Systematic Analysis of Mycobacterial Acylation Reveals First Example of Acylation-mediated Regulation of Enzyme Activity of a Bacterial Phosphatase

Background: Post-translational modifications regulate Mycobacterium tuberculosis physiology and pathogenesis.

Results: Several novel signal transduction proteins are regulated by lysine acylation, including the virulence-associated protein-tyrosine phosphatase PtpB.

Conclusion: M. tuberculosis lysine acylation-regulated pathways affect other signal transduction modules.

Significance: This is the first report of acylation-dependent regulation of enzyme activity of a bacterial phosphatase.

Abstract

Protein lysine acetylation is known to regulate multiple aspects of bacterial metabolism. However, its presence in mycobacterial signal transduction and virulence-associated proteins has not been studied. In this study, analysis of mycobacterial proteins from different cellular fractions indicated dynamic and widespread occurrence of lysine acetylation. Mycobacterium tuberculosis proteins regulating diverse physiological processes were then selected and expressed in the surrogate host Mycobacterium smegmatis. The purified proteins were analyzed for the presence of lysine acetylation, leading to the identification of 24 acetylated proteins. In addition, novel lysine succinylation and propionylation events were found to co-occur with acetylation on several proteins. Protein-tyrosine phosphatase B (PtpB), a secretory phosphatase that regulates phosphorylation of host proteins and plays a critical role in Mycobacterium infection, is modified by acetylation and succinylation at Lys-224. This residue is situated in a lid region that covers the enzyme's active site. Consequently, acetylation and succinylation negatively regulate the activity of PtpB.

Introduction

Regulatory mechanisms in prokaryotic cells confer complexity to their relatively simpler genomes and proteomes. Post-translational modifications (PTMs)⁷ add diversity to the proteome and regulate a myriad of cellular processes. The two extensively studied and arguably the most important PTMs are phosphorylation and acetylation. Mycobacterium tuberculosis, the causative agent of tuberculosis, employs several mechanisms to evade the host immune system; and PTMs play a significant role in this process. So far, the most widely known PTM in Mycobacterium is serine/threonine (Ser/Thr) phosphorylation, which is shown to be important for its survival and virulence (1–4). In addition, Mycobacterium encodes two tyrosine phosphatases (PtpA and PtpB) that are secreted in the host phagosome during infection and are critical for pathogenesis (5, 6).

Lysine acetylation is a ubiquitous modification that is conserved from eukaryotes to prokaryotes (7–10). The initial evidence of protein lysine acetylation in mycobacteria came with the characterization of a NAD⁺-dependent deacetylase (11) and the identification of the first mycobacterial acetyltransferase (12). The acetyltransferases in Mycobacterium smegmatis (MSMEG_5458) and M. tuberculosis (Rv0998) contain a cyclic AMP (cAMP)-binding domain that is fused to a Gcn5-related N-acetyltransferase (GNAT)-like protein acetyltransferase domain (12). The mycobacterial acetyltransferase and deacetylase enzymes regulate the fatty acid metabolism via reversible acetylation of acetyl-CoA synthetase and fatty acyl-CoA synthetases (13–15). However, acetylation substrates in other biological pathways have yet to be revealed. The presence of extensive protein acetylation in several pathogenic and commercially utilized bacteria (10, 16–21) and the prediction of multiple acetyltransferases in Mycobacterium (Ref. 22 and the UniProt Consortium of 2010) led us to explore the abundance of acetylation in this bacteria.

In this study, we investigated the prevalence of lysine acetylation in mycobacteria and its correlation with various physiological processes. To achieve this, we applied an exhaustive approach to overexpress M. tuberculosis proteins and identify their lysine acetylation status in the surrogate host M. smegmatis. For this, 179 protein-coding genes were chosen, with an emphasis on regulatory proteins. This approach helped us to detect important acetylated proteins from different pathways, without being biased for the abundant proteins prominently selected in other proteomics approaches. Furthermore, the mass spectrometric analysis helped us to identify co-occurrence and the inter-relationship of other lysine modifications (succinylation and propionylation). The role of such lysine modifications in regulating enzyme activity has been exemplified using PtpB.

Experimental Procedures

In Silico Analysis

Gene names, protein names, and functional categories were gathered from the Tuberculist database and the UniProt database. Protein functional classes were obtained as described earlier by Camus et al. (23). Gene essentiality data were procured from Tuberculist and the previous studies documenting gene essentiality during different conditions, in vitro growth, infection, and growth on cholesterol-containing media (24–26).

Bacterial Strains and Gene Manipulation

Escherichia coli cells were grown and maintained with constant shaking (220 rpm) at 37 °C in LB medium supplemented with 25 μg/ml kanamycin or 100 μg/ml ampicillin as required. M. smegmatis MC² 155 cells were grown in Middlebrook 7H9 broth supplemented with 0.5% glycerol, 1% ADC (albumin/dextrose/catalase), and 0.05% Tween 80.

The genes coding for 179 proteins (supplemental Table S1) were PCR-amplified using M. tuberculosis H37Rv genomic DNA and primers containing NdeI and HindIII restriction sites. The amplicons were digested with the corresponding restriction enzymes and ligated to the E. coli-Mycobacterium shuttle vector pVV16, previously digested with the same enzymes. The ligated products were transformed into E. coli DH5α and were screened using restriction digestion and DNA sequencing. All the confirmed clones were then electroporated in the electro-competent M. smegmatis MC² 155 cells for overexpression and protein purification.

For co-expression, the genes coding for SahH (rv3248c, 1488 bp), PtpB (rv0153c, 831 bp), and PknG (Rv0410c, 2253 bp) were cloned in E. coli expression vectors pET28a (Novagen) or pGEX-5x-3 (GE Healthcare). M. tuberculosis acetyltransferase Rv0998 (rv0998, 1002 bp) and deacetylase Rv1151c (rv1151c, 714 bp) were cloned in MCS2 of pACYCDuet-1 vector in NdeI and XhoI restriction sites. pET28a-sahH, pET28a-ptpB, and pGEX-5x-3-pknG were co-transformed with either pACYCDuet-rv0998 or pACYCDuet-rv1151c in E. coli BL-21 cells to generate the acetylated (Ac-SahH/Ac-PtpB/Ac-PknG) and deacetylated (DeAc-SahH/DeAc-PtpB/DeAc-PknG) proteins, respectively. For in vitro acetylation assays, Rv0998 was cloned in pGEX-5x-3, and SigA (rv2703, 1587 bp) was cloned in pET28a.

Protein Purification

M. smegmatis MC² 155 transformants were grown in 200 ml of 7H9-ADC medium using kanamycin until the A₆₀₀ reached ∼1.0 (log phase). E. coli transformants were grown until A₆₀₀ ∼0.6–0.8 and induced overnight at 16 °C using 1 mm isopropyl 1-thio-β-d-galactopyranoside. The cells were harvested and disrupted by sonication in lysis buffer (50 mm Tris-Cl (pH 8.5), 300 mm NaCl, 5 mm β-mercaptoethanol, 1 mm phenylmethylsulfonyl fluoride (PMSF), and 1× protease inhibitor mixture (Roche Applied Science)). Proteins from inclusion bodies were purified using solubilization buffer (1.5% N-laurylsarcosine, 25 mm triethanolamine, and 2% Triton X-100 dissolved in lysis buffer). The supernatant obtained after removing the cell debris was incubated with Ni²⁺-NTA affinity resin (Qiagen) previously equilibrated with the equilibration buffer (50 mm Tris-Cl (pH 8.5), 300 mm NaCl, 5 mm β-mercaptoethanol, 1 mm PMSF, 10% glycerol, and 20 mm imidazole) for 2 h. After washing the resin with the wash buffer (50 mm Tris-Cl (pH 8.5), 1 m NaCl, 5 mm β-mercaptoethanol, 1 mm PMSF, 10% glycerol, and 20 mm imidazole), elution was carried out in the elution buffer (50 mm Tris-Cl (pH 8.5), 10% glycerol, 150 mm NaCl, 1 mm PMSF, and 200 mm imidazole). For glutathione S-transferase (GST)-tagged PknG and Rv0998 purification, previously published protocol was used (27). The purified proteins were resolved by SDS-PAGE and confirmed by immunoblotting.

Immunoblotting

The proteins belonging to different M. tuberculosis cellular fractions (cell membrane, cell wall, cytosol, and culture filtrate) were procured from Colorado State University (now BEI Resources), under the TB Vaccine Testing and Research Material Contract.

For immunoblotting, a similar protocol was followed as described previously (28). Briefly, the proteins were resolved by SDS-PAGE and transferred onto nitrocellulose membrane (Millipore). After 1.5 h of blocking the membrane with 3% BSA (Sigma) in PBST (phosphate-buffered saline (pH 7.2) containing 0.1% Tween 20) at room temperature, the blot was incubated for 1 h with primary antibodies. After five washes with PBST, the blot was incubated with secondary antibodies. After five washes, the blots were developed using Immobilon^TM western chemiluminescent HRP substrate kit (Millipore), according to the manufacturer's instructions. Histone (Sigma) and GST proteins were used as positive and negative controls, respectively, for lysine modifications. The antibodies and dilutions used were as follows: HRP-tagged anti-His₆ tag antibody (Abcam; 1:20,000 dilution); rabbit monoclonal anti-acetyl-lysine antibody (Cell Signaling; 1:2,000 dilution); mouse monoclonal anti-acetyl-lysine antibody (Cell Signaling; 1:5,000 dilution); pan anti-succinyl-lysine antibody (PTM Biolabs; 1:2,000 dilution); and pan anti-propionyl-lysine antibody (PTM Biolabs; 1:2,000 dilution). According to the manufacturer, the antibodies generated against the PTMs are specific for the particular lysine modification and do not cross-react (29–31).

Mass Spectrometric Analysis

Protein samples were resolved on SDS-PAGE and bands were visualized by Coomassie Brilliant Blue staining. The desired protein bands were cut from the gel and destained. The proteins were digested, and the peptides were loaded by an EASY-nLC II system (Thermo Scientific, Waltham, MA) onto a self-packed column as described by Bonn et al. (32). Separation of peptides was done using buffer A (0.1% (v/v) acetic acid) and buffer B (99.9% (v/v) acetonitrile and 0.1% (v/v) acetic acid) with a flow rate of 300 nl/min. All the samples were measured using a 45-min gradient (0 min-1% acetonitrile, 3–5, 25–25, 28–75, 29–99, 33–99, 34–1, 45–1), and to improve the separation of the peptides, selected samples (PpiA, PtpB, Pgk, and SahH) were also measured with a 100-min gradient (0 min-1% acetonitrile, 3–5, 69–25, 79–75, 80–99, 84–99, 85–1, 100–1). The LTQ Orbitrap Velos (Thermo Scientific) was operated in data-dependent MS/MS mode using enabled lock mass option. After full scan at a resolution of 30,000 in the Orbitrap, the 20 most intense precursor ions were selected for collision-induced dissociation, and fragment ions were recorded in the linear ion trap.

The acquired data were searched with the Sorcerer^TM-SEQUEST® 4 software (Sage-N Research, Milpitas, CA) against the M. tuberculosis decoy database. This database was composed of all the protein sequences of M. tuberculosis H37Rv (4,036 entries), downloaded from the Uniprot database. Using Bioworks Browser 3.2 EF2, a set of common contaminants as well as the reversed sequences were added to the database. The following parameters were applied: digestion by trypsin, up to two missed tryptic cleavages were allowed; peptide tolerance 10 ppm. Methionine oxidation (+15.99492 Da), cysteine carbamidomethylation (+57.021465 Da), as well as acetylation (+42.010571 Da), succinylation (+100.016044 Da), and propionylation (+56.026215 Da) on lysine were set as variable modifications. Proteins were considered as identified if at least two unique peptides matched solid quality criteria: ΔCn >0.1; XCorr >2.2, 3.3, and 3.75 for doubly, triply, or higher charged peptides.

Analysis of Orthologs of Acetylated Proteins

Acetylome datasets of different species were retrieved from individual studies as follows: Bacillus subtilis (10); E. coli (16, 33); Salmonella enterica (18); Rhodopseudomonas palustris (34); Plasmodium falciparum (35); Toxoplasma gondii (36, 37); Saccharomyces cerevisiae (38); Drosophila melanogaster (39); mouse (40–42); and human (43, 44). Amino acid sequences of orthologs of mycobacterial acetyl proteins were extracted from NCBI Protein Database and employed for multiple sequence alignment with the help of ClustalW2.

Structural Analysis of Acetylated Proteins

Protein structures of Rv0009 (PDB code 1W74) (45), Rv3846 (PDB code 1IDS) (46), Rv3248c (PDB code 2ZIZ) (47), Rv0467 (PDB code 1F61) (48), Rv0410c (PDB code 2PZI) (49), Rv0153c (PDB code 1YWF) (50), Rv1240 (PDB code 4TVO) (51), and Rv1438 (PDB code 3TA6) (52) were retrieved from the PDB database. For other proteins, homology modeling was performed as described earlier (53, 54). The templates were chosen for each protein, Rv1436 (PDB code 1VC2), Rv1437 (PDB code 1V6S), and Rv2889c (PDB code 1EFU), based on the sequence identities of 61, 49, and 39%, respectively. The three-dimensional protein models were generated using MODELLER 9 version 1. Initially, 50 models for each protein were generated, and their general features were evaluated on the basis of the MODELLER's energy and DOPE score. The detailed reliability indices were obtained by the SAVES server. Best models were chosen, and further refinements were carried out with the AMBER force field using 1,000 energy steps of minimization. Three-dimensional images of all the proteins were visualized using UCSF Chimera (55, 56), and the modified residues were located.

In Vitro Acetylation Assay

GST-tagged Rv0998 was used to catalyze the acetyl group transfer using the protocols of Nambi et al. (12, 13). Briefly, 3 μg of Rv0998 protein was incubated with 4–6 μg of substrates (His-PtpB, GST-PknG, or His-SigA) in acetylation buffer (25 mm Tris-Cl (pH 7.4), 100 mm NaCl, 5 mm EDTA, 1 mm cAMP, and 50 μm acetyl-CoA). The reactions were incubated at 25 °C for 30 min followed by inactivation using 5× SDS loading dye. The proteins were resolved using 10 or 12% SDS-PAGE. Immunoblotting was performed using rabbit anti-acetyl-lysine antibody to detect the acetylated proteins.

In Vivo Acetylation of PtpB

Deletion strains of Rv0998 homologs in Mycobacterium bovis BCG (ΔBCG_1055) and M. smegmatis (ΔMSMEG_5458) were obtained from previous studies (12, 13). pVV-ptpB was transformed in wild type and deletion strains, and His-PtpB was purified. The purified proteins were subjected to immunoblotting with anti-acetyl-lysine antibody followed by anti-His tag antibody.

Enzyme Activity Assays

p-Nitrophenol phosphate (pNPP) hydrolysis assay was performed to measure the activity of PtpB with a modified protocol of Sajid et al. (57). Briefly, 2 μg each of acyl-PtpB and DeAc-PtpB was added to a reaction containing 50 mm sodium acetate buffer (pH 5.5) and varying concentrations of substrate pNPP in a 96-well plate. The pH optima for PtpB activity is 5.5 (58); therefore, the reaction was carried out at this pH. During incubation at 37 °C, the increase in absorbance was monitored at 405 nm. Kinetic parameters were calculated by non-linear curve fitting using GraphPad Prism. Average values from three independent experiments were represented with the standard error.

For assessing the effect of PtpB on ERK1/2 phosphorylation, A549 (non-small lung carcinoma) and MIA PaCa-2 (pancreatic adenocarcinoma) cells were used. The cells were maintained in DMEM containing 10% (v/v) fetal bovine serum (FBS), 100 units/ml penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin-B in a humidified 5% CO₂ atmosphere. Cells growing in logarithmic phase were trypsinized and pelleted and were then subjected to lysis using cell lytic buffer (Sigma) containing protease inhibitor mixture (without phosphatase inhibitor). Equal amounts (20 μg) of lysates were incubated with different amounts of His-PtpB (0–2.5 μm) in dephosphorylation buffer, as mentioned above, for 30 min. The reaction was terminated using 5× SDS loading dye, and products were resolved on SDS-PAGE followed by immunoblotting using antibodies against pERK1/2, as explained earlier (59). Tubulin was used as a loading control.

Lactate dehydrogenase assay was performed for analyzing the activity of LldD1 using a modified protocol of Arora et al. (60). The reaction mixture contained 2 mm pyruvate and 2 mm NADH in 20 mm potassium phosphate buffer (pH 7.0). The reaction was started with the addition of 5 μg of purified LldD1, and a decrease in absorbance was measured at 340 nm. In vitro kinase assay of PknG was done as reported previously (61). Enzyme assay of SahH was essentially performed as described earlier (62) using 2 μg each of Ac-SahH and DeAc-SahH.

Results

Abundance of Acetylation in Mycobacteria

To ascertain the existence of lysine acetylation in mycobacteria, we performed immunoblotting experiments. Proteins from different cellular fractions of M. tuberculosis, cytosolic, cell membrane, cell wall, and culture filtrate, were probed with rabbit monoclonal anti-acetyl-lysine antibody. Multiple proteins were found to be acetylated in all the protein fractions with a relatively high prevalence in the cytosolic and cell membrane fractions (Fig. 1A). This experiment suggests the presence of widespread lysine acetylation in the mycobacterial proteins and led us to further explore its characteristics.

FIGURE 1.

Protein lysine acetylation in mycobacteria. A, immunoblot analysis of M. tuberculosis proteins representing different cellular fractions using rabbit monoclonal anti-acetyl-lysine antibody. B, functional classification of total M. tuberculosis proteins, protein targets selected, and proteins purified in this study. C, flow chart depicting stepwise experimental strategy followed throughout this study.

Selection of Target Proteins

To identify the acetylated proteins in mycobacteria, it was important to apply a comprehensive approach that can identify PTMs even in less abundant proteins. In M. tuberculosis, the proteins involved in metabolism, respiration, and cell wall-related processes form the majority of functional proteome as compared with the regulatory and signaling proteins (Fig. 1B) (23). Moreover, it has been shown in E. coli that the proteins involved in regulation and signaling represent the low copy number group as compared with the proteins involved in translation, protein folding, and other constitutive functions (63). Therefore, for our study, we selected 179 protein-coding genes from M. tuberculosis, a large number of which belonged to the less abundant regulatory and signaling proteins (Fig. 1B and supplemental Table S1). A schematic flow chart of the experimental strategy followed in this study is shown in Fig. 1C.

Expression and Purification of Recombinant Proteins from M. smegmatis

Selected genes from M. tuberculosis H37Rv were cloned in E. coli-Mycobacterium shuttle vector pVV16, allowing the constitutive expression of recombinant proteins with a C-terminal His₆ tag. After confirming the integrity of clones by DNA sequencing, the clones were used for protein overexpression in M. smegmatis MC² 155. For purification, the cell-free extracts were prepared by cell disruption in lysis buffer at 4 °C to maintain protein modifications, followed by Ni²⁺-NTA affinity chromatography. To confirm size and integrity of the purified proteins, the proteins were resolved on SDS-PAGE followed by immunoblotting using anti-His₆ tag antibody. With this method, we confirmed the expression of 71 proteins at their expected sizes, and most of the proteins were found to be purified to homogeneity (Fig. 2). The M. smegmatis 60-kDa chaperone (GroEL, MSMEG_1583) was found as an additional band in most of the protein purifications. Because of the presence of a C-terminal histidine-rich region, GroEL tends to co-purify with the overexpressed His-tagged proteins (64, 65). We checked the enzyme activities of three proteins to show that the proteins purified from M. smegmatis are properly folded and functionally active. These proteins, Rv0153c (PtpB), Rv0694 (LldD1), and Rv0410c (PknG), were found to be enzymatically active (data not shown).

FIGURE 2.

Purification of proteins from M. smegmatis. M. tuberculosis proteins were overexpressed in the surrogate host M. smegmatis and purified by Ni²⁺-NTA affinity chromatography. Immunoblotting was performed using antibodies against the His₆ tag. Immunoblots confirmed the identity of 71 His₆-tagged proteins (marked with blue arrows). Corresponding gene and protein names are labeled above the images. In many purifications, the 60-kDa chaperone GroEL was co-purified and recognized by anti-His₆ antibody due to the presence of a histidine- rich C-terminal region.

The functional categorization of 71 purified proteins is presented in Fig. 1B. Depending on the solubility and expression, the protein yield varied from 0.25 to 2.5 mg/liter (supplemental Table S1). Proteins such as PpiA (Rv0009), RpsF (Rv0053), GrpE (Rv0351), SenX3 (Rv0490), RplW (Rv0703), Mdh (Rv1240), Tpx (Rv1932), NdkA (Rv2445c), SpoU (Rv3366), and Rv3249c were obtained in higher amounts; and lower yields were observed for proteins LldD1 (Rv0694), EmbR (Rv1267c), LldD2 (Rv1872c), PpiB (Rv2582), VirS (Rv3082c), and Lsr2 (Rv3597c) (Fig. 2 and supplemental Table S1). Overall, we successfully purified 71 mycobacterial proteins (candidate targets) that were subsequently used for the assessment of lysine acetylation.

Acetylation Status of Candidate Targets

All the 71 candidate targets were investigated for the presence of acetylated lysine residues. The purified proteins were resolved on SDS-PAGE followed by immunoblotting using rabbit monoclonal anti-acetyl-lysine antibody. Among the candidate targets, we found 24 to be acetylated on lysine residues (Fig. 3A and Table 1). Selected protein substrates were further validated using a mouse monoclonal anti-acetyl-lysine antibody, which shows similar results (Fig. 3B). All the acetylated proteins (acetylome) were then classified according to their functional classes, and the percent abundance of each class in the total acetylome of 24 proteins was plotted (Fig. 3C, green bars). For comparison, the abundance of each functional class within the 71 candidate targets was also plotted (Fig. 3C, red bars). This classification showed that in the mycobacterial acetylome, there is an over-representation of proteins involved in “intermediary metabolism and respiration” (37.5%), “information pathways” (25%), and “virulence, detoxification, and adaptation” (12.5%) as compared with the candidate targets (18.3, 23.9, and 9.9%, respectively). In contrast, “regulatory proteins” (12.5%), “cell wall and cell processes” (8.3%), and “lipid metabolism” (4.2%) proteins were under-represented in the mycobacterial acetylome. Careful inspection indicated that 17 mycobacterial acetylated proteins (∼71%) are essential for M. tuberculosis survival and/or pathogenesis (supplemental Table S1)(24–26). Because there were only 37 (∼52%) essential proteins in the list of 71 candidate targets, enrichment of the essential proteins in the mycobacterial acetylome is not due to the selection bias.

FIGURE 3.

Lysine acetylation analysis of mycobacterial proteins. A, immunoblotting of the candidate targets was performed using rabbit monoclonal anti-acetyl-lysine antibody, and immunoblots are shown. Of these, 24 proteins were found to be acetylated at lysine residues (blue). Few of the non-acetylated proteins are also shown (red). Histone (black) and GST (red) were used as positive and negative controls, respectively. Corresponding gene and protein names are labeled above the images (see also Table 1). B, immunoblotting of candidate target proteins using mouse monoclonal anti-acetyl-lysine antibody. Acetylated substrates are marked blue. Histone (black) and GST (red) are used as positive and negative controls, respectively. C, acetylated proteins (green bars) were classified in different functional classes (y axis), and percent representation of each class (x axis) in the total acetylome is plotted. For comparison, all the 71 candidate targets were also classified in the same way and included in the plot (red bars).

TABLE 1.

List of 24 proteins modified by lysine modifications

The abbreviations used are as follows: Ac (acetylation), Suc (succinylation), and Pro (propionylation); TCA cycle, tricarboxylic acid.

Gene name	Protein name	Modifications	Essentiala	Functional category	Specific pathway
Rv0009	PpiA	Ac, Suc	C	Information pathways	Protein folding
Rv0053	RpsF	Ac, Suc	Y	Information pathways	Protein synthesis
Rv0153c	PtpB	Ac, Suc	C	Regulatory proteins	Protein modification
Rv0351	GrpE	Ac, Suc, Pro	Y	Virulence, detoxification, adaptation	Protein folding
Rv0410c	PknG	Ac, Suc, Pro	Y	Regulatory proteins	Protein modification
Rv0467	Icl	Ac, Suc, Pro	Y	Intermediary metabolism and respiration	Glyoxylate
Rv0491	RegX3	Ac	N	Regulatory proteins	Two-component system
Rv0685	EF-Tu	Ac, Suc	Y	Information pathways	Protein synthesis
Rv0694	LldD1	Ac, Suc, Pro	C	Intermediary metabolism and respiration	Glycolysis
Rv1240	Mdh	Ac, Suc, Pro	N	Intermediary metabolism and respiration	TCA cycle
Rv1436	Gap	Ac, Suc	Y	Intermediary metabolism and respiration	Glycolysis
Rv1437	Pgk	Ac, Suc, Pro	Y	Intermediary metabolism and respiration	Glycolysis
Rv1438	Tpi	Ac, Suc	Y	Intermediary metabolism and respiration	Glycolysis
Rv1617	PykA	Ac, Suc, Pro	Y	Intermediary metabolism and respiration	Glycolysis
Rv1694	TlyA	Ac, Suc	N	Virulence, detoxification, adaptation	Virulence
Rv1872c	LldD2	Ac, Suc, Pro	N	Intermediary metabolism and respiration	Glycolysis
Rv2611c	Acyl-transferase	Ac, Suc	Y	Lipid metabolism	Phospholipid synthesis
Rv2703	SigA	Ac, Suc, Pro	Y	Information pathways	Transcription
Rv2875	Mpt70	Ac, Suc	N	Cell wall and cell processes	Surface antigen
Rv2889c	EF-Ts	Ac, Suc	Y	Information pathways	Protein synthesis
Rv3248c	SahH	Ac, Suc	Y	Intermediary metabolism and respiration	Sulphur metabolism
Rv3366	SpoU	Ac, Suc	N	Information pathways	RNA modification
Rv3846	SodA	Ac, Suc, Pro	Y	Virulence, detoxification, adaptation	Detoxification
Rv3875	EsxA	Ac, Suc	N	Cell wall and cell processes	Surface antigen

^a Y indicates essential for in vitro growth; C indicates essential during growth on cholesterol as a carbon source; and N indicates non-essential.

Identification of Acetylation Sites and Their Conservation

The purified acetylated proteins were subjected to mass spectrometry for the identification of acetylation sites. Mass spectrometry identified 18 acetylation sites on 11 proteins (Table 2). All the proteins were modified at 1–2 lysine residues (Table 2 and supplemental Fig. S1).

TABLE 2.

List of acetylated proteins as identified by mass spectrometry

The corresponding peptide sequences and the acetylation sites have been shown.

Gene name	Protein name	Acetylated peptidea
Rv0009	PpiA	(K)TVANFVGLAQGTKDYSTQNASGGPSGPFYDGAVFHR(V) (Lys-50)
Rv0153c	PtpB	(R)FDTELAPEVVTFTKAR(L) (Lys-224)
Rv0410c	PknG	(R)ALLDLGDVAKATR(K) (Lys-520), (R)KLDDLAER(V) (Lys-524)
Rv0467	Icl	(K)KHLDDATIAK(F) (Lys-322), (R)GYTATKHQR(E) (Lys-392)
Rv1240	Mdh	(K)TGAAVTDIKK(M) (Lys-180), (R)IDKSTAELADER(S) (Lys-311)
Rv1436	Gap	(K)AIGLVMPQLKGK(L) (Lys-230), (K)GKLDGYALR(V) (Lys-232)
Rv1437	Pgk	(K)FAADSPPQTVDVGAVPNGLMGLDIGPGSIKR(F) (Lys-309)
Rv1438	Tpi	(K)IAFSLPDKYYDR(V) (Lys-34), (R)SVQTLVDGDKLR(L) (Lys-61)
Rv2889c	Ef-Ts	(R)AEGKPEQALPK(I) (Lys-214)
Rv3248c	SahH	(R)EYAEVQPLKGAR(I) (Lys-60), (K)HLDEKVAR(I) (Lys-456)
Rv3846	SodA	(K)HHATYVKGANDAVAK(L) (Lys-38), (R)AKEDHSAILLNEK(N) (Lys-53)

^a Acetylated lysine residue is underlined, and the corresponding residue number is written in parentheses.

Orthologs of the mycobacterial acetyl proteins were mined in the acetylomes of the organisms belonging to different prokaryotic and eukaryotic classes as follows: bacteria (B. subtilis, E. coli, S. enterica, and R. palustris); parasite (P. falciparum and T. gondii); fungus (S. cerevisiae); insect (D. melanogaster); mouse, and human. Of the 24 acetylated mycobacterial proteins, 15 (∼60%) were found to have their homologs in the acetylomes of other species (Table 3). Acetylation of several proteins was ubiquitous, including the metabolic proteins glyceraldehyde 3-phosphate dehydrogenase (Gap), malate dehydrogenase (Mdh), phosphoglycerate kinase (Pgk), and pyruvate kinase (PykA); translational elongation factors Ef-Tu and Ef-Ts; and superoxide dismutase (SodA). Acetylation of mycobacterial LldD2 is novel for bacteria as lactate dehydrogenase is documented to be acetylated only in the eukaryotic acetylomes, such as protozoan and humans (Table 3) (35, 37, 39). Nine proteins were found to be uniquely acetylated in mycobacteria, including PtpB, Ser/Thr protein kinase G (PknG), secreted immunogenic protein Mpt70, transcriptional regulatory protein RegX3, and secretory antigenic target EsxA.

TABLE 3.

List of organisms in which the orthologs of mycobacterial acetyl-proteins are acetylated

The abbreviations used are as follows: Tgo, T. gondii (36, 37); Pfa P. falciparum (35); Hsa, Homo sapiens (43, 44); Mmu, Mus musculus (40–42); Bsu, B. subtilis (10); Rpa, R. palustris (34); Sen, S. enterica (18); Eco, E. coli (16, 33); Sce, S. cerevisiae (38); and Dme, Drosophila melanogaster (39).

Proteins	Organisms in which orthologs are acetylated
Rv0009 (PpiA)	Tgo, Sce, Hsa, and Mmu
Rv0053 (RpsF)	Bsu and Eco
Rv0351 (GrpE)	Eco, Pfa, Sce, Hsa, and Mmu
Rv0467 (Icl)	Sen and Eco
Rv0685 (Ef-Tu)	Bsu, Eco, Tgo, Pfa, Sce, Hsa, and Mmu
Rv1240 (Mdh)	Bsu, Sen, Eco, Sce, Dme, Hsa, and Mmu
Rv1436 (Gap)	Bsu, Rpa, Sen, Eco, Tgo, Pfa, Sce, Dme, and Hsa
Rv1437 (Pgk)	Bsu, Rpa, Sen, Eco, Tgo, Pfa, Sce, Dme, Hsa, and Mmu
Rv1438 (Tpi)	Sen, Pfa, Sce, Hsa, and Mmu
Rv1617 (PykA)	Bsu, Rpa, Sen, Eco, Tgo, Pfa, Dme, Hsa, and Mmu
Rv1872c (LldD2)	Tgo, Pfa, Hsa, and Mmu
Rv2703 (SigA)	Sen
Rv2889c (Ef-Ts)	Bsu, Eco, Tgo, Pfa, and Sce
Rv3248c (SahH)	Rpa, Sce, Dme, Hsa, and Mmu
Rv3846 (SodA)	Bsu, Sen, Eco, Tgo, Pfa, Sce, Dme, Hsa, and Mmu

Next, we performed multiple sequence alignment of the selected mycobacterial acetyl proteins and their homologs in the diverse species (Fig. 4). Sequence analysis of Gap suggests that acetylation of Lys-230 is conserved among other bacterial acetylomes such as B. subtilis and R. palustris (Fig. 4A). At position 232, although the corresponding lysine residue is present in different species, acetylation is only found in mycobacteria. Gap-Lys-230 and Ef-Ts-Lys-214 are present only in bacteria, and their analogous lysine residues are absent in eukaryotes, implying their specific role in bacteria (Fig. 4, A and B). Pgk acetylation occurs at a completely novel residue Lys-309, and analogous lysine residues are absent in other organisms (Fig. 4C). Of the two acetylation sites each in isocitrate lyase (Icl) and S-adenosylhomocysteine hydrolase (SahH), one site (Lys-322 and Lys-60, respectively) is conserved, although the other (Lys-392 and Lys-456, respectively) is not conserved among other bacterial species (Fig. 4, D and E). These analyses show that although some of the mycobacterial acetylation sites are preserved in other organisms, the bacterium has evolved some unique protein acetylation events, probably to better adapt to the surrounding host environment.

FIGURE 4.

Homologs of acetylated mycobacterial proteins in other species. Multiple sequence alignments of acetylated mycobacterial proteins as follows: A, Gap; B, Ef-Ts; C, Pgk; D, Icl; and E, SahH. Abbreviations of different organisms are as follows: Mtb M. tuberculosis; Eco, E. coli; Bsu, B. subtilis; Rpa, R. palustris; Sty, S. enterica; Tgo, T. gondii; Sen, S. enterica; and Pfa, P. falciparum. Acetylated lysine residues in mycobacteria are colored green (numbered near the gray boxes), and acetylated lysine residues in other bacterial species are colored red. Conserved regions are shown in gray rectangular boxes.

Structure-Function Correlation of Acetylated Residues

To understand the relevance of acetylation sites with respect to the protein structure, we tried to establish the structure-function relationship for the 11 acetyl proteins in which acetylation sites were identified. Among these, the structures are known for eight proteins. For the remaining three proteins, we generated the three-dimensional structures using homology modeling. Acetylated lysine residues were located on each protein structure using UCSF Chimera. The analysis unequivocally shows that 9 out of 18 acetylated lysine residues (50%) exist in the loop regions connecting the secondary structures (such as SahH, PpiA, Gap, Ef-Ts, Icl, and Tpi) (Fig. 5). The other nine acetyl-lysines are present exclusively in α-helices. Additionally, the acetylated lysine residue was always found to be surface-exposed. Overall, this suggests that acetylation may play an important role in regulating protein structure, function, and binding with other interacting partners, a phenomenon that needs to be studied further.

FIGURE 5.

Structural analysis of acetylated proteins. Ribbon diagram representation of acetylated proteins showing the position of acetyl-lysine residue. Secondary structure depiction is as follows: helix (blue), strands (pink), and loops and turns (green). Acetylated lysine residue has been colored red. 9 out of 18 acetyl-lysine residues are present in the loop regions (numbered in blue).

Identification of Other Lysine Modifications on Acetylated Proteins

While analyzing the proteins by mass spectrometry for the presence of acetyl-lysines, we encountered a few aberrant mass shifts at lysine residues in certain peptides. Upon correlating them with other lysine modifications, we discovered the presence of succinylation and propionylation in these peptides (Table 4 and supplemental Fig. S2 and S3). Out of the 14 succinylation sites on 7 proteins, 8 were common with the acetylation sites on the same proteins. Additionally, in Pgk, propionylation and acetylation were found to be present at a common site.

TABLE 4.

List of succinylated and propionylated proteins as identified by mass spectrometry

The corresponding peptide sequences and the modification sites are shown. ^# indicates succinylated lysine residue. * indicates propionylated lysine residue.

Gene name	Protein name	Modified peptidea
Rv0009	PpiA	(K)IALFGNHAPK#TVANFVGLAQGTK(D) (Lys-37), (K)TVANFVGLAQGTK#DYSTQNASGGPSGPFYDGAVFHR(V) (Lys-50)
Rv0153c	PtpB	(R)FDTELAPEVVTFTK#AR(L) (Lys-224)
Rv1240	Mdh	(R)IDK#STAELADER(S) (Lys-311)
Rv1437	Pgk	(R)ETSK#NDDDR(R) (Lys-130), (R)VLEQLTSSTQRPYAVVLGGSK#VSDK(L) (Lys-208), (K)FAADSPPQTVDVGAVPNGLMGLDIGPGSIK*R(F) (Lys-309)
Rv1438	Tpi	(K)IAFSLPDK#YYDR(V) (Lys-34), (R)SVQTLVDGDK#LR(L) (Lys-61),(R)K#ELASLASPR(I) (Lys-197)
Rv2889c	Ef-Ts	(R)AEGK#PEQALPK(I) (Lys-214)
Rv3846	SodA	(K)HATYVK#GANDAVAK(L) (Lys-38), (K)GANDAVAK#LEEAR(A) (Lys-46), (R)AK#EDHSAILLNEK(N) (Lys-53), (K)NVK#VDFAK(A) (Lys-174)

^a The modified residue number is written at the end of the peptides.

Acyl group transfer from succinyl-CoA and propionyl-CoA frequently appears in eukaryotes and has also emerged in a few bacterial species (66–68). To further elucidate and validate the presence of these modifications in mycobacteria, we examined the mycobacterial acetylome for the presence of these modifications. Purified proteins were resolved on SDS-PAGE followed by immunoblotting with anti-succinyl-lysine and anti-propionyl-lysine antibodies. To our surprise, 23 out of 24 acetyl proteins demonstrated the presence of lysine succinylation (Fig. 6A), whereas 10 proteins exhibited lysine propionylation (Fig. 6B). The data strongly suggest the presence of these novel modifications in mycobacteria and the possibility of cross-talk between PTMs mediated by different acyl-CoAs.

FIGURE 6.

Succinylation and propionylation of acetylated mycobacterial proteins. The acetylated mycobacterial proteins were assessed for the presence of other lysine modifications. Immunoblots were probed using specific antibodies as anti-succinyl-lysine (A) and anti-propionyl-lysine are shown (B). Histone and GST were used as positive and negative controls, respectively. As evident, many acetylated proteins were found to be regulated by succinylation and propionylation.

Validation of Rv0998-mediated Lysine Acetylation and Role of Acetylation in Regulating Enzyme Activity

To validate the acetylation of proteins catalyzed by Rv0998, in vitro acetylation assays were performed. Purified GST-tagged Rv0998 was used to assess the acetylation of PtpB, PknG, and SigA using acetyl-CoA. The resulting proteins were analyzed by immunoblotting using rabbit anti-acetyl-lysine antibody. The results show that Rv0998 is able to catalyze the acetyl group transfer on these proteins (Fig. 7A).

FIGURE 7.

Effect of Rv0998-mediated acylation on PtpB activity. A, in vitro acetylation using Rv0998. Acetylation assays were carried out using PtpB, PknG, and SigA. Bands corresponding to different proteins are marked in red boxes. B, Michaelis-Menten plot of acyl-PtpB and DeAc-PtpB with the concentration of substrate pNPP at the horizontal axis. Relative values of V_max are also plotted considering the V_max of DeAc-PtpB as 100%. The kinetic parameters are mentioned in the table at the bottom. Acylation of PtpB decreases the V_max, but the K_m value remains unaltered. The error bars represent standard error of three independent values (p value = 0.0125). C, whole cell lysates of A549 and MIA PaCa-2 cell lines were treated with PtpB, and phosphorylation of ERK1/2 was detected using specific antibodies. Tubulin was used as a loading control. D, PtpB was overexpressed and purified from M. smegmatis (MS) and M. bovis BCG (MB), wild type (WT), and acetyltransferase deletion strains (KO). This was followed by immunoblotting with anti-acetyl-lysine and anti-His tag antibodies.

To further validate these results, three proteins, SahH, PtpB, and PknG, were co-expressed either with M. tuberculosis acetyltransferase (Rv0998) or deacetylase (Rv1151c) in E. coli. Similar co-expression system has been efficiently used to assess the phosphorylation in previous studies (54, 57, 62, 69). The purified proteins were analyzed for the presence of acetylation by mass spectrometry. The analysis identified lysine acetylation at Ac-SahH, Ac-PtpB, and Ac-PknG (co-expressed with acetyltransferase) at the same sites, although no acetylated residue was identified on DeAc-SahH, DeAc-PtpB, and DeAc-PknG (co-expressed with deacetylase) (Table 2). Interestingly, Ac-PtpB was also found to be succinylated at the same residue Lys-224 (Table 4) and was therefore renamed acyl-PtpB. Rv0998 has been previously reported to catalyze propionyl group transfer apart from acetyl transfer (13, 15); the present result suggests that it may also catalyze succinyl group transfer to PtpB.

Next, we analyzed the effect of acylation on the enzyme activity of PtpB. The phosphatase activity of PtpB was measured using pNPP as a substrate, and kinetic parameters were calculated. We found that V_max of acyl-PtpB was significantly decreased (to ∼45%) as compared with DeAc-PtpB, although the K_m value remained unaltered (Fig. 7B). This suggests that acylation at PtpB-Lys-224 decreases the rate of reaction, although the affinity for the substrate does not change. Because signaling proteins show amplification of signals they perceive, any minor change in their activity may be reciprocated as a major phenotypic change. This is supported by the analysis of dephosphorylation activity of PtpB using its physiological substrates ERK1/2 (70). Over small changes in PtpB concentration (0–2.5 μm), a significant difference in ERK1/2 dephosphorylation was observed (Fig. 7C), which is a phenomenon consistent in pancreatic carcinoma cells (MIA PaCa-2) as well as in lung cancer cells (A549). This indicates that PtpB concentration may be a rate-limiting step in ERK1/2 dephosphorylation, and any minute changes in activity of PtpB would significantly affect ERK1/2 phosphorylation status and thus IFN-γ-mediated IL-6 production (70).

SahH and PknG activities remain unaffected by their acetylation status (data not shown), suggesting a specific role of acetylation and/or succinylation in regulating PtpB activity. To the best of our knowledge, this is the first report showing acylation-mediated regulation of enzyme activity of a bacterial phosphatase.

Although the data suggest that Rv0998 is able to acetylate protein substrates, Rv0998-independent mechanisms may still exist as has been seen in E. coli (71). Therefore, the effect of deletion of rv0998 homologs in M. smegmatis and M. bovis BCG on the acetylation status of PtpB was analyzed. The results show that PtpB acetylation remains unaltered in the absence of a functional acetyltransferase homolog (Fig. 7D). It suggests the presence of other acetyltransferase(s) in Mycobacterium. Overlapping substrate specificity of enzymes catalyzing PTMs is also a common feature of Mycobacterium Ser/Thr protein kinases (4) indicating tight regulation of these modifications and importance for bacterial survival and pathogenesis. Further studies will be needed to identify the uncharacterized lysine acetyltransferase(s) responsible for PtpB acetylation in the absence of Rv0998 homologs and to characterize their role in physiology.

Co-occurrence of Other PTMs with Acetylation

We further scanned other modifications on the mycobacterial acetyl proteins by mining the previously published data of phosphorylation and ubiquitylation. Fifteen acetylated proteins (∼60% of the total acetylome) from this study were found to covalently associate with the prokaryotic ubiquitin-like protein (Pup) in M. tuberculosis (Table 5) (72). Pupylation of proteins at a specific lysine residue is a mechanism used by bacterial proteasomal machinery for protein degradation; thus, lysine modification may perform a pivotal role in protein stability by interfering with pupylation. Apart from pupylation, Ser/Thr phosphorylation is present on eight proteins (∼33% of the total acetylome) (Table 5) (4, 28, 62, 69, 73, 74). One such M. tuberculosis acetyl protein with known phosphorylation and pupylation sites is SahH (62, 72, 75). To understand the interaction among these PTMs, all the modification sites of SahH were located in its crystal structure, except Thr-2 that is not a part of the available crystal structure (Fig. 8). The figure shows that all the phosphorylation sites are away from the acetyl-lysines. The pupylation site (Lys-474) is present in the same domain as the acetylated Lys-456 residue. Together, these observations suggest the existence of an intricate regulatory network among different PTMs in Mycobacterium.

TABLE 5.

List of previously known phosphorylation and pupylation on M. tuberculosis acetyl proteins

Gene name	Protein name	Previously known modifications
Rv0410c	PknG	Phosphorylation (74), pupylation (72)
Rv1617	PykA	Phosphorylation (28, 98), pupylation (72)
Rv0685	Ef-Tu	Phosphorylation (69), pupylation (72)
Rv0351	GrpE	Phosphorylation (4), pupylation (72)
Rv2875	Mpt70	Phosphorylation (4)
Rv1240	Mdh	Phosphorylation (4), pupylation (72)
Rv1872c	LldD2	Phosphorylation (28), pupylation (72)
Rv0467	Icl	Pupylation (72)
Rv3846	SodA	Pupylation (72)
Rv3248c	SahH	Phosphorylation (62, 75), pupylation (72)
Rv0053	RpsF	Pupylation (72)
Rv2889c	Ef-Ts	Pupylation (72)
Rv1436	Gap	Phosphorylation (73), pupylation (72)
Rv1437	Pgk	Pupylation (72)
Rv1438	Tpi	Pupylation (72)
Rv0694	LldD1	Pupylation (72)

FIGURE 8.

Crystal structure of M. tuberculosis SahH and localization of modification sites (color-coded). The residue numbers are labeled. Secondary structure depiction is as follows: helix (blue), strands (pink), and loops and turns (green).

Discussion

There are more than 450 different types of PTMs (The UniProt Consortium of 2014), and among them, protein lysine acetylation is the second most abundant reversible PTM after protein phosphorylation (76). In mycobacteria, only few in vivo acetylation targets had been identified such as MSMEG_4207 in M. smegmatis and FadD13 in M. bovis BCG (12, 13). This study deals with the identification of lysine acylation of M. tuberculosis proteins, where each protein has been independently assessed for being modified. Unlike M. tuberculosis acetyltransferase, which is a cAMP-dependent enzyme, M. smegmatis acetyltransferase is only a cAMP-regulated enzyme and possesses considerable activity even in the absence of cAMP (77). Therefore, our strategy imparts a favorable environment to the M. tuberculosis proteins for their acetylation in an un-induced native condition. With this strategy, we observed 24 acetylated proteins among the 71 candidate targets. The two pathways that represent a majority of the M. tuberculosis acetylome were (i) central carbon metabolism and (ii) protein synthesis and modification (Table 1). Eight acetylated proteins are known to be involved in central carbon metabolism, with six glycolytic enzymes, one glyoxylate cycle enzyme, and one citric acid cycle enzyme. Acetylation-mediated regulation of enzymes involved in carbon metabolism is a highly conserved phenomenon among different organisms (Table 3), and its presence in mycobacteria points toward an additional layer of regulation. Besides central metabolism, five acetylated proteins are involved in protein synthesis (Ef-Tu, Ef-Ts, and RpsF), folding, and stability (PpiA and GrpE). The virulence-associated proteins, PknG and PtpB, were also found to be acetylated, connoting an ancillary means for their regulation in addition to phosphorylation. Interestingly, secretory proteins like EsxA, Mpt70, and PpiA were also found to be acetylated.

The structural analysis suggests that 50% of acetylation sites occur in loop regions connecting the secondary structures. Interestingly, the acetylated Lys-50 of PpiA occurs in a loop region that is unique to the mycobacterial protein and is part of an immunogenic peptide (78). The acetylated/succinylated Lys-224 of PtpB is part of a large lid region where Phe-222 obstructs the entry of substrate to the active site (50). This lid needs to move out for the protein to become active, and Lys-224 may have a role in regulating this activation. Decreased V_max of acyl-PtpB strengthens this hypothesis and speculates that lysine acylation could be a novel regulatory mode for PtpB function.

Lysine residues in proteins also undergo other types of acylations such as succinylation, propionylation, and butyrylation (66, 79). Recently, a proteomic study explored more than 1300 lysine succinylation sites in E. coli, yeast, human, and mouse proteomes. The study not only demonstrated widespread lysine succinylation, but it also discerned that 66, 56, 27, and 57% of the succinylation sites overlap with acetylation sites in these organisms, respectively (67). In correlation, we found more than 40% (10 proteins) of the mycobacterial acetylome to be modified either by propionylation or succinylation, indicating their inter-relationship. Moreover, our results denoted that ∼95% of succinylated proteins and ∼55% of succinylation sites overlap with that of acetylation. Different lysine modifications may be catalyzed by similar enzymes as M. tuberculosis acetyltransferase Rv0998 has been shown to propionylate FadD13 and acyl-CoA synthetase at the same lysine residue that is acetylated (13, 15). Succinylation of PtpB co-expressed with Rv0998 supports this argument.

Propionyl-CoA is the by-product of β-oxidation of odd-chain fatty acids, branched-chain fatty acids, branched-chain amino acids and cholesterol, and it is toxic if accumulated (80). Three different mechanisms exist for the alleviation of this toxicity as follows: (i) the methylcitrate cycle (81, 82), (ii) the methylmalonyl pathway (83), and (iii) the incorporation into cell wall lipids (84). Additionally, propionylation of acyl-CoA synthetase in M. bovis BCG leads to diminished synthesis of propionyl-CoA from propionate (13). Identification of a considerable number of propionylated proteins in our study defines another escape route to relieve the toxicity where these proteins may act as an additional sink for propionyl-CoA. Moreover, because 2-methylisocitrate lyase activity of Icl (a key component of glyoxylate pathway) is essential for the methylcitrate cycle (82, 85), propionylation of Icl may have a dual implication, i.e. to modulate Icl activity in response to the propionyl-CoA concentration and simultaneously to sequester excess of propionyl-CoA to relieve the toxicity. Succinyl-CoA is synthesized as a by-product of the methylmalonyl pathway (83), and it is suggested that at slow growth rates there is an increased flux toward succinyl-CoA (86). A GNAT-homologous succinyltransferase, Rv0802c, has been structurally characterized in M. tuberculosis, and it is possible that under these conditions Rv0802c utilizes succinyl-CoA and transfers it to the protein targets (87).

Very recently, a few other parallel studies reported acetylation or succinylation in mycobacterial species thus reaffirming the importance of acylation in Mycobacterium (88–91). ∼65% of the present acetylome overlaps with the proteins identified in either of these studies. Interestingly, out of the 24 acetyl proteins, eight protein, PtpB, RegX3, LldD1, TlyA, Rv2611c, Mpt70, SpoU, and EsxA, were found to be unique to this study. All the other studies were based on proteomics approaches, and thus, identification of novel substrates suggests the potential of the present strategy in recognizing the PTMs in less-abundant proteins, including PtpB. Although this study utilizes a smaller dataset (71 candidate targets), a comparison with the recent studies suggests that similar features of mycobacterial protein lysine modifications are identified in our study (88–91). A majority of modified proteins are found to be associated with metabolism and are cytosolic. Also, negative regulation of protein function was found in three such studies (88, 90, 91).

Analysis of the mycobacterial acetylome for the presence of other modifications such as pupylation and phosphorylation reveal novel signaling nodes in Mycobacterium. Mycobacterial signaling machinery may be viewed as having individual modules such as phosphorylation (4), pupylation (72), and nitrosylation (95). Although pupylation is specifically responsible for protein degradation (72), nitrosylation could occur during infection of the host (95). Phosphorylation is the only known protein modification in Mycobacterium that is multifaceted and regulates diverse processes ranging from cell shape to adaptation inside the host. However, this module is convoluted with the presence of only one phosphatase against 11 serine/threonine protein kinases and by having overlapping specificities of the serine/threonine protein kinases (4, 96). Identification of widespread lysine modifications offers an alternative mechanism to regulate the impact of protein phosphorylation, either as a rival or an ally, as first proposed by Kouzarides (97). M. tuberculosis harbors two tyrosine phosphatases (PtpA and PtpB), both of which are secreted into the macrophage phagosomes. PtpA inhibits the phagosome-lysosome fusion (5). PtpB subverts the innate immune signaling by blocking ERK1/2 and p38-mediated pathways (70). PtpA has been previously shown to be regulated by PTMs such as S-nitrosylation (92) and phosphorylation (93, 94), but no modifications were known to occur on PtpB. This is the first study that depicts the regulation of M. tuberculosis PtpB enzyme activity by any PTM. Thus, the role of PTMs in regulating PtpB function and thus M. tuberculosis pathogenesis demands further research. Such cross-talk between phosphorylation and acetylation may help in efficient inter- and intracellular communication mechanisms. Identification of an acetylation module in the mycobacterial signaling establishes a framework to understand the regulatory networks and a detailed analysis is required to completely understand the biology of the pathogen.

Author Contributions

A. Singhal and G. A. conceptualized the study, designed the experiments, and wrote the manuscript. A. Singhal, G. A., R. V., P. K., T. K., A. Sajid, and R. M. performed the experiments. J. J. generated the protein structures using homology modeling. V. Y. helped with the experiment in Fig. 7C. S. S. performed the experiments with M. smegmatis and M. bovis BCG knockout strains provided by S. S. V. C. H., D. B., and U. G. performed and analyzed the mass spectrometric data of protein modifications. N. S., A. K. P., S. S. V., U. G., and Y. S. provided the funding and useful resources. All authors read and approved the final version of the manuscript.