A Mycobacterial Phosphoribosyltransferase Promotes Bacillary Survival by Inhibiting Oxidative Stress and Autophagy Pathways in Macrophages and Zebrafish*

Background: Several Mycobacterium tuberculosis glycoproteins are involved in tuberculosis pathogenesis. Results: Mycobacterium tuberculosis Rv3242c and Mycobacterium marinum mimG enhance bacillary survival by inhibiting oxidative stress and autophagy pathways in macrophages and zebrafish. Conclusion: Rv3242c and mimG aid intracellular bacterial persistence by modulating host immune responses. Significance: This study has identified a novel virulence factor, which can be considered as drug target for tuberculosis treatment. Mycobacterium tuberculosis employs various strategies to modulate host immune responses to facilitate its persistence in macrophages. The M. tuberculosis cell wall contains numerous glycoproteins with unknown roles in pathogenesis. Here, by using Concanavalin A and LC-MS analysis, we identified a novel mannosylated glycoprotein phosphoribosyltransferase, encoded by Rv3242c from M. tuberculosis cell walls. Homology modeling, bioinformatic analyses, and an assay of phosphoribosyltransferase activity in Mycobacterium smegmatis expressing recombinant Rv3242c (MsmRv3242c) confirmed the mass spectrometry data. Using Mycobacterium marinum-zebrafish and the surrogate MsmRv3242c infection models, we proved that phosphoribosyltransferase is involved in mycobacterial virulence. Histological and infection assays showed that the M. marinum mimG mutant, an Rv3242c orthologue in a pathogenic M. marinum strain, was strongly attenuated in adult zebrafish and also survived less in macrophages. In contrast, infection with wild type and the complemented ΔmimG:Rv3242c M. marinum strains showed prominent pathological features, such as severe emaciation, skin lesions, hemorrhaging, and more zebrafish death. Similarly, recombinant MsmRv3242c bacteria showed increased invasion in non-phagocytic epithelial cells and longer intracellular survival in macrophages as compared with wild type and vector control M. smegmatis strains. Further mechanistic studies revealed that the Rv3242c- and mimG-mediated enhancement of intramacrophagic survival was due to inhibition of autophagy, reactive oxygen species, and reduced activities of superoxide dismutase and catalase enzymes. Infection with MsmRv3242c also activated the MAPK pathway, NF-κB, and inflammatory cytokines. In summary, we show that a novel mycobacterial mannosylated phosphoribosyltransferase acts as a virulence and immunomodulatory factor, suggesting that it may constitute a novel target for antimycobacterial drugs.

by macrophages, whereas pathogenic strains overcome the host immune responses and eventually cause the disease.
Invading non-pathogenic mycobacteria are first engulfed by macrophages by means of phagosomes, where they are exposed to antibacterial effector molecules, such as reactive nitrogen intermediates and reactive oxygen species (ROS). Then the phagosomes mature and sequentially fuse with increasingly hostile organelles of the endocytic pathway. In the highly acidic and hydrolytic milieu of lysosomes, essential components of the bacteria are degraded, thereby causing bacterial death (4). By contrast, pathogenic mycobacteria are able to survive in macrophages for extended period of time by manipulating their immune functions (5). These manipulative strategies include prevention of phagolysosome fusion, synthesis of virulence factors, inhibition of phagosome acidification due to depletion of vesicular proton-ATPase, evasion from toxic effects of nitric oxide (NO) and ROS, suppression of protective cytokine synthesis and T-helper 1 responses, and inhibition of apoptosis (6 -9). Moreover, the alarming emergence of multidrug-resistant forms of mycobacteria has rendered the existing therapeutic arsenals ineffective. One reason for this paradox is that the molecular basis of the pathogenicity of M. tuberculosis is still poorly understood. Numerous studies have dissected some of these mechanisms in detail, whereas others require further understanding. In this context, basic understanding of the metabolic pathways that are critical for the survival and adaptation of M. tuberculosis under different patho-physiological conditions of the host is important.
Mycobacterium marinum, which causes chronic TB in fish, is a model species for M. tuberculosis infection (10). Less commonly, it is also able to cause disease in humans in the form of skin lesions (11). M. marinum is the closest genetic relative to the M. tuberculosis complex and shares virulence factors with M. tuberculosis. Zebrafish and mice infected with M. marinum develop chronic granulomatous disease that is very similar to what is found in human pulmonary disease (11). Due to their genetic and pathological similarities, it is thought that the two organisms share similar mechanisms of establishing disease and modulate host immune responses.
In prokaryotic systems, glycosylated proteins are known to play critical roles in immunogenicity and pathogenicity (12)(13)(14)(15)(16). However, in the context of TB, the role of protein glycosylation is still ill-defined (17). To date, quite a few M. tuberculosis glycoproteins have been identified and characterized as to their role in pathogenesis (18). Among them, the mannose-containing glycoconjugates mannose lipoarabinomannan, the 60-kDa glycoprotein Apa, and Mpb83 of mycobacteria were reported to play a role in host-pathogen interactions and to facilitate the entry of pathogens into phagocytes (19 -23). A few mycobacterial cell wall glycolipids, such as lipoarabinomannans, mannose lipoarabinomannans, and phosphatidylinositol mannosides, play major roles in blocking phagosomal maturation (24).
Purine phosphoribosyltransferases (PRTs) are important enzymes in purine salvage pathways, which are essential for the survival of a number of bacterial species, including mycobacteria (25, 26). PRTs catalyze the reversible transfer of a phosphoribosyl group from phosphoribosylpyrophosphate to a purine base (27-29). Free-living organisms can produce purine nucle-otides either by de novo synthesis or by the salvage of preformed bases. In contrast, many parasitic organisms are unable to synthesize purines de novo and thus depend on enzymes of salvage pathways for the synthesis of purine nucleotides (30). For this reason, such enzymes, including PRTs, were proposed as potential targets for the treatment of parasitic diseases. Recently, the annotation of the M. tuberculosis genome suggested the presence of about 19 putative PRTs (31), most of which still have to be studied by experiment. One of the PRTs characterized in detail is hypoxanthine-guanine phosphoribosyltransferase, encoded by gene Rv3624 (31). In M. tuberculosis, the PRTs are especially significant because they play a role in the biosynthesis of arabinose, which is the main sugar in M. tuberculosis cell walls (32). Because the complex, multilayered cell wall of M. tuberculosis is a major virulence factor and also contributes to the development of drug resistance, special attention is being turned to the development of drugs that inhibit bacterial cell wall biosynthesis.
Mitogen-activated protein kinases (MAPKs) are involved in relaying extracellular signals to intracellular responses. Several studies suggest that the MAPK pathway also affects mycobacterial pathogenesis (33). Thus, it was shown that the intracellular growth of Mycobacterium avium in macrophages depends on the extent of MAPK phosphorylation, indicating a role of the pathway in macrophage activation. The MAPK family includes a large number of kinases (e.g. ERK, p38 MAPK, and c-Jun N-terminal kinase) (34). Activation of MAPK is induced by infection with M. tuberculosis and is essential for the mycobacterium-induced production of proinflammatory cytokines (33-35). In addition, autophagy, a fundamental process in eukaryotic cells, can also capture and eliminate intracellular pathogens, including M. tuberculosis. This activity appears to depend on the interaction between pathogen-associated molecular patterns and pathogen recognition receptors (36).
Our recent in silico studies suggested that the M. tuberculosis cell wall may contain several more uncharacterized glycoproteins and glycosyltransferases (17, 37). As previously demonstrated by several other studies, mannosylation is the most common glycosylation pattern present in the cell wall of M. tuberculosis (38). Moreover, several mannosylated glycoconjugates have been implicated as playing a pivotal role in M. tuberculosis pathogenesis (39, 40). The main objective of the present work was to identify a novel glycoprotein(s) encoded by the M. tuberculosis genome and to elucidate its role(s) in the pathogenesis of mycobacteria. To this end, we employed a glyco-catch method to capture mannose-containing proteins from purified cell walls of M. tuberculosis. A number of studies have shown that such proteins contribute to pathogenicity of mycobacteria and are essential for mycobacterial survival (22,41,42). By mass spectrometry, one of the mannose-binding proteins, encoded by the Rv3242c gene, was identified as phosphoribosyltransferase. By expression and deletion of this gene, we were able to show that PRT functions as an important virulence factor by modulating the innate immune responses in macrophages and in a zebrafish tuberculosis model.

Ethics Statement
Mice used for the isolation of peritoneal macrophages were housed at the School of Biotechnology (KSBT), KIIT University, in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA) in India. Mouse peritoneal macrophage isolation was conducted in accordance with a procedure approved by the KSBT Institutional Animal Ethical Committee (KSBT/IAEC/2013/MEET1/A11). Every effort was made to minimize suffering and ensure the highest ethical and humane standards. Zebrafish were handled in compliance with local animal welfare regulations and maintained according to standard protocols (see the Zebrafish Model Organism Database Web site). All of the zebrafish use and experimental protocols in this research were reviewed and approved by the Institutional Animal Care and Use Committee of the Indian Institute of Science. Fish were anesthetized in an aqueous solution of tricaine (Sigma, catalog no. E10521-50g).

ConA Affinity Purification and Lectin Hybridization
Delipidation of M. tuberculosis cell walls was performed as described previously (44). The delipidated pellet was resuspended in 200 l of lectin-binding buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1 mM CaCl 2 , and 1 mM MgCl 2 ) and stored at Ϫ20°C for further use. To 100 l of agarose-conjugated ConA lectin, 1 ml of equilibration buffer (1ϫ PBS, 1 mM EDTA, 2 mM CaCl 2 ) was added. The beads were mixed gently and centrifuged at 2000 rpm for 5 min. The above steps were repeated twice. Then 100 l of the delipidated cell wall sample was added to the equilibrated agarose-conjugated lectin, allowed to bind for 4 -6 h at room temperature with gentle shaking, and then centrifuged at 2000 rpm for 5 min. The beads were then washed with 1 ml of 1ϫ PBS 3-4 times and centrifuged at 2000 rpm for 5 min. The bound proteins were eluted by the addition of 50 l of 0.4 mM competitive sugar solution. The samples were separated on 10% SDS-polyacrylamide gels and transferred to a PVDF membrane. The membrane was blocked with 2% Triton X-100 in 1ϫ PBS for 10 min and then washed three times with 0.5% Triton X-100 in 1ϫ PBS for 5 min. 10 l of ConA-HRP or biotinylated lectins in 5 ml of wash buffer was added on the membrane and incubated for 4 -5 h on a rocker (overnight in the case of biotinylated lectins) at room temperature. 5 l of streptavidin-HRP in 5 ml of wash buffer was added to the membrane and incubated at room temperature for 1 h with gentle shaking. The blots were washed 3-5 times with wash buffer for 10 min. Finally the bands were visualized using chemiluminescence detection kit (Thermo Scientific, Rockford, IL). Gels were stained for glycoprotein using the ProQ Emerald glycoprotein stain according to the manufacturer's instructions (Molecular Probes, Inc.).

Mass Spectrometric Analysis
Sample Preparation-The protein bands were excised from the gel and washed with dH 2 O, and the pieces were chopped into ϳ1-mm slices, which were then rehydrated in 50 mM NH 4 HCO 3 and acetonitrile (1:1) for 15 min. After removal of solvents, the gel pieces were dehydrated by the addition of acetonitrile and dried under vacuum. Then gel particles were reduced by the addition of freshly prepared 10 mM dithiothreitol (DTT) for 45 min at 56°C. Freshly prepared 50 mM iodoacetamide in 50 mM NH 4 HCO 3 was added, and the suspension was incubated for 30 min in the dark at room temperature. The particles were washed twice with a 50 mM NH 4 HCO 3 and acetonitrile mixture (1:1) for 15 min. Finally, the material was once more dehydrated in acetonitrile followed by vacuum drying.
For in-gel digestion, gel pieces were incubated in 20 ng/l trypsin (Himedia, Mumbai, India) prepared in 25 mM NH 4 HCO 3 at 37°C for 16 h. Following digestion, the peptides were extracted by the addition of extraction buffer (50% acetonitrile in 0.1% trifluoroacetic acid (TFA)) followed by sonication. The digested samples were kept at Ϫ20°C until further analysis.
LC-ESI and MALDI Mass Spectrometry-For LC-ESI analysis, digested samples were fractionated by HPLC (Dionex Ultimate 3000) using an Acclaim 120 A 0 C18 column (4.6 ϫ 150 mm, 3-m particle size). The gradient used was as follows: solvent A: 0.1% formic acid in H 2 O; solvent B: 0.1% formic acid in ACN; gradient: 0 min 95% A; 0 -40 min 95-5% A (linear); 40 -45 min 5% A (isocratic); 45-50 min 5-95% A. Fractions were collected at a flow rate of 0.2 ml/min. Then the fractions were injected into the Fourier transform ion cyclotron resonance ESI-mass spectrometer (7.0 T Apex Ultra, Bruker Daltonics, Bremen, Germany) at a flow rate of 120 l/h for analysis. The data were analyzed using DataAnalysis software version 4.0 (Bruker Daltonics). The obtained spectra were deconvoluted and searched in MASCOT using BioTools version 3.2 software (Bruker Daltonics). The SIB-FindMod Web server was used to enrich the peptide masses relevant for the protein of interest.
For MALDI analysis, 1 l of ␣-cyanocinnamic acid matrix (10 mg/ml in 50% acetonitrile and 0.1% TFA) was mixed with 1 l of digested protein samples. The mixture was spotted on a MALDI plate and allowed to form the crystals, and the peptide spectrum was acquired in an AUTOFLEX III Smartbeam MALDI TOF/TOF instrument (Bruker Daltonics) having an Nd:YAGsmart laser beam of 335-nm wavelength. External calibration was done with a peptide calibration standard supplied by Bruker, with masses ranging from 1046 to 3147 Da. The obtained spectra were acquired and analyzed using FlexControl version 3.0 software in reflectron ion mode with an average of 2000 laser shots at a mass detection range between 750 and 3100 m/z. The data were analyzed using flexAnalysis software version 3.0 (Bruker Daltonics) and searched in MASCOT using BioTools version 3.2 software (Bruker Daltonics).

In Silico Protein Function Prediction
The protein functions were predicted as described previously (37). Briefly, the protein sequence homology analysis was carried out using PSI-BLAST, CDD, COG, ScanProsite, and PFAM (one-dimensional analysis). Conserved motif pattern analysis (two-dimensional analysis) was performed by ClustalW multiple-sequence alignment, and a phylogenetic tree was constructed using the PHYLIP software.
The presence of specific functional folds in the given sequence was analyzed by PHYRE fold recognition analysis (45). The transmembrane prediction was done using the TMHMM, HMMTOP, and the TMPred server. Glycosylation sites were predicted using NetNGlyC, NetOGlyc, Ensemblegly, ProGlycProt, and the GlycoPP Web server. Dompred domain prediction was used to check the number of domains. The Glycomod, Glycoworkbench, and SimGlycan tools were used for the prediction of possible glycan structure. DNA binding prediction was done using DNA Binder.

Molecular Simulation
The ab initio modeling of the Rv3242c was carried out using the I-TASSER Web server (46). The homology modeling of the protein sequence was done using Protein Data Bank entry 1ORO-A as a template. Modeler 9v9 was used to build the PRT domain of the protein using 1ORO-A as a template. The quality of the modeled protein structure was analyzed using the SAVES Web server. The stability of modeled protein was checked by molecular dynamics simulation using the Gromacs software package. First, the modeled protein was energy-minimized using the OPLS (optimized potentials for liquid simulations all-atom) force field. The modeled protein was solvated in water molecules in order to mimic the physiological behavior of the molecules. Finally, molecular dynamics simulation was carried out to examine the quality of the modeled structures by checking their stability using 1-ns simulation, and calculating the lowest energy structure found during the simulation.

Cloning and Expression of Rv3242c
M. tuberculosis Rv3242c and Rv3242c orthologue in M. smegmatis (MSM_1877) were PCR-amplified using gene-specific primers (Table 2) using M. tuberculosis and M. smegmatis genomic DNA as templates. The PCR-amplified products were gel-purified, sequentially digested with PstI and HindIII, and cloned into pSMT3 shuttle vector. The recombinant constructs were transformed into competent E. coli XL-10 gold. The positive colonies were selected on LB agar plates supplemented with 20 g/ml tetracycline and 50 g/ml hygromycin. The positive colonies were confirmed by colony PCR and sequencing using gene-specific primers. Finally, the recombinant constructs were transformed into electrocompetent M. smegmatis. The positive colonies were selected on 7H9 medium containing 50 g/ml hygromycin. The positive transformants were confirmed by colony PCR and sequencing using gene-specific primers.

Phosphoribosyltransferase Assay
Wild type (WT) M. smegmatis, M. smegmatis pSMT3 vector control (pSMT3), recombinant MsmRv3242c, and MSM_1877 strains were grown as described above. 10 8 cfu were harvested at 0, 4, 12, and 24 h; resuspended in 0.5 ml of 0.5ϫ TBE buffer; and sonicated at 35 kHz for 10 min with interruptions every 2 min for cooling. The samples were analyzed for PRT activity using a phosphoribosyl pyrophosphate assay kit (Novocib, Lyon, France) according to the manufacturer's instructions.

Invasion Assay
M. smegmatis culture in midexponential growth phase was pelleted, washed in PBS, pH 7.4, and resuspended in DMEM to a final A 600 of 0.1. Bacterial clumps were removed by ultrasonication of bacterial suspensions for 5 min, followed by a low speed centrifugation for 2 min. Non-phagocytic human epithelial HCT-116 cells (2 ϫ 10 5 cells/well) were seeded onto 24-well tissue culture plates. The next day, cells were washed with 1ϫ PBS, and fresh DMEM without antibiotics was added to the cells 2 h prior to infection. Then the cells were infected with M. smegmatis WT, pSMT3, MsmRv3242c, and M. smegmatis expressing MSM_1877 (an orthologue of Rv3242c in M. smegmatis) strains at a multiplicity of infection (MOI) of 10 for 2 h. After infection, extracellular bacteria were killed by the addition of 20 g/ml gentamycin. Then cells were washed with 1ϫ PBS and lysed with 0.5% Triton X-100 at the indicated time points, and intracellular bacterial count was estimated by plating serially diluted samples on 7H10 plates. M. smegmatis colonies were enumerated after 3 days.

Cell Adhesion Assay
Human colon colorectal carcinoma cells (HCT116) were seeded onto coverslips in a 24-well polystyrene plate (50,000 cells/well). Next day, cells were washed with 1ϫ PBS, and fresh DMEM without antibiotics was added to the cells 4 h prior to the infection. Then the cells were infected with M. smegmatis pSMT3 and MsmRv3242 labeled with 100 M carboxyfluorescein diacetate stain in PBS for 1 h at 37°C as described above with an MOI of 10. After 10 min and 1 h of infection, cells were washed with 1ϫ PBS to remove unadhered bacteria. Then the cells were fixed with 4% paraformaldehyde for 1 h at 37°C. The fixed cells were stained with rhodamine-phalloidin. After 1 h of staining cells were washed with 1ϫ PBS. The coverslips were mounted onto grease-free glass slides with mounting solution containing DAPI. The mounted coverslips were observed under a confocal laser scanning microscope (ZEISS LSM 780). The acquired images were analyzed with Zeiss 2009 image software.

Isolation of Mouse Peritoneal Macrophages
Peritoneal macrophages were collected from 4 -6-week-old female BALB/c mice by intraperitoneal injection of 10 ml of phosphate-buffered saline supplemented with 10% FBS and centrifuged at 400 ϫ g for 10 min at 4°C. The pellet was resuspended in DMEM supplemented with 10% FBS, 2 mM L-glutamate, and 1% penicillin-streptomycin solution. Approximately 2 ϫ 10 5 cells/well were plated in a 24-well plate and incubated at 37°C in 5% CO 2 . After overnight incubation, the non-adherent cells were washed off with warm phosphate-buffered saline. The cells were then cultured for 3 days, and on the fourth day, the cells were subjected to infection.

Intracellular Survival Assay
Intracellular survival assay was performed in mouse peritoneal, RAW264.7, and human THP-1 macrophages. Cells (2 ϫ 10 5 cells/well) were seeded on 24-well tissue culture plates and grown for 18 -20 h. THP-1 cells were treated with phorbol 12-myristate 13-acetate as described above. Cells were infected with M. smegmatis WT, pSMT3, MsmRv3242c, MSM_1877, M. marinum, M. marinum ⌬mimG (Rv3242c orthologue in M. marinum), and complemented (⌬mimG:Rv3242c) strains at an MOI of 10; intracellular survival was determined by lysing the infected macrophages at different time points; and bacterial survival was determined by plating the serially diluted samples as described above. M. marinum infection assays were carried out by incubating the infected macrophages at 30°C. The equal input and T0 counts of infecting bacilli were determined to calculate the percentage survival (% survival ϭ cfu at required time/cfu of bacteria added ϫ 100).

Quantification of Rate of Phagocytosis
To determine the rate of phagocytosis, M. smegmatis WT, pSMT3, and MsmRv3242c strains were labeled with SYTO-9 (200 nM) by incubating bacteria with a dye on a rocker in the dark for 30 min. Stained bacterial cultures were centrifuged at 3500 ϫ g for 5 min, the supernatant was discarded, and the pellet was suspended in 1ϫ PBS. Finally, the pellet was resuspended in infection medium (RPMI without antibiotics), and the absorbance was adjusted to 0.1. Then THP-1 cells were infected with labeled bacteria (MOI ϭ 10) for 2 h. Extracellular bacteria were removed by washing the monolayer cells with 1ϫ PBS. Cells were harvested at 2 and 24 h postinfection to quantify the internalized bacteria using a BD FACSCanto-II flow cytometer and FACS Diva software.

Zebrafish Infections
Wild type zebrafish (Danio rerio) were obtained from local vendors in India. The fish were kept in a 50-liter glass aquarium and kept in continuously well aerated water containing 2 mg/liter instant ocean salt at ϳ28°C under a 14-h/10-h light/dark photoperiod. Zebrafish were fed with commercial food twice a day with flakes (Basic Flake, Fujian, China) or commercial fish food (Tetra, Melle, Germany) in the morning and live Artemia nauplii (Inve Aquaculture Nutrition, Nonthaburi, Thailand) in the evening. The pH, dissolved oxygen content, and total hardness of the aquarium water were maintained as standard conditions for housing zebrafish (see the Zebrafish Model Organism Database Web site).
Infection protocols approved by the institute ethical committee were used for the zebrafish infection studies. Fish were housed separately in a 1-liter fish tank in autoclaved fish water. Water changes were performed daily following feeding. Adult healthy fish were used throughout this study. Infection experiments were performed as described previously (47). Fish were inoculated after anesthetizing first in an aqueous solution of tricaine (Sigma, catalog no. E10521-50g). 10 l of bacterial suspension in PBS were injected intraperitoneally and observed for signs of trauma following injection. Injected fish were housed separately (one fish per tank).

Bacterial Quantification
Recovery of three different bacterial strains from infected fish for colony count were performed by homogenizing the fish in 4 ml of sterile PBS and then plating the dilutions on 7H10 agar plates. The plates were incubated at 30°C for ϳ3-5 days until colonies were visible. Randomly selected colonies were stained with modified Ziehl-Neelsen stain to overrule any other bacterial contamination.

Zebrafish Pathology
Fish were weighed every day at the same time point to monitor the changes in body weight and also checked for gross signs of infection and death. To obtain the tissue samples, fish were euthanized and stored at Ϫ80°C in a freezer and then embedded in paraffin. The fish were sectioned longitudinally in 5-m sections, stained with modified Ziehl-Neelsen stain, and observed under light microscopy at ϫ40 and ϫ63 magnifications.

Estimation of ROS Production
Generation of superoxide anions (O 2 Ϫ ) and hydrogen peroxide (H 2 O 2 ) in RAW264.7 was determined by using dihydroethidium (DHE) and DCFH-DA, respectively. The superoxide indicator DHE, also known as hydroethidine, exhibits blue fluorescence after oxidation in the presence of superoxide anions (O 2 . ) and intercalates to DNA and gives red florescence. Similarly, DCFH-DA, a lipid-permeable non-fluorescent compound, when oxidized by intracellular H 2 O 2 in the presence of cellular esterase, forms the fluorescent compound 2Ј,7Ј-dichlorofluorescein. RAW264.7 cells (2 ϫ 10 5 ) were seeded on a 24-well tissue culture plate. The next day, the cells were infected with M. smegmatis WT, pSMT3, MsmRv3242c, MSM_1877, M. marinum, ⌬mimG, and ⌬mimG:Rv3242c strains at MOI 10 for 24 h as described above. Uninfected cells were used as control. For superoxide anion estimation, cells were harvested and incubated with 5 g/ml of DHE in 1ϫ PBS for 10 min. For H 2 O 2 estimation, 10 M DCFH-DA was added to the cells and incubated at 37°C and 5% CO 2 for 4 h. Then the DCFH-DA-containing medium was removed, the cells were rinsed with 1ϫ PBS, and the fluorescence intensity of both DHE and DCFH was read in a flow cytometer using a BD FACSCanto-II flow cytometer and FACS Diva software using a specific filter. Similarly, ROS production in zebrafish was measured as described previously with slight modifications (48, 49). Briefly, zebrafish were injected with M. marinum, ⌬mimG, and ⌬mimG:Rv3242c and PBS, and the level of ROS was measured using DCFH-DA. Five fish from each group were washed with cold PBS buffer (pH 7.4) and homogenized in 2 ml of cold PBS. The homogenate was centrifuged at 15,000 rpm at 4°C for 15 min, and the supernatant was transferred to a fresh tube for further analysis. 10 l of the supernatant was added to a transparent 96-well plate (Nunc), and 190 l of PBS and 3 l of a stock solution of 10 mM DCFH-DA (dissolved in DMSO) were added to each well. The plate was incubated at 37°C for 30 min in the dark. The fluorescence intensity was measured using an Infinite M200 Pro microplate reader (Tecan, Männedorf, Switzerland) with excitation and emission at 495 and 530 nm, respectively. The ROS values were normalized by measuring the protein content of the supernatants.

Superoxide Dismutase Assay
Superoxide dismutase (SOD) converts superoxide anions (O 2 . ) to hydrogen peroxide (H 2 O 2 ). SOD was detected by a nitrite method that involves generation of superoxide radical by photoreduction of riboflavin and its detection by nitrite formation from hydroxylamine hydrochloride at 543 nm using Griess reagent. Briefly, a mixture of 80 mM Tris buffer (pH 8.0), 1 mM L-methionine, 1% Triton-X-100, 0.5 mM hydroxylamine hydrochloride, and 7 M EDTA was prepared. Proteins were quantified by the Bradford method (50). 80 g of protein extracts prepared from uninfected and pSMT3-, MsmRv3242c-, M. marinum-, ⌬mimG-, and ⌬mimG:Rv3242c-infected macrophages were mixed with mixture and 4 M riboflavin. Control sample contained mixture and 4 M riboflavin only. Then the samples were exposed to photoreduction using a 220-watt bulb for 10 min. The superoxide anions produced from riboflavin will react with NO and form peroxynitrite, which can be quantified by adding an equal volume of Griess reagent (1% sulfanilamide and 0.1% naphthylethylenediamine dihydrochloride in 5% concentrated H 3 PO 4 ), and the color formed was measured at 543 nm. The intensity of color is inversely proportional to the level of SOD.

Catalase Assay
RAW 264.7 (1 ϫ 10 6 ) cells were infected with pSMT3 and MsmRv3242c strains for 24 h. Then cells were harvested and lysed by using lysis buffer (5 mM EDTA, 5 mM EGTA, 1 mM PMSF, protease inhibitor mixture, 10 mM sodium fluoride, 1 mM DTT, and 1 mM sodium orthovanadate). Hydrogen peroxide was diluted in 0.05 M phosphate buffer (pH 7.0) to a final concentration of 5 mM and mixed with 50 g/ml protein extracts. 100 l of freshly prepared substrate solution was added to each well in a microtiter plate. The absorbance was measured immediately at 240 nm every 10 s for 5 min at 22°C. Catalase activity was calculated based on the rate of decomposition of hydrogen peroxide, which is proportional to the reduction of the absorbance at 240 nm.

Real-time PCR Analysis
To determine the expression of Rv3242c as a function of growth, pSMT3 and MsmRv3242c strains were grown in 7H9 medium, and the bacteria were harvested at different time points (4, 12, and 24 h). Total RNA was isolated using TRIzol reagent as per the manufacturer's protocol. cDNA synthesis was performed using a cDNA synthesis kit. The synthesized cDNA was used as a template for RT-PCR amplification using gene-specific primers. All reactions were performed in a total reaction volume of 10 l using SYBR Green PCR Mastermix (Applied Biosystems) and carried out in a Real Plex master cycler (Eppendorf, Germany) with initial denaturation at 95°C for 10 min, final denaturation at 95°C for 30 s, annealing at 52°C for 30 s, and extension at 72°C for 30 s to generate 200-bp amplicons. Similarly, the expression of cytokines (TNF-␣, IL-10, and IL-12), autophagy markers (Beclin-1 and Atg-5), NAMPT-1, and inducible nitric-oxide synthase was quantified by isolating total RNA from M. smegmatis pSMT3-and MsmRv3242c-infected macrophages using gene-specific primers (Table 2) as described above. The mRNA levels were normalized to the transcript levels of sigA and GAPDH, respectively, and the relative -fold changes were calculated. Total RNA from whole fish was prepared as described above. The expression of TNF-␣, IL-1␤, IL-10, and IFN-␥ was determined using real-time PCR (ABi Step One qPCR). Each primer pair (Table 2) was designed using NCBI/Primer-BLAST. The expression of each gene was normalized to that of the ␤-actin transcript and expressed as -fold change relative to the expression seen in PBS-injected zebrafish. PCR conditions were as follows: 95°C for 5 min (once), 95°C for 20 s, 55°C for 30 s, 72°C for 30 s (40 cycles) using the DyNAmo ColorFlash SYBR Green qPCR kit (Thermo Scientific). All quantitative PCRs were performed for three biological replicates, and the data for each sample were expressed relative to the expression level of the ␤-actin gene by using the 2 Ϫ⌬⌬CT method.

LC3-II Staining for Autophagosomes
To analyze the autophagy, RAW264.7 cells were infected with M. smegmatis WT, pSMT3, and MsmRv3242c in the presence and absence of p38 blocker SB203580 and pERK blocker U0126. After 24 h of infection, cells were washed with lukewarm 1ϫ PBS, fixed with 4% paraformaldehyde, and then blocked with 5% BSA, 0.2% saponin, and 100 mM glycine in TBST for 1 h at room temperature. Then 1:1000 diluted LC3-II antibody (catalog no. 2775S, Cell Signaling) was added and incubated for 2 h at room temperature followed by a 1-h treatment with secondary Alexa Fluor goat anti-rabbit 594 antibody (catalog no. A-11037, Life Technologies, Inc.).

M. tuberculosis Cell Wall Contains Mannosylated
Glycoproteins-We limited our investigation to mannosylated proteins from the M. tuberculosis cell walls using lectin ConA, which specifically binds to mannose residues. Purified M. tuberculosis cell walls were delipidated by chloroform/methanol extraction. The proteins present in the delipidated cell wall fraction were then separated by SDS-PAGE. As shown in Fig. 1A (lane 1), blots based on affinity to ConA showed proteins with masses ranging from 20 to 80 kDa. Hybridization with biotinylated ConA detected three protein bands of ϳ23, 27, and 37 kDa (Fig.   1A, lane 2). Further enrichment by chromatography on ConAagarose columns yielded a single protein band with an apparent mass of 23 kDa (Fig. 1A, lane 3).
To confirm the specificity of recognition by ConA, the protein eluted from ConA-agarose was once again run on SDS-PAGE and probed with HRP-ConA. Consistent with the above data, a single protein band of 23 kDa hybridized with the HRPconjugated lectin (Fig. 1A, lane 4), indicating that the enriched protein may be mannosylated.
The Pro-Q Emerald stain is widely used for staining glycoproteins in gels or on blots. It reacts with periodate-oxidized carbohydrate groups, yielding a bright green fluorescent signal with glycoproteins. As shown in Fig. 1A (lane 5), a single protein band of 23 kDa reacted with ProQ Emerald, again indicating that the captured protein is a glycoprotein.
Protein Identification by Mass Spectrometry-Further identification of the 23-kDa protein was performed by both MALDI-TOF/TOF and LC-MS/MS (ESI-MS) mass spectrometry. The ConA-enriched 23 kDa band was excised from the gel and subjected to in-gel trypsin digestion. The resulting proteolytic peptides were recovered from the gel, fractionated by HPLC, and analyzed by ESI-MS. In addition, the peptides were also subjected to MALDI-TOF/TOF using ␣-cyanocinnamic acid as the matrix. Both techniques identified the 23-kDa protein as the Rv3242c gene product (Fig. 1B), previously annotated as M. tuberculosis hypothetical protein. The experimental peak masses as well as the corresponding theoretical masses and sequence matches are summarized in Fig. 1, C and D.
To further confirm that Rv3242c is indeed glycosylated, we determined glycosylation signatures in Rv3242c using different Web-based tools, such as EnsembleGly, Net-N-Glyc, Net-O-Glyc, and prokaryotic glycoprotein prediction programs. All of them predicted a glycosylation signature motif between residues 161 and 163. The Glycomod Web server and SimGlycan predicted the presence of a terminal mannose-containing glycopeptide (data not shown).
Rv3242c Encodes a Phosphoribosyltransferase-The initial annotation from the TubercuList Web server showed that Rv3242c contains a PRT domain. Using sequence-based homology searches, motif analysis, and PHYRE2 fold recognition, we confirmed that the Rv3242c indeed contains a PRT domain ( Table 1). The conserved sequence signature " 177 VVLVDDIIT-TGAT 189 " (Fig. 1E) is only found in PRTs involved in nucleotide biosynthesis and salvage pathways (51, 52), whereas PRTs involved in tryptophan, histidine, and nicotinamide biosynthesis usually lack this motif (51, 52). The phylogenetic analysis also placed Rv3242c near nucleotide salvage pathway biosynthetic PRTs (Fig. 1F). A homology search showed the presence of orthologues in M. smegmatis (MSMEG_1877) and M. marinum (mimG) genomes.
The PRT domain of Rv3242c was modeled using E. coli orotate PRT (Protein Data Bank entry 1ORO) as a template. The alignment showed 34% sequence identity and 44% query coverage (Fig. 1G). In silico glycosylation analysis predicted the presence of a glycosylation site exposed on the surface of the protein (Fig. 1H). As shown in Fig. 1I, 457 residual atoms were aligned between the predicted PRT (Rv3242c) model and the template E. coli orotate PRT.

An M. smegmatis Strain Expressing Rv3242c Exhibits Higher
Phosphoribosyltransferase Activity-To investigate the role of Rv3242c in mycobacterium pathophysiology and to further confirm that Rv3242c encodes for a PRT, we expressed M. tuberculosis Rv3242c in a non-pathogenic M. smegmatis strain using the pSMT3 shuttle vector (53). Several previous studies have already used M. smegmatis as a surrogate model to study the functions of proteins from pathogenic mycobacteria (54). The schematic genetic representation of Rv3242c in the M. tuberculosis H37Rv genome is given in Fig.  2A.
To validate in silico data, we measured PRT activities at 0, 4, 12, and 24 h in M. smegmatis WT, vector control (pSMT3), recombinant MsmRv3242c, and strain MSM_1877 expressing M. smegmatis MSMEG_1877, an Rv3242c orthologue in M. smegmatis. As shown in Fig. 2B, significantly more PRT activity was detected in recombinant MsmRv3242c strain as compared with M. smegmatis WT, M. smegmatis pSMT3, and MSMEG_1877 strains. No differences in the growth kinetics of these strains were observed by either OD measurement (Fig.  2C) or cfu counting (data not shown), suggesting that overexpression of Rv3242c does not lead to growth defects. In conclusion, the above data again confirm that Rv3242c indeed encodes an active phosphoribosyltransferase. In all subsequent experiments, the M. smegmatis strain expressing the empty vector pSMT3 was used as a baseline control.

M. smegmatis Strain Expressing Rv3242c Inhibits NAMPT-1 (Visfatin) Level in Infected
Macrophages-Nicotinamide phosphoribosyltransferase (NAMPT), also known as visfatin, is responsible for NAD ϩ biosynthesis in mammals and is also known to play a major role in regulation of autophagy (55) and ROS production (56). Therefore, we measured the nampt-1 expression after 24 h of infection with M. smegmatis pSMT3 and MsmRv3242c in RAW264.7 cells. We observed an ϳ4-fold decrease in nampt-1 expression in MsmRv3242c-infected cells when compared with pSMT3-infected cells (Fig. 2D).  (Fig. 2F). The adherence rate was quantified by counting the number of bacteria adhered over 100 cells. MsmRv3242c showed significantly more adherence when compared with M. smegmatis WT and pSMT3 (Fig. 2G). These results indicate that Rv3242c is involved in host cell adhesion and invasion. The equal input as well as time 0 counts of infecting bacilli indicated that the observed difference is not due to the different initial bacterial count used for infection.

Expression of Rv3242c by M. smegmatis Enhanced Adhesion and Invasion into Human Epithelial
Then we measured Rv3242c expression after isolating mRNA from recombinant MsmRv3242c and control strains grown under in vitro conditions (extracellular) and inside macrophages (intracellular) using gene-specific primers (Table  2). Quantitative real-time PCR analysis of the recombinant MsmRv3242c strain showed higher expression of Rv3242c under both extracellular (Fig. 2H) and intracellular (Fig. 2I) conditions as compared with control strains. Interestingly, the expression of Rv3242c was significantly higher under intracellular conditions 24 h after infection, whereas, as expected, no amplification products were observed in the negative controls, suggesting that the transcript levels estimated were specific to M. tuberculosis Rv3242c. Similarly, the expression of MSM_ 1877 was analyzed in the M. smegmatis pSMT3 strain, which The PRT domain from Rv3242c (blue) was modeled using E. coli orotate PRT (green) as template and corresponding superimposition with the same structure. H, the ab initio protein structure and modeling were done using the I-TASSER Web server; the structure is stabilized by a 1-ns simulation using the Gromacs molecular dynamics simulation package, which shows that the glycosylation site Asn 161 -Asn 163 (blue) was exposed to the surface. I, the modeled Rv3242c (blue) PRT domain was superimposed with the template orotate PRT (red). Both the model and template show structural characteristics similar to those of the PRT domain.

TABLE 1 Functional genomic prediction(s) result(s) for Rv3242c
Multiple-sequence-based homology search and motif analysis showed that Rv3242c is a PRT protein.  2J) and intracellular conditions (Fig. 2K). Quantitative RT-PCR data showed no significant differences in MSM_1877 expression w.r.t. Ϯ Rv3242c. Gene-specific transcript levels were normalized to M. smegmatis sigA.

PSI-BLAST
Rv3242c Is Involved in Prolonged Bacterial Survival in Mouse Peritoneal, RAW 264.7, and Human THP-1 Macrophages-The above results suggested that Rv3242c may be involved in M. tuberculosis virulence. We hypothesized that if this is indeed the case, the recombinant MsmRv3242c strain should survive longer in macrophages. We therefore compared the intracellular persistence of M. smegmatis WT, M. smegmatis pSMT3, and recombinant MsmRv3242c strains in phorbol 12-myristate 13-acetate-differentiated human THP-1, murine RAW264.7, and mouse peritoneal macrophages. The recombinant MsmRv3242c strain showed a significantly higher bacillary burden in RAW264.7 (Fig. 3A) 1, 8, and 24 h postinfection, in comparison with M. smegmatis WT and pSMT3 control strains. The same phenotype was observed in THP-1 macrophages as well; however, overall cfu counts were higher in THP-1 macrophages as compared with RAW264.7 cells 24 h postinfection (Fig.  3B). We also compared intracellular survival of these three strains in mouse peritoneal macrophages. Consistently, the recombinant Rv3242c strain showed increased intracellular persistence when compared with strains M. smegmatis WT and pSMT3 (Fig. 3C). MSMEG_1877 showed no significant difference in intracellular persistence as compared with the pSMT3 control strain during 8 h of infection, whereas a moderate increase in bacterial count was observed 24 h postinfection (Fig. 3D).
To further corroborate the above results, the M. smegmatis WT, M. smegmatis pSMT3, and recombinant MsmRv3242c strains were fluorescence-stained with SYTO-9 and presented to THP-1 macrophages simultaneously. SYTO-9 is a cell-permeant nucleic acid stain that shows green fluorescence upon binding to nucleic acids. Twenty-four hours after infection, THP-1 cells were washed thoroughly to remove extracellular bacteria, and FACS was then used to detect infected macrophages. The analysis revealed that the macrophages infected with recombinant strain MsmRv3242c had a higher intracellular count (27.3%) (marked with an arrow), as compared with M. smegmatis WT (17%) and pSMT3 control (18%) strains, indicating that Rv3242c indeed enhances intracellular persistence of M. smegmatis (Fig. 3E).
We investigated whether the differential intracellular bacterial survival is due to differences in macrophage cell viability in response to infection. A cytotoxicity assay showed no differences in macrophage viability following infection with M. smegmatis WT, pSMT3, and recombinant MsmRv3242c strains, as compared with uninfected (control) cells (Fig. 3F). mimG Is Required for Virulence of M. marinum in Zebrafish-To assess whether mimG is also required for virulence of M. marinum in vivo, we infected adult zebrafish (n ϭ 15) and followed their survival up to 30 days postinfection. Zebrafish infected with wild type M. marinum showed significant death between 15 and 30 days postinfection, as did fish infected with ⌬mimG:Rv3242c complemented strain. However, significantly more fish survival was observed until 30 days postinfection in ⌬mimG-infected fish (Fig. 4A). At the time of death, the abdomens of the zebrafish infected with wild type   4C) and complemented ⌬mimG:Rv3242c M. marinum (Fig. 4E) strains were visibly hemorrhaging, indicating an inflammatory response. However, significantly less hemorrhaging was detected in zebrafish infected with ⌬mimG bacilli (Fig. 4D), whereas no such phenotypes were observed in PBStreated zebrafish (Fig. 4B). M. marinum infection causes severe emaciation, shredding of scales, and skin lesions in zebrafish (47, 59). Similar symptoms were observed upon wild type M. marinum and ⌬mimG:Rv3242c infection (Fig. 4F). However, the knock-out strain ⌬mimG caused a less severe infection in the zebrafish. Moreover, the ⌬mimG-infected fish showed significantly reduced body weight loss as compared with the wild type and the complemented M. marinum strains (Fig. 4G). Histological analysis of the whole fish was performed 30 days postinfection. Whereas the zebrafish infected with the wild type M. marinum showed a similar bacterial deposition pattern throughout the body as with the ⌬mimGand ⌬mimG: Rv3242c-infected fish, significantly less inflammation and pathological characteristics were observed in ⌬mimG-infected fish when compared with wild type M. marinum and ⌬mimG: Rv3242c infected fish (Fig. 4H). Thus, mimG in M. marinum is required for virulence in zebrafish. mimG Mutant Showed Reduced Survival in Infected Zebrafish-To determine whether the ⌬mimG mutant failed to kill the fish because it is more easily cleared by the host, cfu were enumerated from the whole fish at different time points postinfection. For this, fish were injected intraperitoneally with ϳ1.3 ϫ 10 4 bacilli. Five fish were randomly chosen from each batch to determine the bacterial cfu count after 10, 20, and 25 days. Ten days postinfection, the bacterial count was found to be almost the same as at day 0. It exponentially increased 10 days postinfection in all of the fish, regardless of the M. mari- num genotype. Differences in growth patterns among the genotypes were observed from the 20th day such that a significant decrease in the cfu of ⌬mimG mutant was observed after 25 days as compared with wild type M. marinum and ⌬mimG: Rv3242c complemented strains, suggesting that the ⌬mimG strain is more easily cleared by the host (Fig. 4I).

Deletion of mimG, an Rv3242c Orthologue Present in M. marinum, Decreased Intracellular Bacterial
The Increase in Intracellular Survival by Rv3242c and mimG Is Due to the Inhibition of ROS Production, Modulation of Cytokine Expression, and Impaired Superoxide Dismutase and Catalase Activities-To understand the mechanisms responsible for enhanced intracellular survival of recombinant MsmRv3242c strain, we measured the levels of SOD, catalase, and oxygen radicals, such as ROS and NO, which are key determinants of intracellular mycobacterial burden in infected macrophages (60). It has been proven that pathogenic mycobacteria tend to inhibit the production of ROS and NO to avoid killing by the host cells (61,62).
Next we measured the level of superoxide anions (O 2 . ) in infected macrophages by DHE staining. As shown in Fig. 5A, no significant difference in the level of superoxide anions was observed in M. smegmatis pSMT3-infected (11.6%) and recombinant MsmRv3242c-infected (11.3%) macrophages 24 h postinfection. Then we measured the generation of ROS using DCFH staining. A significant decrease in ROS production was observed in MsmRv3242c-infected (28.6%) macrophages as compared with M. smegmatis pSMT3-infected (36.9%) and MSM_1877-infected (33.7%) cells (Fig. 5B). Consistent with the in vivo data, wild type M. marinum (19.4%) and complemented ⌬mimG:Rv3242c (14.1%) also inhibited ROS production when compared with ⌬mimG (23.2%) mutant after 24 h of infection (Fig. 5C, top). This effect was more pronounced at 48 h after infection, in which wild type M. marinum (15.5%) and complemented ⌬mimG:Rv3242c (11.6%) down-regulated more ROS production as compared with ⌬mimG (23.5%) mutant (Fig. 5C, bottom). The above data clearly suggested that mycobacteria PRT is responsible for decreased ROS production, which leads to increased intracellular bacillary persistence in macrophages.
Superoxide anions (O 2 . ) are converted to hydrogen peroxide (H 2 O 2 ) in the presence of SOD (64). Next we measured the SOD activity in M. smegmatis pSMT3 and recombinant MsmRv3242c-infected macrophages as described under "Experimental Procedures." In this assay, photoreduction of riboflavin is carried out to generate superoxide anion, which in turn react with nitrogen intermediate NO and form peroxynitrite (ONOO Ϫ ) (65). The amount of ONOO Ϫ formed is quantified by the addition of Griess reagent. The absorbance read is inversely proportional to the level of SOD (66). As shown in Fig.  5D, MsmRv3242c infection resulted in a higher ONOO Ϫ formation, which indicates a lower level of SOD as compared with pSMT3-infected cells. Similar results were obtained in M. marinum-infected macrophages, in which more ONOO Ϫ production was observed in wild type M. marinum-and ⌬mimG:Rv3242c-infected macrophages as compared with ⌬mimG-infected cells (Fig. 5E). The H 2 O 2 formed in the cells is further converted to H 2 O and O 2 in the presence of catalase (64). The catalase activity was determined by preparing the protein extracts from macrophages infected with M. smegmatis pSMT3 and recombinant MsmRv3242c, and the rate of decomposition of H 2 O 2 , which directly reflects catalase activity, was determined. Significantly reduced catalase activity was observed in recombinant MsmRv3242c-infected macrophages as compared with M. smegmatis pSMT3-infected cells (Fig. 5F). We also measured the levels of inducible nitric-oxide synthase, which catalyzes the production of NO. No changes in the nitric-oxide synthase transcripts (iNOS; Fig. 5G) and NO production (data not shown) were observed in infected macrophages. Taken together, our data suggest that Rv3242c and mimG aid the intracellular persistence of bacilli by inhibiting the ROS production due to decreased SOD and catalase activities.
Rv3242c and mimG Inhibit Autophagy in Mouse Macrophages-It has been reported that autophagy can kill intracellular pathogens, including M. tuberculosis (67). However, M. tuberculosis is known to escape the killing within macrophages by inhibiting autophagy (68). One of the hallmarks of autophagy is the induction of microtubule-associated protein 1 light chain 3 (LC3) (69). The cytosolic 18-kDa LC3 (also termed LC3-I) form is converted to the autophagosome-associated 16-kDa LC3-II form. The conversion of LC3-I to LC3-II is considered as a reliable marker of autophagy (69). We examined the expression of autophagy markers LC3-II, Beclin-1, and Atg5 in macrophages infected with M. smegmatis WT, pSMT3, recombinant MsmRv3242c, wild type M. marinum, ⌬mimG mutant, and ⌬mimG:Rv3242c strains 24 h postinfection. Western blot analysis indicated a significant decrease in the level of LC3-II in recombinant MsmRv3242c-infected macrophages as compared with other control strains (Fig. 6A, top). We also determined the expression of p62, which is accumulated upon inhibition of autophagy (69). We found significantly higher expression of p62 in MsmRv3242c-infected cells (Fig. 6A, bottom). The expression of Atg5 was also found to be decreased at both translational (Fig. 6B) and transcriptional levels (Fig. 6D). Surprisingly, no difference in the level of beclin-1 was found in recombinant MsmRv3242c-infected macrophages (Fig. 6C). Similarly, M. marinum wild type and ⌬mimG:Rv3242c infection also decreased LC3-II expression in macrophages as compared with ⌬mimG-infected cells (Fig. 6E). These observations clearly suggest that PRT is mainly responsible for down-regulation of autophagy.
During autophagy, acidic autophagic vacuoles, also called autophagolysosomes, are formed as a result of fusion of autophagosomes with lysosomes and are considered as a characteristic feature of cells engaged in autophagy (70). The MAPK/ERK pathway has been shown to inhibit autophagy in response to a number of signals (71,72). We observed activation of MAPK/ERK pathway (shown below) in response to recombinant MsmRv3242c infection in macrophages. Therefore, we checked whether inhibition of p38 and pERK could activate autophagy in MsmRv3242c-infected macrophages. Indeed, the addition of p38 (Fig. 6F) and pERK (Fig. 6G) blockers increased the expression of LC3-II in MsmRv3242c-infected macrophages when compared with MsmRv3242c-infected macrophages without p38 and pERK blockers (Fig. 6A).
Formation of autophagolysosomes was also studied by fluorescence microscopy following LC3-II staining. M. smegmatis WT-and pSMT3-infected macrophages showed pronounced red fluorescence, an indicator of autophagy expression, as compared with recombinant MsmRv3242c-infected macrophages (Fig. 6H), suggesting less vacuole acidification and autophagolysosomal formation during recombinant MsmRv3242c infection. Consistent with the above data, the addition of p38 (Fig.  6I) and pERK (Fig. 6J) blockers increased LC3-II expression in recombinant MsmRv3242c-infected macrophages, indicating p38-and pERK-dependent inhibition of autophagy by MsmRv3242c. The number of LC3 puncta per cell was also quantified. For this, 100 cells were analyzed per assay, which confirmed more LC3 puncta in pSMT3 when compared with MsmRv3242c-infected macrophages. However, after the addition of p38 and pERK blockers, no significant differences in the count of LC3 puncta were observed (Fig. 6K).
To study whether stimulation or inhibition of autophagy can manipulate intracellular bacterial burden, we induced autophagy by exposing the cells to serum starvation for 36 h. First, the starvation was confirmed by the up-regulation and down-regulation of LC3-II and p62 expression, respectively, in RAW264.7 cells (Fig. 6L). In another set of conditions, autophagy was inhibited by treating the cells with the well known autophagy inhibitor 3-methyladenine. Then we compared the survival of bacilli in artificially autophagy-induced and -inhibited cells. MsmRv3242c showed prolonged survival in serum-starved as well as serum-starved plus 3-methyladenine-treated cells. On the other hand, 3-methyladenine treatment alone showed a similar intracellular bacillary burden of M. smegmatis WT, pSMT3 vector control, and recombinant MsmRv3242c in macrophages (Fig. 6M). Moreover, in Atg5silenced (siAtg5) macrophages (Fig. 6N), a similar bacillary burden of MsmRv3242c and pSMT3 strains was observed, whereas in scrambled siRNA-transfected cells, greater survival of MsmRv3242c was observed with respect to pSMT3 (Fig. 6O).
Recombinant Rv3242c Infection Causes NF-B Activation and Altered Cytokine Production in Macrophages-Besides phagocytosis, the interaction between mycobacteria and macrophages induces several signaling pathways that converge at the level of p-NF-B activation. Therefore, we investigated the effect of infection by recombinant MsmRv3242c on activation of p-NF-B in macrophages. Greater expression of p-NF-B was observed in recombinant MsmRv3242c-infected macrophages as compared with M. smegmatis WT-and pSMT3infected cells (Fig. 7A), whereas no difference in the expression of total NF-B was observed in cells infected with these three strains (Fig. 7A).
To assess the immunomodulatory activity of Rv3242c, we determined the levels of several cytokines known to regulate the intracellular fate of M. tuberculosis (6). We observed an increase in transcript levels of proinflammatory TNF-␣ (Fig.  7B) and IL-12 (Fig. 7C) and the anti-inflammatory IL-10 cytokines in recombinant MsmRv3242c-infected macrophages. IL-10 is the key cytokine involved in the p38 pathway (73) and is also known to inhibit autophagy in infected macrophages (74). Therefore, we checked the expression of IL-10 in p38-and pERK-blocked macrophages. In MsmRv3242c-infected macrophages, significantly less expression of IL-10 was observed in both p38-and pERK-blocked macrophages with respect to M. smegmatis pSMT3-infected cells (Fig. 7D). These results suggest that higher expression of IL-10 was due to activation of p38 and pERK.
Rv3242c Activates the MAPK Pathway in Macrophages-The MAPK/ERK signaling pathway involves a series of proteins that communicates a signal from the receptor to the DNA in the nucleus. In the case of M. tuberculosis infection, MAPK plays a decisive role in the production of TNF-␣ and IL-10 (75). Therefore, we investigated the role of Rv3242c in stimulating the phosphorylation of ERK (pERK), JNK (p-JNK), and p38 (p-p38), the primary effectors in this signaling pathway. Infection of macro-phages with recombinant MsmRv3242c strain led to induction of the ERK, JNK, and p38 pathways, as indicated by increased phosphorylation of proteins 24 h postinfection (Fig. 7E). These results correlate well with the increase in the expression of TNF-␣ and IL-10, which are known to be activated due to the activation of the MAPK pathway. Fig. 7F shows a schematic representation of regulation of different pathways by Rv3242c, which leads to increased bacillary survival in macrophages.

Discussion
M. tuberculosis glycoproteins are attributed roles in virulence as well as in antigenic processes due to their ability to manipulate host antibacterial effector mechanisms (76,77). For example, M. tuberculosis Rv1174c acts as an immune-dominant T-cell antigen (78); Rv1860 mediates interaction with host cells through PSP-A (79); Rv3763 functions as a major adhesion protein by interacting with the host mannose receptors (80); mannose lipoarabinomannan is involved in phagosome maturation (81,82); and several other glycoproteins, such as Rv0129c, Rv3804c, Rv1886C, and Rv3803C, which are part of the antigen 85 complex, are implicated in various pathogenic processes (83)(84)(85). However, considering the complexity of the M. tuberculosis genome, the function of many glycoproteins is still undefined. In this study, we identified and proved the function of a novel lectin-captured protein and studied its role in mycobacteria virulence by expression and deletion in M. smegmatis and M. marinum, respectively. Numerous studies have used M. smegmatis as a model host to study the function of M. tuberculosis proteins (54) and to successfully identify genetic loci implicated in intracellular survival (86). Similarly, M. marinum is known as the genetically closest neighbor and shares many virulence factors and mechanisms with M. tuberculosis (87).
Previously, several M. tuberculosis mannosylated proteins have been identified by ConA lectin affinity chromatography (88,89). These studies also indicated that protein mannosylation is one of the major forms of glycosylation in M. tuberculosis. In the present study, we used several mannose-specific lectins to capture the glycoproteins from the M. tuberculosis cell wall. The initial Western blot with ConA-HRP showed the presence of three proteins, which could be mannosylated. Further enrichment steps of the captured sample in the ConA column showed the presence of only one protein band of 23 kDa, which supports the idea that sugar-specific lectins can be used to catch glycoproteins from a mixture of proteins. Two independent mass spectrometry analyses identified it as a novel protein, Rv3242c, which was previously annotated as a "hypothetical protein" in the tuberculosis database. The presence of carbohydrates on Rv3242c was confirmed by ProQ Emerald glycoprotein staining, which is a sensitive method for detecting the aldehydes produced after periodate oxidation of carbohydrates. Further, our phylogenetic analysis, multiple-sequence alignment, and molecular simulation studies as well as the presence of conserved PRT sequence (VVLVDDIIT-TGAT) suggested that Rv3242c could be a putative PRT. Indeed, we observed significantly higher PRT activity in strain MsmRv3242c, which further confirmed that Rv3242c encodes for a PRT enzyme in M. tuberculosis. An intermediate PRT activity was detected in MSMEG_1877 strain, suggesting that M. tuberculosis PRT is catalytically more active than M. smegmatis PRT. Although a previous genome annotation of M. tuberculosis reflected the presence of 19 -22 putative PRTs, this is the first report that has led to identification of a mannosylated PRT in M. tuberculosis. However, further comparative mass spectrometry analysis is required to analyze sugar content and the sites of glycosylation in M. tuberculosis and M. smegmatis PRTs.
A common metabolic pathway crucial for the survival and replication of the pathogens is the synthesis of purine and pyrimidine nucleotides essential for the production of DNA/ RNA molecules. Key enzymes in this pathway are several phosphoribosyltransferases. Previously, de novo pyrimidine biosynthesis has been shown to be required for cell invasion and virulence of parasites (90,91), in which disruption of this pathway led to an avirulent phenotype. Although the metabolism required for sustaining mycobacterial survival during latency is still poorly understood, experimental evidence has suggested the essentiality of nucleotide synthesis in latent bacilli (92). Therefore, selective inactivation of these enzymes could prove to be a crucial drug target for the control of mycobacterial infection.
In this study, we examined the role of identified PRTs from two pathogenic strains, M. tuberculosis H37Rv (Rv3242c) and M. marinum (mimG) in the modulation of macrophage, a primary host cell for M. tuberculosis (43), immune responses. M. smegmatis is unable to invade epithelial cells; however, previous studies have shown that expression of different invasive M. avium genes in M. smegmatis helps M. smegmatis to invade HT-29 cells (93). In this study, using non-phagocytic human colon epithelial HCT-116 cells, we also report that expression of Rv3242c in M. smegmatis aids host cell invasion and adhesion. Previously, M. tuberculosis cell wall-associated protein adhesin was shown to facilitate the entry of pathogen inside the host (57). Because Rv3242c is a cell wall-associated protein, it is presumed that Rv3242c could be involved in mediating mycobacterium-host interactions leading to host cell invasion. The intracellular survival kinetics showed that recombinant strain MsmRv3242c is responsible for enhanced survival in mouse peritoneal, RAW264.7, and THP-1 macrophages, indicating a potential role of PRT in intramacrophage bacterial persistence. It is interesting to note that the expression of orthologue MSM_1877 also moderately increased, although less than Rv3242c, bacillary persistence in macrophages at 24 h postinfection. Transcription of MSMEG_1877 was detected in M. smegmatis WT, suggesting that it is not simply the presence of MSMEG_1877 transcript that is important for promoting intracellular survival, but perhaps its relative abundance is responsible for the observed phenomenon. As discussed above, the M. smegmatis MSMEG_1877 strain exhibited less PRT activity as compared with MsmRv3242c. This gives a possible hint about the proportional relationship between PRT activity and severity of virulence. It could also be attributed partly to differences in the glycosylation pattern of Rv3242c and MSMEG_1877 because, as reported earlier, the presence of mannose-capped lipoarabinomannans, in contrast to uncapped lipoarabinomannans in non-pathogenic mycobacteria, is an important structural feature of virulent mycobacteria to subvert the immunological defense of the host (24). The differences in virulence patterns cannot be attributed to growth rate because all strains showed a similar growth pattern under in vitro conditions. Indeed, deletion of the PRT-encoding gene (mimG) demonstrated that this gene is required for full expression of disease by M. marinum during infection of both zebrafish and macrophage cells, whereas complementation of this gene restored the loss of function. In a recent study, the deletion of the mptC gene in M. marinum altered the lipoarabinomannan mannosylation pattern, which strongly attenuated the virulence of M. marinum in the zebrafish model of infection (94). This study has found that deletion of mptC resulted in loss of the ␣(132)-linked mannose cap of lipoarabinomannans, and complementation of the M. marinum mutant with the M. tuberculosis mptC orthologue restored the loss of function. Similarly, in our study, the complementation of mimG with the Rv3242c restored the virulence defect of the mimG mutant, indicating that the two genes are functionally conserved. These results are also in agreement with previous observations that deletion of PRT leads to impaired survival in M. tuberculosis (26) and severely compromises the infection process of parasites (91). Although the strong binding affinity of Rv3242c with ConA suggested the presence of a mannose sugar in Rv3242c, it remains to be demonstrated experimentally whether Rv3242c and mimG indeed contain mannosylated residues. Thus, these results unambiguously indicate that the role of PRT as a virulence factor is conserved in pathogenic mycobacteria.
ROS plays a crucial role in the control of mycobacterial growth in macrophages (95,96). Pathogenic mycobacteria suppress free radical production to subvert macrophage control of intracellular bacilli load. Macrophages infected with recombinant MsmRv3242c strain showed significantly lower levels of SOD, catalase, and ROS production when compared with pSMT3-infected cells, whereas ⌬mimG mutant-infected macrophages produced high levels of ROS and SOD as compared with M. marinum wild type and ⌬mimG:Rv3242c-infected cells. In vivo ROS measurement also revealed increased ROS production in ⌬mimG mutant-infected zebrafish, whereas a significantly lower level of ROS production was observed in M. marinum-and ⌬mimG:Rv3242c-infected zebrafish. Under physiological conditions, SOD converts O 2 . into H 2 O 2 , which is further converted into H 2 O and O 2 by catalase. Hence, the decreased ROS production was due to down-regulation of SOD and catalase activities by Rv3242c and mimG. Moreover, we observed greater production of peroxynitrite in recombinant MsmRv3242c-, M. marinum wild type-, and ⌬mimG:Rv3242cinfected cells when compared with control and mimG deletion strains. In living organisms, O 2 . is scavenged by SOD. O 2 . interacts with NO to form peroxynitrite. Therefore, due to low SOD activity, more O 2 . is available for interaction with NO, which led to increased production of peroxynitrite, indicating that PRT inhibits oxidative stress responses, which promoted bacterial survival in macrophages. Another mechanism that regulates the mycobacterial growth is the induction of autophagy. M. tuberculosis has evolved mechanisms to avoid autophagymediated killing by inhibiting the processes of autophagosome induction and phagosomal acidification by disrupting the delivery of H ϩ -ATPase subunits. Infection with recombinant MsmRv3242c, M. marinum wild type, and ⌬mimG:Rv3242c down-regulated the expression of autophagic markers LC3-II, Atg5, and Beclin-1 and up-regulated the expression of p62, whereas deletion of mimG in M. marinum up-regulated LC3-II expression. Moreover, the presence of fewer LC3II puncta and more p62 in MsmRv3242c-infected cells clearly indicates that PRT inhibits the programmed autophagy mechanism. p38 and pERK are known to play a major role in control of both the autophagy and MAPK pathways (97). Therefore, in our study, we blocked both p38 using SB203580 and pERK using U0126 and analyzed the expression of LC3II in infected macrophages. Blocking of p38 and pERK increased the expression of LC3-II in MsmRv3242c-infected macrophages. These results suggest that inhibition of autophagy by Rv3242c was mediated through activation of p38 and pERK signaling molecules. These results, in addition to the previous observation (98), further confirm the role of M. tuberculosis glycosylated proteins in evasion of host immune responses to aid bacillary persistence in macrophages. Accumulating evidence suggests that autophagy is a coordinated biological response and that several signaling pathways can act in concert to bring about directed antibacterial autophagy. Our current study provides evidence that Rv3242c activates the MAPK pathway by inducing phosphorylation of MAPK intermediate molecules like pERK, p-P38, and p-JNK in infected macrophages. In fact, M. tuberculosis glycolipid lipoarabinomannan is known to activate MAPK pathway to attenuate host immunity by delaying phagosomal maturation and reducing CD1 expression and presentation of mycobacterial antigens to T cells (99 -101). MAPK activation is also associated with the increased production of proinflammatory cytokines TNF-␣ and IL-12 (102,103). Consistent with these observations, we also observed significant up-regulation of TNF-␣ and IL-12 in recombinant MsmRv3242c-infected macrophages. In vivo data also revealed increased expression of TNF-␣, IFN-␥, and IL-1␤ in M. marinum-and ⌬mimG:Rv3242c-infected zebrafish. However, the transcript of anti-inflammatory cytokine IL-10, which is a part of a Th2 response that favors survival of the mycobacteria, was also increased (60). This was consistently observed in in vivo experiments, in which increased expression of IL-10 was observed in M. marinum and ⌬mimG:Rv3242c-infected zebrafish as compared with ⌬mimG-infected fish. Previously, IL-10 has been shown to inhibit autophagy induction (74). Therefore, the observed autophagy inhibition by Rv3242c could also be attributed to increased IL-10 production. p38 and pERK activation also induce expression of IL-10. In agreement with these results, we also observed down-regulation of IL-10 in the presence of p38 and pERK inhibitors in MsmRv3242c-infected cells. These results clearly indicate that Rv3242c induces IL-10 via p38 and pERK, which in turn results in down-regulation of LC3-II.
Mycobacteria are known to release active membrane vesicles (104). These vesicles contain proteins, lipids, lipopolysaccharide, and periplasmic components, which modulate various host immune responses and signaling pathways (104,105). Previous studies have shown that Visfatin (NAMPT-1), which is a rate-limiting enzyme of the host, regulates ROS production (56) and autophagy expression (55). We also observed significant down-regulation of nampt-1 in Rv3242c-infected macrophages. We hypothesize that decreased ROS production and autophagy inhibition could be due to down-regulation of nampt-1 by Rv3242c. However, further study is required to investigate the mechanisms involved in regulation of NAMPT-1 by Rv3242c.
Similarly, NF-B activity is important to coordinate the immune response for resistance to a wide range of pathogens. NF-B inhibition induces autophagy by increasing the formation of autophagosomes (106), and MAPK triggers the expression of NF-B (107,108). This suggests that MsmRv3242c infection activates the MAPK pathway, which subsequently activates p-NF-B and IL-10, resulting in inhibition of autophagy. However, it remains to be examined whether PRT itself exerts these immune-evading activities or if it just acts as a superantigen eliciting responses to host cells. Based on our experimental findings, we propose a mechanism by which PRT facilitates intramacrophagic bacillary persistence by down-reg-ulating host NAMPT-1 expression, which subsequently downregulates ROS and autophagy via p38, pERK, and IL-10 in infected macrophages. Collectively, our findings identified a new PRT that could be used by pathogenic mycobacteria, such as M. tuberculosis and M. marinum, to alter host immune response, thus presenting attractive targets for new drug therapies.