to room temperature, acetaldehyde (.5 mL) was added and

stirring was continued for 15 min. The solution was concentrated in vacuo and the pinene was removed by drying the reaction mixture on a vacuum pump at 40 °C overnight. After placing the flask under an argon atmosphere, diethyl ether (15 mL) was added with stirring and the solution was then cooled to 0 °C. Ethanolamine (.885 mL) was added and the mixture was stirred for 30 min. The mixture was diluted with ether (25 mL), then washed with water (2 x 30 mL), and the aqueous washes back-extracted with ether (3 x 50 mL). The combined organic layers were washed with brine, dried over anhydrous MgS 0 4 , and concentrated in vacuo. Column chromatography (200 g silica gel, methylene chloride/hexane, 9:1) afforded the desired product as a colorless solid (1.4 g, 60% yield);

MS m/i 237 (M+, .7), 178 (100).

(13R)-9,10-Ethanoanthracene-11 -[13-2H 1 ]methan-13-ol (13b) was prepared using the same procedure with S-Alpine-BoraneR. 139

(S)-2-[1-2H-i]Propen-1-ol (15a) and (R)-2-[1-2Hi]Propen- 1-ol (1 5 b )

Alcohol 13a (2.8 g, 11.8 mmol) was placed in a 100 mL pear shaped flask which was then connected to a short path distillation head with a 10 cm trap attached. The trap was cooled in a salt/ice water bath while the pear shaped flask was heated to 350 °C in a

Woods metal bath. The product distilled over as the reaction proceeded and heating was maintained for 30 to 40 min115.

(R)-2-[1-2 H-|]propen-1-ol (15b) was prepared by the same method from alcohol 13b (2.3 g, 9.7 mmol). The (S)- and (R)-allyl alcohol samples were recovered from the trap in yields of 8 8 % and

77%, respectively.

(S)-2-[1-2H-|]Propenyl (S)-(+)-0-Acetylmandelate (16a)

A solution of (S)-(+)-0-acetylmandelic acid (194 mg, 1 mmol) and 4-dimethylaminopyridine (5 mg) in methylene chloride (10 mL) was cooled to -10 °C. With stirring, 15a (87 mg, .106 mL, 1.5 mmol) and dicyclohexylcarbodiimide (206 mg, 1 mmol) were added and the reaction was allowed to warm to room temperature over 6 h.

After filtration of the precipitated urea, methylene chloride was removed in vacuo and the product was purified by preparative TLC

(silica gel, 2 mm preparative plates, ethyl acetate/hexane, 1:2, Rf

.36) to give a colorless liquid (90 mg, 38% yield ) 122 1H NMR (300

MHz, C6 D6) 8 7.46-7.41 (m, 2 H), 7.09-6.8 (m, 3 H), 6.07 (s, 1 H), 5.50 140

(ddd, J=17.30, 10.58, 5.57 Hz, 1 H), 4.91 (ddd J=17.30, 1.49, 1.44 Hz,

1 H), 4.81 (ddd, J=10.58, 1.44, 1.42 Hz, 1 H), 4.29 (m, 1 H), 1.73 (s, 3

H); 2H NMR (46 MHz, C 6 D6) 8 4.33 (.92 2 H), 4.29 (.08 2H); MS m/i 235

(M+, 4.4), 176 (8.0), 149 (18), 107 (63), 79 (16), 77 (14), 51 (6.3), 43

(100).

( R ) - 2 - [1 - 2 Hi]Propenyl (S)-(+)-0-Acetylmandelate ( 1 6 b l

This was prepared by the same procedure, however, using alcohol

15b in place of 15a; 1H NMR (300 MHz, C 6 D6) 8 7.46-7.41 (m, 2 H),

7.09-6.8 (m, 3 H), 6.07 (s, 1 H), 5.50 (ddd, J=17.30, 10.58, 5.35 Hz, 1

H), 4.91 (ddd, J=17.30, 1.51, 1.44 Hz, 1 H), 4.81 (ddd, 10.58, 1.44,

1.44 Hz, 1 H), 4.34 (m, 1 H), 1.73 (s, 3 H); 2h NMR (46 MHz, C 6 D6) 8

4.34 (.08 2H), 4.30 (.92 2H).

2-[1-2H2]Propen-1-ol (15)

Using the procedure previously described, alcohol 13 (3 g, 12.6 mmol) afforded 15 in 82% yield; 1H NMR (300 MHz, C 6 D 6 ) 8 5.79 (dd,

J=17.6, 11.02 Hz, 1 H), 5.16 (dd, J=17.6, 1.80 Hz, 1 H), 4.98 (dd,

J=11.02, 1.80 Hz, 1 H), 1.39 (s, 1 H).

2-[1-2H2]Propenyl p-toluenesulfonate (171

To a suspension of NaH (87 mg, 60% suspension in oil, 2.1 mmol) in freshly distilled tetrahydrofuran (5 mL), 15 (116 mg, .136 mL, 2 mmol) was added under an argon atmosphere and the reaction was 141 stirred for 15 h at 50 °C. The reaction mixture was cooled to -78 °C and a solution of p-toluenesulfonyl chloride (381 mg, 2 mmol) in dry tetrahydrofuran (2 mL) was added dropwise. The reaction was kept at -10 to -30 °C for 1 h and for 2.5 h at 0 to 4 °C. Water (5 mL) was added to quench the excess NaH and then the mixture was extracted with CHCI3 (3x5 mL). The combined organic extracts were dried over anhydrous MgS 0 4 and concentrated in vacuo. The product was purified by TLC (silica gel, 2 mm preparative plates, hexane/ethyl acetate, 9:1, Rf .3) affording the desired tosylate 17 as a colorless oil (127 mg, 30% yield); MS mJz. 214 (M+, 3.3), 155 (31), 91 (1 0 0 ), 65

(36), 59 (8.5), 43 (20), 39 (21).

2-[1-2H2]Propenyl Coenzyme M (1_g.)

A solution of concentrated NH 4 OH and a flask containing 2- mercaptoethanesulfonic acid (coenzyme M) [75 mg, .456 mmol] were purged with argon for 15 min. While still under argon, the NH 4 O H

(5mL) was added via a syringe to the flask containing the coenzyme

M. In a separate flask, a solution of freshly distilled tetrahydrofuran (1 mL) and 17 (127 mg, .593 mmol) was stirred and purged with argon for 15 min. The NbUOH/coenzyme M mixture was added via a syringe to the THF solution of 17 under argon, and the reaction was stirred overnight at 0 to 4 °C. The solution was concentrated to dryness, dissolved in a minimal amount of methanol/water (1:1) and applied to a XAD-2 column (11 cm x 1.5 142

cm). The column was eluted with methanol/water (1:1) and the

eluent (80 mL) was concentrated in vacuo to afford a colorless solid

which was used without further purification.

9,10-Ethan o anthracene-11-[13-2H 2 ] methyl-13-p- toluenesulfonate (191

To a suspension of NaH (1 g, 60% suspension in oil, 26.5 mmol) in

freshly distilled tetrahydrofuran (50 mL), 13 (6 g, 25.2 mmol) was

added under an argon atmosphere and the reaction was stirred for 7

h at 50 °C. The solution was cooled to -78 °C and a solution of p- toluenesulfonyl chloride (4.7 g, 24.6 mmol) in dry tetrahydrofuran

(8 mL) was added dropwise. The reaction was kept at - 78 °C for 1 h

and then at 0 to 4 °C overnight. Water (25 mL) was added to quench excess NaH and the mixture was extracted with diethyl ether (3 x 50

mL). The combined ether extracts were dried over anhydrous MgS 0 4 , and concentrated in vacuo. Column chromatography (350 g silica gel,

methylene chloride/hexane, 3:2) afforded the desired tosylate 19 as a colorless oil (5 g, 51% yield);(methylene chloride/hexane, 3:2, Rf

.38); MS mJZ. 393 (M+, .2), 178 (100), 91 (16).

13-(Benzylthio)-9,10-Ethanoanthracene-11-[13-

2H 2 ]methane (201

A suspension of NaH (225 mg, 60% suspension in oil, 5.63 mmol) in freshly distilled tetrahydrofuran (15 mL) was cooled to 0 °C and 143 benzyl mercaptan (692mg, .657 mL, 5.63mmol) was added dropwise via a syringe under an argon atmosphere. The solution was stirred for 30 min at room temperature and then cooled to 0 °C. A solution of 19 (1.5 g, 3.8 mmol) in tetrahydrofuran (15 mL) was then added at

0 °C and the reaction was slowly warmed to room temperature and stirred at 50 °C for 29 h. Water (20 mL) was added to quench the excess NaH and the mixture was then extracted with ether (3 x 50 mL) . The combined organic extracts were washed with 45% potassium hydroxide (1 x 50 mL), water (2 x 50 mL), dried over anhydrous MgS 0 4 , and concentrated in vacuo. The crude product was purified by sucessive chromatography on two silica gel columns

( 2 0 0 g each), the first being eluted with methylene chloride/hexane

(3:2) to separate the desired product from unreacted starting material, the second being eluted with methylene chloride/hexane

(1:4) to separate 20 from any disulfide formed from the benzyl mercaptan during work up. This procedure gave a colorless oil (712 mg, 54% yield); (TLC on silica gel, methylene chloride/hexane, 3:2,

Rf .9; methylene chloride/hexane, 1:4, Rf .17); 1H NMR (300 MHz,

CDCI3) 5 7.29-7.16 (m, 9 H), 7.11-7.02 (m, 4 H), 4.32 (d, J=2.06 Hz, 1

H), 4.2 (t, J=2.80 Hz, 1 H), 3.63 (s, 2 H) 2.14-2.05 (br m, 1 H), 2.0

(ddd, J=11.2, 4.5, 2 93 Hz, 1 H), 1.16 (ddd, J=11.2, 9.70, 2.93 Hz, 1 H);

2H NMR (46 MHz, CDCI 3 ) 8 2.09 (1 2H), 2.02 (1 2H); 13q NMR (75.4 MHz,

CDCI3 ) 8 143.92, 143.76, 143.55, 140.15, 138.50, 128.78, 128.45,

125.89, 125.93, 125.72, 125.63, 125.54, 123.52, 123.35, 123.02, 144

47.70, 44.32, 37.52, 36.88, 36.27, 34.97; MS mJz. 344 (M+, 2.4), 178

(100), 91 (7.3). APPENDIX CONFIGURATIONAL ANALYSIS OF CHIRAL METHYL GROUPS

On a conceptual level the chirality analysis is achieved by

replacing one of the methyl hydrogens in acetate with a another

group. The replacement is done by an irreversible reaction of known

steric course which involves a significant primary kinetic isotope

effect in the hydrogen abstraction. It leaves the remaining two

hydrogens in a stereoheterotopic relationship, allowing their

discrimination by another, stereospecific hydrogen abstraction

reaction of known steric course. The isotope effect in the first

reaction causes an unequal distribution of the remaining tritium

between the two methylene hydrogens of the product, and this

tritium distribution reports the configuration and chiral purity of

the methyl group.

In practice, this transformation involves a series of enzymatic

reactions of known stereospecificity which convert acetate first

into malate and then into an equilibrium mixture of fumarate and

malate18*19. The sample of chiral acetate is mixed with a known amount of [ 1 4 C]acetate to adjust the 3 H/14C ratio to approximately

4. The acetate is then converted enzymatically to acetyl coenzyme

A using acetate kinase, ATP, coenzyme A, and phosphotransacetylase. The acetyl coenzyme A is condensed with

145 146 glyoxylate in an irreversible reaction catalyzed by malate synthase.

If, for example we started with chirally pure (S)-acetate (see

Figure 46) and the isotope effect of the malate synthase reaction were infinitely large, only the species of malate shown in the box would result from the enzymatic reaction.

The isotope effect, however, is not infinitely large; it was found to be (kn/ko) 3.7-3.8. Therefore, three species result from the reaction, one in which the proton, one in which the deuterium, and one in which the tritium is removed in the condensation with glyoxylate. Since the final measurement is the analysis for tritium radioactivity, the species which does not contain tritium will be transparent to the chirality analysis and can be ignored. The malate species which results from the removal of the hydrogen, shown in the box, is the major product. The malate is purified by ion exchange chromatography, concentrated in vacuo, and the 3 h / 14C ratio is measured. The malate is then incubated with fumarase. The fumarase reaction involves an anti dehydration/hydration in which the pro-3R hydrogen of S-malate is stereospecifically removed and/or equilbrated with solvent protons. This reaction affords an equilibrium mixture of malate and fumarate in which the pro-3S hydrogen remains carbon-bound and the pro-3R hydrogen is equilibrated with water. The fumarate and malate shown in the box result from the loss of tritium which was in the pro-3R position.

After a series of lyophilizations to remove the water from the fumarase reaction mixture, the 3 H/14C ratio of the residue is Tv^COjH ACETATE KINASE.ATP H MALATE PHOSPHOTRANSACETYLASE, H SYNTHASE «T COASH p JLOH Dr k Kh /Ko = 3.8 ho2c ^ h T C09H COSCoA <50% ♦ ho2c , >0 H ... OH

>50%

D ^ C O jH d COjH FUMARASE T C02H T CO9H ♦ Hj HTO "to. • r L OH 50% -H 20 H02c h ho2c h H0*C H H02C' H

>50% <50% T/C FUMARASE RXN Fs50RACEMIC F = X100 F = 79(R) T/C MALATE F=21 (S)

150-F | (ENANTIOMERICEXCESS)% e * ~ '------x 100 ; EXPTLERROR +2FVALUES = + 7%e.e. 29

Figure 46. Configurational analysis of chiral methyl groups. 148

measured. The percent tritium retention in the fumarase reaction,

which is related to the configuration and chiral purity of the

acetate, is referred to as the F value. The F value can be calculated

by dividing the 3 H/14C ratio of the fumarase reaction residue by the

3 H/14C ratio of the malate multiplied by 100. Based on the isotope

effect of the malate synthase reaction and the known

stereospecificity of the enzymatic reactions, chirally pure (S)-

acetate gives an F value of 21 and chirally pure (R)-acetate gives an

F value of 79. These values are based on the work of Eggerer and coworkers who have shown that the amplitude of the analysis is 50

+ 29. Therefore, a racemic mixture of acetate will result in a

F value of 50. The F value can be converted into percent enantiomeric excess (% e.e.) by the following equation: % e.e. =( | F-

50 | divided by 29 ) times 100.

The reproducibility of the analysis in our laboratory is +2 F values or +7% e.e. The amount of acetate which can be converted into malate is limited to 10 umol per enzyme incubation. Sucessful analyses have been performed with as little radioactivity as 2 0 0 0 dpm 3 H/10 umol acetate, although it is better to have at least 104-

105 dpm 3 H/10 umol. The experimental details of the chirality analysis carried out in our laboratory are described in the following reference125-126. LIST OF REFERENCES

1 Fontaine, E. F.; Peterson, W. H.; McCoy, E.; Johnson, M. J.; Ritter, G. J._J„.Bacterial-1942,42, 701.

2. Barker, H. A.; Kamen, M. D. Proc. Natl. Acad. Sci. U.S.A. 1945, 21 219.

3. Wood, H. G. J. Biol. Chem. 1952, 1 2 &. 579.

4. Poston, J. M.; Kuratomi, K.; Stadtman, E. R. Ann. N.Y. Acad. Sci. 1964, 112. 804.

5. Ljungdahl, L.; Irion, E.; Wood, H. G. Biochemistry 1965. 4. 2771.

6 . Parker, D. J.; Wu, T. F.; Wood, H. G. J. of Bacteriol. 1971, 143. 471.

7. Ghambeer, R. K.; Wood, H. G.; Schulman, M.; Ljungdahl, L. Arch. Biochem. Biophys. 1971. 143. 471.

8 . Ljungdahl, L. G.; Wood, H. G. in "Vitamin B 12" Dolphin, D., Ed.; John Wiley and Sons: New York, N. Y.,1982; pp 166-202.

9. Clark, J. E.; Ljungdahl, L. G. J. Biol. Chem. 1984, 259. 10845.

10. Hu. S. I.; Drake, H. L.; Wood, H. G. J. Bacteriol. 1982, 149. 440.

11. Diekert, G. B.; Thauer, R. K. J. Bacteriol. 1978. 136. 597.

12. Ragsdale, S. W.; Clark, J. E.; Ljungdahl, L. G.; Lundie, L. L.; Drake, H. L. J. Biol. Chem . 1983, 2 5 2 . 2364.

13. Ragsdale, S. W.; Ljungdahl, L. G.; Der Vartanian, D. V. J. Bacteriol. 19 83 . 1 5 5 . 1224.

149 150

14. Ragsdale, S. W.; Wood, H. G. J. Biol. Chem. 1985, 200., 3970.

*15. Pezacka, E.; Wood, H. G. Proc. Natl. Acad. Sci. U.S.A. 1984. fil, 6261.

16. Ragsdale, S. W.; Wood, H. G.; Antholine, W. E. Proc. Natl. Acad. Sci. U.S.A. 1985.82. 6811.

17. Kohara, T.; Kamiya, S.; Yamamoto, T.; Yamamoto, A. Chem. Lett. 1979, 1513.

18. Luthy, J; Retey, J.; Arigoni, D. Nature (London) 1969. 221. 1213.

19. Cornforth, J. W.; Redmond, J. W.; Eggerer, H.; Buckel, W.; Gutschow, C.E. Eur. J. Biochem. 1970, l i 1 .

20. Arigoni, D. Ciba FP.Und, Symp. 1978, 243.

21. Floss, H. G. ACS Svmp. Ser. 1982, 185. pp 229-243.

22. Collman, J.; Hegedus, L. in "Principles and Applications of Organotransmission Metal Chemistry"; University Science Books: Mill Valley, Ca.

23. Kobayashi, K.;Jadhav, P. K.; Zydowsky, T. M.; Floss, H. G. J. .Qrq. Chem. 1983, 3510.

24. Raybuck, S.; Bastein, N.; Zydowsky, L.; Kobayashi, K.; Floss, H. G.; Orme-Johnson, W. H.; Walsh, C. J. Am. Chem. Soc. 1987, 109. 31 73 .

25. Zydowsky, T. M.; Courtney, L. F.; Frasca, V.; Kobayashi, K.; Shimizu, H.; Yuen, L. D.; Matthews, R. G.; Benkovic, S. J.; Floss, H.G. J. Am.Chem Soc. 1986, 1Q&, 3152.

26. Lebertz, H.; Simon, H.; Courtney, L. F.; Benkovic, S. J.; Zydowsky, L. D.; Lee, K.; Floss, H. G. J. Am. Chem. Soc. 1987, 109. 3173. 151

27. Bock, P. L.; Boschetto, D. J.; Rasmussen, J. R.; Demers, J. P.; Whitesides, G. M. J. Am. Chem. Soc. 1974, 2814.

28. Drake, H. L.; Hu, S. I.; Wood, H. G. J. Biol. Chem . 1981, 256. 11137.

29. Barker, H. A. in "Bacterial Fermentations"; John Wiley and Sons: New York, NY, 1956, pp. 1-27.

30. Blaylock, B. A.; Stadtman, T. C. Biochem. Biophvs. Res. Commun. 1963.11, 34.

31. Blaylock, B. A.; Stadtman, T. C. Biochem. Biophvs. Res. Commun. 1964.12, 475.

32. Blaylock, B. A. Arch. Biochem. Biophvs. 1968.124. 314.

33. McBribe, B. G.; Wolfe, R. S. Biochemistry 1971. 1 0 . 2317.

34. Taylor, C. D.; Wolfe, R. S. J. Biol. Chem. 1974. £42, 4879.

35. Romesser, J. A.; Wolfe, R. S. J. Bacteriol. 1982. 152. 840.

36. Leigh, J. A.; Wolfe, R. S. J. Biol. Chem. 1983, 258. 7536.

37. Keljens, J. T.; Daniels, L. Jannsen, H, G,; Borm, R. J. ; Vogels G. D. Eur. J. Biochem. 1983, 130. 545.

38. van Beelen, Ph.D. Dissertation., University of Nijmegen, 1984.

39. Leigh, J. A.; Rinehart, K. L.; Wolfe, R. S. J. Am. Chem. Soc. 1984, 106. 3636.

40. Leigh, J. A.; Rinehart, K. L.; Wolfe, R. S. Biochemistry. 1985. 24 995.

41. Escalante-Semerena, J. C.; Rinehart, K. L.; Wolfe, R. S. J. Biol. Chem. 1984.259. 9447. 152

42. van der Meijden, P.; Heythuysen, H. J.; Sliepenbeek, H. T.; Houwen, F. P.;van der Drift, C.; Vogels, G. D. J Bacteriol. 1983. 153. 6 .

43. Nauman, E.; Fahlbusch, K.; Gottschalk, G Arch. Microbiol. 1984.138. 79.

44. Lovely, D. R.; White, R. H.; Ferry, J. G. J Bacteriol. 1984. 160. 521.

45. Krzycki, J. A.; Lehman, L. J; Zeikus, J. G. J. Bacteriol. 1985. 1 6 3 . 1000.

46. Eikmann, B.; Thauer, R. K. Arch. Microbiol. 1984, 13 8 . 365.

47. Laufer, K.; Eikmanns, B.; Frimmer, U.; Thauer, R. K. Biosciences 1987. 42. 3 6 1 .

48. Hook, L. A.; Corder, R. E.; Hamilton, P. T.; Frea, J. I.; Reeve, J. N. in "Microbial Chemoautrophy"; Strohl, W. R., and Tuovinen, O. H., Eds.; The Ohio State University: Columbus, OH, 1984; pp 275- 289.

49. Nauman, E. Ph.D. Dissertation, Universitat Gottingen, 1983.

50. Walther, R.; Fahlbusch, K.; Sieved, R.; Gottschalk, G.. J,^Bac.tefjQl 1981, 1 4 J L 371.

51. Heil, A.; Zillig, W. FEBS Lett. 1970, H , 165.

52. Austin, A. T. Nature 1960. 186. 1086.

53. Smith, R. H.; Koepke, S. R.; Tondeur, Y., Denlinger, C. L.; Michejda, C. J. J. Chem. Soc. Chem. Comm. 1985, 936.

54. Kobayashi, K.; Shen, S. J. unpublished data.

55. Spratt, T.; Zydowsky, L. D. unpublished data.

56. Shen, S. J.; Ph.D. Dissertation, Purdue University, 1983. 153

57. DeChristopher, P.J.; Lyon, G. D.; Admek, J. P.; Swedo, R. J.; Klein, S. A.; Baumgarten, R. J. 161st Nat'l Meeting A.C.S. 1971.

58. Anderson, N. H.; Uh, H. S. Svn Comm. 1972. 297.

59. Woodard, R. W.; Mascaro, L.; Horhammer, R.; Eisenstein, S.; Floss, H. G. J. Am. Chem. Soc. 1980 1 2 2 , 6314.

60. Curtis, V. A.; Schwartz, H. S.; Hartman, A. F.; Pick, R. M.; Kolar, L. W.; Baumgarten, R. J. Tett Lett 1977. 1969.

61. The (R)- and (S)-[ 2 H i, 3 H]acetic acid was synthesized by modifications of a published procedure 23 by Dr. Thomas Zydowsky and had a chiral purity of 99% ee R and 95% ee S.

62. Arigoni, D.; Lynen, F.; Retey, J. Helv. Chim. Acta. 1966. 49. 311.

63. Shaw, J. E.; Kunerth, D. C.; Sherry, J. J. Tett Lett. 1973. 689.

64. Liotta, C. L.; Harris, H. P.;McDermott, M.; Gonzalez, T.; Smith, K. Tett Lett. 1974. 2417.

65. Liotta, C. L.; Grisdale, E. E. Tett Lett. 1975, 4205.

6 6 . Muller, P.; Seigfried, B. Helv. Chim. Acta. 1974. 57. 987.

67. Muller, P.; Seigfried, B. Tett Lett. 1 9 73 , 3565.

6 8 . Elisinger, F.; Schreiber, J.; Eschenmoser, A. Helv. Chim. Acta. 1960, 42^ 113.

69. Kito, T.; Yamamoto, K.; Hirao, I. Bull. Chem. Soc. Japan 1977. £2, 706.

70. Glass, R. S.; Swedo, R. J. J. Qrq. Chem. 1978, HL 2291.

71. Glass, R. S. J, Chem. Soc. Chem. Comm. 1971, 1546.

72. Townsend, C. A.; Theis, A. B.; J. Org. Chem. 1980, 4JL. 1697. 154

73. Olson, A.R.; Miller, R. J. J. Am. Chem. Soc. 1938. 60. 2690.

74. Barclay, L. R.; Hall, N. D.; Cooke, G. A. Can. J. Chem. 1962, 40. 1981.

75. Gautier, A. E. Ph.D Dissertation, No. 6583, E. T. H., 1980.

76. Shapiro, S.; Wolfe, R. S. J Bacteriol. 1980. 141. 728.

77. Zydowsky, L. D.; Zydowsky, T. M.; Haas, E. S.; Brown, J. W.; Reeve, J. N.; Floss, H. G. J. Am. Chem. Soc. 1987. 109. 7922.

78. Virchaux, P. N. H. Ph.D Dissertation, No. 6739, E. T. H., 1981.

79. Asano, Y.; Woodard, R. W.; Houck, D. R., Floss, H. G. Arch. Biochem. Bioohvs. 1984. 231. 253.

80. Spratt, T.; Zydowsky, T.M. unpublished data.

81. Nagle, D. P.; Wolfe, R. S. Proc. Natl. Acad. Sci. U. S. A. 1983, 80, 2151.

82. Rouvier, P.; Wolfe, R. S. J. Bacteriol. 1983. 152. 840,

83. Ellefson, W. L.; Whitman, W. B. ; Wolfe, R. S. Proc. Natl. Acad. Sci. U. S. A. 1982.79. 3707.

84. Hartzell, P. L.; Wolfe, R. S. Proc. Natl. Acad. Sci. U. S. A. 1986. fi3L 6726.

85. Noll, K.; Rinehart, K. L.; Tanner, R. S.; Wolfe, R. S. Proc. Natl. Acad. Sci. U. S. A. . 1986, & L 4238.

8 6 . Ankel-Fuchs, D. ; Bocher, R.; Thauer, R. K.; Noll, K. M. Wolfe, R. S. EEBS -Lett-1987, m . 123.

87. Kobet, A.; Pfaltz, A.; Ankel-Fuchs, D.; Thauer, R. K. FEBS Lett. 1987, 214. 265. 155

8 8 . Nagel, D. P.; Wolfe, R. S. Proc. Natl. Acad. Sci. U. S. A. 1983. 80. 2151.

89. Ellefson, W. L. ; Wolfe, R. S. J. Biol. Chem. 1981, 256. 4259.

90. Keltjens, J. T.; Whitman, W. B.; Caerteling, C. G.; van Kooten, A. M.; Wolfe, R. S.; Vogels, G. D. Biochem. Biophys. Res. Commun. 1982, 108. 485.

91. Ankel-Fuchs,D.; Jaenchen, R.; Gebhardt, N. A.; Thauer, R. Arch. Microbiol. 19 8 4 , 3,3.9... 332.

92. Huster, R.; Gilles, H. H.; Thauer, R. K. Eur. J. Biochem. 1985, 148. 107.

93. Ankel-Fuchs, D.; Thauer, R. K. Eur. J. Biochem. 1986. 156. 171.

94. Pfaltz, A.; Juan, B.; Fassler, A.; Eschenmoser, A.; Jaenchen, R.; Gilles, H. H.; Diekert, G.; Thauer, R. Helv. Chim. Acta. 1982. 65. 828.

95. Pfaltz, A.; Livingston, D. A.; Juan, B.; Diekert, G.; Thauer, R.; Eschenmoser, A. Helv. Chim. Acta. 1985, 1338.

96. Kratky, C.; Fassler, A.; Pfaltz, A.; Krautler, B.; Juan, B.; Eschenmoser, A. J. Chem. Soc. Chem. Commun.. 1984. 1368.

97. Waditschatka, R.; Kratky, C.; Juan, B.; Heinzer, J.; Eschenmoser, A. J. Chem. Soc. Chem. Commun.. 1985, 22, 1604.

98. Diakum, G. P.; Piggott, B.; Timton, H. J.; Ankel-Fuchs, D.; Thauer, R. K. Biochem. J. 1985. 232, 281.

99. Eidsness, M. K.; Sullivan, R. J.; Schwartz, J. R.; Hartzell, P. L.; Wolfe, R. S.; Flank, A. M.; Cramer, S. P.; Scott, R. A. J. Am. Chem. Soc. 1986. 108. 3120.

100. Shiemke, A. K.; Scott, R. A. J. Am. Chem. Soc. 1988. 110. 1645. 156

101. Albracht, S. P. J.; Ankel-Fichs, D.; Van der Zwaan, J. W.; Fontijn, R. D.; Thauer, R. K. Biochim. Biophvs. Acta. 1986. 50.

102. Juan, B.; Pfaltz, A. J. Chem. Soc. Chem. Commun. 1986. 1327.

103. Juan, B.; Pfaltz, A. J. Chem. Soc. Chem. Commun. 1988. 293.

104. D'Aniello, J. J.; Barefield, E. K. J. Am. Chem. Soc. 1976. 98. 1610.

105. Taylor, R. T.; Weissbach, H. Arch. Biochem. Biophvs. 1969. 129. 728.

106. Walsh, C. T.; Orme-Johnson, W. H. Biochemistry. 1987. 26. 4901.

107. Ellermen, J.; Kobelt, A.; Pfaltz, A.; Thauer, R. K. FEBS Lett. 1987,225., 358.

108. Keltjens, J. T.; Whitman, W. B.; Caertling, S. G.; Van Kooten, A. M.; Wolfe, R. S.; Vogels, G. D. Biochem. Biophvs. Res. Commun. 1985,15&, 495.

109. Wackett, L. P.; Honek, J. F.; Begley, T. P.; Wallace, V.; Orme-Johnson, W. H.; Walsh, C. T. Biochemistry 1987. 26. 6012.

110. Jolly, P. W.; Wilke, G. in "The Organic Chemistry of Nickel"; Maitlis, P. M.; Stone, F. G. A.; West, R Eds; Academic, New York, 1974; pp 139-243.

111. Gunsalus, , R. P.; Romesser, J. A.;Wolfe, R. S. Biochemistry. 1978,1Z, 2374

112. Kim, S. U.; Floss, H. G. unpublished results.

113. Amundsen, L. A.; Nelson, L. S J. Am. Chem. Soc. 1950. 73. 242.

114. Hochstein, F. A; Brown, W. G. J. Am. Chem. Soc. 1948. 70. 348. 157

115. Golfier, M.; Prange, T. J. Labeled Compels. 1972, & 623.

116. Kalvin, D. M.; Woodard, R. W. Tetrahedron. 1984r 3387.

117. Barlett, P.; Tate, F. J. Am. Chem. Soc. 1953, Z5* 91.

118. Pasto, D. J.; Huang, N. Z. J. Org. Chem. 1986. 51 r 412.

119. Kim, J. K.; Caserio, M. C.: J. Org. Chem. 1979. 44. 1897.

120. Hooz, J.; Giliani, S. S. H. Can. J. Chem. 1968, 86.

121. Midland, M. M.; in "Asymmetric Synthesis"; Morrison, J. D.,Ed.; Academic Press: New York, New York, 1983; Vol. 2, pp 45-70.

122. Parker, D. J. J. Chem. Soc. Perkin Trans. II 1983. 83.

123. Fujisaki, S.; Fujiwara, I.; Norisue, Y.; Kajigaeshi, S. Bull. Chem. Soc. Japan. 1985. 2429.

124. Cheng, Y. S.; Liu, W. L.; Chen, S. Synthesis 1980. 223.

125. Floss, H. G.; Tsai, M. D.; Adv Enzvmol. 1979, 243.

126. Floss, H. G. Methods Enzymol. 1982. 87. 126.