Methylation of two-component response regulator MtrA in mycobacteria negatively modulates 1 its DNA binding and transcriptional activation. 2

Posttranslational modifications such as phosphorylation, nitrosylation, and pupylation modulate multiple cellular processes in Mycobacterium tuberculosis . While protein methylation at lysine and arginine 20 residues is widespread in eukaryotes, to date only two methylated proteins in Mtb have been identified. Here we report the identification of methylation at lysine and/or arginine residues in nine mycobacterial 22 proteins. Among the proteins identified, we chose MtrA, an essential response regulator of a two- 23 component signaling system, which gets methylated on multiple lysine and arginine residues to examine 24 the functional consequences of methylation. While methylation of K207 confers a marginal decrease in 25 the DNA binding ability of MtrA, methylation of R122 or K204 significantly reduces the interaction with 26 the DNA. Overexpression of S-adenosyl homocysteine hydrolase (SahH), an enzyme that modulates the 27 levels of S-adenosyl methionine in mycobacteria decreases the extent of MtrA methylation. Most 28 importantly, we show that decreased MtrA methylation results in transcriptional activation of mtrA and 29 sahH promoters. Collectively, we identify novel methylated proteins, expand the list of modifications in 30 mycobacteria by adding arginine methylation, and show that methylation regulates MtrA activity. We 31 propose that protein methylation could be a more prevalent modification in mycobacterial proteins. methylation other metabolites biotin. disruption of one-carbon significance of studying regulators of one-carbon metabolism highlight In summary, the present study provides a framework for elucidation of protein methylation in mycobacteria. We report the addition of protein arginine methylation to the growing list of regulatory PTMs in mycobacteria and suggest that methylation of MtrA at lysine and arginine residues regulates its activity. This study provides an orchestration of methylation and TCS signaling and therefore illuminates the critical role of methylation in bacterial physiology.


INTRODUCTION 33
Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis, is responsible for nearly one

46
In eukaryotes, methylation of histone proteins at specific lysine residues regulates chromatin 47 architecture and transcription, and aberrant methylation is associated with aging and cancer [12]. Arginine 48 methylation is the most extensively studied protein modification in eukaryotes and its role in DNA repair,

49
RNA metabolism, and transcriptional repair is well established [13]. Guanidino group of arginine is 50 involved in the interaction with DNA; the addition of methyl group directly affects the activity of 51 proteins. Methylation of Sam68 (an adapter protein for Src kinases during mitosis) at arginine residue 52 restrains it's binding to Src homology 3 (SH3) domain of phospholipase Cγ-1 and methylation at arginine 53 and lysine residues of CHD1 (chromo-helicase/ATPase DNA-binding protein 1) results in a significant 54 decrease in its binding affinity to DNA [8]. Several non-histone proteins, mainly transcription factors and 55 histone-or chromatin-associated proteins are also regulated by methylation [12,14].

56
In bacteria, however, our understanding of the functional role of lysine or arginine methylation is 57 limited [9]. Lysine methylation is associated with bacterial cell motility of Synechocystis sp. and with host 58 colonization and disease initiation by Pseudomonas aeruginosa [15]. A recent proteomics study has identified 59 abundant lysine and arginine methylation in Escherichia coli [16]. In Mtb, lysine residues of Heparin-Binding

89
Generation of plasmid constructs.

90
We selected 180 protein-coding genes from Mtb genome representing a random set across 91 various functional classes ( Fig S1). Genes involved in regulation and information processing were over-

107
For analyzing the effect of SahH on MtrA methylation, the genes encoding these proteins were 108 co-expressed in Msm. Mtb sahH was cloned in mycobacterial integrative vector pSET152 [21]. For this, sahH and the positive clones were selected on apramycin and kanamycin. His 6 -MtrA was purified from cells containing both pSET-sahH and pVV16-mtrA and used for Western blotting.

116
Expression and purification of recombinant proteins.
For expression and purification of proteins from Msm, recombinant clones (2 µg) in pVV16 118 vector were electroporated and recombinants were selected on kanamycin. Expressed proteins were

139
The concentration of purified proteins was estimated by Bradford assay (Bio-Rad). Purified proteins were 140 resolved on SDS-PAGE and analyzed by staining with coomassie brilliant blue R250 (Sigma-Aldrich).

141
To analyze the effect of Hcy on MtrA methylation, Msm cells harboring pVV16-mtrA were 142 grown in Sauton's medium containing 0.4 mM Hcy. Cells were grown up to A 600 ~ 0.8, harvested, and subjected to Ni 2+ -NTA chromatography for purification of His 6 -MtrA, as described above. The proteins were later analyzed by Western blotting. For normalization, His 6 -MtrA was purified from cells grown

218
Varying amounts of His 6 -MtrA and its site-specific mutants were phosphorylated using 2 µg MBP-tagged

RESULTS
In Mtb, the proteins involved in metabolism, respiration, and cell wall-related processes form the 245 majority of functional proteome compared with regulatory and signaling proteins ( Fig S1a) [28].
functional classes (Fig S1a & b and Table S1) with a focus on less prevalent regulatory and signaling 250 proteins. We selected very few genes from "PE and PPE proteins" and "conserved hypotheticals" and 251 did not select any genes from the categories "Stable RNAs", "Insertion sequences and phages", and few other contaminating proteins (Fig 1a). Among the 72 recombinant proteins, ten proteins were 262 recognized by the anti-methyllysine antibody, suggesting the presence of methylation on lysine residues 263 (Fig 1a). However, we could not detect a distinct methylated band for PykA. To validate the protein 264 identity and the methylation of the ten western blot-positive candidate proteins, we performed high-265 resolution mass spectrometry. We were successful in detecting 20 methyllysine sites belonging to 7 266 candidate proteins except LldD1, LldD2, and Tpi (Fig 1b, Fig S2, Table S2). Interestingly, we also 267 detected 18 methylarginine sites in 7 candidate proteins, which included LldD2 and Tpi (Fig 1b, Fig S2, 268 Table S2). Together, we identified 20 methyllysine and 18 methylarginine sites in 9 out of 10 western blot-  the results obtained in Msm (Fig 1a), MtrA was found to be methylated in Mtb (Fig 2b).
MtrA is a 228 amino acid (aa) long protein with a 102 aa long N-terminal response regulator 280 domain and a 93 aa long C-terminal winged helix-turn-helix DNA binding domain homologous to E. coli 281 OmpR (Fig 2a & Fig S3) [42, 43]. Mass spectrometry data showed that MtrA was methylated on six 282 arginine residues and seven lysine residues (Fig 1b). Analysis of these 13 methylated residues showed that 283 R122 is a conserved residue present in the linker region, and K204 and K207 are adjacent to the DNA 284 recognition helix (Fig 2c & Fig S3). Therefore, we examined the roles of R122, K204, and K207 by 285 mutating them individually to methionine residues, the closest structural mimic to dimethyllysine [44].

286
Wild type and mutant MtrA proteins were expressed in Msm and the purified proteins were probed with 287 anti-methyllysine antibody to compare their relative methylation (Fig 3a). Densitometric analysis of blots  Table S3. The 296 comparison shows that mutating any of these residues negatively affects the methylation at other sites;

297
K204M or K207M completely abolish the methylation at four other sites (Fig. 3c). We observed a 298 background signal for amino acid position 207 in the MtrA K207M mutant, which was due to the 'match 299 between run' event rather than a bonafede fide MS/MS signal. Moreover, the signal was only 0.4% 300 compared with that in MtrA-WT signal, suggesting that it most likely represents the noise (Table S3).

301
Similarly, background signal was also observed for amino acid position 122 in MtrA R122M mutant, which 302 may be due to noise but could not be attributed to 'match between run' event.  (Fig 4a; upper band). In addition to the autophosphorylated EnvZ, we detected efficient differentiate between these substrates (Fig 4a).
that there is no DNA: protein complex formation if either MtrA or EnvZ is absent (Fig 4b). We observed 322 DNA binding only upon incubation of phosphorylated MtrA with radiolabeled oriC DNA fragment and 323 the binding efficiency was dependent on the concentration of MtrA (Fig 4b). These results show that,

346
To test our hypothesis, we first evaluated the effect of increasing Hcy on the growth of Msm.

347
Bacteria were grown in minimal growth medium containing varying concentrations of Hcy and their 348 growth was measured. We found that increasing concentration of Hcy negatively affects bacterial growth 349 in a concentration-dependent manner (Fig 5b & 5c). Results suggested that higher than 0.4 mM Hcy

355
Next, we addressed the influence of overexpressing SahH on MtrA methylation. We analyzed 356 methylation of MtrA purified from Msm containing an integrated copy of Mtb SahH. We observed that 357 overexpression of SahH also resulted in a ~70% decrease in MtrA methylation levels, presumably because 358 of perturbed SAH levels as SAH is a potent inhibitor of methyltransferases (Fig 5f and 5g). Collectively,

359
the data suggest that perturbation of metabolic intermediates negatively modulates MtrA methylation.

361
In Fig 4,     residues was found to negatively regulate its DNA binding function (Fig 4). Arginine methylation 439 regulates several mammalian processes associated with gene expression but is largely unrecognized in

468
In summary, the present study provides a framework for elucidation of protein methylation in 469 mycobacteria. We report the addition of protein arginine methylation to the growing list of regulatory              c.