MIAMI UNIVERSITY The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation

of

Adam John Creighbaum

Candidate for the Degree

Doctor of Philosophy

______Dr. Donald J. Ferguson Jr, Director

______Dr. Annette Bollmann, Reader

______Dr. Xin Wang, Reader

______Dr. Rachael Morgan-Kiss

______Dr. Richard Page, Graduate School Representative

ABSTRACT

EXAMINATION AND RECONSTITUTION OF THE GLYCINE BETAINE- DEPENDENT METHANOGENESIS PATHWAY FROM THE OBLIGATE METHYLOTROPHIC VULCANI B1D

by

Adam J. Creighbaum

Recent studies indicate that environmentally abundant quaternary amines (QAs) are a primary source for methanogenesis, yet the catabolic are unknown. We hypothesized that the methanogenic archaeon Methanolobus vulcani B1d metabolizes glycine betaine through a corrinoid-dependent glycine betaine:coenzyme M (CoM) methyl transfer pathway. The draft genome sequence of M. vulcani B1d revealed a gene encoding a predicted non- pyrrolysine MttB homolog (MV8460) with high sequence similarity to the glycine betaine methyltransferase encoded by Desulfitobacterium hafniense Y51. MV8460 catalyzes glycine betaine-dependent methylation of free cob(I)alamin indicating it is an authentic MtgB . Proteomic analysis revealed that MV8460 and a corrinoid binding protein (MV8465) were highly abundant when M. vulcani B1d was grown on glycine betaine relative to growth on trimethylamine. The abundance of a corrinoid reductive activation enzyme (MV10335) and a methylcorrinoid:CoM methyltransferase (MV10360) were significantly higher in GB-grown B1d lysates compared to other homologs. The glycine betaine:CoM pathway was fully reconstituted in vitro using recombinant MV8460, MV8465, MV10335, and MV10360. Demonstration of the complete glycine betaine:CoM pathway expands the knowledge of direct QA-dependent methylotrophy and establishes a model to identify additional ecologically relevant anaerobic quaternary amine pathways.

EXAMINATION AND RECONSTITUTION OF THE GLYCINE BETAINE- DEPENDENT METHANOGENESIS PATHWAY FROM THE OBLIGATE METHYLOTROPHIC METHANOGEN METHANOLOBUS VULCANI B1D

A DISSERTATION

Presented to the Faculty of

Miami University in partial

fulfillment of the requirements

for the degree of

Doctor of Philosophy

Department of Microbiology

by

Adam J. Creighbaum

The Graduate School Miami University Oxford, Ohio

2020

Dissertation Director: Donald J. Ferguson Jr., Ph. D.

©

Adam John Creighbaum

2020

TABLE OF CONTENTS

LIST OF TABLES iv

LIST OF FIGURES v

LIST OF COMMON ABBREVIATIONS viii

DEDICATION

ACKNOWLEDGEMENTS

INTRODUCTION 1

CHAPTER 1. Examination of the glycine betaine-dependent 30 methylotrophic methanogenesis pathway: insights into anaerobic quaternary amine methylotrophy

Chapter 1.1. Genomic and proteomic analysis of Methanolobus 31 vulcani B1d

Chapter 1.2. Screening the function of MV8460, MV8465, 57 MV10335, and MV10360 from Methanolobus vulcani B1d

Chapter 1.3. In vitro reconstruction of the glycine betaine:CoM 86 methylotrophic pathway from Methanolobus vulcani B1d

APPENDIX I. Analyzing the interchangeability of the MtaA and RamM 105 with homologs from Methanococcoides methylutens Q3c, Methanosarcina acetivorans WWM73, Methanosarcina barkeri Fusaro, Methanomethylovorans hollandica to reconstruct the glycine betaine:CoM methyl transfer pathway from Methanolobus vulcani B1d.

REFERENCES 135

iii LIST OF TABLES

Table Title Page

1 Methylotrophy-associated proteins encoded 35 within Methanolobus vulcani B1d

2 Primers and plasmids 59

3 Proteins selected from M. barkeri Fusaro based on 113 transcriptomic data (López Muñoz et al., 2015)

4 Proteins selected from M. acetivorans WWM73 based on 114 transcriptomic data (Peterson et al., 2016)

5 Proteins selected from M. hollandica based on shotgun 115 Proteomic data and genomic analysis (Chen et al., 2017)

6 Primer sequences of the MtxAs and Rams 117

7 Current status of cloning, production, and activity testing 118 of the proteins selected for this study

iv LIST OF FIGURES

Figure Page

1 Schematic pathway of the three methanogenesis pathways 2

2 Three component system depicting demethylation 7 of a

3 Interaction of MtaB and MtaC with the methyl 11 donor methanol

4 Phylogenetic tree of the COG5598 superfamily of enzymes 15

5 Proposed mechanism utilized by Desulfitobacterium 17 hafniense Y51 to demethylate glycine betaine and methylate tetrahydrafolate

6 Schematic depicting the Stickland reaction on glycine, 20 sarcosine, and betaine

7 Growth curve of Methanolobus vulcani B1d 25

8 Hypothetical methanogenesis pathways for the breakdown 27 quaternary amines

9 The genome of M. vulcani B1d encodes a single 36 homologous MttB that lacks pyrrolysine, mtgB (MV8460)

10 1. Proteomic analysis of likely candidate proteins for 38 glycine betaine-dependent CoM methylation

11 2. Proteomic analysis of likely candidate proteins for 40 glycine betaine-dependent CoM methylation

12 The genome of M. vulcani B1d encodes for an entire 44 methanol pathway (MeOH1) with all the essential genes within proximity of each other

13 The genome of M. vulcani B1d has three pairs of mtmBCs 51 and two pairs of mtbBCs

14 Examples of M. barkeri MS and M. acetivorans enzymes 53 involved in methylotrophy from methylated thiols

v 15 The genome of M. vulcani B1d contains a mtsD/H/F 56 (MV10015) that could encode for a functional protein that could demethylate a methylated thiol compound

16 Gene sequence of the optimized gene encoding MV10360 62 from GenScript

17 A 12% acrylamide SDS-PAGE gel followed by Coomassie 69 blue staining of purified recombinant proteins used to reconstitute the glycine betaine:CoM methyl transfer pathway

18 predictions of DhMtgB and MV8460 76

19 Predicted structural model of MV10350 compared to 78 known MtaB from Methanosarcina barkeri Fusaro

20 Glycine betaine:cob(I)alamin methyl-transfer activity 80 of MV8460

21 Reductive activation of MV8465 by MV10335 82

22 Methylcob(III)alamin:CoM methyl-transfer activity 84 by MV10360

23 Approximate-maximum likelihood representation of 93 the COG5598 MttB superfamily

24 Reconstitution of glycine betaine:CoM activity in vitro 95 with purified recombinant proteins

25 Glycine betaine:CoM activity in vitro using crude extracts 98

26 Representative figure of methanogenesis assays 100 performed on M. vulcani B1d

27 Proposed model of glycine betaine-dependent 104 CoM methylation

28 Relative activities of the MtbA and MtaA enzymes with 109 MttB from M. barkeri during TMA:CoM assays

29 Methylcob(III)alamin:CoM methyl-transfer activity 125 by MV1575 and MV1695

30 Methylcob(III)alamin:CoM methyl-transfer activity 127 by MM0619

vi

31 Representative figure of reconstitution of glycine 129 betaine:CoM activity in vitro with purified recombinant proteins 32 Approximate-maximum likelihood representation of the 132 MtxA phylogenetic tree

33 Approximate-maximum likelihood representation of the 134 Ram phylogenetic tree

vii LIST OF COMMON ABBREVIATIONS Name Abb.

Carbon dioxide CO2 Tetramethylammonium QMA Methyl-coenzyme M reductase Mcr Coenzyme M CoM

Hydrogen H2

Reduced ferredoxin Fdred Methanofuran MFR

Tetrahydromethopterin H4MPT Free thiol SH Coenzyme M methyltransferase Mtr Heterodisulfide reductase Hdr

Oxidized ferredoxin Fdox Coenzyme B CoB Coenzyme A CoA

Tetrahydrosarcinapterin H4SPT Trimethylammonium TMA Dimethylammonium DMA Monomethylammonium MMA Methylthiol:Coenzyme M Methyltransferase MtsA Methylcorrinoid:Coenzyme M Methyltransferase MtxA x: a = methanol; t = trimethylammonium; b = dimethylammonium m = monomethylammonium; s = methylated sulfurs, g = glycine betaine; q = tetramethylammonium

Oxygen O2 Dissolved inorganic carbon DIC Triosephosphate TIM

viii L-Pyrrolysine Pyl L-Pyrrolysine lacking non-Pyl Tetrahydrofolate THF Choline-TMA CutC CutC activation enzyme CutD Glycine betaine GB Corrinoid activation enzyme Ram Glycine betaine transporter OpuD Kildalton kDa Methanol MeOH Calf-intestinal phosphatase CIP Multiple cloning site MCS Rotations per minute RPM Isopropyl β-D-1-thiogalactopyranoside IPTG

Nitrogen N2 Genomic DNA gDNA

ix DEDICATION I would like to dedicate this dissertation to Grace Eib. For those of you that do not know her, she is my high school sweetheart. She has been with me every step of the way since I left for my undergraduate studies at Manchester University. When I reflect on graduate school and my research that eventually resulted in this dissertation, I feel like Grace and I were in this together. Grace has been in college for the long haul too (Child Neurologist incoming!), and I think it is safe to say we have worked together through the pains of achieving our higher degrees even though our interests and paths were different. This work presented here focuses on a unique metabolic pathway from a methanogen, and Grace listened to this topic for 6 years(!) and never complained. I am no physician, but I do not think methanogenesis has much to do with Child Neurology. Instead, she was supportive and loving the entire time. I know for a fact I would have not finished the work that went into this dissertation had she not been there for me. I guess you could say that I am a pretty lucky person to have someone so remarkable in my life. This dissertation is for you, Grace.

x

ACKNOWLEDGEMENTS

I would like to thank Dr. Joe Krzycki for many valuable discussions. I thank Dr. Annette Bollman and Dr. Xin Wang for their willingness to be readers for this dissertation and their valuable input. I thank Dr. Rachael Morgan-Kiss and Dr. Richard Page, along with the other members of my committee, for their challenging questions to help expand my dissertation project. I also acknowledge and thank the staff (Dr. Andor Kiss and Ms. Xiaoyun Deng) of the Center for Bioinformatics and Functional Genomics (CBFG) at Miami University for instrumentation and computational support. Lastly, I would like to thank my advisor Dr. D.J. Ferguson for accepting me into his lab and guiding me through this project and helping shape me into a microbiologist. I look at D.J. not just as a mentor but as a friend that I could go to for various discussions pertaining to nearly any topic. He introduced me to some of the best whiskeys that are currently in my liquor cabinet. I will miss our whiskey hours together that sparked many of the discussions that turned into creative ideas that could expand my project. D.J. allowed for me to explore and attempt projects or experiments without ridicule even though he probably knew they were destined to fail from the beginning. He presented to me the perfect environment where I could critically think and learn by my own errors, and for that, I am grateful. I truly could not have asked for a better mentor. Thank you for helping make my time in graduate school rememberable!

xi Introduction

Biological production of methane by methanogenic () is a strictly anoxic process and the organisms performing this metabolism fall into the phylum (Thauer et al., 2008). Overall methane levels released into the environment by methanogens are estimated at 1 gigaton per year with future predictions suggesting an increase in this number (Thauer, 2008; Yvon-Duroche et al., 2014). The global warming potential of methane has been estimated to range upwards to 28 times greater than that of carbon dioxide (CO2) (Myhre et al., 2013; Wallace et al., 2014; and Dean et al., 2018), which is arguably the main gas that comes to mind during debates over the greenhouse effect. Since the natural breakdown of methane in the atmosphere is CO2, understanding methanogenic potentials and the metabolisms involved by these microorganisms is essential in order to make more accurate estimations pertaining to global warming. Methanogenesis as a whole is comprised of three major known pathways: hydrogenotrophic, acetoclastic, and methylotrophic (Ferry, 2011) (Fig. 1). Currently all bona-fide methanogens possess a methyl-coenzyme M reductase (Mcr). Mcr functions to reduce methylated coenzyme M (CoM), causing the release of methane. To our knowledge, the majority of methanogenic organisms are hydrogenotrophic, however the majority of global methane released biologically is due to acetoclastic methanogenesis (Thauer et al, 2008; Ferry, 2011). Although according to studies it appears that hydrogenotrophic and acetoclastic pathways dominate many environments, it has recently been acknowledged through geochemical profiling that methylotrophy is the primary mechanism for methane production in hypersaline and sulfate reduction zones in marine sediments (Liu and Whitman, 2008; Lazar et al., 2011; Zhuang et al., 2016; Maltby et al., 2018; Zhuang et al., 2018; and Yin et al., 2019). With the recent discovery of a series of obligate methylotrophic methanogens capable of using a new class of methylotrophic substrates known as quaternary amines (except for work by Tanaka (1994) on tetramethylammonium; QMA) as their sole carbon source (Watkins et al., 2012; Watkins et al., 2014; and Ticak et al., 2015), arises the speculation that unknown methylotrophic methanogenic metabolisms exist, and their pathways need to be -

1 Figure 1. Schematic pathway of the three methanogenesis pathways. Each converge at methylation of coenzyme M (CoM) followed by subsequent methane release. The aceticlast pathway is blue, hydrogenotrophic pathway is black, and methylotrophic pathway is green. Arrows correspond to the colored pathway indicating which direction the carbon is going as the pathway proceeds (Lambie et al., 2015).

2

Figure 1

3 determined. Discovery of these new forms of metabolism aids in future efforts to determine how widespread methanogenesis is and can increase our estimation capabilities for predicting methane release. This dissertation focuses on one such metabolic pathway that is being used by the obligate methylotrophic methanogen Methanolobus vulcani B1d that produces methane from the quaternary amine glycine betaine.

The acetoclastic and hydrogenotrophic pathway

Most methanogens that grow via the hydrogenotrophic pathway reduce CO2 to methane using molecular hydrogen (H2) or formate as the initial electron donor, although, through electrochemistry studies there have been reports that show electrons can be gained from direct interaction with cathodes or environmental metals such as iron (Uchiyama et al., 2010; and Rowe et al., 2019). The CO2 reduction pathway proceeds through a stepwise manner which requires a reduced ferredoxin (Fdred) and a formylmethanofuran dehydrogenase for the initial reaction between CO2 and methanofuran (MFR) to form formyl-MFR. The formyl group is then transferred to tetrahydromethanopterin (H4MPT) to form formyl-H4MPT. The exception to utilization of

H4MPT is Methanosarcina barkeri which utilizes tetrahydrosarcinapterin (H4SPT), the analog to H4MPT that is traditionally thought to be used for acetoclastic methanogenesis (discussed below). The formyl group is then dehydrated to form methenyl- H4MPT followed by sufficient electron addition from reduced coenzyme F420 to generate methyl-H4MPT. Energy is then conserved in the subsequent transfer of the methyl group to the free thiol (SH) of CoM by methyl-H4MPT coenzyme M methyltransferase (Mtr) yielding methyl-S-CoM. The presence of a membrane bound Mtr is essential for generation of a Na+ ion gradient during all three methanogenesis pathways. During hydrogenotrophic and acetoclastic methanogenesis the transfer of the methyl group from methyl-H4MPT to CoM is an exergonic reaction and contributes to the overall ion motive force by allowing 2 Na+ ions to transport across the membrane (Mand and Metcalf, 2019). Generation of methyl-S-CoM is thus far known to be required for all three methanogenic pathways as methyl-S-CoM acts as a substrate for Mcr for release of methane. The electrons required for reduction of methyl-SCoM are received from reduced coenzyme B (CoB) and ultimately a heterodisulfide CoM-S-S-CoB is

4 formed and methane is released. A heterodisulfide reductase complex (Hdr) breaks the bond between CoM and CoB in an energy conservation step allowing for recycling of the two coenzymes (Thauer et al., 2008). The energy is conserved by Hdr generating a proton motive force by pumping protons out of the cell (Schlegel and Müller, 2011). The previously mentioned generation of the Na+ ion gradient by Mtr along with the proton motive force generated by Hdr allows for methanogens to simultaneously generate a proton and Na+ ion gradient. However, the link of the gradients generated by these proteins to ATP synthesis is still not clear, however a known A1AO ATP synthase is required for generation of ATP (Schlegel and Müller, 2011). Traditionally it is represented that the Hdr is the last step and that hydrogenotrophy is a linear process starting with electrons being gained from H2 to initiate reduction of ferredoxin. However, Richards and coworkers (2016) have suggested a route that links Hdr to the first step of reduction of CO2 through electron bifurcation where the electrons from H2 are used to reduce oxidized ferredoxin (Fdox) and also are part of the reaction performed by Hdr to separate CoM and CoB yielding the following reaction: CoB−S−S−CoM + 2H2 + Fdox ⇌ + HS−CoB + HS−CoM + 2H + Fdred. Acetoclastic methanogenesis is initiated by an kinase phosphorylating acetate to form acetyl-Pi. The phosphate group is then removed and replaced with coenzyme A (CoA) by phosphate acetyltransferase to form acetyl-CoA. The two-carbon acetyl group then serves as both a methyl source for direct methylation of H4SPT by acetyl-CoA synthetase to form methyl-H4SPT and release of carbon monoxide via carbon monoxide dehydrogenase which is then further oxidized to CO2. The oxidation to

CO2 yields electrons which can be used for other cellular processes during the reduction of methyl-S-CoM to release methane (Ferry, 2011). In general, this is the main process used by organisms, such as Methanosarcina barkeri, to grow solely using the acetoclastic pathway. However, there are examples such as Methanococcus maripaludis that can only assimilate acetate but not generate energy. The energy for M. maripaludis in this instance is derived from hydrogenotrophy and it is unknown as to why energy cannot be obtained from acetate alone (Richards et al., 2016).

5

Methylotrophic pathways from methanogens Methanogenesis from simple methylated amines [trimethylamine (TMA), dimethylamine (DMA), and monomethylamine (MMA)], methylated thiols, and methanol centers around the idea that a methyl group is removed from these compounds and transferred to CoM (Fig. 1). These systems have been studied extensively in organisms from the family (Ferguson et al., 1996; Ferguson and Krzycki, 1997; Sauer, et al., 1997; Sauer and Thauer, 1998, Ferguson et al., 2000; Tallant et al., 2001). With the exception of methylthiols, which can be broken down either by a single methyltransferase (methylthiol:Coenzyme M methyltransferase; or MtsA) or, depending on the type of methylated thiol, by a single-subunit from the MtsD/F/H family, methylotrophic methanogens utilize a three component system for successful methylation of CoM (Fig. 2). In general, these three component systems consist of a substrate-specific methyltransferase (MtxB), a cognate corrinoid binding protein (MtxC), and a secondary methylcorrinoid:CoM methyltransferase (MtxA) (Fig. 2 ). The x represents a substrate- specific designation for each protein (a = methanol, t = trimethylamine (TMA), b = dimethylamine (DMA), m = monomethylamine (MMA), and s = methylated sulfurs), except in the case of MtbA which can function in each of the three simple methylamine pathways (Ferguson et al., 1996; Ferguson and Krzycki, 1997). Additionally, an ATP- dependent activation enzyme such as RamA is intermittently required for reductive activation of the cobalt of the MtxC to the catalytically active Co(I) state in the event MtxC is exposed to an oxidant such as O2 (Ferguson et al., 2009). Once activated, the reduced cobalt atom acts as a nucleophile which accepts a methyl group from the substrate whose transfer is catalyzed by MtxB. The methylated MtxC then acts as a substrate for MtxA for the transfer of the methyl group to CoM (Ferry, 2011). During methylotrophic methanogenesis, reducing equivalents for reducing the methyl group on methyl-CoM to methane are typically gained by oxidizing one out of every four methyl groups to CO2, however some methanogens use H2 as an energy source to drive the pathway (Borrel et al., 2013; Enzmann et al., 2018). The enzymes involved in oxidizing

6 a methyl group to CO2 are the same enzymes used by hydrogenotrophic methanogens, except obligate methylotrophs are incapable of utilizing H2/CO2 as a growth substrate.

The oxidation of a methyl group to CO2 for reducing equivalents involves generation of methyl-H4MPT or methyl-H4SPT. There are discrepancies about the methyl group transitioning back to generate acetyl-CoA for the purpose of assimilating the carbon into building other essential materials for biomass formation (Yin et al., 2019). Interestingly, studies by Yin and coworkers (2019) has shown that the obligate methylotrophic organism Methanococcoides methylutens produced approximately 12% of its methane from CO2 even though the organism lacks the essential hydrogenase to generate the electrons for the initial reduction of CO2. Additionally, it was shown that 60- 86% of dissolved inorganic carbon (DIC) was incorporated into lipids by M. methylutens, however, methane derived from methanol was still the primary metabolism being performed. This phenomenon of incorporating DIC into biomass and methane produced from CO2 occurred at lower temperatures when overall rates of methane were beginning to lower, which is typical in marine sediments and would not be observed in isolated cultures or optimal growth conditions (Yin et al., 2019). These findings are relatively new and need further investigation in order to understand whether other methylotrophic methanogens can perform a similar form of metabolism using compounds such as methylated amines. As previously stated, the initial degradation step of the methylated substrates begins by interaction with the MtxB enzyme. Per our current knowledge, MtxB enzymes contain a triosephosphate isomerase (TIM) barrel fold that forms a cavity where the active site is located to which the methyl donor would enter before demethylation occurs (Ragsdale, 2008). MtaB, the methanol:corrinoid methyltransferase, binds 1 mol of zinc per 1 mol of polypeptide. The zinc(II) atom is bound at a novel location within the TIM barrel and mutations to the residues predicted to be involved with binding result in dramatic loss of enzymatic activity (Krüer et al., 2002; and Hagemeier et al., 2006). Further analysis of MtaB and its cognate corrinoid binding partner, MtaC, supported that the complex formed between the two proteins resulted in MtaC reaching into the catalytic pocket of MtaB forming a binding location for methanol (Fig. 3a) (Hagemeier et al., 2006). As methanol enters into the catalytic pocket it is polarized by the acidic

7 density of the residues surrounding the Zn(II) which causes an increase in charge density of Zn(II). Zn(II) then functions as an electrophile that activates the C-O bond of the compound allowing for nucleophilic attack by the reduced corrinoid of MtaC (Fig. 3b) (Hagemeier et al., 2006). No further crystal structure evidence is available to understand the mechanism of demethylation by MtsA or the enzymes responsible for demethylation of simple methylated amines, at this time.

8 Figure 2 Three component system depicting demethylation of a substrate. The substrate is demethylated by MtxB (see exception for methylated thiols in-text) and MtxC is methylated. MtxC is then demethylated by MtxA and coenzyme M (CoM) is then methylated. Intermittent oxidation of MtxC can occur to the inactive Co(II), which is then reactivated to Co(I) by Ram with the consumption of ATP.

9

Figure 2

10 Figure 3 Interaction of MtaB and MtaC with the methyl donor methanol. A. The placement of the cobalt that is bound by MtaC can be seen in relative proximity to the bound Zn(II) atom by MtaB. The Zn(II) is surrounded by acidic residues that causes a charge density increase of Zn(II) and ultimately placing methanol in an optimal position to be demethylated. B. The Zn(II) acts an electrophile to the methanol molecule allowing for nucleophilic attack of the Co(I) that is bound to MtaC. The methyl group is successfully transferred to MtaC generating methyl-Co(III).

11 Figure 3

12 Pyrrolysine in methanogenesis and the COG5598 superfamily of enzymes A fortunate consequence of the study of methanogenesis from methylamines was the discovery of the 22nd genetically encoded amino acid L-pyrrolysine (Pyl) (Krzycki, 2005). It is suggested that Pyl evolved independently in each of the predicted active sites of MttB, MtbB, and MtmB as they are non-homologous enzymes (Paul et al., 2000), and removal of Pyl abolishes methanogenesis from each of the methylamine substrates (Krzycki, 2005). The synthesis of Pyl requires the five genes pylTSBCD, where gene products PylBCD are required for biosynthesis, while PylT is the tRNA and PylS is the tRNA synthetase that charges the amino acid onto the tRNA (Srinivasan et al., 2002; Blight et al., 2004; Longstaff et al., 2007; and Jiang and Kzrycki, 2012). No crystal structure for MttB, MtbB, or MtmB has been published and there is an ongoing effort to understand the reason for Pyl’s presence in the catalytic pocket of these enzymes. However, none of these enzymes bind zinc and therefore it is possible that Pyl is utilized to stabilize the methyl donor compound prior to demethylation. The differences in the catalytic pockets has yet to be distinguished and therefore it is not known if overall structure of the pocket itself plays a role in selection of the methyl donor. The TMA methyltransferase, MttB, is the namesake member of the widespread MttB COG5598 superfamily of enzymes, which has members spanning hundreds of from both the Bacteria and Archaea domains (Ticak et al., 2014) (Fig. 4). Interestingly, most members of COG5598 lack Pyl and are therefore likely not functional TMA methyltransferases. This observation lead to the speculation that perhaps the COG5598 superfamily was involved with catabolizing higher order methylated ammonium compounds. Further work by Ticak and co-workers confirmed that that one non-Pyl (or Pyl-lacking) MttB homolog from Desulfitobacterium hafniense Y51 (MtgB) is a glycine betaine methyltransferase (Ticak et al., 2014). Moreover, a mechanism for demethylation of glycine betaine with subsequent methylation of tetrahydrofolate (THF), the C1-carrying coenzyme that is used by anaerobic respiring methylotrophic bacteria, was proposed but not fully elucidated (Fig. 5). Since the initial work involving MtgB, another group has published a fully intact demethylation pathway from the human intestinal isolate Eubacterium limosum that utilizes the quaternary amine proline betaine

13 (Picking et al., 2019). Their work supported the glycine betaine:THF proposed model in that it followed a three component pathway. However, despite the work published thus far on methylotrophic metabolism by bacteria from quaternary amines, it has yet to be determined if methanogens follow an analogous CoM-methylating pathway utilizing a Pyl-lacking MttB.

14 Figure 4 Phylogenetic tree of the COG5598 superfamily of enzymes. The namesake of the superfamily comes from MttB which is responsible for the demethylation of TMA. Majority of the members of the superfamily lack L-pyrrolysine (Pyl) and therefore their functions are unknown (blue). The green coloring (light green, bacterial; dark green, archaeal) are those enzymes that are predicted to contain Pyl. The red colored clade contains the Pyl-lacking MtgB enzyme shown to demethylate glycine betaine (Ticak et al., 2014).

15

Figure 4

16 Figure 5 Proposed mechanism utilized by Desulfitobacterium hafniense Y51 to demethylate glycine betaine and methylate tetrahydrafolate (THF). The enzymes are named based on their locus found within the genome of D. hafniense Y51. DSY3156 is MtgB and is a demonstrated glycine betaine methyltransferase. DSY3155 is the cognate corrinoid binding protein, MtgC. DSY3157 is the MtgA that would demethylate MtgC and subsequently methylate THF.

17

Figure 5

18 The Stickland Reaction on glycine betaine Glycine betaine was initially thought to only contribute to methanogenesis indirectly for its role as a source of TMA due to breakdown by fermentative bacteria (King, 1984). This is particularly important in marine environments where methylotrophic methanogens compete well, due to high sulfate concentrations in the sediments (Purdy et al., 2003). Organisms from the genera Clostridium; and additionally, Eubacterium acidaminophilum; can encode for the Stickland Reaction (Stickland, 1934; Stickland, 1935; Nisman, 1954; Barker,1981; and King, 1984). Successful breakdown of glycine betaine anoxically to generate TMA and acetate was first characterized by Naumann and co-workers (1983) using C. sporogenes and Methanosarcina barkeri in a coculture. The mechanism of action for successful reduction of glycine betaine (or sarcosine or glycine) is comprised of three different protein fractions: A, B, and C (Fig. 6) (Andreesen, 2004). Protein fraction A is comprised of the substrate-binding enzyme that differs based on substrate and then also whether it is from Clostridium or E. acidaminophilum. The substrate-binding selenocysteine protein for glycine betaine from Clostridium is GrdHI, and it functions to bind glycine betaine through an unknown mechanism and ultimately forms GrdHI-carboxymethyl (acetyl)-selenoether. It is supported that the proteins involved with sarcosine or glycine create a temporary Schiff formation. This is not possible with glycine betaine because the of polarized quaternary amine. The generation of an acetyl-selenoether occurs during of glycine betaine, sarcosine, or glycine and therefore the proteins found in fraction A and C are the same regardless of the starting substrate. GrdA from fraction A is also a selenocysteine protein and accepts the acetyl group from GrdHI to form GrdA-acetyl- selenoether. GrdA is then acted on by GrdCD from protein fraction C through a poorly understood mechanism to transfer the acetyl group onto GrdCD forming an acetyl thioester which is subsequently cleaved off releasing acetyl phosphate. Work thus far supports that the end products of the Stickland reaction from glycine betaine are acetate and TMA, but when the TMA is released during the mechanism does not appear to be fully known (Naumann et al., 1983; Meyer et al., 1995; and Andreesen, 2004).

19 Figure 6 Schematic depicting the Stickland reaction on glycine, sarcosine, and betaine. The enzymes required for interaction with the substrate are found in protein fraction B. Protein fraction A contains GrdA which receives the acetyl group from protein A. The last step for transfer of the acetyl group is an interaction between GrdA and protein fraction C in an energy saving step (Andreesen, 2004).

20

Figure 6

21 Methanogenesis from quaternary amines When studying metabolisms in communities by using microcosms that mimic coastal intertidal sediments it was estimated that up to 90% of the methane released was from glycine betaine and other structurally similar quaternary amines (Oremland et al., 1982; King, 1984; and Jones et al., 2019). Given that quaternary amines are a major source of TMA in these environments it can be assumed that majority of the methane released would follow initial catalysis of the quaternary amine. There are currently no suitable methods for determining the concentrations of specific quaternary amines in the environment, however, TMA has been measured at nanomolar levels in oceanic environments to millimolar concentrations in marine and brackish environments (Gibb et al., 1999; Fitzsimons et al., 2001; Gibb and Hatton, 2004; and Beale and Airs, 2016). Given this knowledge it is not currently possible to determine which specific quaternary amine is responsible for the majority of the methane released from these environments, or if this can be contributed to a single type of quaternary amine. Additionally, it is not known what proportion of methane is due to direct quaternary amine utilization versus initial cleavage then utilization of TMA. Besides the recently described GrdHI from the Stickland reaction on glycine betaine, another set of enzymes responsible for quaternary amine degradation to yield TMA includes choline-TMA lyase (CutC) and the CutC activation enzyme (CutD) (Craciun and Balskus, 2012). Briefly, CutC uses a mechanism dramatically different from GrdHI in that it uses a glycyl radical to successfully cleave the C-N bond of choline to generate TMA and acetaldehyde (Hayward and Stadtman, 1959; Hayward and Stadtman 1960; Hayward, 1960; and Craciun and Balskus, 2012). Even though it has been known that glycine betaine and choline both can serve as precursors for TMA, utilizing knowledge of the GrdHI and CutCD enzymes in combination of metagenomic sequencing, researchers have started to pinpoint the specific bacteria in these environments that are responsible for catalysis of glycine betaine and choline (Jameson et al., 2018; and Jones et al., 2019). Granted that majority of microorganisms in the world are uncultured at the genera level (Hug, 2018) and therefore it is unknown how they are involved in carbon cycling, one major example of combining the aforementioned items was the discovery of the novel family Candidatus ‘Betainaceae’

22 fam. nov. in Clostridiales and its involvement in generation of TMA in salt marshes (Jones et al., 2019). This arguably presses the issue further that although quaternary amines have been linked to methanogenesis for decades there are still gaps in fully understanding their impact on the global carbon cycle. Outside of the work performed by Asakawa and co-workers (1998) on the methanogenic organism Methanococcoides sp NaT1 (NaT1) (Tanaka, 1994) and its ability to directly utilize QMA for methanogenesis, studies determining how quaternary amines can be used directly by methanogens has been scarce. Therefore, understanding the contribution of methane release within communities found naturally in these environments can be grossly mistaken without knowing the metabolic potentials of the organisms present. Methanogenesis from QMA was shown to proceed via a methylotrophic pathway using proteins apparently unique to its pathway (Asakawa et al., 1998). Unfortunately, NaT1 was lost and no genomic information was reported, leaving many unanswered questions regarding quaternary amine-dependent methanogenesis. Particularly, the sequence identity of the QMA:corrinoid methyltransferase (MtqB) compared to other analogous methyltransferases is unknown and whether MtqB contained Pyl is also unknown.

Methanolobus vulcani B1d and growth from glycine betaine Nearly two decades after NaT1 was discovered, more methanogens were reported to utilize quaternary amines as a direct substrate, including M. vulcani B1d, which can grow on glycine betaine (Watkins et al., 2012; Watkins et al., 2014; Ticak et al., 2015). M. vulcani B1d was isolated from brackish sediments from the Southwest Branch Back River, Virginia, USA, while using glycine betaine as the sole carbon source for enrichment. M. vulcani B1d is highly similar (99.1% identity with 16s rRNA) to Methanolobus vulcani PL-12MT which belongs to the order and more specifically to the family Methanosarcinaceae (Stetter, 1982; and Ticak et al., 2015). Despite the high similarity, M. vulcani PL-12MT cannot grow on glycine betaine as a sole carbon source. The Methanolobus is one of few methanogenic genera that have been indicated as obligate methylotrophs in that they receive all of their energy from performing methylotrophy (Fig. 1) (Yin et al., 2019). It was therefore

23 predicted that growth from glycine betaine was most likely due to methylotrophy as the Stickland reaction has thus far only been found in bacteria. Carbon equivalence studies of methane produced versus glycine betaine consumed were in a ratio of 0.71:1, which is within the ratio range to suggest a single carbon was being used from glycine betaine (Fig. 7) (Ticak et al., 2015). Additionally, from this work was the isolation of Methanococcoides methylutens Q3c, another obligate methylotrophic methanogen that can use the quaternary amines QMA and choline. No further discussion of M. methylutens Q3c will be done in this dissertation, but this organism is worth mentioning because its metabolism of the two quaternary amines helped generate the proposed schematic for quaternary amine catalysis from M. vulcani B1d and M. methylutens Q3c (Fig. 8) (Ticak et al., 2015). In general, the quaternary amine would be internalized by a transporter and would then follow one of the three potential pathways. 1.) The quaternary amine could be demethylated through a three-component system that would be analogous to TMA demethylation resulting in methylation of CoM. 2.) Similarly to the first idea, quaternary amines could serve as methyl donors for direct methylation of H4SPT. 3.) The quaternary amine could be initially degraded to form formaldehyde through an unknown quaternary amine dehydrogenase and then a condensation reaction could occur causing formation of methylene-H4SPT. This third pathway would follow a mechanism that was proposed by Welander and Metcalf (2008) and ultimately bypass Mtr/Mer (Fig. 1). Despite the possibility of this third pathway being an option for quaternary amine degradation, Ticak et al. (2015) indicated that this possibility was unlikely due to previous studies on a bypass system from methanol in Methanosarcina barkeri that did not support growth (Welander and Metcalf, 2005). No further analysis was performed in the study of M. vulcani B1d or M. methylutens Q3c in terms of genomic analysis or determining the metabolic pathway involved with breakdown of the quaternary amines.

24 Figure 7 Growth curve of Methanolobus vulcani B1d. The proportion of methane produced to glycine betaine (GB) consumed was 0.71:1, which was attributed to a demethylation mechanism (Ticak et al., 2015).

25

Figure 7

26 Figure 8 Hypothetical methanogenesis pathways for the breakdown quaternary amines. The traditional three component pathway would stem from pathway I with direct methylation of coenzyme M (CoM). An alternative pathway (pathway II) would utilize direct methylation of tetrahydrosarcinapterin (H4SPT) which the methyl group would then be transferred to CoM for subsequent release of methane. Lastly, the third pathway (pathway III) requires formation of a formaldehyde intermediate prior to formation of methylene-H4SPT (Ticak et al., 2015).

27

Figure 8

28 Goal of this study At the start of this project, the only known information regarding the physiological pathway for methanogenesis from quaternary amines was the work on NaT1 and QMA. The discovery of Pyl came after the work with NaT1, yet it was not until 2014 when the first methyltransferase from the COG5598 superfamily was determined to function on a quaternary amine. Despite the knowledge gained from studying MtgB from D. hafniense Y51 and the isolation of new methanogens capable of growing on quaternary amines, no progress to determine a methanogenesis pathway from these compounds had been made. Much work had shifted from three component systems in terms of methylated amines because those compounds of higher order were always thought to be precursors for methanogenesis. The isolation of M. vulcani B1d using glycine betaine as its carbon source sparked interest in our laboratory to determine the metabolic pathway used by this organism to grow on glycine betaine. This would allow for a renewed study of the metabolism of ecologically relevant compounds that play a significant role in the global carbon cycle. We therefore embarked on the primary goal of identifying and then demonstrating the functionality of the enzymes that initiate methanogenesis from glycine betaine.

29

Chapter 1 Examination of the glycine betaine-dependent methylotrophic methanogenesis pathway: insights into anaerobic quaternary amine methylotrophy

Adam J. Creighbaum, Tomislav Ticak, Shrameeta Shinde, Xin Wang, Donald J. Ferguson Jr.

Frontiers in Microbiology, 10:2572

30 Chapter 1.1 Genomic and proteomic analyses of Methanolobus vulcani B1d

Overview This section focuses on the genomic and proteomic analyses that were performed on Methanolobus vulcani B1d. The genome analysis utilizes previous work from Dr. Tomislav Ticak (Ticak dissertation, 2015) as a foundation for the project presented in this section and expands upon it to include updated information that was published following his initial work. The isolation of M. vulcani B1d using glycine betaine as the sole carbon source was one of few methanogenic organisms shown to grow directly on quaternary amines. Dr. Ticak’s screening of the M. vulcani B1d draft genome resulted in finding an encoded Pyl-lacking MttB (MV8460) that has high identity and sequence similarity to the glycine betaine demethylating MtgB that is produced by Desulfitobacterium hafniense Y51. Given that M. vulcani B1d is an obligate methylotrophic methanogen, MV8460 was the initial target protein thought to initiate the metabolism of glycine betaine. Additionally, a cognate corrinoid binding partner protein was found to be encoded directly next to MV8460. These early observations allowed for development of a core hypothesis that M. vulcani B1d utilizes MV8460, which shares high identity and similarity to MtgB from D. hafniense Y51, to demethylate glycine betaine and then methylate the corrinoid binding protein. Any major items that were re- confirmed in this document that were from Dr. Ticak’s initial analysis are stated in the text; along with the differences that were discovered following his work.

31 Materials and Methods Methanolobus vulcani B1d growth conditions Methanolobus vulcani B1d was routinely cultivated under strict anaerobic conditions in brackish medium using either glycine betaine (80 mM), methanol (62.5 mM), or trimethylammonium (TMA) (40 mM) as the growth substrate (Ticak et al., 2015). Cells were kept static at 37°C to avoid any potential introduction of oxygen to the media. The growth phases were monitored at OD600 to ensure similar growth curves were observed as originally published by Ticak et al., 2015. After several passages if cells started to shift from the original growth pattern, frozen stock cultures were revived and re-passaged for consistency. Cultures were passaged a minimum of three times when transitioning to a new carbon source to allow for complete acclimation.

Genomic analysis and deposition to NCBI As stated previously, the draft genome of M. vulcani B1d (accession number VIAQ00000000) was used to find genes of interest. The annotation process performed by GenBank numbered the gene locus tags in increments of five. The MtgB (DSY3156) from Desulfitobacterium hafniense Y51 was used to search for potential Pyl-lacking MttB homolog candidates from M. vulcani B1d. Additionally, the M. vulcani B1d genome was screened for potential gene products using MttB (accession number P0C0W7.3),

MttC (accession number AAD14631.1), MtaB (accession number Q46EH3.1), MtaC (Q46EH4.1), MtbA (accession number O30640.3), MtaA (Q48949.1), and RamA

(accession number ACL01114.1) from Methanosarcina barkeri MS. Putative operons (partial and potentially complete) that could encode for enzymes to degrade methanol, TMA, DMA, MMA, or glycine betaine were generated and used for further analysis when determining which MtxA and Ram enzyme was used by M. vulcani B1d for degradation of glycine betaine. Additionally, screening for potential genes that would encode for demethylation of methylated thiols included: MtsA (accession: AAC46230.1) and MtsB (accession number AAC46231.1) from M. barkeri MS and MtsD (accession number AAM04298.1), MtsF (accession number AAM07726.1), and MtsH (accession number AAM07897.1) from Methanosarcina acetivorans.

32 Proteomic Analysis M. vulcani B1d cells grown to mid-log phase on either glycine betaine (80 mM), methanol (62.5 mM), or TMA (40 mM) were used for proteomic analysis following similarly to previously described methods (Ticak et al., 2015). Anoxically harvested cell pellets were resuspended in 12 mL of buffer (50 mM Tris-HCl, 10 mM CaCl2, 0.1 % n- Dodecyl β-D-maltoside, pH 7.6), followed by cell lysis via French press, as described below. Cell lysates were centrifuged at 40,000 x g for 45 min at 4°C and the supernatants were collected and used for proteomic analysis. Protein concentrations estimates were determined using the Bradford assay (Bradford, 1976) (Thermo Scientific). For each sample, 100 µg of total protein was denatured in 8 M urea supplemented with 5 mM dithiothreitol. The protein solution was diluted to 2 M urea using the same Tris buffer, followed by a 1:100 w/w Trypsin Gold (Promega) protein digestion at 37°C for 18 hrs. Digested peptides were desalted using Sep-Pak C18 columns following the manufacturer’s protocol (Waters Corporation), followed by peptide fractionation using a Pierce High pH Reverse-Phase Peptide Fractionation Kit (Thermo Scientific). Eight peptide fractions of each sample were separated using a capillary C18 column on an EASY-nLC 1000 liquid chromatograph coupled to a Thermo LTQ Orbitrap XL mass spectrometer for MS analysis. The peptides were scanned in the range of 350-1800 m/z at a resolution of 30,000 operating in the data-dependent mode. For each scan, the 12 most abundant peaks were selected and subjected to MS/MS analysis by collision induced dissociation fragmentation. The peptide identities were searched against the M. vulcani B1d database using pipeline programs integrated in PatternLab for Proteomics (version 4.1.0.17) and normalized spectral abundance factor (NSAF) was used to compare protein abundances. Raw data and the searched .sqt files have been deposited to the MassIVE repository with the identifier MSV000084013.

33 Results and Discussion Genomic and Proteomic Analysis Significant gaps in knowledge remain regarding the pathways for methanogenesis from quaternary amines. To begin addressing these gaps we analyzed the draft genome of M. vulcani B1d for genes of interest related to methylotrophic metabolisms from methanol, TMA, DMA, MMA, glycine betaine, or methylated thiols (Table 1). We also undertook a proteomic analysis of M. vulcani B1d during growth on methanol, TMA, or glycine betaine (Fig. 1 and 2). M. vulcani B1d was never tested for growth on methylated thiol compounds and therefore only a genomic analysis was performed to report the putative genes predicted to be involved. Our reasoning for only using TMA and not its degradation products (DMA and MMA) is that it has been noted that TMA metabolism upregulates enzymes needed for DMA and MMA metabolism during growth of M. vulcani B1d on TMA (Dr. Bill Metcalf, personal communication). At the start of this project, MtgB, from D. hafniense Y51 (DhMtgB) was the only Pyl-lacking MttB homolog of the COG5598 superfamily of enzymes that showed functionality, and it was determined to demethylate glycine betaine (Ticak et al., 2014). Congruent with work that was noted in Dr. Ticak’s dissertation (2015) and following genomic annotation by GenBank, there was only a single encoded Pyl-lacking MttB homolog (MV8460) found within the genome. MV8460 is directly flanked by an encoded putative cognate corrinoid binding protein homolog (MV8465) and a putative glycine betaine transporter, OpuD, (MV8455) (Fig. 9). MV8460 was found to be 65% identical and has 83% sequence similarity to DhMtgB. Given the information thus far, we considered the genes encoding MV8455, MV8460, and MV8465 as a putative operon and potential candidates for being involved in glycine betaine metabolism. Glycine betaine-grown M. vulcani B1d compared to methanol- or TMA-grown M. vulcani B1d showed significant increases in both MV8460 and MV8465 (Fig. 10). MV8455 was present only during growth on glycine betaine (Fig. 11). In addition to MV8455 there is another putative OpuD (MV8440) that is 64% identical and 80% similar to MV8455 and is encoded upstream of the putative glycine betaine degradation operon. However, there are two genes, a sensory transduction histidine kinase (MV8445) and an encoded putative MtaC (MV8450), between MV8440 and the potential operon (Fig. 9). MV8440

34 Table 1 Methylotrophy-associated proteins encoded within Methanolobus vulcani B1d

Protein Gene Annotated Putative Function Locus MtbA 1575 Methylamine specific methylcobalamin:CoM methyltransferase MtmB 1580 MMA specific methyltransferase MtmB 1585 MMA specific methyltransferase MtmC 1590 MMA pathway corrinoid binding protein MtbB 1595 DMA specific methyltransferase MtbC 1600 DMA pathway corrinoid binding protein LicB 1610 Choline transporter MttC 1615 TMA pathway corrinoid binding protein MttB 1620 TMA specific methyltransferase MtmB 1650 MMA specific methyltransferase MtmC 1655 MMA specific methyltransferase DMA permease 1675 DMA permease MtbB 1680 DMA specific methyltransferase MtbC 1685 DMA pathway corrinoid binding protein MtbA 1695 Methylamine specific methylcoblamin:CoM methyltransferase MtmB 1720 MMA specific methyltransferase MtmC 1725 MMA specific methyltransferase Pyl tRNA synthetase 1750 Decoding amber codon (UAG) PylB 1755 Pyrrolysine biosynthesis PylC 1760 Pyrrolysine biosynthesis PylD 1765 Pyrrolysine biosynthesis RamA 1770 Corrinoid activation enzyme MtaA 2245 Methanol specific cobalamin:CoM methyltransferase MtaC 375 Methanol pathway corrinoid binding protein MtaB 380 Methanol specific methyltransferase OpuD 8440 Glycine betaine transporter Sensory Transduction Histidine 8445 Histidine kinase Kinase MtaC 8450 Methanol pathway corrinoid binding protein OpuD 8455 Glycine betaine transporter MtgB 8460 Pyl-lacking glycine betaine methyltransferase MtgC 8465 Glycine betaine pathway corrinoid binding protein RamA 10000 Corrinoid activation enzyme GroEL 10005 Chaperone proteins Hydantoinase 10010 Functionality unknown in terms of methylotrophy MtsD/F/H 10015 Methylated thiol methyltransferase RamM 10335 Corrinoid activation enzyme Metal Chaperone Zn 10340 Zinc homeostasis Homeostasis GTPase (CobW) MtaC 10345 Methanol pathway corrinoid binding protein MtaB 10350 Methanol methyltransferase Hydantoinase 10355 Functionality unknown in terms of methylotrophy MtaA 10360 Methanol specific cobalamin:CoM methyltransferase

35 Figure 9 The genome of M. vulcani B1d encodes a single homologous MttB that lacks pyrrolysine, mtgB (MV8460). Directly upstream of mtgB is the predicted glycine betaine transporter, opuD (MV8455), and downstream is the cognate corrinoid binding protein, mtgC (MV8465). The second opuD (MV8440), the histidine sensory kinase hsk (MV84415), and mtaC (MV8450) are also shown due to proximity of mtgB and mtgC.

36

Figure 9

37 Figure 10 Proteomic analysis of likely candidate proteins for glycine betaine-dependent CoM methylation. (A) MtxBs, (B) MtxCs, (C) MtxAs, and (D) Ram enzymes during M. vulcani B1d growth on glycine betaine, methanol, or TMA as the sole carbon source. Any proteins that did not appear during growth on glycine betaine were omitted and can be found in Fig. 11. Protein abundances were estimated by using the normalized spectral abundance factor (NSAF). Significant differences are indicated as follows when comparing individual protein levels between substrates: p ≤ 0.05 are shown by (∗), p ≤ 0.01 are shown by (∗∗), and p ≤ 0.001 are shown by (∗∗∗). The green bar indicates a significant difference between levels of analogous enzymes produced when grown on glycine betaine. GB = glycine betaine, TMA = trimethylammonium, and MeOH = methanol.

38

Figure 10

39 Figure 11 Proteomic analysis of likely candidate proteins for glycine betaine-dependent CoM methylation. (A) MtxBs, (B) MtxCs, (C) MtxAs, and (D) OpuD (MV8455) during M. vulcani B1d growth on glycine betaine, methanol, or TMA as the sole carbon source. Protein abundances were estimated by using the normalized spectral abundance factor (NSAF). Significant differences, based on standard deviations (n = 3), are indicated as follows when comparing individual protein levels between substrates: p ≤ 0.05 are shown by (*), p ≤ 0.01 are shown by (**), and p ≤ 0.001 are shown by (***). The green bar indicates a significant difference between levels of analogous enzymes produced when grown on glycine betaine. GB = glycine betaine, TMA = trimethylammonium, and MeOH = methanol.

40

Figure 11

41 was not present during growth on any of the substrates that were tested and therefore we do not think it has a role in glycine betaine metabolism. OpuD transporters are ABC transporters and it has been noted that ABC transporters are sometimes misannotated or that they can appear similar, yet their functionality is unknown (Galagan et al., 2002). More work needs to be performed on the transporters responsible for moving simple methylated amines and quaternary amines into the cell to better differentiate the potential roles of the transporters. MV8450 is the last protein to be mentioned due to it being encoded nearby the putative glycine betaine degradation operon. MV8450 is 52% identical and has 77% sequence similarity to MV10345, another putative MtaC, which is encoded within a cluster of genes that encode for the various predicted proteins responsible for methylotrophy from methanol (Fig. 12). MV10345 and the other encoded proteins near MV10345 that are predicted to be involved with methanol metabolism will be discussed later. Out of the three proteomic replicates performed on M. vulcani B1d growth from glycine betaine, two indicated the presence of MV8450, but it was not within the top 900 proteins detected in either replicate (data not shown). We therefore do not think MV8450 is involved with glycine betaine metabolism as it is typical for methylotrophic enzymes to be among the most abundant during growth on the substrate they are metabolizing (Fu and Metcalf, 2015; López Muñoz et al., 2015; and Peterson et al., 2016). Given our current knowledge that methylotrophy from TMA, DMA, MMA, or methanol each require a methylcobalamin:CoM methyltransferase (MtxA) and a corrinoid activation enzyme (Ram) (Ferguson et al., 2009), we predicted that homologs of these enzymes would be involved with glycine betaine metabolism. We found no MtxA or Ram enzyme near the glycine betaine degradation operon. However, M. vulcani B1d does encode elsewhere four MtxA enzymes, two of which were annotated as MtaAs (MV2245 and MV10360) for methanol metabolism, and two that were annotated as MtbAs (MV1575 and MV1695) for methylated amine metabolism. There is one Ram enzyme that is annotated as a RamM (MV10335) for methanol metabolism, and two that are annotated as RamAs (MV1770 and MV10000) for methylated amine metabolism. MV10360 and MV10335 are encoded near MV10345, the putative MtaC that was mentioned previously (Fig. 12). MV10335 was produced in significantly higher

42 amounts during growth on methanol compared to TMA, but levels on glycine betaine were not significantly different to methanol or TMA. MV10360 was produced at equal levels during growth on each substrate (Fig. 10). Both MV1575 and MV1695 were produced during all the conditions tested but MV1575 was significantly more abundant than MV1695 and was most abundant during growth on TMA. MV10360 is significantly more abundant than MV1575 during growth on glycine betaine and therefore we predicted that MV10360 is the main MtxA that is used during methylotrophy from glycine betaine. Similarly, to MV10360, MV10335 was produced ubiquitously regardless of the substrate, however it had significantly higher abundance during growth on methanol compared to TMA or glycine betaine. MV1770 was produced at significantly higher levels when M. vulcani B1d was grown on TMA compared to methanol or glycine betaine, while MV10000 was not produced during any conditions. When comparing the abundance of MV10335 to MV1770 during growth on glycine betaine, we concluded that MV10335 was significantly more abundant than MV1770. Therefore, we predict that if an MtxA and Ram enzyme were needed for growth on glycine betaine, that MV10360 and MV10335 would be used by M. vulcani B1d even though they are not encoded near MV8460 and MV8465.

43 Figure 12 The genome of M. vulcani B1d encodes for an entire methanol pathway (MeOH1) with all the essential genes within proximity of each other. The mtaB (MV10350) and mtaC (MV10345) are directly next to each other, while the ramM (MV10335) and mtaA (MV10360) are separated by a cobW (MV10340) and a hyuA (MV10335), respectively. As seen from Fig. 11 and 12, the methylotrophic protein products of the genes are present during all types of growth.

44

Figure 12

45 Further genomic and proteomic analysis suggested a complete putative methanol degradation pathway (MeOH1) (Fig. 12) and a second partial methanol degradation pathway (MeOH2) (data not shown). Two of the proteins from MeOH1, MV10360 and MV10335, were discussed previously and found to be produced during all growth conditions. Encoded adjacent to MV10335 is a putative metal chaperone GTPase that is involved with zinc homeostasis. MV10335 was produced during all growth conditions with no apparent significant differences between the conditions. The presence of this protein on various growth conditions and being encoded near other proteins that can metabolize methanol is not surprising as both MtaB and MtaA are zinc binding enzymes (Sauer and Thauer, 1997). Encoded next to MV10360 is a putative hydantoinase (MV10355) and it was produced at equal levels during all growth conditions. There is no current evidence supporting whether hydantoinases play a role in methylotrophy, and therefore no in-depth analysis will be performed on these enzymes. However, there are several hydantoinases encoded within the M. vulcani B1d genome that are located near enzymes responsible for TMA, DMA, and MMA degradation, therefore they could constitute an avenue for future work. Additionally, MV10345, the MtaC, was produced during all growth conditions but showed significantly more abundance during growth on methanol compared to glycine betaine or TMA. There are two MtaBs (MV10350 and MV0380) encoded in the M. vulcani B1d genome and MV10350 is the final protein to be discussed that is part of MeOH1. MV10350 is produced during growth on all tested conditions, but it is significantly in higher abundance during growth on methanol compared to glycine betaine or TMA (Fig. 10). The MeOH2 only contains MV0380 and a MtaC (MV0375). Both proteins are produced during all growth conditions and at significantly higher abundances during growth on methanol compared to glycine betaine or TMA (Fig. 11). MV0380 and MV0375 were produced at significantly lower abundances compared to their homologs; MV10350 and MV10345, respectively; during growth on methanol. Therefore, given our current data, if MV0380 and MV0375 were involved with methanol metabolism, their role would likely be minor compared to MV10345 and MV10350. Interestingly, when comparing abundances of MV8460, the Pyl-lacking MttB homolog, and MV10350, MV10350 is at a statistically greater abundance during growth on glycine betaine. This

46 is also seen with MV10345, the MtaC, when comparing to MV8465 during growth on glycine betaine. These observations are discussed in Section 1.2. In conclusion, the proteomic data revealed constitutive production of the proteins involved in methanol- dependent methanogenesis (Fig. 10 and 11) at varying levels depending on the substrate. This suggests that methanol could be M. vulcani B1d’s most commonly available carbon source in its environment, which would make it advantageous for the organism to maintain production of the enzymes under all conditions. The only other COG5598 family member found to be encoded in the genome is MV1620 a Pyl-containing MttB with 30% identity and 50% similarity to MV8460. Encoded directly next the MV1620 is a putative MttC (MV1615). There are other MttCs encoded within the genome of M. vulcani B1d, but none of these appear to be located near other putative encoded enzymes that would be used for methylotrophy and therefore they will not be discussed further at this time. MV1615 and MV1620 were both produced during all growth conditions but were statistically more significant during growth on TMA. Encoded nearby MV1615 is an encoded putative LicB protein (MV1610) that is annotated as a choline permease. Traditionally uptake and activation of exogenous choline requires LicABC. There is no LicAC encoded near LicB, however, LicB is present in low abundance only during growth on TMA but not glycine betaine or methanol. As previously stated, regarding OpuD transporters, more work needs to be performed concerning methylamine and quaternary amine transporters to properly distinguish them and their functionalities. When analyzing the genome for genes that would encode metabolic enzymes for MMA, it was determined there were three pairs of MtmBs and MtmCs (Table 1). Further gene analysis of the MtmBs indicated that they appear to be potential gene duplications within the genome and therefore we cannot definitively state which was used during growth on MMA. The MtmBs were greater than 90% identical with greater than 94% sequence similarity, while the MtmCs were greater than 91% identical with greater than 92% sequence similarity. The proteomic analysis showed that these MtmBC pairs were significantly produced at higher levels during growth on TMA versus glycine betaine or methanol (Fig. 11). As stated earlier, this is not surprising as an eventual breakdown product of TMA is MMA. Worth noting is a fourth putative MtmB (MV1580) that does not

47 appear to have a MtmC partner and is encoded next to the MtmB (MV1585) that is part of one of the previously mentioned MtmBC pairs (Fig. 13). Interestingly, there is only 31% identical with 53% amino acid sequence similarity between MV1580 and MV1585, and unlike the other MtmBs, MV1580 does not contain the critical Pyl residue. Trace amounts of MV1580 were detected during the proteomic analysis on all conditions tested and it appears to be in equal abundance during growth with TMA or methanol as the substrate. The gene that encodes for MV1580 faces in the opposite direction as the rest of the surrounding genes (Fig. 13). The trace levels that can be detected during the proteomics suggests that M. vulcani B1d actively generates RNA specifically for this gene because of its orientation. Pyl-lacking MtmBs and MtbBs are encoded periodically within the genomes of various methylotrophic methanogens, however, their functionalities are unknown and therefore future work could be performed to understand the functionality of MV1580. Similarly, to the MtmBC pairs, there are two pairs or MtbBCs encoded within the genome of M. vulcani B1d. Upon further analysis it was determined that the MtbBs are 92% identical with 97% sequence similarity. The MtbCs are also 92% identical with 97% sequence similarity. Therefore, we considered this to be gene duplication and could not determine which pair is more prevalent during growth conditions. The MtbBC pairs are statistically more significant during growth on TMA as compared to growth on glycine betaine or methanol. Further experimentation such as RNA sequencing could be performed to try and better determine which MtmBC and MtbBC pairs are present during growth on TMA. The final two enzymes needed for successful methylotrophy from TMA, DMA, or MMA, is a RamA and a MtbA. As stated previously, MV1575 and MV1770 were the MtbA and RamA enzymes that were produced at higher abundance during growth on TMA compared growth on glycine betaine or methanol. We therefore predict that these two enzymes are the ones primarily used during methylotrophic growth from TMA. Since M. vulcani B1d grows on TMA, DMA, and MMA, it requires the presence of Pyl. The entire machinery for Pyl synthesis and processing (PylTS and PylBCD) is encoded directly before MV1770. Except for PylD, the rest of the enzymes required for production of Pyl are present during all growth conditions and are statistically in higher abundance during growth on TMA. PylD was only detected during growth on glycine

48 betaine and TMA with a higher abundance during TMA growth. However, Pyl-containing MtmBs, MtbBs, and MttBs are all detectable during growth on methanol. This suggests potentially PylD is either not completely needed for Pyl synthesis, which is unlikely (Srinivasan et al., 2002; Blight et al., 2004; Longstaff et al., 2007; and Jiang and Kzrycki, 2012), or that it is present but at undetectable levels. As indicated, M. vulcani B1d was not tested for growth using methylated thiols as a substrate. M. acetivorans and Methanosarcina siciliae can grow using methylated thiol compounds, but M. barkeri MS will only convert the methylated thiol to methane and requires another carbon source such as acetate for growth (Sowers et al., 1984; van der Maarel et al., 1995; and Tallant et al., 2001). The reduction of methyl-CoM and subsequent release of methane is linked to an energy conservation step and therefore it is predicted M. barkeri would perform methylotrophy from methylated thiols purely for acquisition of energy. The two possible mechanisms for demethylation of dimethylsulfide, as discussed earlier, would either be by MtsAB found in M. barkeri MS or through using MtsD from the family of single-subunit CoM methyltransferases (MtsD/F/H) from M. acetivorans (Tallant et al., 2001; and Fu and Metcalf, 2015) (Fig. 14) . These proteins are considered single-subunit CoM methyltransferases because they contain both a C-terminal “MtxA” domain and a corrinoid binding N-terminal domain that forgoes the need for an additional corrinoid binding protein. The MtsF can metabolize methanethiol, and MtsH can bind either substrate (Fu and Metcalf, 2015). There is a single protein (MV10015) encoded within the M. vulcani B1d genome that is annotated as both a dimethylsulfide methyltransferase corrinoid protein and dimethylsulfide:corrinoid methyltransferase (Fig. 15). Dr. Ticak’s dissertation (2015) noted this fused protein and suggested potential growth of M. vulcani B1d on dimethylsulfide. A protein-protein BLAST analysis of the proteins encoded from the M. vulcani B1d genome using MtsA (accession number AAC46230.1) from M. barkeri MS as a query showed a 21% identity with 39% sequence similarity to MV10015. In fact, MtsA has higher identity and sequence similarity to MV1695 at 34% and 53%, respectively. MV10015 is 34% identical and shares 56% sequence similarity to MtsB (accession number AAC46231.1). However, comparisons between MV10015 and MtsD (accession number AAM04298.1) showed a 79% identity with 90% sequence similarity,

49 compared to MtsF (accession number AAM07726.1) there is a 54% identity with 71% sequence similarity, and a 68% identity with 83% sequence similarity to MtsH (accession number AAM07897.1). With the high percentages with each of the queries it is difficult to predict the function of MV10015 with the current knowledge in the field.

50 Figure 13 The genome of M. vulcani B1d has three pairs of mtmBCs and two pairs of mtbBCs. Depicted here is one of the mtmBC pairs, one mtbBC pair, a methylcobalamin:CoM methyltransferase mtbA (MV1575), and the single Pyl-lacking mtmB (MV1580). MV1580 faces in the opposite direction as the rest of the other methylotrophic enzymes.

51

Figure 13

52 Figure 14 Examples of (A) M. barkeri MS and (B) M. acetivorans enzymes involved in methylotrophy from methylated thiols. (A) Dimethylsulfide (DMS) and methylmercaptopropionate (MMPA) breakdown to methanethiol (MeSH) and mercaptopropionate (MPA), respectively, by M. barkeri MS requires the methyltransferase MtsA and corrinoid-binding protein MtsB for transfer of the methyl group to CoM. (B) Putative DMS:CoM mechanism based on reassessment of MtsD functionality during DMS metabolism by M. acetivorans. This figure was adapted from work published by Fu and Metcalf, (2015).

53

Figure 14

54 Since each of the MtsD/F/H proteins binds a corrinoid, a reduction enzyme is required for activation of that corrinoid containing domain. It was determined that the so- called RamS (MA0849) (accession number AAM04288.1) from M. acetivorans is 61% identical with 78% sequence similarity to MV10000, encoded near MV10015 (Fig. 15), although functionality of the RamS has not been confirmed (Fu and Metcalf, 2015). Transcriptomic data, however, does show MA0849 is strongly upregulated during growth of M. acetivorans on dimethylsulfide and methanethiol (Fu and Metcalf, 2015). Interestingly, MV10000 shares lower percentages with MV1770, the other RamA, at 55% identical with 74% sequence similarity. This suggests that although MV10000 is annotated as a RamA, it is more closely related to the RamS enzyme from M. acetivorans and therefore may be utilized during growth that involves methylated thiols. Additionally, nearby MV10015 are encoded putative GroEL and hydantoinase, neither of which help hint towards utilization of methylated thiol compounds. Further growth experiments and biochemical tests need to be performed to determine if M. vulcani B1d can grow utilizing methylated thiols and to assess what roles MV10015 and MV10000 may have with the metabolism of M. vulcani B1d.

55 Figure 15 The genome of M. vulcani B1d contains a mtsD/H/F (MV10015) that could encode for a functional protein that could demethylate a methylated thiol compound. Given that the protein of interest here contains a fused corrinoid domain, it is not surprising to find a ramS (MV10000) nearby. The hyuA (MV10010) and groEL (MV10005) have no known functionality in terms of methylotrophy, but homologs do periodically appear next to other encoded methylotrophic proteins.

56

Figure 15

57 In conclusion we were able to determine key proteins that were abundant during different growth conditions that we think play a role during the metabolisms of methanol, TMA, and glycine betaine. The key proteins we predict to be involved for glycine betaine metabolism are the Pyl-lacking MtxB, MV8460, from the COG5598 superfamily of enzymes, the corrinoid binding MtxC, MV8465, the corrinoid activation enzyme RamM, MV10335, and the methylcorrinoid:CoM enzyme MtxA, MV10360. Although we predict the original purpose of MV10335 and MV10360; along with the MtaC, MV10345, and MtaB, MV10350; would have been for methanol metabolism, M. vulcani B1d has adapted to be able to use the first two for glycine betaine metabolism. However, further genomic and evolution experiments would need to be performed to determine whether this phenomenon could occur. The biochemical testing to determine functionality of these four proteins are presented in the following sections 1.2 and 1.3. Additionally, we predict that gene duplication occurred regarding the MtmBs, MtmCs, MtbBs, and MtbCs, and therefore we were unable to distinguish which ones play the primary role during DMA and MMA metabolism. The presence of these enzymes during M. vulcani B1d growth on DMA and MMA is expected as these pathways have been extensively studied in the past (Ferguson et al., 2000). The remaining enzyme of interest is the Pyl- lacking MV1580, which differs from the other MtmB gene duplicates based on identity and sequence similarity, but nonetheless, does appear to be homologous. To our knowledge the only Pyl-containing and Pyl-lacking homologs we can compare based on functionality are MttBs and Pyl-lacking MttBs from the COG5598 superfamily, where examples of the latter demethylate glycine betaine or proline betaine (Ticak et al., 2014; and Picking et al., 2019). Given no functionality has been shown for Pyl-lacking MtmBs and that abolishing Pyl synthesis causes loss of methanogenesis from TMA, DMA, and MMA, we can speculate that MV1580 would not demethylate MMA but might function on a higher order methylated amine or different methyl donor altogether.

58 Chapter 1.2 Screening the function of MV8460, MV8465, MV10335, and MV10360 from Methanolobus vulcani B1d

Overview Based on the predictions made using the proteomic and genomic data gathered from M. vulcani B1d, we began testing the function of the individual proteins we predict to be responsible for glycine betaine metabolism. Those proteins included the Pyl- lacking MttB homolog (MV8460), the corrinoid binding protein (MV8465), a methylcob(lll)alamin:CoM methyltransferase (MV10360), and a corrinoid activation enzyme (MV10335). Each of these were present during the growth on glycine betaine according to our proteomic analysis performed in section 1.1. In addition to testing functionality of what we predict to be the glycine betaine methyltransferase, we took a modeling approach to compare a bona-fide MtgB from D. hafniense Y51 to MV8460 as they come from two separate domains of life. MV10350 was also modeled as it was also present during growth on glycine betaine. Since it is a MtxB and would be the substrate-specific methyltransferase, we wanted to start to determine if it plays a role in glycine betaine metabolism. Our reasoning for the modeling was two-fold: 1) MV8460 is quite different than the Pyl-containing MttB found in M. vulcani B1d and therefore we wanted to compare how similar MV8460 is to the bacterial MtgB and 2) MV10350 is considered methanol-specific and therefore we wanted to determine through modeling if glycine betaine could dock into the active site of MV10350. The testing of each protein individually is advantageous as we build upon understanding how M. vulcani B1d is metabolizing glycine betaine.

59 Materials and Methods Cloning and Expression Vectors B1d genomic DNA (gDNA) was extracted following a phenol-chloroform method (Sambrook et al. 1989) and the gene encoding MV8460 was amplified using primers shown in Table 2. The PCR product was digested with SacII and XhoI (NEB) for 3 hours and treated with calf-intestinal alkaline phosphatase (CIP) (New England Biolabs) following manufacturer’s recommendations before ligation into pASK-IBA43(+) (IBA Life Sciences) using T4- (New England Biolabs) to yield pASK_MV8460. The manufacturer suggests a dilution of the T4-Ligase, however, we found increased success by not diluting the enzyme. The construct was maintained in Escherichia coli DH5α that were made competent by following the established TOP10 chemically competent cells that is available on OpenWetWare (https://openwetware.org/wiki/TOP10_chemically_competent_cells). The cloning of pASK_MV10335 followed similarly to pASK_MV8460 using different primers (Table 2). The expression vector pET28(+) (Addgene) was modified to have a larger multiple cloning site (MCS). A DNA fragment containing an E. coli BL21(DE3) codon optimized version of the gene encoding MV10360 was purchased from Genscript in pUC57. The fragment contained flanking restriction sites that would allow for removal of MV10360 from pUC57 (Fig. 16). The fragment was removed from pUC57 via restriction digest with NcoI and BamHI (NEB) for 3 hours and ligated into linearized CIP-treated pET28(+) yielding pET28_MV10360_Opt. The optimized MV10360 gene was then removed from pET28_MV10360_Opt using DraI and PmlI (New England Biolabs) which resulted in a linearized pET28(+) that now included the new MCS. The digest was separated via DNA gel electrophoresis followed by purification of the linearized plasmid using a Wizard® Plus SV Minipreps DNA Purification System (Promega) following manufacturer’s recommended protocol. The plasmid was then ligated onto itself, resulting in pETAC17a. The gene encoding MV10360 was amplified from M. vulcani B1d gDNA using a Phusion® High-Fidelity PCR Kit (New England Biolabs) following manufacturer’s recommended protocol before cloning into pETAC17a yielding pETAC_MV10360. Previous work with recombinant MtxAs has proved to be difficult.

60 Table 2 Primers and plasmids Primer name Plasmid Sequence Product pASK- ATGCCCGCGGAATGATAC MV8460 F IBA43plus CAAAATTCGATG pASK_MV8460 pASK- ATCCCTCGAGTTTTTTCAA MV8460 R IBA43plus TTCCGCAAACC GGCACTCGAGTAGGTGAC CAGTCCCAAAATGATTTTA MV8465 + AsiSI F pDL03c ATAAATTAAGGAGCGATC GCATATGGTTACACAGGA TGAAATTAATTC CGCTTGGAAGTACAGGTT pDLAC03_MV8465 MV8465 + Tev R pDL03c TTCACATTCACCTGCAGC TGCC GCATCCGCGGTCATTAGT GGTGGTGGTGGTGGTGC Tev + His R pDL03c GCTTGGAAGTA CAGGTTTTC pASK- ATGCCCGCGGAATGAAAA MV10335 F IBA43plus TTGGAGTTGCAATC pASK_MV10335 pASK- ATGCCTCGAGAGCTTCCT MV10335 R IBA43plus GTTCCATAACC ATGCCCGCGGAATGAACA MV10360 F pETAC17a TGAAAGAAAGATTACTC pETAC_MV10360 ATGCCTCGAGTTATTATG MV10360 R pETAC17a CGTAGTAATCGTCTC pETAC17 Seq F pETAC17a ATGCGTCCGGCGTAGAG pETAC17 Seq R pETAC17a GTTATTGCTCAGCGGTGG CACTAGTGATCTAGATGC pDL05c Seq F pDL05c ATG CGACGTTGTAAAACGACG pDL05c Seq R pDL05c G pDLAC03 + pDLAC05_MV8465 pDL05c

61 Figure 16 Gene sequence of the optimized gene encoding MV10360 from GenScript. The bolded areas are the flanking regions that make up the multiple cloning site that was generated following removal of MV10360 from pET28_MV10360_Opt using DraI and PmlI to generate pETAC17a.

62 CCATGGAACATCATCACCACCATCACGAAAACCTGTACTTCCAAGGCTACCGGT ACCGCGGACTGCAGTTTAAAATGAATATGAAGGAGCGTCTGCTGAAGGCGCTGAA AGGCGAAGAAGTTGACAAAGTTCCGGTTTGCACCGTTACCCAGAGCGGCATCGTT GAGCTGATGGACAAGACCGGTGCGAGCTGGCCGGAAGCGCACAGCGACGCGAAA ATGATGGCGGATCTGGCGTACGCGAGCTATGCGGAGTGCGGTCTGGAAGGCGTG CGTGCGCCGTACTGCCTGACCGTTCTGGCGGAGGCGATGGGTTGCACCATTAACA TGGGCACCAAGAACCGTCAGCCGAGCGTTACCGATCACCCGTATCCGAAAGGTGT GGACGATCTGGCGATGCCGGAGGACCTGCTGAGCCAAGGCCGTATCCCGGTGGT TATGGAAGCGCTGGGTATTCTGCGTGAGAAGTGCGGCGATGAAGTGCCGGTTATC GCGGGTATGGAAGGTCCGGTTACCCTGGCGAGCGACCTGGCGAGCGTGAAGAAA TTCATGAAATGGAGCATCAAGAAACCGGAGGATTTCCAGACCATTCTGGACTTTGC GTGCGATGCGTGCATCGAATATGCGAACGCGATGCTGGCGGCGGGTGCGGACGT GATTAGCGTTCCGGACCCGGTGGCGAGCCCGGACCTGATGGCGCCGGATGTGTT CGATAAGATTCTGAAACCGGTTCTGCAACGTTTTGCGGACGGTGTGAACGGCCCG ATGATCCTGCACGTGTGCGGCGATGTTACCGCGATCATTGAGATGATGGCGGACT GCCACTTCGAAAGCATCAGCATTGAGGAAAAGGTTAAAGATCTGAAGGGTGCGAA GGCGAAAGTGGGCGACAAATGCACCATCTGCGGTAACGTGAGCAGCCCGTTTGTT CTGCTGGCGGGTGATGAAGCGGCGGTTAAAGCGGCGGCGAAACAGGCGCTGGAT GATGGTATTGATGTGCTGGCGCCGGGTTGCGGCATCGCGCCGGATACCCCGGTT GCGAATCTGAAGGCGATGGTTGAAGCGCGTGATGACTACTATGCGTAACACGTGG GATCC

Figure 16

63 We wanted an alternative plan if MV10360 did not produce in E. coli BL21 using a clone that was generated from M. vulcani B1d gDNA. Therefore, during construction of the new pETAC17a vector with a larger MCS, we had DNA encoding MV10360 synthesized and optimized for codon utilization for production in E. coli BL21 as an intermediate step in construction of pETAC17a. Cloning of MV8465 into pDL05c followed established methods (Longstaff et al. 2007) using the MV8465+AsiSI F and MV8465+Tev R primers (Table 2). A second amplification using MV8465+AsisI F and Tev+His R added a 3ʹ hexahistidine tag. The fragment was digested with XhoI and SacII (New England Biolabs) and ligated into pDL03c yielding pDLAC03_MV8465 before cloning into pDL05c resulting in pDLAC05_MV8465 and was maintained in E. coli EC100 pir+ cells that were made chemically competent as stated previously. The E. coli EC100 pir+ cells were used because the pDL03c and pDL05c plasmids contain the pir-dependent R6K γ replication origin (Metcalf et al., 1997). The original design of the pDL05c required the initial step of cloning into pDL03c as the promoter region required for gene expression came from the pDL03c and was subsequently placed into pDL05c along with the gene. The construction of pDLAC05_MV8465 included a restriction site (AsiS1) directly upstream of the location of the gene and small MCS on the 3’ end. This was intended to allow for pDLAC05_MV8465 to be a tool where MV8465 could be removed and a new gene inserted, which would forgo the need to transform into pDL03c first. However, the AsiS1 enzyme is not entirely efficient and we were not able to clone into pDLAC05_MV8465 following its construction. Transformation of pDLAC05_MV8465 into M. acetivorans WWM73 followed closely to previously established polyethylene glycol transformation methods (Ladapo 1990; and Oelgeschläger and Rother, 2009).

Production of Recombinant Proteins Production of MV8460 was carried out using established methods with E. coli BL21 as the production strain (Ticak et al. 2014). A 3% (v/v) inoculum was used, and cells were grown at 37°C shaking at 250 RPMs to an OD600 between 0.4 - 0.5 and induced for protein production with anhydrous tetracycline (1 mg/L) (Sigma-Aldrich).

64 The anhydrous tetracycline was dissolved in ethanol with no adverse detriments to the E. coli or degradation of the protein being produced. The cells were incubated under the same conditions for 4 h and then harvested by centrifugation at 7,500 x g for 15 min at 4°C and the cell pellet stored at -80°C. E. coli ArcticExpress (Agilent) was used to produce MV10360, following manufacturer’s protocol with modifications. The E. coli ArcticExpress contains a ColE1-compatible, pACYC-based plasmid that constitutively expresses the genes that encode for the cold-adapted chaperonins Cpn10 and Cpn60 from Oleispria antarctica (Agilent, https://www.agilent.com/cs/library/usermanuals/Public/230191.pdf), which were intended to help with proper folding of MV10360. A starter culture of E. coli ArcticExpress cells harboring pETAC_MV10360 were grown overnight at 37°C shaking at 250 RPM in LB containing 20 µg/mL gentamycin for selection of pACYC-based plasmid and 50 µg/mL kanamycin for selection of pETAC_MV10360. A 3% (v/v) inoculum was then transferred to fresh LB that lacked antibiotics and allowed to incubate at 32°C shaking at 250 RPMs for 3 h. Following this initial growth period and prior to induction, cells were chilled statically at 4°C for 3 h with periodic mixing to help remove the trapped heat. Cells were induced with 1 mg/L isopropyl β-D-1- thiogalactopyranoside (IPTG) and incubated at 10°C shaking at 125 RPM for 24 h. The cells were harvested and stored using the same conditions as previously stated. Production of MV10335 was done under strict anoxic conditions using E. coli SG13009, a gift from Dr. Joseph Krzycki (The Ohio State University). E. coli SG13009 was recommended for production of proteins under anoxic conditions, based on empirical evidence in the Krzycki laboratory. Cells were grown in anoxic LB supplemented with sodium phosphate, pH 7.2, (44 mM) prior to autoclaving and glucose (80 mM) and fumarate (80 mM) after autoclaving. The medium was aseptically bubbled with pure N2 for 30 min, after sterilization, then the flask was quickly stoppered and secured with copper wire. Additionally, 160 mL serum bottles were prepared containing 42 mL of the sodium phosphate amended LB medium and made anoxic by stoppering the top followed by multiple gas/evacuation cycles using N2, then sterilized by autoclaving and then further amended with filter sterilized anoxic glucose and fumarate. Starter cultures [50 mL total or 5% (v/v)] were grown statically overnight at 34°C. The

65 starter cultures were transferred to the amended LB via needle and syringe and grown at 34°C shaking at 125 rpm to an OD600 between 0.3 – 0.4 and then supplemented with cysteine (5 mM) and ferrous ammonium chloride (0.1 mM) and allowed to incubate for 2 min. Production was induced with IPTG (1 mM) followed by a 6 h incubation at 34°C shaking at 125 RPM. Cysteine and ferrous ammonium chloride were added again using the same concentrations after the 6 h incubation and the culture was further incubated for 2 h. To achieve an anoxic harvest of the cells, 500 mL polycarbonate bottles with an O-ring cap assembly (Beckman) were placed into an anaerobic chamber (Coy Laboratory Products, Inc.) three days prior to use to ensure removal of oxygen from the plastic. Transfer of the culture into the anoxic bottles was performed in the anaerobic chamber, while balancing with water-filled aerobic bottles and subsequent centrifugation at 7,500 x g for 15 min was performed outside of the chamber. Spent media was decanted in the anaerobic chamber, the bottles were sealed, and the cell pellet was stored at -80°C until ready for protein isolation. Storage of MV10335 at -80°C helped keep the bottles anoxic until they were needed. MV8465 production was done in M. acetivorans WWM73 (Longstaff et al., 2007). Cells producing MV8465 were grown anaerobically in 1-L of high-salt mineral media.

The medium mixes A (400 mM NaCl, 45 mM NaHCO3, 13 mM KCl, 0.1% [v/v] resazurin,

DSMZ 141 trace element solution, pH 6.8) and B (54 mM MgCl and 2 mM CaCl2) were prepared and autoclaved separately, combined into a side-arm flask, made anoxic by a steady stream of sterile N2/CO2 (20% and 80%, respectively) gas, capped with a stopper, which was then secured with wire. This combined solution was then transferred to the anoxic chamber where it was amended with filter sterilized KH2PO4 (5 mM),pH 7.2, ammonium chloride (19 mM), and cysteine hydrochloride (2.8 mM) and allowed to incubate statically until the media turned colorless, which indicated complete reduction of the resazurin indicator. Once colorless, the media was then amended with Na2S (0.4 mM), anoxic methanol (62.5 mM), acetate (40 mM), ampicillin (100 µg/L), and puromycin (2 µg/L). A starter culture of WWM73 pDL05cMV8465 was grown statically at

37°C to OD600 ~0.5 and a 5% (v/v) inoculum was transferred to the 1-L of high-salt media. The culture was grown statically at 37°C to stationary phase with periodic

66 venting to reduce pressure buildup. Cells were harvested aerobically and stored as stated above.

Protein Purification Purification of His-tagged MV8460, MV10360, and MV8465 followed closely to established methods (Ticak et al., 2014), with additional steps. Cells were lysed via French press at 20,000 psi and soluble lysate generated by centrifugation at 250,000 x g at 4 °C for 1.5 h, and then passed through a 0.22 µm syringe filter to remove any remaining cellular debris. All proteins were partially purified using a 1 mL HisTrap™ HP using an ÄKTA Prime Plus (GE Healthcare) that was located inside of an anaerobic chamber (Coy Laboratories, Inc). Lysates were prepared aerobically and loaded onto the 1 mL columns and equilibrated with a mix of anoxic 95% Buffer A (50 mM sodium phosphate, 500 mM NaCl, 40 mM imidazole, pH 7.2) and 5% Buffer B (50 mM sodium phosphate, 500 mM NaCl, 500 mM imidazole, pH 7.2) (Ticak et al. 2014) for 10 column volumes and subjected to a 50-mL linear gradient from 5% to 100% Buffer B at 1 mL/min. MV8460, MV8465, and MV10360 each eluted as single peaks at 80-175 mM, 85-145 mM, and 65-145 mM imidazole, respectively. The HisTrap™ HP columns were then washed with 10 column volumes of 100% Buffer B to ensure no remaining proteins were bound. Protein peaks were pooled and diluted 1:5 with Buffer C (20 mM MOPS, 0 M NaCl, pH 7.8). Samples were loaded onto 1-mL Bio-Scale™ Mini UNOsphere™ Q cartridges (BIO-RAD) for an additional purification step. The protein-bound columns were equilibrated with 10 column volumes of Buffer C followed by a 0% to 70% Buffer D (20 mM MOPS, 1 M NaCl, pH 7.8) linear gradient. MV8460, MV8465, and MV10360 each eluted as single peaks at 330-420 mM, 240-430 mM, and 300-500 mM NaCl, respectively. MV8465 required an additional third purification step. MV8465 was diluted 1:5 with Buffer E (5 mM sodium phosphate, pH 7.0), loaded onto a 5-mL Bio-Scale Mini CHT Type I Cartridge (BIO-RAD), and the column equilibrated with 5 column volumes of Buffer E followed by a 0% to 100% Buffer F (500 mM sodium phosphate, pH 7.0) linear gradient following the manufacturer's protocol. MV8465 eluted from the CHT column in a single peak at 185-210 mM phosphate. Protein purity was assessed via SDS-PAGE (BIO-RAD) followed by Coomassie blue staining (Fig. 17). Protein concentrations were

67 determined by the Bradford protein assay (Bradford 1976) and frozen at -20 °C without glycerol. Proteins were used within 6 months of freezing. Materials needed for MV10335 purification were made anoxic and stored in the anaerobic chamber three days prior to use. The frozen cell pellet was quickly moved to the anaerobic chamber and resuspended in anoxic Buffer A and lysed via an anoxically- adapted French press cell at 20,000 PSI. The anoxic lysate was spun at 250,000 x g at 4°C for 1.5 h, and then passed through a 0.22 µm syringe filter. MV10335 was purified as described above requiring both a 1 mL HisTrap™ HP and a 1-mL Bio-Scale™ Mini UNOsphere™ Q cartridge. MV10335 eluted as a single peak at 77-175 mM imidazole and as a single peak at 280-420 mM NaCl. Relevant fractions were pooled and then filtered into an anoxic Wheaton serum bottle. MV10335 degrades upon freeze-thaw cycles and therefore was stored at 4°C and used within five days after purification.

Homology Modeling Prediction of MV8460 and Molecular Docking of Glycine Betaine The apo-structure of DhMtgB chain A (PDB – 2QNE) was used as a template to generate two models of MV8460 using both MODELLER (Webb and Sali, 2017) and I- TASSER (Zhang, 2008) for comparison. MetaPocket 2.0 (Huang, 2009) was used to predict possible (s) for glycine betaine in DhMtgB and MV8460. Probable binding pockets and molecular docking predictions were performed using AutoDockTools and AutoDockVina, respectively (Trott and Olson, 2010). Models with glycine betaine docked in the active site were visualized with PyMol v2.3. Aligned models were inspected for residues within 4–5 Å from glycine betaine to identify proposed active site signatures, generated with WebLogo3 (Sharma et al., 2012).

68 Figure 17 A 12% acrylamide SDS-PAGE gel followed by Coomassie blue staining of purified recombinant proteins used to reconstitute the glycine betaine:CoM methyl transfer pathway. Lane 1 is the molecular weight ladder. Lane 2 is MV8465 at ~28 kDa. Lane 3 is MV10360 at ~37 kDa. Lane 4 is MV8460 at ~53 kDa. Lane 5 is MV10335 at ~59 kDa. Lanes 2-5 each contained approximately 3 µg of protein. The image was digitally modified (seen between lanes 3 and 4) to remove a lane that consisted of a replicate of MV10360.

69

Figure 17

70 Homology modeling prediction of MV10345 and molecular docking of methanol and glycine betaine. The predicted methanol methyltransferase, MV10345, amino acid sequence was threaded using I-TASSER (Zhang, 2008) to the methanol methyltransferase portion of the methanol:cobalamin methyltransferase complex, MtaBC (PDB - 2I2X) (Hagemeier, et al. 2006). The predicted zinc-bound, and cobalamin-bound forms of MV10350 were aligned to 2I2X, chain A with the cognate corrinoid ligand of MtaC, zinc ions, and predicted potassium ion. The proposed active site of the Methanosarcina barkeri MtaB (Hagemeier, et al. 2006) was used to guide the docking studies with AutoDockVina (Trott and Olson, 2010) in MV10345. All structures were visualized with PyMol v2.3, both separately, and aligned with the M. barkeri MtaB.

MV8460 and MV10360 Methyltransferase Activity Assays monitoring the activity of MV8460 followed closely to previously described methods (Ticak et al., 2014). An initial reaction mixture consisting of 50 μg of recombinantly purified MV8460 (950 pmol), hydroxycobalamin (1.75 mM), and Ti(III) citrate (16.25 mM) in Mops (50 mM), pH 6.5, totaling 585 uL was added to an anaerobic

0.2-cm quartz cuvette that was under a steady stream of ultra-high purity N2 gas. A HP 8453 photodiode array UV-Visible Spectrophotometer (Agilent Technologies) instrument was blanked against a reaction mixture where Mops buffer was substituted for the hydroxycobalamin. Complete reduction of the hydroxycobalamin to cob(I)alamin was monitored spectrophotometrically by measuring absorbance changes at 540 nm and 578 nm. Cob(I)alamin and cob(II)alamin share an isosbestic point at 540 nm, and cob(I)alamin and methylcob(III)alamin share an isosbestic point at 578 nm (Ferguson et al., 2011). The assay was initiated by the addition of 15 uL of 2 M glycine betaine (50 mM final) to a final reaction volume of 600 uL. Conversion of reduced cob(l)alamin to methylcob(III)alamin was monitored by measuring the absorbance at 540 nm every 10 s for 450 s. The controls consisted of omitting MV8460 or glycine betaine; and using TMA, QMA, or choline as methyl donors. All assays were performed under dim red light to avoid photolysis of the methyl-cobalt bond (Ferguson et al., 2011).

71 MV10360 activity assays followed similarly to described methods under red light (Ferguson et al., 2011). Briefly, a 0.2-cm quartz cuvette was stoppered and made anoxic. An anoxic reaction mixture of methylcob(III)alamin (0.5 mM) in phosphate buffer, pH 7.2, was added to the cuvette to a volume of 595 uL along with 5 uL of 1 M CoM (8.3 mM final) and incubated at 37°C for 10 min. A cuvette containing only phosphate buffer, MV10360, and CoM was used to blank the Spectronic 20D+ Spectrophotometer (Thermo Scientific). The reaction was initiated by addition of 50 μg MV10360 and incubated at room temperature while monitoring continuously at 540 nm until there was no detectable absorbance change. Demethylation of methylcob(III)alamin will yield either cob(I)alamin or cob(II)alamin both of which share the isosbestic point at 540 nm (Kreft and Schink, 1994). The controls consisted of omitting either MV10360 or CoM.

Reductive Activation of MV8465 by MV10335 Followed by Methylation of MV8465 by MV8460 Monitoring reduction of MV8465 was performed under strict anaerobic conditions following previously described methods (Ferguson et al., 2009). Reaction mixtures were assembled in stoppered anoxic 0.2-cm quartz cuvettes containing ATP (12.5 mM),

MgCl2 (25 mM), Ti(III)-citrate (4 mM), and MV8465 (800 μg/ml) in 50 mM MOPS, pH 7.2. A cuvette containing all components excepts MV8465 was used to blank a HP 8453 photodiode array UV-Visible Spectrophotometer (Agilent Technologies). Reactions were incubated at 37 °C monitoring change at 378 nm until the spectrum stabilized allowing the Ti(III)-citrate to partially reduce MV8460. MV10335 (216.5 μg/ml) was added and the 400 uL reaction was incubated for 45 min monitoring changes at 378 nm and 475 nm every 2 min. Upon stabilization of the spectrum, 3.75 μl each of MV8460 (144 μg/ml) and glycine betaine (18.75 mM) were added and the cuvette incubated for 30 min, taking measurements every 2 min. The controls consisted of omitting MV8465, MV8460, or MV10335. Methylation of MV8465 was only achievable when all components were present and glycine betaine was used as the methyl donor. Other methyl donating compounds tested included TMA, QMA, and choline.

72 Results Structural Modeling of MV8460 Based on the draft M. vulcani B1d genome, MV8460 is the sole Pyl-lacking MttB homolog found in M. vulcani B1d, which is 65% identical and has 83% sequence similarity to DhMtgB. We therefore generated models of MV8460 using the apo-crystal structure of DhMtgB (PDB – 2QNE) as a template and compared the structures and predicted active sites of the two enzymes (Fig. 18). The model of MV8460 generated using I-TASSER was the most accurate [C-score (2), TM-score (0.99 ± 0.04), and RMSD (2.9 ± 2.1 Å)] when compared to 2QNE. The structural motifs of DhMtgB and MV8460 are α/β TIM-barrel folds, much like other methyltransferase enzymes (Hao et al., 2002; Hagemeier et al., 2006). MetaPocket 2.0 (Huang, 2009) highlighted an eight β-sheet orientation in the center of the enzyme which generates a deep funnel that was indicated as a possible region for glycine betaine interaction. Glycine betaine was docked at this location in DhMtgB with a predicted kCal/mol of −3.9 and MV8460 with a predicted kCal/mol of −3.7 and their global structures overlaid (Figures 18 A,B). The binding sites of the overlapped structures were analyzed for conserved or semi-conserved residues reported to bind glycine betaine and an active site logo was generated (Fig. 18 C). Various crystal structures of known glycine betaine-binding enzymes; 1R9L (Schiefner et al., 2004), 6EYG, 3TMG (Li et al., 2015), 1SW2 (Li et al., 2015), 2B4L (Horn et al., 2006), 4MJW (Salvi et al., 2014), and 3L6H (Wolters et al., 2010); not related to the COG5598 superfamily, were used to validate the modeled glycine betaine ligand. The predicted binding site of glycine betaine in DhMtgB and MV8460 is most comparable to the glycine betaine-bound 4MJW crystal. The S101, H466, and N510 residues in 4MJW coordinate the carboxyl moiety of glycine betaine for cation-π interactions with the surrounding aromatics. This suggests that four predicted active site residues may interact with glycine betaine as they are conserved between DhMtgB and MV8460: Y94, N199, R309, and H345.

73 MtaB Modeling Due to the presence of MV10350 during growth on glycine betaine (Fig. 10 and 11), we modeled MV10350 to determine if glycine betaine could fit into the catalytic active site (Fig. 19). MV10350 shares 72.77% sequence identity and 99% query coverage with M. barkeri Fusaro MtaB (UniProtKB – Q46EH3). I-TASSER generated a homology model of MV10350 with a C-score (2), TM-score (0.99 ± 0.03), and RMSD (2.5 ± 1.9Å) to M. barkeri MtaB (PDB – 2I2X) (Hagemeier et al., 2006) (Fig. 19 A). We used the proposed active site motif of MtaB (Fig. 19 B) for predictive docking of methanol and glycine betaine to MV10350. The hydroxyl-group of methanol is located either near zinc, C219 or E312 in MV10350 (Hagemeier et al., 2006). Attempts to dock glycine betaine resulted in poor ligand positioning to cobalamin or steric hindrance due to zinc or the proposed potassium ion. Key residues involved with cation-π interactions for quaternary amine interactions are also lacking in MV10350 and therefore we could not accurately model glycine betaine into the active site of this enzyme.

MV8460 Glycine Betaine:Cob(l)alamin Activity We hypothesized that the MV8460 was responsible for initiating methanogenesis from glycine betaine by catalyzing the corrinoid-dependent demethylation of the substrate, analogous to the function of DhMtgB (Ticak et al., 2014). We detected methylation of free cob(I)alamin by MV8460 using glycine betaine as the methyl donor, exhibited by an increase at 540 nm (Fig. 20). Cob(I)alamin and methylcob(III)alamin share an isosbestic point at 578 nm which is disrupted in the presence of cob(II)alamin (Kreft and Schink, 1993; Ticak et al., 2014). Inadvertent oxidation of cob(I)alamin can cause a false positive due to the formation of cob(II)alamin, causing increases at 540 and 578 nm (Kreft and Schink, 1993; Ticak et al., 2014). Absorbance at 578 nm remained unchanged throughout the assay, suggesting a direct conversion of cob(l)alamin to methylcob(III)alamin (Fig. 20 B). The specific activity of the recombinant MV8460 under the conditions tested was 0.21 μmol min–1 mg–1. Activity was not detected when choline, TMA, or QMA were used as methyl donor substrates.

74 MV8465 Reduction by MV10335 Followed by Methylation of MV8465 by MV8460 In Methanosarcina barkeri, reduction of the corrinoid binding proteins in the TMA, DMA, and MMA pathways relies on the RamA enzyme, and likewise, reduction of the corrinoid binding proteins in the methanol pathway relies on RamM (Ferguson et al., 2009). MV8465 displayed the characteristic UV-visible spectrum of a corrinoid binding protein (Fig. 21A) (Ferguson and Krzycki, 1997). We tested the ability of MV10335 to reduce the bound corrinoid of MV8465 from a Co(II) to a Co(I) state. MV8465 was partially reduced to the Co(II) form with excess Ti(III)-citrate to a stabilized absorbance spectrum, consistent with the corrinoid being in the Co(II) state (Kreft and Schink, 1993) (Fig. 21). The addition of MV10335 resulted in a significant increase at 386 nm, indicative of the Co(I) state. Following reduction of MV8465 by MV10335 we added MV8460 and glycine betaine as the methyl donor, which resulted in an observed decrease at 386 nm with a concomitant increase at 540 nm (Fig. 21 B).

MV10360 Methylcob(III)alamin:CoM Activity The penultimate step to methanogenesis is the methylation of CoM, and achievement of this step in the TMA degradation pathway is through methylcob(III)alamin:CoM methyltransferase (MtxA) (Ferguson et al., 1996; Ferguson and Krzycki, 1997). CoM methylase activity of MV10360 was confirmed by monitoring change at 540 nm. The specific activity of the recombinant MV10360 under the conditions tested was 1.6 μmol min–1 mg–1. Cob(II)alamin is generated during methylcob(III)alamin:CoM methyl transfer, causing a decrease at 540 nm (Fig. 22) (Ferguson et al., 2011).

75 Figure 18 Active site predictions of DhMtgB and MV8460. (A) The models represent the aligned global structure of DhMtgB (green) docked to glycine betaine (green mesh) and homology model of MV8460 (cyan) docked to glycine betaine (cyan mesh). (B) The proposed interacting residues of both DhMtgB and MV8460 are shown within the TIM- barrel funnel which was highlighted with MetaPocket 2.0 (B 2009). Black labeled residues indicate those shared between both structures while divergent amino acids are coded by either green (DhMtgB) or cyan (MV8460). (C) Active site logo generated after aligning both DhMtgB and MV8460 residues within 4-5 Å of the proposed docking site of glycine betaine were highlighted and aligned from N-terminus to C-terminus. Yellow arrows indicate aromatic residues which may interact in cation-π interaction with the methyl moiety of glycine betaine.

76

Figure 18

77 Figure 19 Predicted structural model of MV10350 compared to known MtaB from Methanosarcina barkeri Fusaro. (A) The global monomeric structure of MtaB (green) and predicted structure of MV10350 (cyan) are aligned. MtaB is bound with two zinc ions (gray), a predicted potassium ion (purple), and the cognate corrinoid ligand of MtaBC complex (red), while MV10350 is bound to zinc (grey) and cobalamin (tv-red). (B) The active site residues for zinc-binding are shown for MtaB (green) and MV10350 (cyan) relative to the cobalamin ligand (red, MtaB; tv-red, MV10350). Molecular docking of MV10350 for methanol using AutoDockVina (Trott and Olson, 2010) is highlighted with a black circle.

78 Figure 19

79 Figure 20 Glycine betaine:cob(I)alamin methyl-transfer activity of MV8460. Over a period of 450 s, readings were taken every 10 s using a HP 8453 photodiode array spectrophotometer to monitor conversion of cob(I)alamin to methylcob(III)alamin. (A) Each spectrum represents a separate time point. The black bold arrows indicate that as the reaction progresses the absorbance increases at 540 nm while decreasing at 620 nm. Methylation of cob(I)alamin results in an increase at 540 nm and decrease at 620 nm. (B) No apparent changes at 578 nm with increases at 540 nm suggests direct conversion of cob(I)alamin to methylcob(III)alamin.

80

Figure 20

81 Figure 21 Reductive activation of MV8465 by MV10335. In order for methylotrophic methanogenesis pathways to function, the corrinoid binding protein must be in the active Co(I) state. MV10335 functions to reduce MV8465 to this active state. Ti(III)- citrate serves as an electron source for MV10335 but can also reduce MV8465 to Co(II), and therefore was added prior to the addition of MV10335. (A) Each spectrum is a time point with the black arrows indicating an increase absorbance at 378 nm and a decrease in absorbance at 478 nm, which is characteristic of reduction from Co(II) to Co(I). (B) MV8465 was initially reduced to Co(II) by Ti(III)-citrate to a stabilized spectrum (black line) and was then reduced to a Co(I) by the addition of MV10335 (red line). The major peak seen at 386 nm is characteristic of a Co(I) species. Addition of MV8460 and glycine betaine resulted in a decrease at 386 nm and an increase at 540 nm (blue line) indicating the formation of methylcob(III)alamin.

82

Figure 21

83 Figure 22 Methylcob(III)alamin:CoM methyl-transfer activity by MV10360. The reaction contained 0.5 mM methylcob(III)alamin, 50 mM phosphate buffer at pH 7.2, 5 mM CoM, 40 µg MV10360, and was performed anoxically under dim red light. Readings were taken every 15 s until completion. Demethylation of methylcob(III)alamin was measured by a decrease in absorbance at 540 nm. Error bars represent standard deviations (n = 3).

84

Figure 22

85 Discussion Until recently, the only characterized Pyl-lacking COG5598 homolog was from D. hafniense Y51 and it functions to demethylate glycine betaine (Ticak et al., 2014). The high similarity between MV8460 and DhMtgB coming from different domains of life piqued our interest to search for conserved amino acids and model DhMtgB and MV8460 for basic analysis of the catalytic pocket and docking of glycine betaine (Fig. 18). Interestingly, in many of the crystal structures of enzymes which bind glycine betaine or other quaternary amines (Schiefner et al., 2004; Horn et al., 2006; Wolters et al., 2010; Salvi et al., 2014; Li et al., 2015), the methyl moiety is commonly flanked with aromatic compounds which are involved in cation-π and π-π stacking. The residues likely involved with either cation-π or π-π stacking are Y94 and F353/Y353 while the interaction of H345 is more complex. Given physiological pH and the pKa of histidine, H345 is likely protonated and acts as a hydrogen donor to the carboxyl moiety of glycine betaine much like 4MJW (Salvi et al., 2014). Additionally, given the distance (4.1 Å, DhMtgB; 4.6 Å, MV8460) and the positioning of H345 it is unlikely to interact in cation-π with the methyl moiety. It is more likely that H345 would be involved in π-π stacking with the nearby F353/Y353 if H345 was unprotonated. The position of glycine betaine within the predicted funnel would also involve R309 for positioning of the carboxylic moiety in both DhMtgB and MV8460 for helping coordinate the molecule and methyl group for catalytic attack by the Co(I) species from MV8465. This positioning of the methyl group for catalytic attack is analogous to what has been seen previously in MtaBC (Hagemeier et al., 2006). Amongst the compounds tested, MV8460 appeared to only interact with glycine betaine, which is consistent with previous work on methylotrophic methyltransferases (Burke and Krzycki, 1995; Ferguson and Krzycki, 1997; Sauer et al., 1997; Sauer and Thauer, 1998; Ferguson et al., 2000; Tallant et al., 2000; Pritchett and Metcalf, 2005; Ticak et al., 2014). In-depth analysis of the biochemical interactions between MV8460 and MV8465 was beyond the scope of this work and therefore we did not determine if MV8460 could interact with another corrinoid binding protein from a different methylotrophic pathway. However, productive interaction of methylotrophic methyltransferases with non-cognate corrinoid binding proteins has never been

86 reported, to our knowledge. Therefore, interaction of MV8460 with another corrinoid protein seems unlikely. As previously stated in section 1.1 during the proteomic analysis, when comparing abundances of MV8460 and MV10350, MV10350 is at a statistically greater abundance during growth on glycine betaine (Fig. 12). Therefore, we considered the possibility that MV10350 could be involved with glycine betaine metabolism. We modeled MV10345 to determine if glycine betaine could fit into the catalytic active site. MV10350 and MV8460 are not homologous and our modeling data of MV10350 using the crystal structure of the M. barkeri MtaB (Hagemeier et al., 2006) as a guide, suggests that glycine betaine would not fit into the catalytic pocket of MV10350. The MtaB from M. barkeri Fusaro forms a tight complex with MtaC and supporting evidence suggests this complex is only active for methanol degradation (Sauer et al., 1997; Sauer and Thauer, 1998; Pritchett and Metcalf, 2005; Hagemeier et al., 2006). Given this knowledge, the MtaB from M. vulcani B1d was not produced and isolated for functionality testing. Further supporting evidence for why we think MtaB is unlikely to be involved with glycine betaine metabolism will be discussed in the following Section 1.3.

87 Chapter 1.3 In vitro reconstruction of the glycine betaine:CoM methylotrophic pathway from Methanolobus vulcani B1d

Overview Section 1.3 is focused on in vitro reconstitution of the glycine betaine:CoM methyl transfer pathway using recombinant MV8460, MV8465, MV10335, and MV10360. Our findings from performing these assays were then complemented with additional activity assays from crude extracts of M. vulcani B1d cells grown either on glycine betaine, TMA, or methanol. One finding we found striking was the lack of activity of methanol:CoM activity in the extracts. To address this issue, we performed methanogenesis assays with whole cells to help continue to understand if the MtaB (MV10350) may be interacting with glycine betaine. We also expanded the MttB COG5598 superfamily of enzymes phylogenetic tree to show where MV8460 clusters, relative to other Pyl-MttBs encoded from organisms that have been reported to grow anoxically using glycine betaine as the sole carbon source. At the conclusion of this project we were able to support our hypothesis that M. vulcani B1d performs methanogenesis from glycine betaine through a corrinoid-dependent methyltransferase pathway that is initiated by the Pyl-lacking COG5598 enzyme MV8460.

88 Materials and Methods In vitro reconstitution of glycine betaine:CoM methyl transfer pathway Testing for functionality of individual proteins that were used for in vitro reconstitution of the glycine betaine:CoM pathway followed similarly to established methods and are summarized in Section 1.2. All assays were performed as previously described (Ferguson et al., 2009; Ferguson et al., 2011; Ticak et al., 2014), with minor modifications. A 5x reaction mixture containing CoM (15 mM), ATP (62.5 mM), and

MgCl2 (125 mM) was prepared in 50 mM MOPS, pH 7.2. The assay was performed in a stoppered anoxic quarts cuvette with a final reaction volume of 250 µL. Ti(III)-citrate (Seefeldt and Ensign, 1994; Ferguson et al., 2011) amended MOPS was added followed by MV8460 (10 µg/190 pmol), MV10360 (5 µg/139 pmol), MV8465 (5 µg/174 pmol), MV10335 (5 µg/85 pmol), and then the reaction mixture was added. Ten µL of Ti(III)-citrate (~4 mM) was added as a source of reducing potential and the cuvette incubated at 37°C for 45 min. The reaction was initiated by adding 10 µL of the methyl donor (16 mM). Samples were taken every 4 min for 24 min and mixed with Ellman’s reagent (Ellman, 1958) in a 96-well round bottom plate and measured at 412 nm using a MolDev FilterMax F5 plate reader. Ellman’s reagent [5,5’-dithiobis-(2-nitrobenzoic acid) or DTNB] reacts with free thiol groups causing a yellow coloration that can be measured at absorbance 412 nm to determine the concentration of free thiol. CoM contains a thiol group that is methylated by the MtxA enzyme, and if methylated, the reaction between CoM and Ellman’s reagent cannot occur and therefore coloration measurements at later timepoints were not as intense, which indicates a decrease in the amount of free thiol, due to CoM methylation. The lower limit of detection for the free thiol on CoM with Ellman’s reagent, in our hands, is a loss of 0.3 mM free thiol over the course of a 40 min assay. The controls to determine the necessity for each protein consisted of assays that either lacked MV8460, MV8465, MV10360, or MV10335. Additionally, other controls consisted of using TMA, choline, or QMA as the methyl donors, and performing the assay aerobically. Each protein was required for reconstitution of the pathway and successful methylation of CoM was only achieved when glycine betaine was the methyl donor.

89 Methanolobus vulcani B1d substrate dependent CoM methylation in extracts M. vulcani B1d cells were grown either on glycine betaine (80 mM), methanol (62.5 mM), or TMA (40 mM) on brackish media and anoxically harvested during mid-log phase (Ticak et al., 2015) (Introduction: Fig. 7). Cells were resuspended with anoxic Mops (50 mM), pH 7.2, and anoxically lysed, as described in Section 1.2. Lysed cells were centrifuged, and lysates were filtered into sterilized anoxic Wheaton serum bottles, and the headspace was exchanged with H2. The assay followed similarly to previously established methods (Ferguson et al., 2011). Stoppered anoxic quarts cuvette were prepared as previously described except under a headspace of H2 instead of N2. An assay mixture of ATP (62.5 mM), MgCl2 (125 mM), CoM (30 mM), and bromoethanesulfonate (BES) (20 mM) in Mops buffer (50 mM), pH 7.0, was added to the cuvettes, with a starting assay volume of 25 μl. The BES serves to inhibit the MCR that would naturally be present in the crude extracts. A false negative will occur without the addition of BES because MCR will cause the methyl group from CoM to be released in the form of methane which would generate a free thiol that can react with Ellman’s reagent. To the mixture, 90 ul of crude cellular extract was added followed by 5 μl Ti(III)- citrate (~4 mM). The reaction was initiated by addition of 5 ul of the methyl-donor (400 mM) to a final volume of 125 μl. Reactions were then incubated at 37°C with 3 ul samples taken every 4 minutes and added to 250 μl of Ellman’s reagent (0.5 mM).

Ti(III)-citrate (~4 mM) and a headspace of H2 were both required for substrate demethylation activity. The controls consisted of omitting the methyl-donor substrate along with exposing the assays to oxygen.

Sequence acquisition and phylogenetic construction of MttB superfamily. The DhMtgB amino acid (aa) sequence (UniProtKB–Q24SP7) was used as a query with PSI-BLAST (Altschul et al., 1997) with a 1E-10 cutoff, for the non-redundant protein sequence (nr) database. Similarly, the MttB of M. barkeri (UniProtKB – O93658) was used to acquire Pyl-encoding MttB sequences via tBLASTn (Gertz et al., 2006). Sequences varying more than one standard deviation from the mean aa length and those with identities less than 30% of the query were removed. The remaining 5517 sequences were filtered using CD-Hit (Li and Godzig, 2006) to remove those with

90 greater than 90% identity providing a final dataset of 2356 sequences averaging 504 aa. MV8460 and MttB10 (UniProtKB – H6LKF8) were added to the dataset from B1d and Acetobacterium woodii, respectively. The dataset was aligned with MUSCLE (Edgar, 2004) using default settings. Phylogenetic analysis was performed using approximately- maximum-likelihood with FastTree 2 (Price et al., 2010), using JTT+CAT (Jones et al., 1992), WAG+CAT (Whelan and Goldman, 2001), and LG+CAT (Le and Gascuel, 2008), with and without Gamma distribution. Phylogenetic trees were generated using the interactive Tree of Life (iTOL) (Letunic and Bork, 2016).

Methanogenesis Assays Methanogenesis assays were performed similarly as previously described (Tallant and Krzycki, 1997). Briefly, M. vulcani B1d was grown on either glycine betaine, TMA, or methanol to mid-log phase as previously described and then chilled for 30 min. The cells were centrifuged at 8,500 x g for 20 min at 4°C in an anoxic centrifuge bottles (Beckman). Cells were washed with anoxic ice-cold 50 mM MOPS buffer, pH 7.2, amended with NaCl (250 mM) and centrifuged again. The cells were resuspended in 4 mL of the same buffer and evenly divided into four 13.1 mL anoxic-Wheaton serum bottles. To the bottles, either glycine betaine, TMA, or methanol was added (25 mM) and incubated in a 37°C water bath shaking at 125 rpm. Methane was sampled every 6 min and ran at 5 mL/min through a TG-BOND Q 30 mm x 0.53 mm x 20 µm column into a Flame Ionization Detector (FID) that was attached to a Trace 1300 Gas Chromatograph Split/Splitless Injector instrument (Thermo Scientific). Controls consisted of omitting the methyl-donor substrate.

91 Results Phylogenetic tree and sequence acquisition We generated an updated COG5598 phylogenetic tree (Fig. 23), from our prior analysis (Ticak et al., 2014). The WAG+CAT (Whelan and Goldman, 2001) tree provided the fewest bad-splits (17/2358) and the best log-likelihood (-1119437.408) with a Δ log-likelihood (7.948). The clade that contains the bona-fide MtgB from D. hafniense Y51 (DhMtgB) was analyzed for organisms reported to grow anaerobically with glycine betaine. The majority of the reported organisms capable of anaerobic GB-dependent growth from the clade are Clostridiales, with only two archaeal members present: Methanococcoides vulcani SLH33(T) and B1d (Muller et al., 1981; Möller et al., 1984; Dehning et al., 1989; Finster et al., 1997; Kuhner et al., 1997; Nielsen et al., 2006; Sattley and Madigan 2007; Sikorski et al., 2010; L'Haridon et al., 2014; Ticak et al., 2014; Poehlein et al., 2015; Ticak et al., 2015; Lechtenfeld et al., 2018)

Glycine betaine:CoM reconstitution and methanogenesis assays Following enzymatic confirmation of individuals enzymes activities (MV8460, MV8465, MV10335, and MV10360) predicted to be involved in the proposed glycine betaine:CoM pathway, we tested our hypothesis that glycine betaine-dependent methanogenesis occurs through a corrinoid-dependent methyl transfer pathway initiated via MV8460. We successfully reconstituted glycine betaine:CoM methyl transfer using highly purified recombinant proteins (Fig. 24). MV8460, MV8465, MV10335, and MV10360 are each required for methylation of CoM. No CoM methylation was detected when using choline, QMA, or TMA as methyl donors.

92 Figure 23 Approximate-maximum likelihood representation of the COG5598 MttB superfamily. The above phylogenetic tree (A) shows the proposed evolutionary relationship between both trimethylamine methyltransferases (dark green, archaeal; light green, bacterial) and glycine betaine methyltransferases (red). A portion of the glycine betaine clade is expanded (B) with bootstrap values (greater than 80) being positioned at nodes. Bold names represent members previously reported to utilize glycine betaine while red names indicate biochemically demonstrated enzymes, accession numbers for the enzymes are in parentheses. The scale bars represent amino acid substitutions per site in both (A,B).

93

Figure 23

94 Figure 24 Reconstitution of glycine betaine:CoM activity in vitro with purified recombinant proteins. Loss of the free thiol group on CoM was monitored at 412 nm, using Ellman’s reagent. Addition of glycine betaine resulted in a significant decrease in the amount of HS-CoM that is detectable (blue diamond), indicating an intact glycine betaine:CoM methyl transfer pathway. Methylation of CoM was not detectable when TMA (red square) or choline (gray triangle) served as the methyl donors. The assays contained 10 μg of MV8460 and 5 μg each of the remaining proteins. Assays in which MV8460, MV8465, MV10335, or MV10360 were omitted lacked glycine betaine:CoM activity (data not shown). Error bars represent standard deviations (n = 3). TMA = trimethylammonium

95

Figure 24

96 To confirm that the activity we observed was not an artifact of recombinant proteins, we performed in vitro glycine betaine:CoM activity assays using crude extracts from B1d (Fig. 25). Extracts from glycine betaine-grown cells methylated CoM when glycine betaine was the methyl donor, at a rate of 0.27 umol•min-1•mg-1, but not with TMA or methanol. Extracts from TMA-grown cells methylated CoM when TMA was the methyl donor, at a rate of 0.31 µmol•min-1•mg-1, but not with glycine betaine, choline, or methanol. Surprisingly, CoM methylation was undetectable in extracts from methanol- grown cells using methanol, TMA, or glycine betaine as methyl donors, despite the extracts having been prepared identically to extracts of glycine betaine- or TMA-grown cells. The lower limit of detection for this assay was a loss of 0.3 mM free CoM over the course of a 40 min assay. Extracts prepared from M. vulcani B1d cells grown on glycine betaine, TMA, or methanol lacked detectable methanol:CoM activity, we therefore tested for methanogenesis from prepared live cells, following previously established methods (Tallant and Krzycki, 1997). Washed cells grown on glycine betaine, TMA, or methanol showed rapid methane production when provided the same substrate, but a lag was detected if the cells were provided a different substrate (Fig. 26).

97 Figure 25 Glycine betaine:CoM activity in vitro using crude extracts. Loss of the free thiol group on CoM was monitored using Ellman’s reagent at 412 nm. (A) Addition of glycine betaine to crude extracts from glycine betaine grown B1d cells resulted in a significant decrease in the amount of detectable HS-CoM (blue triangle), indicating an intact glycine betaine:CoM methyl transfer pathway. Detection of HS-CoM persisted when TMA was used as a methyl donor (red square), indicating a lack of a TMA:CoM methyl transfer pathway. (B) Addition of TMA to crude extracts from TMA grown M. vulcani B1d cells resulted in a significant decrease in the amount of detectable HS-CoM (red square), indicating an intact TMA:CoM methyl transfer pathway. Detection of HS-CoM persisted when glycine betaine was used as a methyl donor (blue triangle), suggesting a lack of a glycine betaine:CoM methyl transfer pathway. Error bars represent standard deviations (n = 3). No detectable decrease in HS-CoM was observed in any crude extracts when using methanol as the methyl donor, including from cells that were grown on methanol (data not shown). GB = glycine betaine, TMA = trimethylammonium

98

Figure 25

99 Figure 26 Representative figure of methanogenesis assays performed on M. vulcani B1d after initial growth either on (A) TMA, (B) glycine betaine, or (C) methanol. The cells were anoxically pelleted, washed with 50 mM Mops buffer, and then supplemented either TMA (blue line), glycine betaine (grey line), or methanol (orange line) and then incubated at 37 °C for 24 min with methane readings being taken every 6 min. Assays were performed multiple times but methane percentages were not similar, however, the trend observed in this representative figure was the same for each assay.

100 A

0.40 0.30 0.20

0.10 Methane (%) Methane 0.00 0 6 12 18 24 30 Time (min) B

0.2 0.18 0.16 0.14

Methane (%) Methane 0.12 0 6 12 18 24 30 Time (min)

C

0.18 0.17 0.16 0.15 0.14

Methane (%) Methane 0.13 0 6 12 18 24 30 Time (min)

101 Figure 26 Discussion Methanogenesis by M. vulcani B1d from glycine betaine as the sole carbon source results in approximately 0.75:1 stoichiometry of methane produced to glycine betaine consumed (Ticak et al., 2015). This suggested initial breakdown of glycine betaine is through a single demethylation reaction and not through the Stickland reaction that results in the formation of TMA and requires a betaine reductase (not encoded in the M. vulcani B1d draft genome) (Naumann et al., 1983). This work reports in vitro reconstitution of the methylotrophic glycine betaine:CoM pathway from M. vulcani B1d using recombinant enzymes. Our work is consistent with published work on QMA-dependent methanogenesis (Asakawa et al., 1998), but expands it to include the identities of the genes encoding the enzymes of the pathway, the involvement of a Pyl- lacking COG5598 enzyme, and focuses on a likely more ecologically relevant quaternary amine. This work is also consistent with a recently published study showing the involvement of another Pyl-lacking COG5598 enzyme and cognate corrinoid binding protein during methylotrophic growth of the gut bacterium Eubacterium limosum on the quaternary amine proline betaine (Picking et al., 2019). Given this knowledge, we propose the following model for metabolism of glycine betaine from M. vulcani B1d (Fig. 27). The glycine betaine:CoM model (Fig. 27) is analogous to the simple methyl- amine degradation pathways (Introduction: Fig. 2) except the substrate-specific methyltransferase MtxB lacks Pyl. The demethylation of glycine betaine yields N,N- dimethylglycine which to date has yet to be determined if it can ever serve as a direct substrate for methanogenesis. The ΔG°’ for demethylation of N,N-dimethylglycine is - 67.8 kJ/mol of N,N-dimethylglycine (Watkins et al., 2013). Both of the remaining methyl groups from N,N-dimethylglycine would yield energy for the methanogen if demethylated and those values are similar to methylamines (-43.0 kJ/mol) (Watkins et al., 2014). However, similarly to glycine betaine, N,N-dimethylglycine has been shown to act as a salt and temperature stress protectant for Bacillus subtilis but this is to a lesser degree compared to glycine betaine (Bashir et al., 2014). The demethylation product of N,N-dimethylglycine is monomethylglycine (sarcosine), which does not serve

102 appear to serve as an osmoprotectant. Although the work was performed on B. subtilis, which does not use glycine betaine, N,N-dimethylglycine, nor sarcosine as a source for energy, there are no current studies to support how efficient N,N-dimethylglycine serves as an osmoprotectant in a methanogen. The demethylated products of other quaternary amines, such N,N-dimethylethanolamine from choline, can serve as a direct substrate for methanogens, but the enzyme responsible for this metabolism is still unknown (Watkins et al., 2012). Conversely to glycine betaine and N,N-dimethylglycine, choline and N,N-dimethylethanolamine do not appear to have osmoprotectant properties (Pocard et al., 1997; and Boncompagni et al., 1999). Further studies concerning glycine betaine and N,N-dimethylglycine osmoprotectant properties in M. vulcani B1d need to be conducted to better understand why M. vulcani B1d evolved to use glycine betaine as a carbon source, but not dimethylglycine. As stated in section 1.2, an analysis of MV8460 and MV10350 was performed and from our structure data we do not think that MV10360 would interact with glycine betaine. To further support this claim we ran methanol:CoM methyl transfer assays on crude cell extracts from cells grown on glycine betaine, methanol, or TMA. Given that a putative methanol pathway is present regardless of the growth substrate, we anticipated this would be a rapid way to assess MV10360 functionality with glycine betaine. These assays performed with glycine betaine or TMA with crude extracts showed intact substrate:CoM pathways, as expected, which supports that our assay setup according to Ferguson et al. (2011) was correct. Interestingly, we could not detect methanol:CoM methyl transfer activity in any of our extracts even though all of the proteins required for methanol:CoM functionality are present. We were not able to determine why the extract showed no functionality. However, our methanogenesis assays on methanol grown cells when they were fed glycine betaine showed no apparent activity within a 24 min assay (Fig. 26 C), indicating that the enzymes of the intact methanol pathway did not readily interact with glycine betaine. Given our modeling, methyl transfer activity, and methanogenesis data, we suggest that if interactions between MV10350 and glycine betaine were possible, it would not be active at biologically relevant levels.

103 Figure 27 Proposed model of glycine betaine-dependent CoM methylation. Proteins are represented by colored circles/ovals and locations of genes that encode for them are also represented. The first methyltransferase, MtgB (blue), is a Pyl-lacking member of the COG5598 superfamily. MtgB catalyzes transfer of a methyl group from glycine betaine to the cognate corrinoid-binding protein, MtgC (red). The second methyltransferase, MtaA (green), catalyzes methyl transfer from Me-MtgC to CoM. The corrinoid reductive activation enzyme, RamM (purple), reduces the bound corrinoid of MtgC, when required. DMG = dimethylglycine.

104

Figure 27

105 Given the homology between MtgB and MttB, we predicted the involvement of a RamA and MtbA in the glycine betaine pathway. However, given that MtaA from the methanol pathway from M. barkeri can function in the methanogenesis pathway for TMA (Ferguson et al., 1996; Ferguson and Krzycki, 1997), it is not surprising that MV10335 (RamM) and MV10360 (MtaA) are used during methanogenesis from glycine betaine in M. vulcani B1d. The genetic regulation of the methanol operon that encodes these two enzymes is not currently known and may be subject to future study. In available genomes, the first methyltransferase of methylotrophic pathways is consistently encoded adjacent to its cognate corrinoid binding partner (Galagan et al., 2002; Maeder et al., 2006; Webster et al., 2019). It is interesting to speculate that methylotrophic metabolism of glycine betaine, or potentially other quaternary amines, could be achieved with the acquisition of two genes that encode a methyltransferase and a cognate corrinoid binding partner, yet methanogenic pathways for other quaternary amines as direct carbon sources have yet to be established. The methylotrophic methanogenic pathway for glycine betaine is now known, and in M. vulcani B1d it involves two proteins associated with a methanol methanogenic pathway. As stated above, many microorganisms in brackish or marine environments utilize glycine betaine transporters to internalize glycine betaine to use it as an osmoprotectant or for biosynthesis (Ticak et al., 2014; Ticak et al., 2015). Therefore, in nutritionally depleted environments due to competition, an advantage may be gained by the acquisition of genes encoding the enzymes for glycine betaine methylotrophy. This suggests a role for horizontal gene transfer with these forms of metabolism and potentially explains the clustering of the Pyl-lacking COG5598 enzymes from the glycine betaine-utilizing B1d and SLH33(T) archaea in a bacterial dominated clade (Fig. 23 B). Our work here expands the known knowledge of methanogenesis and therefore directly ties to the outlined need for understanding metabolism and physiology of methane producing organisms (Mand and Metcalf, 2019). The biological production of methane and other biogases can serve as alternative fuel sources to fossil natural gas and have potential to provide power to more than 3.5 million homes (US Department of Agriculture, US Environmental Protection Agency, and US Department of Energy, 2014). Production of methane at high amounts is achieved through enrichment

106 processes which allows for integration of this fuel source into the existing energy supply. This enrichment process relies on understanding the physiology of methanogens and what carbon sources they metabolize (Mand and Metcalf, 2019). Glycine betaine is a naturally occurring compound and has been shown through this work to be metabolized through a methylotrophic pathway. It is interesting to speculate that the metabolism of glycine betaine can be used in tandem with other currently known methanogenic pathways to increase methane production for human use as a power source.

107 Appendix I

Analyzing the interchangeability of the MtaA and RamM with homologs from Methanococcoides methylutens Q3c, Methanosarcina acetivorans WWM73, Methanosarcina barkeri Fusaro, Methanomethylovorans hollandica to reconstruct the glycine betaine:CoM methyl transfer pathway from Methanolobus vulcani B1d.

Overview and Background The information supplied in this appendix are projects that were started but were not completed in full by the time of writing this dissertation. The status of completion for each project is noted. The data presented here are the initial steps to better understand how M. vulcani B1d acquired the ability to perform methanogenesis from the quaternary amine glycine betaine. As shown in the previous sections, we determined that M. vulcani B1d is using a corrinoid-dependent methyltransferase system (Section 1.3: Fig. 27) to demethylate glycine betaine to presumably N,N-dimethylglycine with methane being the terminal product. The single demethylation of glycine betaine is supported based on carbon equivalents studies that were previously performed along with M. vulcani B1d’s inability to grow using N,N-dimethylglycine as a carbon source (Introduction: Fig 7) (Ticak et al., 2015). One major observation to note is that M. vulcani B1d upregulates a RamA and MtaA enzyme (MV10335 and MV10360, respectively) (Section 1.1: Fig. 10) during growth on glycine betaine which we concluded were used to reduce and subsequently demethylate the corrinoid binding protein, respectively (Section 1.2: Fig. 21, 22; and Section 1.3: Fig. 27). Earlier work performed on the TMA methanogenesis pathway from Methanosarcina barkeri supported that both MtaA and MtbA enzymes were capable of functioning in the TMA:CoM pathway to demethylate MttC (Ferguson and Krzycki, 1996). However, MttC affinity for MtbA is greater than MtaA on a molar basis when comparing relative activities (Fig. 28) (Ferguson and Krzycki, 1997). MtaA is considered to be used during methanol metabolism while MtbA functions on simple methylamine (TMA, DMA, and MMA) metabolism (Harms and Thauer, 1996; and Leclerc and

108 Grahame, 1996; Ferguson et al., 1996; Ferguson and Krzycki 1997; and Ferguson et al., 2000). Interestingly, although MtaA can functionally interact with MttC, it is not capable of interacting with the MtbC and MtmC, while MtbA can serve to demethylate any of these three corrinoid binding proteins (Ferguson et al., 1996).

109 Figure 28 Relative activities of the MtbA and MtaA enzymes with MttB from M. barkeri during TMA:CoM assays. To determine the relative activities, the MtbA and MtaB were varied over the indicated range (X – axis) while MttB remained constant at 35 µg of protein. At the time of this work, MtbA and MtaA were known as MT2-A and MT2-M, respectively, as they were the second methyltransferases involved during methylotrophy from methylamines and methanol, respectively (Ferguson and Krzycki, 1997).

110

Figure 28

111 Prior to the work presented in this document, few studies had been performed on the Ram enzymes that activate the corrinoid binding proteins (Daas et al., 1996 A,B; Wassenar et al., 1996; and Ferguson et al., 2009). The work performed by Daas and coworkers (1996 A,B) supported that a methyltransferase activation protein (MAP) was capable of activating the MtaC from the methanol methyl transfer system found in M. barkeri. However, there was no protein sequence reported and therefore the gene responsible for the functionality reported was still not known at the time. Additional work by Wassenaar and coworkers (1996) supported that the same MAP protein from M. barkeri was activating a methylamine:CoM methyl transfer system. This work was challenged later by suggesting the extract preparations that were used to attempt to determine MAP activity contained many cellular proteins that potentially could have been involved in the observed methylamine:CoM function (Ferguson et al., 2009). Further work on methylamine:CoM systems by Ferguson and coworkers (2009) showed the corrinoid activating enzyme RamA, which is an iron-sulfur containing protein, functions in these pathways. RamA was not reported to function in the methanol:CoM pathway but, like MtbA, it can function in each of the methylamine:CoM pathways. MAP does not appear to bind the iron-sulfur cluster and the reported functionality of RamA supported that MAP and RamA are not the same enzyme (Ferguson et al., 2009). However, in addition to the work on RamA, the researchers also reported that RamA homologs were present in M. acetivorans, M. mazeii, and M. barkeri next to genes that would encode for methanol metabolism and designated these RamA homologs as ‘RamM’ (Ferguson et al., 2009). Our work on the glycine betaine:CoM methyl transfer system from M. vulcani B1d is the first to report the functionality of a purified RamM enzyme (MV10335) which is encoded within MeOH1 and is present during growth on all substrates tested, and is most abundant on glycine betaine when comparing to the RamA homolog (MV1770) (Section 1.1: Fig. 10). Given that glycine betaine is a quaternary amine compound and that methanogenesis from this compound is initiated from a Pyl-lacking MttB homolog from the COG5598 superfamily, we predicted that a MtbA and RamA would be involved. Thus far interactions between MtxBs and MtxCs seem to be substrate-specific and, in some cases, the MtxA is more promiscuous, however, it is uncertain whether this is true

112 for Ram enzymes. Therefore, it is interesting that a RamM (MV10335) and MtaA (MV10360) are the more abundant enzymes during glycine betaine growth when M. vulcani B1d encodes for a RamA (MV1770) and two MtbAs (MV1595 and MV1680), and additionally MtgB (MV8460) is a homolog of MttB. The MtgC (MV8465) that is used for glycine betaine metabolism is more closely related to MtbC (MV1600) than to MttC (MV1615). This raises the question of whether MV10360 is involved during growth on glycine betaine since MtaAs have only previously been shown to function on methanol:CoM and TMA:CoM systems but no other simple amine pathways. Our preliminary data supports that the MtbAs (MV1595 and MV1695) can replace MV10360 and reconstitute the glycine betaine:CoM pathway (discussed below).

Goal and hypothesis of appendix I The goal of the work in this section is to determine if MtxAs and Ram enzymes from other methanogenic archaea can serve as surrogates to reconstitute the glycine betaine:CoM methyl transfer pathway from M. vulcani B1d. Our hypothesis is that MtxAs and Ram enzymes annotated or demonstrated to be involved with methanol or TMA methylotrophy from select methanogenetic archaea (Methanococcoides methylutens Q3c, Methanosarcina acetivorans WWM73, Methanosarcina barkeri Fusaro, Methanomethylovorans hollandica) can replace MV10335 and MV10360, respectively, in the glycine betaine:CoM methyl transfer pathway from M. vulcani B1d. Our rationale is that we predict that M. vulcani B1d acquired the genes encoding MV8455, MV8460, and MV8465 and exploited portions of an existing methanol:CoM system, that is constitutively present, to gain the ability to metabolize glycine betaine. We are interested in further understanding this potential phenomenon by using enzymes from various methylotrophic methanogenic organisms during growth on methanol or TMA.

113 Materials and Methods MtxA and Ram selection from Methanococcoides methylutens Q3c, Methanosarcina acetivorans WWM73, Methanosarcina barkeri Fusaro, and Methanomethylovorans hollandica The selection of the MtxAs and Ram enzymes was based on adhering to one of the following criteria: 1) the MtxA or Ram enzyme must have been detected during growth of the organism on either methanol or TMA; 2) if not shown to be present, the MtxA or Ram enzyme must be encoded near other methylotrophic proteins that are present during methanol or TMA metabolism and therefore would likely be required to complete the methylotrophic pathway; or 3) be the sole MtxA or Ram enzyme of its type found within the genome. The selected proteins and their comparisons to the four MtxAs (MV1575, MV1695, MV2245, and MV10360) or the three Ram homologs (MV1770, MV10000, and MV10335) from B1d are summarized in Table 3, 4, and 5. For Methanosarcina acetivorans C2a, Methanosarcina barkeri Fusaro, and Methanomethylovorans hollandica, we searched published data that included either proteomic or transcriptomic data that involved growing these organisms either on TMA or methanol. We found transcriptomic data for M. acetivorans C2a and M. barkeri Fusaro (López Muñoz et al., 2015; Peterson et al., 2016). The M. acetivorans C2a data supplied information for both TMA and methanol, but the transcriptomics data for M. barkeri Fusaro only had methanol data that was useful to our needs. To overcome this issue, we searched the M. barkeri Fusaro genome to determine how many MtbA and RamA enzymes were encoded. Fortunately, there is only a single MtbA (Mbar_A0841) and a single RamA (Mbar_A0840) encoded within the genome and therefore they were selected even though we lacked the transcriptomic data to support their involvement with TMA metabolism. Additionally, work by Ferguson et al. (1996) showed that both MtaA and MtbA are present during growth of M. barkeri. Preliminary proteomic data for Methanococcoides methylutens Q3c grown on either on TMA or methanol was performed by Mr. Jeff Ringiesn and Ms. Jyoti Kashyap with assistance from Dr. Xin Wang and coworkers (data not shown). The methods followed similarly to those that were previously stated in Section 1.1. The MtxA and Ram enzymes that appeared with highest abundance during growth of M. methylutens

114 Q3c on TMA or methanol were then selected. A protein-protein BLAST analysis of the proteins encoded from the genome M. vulcani B1d was then performed using proteins found from the searched data as queries. We were only able to find proteomic data on M. hollandica (Chen et al., 2017). Unfortunately, this work relied on doing shotgun proteomics of a mixed mud sample that had been enriched with the quaternary amine QMA. The strain that was specifically in the mud sample was not isolated, however, the researchers did supply information of the methylotrophic enzymes that were present during the enrichment by matching their peptide hits and coverage to the publicly available genome of M. hollandica (DSM 15978). Given our knowledge of the work performed by Asakawa and coworkers on methylotrophy from QMA, it is possible that M. hollandica could be using a corrinoid- dependent demethylation mechanism to metabolize QMA. We selected the MtbA (Metho_0007) and MtaA (Metho_0776), that were detected by the proteomics. Lastly, according to our M. vulcani B1d proteomic data, it appears that Ram enzymes are not overly abundant compared to the other proteins involved with substrate:CoM methylotrophy (Section 1.1: Fig. 10). Chen and coworkers (2017) did not report the presence of Ram enzymes. We searched the proteins encoded near Metho_0007 and Metho_0776 for potential Ram enzymes and we found a RamA (Metho_0004) and a RamM (Metho_0772). Given the proximity of the Ram enzymes to the MtxAs and their annotated functions, we decided to proceed with them despite a lack of data supporting their presence during growth on methanol or TMA.

115 Table 3 Proteins selected from M. barkeri Fusaro based on transcriptomic data (López Muñoz et al., 2015).

Name / TMA Methanol Highest Identities and Gene ID Protein Accession Normalized Normalized Match Similarities NCBI Mean Mean MV1770: Identities = 368/541 (68%), Similar = 443/541 (81%) MV10000: Identities = MV1770

Mbar_A0840 RamA AAZ69816.1 N/A 135.40 310/522 (59%), Similar (E = 0.0) = 397/522 (76%) MV10335: Identities = 246/528 (46%), Similar = 335/528 (63%) MV1770: Identities = 244/540 (45%), Similar = 346/540 (64%) MV10000: Identities = MV10335

Mbar_A1055 RamM AAZ70023.1 N/A 289.36 222/511 (43%), Similar (E = 0.0) = 319/511 (62%) MV10335: Identities = 357/536 (66%), Similar = 439/536 (81%) MV1575: Identities = 129/343 (37%), Similar = 190/343 (55%) MV1695: Identities = 135/343 (39%), Similar MV10360 = 195/343 (56%)

Mbar_A1054 MtaA Q48949.1 N/A 491.02 (E = e-150) MV2245: Identities = 153/340 (45%), Similar = 218/340 (64%) MV10360: Identities = 208/336 (61%), Similar = 256/336 (76%) MV1575: Identities = 135/346 (39%), Similar = 195/346 (56%) MV1695: Identities = 135/342 (39%), Similar MV2112 = 193/342 (56%)

Mbar_A3639 MtaA AAZ72503.1 N/A 406.19 (E = e-159) MV2245: Identities = 153/340 (45%), Similar = 218/340 (64%) MV10360: Identities = 215/336 (63%), Similar = 267/336 (79%) MV1575: Identities = 245/339 (72%), Similar = 277/339 (81%) MV1695: Identities = 230/339 (67%), Similar MV0221 = 272/339 (80%)

Mbar_A0841 MtbA AAZ69817.1 N/A 197.09 (E = 0.0) MV2245: Identities = 118/335 (35%), Similar = 185/335 (55%) MV10360: Identities = 128/337 (37%), Similar = 186/337 (55%)

116 Table 4 Proteins selected from M. acetivorans WWM73 based on transcriptomic data (Peterson et al., 2016). These proteins were used a query for during building of the MtxA and Ram phylogenetic trees.

Name / Highest TMA RNA Methanol Identities and Gene ID Protein Accession Match counts RNA counts Similarities NCBI MV1770: Identities = 364/541 (67%), Similar = 443/541 (81%) MV10000: Identities = MV0274

MA_0150 RamA AAM03603.1 4.1 3.5 304/522 (58%), Similar (E = 0.0) = 398/522 (76%) MV10335: Identities = 244/539 (45%), Similar = 346/539 (64%) MV1770: Identities = 238/537 (44%), Similar = 343/537 (63%) MV10000: Identities = MV2107

MA_4380 RamM AAM07722.1 5.9 6.9 215/511 (42%), Similar (E = 0.0) = 315/511 (61%) MV10335: Identities = 360/536 (67%), Similar = 440/536 (82%) MV1575: Identities = 238/339 (70%), Similar = 279/339 (82%) MV1695: Identities = 233/339 (68%), Similar MV0221 = 280/339 (82%)

MA_0146 MtbA P58869.2 136.5 15.8 (E = 0.0) MV2245: Identities = 126/341 (36%), Similar = 191/341 (56%) MV10360: Identities = 125/334 (37%), Similar = 185/334 (55%) MV1575: Identities = 131/343 (38%), Similar = 197/343 (57%) MV1695: Identities = 137/343 (39%), Similar MV2112 = 200/343 (58%)

MA_4379 MtaA AAM07721.1 114.7 149.6 (E = e-155) MV2245: Identities = 149/340 (43%), Similar = 218/340 (64%) MV10360: Identities = 210/336 (62%), Similar = 262/336 (77%)

117 Table 5 Proteins selected from M. hollandica based on shotgun proteomic data and genomic analysis (Chen et al., 2017).

Name / Relative Highest Coverage Gene ID Protein Accession Abundance Identities and Similarities Match (%) NCBI (%) MV1575: Identities = 265/339 (78%), Positives = 305/339 MV1695: Identities = 279/339 (82%), Positives = MV1575 310/339 (91%) Metho_0007 MtbA AGB48305.1 28.61 0.19 (E = 0.0) MV2245: Identities = 124/335 (37%), Positives = 191/335 (57%) MV10360: Identities = 131/331 (39%), Positives = 186/331 (56%) MV1575: Identities = 131/344 (38%), Positives = 195/344 (56%) MV1695: Identities = 139/342 (40%), Positives = MV10360 197/342 (57%) Metho_0776 MtaA AGB49025.1 43.07 0.93 (E = e-177) MV2245: Identities = 150/341 (43%), Positives = 212/341 (62%) MV10360: Identities = 245/335 (73%), Positives = 274/335 (81%) MV1770: Identities = 417/540 (77%), Positives = 475/540 (87%) MV10000: Identities = MV1770 Metho_0004 RamA AGB48302.1 N/A N/A 304/522 (58%), Positives = (E = 0.0) 396/522 (75%) MV10335: Identities = 247/537 (45%), Positives = 350/537 (65%) MV1770: Identities = 247/540 (45%), Positives = 363/540 (67%) MV10000: Identities = MV10335 Metho_0772 RamM AGB49021.1 N/A N/A 220/523 (42%), Positives = (E = 0.0) 323/523 (61%) MV10335: Identities = 426/537 (79%), Positives = 486/537 (90%)

118 Cloning and expression vectors All primers used during the process of cloning can be found in Table 6 along with the status of the construct. Cloning of the genes from M. vulcani B1d that encode for MV1595 and MV1695 into pASK-IBA43plus followed the same methods used to clone MV8460, while cloning of MV2245 into pETAC17a followed the same methods used to clone MV10360 as described in Section 1.2. Cloning of the genes that encode MtbA (MM1399) and MtaA (MV0619) from M. methylutens Q3c into pETAC17a was performed by Mr. Peter Brechting and followed traditional cloning methods that were described in Section 1.2. Cloning of the remaining gene candidates into pETAC17a was performed by Mr. Trevor Powell using the Gibson Assembly® Protocol (New England Biolabs), following manufacturers recommended protocol. During the writing of this document, not all constructs were completed, and the current status of these clones is summarized in Table 7.

Production and purification of recombinant proteins Production of MM1399 from M. methylutens Q3c and MA0146 and MA4379 from M. acetivorans WWM73 MA was identical to the techniques used to produce MV8460 and MV1360, respectively. Production of MM2121 followed techniques identical to anoxic production of MV10335. Thus far only MM1399 and MM0619 have been purified. Purification followed similarly to what was described in Section 1.2 except only a single HisTrap™ HP column (GE Healthcare) was required. Status for production of the remaining proteins is summarized in Table 7. As stated in Section 1.1, there are three Ram homologs encoded within the M. vulcani B1d genome (Section 1.1: Table 1). We successfully reconstituted the glycine betaine:CoM pathway using MV10335 as the enzyme to activate MV8465 (Section 1.2 and 1.3: Fig. 21 and 24, respectively). As part of the support for the work presented here in Appendix I, we attempted to produce the other two Ram homologs from M. vulcani B1d, but we were unsuccessful. Therefore, were not able to test whether MV1770 or MV10000 could replace MV10335 in the glycine betaine:CoM pathway. Further attempts to produce these enzymes will be done at a later time.

119 Table 6. Primer sequences of the MtxAs and Rams

Primer name Sequence MA0146 F AACCTGTACTTCCAAGGCTATATGACAGAATATACCCCAAAA MA0146 R ACAAACTGCAGTCCGCGGTATCAGTACTTGTAGTTTTTAGC MA0150 F AACCTGTACTTCCAAGGCTATATGTACGGAATAGCACTTGAC MA0150 R ACAAACTGCAGTCCGCGGTATTATTTTTCCTTTATTTTGAGTTTC MA4379 F AACCTGTACTTCCAAGGCTATATGACCGATATGAGCGAATTC MA4379 R ACAAACTGCAGTCCGCGGTATCAGGCGTAGAATTCGTTTC MV4380 F AACCTGTACTTCCAAGGCTATATGAGAACAGGAGTTGCAATT MV4380 R ACAAACTGCAGTCCGCGGTATCAGAGTCCTGCAACGGT Mbar_A0840 F AACCTGTACTTCCAAGGCTATATGTATGGAATAGCACTTG Mbar_A0840 R ACAAACTGCAGTCCGCGGTATTATTTCGCTGTGATTTTCAG Mbar_A0841 F AACCTGTACTTCCAAGGCTATATGGCAGAATATACCCCAAAAG Mbar_A0841 R ACAAACTGCAGTCCGCGGTATTAGTATGTGTGGCTTTTTGCG Mbar_A1054 F AACCTGTACTTCCAAGGCTATATGAGCGAATTTACACTTAAAAC Mbar_A1054 R ACAAACTGCAGTCCGCGGTATCAGGCGTAGTACTCGTC Mbar_A1055 F AACCTGTACTTCCAAGGCTATATGAGAATAGGAGTTGCAATTG Mbar_A1055 R ACAAACTGCAGTCCGCGGTATTACTTTCCTGCGACTGTC Mbar_A3639 F AACCTGTACTTCCAAGGCTATATGAATGAGATGGCACTTAAAG Mbar_A3639 R ACAAACTGCAGTCCGCGGTATCAAGCGTAATATTCGTCTC Metho_0004 F AACCTGTACTTCCAAGGCTATATGCTTTTTGAAGGTATCTTC Metho_0004 R ACAAACTGCAGTCCGCGGTATTAACGCTTCACCTTCAAAATTG Metho_0007 F AACCTGTACTTCCAAGGCTATATGTCCGATTACACCCCTAAAG Metho_0007 R ACAAACTGCAGTCCGCGGTATCAATACTTATGTGCTTTCGC Metho_0772 F AACCTGTACTTCCAAGGCTATATGAGACTAGGTGTTGCAG Metho_0772 R ACAAACTGCAGTCCGCGGTATCAGATCTCCATTACCTTC Metho_0776 F AACCTGTACTTCCAAGGCTATATGAGCGACTTAAACATG Metho_0776 R ACAAACTGCAGTCCGCGGTATTACTTGTAGTATTCGTTCC MM620 F AACCTGTACTTCCAAGGCTATATGCGAACGGGAATTGCAATAG ACAAACTGCAGTCCGCGGTATTAACCCAGTACTTTGAGATTATTC MM620 R C MM2121 F AACCTGTACTTCCAAGGCTATATGTATGGTATTGCATTGGATC ACAAACTGCAGTCCGCGGTATCATTTCTTAATAGAAATGGTCTTA MM2121 R T

120 Table 7 Current status of cloning, production, and activity testing of the proteins selected for this study Protein Gene Protein Activity Organism Gene locus Annotation Cloned Produced Tested MV1575 MtbA Y Y Y Methanolobus vulcani B1d MV1695 MtbA Y Y Y MV2245 MtaA Y Y Y MM0619 MtaA Y Y Y Methanococcoides methylutens Q3c MM1399 MtbA Y Y N MA0146 MtbA Y Y N Methanosarcina acetivorans WWM73 MA4379 MtaA Y Y N Metho_0007 MtbA N N N Methanomethylovorans hollandica Metho_0776 MtaA N N N Mbar_A1054 MtaA N N N Methanosarcina barkeri Fusaro Mbar_A3639 MtaA N N N Mbar_A0841 MtbA N N N MM0620 RamM Y N N Methanococcoides methylutens Q3c MM2121 RamA Y Y N MA0150 RamA Y N N Methanosarcina acetivorans WWM73 MA4380 RamM Y N N Metho_0772 RamM N N N Methanomethylovorans hollandica Metho_0004 RamA N N N Mbar_A0840 RamA N N N Methanosarcina barkeri Fusaro Mbar_A1055 RamM N N N

121 MtxA methyltransferase activity and in vitro reconstruction of glycine betaine:CoM pathway using MtxA surrogates Methods for testing the function of the MtxAs can be found in Section 1.2 under the heading “MV10360 methyltransferase activity”. All variables and conditions were identical. Thus far only MV1575, MV1695, and MM0619 have been tested. The methods for reconstruction of the glycine betaine:CoM pathway can be found in Section 1.3 under the heading “In vitro reconstitution of glycine betaine:CoM methyl transfer pathway”. The only difference was that MV10360 was replaced with either MV1575 or MV1695, while all other conditions were identical.

Reconstruction of glycine betaine:CoM by addition of MV8460 and MV8465 to Methanococcoides methylutens Q3c extracts M. methylutens Q3c was grown either on methanol (62.5 mM) or TMA (40 mM) to mid-log phase and anoxically harvested, as described previously (Ticak et al., 2015). Crude extracts were prepared as described in Section 1.3. Glycine betaine-dependent CoM methylation assays in the M. methylutens Q3c extracts were prepared as described in Section 1.3 with the following additional steps. To each of the reaction mixtures, MV8460 (30 ug) and MV8465 (15 ug) were added and glycine betaine was used as the methyl donor. Controls consisted of omitting MV8460, MV8465, or the methyl donor; and using methanol, TMA, or QMA with MV8460 and MV8465 added to extracts that were prepared from cells not grown on the methyl donor as a substrate.

Sequence acquisition and phylogenetic construction of the MtxA and Ram trees The amino acid sequences for MtbA and MtaA from M. barkeri Fusaro (Table 4) and each of the four MtxAs from M. vulcani B1d were used as a query with protein- protein BLAST with default settings, for the non-redundant protein sequence (nr) database. Similarly, the RamA and RamM from M. barkeri Fusaro along with the three Ram homologs from M. vulcani B1d were used as a query for acquisition of the Ram sequences. The starting dataset was 600 sequences for the MtxAs and 500 sequences for the Rams. To both data sets, duplicates with 90% or higher identity were removed using the publicly available ElimDupes tool from the HIV databases website

122 (www.hiv.lanl.gov). The sequences from the datasets were then separately aligned with MUSCLE using default settings. Those sequences that were annotated as ‘partial’ according to NCBI annotation were manually removed from both datasets and the sequences were re-aligned. Phylogenetic analysis was performed in MEGA using approximate-maximum-likelihood with 500 bootstrap replications using JTT (Jones et al., 1992), WAG (Whelan and Goldman, 2001), or LG (Le and Gascuel, 2008) with Gamma distribution set to 5 discrete Gamma categories. Phylogenetic trees were generated using the interactive Tree of Life (iTOL) (Letunic and Bork, 2016).

123 Results and Discussion Reconstruction of glycine betaine:CoM activity by addition of MV8460 and MV8465 to Methanococcoides methylutens Q3c extracts Our initial experimental design was to test the interchangeability of MtxA and Ram homologs using extracts prepared from our selected organisms that were grown using methanol or TMA as the substrate. Since M. methylutens Q3c is an obligate methylotrophic methanogen that grows using methanol or TMA, it would require the use of a MtaA and a RamM or a MtbA and a RamA, respectively. Therefore, by adding MV8460 and MV8465 to anoxic extracts from M. methylutens Q3c, we could test the potential of the MtxA and Ram homologs present in the extracts to replace MV10360 and MV10335 to reconstruct the glycine betaine:CoM pathway. We were able to reconstitute the pathway, but our results were not consistent based on time to completion (data not shown). The time to completion ranged from 40 min to 80 min and therefore we considered this technique to not be replicable for statistical purposes. We predict that the inconsistencies were at the level of the extracts and not due to MV8460 or MV8465. Even though we attempt to prepare the extract replicates identically, we are unable to determine the amount of MtxA or Ram in each specific extract at the time of performing the assays. Given that we were able to detect some activity, albeit slow, this demonstrated the proof of concept that the MtxA and Ram enzyme present in the extract could function to interact with MV8465, which supported that MV10335 and MV10360 had successfully been substituted. Future experiments will use recombinant versions of the MtxA and Ram homologs (Table 7) in combination with MV8460 and MV8465 to attempt to reconstitute the glycine betaine:CoM pathway using the same assay that was described in Section 1.3 (Fig. 24).

MV1575, MV1695, MV2245, and MM0619 methylcob(III)alamin:CoM activity Similar to our rationale regarding the functionality of MV10360 (Section 1.2: Fig 22), it was necessary to confirm that MV1575, MV1695, and MM0619 were functional enzymes that could demethylate methylcob(III)alamin before incorporating them as surrogates for MV10360 in the glycine betaine:CoM pathway. CoM methylase activity of MV1575, MV1695, and MM0619 was confirmed by monitoring change at 540 nm.

124 Cob(II)alamin is generated during methylcob(III)alamin:CoM methyl transfer, causing a decrease at 540 nm (Ferguson et al., 2011; Fig. 29 AB, Fig. 30). The assays for MM0619 were performed by Peter Brechting.

Glycine betaine:CoM reconstruction using MV1595 and MV1695 surrogates for MV10360 We hypothesized that MtxAs from M. methylutens Q3c, M. acetivorans WWM73, M. barkeri Fusaro, and M. hollandica could serve as surrogates for MV10360 to reconstitute the glycine betaine:CoM pathway from M. vulcani B1d. Our preliminary experiments to justify future testing of MtxAs from these organisms was to test the MV1595 and MV1695 from M. vulcani B1d as they are MtbAs and MV10360 is a MtaA. Using recombinant enzymes, we successfully reconstituted the glycine betaine:CoM pathway when MV10360 was substituted with MV1595 or MV1695 (Fig. 31). These data are preliminary and need to be replicated. No CoM methylation was detected when choline or TMA was used as the methyl donor. The remaining MtaA from M. vulcani B1d (MV2245) was tested for activity with methylcob(III)alamin, but we were not able to test if it is able to replace MV10360. Although we have shown activity of MM0619 using methylcob(III)alamin (Fig. 30), we have yet to determine if it will serve as a surrogate of MV10360. We will test MM0619 once we have successfully produced and confirmed the activity of the RamM (MM0620) from M. methylutens Q3c. Our reasoning for waiting is that M. vulcani B1d appears to have gained the ability to perform methylotrophy from glycine betaine by exploiting the presence of MV10335 and MV10360, therefore it would seem logical to attempt substitutions of both MV10335 and MV10360 simultaneously. Efforts to produce MM0620 are currently underway and it will be examined at a later time.

125 Figure 29 Methylcob(III)alamin:CoM methyl-transfer activity by (A) MV1575, (B) MV1695, and (C) MV2245 . The reaction contained 0.5 mM methylcob(III)alamin, 50 mM phosphate buffer at pH 7.2, 5 mM CoM, 50 µg either MV1575 or MV1695, and was performed anoxically under dim red light. Readings were taken every 15 s until completion. Demethylation of methylcob(III)alamin was measured by a decrease in absorbance at 540 nm. Error bars represent standard deviations (n = 3).

126 A 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25

0.20 Absorbance (540 nm) (540 Absorbance 0.15 0.10 0 50 100 150 200 250 Time (sec)

B 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25

Absorbance (540 nm) (540 Absorbance 0.20 0.15 0.10 0 50 100 150 200 250 Time (sec)

C 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25

0.20 Absorbance (540 nm) (540 Absorbance 0.15 0.10 0 50 100 150 200 250 Time (sec)

Figure 29

127 Figure 30 Methylcob(III)alamin:CoM methyl-transfer activity by MM0619. The reaction contained 0.5 mM methylcob(III)alamin, 50 mM phosphate buffer at pH 7.2, 5 mM CoM, 40 µg MM0619, and was performed anoxically under dim red light. Readings were taken every 15 s until completion. Demethylation of methylcob(III)alamin was measured by a decrease in absorbance at 540 nm. Error bars represent standard deviations (n = 3).

128

Figure 30

129 Figure 31 Representative figure of reconstitution of glycine betaine:CoM activity in vitro with purified recombinant proteins. Loss of the free thiol group on CoM was monitored at 412 nm, using Ellman’s reagent. Addition of glycine betaine with MV1575 (blue diamond) or MV1695 (red square) in place of MV10360 resulted in a significant decrease in the amount of HS-CoM, indicating an intact glycine betaine:CoM methyl transfer pathway. Controls consisted of: MV1575 and MV10335 with QMA as the methyl donor (green triangle), MV1575 without MV10335 with glycine betaine as the methyl donor (light blue star), MV1695 and MV10335 with QMA as the methyl donor (purple X), and MV1695 without MV10335 with glycine betaine as the methyl donor (orange circle).

130 Figure 31

131 Phylogenetic trees of MtxA and Ram proteins We generated phylogenetic trees from the MtxA and Ram sequences we acquired and processed as stated in the methods (Fig. 32 and 33). For the MtxA tree (Fig. 32), the WAG model (Whelan and Goldman, 2001) and the best log-likelihood (- 21328.89) with a +G, parameter (2.5820). Although many of the nodes on the tree have bootstrap values less than 80, it can be seen from the root that there are three distinct major branches that form with confidence being greater 80 at the node. As expected, the MtbAs and MtaAs cluster separately along two different major branches. The MV2245 from M. vulcani B1d clusters separately and distinctly from both the MtbA and the MtaA branches. Interestingly, MV2245 is annotated as a MtaA but only shares moderate identity (45%) and sequence similarity (65%) with MV10360. For the Ram tree, the JTT matrix-based model (Jones et al., 1992) provided the best log-likelihood (- 37759.63) with +G, parameter (0.7374).The RamM and RamA homologs appear to cluster into distinctly different branches, as was expected. Given the model that was generated, MV10000 diverges from the other RamA homologs, however, the node bootstrap value is below 80. However, following further down the branches there are high bootstrap values which supports clustering of MV10000 from the other RamAs. MV10000 is more closely clustered to the so-called RamS (MA0489) from M. acetivorans, which is predicted to be involved in methylated sulfur metabolism. However, lacking activity data on these enzymes, we cannot definitively state that MV10000 or MA0849 function differently than bona-fide RamAs (Fu and Metcalf, 2015).

132 Figure 32 Approximate-maximum likelihood representation of the MtxA phylogenetic tree. The phylogenetic tree shows the proposed evolutionary relationship between the MtbAs and MtaAs (Table 7) selected for this study. Branch coloring indicates confidence of splits at the nodes with coloration cutoff being dependent on bootstrap values greater than 80. Values closer to 80 are red branches and as the values increase towards 100 the branches transition to green. Branches that are black are all values that are less than 80. The scale bars represent amino acid substitutions per site.

133

Figure 32

134 Figure 33 Approximate-maximum likelihood representation of the Ram phylogenetic tree. The phylogenetic tree shows the proposed evolutionary relationship between the RamM and RamA homologs (black lettering) (Table 7) selected for this study. Additionally, shown is the RamA homolog from M. acetivorans (MA0489) (blue lettering) that was labeled as a ‘RamS’ (Fu and Metcalf, 2015). Branch coloring indicates confidence of splits at the nodes with coloration cutoff being dependent on bootstrap values greater than 80. Values closer to 80 are red branches and as the values increase towards 100 the branches transition to green. Branches that are black are all values that are less than 80. The scale bars represent amino acid substitutions per site.

135

Figure 33

136 In conclusion, our preliminary data support that other enzymes can substitute for MtaA (MV10360) from M. vulcani B1d to reconstitute the glycine betaine:CoM pathway in vitro (Fig. 31). Initial assays in which anoxic extracts, from M. methylutens Q3c grown on methanol or TMA were amended with purified MV8460 and MV8465, were successful but lacked consistency and reproducibility. Therefore, assays using purified recombinant proteins will be used for future experiments. We have constructed MtxA and Ram phylogenetic trees so the relative clustering of MtaAs, MtbAs, RamAs, or RamMs can now be visualized. These phylogenetic trees are important for this study as we attempt to determine the interchangeability of MtxAs and Rams in the glycine betaine:CoM pathway. Once the presence or absence of activity of the interchanged enzymes in the pathway has been determined, we will be able to visualize where they cluster in the trees and then analyze those specific lineages to look for evolutionary patterns. The overall hypothesis will be supported if the MtxAs and Rams from M. methylutens Q3c, M. acetivorans WWM73, M. barkeri Fusaro, and M. hollandica can serve to substitute MV10360 and MV10335, respectively. Given that MtaA interacts with both MttC and MtaC but not MtbC suggests limited specificity. The MtgC (MV8465) from M. vulcani B1d is more closely related to MtbC than MttC, and our data presented here supports that a MtaA (MV10360) is the primary MtxA used during glycine betaine:CoM metabolism. These results can then be combined with the phylogenetic analysis for future work on understanding more about MV8460 and how it differs from other MttCs or MtbCs. The potential for a wider range of MtxAs (and perhaps Rams) interacting MtgB suggests that methylamine methylotrophy is still evolving and that an organism’s acquisition of a new carbon source could stem from adaptations of MtxCs. The overall goal is to gain a better understanding of how M. vulcani B1d was able to evolve to use glycine betaine as a carbon source via methylotrophy. The initial step, which we hope to explore further, is determining the minimum number of genes that would be required during a horizontal gene transfer event in order to acquire the ability to perform methylotrophy from glycine betaine. Thus far it appears that MV8460 and MV8465 may be the only genes required. Although, it could also be dependent on what, if any, quaternary amine transporters are encoded in the genome of the organism.

137 References

Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., et al. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acid Res. 25, 3389–3402. doi: 10.1093/nar/25.17.3389

Andreesen, J. R. (2004). Glycine reductase mechanism. Curr. Opin. Chem. Biol. 8, 454-461. doi: 10.1016/j.cbpa.2004.08.002

Asakawa, S., Sauer, K., Liesack, W., and Thauer, R. K. (1998). Tetramethylammonium: coenzyme M methyltransferase system from Methanococcoides sp. Arch. Microbiol. 170, 220–226. doi: 10.1007/s002030050636

Auffret, M. D., Stewart, R., Dewhurst, R. J., Duthie, C. A., Rooke, J. A., Wallace, R. J., et al. (2017). Identification, comparison, and validation of robust rumen microbial biomarkers for methane emissions using diverse Bos taurus breeds and basal diets. Front. Microbiol. 8:2642. doi: 10.3389/fmicb.2017.02642

Barker, H.A. (1981). Amino acid degradation by anaerobic bacteria. Ann. Rev. Biochem. 50, 23–40. doi: 10.1146/annurev.bi.50.070181.000323

Bashir, A., Hoffmann, T., Smits, S. H., and Bremer, E. (2014). Dimethylglycine provides salt and temperature stress protection to Bacillus subtilis. Appl. Environ. Microbiol. 80, 2773-2785. doi: 10.1128/AEM.00078-14

Beale, R. and Airs, R. (2016). Quantification of glycine betaine, choline and trimethylamine N-oxide in seawater particulates: Minimisation of seawater associated ion suppression. Anal. Chim. Acta.938, 114-122. doi: 10.1016/j.aca.2016.07.016

138 Boncompagni, E., Osteras, M., Poggi, M. C., Le Rudulier, D. (1999). Occurrence of choline and glycine betaine uptake and metabolism in the family Rhizobiaceae and their roles in osmoprotection. Appl. Environ. Microbiol. 65, 2072–2077. PMCID: PMC91300

Borrel, G., O’Toole, P. W., Harris, H. M., Peyret, P., Brugere, J. F., and Gribaldo, S. (2013). Phylogenomic data support a seventh order of Methylotrophic methanogens and provide insights into the evolution of methanogenesis. Genome Biol. Evol. 5, 1769– 1780. doi: 10.1093/gbe/evt128

Blight, S. K., Larue, R. C., Mahapatra, A., Longstaff, D. G., Chang, E., Zhao, G., et al. (2004). Direct charging of tRNA(CUA) with pyrrolysine in vitro and in vivo. Nature. 431, 333–5. doi:10.1038/nature02895

Bradford, M. M. (1976). A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. doi: 10.1006/abio.1976.9999

Burke, S. A., and Krzycki, J. A. (1995). Involvement of the “A” isozyme of methyltransferase II and the 29-kilodalton corrinoid protein in methanogenesis from monomethylamine. J. Bacteriol. 177, 4410–4416. doi: 10.1128/jb.177.15.4410- 4416.1995

Chen, W. Y., Kraková, L., Wu, J. H., Pangallo, D., Jeszeová, L., Liu, B., and Yasui, H. (2017). Community and proteomic analysis of anaerobic consortia converting tetramethylammonium to methane. Archaea. 2017, 2170535. doi: 10.1155/2017/2170535

Craciun, S., and Balskus, E. P. (2012). Microbial conversion of choline to trimethylamine requires a glycyl radical enzyme. Proc. Natl. Acad. Sci. U.S.A. 109, 21307–21312. doi: 10.1073/pnas.1215689109

139 A. Daas, P. J. H., Hagen, W. R., Keltjens, J. T., van der Drift, C., and Vogels, G. D. (1996). Activtaion mechanism of methanol:5-hydroxybenzimidazolylcobamide methyltransferase from Methanosarcina barkeri. J. Biol. Chem. 271, 22346-22351. doi: 10.1074/jbc.271.37.22346

B. Daas, P. J. H., Wassenaar, R. W., Willemsen, P., Theunissen, R. J., Keltjens, J. T., van der Drift, C., and Vogels, G. D. (1996). Purification and properties of an enzyme involved in the ATP-dependent activation of the methanol:2- mercaptoethanesulfonic acid methyltransferase reaction in Methanosarcina barkeri. J. Biol. Chem. 271, 22339-22345. doi: 10.1074/jbc.271.37.22339

Dean, J. F., Middelburg, J. J., Röckmann, T., Aerts, R., Blauw, L. G., Egger, M., et al. (2018). Methane feedbacks to the global climate system in a warmer world. Rev. Geophys. 56, 207– 250. doi: 10.1002/2017RG000559

Dehning, I., Stieb, M., and Schink, B. (1989). Sporomusa malonica sp. nov., a homoacetogenic bacterium growing by decarboxylation of malonate or succinate. Arch. Microbiol. 151, 421–426. doi: 10.1007/bf00416601

Deng, J., McCalley, C. K., Frolking, S., Chanton, J., Crill, P., Varner, R., et al. (2017). Adding stable carbon isotopes improves model representation of the role of microbial communities in peatland methane cycling. J. Adv. Model. Earth Syst. 9, 1412– 1430. doi: 10.1002/2016ms000817 Edgar, R. C. (2004). MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32, 1792–1797. doi: 10.1093/nar/gkh340

Ellman, G. L. (1958). A colorimetric method for determining low concentrations of mercaptans. Arch. Biochem. Biophys. 74, 443–450. doi: 10.1016/0003-9861(58)90014- 6

140 Enzmann, F., Mayer, F., Rother, M., and Holtmann, D. (2018). Methanogens: biochemical background and biotechnological applications. AMB Express 8:1. doi: 10.1186/s13568-017-0531-x

Ferguson, D. J. Jr., Gorlatova, N., Grahame, D. A., and Krzycki, J. A. (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. doi: 10.1074/jbc.m910218199

Ferguson, D. J. Jr., and Krzycki, J. A. (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. doi: 10.1128/jb.179.3.846-852.1997

Ferguson, D. J. Jr., Krzycki, J. A., and Grahame, D. A. (1996). Specific roles of methylcobamide:coenzyme M methyltransferase isozymes in metabolism of methanol and methylamines in Methanosarcina barkeri. J. Biol. Chem. 271, 5189–5194. doi: 10.1074/jbc.271.9.5189

Ferguson D.J. Jr, Gorlatova N, Grahame D.A., Krzycki J.A. (2000). Reconstitution of dimethylamine:coenzyme M methyl transfer with a discrete corrinoid protein and two methyltransferases purified from Methanosarcina barkeri. J. Biol. Chem. 275, 29053-60. doi: 10.1074/jbc.M910218199

Ferguson, D. J. Jr., Longstaff, D. G., and Krzycki, J. A. (2011). Assay of methylotrophic methyltransferases from methanogenic archaea. Methods Enzymol. 494, 139–158. doi: 10.1016/B978-0-12-385112-3.00008-1

Ferguson, T. K., Soares, J. A., Lienard, T., Gottschalk, G., and Krzycki, J. A. (2009). RamA, a protein required for reductive activation of corrinoid-dependent

141 methylamine methyltransferase reactions in methanogenic archaea. J. Biol. Chem. 284, 2285–2295. doi: 10.1074/jbc.M807392200

Ferry, J. G. (2011). Fundamentals of methanogenic pathways that are key to the biomethanation of complex biomass. Curr. Opin. Biotechnol. 22, 351–357. doi: 10.1016/j.copbio.2011.04.011

Finster, K., Liesack, W., and Tindall, B. J. (1997). Desulfospira joergensenii, gen. nov., sp. nov., a new sulfate-reducing bacterium isolated from marine surface sediment. Syst. Appl. Microbiol. 20, 201–208. doi: 10.1016/s0723-2020(97)80066-5

Fitzsimons, M. F., Kahni-danon, B., Dawitt, M., (2001). Distributions and adsorption of the methylamines in the inter-tidal sediments of an East Anglian Estuary. Environ. Exp. Bot. 46, 225-236. doi: 10.1016/S0098-8472(01)00102-2 Fu, F. and M, W. W. (2015). Genetic basis for metabolism of methylated sulfur compounds in Methanosarcina species. J. Bacterol. 197, 1515-1524. doi: 10.1128/JB.02605-14

Galagan, J. E., Nusbaum, C., Roy, A., Endrizzi, M. G., Macdonald, P., FitzHugh, W., et al. (2002). The genome of Methanosarcina acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 12, 532–542. doi: 10.1101/gr.223902

Gertz, E. M., Yi-Kuo, Y., Agarwala, R., Schäffer, A. A., and Altschul, S. F. (2006). Composition-based statistics and translated nucleotide searches: improving the TBLASTN module of BLAST. BMC Biol. 4:41. doi: 10.1186/1741-7007-4-41

Gibb, S. W., Mantoura, R. F. C., Liss, P. S. (1999). Ocean-atmosphere exchange an atmospheric speciation of ammonia and methylamines in the region of the NW Arabian Sea. Global Biogeochem. Cy. 13, 161-178. doi: 10.1029/98GB00743

142 Gibb, S. W. and Hatton, A. D. (2004). The occurrence and distribution of trimethylamine-N-oxide in Antarctic coastal water. Mar. Chem. 91, 65-75. doi: 10.1016/j.marchem.2004.04.005

Hagemeier, C. H., Krer, M., Thauer, R. K., Warkentin, E., and Ermler, U. (2006). Insight into the mechanism of biological methanol activation based on the crystal structure of the methanol-cobalamin methyltransferase complex. Proc. Natl. Acad. Sci. U.S.A. 103, 18917–18922. doi: 10.1073/pnas.0603650103

Hao, B., Gong, W., Ferguson, T. K., James, C. M., Krzycki, J. A., and Chan, M. K. (2002). A new UAG-encoded residue in the structure of a methanogen methyltransferase. Science 296, 1462–1466. doi: 10.1126/science.1069556

Harms, U., and Thauer, R. K. (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. 234, 653-9. doi: 10.1111/j.1432-1033.1996.00653.x

Hayward, H. R., and Stadtman, T. C. (1959). Anaerobic degradation of choline. I. Fermentation of choline by an anaerobic, cytochrome-producing bacterium, Vibrio cholinicus nov. sp. J. Bacteriol. 78, 557– 561. PMCID: PMC290585

Hayward, H. R., and Stadtman, T. C. (1960). Anaerobic degradation of choline. II. Preparation and properties of cell-free extracts of Vibrio cholinicus. J. Biol. Chem. 235, 538-543.

Hayward, H. R. (1960). Anaerobic degradation of choline. III. Acetaldehyde as an intermediate in the fermentation of choline by extracts of Vibrio cholinicus. J. Biol. Chem. 235, 3592-3596.

143 Horn, C., Sohn-Bösser, L., Breed, J., Welte, W., Schmitt, L., and Bremer, E. (2006). Molecular determinants for substrate specificity of the ligand-binding protein OpuAC from Bacillus subtilis for the compatible solutes glycine betaine and proline betaine. J. Mol. Biol. 357, 592–606. doi: 10.1016/j.jmb.2005.12.085 Huang, B. (2009). metaPocket: a meta approach to improve protein ligand binding site prediction. OMICS 13, 325–330. doi: 10.1089/omi.2009.0045

Hug., L A. (2018). Sizing up the uncultured microbial majority. mSystems. 3, e00184- 18. doi: 10.1128/mSystems.00185-18

James, E., Quareshy, M., and Chen., Y. (2018). Methodological considerations for the identification of choline and carnitine-degrading bacteria in the gut. Methods. 149, 42- 48. doi: 10.1016/j.ymeth.2018.03.012

Jiang, R., and Krzycki, J. A. (2012). PylSn and the homologous N-terminal domain of pyrrolysyl-tRNA synthetase bind the tRNA that is essential for the genetic encoding of pyrrolysine. J. Biol. Chem. 287, 32738–46. doi:10.1074/jbc.M112.396754

Jones, D. T., Taylor, W. R., and Thornton, J. M. (1992). The rapid generation of mutation data matrices from protein sequences. Comput. Appl. Biosci. 8, 275–282. doi: 10.1093/bioinformatics/8.3.275

Jones, H. J., Kröber, E., Stephenson, J., Mausz, M. A., Jameson, E., Millard, A., et al. (2019). A new family of uncultivated bacteria involved in methanogenesis from the ubiquitous osmolyte glycine betaine in coastal saltmarsh sediments. Microbiome. 7, 120. doi: 10.1186/s40168-019-0732-4

King, G. M. (1984). Metabolism of trimethylamine, choline, and glycine betaine by sulfate-reducing and methanogenic bacteria in marine sediments. Appl. Environ. Microbiol. 48, 719–725.

144 König, H., Stetter K. O. (1982). Isolation and characterization of , sp. nov., a coccoid methanogen growing only on methanol and methylamines. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 3, 478- 490. doi: 10.1016/S0721-9571(82)80005-7

Kreft, J. U., and Schink, B. (1993). Demethylation and degradation of phenylmethylethers by the sulfide-methylating homoacetogenic bacterium strain TMBS 4. Eur. J. Biochem. 226, 945–951.

Krüer, M., Haumann, M., Meyer-Klaucke, W., Thauer, R. K., and Dau, H. (2002). The role of zinc in the methylation of the coenzyme M thiol group in methanol:coenzyme M methyltransferase from Methanosarcina barkeri. Eur. J. Biochem. 269, 2117–2123. doi:10.1046/j.1432-1033.2002.02860.x

Krzycki, J. A. (2005). The direct genetic encoding of pyrrolysine. Curr. Opin. Microbiol. 8, 706–712. doi: 10.1016/j.mib.2005.10.009

Kuhner, C. H., Frank, C., Griesshammer, A., Schmittroth, M., Acker, G., Gössner, A., et al. (1997). Sporomusa silvacetica sp, nov., an acetogenic bacterium isolated from aggregated forest soil. Int. J. Syst. Evol. Microbiol. 47, 352–358. doi: 10.1099/00207713-47-2-352

Ladapo, J., and Whitman, W. B. (1990). Method for isolation of auxotrophs in the methanogenic archaebacteria: role of the acetyl-CoA pathway of autotrophic CO2 fixation in Methanococcus maripaludis. Proc. Natl. Acad. Sci. U.S.A. 87, 5598–5602. doi: 10.1073/pnas.87.15.5598 Lambie, S. C., Kelly, W. J., Leahy, S. C., Li, D., Reilly, K., McAllister, T. A., Valle, E. R., et al. (2015). The complete genome sequence of the rumen methanogen Methanosarcina barkeri CM1. Stand. Genomic. Sci. 10, 574. doi: 10.1186/s40793-015- 0038-5

145 Lazar, C. S., Parkes, R. J., Cragg, B. A., L’Haridon, S., and Toffin, L., (2011). Methanogenic diversity and activity in hypersaline sediments of the centre of the Napoli mud volcano, Eastern Mediterranean Sea. Environ. Microbiol. 13, 2078-2091. doi:10.1111/j.1462-2920.2011.02425.x

Le, S. Q., and Gascuel, O. (2008). An improved general amino acid replacement matrix. Mol. Biol. Evol. 25, 1307–1320. doi: 10.1093/molbev/msn067

Lechtenfeld, M., Heine, J., Sameith, J., Kremp, F., and Müller, V. (2018). Glycine betaine metabolism in the acetogenic bacterium Acetobacterium woodii. Environ. Microbiol. 20, 4512–4525. doi: 10.1111/1462-2920.14389

LeClerc, G. M., and Grahame, D. A. (1996). Methylcobamide:coenzyme M methyltransferase isozymes from Methanosacrina barkeri. Physiochemical characterization, cloning, sequence analysis, and heterologous gene expression. J. Biol. Chem. 271, 18725-31. doi: 10.1074/jbc.271.31.18725

Letunic, I., and Bork, P. (2016). Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Res. 44, W242– W245. doi: 10.1093/nar/gkw290

L’Haridon, S. M., Chalopin, M. M., Colombo, D. M., and Toffin, L. (2014). Methanococcoides vulcani sp. nov., a novel marine methylotrophic methanogen; using betaine, choline and N,N-dimethylethanolamine for methanogenesis, isolated from the napoli mud volcano in the eastern mediterranean Sea; and emendation of the genus Methanococcoides. Int. J. Syst. Evol. Microbiol. 64, 1978–1983. doi: 10.1099/ijs.0.058289-0

Li, C. Y., Chen, X. L., Shao, X., Wei, T. D., Wang, P., Xie, B. B., et al. (2015). Mechanistic Insight into trimethylamine N-oxide recognition by the marine

146 bacterium Ruegeria pomeroyi DSS-3. J. Bacteriol. 197, 3378–3387. doi: 10.1128/JB.00542-15

Li, W., and Godzig, A. (2006). Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 22, 1658–1659. doi: 10.1093/bioinformatics/btl158

Liu, y., and Whitman, W.B. (2008). Metabolic, phylogenetic, and ecological diversity of the methanogenic archaea. Ann. N. Y. Acad. Sci. 1125, 171-89. doi: 10.1196/annals.1419.019.

López Muñoz, M. M., Schönheit, P., and Metcalf, W. W. (2015). Genetic, genomic, transcriptomics studies on pyruvate metabolism in Methanosarcina barkeri Fusaro. J. Bacteriol. 197, 3592-3600. doi:10.1128/JB.00551-15

Longstaff, D. G., Blight, S. K., Zhang, L., Green-Church, K. B., and Krzycki, J. A. (2007). In vivo contextual requirements for UAG translation as pyrrolysine. Mol. Microbiol. 63, 229–241. doi: 10.1111/j.1365-2958.2006.05500.x

Maeder, D. L., Anderson, I., Brettin, T. S., Bruce, D. C., Gilna, P., Han, C. S., et al. (2006). The Methanosarcina barkeri genome: comparative analysis with Methanosarcina acetivorans and Methanosarcina mazei reveals extensive rearrangement within methanosarcinal genomes. J. Bacteriol. 188, 7922–7931. doi: 10.1128/jb.00810-06

Maltby, J., Steinle, L., Löscher, C. R., Bange, H. W., Fischer, M. A., Schmidt, M., and Treude, T. (2018). Microbial methanogenesis in the sulfate-reducing zone of sediments in the Eckernförde Bay, SW Baltic Sea. Biogeosciences 15, 137-157. doi: 10.5194/bg-15-137-2018

147 Mand, T. D., and Metcalf, W. W. (2019). Energy Conservation and Hydrogenase Functions in Methanogenic Archaea, in Particular the Genus Methanosarcina. Microbiol. Mol. Biol. Rev. 83, e00020-19. doi: 10.1128/MMBR.00020-19

Mausz, M. A., and Chen, Y. (2019). Microbiology and ecology of methylated amine metabolism in marine ecosystems. Curr. Issues Mol. Biol. 33, 133–148. doi: 10.21775/cimb.033.133

Metcalf, W. M., Zhang, J. K., Apolinario, E., Sowers, K. R., and Wolfe, R. S. (1997). A genetic system for Archaea of the genus Methanosarcina:liposome mediated transformation and contruction of shuttle vectors. Proc. Natl. Acad. Sci. U.S.A. 94, 2626–2631. doi: 10.1073/pnas.94.6.2626

Meyer, M., Granderath, K., and Andreesen, J. R. (1995). Purification and characterization of protein PB of betaine reductase and its relationship to the corresponding proteins glycine reductase and sarcosine reductase from Eubacterium acidaminophilum. Eur. J. of Biochem. / FEBS. 234, 184– 191. doi: 10.1111/j.1432- 1033.1995.184_c.x

Möller, B., Ossmer, R., Howard, B. H., Gottschalk, G., and Hippe, H. (1984). Sporomusa, a new genus of Gram-negative anaerobic bacteria including Sporomusa sphaeroides spec. nov. and Sporomusa ovata spec. nov. Arch. Microbiol. 139, 388–396. doi: 10.1007/bf00408385

Muller, E., Fahlbusch, K., Walther, R., and Gottschalk, G. (1981). Formation of N,N- dimethylglycine, acetic acid, and butyric acid from betaine by Eubacterium limosum. Appl. Environ. Microbiol. 42, 439–445.

Myhre, G., Shindell, D., Bréon, F.‐M., Collins, W., Fuglestvedt, J., Huang, J., et al. (2013). Anthropogenic and natural radiative forcing. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report

148 of the Intergovernmental Panel on Climate Change (pp. 659– 740). Cambridge: Cambridge University Press.

Naumann, E., Hippe, H., and Gottschalk, G. (1983). Betaine: new oxidant in the Stickland reaction and methanogenesis from betaine and L-alanine by a Clostridium sporogenes- Methanosarcina barkeri coculture. Appl. Environ. Microbiol. 45, 474–483.

Nielsen, M. B., Kjeldsen, K. U., and Ingvorsen, K. (2006). Desulfitibacter alkalitolerans gen. nov., sp. nov., an anaerobic, alkalitolerant, sulfite-reducing bacterium isolated from a district heating plant. Int. J. Syst. Evol. Microbiol. 56, 2831–2836. doi: 10.1099/ijs.0.64356-0

Nisman, B. (1954). The Stickland reaction. Bacteriol. Rev. 18, 16-42. PMCID: PMC180783

Oelgeschläger, E. and Rother, W. (2009) In vivo role of three fused corrinoid/methyl transfer proteins in Methanosarcina acetivorans. Mol. Microbiol. 72, 1260-1272. doi: 10.1111/j.1365-2958.2009.06723.x

Olefeldt, D., Turetsky, M. R., Crill, P. R., and McGuire, A. D. (2013). Environmental and physical controls on northern terrestrial methane emissions across permafrost zones. Glob. Change Biol. 19, 589–603. doi: 10.1111/gcb.12071

Oremland, R. S., Marsh, L. M., and Polcin, S. (1982). Methane production and simultaneous sulphate reduction in anoxic, salt marsh sediments. Nature. 296, 143– 145. doi:10.1038/296143a0

Paul, L., Ferguson, D. J., and Krzycki, J. A. (2000). The trimethylamine methyltransferase gene and multiple 178 dimethylamine methyltransferase genes of Methanosarcina barkeri contain in-frame and read- through amber codons. J. Bacteriol. 182, 2520–2529. doi:10.1128/JB.182.9.2520-2529.2000

149

Peterson, J. R., Thor, S., Kohler, L., Kohler, P. R., Metcalf, W. W., and Luthey- Schulten, Z. (2016). Genome-wide gene expression and RNA half-life measurements allow predictions of regulation and metabolic behavior in Methanosarcina acetivorans. BMC Genomics. 16, 924. doi: 10.1186/s12864-016-3219-8

Picking, J. W., Behrman, E. J., Zhang, L., and Krzycki, J. A. (2019). MtpB, a member of the MttB superfamily from the human intestinal acetogen Eubacterium limosum, catalyzes proline betaine demethylation. J. Biol. Chem. 294, 13697–13707. doi: 10.1074/jbc.RA119.009886 Pocard, J. A., Vincent, N., Boncompagni, E., Smith, L. T., Poggi, M. C., and Le Rudulier, D. (1997). Molecular characterization of the bet genes encoding glycine betaine synthesis in Sinorhizobium meliloti 102F34. Microbiol. 143, 1369–1379. doi: 10.1099/00221287-143-4-1369

Poehlein, A., Cebulla, M., Ilg, M. M., Bengelsdorf, F. R., Schiel-Bengelsdorf, B., Whited, G., et al. (2015). Complete genome sequence of Rnf- and cytochrome- containing autotrophic acetogen clostridium aceticum DSM 1496. Genome Announc. 3:e786-15. doi: 10.1128/genomeA.00786-15

Price, M. N., Dehal, P. S., and Arkin, A. P. (2010). FastTree 2–approximately maximum-likelihood trees for large alignments. PLoS One 5:e9490. doi: 10.1371/journal.pone.0009490

Pritchett, M. A., and Metcalf, W. W. (2005). Genetic, physiological and biochemical characterization of multiple methanol methyltransferase isozymes in Methanosarcina acetivorans C2A. Mol. Microbiol. 56, 1183–1194. doi: 10.1111/j.1365- 2958.2005.04616.x

Purdy, K. J., Nedwell, D. B., and Embley, T. M. (2003). Analysis of the sulfate- reducing bacterial and methanogenic archaeal populations in contrasting antarctic

150 sediments. Appl. Environ. Microbiol. 69, 3181–3191. doi: 10.1128/aem.69.6.3181- 3191.2003

Ragsdale, S. W. (2008). Catalysis of methyl group transfers involving tetrahydrofolate and B12. Vitam. Homr. 79, 293-324. doi:10.1016/S0083-6729(08)00410-X Rebouche, C. J., and Seim, H. (1998). Carnitine metabolism and its regulation in microorganisms and mammals. Annu. Rev. Nutr. 18, 39–61. doi: 10.1146/annurev.nutr.18.1.39

Richards, M. A., Lie, T. J., Zhang, J., Ragsdale, S. W., Leigh, J. A., Price, N. D. (2016). Exploring Hydrogenotrophic Methanogenesis: a Genome Scale Metabolic Reconstruction of Methanococcus maripaludis. J. Bacteriol. 198, 3379-3390. doi: 10.1128/JB.00571-16

Rowe, A. R., Xu, S., Gardel, E., Bose, A., Girguis, P., Amen, J. P., El-Nagger, M. Y. (2019). Methane-linked mechanisms of electron uptake from cathodes by Methanosarcina barkeri. Appl. Environ. Microbiol. 10, e02448-18. doi: 10.1128/mBio.02448-18

Salvi, F., Wang, Y. F., Weber, I. T., and Gadda, G. (2014). Structure of choline oxidase in complex with the reaction product glycine betaine. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 70, 405–413. doi: 10.1107/s1399004713029283

Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. New York, NY: Cold Spring Harbor Laboratory.

Sattley, W. M., and Madigan, M. (2007). Cold-active acetogenic bacteria from surficial sediments of perennially ice-covered Lake Fryxell, Antarctica. FEMS Microbiol. Lett. 272, 48–54. doi: 10.1111/j.1574-6968.2007.00737.x

151 Sauer, K., and Thauer, R. K. (1997). Methanol:coenzyme M methyltransferase from Methanosarcina barkeri. Zinc dependence and thermodynamics of the methanol:cob(I)alamin methyltransferase reaction. Eur. J. Biochem. 249, 280-5. doi: 10.1111/j.1432-1033.1997.t01-1-00280.x

Sauer, K., Harms, U., and Thauer, R. K. (1997). Methanol:coenzyme M methyltransferase from Methanosarcina barkeri. Purification, properties and encoding genes of the corrinoid protein MT1. Eur. J. Biochem. 243, 670–677. doi: 10.1111/j.1432- 1033.1997.t01-1-00670.x

Sauer, K., and Thauer, R. K. (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. doi: 10.1046/j.1432-1327.1998.2530698.x

Schiefner, A., Breed, J., Bosser, L., Kneip, S., Gade, J., Holtmann, G., et al. (2004). Cation-pi interactions as determinants for binding of the compatible solutes glycine betaine and proline betaine by the periplasmic ligand-binding protein prox from Escherichia coli. J. Biol. Chem. 279, 5588–5596. doi: 10.1074/jbc.m309771200

Schlegal, K. and Müller, V. (2011). Sodium Ion Translocation and ATP Synthesis in Methanogens. Methods Enzymol. 494, 233-255. doi: 10.1016/B978-0-12-385112- 3.00012-3

Seefeldt, L. C., and Ensign, S. A. (1994). A continuous spectrophotometric activity assay for using the reductant titanium (III) citrate. Anal. Biochem. 221, 379– 386. doi: 10.1006/abio.1994.1429

Sharma, V., Murphy, D. P., Provan, G., and Baranov, P. V. (2012). CodonLogo: a sequence logo-based viewer for codon patterns. Bioinformatics 28, 1935–1936. doi: 10.1093/bioinformatics/bts295

152 Sikorski, J., Lapidus, A., Chertkov, O., Lucas, S., Copeland, A., Glavina Del Rio, T., et al. (2010). Complete genome sequence of Acetohalobium arabaticum type strain (Z- 7288T). Stand. Genomic Sci. 3, 57–65. doi: 10.4056/sigs.1062906

Sowers, K. R., Baron, S. F., and Ferry, J. G. (1984). Methanosarcina acetivorans sp. nov., an acetotrophic methane-producing bacterium isolated from marine sediments. Appl. Environ. Microbiol. 47, 971-8. PMID: 16346552

Srinivasan, G., James, C. M., and Krzycki, J. A. (2002). Pyrrolysine encoded by UAG in Archaea: charging of a UAG-decoding specialized tRNA. Science. 29, 1459–1462. doi:10.1126/science.1069588

Stickland, L. H. (1934). Studies in the metabolism of the strict anaerobes (genus Clostridium): The chemical reactions by which Cl. sporogenes obtains its energy. Biochem. J. 24, 1746-1759. doi: 10.1042/bj0281746

Stickland, L. H. (1935). Studies in the metabolism of the strict anaerobes (genus Clostridium): The reduction of proline by Cl. sporogenes. Biochem. J. 29, 288-290. doi: 10.1042/bj0290288

Tallant, T. C., and Krzycki, J. A. (1997). Methylthiol:coenzyme M methyltransferase from Methanosarcina barkeri, an enzyme of methanogenesis from dimethylsulfide and methylmercaptopropionate. J. Bacteriol. 179, 6902–6911. doi: 10.1128/jb.179.22.6902- 6911.1997

Tallant, T. C., Paul, L., and Krzycki, J. A. (2001). The MtsA subunit of the methylthiol:coenzyme M methyltransferase of Methanosarcina barkeri catalyses half- reactions of corrinoid dependent dimethylsulfide:coenzyme M methyl transfer. J. Biol. Chem. 276, 4485–4493. doi: 10.1074/jbc.m007514200

153 Tanaka, K. (1994). Anaerobic degradation of tetramethylammonium by a newly isolated marine methanogen. J. Ferment. Bioeng. 78, 386–388. doi: 10.1016/0922- 338x(94)90287-9

Thauer, R. K., Kaster, A. K., Seedorf, H., Buckel, W., and Hedderich, R. (2008). Methanogenic archaea: ecologically relevant differences in energy conservation. Nat. Rev. Microbiol. 6, 579-591. doi: 10.1038/nrmicro1931

Ticak, T., Hariraju, D., Bayron, M., Arivett, B. A., Fiester, S. E., and Ferguson, D. J. Jr. (2015). Isolation and characterization of a tetramethylammonium degrading Methanococcoides strain and a novel glycine betaine utilizing Methanolobus strain. Arch. Microbiol. 197, 197–209. doi: 10.1007/s00203-014- 1043-6

Ticak, T., Kountz, D., Girosky, K., Krzycki, J. A., and Ferguson, D. J. Jr. (2014). A non-pyrrolysine member of the widely distributed trimethylamine methyltransferase family is a glycine betaine methyltransferase. Proc. Natl. Acad. Sci. U.S.A. 111, E4668– E4676. doi: 10.1073/pnas.1409642111

Trott, O., and Olson, A. J. (2010). AutoDock vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31, 455–461. doi: 10.1002/jcc.21334 Uchiyama, T., Ito, K., Mori, K., Tsurumaru, H., Harayama, S. (2010). Iron-corroding methanogen isolated form a crude-oil storage tank. Appl. Environ. Microbiol. 76, 1783- 1788. doi: 10.1128/AEM.00668-09

US Department of Agriculture, US Environmental Protection Agency, US Department of Energy. (2014). Biogas opportunities roadmap. USDA, EPA, DOE, Washington, DC.

154 van der Maarel, M., Jansen, M., and Hansen T. A. (1995). Methanogenic conversion of 3-S-methylmercaptopropionate to 3-mercaptopropionate. Appl. Environ. Micrbiol. 61, 48-51.

Wallace, R.J., Rooke, J.A., McKain, N., Duthie, C.A., Hyslop, J. J., Ross, D. Q., Waterhouse, A., Watson, M., and Roeche, R. (2015). The rumen microbial metagenome associated with high methane production in cattle. BMC Genomic 16, 839. Doi: 10.1186/s12864-015-2032-0

Wassenaar, R. W., Daas, P. J., Geerts, W. J., Keltjens, J. T., and van der Drift, C. (1996). Involvement of methyltransferase-activating protein and methyltransferase 2 isoenzyme II in methylamine:coenzyme M methyltransferase reactions in Methansarcina barkeri Fusaro. J. Bacteriol. 178, 6937-6944. doi: 10.1128/jb.178.23.6937-6944.1996

Watkins, A. J., Roussel, E. G., Parkes, R. J., and Sass, H. (2014). Glycine betaine as a direct substrate for methanogens (Methanococcoides spp.). Appl. Environ. Microbiol. 80, 289–293. doi: 10.1128/AEM.03076-13

Watkins, A. J., Roussel, E. G., Webster, G., Parkes, R. J., and Sass, H. (2012). Choline and N,N-dimethylethanolamine as direct substrates for methanogens. Appl. Environ. Microbiol. 78, 8298–8303. doi: 10.1128/AEM.01941-12 Webb, B., and Sali, A. (2017). Protein structure modeling with modeller. Meth. Mol. Biol. 1654, 39–54. doi: 10.1007/978-1-4939-7231-9_4

Webster, G., Mullins, A. J., Watkins, A. J., Cunningham-Oakes, E., Weightman, A. J., Mahenthiralingam, E., et al. (2019). Genome sequences of two choline-utilizing methanogenic archaea, Methanococcoides spp., isolated from marine sediments. Microbiol. Resour. Announc. 8:e342-19. doi: 10.1128/MRA.00342-19

155 Welander, P. V, and Metcalf, W. W. (2005). Loss of the mtr operon in Methanosarcina blocks growth on methanol, but not methanogenesis, and reveals an unknown methanogenic pathway. PNAS. 102, 10664–9. doi:10.1073/pnas.0502623102

Welander, P. V., and Metcalf, W. W. (2008). Mutagenesis of the C1 oxidation pathway in Methanosarcina barkeri: New insights into the Mtr/Mer bypass pathway. J. Bacteriol. 190, 1928–1936. doi:10.1128/JB.01424-07

Whelan, S., and Goldman, N. (2001). A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approach. Mol. Biol. Evol. 18, 691–699. doi: 10.1093/oxfordjournals.molbev.a003851

Wolters, J. C., Berntsson, R. P.-A., Gul, N., Karasawa, A., Thunnissen, A. M., Slotboom, D. J., et al. (2010). Ligand binding and crystal structures of the substrate- binding domain of the ABC transporter OpuA. PLoS One 5:e10361. doi: 10.1371/journal.pone.0010361

Yin, X., Wu, W., Maeke, M., Richter-Heitmann, T., Kulkarni, A. C., Oni, O. E., et al. (2019). CO2 conversion to methane and biomass in obligate methylotrophic methanogens in marine sediments. ISME 13, 2107-2119. doi: 10.1038/s41396-019- 0425-9

Yvon-Durocher, G., Allen, A. P., Bastviken, D., Conrad, R., Gudasz, C., St. Pierre, A., et al. (2014). Methane fluxes show consistent temperature dependence across microbial to ecosystem scales. Nature 507, 488–491. doi: 10.1038/nature13164

Zhang, Y. (2008). I-TASSER server for protein 3D structure prediction. BMC Bioinformatics 9:40. doi: 10.1186/1471-2105-9-40

156 Zhuang, G. C., Elling, F. J., Nigro, L. M., Samarkin, V., Joye, S. B., Teske, A., and Hinrichs, K. U. (2016). Multiple evidence for methylotrophic methanogenesis as the dominant methanogenic pathway in hypersaline sediments from the Orca Basin, Gulf of Mexico. Geochim. Cosmochim. Acta. 187, 1-20. doi: /10.1016/j.gca.2016.05.005

Zhuang, G. C., Heuer, V. B., Lazar, C. S., Goldhammer, T., Wendt, J., Samarkin, V. A., et al. (2018). Relative importance of methylotrophic methanogenesis in sediments of the Western Mediterranean Sea. Geochim. Cosmochim. Acta. 224, 171-186. doi: 10.1016/j.gca.2017.12.024

157