Studies of the Genetic Encoding of Pyrrolysine From

Home , Deoxyribonucleotide, Methanogenium frigidum, Methanosarcina, Methanosarcinaceae, Methanosarcina barkeri, Pyrrolysine

STUDIES OF THE GENETIC ENCODING OF PYRROLYSINE FROM

METHANOGENIC ARCHAEA

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of

Philosophy in the Graduate School of The Ohio State University

Anirban Mahapatra, M.Sc.

*****

The Ohio State University

2007

Dissertation Committee: Dr. Joseph A. Krzycki, Advisor Approved by Dr. Juan D. Alfonzo Dr. Charles J. Daniels Dr. Kurt L. Fredrick ______Advisor, Graduate Program in Microbiology

ABSTRACT

Pyrrolysine is the 22nd genetically encoded amino acid to be found in nature.

Co-translational insertion at in-frame UAG codons proceeds so that a single pyrrolysine residue is found in the active site of all methylamine methyltransferases required for methylamine metabolism in Methanosarcina spp.

This study examines processes central to the translation of UAG as pyrrolysine.

The pylT gene encoding the tRNA for pyrrolysine, tRNAPyl, is part of the pyl

operon of Methanosarcina spp. which also contains the pylS gene, encoding a

class II aminoacyl-tRNA synthetase; and the pylB, pylC, and pylD genes proposed to catalyze the synthesis of pyrrolysine from metabolic precursors.

Here, in the first study, by characterizing a Methanosarcina acetivorans mutant with the pylT gene region deleted, the essentiality of pyrrolysine incorporation in

methanogenesis from methylamines is tested. The mutant lacks detectable

tRNAPyl but grows similar to wild-type on methanol or acetate for which

translation of UAG as pyrrolysine is likely not to be essential. However, unlike

wild-type, the mutant can not grow on any methylamine or use monomethylamine

as the sole source of nitrogen. Monomethylamine methyltransferase activity is

detectable in wild-type cells, but not in mutant cells during growth on methanol.

Further, the pyrrolysine-containing monomethylamine methyltransferase is

absent from the mutant. This study is the first genetic analysis of UAG translation

as pyrrolysine, and it reveals the phenotype of a Methanosarcina strain that can

not decode UAG as pyrrolysine.

PylS is an aminoacyl-tRNA synthetase encoded by the pylS gene adjacent to

pylT. The data presented here in the second study shows that PylS catalyzes the

ATP-dependent activation of synthetic pyrrolysine. PylS-catalyzed activation is

highly specific for cognate amino acid, pyrrolysine, and does not require tRNA.

Taken together with results obtained by others showing the ligation of pyrrolysine

to tRNAPyl and the ability of pylS and pylT gene products to suppress amber codons in a recombinant system, the data presented indicates that PylS is a pyrrolysyl-tRNA synthetase capable of directly ligating pyrrolysine to tRNAPyl in

vitro and in vivo. These results prove that PylS is the first aminoacyl-tRNA

synthetase to be discovered from nature that is capable of attaching a

genetically-encoded amino acid, not one of the common twenty, to cognate tRNA. Further, a metabolite from an Escherichia coli strain with the pylBCD

genes expressed is shown to serve as a substrate for PylS activation. Along with

results obtained by others showing PylS-catalyzed ligation of this metabolite to

tRNAPyl, the data indicates that the pylB, pylC, and pylD gene products are

sufficient for pyrrolysine synthesis from metabolic precursors common to M. acetivorans and E. coli.

Earlier, in vitro studies by others indicated the two canonical lysyl-tRNA

iii

synthetases found in Methanosarcina spp. form a complex and slowly attach lysine to tRNAPyl. Owing to the implicit problem of substrate selection between

lysine-tRNAPyl and pyrrolysine-tRNAPyl in translation, the relevance of the

complex catalyzed ligation required testing in vivo. This is the focus of the third

study. Data presented indicates that intact lysyl-tRNA synthetase genes are not

required for methanogenesis from methylamines in Methanosarcina. Further,

levels of monomethylamine methyltransferase and tRNAPyl aminoacylation are

not diminished in lysyl-tRNA synthetase mutants. Taken together, the data

presented here demonstrates that the indirect route of tRNAPyl aminoacylation

involving complex formation of the two lysyl-tRNA synthetases is not essential for

UAG translation as pyrrolysine.

In the final study, the substrate specificity of PylS is probed using analogs of

pyrrolysine. Features of the pyrroline ring of pyrrolysine that are determinants of

PylS catalyzed activation are identified. Analogs used in functional probing in

vitro are then incorporated in a monomethylamine methyltransferase in an

Escherichia coli reporter system.

Dedicated to the loving memory of my grandmother,

Mahamaya Mahapatra

ACKNOWLEDGMENTS

As I reflect on the last few years of my life, I feel a deep sense of gratitude to

all who have made this work possible. First, I thank my advisor, Joseph Krzycki,

for handing me a set of fascinating scientific problems to work on, and for

patiently waiting as I developed the skills to analyze and interpret them. On a

scientific level, he has been my guru in the original Sanskrit connotation of the

word. I am also indebted to the members of my committee, Juan Alfonzo,

Charles Daniels, Kurt Fredrick for their time and input on all aspects of my

research and career development. Much of my work would have been delayed or

simply, not possible without collaboration, and I thank Michael Chan and Bill

Metcalf and members of their labs at Ohio State and the University of Illinois,

Urbana-Champaign, respectively for supporting my project.

I thank David Longstaff for befriending me when I knew no one else in the

department. Jitesh Soares has become one of my closest friends, and I cannot

thank him enough for always being there. I will always associate the best of times

with Ross Larue’s boisterous laughter and I thank him heartily for his friendship

through the years. I also thank Carey James, Sherry Blight, Jodie Lee, Ruisheng

Jiang, Marsha Thalhofer, Jenee Smith, and Jess Dsyzel, among others, for making the ninth floor of the Riffe building such an agreeable work environment.

Outside of work, I am grateful to Indrajit Mukherjee for sharing my sense of

humor. I thank Rituparna Roy, Pushan Pahari, and Sudhakar Panda for the

support over the years and miles. I thank my cousin, Prasun Pahari for being a

co-conspirator in all creative endeavors, and my cousin, Atalanta Kar, for being a dependable “younger brother.” I am grateful to my in-laws - Tarasankar and

Anjali Panigrahi; Deblina, and Nilanjana; and to my aunts - Jayasri Mahapatra,

Asita Sarangi, and Manjusri Kar for their support.

It is with sorrow that I remember personal losses. It is hard to believe my

uncle, Prasanta Kar, is no more. I am inspired that I met such a staunchly

idealistic person in a largely cynical world. I will deeply miss my uncle

Arunangshu Panda and my aunt Sunita Acharya, and am grateful to have known

them both.

I am humbled when I realize that every single aspect of my life has been built

on the sacrifices and prayers of my family members and well-wishers. I

apologize for the impudence of thanking them, for without their support, I would be nothing. Theirs is a debt that I can never repay. My grandparents are not alive today, but I know their blessings are with me. My younger sister, Rita, has been a reliable source of common sense through the years. I am deeply grateful to my father for instilling an appreciation of the natural world. My mother has

vii

been a source of unconditional love my entire life, and I hope she is proud of how her son turned out. Last but not least, this work would never have been possible without the undying support of my wife, Rituparna, who sacrificed her own aspirations so that I could come to America to pursue mine.

viii

VITA

December 3, 1974………………….Born, New Delhi, India

1997………………………………….B.Sc. (Hons) in Botany, Vidyasagar University

1999………………………………….M.Sc. in Botany, Vidyasagar University

2001-2007…………………………...Graduate Teaching and Research Associate,

The Ohio State University

PUBLICATIONS

1. Mahapatra A., Srinivasan G., Richter K.B., Meyer A., Lienard T., Zhang J.K., Zhao G., Kang P.T., Chan M., Gottschalk G., Metcalf W.W., and J.A. Krzycki (2007) Class I and class II lysyl-tRNA synthetase mutants and the genetic encoding of pyrrolysine in Methanosarcina spp. Mol. Microbiol. 64:1306-18.

2. Longstaff D.G., Larue R.C., Faust J.E., Mahapatra A., Zhang L., Green-church K.B., and J.A. Krzycki. (2007) A natural genetic code expansion cassette enables transmissible biosynthesis and genetic encoding of pyrrolysine. Proc. Natl. Acad Sci USA. 104:1021-1026.

3. Mahapatra A., and J.A. Krzycki (2007) The Genetic Code In: McGraw-Hill Yearbook of Science & Technology, McGraw-Hill, New York. 93-98.

4. Mahapatra A., Patel A., Soares J.A., Larue R.C., Zhang J.K., Metcalf W.W., and J.A. Krzycki (2006) Characterization of a Methanosarcina acetivorans mutant unable to translate UAG as pyrrolysine. Mol. Microbiol. 59: 56-66.

5. Blight, S. K., Larue, R. C., Mahapatra, A., Longstaff, D. G., Chang, E., Zhao, G., Kang, P. T., Green-Church, K. B., Chan, M. K., and J. A. Krzycki. (2004) Direct charging of tRNACUA with pyrrolysine in vitro and in vivo. Nature. 431: 333- 335.

FIELDS OF STUDY

Major Field: Microbiology

TABLE OF CONTENTS

Page

Abstract...... ii Dedication……………………………………………………………………………...... v Acknowledgments………………………………………………………………….…...vi Vita…………………………………………………………………………………...…..ix List of Tables…………………………………………………………………………...xv List of Figures…………………………………………………………………...... xvi List of Symbols and Abbreviations………………………………………………..…xix

Chapters: 1. Introduction………………………………………………………………….....1 1.1 The metabolic diversity of methanogenic archaea………………...1 1.1.1 Archaea……………………………………………………...1 1.1.2 Methane production and methanogens……………..…...2 1.1.3 Methanogenesis……………………………………………6 1.1.3.1 Hydrogenotrophic methanogenesis……...... 6 1.1.3.2 Acetoclastic methanogenesis………………...6 1.1.3.3 Methylotrophic methanogenesis…………...... 7 1.2. Translation…………………………………………………………...... 9 1.2.1 Translation termination………………………………..…...9 1.2.2 Suppressor tRNAs in translation………………………..12 1.2.3 Aminoacyl-tRNA synthesis……………………………....15 1.2.3.1 Aminoacyl-tRNA synthetases……………....15 1.2.3.2 Specificity of amino acid activation……...... 18 1.2.3.3 Specificity of amino acid attachment to cognate tRNA………………………………...21 1.2.3.4 Indirect aminoacylation mechanisms…...….22 1.3 Selenocysteine: the 21st genetically encoded amino acid………..25 1.4 Expansion of the genetic code with pyrrolysine…………………...28 1.4.1 The methylamine methyltransferase genes possess single in-frame amber codons………...... 28 1.4.2 Structural analysis of Monomethylamine methyltransferase………………………………………….31 1.4.3 tRNAPyl: an amber-decoding tRNA………….…………..32 xi

1.4.4 Aminoacylation of tRNAPyl……………………………….34 1.4.4.1 PylS is a Class II aminoacyl-tRNA synthetase……………………………….…...35 1.4.4.2 Non-homologous lysyl-tRNA synthetases in Methanosarcina spp. ….….37 1.4.5 Contextual requirements for pyrrolysine insertion……..39 1.5 Artificial expansion of the genetic code with unnatural amino acids…………………………………………………………..40 1.6 Overview of work presented in following chapters…………….…43

2. Characterization of a mutant unable to translate UAG as pyrrolysine…………………………………………………………………….57 2.1 Introduction…………………………………………………………..57 2.2 Experimental Procedures…………………………………………..58 2.2.1 Organisms, growth conditions and reagents………….58 2.2.2 Construction of mutant strain………………………...... 59 2.2.3 Isolation of DNA and tRNA…………………………...... 59 2.2.4 Preparation of cell extracts…………………………...... 61 2.2.5 Southern hybridization analysis………………….….....61 2.2.6 Acid-urea gel electrophoresis and Northern hybridization analysis…………………………………....62 2.2.7 Immunoblot analysis………………………………….….63 2.2.8 MMA:CoM methyltransferase activity assays…………64 2.3 Results……………………………………………………………...... 65 2.3.1 The ∆ppylT mutant……………………………..……….65 2.3.2 tRNAPyl is not detectable in ∆ppylT………………..…..67 2.3.3 The ∆ppylT mutant is lethal on specific growth substrates……………………………………………..….68 . 2.3.4 Growth using MMA as a sole nitrogen source…….…69 2.3.5 MMA:CoM methyltransfer activity is not detectable in ∆ppylT………………………………………………..……70 2.3.6 The ∆ppylT strain does not produce detectable MtmB protein……………………………………………………..71 2.3.7 The ∆ppylT strain does not produce detectable MtbA protein………………………………………………….....72 2.4. Discussion…………………………………………………...... …..73

3. Pyrrolysine ligation to cognate tRNA: Activation of pyrrolysine by a novel aminoacyl-tRNA synthetase……………………………………………..…87 3.1 Introduction………………………………………………………..…87 3.2 Experimental Procedures………………………………………..…88 xii

3.2.1 Strains and plasmids……………………………………..88 3.2.2 PylS purification………………………………………...... 89 3.2.3 PylS substrates…………………………………….…...... 89 32 3.2.4 PPi-ATP exchange with PylS………………………….90

3.3 Results……………………………………………………………..…...92 32 3.3.1 The PPi-ATP exchange reaction catalyzed by PylS is pyrrolysine dependent……………………………..…...92 32 3.3.2 The PPi-ATP exchange assay confirms that pyrrolysine synthesis is dependent on pylBCD expression……………………………………………...…..92 3.4. Discussion………………………………………………………..…...93

4. Class I and Class II lysyl-tRNA synthetase mutants and the genetic encoding of pyrrolysine……………………………………....103 4.1 Introduction………………………………………………………....103 4.2 Experimental Procedures………………………………………....105 4.2.1 Organisms, growth conditions and reagents………....105 4.2.2 ∆lysKc::mm and ∆lysS::pac strains…………………....106 4.2.3 Southern hybridization………………………………….106 4.2.4 Anti-MtmB immunoblotting in wild-type and mutant strains…………………………………...... 107 4.2.5 Isolation of genomic DNA and tRNA……………...... 107 4.2.6 Acid-urea gel electrophoresis and Northern hybridization………………………………….107 4.3 Results……………………………………………………………...108 4.3.1 Construction of M. acetivorans mutants lacking LysRS1 or LysRS2…………………………....108 4.3.2 Catabolic range of strains bearing lysK or lysS disruptions………………………………………...110 4.3.3 UAG translation in lysS and lysK deletion strains….111 4.3.4. Aminoacylation of tRNAPyl and tRNALys in mutant strains…………………………………………………....111 4.4 Discussion………………………………………………………...... 114

5. The amino acid specificity of pyrrolysyl-tRNA synthetase………….....127 5.1 Introduction…………………………………………………………….……..127 5.2 Experimental Procedures…………………………………...…….129 5.2.1 General…………………………………………...……....129 32 5.2.2 PPi-ATP exchange with PylS…………………………130 5.2.3 Ligation of amino acid to tRNAPyl catalyzed by PylS...131 5.2.4 In vivo incorporation of analogs in E. coli……….…....132 xiii

5.3 Results……………………………………………………………...... 133 5.3.1 Activation and aminoacylation of tRNAPyl with pyrrolysine analogs………………………………...... 133

5.3.2 Incorporation of pyrrolysine analogs in E. coli proteins………………………………………….137 5.4 Discussion…………………………………………………………...138

List of References………………………………………………………...... 155

xiv

LIST OF TABLES

Table Page

1.1 Aminoacyl-tRNA synthetases for the twenty common amino acids…………………………………………………………………...... 55

1.2 Distribution of methylamine methyltransferase genes among select members of Methanosarcinaceae………………………………………..….56

2.1 Methanol:CoM methyl transfer and MMA:CoM methyl transfer activity…86

32 3.1 The PPi-ATP exchange reaction catalyzed by PylS is pyrrolysine specific………………………………………………………………………...100

3.2 The ATP dependent activation of pyrrolysine by PylS is not dependent on tRNA…………………………………………………………………………...101

32 3.3 Kinetic parameters of the PylS catalyzed PPi-ATP exchange reaction102

4.1 Growth of lysyl-tRNA mutants with various growth substrates………….125

4.2. In vivo levels of tRNAPyl and tRNALys aminoacylation in wild type, ΔlysK and ΔlysS strains……………………………………………………………..126

32 5.1 Steady-state PPi-ATP exchange kinetics of M. barkeri PylS with pyrrolysine analogs………………………………………………………..…154

LIST OF FIGURES

Figure Page

1.1 Distribution of methanogens within Domain Archaea…...... 44

1.2 Cells of the methanogen Methanosarcina barkeri………………………....45

1.3 Pathways of methylotrophic methanogenesis……………………………...46

1.4 Common and uncommon amino acids specified by the genetic code…..47

1.5 Structural variations between the two classes of aminoacyl-tRNA synthetases………………………………………………………………….... 48

1.6 Comparison of the structure of tRNA bound to ternary complex with that of bacterial release factor………………………………………………….....49

1.7 Genetic encoding of selenocysteine………………………………………...50

1.8 Resolved x-ray crystallographic structures of monomethylamine methyltransferase……………………………………………………………...51

1.9 Cloverleaf secondary structure of tRNAPyl…………………………………..52

1.10 Three leading possibilities for tRNAPyl synthesis and aminoacylation…...53

1.11 Alignments of motif 2 and motif 3 sequences from select class II aminoacyl-tRNA synthetases…………………………………………….…..54

2.1 The pyl and mtmB gene clusters of M. acetivorans………………………..78

2.2 Southern analysis of restriction digested wild-type and ΔppylT genomic DNA………………………………………………………….79

2.3 Northern analysis of tRNAPyl from wild-type and ΔpylT…………………...80

xvi

2.4 Growth curves of wild-type (closed symbols) and ΔpylT (open symbols) on methanol and methylamines………………………………………………….81

2.5 Growth of wild-type and ΔpylT strains on medium lacking ammonia…....82

2.6 Detection of MtmB by immunoblotting……………………………………....83

2.7 SDS-Polyacrylamide gel electrophoresis of fractionated soluble proteins from wild-type and ΔppylT cells grown on methanol and MMA…………..84

2.8 The amine specific methylase, MtbA is not detectable from extracts of ΔppylT grown on methanol-MMA medium……………………………….…85 . 32 3.1 The PPi-ATP exchange reaction mediated by PylS is pyrrolysine dependent………………………………………………………………………97

3.2 Hanes-Woolf plots to determine the apparent kinetic parameters of activation by PylS……………………………………………………………...98

32 3.3 Cellular amino acid pools tested in the PylS-mediated PPi-ATP exchange assay………………………………………………………………..99

4.1 The ∆lysKc::mm and ∆lysS::pac mutant strains of M. acetivorans lack functional LysRS1 and LysRS2 respectively……………………………...120

4.2 Confirmation of ∆lysS::pac and ∆lysKc::mm mutations in M. acetivorans by Southern analysis…………………………………………………………121

4.3 Anti-MtmB immunoblotting of total soluble protein from wild-type, ∆lysKc::mm and ∆lysS::pac strains………………………………………..122

4.4 Relative tRNAPyl aminoacylation levels in wild-type, ∆lysKc::mm and ∆lysS::pac strains…………………………………………………………….123

4.5 Relative tRNALys aminoacylation levels in wild-type, ∆lysKc::mm and ∆lysS::pac strains…………………………………………………………….124

5.1 Commercially available amino acids tested for tRNAPyl aminoacylation by PylS………………………………………………………………………..…..145

5.2 Pyrrolysine analogs……………………………………………………..…...146

5.3 Activation of pyrrolysine and analogs by PylS……………………..……..147

xvii

5.4 Aminoacylation in vitro as determined by acid-urea gel electrophoresis and Northern blotting to detect tRNAPyl…………………………………….149

5.5 Anti-MtmB immunoblot of extracts from E. coli…………………………...150

5.6 Comparison of incorporation of analogs in MtmB………………………...151

5.7 The active site of Methanosarcina mazei PylS……………………………152

5.8. The deduced active site of Methanosarcina barkeri PylS……………....153

xviii

LIST OF SYMBOLS AND ABBREVIATIONS

Å angstrom (unit)

AARS aminoacyl-tRNA synthetase aa-tRNAaa aminoacyl-tRNA

AMP adenosine 5'-monophosphate

ATP adenosine 5'-triphosphate

bp base pair

°C degree Celsius (unit)

CoB coenzyme B

CoM coenzyme M

Ci curie (unit)

Cyp 2-amino-6-(cyclopentanecarboxamino)hexanoic acid

Cyc Nε-cyclopentyloxycarbonyl-L-lysine

Δ delta (deletion)

DMA dimethylamine

DNA deoxyribonucleic acid

DNase deoxyribonuclease

DTT dithiothreitol

xix

E. coli Escherichia coli

EDTA ethylenediaminetetraacetic acid

EF-Tu elongation factor-Tu g gram (unit) x g times gravitational constant

GTP guanosine triphosphate

HEPES (N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid])

H SPT tetrahydrosarcinapterin 4

IPTG isopropyl-β-D-thiogalactoside

KF potassium fluoride kcat enzyme turnover number

KCl potassium chloride kDa kilodalton (unit)

Km Michaelis constant

L Liter (unit)

LB Luria-Bertani

Lys L-lysine m mili

M molar (unit)

μ micro

M. acetivorans Methanosarcina acetivorans

M. barkeri Methanosarcina barkeri

MgCl magnesium chloride 2

MCR methyl-CoM reductase

MMA monomethylamine

MOPS 3-[N-morpholino]propanesulfonic acid

mRNA messenger RNA

NaCl sodium chloride

NH Cl ammonium chloride 4 nt nucleotide

OD optical density

ORF open reading frame

PAGE polyacrylamide gel electrophoresis

PCR polymerase chain reaction

PPi inorganic pyrophosphate

Pyl L-pyrrolysine

PYLIS pyrrolysine insertion sequence

PylS pyrrolysyl-tRNA synthetase

RF release factor

RNA ribonucleic acid

RNase ribonuclease

s second (unit)

SDS sodium dodecyl sulfate

SECIS selenocysteine insertion sequence

xxi

Thf 2-amino-6-((R)-tetrahydrofuran-2-carboxamido)hexanoic acid

TMA trimethylamine

Tris-HCl tris-(hydroxylmethyl) aminomethane hydrochloride tRNA transfer RNA

UV ultraviolet

V volt (unit) v volume

V theoretical maximum reaction velocity max

w weight

xxii

CHAPTER 1

INTRODUCTION

1.1 The metabolic diversity of methanogenic archaea

1.1.1 Archaea

One of the most significant accomplishments of modern microbiology has been recognizing that organisms belonging to the new grouping, Archaea (also known as Archaebacteria), constitute a truly unique lineage of cellular life distinct from both Bacteria and Eukarya (Figure 1.1; Woese et al, 1990). Prior to this discovery, these organisms were known as archaebacteria (“old” bacteria) and lumped together with eubacterial (“true” bacterial) members of Bacteria.

Members of Archaea, the so-called “third domain” of life, do display many similarities to eubacteria, most notably with regard to cellular morphology and genomic organization (Deppenmeier, 2002). However, similarities to eukaryotes

(members of Eukarya) with respect to the replication, transcription, and translation protein machinery, hint at a shared ancestry. Nevertheless, there are several unique archaeal characteristics, including distinct cell membranes with isoprenyl ether lipids in place of acyl ester lipids found in bacteria, and

uncommon stereochemistry of the glycerophosphate backbone of membrane lipids (Deppenmeier, 2004). With the sequencing of the entire archaeal genomes, it became apparent that many genes in these organisms have no known equivalent genes in either bacterial or eukaryotic organisms. Currently, it is thought that all three clusters of organisms shared a common ancestor before the lineages diverged approximately 3 billion years ago (Gribaldo et al, 2006).

At present, domain Archaea is divided into four phylogenetically discrete groups namely, the Crenarchaeota comprising thermophilic and thermoacidophilic organisms; the Euryarchaeota consisting of methanogens and halophiles; the Korarchaeota containing uncultured organisms found in terrestrial hot springs (Woese et al, 1990; Barns et al, 1996); and the Nanoarchaeota consisting of a recently described species Nanoarchaeum equitans (Huber et al,

2002).

1.1.2 Methane production and methanogens

Considering the role in the global carbon cycle, the production of methane is definitely a significant geochemical part of the global carbon cycle. Over one billion tons of this gas is produced every year (Ferry, 1999). Much of the methane produced is biogenic in origin (Khalil and Rasmussen, 1994). Organisms that produce methane as an end-product of their metabolism are known as methanogens. All verified methanogens are obligate anaerobes which are widely distributed in diverse anoxic environments such as the fresh-water sediments of

lakes and rivers, saline coastal marine sediments, frigid tundra areas, boiling hydrothermal vents, rice paddies, anaerobic sludge digesters, the gastrointestinal tract of ruminants and termites, as well as the lower gastrointestinal tract of humans and other monogastric mammals (Ferry 1999; Deppenmeier, 2002).

The anaerobic food chain is a four-step process by which organic materials are mineralized in anoxic environments. A group of hydrolytic microbes breaks down complex biological polymers such as polysaccharides, lipids, proteins, and nucleic acids to their respective monomers. Then, these monomers are broken down to simple carboxylic acids, alcohols, ketones, and other compounds such as hydrogen gas and carbon dioxide by fermentative bacteria. Many fatty acids and alcohols are oxidized to acetates and one-carbon derivatives by acetogenic or syntrophic bacteria (Diekert and Wohlfarth, 1994; Drake et al, 1997; Schink,

1997). Methanogens can utilize many of these compounds such as hydrogen and carbon dioxide; acetate; and one-carbon compounds as carbon and energy sources resulting in the production of methane and carbon dioxide as end- products (Zinder, 1993). In the process, methanogens remove hydrogen that builds up in the anaerobic food chain as a result of the degradation of fatty acids and alcohols (Conrad, 1996; Deppenmeier, 2002).

Methanogens also have a profound effect on contemporary anthropocentric issues. Methanogens are responsible for most of the natural gas reserves in use today (Thauer, 1998). After carbon dioxide, methane is the most important greenhouse gas and contributes to an estimated 16% of the greenhouse effect

(Conrad, 1996). Studies on ice cores indicate a steady increase in methane over

the past 150 years predominantly arising from anthropogenic sources such as flooded rice paddies (Conrad, 1996; Deppenmeier, 2002). Additional reservoirs of methane, also presumably of biogenic origin, are the methane hydrates found in deep sea sediments (Reed et al, 2002). It has been suggested that these ice- like methane hydrates, constantly subjected to high pressures and low temperatures, are a potentially exploitable source of natural gas reserves

(Deppenmeier, 2002).

The predominant methane-producers in monogastric mammals, including humans, are species of Methanobrevibacter (Eckburg, et al., 2005). A single species Methanobrevibacter smithii comprises as much as 10% of all anaerobic organisms in the normal human intestinal flora (Eckburg, et al., 2005). Apart from

M. smithii, another methanogen, Methanosphaera stadtmanae, also a member of the Methanobacteriales, has been isolated from the human gastrointestinal tract

(Fricke, et al., 2006). Human digestion of complex dietary polysaccharides is facilitated by bacterial fermentation in the gastrointestinal tract. Interestingly, it has been demonstrated that methanogens boost bacterial fermentation in the gastrointestinal tract by eliminating hydrogen, one of the end products of fermentation (Samuel and Gordon, 2006). Enhanced fermentation has been linked with obesity, which in turn, has been demonstrated to correlate with an enhanced presence of methanogens in complex gut microbial communities

(Samuel and Gordon, 2006; Turnbaugh et al, 2006). Recently, selective targeting

of methanogens by inhibiting steps of the methane-yielding pathway has been recommended to reduce energy yield in obese humans (Samuel et al., 2007). If implemented this would be the first targeting of any archaeon for therapeutic purposes.

Bearing in mind the end-product of the sole-energy yielding process for methanogens is methane, one might assume that the term “methanogen” represents a few organisms, narrow in range with respect to morphology and phylogeny. However, this is far from the truth. Methanogenic species that are cocci, rods, or spirilla have been described, as have organisms that are arranged in packets or sarcina (Figure 1.2; Sowers et al, 1984; Sowers et al, 1993;

Deppenmeier, 2002). In terms of cell wall composition, methanogens have been described with pseudomurein, heteropolysaccharides, or protein subunits (Jones et al, 1987). Methanogens may be psychrophilic, mesophilic, thermophilic, or hyperthermophilic (Garcia et al, 2000). In all, over 80 species have been described that are classified into five distinct orders, Methanobacteriales,

Methanococcales, Methanomicrobiales, Methanopyrales, and

Methanosarcinales (Garcia et al, 2000). Species belonging to the first four orders have somewhat restricted substrate specificity when compared with members of the fifth, the Methanosarcinales, since they use hydrogen plus carbon dioxide

(and often formate) as substrates (Thauer, 1998). In contrast, members of the

Methanosarcinales can utilize hydrogen plus carbon dioxide (hydrogenotrophic methanogenesis); acetate (acetoclastic methanogenesis); methanol,

methylamines, and other methylated one-carbon compounds (methylotrophic methanogenesis); and carbon monoxide (Thauer, 1998). A discussion of methanogenesis from each class of compounds is discussed in subsequent sections.

1.1.3 Methanogenesis

1.1.3.1 Hydrogenotrophic methanogenesis

The hydrogenotrophic pathway of methanogesis entails the reduction of carbon dioxide with the concomitant oxidation of hydrogen gas to provide reducing

equivalents through hydrogenases (DiMarco et al, 1990). The change in Gibb’s free energy (∆G’o) for this process is -131 kJ per mol of methane gas

(Deppenmeier, 2002). Approximately half of all described methanogens can also

utilize formate for methanogenesis with change in Gibb’s free energy (∆G’o) of -

106 kJ per mol of methane gas (Deppenmeier, 2002). Species of

Methanosarcina, however, do not utilize formate, and one, Methanosarcina acetivorans, can not perform hydrogenotrophic methanogenesis due to lack of two key hydrogenases (Galagan et al, 2002).

1.1.3.2 Acetoclastic methanogenesis

Notwithstanding the fact that the major portion of methane produced biologically is from acetate, only two described genera, Methanosarcina and

Methanosaeta can use this substrate for methanogenesis (Ferry, 1992;

Deppenmeier, 2002; Galagan, 2002). The pathway invoked by these organisms,

known as the acetoclastic pathway, occurs via the dismutation of acetate (Ferry,

1992). Acetate is activated to acetyl-CoA with the carbonyl group subsequently oxidized to carbon dioxide to provide reducing equivalents. The methyl group enters the general methanogenic pathway by transfer to tetrahydromethanopterin or tetrahydrosarcinopterin (in Methanosarcina) leading ultimately to reduction to methane (Thauer, 1998). The energy yield from the acetoclastic pathway (∆G’o =

-36 kJ per mol of methane) is significantly less than that from the

hydrogenotrophic pathway (Ferry 1992; Deppenmeier, 2002).

1.1.3.3 Methylotrophic methanogenesis

Although the largest number of organisms utilize hydrogen/carbon dioxide and

most of the methane of biological origin is derived from acetate, the most widely

used class of substrates for methanogenesis in marine and brackish environment

are methylated compounds which can be used only by members of

Methanosarcinaceae (Blaut, 1994; Galagan et al, 2002; Li et al, 2005). Such

compounds include methanol, methylthiols, and methylamines (Thauer, 1998).

The methylamines and methythiols originate from the anaerobic degradation of

widespread cellular osmolytes such as betaine, trimethylamine-N-oxide, and

dimethylsulfoniopropionate found in marine plants and phytoplankton

(Deppenmeier, 2002).

The general pathway by which methylotrophic methanogenesis is initiated is similar regardless of which substrate is being utilized and involves the series of methyltransfer reactions outlined in Figure 1.3. Research in M. barkeri indicates that similar mechanisms exist for methanogenesis from methanol, trimethylamine

(TMA), dimethylamine (DMA), and monomethylamine (MMA) (Grahame, 1989;

Sauer and Thauer, 1998; Sauer and Thauer, 1997; Burke and Krzycki, 1995;

Ferguson et al, 1996; Ferguson and Krzycki, 1997; Burke and Krzycki, 1997;

Tallant and Krzycki, 1997; Li et al, 2005). Briefly, the transfer of the methyl moiety of the methylated compounds to a substrate specific corrinoid-protein is catalyzed by a methyltransferase designated as MT1.The second methyltransferase MT2 catalyzes the transfer of the methyl group from the corrinoid protein to coenzyme M, which is the direct precursor of methane in all known methods of methanogenesis (Ermler, 1997; Li et al, 2005). For every three methyl moieties reduced to methane, one methyl moiety is oxidized to carbon dioxide via the reverse carbon dioxide-reduction pathway to provide six reducing equivalents (Welander and Metcalf, 2005). Due to the similarity in metabolic events required to produce methane from known methylated compounds and the availability of complete genomes of Methanosarcina, many putative methyltransferase systems have been hypothesized to function in the utilization of yet undiscovered growth substrates (Galagan et al, 2002).

The methanol-specific MT1 enzyme is designated MtaB, the corrinoid protein is MtaC, and MT2 - the methanol specific CoM methylase is MtaA (van der

Meijden et al, 1983; van der Meijden et al, 1984; Sauer and Thauer, 1997; Sauer

and Thauer, 1998). For TMA, DMA, and MMA utilization, the specific MT1

enzymes are MttB, MtbB, and MtmB respectively (Ferguson and Krzycki, 1997;

Paul et al, 2000; Burke and Krzycki, 1997). In similar fashion, the cognate

corrinoid proteins for TMA, DMA, and MMA utilization are MttC, MtbC, and MtmC

(Ferguson and Krzycki, 1997; Paul et al, 2000; Burke and Krzycki, 1997). All

three methylated corrinoid proteins can transfer a methyl moiety to the amine- specific CoM methylase (MT2), MtbA (LeClerc and Grahame, 1996).

Interestingly, while members of Methanosarcinales are the only methanogens

that can metabolize methylated compounds, the two methanogens found in the

human gastrointestinal tract, Methanobrevibacter smithii and Methanosphaera

stadtmanae, contain the mtaB, mtaC, and mtaA genes required for methylotrophic methanogenesis from methanol (Samuel et al., 2007; Fricke, et al., 2006). It may be noted, however, that genes encoding proteins required for methanogenesis from methylamines have not been discovered in either species

(Samuel et al., 2007; Fricke, et al., 2006).

1.2. Translation

1.2.1 Translation termination

The flow of information from genes to proteins involves the transcription of genes to RNA and the subsequent translation of RNA to proteins. The process of translation in turn has four discrete steps. In initiation, the first step, the subunits

of the ribosome join the mRNA to be translated and start protein synthesis at the

AUG initiator codon. In repeated cycles of the second step, elongation, sense

codons are read in triplets, and the polypeptide chain is synthesized. The third

step, termination, allows the release of the nascent polypeptide at one of three

termination codons, UAA, UAG, or UGA. A fourth step, ribosome recycling,

allows the regeneration of ribosomal subunits for iterative rounds of translation

(reviewed in Ramakrishnan, 2002).

At any moment, there are two tRNAs in the ribosome, one at the ribosomal A- site and the other at the ribosomal P-site. The termination reaction involves release of the polypeptide chain from the last tRNA (reviewed in Kisselev and

Buckingham, 2000). Standard termination of translation occurs at UAA, UAG, or

UGA codons which are not decoded by standard tRNAs. In E. coli, termination codons are recognized by class I release factors (RFs). The proteins decode the termination codons present at the ribosomal A-site (Kapp and Lorsch, 2004).

RF1 recognizes UAA and UAG, and RF2 recognizes UAA and UGA (Scolnick et

al, 1968). RF1 and RF2 both activate the ribosome to hydrolyze the peptidyl-

tRNA at the peptidyl transferase center of the ribosome (Nissen et al, 2000).

Cleavage of the polypeptide occurs in a reaction analogous to peptidyl transfer;

instead of aminoacyl-tRNA, however, the acceptor is H2O (Nissen et al, 2000).

Class 2 RFs are GTPase enzymes that facilitate the release of class I RFs, and

hence they are not codon-specific (Kapp and Lorsch, 2004). In E. coli, RF3 is the

class 2 RF (Grentzmann et al, 1994; Frolova et al, 1994). RF3-GDP binds to the

ribosome prior to translation termination where GTP replaces GDP (Freistroffer

et al, 1997). In reactions performed in vitro using transcripts with a high percentage of stop codons, RF3 is not essential (Frolova et al, 1994), though it seems to be essential for termination using transcripts with a lower percentage of stop codons (Zhouravleva et al, 1995).

It is generally thought that molecular mimicry allows RF1/RF2 complexed with

RF3 to enter the ribosomal A-site, as this complex is thought to resemble Ef-Tu-

GTP-aminoacyl tRNA ternary complex (Ito et al, 2000). Structurally, class 1 RFs resemble tRNAs although the former are proteins and the latter RNA molecules

(Figure 1.6). A putative anticodon region of peptide has been identified in both

RF1 and RF2 (Ito et al, 2000). The peptide anticodon and the termination codon

are hypothesized to form an interaction akin to that of the codon and anticodon

pairing of mRNA and tRNA, respectively (Kapp and Lorsch, 2004). However, this

interpretation is open to debate as recent cryo-electron microscopy has revealed

that RF2 assumes a different structure in the presence of ribosomal complex

from that deduced from X-ray crystallography (Nakamura and Ito, 2003).

In eukaryotes, there is a single class 1 RF, eRF1, which allows hydrolysis

from any one of the three termination codons (Konecki et al, 1977). A class 2 RF,

eRF3, is essential for the release of eRF in eukaryotes (Mikuni et al, 1994). The

archaeal mechanism of translation termination is not well understood but studies

indicate that the mechanism resembles that in eukaryotes (Dontsova et al, 2000).

A class I RF, aRF1, appears to be homologous to eRF1; however no archaeal

homolog of eRF3 has been discovered to date (Dontsova et al, 2000; Lecompte et al, 2002). Recently, information gleaned from a bioinformatics analysis suggests that species of Methanosarcina may possess two aRF1 homologs

(Zhang et al, 2005)

Regarding sequences surrounding the termination codon, studies have indicated that the strength of termination is partly influenced by sequences downstream of the termination codon (Buckingham, 1994; McCaughan et al,

1995). There is a strong correlation of termination with the nucleotide following the termination codon in both eubacteria and eukaryotes (Brown et al, 1990a;

Brown et al, 1990b; Poole et al, 1995; Poole et al, 1998; Ozawa et al, 2002). This base is often referred to as being in the +4 position (Buckingham, 1994; Poole et al, 1995; Tate et al, 1995). Studies indicate that bases within +1 to +6 directly contact RFs and are involved in the strength of termination (Tate et al, 1999).

Apart from the three bases of the termination codon, probably the most influential base is the one immediately downstream; studies indicate it also influences the chances of frameshifting and termination codon readthrough occurring

(Buckingham, 1994).

1.2.2 Suppressor tRNAs in translation

Usually, translation occurs in a standard procedure in which discrete unambiguous codon triplets are read one after the other. However, a few aberrations have been elucidated in great detail. One of them, called

programmed frameshifting, requires the ribosome to switch to an alternate reading frame one base forward or behind ( Brierley et al, 1989; Fu and Parker,

1994; Gesteland and Atkins, 1996). One example of frameshifting in the +1 direction is in the decoding of the E. coli prfB transcript which gives rise to RF2

(Craigen et al, 1985) while an example of -1 frameshifting in the opposite direction involves the dnaX transcript giving rise to the γ and τ subunits of DNA polymerase III (Namy et al, 2004). A second non-standard translation event is translational bypassing in which the ribosome jumps over noncoding sequences and resumes translation further downstream. The only well-documented occurrence of translational bypassing is in the translation of transcripts originating from gene 60 in bacteriophage T4 (Herr et al, 2000).

Far more widespread than translational frameshifting or translational bypassing is the redefinition of sense and termination codons (Gesteland and

Atkins, 1996). Since each of the 64 codons is defined in the universal genetic code as either signaling the insertion of one of the twenty common amino acids or signaling termination, redefinition requires one of these codons to be reprogrammed to insert a different amino acid. Since tRNA molecules act as adaptors to effectively insert amino acids at specific codons, assigning new meaning involves mutating the anticodon of the tRNA (Pure et al, 1985; Weiss and Friedberg, 1986; Murgola, 1985). When the anticodon of the redefined tRNA recognizes a codon designated for a different amino acid in the universal genetic code it is known as a missense suppressor tRNA and when it recognizes the

termination codons it is known as a nonsense suppressor tRNA (Murgola, 1985).

Nonsense suppressor tRNAs that recognize all three termination codon have

been isolated from E. coli (Murgola, 1985). A nonsense suppressor tRNA faces a unique situation compared to a canonical tRNA in that it has to compete with RF to define the coding of the termination codon. UGA is the least efficient of the natural termination codons and is leaky at 1-3% (Atkins and Gesteland, 1996).

However, UAG is used infrequently in E. coli and suppression of this termination codon is relatively efficient (10%-50%); at least six different amber suppressor tRNA species (all with the CUA anticodon) have been isolated (Murgola, 1985;

Atkins and Gesteland, 1996). Well-characterized examples, including the supD, supE, and supF loci encode amber suppressor tRNA species that insert serine, glutamine, and tyrosine, respectively (Eggertsson and Söll, 1988). Further examples abound. For example, suppression of UAG to insert glutamine allows readthrough in retroviral genes which lie in the same reading frame such as the gag and pol genes of murine leukemia virus (Philipson et al, 1978; Yoshinaka et al, 1985). Apart from the canonical glutamine encoding tRNA, the ciliate

protozoan Tetrahymena thermophila has two glutamine inserting tRNAs that

recognize UAG and UAA, respectively (Schull and Beier, 1994). An example

from a mammalian cell is that of the leucine-encoding UAG suppressor tRNA

isolated from calf liver (Valle, et al, 1987)

1.2.3 Aminoacyl-tRNA synthesis

1.2.3.1 Aminoacyl-tRNA synthetases

To function as adapters, tRNA molecules must be aminoacylated or “charged” to form aminoacyl-tRNAs. This process, known as aminoacyl-tRNA synthesis, is catalyzed by a family of ubiquitous enzymes known as the aminoacyl-tRNA synthetases (AARS). Aminoacyl-tRNA synthetases catalyze the covalent ligation of amino acids, for which they are specific, to the terminal adenosine at the 3’ end of cognate tRNAs. Since errors in aminoacylation lead to inaccurate amino acids being inserted into peptides during protein synthesis, there is selective pressure to ensure that the correct amino acid is ligated to the cognate tRNA.

The aminoacyl-tRNA synthetases must maintain the fidelity of the reaction, but face a biochemical dilemma in that cognate and incorrect noncognate amino acids are small molecules with similar chemistries. Further, tRNA species are also similar to each other in terms of general structure and the chemistry of constituent bases. Despite these constraints, the error rate of aminoacylation of tRNA is less than 1 in 5000 for most aminoacyl-tRNA synthetases (First, 2005;

Jakubowski, 2005). This remarkable rate of accuracy is achieved through tight coupling between substrate specificity and catalysis (First, 2005).

The aminoacylation reaction involves a two-step mechanism that occurs in a single-active site within the enzyme. In the first step, the amino acid is activated by ATP to form the enzyme-bound aminoacyl-adenylate intermediate (Arnez and

Moras, 1997; First, 2005). Activation involves nucleophilic attack on the α- phosphorus of ATP by the oxygen of α-carboxylate of the amino acid. Cleavage of pyrophosphate from the ATP provides energy for this step (∆G°’=33.4 kJ mol-

1). In the second step, the amino acid is transferred to the 2’- or 3’- hydroxyl of

the terminal adenosine of the cognate tRNA (Ibba and Söll, 2000). This step

involves the nucleophilic attack of the 2-‘ or 3’-hydroxyl of the terminal adenosine on the α-carbonyl carbon on the aminoacyl adenylate. Energy for this step of the reaction is derived from cleavage of the mixed anhydride bond between the α-

carbonyl and the α-phosphate of the aminoacyl-adenylate. The resulting aminoacyl-tRNA and AMP products are released by the AARS, allowing the

enzyme to proceed in further cycles of catalysis (Schimmel and Söll, 1979). Four

AARS enzymes, glutaminyl-, glutamyl-, arginyl-, and the class I lysyl-tRNA synthetase require that the cognate tRNA be present during the activation step

(Ibba and Söll, 2000; First 2005). In these examples, the binding of cognate tRNA causes a conformational change within the active-site to facilitate amino acid activation (Francklyn et al, 2002; First, 2005).

Although AARS enzymes catalyze similar reactions, they can be divided into two unrelated structurally-distinct classes (Table 1.1; Eriani et al, 1990). Class I aminoacyl-tRNA synthetases enzymes share the “HIGH” and “KMSKS” signature sequence motifs within an amino-terminal Rossmann fold domain (Cusack, 1997) as shown in Figure 1.5A. The Rossmann fold domain consists of alternating α-

helices and β-strands (Arnez and Moras, 1997; Ibba and Söll, 2000; First 2005)

and is the site of activation and aminoacyl-tRNA formation in class I AARS. The connective peptide insertion domain (CP1), which recognizes the acceptor stem of tRNA, splits the Rossmann fold domain into an amino-terminal half and a carboxy-terminal half. All class I AARS enzymes facilitate the attachment of amino acids to the 2’- hydroxyl of tRNA (First, 2005).

A divergent class of aminoacyl-tRNA synthetases, the class II aminoacyl- tRNA synthetase enzymes share a common catalytic core consisting of a seven- stranded β-sheet surrounded by several α-helices (Figure 1.5B; Cusack et al,

1990). All class II AARS enzymes share three conserved sequence motifs. Motif

1 assists in dimerization and the orientation of the other two motifs; motif 2 is involved in the binding of ATP, the amino acid, and the acceptor end of tRNA; and motif 3 interacts with ATP (First, 1995). All Class II AARS enzymes attach the amino acid to the 3’-hydroxyl of the terminal adenosine of tRNA except for the phenylalanyl-tRNA synthetase, which attaches phenylalanine to the 2’- hydroxyl (Moor et al, 1992). A further distinction between the class I and class II

AARS enzymes can be made with respect to tRNA interactions. Class I AARS enzymes bind the acceptor stem of tRNA from the minor groove side, whereas class II AARS enzymes bind from the major groove side.

The two classes of AARS enzymes are further divided into subclasses based on additional domains present (apart from the catalytic core), and quaternary structure (Arnez and Moras, 1997; First 2005). Among class I aminoacyl-tRNA synthetases, members of subclass Ia and Ib are monomers, while members of

subclass Ic are homodimers. Class II aminoacyl-tRNA synthetases have also been divided into three subclasses, but these differ vis-à-vis the location and topology of the tRNA anticodon binding domains. In subclass IIa, the anticodon binding domain is a mixed β-sheet located on the carboxy-terminal end of the catalytic core, except in alanyl-tRNA synthetase and seryl-tRNA synthetase, neither of which has a discernable anticodon binding domain (Biou et al,1994). In subclass IIb, the anticodon binding domain has a β-barrel structure with an oligonucleotide binding (OB) fold motif (Onesti et al, 1995). This domain is located on the amino-terminal end of the catalytic core. The catalytic and anticodon binding domains of phenylalanyl-tRNA synthetase, an AARS belonging to class IIc, are located on separate subunits (Mosyak et al, 1995). A non- canonical AARS recently discovered from certain members of Archaea,

Phosphoseryl-tRNA synthetase, has also been placed in class IIc (Sauerwald et al, 2005; Kamtekar et al, 2007).

1.2.3.2 Specificity of amino acid activation

Binding of ATP and cognate amino acid in the AARS active site occurs by a random mechanism, although for the aspartyl-tRNA synthetase, it has been suggested that ATP binding occurs prior to amino acid binding (First, 2005;

Eldred and Schimmel, 1972). For aminoacyl-tRNA synthetases, specificity of activation is facilitated through the use of differential binding energies for cognate and noncognate amino acids (Fersht et al, 1985; Fersht, 1998). In general,

isosteric and smaller amino acids can enter the binding pocket but binding is minimized through unfavorable hydrophobic and electrostatic interactions. Larger amino acids, on the other hand, are excluded sterically (Fersht, 1998). When these mechanisms are inadequate to differentiate cognate and non-cognate amino acids, editing mechanisms are employed to prevent mischarging with noncognate amino acids (Jakubowski, 2005). Pre-transfer editing occurs through hydrolysis of an enzyme-bound aminoacyl-adenylate intermediate formed with a non-cognate amino acid in many AARS enzymes that cannot sterically exclude or electrostatically repel these amino acids (Jakubowski, 2005). Another proofreading step known as post-transfer editing occurs when hydrolysis of the misacylated enzyme-bound tRNA occurs (Jakubowski, 2005). Both mechanisms fit the double sieve model proposed by A. Fersht (1979) in which exclusion is the first sieve and editing is the second.

For class I AARS enzymes, the amino acid binding pockets assume an open and relaxed confirmation, while the amino acid pockets in class II AARS enzymes are bound by rigid templates (Arnez and Moras, 1997). The rest of this section will focus on specific details of activation with respect to class II AARS enzymes. For detailed discussion of determinants of amino acid activation for specific class I AARS enzymes refer to recent reviews (Ibba and Söll, 2000; First,

2005).

In all class II AARS enzymes, the ATP binding pocket is well conserved. In the presence of ATP, the adenine ring of ATP stacks between the invariant motif

2 phenylalanine and the motif 3 arginine; further, motif 2 and motif 3 arginines

interact with the α- and γ-phosphates of ATP, respectively (First, 2005; Biou,

1994; Belrhali et al, 1994). Three Mg2+ ions stabilize the bent confirmation of the

ATP phosphates (First, 2005). A hydrogen bonding interaction between the

nitrogen at position 6 of ATP and a carbonyl of a main chain residue confers

specificity since this interaction can not occur with GTP (Belrhali et al, 1994;

Onesti et al, 1995; Cavarelli et al, 1994).

The amino acid binding pocket is directly adjacent to the ATP binding pocket.

The protonated α-amino group of the amino acid forms a strong hydrogen bond

with a conserved acidic residue (First, 2005; Belrhali et al, 1994). The invariant

motif 2 arginine bridges the α-carboxyl of the amino acid and the α-phosphate of

ATP in many class II aminoacyl-tRNA synthetases (Belrhali et al, 1994; Arnez et al, 1999). Specificity is achieved through extensive electrostatic and hydrophobic interactions. For polar amino acids, an extensive hydrogen binding network excludes hydrophobic interactions. Conversely, the class II AARS, phenylalanyl- tRNA synthetase achieves specificity for the hydrophobic amino acid phenylalanine through edge-to-face interactions with conserved phenylalanine residues and hydrophobic interactions with conserved glycines (Mosyak et al,

1995; Fishman et al, 2004). The hydrophobic nature of the deep phenylalanine binding pocket, and to a lesser extent, steric hindrance, prevents tyrosine from interacting (Fishman et al, 2004). Histidinyl-tRNA synthetase appears to be unique in that it uses an induced-fit mechanism as opposed to the “lock-and-key”

mechanism found among all other class II AARS enzymes. Although loops are

present other class II AARS enzymes, binding of histidine seems to invoke major

conformational changes in a number of loops to facilitate recognition and

specificity (Yaremchuk et al, 2001). A further method by which AARS enzymes

establish specificity is through metal ion mediated coordination. For example, the

threonyl-tRNA synthetase contains a catalytic Zn2+ ion that is responsible for

discriminating between threonine and the non-cognate amino acid, valine (Nureki

et al, 1993).

1.2.3.3 Specificity of amino acid attachment to cognate tRNA

Once an enzyme-bound aminoacyl-adenylate intermediate is formed with the

cognate amino acid and ATP substrates, the amino acid must be attached to the

correct tRNA. All tRNA species conform to the same general shape since

selective pressures require that they are recognized by elongation factor and the

ribosome (First, 2005; Schimmel and Söll, 1979). The initial binding interaction

between tRNAs and the aminoacyl-tRNA synthetases is the displacement of

bound water molecules by the phosphate backbone of tRNA which is relatively

nonspecific. Further, the region of interaction is extensive, minimizing the effect of a single base-specific interaction on the cumulative binding affinity. A low binding affinity ensures that the aminoacylated tRNA is released from the

enzyme (First, 2005; Schimmel and Söll, 1979). It is believed that differential binding affinities are inadequate to explain the specificity exhibited for cognate

tRNA molecules and that essential conformational changes are induced only upon binding of the cognate tRNA (Yarus and Berg, 1969; First, 2005). However, the binding energy that results from recognition of the cognate tRNA is used to catalyze the transfer of amino acid to the tRNA. The observation that AARS enzymes couple aminoacylation of tRNA to recognition of the cognate tRNA is consistent with this observation. It is likely that stabilization of the transition state of tRNA aminoacylation in AARS-tRNA complexes enhances specificity (Fersht,

1979).

Identity elements are specific nucleotides in the cognate tRNA which allow

AARS enzymes to distinguish cognate versus non-cognate tRNAs (McClain,

1993; Giege et al, 1998). Such identity elements include recognition elements which enhance recognition and antideterminants which hamper it. Generally, identity elements are situated in two hotspots, namely the anticodon stem and loop; and the acceptor stem (McClain, 1993; Giege et al, 1998). For some tRNA species, post-translational modification of bases of considered key identity elements confers specificity (Gangloff and Dirheimer, 1973). In AARS enzymes, the catalytic domain is often distant from the anticodon binding domain and/or other tRNA recognition elements. This spatial separation implies that there is communication between these distant regions to achieve the high level of accuracy of aminoacylation (Ibba and Soll, 2000; First 2005).

1.2.3.4 Indirect aminoacylation mechanisms

According to Crick’s adaptor hypothesis, there are twenty different aminoacyl-

tRNA synthetases each of which charges a cognate tRNA species with one of

the twenty common natural amino acids (Crick, 1958). The assumption that there

are twenty AARS enzymes holds true for a wide-range of organisms including E. coli, Saccharomyces cerevisiae, and humans. However, with the sequencing of a number of microbial organisms, it became clear that the paradigm of twenty did

not account for the apparent absence of key aminoacyl-tRNA synthetases in

many taxa. Two of the earliest sequenced organisms Methanocaldococcus

jannaschii and Methanothermobacter thermoautotrophicus lack homologs to

known glutaminyl-, asparaginyl-, lysyl-, and cysteinyl-tRNA synthetases even

though the respective amino acids are found in proteins from these organisms

(Tumbula et al, 1999; Francklyn et al, 2002; Prætorius-Ibba and Ibba, 2003). A

class I lysyl-tRNA synthetase (LysRS1) was discovered explaining the absence

of the class II lysyl-tRNA synthetase (LysRS2), which was the only known lysyl-

tRNA synthetase at the time (Ibba et al, 1997; Ibba et al, 1999). The mechanisms

by which organisms circumvent the need for the canonical glutaminyl-, asparaginyl-, and cysteinyl-tRNA enzymes have since been elaborated, and are the focus of this section. Aminoacylation of the initiator tRNA for methionine in eubacteria, mitochondria, and chloroplasts, known as tRNAfMet, is also discussed

briefly, as it is a known example of a modified amino acid inserted into proteins

co-translationally. The indirect route of aminoacylation for selenocysteine, a true

expansion of the genetic code, is discussed in Section 1.3. Most bacteria and all

sequenced archaea lack the canonical glutaminyl-tRNA synthetase (GlnRS) but

can synthesize glutaminyl-tRNAGln through an indirect pathway (Prætorius-Ibba

and Ibba, 2003). A non-discriminating glutamyl-tRNA synthetase (ND-GluRS)

with relaxed tRNA specificity misacylates tRNAGln with glutamate. The glutamyl

moiety is then converted to glutaminyl in a reaction catalyzed by the glutamine-

dependent glutamyl-tRNAGln amidotransferase (Ibba and Söll, 2005; Cathopoulis

et al, 2007). An analogous mechanism exists for the indirect aminoacylation and

biosynthesis of asparaginyl-tRNAAsn. In this case a non-discriminating aspartyl-

tRNA synthetase (ND-AspRS) mischarges tRNAAsn with aspartate. Aspartate is converted to asparagine on tRNAAsn in a reaction mediated by the asparagine-

dependent aspartyl-tRNAGln amidotransferase (Ibba and Söll, 2005; Cathopoulis

et al, 2007).

Organisms that lack a canonical cysteinyl-tRNA synthetase (CysRS) utilize a

recently discovered novel class II AARS enzyme (Sauerwald et al, 2005). The O-

phosphoseryl-tRNA synthetase (SepRS) aminoacylates tRNACys with O-

phosphoserine; subsequently, O-phosphoserine attached to tRNACys is converted

to cysteine in a reaction catalyzed by the O-selenophosphate-tRNA:cysteine-

tRNA synthase enzyme (Sauerwald et al, 2005). This pathway may also be the

only method for synthesizing cysteine in organisms lacking CysRS (Sauerwald et

al, 2005).

In the case of eubacterial, mitochondrial, and plastid translation,

formylmethionyl- tRNAfMet is used at initiation codons (Ibba and Söll, 2004).

Methionyl-tRNA synthetase (MetRS) aminoacylates this tRNA with methionine;

this moiety is then formylated to provide formylmethionyl-tRNAfMet for initiation

(Ibba and Söll, 2004). Once inserted into proteins cotranslationally, the N-formyl group of formylmethionine is cleaved by peptide deformylase (Giglione and

Meinnel, 2001). Since methionine, and not formylmethionine, is found in mature proteins, by convention, only the former is considered a genetically encoded amino acid (Ibba and Söll, 2004).

1.3 Selenocysteine: the 21st genetically encoded amino acid

Selenocysteine (Sec) is considered the 21st genetically encoded amino acid

(Figure 1.4A). It is the predominant biological form of selenium, a trace element

required for development in humans. Although relatively uncommon in proteins, it

is widespread in distribution across phyla, and is found in all many organisms of

all three Domains of life (Zinoni et al, 2000; Wilting et al, 1997; Hoffman and

Berry, 2005). Structurally, selenocysteine resembles cysteine, one of the twenty

common amino acids. The only difference is that in selenocysteine a selenium

atom replaces the sulfur of cysteine. Many selenium-containing proteins, also

known as selenoproteins, are enzymes that catalyze oxidation-reduction

reactions (Stadtman, 1996). In eukaryotes, selenoproteins serve two well-

characterized functions, namely, participation in anabolic processes and the

protection of cells from oxidative damage (Hatfield and Gladyshev, 2002).

Selenocysteine is found in the active site of selenium-containing enzymes and

carries out reactions catalyzed by these enzymes at a faster rate than would be possible if substituted with cysteine (Stadtman, 1996).

Universally, the mechanism of decoding the opal codon, UGA, as Sec requires a context-specific translational reprogramming event because UGA is one of the three canonical stop codons (Driscoll and Copeland, 2003). Another difference between selenocysteine and the twenty amino acids is that the biosynthesis of selenocysteine occurs on a specific tRNA (Commans and Böck,

1999; Hatfield and Gladyshev, 2002). Further, aminoacylation of this tRNA and insertion of the amino acid proceed in a manner different from the twenty amino acids. The twenty amino acids are cytosolic metabolites that are ligated on to specific tRNA species by cognate AARS enzymes. In contrast, there is no specific selenocysteine-tRNA synthetase; the formation of selenocysteine attached to the corresponding tRNA proceeds through the indirect route described below.

As shown in Figure 1.7, in Bacteria, four gene products are required for

synthesis and incorporation of selenocysteine (reviewed in Driscoll and

Copeland, 2003). The selC gene product is a highly specialized selenocysteinyl

tRNA (tRNASec) that is aminoacylated with serine by the conventional seryl-tRNA

synthetase (Heider et al, 1992; Commans and Böck, 1999). Conversion of seryl-

tRNASec to selenocysteinyl-tRNASec (Sec-tRNASec) is catalyzed by selenocysteine synthase (SelA), which uses an activated selenium moiety in the form of selenophosphate to modify the serine residue into Sec (Forchhammer and Böck,

1991). The synthesis of this selenophosphate is catalyzed by selenophosphate

synthetase (SelD) (Commans and Böck, 1999). Once charged with

selenocysteine, Sec-tRNASec is bound to a alternative elongation factor, bacterial

SelB (bSelB), for which it is highly specific; bSelB also binds GTP and the cis- acting mRNA stem-loop selenocysteine insertion sequence (SECIS) structure downstream of the UGA codon in the coding region of the mRNA (Forchhammer et al, 1989; Forchhammer et al, 1990; Förster et al, 1990; Berry et al, 1991). By a conformational change in this quaternary complex, GTP hydrolysis, and the partial unwinding of the SECIS element for translation of the downstream region, bSelB allows interaction of Sec-tRNASec with the ribosome (Huttenhofer and

Böck, 1998).

The insertion of selenocysteine in Eukarya involves a higher degree of

complexity (Driscoll and Copeland, 2003). Eukaryotic SECIS elements show little

nucleotide conservation, but all form a structurally similar hairpin present in the 3’

untranslated region (UTR) of selenoprotein mRNAs, sometimes several

kilobases downstream from the selenocysteine UGA codon (Shen et al, 1997).

Two proteins, namely SECIS Binding Protein 2 (SBP2) that binds the SECIS

Sec element and the tRNA specific elongation factor EF-Sec recruited by SBP2, replace the bSelB protein in eukaryotes (Fagegaltier, et al, 2000; Copeland et al,

2000; Copeland, 2003). At least one advantage of evolving a more complex system than found in eubacteria is the ability to regulate UGA recoding with more flexibility and to insert more than one selenocysteine into a single polypeptide

(Berry et al, 2005).

Research of the co-translational insertion of selenocysteine in Archaea

increased after the sequencing of the M. jannaschii genome (Bult et al, 1996).

Insertion of selenocysteine in archaeal species studied thus far have shown

striking similarities with the eukaryotic mode of insertion (Rother et al, 2001). In

M. maripaludis an open reading frame with two in frame UGA codons and a

SECIS element in the 3’ UTR has been found suggesting that more than one

selenocysteine may be inserted into a single polypeptide in archaea, as in some eukaryotic selenoproteins (Rother et al, 2001). Similarly, in all studied selenoproteins, the SECIS is in the UTR (in the 3’UTR with one exception in the

5’ UTR) (Böck et al, 2004). Archaeal SelB (aSelB) homologs lack the C-terminal

SECIS binding structural element in bSelB and do not bind SECIS elements in vitro (Rother et al, 2000). However, in vitro aSelB binds Sec-tRNASec to the

exclusion of seryl-tRNASec and other charged tRNAs suggesting a role similar to

that of Eukaryal EF-Sec (Rother et al, 2000). In fact, EF-Sec (also known as

eSelB) was discovered from sequence similarity with aSelB (Driscoll and

Copeland, 2003). It has also been demonstrated in vivo that aSelB is a

translation factor in Archaea (Rother et al, 2003).

1.4 Expansion of the genetic code with pyrrolysine

1.4.1 The methylamine methyltransferase genes possess single in-frame

amber codons.

As mentioned in Section 1.1.4.3, the process of methanogenesis from TMA,

DMA, or MMA is initiated by substrate-specific methyltransferases for each

methylamine. However, there is no obvious sequence similarity between

methyltransferases for different methylamines. Therefore, on sequencing the

methylamine methyltransferase genes from M. barkeri in our lab, it came as a

surprise that all methylamine methyltransferase genes shared a single in-frame

amber codon (TAG in DNA) within the open reading frames (Burke et al., 1998;

Paul et al., 2000). This in-frame amber codon is specific to methylamine methyltransferase genes, since it has not been found in the functionally analogous methanol methyltransferase genes (Sauer et. al, 1997; Deppenmeier et al., 2002; Galagan et al., 2002).

Table 1.2 indicates the various methylamine methyltransferases found in a few species of Methanosarcinaceae for which the genomes have been sequenced. A key feature of these genes is that they are present in multiple, nearly identical copies in the Methanosarcinaceae. Two copies of the TMA methyltransferase gene mttB have been found in M. acetivorans (Galagan et al.,

2002). Likewise, three copies of the DMA methyltransferase gene mtbB have been discovered in M. barkeri (Paul et al., 2000). The two genes encoding the

MMA methyltransferase in M. barkeri are denoted mtmb1 and mtmb2 and share

98% identity in amino acid deduced from sequence. In both MMA

methyltransferase genes from M, barkeri (the most-thoroughly studied members

of methylamine methyltransferases), the single in-frame amber codon is located

at codon position 202 followed by a downstream ochre (TAA) stop codon at

codon position 458 (Burke et al., 1998).

The cold-adapted Antarctic methanogen Methanococcoides burtonii also possesses methylamine methyltransferase genes each with a single in-frame amber codon (Saunders et al., 2003; Goodchild et al., 2004). Surprisingly, a non- methanogenic Gram-positive bacterium, Desulfitobacterium halfniense, also possesses an mttB homolog with an in-frame amber codon deduced to correspond to the same position as those experimentally validated in other mttB

genes in methanogens (Srinivasan et al., 2002). For each methylamine

methyltransferase gene, the in-frame amber codon is found at exactly the same

position for that methyltransferase in all organisms for which genomes are

available (Burke et al., 1998; Paul et al., 2000; Galagan et al., 2002;

Deppenmeier et al., 2002; Saunders et al., 2003).

On analyzing the isolated TMA, DMA, and MMA methyltransferase

monomers, it was discovered that each had a molecular weight of approximately

50 kDa. This mass would only be obtained if translation did not cease at the

internal in-frame amber codon but at a subsequent ochre or opal stop codon. If translation had stopped at the internal in-frame amber codon, a 23-kDa MMA

methyltransferase, 38-kDa DMA methyltransferase, or 32-kDa TMA

methyltransferase protein would have resulted. The 50-kDa TMA, DMA, or MMA

methyltransferase is thus referred to as the full-length readthrough product; the

smaller fragments are truncated gene products. Consistent with readthrough of

the in-frame amber codon and with the active isolated protein, little or no UAG-

truncated 23-kDa mtmB gene product was detectable from M. barkeri cells

(James et al., 2001).

1.4.2 Structural analysis of Monomethylamine methyltransferase

With the isolation of the monomethylamine methyltransferase, MtmB from M.

barkeri, it became possible to examine the residue corresponding to the in-frame

UAG codon within the protein. One of the possibilities suggested early on was

that the codon specified a specialized residue (Burke et al, 1998). Digesting

MtmB with trypsin and analyzing the resulting fragments by mass spectrometry

and Edman degradation revealed that the UAG corresponds to lysine (James et

al, 2001). A final identification of the residue was not possible, however, due to

the harsh conditions of peptide isolation; it was possible that the residue was not

lysine, but in fact a modified lysine. To resolve this issue, the X-ray crystallographic structure of MtmB was determined. Two forms of MtmB were

resolved to 1.55 Å and 1.7 Å in NaCl and (NH4)2SO4, respectively (Figure 1.8;

Hao et al, 2002).

MtmB is a homohexamer formed from a dimer of trimers. Each subunit

exhibits an α/β TIM barrel fold (Hao et al, 2002). The eight-stranded β-barrel

forms a deep anionic cavity. Positioned within this cavity is the UAG-encoded

residue. Analysis of electron density maps for the two crystallized forms led to

the assignment of this amino acid as (4R, 5R)-4-substituted-pyrroline-5-

carboxylate in amide linkage to the epsilon nitrogen of L-lysine (Hao et al, 2002).

This amino acid was named L-pyrrolysine (Hao et al, 2002). Taken together with

the UAG-decoding machinery discovered in M. barkeri (discussed in the following

chapter), pyrrolysine became known as the 22nd genetically encoded amino acid

(Atkins and Gesteland, 2002).

An unresolved matter concerned the identity of the substituent at C4 position

of the pyrroline ring. Analysis of derivatized pyrrolysine in crystals of MtmB led to

the assignment of this substituent as a methyl group (Hao et al, 2004). Accurate

mass of pyrrolysine from tandem mass spectrometry has confirmed that the substituent is a methyl group (Soares, 2005). With the identity of pyrrolysine established, it has been possible to chemically synthesize this amino acid (Hao et al, 2004).

Mass spectrometric analysis has revealed that the UAG-encoded residue in

MttB and MtbB is also pyrrolysine (Soares, 2005). Proteomic evidence from M. burtonii suggests that the UAG-encoded residue in MttB from this organism is also pyrrolysine (Goodchild et al, 2004). The distribution among all methylamine methyltransferases is consistent with a proposed role in catalysis. Briefly, it has been proposed that pyrrolysine helps activate and orient methylamines for methyltransfer (Hao et al, 2002; Hao et al, 2004; Krzycki, 2004).

1.4.3 tRNAPyl: an amber-decoding tRNA

An absolute requirement for the genetic encoding of an amino acid is that a tRNA specific for this amino acid brings it to the ribosome for incorporation at a

specific mRNA codon. Since the codon corresponding to the novel amino acid

pyrrolysine is the amber codon (UAG) that usually indicates stop, contigs of the

genome Methanosarcina barkeri Fusaro were scanned for an amber-decoding

tRNA with a complementary CUA anticodon (Srinivasan et al, 2002). A

bioinformatics approach yielded the pylT gene whose predicted product is the

UAG-decoding tRNA. This tRNA was detectable from unfractionated tRNA

samples by Northern blotting (Srinivasan et al, 2002). As mentioned in a

subsequent section, this tRNA is involved in the translation of UAG as pyrrolysine

Pyl and is known as tRNA or tRNACUA. The pylT gene that encodes this tRNA is an

indicator of pyrrolysine insertion in organisms such as M. barkeri, M. acetivorans,

M mazei, and M. burtonii. Remarkably, this gene has also been found in the

gram-positive eubacterial species, Desulfitobacterium halfniense and from

metagenomic libraries of a symbiont found in a gutless oligochaete (Srinivasan et

al, 2002; Nonaka et al, 2006; Zhang and Gladyshev, 2007) .

As shown in Figure 1.9, the clover-leaf structure of tRNAPyl displays a number

of unusual features. The anticodon stem forms from six base pairs, rather than

the more common five (Srinivasan et al, 2002; Theobald-Dietrich et al, 2004). A

variable arm that is smaller than the norm is also found in tRNAPyl. Further, many

bases generally conserved such as the GG sequence of the D-loop and TψC

sequence of the TψC loop are absent (Srinivasan et al, 2002; Theobald-Dietrich

et al, 2004). The tRNA with the structure most similar to that of tRNAPyl is the

bovine mitochondrial serine tRNA (Theobald-Dietrich et al, 2004). In vitro assays

indicated that lysylated-tRNAPyl with a mutated anticodon allowing the attachment of lysine, is recognized by the canonical bacterial elongation factor Ef-Tu

(Theobald-Dietrich et al, 2004). Only two modified bases have been detected

from tRNAPyl from mass spectrometric analysis, 4-thiouridine at position 8 and 1-

methyl pseudouridine at position 50; the lack of modifications is thought to

explain the suitability of transcripts of tRNAPyl in aminoacylation (Polycarpo et al,

2004). Recently, the discriminator base G73 and the first base pair in the

acceptor stem were shown to be major identity elements in D. halfniense tRNAPyl

(Herring et al, 2007).

1.4.4 Aminoacylation of tRNAPyl

As pyrrolysine is a modified lysine, lysylation of tRNAPyl by an AARS was

postulated to be the feasible first step in pyrrolysine biogenesis (Srinivasan et al,

2002). As discussed in greater detail in the next two sections, PylS and the canonical lysyl-tRNA synthetases were suggested as candidate enzymes for catalyzing the ligation of lysine to tRNAPyl. Three leading possibilities for

aminoacylation reactions are shown in Figure 1.10. Since pyrrolysine is found in

proteins, the formation of lysyl-tRNAPyl would indicate that prior to insertion, an indirect tRNA-dependent modification of lysine to pyrrolysine is required in order to safeguard translational accuracy (Srinivasan et al, 2002; Polycarpo et al,

2003).

1.4.4.1 PylS is a class II aminoacyl-tRNA synthetase

With the discovery of a UAG-decoding tRNA gene (pylT) near the MMA methyltransferase gene cluster in M. barkeri, questions arose vis-à-vis aminoacylation of this tRNA. What is the amino acid ligated to this tRNA? Is one of the common twenty amino acids attached followed by biosynthesis of pyrrolysine, or is pyrrolysine directly attached? Is a previously described AARS enzyme responsible for catalyzing this reaction or is this function accomplished by a hitherto unknown enzyme?

Sequencing of M. barkeri MS DNA was a first step in resolving these questions. Three open reading frames (ORFs) were found immediately following the pylT gene and were denoted pylS, pylB, and pylC (Srinivasan et al, 2002).

Co-transcription of these ORFs was thought a likely possibility since Northern blotting with probes specific for pylT, pylS, pylB, and pylC resulted in the detection of a 4.2 kilobase transcript (Srinivasan et al, 2002). An ORF found further downstream was named pylD. The pylB, pylC, and pylD gene products were proposed to play a role in the biosynthesis, based on sequence similarity with proteins of known function. Interestingly, one of the genes identified, pylS, revealed a predicted protein product with a domain resembling that of the core catalytic domain of a number of Class II AARS enzymes (Figure 1.11).

Assignment of PylS as a canonical AARS belonging to one of the known

subclasses of Class II enzymes was not possible due to significant predicted

sequence similarity between the core domain and that of enzymes from three

different subclasses. (Srinivasan et al, 2002).

To test the activity of the pylS gene product, the gene was cloned and

overexpressed in E. coli. The protein resulting from the heterologous expression

of pylS was engineered with an amino-terminal hexahistidine tag to facility

isolation by nickel-affinity chromatography. In filter-binding aminoacylation

assays, PylS attached radiolabeled lysine the UAG-decoding tRNA, thus

prompting it to be denoted the third lysyl-tRNA synthetase from M. barkeri

(Srinivasan et al, 2002). Searches of sequenced genomes revealed that the pylS

gene is always associated with pyrrolysine insertion (Srinivasan et al, 2002;

Galagan et al, 2002; Deppenmeier, et al 2002; Goodchild et al, 2002; Zhang et

al, 2005; Zhang and Gladyshev, 2007). Along with pylT and the other pyl genes,

pylS is found in all sequenced Methanosarcina species, Methanococcoides, and a gram-positive bacterium Desulfitobacterium halfniense. Remarkably the pylS

gene is split in D. halfniense into two halves called pylSn and pylSc (Srinivasan

et al, 2002).

The intact PylS protein (from M. barkeri MS pylS) is notoriously unstable when

purified from E. coli with an amino-terminal hexahistidine tag; thus, a number of

pylS genes from diverse organisms with sequences cloned to add different tags

have been cloned and expressed (Polycarpo et al, 2003; Blight et al, 2004;

Polycarpo et al, 2004; Herring et al, 2006; R. Jiang, personal communication). A

reason for this instability may be the disordered nature of the linker region found joining the amino-terminal domain with the core catalytic domain in intact PylS

(R. Jiang, personal communication).

Recently, crystal structures of the core catalytic domain of PylS of M. mazei

associated with various ligands has been solved to 1.8-2.5 Ǻ (Kavran et al,

2007). It has also been suggested from phylogenetic analysis of structure-based alignments, that PylS is a class IIc AARS that evolved from an ancestral phenylalanyl-tRNA synthetase and that this split occurred prior to the last universal common ancestor state (Kavran et al, 2007).

1.4.4.2 Non-homologous lysyl-tRNA synthetases in Methanosarcina spp.

As mentioned in the preceding section, along with the discovery of tRNAPyl, it was reported that recombinantly expressed PylS aminoacylates this tRNA with lysine in vitro. However, the unstable nature of the recombinant PylS used in that study prompted extensive studies using more stable variants of recombinant

PylS (Polycarpo et al, 2003; Blight et al, 2005; Polycarpo et al, 2005).

As pointed out in a preceding section, all AARS enzymes are categorized into two mutually exclusive classes, termed class I and class II, based on structurally distinct catalytic domains. This dichotomy is nearly universally conserved with the notable exception of the lysyl-tRNA synthetases (LysRS), which have recently been found in both class I and class II. On analysis of the genome sequence of

Methanocaldococcus janaschii, a class I LysRS (LysRS1) was discovered

whereas all LysRS enzymes known up until then were class II LysRSs (LysRS2)

(Ibba et al, 1997). Since that discovery, analysis of sequenced genomes has

shown that class I LysRSs are found in most archaea and a few bacteria (Ibba et

al, 1999; Ambrogelly et al, 2002). Detailed analysis of LysRS1 and LysRS2

indicates that these enzymes display differential specificity for non-cognate

amino acids and differential resistance to inhibitors; this may explain why two

unrelated LysRS enzymes exist (Levengood et al, 2004). Although LysRS1 and

LysRS2 are usually not found together, species of Methanosarcina have been

shown to possess both (Ambrogelly et al, 2002). The only other well-documented

example of both LysRS1 and LysRS2 being present in the same organism is in

the non-pyrrolysine containing Bacillus cereus, where they have been shown to

lysylate tRNAother (Ataide et al, 2005). Aminoacylation of tRNAPyl with lysine

suggested that PylS was the third LysRS from that organism, although later

studies both in vivo and in vitro show that this observation is not physiologically relevant (Polycarpo et al, 2003; Blight et al, 2004).

There are two tRNALys isoacceptors in species of Methanosarcina, one with a

UUU anticodon and another with a CUU anticodon (Galagan et al, 2002;

Deppenmeier et al, 2002; Maeder et al, 2006). In aminoacylation reactions

performed in vitro, it was shown that M. barkeri LysRS2 aminoacylates both

Lys equally well, whereas LysRS1 aminoacylates tRNA CUU well but aminoacylated

Lys tRNA UUU poorly (Ambrogelly et al, 2002; Polycarpo et al, 2003). It was

suggested from in vitro observations that M. barkeri LysRS1 and LysRS2 charge

tRNAPyl with lysine when present together, and that neither PylS nor the canonical LysRS enzymes can charge tRNAPyl alone (Polycarpo et al, 2003).

Structural modeling studies showed that both LysRS1 and LysRS2 could be

docked on to the tRNAPyl species with minimal steric hindrance. Selective

inhibition suggested that LysRS2 catalyzed the ligation of lysine to tRNAPyl,

whereas LysRS1 refolded tRNAPyl to an active conformation (Polycarpo et al,

2003). In B. cereus aminoacylation of a small RNA of unknown function, tRNAother, is mediated by the concerted action of LysRS1 and LysRS2 in much

the same way that tRNAPyl is lysylated in Methanosarcina, even though the former does not have any of the components of the pyrrolysine insertion pathway

(Ataide et al, 2005).

1.4.5 Contextual requirements for pyrrolysine insertion

Like selenocysteine, pyrrolysine is translated in response to a canonical stop codon. As noted in section 1.3, the insertion of selenocysteine is in response to a cis-acting stem-loop structure referred to as the SECIS element. The SECIS element is absolutely essential for the translation of UGA as selenocysteine.

Along with the discovery of pyrrolysine, theoretical explanations of how contextual elements might be required for UAG translation of this amino acid abounded (Atkins and Gesteland, 2002; Namy et al, 2004, Beebe and Schimmel,

2004). From genomes, it was predicted that only 4% of ORFs terminated with

UAG in species of Methanosarcina. Bioinformatics approached revealed a

putative stem-loop structure immediately downstream of the UAG codon in the

mtmB1 transcript (Pottenplackel, 1999; Namy et al, 2004). It was suggested this

element, named the PYLIS element, performed a function similar to the SECIS

element that is required for selenocysteine insertion (Namy et al, 2004).

Structural probing of mtmB transcripts revealed that this structure forms in vitro;

however, the biological relevance of this element was not ascertained (Theobald-

Dietrich et al., 2005). Stem-loops were subsequently found in the transcript of the

mttB homolog from D. halfniense (Ibba and Söll, 2004). Another stem-loop was also discovered in the transcript of mtbB. Unlike the SECIS element, the various forms of the PYLIS-like elements show no obvious structural similarity to the mtmb1 PYLIS. Yet another bioinformatics study predicted that pyrrolysine and selenocysteine used dissimilar strategies for suppression of termination during translation (Zhang et al, 2005).

A recent study from our lab provided the first experimental examination of mRNA contextual signals in Methanosarcina (Longstaff et al, 2007a). It was established that the PYLIS element is dispensable for the translation of UAG as pyrrolysine. However, through mutational analysis, it was demonstrated that the

PLYIS element enhances the incorporation of pyrrolysine at the expense of

termination at UAG. Because pyrrolysine is proposed to play an essential role in methanogenesis from methylamines, it is suggested that such an enhancement

might be beneficial from a physiological standpoint. (Longstaff et al, 2007a).

1.5 Artificial expansion of the genetic code with unnatural amino acids

Why did nature allow the expansion of the genetic code to include pyrrolysine and selenocysteine? Analysis of the chemical properties of these two amino acids may hold a key clue to answering this question. Both selenocysteine and pyrrolysine have unique chemistries that underlie the presence of these amino acids in key enzymes. For modern organisms, the cost in terms of energy required to encode these amino acids is more than offset by the benefits of the enhanced or expanded metabolism. Clearly, selection pressures act in these organisms to ensure that the complex decoding machinery required for selenocysteine or pyrrolysine insertion is retained.

While selenocysteine and pyrrolysine presage the discovery of additional components to the genetic code that might serve to execute specific roles in specialized niches, it is already feasible to insert unnatural amino acids in vivo by manipulating the translational system of the host organism (Wang et al, 2001;

Wang and Schultz, 2004; Wang et al, 2006; Hendrickson et al, 2004; Tang and

Tirrell, 2002; Kiick et al, 2001). Often an auxotrophic strain can grow in media in which a natural amino acid is replaced by an unnatural structurally analagous amino acid. In this case, the unnatural amino acid substitutes for the naturally occurring one (Hendrickson et al, 2004). Another method involves relaxing the amino acid specificity of the aminoacyl-tRNA synthetase, often by diminishing proofreading activity (Doring et al, 2001; Hendrickson et al, 2004). In Xenopus laevis oocytes, microinjection of chemically misacylated tRNAs and mutant

mRNA allows the incorporation of unnatural amino acids into translated protein

(Ho and Kan, 1987).

Recently, it has become possible to co-translationally insert unnatural amino

acids by the introducing cognate tRNA/AARS pairs into host cells. Such a

methodology requires the tRNA/AARS pair to be orthogonal and that there be

minimal interaction of the tRNA and AARS with the endogenous AARS or tRNA,

respectively, of the host cell. However, the aminoacylated tRNA must interact

with the host elongation factor and ribosome. The natural amino acid specificity

of the orthogonal pair can be changed by mutation to the desired unnatural

amino acid (Wang et al, 2001; Wang and Schultz, 2004; Wang et al, 2006;

Hendrickson et al, 2004; Tang and Tirrell, 2002; Kiick et al, 2001). The anticodon

of the tRNA is often changed to recognize rarely used stop or sense codons.

Orthogonal tRNA species that recognize four or five nucleotide frame-shift

codons have also been designed (Hohsaka and Sisido, 2002). The UAG stop

codon is widely used in many organisms due to the infrequent use of this codon

to signal termination. Specificity of the AARS for the unnatural amino acid (and not for a natural one) is achieved through cycles of mutagenesis and selection.

When the genes for the orthogonal pair are introduced into the host organism and a source of the amino acid is established (either exogenously as supplied or endogenously by host biosynthesis), the amino acid is inserted at codons complementary to the anticodon of the orthogonal tRNA (Wang et al, 2001; Chin et al, 2003).

Unnatural amino acids that are fluorescent, glycosylated, or chemically

reactive have been incorporated throughout proteins. The artificial expansion of

the genetic code may allow improved studies of protein structure and function, as

well as generate proteins that have novel applications in biotechnology and

medicine (Wang and Schultz, 2004).

1.6 Overview of work presented in following chapters

The work presented here examines key aspects of UAG translation as

pyrrolysine. The first study presents proof of a direct association between the pyl genes and methylamine metabolism through the analysis of a Methanosarcina acetivorans mutant lacking a section of the genome containing the gene encoding for tRNAPyl. Work presented in the following study, in conjunction with work done by others, establish that PylS is a class II aminoacyl-tRNA synthetase specific for pyrrolysine. This discovery, established PylS as the first aminoacyl- tRNA synthetase with specificity for a natural amino acid that is outside of the

common twenty. Work presented in the following study examines the in vivo relevance of the indirect route of aminoacylation employing the class I and class

II lysyl-tRNA synthetases to translation of pyrrolysine. In the final study, the

amino acid specificity of PylS is probed through the use of synthetic analogs of

pyrrolysine.

Figure 1.1 A phylogenetic representation based on 16S rRNA sequences. A stylized representation of the Euryarchaeota with representative methanogens is shown. (Figure courtesy of J.A. Krzycki).

Figure 1.2 Cells of the methanogen Methanosarcina barkeri. The cells were made visible with UV light in a fluorescence microscope. Autofluorescence is due to the presence of the electron carrier F420. An individual cell is approximately 1.7 µm in diameter. (Image courtesy of J.G. Zeikus).

Figure 1.3 Pathways of methylotrophic methanogenesis. The oxidative branch of the pathway which yields reducing equivalents (2 e-) indicated by red arrows. The green arrow indicates the reductive step, namely the formation of methane from methyl-CoM. For clarity of representation, enzymes catalyzing these key steps have been omitted from the schematic. Both the oxidative and reductive steps of methanogenesis are used in methylotrophic methanogenesis from all known methylated substrates. The initial reactions of methyltransfer from methylamines (blue) and methanol (gray) are indicated with enzymes demonstrated to catalyze the steps indicated above the corresponding arrows. (Schematic based on Burke and Krzycki, 1995; Ferguson et al, 1996; Ferguson and Krzycki, 1997; Thauer 1998; Ferguson and Krzycki 2000; Deppenmeier, 2002; Galagan et al, 2002).

Figure 1.4 Common and uncommon amino acids specified by the genetic code. Three of the original set of twenty amino acids (left side) are compared with new additions to the genetic code (right side). The carbon-backbone is indicated in green, oxygen atoms in red and nitrogen atoms in blue. Panel A shows stick-figure representations of cysteine and selenocysteine. Cysteine is similar to selenocysteine except the sulfur atom (yellow) has been replaced by selenium (orange). Panel B shows stick-figure representations of lysine and pyrrolysine. Pyrrolysine resembles lysine except for the addition of a chemically- reactive five-membered pyrroline ring. Panel C shows stick-figure representations of phenylalanine and the unnatural amino acid benzyoylphenylalanine which was added to the genetic code by artificial means (Chin, et al, 2003). Each of the new additions bring new reactivities previously lacking in other genetically-encoded amino acids. The common amino acids are for comparison and no precursor relationship should be inferred. (Figure modified from Mahapatra and Krzycki, 2007)

Figure 1.5 Structural variations between the two classes of aminoacyl-tRNA synthetases. The structure of a monomer of a class I AARS, LysRS1 (Panel A), is compared with that of a monomer of a class II AARS, LysRS2 (Panel B). The upper arrows indicate the Rossman-fold domain that is characteristic of class I AARS enzymes compared with the class II AARS core which consists of a seven-stranded β-sheet flanked by α-helices. The lower arrows indicate the different domains responsible for tRNA anticodon binding. In LysRS1 this domain is known as the SC-domain, whereas in LysRS2 (and in all other class IIb members) it is the OB-fold domain. The x-ray crystallographic structure of LysRS1 was resolved by Terada, et al (2002) and that of LysRS2 by Onesti, et al (2000) to 2.6 and 2.7 Å, respectively. Coordinates for LysRS1 (PDB Id: 1irx) and LysRS2 (PDB Id: 1bbu) were obtained from the Protein Data Bank. Structures were visualized using PyMOL software (documentation at DeLano, 2002).

Figure 1.6 Comparison of the structure of tRNA bound to ternary complex with that of bacterial release factor. The structure of Thermus aquaticus ternary complex at 2.6 Å resolution (PDB Id: 1b23; Nissen et al, 1999) in Panel A is compared with the structure of T. thermophilus release factor-2 at 2.5 Å resolution (PDB Id: 2ihr; Dobbek et al, 2007) in Panel B. Coordinates were downloaded from Protein Data Bank and structures were visualized using PyMOL (DeLano, 2002).

Figure 1.7 Genetic encoding of selenocysteine. The reactions catalyzed by bacterial enzymes are designated. Once made, selenocysteinyl-tRNA binds to a specific elongation factor (SelB and homologs) that rejects all other aminoacyl- tRNA species. The binding of the specific elongation factor to the SECIS allows selenocysteinyl-tRNA to compete with a translation release factor that would otherwise bind to the UGA codon and cause termination of protein synthesis (Figure modified from Mahapatra and Krzycki, 2007).

Figure 1.8 Resolved x-ray crystallographic structures of monomethylamine methyltransferase. Structures of monomers of MtmB crystallized using NaCl (Panel A) and (NH4)2SO4 (Panel B) were resolved to 1.55 Å and 1.7 Å, respectively (Hao et al, 2002). The surface structure is shown with residue 202, named pyrrolysine, indicated in yellow. Pyrrolysine is inside a deep cavity in the enzyme. Coordinates (PDB Ids: 1irx and 1l2q) were downloaded from Protein Data Bank and structures visualized using PyMOL (DeLano, 2002).

Figure 1.9 Cloverleaf secondary structure of tRNAPyl. Bases conserved among identified forms from cultivable species are in blue with variant bases in lavender. Numbering of bases in red is based on structural analysis (Theobald- Dietrich et al, 2004) and according to the system of Sprinzl, et al (1987). Bases commonly found in other tRNA species that are missing from tRNAPyl are indicated by black circles. The invariant acceptor stem CCA, the discriminator base and the anticodon are boxed. (Figure based on Srinivasan et al, 2002; Theobald-Dietrich et al, 2004).

Figure 1.10 Three leading possibilities for tRNAPyl synthesis and aminoacylation. Since pyrrolysine resembles lysine, two possible scenarios of how tRNAPyl is aminoacylated focus on the attachment of lysine to the tRNA, followed by synthesis of pyrrolysine on the tRNA (Scheme 1 and 2). These indirect routes are analogous to the aminoacylation and synthesis of selenocysteine (Driscoll and Copeland, 2003). A third possibility is that pyrrolysine is synthesized prior to aminoacylation and attached to tRNAPyl by a non-canonical aminoacyl-tRNA synthetase such as PylS (Scheme 3). A number of enzymes have been proposed to catalyze aminoacylation including the non- canonical aminoacyl-tRNA synthetase, PylS (Scheme 1 and 3; Srinivasan et al, 2002), and also a complex of the canonical lysyl-tRNA synthetases (Scheme 2; Polycarpo et al, 2003). The PylB, PylC, and PylD enzymes have been hypothesized to catalyze steps in the biosynthesis of pyrrolysine (Srinivasan et al, 2002) from lysine bound to tRNAPyl (Scheme 1 and 2) or as a free cytosolic metabolite (Scheme 3) (Krzycki, 2005).

Figure 1.11 Alignments of motif 2 and motif 3 sequences from select class II aminoacyl-tRNA synthetases. Motifs 2 and 3 are located in loops close to the active site of class I AARS enzymes and participate in binding of ATP, amino acid substrate, and the acceptor end of tRNA. Motif 1 (not shown here) is another conserved sequence motif that forms part of the dimer interface. Here X is any amino acid; + represents R, K, H; - is D or E; φ represents a hydrophobic amino acids (I, L, M, F, W, V); and λ represents small amino acids (A, C, G, P, S, T). Organism abbreviations are as follows: MbMS, M. barkeri MS; Ma, M. acetivorans; Dh, D. halfniense; Ec, E. coli; Pk, Pyrrococcus kodakaraensis; and Tt, Thermus thermophilus. The data in this table was collated from the published sequences and structures of the following enzymes: PylS (Srinivasan et al, 2002; Krzycki, 2005); LysRS2 (Desogus et al, 2000); AspRS (Schmitt et al, 1998); AsnRS (Berthet-Columinas et al, 1998); HisRS (Arnez et al, 1997); and the α- subunit of PheRS (Mosyak et al, 1995).

Class Subclass AARS Tertiary Structure Class I Ia ArgRS α

CysRS α IleRS α LeuRS α

MetRS α2 ValRS α

Ib GlnRS α GluRS α LysRS1 α

Ic TrpRS α2

TyrRS α2

Class II IIa GlyRS α 2 HisRS α2

ProRS α2

SerRS α2 ThrRS α2

IIb AspRS α 2 AsnRS α2

LysRS2 α2

IIc AlaRS α4 GlyRS α2β2 PheRS α β 2 2

Table 1.1 Aminoacyl-tRNA synthetases for the twenty common amino acids. (Table adapted from data taken from First, 2005; Ataide and Ibba, 2006).

Organism Substrate Gene Reference M. barkeri MS TMA mttB (Paul et al, 2000) DMA mtbB1, mtbB2, mtbB3 (Paul et al, 2000) MMA mtmB1, mtmB2 (Burke et al, 1998) M. acetivorans TMA mttB1, mttB2 (Galagan et al, 2002) DMA mtbB1, mtbB2, mtbB3 (Galagan et al, 2002) MMA mtmB1, mtmB2 (Galagan et al, 2002) M. mazei TMA mttB1, mttB2 (Deppenmeier et al, 2002) DMA mtbB1, mtbB2, mtbB3 (Deppenmeier et al, 2002) MMA mtmB (Deppenmeier et al, 2002) M. burtonii TMA mttB1, mttB2 (Goodchild et al, 2004) DMA mtbB1, mtbB2 (Goodchild et al, 2004) MMA mtmB1, mtmB2 (Goodchild et al, 2004)

Table 1.2 Distribution of methylamine methyltransferase genes among select members of Methanosarcinaceae. Genes identified or predicted as bearing in-frame UAG codons are listed. The data in the table is collated from the references specified.

CHAPTER 2

CHARACTERIZATION OF A MUTANT

UNABLE TO TRANSLATE UAG AS PYRROLYSINE

2.1 Introduction

As mentioned in Chapter 1, pylT, which encodes a UAG-decoding tRNA, and pylS, which encodes a class II aminoacyl-tRNA synthetase, were discovered near the mtmb1 gene of M. barkeri (Srinivasan et al, 2002). The same genes are maintained in M. acetivorans (Figure 2.1). As also discussed in Chapter 1, the pylT and pylS genes are adjacent to three other reading-frames designated pylB, pylC and pylD. In at least one species of Methanosarcina, the pylTSBC genes form a transcriptional unit. Homologs of these genes are also found in D. halfniense in the same order found in species of Methanosarcina indicating that lateral gene transfer of the entire pyl genes is likely and indicating that pylD is probably associated with the pyl genes (Srinivasan et al, 2002).

Despite the fact that tRNAPyl is central to theories on how pyrrolysine is

encoded by UAG, there has been no previous in vivo demonstration of the

requirement for tRNAPyl in the process. This was not originally possible due to the

lack of genetic tools for species of Methanosarcina. As tools have been recently

developed for M. acetivorans (Metcalf et al, 1997; Rother and Metcalf, 2005) and

the entire genome has been sequenced (Galagan et al, 2002), we proceeded to

study genetic encoding of pyrrolysine in this organism. Here, we show that a

direct replacement of the pylT gene with an antibiotic resistance marker gene in

M. acetivorans is not lethal, but does lead to pleiotropic lesions with respect to

growth on methylamines. The results we obtain here provide a ready

physiological rationale for the retention or recruitment of pyrrolysine to the genetic code of species of Methanosarcina, and demonstrate the feasibility of

performing genetic analyses of pyrrolysine metabolism.

2.2 Experimental Procedures

2.2.1 Organisms, growth conditions and reagents

Methanosarcina acetivorans C2A (Sowers et al, 1984) was cultured

anaerobically at 37°C in DSM medium 304 (Sowers and Schreier, 1995) with the

following modifications: 10 mL of vitamin solution from DSM medium 141

(Sowers and Schreier, 1995) was substituted for yeast extract; Na2CO3 was

replaced with 20 mM 3-[N-morpholino]propanesulfonic acid (MOPS) pH 7.0 as

the buffer; and N2 gas was used instead of N2:CO2 gas mix. Cultures were grown

in 10 mL of this modified DSM medium in 27 mL rubber-stoppered test tubes

for growth studies; in 50 mL medium in rubber-stoppered serum-vials for isolation

of nucleic acids and enzyme assays; or in 15 L medium in a 20 L stirred carboy

for protein extraction for immunoblotting. The growth substrates were 100 mM

methanol, 40 mM MMA, 40 mM DMA, 40 mM TMA or 40 mM sodium acetate.

The methylamines were added as hydrochloride salts. For growth in carboys, a

combination of both 100 mM methanol and 40 mM MMA was added as growth

substrates. For testing nitrogen utilization, cultures in exponential growth phase

on methanol and ammonia were transferred to medium lacking added ammonia

and prepared under argon gas instead of the standard N2 gas mentioned above.

Growth substrates were 100 mM methanol, 40 mM MMA, or a combination of

both.

2.2.2 Construction of mutant strain

The mutant strain used in this study (hereafter designated as ∆ppylT)

constructed by Asmita Patel is described in detail elsewhere (Mahapatra et al,

2006). Liposome-mediated transformation in M. acetivorans was performed in

the lab of William Metcalf using established procedures (Metcalf et al, 1997;

Rother and Metcalf, 2005).

2.2.3 Isolation of DNA and tRNA

Isolation of genomic DNA from 50 mL cultures of M. acetivorans strains in

late-exponential or stationary phase followed the method of Sowers (1995). Cells

were pelleted at 5000 g and resuspended in 10 mL of 10 mM Tris HCl, 1 mM

EDTA, 0.4 M NaCl, and 0.04 M MgSO4, at pH 8.0. Cells were lysed by adding 0.3

mL of 10% SDS. DNA was extracted twice with phenol : chloroform : isoamyl

alcohol in the ratio 25:24:1 (Ambion, Austin, TX), once with chloroform, and

ethanol precipitated and resuspended in TE buffer (Sambrook et al, 1989) prior

to treatment with RNase A (Roche, Pleasanton, CA). The extraction step was

repeated to remove RNase A and the DNA was ethanol precipitated and

resuspended in TE buffer.

Unfractionated tRNA was isolated from exponentially-growing cells in medium

containing 100 mM methanol and 40 mM MMA according to methods described

previously (Jester et al, 2003; Blight et al, 2004) with the exception that the glass

bead mixing step in Jester et al, 2003 was omitted. Cells from 50 mL cultures

were anaerobically pelleted by centrifugation at 15,000 g for 10 minutes at 4°C.

All subsequent steps were also carried out on ice or at 4°C. The pellet obtained was rinsed with 300 µL of chilled 0.3 M sodium acetate and 10 mM EDTA, pH

4.5. The pellet was then resuspended in 300 µL of the same solution. The RNA was extracted twice with acid phenol : chloroform (5:1) buffered at pH 4.5

(Ambion). A cycle of vortexing four times for 30 s each was followed by centrifugation at 18,000 g for 15 min at 4°C. The RNA was ethanol precipitated

twice and resuspended in 0.3 M cold sodium acetate (pH 4.5) in between each

precipitation step. After decanting the supernatant and brief air-drying, the pellet

was resuspended in 10 mM cold sodium acetate (pH 4.5).. An aliquot of tRNA

was deacylated by adding an equal volume of a solution of 100 mM Tris-HCl and

100 mM NaCl (pH 9.5) and incubating at 70°C for 30 minutes.

2.2.4 Preparation of cell extracts

Cell extracts were prepared by anoxic lysis of harvested cells suspended in

50 mM MOPS buffer with a French pressure cell operated at 18,000 psi. Total soluble proteins were separated from cellular debris by centrifugation at 150,000 g for 30 min. Protein concentrations were determined using the bicinchoninic assay method (Smith et al, 1985) with bovine serum albumin as the standard.

For enzyme assays, cells from three pooled 50 mL cultures were harvested anoxically at late exponential phase (OD600 = 0.8). For immunoblotting, cells from

15 L cultures of both mutant and wild-type strains were harvested during various phases of growth.

2.2.5 Southern hybridization analysis

Genomic DNA from both wild-type and mutant strains of M. acetivorans were

restriction digested using HindIII and PstI and electrophoresed on 1% agarose

gels at 40 V for 16 h. After visualizing under ultraviolet (UV) light, aligning wells to

a fluorescent ruler and removing the marker lane, the gels were denatured. Gels were gently shaken in a solution of 1 M NaCl/0.5 M NaOH twice for 15 min each and rinsed briefly with double-distilled water. The gels were then gently shaken in a solution of 1.5 M NaCl/0.5 M Tris-HCl (pH 7.4) for 30 minutes in two changes of buffer, and blotted onto Hybond N+ membrane (Amersham Biosciences,

Piscataway, NJ) overnight by using a blotting stack. Following UV-crosslinking at

150 mJ, membranes were pre-hybridized for 30 minutes at 54°C in a solution of

0.25 M sodium phosphate and 5% SDS (pH 7.2). For probing of the blots, either

probe UPT (5′-GGAGCTCTTTGCGGACATGGAGG-3′) complementary to a

region 667 bases upstream of the start of the pylT (171 bases upstream of the

deleted region in the mutant) was used; or probe PT, comprising the 72

deoxyribonucleotides corresponding to the sense strand of pylT, was employed.

Ten pmol of either oligodeoxyribonucleotide were labelled with [γ-P32]ATP (ICN

Biomedicals) using polynucleotide kinase (Roche Diagnostics, Indianapolis, IN)

according to the manufacturer’s protocol. Prior to adding to the membrane, the

reactions were subjected to size exclusion chromatography in 1 mL volume spun columns packed with Sephadex G-25 gel permeation matrix to remove excess nucleotides. After 24 h incubation at 54°C, membranes were washed for 10 min each in 20 mM sodium phosphate, 5% SDS (pH 7.2); and 20 mM sodium phosphate, 0.5% SDS (pH 7.2). Membranes were visualized with a STORM phosphorImager (G.E. Healthcare, Piscataway, NJ).

2.2.6 Acid-urea gel electrophoresis and Northern hybridization analysis

Aminoacylated and deacylated tRNA species were detected following the method of Varshney et al (1991) with modifications (Jester et al, 2003).

Aminoacylated and deacylated samples were loaded on to 14% denaturing 7.0 M urea gels (pH 5.0) and were electrophoresed at constant voltage of 50 V at 4°C for 24 h in 0.3 M sodium acetate (pH 5.0) running buffer. The running buffer was

changed every 8 h to minimize deacylation of samples due to shifts in pH. Gels

were electroblotted on to Hybond N+ membranes (Amersham Biosciences) at 40

V for 2 h at 4°C in a blotting solution with a final concentration of 10 mM Tris- acetate, 5 mM sodium acetate and 0.5 mM EDTA (pH 8.0) using a Trans-Blot

Cell blotting apparatus (Bio-Rad Laboratories, Hercules, CA). The membranes were then UV cross-linked and stored at 4°C for subsequent hybridization.

Membranes were incubated at 54°C in a hybridization solution containing 0.25 M sodium phosphate and 5.0% SDS, pH 7.2. Ten pmol of a 72 base deoxyribonucleotide complementary to M. acetivorans tRNAPyl were labeled with

[γ-P32]ATP (ICN Radiochemicals, Costa Mesa, CA) catalyzed by polynucleotide

kinase (Roche Diagnostics, Indianapolis, IN) according to the manufacturer’s

protocol. Membranes were incubated at 54°C for 12–16 h in the same

hybridization solution to which labeled probe had been added. Following

hybridization, membranes were washed for 5 min each in 1% SDS in standard

saline-citrate (SSC) (Sambrook et al, 1989); followed by two 5 min washes in 1%

SDS in 0.5 × SSC, with all solutions at 54°C. Radioactive bands on the blots

were visualized by phosphor imaging.

2.2.7 Immunoblot analysis

Proteins in a cellular extract were separated by 12.5% SDS polyacrylamide-

gel electrophoresis (Laemmli, 1970). The gel was split in half following

electrophoresis and half was stained with Coomassie brilliant blue. The other half

was electroblotted on to Hybond-P (GE Healthcare, Piscataway, NJ) after which

immunoblotting was performed as described (James et al, 2001). Affinity-purified polyclonal rabbit antibodies raised against Methanosarcina barkeri MtmB (James

et al, 2001) were used as the primary antibody to probe the blot in anti-MtmB immunoblots, while affinity-purified ovine polyclonal antibodies raised against M.

barkeri MtbA and MtaA (Ferguson et al, 1996) were used in anti-MtbA and anti-

MtaA blots, respectively. Following further incubation with secondary antibodies

conjugated to horseradish peroxidase (GE Healthcare, Piscataway, NJ), reacting

protein bands were visualized with 4-chloro-1-napthol and hydrogen peroxide.

2.2.8 MMA:CoM methyltransferase activity assays

CoM-methyltransferase activity assays were performed as previously described

(Burke and Krzycki, 1997). Assays were carried out in 13.8 mL rubber-stoppered

serum vials that had been repeatedly evacuated and flushed with H2. The

reaction mixture contained 50 mM MOPS at pH 7.0, 4 mM titanium citrate, 10

mM ATP, 20 mM MgCl2, 3.2 mM 2-bromoethanesulfonic acid, 2 mM CoM, cell extract (1.5 mg protein), and 100 mM of either methanol or MMA. The mixture

was incubated at 37°C, samples of the reaction mixture were periodically

removed, and the remaining free thiol of CoM was detected by Ellman’s reagent.

2.3 Results

2.3.1 The ∆ppylT mutant

On examination of the genome sequence of M. acetivorans, it was found that the pyl genes are not annotated with the pyl functions. Upon further analysis of the genome, it was discovered that ORF MA0155 (Genbank accession number

AAM03608) yields a gene product with 86 sequence similarity to the annotated and experimentally validated M. barkeri Fusaro PylS. MA0155 is probably mis- annotated and extends another 117 bases 5’ of the annotated ATG start site based on homology to the M. barkeri pylS start site. This ORF was also introduced in E. coli and tested in vivo (Blight et al, 2004; Chapter 5). As mentioned in Chapter 1, the M. acetivorans pylT gene encoding tRNAPyl is

identical to the M. barkeri Fusaro pylT gene (Srinivasan et. al., 2002). Similar to the scenario in M. barkeri Fusaro, the M. acetivorans pylTS genes are followed by pylB, pylC, and pylD genes. However, the orientation of the pylTSBCD gene cluster is flipped when compared to the same genes with respect to the mtmCBP and mtbA genes in the two Methanosarcina species.

A 761 base-pair segment of chromosomal DNA encompassing the pylT gene of M. acetivorans was deleted by replaced with the Streptomyces alboniger puromycin acetyltransferase cassette (pac) by A. Patel (Mahapatra et al, 2006).

Briefly, she introduced the pac on plasmid pAP1 which has a 2 kb region of the

M. acetivorans genome with pylT as the central locus. However, the pylT gene and 193 and 496 base pair downstream and upstream of pylT were deleted and

replaced with pac. This plasmid was linearized and transformed into M.

acetivorans in the laboratory of W. Metcalf. Thus, the pylT gene and flanking

regions including putative promoters of the pyl genes were deleted and replaced

by pac.

Puromycin-resistant mutants were obtained on high-salt methanol acetate

medium. We used Southern hybridization to screen the mutants obtained. A

strain with the pylT gene and putative promoters replaced was designated

∆ppylT (Figure 2.2). DNA from puromycin-resistant mutants and wild-type was extracted, digested, electrophoresed, blotted, and hybridized to probe UPT,

which is complementary to sequence 667 bases upstream of the pylT gene

(corresponding to 167 bases outside the pyl cassette replacement region). The

0.75 kb SmaI/BamHI genomic restriction fragment found in wild-type that is

complementary to probe UPT is replaced in the desired mutant with the 1.3 kb

pac, which bears no HindIII site. The ∆ppylT strain gives rise to a HindIII restriction fragment that is 0.5 kb larger than the wild-type strain. Introduction of the pac gives rise to a new PstI site in the chromosome immediately adjacent to the new BamHI site. This causes the hybridizing genomic PstI fragment to decrease in size from approximately 2 kb in wild-type to 1 kb in ∆ppylT.

Restriction digest band pattern observed from the mutant was consistent with a double crossover that resulted in the pac being transcribed opposite to the orientation of the pylT and pylS genes. No restriction digest band patterns corresponding to a single crossover event were observed from the strain that had

resulted from the transformation of M. acetivorans with pAP1.

The same PstI and HindIII restriction fragments observed hybridizing to probe

UPT were observed with probe PT, which is complementary to the entire pylT

gene in M. acetivorans. While these fragments were detectable in wild-type they

were not observed even under hybridization conditions of low stringency. Taken

together, these results show that the ∆ppylT mutant bears the direct replacement of the 0.7 kb region of the genome with the pylT gene.

2.3.2 tRNAPyl is not detectable in ∆ppylT

Because the ∆ppylT mutant lacks the pylT gene, it should not be able to synthesize tRNAPyl. To validate this assertion, unfractionated tRNA was isolated

from exponential-phase wild-type and ∆ppylT cells growing on methanol,

subjected to acid-urea polyacrylamide gel electrophoresis, blotted, and

hybridized with a probe complementary to the entire length of tRNAPyl. We

observed that wild-type cells growing on methanol also express pylT (Figure 2.3).

Both aminoacylated and unacylated/deacylated tRNAPyl species were detected in

the isolated M. acetivorans unfractionated tRNA (Figure 2.3, Panel A). We were

unable to detect the corresponding aminoacylated or unacylated/deacylated

tRNAPyl species from unfractionated ∆ppylT confirming that ∆ppylT lacks this tRNA. As a control for the purification procedure and the probing reaction, the same membrane was stripped of radioactivity and re-probed with an oligo

Lys complementary to tRNAUUU isoform of tRNA (Figure 2.3, Panel B).

2.3.3 The ∆ppylT mutant is lethal on specific growth substrates

As mentioned in Chapter 1 and Section 2.1 of this chapter, species of

Methanosarcina including M. acetivorans have the capacity to utilize a number of methylated compounds as substrates for methanogenesis and carbon assimilation. We hypothesized that the loss of tRNAPyl would lead to a defect in

growth on the substrates that required the translation of the amber codon to

proceed to produce intact, functional proteins key to the utilization of these

specific substrates. As a test of this hypothesis, we screened the ability of the

∆ppylT mutant to grow on some of its known substrates.

The mutant was selected on growth medium supplemented with methanol and acetate. Thereafter, the ability to grow on these two substrates was tested. Wild-

type and mutant cells showed generations times of 23 and 25 hours respectively,

when grown on 100 mM methanol (Figure 2.4, Panel A). Wild-type and mutant

cells pre-grown on methanol and inoculated in to medium with 40 mM acetate

had generation times of 35 and 38 hours respectively. It should be noted that both wild-type and mutant strains exhibited extended lag periods prior to growth.

The differences in rates of growth in wild type and mutant strains were not

deemed to be statistically significant. Taken together, these data indicate that the decoding of UAG requiring tRNAPyl is non-essential for M. acetivorans growing on

methanol or acetate. Further, since the mutant strain remained viable when

transferred on to either substrate over two years, no deleterious effects of the

∆ppylT mutation during growth on these substrates is expected.

Wild-type and mutant cells, pre-grown on methanol were inoculated into growth medium containing 40 mM of TMA, DMA, or MMA. Upon transfer to the methanogenic medium supplemented with the methylamines, only wild-type cells exhibited logarithmic cellular growth. Wild type cells had generation times of 25,

32, or 45 hours on TMA, DMA, or MMA respectively (Figure 2.4, Panel B). The

∆ppylT mutant strain did not grow on any methylamine even on incubating for extended periods of time up to six months.

2.3.4 Growth using MMA as a sole nitrogen source

M. acetivorans can use either dinitrogen or ammonia as a source of nitrogen

(Galagan et al, 2002). We show that wild-type M. acetivorans can also use MMA as a sole nitrogen source while growing on methanol (Figure 2.5). However, the

∆ppylT mutant cannot utilize MMA as the sole nitrogen source for growth. Wild- type cells in medium with a combination of 40 mM MMA and 100 mM methanol under argon phase grew normally with a generation time of 26 hours. This generation time is very similar to that of wild-type cells growing on methanol with ammonia under dinitrogen. Wild-type cells also grew under medium in argon gas phase lacking methanol, ammonia, or dinitrogen in the presence of 40 mM MMA.

However, growth occurred only after a lag period of 10 days and also with a slower generation time of 50 hours. In contrast, ∆ppylT did not grow on MMA as a sole energy and nitrogen source. Also, the mutant did not utilize MMA as a

nitrogen source in methanol-added medium lacking ammonia and dinitrogen. In methanol-supplemented medium, in the absence of any added nitrogen source,

wild-type cells showed an initial increase in OD600 before early cessation of

growth. A definitive explanation for this observation is lacking. However, it has

shown that MMA can be obtained from stationary-phase cells of M. barkeri grown

only on methanol (Srinivasan, 2005). It is possible that such an event is occurring

in both M. acetivorans wild-type and mutant strains with the exception that the

wild-type strain can use this MMA for a short burst of nitrogen utilization and growth while mutant strain can not.

2.3.5 MMA:CoM methyltransfer activity is not detectable in ∆ppylT

To determine if there is a direct link between the defect in the ability to translate UAG as pyrrolysine (due to the lack of tRNAPyl) and the ability to

metabolize methylamines, we compared the MtmB activity of extracts from mutant with extracts from wild-type in MMA:CoM methyl transfer. MtmB was first isolated in biochemical reconstitution assays of the MMA:CoM methyl transfer activity using M. barkeri MS (Burke and Krzycki, 1997). A difficulty of testing

∆ppylT for this activity is that cells in exponential-phase growing on methanol only produce minimal levels of MMA:CoM methyltransferase activity. We were able to circumvent this difficulty by growing wild-type cells in medium containing both methanol and MMA. Cells grown under such conditions exhibit relatively high specific activities of MMA:CoM methyl transfer activity (Table 2.1).

Methanol:CoM methyltransferase activity was detectable from extracts of both

wild-type and mutant strains. However, consistent with the hypothesis that there

is a link between UAG translation and methylamine-dependent methyltransfer,

we were unable to detect MMA:CoM methyltransferase activity from extracts of

∆ppylT strain.

2.3.6 The ∆ppylT strain does not produce detectable MtmB protein.

It is possible that the loss of MMA methyltransferase activity demonstrated in the preceding section could be the result of the MtmB protein being inactivated instead of interference with the synthesis of MtmB. Inactivation could result from a scenario in which an alternative method of translating the UAG existed in M.

acetivorans. In addition, the 23 kDa product might be observed due to the

truncation of the protein due to translational stopping at the UAG codon. To

examine these scenarios, we performed immunoblotting of wild-type and mutant

extracts grown on 100 mM methanol and 40 mM MMA for MtmB, using a polyclonal antibody previously shown to allow the detection of both the 50 kDa full-length and the 23 kDa truncated MtmB products from M. barkeri (James et al,

2001). Since it is possible that MtmB might be synthesized in different phases of growth in either mutant or wild-type, or that aberrant MtmB might have a faster turnover, extracts were examined from different phases of growth in both wild- type and mutant strains. An intense 50 kDa band that comigrated with purified

MtmB standard was detectable in lanes with extracts from cells harvested during

exponential phase wild-type M. acetivorans (Figure 6). The intensity of this band

was found to diminish in later phases in wild-type. In contrast, regardless of the

phase of growth in question, no MtmB (either full-length or a truncated product)

was detectable in lanes in which extracts of ∆ppylT had been loaded.

2.3.7 The ∆ppylT strain does not produce detectable MtbA protein.

Analysis of total soluble proteins from wild-type and ∆ppylT in Coomassie-

stained SDS-polyacrylamide gels revealed differences in band intensities

corresponding to different molecular weights in the two strains on 100 mM

methanol and 40 mM MMA (Figure 2.7). In methanogenesis from methylamines

in M. barkeri, a methylamine methyltransferase specific for TMA, DMA, or MMA

(MttB, MtbB, and MtmB respectively) transfers the methyl group to a cognate

TMA, DMA, or MMA specific corrinoid protein (MttC, MtbC, and MtmC

respectively) (Burke et al, 1998). In the subsequent step common for all

methylamines, the methyl group is transferred from the specific corrinoid protein

to CoM by a 36 kDa zinc-containing CoM methylase, MtbA; methyl transfer from methanol proceeds via an analogous methyltransferase system consisting of three enzymes MtaB, MtaC, and the methanol specific CoM methylase, MtaA

(Harms and Thauer, 1996). In M. barkeri, methanol-utilizing cells contain primarily MtaA and that expression of the gene which encodes for MtbA is repressed during growth on methanol (Harms and Thauer, 1996). Studies have revealed that the mtbA is transcribed into a monocistronic 1.1 kb transcript that is

preferentially transcribed during growth on TMA or H2/CO2 (Harms and Thauer,

1996; Burke et al, 1998). Here, we examined soluble proteins from both strains

for the presence of MtbA, the methylamine-specific CoM methylase. We were

unable to detect MtbA in cells of ΔppylT even though mtbA does not possess an

in-frame amber codon (Fig 2.8). An explanation for this observation is currently

not known. However, it is tempting to speculate that lack of detectable MtbA in

∆ppylT is due to transcriptional or translational regulation.

2.4. Discussion

The presence of in-frame UAG codons in genes encoding methylamine methyltransferases was the first indication that translation of UAG in these genes might be required for methylamine metabolism (Burke et al, 1998; Paul et al,

2000). Similar in-frame UAG codons were not found in other sequenced genes required for methanogenesis (Burke et al, 1998; Paul et al, 2000). On examination of the genomes of species of Methanosarcina, it was noted that the methylamine methyltransferase family were the only family of gene products with

UAG translation, with the possible exception of a family of putative in-frame UAG codon-containing transposases (Deppenmeier et al, 2002; Galagan et al, 2002).

A gram-positive organism, Desulfitobacterium halfniense, was also found to have the pyl genes along with a putative in-frame UAG codon-containing TMA methyltransferase homolog (Srinivasan et al, 2002). An extensive bioinformatics approach has indicated that only the methyltransferase and transposase gene

families of Methanosarcina, Methanococcoides, and Desulfitobacterium possess in-frame amber codons (Zhang et al, 2005). In such a situation, the expected outcome of loss of the pylT gene would be the loss of the ability to make full- length methylamine methyltransferases. Such a loss would be predicted to cause a cessation of methylamine metabolism for catabolism. Because of the limited distribution of the in-frame UAG codon, such a loss would also be predicted to not adversely affect metabolism of other metabolites. In fact, the ∆ppylT strain showed a pleiotropic phenotype in that it was unable to use TMA, DMA, or MMA as a growth substrate. The observed growth defect was concomitant with loss of the ability to produce MMA methyltransferase. Full-length 50 kDa MMA methyltransferase was produced by wild-type cells; however, the ∆ppylT strain did not produce it in any phase of growth. Consistent with our previous biochemical studies that indicated MMA:CoM methyltransferase activity only mediated by MtmB present with the cognate corrinoid protein and methyltransferase (Burke and Krzyki, 1997), we show here that loss of MtmB in a cellular milieu correlates with loss of detectable MMA:CoM methyltransferase activity.

A truncated MtmB product would be produced if UAG-directed termination occurred instead of translation. We note with interest that ∆ppylT did not produce a truncated MtmB product. Plausible reasons for the absence of detectable truncated MtmB product include instability of a produced truncated product, decrease or cessation of transcription of mtmB in the mutant, or inefficient

termination of UAG in the context of the mtmB transcript. Since we have been unsuccessful in analyzing mRNA from wild-type M. acetivorans, we have been unable to examine levels of transcription.

Deletion of 761 bases in the mutant might possibly have affected loci other than pylT. ORF MA0156 is partially truncated by the deletion. Although the function of this ORF is not annotated in the genome of M. acetivorans, it has been annotated in the M. barkeri genome as a squalene cyclase (Maeder et al,

2006). Although it is a formal possibility that truncation of this ORF might result in the pleiotropic loss of methylamine metabolism while not affecting tolerance to methylamines, it is unlikely given the assignment. More likely to effect methylamine metabolism are the other genes of the pyl gene cluster. The putative TTTATA transcriptional promoters immediately upstream of pylT and pylS are deleted in the mutant, possibly eradicating expression of the pyl genes.

In M. barkeri, pylTSBCD forms a transcriptional unit (Srinivasan et al, 2002); however, it is unclear if such a transcriptional unit is formed in M. acetivorans.

The pylS gene product has been demonstrated to act as a class II AARS enzyme, while the pylB, pylC, and pylD gene products have been suggested to have roles in biosynthesis of pyrrolysine based on functional homology

(Srinivasan et al, 2002). Therefore, the failure to express any of the pyl genes might have the same effect on UAG translation as pyrrolysine as only deleting pylT. Others in the lab are now examining the pyl region using recently developed techniques for markerless deletion (Pritchett and Metcalf, 2005). This

should enable precise deletions of individual ORFs to be constructed without

polarity effects. Such an approach will also be useful since it will allow

complementation with puromycin resistance as a marker, which unfortunately, is the only documented commercially-available resistance marker in species of

Methanosarcina.

The results obtained with ∆ppylT indicate that detailed genetic investigations

of UAG translation as pyrrolysine are feasible. The deletion of pylT appears to be

only conditionally lethal, since ∆ppylT grows normally on methanol and acetate,

but does not grow on any methylamine. In-frame UAG codons are not observed

in genes that encode the key steps of methanogenesis from methanol or acetate

(Maupin-Furlow and Ferry, 1996; Sauer et al, 1997; Galagan et al, 2002). This

indicates that the central enzymes of methanogenesis and proteins required for

energy generation are not dependent on pyrrolysine. The necessity of pyrrolysine

in the H2/CO2 dependent methanogenic pathway could not be determined in this

species of Methanosarcina, since it is unable to grow using H2/CO2 as methane

precursors (Sowers et al, 1984). An additional observation is that essential

anabolic reactions involving ammonia are not compromised in ∆ppylT. We note,

however, that ∆ppylT is unable to use MMA as a sole nitrogen source in the

presence of methanol. This observation indicates that assimilation of ammonia

occurs through the liberation of ammonia from MMA and ceases in cells unable to make MMA methyltransferase. Taken together, our conclusions suggest that

M. acetivorans does not possess any alternative mechanisms for methylamine

utilization than MMA:CoM methyltransfer.

The phenotype exhibited by the mutant described in this chapter is suggestive

of that of a deletion mutant of Methanococcus maripaludis in which the selB gene

encoding the selenocysteine-specific elongation factor, SelB, was lacking (Rother

et al, 2003). The selB mutant was viable when forming methane from H2/CO2, but

not in the presence of formate as loss of SelB was associated with loss of formate dehydrogenase activity (Rother et al, 2003). It should be noted that both formate dehydrogenases of M. maripaludis are predicted to contain selenocysteine (Wood et al, 2003). In the current study, we demonstrate that loss of pylT-dependent UAG translation of pyrrolysine in M. acetivorans results in a mutant unable to grown on methylamines due to loss of the ability to form MMA methyltransferase. Taken together, these phenotypes indicate that the 21st and

22nd amino acids play vital roles in expanding the limited substrate range of

methanogenic members of Archaea.

Figure 2.1 The pyl and mtmB gene clusters of M. acetivorans. The ORFs found in the M. acetivorans genome extending from nucleotide 169234 to 188921 (Genbank accession number AE010299). Below each ORF are shown the locus id numbers given by the Broad Institute (the MA prefix of each number is not shown). The genes whose products have been demonstrated to be involved in CoM methylation with MMA are shown in yellow. The mtmP gene may be involved in MMA transport, while the ramA product has been shown to be involved in activation of the methyl transfer reaction (Ferguson et al, in preparation). In black are shown ORFs with no demonstrated role in MMA:CoM methylation, and whose annotation indicate possible involvement in electron transport (MA0147) or rRNA modification (MA0148). Others shown in black are annotated as a psuedogene (MA4684), or ORF of unknown function (MA0149, MA0156, MA0157). MA0156 Is likely to be monocistronic, as MA0157 is divergently transcribed. The genes of the pyl gene cluster are shown in blue and red. The location of the deleted segment of the wild-type genome that was replaced by the pac cassette in the ΔpylT genome is shown by the thin black line.

Figure 2.2 Southern analysis of restriction digested wild-type and ΔppylT genomic DNA. DNA from either wild-type (lanes 1 and 3) or ΔppylT (lanes 2 and 4) were restriction digested with PstI (lanes 1 and 2) or HindIII (lanes 3 and 4) electrophoresed, blotted and probed with 32P-labelled deoxyoligonucleotide that was complementary to either (Panel A) the region upstream of pylT (probe UPT); or (Panel B) the pylT gene (probe PT).

Figure 2.3 Northern analysis of tRNAPyl from wild-type and ΔppylT. (Panel A) The cellular tRNA pool was isolated from exponentially growing wild-type (lanes 1 and 2) and ΔppylT (lanes 3 and 4), electrophoresed on acid-urea gels, and transferred to a membrane prior to probing with a specific oligodeoxyribonucleotide probe complementary to the tRNAPyl. (Panel B) The same blot was then stripped of all radioactivity, then probed with an oligo complementary to tRNALys. Lanes 2 and 4 were as isolated; for lanes 1 and 3 the tRNA pool was subjected to mild alkaline hydrolysis of the aminoacyl-tRNA ester linkage. The bands corresponding to uncharged and aminoacylated tRNAs are indicated.

Figure 2.4 Growth curves of wild-type (closed symbols) and ΔppylT (open symbols) on methanol and methylamines. Panel A is a comparison of the growth of wild-type (●) and ΔppylT (○) on 100 mM methanol, while Panel B shows growth of wild-type and ΔppylT on 40 mM methylamines: wild-type on TMA (●), DMA (▲), and MMA (■); ΔppylT on TMA (○), DMA (Δ) and MMA (□). A representative growth curve from each condition tested is shown. Duplicate tubes were used for each experiment, and the experiment repeated four times with the same results.

Figure 2.5 Growth of wild-type and ΔppylT strains on medium lacking ammonia. Both strains were transferred to modified DSM304 media made without detectable ammonia and prepared under argon gas instead of the standard N2 gas (see Experimental Procedures). Growth substrates were 100 mM methanol (triangles), 40 mM MMA (squares), or a combination of both (circles), and the tubes were inoculated with either wild-type (closed symbols), or ΔppylT (open symbols). Representative growth curves from one of three replicate cultures are shown. As a positive control, the same ΔppylT culture was simultaneously inoculated into ammonia- and methanol-containing medium and normal levels of growth were observed.

Figure 2.6 Detection of MtmB by immunoblotting. (Panel A) Growth curves of wild-type (●) and ΔppylT (○) in 15 L carboys. Both cultures were grown in the presence of methanol and MMA. The variability of lag times is not a consistent property of the wild-type or mutant, and variable lag times for M. acetivorans are often seen after inoculation. (Panel B) Time-points of the cultures above were harvested and subjected to electrophoresis in 12.5 % polyacrylamide SDS gels, then immunoblotted with rabbit polyclonal antibodies raised against M. barkeri MS MtmB. The smaller 42 kDa band in the standard lane has been observed previously and is a degradation product of the 50-kDa MtmB protein (James, et al, 2001). Lane 1 was loaded with buffer only. Lanes 2-5 were loaded with 500 µg protein from wild-type extracts harvested at day 6 (lane 2), day 7 (lane 3), or day 8 (lane 4), or day 9 (lane 5) of growth; lanes 6-8, 500 µg protein from extracts of wild-type M. acetivorans harvested at day 4 (lane 6), day 5 (lane 7), or day 6 (lane 8). Lane 9 was loaded with 1 µg of purified M. barkeri MtmB standard.

Figure 2.7 SDS-Polyacrylamide gel electrophoresis of fractionated soluble proteins from wild-type and ΔppylT cells grown on methanol and MMA Proteins were harvested from late-exponential growth-phase cultures of wild-type and ΔppylT grown on 100 mM methanol and 40 mM MMA. Samples with 150 µg of soluble fraction of proteins from wild-type (lane 1) and ΔppylT (lane 2) were loaded on to the gel. In the rightmost lane, commercially available molecular weight size standards were run with the matching sizes indicated.

Figure 2.8 The amine specific methylase, MtbA is not detectable from extracts of ΔppylT grown on methanol-MMA medium. Panel A shows a representative immunoblot performed to detect the methanol-specific methylase using anti-MtaA antibodies, while Panel B was performed to detect the presence of the methylamine-specific methylase using anti-MtbA-antibodies (Panel B). In both panels, lane 1 was loaded with 2 µg of isolated M. barkeri MtaA; lane 2 was loaded with 300 µg of fractionated soluble proteins from late-exponential growth- phase M. acetivorans wild-type cells; lane 3 was loaded with 300 µg of fractionated soluble proteins from late-exponential growth-phase ΔppylT cells; and lane 4 was loaded with 2 µg of isolated M. barkeri MtbA.

Strain Substrate Specific Activity

(nmol min-1 mg-1) wild-type methanol 42

ΔppylT methanol 96 wild-type MMA 101

ΔppylT MMA N.D.*

Table 2.1 Methanol:CoM methyl transfer and MMA:CoM methyl transfer activity. Results are average values of two experiments performed in duplicate.

(*N.D. -not detectable).

CHAPTER 3

PYRROLYSINE LIGATION TO COGNATE tRNA:

ACTIVATION OF PYRROLYSINE BY A NOVEL AMINOACYL-tRNA

SYNTHETASE

3.1 Introduction

Aminoacylation involves a two-step mechanism (First, 2005). In the first step, the amino acid is activated by ATP to form an enzyme-bound aminoacyl adenylate intermediate. The second step involves the ligation of the amino acid to the cognate tRNA. Both reactions occur within a single active site.

As mentioned in Chapter 1 the pylT gene encodes an unusual amber- decoding tRNAPyl, whereas the pylS gene encodes an aminoacyl-tRNA

synthetase, PylS, that was previously observed to charge a tRNAPyl transcript

with 3H-lysine using acid precipitation and filter capture as the assay (Srinivasan et al, 2002). Although the possibility that PylS aminoacylated tRNAPyl with

pyrrolysine directly was also mentioned in that study, it was not examined to lack

of available synthetic pyrrolysine (Srinivasan et al, 2002).

In this study, we examine the possibility of directly ligating chemically synthesized pyrrolysine (Hao et al, 2004) to tRNAPyl in vitro. Here, the specificity

32 and kinetics of the amino acid-dependent isotopic exchange of PPi into ATP are examined. Together with the studies on overall aminoacylation performed collaboratively by S. Blight and R. Larue (Blight et al, 2004), our data suggest that the direct ligation of pyrrolysine to tRNAPyl is feasible.

3.2 Experimental Procedures

3.2.1 Strains and plasmids

A strain of E. coli BL21 (DE3) bearing the ppylSH6 plasmid was constructed by S. Blight (Blight et al, 2004). The ppylSH6 plasmid is a pET 22b-derived plasmid (Novagen, Madison, Wisconsin) with the M. barkeri MS pylS gene cloned to facilitate purification of PylS with a hexahistidine tag at the C-terminus

(Blight et al, 2004) (Blight et al, 2004).

The pylB, pylC, and pylD genes were amplified by PCR from M. acetivorans

C2A genomic DNA using Ex-taq (Takara Mirus Bio, Madison, WI) and were subsequently ligated into pCR2.1-Topo (Invitrogen, Carlsbad, CA) by D.

Longstaff and J. Faust (Longstaff et al, 2007b). The pk13, pk14, pk15, and pk16 vectors bear combinations of the pylBCD genes cloned the pACYC-Duet

(Novagen, Madison, Wisconsin) backbone (Longstaff et al, 2007b). Briefly, pk13 bears pylB, pylC and pylD; pk14 bears pylB and pylD; pk15 bears pylC and pylD; and pk16 bears pylB and pylC.

3.2.2 PylS purification

PylS was purified from cell extracts of the E. coli BL21 (DE3) strain with

ppylSH6 as described (Blight et al, 2004). Cells were lysed in 20mM sodium

phosphate, 500mM NaCl, 10mM imidazole pH 7.4 and PylS was extracted by Ni-

activated trap chelating HP column (GE Healthcare, Piscataway, New Jersey).

PylS-His6 eluted at 240mM imidazole during the application of 10–500mM

imidazole in the same buffer to the column. Protein concentration was

determined as described in Chapter 1. Purification to homogeneity (98%) was established by running 5 µg of fractions on an SDS-PAGE gel with the detection of a single band corresponding to 50 kDa detectable upon Coomassie staining.

3.2.3 PylS substrates

L-Pyrrolysine was synthesized and characterized with the use of 13C and 1H

NMR by Hao et al, 2004. The pyrrolysine used in experiments was further analyzed by electrospray mass spectrometry and revealed two predominant peaks with m/z 256.16 (M þ H) and 278.14 (M þ Na), where M is L-pyrrolysine

(Soares et al, 2006). The cellular tRNA pool was isolated from M. acetivorans

C2A (OD600 0.6–0.7) growing on trimethylamine at 37°C in modified DSM media as described in Chapter 2. Lysine and the other nineteen common amino acids were purchased from Sigma (St Louis, MO).

Cells carrying pK13–16 were grown on LB with chloramphenicol, induced at

an OD600 of 0.3 with 1 mM IPTG and incubated for 4 h at 37°C. Harvested cell

pellets were then extracted with 2 mL of methanol per gram of cell wet weight for

10 min at room temperature. The supernatant was passed through a YM-3 filter

(Millipore, Bedford, MA) to completely remove any remaining high-molecular-

weight material (R. Larue; Longstaff et al, 2007b).

32 3.2.4 PPi-ATP exchange with PylS

32 The PPi-ATP exchange reaction was performed under standard conditions

as described previously (Cole and Schimmel, 1970). All reactions were

performed in duplicate at 37°C with at least two batches of purified recombinant

PylS. Standard reactions (100 µL unless otherwise noted) were set up with 20

mM HEPES-KOH (pH 7.2), 10 mM MgCl2, 25 mM KCl, 1 mM KF, 4 mM DTT, 2 mM ATP except for kinetic analyses, 2 mM [32P]PPi (4-22 dpms/pmol as

mentioned, Perkin Elmer, Waltham, MA). Concentrations of synthetic L-

pyrrolysine were 100 µM in the standard assay, except when determining kinetic

parameters or unless otherwise mentioned. The concentration of protein was 0.3

to 1 µM.

At specific times, 25 µL aliquots were removed and quenched in 500 µL of a stop-reaction suspension containing 1.6% activated charcoal w/v (Sigma, St.

Louis, MO), 80 mM sodium tetrapyrophosphate, plus 3.5% perchloric acid v/v.

Samples were filtered through Whatman GF/C glass-fiber filters and washed

three times with 5 mL of 40 mM sodium tetrapyrophosphate solution containing

1.7% perchloric acid v/v and once with 5 mL of 100% ethanol. Filters were dried

in 80°C for 5 minutes. The filters were placed in 20 mL scintillation vials with 10

mL of Ultima Gold F scintillation cocktail (Perkin Elmer, Waltham, MA), shaken

once and counted in a Packard TriCarb 2100-TR liquid scintillation analyzer.

Identical reactions without amino acid, ATP, or enzyme were also set up as

negative controls in duplicate. Reactions in the absence of amino acid were used

in determining specific activities, which were calculated for all assays performed.

The dependence of the reaction rate on both pyrrolysine and ATP was verified

with two independent experiments performed with two different batches of

purified enzyme. One substrate was kept at a constant concentration much

higher than its respective apparent Km value. [Pyrrolysine] was ~2 Km for variation

of ATP, and [ATP] was >10 Km for variation of pyrrolysine. In each case, the

concentration of the second substrate was varied over the range Km/5 to 5-8 Km

(Fersht, 1985). Km and Vmax values were determined using Haanes-Wolf plots.

To assay methanolic extract from cells bearing the pk13-16 plasmids, a

method was developed by R. Larue in which aliquots were dried during

centrifugation under vacuum, and the extracted metabolites were resuspended in

the pyrophosphate exchange reaction assay buffer (Longstaff et al, 2007b).

Assays were carried out in the presence of the metabolites extracted from the

equivalent of 10 mg of dry weight of E. coli. Assays were performed with 5.5 µM

PylS in 100 µL with 20 µL aliquots taken at each time point.

3.3 Results

32 3.3.1 The PPi-ATP exchange reaction catalyzed by PylS is pyrrolysine

dependent.

The rate at which amino acid activation occurs can be assayed by the reverse

32 32 isotopic exchange of PPi-ATP over time. We noted that the PPi-ATP

exchange reaction catalyzed by PylS is pyrrolysine dependent (Figure 3.1). No

isotopic exchange above background was observed for reactions lacking PylS,

32 ATP, or synthetic pyrrolysine. We also noted that the PPi-ATP exchange

reaction is pyrrolysine specific, as a combination of the standard 20 amino acids

at 100 µM or lysine at 1 mM could not substitute for pyrrolysine (Table 3.1). Like

all other class II aminoacyl-tRNA synthetase enzymes, PylS dependent activation

does not require the presence of tRNA (Table 3.2).

We then determined the apparent kinetic parameters of activation by PylS by

assaying with varying concentrations of substrate and constructing Hanes-Wolf

plots (Figure 3.2). We obtained an apparent Km for ATP of 2 µM and an apparent

-1 -1 Km of 53 µM for pyrrolysine giving an apparent Vmax of 120 nmol min mg

-1 (apparent kcat of 6 min ) (Table 3.3)

32 3.3.2 The PPi-ATP exchange assay confirms that pyrrolysine synthesis is

dependent on pylBCD expression.

Having established that none of the 20 standard amino acids can substitute

32 for pyrrolysine in PylS-mediated PPi-ATP exchange reactions, we assayed for

pyrrolysine in E. coli bearing pylBCD. We extracted the methanol-soluble low-

molecular-weight fraction from E. coli bearing pACYCDuet-1 and found it did not

Pyl 32 possess a substrate for tRNA aminoacylation or PPi-ATP exchange by PylS.

32 In contrast, a PPi-ATP exchange reaction, depended upon addition of an

extracted metabolite pool from E. coli containing pylBCD on plasmid pK13, was

detected (Figure 3.3). Further, Northern blots of acid-urea gels revealed

aminoacylation of tRNAPyl in the presence of that same metabolite extract

(Longstaff et al, 2007b; Ross Larue). Using our determined kinetic parameters of

32 PylS in the pyrrolysine-dependent PPi-ATP exchange assay, we estimated from

the rate of the exchange reaction that E. coli bearing pK13 produced 0.1–0.5

µmol pyrrolysine per gram of dry weight. Production of the PylS substrate was found only with E. coli bearing pylBCD. E. coli transformed with pK14, pK15, or pK16, each of which lacks only one of the pylBCD genes, did not produce a PylS

substrate. Significantly, the PylS substrate was produced in E. coli lacking pylT and pylS, consistent with the synthesis of pyrrolysine in the absence of tRNAPyl or the pyrrolysyl-tRNA synthetase.

3.4. Discussion

We show here that PylS carries out the first stage of tRNA aminoacylation with pyrrolysine. Along with the aminoacylation assays done in our laboratory

and the ability of pylS and pylT gene products to suppress amber codons in a recombinant system (Blight et al, 2004), our current results strongly suggest that

PylS is a pyrrolysyl-tRNA synthetase capable of directly ligating pyrrolysine to tRNAPyl in vitro. Another study demonstrated similar results in vitro (Polycarpo et al, 2004). Using a racemic mixture of an enamine analog of pyrrolysine, a similar apparent kcat value for PylS dependent activation was obtained; however a much higher apparent Km of 1100 µM was reported, perhaps due to the impurities in the preparation, competitive inhibition with a compound in the mixture, and/or inefficient substrate binding using the analog preparation (Polycarpo et al, 2004).

PylS acts as expected for an aminoacyl-tRNA synthetase with specificity for

32 pyrrolysine in the PPi-ATP exchange assay. The values for the apparent Km of

ATP and amino acid for the exchange reaction obtained in our study are comparable and in some cases lower than rates determined in other organisms for the canonical aminoacyl-tRNA synthetases. However, to the best of our knowledge the value of apparent kcat for exchange kinetics is lower than values reported for canonical aminoacyl-tRNA synthetases from archaea.

PylS shows faint sequence similarity to a number of class II aminoacyl-tRNA synthetases, and has the recognizable ATP binding motifs for a class II enzyme.

Recently, the X-ray crystallographic structure of a truncated version of PylS revealed that extensive hydrophobic and electrostatic interactions stabilize the pyrrolysyl-adenylate in the active site (Kavran et al, 2007). Structure-based sequence analysis suggests that PylS is most similar to the canonical α-subunit

of phenylalanyl tRNA-synthetase (Kavran et al, 2007). The exchange assay of

class II enzymes is independent of the presence of tRNA. Therefore, the

behavior of PylS in the pyrrolysine activation assay is that expected of a class II

aminoacyl-tRNA synthetase.

The pylBCD gene products were previously hypothesized to catalyze the

synthesis of pyrrolysine (Srinivasan et al, 2002; Krzycki, 2005). Here, we

demonstrate that a metabolite from E. coli with the heterologously expressed

pylBCD genes does indeed serve as a substrate for PylS dependent activation. It

was demonstrated that the same metabolite serves as a substrate for overall

aminoacylation and is incorporated into MtmB in vivo in an E. coli reporter

system (Longstaff et al, 2007b). These results are striking for a number of

reasons. First, they indicate that the genes of the pyl operon confer both the

ability to synthesize and encode pyrrolysine in organisms as diverse as M.

barkeri and E. coli. Second, they imply that metabolic precursors used by the

pylBCD gene products are found in diverse organisms in both the archaeal and

bacterial domains of life. Lastly, our data suggests that the genetic code of a

number of diverse organisms such as D. halfniense might have been expanded

to include pyrrolysine by the horizontal gene transfer of the pyl genes and that

this process is unlikely to be an isolated occurrence.

PylS is highly specific for pyrrolysine in assays for aminoacylation (Blight et al,

32 2004; Polycarpo et al, 2004). This same specificity is reflected in the PPi-ATP exchange assay for the activation of the amino acid, the first stage of

aminoacylation. Our results support the contention that PylS is the first aminoacyl-tRNA synthetase to be discovered with specificity for an amino acid that is not one of the common set of twenty. Our results further indicate that

pyrrolysine is biosynthesized as a free amino acid in the cytoplasm of

32 Methanosarcina spp. Indeed, the ability to perform the PPi-ATP exchange

assay suggests that the method of decoding of UAG as pyrrolysine in MMA

methyltransferase proceeds similar to the direct decoding for the 20 canonical

amino acids and that this mechanism is unlike that utilized for selenocysteine

decoding.

32 Figure 3.1 The PPi-ATP exchange reaction mediated by PylS is pyrrolysine dependent. Aliquots were removed at the specified time-points from a 200 µl reaction in order to estimate the amount of 32P-ATP formed. In the figure are illustrated averaged duplicate reactions that were complete (●); or lacking PylS (◊), ATP (▲ ), or synthetic pyrrolysine (X).

Figure 3.2 Haanes-Wolf plots to determine the apparent kinetic parameters of activation by PylS. Separate experiments were performed to determine the apparent Km and Vmax for ATP (Panel A) and pyrrolysine (Panel B). Concentrations indicated are in µM.

32 Figure 3.3 Cellular amino acid pools tested in the PylS-mediated PPi-ATP 32 exchange assay. Exchange of P-PPi into ATP was monitored in 100-µL reactions containing 5.5 µM PylS, 10 mM MgCl2, 25 mM KCl, 1 mM potassium 32 fluoride, 4 mM DTT, 2 mM ATP, and 2 mM P-PPi (12 dpm/pmol) in 20 mM HEPES-KOH (pH 7.2) and incubated at 37°C. Aliquots were removed at the time points indicated, and the amount of radiolabel bound to acid-washed activated charcoal was quantified to estimate the amount of 32ATP formed. Shown are illustrated results from averaged duplicate reactions that were supplemented with the extracted amino acid pool from pK13 (●), pK14 (X), pK15 (∆), pK16 (◊), or pACYCDuet-1 (○).

Pyl standard 20 amino Specific activity

(100 µM) acids (100 µM) (nmol min-1 mg-1)

+ - 65

+ + 55

- + N.D.*

- - N.D.*

32 Table 3.1 The PPi-ATP exchange reaction catalyzed by PylS is pyrrolysine specific. An amino acid preparation containing 100 µM of each standard natural amino acid was prepared and tested for activity. Incubation for as long as 30 minutes resulted in no detectable isotopic activity into ATP above background

(N.D.*- none detected). Testing lysine at 1 mM concentration also resulted in no detectable isotopic activity above background.

100

PylS tRNAPyl Pyl ATP Specific activity

(1.2 µM) (1.5 µM) (50 µM) (2 mM) (nmol min-1mg-1)

+ + + + 36.2

+ - + + 36.4

- + + + N.D.*

+ + + - N.D.*

+ + - + N.D.*

Table 3.2 The ATP dependent activation of pyrrolysine by PylS is not dependent on tRNA. (N.D.*- none detected)

101

Substrate Apparent Concentration Concentration

Km (µM) of Pyl (µM) of ATP (µM)

ATP 2 100 µM N.A.*

Pyl 53 N.A.* 2 mM

-1 -1 -1 Apparent Vmax = 120 nmol min mg (apparent kcat of 6 min )

32 Table 3.3 Kinetic parameters of the PylS catalyzed PPi-ATP exchange reaction. The dependence of the reaction rate on both pyrrolysine and ATP was verified with two independent experiments performed with two different batches of purified enzyme. One substrate was kept at a constant concentration at least two times higher than its respective apparent Km value. In each case, the

concentration of the second substrate was varied over the range Km/5 to 5-8 Km.

Averaged apparent Km and Vmax values were determined using Haanes-Wolf

plots such as the ones shown in Figure 3.2. (*N.A. – not applicable)

102

CHAPTER 4

CLASS I AND CLASS II LYSYL-tRNA SYNTHETASE MUTANTS AND THE

GENETIC ENCODING OF PYRROLYSINE

4.1 Introduction

As indicated in Chapter 2, deleting the pylT gene and the pylTSBCD promoter

gave rise to a mutant unable to translate UAG as pyrrolysine. This mutant can

not make MtmB or utilize any methylamine as a source of energy or as a cellular

nitrogen precursor. As demonstrated in Chapter 3, the product of the pylS gene

is a Class II AARS with specificity for pyrrolysine. Other natural amino acids do

not serve as substrates for this pyrrolysyl-tRNA synthetase.

A second route for the aminoacylation of tRNAPyl exists in vitro. Species of

Methanosarcina have been shown to possess two non-homologous lysyl-tRNA synthetase (LysRS) genes (Deppenmeier et al, 2002; Galagan et al, 2002). Apart from Methanosarcina, only a few other isolated species in other genera, such as

Nitrosococcus oceani (Klotz et al), Bacillus cereus (Ivanova et al, 2003, Ataide et

al, 2005) and Treponema pallidum (Fraser et al, 1998) have been shown to

possess both LysRS genes. The lysK gene encodes a Class I lysyl-tRNA

103

synthetase (LysRS1) which was first discovered in the methanogen

Methanococcus maripaludis (Ibba et al, 1997) and has subsequently been

discovered in many archaeal and a few bacterial genera (Woese et al, 2000;

Ambrogelly et al, 2002). The lysS gene encodes a Class II lysyl-tRNA synthetase

(LysRS2) gene. LysRS2 enzymes are the only lysyl-tRNA synthetase enzymes in

most bacterial and eukaryotic species.

A complex of LysRS1 and LysRS2 with tRNAPyl has been shown to attach

lysine to tRNAPyl slowly in vitro (Polycarpo et al, 2003). This has been proposed

to be the first step of a pathway in which lysyl-tRNAPyl is converted to pyrrolysyl-

tRNAPyl (Polycarpo et al, 2003; Ibba and Söll, 2004). This proposed indirect route

has precedence in the synthesis of selenocysteinyl-tRNASec from seryl-tRNASec

(Böck et al, 2004). Another possibility is that the lysylation of tRNAPyl plays an

unknown role in the metabolism of the methanogen. The in vivo significance of the lysylation of tRNAPyl has not been tested in any study prior to this one.

Recent reports have mentioned the presence of both the indirect and direct

tRNAPyl charging routes (Namy et al, 2004; Veit et al, 2005; Klotz et al, 2006

Stortchevoi 2006; Delarue, 2007). The need to test the relevance of the

LysRS1/LysRS2 pathway in vivo has also been mentioned (Stortchevoi 2006).

Alternatively, the presence of both lysK and lysS in Methanosarcina may have

no relation to UAG translation as pyrrolysine. It has been hypothesized that the

both the LysRS1 and the LysRS2 enzymes have been retained in vivo due to differences in kinetic performance. For example, differences in resistance to

104

tRNALys misacylation with substrate analogs, resistance to inhibitory effects, or even the intrinsic rate kinetics can account for retention of the non-homologous enzymes (Jester et al, 2003; Levengood et al, 2004; Shaul et al, 2006).

In this chapter, deletion mutants of lysK or lysS were characterized to examine the relevance of these hypotheses in vivo.

4.2 Experimental Procedures

4.2.1 Organisms, growth conditions and reagents

Methanosarcina acetivorans C2A (Sowers et al, 1984) was cultured

anaerobically using DSM medium 304 (Sowers and Schreier, 1995) at 37°C with the modifications mentioned in Chapter 2. A further modification, however, was that nitrogen gas was replaced by 5% CO2 in the nitrogen gas phase. Cultures of

M. acetivorans were grown in 10 mL medium in 27 mL rubber-stoppered serum

vials for growth studies, or in 50 mL medium in rubber-stoppered serum vials for nucleic acid and protein extraction. The growth substrates were 100 mM methanol, 40 mM MMA, 40 mM DMA, 40 mM TMA, or 40 mM sodium acetate.

The methylamines were added as hydrochloride salts. For growing the two mutant strains containing the pac cassette (Gernhardt et al, 1990; Rother and

Metcalf, 2005; Sandbeck and Leigh, 1991) puromycin was added to this modified

DSM media at a final concentration of 2 µg/mL (Zhang et al, 2000).

105

4.2.2 ∆lysKc::mm and ∆lysS::pac strains

The ∆lysKc::mm strain was obtained from the lab of W. Metcalf. This strain possesses the first 421 bases of the 1533 bases of lysK, while the remainder has been deleted. The ∆lysS::pac strain was constructed by K. Richter in our lab.

Briefly, the lysS gene and 200 bp upstream and downstream were replaced with the pac cassette. A meticulous description of the construction of both strains is detailed in Mahapatra et al (2007).

4.2.3 Southern hybridization

Southern hybridization was performed essentially as described in Chapter 2.

Genomic DNA from wild- type or ∆lysKc::mm was restriction digested with EcoRI or PstI, electrophoresed, transferred to a Hybond nylon membrane (GE

Healthcare, Piscataway, NJ), and probed with an oligonucleotide that had been labelled with [γ-32P]-ATP and polynucleotide kinase. The probe used for comparing the wild-type and the lysK strains was

GCTGACAACTACGACCCTCTGCGCAAG GTTTACCCTTTCC, which is nucleotide 181 to 220 of the reading frame M. acetivorans lysK gene (Genbank

Accession No. NC_003552, region: 624116–625717). The probe used for comparison of wild-type and the lysS strains was

ATGCACACAGTCATTTCAATAATGCATT, which begins 691 nucleotides after the reading frame of lysS (NC_003552, region 885884–887419).

106

4.2.4 Anti-MtmB immunoblotting in wild-type and mutant strains

Cell extracts from two to three combined 50 mL cultures were prepared as described in Chapter 2, except the total volume of soluble proteins obtained after ultracentrifugation was reduced in half by using an Amicon Centricon YM-30

(Millipore, Billerica, MA). The retained protein was dialyzed against 50 mM

MOPS, pH 7.0 at 4°C overnight. All protein concentrations were determined using the bicinchoninic acid method with bovine serum albumin as the standard.

Immunoblotting of mtmB gene products in soluble proteins from M. acetivorans strains was then performed as described in Chapter 2.

4.2.5 Isolation of genomic DNA and tRNA

DNA from M. acetivorans was isolated from stationary phase cells as described in Chapter 2. Isolation of the in vivo aminoacylated unfractionated

tRNA for assessment of relative amounts of charging was from two collected 50

mL cultures grown on TMA to mid-exponential phase of growth by the procedure outlined in Chapter 2. Nucleic acids were quantified spectrophotometrically.

Quality of DNA and tRNA preparations was examined by visualization of bands

after electrophoresis in 1% agarose gels.

4.2.6 Acid-urea gel electrophoresis and Northern hybridization

Charged and uncharged tRNA species (Varshney et al, 1991) were detected

from M. acetivorans as detailed in Chapter 2. Briefly, following electrophoresis in

107

acid-urea denaturing polyacrylamide gels, the samples were electroblotted to

Hybond-N+ membranes (GE Healthcare, Piscataway, NJ). Probing with probes

complementary to the entire length of tRNAPyl; or to 23 nucleotides

Lys complementary to a portion of tRNA CUU (TAGGCCGACTGGTTAAAAGCCAG);

Lys or tRNA CUU (CGGGCCTAAGGATTAAGAGTCCA), was performed at 54°C for

16 h. The radioactive bands were visualized on a STORM phosphorImager (GE

Healthcare, Piscataway, NJ) subsequent to washing of the membrane. Band

intensity was calculated using ImageQuant software (GE Healthcare) from the

ratio of the intensity of the band corresponding to the acylated tRNA in each lane

to the summation of the intensities of the band corresponding to the acylated and

deacylated tRNA in the same lane, with both samples corrected for background.

4.3 Results

4.3.1 Construction of M. acetivorans mutants lacking LysRS1 or LysRS2

The observation that the lysK gene, encoding LysRS1, is differentially

expressed in M. barkeri on methanol and TMA, and that the expression patterns

closely match that of mtmB1 on the same substrates, suggested a connection

between methylamine metabolism and lysK expression (Srinivasan, 2005). A

direct explanation for this correlation was suggested by the observation that

LysRS1 and LysRS2 (the lysK and lysS gene products) can slowly aminoacylate

tRNAPyl with lysine in vitro only when present together (Polycarpo et al, 2003).

Since a genetic system has not been developed for M. barkeri MS, we used M.

108

acetivorans to test the necessity of both the lysK and lysS genes for growth on

methylamines. M. acetivorans has emerged, in recent years, as the species of choice for genetic manipulation (Rother and Metcalf, 2005).

A lysK mutant strain was envisioned that would lack a functional gene encoding for LysRS1 and would rely on the LysRS2 enzyme for cellular processes including the lysylation of tRNAPyl (Figure 1). Conversely, the lysS

strain would lack the lysS gene encoding for LysRS2 and would rely solely on

LysRS1 for the formation of lysyl- tRNAPyl.

The ∆lysK421::minimar strain (denoted hereafter as ∆lysKc::mm) had 70% of

the 3’ of the lysK gene deleted. This strain was created in the lab of W. Metcalf

from a previously described transposition of a modified mini-mariner element

containing the puromycin acetyltransferase (pac) cassette (Metcalf et al, 1997)

between the open reading frames of lysS and lysM (Zhang et al, 2000). The

transposon and adjacent DNA was restriction digested and recovered as a self-

replicating plasmid in E. coli. Finally, a BfaI deletion eliminated 70% of the 3’

region of lysK from the plasmid, which was subsequently re-ligated and

linearized with EcoRI for liposome-mediated transformation (Metcalf et al, 1997)

into M. acetivorans. All cloning and transformation steps in the construction of

∆lysKc::mm were done by J.K. Zhang in the lab of W. Metcalf.

A strain with the entire monocistronic lysS gene deleted and replaced with pac

was constructed by K. Richter. This strain was designated ∆lysS::pac. No other

109

annotated genes are disrupted in ∆lysS::pac. Details of the construction of the

∆lysKc::mm and ∆lysS::pac strain are published elsewhere (Mahapatra et al,

2007).

Confirmation of integration was accomplished by Southern hybridizations by probing for sequence regions flanking the targeted genes (Figure 4.2). Additional

Southern hybridization directly for lysS demonstrated that the gene was undetectable in ∆lysS::pac.

4.3.2 Catabolic range of strains bearing lysK or lysS disruptions

As detailed in Chapter 2, the ∆ppylT mutant lacking the pylT gene encoding tRNAPyl and the pyl promoter is unable to utilize methylamines for methanogenesis or nitrogen metabolism. However, this mutant grows normally on methanol and acetate. The phenotype of this strain shows a direct link between the inability to translate UAG as pyrrolysine and a specific defect in methylamine metabolism. In contrast to the ∆ppylT strain, both the ∆lysKc::mm and ∆lysS::pac strains described here utilized all tested substrates. We observed statistically similar growth for mutant and wild-type strains with methanol, TMA, DMA, MMA and acetate (Table 4.1). Since the mutants tested here lack intact lysyl-tRNA synthetase genes, we conclude that either LysRS1 or

LysRS2 can serve as the sole lysyl-tRNA synthetase in M. acetivorans, and that no growth condition tested in this study requires the presence of both LysRS1 and LysRS2.

110

4.3.3 UAG translation in lysS and lysK deletion strains

As described in Chapter 2, deleting the pylT gene and promoter resulted in

the loss of detectable MtmB. To test what effects mutating lysyl-tRNA synthetase

genes would have on MtmB production, we performed immunoblotting of cellular

lysates from wild-type and ∆lysKc::mm and ∆lysS::pac mutants with antibodies

raised against MtmB. In wild-type, the 50 kDa MtmB protein is produced as a

result of translation continuing past the UAG codon found in the mid-coding

region of mtmB transcripts. The ∆lysKc::mm and ∆lysS::pac strains both

produced full-length MtmB protein at levels mimicking those observed with the

wild-type strain (Figure 4.3).

When the pyrrolysine insertion element (PYLIS) found in mtmB transcripts is removed in M. acetivorans, an increase in termination at the internal in-frame

UAG codon is observed resulting in a truncated 23 KDa product (Longstaff et al,

2007b). No such product was detectable in either ∆lysKc::mm and ∆lysS::pac strains (Figure 4.3), indicating that the absence of either LysRS1 or LysRS2 did not cause an enhancement of termination at the internal in-frame UAG codon.

4.3.4. Aminoacylation of tRNAPyl and tRNALys in mutant strains

Three tRNA species are potential substrates of LysRS1 and LysRS2 in

Methanosarcina. In addition to tRNAPyl which can be lysylated slowly in vitro in

111

the presence of both enzymes (Polycarpo et al, 2003), two isoforms of tRNALys

with the UUU or CUU anticodons are present. These two isoacceptors of tRNALys have been demonstrated in vitro to be substrates of LysRS1 and LysRS2

(Ambrogelly et al, 2002; Polycarpo et al, 2003). We examined levels of cellular tRNA to determine if aminoacylation was compromised in vivo due to the lack of functional LysRS1 or LysRS2. Cellular tRNA was isolated from mid-exponential cultures of the indicated M. acetivorans strains grown on TMA. Charged and uncharged tRNAs were separated by acid-urea polyacrylamide gel electrophoresis as described in Chapter 2 and elsewhere (Varshney et al, 1991;

Jester et al, 2003; Blight et al, 2004). After separation, charged and uncharged species were detected by Northern hybridization using probes specific for

Pyl tRNA , tRNACUU or tRNAUUU. The upper band in each lane corresponds to

aminoacyl-tRNA, while the lower band corresponds to unacylated tRNA, as

shown by a similar migration pattern obtained upon mild alkaline hydrolysis of the ester bond of the aminoacyl-tRNA. Images were scanned using a phosphorImager and percent charging of each tRNA was determined using

ImageQuant software as described in Experimental Procedures. Values were calculated using ImageQuant software from the ratio of the intensity of the band corresponding to the acylated tRNA in each lane to the intensity of the band corresponding to the acylated tRNA plus deacylated tRNA in the same lane minus a gel background. To minimize deviation, three to six replicate analyses were performed in testing levels of aminoacylation for each tRNA species.

112

Neither the lysK nor the lysS deletion strain showed any statistically significant

deviation in aminoacylation of tRNAPyl when compared with wild-type (Figure 4.4,

Table 4.2). From Northern blots, we observed that the ratios of charged to uncharged tRNAPyl were similar. The differences in tRNAPyl aminoacylation in

replicate sets between wild-type, ∆lysKc::mm and ∆lysS::pac were deemed not to be statistically significant by Student’s t-test at the 95% confidence interval.

It has been demonstrated in vitro that both LysRS1 and LysRS2 aminoacylate tRNACUU well, but LysRS1 aminoacylates tRNAUUU poorly (Polycarpo et al, 2003).

To examine if reliance solely on LysRS1 or LysRS2 would influence the

aminoacylation of either tRNALys isoform, we calculated levels in wild-type,

∆lysKc::mm and ∆lysS::pac, subsequent to Northern blotting of acid urea-gels.

Probes specific for tRNAUUU and tRNACUU were employed. Between the two

tRNALys probes there is <50% sequence similarity. Representative Northern blots

are shown in Figure 4.5 and show that both isoforms of tRNALys are

aminoacylated in wild-type and mutant strains.

The ∆lysKc::mm and wild-type strains showed statistically similar amounts of

tRNAUUU and tRNACUU lysylation indicating that the LysRS2 had no preference

for either tRNA substrate and that the lack of a functional LysRS1 did not hamper

aminoacylation of lysyl tRNAs. However, ∆lysS::pac possessed statistically significant lower levels of both tRNAUUU and tRNACUU aminoacylation, showing

that reliance solely on LysRS1 led to lower amounts of lysyl-tRNA.

113

4.4 Discussion

It has been hypothesized that the lysS gene became part of the genome of a

Methanosarcina ancestor that possessed only the lysK gene through horizontal gene transfer (Shaul et al, 2006). The results presented in this chapter, when taken with expression studies conducted by G. Srinivasan (2005), indicate the extent to which this laterally-transferred lysS gene has become a part of the methanosarcinal metabolism. Surprisingly, the deletion of lysK or lysS does not affect growth in this methanogen. This suggests that the specific requirements of either gene are not essential for any process required growth in the laboratory on methylamines, methanol, or acetate, including the genetic encoding of pyrrolysine.

The correlation between expression of lysK and mtmb1 in M. barkeri first suggested a connection between the use of LysRS1 and growth on methylamines (Srinivasan, 2005). However, the regulation of lysK expression might not be a common feature of all species of Methanosarcina, since in this study, a mutant of M. acetivorans that lacked lysS was found to grow normally.

The assumption is that at least in this mutant, lysK is constitutively expressed or compensated by secondary mutation. A recent study of mRNA expression in M. mazei from mid-logarithmic phase growth came to the conclusion that there is no difference in expression levels of either lysK or lysS mRNA in that organism (Veit et al, 2005).

114

As examined in Chapter 2, loss of tRNAPyl results in a deficiency in the ablility

to utilize methylamines for growth and nitrogen metabolism. The coexpression of

lysK and mtmB might suggest a supporting role for the LysRS1/LysRS2 complex in aminoacylating tRNAPyl. Therefore, it was essential to test this hypothesis in

vivo. We have shown here that the formation of lysyl-tRNAPyl as either an end-

product or in the indirect synthesis of pyrrolysine is not essential, since mutants

lacking functional genes for LysRS1 or LysRS2 grew normally on methylamine,

methanol, and acetate with respect to wild-type. Both mutant strains possessed

levels of the pyrrolysyl-protein, MtmB, at levels similar to that of wild-type. The in vivo levels of aminoacylated tRNAPyl as isolated from wild-type and mutants

lacking functional LysRS1 or LysRS2 were comparable. The LysRS1 enzyme has

been proposed to function as a tRNA chaperonin for lysylation of tRNAPyl;

LysRS2 is proposed to have to catalyze the lysylation of tRNAPyl (Polycarpo et al,

2003). Formally, the lysK mutant characterized here retains the ability to encode

a small portion of the non-catalytic tRNA-binding domain of LysRS1 which may

as yet, play a role as a tRNA chaperonin. However, the ∆lysS::pac mutant has the entire lysS gene deleted. Therefore, if there was a role for the

LysRS1/LysRS2 complex in the genetic encoding of pyrrolysine, lysylation of tRNAPyl should not occur in this strain. However, our results demonstrate that

tRNAPyl is in fact aminoacylated in the ∆lysS::pac mutant strain at levels similar to that of wild-type. The results presented in this chapter are consistent with sequence data derived from non-methanosarcinal organisms whose genomes

115

possess the pylTSBCD gene cluster, but do not encode annotated Class I and

Class II LysRS-encoding genes. The methanogenic archaeon M. burtonii

(Goodchild et al, 2004) has been shown to possess the lysK gene for LysRS1,

but not lysS encoding for LysRS2. Conversely, the gram-positive bacterium, D. halfniense Y51 possesses the lysK gene, but lacks an identified lysS gene

(Nonaka et al, 2006). Taken together, our data and the absence of key genes

from annotated genomes, clearly demonstrate that an indirect pathway requiring the lysylation of tRNAPyl as a route to forming pyrrolysyl-tRNAPyl is at best a minor

route in the translation of UAG as pyrrolysine. These data also indicate that the in

vivo lysylation of tRNAPyl by LysRS1/LysRS2 for any other cellular activities is

dispensable in organisms which retain the pyl gene cluster.

Lysyl-tRNAPyl can bind EF-Tu from E. coli (Théobald-Dietrich et al, 2004). A.

Meyer has demonstrated that LysRS1/LysRS2 does not substitute for PylS aminoacylation of tRNAPyl to allow amber-suppression in the E. coli recombinant system (Mahapatra et al, 2007). This result is consistent with the data presented here, and also indicates that PylS aminoacylation of tRNAPyl is more robust that

the concerted action of LysRS1 and LysRS2 under in vivo conditions.

Regardless, given that co-expression patterns of lysK and mtmb1 have been shown to match (Srinivasan, 2005), it remains possible that lysK expression in M. barkeri MS is associated with the lysylation of tRNAPyl by a complex of

LysRS1/LysRS2. It has been suggested that lysyl-tRNAPyl itself is utilized in the

insertion of lysine in MMA methyltransferase in the unanticipated scenario that

116

pyrrolysine is not synthesized and full-length MMA methyltransferase is required.

There are a number of reasons why this scenario is unlikely. Methylamine

methyltransferases are abundant cellular proteins when methylamines are

utilized by the methanogens. On the other hand, the lysylation reaction is slow in

vitro and may not be able to cope with the required demand for intact functional

methyltransferases. Further, the chemical reactivity of pyrrolysine with ammonia

and other amines, along with structural alignments argue strongly for a role for

pyrrolysine in the catalysis of methyltransfer by methylamine methyltransferases

(Hao et al, 2004; Krzycki, 2004). It seems unlikely that the catalytic role of

pyrrolysine is substituted by lysine insertion in these methyltransferases. Recent

data from our laboratory further augment a unique role for pyrrolysine in catalysis

by methylamine methyltransferases (D. Longstaff and J. Soares, personal

communication). Another possibility is that UAG translation with lysyl- tRNAPyl

serves to prevent ribosomal stalling under pyrrolysine limiting conditions.

However, our laboratory has recently demonstrated that UAG can function as a

stop codon within Methanosarcina, obviating such a role for lysyl-tRNAPyl

(Longstaff et al, 2007a). Also, we did not observe increased termination at UAG in the absence of either lysK or lysS. A plausible hypothesis is that the slow assembly of lysyl-tRNAPyl allows its use in a hitherto unknown regulatory role.

Scenarios in which this molecule serves to signal the cellular status of some

aspect of pyrrolysine availability or UAG translation are possible. A caveat is that

the concentration of lysyl-tRNAPyl must be relatively low under normal

117

circumstances, since otherwise it would serve as a competitor for pyrrolysyl-

tRNAPyl in translation. It has been demonstrated that lysyl-tRNAPyl and pyrrolysyl-

tRNAPyl migrate differently in denaturing acid-urea gels (Polycarpo et al, 2004). If

both species were present, one might expect an additional acylated band on visualizing unfractionated tRNAs. Yet, we did not observe any additional bands in gels with unfractionated tRNA from in vivo samples. A final assessment of the relevance of the role of lysylated tRNAPyl will be possible in strains lacking pylS

and the pylBCD biosynthetic genes.

The observed differences in the lysK and lysS deletion mutants are consistent

with different roles for LysRS1 and LysRS2 in cellular metabolism. LysRS1 is

more discriminating with respect to substrate specificity than LysRS2, and this

observation has been hypothesized to be a reason why lysK is retained in many

genomes (Jester et al, 2003; Levengood et al, 2004; Wang et al, 2006). In

Methanosarcina, we show that in the absence of the lysK gene, any possible

misacylation by the remaining lysS gene encoding the more promiscuous

LysRS2 is not a severe disadvantage to the cell under tested conditions. It is

possible, however, that in environments where methylamines predominate, lysine

analogs may require the possession of the LysRS1 encoding gene.

It has been demonstrated in vitro that LysRS1 and LysRS2 differ in their ability to acylated the two tRNALys isoacceptors (Ibba et al, 1999; Söll et al,

2000). We have demonstrated here that this is not the case in vivo in

Methanosarcina. The deletion of either the LysRS1 or LysRS2 encoding gene did

118

not abolish the aminoacylation of either of the two isoacceptors of tRNALys. This is consistent with what has been observed in other studies in bacteria in which the replacement of lysS by lysK has not affected isoacceptors aminoacylation levels in vivo (Jester et al, 2003). We did observe, however, that the deletion of the LysRS2 encoding lysS gene led to a statistically significant drop in the aminoacylation levels of both isoacceptors of tRNALys. In contrast, deletion of the

LysRS1 encoding lysK gene had no effect on aminoacylation of tRNALys. This finding suggests that the LysRS1 enzyme cannot fully compensate for the loss of the LysRS2 enzyme. The favorable in vitro kinetics of LysRS2 as compared to

LysRS1 has prompted Shaul et al (2006) to suggest this as the reason why archaea such as Methanosarcina have recruited and retained the Class II version of these enzymes. The data presented in this study does not allow for a direct test of this hypothesis. However, we show that lysS, a gene probably horizontally transferred from a bacterial lineage to an ancestral archaeon, now plays the main role in aminoacylation of tRNALys in M. acetivorans.

119

Figure 4.1 The ∆lysKc::mm and ∆lysS::pac mutant strains of M. acetivorans lack functional LysRS1 and LysRS2 respectively.

120

Figure 4.2 Confirmation of ∆lysS::pac and ∆lysKc::mm mutations in M. acetivorans by Southern analysis. DNA from wild-type and mutant strains were restriction digested, electrophoresed, blotted, and probed with a radiolabelled oligonucleotide complementary to a region upstream of the target gene in wild type that is also preserved in section upstream in the desired deletion strain. (Panel A) Wild-type (wt) and ∆lysKc::mm (DK) DNA digested with EcoRI; restriction fragments near the predicted sizes of 4.6 (wild type) and 6.2 (∆lysKc::mm) kb were detected. Panel B. Wild-type (wt) and ∆lysS::pac (DS) DNA digested with PstI; restriction fragments near the predicted sizes of 4.3 (wild type) and 2.8 (∆lysS::pac) kb were seen.The migration position and size in kb of DNA standards are given on the side of each membrane.

121

Figure 4.3 Anti-MtmB immunoblotting of total soluble protein from wild- type, ∆lysKc::mm and ∆lysS::pac strains. Cells were grown on MMA and lysed to compare relative MtmB abundance. (Panel A) Lane 1 was loaded with 2 mg of purified MtmB protein, while 200 mg of soluble protein from wild-type or ∆lysKc::mm strains was loaded, respectively, in lanes 2 and 3. (Panel B) Lane 1 was loaded with 2 mg of purified MtmB, while 50 mg of soluble protein from wild- type or ∆lysS::pac strains was loaded, respectively, in lanes 2 and 3.

122

Figure. 4.4 Relative tRNAPyl aminoacylation levels in wild-type, ∆lysKc::mm and ∆lysS::pac strains. Cellular tRNA was isolated from exponential wild-type (Panels A and B; lanes 1 and 2), ∆lysKc::mm( Panel A, lanes 3 and 4) and ∆lysS::pac (Panel B, lanes 3 and 4) strains, electrophoresed on acid-urea agarose gels, electroblotted, and probed with a specific oligonucleotide probe complementary to the entire length of tRNAPyl. Lanes 2 and 4 were loaded with the cellular tRNA as isolated, while lanes 1 and 3 were subjected to mild alkaline hydrolysis in order to deacylate tRNAPyl as a control.

123

Figure 4.5 Relative tRNALys aminoacylation levels in wild-type, ∆lysKc::mm and ∆lysS::pac strains. Both Panel A and Panel B are Northern blots of acid- urea gels. Lanes 2, 4 and 6 in both Panel A and Panel B were untreated; lanes 1, 3 and 5 were subjected to mild alkaline hydrolysis. (Panel A) Northern analysis of tRNAUUU from wild-type, ∆lysKc::mm and ∆lysS::pac. Unfractionated mature tRNA was isolated from exponential growth- phase wild type (lanes 1 and 2), ∆lysKc::mm (lanes 3 and 4) and ∆lysS::pac (lanes 5 and 6), electrophoresed on acid-urea gels and transferred prior to probing. Hybridization was performed with a tRNAUUU-specific oligonucleotide probe complementary to the 23 nucleotides, including the anticodon but with < 50% identity to the sequence of tRNACUU. (Panel B) Northern analysis of tRNACUU from wild type, ∆lysKc::mm and ∆lysS::pac. Unfractionated mature tRNA was isolated from exponential growth- phase wild type (lanes 1 and 2), ∆lysKc::mm (lanes 3 and 4) and ∆lysS::pac (lanes 5 and 6), electrophoresed on acid-urea gels and transferred prior to probing. For hybridization, a tRNACUU-specific oligonucleotide probe complementary to the 15 nucleotides including the anticodon but with < 50% similarity to the sequence of tRNAUUU was employed.

124

Growth Genotype Generation time (h) substrate

Methanol wild-type 23 +/- 9

∆lysKc::mm 22 +/- 10

∆lysS::pac 20 +/- 13

TMA wild-type 22 +/- 5

∆lysKc::mm 16 +/- 2

∆lysS::pac 19 +/- 5

DMA wild-type 24 +/- 3

∆lysKc::mm 24 +/- 5

∆lysS::pac 24 +/- 5

MMA wild-type 38 +/- 15

∆lysKc::mm 50 +/- 17

∆lysS::pac 42 + 14

Table 4.1 Growth of lysyl-tRNA mutants with various growth substrates.

The optical density at 600 nm was followed in 27 mL anaerobic culture tubes.

The values for exponential generation time are mean values with standard deviation for 5-9 replicates.

125

tRNA Genotype Percentage charging pyrrolysyl tRNA wild-type 26 +/- 8 tRNACUA ∆lysKc::mm 29 +/- 8

∆lysS::pac 32 +/- 12 lysyl tRNA wild-type 71 +/- 3 tRNAUUU ∆lysKc::mm 70 +/- 3

∆lysS::pac 55 +/- 4 lysyl tRNA wild-type 72 +/- 5 tRNACUU ∆lysKc::mm 68+/- 7

∆lysS::pac 59 +/- 8

Table 4.2. In vivo levels of tRNAPyl and tRNALys aminoacylation in wild type,

ΔlysK and ΔlysS strains. The values are mean values with standard deviation for 3-7 replicates.

126

CHAPTER 5

AMINO ACID SPECIFICITY OF PYRROLYSYL-tRNA SYNTHETASE

5.1 Introduction

In Chapter 3, PylS was demonstrated to directly and robustly aminoacylate

tRNAPyl with chemically-synthesized pyrrolysine (Figure 5.1; compound 1) both in

vitro and in vivo (Hao et al, 2004; Blight et al, 2004). In another study, PylS was

shown to aminoacylate tRNAPyl with an enamine analog of pyrrolysine (Polycarpo

et al, 2004). Unfortunately, detailed analysis of the kinetics of PylS catalyzed

activation of enamine pyrrolysine was not possible because of the racemic nature

of the preparation (Polycarpo et al, 2004). Subsequently, two other pyrrolysine

analogs, Nε-D-prolyl-L-lysine (D-prolyl-lysine) and the commercially-available, Nε

-cyclopentyloxycarbonyl-L-lysine (Cyc; Figure 5.2, compound 7) were shown to serve as PylS substrates (Polycarpo et al, 2006). Cyc, is larger than pyrrolysine and other previously tested analogs as it contains a carbonyl group between the pentane ring and the epsilon nitrogen of lysine. Therefore, PylS-catalyzed activation of Cyc would be predicted to proceed at a much slower rate than the

127

activation of substrates with geometries more reminscient of pyrrolysine.

Unfortunately, the catalytic efficiency of PylS in activation using any alternative

substrate has not been compared to the efficiency of PylS activation of pyrrolysine in any report prior to this one. Further, a definitive probing of the amino acid specificity of PylS had not been possible, as the structure of PylS had

not been resolved.

Recently, three crystal structures of truncated PylS from Methanosarcina

mazei containing the catalytic domain with AMP-PNP; Cyc; and enamine

pyrrolysine-AMP plus pyrophosphate, respectively were resolved (Kavran et al,

2007). Co-crystallization of enamine analog of pyrrolysine with PylS indicated

that that the ring of this analog was coordinated in a deep hydrophobic pocket of

PylS (Kavran et al, 2007). Two residues of this pocket, namely a tyrosine at

position 384 and an asparagine at position 346 of PylS were postulated to be the

primary determinants for PylS recognition of enamine pyrrolysine through specific

hydrogen bonding interactions (Kavran et al, 2007). Both residues are conserved

among all known PylS variants. One of the conserved residues, Tyr384, was

hypothesized to hydrogen bond with the ring nitrogen of enamine pyrrolysine

(Kavran et al, 2007); interestingly, this interaction is not possible with Cyc,

because of the lack of an analagous hydrogen bonding partner in the

cyclopentane ring. Unfortunately, the structure of Cyc in the hydrophobic pocket

of PylS was disordered (Kavran, et al, 2007). Further, the location of the

hydrogen bonding partners that are associated in hydrogen bonding interactions

128

with the other specificity-determining residue identified in that study, Asn346, are different in enamine pyrrolysine and Cyc (Kavran, et al, 2007).

Our goal in the present study is to gain insights into the specificity of PylS by the functional probing of the active site. We have designed and synthesized analogs of pyrrolysine, which we have tested in PylS-catalyzed pyrrolysine activation and tRNA aminoacylation. By comparing PylS catalysis using different substrates, we have elucidated key features of pyrrolysine recognition. Our results indicate that a hydrophobic cyclopentane ring structure is essential for

PylS recognition. Our results also suggest that hydrogen bonding between the nitrogen at the C2 position of the ring with a residue of PylS is required for efficient activation. In agreement with our in vitro data, amino acid substrates of

PylS also incorporated in MtmB1 in an Escherichia coli reporter system with high efficiency. One of the compounds we tested, 2-amino-6-((R)-tetrahydrofuran-2- carboxamido)hexanoic acid (Thf) shows particular promise as a lead for synthesizing other amino acids with novel functionalities for inserting in proteins.

5.2 Experimental Procedures

5.2.1 General

The commercially available amino acids tested for tRNAPyl aminoacylation by

PylS: L-lysine (Figure 5.1, compound 2), D-proline (Figure 5.1, compound 3), Nε- acetyl-L-lysine (Figure 5.1, compound 4) and Nε-benzoyl-L-lysine (Figure 5.1,

compound 5) were purchased from Sigma (Saint Louis, MO). The previously

129

described PylS substrate, (S)-2-amino-6-(cyclopentoxycarbonylamino)hexanoic acid, also referred to as Nε -cyclopentyloxycarbonyl-L-lysine (Cyc; Figure 5.2,

compound 7) (Polycarpo et al, 2006) was also obtained from Sigma (Saint Louis,

MO). L-pyrrolysine was synthesized as described in the lab of M. Chan (Hao et

al, 2004). For this study, two additional Pyl analogs, namely, 2-amino-6-

(cyclopentanecarboxamino)hexanoic acid, (Cyp; Figure 5.2, compound 2); and 2-

amino-6-((R)-tetrahydrofuran-2-carboxamido)hexanoic acid, (Thf; Figure 5.2, compound 3) were synthesized in the lab of M. Chan. Finally, 2-amino-6((R)-

tetrahydrofuran-3-carboxamido)hexanoic acid, (OxR Figure 5.2, compound 4); 2-

amino-6-((S)-tetrahydrofuran-3-carboxamido)hexanoic acid, (OxS Figure 5.2,

compound 5); and 2-amino-6-(thiopene-2-carboxamido)hexanoic acid, (Thp

Figure 5.2, compound 6) were synthesized by 21st Century Biochemicals

(Marlboro, MA).

32 5.2.2 PPi-ATP exchange with PylS

32 The PPi-ATP exchange reaction was performed as described previously

(Chapter 3; Blight et al, 2004; Longstaff et al, 2007). All reactions were performed

in duplicate at 37°C with at least two batches of purified recombinant PylS.

Standard reactions with a final volume of 100 µL were set up with 20 mM

HEPES-KOH (pH 7.2), 10 mM MgCl2, 25 mM KCl, 1 mM KF, 4 mM DTT, 2 mM

32 ATP, and 2 mM [ P]PPi (10-22 dpms/pmol, Perkin Elmer, Waltham, MA). Amino

acid concentrations in the standard assay were 100 µM for pyrrolysine and 2 mM

130

for all analogs tested, except when determining kinetic parameters. The concentration of PylS was 0.3 to 1 µM.

At specific times, 20 µL aliquots were removed and quenched in 500 µL of a stop-reaction suspension and counted in a liquid scintillation counter as decribed in Chapter 3.

For experiments determining kinetic parameters, independent reactions in duplicate were performed as described, with amino acid concentrations varied over the range Km/5 to 5-8 Km (Chapter 3; Blight et al, 2004). Apparent kinetic parameters were obtained by non-linear regression using Prism software

(GraphPad, San Diego, CA).

5.2.3 Ligation of amino acid to tRNAPyl catalyzed by PylS

Aminoacylation assays were performed essentially as described previously for

PylS (Blight et al., 2004; Mahapatra et al, 2007) in 10 mM HEPES-KOH pH 7.2 using M. acetivorans unfractionated tRNA (8 µg) that had been purified as described previously (Blight et al., 2004; Mahapatra et al, 2007). The concentrations of amino acids used in each experiment are supplied in the figure legends. Reactions were initiated with the addition of 0.3–1.1 µM of purified recombinant PylS. Reactions were incubated at 37°C for 50 min and then terminated with an equal volume of 0.3 M sodium acetate, 8 M urea, pH 5.0 for subsequent analysis by the acid-urea Northern blotting procedure (Varshney et al., 1991) with modifications (Jester et al., 2003). After samples were

131

electrophoresed in acid-urea denaturing polyacrylamide gels, the samples were

electroblotted to Hybond-N+ membranes (GE Healthcare, Piscataway NJ). A

probe complementary to the entire length of tRNAPyl, which had previously been

used in the studies documented in Chapters 2 and 3, was used in the detection

of charged (upper band) and uncharged (lower band) tRNA species (Blight et al;

2004; Mahapatra et al; 2006).

5.2.4 In vivo incorporation of analogs in E. coli

A strain of E. coli BL21 (DE3) containing the pEC03 plasmid was used in this

study (Blight et al., 2004). The pEC03 plasmid contains M. barkeri mtmb1 and

pylS; and M. acetivorans pylT genes cloned into the pETDuet-1 vector (EMD

Biosciences). We tested for in vivo incorporation following previously described

methods (Blight et al., 2004). Briefly, an overnight culture of a strain containing

the pEC03 plasmid was grown in 3 mL of Luria–Bertani broth with 100 mg mL-1 ampicillin. Subsequently, 20 mL of the overnight culture was inoculated into 1 mL of fresh medium and grown to an optical density OD600 of 0.6. Cultures were

transferred to polypropylene tubes and induced with 1 mM IPTG for 3 h with

shaking in the presence or absence of 1 mM of pyrrolysine or 1-10 mM of

pyrrolysine analog. Lysates were normalized for cell growth prior to anti-MtmB

immunoblot analysis to determine relative abundance of mtmB1 UAG-termination

and UAG-translation products. Immunoblotting of mtmB gene products from E.

coli cellular lysates was then performed as described (Blight et al., 2004).

132

5.3 Results

5.3.1 Activation and aminoacylation of tRNAPyl with pyrrolysine analogs

As a starting point for a comprehensive analysis of the amino acid specificity of PylS, we tested the commercially-available analogs, L-lysine, D-proline, and

Nε-acetyl-L-lysine (Figure 5.1). None of these amino acids substituted for

pyrrolysine in PylS-mediated tRNAPyl aminoacylation. Our observation is

consistent with previous biochemical data on PylS activity (Polycarpo et al, 2003;

Blight et al, 2004; Polycarpo et al, 2006). It has been observed that the amino

acid binding pocket of PheRS and that of PylS are both hydrophobic and similar

in organization; in addition, structural phylogeny indicates that PylS is derived

from PheRS (Kavran et al, 2007). We therefore tested Nε-benzoyl-L-lysine to see

if the pyrroline ring could be substituted with a benzene ring for PylS recognition.

We did not detect any aminoacylated tRNAPyl in PylS catalyzed aminoacylation

reactions using this analog (Figure 5.1).

Although interactions between the methyl moiety at the C4 of the pyrroline

ring and the hydrophobic active site of PylS might improve recognition of

pyrrolysine, this moiety has been shown not to be essential for aminoacylation

(Polycarpo et al, 2006; Kavran et al, 2007). In this study, all the analogs we

tested lacked this methyl group. Hydrogen bonding interactions between specific

residues and amino acid substrates are essential in conferring the amino acid

specificity of aminoacyl-tRNA synthetases for the common amino acids (Fersht,

1985). In addition, hydrophobic interactions are important in determining the

133

specificity of a number of amino acids (Fersht, 1985; Mosyak et al, 1995). As Nε- acetyl-L-lysine and Nε-benzoyl-L-lysine are not PylS substrates, we expected

features of the pyrroline ring to be essential for PylS recognition, as the ring is

absent from both Nε-acetyl-L-lysine and Nε-benzoyl-L-lysine. Within the ring, the

imine bond nitrogen is the only potential hydrogen bonding partner. Based on x-

ray crystallographic data, the nitrogen atom in the ring of enamine pyrrolysine has been hypothesized to hydrogen bond with Tyr-384 of PylS (Kavran et al,

2004). To test if this hydrogen bonding interactions is essential, we synthesized

2-amino-6-((R)-tetrahydrofuran-2-carboxamido)hexanoic acid, (Thf) (Figure 5.2, compound 3). In Thf, the imine nitrogen at the 2-position of the pyrroline ring of

Pyl is replaced with oxygen, which is also electronegative and capable of forming analagous hydrogen bonds.

Based on known interactions between the twenty common amino acids and cognate aminoacyl-tRNA synthetases, we predicted that the cyclopentane ring would be a specificity-conferring determinant. To test the requirement of the ring, we synthesized 2-amino-6-(cyclopentanecarboxamino)hexanoic acid, (Cyp)

(Figure 5.2, compound 2), which bears a saturated cyclopentane ring lacking the imine-bond nitrogen or any analagous hydrogen bonding partner.

32 We tested Thf and Cyp for PylS catalyzed activation in the PPi-ATP

exchange reaction. Isotopic exchange into 32P-ATP above background was

detected for Thf (Figure 5.3, Panel A) and Cyp (Figure 5.3, Panel B), with a

higher rate of reaction obtained with Thf. Having determined that a hydrogen

134

bonding partner in the ring enhances the activation of the analog, we wanted to

determine if this hydrogen bonding partner must be at a specific position in the

ring in order for this enhancement to occur. To test this, we examined 2-amino-

6((R)-tetrahydrofuran-3-carboxamido) hexanoic acid, (OxR) (Figure 5.2, compound 4) and 2-amino-6-((S)-tetrahydrofuran-3-carboxamido)hexanoic acid,

(OxS) (Figure 5.2, compound 5). We did not detect any isotopic exchange above

background with OxR or OxS in the exchange assay (Figure 5.3, Panel C).

However, such activity if present may also be undetected as it might be below

the limit of detection. Nonetheless, what is intriguing about this result is that it suggests that the presence of the electronegative oxygen atom in any other position of the ring hinders activation, possibly due to repulsive force exerted by hydrophobic residues that are in close proximity to this atom in the binding pocket. Next, we tested 2-amino-6-(thiopene-2-carboxamido)hexanoic acid,

(Thp) (Figure 5.2, compound 6). Thp contains an electronegative sulfur atom in place of the the nitrogen of Pyl; however, hydrogen bonding interactions involving sulfur are generally weaker than those involving either nitrogen or oxygen as a partner (Platts et al, 1995). In addition, the thiopene ring of Thp has two unsaturated carbon-carbon bonds that are an additional feature not present in any other analog tested. We were unable to detect activity with Thp. Finally, we tested the commercially-available pyrrolysine analog used in a previous study and in crystallographic studies, Cyc (Figure 5.2, compound 7; Polycarpo et al,

2006; Kavran et al, 2007), and found that it supported PylS- catalyzed activation

135

as previously reported (Polycarpo et al, 2006).

We then determined the steady-state kinetic parameters of activation of all

tested PylS amino acid substrates. We obtained an apparent steady-state

-1 Michaelis rate constant (Km) for Pyl of 50 µM and a kcat of 6 min (Table 1) which was similar to our previously reported values of 53 µM and 6 min-1, respectively

(Blight et al, 2004). We obtained a Km for PylS catalyzed activation of Thf of 375

-1 µM and an apparent kcat of 4 min corresponding to 8% of the catalytic efficiency of PylS mediated activation of Pyl. Cyp was activated at a substantially lower rate

-1 with a Km value of 5497 µM and a kcat of 2 min corresponding to 0.3% of the

activity of PylS for Pyl. In a previous study, the Km of the commercially available

analog Cyc was determined to be 670 µM; however, no kcat value was published

for the activation reaction (Polycarpo et al, 2006). Here, we determined the

catalytic efficiency of this analog. We obtained a 16% lower Km value of 563 µM

-1 compared to the previous report (Poycarpo et al, 2006) and a kcat of 0.8 min

which corresponded to 0.8% of the activity of PylS for Pyl in activation.

Amino acids which are activated in vitro are not always incorporated in cellular proteins. Many amino acids that are activated are removed before transfer to tRNA or post-transfer by several AARS enzymes via editing reactions (Franckyn et al, 2002). Although editing reactions have not been discovered for PylS, we wanted to test if activation of our most the analog showing the highest activity was followed by ligation to tRNAPyl. Uncharged tRNAPyl and aminoacyl-tRNAPyl can be resolved by acid-urea gel electrophoresis and Northern blotting with

136

tRNAPyl-specific probes (Blight et al, 2004; Polycarpo et al, 2004; Polycarpo et al,

2006). In vitro aminoacylation reactions with PylS were perfomed with Thf in

place of Pyl; 2 mM of Thf supported aminoacylation of tRNAPyl at ~55% levels,

compared with ~65% using 50 µM of Pyl (Figure 5.4).

5.3.2 Incorporation of pyrrolysine analogs in E. coli proteins

An E. coli strain bearing the pEC03 plasmid with the pylS and pylT gene

(encoding tRNAPyl) and the mtmb1 reporter gene has been used expand the genetic code of this organism to include pyrrolysine (Blight et al, 2004). Briefly, expressing the pylT and pylS genes and supplementing cultures with pyrrolysine allows the insertion of pyrrolysine at an internal amber codon in mtmB1. Here, we

used this system to test the in vivo incorporation of those analogs that had

served as PylS substrates in vitro. For the incorporation of an exogenously

added amino acid into a protein in vivo, the amino acid must be stable and non- toxic; imported into the cell; recognized and ligated to a specific tRNA by a cognate AARS; and finally, this charged tRNA must be within an acceptable thermodynamic range for ternary complex formation with the host elongation

factor (Ef-Tu) and GTP (Wang and Schultz, 2006). We found that 1 mM Thf was

incorporated into MtmB with efficiency comparable to that obtained with 50 µM

Pyl (50%) (Figure 5.5, Panel A). We were unable to detect incorporation of Cyp

at 1 mM concentration, but observed full length product when the concentration

of Cyp added was increased to 10 mM, which is higher than the Km for PylS

137

catalyzed activation of this analog (Figure 5.5, Panels A and B). Cyc has been previously shown to be incorporated into E. coli proteins at the UAG codon

(Polycarpo et al, 2006). We found here that Cyc is also inserted in MtmB in our

E. coli reporter system (Figure 5.6).

5.4 Discussion

The amino acid specificity of PylS has been a focus of inquiry ever since the identification of pyrrolysine (Srinivasan, et al, 2002; Hao et al, 2002). In the initial report, it was suggested from in vitro data that PylS recognizes lysine (Srinivasan et al, 2002). However, more detailed studies in vitro (Polycarpo et al, 2003;

Polycarpo et al, 2004; Blight et al, 2004) and in vivo (Blight et al, 2004) established that PylS is a class II aminoacyl-tRNA synthetase with specificity for pyrrolysine. The unique chemical structure of Pyl (Hao et al, 2002; Hao et al,

2004; Soares et al, 2005) allows PylS to discriminate against the other natural genetically-encoded amino acids through a number of hydrophobic and electrostatic interactions. A study detailing the PylS catalyzed activation of an enamine analog of pyrrolysine and the subsequent ligation to tRNAPyl in vitro,

suggested that pyrrolysine analogs could be used to study the specific

interactions between PylS and amino acid substrate (Polycarpo et al, 2004). In

prior reports, the apparent steady-state Michaelis rate constant (Km) value for Pyl

was noted to be 53 µM (Blight et al, 2004), while the Km value for the enamine

analog was 1100 µM (Polycarpo et al, 2004); however, the kcat values for both

138

Pyl and enamine pyrrolysine under steady-state conditions were similar (~6 min-

1). Since the enamine analog used by Polycarpo et al, (2004) was only 50% pure, the 20-fold lower Km with enamine pyrrolysine compared to Pyl might have been due to competitive inhibition by an impurity present in the sample; it might also have been due to weaker affinity of PylS for this analog. However, the extent to which these two factors contributed to a higher Km in PylS catalyzed activation of enamine pyrrolysine could not be determined in that study (Polycarpo et al,

2004). A second study reported the PylS recognition of two other Pyl analogs,

Cyc and D-prolyl-lysine (Polycarpo et al, 2006). However, the catalytic efficiency of PylS in the activation of these analogs was not reported (Polycarpo et al,

2006). Kinetic data of the activation step is a good prognosticator of the kinetics of aminoacylation in the absence of editing mechanisms. At the time of publication, kinetics of aminoacylation with pyrrolysine and analogs were lacking.

Until recently, a major hindrance to a systematic study of specificity determinants was the lack of a resolved crystal structure of PylS. The recent resolution of the crystal structures of a fragment of Methanosarcina mazei PylS complexed with enamine pyrrolysine-AMP plus pyrophosphate, and complexed with Cyc has led to a number of interactions between PylS and Pyl being hypothesized as conferring amino acid affinity and specificity (Kavran et al,

2007). Based on co-crystallization with enamine pyrrolysine, a hydrogen bonding network between five residues of PylS with Pyl has been hypothesized to be involved in recognition; however, specificity for pyrrolysine is thought to arise

139

from hydrogen bond interactions between partners in pyrrolysine and two strictly conserved residues, Asn-346 and Tyr-384 (Kavran, et al, 2007). Interestingly,

Tyr-384 is located on a mobile loop that is hypothesized to close the hydrophobic

pocket upon binding of enamine pyrrolysine. However, there is no equivalent

hydrogen bonding partner in Cyc. Further, resolved crystal structures revealed

that binding of this analog was somewhat disordered unlike structures obtained

upon binding of Pyl (Kavran et al, 2007). In addition, the crystal structure of PylS

complexed with Cyc revealed hydrogen-bonding interaction between Asn-346

and the primary carbonyl of Cyc; the additional water-mediated hydrogen- bonding interaction between Asn-346 that is observed for enamine pyrrolysine

binding is thought to be absent (Kavran et al, 2007). Therefore, very little data on

the recogntion of the natural substrate, Pyl could be gleaned from x-ray crystallographic data on the binding of the commercial analog, Cyc.

As the cycloalkane ring is positioned deep within a hydrophobic pocket of the active site of PylS, affinity for the amino acid is also thought to arise from hydrophobic interactions (Figure 5.7). M. barkeri MS PylS superimposed on the deduced M. mazei PylS structure reveals that the overall structure and the hydrophobicity of the amino acid binding pocket are maintained (Figure 5.8). The results presented in this study are in agreement with the hypothesis that hydrophobic interactions are required for recognition.

Among all the analogs tested, we obtained the highest catalytic efficiency

relative to Pyl for PylS catalyzed activation with Thf. This observation is

140

consistent with the hypothesis that the imine bond nitrogen is a hydrogen

bonding partner with a residue of PylS, perhaps Tyr-384. Replacement with an analogous hydrogen bonding partner, oxygen, allows PylS-catalyzed activation to proceed at a high rate when compared to all analogs tested here and elsewhere

(Polycarpo et al, 2004; Polycarpo 2006). We further establish that the position of this hydrogen bonding partner in the ring is essential, since OxR and OxS are not activated by PylS. Thp possesses a sulfur atom in the same position; however, hydrogen bond complexes with sulfur are generally weaker than analogous interactions involving oxygen in Thf and nitrogen in Pyl (Platts et al, 1995). In addition, the effect of the delocalized electrons in the thiopene ring, as well as the larger sulfur atom precludes a definitive evaluation of the hydrophobic and electrostatic interactions within the pocket. As a commercially-available precursor

was not available for the synthesis of a sulfur-containing analog with a saturated ring (J.B. Fishman, 21st Century Biochemicals, Marlboro, MA, personal communication), a more elaborate scheme for synthesis will be required for future studies.

The organization of the cyclopentane ring of pyrrolysine in the hydrophobic pocket of PylS is immediately reminiscent of that of phenylalanine in the similar hydrophobic pocket of PheRS (Kavran et al, 2007). Amino acid binding to the interior surface of the pocket and the subsequent closing of the pocket by an aromatic residue in a mobile loop is common to both PylS and PheRS. However, a notable difference is that the amino acid in the flexible loop of PylS is Tyr as

141

opposed to Phe in PheRS. (Kavran et al, 2007). Another difference is the hydrogen bonding partner in position 2 of the pyrroline ring. No corresponding hydrogen bonding interaction has been observed in Phe-PheRS recognition and is unlikely to occur due to the lack of a partner in the benzene ring. Our results in this study indicate that hydrogen-bonded complex formation with the imine bond nitrogen is not essential as long as the hydrophobicity of the ring is maintained, since the cyclopentane ring-containing analog Cyp is a substrate of PylS.

However, under physiological conditions the hydrogen bonding interaction with a conserved tyrosine might enhance binding of pyrrolysine in the active site of

PylS, as Thf is demonstrated here as being a better substrate.

Additionally, the decrease in catalytic efficiency obtained with Thf with respect to Pyl may be due, in part, to the absence of the hydrophobic C4-methyl group which is lacking in the former. While not essential, complementarity between the C4-methyl group of Pyl and the interior of the hydrophobic pocket might be enhanced through Van der Waals forces that facilitate enhanced binding and catalysis. Additionally, differences in hydrogen bond strength in interactions involving the nitrogen of Pyl vis-à-vis the oxygen of Thf might explain the lower affinity of PylS for Thf. A final difference between the ring of Pyl and that of Thf is that former has a N2-C3 imine bond which is lacking in the latter.

Since aminoacyl-tRNA synthetases are evolved to recognize, activate and charge cognate amino acids with high specificity, aminoacylation is often the rate-limiting step for in vivo incorporation of non-cognate amino acids.

142

Interestingly, the incorporation of Cyp at a UAG in E. coli could not be detected at

1 mM concentration which is well below the Km value we obtained of 5.5 mM for

PylS catalyzed activation of this analog (Figure 5.5, Panel A). However, at a 10

mM concentration (~ 2-fold Km value), incorporation was detected (Figure 5.5,

Panel B). Similar to our observed data on aminoacylation in vitro, in physiological

conditions, aminoacylation with Cyp might not be as efficient as that with Thp,

resulting in cellular levels of tRNAPyl aminoacylated with it to be low.

The efficiency of site-specific incorporation of synthetic amino acids on an amber codon using an orthogonal tRNA/synthetase pair is approximately 20%

(Anderson and Schultz, 2003). Cyc and D-proline-L-lysine have previously been inserted into proteins at 18% and 25% efficiency, respectively (Polycarpo et al,

2006). In a direct comparison, we found that the efficiency of Cyc incorporation into MtmB is less than half that of Thf (Figure 5.6). The maximal efficiency of incorporation of Thf that we obtained is at least as high as that of Pyl in E. coli.

Thus, Thf is promising as a lead compound for introducing amino acids with novel functionalities since the synthesis of this compound is less painstaking than that of Pyl and functional groups can be added to the ring of Thf (M. Chan, personal communication). A recent study demonstrates that the use of an orthogonal tRNA/synthetase pair, as well as introduction of orthogonal ribosome

(with two specific nucleotide changes in 16S rRNA) along with a cognate mRNA increases the incorporation of unnatural amino acids at a single amber codon from 20% to 60% (Wang et al, 2007). Since the Pyl analog, Thf, can be inserted

143

into MtmB with high efficiency, studies to further increase the potential for incorporation using similar methods with an additional orthogonal ribosome/mRNA pair might yield a system that allows the incorporation of Thf derivatives at termination codons to approach the incorporation of the common, natural amino acids at sense codons.

144

Figure 5.1 Commercially available amino acids tested for tRNAPyl aminoacylation by PylS. Panel A shows the structures of pyrrolysine (1), L- lysine (2), D-proline (3), Nε-acetyl-L-lysine (4) and Nε-benzoyl-L-lysine (5). Panel B shows the results of a tRNAPyl specific northern blot of aminoacylation products obtained with 4 µg of unfractionated pool tRNA from M. acetivorans and 1.5 µM PylS. Reactions were incubated for 50 minutes at 37°C with the following amino acids: lane 1,no amino acid; lane 2, 50 µM pyrrolysine; lane 3, 1mM L-lysine; lane 4; 1 mM Nε-Acetyl-L-lysine; lane 5, 1 mM Nε-benzoyl-L-lysine, lanes 6-8 contain 50 µM pyrrolysine and in lane 6, 1 mM Nε-acetyl-L-lysine; lane 7, 1 mM Nε-benzoyl-L-lysine; lane 8, 1mM D-proline. Finally, lane 9 contains 1 mM of both L-lysine and D-proline.

145

Figure 5.2 Pyrrolysine analogs. The chemical structures of L-pyrrolysine (Pyl) (1) and analogs: 2-amino-6-(cyclopentanecarboxamino)hexanoic acid, (Cyp) (2); 2-amino-6-((R)-tetrahydrofuran-2-carboxamido)hexanoic acid, (Thf) (3); 2-amino- 6((R)-tetrahydrofuran-3-carboxamido)hexanoic acid, (OxR) (4); 2-amino-6-((S)- tetrahydrofuran-3-carboxamido)hexanoic acid, (OxS) (5); 2-amino-6-(thiopene-2- carboxamido)hexanoic acid, (Thp) (6); (S)-2-amino-6- (cyclopentoxycarbonylamino)hexanoic acid, (Cyc) (7). Compounds 1-6 were synthesized in this study. Compound 7 is a commercially-available analog that has previously been demonstrated to be aminoacylated to tRNAPyl by PylS at high concentrations (Polycarpo et al, 2006).

146

Continued

Figure 5.3 Activation of pyrrolysine and analogs by PylS. (Panel A): Open circles (○) represent complete reactions lacking amino acid; closed circles (●) supplemented with 100 µM Pyl and closed squares (■) with 2 mM Thf. (Panel B): Closed circles (●) represent complete reactions lacking amino acid; open squares (□) with 2 mM Thf and open circles (○) supplemented with 2 mM Cyp. (Panel C): Open circles (○) represent reactions lacking amino acid; closed squares (■) with 2 mM Thf; closed diamonds (♦) with 2 mM OxS; crosses (X) with 2 mM OxR; and closed triangles (▲) with 2 mM Thp.

147

Figure 1 continued

148

Acylated tRNAPyl

Deacylated tRNAPyl

Figure 5.4 Aminoacylation in vitro as determined by acid-urea gel electrophoresis and Northern blotting to detect tRNAPyl. Aminoacylation reactions were performed as described in Experimental Methods. Standard reactions were with 2 mM Thf (lane 1), 50 µM Pyl (lane 2), or no amino acid (lane 3). The lower band in lanes 1 and 2 corresponds to the uncharged tRNA band in lane 4.

149

Figure 5.5 Anti-MtmB immunoblot of extracts from E. coli. Anti-MtmB western blot of extracts from E. coli. Lanes 1-7 and 8-10 are from different experiments. Extracts in lanes 1-2, 4-5, and 9-10 are from an E. coli strain bearing the pECO3 plasmid with the pylT, pylS and mtmb1 genes, while extracts in 3 and 4 are from a strain with a plasmid with only the mtmb1 gene. Lanes 3 and 4 are from cultures supplemented with 1 mM Cyp, while lane 9 is from a culture containing 10 mM of Cyp; lane 2 with 1 mM Thf; and lane 3 and 5, 50 µM Pyl. Lanes 4, 6, and 10 were from cultures not supplemented with any exogenous pyrrolysine analog. Lanes 7 and 8 were loaded with the purified 50 kDa full-length MtmB control. The lower band in lanes 1-6 and 9-10 is a 23 kDa truncated product resulting from termination at the amber codon.

150

50 kDa

23 kDa

Figure 5.6 Comparison of incorporation of analogs in MtmB. Shown above is an Anti-MtmB immunoblot of extracts from E. coli. Extracts are from an E. coli strain bearing the pECO3 plasmid with the pylT, pylS and mtmb1 genes. Lane 1 was supplemented with 1 mM Thf and lane 2 with 5 mM Cyc. The upper band corresponds to the full-length 50 kDa product, while the lower band corresponds to the truncated 23 kDa product.

151

Figure 5.7 The active site of Methanosarcina mazei PylS. Pyrrolysine (Panel A) and the pyrrolysine analog Thf (Panel B) are modeled in the active site of M. mazei PylS (PDB Id: 2q7h), Based on x-ray crystallographic structures, the strictly conserved residues, Asn-346 and Tyr-384 are hypothesized to hydrogen bond with pyrrolysine to confer amino acid specificity (Kavran et al, 2007). Tyr- 384 is mobile in the structure of PylS unless hydrogen bound to enamine pyrrolysine (Kavran et al, 2007). Structures were visualized using PyMOL (DeLano, 2002).

152

Figure 5.8 The deduced active site of Methanosarcina barkeri PylS. (Panel A) Deduced M. barkeri MS PylS (GenBank: AAQ19545) structure superimposed on M. mazei PylS fragment (PDB Id: 2q7h; Kavran et al, 2007). The theoretical model energy was calculated to be -12461 kcal mol-1. The conserved Tyr and Asn residues proposed to confer specificity are indicated. The superimposed model was generated using Geno3D software (Combet et al, 2002). (Panel B) Active site of M. barkeri PylS generated using the same software. Structures were visualized using PyMOL (DeLano, 2002).

153

-1 -1 -1 Amino KM (µM) kcat (min ) kcat/KM (min µM ) Relative catalytic acid efficiency

Pyl 50 6 0.12 100%

Thf 380 4 0.01 8.3%

Cyc 550 0.8 0.001 0.8%

Cyp 5500 2 0.0004 0.3%

32 Table 5.1 Steady-state PPi-ATP exchange kinetics of M. barkeri PylS with pyrrolysine analogs.

154

LIST OF REFERENCES

Ambrogelly, A., Korencic, D., and M. Ibba. 2002. Functional annotation of class I lysyl-tRNA synthetase phylogeny indicates a limited role for gene transfer. J. Bacteriol. 184: 4594–4600.

Arnez, J. G., Augustine, J. G., Moras, D., and C.S. Francklyn. 1997. The first step of aminoacylation at the atomic level in histidyl-tRNA synthetase. Proc. Natl. Acad. Sci. U.S.A. 94:7144-7149.

Arnez, J. G. and D. Moras. 1997. Structural and functional considerations of the aminoacylation reaction. TIBS. 22: 211-216.

Ataide, S.F., and M. Ibba. 2006. Small molecules – big players in the evolution of protein synthesis. ACS Chem. Biol. 1:285-297.

Ataide, S.F., Jester, B. C., Devine, K. M., and M. Ibba. 2005. Stationary-phase expression and aminoacylation of a transfer-RNA-like small RNA. EMBO Rep. 6: 742-747.

Atkins, J.F. and R.F. Gesteland. 2002. Biochemistry: The 22nd amino acid. Science. 296:1409-1410.

Barns, S. M., Delwiche, C. F., Palmer, J. D., and N. R. Pace. 1996. Perspectives on archaeal diversity, thermophily and monophyly from environmental rRNA sequences. Proc. Natl. Acad. Sci. U.S.A. 93: 9188-9193.

Belrhali, H., Yaremchuk, A., Tukalo, M., Larsen, K., Berthet-Colominas, C., Leberman, R., Beijer, B., Sproat, B., Als-Nielsen, J., Grubel, G., et al. 1994. Crystal structures at 2.5 angstrom resolution of seryl-tRNA synthetase complexed with two analogs of seryl adenylate. Science. 263:1432-1436.

Berry, M. J. 2005. Knowing when not to stop. Nat. Struct. Mol. Biol. 12: 389-390.

155

Berry, M. J., Banu, L., Chen, Y. Y., Mandel, S. J., Kieffer, J. D., Harney, J. W., and P. R. Larsen. 1991. Recognition of UGA as a selenocysteine codon in type I deiodinase requires sequences in the 3' untranslated region. Nature. 6341: 273- 276.

Biou, V., Yaremchuk, A., Tukalo, M., and S. Cusack, 1994. The 2.9 A crystal structure of T. thermophilus seryl-tRNA synthetase complexed with tRNA(Ser). Science. 263:1404-1410.

Blaut, M. 1994. Metabolism of methanogens. Antonie van Leeuwenhoek. 66: 187-208.

Blight, S. K., Larue, R. C., Mahapatra, A., Longstaff, D. G., Chang, E., Zhao, G., Kang, P. T., Green-Church, K. B., Chan, M. K., and J. A. Krzycki. 2004. Direct charging of tRNACUA with pyrrolysine in vitro and in vivo. Nature. 431: 333-335.

Boccazzi, P., Zhang, J. K., and W. W. Metcalf. 2000. Generation of dominant selectable markers for resistance to pseudomonic acid by cloning and mutagenesis of the ileS gene from the Archaeon Methanosarcina barkeri Fusaro. J. Bacteriol. 182: 2611-2618.

Brierley, I., Digard, P. and S. C. Inglis. 1989. Characterization of an efficient coronavirus ribosomal frameshifting signal: requirement for an RNA pseudoknot. Cell. 57: 537-547.

Brown, C. M., Stockwell, P. A., Trotman, C. N., and W. P. Tate. 1990a. Sequence analysis suggests that tetra-nucleotides signal the termination of protein synthesis in eukaryotes. Nucleic Acids Res. 18: 6339-6345.

Brown, C. M., Stockwell, P. A., Trotman, C. N., and W. P. Tate. 1990b. The signal for the termination of protein synthesis in procaryotes. Nucleic Acids Res. 18: 2079-2086.

Buckingham, R. H. 1994. Codon context and protein synthesis: enhancements of the genetic code. Biochimie. 76: 351-354.

Bult, C.J., White, O., Olsen, G.J., Zhou, L., Fleischmann, R.D., Sutton, G.G., Blake, J.A., FitzGerald, L.M., Clayton, R.A., Gocayne, J.D., Kerlavage, A.R., Dougherty, B.A., Tomb, J.F., Adams, M.D., Reich, C.I., Overbeek, R., Kirkness, E.F., Weinstock, K.G., Merrick, J.M., Glodek, A., Scott, J.L., Geoghagen, N.S., and J.C. Venter. 1996. Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii. Science. 273:1058-1073.

156

Burke, S. A., and J. A. Krzycki. 1995. Involvement of the ‘A’ isozymes of methyltransferase II and the 29-kilodalton corrinoid protein in methanogenesis from methylamine. J. Bacteriol. 177: 4410-4416.

Burke, S. A., and J. A. Krzycki. 1997. Reconstitution of monomethylamine: coenzyme M methyl transfer with a corrinoid protein and two methyltransferases purified from Methanosarcina barkeri. J. Biol. Chem. 272: 16570-16577.

Burke, S. A., Lo, S. L., and J. A. Krzycki. 1998. Clustered genes encoding the methyltransferases of methanogenesis from monomethylamine. J. Bacteriol. 180: 3432-3440.

Böck, A., Thanbichler, M., Rother, M., and Resch, A. 2004 Selenocysteine. In The Aminoacyl-tRNA synthetases. Ibba, M., Francklyn, C. and Cusack, S. (eds): Landes Bioscience.

Cathopoulis, T., Chuawong, P., and T.L. Hendrickson. 2007. Novel tRNA aminoacylation mechanisms. Mol. Biosyst. 3: 408-418.

Cavarelli, J., Eriani, G., Rees, B., Ruff, M., Boeglin, M., Mitschler, A., Martin, F., Gangloff, J., Thierry, J. C., and D. Moras. 1994. The active site of yeast aspartyl- tRNA synthetase: structural and functional aspects of the aminoacylation reaction. EMBO J. 13: 327-337.

Chin, J. W., Cropp, T. A., Anderson, J. C., Mukherji, M., Zhang, Z. and P. G. Schultz. 2003. An expanded Eukaryotic genetic code. Science. 301: 964-967.

Cole, F. and P. R. Schimmel. 1970. On the rate law and mechanism of the adenosine triphosphate-pyrophosphate isotope exchange reaction of aminoacyl- transfer ribonucleic acid synthetases. Biochemistry. 9: 480–489.

Combet, C., Jambon, M., Deléage, G., and C. Geourjon. 2002. Geno3D: Automatic comparative molecular modelling of protein. 18: 213-214

Commans, S. and A. Böck. 1999. Selenocysteine inserting tRNAs: an overview. FEMS Microbiol. Rev. 23: 335-351.

Conrad, R. 1996. Soil microorganisms as controllers of atmospheric trace gases (H2, CO, CH4, OCS, N2O and NO). Microbiol. Rev. 60: 609-640.

Copeland, P. 2003. Regulation of gene expression by stop codon recoding: selenocysteine. Gene. 312: 17-25.

157

Copeland, P. R., Fletcher, J. E., Carlson, B. A., Hatfield, D. L., and D. M. Driscoll. 2000. A novel RNA binding protein, SBP2, is required for the translation of mammalian selenoprotein mRNAs. EMBO J. 19: 306-314.

Craigen, W. J., Cook, R. G., Tate, W. P., and C. T. Caskey. 1985. Bacterial peptide chain release factors: conserved primary structure and possible frameshift regulation of release factor 2. Proc. Natl. Acad. Sci. U.S.A. 82: 3616- 3620.

Crick, F. H. C. 1958. On protein synthesis. Symp. Soc. Exp. Biol. 12: 138-163.

Cusack, S. 1997. Aminoacyl-tRNA synthetases. Curr. Opin. Struct. Biol. 7: 881- 889.

Cusack, S., Berthet-Colominas, C., Hartlein, M., Nassar, N., R. Leberman. 1990. A second class of synthetase structure revealed by X-ray analysis of Escherichia coli seryl-tRNA synthetase at 2.5 A. Nature. 347: 249-255.

Cusack, S., Yaremchuk, A., and M. Tukalo. 1996. The crystal structure of the ternary complex of T. thermophilus seryl-tRNA synthetase with tRNA(Ser) and a seryl-adenylate analogue reveals a conformational switch in the active site. EMBO J. 15: 2834-2842.

DeLano, W. L. 2002. The PyMOL Molecular Graphics System. http://www.pymol.org.

Delarue, M. 2007. An asymmetric underlying rule in the assignment of codons: Possible clue to a quick early evolution of the genetic code via successive binary choices. RNA 13: 161-169.

Deppenmeier, U. 2002. The unique biochemistry of methanogenesis. Prog. Nucleic Acid Res. Mol. Biol. 71: 223-283.

Deppenmeier, U., Johann, A., Hartsch, T., Merkl, R., Schmitz, R. A., Martinez- Arias, R., Henne, A., Weizer, A., Baumer, S., Jacobi, C., Bruggemann, H., Lienard, T., Christmann, A., Bomeke, M., Steckel, S., Bhattacharyya, A., Lykidis, A., Overbeek, R., Klenk, H. P., Gunsalus, R. P., Fritz, H. J., and G. Gottschalk. 2002. The genome of Methanosarcina mazei: evidence for lateral gene transfer between bacteria and archaea. J. Mol. Microbiol. Biotechnol. 4: 453-461.

Deppenmeier, U. 2004. The membrane-bound electron transport system of Methanosarcina species. J. Bioenerg. Biomembr. 36: 55-64.

158

Desogus, G., Todone, F., Brick, P., and S. Onesti. 2000. Active site of lysyl-tRNA synthetase: structural studies of the adenylation reaction. Biochemistry. 39: 8418-8425.

Diekert, G. and G. Wohlfarth. 1994. Metabolism of homocetogens. Antonie Van Leeuwenhoek. 66: 209-221.

DiMarco, A. A., Bobik, T. A., and R. S. Wolfe. 1990. Unusual coenzymes of methanogenesis. Annu. Rev. Biochem. 59:355-94.

Dontsova, M., Frolova, L., Vassilieva, J., Piendl, W., Kisselev, L., and M. Garber. 2000. Translation termination factor aRF1 from the archaeon Methanococcus jannaschii is active with eukaryotic ribosomes. FEBS Lett. 472: 213-216.

Döring, V., Mootz, H. D., Nangle, L. A., Hendrickson, T. L., de Crécy-Lagard, V., Schimmel, P., and P. Marlière. 2001. Enlarging the amino acid set of Escherichia coli by infiltration of the valine coding pathway. Science. 292: 501-504.

Drake, H. L., Daniel, S. L., Kusel., K., Matthies., C., Kuhner, C., and S. Braus- Stromeyer. 1997. Acetogenic bacteria: what are the in situ consequences of their diverse metabolic versatilities? Biofactors. 6: 13-24.

Driscoll, D. M. and P. R. Copeland. 2003. Mechanism and regulation of selenoprotein synthesis. Annu. Rev. Nutr. 23: 17-40.

Eckburg, P. B., Bik, E. M., Bernstein, C. N., Purdom, E., Dethlefsen, L., Sargent, M., Gill, S. R., Nelson, K. E., and D.A. Relman. 2005. Diversity of the human intestinal microbial flora. Science. 308: 1635-8.

Eldred, E.W., and P. Schimmel. 1972. Investigation of the transfer of amino acid from a transfer ribonucleic acid synthetase-aminoacyl adenylate complex to transfer ribonucleic acid. Biochemistry. 11:17-23.

Eriani, G., Delarue, M., Poch, O., Gangloff, J., and D. Moras. 1990. Partition of tRNA synthetases into two classes based on mutually exclusive sets of sequence motifs. Nature. 347: 203-206.

Ermler, U., Grabarse W., Shima, S., Goubeaud, M., and R.K. Thauer. 1997. Crystal Structure of Methyl-Coenzyme M Reductase: the Key Enzyme of Biological Methane Formation. Science. 278: 1457-1462.

Fagegaltier, D., Hubert, N., Yamada, K., Mizutani, T., Carbon, P., and A. Krol. 2000. Characterization of mSelB, a novel mammalian elongation factor for selenoprotein elongation. EMBO J. 19: 4796-4805.

159

Ferguson, D. J., Jr., and J. A. Krzycki. 1997. Reconstitution of trimethylamine- dependent coenzyme M methylation with the trimethylamine corrinoid protein and the isozymes of methyltransferase II from Methanosarcina barkeri. J. Bacteriol. 179: 846-852.

Ferguson, D. J., Jr., Krzycki, J. A., and D. A. Grahame. 1996. Specific roles of methylcobamide: coenzyme M methyltransferase isozymes in metabolism of methanol and methylamines in Methanosarcina barkeri. J. Biol. Chem. 271: 5189-5194.

Ferguson, D. J., Jr., Gorlatova, N., Grahame, D. A., and J. A. Krzycki. 2000. Reconstitution of dimethylamine: coenzyme M methyl transfer with a discrete corrinoid protein and two methyltransferases purified from Methanosarcina barkeri. J. Biol. Chem. 275: 29053-29060.

Ferry, J. G. 1992. Methane from acetate. J. Bacteriol. 174: 5489-5495.

Ferry, J. G. 1999. Enzymology of one-carbon metabolism in methanogenic pathways. FEMS Microbiol. Rev. 23: 13-38.

Fersht, A. R., 1985. Enzyme structure and mechanism. (Freeman, San Francisco, CA).

Fersht, A. R., Wilkinson, A. J., Carter, P., and G. Winter. 1985. Fine structure- activity analysis of mutations at position 51 of tyrosyl-tRNA synthetase. Biochemistry. 24:5858-5861.

Fersht, A. R. 1979. In Transfer RNA: structure, properties and recognition. P. R. Schimmel, D. Söll, J. N. Abelson, (eds.). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 247-254.

Fersht, A. R. 1998. Protein structure: sieves in sequence. Science. 280: 541.

First, E.A. 2005. Catalysis of the tRNA Aminoacylation Reaction. In The Aminoacyl-tRNA Synthetases. Edited by Ibba, M., Francklyn, C., and S. Cusack. Landes Bioscience. 328-350.

Förster, C., Ott, G., Forchhammer, K., and M. Sprinzl. 1990. Interaction of a selenocysteine-incorporating tRNA with elongation factor Tu from E. coli. Nucleic Acids Res. 18: 487-491.

Forchhammer, K., Leinfelder, W., and A. Böck. 1989. Identification of a novel translation factor necessary for the incorporation of selenocysteine into protein. Nature. 342: 453-456.

160

Forchhammer, K., Leinfelder, W., Boesmiller, K., Veprek, B., and A. Böck. 1991. Selenocysteine synthase from Escherichia coli: nucleotide sequence of the gene (selA) and purification of the protein. J. Biol. Chem. 266: 6318-6323.

Forchhammer, K., and A. Böck. 1991. Selenocysteine synthase from Escherichia coli. Analysis of the reaction sequence. J. Biol. Chem. 266: 6324-6328.

Francklyn, C., Perona, J. J., Puetz, J. and Y.-M. Hou. 2002. Aminoacyl-tRNA synthetases: versatile players in the changing theater of translation. RNA. 8: 1363–1372.

Fraser, C.M., Norris, S.J., Weinstock, G.M., White, O., Sutton, G.G., Dodson, R., Gwinn, M., Hickey, E.K., Clayton, R., Ketchum, K.A., Sodergren, E., Hardham, J.M., McLeod, M.P., Salzberg, S., Peterson, J., Khalak, H., Richardson, D., Howell, J.K., Chidambaram, M., Utterback, T., McDonald, L., Artiach, P., Bowman, C., Cotton, M.D., Fujii, C., Garland, S., Hatch, B., Horst, K., Roberts, K., Sandusky, M., Weidman, J., Smith, H.O., and J.C. Venter. 1998. Complete genome sequence of Treponema pallidum, the syphilis spirochete. Science 281: 375-388.

Freistroffer, D. V., Pavlov, M. Y., MacDougall, J., Buckingham, R. H., and M. Ehrenberg. 1997. Release factor RF3 in E. coli accelerates the dissociation of release factors RF1 and RF2 from the ribosome in a GTP-dependent manner. EMBO J. 16: 4126-4133.

Fricke, W. Seedorf, H., Henne, A., Kruer, M., Liesegang, H., Hedderich, R., Gottschalk, G., and R.K. Thauer. 2006. The genome sequence of Methanosphaera stadtmanae reveals why this human intestinal archaeon is restricted to methanol and H2 for methane formation and ATP synthesis. J. Bacteriol. 188: 642-58.

Frolova, L., Le Goff, X., Rasmussen, H. H., Cheperegin, S., Drugeon, G., Kress, M., Arman, I., Haenni, A. L., Celis, J. E., Philippe, M., Justesen, J., and L. Kisselev. 1994. A highly conserved eukaryotic protein family possessing properties of polypeptide chain release factor. Nature. 372: 701-703.

Fu, C., and J. Parker. 1994. A ribosomal frameshifting error during translation of the argI mRNA of Escherichia coli. Mol. Gen. Genet. 243: 434-441.

Galagan, J. E., Nusbaum, C., Roy, A., Endrizzi, M. G., Macdonald, P., FitzHugh, W., Calvo, S., Engels, R., Smirnov, S., Atnoor, D., Brown, A., Allen, N., Naylor, J., Stange-Thomann, N., DeArellano, K., Johnson, R., Linton, L., McEwan, P., McKernan, K., Talamas, J., Tirrell, A., Ye, W., Zimmer, A., Barber, R. D., Cann, I., Graham, D. E., Grahame, D. A., Guss, A. M., Hedderich, R., Ingram-Smith, C., Kuettner, H. C., Krzycki, J. A., Leigh, J. A., Li, W., Liu, J., Mukhopadhyay, B., 161

Reeve, J. N., Smith, K., Springer, T. A., Umayam, L. A., White, O., White, R. H., Conway de Macario, E., Ferry, J. G., Jarrell, K. F., Jing, H., Macario, A. J. L., Paulsen, I., Pritchett, M., Sowers, K. R., Swanson, R. V., Zinder, S. H., Lander, E., Metcalf, W. W., and B. Birren. 2002. The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 12: 532-542.

Gangloff J., and G. Dirheimer. 1973. Studies on aspartyl-tRNA synthetase from Baker's yeast. I. Purification and properties of the enzyme. Biochim. Biophys. Acta. 294:263-272.

Garcia, J.-L., Patel, B. K. C., and B. Ollivier. 2000. Taxonomic, phylogenetic, and ecological diversity of methanogenic Archaea. Anaerobe. 6: 205-226.

Gernhardt, P., Possot, O., Foglino, M., Sibold, L., and A. Klein. 1990. Construction of an integration vector for use in the archaebacterium Methanococcus voltae and expression of a eubacterial resistance gene. Mol. Gen. Genet. 221:273–279.

Gesteland, R. F. and J. F. Atkins. 1996. RECODING: Dynamic reprogramming of translation. Annu. Rev. Biochem. 65: 741-768.

Giege, R., Sissler, M., and C. Florentz. 1998. Universal rules and idiosyncratic features in tRNA identity. Nucleic Acids Res. 22:5017-5035. . Goodchild, A., Saunders, N. F. W., Ertan, H., Raferty, M., Guilhaus, M., Curmi, P. M. G., and R. Cavicchioli. 2004. A proteomic determination of cold adaptation in the Anarctic archaeon, Methanococcoides burtonii. Mol. Microbiol. 53: 309-321.

Grahame, D. A. 1989. Different isozymes of methylcobalamin: 2- mercaptoethanesulfonate methyltransferase predominate in methanol- versus acetate-grown Methanosarcina barkeri. J. Biol. Chem. 264: 12890-12894.

Grentzmann, G., Brechemier-Baey, D., Heurgue, V., Mora, L., and R. H. Buckingham. 1994. Localization and characterization of the gene encoding release factor RF3 in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 91: 5848- 5852.

Gribaldo, S. and C. Brochier-Armanet. 2006. The origin and evolution of Archaea: a state of the art. Philos Trans R Soc Lond B Biol Sci 361: 1007-22.

Hao, B., Gong, W., Ferguson, T. K., James, C. M., Krzycki, J. A., and M. K. Chan. 2002. A new UAG-Encoded residue in the structure of a Methanogen methyltransferase. Science. 296: 1462-1466.

Hao, B., Zhao, G., Kang, P. T., Soares, J. A., Ferguson, T. K., Gallucci, J., 162

Krzycki, J. A., and M. K. Chan. 2004. Reactivity and chemical synthesis of L- pyrrolysine – the 22nd amino acid. Chem. Biol. 9: 1317-24.

Harms, U. and R. K. Thauer. 1996. Methylcobalamin: coenzyme M methyltransferase isoenzymes MtaA and MtbA from Methanosarcina barkeri. Cloning, sequencing and differential transcription of the encoding genes, and functional overexpression of the mtaA gene in Escherichia coli. Eur. J. Biochem. 235: 653-659.

Hatfield, D. L., and V. N. Gladyshev. 2002. How selenium has altered our understanding of the genetic code. Mol. Cell Biol. 11: 3565-3576.

Heider, J., Baron, C., and A. Böck. 1992. Coding from a distance: dissection of the mRNA determinants required for the incorporation of selenocysteine into protein. EMBO J. 11: 3759-3766.

Hendrickson, T. L.. de Crecy-Lagard, V., P. Schimmel. 2004. Incorporation of nonnatural amino acids into proteins. Annu. Rev. Biochem. 73:147-176.

Herr, A. J., Atkins, J. F. and R. F. Gesteland. 2000. Coupling of open reading frames by translational bypassing. Annu. Rev. Biochem. 69: 343-372.

Herring, S., A. Ambrogelly, Polycarpo, C. R. and D Soll. 2007. Recognition of pyrrolysine tRNA by the Desulfitobacterium hafniense pyrrolysyl-tRNA synthetase. Nucleic Acids Res 35: 1270-8.

Ho, Y. S. and Y. W. Kan. 1987. In vivo aminoacylation of human and Xenopus suppressor tRNAs constructed by site-specific mutagenesis. Proc. Natl. Acad. Sci. U.S.A. 84: 2185–2188.

Hohsaka, T., and M. Sisido. 2002. Incorporation of non-natural amino acids into proteins. Curr. Opin. Chem. Biol. 6: 809-815.

Hoffmann, P. R. and M. J. Berry. 2005. Selenoprotein synthesis: A unique translational mechanism used by a diverse family of proteins. Thyroid. 15: 769- 775.

Huber, H., Hohn M.J., Rachel R., Fuchs, T., Wimmer, V.C., and K.O. Stetter. 2002. A new phylum of Archaea represented by a nanosized hyperthermophilic symbiont. Nature 417: 63-7.

Huttenhofer, A. and A. Böck. 1998. Selenocysteine inserting RNA elements modulate GTP hydrolysis of elongation factor SelB. Biochemistry (Mosc). 37: 885-890. 163

Ibba, M., Morgan, S., Curnow, A. W., Pridmore, D. R., Vothknecht, U. C., Gardner, W., Lin, W., Woese, C. R., and D. Söll. 1997. A euryarchaeal lysyl- tRNA synthetase: resemblance to class I synthetases. Science. 278: 1119–1122.

Ibba, M., Losey, H. C., Kawarabayasi, Y., Kikuchi, H., Bunjun, S. and D. Söll. 1999. Substrate recognition by class I lysyl-tRNA synthetases: a molecular basis for gene displacement. Proc. Natl. Acad. Sci. U.S.A. 96: 418-423.

Ibba, M. and D. Söll. 2000. Aminoacyl-tRNA synthesis. Annu. Rev. Biochem. 69: 617-650.

Ibba, M. and D. Söll. 2004. Aminoacyl-tRNAs: setting the limits of the genetic code. Genes Dev. 18: 731-738.

Ito, K., Uno, M., and Y. Nakamura. 2000. A tripeptide 'anticodon' deciphers stop codons in messenger RNA. Nature. 403: 680-684.

Ivanova, N., Sorokin, A., Anderson, I., Galleron, N., Candelon, B., Kapatral, V., Bhattacharyya, A., Reznik, G., Mikhailova, N., Lapidus, A., Chu, L., Mazur, M., Goltsman, E., Larsen, N., D'Souza, M., Walunas, T., Grechkin, Y., Pusch, G., Haselkorn, R., Fonstein, M., Ehrlich, S.D., Overbeek, R., and N. Kyrpides, 2003 Genome sequence of Bacillus cereus and comparative analysis with Bacillus anthracis. Nature 423: 87-91.

Jakubowski, H. 1995. Proofreading in vivo. Editing of homocysteine by aminoacyl-tRNA synthetases in Escherichia coli. J. Biol. Chem. 270:17672- 17673.

James, C. M., Ferguson, T. K., Leykam, J. F., and J. A. Krzycki. 2001. The amber codon in the gene encoding the monomethylamine methyltransferase isolated from Methanosarcina barkeri is translated as a sense codon. J. Biol. Chem. 276: 34252-34258.

Jester, B. C., Levengood, J. D., Roy, H., Ibba, M., and K. M. Devine. 2003. Nonorthologous replacement of lysyl-tRNA synthetase prevents addition of lysine analogues to the genetic code. Proc. Natl. Acad. Sci. U.S.A. 100: 14351-14356.

Jones, W. J., Nagle, D. P., and W. B. Whitman. 1987. Methanogens and the diversity of archaebacteria. Microbiol. Rev. 51: 135-177.

Kamtekar, S., Hohn, M. J., Park, H. S., Schnitzbauer, M., Sauerwald, A., Söll, D., and T.A. Steitz. 2007. Toward understanding phosphoseryl-tRNACys formation: the crystal structure of Methanococcus maripaludis phosphoseryl-tRNA synthetase. Proc. Natl. Acad. Sci. U.S.A. 104:2620-2625.

164

Kapp, L. D. and J. R. Lorsch. 2004. The molecular mechanics of eukaryotic translation. Annu. Rev. Biochem. 73: 657-704.

Kavran, J. M., Gundllapalli, S., O'Donoghue, P., Englert, M., Söll, D., and T.A. Steitz. 2007. Structure of pyrrolysyl-tRNA synthetase, an archaeal enzyme for genetic code innovation. Proc. Natl. Acad. Sci. U.S.A. 104: 11268-11273.

Khalil, M. A. K. and R. A. Rasmussen. 1994. Global emissions of methane during the last several centuries. Chemosphere. 29: 833-842.

Kiick, K. L.,Weberskirch, R. and D. A. Tirrell. 2001. Identification of an expanded set of translationally active methionine analogues in Escherichia coli. FEBS Lett. 502: 25–30.

Kisselev, L. L. and R. H. Buckingham. 2000. Translational termination comes of age. TIBS. 25: 561-566.

Klotz, M.G., Arp, D.J., Chain, P.S., El-Sheikh, A.F., Hauser, L.J., Hommes, N.G., Larimer, F.W., Malfatti, S.A., Norton, J.M., Poret-Peterson, A.T., Vergez, L.M., and B.B. Ward. 2006 Complete genome sequence of the marine, chemolithoautotrophic, ammonia-oxidizing bacterium Nitrosococcus oceani ATCC 19707. Appl. Environ. Microbiol. 72: 6299-6315.

Konecki, D. S., Aune, K. C., Tate, W., and C. T. Caskey. 1977. Release factor binding to ribosome requires an intact 16 S rRNA 3' terminus. J. Biol. Chem. 252: 4514-4520.

Krzycki, J. A. 2004. Function of genetically encoded pyrrolysine in corrinoid- dependent methylamine methyltransferases. Curr. Opin. Chem. Biol. 8: 484-491.

Krzycki, J. A. 2005. The direct genetic encoding of pyrrolysine. Curr. Opin. Microbiol. 8: 706-712.

Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227: 680-685.

LeClerc, G. M., and D. A. Grahame. 1996. Methylcobamide:coenzyme M methyltransferase isozymes from Methanosarcina barkeri. Physicochemical characterization, cloning, sequence analysis, and heterologous gene expression. J. Biol. Chem. 271: 18725-31.

Lecompte, O., Ripp, R., Thierry, J. C., Moras, D., and O. Poch. 2002. Comparative analysis of ribosomal proteins in complete genomes: an example of reductive evolution at the domain scale. Nucleic Acids Res. 30: 5382-5390.

165

Levengood, J., Ataide, S.F., Roy, H., and M Ibba. 2004. Divergence in noncognate amino acid recognition between class I and class II lysyl-tRNA synthetases. J. Biol. Chem. 279: 17707-17714.

Li, Q., Li, L., Rejtar, T, Kargar, B.L., and J.G. Ferry. 2005. Proteome of Methanosarcina acetivorans Part I: an expanded view of the biology of the cell. J. Proteome Res. 4:112-28.

Longstaff, D.G., Blight, S.K., Zhang, L., Green-Church, K.B., and J.A. Krzycki. 2007a. In vivo contextual requirements for UAG translation as pyrrolysine. Mol. Microbiol. 63: 229-241.

Longstaff, D.G., Larue, R.C., Faust, J.E., Mahapatra, A., Zhang, L., Green- Church, K.B., and J.A. Krzycki. 2007b. A natural genetic code expansion cassette enables transmissible biosynthesis and genetic encoding of pyrrolysine. Proc. Natl. Acad. Sci. U.S.A. 104: 1021-1026.

Maeder, D.L., Anderson, I., Brettin, T.S., Bruce, D.C., Gilna, P., Han, C.S., Lapidus, A., Metcalf, W.W., Saunders, E., Tapia, R., and K.R. Sowers. 2006. The Methanosarcina barkeri genome: comparative analysis with Methanosarcina acetivorans and Methanosarcina mazei reveals extensive rearrangement within methanosarcinal genomes. J Bacteriol 188: 7922-7931

Mahapatra A., and J.A. Krzycki. 2007. The Genetic Code In: McGraw-Hill Yearbook of Science & Technology, McGraw-Hill, New York. 93-98.

Mahapatra, A., Patel, A., Soares, J. A., Larue, R. C., Zhang, J.-K., Metcalf, W. W. and J. A. Krzycki. 2006. Characterization of a Methanosarcina acetivorans mutant unable to translate UAG as pyrrolysine. Mol. Microbiol. 59: 56-66.

Mahapatra, A., Srinivasan, G., Richter, K. B., Meyer, A., Lienard, T., Zhang, J-K., Zhao, G., Kang, P. T., Chan, M., Gottschalk, G., Metcalf, W. W., and J.A. Krzycki. 2007. Class I and class II lysyl-tRNA synthetase mutants and the genetic encoding of pyrrolysine in Methanosarcina spp. Mol. Microbiol. 64: 1306-1318.

Maupin-Furlow, J.A., and J.G. Ferry. 1996. Analysis of the CO Dehydrogenase/acetyl-coenzyme A synthase operon of Methanosarcina thermophila. J. Bacteriol. 178: 6849–6856.

McCaughan K. K., Brown, C. M., Dalphin, M. E., Berry, M. J., and W. P. Tate. 1995. Translational termination efficiency in mammals is influenced by the base following the stop codon. Proc. Natl. Acad. Sci. U.S.A. 92: 5431-5435.

McClain, W.H. 1993. Rules that govern tRNA identity in protein synthesis. J. Mol. Biol. 234:257-280. 166

van der Meijden, P., Heythuysen, H. J., Pouwels, A., Houwen, F., van der Drift, C., and G. D. Vogels. 1983. Methyltransferases involved in methanol conversion by Methanosarcina barkeri. Arch. Microbiol. 134: 238-242. van der Meijden, P., te Brömmelstroet, B. W., Poirot, C. M., van der Drift, C., and G. D. Vogels. 1984. Purification and properties of methanol: 5- hydroxybenzimidazolylcobamide methyltransferase from Methanosarcina barkeri. J. Bacteriol. 160: 629-635.

Metcalf, W. W., Zhang, J. K., Apolinario, E., Sowers, K. R., and R. S. Wolfe. 1997. A genetic system for Archaea of the genus Methanosarcina: liposome- mediated transformation and construction of shuttle vectors. Proc. Natl. Acad. Sci. U.S.A. 6: 2626-2631.

Mikuni, O., Ito, K., Moffat, J., Matsumura, K., McCaughan, K., Nobukini, T., Tate, W., and Y. Nakamura. 1994. Identification of the prfC gene, which encodes peptide-chain-release factor 3 of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 91: 5798-5802.

Moor, N., Nazarenko, I., Ankilova, V., Khodyreva, S., and O. Lavrik. 1990. Determination of tRNA(Phe) recognition nucleotides for phenylalanyl-tRNA synthetase from Thermus thermophilus. Biochimie. 74:353-356.

Mosyak, L., Reshetnikova, L., Goldgur, Y., Delarue, M., and M.G. Safro. 1995. Structure of phenylalanyl-tRNA synthetase from Thermus thermophilus. Nat. Struct. Biol. 2:537-547.

Murgola. E.J. 1985. tRNA suppression and the code. Annu. Rev. Genet. 19:57- 80

Nakamura, Y. and K. Ito. 2003. Making sense of mimic in translation termination. Trends Biochem. Sci. 28: 99-105.

Namy, O., Rousset, J.-P., Napthine, S., and I. Brierley. 2004. Reprogrammed genetic decoding in cellular gene expression. Mol. Cell. 13: 157-168.

Nissen, P., Hansen, J., Ban, N., Moore, P. B., and T.A. Steitz. 2000. The structural basis of ribosome activity in peptide bond synthesis. Science. 289: 920-30.

Nonaka, H., Keresztes, G., Shinoda, Y., Ikenaga, Y., Abe, M., Naito, K., Inatomi, K., Furukawa, K., Inui, M., and H. Yukawa. 2006. Complete genome sequence of the dehalorespiring bacterium Desulfitobacterium hafniense Y51 and comparison with Dehalococcoides ethenogenes 195. J. Bacteriol. 188: 2262-2274. 167

Nureki, O., Kohno, T., Sakamoto, K., Miyazawa, T., and S.Yokoyama. 1993. Chemical modification and mutagenesis studies on zinc binding of aminoacyl- tRNA synthetases. J. Biol. Chem. 268:15368-15373.

Onesti, S., Desogus, G., Brevet, A., Chen, J., Plateau, P., Blanquet, S., and P. Brick. 2000. Structural studies of lysyl-tRNA synthetase: conformational changes induced by substrate binding. Biochemistry. 39:12853-12861.

Onesti, S. Miller, A. D., and P. Brick. 1995. The crystal structure of the lysyl-tRNA synthetase (LysU) from Escherichia coli. Structure. 3:163-176.

Ozawa, Y., Hanaoka, S., Saito, R., Washio, T., Nakano, S., Shinagawa, A., Itoh, M., Shibata, K., Carninci, P., Konno, H., Kawai, J., Hayashizaki, Y., and M. Tomita. 2002. Comprehensive sequence analysis of translation termination sites in various eukaryotes. Gene. 300: 79-87.

Paul, L., Ferguson, D. J., Jr., and J. A. Krzycki. 2000. The trimethylamine methyltransferase gene and multiple dimethylamine methyltransferase genes of Methanosarcina barkeri contain in-frame and read-through amber codons. J. Bacteriol. 182: 2520-2529.

Philipson, L., Andersson, P., Olshevsky, U., Weinberg, R., Baltimore, D., and R. Gesteland. 1978. Translation of MuLV and MSV RNAs in nuclease-treated reticulocyte extracts: enhancement of the gag-pol polypeptide with yeast suppressor tRNA. Cell. 13: 188-199.

Platts J.A., Howard, S.T., and B.R.F. Bracke. 1995. Directionality of hydrogen bonds to sulfur and oxygen. J. Am. Chem. Soc. 118:2726-2733

Polycarpo, C., Ambrogelly, A., Ruan, B., Tumbula-Hansen, D., Ataide, S. F., Ishitani, R., Yokoyama, S., Nureki, O., Ibba, M., and D. Söll. 2003. Activation of the pyrrolysine suppressor tRNA requires formation of a ternary complex with class I and class II lysyl-tRNA synthetases. Mol. Cell. 12: 287-294.

Polycarpo, C., Ambrogelly, A., Bérubé, A., Winbush, S. M., McCloskey, J. A., Crain, P. F., Wood, J. L., and D. Söll. 2004. An aminoacyl-tRNA synthetase that specifically activates pyrrolysine. Proc. Natl. Acad. Sci. U.S.A. 101: 12450– 12454.

Polycarpo, C. R., S. Herring, Bérubé, A., Wood, J. L., Söll D., and A. Ambrogelly. 2006. Pyrrolysine analogues as substrates for pyrrolysyl-tRNA synthetase. FEBS Lett 580: 6695-700.

168

Poole, E. S., Brown, C. M., and W. P. Tate. 1995. The identity of the base following the stop codon determines the efficiency of in vivo translational termination in Escherichia coli. EMBO J. 14: 151-158.

Poole, E. S., Major, L. L., Mannering, S. A., and W. P. Tate. 1998. Translational termination in Escherichia coli: three bases following the stop codon crosslink to release factor 2 and affect the decoding efficiency of UGA-containing signals. Nucleic Acids Res. 26: 954-960.

Pottenplackel, L. P. 1999. Analysis of the genes encoding the enzymes initiating methanogenesis from methylthiols, trimethylamine and dimethylamine in Methanosarcina barkeri MS. Ph.D. Thesis, The Ohio State University. 153-155.

Prætorius-Ibba, M. and M. Ibba. 2003. Aminoacyl-tRNA synthesis in archaea: different but not unique. Mol. Microbiol. 48: 631-637.

Pritchett, M.A., and W.W. Metcalf. 2005. Genetic, physiological and biochemical characterization of multiple methanol methyltransferase isozymes in Methanosarcina acetivorans C2A. Mol. Microbiol. 56: 1183–1194.

Pritchett, M. A., Zhang, J. K., and W. W. Metcalf. 2004. Development of a markerless genetic exchange method for Methanosarcina acetivorans C2A and its use in construction of new genetic tools for methanogenic archaea. Appl Environ Microbiol. 70: 1425-1433.

Pure, G. A. Robinson, G. W. Naumovski, L., and E.C. Friedberg. 1985. Partial suppression of an ochre mutation in Saccharomyces cerevisiae by multicopy plasmids containing a normal yeast tRNAGln gene. J. Mol. Biol. 183:31-42

Ramakrisnan, V. 2002. Ribosome structure and the mechanism of translation. Cell. 108: 557-572.

Reed D.W., Fujita, Y., Delwiche, M.E., Blackwelder, D.B., Sheirdan, P.P, Uchida, T., and F.S. Colwell. 2002. Microbial communities from methane hydrate-bearing deep marine sediments in a forearc basin. Appl. Environ. Microbiol. 68: 3759- 3770

Rother, M. and W. W. Metcalf. 2004. Anaerobic growth of Methanosarcina acetivorans C2A on carbon monoxide: An unusual way of life for a methanogenic archaeon. Proc. Natl. Acad. Sci. U.S.A. 101: 16929-16934.

Rother, M. and W. W. Metcalf. 2005. Genetic technologies for Archaea. Curr. Opin. Microbiol. 8 :745–751.

169

Rother, M., Mathes, I., Lottspeich, F., and A. Böck, A. 2003. Inactivation of the selB gene in Methanococcus maripaludis: effect on synthesis of selenoproteins and their sulfur-containing homologs. J. Bacteriol. 185: 107–114.

Rother, M., Resch, A., Gardner, W. L., Whitman, W. B., and A. Böck. 2001. Heterologous expression of archaeal selenoprotein genes directed by the SECIS element located in the 3’ non-translated region. Mol. Microbiol. 40: 900-908.

Sambrook, J., Fritsch, E. F., and T. Maniatis. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Samuel, B. S., Hansen, E. E., Manchester, J. K., Coutinho, P. M., Henrissat, B., Fulton, R., Latreille, P., Kim, K., Wilson, R. K., J.I. Gordon. 2007. Genomic and metabolic adaptations of Methanobrevibacter smithii to the human gut. Proc. Natl. Acad. Sci. U.S.A. 104: 10643-8.

Samuel, B. S. and J. I. Gordon. 2006. A humanized gnotobiotic mouse model of host-archaeal-bacterial mutualism. Proc. Natl. Acad. Sci. U.S.A. 103: 10011-6.

Sandbeck, K.A., and J.A. Leigh, J.A. 1991. Recovery of an integration shuttle vector from tandem repeats in Methanococcus maripaludis. Appl. Environ. Microbiol. 57: 2762–2763.

Sauer, K., and R. K. Thauer. 1998. Methanol: coenzyme M methyltransferase from Methanosarcina barkeri--identification of the active-site histidine in the corrinoid-harboring subunit MtaC by site-directed mutagenesis. Eur. J. Biochem. 253: 698-705.

Sauer, K., Harms, U., and R. K. Thauer. 1997. Methanol: coenzyme M methyltransferase from Methanosarcina barkeri. Purification, properties and encoding genes of the corrinoid protein MT1. Eur. J. Biochem. 243: 670-677.

Sauerwald, A., Zhu, W., Major, T. A., Roy, H., Palioura, S., Jahn, D., Whitman, W. B., Yates, 3rd, J. R., Ibba, M., and D. Söll. 2005. Science. RNA-dependent cysteine biosynthesis in archaea. 307: 1969-1972.

Saunders, N. F., Thomas, T., Curmi, P. M., Mattick, J. S., Kuczek, E., Slade, R., Davis, J., Fransmann, P. D., Boone, D., Rusterholtz, K., Feldman, R., Gates, C., Bench, S., Sowers, K., Kadner, K., Aerts, A., Dehal, P., Detter, C., Glavina, T., Lucas, S., Richardson, P., Larimer, F., Hauser, L., Land, M., and R. Cavicchioli. 2003. Mechanisms of thermal adaptation revealed from the genomes of the Antarctic Archaea Methanogenium frigidum and Methanococcoides burtonii. Genome Res. 13: 1580-1588.

170

Schimmel, P and K. Beebe. 2004. Molecular biology: genetic code seizes pyrrolysine. Nature. 431:257-258.

Schimmel, P., and D. Söll. 1979. Aminoacyl-tRNA synthetases: general features and recognition of transfer RNAs. Annu. Rev. Biochem. 48: 601–648.

Schink, B. 1997. Energetics of syntrophic cooperation in methanogenic degradation. Microbiol. Mol. Biol. Rev. 61: 262-280.

Schmitt, E., Moulinier, L., Fujiwara, S., Imanaka, T., Thierry, J. C., and D. Moras. (1998). Crystal structure of aspartyl-tRNA synthetase from Pyrococcus kodakaraensis KOD: archaeon specificity and catalytic mechanism of adenylate formation. EMBO J. 17:5227-5237.

Schull, C. and H. Beier. 1994. Three Tetrahymena tRNA(Gln) isoacceptors as tools for studying unorthodox codon recognition and codon context effects during protein synthesis in vitro. Nucleic Acids Res. 22:1974-1980.

Scolnick, E., Tompkins, R., Caskey, T., and M. Nirenberg. 1968. Release factors differing in specificity for terminator codons. Proc. Natl. Acad. Sci. U.S.A. 61: 768-774.

Shaul, S., Nussinov, R., and T. Pupko. 2006. Paths of lateral gene transfer of lysyl-aminoacyl-tRNA synthetases with a unique evolutionary transition stage of prokaryotes coding for class I and II varieties by the same organisms. BMC Evol. Biol. 6: 22.

Shen, Q., Chu, F. F., and P. E. Newburger. Sequences in the 3' untranslated region of the human cellular glutathione peroxidase gene are necessary and sufficient for selenocysteine incorporation at the UGA codon. 1993. J. Biol. Chem. 268: 11463-11469.

Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and D.C. Klenk. 1985. Measurement of protein using bicinchoninic acid. Anal Biochem 150:76– 85.

Soares, J. A., Zhang, L., Pitsch, R. L., Kleinholz, N. M., Jones, R. B., Wolff, J. J., Amster, J., Green-Church, K. B., and J. A. Krzycki. 2005. The residue mass of L- pyrrolysine in three distinct methylamine methyltransferases. J. Biol. Chem. 280: 36962-36969.

Sowers, K.R., Baron, S.F., and J.G. Ferry. 1984. Methanosarcina acetivorans sp. nov., an acetotrophic methane-producing bacterium isolated from marine sediments. Appl. Environ. Microbiol 47: 971-978. 171

Sowers, K.R., Boone, J.E., and R.P. Gunsalus. 1993. Disaggregation of Methanosarcina spp. and growth as single cells at elevated osmolarity. Appl. Environ. Microbiol. 59: 3832-3839.

Sowers, K. R., and H. J. Schreier. 1995. Media for methanogens. In Archaea, a Laboratory Manual. K. R. Sowers and H. J. Schreier (eds.). Plainview, NY: Cold Spring Harbor Laboratory Press, 459–489.

Sprinzl M., Hartmann, T., Meissner, T., Moll, J., and T. Vorderwüllbecke. 1987. Compilation of tRNA sequences and sequences of tRNA genes. Nucleic. Acids Res. 15:53-188

Srinivasan, G., James, C. M., and J. A. Krzycki. 2002. Pyrrolysine encoded by UAG in Archaea: Charging of a UAG-decoding specialized tRNA. Science. 296: 1459-1462.

Srinivasan, G. 2003. Translation of the amber codon in methylamine methyltransferase genes of a methanogenic archaeon. Ph.D. Thesis, The Ohio State University. 91-92.

Stadtman, T. C. 1996. Selenocysteine. Annu. Rev. Biochem. 65: 83–100.

Stortchevoi, A.A. 2006. Misacylation of tRNA in prokaryotes: a re-evaluation. Cell Mol. Life Sci. 63: 820-831.

Söll, D., Becker, H.D., Plateau, P., Blanquet, S., and M. Ibba. 2000. Context- dependent anticodon recognition by class I lysyl-tRNA synthetases. Proc. Natl. Acad. Sci. U.S.A. 97: 14224-14228.

Tang, Y. and D.A. Tirrell. 2002. Attenuation of the editing activity of the Escherichia coli leucyl-tRNA synthetase allows incorporation of novel amino acids into proteins in vivo. Biochemistry. 41:10635-10645.

Tate, W. P., Mansell, J. B., Mannering, S. A., Irvine, J. H., Major, L. L., and D. N. Wilson. 1999. UGA: a dual signal for “stop” and for recoding in protein synthesis. Biochemistry (Mosc). 64: 1342-1353.

Tate, W. P., Poole, E. S., Horsfield, J. A., Mannering, S. A., Brown, C. M., Moffat, J. G., Dalphin, M. E., McCaughan, K. K., Major, L. L., and D. N. Wilson. 1995. Translational termination efficiency in both bacteria and mammals is regulated by the base following the stop codon. Biochem. Cell Biol. 73: 1095-1103. Terada, T., Nureki, O., Ishitani, R., Ambrogelly, A., Ibba, M., Söll, D., and S. Yokoyama. 2002. Functional convergence of two lysyl-tRNA synthetases with unrelated topologies. Nat. Struct. Biol. 9:257-262. 172

Thauer, R. K. 1998. Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiol. 144: 2377-2406.

Théobald-Dietrich, A., Giegé, R., and J. Rudinger-Thirion. 2005. Evidence for the existence in mRNAs of a hairpin element responsible for ribosome dependent pyrrolysine insertion into proteins. Biochimie. 87: 813-817.

Théobald-Dietrich, A., Frugier, M., Giegé, R., and J. Rudinger-Thirion. 2004. Atypical archaeal tRNA pyrrolysine transcript behaves towards EF-Tu as a typical elongator tRNA. Nucleic Acids Res. 32: 1091-1096.

Tumbula, D., Vothknecht, U. C., Kim, H-s., Ibba, M., Min, B., Li, T., Pelaschier, J., Stathopoulos, C., Becker, H., and D. Söll. 1999. Archaeal aminoacyl-tRNA synthesis: Diversity replaces dogma. Genetics. 152: 1269-1276. 189

Tallant, T. C., and J. A. Krzycki. 1997. Methylthiol:coenzyme M methyltransferase from Methanosarcina barkeri, an enzyme of methanogenesis from dimethylsulfide and methylmercaptopropionate. J. Bacteriol. 179: 6902-11.

Turnbaugh, P. J., Ley, R. E., Mahowald, M. A., Magrini, V., Mardis, E. R., and J.I. Gordon. 2006. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 444: 1027-1031.

Valle, R.P., Morch, M.D., and A.L. Haenni. 1987. Novel amber suppressor tRNAs of mammalian origin. EMBO J. 6:3049-3055

Varshney, U., Lee, C. P. and U. L. RajBhandary. 1991. Direct analysis of aminoacylation levels of tRNAs in vivo. Application to studying recognition of Escherichia coli initiator tRNA mutants by glutaminyl-tRNA synthetase. J. Biol. Chem. 266: 24712–24718.

Veit, K., Ehlers, C., and R.A. Schmitz. 2005. Effects of nitrogen and carbon sources on transcription of soluble methyltransferases in Methanosarcina mazei strain Gö1. J. Bacteriol. 187: 6147-6154.

Wang, L., Brock, A., Herberich, B. and P. G. Schultz. 2001. Expanding the genetic code of Escherichia coli. Science. 292: 498–500.

Wang, K., Neumann, H., Peak-Chew, S. Y., and J.W. Chin. 2007. Evolved orthogonal ribosomes enhance the efficiency of synthetic genetic code expansion. Nat. Biotechnol. 25:770-777

173

Wang, L., and P.G. Schultz. 2004. Expanding the genetic code. Angew. Chem. Int. Ed. Engl. 44:34-66.

Wang, L., Xie, J., and P.G. Schultz. 2006. Expanding the genetic code. Annu. Rev. Biophys. Biomol. Struct. 35:225-249.

Wang, S., Praetorius-Ibba, M., Ataide, S.F., Roy, H., and M. Ibba. 2006. Discrimination of cognate and noncognate substrates at the active site of class I lysyl-tRNA synthetase. Biochemistry 45: 3646-3652. Gln Weiss W.A. and E.C. Friedberg. 1986. Normal yeast tRNA CAG can suppress amber codons and is encoded by an essential gene. J Mol. Biol. 192:725-735.

Welander, P. V., and W. W. Metcalf. 2005. Loss of the mtr operon in Methanosarcina blocks growth on methanol, but not methanogenesis, and reveals an unknown methanogenic pathway. Proc. Natl. Acad. Sci. U.S.A. 102: 10664-10669.

Wilting, R., Schorling, S., Persson, B. C., and A. Böck. 1997. Selenoprotein synthesis in archaea: identification of an mRNA element of Methanococcus jannaschii probably directing selenocysteine insertion. J. Mol. Biol. 266: 637-641.

Woese, C.R., Olsen, G.J., Ibba, M., and D. Söll. 2000. Aminoacyl-tRNA synthetases, the genetic code, and the evolutionary process. Microbiol. Mol. Biol. Rev. 64: 202-236.

Woese, C. R., Kandler, O., and M. L. Wheelis. 1990. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc. Natl. Acad. Sci. U.S.A. 87: 4576-4579.

Wood, G.E., Haydock, A.K., and J.A. Leigh. 2003. Function and regulation of the formate dehydrogenase genes of the methanogenic archaeon Methanococcus maripaludis. J. Bacteriol. 185: 2548–2554.

Yaremchuk, A., Tukalo, M., Grotli, M., and S. Cusack. 2001. A succession of substrate induced conformational changes ensures the amino acid specificity of Thermus thermophilus prolyl-tRNA synthetase: comparison with histidyl-tRNA synthetase. J. Mol. Biol. 309:989-1002.

Yarus M, and P. Berg. 1969. Recognition of tRNA by isoleucyl-tRNA synthetase. Effect of substrates on the dynamics of tRNA-enzyme interaction. J. Mol. Biol. 42:171-189.

174

Yoshinaka, Y., Katoh, I., Copeland, T. D., and S. Oroszlan. 1985. Murine leukemia virus protease is encoded by the gag-pol gene and is synthesized through suppression of an amber termination codon. Proc. Natl. Acad. Sci. U.S.A. 82: 1618-1622.

Zhang, J.K., Pritchett, M.A., Lampe, D.J., Robertson, H.M., and W.W. Metcalf. 2000. In vivo transposon mutagenesis of the methanogenic archaeon Methanosarcina acetivorans C2A using a modified version of the insect mariner- family transposable element Himar1. Proc. Natl. Acad. Sci. U.S.A. 97: 9665- 9670.

Zhang, Y., Baranov, P. V., Atkins, J. F., and V. N. Gladyshev. 2005. Pyrrolysine and selenocysteine use dissimilar decoding strategies. J. Biol. Chem. 280: 20740-20751.

Zhang, Y. and V. N. Gladyshev. 2007. High content of proteins containing 21st and 22nd amino acids, selenocysteine and pyrrolysine, in a symbiotic deltaproteobacterium of gutless worm Olavius algarvensis. Nucleic Acids Res 35: 4952-63.

Zhouravleva, G., Frolova, L., Le Goff, X., Le Guellec, R., Inge-Vechtomov, S., Kisselev, L., and M. Philippe. 1995. Termination of translation in eukaryotes is governed by two interacting polypeptide chain release factors, eRF1 and eRF3. EMBO J. 14: 4065-4072.

Zinder, S. H. 1993. In Methanogenesis. J. G. Ferry (ed.). Chapman & Hall, New York/London. 128.

Zinoni, F., Heider, J., and A. Böck. 1990. Features of the formate dehydrogenase mRNA necessary for decoding of the UGA codon as selenocysteine. Proc. Natl. Acad. Sci. U.S.A. 87: 4660-4664.

Zoldak, G., Redecke, L., Svergun, D. I., Konarev, P. V., Voertler, C. S., Dobbek, H., Sedlak, E., and M. Sprinzl. 2007. Release factors 2 from Escherichia coli and Thermus thermophilus: structural, spectroscopic and microcalorimetric studies. Nucleic Acids Res. 35:1343-1353.

175