The unique N-terminal region of Mycobacterium tuberculosis sigma factor A plays a dominant role in the essential function of this protein

SigA (σA) is an essential protein and the primary sigma factor in Mycobacterium tuberculosis (Mtb). However, due to the absence of genetic tools, our understanding of the role and regulation of σA activity and its molecular attributes that help modulate Mtb survival is scant. Here, we generated a conditional gene replacement of σA in Mtb and showed that its depletion results in a severe survival defect in vitro, ex vivo, and in vivo in a murine infection model. Our RNA-seq analysis suggests that σA either directly or indirectly regulates ∼57% of the Mtb transcriptome, including ∼28% of essential genes. Surprisingly, we note that despite having ∼64% similarity with σA, overexpression of the primary-like σ factor SigB (σB) fails to compensate for the absence of σA, suggesting minimal functional redundancy. RNA-seq analysis of the Mtb σB deletion mutant revealed that 433 genes are regulated by σB, of which 283 overlap with the σA transcriptome. Additionally, surface plasmon resonance, in vitro transcription, and functional complementation experiments reveal that σA residues between 132-179 that are disordered and missing from all experimentally determined σA-RNAP structural models are imperative for σA function. Moreover, phosphorylation of σA in the intrinsically disordered N-terminal region plays a regulatory role in modulating its activity. Collectively, these observations and analysis provide a rationale for the centrality of σA for the survival and pathogenicity of this bacillus.

Mycobacterium tuberculosis (Mtb) enters the body through nasal airways and deposits in the lower lungs, wherein the primarily alveolar macrophages engulf the bacteria (1). In response to the host's dynamic microenvironment, Mtb remodels its gene expression to facilitate its survival within the host. Gaining insights into how Mtb modulates its transcriptional machinery to survive under hostile host conditions is imperative for tackling this deadly pathogen. Gene expression is primarily regulated at the transcription initiation step in bacteria. Transcription initiation involves many diverse molecular interactions that allow the apo RNA polymerase (RNAP) to recognize the promoter and facilitate DNA unwinding around the transcription start-point. Contrary to eukaryotes, wherein three RNAP are present, bacteria encode only one RNAP consisting of 2 subunits of α, one subunit each of β, β 0 , and a ω subunit (2). Even though core RNAP is sufficient for transcriptional elongation, it cannot initiate transcription without a σ factor (3). σ factors play a crucial role in promoter recognition and initiating the melting process of promoter regions (4)(5)(6)(7).
Mtb encodes for one essential and twelve nonessential σ factors, classified into four groups based on their domain architecture (8,9) (Fig. S1). Group 4 mainly comprises ECF (extra cytoplasmic σ factor) σ C to σ M except for σ F . σ F and σ B are the sole members of groups 3 and 2, respectively (9)(10)(11). Essential σ factor, σ A , belong to group 1, containing four α-helical domains. In addition, it harbors a long N-terminal extension (Fig. S1). Studies have previously shown that N-terminal (1.1) region in Escherichia coli σ 70 (RpoD) inhibits its binding to promoters in the absence of RNAP (12). However, there is little sequence similarity between RpoD and Mycobacterium smegmatis σ A (σ A Msm ), especially in the N-terminal region. Furthermore, the N-terminal region of σ A Msm is predicted to be intrinsically disordered (13). The overexpression of sigA enhances bacterial survivability inside macrophages and mice lungs. The expression of anti-sense sigA reduces the bacillary load (14).
The highly conserved amino acid residues in the subdomains 2.4 and 4.2 of σ A directly interact with −10 and −35 regions of the promoter (13). In the case of ECF σ factors, the cognate anti-σ factors sequester these initiation factors, thus regulating their activity (15). However, the regulation of σ A has not been investigated yet. Besides σ factors, mycobacterial transcription is modulated by transcription factors such as RbpA, CarD, and WhiB proteins, specific to actinomycetes (16)(17)(18)(19)(20)(21). These transcription factors interact with σ A to regulate transcription, and disrupting these interactions slow down the transcription process (16,21,22). The presence of RbpA increases the affinity of σ A for RNAP caused by the interaction of σ A and RbpA. Hubin et al. predicted the interacting amino acid residues of σ A Msm from the structure of the RbpA-σ A Msm -DNA complex (13). It was also shown that interaction between σ A and DevR enables bacteria to survive hypoxia (23). The C-terminal domain of σ A interacts with various transcriptional factors such as WhiB. Furthermore, σ A -mediated activation of the eis gene (Rv2416) results in enhanced intracellular growth of the W-beijing strain of mycobacteria (24).
While the σ A expression is constitutive, the expression of σ B is upregulated during stationary conditions (25,26). However, during exponential growth conditions, σ B expression levels are comparable with σ A (27). σ A and σ B share significant sequence similarity (64%) identity between the structured DNA recognition regions of σ A and σ B , including −10 and −35 region-binding amino acid residues. Deletion of sigB renders the pathogen sensitive to heat, oxidative, surface stress, and antibiotics (28). However, it does not impact the survival of Mtb in macrophages or mouse lungs (1,9,26,28). Chip Seq data suggested that several promoters are shared by σ A and σ B (27). σ A and σ B are both phosphorylated (29,30); however, the importance of this modification in regulating their activity remains elusive.
Collectively, our knowledge regarding the structurefunction and regulatory aspects of σ A in Mtb is poor. Here, we set out to answer a few critical questions related to the functionality of σ A in Mtb and its functional correlation with σ B . We sought to answer the following questions: (i) What is the effect of depleting σ A on the survival of Mtb within the host? (ii) What is the effect of σ A depletion and σ B KO on the global transcriptome of Mtb and functional redundancy of σ A and σ B ?, (iii) Can overexpression of σ B rescue phenotypic defects observed upon σ A depletion?, (iv) The role of the unique N-terminal extension of σ A ? (v) How does phosphorylation modulate σ A function? Here, we describe results from experiments designed to address these aspects of σ A -mediated modulation of the expression profile in Mtb. A highlight is the intriguing finding that a disordered extended N-terminal polypeptide, hitherto 'unseen' in structural descriptions of the Mtb RNAP complex, plays a dominant role in σ A activity.
Results σ A is essential for Mtb survival σ A is a 528 amino acid essential protein with an extended 207 amino acid-long disordered N terminal region preceding the four α helical domains (Fig. 1A). Domain 2 interacts with the −10-promoter element, domain 3 with the extended −10 promoter region, and domain 4 interacts with the −35 region (Fig. S1). The function of domain 1, consisting of nonconserved regions and the extended N-terminal region, has not yet been elucidated (Figs. 1A and S1). To investigate structure-function relationships and the in vitro and in vivo role of σ A , we sought to generate a conditional gene replacement mutant (RvΔsigA). sigA was cloned into pFICTO, and the construct was electroporated into Mtb-H37Rv (Rv) to generate merodiploid strain (Rv::sigA) (Fig. 1B). The expression of FLAG-σ A from integrative copy is under anhydrotetracycline (ATc) regulation. In the absence of ATc, the P smyc promoter (31) is active, and upon the addition of ATc, the r-TetR-ATc complex binds to the Tet operator sequences in the promoter, shutting down transcription. Western Blot analysis confirmed that the expression of FLAG-σ A from the merodiploid is significantly lowered upon the addition of ATc (Fig. 1B). Subsequently, sigA at the native loci was replaced with a hyg r -sacB cassette with the help of a specialized transduction methodology to generate RvΔsigA (32). The fidelity of recombination at the native loci was confirmed by PCR and Western blot analysis (Fig. 1, C and D). Analysis of FLAG-σ A levels at different days post-ATc addition revealed a significant reduction from day 3 (Fig. 1E). In vitro growth analysis suggested that while the growth of RvΔsigA was comparable to Rv in the absence of ATc, depletion of FLAG-σ A resulted in compromised growth, eventually resulting in a 4-log fold (10,000-fold) difference between the growth of parental and mutant strain (Fig. 1F). Importantly, depletion of FLAG-σ A (RvΔsigA +ATc) significantly reduced the intracellular survival in the peritoneal macrophages as compared to either Rv or RvΔsigA-ATc-infected cells (Fig. 1G). Together, these results confirm the generation of conditional RvΔsigA mutant and the criticality of σ A for both in vitro and ex vivo survival of the pathogen.
σ A is essential for Mtb survival in vivo We used a murine infection model to evaluate in vivo survival of the RvΔsigA mutant. Mice infected with Rv or RvΔsigA through the aerosolic route and the colony forming units (CFUs) enumerated 24 h post infection (p.i) indicated efficient and equivalent implantation of WT and mutant bacilli in the lungs of mice (Fig. 2, A and B). To examine the impact of depleting σ A , a set of animals infected with RvΔsigA were provided doxycycline (Dox) in the drinking water ( Fig. 2A). The data suggest that depletion of σ A from the start of infection results in a significant reduction (2.5-3.5 log 10 fold) in the bacillary load (RvΔsigA + Dox) at 4 weeks p. i compared with either Rv + Dox or RvΔsigA -Dox in both lungs and spleen of the infected animals ( Fig. 2, B and C). We sought to examine the impact of depleting σ A from an established infection. Mice were infected with Rv or RvΔsigA, and the infection was allowed to be established for 2 weeks (Fig. 2D). Subsequently, RvΔsigA-infected animals were divided into four groups; wherein two groups were given Dox in the water for the subsequent weeks, and the CFUs were enumerated (Fig. 2D). Compared with Rv + Dox or RvΔsigA -Dox, a decrease of 1.5 and 2.5 log 10 fold was observed in the lungs of RvΔsigA +Dox mice at 6-and 10-weeks p. i, respectively (Fig. 2E). A similar trend was observed in the mice spleens (Fig. 2F). Together, this data suggests that while σ A is critical for the survival of Mtb within the host, the impact of depleting σ A is less significant during the chronic stage of infection.
σ A depletion impacts global transcription in Mtb σ A is proposed to regulate majority of the mycobacterial transcriptome under normal growth conditions. To identify Figure 1. σ A is essential for Mtb survival. A, schematic overview of the domain architecture of Mtb SigA. B, schematic depicting homologous recombination between the Allelic Exchange Substrate (AES) and sigA native genomic loci (rv2703). Legitimate recombination was confirmed by performing multiple PCRs. Primer used for the confirmation are depicted. Inlet shows the immunoblot, wherein Rv::sigA merodiploid was grown in the absence or presence of ATc. Whole-cell lysates (WCLs) were resolved on 10% SDS-PAGE, transferred to nitrocellulose membrane, and probed with α-SigA (10,000-fold dilution) 61 and α-GroEL2 (10,000-fold dilution) antibodies generated in house. The band corresponding to endogenous and FLAG-tagged SigA are indicated the regulons and pathways that σ A modulates, we performed RNA-seq of RvΔsigA grown in vitro in the absence and presence of ATc. Towards this, RNA was extracted from RvΔsigA cultured in the absence or presence of ATc for 3 or 4 days. The RNA obtained 4 days post-ATc addition was degraded, and hence we used RNA extracted from RvΔsigA in the absence or presence of ATc on day 3 for the experiment. The experiment was performed in biological triplicates, and the principal component analysis (PCA) showed appropriate clustering of RvΔsigA-ATc and RvΔsigA +ATc samples (Fig. 3A). The data analysis showed that sigA expression was downregulated 0.7 fold in RvΔsigA +ATc samples (padj = 2.79E-08). To identify differentially expressed genes (DEGs), we applied a cut-off as padj <0.05 and absolute log 2 FC >0.5. There were 2320 DEGs, among which 1133 genes were upregulated, and 1187 genes were downregulated upon depletion of σ A (Fig. 3B and Table S1). Heat maps for each biological RvΔsigA − ATc and RvΔsigA + ATc sample show a complete set of upregulated and downregulated genes (Fig. 3C). To validate the RNA-seq analysis results, we selected five upregulated and downregulated genes each and performed a qRT-PCR analysis with RvΔsigA − ATc and RvΔsigA + ATc samples (Fig. 3D). To further validate this observation, Western blot analysis of σ B and GlmU, which are upregulated and downregulated, respectively, in RvΔsigA on different days post-ATc addition were examined (Fig. 3E). These results were consistent with RNA sequence data, wherein we observed an increase in the protein levels of σ B and a decrease in the levels of glmU at the later time points post-ATc addition (Fig. 3E).
Next, we assessed the operonic arrangement of the 2320 protein coding DEGs and found that 39.74% of the total were nonoperonic, while 60.26% were operonic (Fig. 3F). Of the 242 essential genes differentially regulated upon σ A depletion, 57 were nonoperonic and 185 operonic (Fig. 3G). Pathway analysis suggested that many DEGs belong to multiple pathways, including lipid metabolism, intermediary metabolism, and cell wall and cell processes (Fig. 3H). Gene enrichment analysis indicated that critical metabolic pathways such as oxidative phosphorylation (ETC), TCA cycle, nucleic acid biosynthesis, and cell wall synthesis pathways are substantially enriched upon depletion in downregulated genes ( Fig. 3I and Table S2).
Together, it appears that σ A regulates the expression of 1/3rd of the essential genes in Mtb. More importantly, it regulates the expression of many genes involved in cell wall synthesis, oxidative phosphorylation, and TCA cycles. Thus, results suggest that σ A either directly or indirectly regulates 57% of Mtb genes and that encompasses most critical cellular processes. It is thus not surprising that its depletion results in cell death. σ B overexpression fails to rescue the σ A depletion phenotype RNA-seq analysis and Western blot analysis (Fig. 3) suggested that depletion of σ A led to an increase in the expression levels of σ B . σ B is the only group II σ factor in Mtb, with 64% identity with σ A . The significant difference is the absence of the N-terminal extended region in σ B (Fig. 4A). To ascertain whether the higher levels of σ B influence the expression of proteins from σ A -regulated promoters, we performed a reporter assay. The upstream promoter region up to the start codon of sigA (500 bp) was fused with the ORF of luciferase in an integrative shuttle vector. The construct was electroporated into Rv and RvΔsigA (Fig. 4B). The luciferase activity in the lysates was evaluated on different days post ATc in addition to evaluating the impact of σ A depletion on the expression of σ Amodulated promoter. There was a significant reduction in the luciferase activity (compared with -ATc samples) on days 1 and 2 post σ A depletion. However, on the subsequent days, the activity was higher (Fig. 4C). We reasoned that higher expression of σ B is likely responsible for the increased promoter activity upon σ A depletion. This hypothesis aligns with a previous report, which suggested a significant overlap in the promoter utilization between σ A and σ B (27). If this is indeed true, overexpression of σ B should compensate for the absence of σ A . σ B overexpression construct (pNit-F-SigB) was transformed into Rv and RvΔsigA, and the expression of FLAG-σ B was confirmed by Western blotting analysis (Fig. 4D). The overexpression of σ B failed to mitigate the phenotypic defects observed upon depletion of σ A both in vitro and ex vivo (Fig. 4, E and F), suggesting a unique regulatory function for σ A . Thus, we set out to examine the extent of uniqueness/redundancy of these σ factors. To address this question, we generated a mutant of sigB, RvΔsigB, wherein the sigB gene was replaced with a hyg r antibiotic marker (Fig. S2A). The replacement of sigB at the native loci was confirmed by multiple PCRs (Fig. S2B) and Western blot (Fig. 4G). In vitro growth (Fig. 4H) and ex vivo infection ( Fig. 4I) experiments suggested that the presence of σ B was dispensable for Mtb survival. Together, data suggest that the function of σ A is distinctive, and the absence of σ B has no particular impact on growth.
σ B -regulated transcriptome only partially overlaps with σ A Next, we sought to examine the commonality in downstream gene expression between σ A and σ B . RNA-seq analysis was performed in biological duplicates with the samples extracted from the log phase cultures of Rv and RvΔsigB. PCA analysis showed suitable clustering of Rv and RvΔsigB samples (Fig. 5A).Volcano plot represent 433 DEGs using a cut-off of by arrows. C, agarose gel images depicting PCR amplification from genomic DNA isolated from Rv or RvΔsigA with multiple indicated primer sets. Amplification with F1-R1 or F2-R2 primers results in 1.2 kb or 1.6 kb respectively from Rv but none with RvΔsigA, and amplification with F1-R3 or F4-R2 primers results in 1.1 kb or 1.4 kb respectively from RvΔsigA but none with Rv. D, 30 μg of WCLs from Rv or RvΔsigA were resolved on 10% SDS-PAGE, transferred to the membrane, and probed with α-sigA and α-GroEL antibodies. E, Rv or RvΔsigA grown in the absence of ATc were seeded at A 600 0.05, and the cultures were grown in the absence or presence of ATc. WCLs were prepared second day onward post ATc addition. The WCLs were resolved and probed with α-SigA and α-GroEL antibodies. F, Rv or RvΔsigA cultures were seeded at A 600 0.05 in 7H9-ADC, and the growth in vitro was monitored by enumerating CFU every day for 6 days. The data represent mean CFU log 10 /ml ± SD of three independent replicates. ***, p < 0.0001. G, peritoneal macrophage were isolated from BALB/c mice and infected with Rv or RvΔsigA at 1:5 MOI. The cells were washed with RPMI 4 h post infection (p.i) and RPMI with or without ATc. The intracellular bacillary load was enumerated on Day 0, 2, and 4 p.i. Data information: The data represent mean CFU log 10 /ml ± SD of three independent replicates. Statistical significance was drawn in comparison with RvΔsigA using one-way ANOVA (Tukey test; GraphPad prism 9). *, p < 0.01; ***, p < 0.0001. ATc, anhydrotetracycline; CFU, colony forming unit; Mtb, Mycobacterium tuberculosis. i. Each data point indicates CFU log 10 /lung or spleen obtained from one mice, and error bar represents as mean CFU log 10 /ml ± SD. *, p < 0.01; **, p < 0.001; ***, p < 0.0001. D, schematic depicting the outline of experiment. Dox was introduced after establishment of the infection, i.e., at day 14. E and F, BALB/c mice were infected with 100 CFU/mice of Rv or RvΔsigA, and the infection was padj < 0.05 and absolute log 2 FC > 0.5, among which 247 genes were upregulated and the 187 genes were downregulated in RvΔsigB ( Fig. 5B and Table S3). The heat map shows that this trend is maintained for both the biological replicates of Rv or RvΔsigB (Fig. 5C). To validate the data, we performed qRT-PCR analysis of the top five upregulated and downregulated genes. The data obtained is in agreement with RNA-seq trends, even though the data for one sample was statistically insignificant (Fig. 5D). Analysis suggested that 61.9% of genes belong to operons, indicating plausible regulation of their promoters by σ B (Fig. 5E).
Interestingly, among the 417 protein-coding DEGs, only 54 are essential genes (Fig. 5F). While we find genes corresponding to all the cellular processes among DEGs, hypothetical, intermediatory metabolism, and cell wall/synthesis process contribute the most (Fig. 5G). Gene enrichment analysis indicated enrichment of five different processes, including growth, only in upregulated genes ( Fig. 5H and Table S4). Unlike σ A , the absence of σ B did not show enrichment of any cellular processes in downregulated DEGs. Next, we overlapped the DEGs obtained upon sigA depletion with those obtained without sigB. It appears 98 and 28 genes are Red spots indicate the genes that are upregulated and blue spots indicated those are downregulated (padj < 0.05, log 2 fold change = 0.5). C, heat maps showing the normalized read counts of DEGs from three biological replicates of Rv and RvΔsigA. Color intensity indicates relative upregulation (blue) or downregulation (red). D, five each upregulated and downregulated genes from RNA-seq analysis were validated with the help of qRT-PCR. Data information: data were normalized with respect to 16s rRNA and is plotted as mean ± SD (performed in triplicates (n = 3)). Statistical significance was analyzed using Student's t test (two-tailed, unpaired). *, p < 0.01; **, p < 0.001; ***, p < 0.0001. E, Rv and RvΔsigA WCLs were prepared on different days post ATc from cultures grown for 6 days in 7H9 in the presence or absence of 1 μg/ml ATc. WCLs from Rv, RvΔsigA were resolved and probed with α-SigA, α-SigB, and α-glmU. F, the pie chart depicting the percentage of total DEGs that are operonic and nonoperonic. G, bar graph showing the number of operonic and nonoperonic DEGs that belong to essential (E), essential domain (ED), growth defect (GD), growth advantage (GA), non-essential (NE), and uncertain (U) categories. H, bar graphs depicting the numbers of upregulated (purple) and downregulated (blue) DEGs plotted that belong to a various functional categories. I, gene enrichment analysis shows the number of genes in various biological pathways that are upregulated or downregulated upon depletion of SigA (FDR < 0.05). ATc, anhydrotetracycline; Mtb, Mycobacterium tuberculosis; WCL, whole-cell lysate.
established for 14 days. Dox was administrated to Rv and one set of RvΔsigA (n = 5) infected mice, while other set was left untreated. CFU was enumerated on day 1, 2-, 6-, and 10-weeks p.i in the lung (E) and spleen (F) homogenates. Data information: Each data point indicates CFU log 10 /lung or spleen obtained from one mice, and error bar represents the mean CFU log 10 /lung or spleen ± SD. Statistical significance was drawn in comparison with RvΔsigA using oneway ANOVA (Tukey test; GraphPad prism 9). **, p < 0.001; ***, p < 0.0001. CFU, colony forming unit; Mtb, Mycobacterium tuberculosis. Rv and RvΔsigA were electroporated with pSW-luc SApr to generate Rv::luc SApr and RvΔsigA::luc SApr strains. C, luciferase activity was measured from WCLs prepared from cultures on different days post ATc addition. The experiment was performed in triplicates. Data were normalized with respect to 16s rRNA and is plotted as mean ± SD (performed in triplicates (n = 3).Statistical significance was analyzed using Student's t test (two-tailed, unpaired). ***, p < 0.0001. D, Rv and RvΔsigA strains were electroporated with pNiT-3XF and pNiT-3XF-sigB to upregulated or downregulated in the absence of either σ factor. The data in Figure 4 shows upregulation of σ B upon σ A depletion, which may have contributed to upregulated DEGs. Those genes which σ B regulates in the absence of σ A are expected to overlap with downregulated DEGs in the absence of σ B . We observed 91 such genes in the overlap analysis ( Fig. 5I and Table S5). Importantly, 2000 DEGs obtained upon depletion of σ A and 150 DEGs obtained in the absence of σ B were unique. The data thus suggest that σ A and σ B regulate many nonoverlapping genes.

The N-terminal unstructured region is critical for σ A functionality
The data presented in Figures 4 and 5 suggest that σ B cannot complement σ A despite having considerable homology in the four structured promoter recognition domains. As stated earlier, σ A has a long 225 aa and unique N-terminal extension (33). This led us to examine the role of the N-terminal region in modulating the function of σ A . We generated a series of N-terminal truncation mutants, wherein we deleted either 42, 92, 132, or 179 aa from the N-terminus. The mutants were cloned into pNit1 and pET vectors to generate pNiT-sigA Δ42 , pNiT-sigA Δ92 , pNiT-sigA Δ132 , and pNiT-sigA Δ179 and the corresponding pET vector constructs (Fig. 6A). RpbA, CarD, RNAP, σ A , and σ A deletion mutants were expressed and purified from E. coli. CD studies suggested the absence of gross changes in secondary structure among the truncated σ A variants (Fig. S3). σ A is known to bind with the rrnAP3 promoter element (13), and hence, biotinylated-AP3 promoter immobilized on streptavidin (SA) chip was used for determining the binding efficiency using Surface Plasmon Resonance (SPR). We first evaluated the ability of σ A and σ A deletion mutants to interact with the promoter region with the help of SPR. We did not observe statistically significant differences in the binding affinity between WT and deletion mutants of σ A (Figs. 6B and S4). Next, we examined the impact of σ A truncation mutants on their ability to initiate transcription. RNAP holoenzyme, RbpA, and CarD were purified, and in vitro transcription assay was performed in the presence or absence of σ A wt/mut . As expected, in the absence of σ A , the RNAP holoenzyme, RbpA, and CarD could not be recruited to the promoter and hence no product formation was observed (Fig. 6, C-E). Except for σ A Δ179 , σ A and the other σ A deletion mutants could initiate transcription, suggesting that the residues between 132 and 179 are critical for σ A 's ability to initiate the transcription (Fig. 6, D and E). We hypothesized that residues between 132-179 are primarily acidic and may be involved in forming the transcription bubble necessary for transcription initiation. Subsequently, we examined the ability of σ A deletion mutants to complement the function of σ A in vitro and ex vivo. pNiT constructs were electroporated into RvΔsigA, and the expression of σ A or σ A N-terminal deletion constructs was evaluated by Western blot (Fig. 6F). While all the σ A deletion mutants were seen to be expressed, their expression was lower than the full length (Fig. 6F). In line with our in vitro transcription data, all the mutants other than σ A Δ179 complemented the function of σ A in the presence of ATc (Fig. 6G). The results obtained in ex vivo infection experiments were found to be similar to in vitro growth data (Fig. 6H). These results suggest that acidic amino acids between amino acids 132-179 residues play an important role in modulating the activity of σ A .

Phosphorylation impacts σ A function
Phosphorylation of ECF-σ factors by Serine/threonine kinases influences their interaction with anti-σ factors thereby regulating its function (34). High throughput phosphoproteomic studies identified phosphorylation of T69 and T71 residues in the unstructured N-terminal region of σ A (29,30). Depletion of PknA results in decreased phosphorylation of T69 and T71 residues in σ A (29). In addition, S57 in σ A was also identified to be a site for phosphorylation (unpublished data). We sought to examine the impact of σ A phosphorylation on its interactions with the promoter region, in vitro transcription, and its function in Mtb. To evaluate the impact of phosphorylation, we generated σ A phosphoablative mutants σ A Abl (S 57 (Fig. S6). SPR analysis suggested that the binding affinity of σ A Abl mutant is 2 fold higher than σ A (Fig. 7, B and C). The trend was similar to σ A Mim ; however, the values were not statistically significant. Next, we performed in vitro transcription assay, as described earlier to evaluate the impact of σ A phosphorylation on transcription. Compared with the WT σ A , both phosphoablative and mimetic mutants resulted in higher amounts of RNA product (Fig. 7, D-E). To further investigate the effect of SigA phosphorylation, we coexpressed His-SigA and MBP or MBP-PknA or MBP-PknB. We generated a single construct coexpressing both SigA and generate Rv::F-sigB and RvΔsigA::F-sigB. WCLs were prepared from cultures grown for 6 days in 7H9 + 0.2 μM IVN, in the presence or absence of 1 μg/ml ATc. WCLs from Rv, RvΔsigA, and RvΔsigA wt/mut were resolved and probed. E, Rv, Rv::sigB, RvΔsigA, and RvΔSigA::F-sigB cultures grown in 7H9-ADC + 0.2 μM IVN, in the absence or presence of ATc for 6 days and CFUs were enumerated. The data represent mean CFU log 10 /ml ± SD of three independent replicates. F, murine peritoneal macrophage cells were infected with Rv, Rv::sigB, RvΔsigA, and RvΔSigA::F-sigB at 1:5 MOI. 0.2 μM IVN was added to induce the expression of episomally produced SigA, and ATc was added wherein indicated to repress FLAG-SigA expression. CFUs were enumerated at 96 h post infection. The data represent mean CFU log 10 /ml ± SD of three biological replicates (n = 3). G, 30 μg of WCLs from Rv and RvΔsigB were resolved and probed with α-sigB and α-GroEL antibodies. H, Rv and RvΔsigB cultures were inoculated at 0.05 A in 7H9 + 0.2 μM IVN, in the absence or presence of 1.0 μg/ml ATc, and CFUs were enumerated on day 6. The data represent mean CFU log 10 /ml ± SD of three independent replicates. I, murine peritoneal macrophage cells were infected with Rv, Rv::sigB, RvΔsigA, and RvΔSigA::sigB at 1:5 MOI. Post wash, cells were replenished with RPMI + 0.2 μM IVN, ± 1.0 μg/ml ATc, and CFUs were enumerated at 96 h p.i. Data information: the data represent mean CFU log 10 /ml ± SD of three independent replicates. Statistical significance analyzed using one-way ANOVA (Tukey test; GraphPad prism 9). ***, p < 0.0001. ATc, anhydrotetracycline; CFU, colony forming unit; IVN, isovaleronitrile; WCL, wholecell lysate.
Functionality of SigA is dependent on its N-terminal region   Figure S4. C, schematic representation of in vitro transcription assay. 250 nt template DNA-harboring rrnAP3 promoter was mixed with holo RNA polymerase, RbpA, CarD, NTPs, α[ 32 P] UTP, and SigA/SigA Mut . D, in vitro transcription assay performed in the presence or absence of either SigA or SigA Mut were resolved on UREA-PAGE. Representative image of 150 nt-radiolabeled RNA product captured using phosphorimager. E, quantification of three biological replicates. F, RvΔsigA strain kinase with a Histidine and a MBP tag respectively. E. coli was used as a surrogate host to obtain phosphorylated SigA. It is apparent from the Western blot presented in Figure 7F that His-SigA, MBP, MBP-PknA, and MBP-PknB expressed robustly in the lysates (Fig. 7F). Subsequently, we validated the phosphorylation of SigA by PknA and PknB by probing the purified SigA with phosphor-threonine antibodies (Fig. 7F). We note that both PknA and PknB phosphorylate SigA efficiently in vivo. To evaluate the phosphorylation status of SigA, His-SigA was pulled down and probed with α-p-Thr and α-SigA antibodies. α-p-Thr blots indicate that both PknA and PknB phosphorylate SigA (Fig. 7F). In an in vitro transcription assay with phosphorylated SigA, we found that phosphorylated SigA (SigA-MBP-PknA/B) does not show significant change in transcript formation in comparison to unphosphorylated SigA (SigA-MBP) (Fig. 7G).
Subsequently, we subcloned these genes into the pNit-1 vector; the recombinant Mtb strains were evaluated for their ability to complement the depletion phenotype. The expression levels of σ A wt/mut were found to be comparable with endogenous σ A expression (Fig. 7H). While the WT almost wholly rescued the phenotype, both σ A Abl and SigA Mim partially rescued the growth defects in vitro and ex vivo (Fig. 7, I and J). The inability of σ A Abl's to complement may be due to the possible interactions of the hydroxyl group with the initiator complex. On the other hand, the failure of σ A Mim to completely complement may be due to the regulatory role of phosphorylation in the initiator complex. Together, these results suggest a partial role for phosphorylation in regulating the σ A -mediated transcription.

Discussion
The response of Mtb to various host-derived stresses in vivo is coordinated by many regulators such as secretory virulence factors, one-and two-component signaling modules, serine/ threonine protein kinases, and a multitude of transcription factors. Besides protecting Mtb from host immune response, these highly effective stress response strategies also contribute to antibiotic tolerance. Transcription factors, including σ factors, CarD, RbpA, WhiB family protein, and others, contribute to the regulation at the transcriptional level. σ factors are necessary for the recruitment of RNAP to the promoter. Thus, their levels in the cytosol impact the polymerase-promoter interaction and subsequent transcription (8,9). While there is considerable literature on the functions of nonessential σ factors, insights into structure-function features and regulatory aspects of the essential σ factor σ A in vivo remained elusive. This lacuna is in large part due to the nonavailability of genetic tools such as a conditional gene replacement mutant. This report sought to investigate many unanswered questions related to σ A in Mtb by generating a tetracycline-regulatable conditional gene replacement mutant (Fig. 1).
Gene replacement mutants provide valuable information regarding the functional roles as well as the in vitro, ex vivo, and in vivo essentiality of a gene. High-throughput transposon mutagenesis studies suggested that except sigA, the remaining twelve σ factors are nonessential. Of these, except for sigK, gene replacement mutants were generated for the remaining eleven nonessential σ factors (28,(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). Even though these σ factors are nonessential for growth in vitro, deletion of sigC, sigD, sigE, sigF, sigH, and sigL leads to attenuated growth in murine or guinea pig models of infection (35-38, 41, 43, 46, 47). SigB and sigJ are notable omissions, where their deletion had no impact on the growth in vivo (28,42). Here, we created a conditional gene replacement mutant of sigA to determine its role. Results presented in Figures 1 and 2 show that sigA is essential for growth in vitro, ex vivo, and in vivo. This agrees with a previous report, wherein an anti-sense sigA RNA was used to downregulate its expression; the strain showed decreased growth in both macrophage and murine models (14). However, these experiments were performed for only 20 days in murine models. Importantly, conditional gene replacement generated in our study allowed us to evaluate the essentiality of sigA in an established infection (Fig. 2). Our results suggest that depletion of σ A even from an established infection leads to mycobacterial death (Fig. 2). However, it has not escaped our attention that there were no further changes in the mycobacterial load between the sixth and 10th week despite continued depletion of sigA. We speculate that other stress-induced σ factors may partially compensate for the decreased levels of σ A in chronic stages of infection in murine lungs. The same approach has been used previously to investigate the effect of GlmU depletion from fully infected lungs, wherein we observed complete clearance of pathogens (48). As compared to GlmU depletion, σ A depletion from established infection produced a weaker phenotype. Based on this observation, we concluded that σ A depletion may not be as crucial for pathogen survival during chronic infection.
Examining global transcriptional changes in the absence of transcription factors provides information on the subset of genes regulated by the factor. Depleting sigA for 4 days resulted in compromised RNA quality, and depletion for 2 days was insufficient (data not shown). Nevertheless, we compared analyzed RNA-sequ data from second depleted cells with third depleted cells. The large proportion of upregulation or downregulation DEGs, the difference became more pronounced when cells progressed from second to third depletion was electroporated with pNit-sigA, pNit-sigA Δ42 , pNit-sigA Δ92 , pNit-sigA Δ132 , and pNit-sigA Δ179 to generate RvΔsigA::sigA, RvΔsigA::sigA Δ42 , RvΔsigA::sigA Δ92 , RvΔsigA::sigA Δ132 , and RvΔsigA::sigA Δ179 . WCLs were prepared from cultures inoculated at A 0.05 and grown for 6 days in the IVN and presence or absence of ATc. 30 μg of WCLs from Rv, RvΔsigA, and RvΔsigA wt/mut were resolved and probed with α-sigA and α-GroEL antibodies. G, Rv, RvΔsigA, and RvΔsigA wt/mut cultures inoculated A 0.05 in 7H9-ADC + 0.2 mM IVN, in the absence or presence of 1 μg/ml ATc, and CFUs were enumerated on day 6. The data represent mean CFU log 10 /ml ± SD of three independent replicates. H, murine peritoneal macrophage cells were infected with Rv, RvΔsigA, and RvΔsigA wt/mut at 1:5 MOI. 4 h p.i, cells were washed and replenished with RPMI + 0.2 μM IVN media with or without 1 μg/ml ATc, and CFUs were enumerated at 96 h p.i. Data information: the data represent mean CFU log 10 /ml ± SD of three independent replicates. Statistical significance was analyzed using one-way ANOVA (Tukey test; GraphPad prism 9). p < 0.01; **, p < 0.001; ***, p < 0.0001. ATc, anhydrotetracycline; CFU, colony forming unit; IVN, isovaleronitrile; SPR, surface plasmon resonance; WCL, whole-cell lysate.
Functionality of SigA is dependent on its N-terminal region Figure 7. Phosphorylation impacts σ A function. A, line diagram depicting the phosphorylation sites (marked in red) in the N-terminal region of of SigA. S57, T69, and T71 were mutated to A57, A69, and A71 to generate σ A ablative mutant (σ A Abl ) and to D57, E69, and E71, to generate σ A mimetic mutant (σ A  Mim ). B, the SPR sensorgram of rrnAP3 promoter DNA and σ A /SigA Mut interactions. C, binding affinity of various SigA mutants with rrnAP3 promoter DNA derived from the SPR sensorgrams in (B). SPR experiments using σ A -truncated variants and phosphomutants were performed in quadruplets at the same time. WT σ A σ A -rrnAP3 promoter interaction control is the same for both sets (Figs. 6B and 7C). D, in vitro transcriptions assay as described above. The (data not shown). RNA-seq analysis of 3-day-depleted RvΔsigA samples showed differential expression of 2320 genes (Fig. 3). These include genes involved in critical metabolic pathways, nucleic acid biosynthesis, DNA replication, lipid, and protein synthesis. Energy metabolism generates ATP, which fuels cellular processes and thus represents one of the most critical pathways in an organism. In mycobacteria, 10 NADH dehydrogenases (nuo A to J) convert nicotinamide NADH to NAD + (49), and results show that all 10 genes are downregulated upon σ A depletion. Significantly, 30% of all the essential genes are differentially regulated upon σ A depletion, which is likely to be the reason for the cell death. We also observed upregulation of σ B in σ A -depleted samples both at RNA and protein levels (Fig. 3), suggesting that mycobacteria may be compensating for the absence of σ A by upregulating the expression of σ B .
Overexpression of σ B upregulates transcription of PE-PGRS proteins, culture filtrate antigens, ribosomal proteins, the ketoacyl synthase, KasA, and the regulatory proteins WhiB2 and IdeR (50). However, the expression of other σ factor genes is unaltered, suggesting that σ B may serve as a terminal modulator of the σ factor regulatory cascade (50). Chip Seq experiments revealed that in M. smegmatis, 72 out of 200 promoters detected are shared between both σ A and σ B (45). The expression levels of σ B at both RNA and protein levels are higher upon the depletion of σ A (Fig. 3). Despite having considerable overlap at amino acid levels with σ A (64%), σ B failed to rescue the sigA depletion phenotype (Fig. 4). In a previous study, RNA-seq analysis performed with M. smegmatis σ B mutant showed 125 DEGs with 100 genes upregulated and 25 genes downregulated (27). We find that in Mtb, 433 genes are DEGs upon σ B deletion. The differences in the number of DEGs between the two studies are likely due to (a) differences in the strains used for the studies and (b) differences in the cut-offs applied. Among the 433 DEGs obtained upon deletion of Mtb σ B , 283 genes overlapped with σ A DEGs. Among the 1168 downregulation genes σ A regulon, only 28 genes seem to be dependent on higher expression of σ B (Fig. 5). The striking difference between these σ factors is the absence of the N-terminal region in σ B . We made attempts to generate a chimera wherein the N-terminal domain of σ A was fused to σ B . However, we could not detect the expression of the chimeric protein suggesting that fusion protein may be unstable (data not shown).
Orthologs of σ A possesses varying lengths of N-terminal extension. These range from 83 aa in Bacillus subtilis, 95 in E. coli, and 163 in M. smegmatis (Fig. S7). However, the N-terminal polypeptide stretch observed in σ A is significantly longer at 221 aa. Domain σ 1.1 (σR.1.1) is involved in abrogating nonproductive interactions of free σ A with the promoter (51,52). Upon the formation of holoenzyme, domain σ 1.1. occupies RNAP active site channel as a DNA mimic (6,53). In σ A , the 1.1 region is not well defined. Results suggest that deletion of amino acids between 132-179 abrogates σ A functionality (Fig. 6). We detected high levels of negatively charged amino acid residues in regions beyond 132. Based on the data, we speculate that the amino acids beyond 132 may be playing a role akin to σ 1.1 .
Serine/threonine-protein kinase-mediated phosphorylation regulates many cellular processes in mycobacteria, including transcription. An example of immediate relevance is the ECF σ factors σ H and its cognate anti-σ factor RshA. Both are targets of PknB, and the phosphorylation of RshA interferes with σ H -RshA interactions (54,55). PknD-mediated phosphorylation of Rv0516c inhibits its interaction with the anti-anti-σ factor and eventually influences the genes regulated by σ F (56). In the case of the primary σ factor σ A , high throughput phosphoproteomic studies led to the identification of residues T69 and T71 as sites for phosphorylation. Both sites lie in the intrinsically disordered N-terminal stretch in σ A . Subsequently, we identified an additional phosphorylation on S57 in a high throughput phosphoproteomic analysis performed recently (unpublished data). Does phosphorylation alter promoter recognition or transcription initiation? This was clearly not the case as absence of phosphorylation or σ A mutants mimicking phosphorylation had no impact on either DNA binding or in vitro transcription (Fig. 4). Intriguingly, neither mutant could completely restore the function of σ A in vitro growth nor ex vivo (Fig. 4). Despite phosphorylation playing a partial role in ex vivo survival, we did not observe any significant differences in in vitro transcription initiation when we used the phosphorylated form of σ A (Fig. 7H). According to these results, the phosphorylation of σ A may modulate the nature and strength of σ A interactions with other proteins, thus regulating promoter binding.
Together, these studies suggest that the long N-terminal intrinsically disordered polypeptide stretch in σ A performs a distinct functional and regulatory role in this primary σ factor. The primary-like σ B lacks this stretch and cannot functionally replace σ A . The distinctive regulons of these two σ factors also highlight the functional differences. These observations, alongside the finding that depleting sigA in an established Mtb infection leads to mycobacterial death, rationalizes the critical role of this essential initiation factor in the survival and radiolabeled 150 nt mRNA product was resolved on Urea-PAGE and visualized by phosphor scanner and quantified using Quantity One. Figure provided is representative of one of three replicates. E, quantification of three biological replicate data from (D). F, pDuet constructs with MBP, MBP-pknA, or MBP-pknB cloned in MCS2 with or without sigA cloned in MCS1 were transformed in BL21 (DE3) codon plus cells. Lysates and His pulldowns were probed with anti-Thr(P), antiHis, and anti-MBP antibodies. G, quantification of three biological replicate data from above described in vitro transcription assay using phospho-SigA. H, RvΔsigA strain was electroporated with pNit-sigA, pNit-sigA abl , and pNit-sigA Mim to generate RvΔsigA:sigA, RvΔsigA::sigA abl , and RvΔsigA::sigA Mim . WCLs were prepared from cultures inoculated at A 600 of 0.05 and grown for 6 days in the 7H9 media containing 0.2 μM IVN in the presence or absence of 1 μg/ml ATc. 30 μg of WCLs from Rv, RvΔsigA, and RvΔsigA wt/mut were resolved on 10% SDS-PAGE and probed with α-sigA and α-GroEL antibodies. I, Rv, RvΔsigA, and RvΔSigA wt/mut cultures were seeded at A 0.05 in 7H9-ADC + 0.2 mM IVN, in the absence or presence of ATc, and CFUs were enumerated on day 6. The data represent mean CFU log 10 /ml ± SD of three independent replicates. J, murine peritoneal macrophage cells were infected with Rv, RvΔsigA, and RvΔsigA wt/mut as described in Figure 3, and CFUs were enumerated at 96 h p.i. Data information: the data represent mean CFU log 10 /ml ± SD of three independent replicates. Statistical significance was analyzed using one-way ANOVA (Tukey test; GraphPad prism 9). *, p < 0.01; **, p < 0.001; ***, p < 0.0001. ATc, anhydrotetracycline; CFU, colony forming unit; IVN, isovaleronitrile; SPR, surface plasmon resonance; WCL, whole-cell lysate. pathogenicity of Mtb. There are certain sigma factor deletion mutant strains that exhibit attenuated disease progression and prolonged survival in immunocompetent hosts in spite of no growth deficits under in vivo conditions. They may serve as good vaccine candidates if they persist in the host at a high level of infection and stimulate the immune response without inflicting detrimental pathological changes. A number of mycobacterial strains exhibit this kind of phenotype, including sigF, sigH, sigE, sigD, and sigC mutants (35-38, 41, 43, 46, 47). Nevertheless, we have not evaluated the immune response after depleting σ A . Further experiments are needed to determine the therapeutic potential of σ A .

Animal experimentation
Animal experiment protocols were reviewed and approved by the Institutional Animal Ethics Committee of the National Institute of Immunology, New Delhi, India (the approval number is IAEC# 462/18). The experiments were carried out as per the guidelines issued by the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), Govt. of India.

Bacterial strains and culturing
The list of bacterial strains used is given in Table 1. Cultures of Mtb (H37Rv or mutant strains) strains were grown in Middlebrook 7H9 medium (BD Biosciences) supplemented with 10% albumin, dextrose, catalase, NaCl, and 0.2% glycerol (Sigma), or 7H11 agar (BD Biosciences) 10% OADC (oleic acid added to ADC). Log phase cultures grown in the absence of ATc were used for seeding new cultures at A 0.05. The expression of the episomal copy of SigA (from pNit constructs) was induced by adding 0.2 μM isovaleronitrile (IVN). Cultures were grown in the absence or presence of 1 mg/ml ATc, and the growth was monitored by enumerating CFUs every day or 6 days postinoculation depending on the experiment. All the experiments were performed in biological triplicates.
Expression and purification of recombinant M. tuberculosis protein in E. coli system RbpA and CarD from Mtb were cloned into pET-28a vector, and the proteins were overexpressed in E. coli BL21(DE3) cells. The recombinant proteins were purified by Ni-affinity chromatography in 50 mM Tris-HCl pH8.0, 500 mM NaCl, 5% glycerol, and eluted using 20 to 500 mM gradient of imidazole. The eluted fractions were resolved on 15% SDS-PAGE and then concentrated using ultrafiltration (Merck). The proteins were buffer exchanged to 20 mM Tris-HCl pH 8, 150 mM NaCl, 5% glycerol. Subunits of Mtb RNAP were expressed and purified from E. coli BL21(DE3) cotransformed with pETBC and pACYCZA (encoding rpoB, rpoC, rpoZ, and rpoA subunits; a kind gift from Banerjee et al., 2014) (57). Core RNAP was purified as described previously, and final fractions containing RNAP were pooled, concentrated to nearly 1 mg/ml, and stored as aliquots at −80 C. sigA was cloned into MCS-I of MBP/MBP-PknA/MBP-PknB-pDuet vector (58). SigA was coexpressed with MBP, MBP-PknA, or MBPPknB, respectively, in E. coli surrogate host as we described previously (58).

Generation of mutant strain
SigA or sigB was PCR amplified from Mtb H37Rv genomic DNA with phosphorylated forward and reverse harboring NdeI and HindIII sites. Amplicons were purified and cloned into the SmaI digested and dephosphorylated pUC19 vector. The sigA insert was released with NdeI-HindIII digestion and cloned into corresponding sites in Mycobacterial-E. coli shuttle plasmid pFICTO (59). pFICTO-sigA was electroporated into Rv to generate merodiploid strain Rv::sigA, which expresses FLAG-SigA (3X-FLAG) from the L5 loci and sigA from the native loci. Upon the addition of ATc, the expression of FLAG-SigA shuts down as the tet repressor binds with its cognate operator sequences. To generate Allelic exchange substrate (AES), 1 kb left-hand-and right-hand-flanking sequences of sigA were amplified. Amplicons were digested with PflMI and ligated with oriE+ λ cos (1.6 kb) and hygr-sacB (3.6 kb) cassettes obtained after digesting pYUB1474 (32). AES was linearized with PacI and cloned into the corresponding site in the Table 1 Strains used in the study

Strains
Description Source temperature-sensitive phAE159 shuttle phagemid. The gene replacement mutant RvΔsigA was generated with the help of specialized transduction (32) by transducing Rv::sigA merodiploid strain with the temperature-sensitive phagemid. Genomic DNA was prepared from the recombinant colonies, and recombination at the native sigA loci was confirmed by performing multiple PCRs. AES for making RvΔsigB mutant was generated as described above, harboring 1 kb upstream and downstream flanks. The mutant was generated with the help of the recombineering method (60), as described previously. The recombination at the native sigB loci in the resulting colonies was confirmed with the help of multiple PCRs.

Generation of complementation constructs
Site-directed mutants in σ A were generated with the help of overlapping PCRs. sigA or sigA Mut were subcloned into NdeI-HindIII sites in the pNit-1 (ref) vector to generate pNit-SigA or pNit-SigA Mut constructs. The N-terminal deletion constructs of σ A were generated by PCR amplifying, then using different N-terminal forward primers harboring the NdeI site and the reverse primer harboring the HindIII site. They were cloned into a pNit-1 vector to generate pNit-SigA del constructs. Similarly, the sigB gene was cloned into the pNit-3F vector (61) to generate the pNit-SigB construct. The presence of mutations at the desired location was confirmed by sequencing. RvΔsigA was electroporated with pNit-SigA or pNit-SigA Mut , or pNit-SigA del , or pNit-SigB constructs to generate RvΔsi-gA::sigA or RvΔsigA::sigA Mut or RvΔsigA::sigA del or RvΔsigA::-sigB strains, respectively. Various complementation strains generated in the study are described in Table 2. The expression of the episomal copy of SigA or SigA Mut or SigA Del or SigB was induced by adding 0.2 μM IVN.
Lysate preparation, western blots, and growth kinetics Rv and RvsigA::sigA wt/mut/del strains grown in the absence of ATc or IVN were seeded at A 0.05 in the presence of 0.2 μM IVN and absence or presence of 1 μg/ml ATc. In vitro growth Functionality of SigA is dependent on its N-terminal region in the 7H9-ADC medium was performed for either 6 days or as indicated in the figure/legend. The survival was evaluated by enumerating CFUs on 7H11-OADC-agar plates. Whole-cell lysates (WCLs) were prepared as described previously (62). WCLs were estimated using the Bradford protein estimation method, and 30 μg WCLs were resolved on 10% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with α-SigA, α-GroEL1 antibodies generated previously in the lab (62).

Ex vivo peritoneal macrophages infections
BALB/c mice were injected with 4% thioglycolate solution (Difco) in the peritoneal cavity, and 96 h post-injection, peritoneal macrophage was isolated and seeded in RPMI media containing 10% fetal bovine serum. Log phase cultures of Mtb strains passed through a 26 gauge needle to create single-cell suspensions. Peritoneal macrophages were infected with Rv or RvΔsigA or RvΔsigA::sigA or RvΔsigA::sigA Mut or RvΔsi-gA::sigA del or RvΔsigA::sigRv, at MOI of 1:5 (cell: bacteria). Four hours p.i, media was removed and cells were washed thrice to remove extracellular bacteria. The cells were replenished with RPMI media containing 10% fetal bovine serum with or without 0.2 μM IVN or 1 μg/ml ATc. 24, 48, 72, or 96 h p.i, cells were lysed in 100 μl of 0.05% SDS, and CFUs were enumerated different dilutions on OADC-containing 7H11 agar plates.

In vivo mice infections
To investigate the physiological impact of sigA depletion on bacterial survival in the host, we performed murine infection experiments as described previously (62). Dox hydrochloride (1 mg/kg with 5% dextrose in drinking water) was provided to Rv and RvΔsigA-infected mice as indicated in legends/figures, either from the time of the infection (day 1) or after infection establishment (14 days p.i). CFUs were enumerated at the indicated times.

Luciferase assays
Five hundred basepair sequence just upstream of sigA ORF was cloned into an integrative shuttle plasmid (pSW-luc SApr ) at the ScaI-NdeI site. The luciferase (luc) ORF was cloned into NdeI-HindIII sites. Rv and RvΔsigA strains were electroporated with pSWN-sigA pro -luc to generate Rv::sig pro-luc and RvΔsi-gA::sig pro-luc strains. WCL was prepared as described above, and Luciferase assays were performed as described previously (62).

SPR studies
Interaction of σA, its N-terminal truncations, and its phosphoablative and phosphomimetic mutations with rrnAP3 promoter were performed in Biacore 2000 instrument (Biacore,). The sense and anti-sense strand of rrnAP3 promoter DNA (−70 to +20) was PCR amplified from Mtb genome. The sense strand was synthesized with 5 0 biotinylation modification (Sigma Aldrich). The strands were annealed before the immobilization by mixing the sense: anti-sense strand at a molecular ratio of 1:2 in Sodium Saline citrate buffer. Biotinylated rrnAP3 (5 0 ) promoter fragment was immobilized on an SA chip at a surface density of 985 RU ng/mm 2 . The experiment was executed in running buffer comprising 25 mM Hepes (pH 8), 250 mM NaCl, and 5% glycerol. The study used various concentrations of purified His-σ A and His-σ A Mut as analytes. The first channel of SA chip was not immobilized with any biotinylated promoter and was used as a control for the binding studies. BIA-evaluation software program was used to evaluate interaction kinetics. The σ A and rrnAP3 promoter interaction sensograms were used as control for calculating both σ A N-terminal truncation constructs (Fig. 6B) and σ A phosphomutants promoter-binding kinetics (Fig. 7, B and C).

In vitro transcription assay
Transcription was carried out in 5 μl transcription buffer (TB, 10 mM Tris-HCl pH 7.9, 70 mM K-glutamate, 0.1 mM EDTA, 1 mM DTT, 5% glycerol, 5 mM MgCl2, and 2 mM MnCl2). The promoter rrnAP3 (−100 to +150) was PCR amplified from Mtb genomic DNA, and the amplicons were purified. RNAP holoenzyme was assembled by mixing 100 nM core RNAP with 500 nM σ subunit followed by 15 min incubation at 37 C. Subsequently, 2 μM RbpA and 4 μM of CarD were added to the reaction and incubated for 10 min. The above mixture was incubated with a 40 nM rrnAP3 promoter fragment at 37 C for 5 min in the presence of 1 μl heparin (0.5 mg/ml) to inhibit the nonspecific RNA-DNA complex. RNA synthesis was initiated by addition of 1 μl NTP mix (final concentration: 0.1 mM of ATP, GTP, CTP, 10 μM of UTP, and 0.5 μCi α[ 32 P] UTP) and the reaction was carried out for 45 min at 37 C. RNA products were purified by precipitating using 1/10th volume of 3 M sodium acetate (pH 5.2), glycogen (1 μg/μl), and 2.5 volumes of 100% cold ethanol. The samples were centrifuged at 10,000 rpm, the pellet was washed with 70% cold ethanol, and the pellet was eventually resuspended in 20 μl nuclease-free water. 2× RNA loading dye (80% formamide, 10 mM EDTA, 0.01% Bromophenol Blue, 0.01% Xylene Cyanol) was added to the samples, heated at 95 C for 5 min chilled on ice, and resolved on 5% Urea-PAGE. Transcription products were visualized using Phosphorimager and quantified by using Quantity One software (Biorad).

RNA isolation and qRT-PCRs
Total RNA was isolated from exponentially growing bacteria cells. To isolate total RNA, cells were inoculated at 0.05 A in the presence or absence of ATc and allowed to grow for 3 days. Cells were resuspended in TRIzol reagent (Invitrogen), and total RNA was isolated from following mycobacterial cells and performed qRT-PCRs as described perviously (62).

Transcriptomics study
Agilent 2100 Bioanalyser (Agilent RNA 6000 Nano Kit) or 4200 Tape station system was used to determine RNA integrity. RNA-seq was performed using samples with RIN values higher than 7. Sequencing was performed at the Centre for cellular and molecular biology core sequencing facility. Illumina adapters and low-quality reads were eliminated from raw sequencing reads using cutadapt. Low read quality scores (<20) and <36 bp were discarded. The processed reads were then mapped to the Mtb H37RV, downloaded from https://ftp. ncbi.nlm.nih.gov/genomes/refseq/bacteria/Mycobacterium_ tuberculosis/reference/GCF_000195955.2_ASM19595v2/, using hisat2 with default parameters. Uniquely aligned reads were counted using feature Counts of the Subread package. There were 4008 genes in the gtf file, for which we had the count information. The gtf was modified to remove 200 bp from either end of SigA/SigB. Genes with a total read count 10 across all the samples were removed, resulting in 3987 genes for the SigA knockdown experiment and 3999 genes for the SigB KO experiment. DESeq2 was used to analyze differential gene expression. A gene was considered to be differentially expressed if the adjusted p-value was less than 0.05 and the absolute log 2 fold change was greater than 0.5. Raw read counts were normalized with rlog, provided by DESeq2 package for PCA plot and heat map.

Functional enrichment analysis
For functional enrichment analysis, DAVID web services were used (https://david.ncifcrf.gov/). We only used the GO terms and KEGG pathways for enrichment analysis. We only plotted the top 10 enriched GO terms/KEGG pathways based on gene counts.

Statistical analysis
One-way ANOVA was used to determine the significance of the data. The data sets were plotted and statistical significance was calculated with GraphPad Prism version 9 followed by minimal modifications using Adobe Illustrator (2021). The corresponding author can provide the source data for this study upon request.
Supporting information-This article contains supporting information.