Role of VapBC12 Toxin-Antitoxin Locus in Cholesterol-Induced Mycobacterial Persistence

The current TB treatment regimen involves a combination of drugs administered for an extended duration that could last for 6 months to 2 years. This could lead to noncompliance and the emergence of newer drug resistance strains.

IMPORTANCE The current TB treatment regimen involves a combination of drugs administered for an extended duration that could last for 6 months to 2 years. This could lead to noncompliance and the emergence of newer drug resistance strains. It is widely perceived that the major culprits are the so-called nonreplicating and metabolically inactive "persister" bacteria. The importance of cholesterol utilization during the persistence stage of M. tuberculosis infection and its potential role in the generation of persisters is very intriguing. We explored the mechanism involved in the cholesterol-mediated generation of persisters in mycobacteria. In this study, we have identified a toxin-antitoxin (TA) system essential for the generation of persisters during M. tuberculosis infection. This study verified that M. tuberculosis strain devoid of the VapBC12 TA system failed to persist and showed a hypervirulent phenotype in a guinea pig infection model. Our studies indicate that the M. tuberculosis VapBC12 TA system acts as a molecular switch regulating persister generation during infection. VapBC12 TA system as a drug target offers opportunities to develop shorter and more effective treatment regimens against tuberculosis. G lobally, a third of the human population is infected with Mycobacterium tuberculosis, the causative agent of tuberculosis. Being an obligate intracellular pathogen, M. tuberculosis has coevolved with humans for centuries (1)(2)(3). Interestingly, while in most bacterial diseases the standard therapeutic regimen involves a single antibiotic treatment lasting for a few days to weeks, tuberculosis treatment requires administration of multiple drugs for a duration that could extend from 6 months to 2 years depending on the antibiotic susceptibility profile of the infected strain. This phenotype is attributed to a slow-growing metabolically altered subset of the heterogeneous M. tuberculosis population called persisters (4,5). These persisters are refractory to antimycobacterial drugs and can only be targeted using a strict regimen consisting of a combination of drugs for an unusually extended period. As a result, a protracted regimen thus impels noncompliance and results in an increase in the frequency of multidrug-resistant and extensively drug-resistant tuberculosis cases (6)(7)(8)(9).
Although several studies have described the stress-induced generation of persisters under in vitro growth conditions (10)(11)(12)(13), the exact conditions triggering the generation and enrichment of persisters inside the host during a normal course of mycobacterial infection remain unclear. Upon infection, M. tuberculosis induces the formation of lipid-rich foamy macrophages. Lysis of these macrophages results in the formation of the caseous core of a typical "tuberculous granuloma," providing M. tuberculosis with a lipid-rich niche. While residing in this nutrient-deprived granuloma, M. tuberculosis adapts itself to utilize host-derived lipids, including cholesterol as a favored carbon source (14). Surprisingly, M. tuberculosis does not rely on cholesterol as the sole carbon source during infection (15,16); nonetheless, this cholesterol utilization causes inhibition of growth and activation of pathways, leading to the generation of persisters in the mycobacterial population (14,17). M. tuberculosis facilitates the intracellular accumulation of cholesterol by upregulating cholesterol biosynthesis pathways and helps to convert resident macrophages into foamy macrophages (18). These findings imply that by hijacking host pathways M. tuberculosis creates a favorable niche inside the host, facilitating long-term disease persistence (19,20).
Toxin-antitoxin (TA) proteins play a crucial role in generating persisters in several bacterial species (21)(22)(23). These TA systems, known to modulate growth under various growth and stress conditions, are found in a wide range of bacterial and archaeal chromosomes and plasmids (24)(25)(26). Research conducted during the past decade has demonstrated that TA loci act as effectors of dormancy and persistence in several bacterial species (21,22). Each TA locus consists of genes expressing a TA pair. Antitoxin, being more labile, degrades under specific growth and stress conditions, resulting in the activation of the cognate toxin. The activated toxin modulates growth by targeting growth-related genes.
The genome of M. tuberculosis constitutes 88 TA systems, whereas the saprophytic soil-dwelling Mycobacterium smegmatis genome has only one TA locus, clearly highlighting the role of TA systems in bacterial adaptation and survival in a very hostile environment inside the host (27). Based on the mechanism of toxin activation, the TA system is classified into seven different types. The VapBC family, the most characterized of all, belongs to the type II group. The toxin from the type II group targets all forms of RNA, including mRNA, rRNA, and tRNA. Although the type II TA system has been shown to regulate growth in several bacterial species, the exact mechanism regulating the generation of persistence is not well defined. It is hypothesized that each TA pair is required for the survival of bacteria under specific growth and stress conditions (28,29); however, the presence of a very high number of the TA systems in the M. tuberculosis genome also increases the chances of redundancy and the possibility of multiple TA systems regulating specific growth conditions.
In the present study, we identified the role of one such M. tuberculosis RNase toxin, VapC12, to be critical for the cholesterol-induced generation of antibiotic persistence in mycobacteria. Our data conclusively demonstrate that cholesterol activates the RNase toxin by disrupting its binding to the cognate antitoxin VapB12. We further demonstrated that the proT tRNA of M. tuberculosis is a bona fide substrate of the VapC12 RNase toxin and that the toxin-mediated modulation of the proT tRNA regulates the generation and enrichment of the cholesterol-induced persisters in mycobacteria. Finally, we also demonstrated that the vapC12 gene is critical for M. tuberculosis to persist in a guinea pig model of tuberculosis infection.

RESULTS
The vapC12 gene is essential for cholesterol-specific growth modulation in M. tuberculosis. We have previously demonstrated that the utilization of host cholesterol as a carbon source is essential for maintaining persistence during M. tuberculosis infection (14). In this study, we used M. tuberculosis grown in minimal medium with cholesterol as the only carbon source (this will be referred to as cholesterol medium henceforth) as an in vitro model to examine the role of cholesterol in the formation, maintenance, and enrichment of persisters during M. tuberculosis infection. We have initially analyzed the metabolic and replication rates of M. tuberculosis grown in the cholesterol media. In comparison to glycerol, we observed a 10-fold decrease in the growth and a 3-fold decrease in the metabolic rates of wild-type (WT) M. tuberculosis (H37Rv) grown in the cholesterol media ( Fig. 1A and B; see also Fig. S1A in the supplemental material). Surprisingly, the presence of cholesterol was found to have an inhibitory effect on the growth of both M. tuberculosis and M. bovis BCG (BCG) strains (see Fig. S1A to C). An in vitro time-kill curve assay revealed a cholesterol-specific increase in the frequency of the generation of a rifamycin-tolerant subpopulation in WT H37Rv (Fig. 1C). Since the TA loci across bacterial species are known to modulate growth (22,(30)(31)(32), we speculated the aforementioned phenotype to be regulated by one of the M. tuberculosis TA loci. For that, we analyzed the data of a transposon mutagenesis screening performed in M. tuberculosis H37Rv to identify genes essential for growth under the condition where cholesterol is provided as the sole carbon source (33). Through manual curation of the data, we identified transposon insertions in six VapC toxins that imparted the insertion mutant strains a significant growth advantage in cholesterol relative to glycerol media. This finding suggests the role of one or all of these toxins in the cholesterol-mediated growth modulation of M. tuberculosis (see Fig. S2A). Of these six vapC genes, we generated clean deletion mutants for the top two toxins, VapC8 (Rv0665) and VapC12 (Rv1720c), which demonstrated the highest increase in the growth rate. We found that in comparison to the WT strain, the vapC8null strain showed no significant growth difference in the cholesterol media (see Fig. S2B), whereas the mutant strain lacking the vapC12 gene failed to slow down its growth and was found to be metabolically more active in the cholesterol media relative to the glycerol, suggesting essentiality of this gene in cholesterol-mediated growth retardation. The mutant phenotype was found to be gene-specific because adding back the vapBC12 locus in the mutant strain restored the WT phenotype ( Fig. 1D and E). As expected, the cholesterol-mediated increase in the frequency of the generation of the rifamycin-tolerant population, observed in the WT strain, was reversed in the vapC12-null strain, underscoring the role of this putative toxin in the generation of cholesterol-induced antibiotic persistence in mycobacteria ( Fig. 1F; see Fig. S3A in the supplemental material). To gain additional insights, we performed transcriptional profiling of these strains in media containing glycerol and cholesterol as a sole carbon source through transcriptome sequencing (RNA-seq) analysis. As expected, relative to glycerol, both WT and vapC12 mutant strains grown in cholesterol had higher levels of expression of genes involved in cholesterol metabolism, the methyl-citrate cycle, and glyoxylate pathways (34,35). The slowing down of WT M. tuberculosis in the cholesterol media can be attributed to lower expression in the transcript levels of genes involved in respiration (e.g., cytochrome and the ATP synthesis pathway genes). Intriguingly, relative to glycerol, we observed significantly lower expression of genes belonging to the esx3 locus in M. tuberculosis grown in cholesterol media, suggesting the possibility of iron-mediated growth modulation in the cholesterol-rich environment (36,37). In addition, in comparison to the glycerol, we observed an  Table 1; see also Fig. S3B in the supplemental material). Despite the cholesterol-specific differences observed in the growth phenotype between H37Rv and vapC12 mutants, expression profiling data showed few differentially expressed genes between the samples, suggesting that a posttranscriptional regulation mechanism plays a decisive role in inducing cholesterol-mediated growth regulation in M. tuberculosis (Table 2). Since ATP depletion is one of the mechanisms through which bacteria increase their tolerance to antibiotics leading to persistence (21,38,39) and our RNA sequencing data also revealed differential expression of ATP synthesis pathway genes, we quantified intracellular ATP levels in glycerol-and cholesterol-grown BCG cultures. Compared to the glycerol, the cholesterol-grown BCG culture demonstrated a 25-fold decrease in intracellular ATP levels. This cholesterolspecific depletion of intracellular ATP depends on the presence of the vapC12 gene (Fig. 1H).
VapC12 RNase toxin targeting proT is essential for cholesterol-mediated growth regulation in M. tuberculosis. Since the VapC family of toxins targets RNAs (40)(41)(42), the presence of proT-tRNA (proT) gene upstream to the vapC12 gene was intriguing. Thus, we hypothesized that this proT tRNA can be one of the substrates for the toxin ( Fig. 2A). As predicted, we observed a cholesterol-specific decrease in the proT transcript levels in wild-type strain (see Fig. S4A), and this cholesterol-specific decrease was not observed in a vapC12-null strain (Fig. 2B), suggesting that proT tRNA could be one of the major substrates of the VapC12 RNase toxin. To further confirm the proT specificity, we quantified the transcript levels of 10 different tRNAs that had GC-rich anticodon sequences. Surprisingly, we found no cholesterol specific differences in the abundance of any of the tested tRNAs, emphasizing that proT is one of the major substrates of the VapC12 RNase toxin ( Fig. 2B; see also Fig. S4B). To further rule out the role of low growth rate contributing to the above phenotype, we quantified the relative proT expression in a WT strain (M. bovis BCG) grown in palmitate as the sole carbon source (see Fig. S5A), and we did not observe any palmitate-specific changes in the proT transcript levels (Fig. 2B).
In contrast to the findings of previous studies (27,43), we observed that overexpression of the putative vapC12 toxin gene in a WT M. bovis BCG background reduced the proT transcript levels eliciting significant growth defect ( Fig. 2C to E). This phenotype was more pronounced in the presence of cholesterol (see Fig. S5B). Furthermore, the FIG 1 Legend (Continued) points postinoculation. Experiments were performed in triplicates, and data represent the means 6 the standard errors of the mean (SEM). (B and E) Resazurin-based estimation of the metabolic activity of H37Rv (B) and DvapC12 and DvapC12:vapBC12 (E) strains grown in minimal medium supplemented with glycerol and cholesterol. Strains were serially diluted in a 96-well plate in respective media. The experiment was performed in two independent sets, and the plate was incubated at 37°C for 5 days. One set of the experiment was used for recording fluorescence after adding the PrestoBlue reagent at 570 or 585 nm, whereas the other set was used for the enumeration of bacteria present in each well. The metabolic activity calculated for each well is representative of the mean fluorescence readout per bacteria from three independent experiments. Data were analyzed using unpaired Student t test. *, P , 0.05; **, P , 0.01. (C and F) Kill curve of M. bovis BCG (C) and BCGDvapC12 and BCG DvapC12:vapBC12 (F) strains grown in glycerol-and cholesterol-rich media. Log-phase cultures of strains were washed with PBS-tyloxapol and inoculated at an absorbance of 0.05. The cultures were allowed to grow for 4 days before being treated with 5Â MIC of rifamycin. Bacterial enumeration was performed through CFU plating of cultures on 7H111OADC plates at various time points. The kill curve was plotted by calculating the percent survival. The experiment was repeated three times, and data represented as means 6 the SEM. Data were analyzed using unpaired Student t test. *, P , 0.05; **, P , 0.01. (D) Percent wild-type growth of the vapC12 mutant and DvapC12:vapBC12 strains in a minimal media containing 0.1% glycerol and 0.01% cholesterol. Growth was estimated by CFU plating of cultures on 7H111OADC plates at 8 days postinoculation. The experiment was repeated three times, and the data represent the means 6 the SEM. The data were analyzed using unpaired Student t test. *, P , 0.05; **, P , 0.01. (G) Heatmap visualization of differentially expressed transcripts in wildtype H37Rv grown in glycerol and cholesterol media, analyzed by RNA-seq. Expression data of the respective genes based on the adjusted FDR are depicted in the heatmap. The RNA for sequencing was isolated from four different sets of cultures grown in the respective media. (H) Percent estimation of ATP in wild-type BCG and DBCGvapC12 strains grown in a cholesterol-rich media relative to glycerol media. ATP estimated in micromolar concentrations was normalized with per milligram of protein in each sample. The experiment was repeated three times, and the data plotted represent means 6 the SEM. The data were analyzed using an unpaired Student t test. *, P , 0.05; **, P , 0.01. toxin phenotype was reversed if the vapC12 toxin gene was coexpressed along with its cognate vapB12 antitoxin gene ( Fig. 2C) (44). To further validate whether proT is the substrate of VapC12 toxin, we generated recombinant His-tagged VapC12 toxin expressed and purified in a heterologous E. coli expression system (see Fig. S6A). When exposed to in vitro transcribed tRNAs, namely, proT and proU, the purified recombinant toxin specifically cleaved proT ( Fig. 2F and G). Furthermore, we mutated two highly conserved aspartate residues D 5 and D 94 in the PIN domain of the toxin to alanine (see Fig. S6B). An aspartate (D)-to-alanine (A) conversion of the 94th residue of the VapC12 toxin failed to cleave the substrate. This inactivation of the toxin by D 94 A substitution may be due to its inability to bind to Mg 21 , which is a critical cofactor required for its activity (40,45). Although we used proU-tRNA (proU) as our control, we do not rule out the possibility of other tRNAs being a VapC12 substrate. A D 5 A substitution did not affect the RNase activity of the toxin. Activation of the type II toxins is primarily driven by the degradation of the corresponding antitoxin. Therefore, to study cholesterol-specific degradation of the antitoxin, we generated a recombinant M. bovis BCG overexpressing both toxin and antitoxin in which the antitoxin was tagged with 6ÂHis at the N terminus. By Western blot analysis, we observed that in the presence of cholesterol there was a time-dependent dissociation and degradation of the antitoxin (VapB12) protein that possibly resulted in the generation of the active monomeric form of the cognate toxin ( Fig. 2H and I). Surprisingly, the only lysine residue, K 19 , of the antitoxin lost its acetylation in the presence of cholesterol (Fig. 2J), and this we believe could possibly be a signal for cholesterol-induced degradation of antitoxin and the subsequent activation of VapC12 toxin.   We validated this by liquid chromatography-tandem mass spectrometry (LC-MS/MS) wherein we observed that, unlike cholesterol, acetylation of the lysine residue (K 19 ) of the antitoxin protein was observed only in the protein lysates generated from glycerolgrown M. tuberculosis culture (Fig. 2K). Surprisingly, the protein coverage of the antitoxin peptides isolated from M. tuberculosis grown in cholesterol was $95% except for the peptide LHELK with a sequence coverage of $50 , 95 (see Fig. S7 in the supplemental material). This could be attributed to a cholesterol-specific degradation of the antitoxin (Fig. 2I). To confirm this further, we generated a recombinant M. bovis BCG strain overexpressing the antitoxin protein harboring a lysine to alanine (K 19 A) substitution and, as expected due to the absence of the lysine residue, the antitoxin could not be acetylated (Fig. 2J). This resulted in constitutive degradation of the antitoxin leading to growth inhibition, independent of the carbon source, in mycobacteria (Fig. 2L).

Cholesterol-Induced Antibiotic and Disease Persistence
Cholesterol-dependent activation of vapC12 toxin generates and enriches the slow-growing population in the M. tuberculosis culture. To further evaluate cholesterol-induced activation of VapC12 toxin and the subsequent slowdown of M. tuberculosis growth, a log-phase culture of WT M. tuberculosis grown in an enriched (7H91OADC [oleic acid-albumin-dextrose-catalase]), cholesterol, and glycerol medium was subsequently exposed to cholesterol, and the effect on bacterial growth was assessed (Fig. 3A). Interestingly, while there was no change in the growth rate of culture previously exposed to cholesterol, a significant reduction in the growth of M. tuberculosis was observed in cultures that were growing in both glycerol and the enriched media before the cholesterol switch (Fig. 3B). This phenotype was again found to be dependent on the presence of the vapC12 toxin gene (Fig. 3C). These results indicate a vapC12-mediated cholesterol-dependent reduction in M. tuberculosis growth. Compared to the glycerol media, the reduction in the growth rate was more prominent in the enriched media. We also found a reduction in the transcript levels of the proT tRNA in cholesterol-exposed M. tuberculosis culture previously grown in either enriched media or glycerol media (Fig. 3D), further confirming the finding that the difference in growth is due to toxin-mediated degradation of proline-tRNA. To explore the extracellular role of VapC12 toxin in restricting the growth of fast-growing bacteria in a heterogeneous population (46,47), we suspended a log-phase culture of the M. tu- , and mutant VapC12D 5 A toxin protein incubated with proT (F). Each reaction mixture was incubated at 37°C for 3 h. The products of each of the reactions were run on a 3% agarose gel and visualized by adding ethidium bromide, followed by exposure to UV light. (G) Relative density of marked RNA bands in panel F quantified using ImageJ. The experiment (see panel F) was repeated three times, and data were plotted to represent the means 6 the SEM. (H) Schematic representation of the protocol for the experiment to demonstrate cholesterol-specific dissociation of the antitoxin. (I) Western blot for cholesterol-specific dissociation and degradation of the antitoxin from the toxin-antitoxin complex. The His-tagged antitoxin was tracked using an anti-His antibody in the cell lysate of BCG overexpressing His-tagged antitoxin as a part of the toxin-antitoxin complex. Cell lysates were prepared by sampling cultures grown in both glycerol and cholesterol media at different time points and probed with an anti-His antibody. (J) Western blot of the protein lysates prepared from BCG overexpressing the toxin-antitoxin locus (VapBC12) with His-tagged antitoxin VapB12 and BCG strain with N-terminal His tagged VapBC12, wherein the lysine residue of the antitoxin is converted to alanine (VapB K:A C12). Immunoprecipitation was performed with mouse anti-His antibody and probed with rabbit anti-acetyl lysine and anti-His antibodies. To normalize for the amount of the protein, a 3-fold higher concentration of protein was loaded in the cholesterol-grown BCG sample. (K) Mass spectrometry analysis of His-tagged antitoxin protein isolated from BCG overexpressing VapBC12 complex grown in glycerol and cholesterol media. Tryptic digests of immunoprecipitated samples from glycerol-and cholesterol-grown cultures were analyzed by LC-MS/MS (Sciex, 5600 Triple-TOF). A representative MS/MS spectrum of the peptide from the glycerol-grown sample, ELLHELK(Ac)AR, was acetylated and displays a mass shift corresponding to acetylation (m/z 416.26) compared to the unmodified peptide from the cholesterol-grown sample. (L) Growth curve of M. bovis BCG, BCG overexpressing toxin-antitoxin (vapBC12), and lysine mutant (vapB K:A C12) strains in a minimal media plus 0.1% glycerol and minimal media plus 0.01% cholesterol. Bacterial enumeration was performed by CFU plating on 7H111OADC plates at regular intervals. The experiment was performed in triplicates, and the data plotted represent means 6 the SEM. Data were analyzed by using an unpaired Student t test. *, P , 0.05; **, P , 0.01. berculosis vapC12 mutant strain separately in the spent media harvested from either the cholesterol-grown WT or vapC12 mutant strain (Fig. 3E). A decrease in the vapC12 mutant CFU was observed only in culture exposed to supernatant isolated from the cholesterol-grown WT strain, suggesting that either VapC12 toxin directly or a VapC12dependent secretory protein selectively enriches the slow-growing persister population in M. tuberculosis cultures in a cholesterol-rich environment (Fig. 3F). The quantification of proT levels in these cultures further suggested that the observed phenotype was indeed due to differences in the toxin-mediated proT cleavage (Fig. 3G). Furthermore, neutralization of the toxin in WT spent media by adding a purified antitoxin increased the vapC12 mutant CFU; similarly, the addition of the purified toxin in spent media from DvapC12 resulted in a dose-dependent decrease in the CFU of the vapC12-null strain (Fig. 3H). Finally, we demonstrated that VapC12 toxin was detected only in the culture filtrate isolated from cholesterol-grown M. bovis BCG overexpressing Flag-tagged VapC12 and not from the glycerol-grown culture ( Fig. 3I and J).
Significance of vapC12-mediated downregulation of the proT-codon enriched proteins. To investigate the implications of proT degradation in cholesterol-mediated growth modulation, a genomewide in silico analysis was performed to determine the frequency of the proT tRNA codon in each M. tuberculosis gene (see Fig. S8A in the supplemental material). The mycobacterial genome has four designated proline tRNAs (pro-T, pro-Y, pro-U, and pro-X) that incorporate the proline residue to a nascent polypeptide during translation. The results of the in silico analysis revealed that pro-T and pro-Y encode 85.53% of the total proline incorporated into the M. tuberculosis H37Rv proteome, with pro-Y (CCG) and pro-T (CCC) codon usage being 63.6 and 36.4%, respectively (see Fig. S8A). A list of 136 M. tuberculosis genes that had at least 60% of the proline encoded by proT tRNA was identified (Table 3). In contrast, a functional categorization revealed that proline incorporated in the PE-PGRS protein family has a significantly higher percentage (56.2%) of proline codons that require the proT tRNA. Besides, a gradient in the percentage of the proT codon usage was observed (see Fig. S8B). This led to speculation that the expression of antigenic proteins belonging to the PE-PGRS family, which contain various numbers and frequencies of the proT codons, are differentially regulated in a cholesterol-rich environment. We hypothesized that through VapC12 toxin-mediated degradation of proT in cholesterol, M. tuberculosis downregulates the expression of these antigenic proT-rich PE-PGRS proteins. We selected a set of five different PE-PGRS proteins with varying proT codon usage and found that the expression levels of PE-PGRS proteins in wild-type M. bovis BCG grown in cholesterol media were inversely correlated with the frequency of proT encoded proline in the gene (  The vapC12 mutant strain was grown in a medium containing 0.01% cholesterol in triplicate for the first 4 days and then resuspended in spent medium from H37Rv, vapC12 mutant, and fresh cholesterol individually. Bacterial enumeration was performed by plating cultures on 7H111OADC plates. The experiment was repeated three times, and the data plotted represent the means 6 the SEM. Data were analyzed using unpaired Student t test. *, P , 0.05; **, P , 0.01. (G) Expression analysis of proT tRNA through qRT-PCR in the vapC12 mutant strain grown in a cholesterol-spent media at day 10 of the growth curve relative to the culture grown in a fresh cholesterol-rich media at day 4 (see panel F). (H) The BCG vapC12 mutant strain was grown in a media containing 0.01% cholesterol in triplicate for the first 4 days and then resuspended in spent media from wildtype BCG, vapC12 mutant, and wild-type BCG supplemented with purified VapB12 antitoxin (12.2 nM) and VapC12 toxin at two different concentrations (6.9 and 13.8 nM). Bacterial enumeration was performed by CFU plating on 7H111OADC plates. The experiment was repeated three times, and the data plotted represent the means 6 the SEM. Data were analyzed using unpaired Student t test. *, P , 0.05 **, P , 0.01. (I) Schematic representation of the experiment to demonstrate that toxin is secreted out in the culture filtrate of BCG. (J) Western blot showing the Flag-tagged toxin in the culture filtrate of BCG strain overexpressing the toxin-antitoxin complex (VapBC12). The culture filtrate was probed with an anti-Flag antibody to detect the toxin protein, the anti-Ag85B antibody was used as a positive control for the secretory protein, and an anti-GroEL1 antibody was used as a negative control to ensure no lysis of bacterial cells occurred during sample preparation. Lipid metabolism 60 Rv2724c

Cholesterol-Induced Antibiotic and Disease Persistence
Lipid metabolism 60 Rv2982c Lipid metabolism 60 Rv3221c Lipid metabolism 100 Rv0771 Intermediary metabolism and respiration 60 Rv1851 Intermediary metabolism and respiration 60 Rv1826 Intermediary metabolism and respiration 60 Rv2499c Intermediary metabolism and respiration 60 Rv2511 Intermediary metabolism and respiration 60 Rv2250A Intermediary metabolism and respiration 63.63636364 Rv1692 Intermediary metabolism and respiration 64.70588235 Rv1311 Intermediary metabolism and respiration 66.66666667 Rv1555 Intermediary metabolism and respiration 66.66666667 Rv1990A Intermediary metabolism and respiration 66.66666667 Rv2539c Intermediary metabolism and respiration 66.66666667 Rv3145 Intermediary metabolism and respiration 66.66666667 Rv2421c Intermediary metabolism and respiration 70 Rv3624c Intermediary metabolism and respiration 70 Rv2754c Intermediary metabolism and respiration 71.42857143 Rv0558 Intermediary metabolism and respiration 72.72727273 Rv0814c Intermediary metabolism and respiration 75 Rv1305 Intermediary metabolism and respiration 75 Rv3118 Intermediary metabolism and respiration 75 Rv3154 Intermediary metabolism and respiration 77.77777778 Rv2537c Intermediary metabolism and respiration 80 Rv0137c Intermediary metabolism and respiration 81.81818182 (Continued on next page)  Insertion sequences and phages 100 Rv1316c Information pathways 60 Rv2069 Information pathways 60 Rv2906c Information pathways 60 Rv2058c Information pathways 66.66666667 Rv2056c Information pathways 66.66666667 Rv1643 Information pathways 66.66666667 Rv0722 Information pathways 100 Rv2441c Information pathways 100 Rv3053c Information pathways 100 Rv3462c Information pathways 100 Rv1772 Conserved hypotheticals 60 Rv0678 Conserved hypotheticals 60 Rv0607 Conserved hypotheticals 60 Rv3678A Conserved hypotheticals 60 Rv1590 Conserved hypotheticals 60 Rv2283 Conserved hypotheticals 60 Rv2426c Conserved hypotheticals 60 Rv2438A Conserved hypotheticals 60 Rv2558 Conserved hypotheticals 60 Rv3224B Conserved hypotheticals 60 Rv0323c Conserved Conserved hypotheticals 100 Rv0378 Conserved hypotheticals 100 Rv1893 Conserved hypotheticals 100 Rv1993c Conserved hypotheticals 100 Rv2548A Conserved hypotheticals 100 Rv2738c Conserved hypotheticals 100 Rv3440c Conserved hypotheticals 100 Rv0011c Cell wall and cell processes 60 Rv0431 Cell wall and cell processes 60 Rv1463 Cell wall and cell processes 60 Rv1973 Cell wall and cell processes 60 Rv2856 Cell wall and cell processes 60 Rv2936 Cell wall and cell processes 61.53846154 Rv3864 Cell wall and cell processes 61.9047619 Rv3312A Cell wall and cell processes 62.5 Rv0583c Cell wall and cell processes 65 Rv2732c Cell wall and cell processes 66.66666667 (Continued on next page) Cholesterol-Induced Antibiotic and Disease Persistence script level and, to our surprise, we found that, in comparison to glycerol, all other genes (except for Rv1068c) were found to be transcriptionally upregulated in cholesterol. These findings further confirm that the cholesterol-dependent decrease in the expression of these proteins is perhaps regulated at the translation level (see Fig. S8D).
The factor rpfA, one of the four resuscitation-promoting factors (rpfA to rpfD) of M. tuberculosis, has 53.16% of its proline encoded by proT. These Rpf proteins are peptidoglycan glycosidases required for the activation of quiescent bacteria; this is an essential step for the reactivation of M. tuberculosis in an in vitro model (48). In addition, rpfA in M. tuberculosis has been reported to be secretory through a sec-dependent pathway and speculated to be involved in modulating host during the reactivation process (49). Interestingly, in comparison to glycerol, we found a vapC12-dependent decrease in the expression of RpfA protein in cholesterol grown M. bovis BCG strain overexpressing a 6ÂHis-tagged M. tuberculosis RpfA protein. We believe that the cholesterol-dependent regulation of RpfA levels by the VapC12 toxin is a mechanism by which M. tuberculosis sustains latency during infection.
The vapC12 mutant strain of M. tuberculosis demonstrates a hypervirulence phenotype in the guinea pig infection model. Before initiating any infection-related studies, we confirmed the presence of phthiocerol dimycocerosates (PDIM), a virulence-associated surface lipid, in the wild type, vapC12 mutant, and the complemented strains (see Fig. S9A in the supplemental material). To assess the possible role of vapC12 in the host, we first infected mouse bone marrow-derived macrophages (BMDM) with WT and vapC12-null strains. The null strain demonstrated increased replication in BMDM, which was abolished in the complemented strain (see Fig. S9B). Of note, the vapC12-null strain also showed an enhanced growth phenotype when exposed to higher oxidative and nitrosative stress conditions (see Fig. S9C). Next, we infected guinea pigs with WT, vapC12-null, and complemented M. tuberculosis strains. At 7 weeks postinfection, a higher bacterial load was observed in the lungs of guinea pigs infected with the vapC12-null strain compared to animals infected with a WT or complemented strain (Fig. 4B). A similar profile was observed in the spleen (see Fig. S9D). In comparison to the wild type, gross examination of the lungs infected with the mutant strain revealed higher inflammation and enhanced tissue damage. Histological examination of the lungs infected with the wild-type strain at 7 weeks postinfection revealed a typical necrotic granuloma with caseated nuclei surrounded by a dense fibrous layer (Fig. 4C and D). In contrast, at the same time postinfection, lung tissues of guinea pigs infected with the vapC12-null strain displayed an increased number of granulomas ( Fig. 4C and D; see also Fig. S9E), which were predominantly diffused and nonnecrotic, devoid of any outer fibrous layer (Fig. 4E). We also observed that these less-defined granulomas were infiltrated with a greater admixture of cell types, including neutrophils. Surprisingly, animals infected with the vapC12-null strain failed to induce an inflammatory response, as indicated by a decreased mRNA expression of inflammatory cytokines (Fig. 4F). Guinea pigs infected with a complemented strain had a phenotype similar to those infected with a WT strain (Fig. 4C to F). Cell wall and cell processes 75 Rv1881c Cell wall and cell processes 75 Rv2301 Cell wall and cell processes 75 Rv3271c Cell wall and cell processes 75 Rv0900 Cell wall and cell processes 100 Rv0039c Cell wall and cell processes 100 Rv0288 Cell wall and cell processes 100 Rv2520c Cell wall and cell processes 100 Rv3789 Cell wall and cell processes 100 Rv3857c Cell wall and cell processes 100

DISCUSSION
Chronic infections necessitate the etiologic agent to persist inside the host for an extended duration. M. tuberculosis remarkably adapts to a very hostile niche by augmenting its ability to thrive inside the host for decades. The pathogen's ability to modulate host immune response and its capacity to tolerate high concentrations of antimycobacterial drugs are key to its persistence. To do so, M. tuberculosis senses various stage-specific environmental cues and accordingly regulates the expression of various proteins that eventually help the pathogen to attain distinct phenotypes critical for long-term survival. Inside the host, M. tuberculosis encounters an extraordinary challenge of surviving on host-derived nutrients and subsequently creating a niche conducive for its growth. Conversely, the host has developed ways and means to deprive the unwanted guest of critical nutrients, including the much-needed carbon source. Although we have earlier demonstrated that the utilization of host cholesterol is essential for disease persistence in tuberculosis, the role of cholesterol utilization and the subsequent mechanism leading to this phenotype is largely not well defined.
In the present study, we demonstrated that cholesterol utilization results in an increase in the frequency of generation of antibiotic persisters in mycobacteria. This phenotype was abrogated in an M. tuberculosis RNase toxin vapC12-null strain. Mechanistically, we also identified M. tuberculosis proT tRNA as one of the substrates of the VapC12 RNase toxin and that the toxin-mediated modulation of the proT tRNA levels regulate antibiotic persistence in mycobacteria. Finally, using a guinea pig model of M. tuberculosis infection, we demonstrated that a reduction in the frequency of generation of in vitro antibiotic persisters significantly enhances the ability of vapC12-null strain to grow inside the host affecting disease persistence (50). In line with the recently established definitions and guidelines (51), our data suggest that in tuberculosis both antibiotic and disease persistence, either individually or in tandem, influence the disease progression and treatment outcomes. Coevolution for centuries has molded M. tuberculosis to adapt and utilize host-derived lipids, including cholesterol as a preferred carbon source (52,53). Although M. tuberculosis has access to multiple carbon sources inside the host (15,16), the utilization of host cholesterol is essential for long-term persistence. While nutrient-dependent growth modulation is very common (54,55), our data for the first time indicate its effect on antibiotic and disease persistence during M. tuberculosis infection.
Interestingly, a decrease in the expression of esx3 locus, a type VII secretion system critical (39) for iron uptake (36,37), during cholesterol utilization suggests that M. tuberculosis deprives itself of iron to restrict growth under cholesterol-rich condition. We also found that the cholesterol-exposed M. tuberculosis downregulates the expression of genes belonging to the electron transport chain, resulting in a sharp decline in the intracellular ATP levels. Our data are in line with similar studies implicating that a lower ATP concentration leads to disease persistence in several species of bacteria, including M. tuberculosis (38,39,56,57). These findings suggest that the modulation of intracellular ATP levels by a vapC12 gene-encoded protein might have a role in cholesterolspecific growth modulation in M. tuberculosis. Obtaining mechanistic insights into pathways leading to VapC12 toxin-dependent regulation of intracellular ATP levels would be an interesting area for future research. In addition, an increase in the transcript levels of DosR regulon genes in the cholesterol media suggests that, in addition to hypoxia, sensing of intracellular cholesterol by M. tuberculosis can trigger the induction of DosR regulon genes in M. tuberculosis. Surprisingly, despite the cholesterol-specific growth differences observed between the WT and vapC12 mutant strains, the differences observed in the transcript levels were very minimal, implicating posttran- scriptional regulation for the observed phenotype. Our findings demonstrated growth modulation specifically attributed to the abundance of proline tRNA levels modulated by activation of VapC12 RNase toxin. Although studies describing tRNA-dependent growth modulation have been reported (28,58,59), our study describing the mechanism of nutrient-dependent regulation of tRNA abundance modulating growth is a novel finding.
Posttranslational modifications (PTMs) confer diversity to regulatory mechanisms that control various cellular pathways. Two of the most extensively studied PTMs are phosphorylation and acetylation. Together, they are known to regulate the stability and activity of proteins in both eukaryotes and prokaryotes. Lysine acetylation is known to regulate various cellular pathways conserved across species, including mycobacteria (60,61). Our growth curve and mass spectrometry data also suggest that the cholesterol-mediated growth modulation is triggered by deacetylation of the only lysine residue present in the antitoxin. This assumes importance since M. tuberculosis in the host is known to persist inside foamy macrophages rich in cholesterol (18).
Our findings also suggest that the proT-encoded proline-rich proteome of M. tuberculosis, including PE-PGRS proteins, are essential for cholesterol-induced growth modulation, leading to antibiotic persistence. We predict that VapC12-mediated differential expression of these immunoregulatory PE-PGRS proteins critically regulates the disease progression and hence influences the treatment outcomes. Interestingly, our finding suggests that an increased exposure of M. tuberculosis to host cholesterol downregulates the expression of these immunomodulatory proteins, leading to enhanced proinflammatory cytokine secretion. This pathogen-driven proinflammatory response helps M. tuberculosis to establish a long-term niche inside the host in the form of a welldefined structure called granuloma: a hallmark of tuberculosis infection. M. tuberculosis also induces the infiltration of neutrophils in the lung tissues of infected mice. A recent report suggests that a subset of these tissue-infiltrated neutrophils provides a permissive intracellular niche facilitating increased survival of M. tuberculosis inside the host. These granulocytes are also reported to be involved in tissue damage and induce immunoregulatory functions by actively secreting anti-inflammatory cytokines (62). In comparison to the wild type, an increase in neutrophil infiltration observed in the lungs of guinea pigs infected with vapC12 mutant strain possibly explains the enhanced pathology and increase in mutant growth fitness phenotype observed in guinea pigs. Our data possibly indicate that a vapC12-mediated regulation of granulocyte infiltration during infection is essential for disease persistence in tuberculosis. The role of the temporospatial expression of critical M. tuberculosis surface antigens in host immune response and its effect on the recruitment of immune cells to the site of infection would require a more integrative approach and in-depth immunophenotyping studies.
VapC12-mediated proT codon-based differential expression of various M. tuberculosis proteins, including PE-PGRS, is a novel finding. Interestingly, the presence of an unusually high number of the vapBC TA pairs in M. tuberculosis, together with the existing posttranslational modification, we believe, might add to the existing complexity of gene regulation at the posttranscriptional level. A comprehensive analysis of the stage-specific regulation of the expression of M. tuberculosis proteins and its implications on the disease progression will be an interesting area for future research. While pathogen rewiring their metabolic pathways for disease persistence is quite well studied (63,64), there are fewer studies on the modulation of host immune response by the temporospatial expression of M. tuberculosis surface antigens contributing to disease persistence. Our study suggests that both the growth modulation and differential expression of surface antigens by M. tuberculosis might be a pathogen-driven mechanism that contributes to disease persistence. A recent publication in support of this hypothesis suggests how ubiquitination of one of the proT-enriched proline codon proteins belonging to the PE-PGRS family was identified as a signal for the host to eliminate the pathogen (65). This information could be used to design a better and more efficient vaccine against tuberculosis. In light of our current findings, it will be very intriguing to study the role of PE-PGRS proteins in modulating the host response and their role in the disease progression during M. tuberculosis infection. Furthermore, the functional characterization of proT tRNA-encoded proline-rich proteins and their implications in the stage-specific replication and growth rate of M. tuberculosis inside the host should be explored.
The findings support our hypothesis that the VapC12 toxin acts as a molecular switch that regulates growth in the presence of cholesterol. Because an actively growing M. tuberculosis culture is always heterogeneous and has individual bacteria growing at different rates, the rate of growth is directly proportional to the level of the VapBC12 TA protein accumulated in the cytoplasm. Upon exposure to cholesterol, the fate of each bacterium is dictated by the concentration of the activated toxin present inside the cell. Depending on the intracellular levels of the activated toxin, the bacterium is either eliminated or acquires a metabolically less active state. This results in the enrichment of the slow-growing population in M. tuberculosis culture exposed to a cholesterol-rich environment (see Fig. S10) in the supplemental material. Furthermore, the extracellular presence of this toxin ensures the clearance of any rapidly dividing mutant bacteria generated due to the spontaneous incorporation of a genetic lesion. This is the first study to identify a novel mechanism of cholesterol-dependent stochastic enrichment of slow-growing M. tuberculosis during mycobacterial infection.
These findings will help to identify a novel mechanism of the generation of antibiotic persistence and define targets against the persister population. Approaches targeting the persister population will enhance the rate of clearance of the pathogen, resulting in a significant reduction in the duration of treatment. This will help in significantly reducing the risk associated with the current extended regimen extending from 6 months to 2 years. Thus, we have empirically demonstrated that both antibiotic and disease persistence contributes to chronic M. tuberculosis infection, and targeting pathways essential for both could potentially shorten the treatment regimen. This finding is significant since a better understanding of the disease persistence and targeting the M. tuberculosis persister population as a therapeutic strategy will open new paradigms in tuberculosis treatment.

MATERIALS AND METHODS
Bacterial strains and culture. M. tuberculosis mutants were derived from strain H37Rv by using homologous recombination between the suicide plasmid and bacterial genome. Flanking regions (1,000 bp) of the target gene, Rv1720c (vapC12) were cloned in pJM1 suicide vector and electroporated in H37Rv-competent cells using standard M. tuberculosis protocols (Tanya Parish and Neil G. Stroker). The strains were maintained on Middlebrook 7H11 agar or 7H9 broth (Difco Middlebrook 7H11 agar [catalog no. 283810] and 7H9 broth [catalog no. 271310]) supplemented with 10% OADC enrichment. Hygromycin was added at 50 mg/ml. To complement the vapC12 mutant, the loxP-flanked chromosomal hygromycin resistance gene was excised by the expression of Cre recombinase. This strain was transformed with pJEB402 harboring the Rv1720c-1721c (vapBC12) genes. For growth on defined carbon sources, strains were grown in "minimal media" (0.5 g/liter asparagine, 1 g/liter KH 2 PO 4 , 2.5 g/liter Na 2 HPO 4 , 50 mg/liter ferric ammonium citrate, 0.5 g/liter MgSO 4 s²7H 2 O, 0.5 mg/liter CaCl 2 , 0.1 mg/liter ZnSO 4 ) containing 0.1% (vol/vol) glycerol or 0.01% (wt/vol) cholesterol and 50 mg/ml sodium palmitate. Growth was determined by CFU plating at different time points on 7H11 with 10% OADC plates. Similar protocols were followed for maintaining and culturing M. bovis BCG strains Growth curve. Log-phase cultures of wild-type H37Rv, DvapC12, and DvapC12:vapBC12 strains were washed with phosphate-buffered saline plus Tween (PBST) twice and inoculated in minimal media with 0.1% glycerol and 0.01% cholesterol, respectively, at an absorbance of 0.005. Aliquots of the cultures were obtained at different time points and plated on 7H111OADC plates for bacterial enumeration. Similar protocols were followed for M. bovis BCG strains.
Resazurin-based metabolic activity assay. Log-phase cultures of wild-type H37Rv, DvapC12, and DvapC12:vapBC12 strains at an optical density (OD) of 0.5 were washed with PBST twice, and the OD at 600 nm (OD 600 ) was set to 0.05 in glycerol and cholesterol media. These cultures were serially diluted in the respective media in a 96-well plate. The experiment was performed in duplicate, and both plates were incubated at 37°C for 5 days before PrestoBlue cell viability reagent (Invitrogen catalog no. A13261) was added to each well in one set of the plates. The plates were incubated for another 2 days. The fluorescence read-out of the plate with PrestoBlue was taken at 570/585 nm using a Synergy HTX multi-mode microplate reader. For bacterial enumeration in each well, CFU plating was done with a plate with no PrestoBlue. To determine the average metabolic activity, the total fluorescence recorded was normalized for the number of bacteria in the corresponding well.
Antibiotic kill curve. Log-phase cultures of wild-type M. bovis BCG, BCG vapC12 mutant, and BCG DvapC12:vapBC12 strains grown in 7H9-enriched media were washed with PBST twice and inoculated in glycerol and cholesterol media at an absorbance of 0.05. The cultures were allowed to grow for 4 days before being treated with 5Â MIC of rifamycin. Bacterial enumeration was performed using CFU plating of cultures on 7H111OADC plates at various time points. The kill curve was plotted by calculating the percent survival.
In vitro stress assay. Log-phase cultures of wild-type H37Rv and vapC12 mutant strains were washed with PBST twice and inoculated at an absorbance of 0.1 in 7H9-enriched media for each stress condition, keeping an untreated control. The survival was plotted by CFU plating at different time points posttreatment for different stress conditions, viz., oxidative (5 mM H 2 O 2 for 6 h) and nitrosative (200 mM DETA-NO for 24 h).
Bone marrow-derived macrophages. Bone marrow-derived macrophages (BMDM) were isolated by culturing bone marrow cells from C57BL6 mice in Dulbecco modified Eagle medium containing 10% fetal bovine serum, 2 mM glutamine, 10% L929-conditioned media, and 10 mg/ml ciprofloxacin for 5 days. Approximately 24 h prior to infection, differentiated BMDM were detached and seeded on a 24well tissue culture plate at 5 Â 10 5 cells/well in the same medium lacking antibiotic. Macrophages were infected with different strains of M. tuberculosis at an MOI of 1 for 4 h at 37°C and 5% carbon dioxide. Extracellular bacteria were removed by three washes with warm PBS. Intracellular bacteria were quantified by lysing the cells with 0.01% Triton X-100 (Sigma, CAS:9002-93-1) at the indicated time points and plating dilutions on 7H11 agar.
RNA sequencing. Log-phase cultures of H37Rv and DvapC12 were washed with PBST twice and inoculated in glycerol and cholesterol media at an absorbance of 0.005. RNA was isolated from the cultures at day 4 using an RNeasy minikit according to the manufacturer's protocols (Qiagen, catalog no. 74104). The RNA was DNase treated using a Turbo DNA free kit according to the manufacturer's protocol (Thermo Fisher Scientific) to remove any genomic DNA contamination. All mycobacterial total RNAs were analyzed using an Agilent Bioanalyzer (Agilent, Santa Clara, CA) for quality assessment with an RNA integrity number (RIN) range of 5.6 to 9.7 and a median of 7.5. rRNA was depleted from 500 ng of bacterial RNA using a RiboMinus bacterial transcriptome isolation kit (Invitrogen/Thermo Fisher Scientific, Waltham, MA) according to the manufacturer's protocol. cDNA libraries were prepared from the resultant rRNA-depleted RNA and 1 ml of a 1:500 dilution of ERCC RNA Spike in Controls (Ambion/ Thermo Fisher Scientific) using Lexogen SENSE total RNA-seq Library Prep kit (Lexogen GmnH, Vienna, Austria) according to the manufacturer's protocol, except with 21 PCR cycles. The length distribution of the cDNA libraries was monitored using a DNA high-sensitivity reagent kit on a Perkin-Elmer Labchip (Perkin-Elmer, Waltham, MA). All samples were subjected to an indexed paired-end sequencing run of 2 Â 51 cycles on an Illumina HiSeq 2000 system (Illumina, San Diego, CA; 16 samples/lane). Raw reads (FASTQ files) were mapped to the M. tuberculosis H37Rv (GenBank accession number AL123456) using bowtie2 with the default parameters. The gene counts were then counted using featureCounts (part of the Subread package) using the genome annotations provided in the GenBank file. The gene counts were then used in DESeq2 for differential gene expression analysis. Multiple testing correction was performed using the method of Benjamini and Hochberg. P values of ,0.05 were deemed statistically significant. Computations were done using the R statistical language, v3.3.1.
Quantitative RT-PCR. Comparative qRT-PCR was performed on RNA isolated from various strains under different experimental conditions. RNA isolation was done from culture using an RNeasy minikit according to the manufacturer's protocols (Qiagen, 74104). The RNA was DNase treated using a Turbo DNA-free kit according to the manufacturer's protocol (Thermo Fisher Scientific) for making cDNA with an Accuscript Hi-Fidelity cDNA synthesis kit (Agilent). The qRT-PCR was set up using Brilliant III Ultra-Fast SYBRgreen qPCR master mix in an Mx3005P qPCR system (Agilent). The primers used are listed in Table 4. The data analysis was done using MxPro software.
VapC12 expression and protein purification. vapC12 (Rv1720c) was cloned in pET28a vector using the primer sequences listed in Table 4. The SDM mutants D 5 A and D 94 A were generated by DpnI treatment. The clones were transformed in E. coli Rosetta cells. Briefly, the overnight culture was inoculated into 1 liter of fresh LB media (1:100) supplemented with 100 mg/ml of kanamycin and allowed to grow until the OD 600 reached ;0.5. The culture was then induced with 1 mM IPTG (isopropyl-b-D-thiogalactopyranoside) and allowed to grow overnight at 37°C. Cells were harvested by centrifugation at 6,000 Â g for 10 min and checked for expression of wild-type or mutant rRv1720c by SDS-PAGE. Most of the target protein was present in the pellet as inclusion bodies (IBs). Isolation of pure IBs containing rRv1720c was performed by sonication and several washing steps. Briefly, purified rRv1720c IBs (1 ml) were solubilized in 9 ml of buffer (50 mM Tris-HCl [pH 8.0], 300 mM NaCl, 10 mM b-mercaptoethanol, 8 M urea) and incubated at room temperature for 1 h, followed by centrifugation at 15,000 Â g for 20 min at 10°C. The supernatant obtained after centrifugation was used for the purification of recombinant Rv1720c protein by immobilized metal ion affinity chromatography using a HisTrap FF column (GE Healthcare Buckinghamshire, UK) under denaturing conditions. Protein was eluted using buffer (  In vitro transcription of tRNA. The tRNAs were transcribed using a MegaScript kit (Invitrogen, AM1334) according to the manufacturer's protocol, which was followed by phenol-chloroform extraction and isopropanol precipitation for purified transcript. The primers used for in vitro transcription are listed in Table 4.
In vitro tRNA cleavage assay. In vitro RNA cleavage assay was performed using tRNAs produced via T7 transcription. Cleavage reactions using 3 pmol of tRNA were incubated with 30 pmol of recombinant proteins (rRv1720c, rRv1720cD 5 A, and rRv1720cD 94 A) and at 37°C for 3 h in tRNA cleavage buffer (10 mM HEPES [pH 7.5], 15 mM potassium chloride, 3 mM magnesium chloride, 10% glycerol). The samples were mixed with 2Â formamide gel loading buffer (95% [wt/vol] formamide, 50 mM EDTA) and incubated at 95°C for 5 min before running on a 3% agarose gel. The bands were visualized by using ethidium bromide and exposing the gel to UV light.
Persister enrichment assay. Log-phase cultures of wild-type H37Rv and DvapC12 strains (Table 5) were washed with PBST, and an OD of 0.005 was set in 7H9-enriched media, glycerol media, and cholesterol media. The cultures were washed after 4 days, and fresh cholesterol medium was added to all the tubes. Bacterial enumeration was done by CFU plating on 7H111OADC plates at different time points. A similar protocol was followed for M. bovis BCG strains.
Spent medium preparation. Log-phase cultures of H37Rv and DvapC12 were washed with PBST and inoculated at an OD of 0.005 in cholesterol media. The cultures were then grown for a week at 37°C in an incubator shaker. The cultures were pelleted down, and the supernatant was filtered through a 0.2-mm filter. The spent cholesterol supernatants of both strains were used for further experiments. To prepare heat-inactivated supernatant, the spent medium was exposed to 95°C for 30 PE-PGRS/RpfA blots. All selected PE-PGRS genes, along with the rpfA (with high proT and high proY codon usage), were cloned in pMV261 using primers listed in Table 4. The proteins were translationally fused with His tag at the N-terminal by incorporating bases encoding His tag in the forward cloning primers. The clones were transformed in BCG, and the cultures were maintained in 7H9-enriched media. The log-phase cultures of the constructs were washed twice with PBST and inoculated at an OD of 0.005 in glycerol and cholesterol media. The cultures were allowed to grow until reaching an OD of 0.8 and then pelleted, washed, and lysed in PBS. The samples were run on a 10% SDS-PAGE gel and developed with anti-His antibody (Biospecs, BTL1010).
In vivo animal experiments. The animal experiments were approved by the animal ethics committee of ICGEB (approval IAEC/THSTI/2015-1). The animal experiments were performed in accordance with the guidelines of the Committee for Purpose of Control and Supervision of Experiments on Animals (CPCSEA, Government of India). Three-to four-week-old female Dunkin-Hartley guinea pigs (200 to 300 g) were used to check the pathogenicities of various M. tuberculosis strains. The guinea pigs were infected with various M. tuberculosis strains via the aerosol route with 100 bacilli isolated from the logphase culture (OD = 0.8 to 1). The bacterial load was estimated by CFU plating at different time points postinfection. Briefly, animals were sacrificed at day 1, week 4, and week 7 postinfection. Infected lungs and spleens were harvested and homogenized using a tissue homogenizer. After serial dilution, these tissue lysates were plated on 7H111OADC plates. For histopathological analysis, lung sections were fixed in 10% formalin and stained with hematoxylin and eosin. The tissue samples were coded and evaluated for the granulomatous organization by a pathologist who had no prior knowledge of the samples. All granulomas in each section were scored, and the scores were added to obtain a total granuloma score. In addition, sections were semiquantitatively assessed for the percentage of the section occupied by granuloma, which was expressed as the granuloma fraction.
Cytokine profiling and RNA isolation from the spleen. Single-cell suspension of splenocytes was prepared by passing the spleen through a cell strainer (0.45 mm). The back of the syringe plunger was used to macerate the cells through the filter. The cell pellet was incubated with the RBC lysis buffer for 4 to 5 min after, and then the pellet was used to extract RNA. RNA was isolated from both uninfected and infected spleens harvested from guinea pigs at 7 weeks postinfection using an RNeasy minikit according to the manufacturer's protocol. cDNA synthesis and qRT-PCR were performed as described previously using primers listed in Table 4.
Lysine SDM generation. The SDM vapB12 mutant, wherein K 19 (lysine) was mutated to A 19 (alanine), was generated by DpnI enzyme treatment. The vapBC12His pMV261kan plasmid was used as a template for PCR amplification using SDM primers listed in Table 4. The PCR product was PCR purified and treated with DpnI enzyme for 4 h at 37°C. No-template and no-DpnI-treatment control samples were also obtained. All the reactions were transformed in E. coli XL-1 Blue-competent cells. The construct was sequenced and electroporated in M. bovis BCG-competent cells.
Immunoprecipitation. BCG:vapBC12 His-tagged pMV261 and BCG:vapB K19A C12 His-tagged cultures were maintained and grown in 7H9-enriched media, and log-phase cultures of the strains were washed twice with PBST and resuspended in glycerol and cholesterol media at an OD of 0.1. After 48 h, the cultures were pelleted and lysed in 1Â PBS. The protein quantification of the lysates was done using a Pierce BCA protein assay kit (Thermo Fisher Scientific) according to the manufacturer's protocol. A 1-mg portion of the lysate was incubated with a 1:200 dilution of mouse anti-His antibody (Biospecs, BTL1010), followed by incubation overnight at 4°C on a rocker. The Ag-Ab complex was incubated with protein G-agarose beads and kept at 4°C on a rocker for 8 to 10 h. After the incubation, the supernatant was collected and beads were washed three times with 1Â TBST. The final elution was done with 100 mM glycine (pH 2.0), which was neutralized later with Tris-Cl (pH 8.8). The eluted samples were run on the gel and blotted with rabbit anti-His (Santa Cruz, H-15:sc-803) and rabbit anti-acetylated lysine antibody (Cell Signaling, 9441L).
Sample processing protocol for mass spectrometry. BCG strain overexpressing the toxin-antitoxin (vapBC12) locus where antitoxin is His tagged was used for sample preparation. A log-phase culture of the overexpressed strain was washed with PBS and inoculated into minimal medium with 0.1% glycerol and minimal medium with 0.01% cholesterol. The cultures were allowed to grow for 48 h, and cell lysate was prepared. Immunoprecipitation was performed with an anti-His antibody (BTL1010).
In-solution digestion was carried out using 10 mg of the proteins from each condition. The samples were subjected to reduction and alkylation using 5 mM dithiothreitol (60°C for 45 min) and alkylation using 10 mM iodoacetamide. Trypsin (Gold Mass-Spectrometry Trypsin; Promega, Madison, WI) digestion was carried out at 37°C for 10 to 12 h. The peptides were vacuum-dried and stored at 280°C until LC-MS/MS analysis.
LC-MS/MS analysis. All fractions were evaluated by using a 5600 Triple-TOF mass spectrometer that was directly linked to a reversed-phase high-pressure liquid chromatography Ekspert-nanoLC 415 system (Eksigent, Dublin, CA). Formic acid (0.1%) in water was used as mobile phase A, and mobile phase B was 0.1% formic acid in acetonitrile. All fractions were eluted from the analytical column at a flow rate of 250 nl/min using an initial gradient elution of 10% B from 0 to 5 min, transitioned to 40% over 120 min, ramping up to 90% B for 5 min, holding 90% B for 10 min, followed by reequilibration of 5% B at 10 min with a total run time of 150 min. Peptides were injected into the mass spectrometer using a 10-mm SilicaTip electrospray PicoTip emitter. Mass spectra (MS) and tandem mass spectra (MS/MS) were recorded in positive-ion and high-sensitivity mode with a resolution of ;35,000 full-width half-maximum. Before testing samples on the mass spectrometer, calibration of spectra occurred after the acquisition of every sample using dynamic LC-MS and MS/MS acquisitions of 100 fmol of b-galactosidase. The