Proteomic insights into a M. tuberculosis clinical isolate with an increased propensity to form viable but non-replicating subpopulations during acid stress

27 Phagosome acidification is one of the challenges faced by Mycobacterium tuberculosis during 28 infection. This intracellular pathogen is known to adapt to its stressful environment through stress 29 response pathways and by secreting proteins to modify the host immune response for survival and 30 proliferation. However, M. tuberculosis also holds the potential to form viable but non-replicating 31 (VBNR) and antibiotic tolerant persisters in response to environmental stress, including acid stress. 32 In this study we used a in vitro acid stress model to stimulate the formation of a VBNR 33 subpopulation in a M. tuberculosis clinical isolate with an increased propensity to form VBNR 34 bacteria. Mass spectrometry-based proteomics was used to characterize the cellular proteome and 35 culture filtrate proteome of actively replicating (pH 6,5) and VBNR enriched (pH 4,5) cultures. We 36 show that in response to acid stress, M. tuberculosis S169 increases the expression of known stress 37 response proteins, including the methyltransferase Rv1405c and the acid stress response two-38 component regulatory protein TcrX. Interestingly, we found that the dormancy response regulon 39 components were less abundant in acid stressed M. tuberculosis S169. Our protein aggregation 40 capture culture filtrate proteomic approach revealed that the culture filtrates of low pH stressed M. 41 tuberculosis S169 contained less proteins than that of actively replicating cultures. We identified 42 several proteins previously implicated in M. tuberculosis persistence, including toxin-antitoxin 43 proteins (VapC51 and VapB10), the chorismate mutase (Rv1885c), and several uncharacterized 44 proteins. The observed differences identified in the characterisation of this clinical isolate in 45 comparison to published M. tuberculosis H37Rv highlights the need to investigate M. tuberculosis 46 clinical isolates for a more representative understanding of the tuberculosis stress response. 47

Tuberculosis, TB, persister, VBNR, heterogenous, acid stress 48 49 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint Author Summary 50 Tuberculosis is caused by Mycobacterium tuberculosis and this pathogen can form a subpopulation 51 of viable but non-replicating (VBNR) cells that are recalcitrant to antibiotic treatment. These 52 persister bacteria increases the risk of treatment failure and tuberculosis recurrence following 53 treatment. Stimulation of a persister population through triggered persister formation can be achieved 54 by environmental stress factors such as low pH, nutrient starvation, hypoxia, and antibiotic exposure. 55 In this study we investigate the cellular and culture filtrate proteomes of a high persister forming 56 clinical isolate, M. tuberculosis S169, in response to acid stress. We show that following the 57 stimulation of a VBNR subpopulation in response to acid stress, several known acid stress response 58 proteins are more abundant in VBNR enriched cultures. Interestingly, we found that stress response 59 proteins were less abundant. Using a protein aggregation capture approach we successfully 60 characterized culture filtrates M. tuberculosis cultures, reducing the bacterial culture amount required 61 for these experiments. Culture filtrates differed between actively replicating and VBNR enriched 62 cultures. Several immunogenic proteins were identified in a higher abundance in the culture filtrates 63 of VBNR enriched cultures. 64 65

66 Mycobacterium tuberculosis , the pathogen which causes tuberculosis (TB), primarily infects 67 macrophages. During infection, M. tuberculosis is exposed to high levels of reactive oxygen and 68 nitrogen intermediates, reduced oxygen and nutrient availability, and low pH (1). M. tuberculosis 69 can replicate under these conditions or persist in a viable, but non-replicating (VBNR) state (2). 70 Persister M. tuberculosis , can coexist with replicating mycobacteria during infection, but these 71 bacteria are transiently insensitive to antibiotic treatment due to their non- or slow-replicating nature 72 (2,3). These antibiotic recalcitrant bacteria contribute to the length of TB treatment regimens, where 73 multiple antibiotics are required to treat infections for extended periods of time (4,5). Persister 74 subpopulations may resume growth under favourable conditions, increasing the possibility for 75 recurrent disease following treatment (6,7). 76 Bacterial persisters are thought to arise from spontaneous persistence or triggered persistence (8). 77 Spontaneous persistence describes the stochastic formation of persisters at a rate that is constant 78 during growth, accounting for approximately 1% of the bacterial population in stationary phase 79 (9,10). Triggered persistence refers to the formation of a persister subpopulation in response to a 80 stress signal such as starvation, population density, pH stress, immune factors and drug treatment (8). 81 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint In other bacterial species, mechanisms of persister formation include metabolic slow down, 82 modulation of nucleoid-associated proteins, expression of drug efflux pumps, activation of toxin-83 antitoxin modules, and upregulation of stress response genes (11–15). Similarly, a multiple stress 84 dormancy model for M. tuberculosis induced lower energy metabolism, reduced transcription and 85 translation, and increased expression of stress response genes as mechanisms of dormancy (2). 86 Several studies have highlighted the upregulation of the dormancy response regulon (DosR) in 87 response to low oxygen and nitric oxide exposure, further emphasizing the role of stress response 88 proteins in bacterial dormancy (16–18). Toxin-antitoxin modules were also more abundant in the 89 proteomes of nutrient starved M. tuberculosis, highlighting the cross-species similarities in persister 90 linked pathways (19). 91 The intracellular pathogen M. tuberculosis interacts with the host by presenting various molecules on 92 its cell surface, but also through the secretion of molecules (20,21). M. tuberculosis secretion 93 substrates play a role in nutrient acquisition, host epigenetic modification, prevention of phagosome 94 maturation, modulation of cytokine response, autophagy, redox regulation, and necrosis and bacterial 95 dissemination (20,22). Identification of proteins secreted by M. tuberculosis , especially in response 96 to environmental stress, is crucial in understanding how M. tuberculosis subverts killing by host 97 macrophages (23,24). Salmonella persisters have been shown to not only maintain a metabolically 98 active state, but to secrete effector proteins which reprogrammed the host cell polarization form a 99 pro-inflammatory to anti-inflammatory state (25). Furthermore, advances in immunopeptidomics 100 have highlighted the need to investigate proteins identified in the extracellular region of M. 101 tuberculosis by demonstrating that ESX secretion substrates are the prominent source of MHC-I 102 presented M. tuberculosis peptides (26). This finding highlights the importance of investigating M. 103 tuberculosis secreted proteins as anti-TB vaccine development candidates. 104 In this study we sought to characterize both the cellular proteome and the culture filtrate of a M. 105 tuberculosis clinical isolate which has an increased propensity to form VBNR subpopulations (7). 106 This clinical isolate, M. tuberculosis S169, was obtained from an HIV-negative patient who failed 107 treatment following standard 6-month anti-TB treatment (27). M. tuberculosis S169 is susceptible to 108 anti-TB treatment and no drug-resistance conferring mutations were identified using whole genome 109 sequencing (7). We hypothesize that VBNR M. tuberculosis may secrete a different subset of 110 proteins to that of actively replicating M. tuberculosis . Given the abundance of VBNR M. 111 tuberculosis in bacterial culture, we opted to use this clinical isolate with an increased propensity to 112 form VBNR M. tuberculosis, to increase the likelihood of identifying VBNR secreted proteins (10). 113 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint We used a low pH stress model to trigger the formation of a VBNR subpopulation and verified the 114 enrichment of a VBNR subpopulation using a dual replication reporter plasmid (28,29). To 115 characterize the culture filtrates of actively replicating and VBNR enriched cultures, we made use of 116 a protein aggregation capture approach (30). Our culture filtrate mass spectrometry approach 117 required smaller amounts of bacterial culture, reducing experimental time, technical variability from 118 pooling multiple cultures, experimental cost, and in the case of M. tuberculosis, biohazardous risk. 119 120

121 Bacterial culture and acid stress exposure 122 M. tuberculosis clinical isolate S169 obtained from Dr Stephanus Malherbe was collected by the 123 Catalysis TB – Biomarker Consortium, Ethical approval granted by Stellenbosch University Human 124 Research Ethics Committee (registration number N10/01/013) (27). Informed consent was obtained 125 from all study participants (27). All reagents were purchased from Sigma Aldrich unless otherwise 126 stated. All experiments were performed in biological triplicate. 127 Liquid mycobacterial cultures were grown in Middlebrook 7H9 supplemented with 10% dextrose-128 catalase (DC), 0.2% glucose and 0.05% Tween-80 (7H9-DC). The clinical isolate was transformed 129 with the replication reporter plasmid pTiGc and transformants were cultured in the presence of 25 130 µg/ml kanamycin at 37°C. For acid stress, M. tuberculosis S169::pTiGc was sub-cultured in 7H9-DC 131 pH 6.5 supplemented with 4mM theophylline to an OD 600 of ~1.0 prior to washing with culture 132 media. Cultures were resuspended in fresh 7H9-DC at pH 6.5 and pH 4.5 and incubated at 37°C for 133 48h. Fluorescence dilution was used for the identification of VBNR mycobacteria as previously 134 described (28). Aliquots of bacterial culture pre- and post-acid stress were plated to verify bacterial 135 viability following acid stress. 136 Imaging flow cytometry sample preparation, acquisition, and analyses 137 Cultured M. tuberculosis S169::pTiGc was sonicated for 12 minutes at 36kHz (Zues, Sonicator bath) 138 before filtering through a 40 µm filter. Cells were fixed with 4% formaldehyde in phosphate buffered 139 saline (PBS) and 0.05% Tween-80 for 30 minutes. Cells were washed and resuspended in PBS prior 140 to analysis on the Amnis® ImageStream® X Mark II Imaging Flow Cytometer. A minimum of 20,000 141 in-focus events were captured per sample and the data were analysed using IDEAS 6.2 software. 142 Signals from bright field, GFP and TurboFP635 were collected from channels 1, 2 and 4, 143 respectively. The spot count feature was used to identify a population of single cells (identified as 1 144 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint object) after which the GFP best in focus population was selected for viable cells. A histogram of the 145 normalized frequency against the intensity of TurboFP635 was generated for different pH conditions 146 and time points and overlapped to demonstrate the formation of a viable, but non-replicating 147 population. 148 VBNR bacteria can be identified by the retention of high TurboFP635. A gate was set on the top 50th 149 percentile of the induced TurboFP635 (pH 6.5 0h) culture, termed “high red”. This gate was used to 150 determine the frequency of “high red” VBNR bacteria in pH 6.5 and pH 4.5 cultures at 48h. The 151 percentage of VBNR subpopulation was calculated by the following equation: 152 % VBNR subpopulation = % high red pH 4.5 48h - % high red pH 6.5 48h 153 Protein extraction 154 Following acid stress exposure of M. tuberculosis S169::pTiGc, the cell biomass was separated from 155 the culture supernatant by centrifugation at 4000 rpm for 20 minutes. 156 Culture supernatants were sterilized by filtering twice using 0.22 µm Steriflip ® filter (Sigma 157 Aldrich) and kept at under ice-cold conditions. Culture filtrates from pH 4.5 cultures were 158 neutralized with sodium hydroxide to a pH of 6.5. All culture filtrates were concentrated using 159 Amicon Ultra 3KDa spin columns (Sigma Aldrich) from 25 mL to 500 µL. SDS was added to a final 160 concentration of 10% prior to incubation at 60°C for 1 hour and storage at -20°C. 161 Cell pellets from 25 mL cultures were stored at -20°C prior to resuspension in 1 mL lysis buffer (20 162 mM Tris-HCl, 1% SDS) containing protease inhibitors (Roche c0mpleteTM mini EDTA-free protease 163 inhibitor cocktail). Cells were washed once for 2 minutes at 14 000 rpm, 4°C. Cell pellets were 164 resuspended in 300 µL lysis buffer containing protease inhibitors and 10 µL DNAseI (RNAse-free) 165 prior to cell disruption by mechanical bead-beating (Fast-prep-24 TM, MPbio) using 300 µL acid-166 washed glass beads (425-600 µm, Sigma Aldrich). Bead-beating was performed 8 times for 20 167 seconds, at 4 m.s-1 with 1 minute intervals on ice. The cytosolic lysate was cleared by centrifugation 168 at 12 000 rpm for 10 minutes at 4°C, the supernatant was recovered then clarified again by 169 centrifugation as before. Clarified supernatants were filtered through 0.22 µm Acrodisks (Sigma 170 Aldrich) and stored at -20°C. 171 Mass spectrometry sample preparation 172 Culture filtrate and cell lysate proteins were reduced and alkylated prior to capture on MagReSyn ® 173 HILIC beads (Resyn Biosciences). Briefly, the concentrated protein sample was reduced with 5 mM 174 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint TCEP at room temperature for 1 hour and subsequently alkylated at 13 mM iodoacetamide for 1 hour 175 in the dark. Iodoacetamide was quenched with 11 mM DTT. Cell lysate proteins were digested on 176 MagReSyn® HILIC beads as recommended by the manufacturer before storage at -20 °C. 177 Culture filtrate proteins were captured onto 150 µL prepared MagReSyn ® HILIC beads, overnight at 178 37°C. Beads were washed as per manufacturer recommendations and resuspended in 50 mM TEAB 179 with 1 µg sequence modified trypsin. An on-bead digest was performed at 37°C for 18 hours, 180 shaking at 800 rpm. Peptides were recovered and beads were incubated with 100 µL 1% TFA for 30 181 seconds at 800 rpm. The 1% TFA solution was recovered and pooled with recovered peptides from 182 overnight digest. To ensure that bead captured proteins were digested, the protein digest was 183 repeated with 100 µL TEAB and 1 µg sequence modified trypsin as before and recovered peptides 184 were pooled and stored at -20°C. 185 Recovered supernatants were dried using a Concentrator plus (Eppendorf) and desalted as previously 186 described (31). Peptide concentrations were determined against a 1 mg/mL peptide solution (Pierce) 187 using a spectrophotometer (Jenway 7415). 188 LC-MS/MS 189 Liquid chromatography was performed on an Ultimate 3000 RSLC equipped with a 20mm × 100 μ m 190 C18 trap column (Thermo Scientific) and a CSH 25cm × 75 μ m, 1.7 μ m particle size C18 column 191 (Waters). Solvent A (2% acetonitrile:water and 0.1% formic acid) was used to load samples onto the 192 trap column from an autosampler (set to 7°C) at a flow rate of 2µL/min, for 5 minutes before sample 193 elution onto the analytical column. A defined flow rate of 300 nL/min with the following gradient 194 with Solvent B (100% acetonitrile, 0.1% formic acid) was used: 5–30% B over 60 min and 30–50% 195 B from 60–80 min at 45°C. 196 Culture Filtrate samples were analysed in Data Dependent Acquisition (DDA) mode. Cell lysate 197 samples were analysed in Data Independent Acquisition (DIA) mode. 198 DDA mass spectrometry analysis was performed using a Orbitrap Fusion TM Tribird TM Mass 199 spectrometer (Thermo Scientific) equipped with a Nanospray Flex ionization source on positive 200 mode with spray voltage set to 2 kV and ion transfer capillary at 290°C. Spectra were internally 201 calibrated using polysiloxane ions at m/z = 445.12003. For MS1 scans the Orbitrap detector was set 202 to a resolution of 60,000 over a scan range of 375–1500 with the AGC target at 4E5, and maximum 203 injection time of 50 ms. Data was acquired in profile more. MS2 acquisitions were performed using 204 monoisotopic precursor selection with ion charges +2 - +7 with an error tolerance of +/-10 ppm. 205 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint Precursor ions were excluded from repeat fragmentation for 60 s. Precursor ions were selected for 206 fragmentation in HCD mode using the quadrupole mass analyser at an HCD energy of 30%. 207 Fragments ions were detected in the Orbitrap mass analyzer with a resolution of 30,000. The AGC 208 target was set to 5E4 and a maximum injection time of 60 ms. Data was acquired in centroid mode. 209 DIA mass spectrometry analysis was performed on the same instrument as for DDA analysis. The 210 resolution of MS1 scans were set to 60,000 over a scan range of 375–1500. The AGC target was set 211 to standard and a maximum injection time of 100 ms. Data was acquired in profile more. Precursor 212 ions were selected for fragmentation in HCD mode using the quadrupole mass analyser at an HCD 213 energy of 30%. Precursor ions were scanned in three windows, 355-555, 555-755, and 755-955 m/z 214 and a 10 m/z isolation window with a 1 m/z overlap. Ions were detected in the Orbitrap mass 215 analyzer set to 30,000 resolution and the AGC and maximum injection time set to custom. Data were 216 acquired in centroid mode. 217 Protein identification 218 MaxQuant 2.2.0.0 was used to analyze DDA tandem mass spectrometry data using the M. 219 tuberculosis database (UP000001584) downloaded from Uniprot ( https://www.uniprot.org/) in 220 February 2023 (32,33). Carbamidomethyl cysteine was set as a fixed modification and oxidated 221 methionine and N-terminal acetylation of proteins were selected as variable modifications. A 222 maximum of 2 missed tryptic cleavages were allowed and proteins were identified with a minimum 223 of 1 unique peptide. The protein and peptide false discovery rate (FDR) threshold was less than 0.01. 224 Relative quantification was performed for identified protein groups using the MaxQuant LFQ 225 algorithm and the “match between runs” algorithm was used to detect peptides which were not 226 selected for MS/MS analysis in other replicate experiments. 227 Secreted proteins were identified using LFQ intensity data using Perseus (Figure 3). Potential 228 contaminants, reverse hits, only identified by site potential contaminants, and proteins only identified 229 with one unique peptide were removed. Proteins were considered true identifications if identified in 230 at least two of the three biological replicate experiments for a particular condition. Proteins unique to 231 either pH 6.5 or pH 4.5, where identifications were only made in one of the two conditions, were 232 identified. For the remaining proteins, the LFQ intensity data was establish if proteins were 233 significantly differentially abundant in the culture filtrates of pH 4.5 versus pH 6.5 cultures. The data 234 was imputed by replacing missing values from the normal distribution of log2 transformed data for 235 each experiment. Statistically significant differentially abundant proteins were identified following a 236 paired student’s t-test and a Benjamini-Hochberg FDR correction of 0.05. 237 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint Cell lysate DIA data was analysed using FragPipe v22.0 (34–43). The DIA_SpecLib_Quant 238 workflow was selected which built a spectral library using MSFragger-DIA against the M. 239 tuberculosis UP000001584 database with added decoy and contaminant proteins using default 240 settings. The library was filtered to 1% FDR at protein and peptide level. The spectral library was 241 applied to quantify the DIA data using DIA-NN at 1% FDR. FragPipe-Analyst was used for the 242 analysis and visualization of DIA data using the DIA workflow (44). Variance stabilizing 243 normalization for DIA data was used for normalization and a Perseus-type imputation was selected 244 for differential expression analysis. Differentially expressed genes were identified using a log fold 245 change of 2 and an Benjamini Hochberg corrected p-value of 0.05. Protein groups only identified in 246 a single biological replicate and contaminant protein groups were excluded from the identified 247 protein list. 248 M. tuberculosis gene annotations, protein names and gene ontology information was obtained from 249 Uniprot ( http://www.uniprot.org/) and KEGG (https://www.genome.jp/kegg/) (33,45). Gene 250 ontology enrichment analysis was done using the ShinyGO gene-set enrichment tool 251 (https://bioinformatics.sdstate.edu/go/) (46). 252

253 Acid stress promotes the formation of a viable, but non-replicating M. tuberculosis population 254 Clinical isolate M. tuberculosis S169 was obtained from a patient who remained culture positive 255 following 6 months of TB treatment (27). Whole genome sequencing did not identify any known 256 anti-TB treatment resistance conferring mutations and drug susceptibility was confirmed by 257 phenotypic testing. Fluorescence dilution demonstrated an increased ability of this isolate to form 258 VBNR M. tuberculosis in a macrophage infection model (7,28). We set out to establish if we could 259 replicate the VBNR formation observed for M. tuberculosis S169 in a macrophage infection model 260 using an in vitro low pH stress model (29). Briefly, M. tuberculosis S169 transformed with the 261 replication reporter plasmid, pTiGc, was cultured in the presence of theophylline to induce the 262 expression of TurboFP635. Cells were transferred into theophy lline free culture media at either pH 263 6.5 or pH 4.5 and incubated for 48h (Figure 1A-B). Imaging flow cytometry demonstrated active 264 replication of M. tuberculosis S169 at pH 6.5 with continued high levels of GFP expression, but 265 reduced levels of red fluorescence following the removal of the inducer theophylline (Figure 1C). 266 Following 48h of acid stress at pH 4.5, M. tuberculosis S169 continued to express high levels of 267 GFP, but had reduced red fluorescence intensity, suggesting a decrease in replication (Figure 1C). 268 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint Low pH stress induced the formation of 17.5% (+/- 4.5) VBNR M. tuberculosis S169, consistent 269 with previous results investigating VBNR subpopulations using a macrophage infection model (7). 270 Low pH stress of M. tuberculosis S169 results in down-regulation of DosR 271 The cell biomass recovered from three independent low pH stress experiments was analysed using 272 mass spectrometry to investigate the cellular stress response of the high VBNR M. tuberculosis S169 273 clinical isolate. We identified 2959 protein groups in the cell lysates of actively replicating and acid 274 stressed M. tuberculosis S169 which mapped to 2924 proteins in the KEGG database (Table S1) 275 following the removal of contaminant proteins and proteins only identified in a single biological 276 replicate. A comparison of actively replicating and low pH stressed cell lysates revealed that 46 277 proteins were only identified in the cell lysates of actively replicating M. tuberculosis S169 (Table 278 S2) and an additional 14 proteins were only identified in the cell lysates of VBNR-enriched M. 279 tuberculosis S169 (Table S3). Differential analysis of the cell lysate data revealed that 77 proteins 280 were significantly more abundant, and 269 proteins were significantly less abundant (adjusted p-281 value 1) in the cell lysates of low pH stressed M. tuberculosis S169 when 282 compared to that of actively replicating M. tuberculosis S169 (Table S1, Figure 2A). 283 Gene ontology (GO) enrichment analysis of proteins significantly more abundant in the cell lysates 284 of acid stressed cultures did not reveal any significant results. Regardless, the two-component system 285 TcrXY component TcrX was significantly more abundant in acid stressed M. tuberculosis S169 286 (Table S1). The TcrXY two component system has previously been shown to be upregulated in 287 response to acid stress (47). S-adenosyl methionine (SAM)-dependent methyltransferase proteins 288 Rv1403c and Rv1405c were also significantly more abundant in acid stressed M. tuberculosis S169 289 (Table S1). These SAM-dependent methyltransferases have previously been reported to be 290 upregulated in response to low pH (48–50). 291 A GO enrichment analysis of significantly less abundant proteins in cell lysates of acid stressed 292 cultures suggested a down regulation of GO terms associated with universal stress response proteins 293 (Figure 2B, Table S4). The dormancy response regulon, under the control of the two-component 294 system DevR/DevS, has previously been shown to be upregulated in response to acid stress (29,51). 295 The DosR regulon is composed of 47 genes (52). In our study, we identified 38 DosR regulon 296 encoded proteins of which 34 were significantly less abundant in the cell lysates of acid stressed M. 297 tuberculosis S169 (Table 1). 298 299 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 300 301 Table 1. Abundance of DosR proteins in M. tuberculosis S169 in response to acid stress 302 Gene Name log2FC adj. pvalue significant Gene Name log2FC adj. pvalue significant Rv0079 -3.52 0.000511 TRUE fdxA -2.08 0.0308 TRUE Rv0080 -2.51 0.0029 TRUE Rv2028c -3.43 0.00107 TRUE Rv0081 -1.12 0.0199 TRUE pfkB -6.13 0.00298 TRUE Rv0569 -5.16 0.00249 TRUE Rv2030c -4.55 0.00135 TRUE nrdZ -2.42 0.0181 TRUE hspX -5.5 0.000573 TRUE Rv0571c -1.48 0.0378 TRUE acg -4.71 0.00126 TRUE Rv0572c -1.6 0.0167 TRUE Rv2623 -4.24 0.00342 TRUE pncB2 -2.18 0.0273 TRUE Rv2624c -2.86 0.00249 TRUE Rv0574c -1.93 0.0138 TRUE Rv2625c Not detected N/A N/A Rv1733c Not detected N/A N/A Rv2626c Not detected N/A N/A Rv1734c Not detected N/A N/A Rv2627c -4.97 0.00257 TRUE Rv1735c Not detected N/A N/A Rv2628 Not detected N/A N/A narX -2.79 0.00278 TRUE Rv2629 -1.83 0.00497 TRUE narK2 -3.86 0.000717 TRUE Rv2630 -0.0868 0.905 FALSE Rv1738 -4.67 0.00242 TRUE rtcB Not detected N/A N/A Rv1812c 0.123 0.148 FALSE Rv3126c Not detected N/A N/A Rv1813c -2.89 0.00774 TRUE Rv3127 -4.67 0.00316 TRUE Rv1996 -3.13 0.0104 TRUE Rv3129 Not detected N/A N/A ctpF -4.78 0.00974 TRUE tgs1 -5.32 0.00156 TRUE Rv1998c -0.121 0.726 FALSE Rv3131 -4.43 0.00228 TRUE Rv2003c -1.93 0.012 TRUE devS -1.67 0.0153 TRUE Rv2004c -3.14 0.00576 TRUE devR -2.11 0.0301 TRUE Rv2005c -3.32 0.00426 TRUE Rv3134c -5.4 0.00139 TRUE Rv2006 0.053 0.889 FALSE .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 12 Culture filtrates of VBNR enriched cultures showed a higher abundance of lipoproteins 303 In total, we identified 461 protein groups of which 192 protein groups were only identified in the 304 culture filtrates of pH 6.5 cultures and 45 protein groups were only identified in the culture filtrates 305 of pH 4.5 cultures (Figure 3). Abundance data of protein groups identified in both test conditions 306 revealed that 83 protein groups were differentially abundant between the conditions tested (q-value /<0.05) (Figure 3, Table S5-6). Of the 83 significantly differentially 308 abundant proteins, 43 protein groups were less abundant, and 40 proteins were more abundant in the 309 culture filtrates of VBNR enriched M. tuberculosis S169 (Table S5-6). In total, we identified 275 310 protein groups, which mapped to 274 proteins in the KEGG database, in the culture filtrates of 311 actively replicating M. tuberculosis S169 cultures (Table S5). The 128 protein groups identified in 312 the culture filtrates of VBNR-enriched cultures mapped to 128 proteins in the KEGG data base 313 (Table S6). 314 GO enrichment of proteins identified in the culture filtrates of actively replicating and VBNR 315 enriched M. tuberculosis S169 cultures confirmed the enrichment of pathways extracellular region, 316 external encapsulating structure, and secreted (Figure 4, Table S7-8). Other enriched pathways 317 included cell wall, cell periphery, plasma membrane, and membrane (Figure 4, Table S7-8). The 318 lipoprotein pathway was revealed to be enriched in the culture filtrates of pH 4.5 cultures (Figure 4B, 319 Table S8). Several lipoproteins were only identified in VBNR enriched culture filtrates (LpqG, 320 DppA, LpqO, FecB2, LppL, LppM, Subl, GlnH, and Rv2585c) or identified with a higher relative 321 abundance in pH 4.5 culture filtrates (FecB, LpqB, LprG, and LprA) (Table S6). Zymogen binding, 322 preceding the proteolytic cleavage of enzymes to an active state, was also enriched in the culture 323 filtrates of acid-stressed M. tuberculosis S169 (Figure 4B). Zymogen binding proteins were more 324 abundant in the culture filtrates of VBNR enriched cultures, including MetK, LpdC, Mpt64, GroES, 325 FbpA and FpbB (Table S6, S8). Proteases were also enriched within the culture filtrates of acid 326 stressed M. tuberculosis S169 and included proteases HtrA1, PepA, PepD, Rv3671c, Clp1, Clp2, 327 MycP3 and Rv2672 (Table S6, S9). 328

329 Phagosome acidification is an environmental stress faced by M. tuberculosis during host infection 330 (53). In this study, we took advantage of a low pH stress model to trigger the formation of a M. 331 tuberculosis VBNR subpopulation (29). To increase the probability of identifying VBNR secreted 332 proteins, we made use of a clinical isolate, M. tuberculosis S169, which we previously showed to 333 form high proportions of VBNR bacteria (7,27). The fluorescence dilution replication plasmid 334 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 13 enabled the tracking of bacterial replication in response to low pH stress (Figure 1A) (28,29). 335 Following acid-stress for 48h at pH 4.5 (Figure 1B), imaging flow cytometry confirmed the 336 formation of a large VBNR subpopulation of M. tuberculosis S169 (Figure 1C). 337 Proteomic characterization of the cell lysate revealed the differential abundance of 346 proteins in 338 response to acid-stress (Figure 2A, Table S1). Of the 77 proteins significantly more abundant 339 (adjusted p-value 2) in the cell lysates of acid stressed M. tuberculosis S169, no 340 enriched pathways were identified using GO enrichment analysis. However, in agreement with 341 previous reports, the TcrXY two-component system components were more abundant in acid-342 stressed M. tuberculosis S169, with the response regulator TcrX significantly more abundant (Table 343 S1) (47). TcrX is required for M. tuberculosis survival during chronic infection (47). Similarly, the 344 acid stress-induced methyltransferases Rv1403c and Rv1405c were also significantly more abundant 345 in acid stressed M. tuberculosis S169 (Table S1), as previously reported for M. tuberculosis 346 (49,50,54). Interestingly, Rv1405c has also been reported to be upregulated during the enduring 347 hypoxic response and nitrosative stress (55,56). Even though Rv1405c is not essential for in vitro 348 survival, it is required for survival in C57BL/6J mice (57,58). The role of the Rv1405c 349 methyltransferase during infection remains unknown. 350 Two-component systems are required by the bacteria to respond to environmental changes. The PhoP 351 component from the PhoPR two-component system is known to positively regulate the aprABC 352 operon in response to acidic pH (59–61). In this study, the AprA protein was only identified in acid 353 stressed M. tuberculosis S169 (Table S1, S3) and AprB and AprC proteins were not detected (Table 354 S1). Interestingly, despite detection of AprA in VBNR enriched cultures, PhoPR components were 355 found to be less abundant in the cell lysates of acid stressed bacteria (Table S1). The two-component 356 system, KdpD/KdpE, has been suggested to play a role in the evasion of phagocytic killing and 357 enabling bacterial persistence (62). In this study, KdpA, KdpB and KdpD were all significantly more 358 abundant in the cell lysates of VBNR-enriched cultures (Table S1). KdpE was found to be more 359 abundant, but not significantly (Table S1). Interestingly, KdpA has been suggested to be required for 360 ATP homeostasis and persister formation in Mycobacterium marinum (63). 361 In response to acid stress, 269 cell lysate proteins were significantly less abundant (adjusted p-value 362 2) than in the of actively replicating M. tuberculosis S169 (Table S1). Gene 363 Ontology enrichments revealed a lower abundance of universal stress proteins (Figure 2B, Table S4), 364 including components from the two-component system DosR, also known as the dormancy response 365 regulon. DosR is known to be upregulated in response to hypoxia, starvation and low pH (29,51,64). 366 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 14 Interestingly, in this study, 34 components of the DosR regulon were significantly less abundant in 367 the acid-stressed M. tuberculosis S169 (Table 1). These results contrast with some previous reports, 368 which largely show an upregulation of DosR in response to environmental stress. However, a lower 369 induction for DosR in response to low pH has been reported in comparison to hypoxia, starvation 370 and stationary phase growth (51). In agreement with our results, DosR components Rv0080, NarX, 371 Rv2030c, and Rv1813c have previously been shown to be downregulated in response to acid stress 372 (51). The DosR regulon is largely down regulated in a M. tuberculosis pellicle biofilm model (65). 373 More recently, another pellicle biofilm study showed the down regulation of DosR genes in five of 374 the six M. tuberculosis lineage 4 clinical isolates studied (66). The clinical isolate investigated in this 375 study, M. tuberculosis S169, belongs to lineage 4 (7). Interestingly, M. tuberculosis H37Rv, in which 376 the DosR regulon is upregulated in response to low pH, also belongs to lineage 4 (29). These 377 findings highlight the need to investigate the response of M. tuberculosis clinical isolates to 378 physiologically relevant stress conditions for a more comprehensive understanding of the 379 mycobacterial stress response. 380 RocA, EspA, EspC, Rv2390c, and PE34 were significantly more abundant in the cell lysates of acid 381 stressed M. tuberculosis S169, as previously reported (Table S1) (67). ESX-1 is important for M. 382 tuberculosis virulence and EspA and EspC are ESX-1 secretion associated proteins (68–71). Several 383 other ESX-1 proteins were more abundant in the cell lysates of acid stressed cultures (Table S1) (72–384 74). Despite the increased abundance of EspA, EspB, EspD in acid stressed cell lysates, these 385 proteins were not detected in the culture filtrates of acid stressed M. tuberculosis S169 (Table S6). 386 EspF, EspC, EspH, EspR, and EspK proteins were present in the culture filtrates of actively 387 replicating M. tuberculosis S169 (Table S5). In agreement with previous reports, Rv0516c, LipL, and 388 PPE59 were less abundant in response to acid stress (67). Interestingly, PPE22 has not previously 389 been reported to be upregulated in response to acid stress, however, in this study PPE22 was 390 significantly more abundant in the cell lysates of acid stressed M. tuberculosis S169 (Table S1) (67). 391 Moreso, PPE22 was only identified in the culture filtrates of acid stressed cultures (Table S6). PPE22 392 has been previously been detected in guinea pig lungs at 30 days post infection and more recently 393 has been shown to induce a protective immune response in BALB/c mice, showing promise as a 394 vaccine development candidate (75,76). 395 Several cell division proteins were less abundant in the cell lysates of VBNR-enriched cultures 396 including WhiB2, MtrA, SepF, FtsZ, FtsK, FtsQ, FtsW, FtsE, CwsA, and CrgA (Table S1). We also 397 detected a lower abundance of DNA replication and repair proteins, including ImuA, RecA, RecR, 398 RecN, DnaB, and Rv1277 (Table S1). The downregulation of DNA replication and repair proteins 399 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 15 and cell division protein aligns with reduced bacterial replication, as observed with the fluorescence 400 dilution experiments (Figure 1). Low pH stress induced the formation of a VBNR subpopulation 401 (Figure 1C). Interestingly, resuscitation-promoting factors (Rpf) RipA, RipB, and RpfC were 402 significantly less abundant in the cell lysates of VBNR enriched M. tuberculosis S169 cultures 403 (Table S1). RipB was detected in the culture filtrates of both actively replicating and VBNR enriched 404 cultures, but at a significantly lower abundance in VBNR enriched culture filtrates (Table S5-6). 405 Other Rpf proteins were identified in the culture filtrates of both test conditions, but with no 406 significant difference in abundance (Table S9). Rpf muralytic enzymes stimulate growth of dormant 407 M. tuberculosis and the loss of Rpfs results in an impaired ability of M. tuberculosis to resuscitate 408 from a non-culturable state (77). 409 Culture supernatants have low protein concentrations, often resulting in the need to pool culture 410 supernatants from multiple cultures to a obtain enough protein. This practice increases the possibility 411 of introducing inter-culture variation. To overcome this limitation, we applied a protein aggregation 412 capture approach to study the culture supernatant of a single bacterial culture per replicate 413 experiment. A single culture has the benefit of reducing time, cost, and biohazardous risk in addition 414 to limiting technical and biological variation from multiple cultures. Applying this approach, we 415 showed that the culture filtrates of actively replicating M. tuberculosis S169 contained 274 proteins 416 compared to the 128 proteins identified in the culture filtrates of VBNR enriched M. tuberculosis. 417 GO pathway enrichment analysis confirmed the enrichment of extracellular region and secreted 418 pathways (Figure 4). Zymogen binding, lipoprotein, protein folding, and protease pathways were 419 enriched from proteins identified in the extracellular fraction of low pH stressed M. tuberculosis 420 S169 (Figure 4B). Zymogens are the inactive precursors of enzymes which get converted to active 421 forms by proteolysis. Lipoproteins have been implicated in M. tuberculosis virulence and immune 422 modulation (78). Other enriched pathways included the external encapsulating structure, cell 423 periphery, and plasma membrane which may be the result of culturing M. tuberculosis S169 in media 424 containing the detergent Tween-80 to prevent bacterial clumping. The inclusion of Tween-80 in M. 425 tuberculosis culture media has been speculated to result in the solubilization of lipids and the 426 shedding of surface adhered molecules (79,80). As indicated by the pathway enrichment analysis, 427 cytosolic proteins including RNA polymerase subunits were identified in actively replicating M. 428 tuberculosis S169 culture filtrates (Table S5). Small ribosomal subunits were also identified in the 429 culture filtrates of both actively replicating and VBNR enriched M. tuberculosis S169 cultures (Table 430 S5 and S6). The identification of these proteins outside the cell may suggest some cell lysis occurred. 431 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 16 The culture filtrates of VBNR enriched M. tuberculosis S169 cultures contained 45 proteins not 432 identified in the extracellular fraction of actively replicating bacteria (Table S6). These proteins and 433 the 83 significantly differentially abundant proteins (40 more abundant and 43 less abundant) 434 identified in the culture filtrates of acid stressed M. tuberculosis S169 suggest that M. tuberculosis 435 may secrete a different subset of proteins in response to low pH stress. The VBNR subpopulation 436 only accounted for 17.5% (+/- 4.5) of the bacterial population investigated at pH 4.5, however, we 437 speculate that VBNR protein secretion contributed to the differences observed in the culture filtrates 438 between pH 6.5 and pH 4.5 cultures. Although not investigated in this study, differences in the 439 proteins found in the extracellular region of M. tuberculosis S169, may result in a different immune 440 response during infection. Culture filtrates from VBNR enriched cultures included proteins from 441 Toxin-antitoxin (TA) systems, VapC51 and VapB10 (Table S6). TA systems have been implicated in 442 the adaptation to environmental stress and bacterial persistence. Type II TA systems are highly 443 abundant in M. tuberculosis genomes (10,81). Interestingly, the chorismate mutase Rv1885c was 444 only identified in the secreted fraction of VBNR enriched cultures, and was recently suggested 445 contribute to Mycobacterium bovis BCG pathogenesis by inhibiting mitochondria-mediated cell 446 death of macrophages (82). Immunogenic proteins more abundant in the culture filtrates of VBNR 447 enriched cultures included FbpA (Mpt44), Mpt53, Mpt64 and Mpt63 (Table S6). 448 In this study we investigated the cellular proteome and the extracellular region of a clinical isolate 449 with an increased propensity to form VBNR bacteria in response to low pH stress (7). We 450 acknowledge that our study was limited by only investigating a single clinical isolate, however, this 451 isolate was chosen to increase the likelihood of identifying changes in the proteome because of its 452 increased propensity to form VBNR bacteria. We demonstrated that this clinical isolate did form a 453 viable but non-replicating population in response to in vitro low pH stress. Cell lysate proteomics 454 revealed increased abundance of known acid stress proteins, however, in contrast to what has been 455 published previously, several proteins of the DosR response regulon were significantly less abundant 456 in low pH stressed M. tuberculosis S169. This study highlights the need to investigate the cellular 457 response of clinical isolates, specifically clinical isolates obtained from individuals with 458 unfavourable outcomes, to improve our understanding of factors which may contribute to treatment 459 failure. Using our culture filtrate mass spectrometry approach, we demonstrated that the culture 460 filtrate composition of actively replicating and low pH stressed VBNR enriched cultures had 461 different compositions. While we cannot definitively demonstrate secretion of proteins by VBNR 462 bacteria, several proteins identified in the culture filtrates of VBNR enriched cultures have 463 implicated roles in bacterial persistence. Importantly, the culture filtrate approached used in this 464 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 17 study has the potential to be used not only to investigate M. tuberculosis extracellular fractions but 465 can be adapted to study extracellular proteins in other bacteria. 466 467 Declarations 468 Ethics approval statement 469 Ethics approval was obtained from the Human Research Ethics Committee (N10/01/013) and the 470 Biological and Environmental Safety Committee (BES-2023-13049) at Stellenbosch University. 471 Consent for publication 472 Not applicable. 473 Availability of data and materials 474 Imaging flow cytometry data are available from the corresponding author upon request. Mass 475 spectrometry proteomics data are available from the ProteomeXchange Consortium via the PRIDE 476 partner repository with the identifiers PXD068623 and PXD068720 (83). 477 Competing interests 478 Authors declare that the research reported in this manuscript was completed in the absence of any 479 commercial or financial relationships which could constitute a potential conflict of interest. 480 Funding 481 This research was supported by the VALIDATE Network which was funded by Gates Foundation 482 (INV-031830) and the South African government through the National Research Foundation of 483 South Africa (NRF) and the South African Medical Research Council (SAMRC). NK acknowledges 484 research and salary support from the VALIDATE Network, which was funded by the Gates 485 Foundation (INV-031830). SS is funded by the South African Research Chairs Initiative of the 486 Department of Science and Technology and National Research Foundation (NRF) of South Africa, 487 award number UID 86539. 488 The authors are all affiliated with the with the DSI-NRF Centre of Excellence for Biomedical 489 Tuberculosis Research; South African Medical Research Council Centre for Tuberculosis Research; 490 Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, 491 Stellenbosch University, Cape Town. 492 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 18 Authors contributions 493 NK, JC, JM, and SS assisted with experimental design and conceptualization. NK and JC performed 494 the experimental work and NK analyzed the results. NK drafted the manuscript, tables, and figures. 495 All authors contributed to this manuscript and approved the submitted version. 496

497 We acknowledge Maré Volk from the Central Analytical Facilities at Stellenbosch University for 498 technical assistance for mass spectrometry. 499 Figure 1A was created in BioRender. Sampson, S. (2025) https://BioRender.com/149jzej. 500 Figure 1B was created in BioRender. Sampson, S. (2025) https://BioRender.com/fyvj3w9. 501 502

503 1. Gengenbacher M, Kaufmann SHE. Mycobacterium tuberculosis: Success through dormancy. 504 FEMS Microbiology Reviews. 2012;36(3):514–32. 505 2. Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, et al. A Novel In Vitro 506 Multiple-Stress Dormancy Model for Mycobacterium tuberculosis Generates a Lipid-Loaded, 507 Drug-Tolerant, Dormant Pathogen. Ahmed N, editor. PLoS ONE. 2009 Jun 29;4(6):e6077. 508 3. Michaux C, Ronneau S, Giorgio RT, Helaine S. Antibiotic tolerance and persistence have 509 distinct fitness trade-offs. Monack DM, editor. PLoS Pathog. 2022 Nov 14;18(11):e1010963. 510 4. Zhang Y, Yew WW, Barer MR. Targeting Persisters for Tuberculosis Control. Antimicrob 511 Agents Chemother. 2012 May;56(5):2223–30. 512 5. World Health Organization. Global Tuberculosis Report 2022 [Internet]. [cited 2022 Dec 5]. 513 Available from: https://www.who.int/teams/global-tuberculosis-programme/tb-reports/global-514 tuberculosis-report-2022 515 6. Huemer M, Mairpady Shambat S, Brugger SD, Zinkernagel AS. Antibiotic resistance and 516 persistence—Implications for human health and treatment perspectives. EMBO Reports. 2020 517 Dec 3;21(12):e51034. 518 7. Coetzee JL, Kriel NL, Loubser J, Dippenaar A, Sampson SL, Malherbe ST, et al. Assessing the 519 propensity of TB clinical isolates to form viable but non-replicating subpopulations. Sci Rep. 520 2024 Nov 12;14(1):27686. 521 8. Balaban NQ, Helaine S, Lewis K, Ackermann M, Aldridge B, Andersson DI, et al. Definitions 522 and guidelines for research on antibiotic persistence. Nat Rev Microbiol. 2019 Jul;17(7):441–8. 523 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 19 9. Keren I, Shah D, Spoering A, Kaldalu N, Lewis K. Specialized Persister Cells and the 524 Mechanism of Multidrug Tolerance in Escherichia coli. J Bacteriol. 2004 Dec 15;186(24):8172–525 80. 526 10. Keren I, Minami S, Rubin E, Lewis K. Characterization and Transcriptome Analysis of 527 Mycobacterium tuberculosis Persisters. Davies JE, editor. mBio. 2011 Jul;2(3):e00100-11. 528 11. Shan Y, Brown Gandt A, Rowe SE, Deisinger JP, Conlon BP, Lewis K. ATP-Dependent 529 Persister Formation in Escherichia coli. Bush K, editor. mBio. 2017 Mar 8;8(1):e02267-16. 530 12. Dörr T, Vulić M, Lewis K. Ciprofloxacin Causes Persister Formation by Inducing the TisB toxin 531 in Escherichia coli. Waldor MK, editor. PLoS Biol. 2010 Feb 23;8(2):e1000317. 532 13. Radzikowski JL, Vedelaar S, Siegel D, Ortega ÁD, Schmidt A, Heinemann M. Bacterial 533 persistence is an active σ S stress response to metabolic flux limitation. Molecular Systems 534 Biology. 2016 Sep;12(9):882. 535 14. Pu Y, Zhao Z, Li Y, Zou J, Ma Q, Zhao Y, et al. Enhanced Efflux Activity Facilitates Drug 536 Tolerance in Dormant Bacterial Cells. Molecular Cell. 2016 Apr;62(2):284–94. 537 15. Harms A, Fino C, Sørensen MA, Semsey S, Gerdes K. Prophages and Growth Dynamics 538 Confound Experimental Results with Antibiotic-Tolerant Persister Cells. Vogel J, editor. mBio. 539 2017 Dec 29;8(6):e01964-17. 540 16. Voskuil MI. Mycobacterium tuberculosis gene expression during environmental conditions 541 associated with latency. Tuberculosis. 2004 Jan;84(3–4):138–43. 542 17. Honaker RW, Leistikow RL, Bartek IL, Voskuil MI. Unique Roles of DosT and DosS in DosR 543 Regulon Induction and Mycobacterium tuberculosis Dormancy. Infect Immun. 2009 544 Aug;77(8):3258–63. 545 18. Schubert OT, Ludwig C, Kogadeeva M, Zimmermann M, Rosenberger G, Gengenbacher M, et 546 al. Absolute Proteome Composition and Dynamics during Dormancy and Resuscitation of 547 Mycobacterium tuberculosis. Cell Host & Microbe. 2015 Jul;18(1):96–108. 548 19. Albrethsen J, Agner J, Piersma SR, Højrup P, Pham TV, Weldingh K, et al. Proteomic Profiling 549 of Mycobacterium tuberculosis Identifies Nutrient-starvation-responsive Toxin–antitoxin 550 Systems. Molecular & Cellular Proteomics. 2013 May;12(5):1180–91. 551 20. Pal R, Bisht MK, Mukhopadhyay S. Secretory proteins of Mycobacterium tuberculosis and their 552 roles in modulation of host immune responses: focus on therapeutic targets. The FEBS Journal. 553 2022 Jul;289(14):4146–71. 554 21. Vinod V, Vijayrajratnam S, Vasudevan AK, Biswas R. The cell surface adhesins of 555 Mycobacterium tuberculosis. Microbiological Research. 2020 Feb;232:126392. 556 22. Siegrist MS, Unnikrishnan M, McConnell MJ, Borowsky M, Cheng TY, Siddiqi N, et al. 557 Mycobacterial Esx-3 is required for mycobactin-mediated iron acquisition. Proc Natl Acad Sci 558 USA. 2009 Nov 3;106(44):18792–7. 559 23. Augenstreich J, Briken V. Host Cell Targets of Released Lipid and Secreted Protein Effectors of 560 Mycobacterium tuberculosis. Front Cell Infect Microbiol. 2020 Oct 23;10:595029. 561 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 20 24. Briken V, Miller JL. Living on the edge: inhibition of host cell apoptosis by Mycobacterium 562 tuberculosis. Future Microbiology. 2008 Aug;3(4):415–22. 563 25. Stapels DAC, Hill PWS, Westermann AJ, Fisher RA, Thurston TL, Saliba AE, et al. Salmonella 564 persisters undermine host immune defenses during antibiotic treatment. Science. 2018 Dec 565 7;362(6419):1156–60. 566 26. Leddy O, White FM, Bryson BD. Immunopeptidomics reveals determinants of Mycobacterium 567 tuberculosis antigen presentation on MHC class I. eLife. 2023 Apr 19;12:e84070. 568 27. the Catalysis TB–Biomarker Consortium, Malherbe ST, Shenai S, Ronacher K, Loxton AG, 569 Dolganov G, et al. Persisting positron emission tomography lesion activity and Mycobacterium 570 tuberculosis mRNA after tuberculosis cure. Nat Med. 2016 Oct;22(10):1094–100. 571 28. Mouton JM, Helaine S, Holden DW, Sampson SL. Elucidating population-wide mycobacterial 572 replication dynamics at the single-cell level. Microbiology. 2016 Jun 1;162(6):966–78. 573 29. Mouton JM, Heunis T, Dippenaar A, Gallant JL, Kleynhans L, Sampson SL. Comprehensive 574 Characterization of the Attenuated Double Auxotroph Mycobacterium 575 tuberculosisΔ leuDΔ panCD as an Alternative to H37Rv. Front Microbiol. 2019 Aug 20;10:1922. 576 30. Batth TS, Tollenaere MaximAX, Rüther P, Gonzalez-Franquesa A, Prabhakar BS, Bekker-577 Jensen S, et al. Protein Aggregation Capture on Microparticles Enables Multipurpose Proteomics 578 Sample Preparation*. Molecular & Cellular Proteomics. 2019 May;18(5):1027–35. 579 31. Kriel NL, Heunis T, Sampson SL, Gey Van Pittius NC, Williams MJ, Warren RM. Identifying 580 nucleic acid-associated proteins in Mycobacterium smegmatis by mass spectrometry-based 581 proteomics. BMC Mol and Cell Biol. 2020 Dec;21(1):19. 582 32. Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range 583 mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008 584 Dec;26(12):1367–72. 585 33. The UniProt Consortium. UniProt: a hub for protein information. Nucleic Acids Research. 2015 586 Jan 28;43(D1):D204–12. 587 34. Yu F, Teo GC, Kong AT, Fröhlich K, Li GX, Demichev V, et al. Analysis of DIA proteomics 588 data using MSFragger-DIA and FragPipe computational platform. Nat Commun. 2023 Jul 589 12;14(1):4154. 590 35. Teo GC, Polasky DA, Yu F, Nesvizhskii AI. Fast Deisotoping Algorithm and Its Implementation 591 in the MSFragger Search Engine. J Proteome Res. 2021 Jan 1;20(1):498–505. 592 36. Yang KL, Yu F, Teo GC, Li K, Demichev V, Ralser M, et al. MSBooster: improving peptide 593 identification rates using deep learning-based features. Nat Commun. 2023 Jul 27;14(1):4539. 594 37. Käll L, Canterbury JD, Weston J, Noble WS, MacCoss MJ. Semi-supervised learning for peptide 595 identification from shotgun proteomics datasets. Nat Methods. 2007 Nov;4(11):923–5. 596 38. Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A Statistical Model for Identifying Proteins by 597 Tandem Mass Spectrometry. Anal Chem. 2003 Sep 1;75(17):4646–58. 598 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 21 39. Philosopher: a versatile toolkit for shotgun proteomics data analysis | Nature Methods [Internet]. 599 [cited 2024 Jun 14]. Available from: https://www.nature.com/articles/s41592-020-0912-y 600 40. Demichev V, Szyrwiel L, Yu F, Teo GC, Rosenberger G, Niewienda A, et al. dia-PASEF data 601 analysis using FragPipe and DIA-NN for deep proteomics of low sample amounts. Nat Commun. 602 2022 Jul 8;13(1):3944. 603 41. Li K, Vaudel M, Zhang B, Ren Y, Wen B. PDV: an integrative proteomics data viewer. Wren J, 604 editor. Bioinformatics. 2019 Apr 1;35(7):1249–51. 605 42. MacLean B, Tomazela DM, Shulman N, Chambers M, Finney GL, Frewen B, et al. Skyline: an 606 open source document editor for creating and analyzing targeted proteomics experiments. 607 Bioinformatics. 2010 Apr 1;26(7):966–8. 608 43. Kong AT, Leprevost FV, Avtonomov DM, Mellacheruvu D, Nesvizhskii AI. MSFragger: 609 ultrafast and comprehensive peptide identification in mass spectrometry–based proteomics. Nat 610 Methods. 2017 May;14(5):513–20. 611 44. Hsiao Y, Zhang H, Li GX, Deng Y, Yu F, Kahrood HV, et al. Analysis and visualization of 612 quantitative proteomics data using FragPipe-Analyst [Internet]. 2024 [cited 2024 Jun 13]. 613 Available from: http://biorxiv.org/lookup/doi/10.1101/2024.03.05.583643 614 45. Kanehisa M, Goto S. KEGG: Kyoto Encyclopedia of Genes and Genomes. 615 46. Ge SX, Jung D, Yao R. ShinyGO: a graphical gene-set enrichment tool for animals and plants. 616 Valencia A, editor. Bioinformatics. 2020 Apr 15;36(8):2628–9. 617 47. Stupar M, Tan L, Kerr ED, De Voss CJ, Forde BM, Schulz BL, et al. TcrXY is an acid-sensing 618 two-component transcriptional regulator of Mycobacterium tuberculosis required for persistent 619 infection. Nat Commun. 2024 Feb 22;15(1):1615. 620 48. Golby P, Hatch KA, Bacon J, Cooney R, Riley P, Allnutt J, et al. Comparative transcriptomics 621 reveals key gene expression differences between the human and bovine pathogens of the 622 Mycobacterium tuberculosis complex. Microbiology. 2007 Oct 1;153(10):3323–36. 623 49. Rohde KH, Abramovitch RB, Russell DG. Mycobacterium tuberculosis Invasion of 624 Macrophages: Linking Bacterial Gene Expression to Environmental Cues. Cell Host & Microbe. 625 2007 Nov;2(5):352–64. 626 50. Healy C, Golby P, MacHugh DE, Gordon SV. The MarR family transcription factor Rv1404 627 coordinates adaptation of Mycobacterium tuberculosis to acid stress via controlled expression of 628 Rv1405c, a virulence-associated methyltransferase. Tuberculosis. 2016 Mar;97:154–62. 629 51. Vilchèze C, Yan B, Casey R, Hingley-Wilson S, Ettwiller L, Jacobs WR. Commonalities of 630 Mycobacterium tuberculosis Transcriptomes in Response to Defined Persisting Macrophage 631 Stresses. Front Immunol. 2022 Jul 1;13:909904. 632 52. Bartek IL, Rutherford R, Gruppo V, Morton RA, Morris RP, Klein MR, et al. The DosR regulon 633 of M. tuberculosis and antibacterial tolerance. Tuberculosis. 2009 Jul;89(4):310–6. 634 53. Gengenbacher M, Kaufmann SHE. Mycobacterium tuberculosis/i5 : success through dormancy. 635 FEMS Microbiol Rev. 2012 May;36(3):514–32. 636 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 22 54. Golby P, Hatch KA, Bacon J, Cooney R, Riley P, Allnutt J, et al. Comparative transcriptomics 637 reveals key gene expression differences between the human and bovine pathogens of the 638 Mycobacterium tuberculosis complex. Microbiology. 2007 Oct 1;153(10):3323–36. 639 55. Rustad TR, Harrell MI, Liao R, Sherman DR. The Enduring Hypoxic Response of 640 Mycobacterium tuberculosis. Bähler J, editor. PLoS ONE. 2008 Jan 30;3(1):e1502. 641 56. Birhanu AG, Gómez-Muñoz M, Kalayou S, Riaz T, Lutter T, Yimer SA, et al. Proteome 642 Profiling of Mycobacterium tuberculosis Cells Exposed to Nitrosative Stress. ACS Omega. 2022 643 Feb 1;7(4):3470–82. 644 57. Sassetti CM, Boyd DH, Rubin EJ. Genes required for mycobacterial growth defined by high 645 density mutagenesis: Genes required for mycobacterial growth. Molecular Microbiology. 2003 646 Mar 25;48(1):77–84. 647 58. Sassetti CM, Rubin EJ. Genetic requirements for mycobacterial survival during infection. Proc 648 Natl Acad Sci USA. 2003 Oct 28;100(22):12989–94. 649 59. Abramovitch RB, Rohde KH, Hsu F, Russell DG. aprABC/i5 : a Mycobacterium tuberculosis 650 complex/i5 specific locus that modulates pH/i5 driven adaptation to the macrophage phagosome. 651 Molecular Microbiology. 2011 May;80(3):678–94. 652 60. Walters SB, Dubnau E, Kolesnikova I, Laval F, Daffe M, Smith I. The Mycobacterium 653 tuberculosis PhoPR two/i5 component system regulates genes essential for virulence and complex 654 lipid biosynthesis. Molecular Microbiology. 2006 Apr;60(2):312–30. 655 61. Baker JJ, Johnson BK, Abramovitch RB. Slow growth of MYCOBACTERIUM TUBERCULOSIS at 656 acidic PH is regulated by PHOPR and host/i5 associated carbon sources. Molecular Microbiology. 657 2014 Oct;94(1):56–69. 658 62. Vlisidou I, Eleftherianos I, Dorus S, Yang G, ffrench-Constant RH, Reynolds SE, et al. The 659 KdpD/KdpE two-component system of Photorhabdus asymbiotica promotes bacterial survival 660 within M. sexta hemocytes. Journal of Invertebrate Pathology. 2010 Nov;105(3):352–62. 661 63. Liu X, Wang C, Yan B, Lyu L, Takiff HE, Gao Q. The potassium transporter KdpA affects 662 persister formation by regulating ATP levels in Mycobacterium marinum. Emerging Microbes & 663 Infections. 2020 Jan;9(1):129–39. 664 64. Park H, Guinn KM, Harrell MI, Liao R, Voskuil MI, Tompa M, et al. Rv3133c/dosR is a 665 transcription factor that mediates the hypoxic response of Mycobacterium tuberculosis. 666 Molecular Microbiology. 2003 May;48(3):833–43. 667 65. Richards JP, Cai W, Zill NA, Zhang W, Ojha AK. Adaptation of Mycobacterium tuberculosis to 668 Biofilm Growth Is Genetically Linked to Drug Tolerance. Antimicrobial Agents and 669 Chemotherapy. 2019;63(11). 670 66. Youngblom MA, Smith TM, Murray HJ, Pepperell CS. Adaptation of the Mycobacterium 671 tuberculosis transcriptome to biofilm growth. Sassetti CM, editor. PLoS Pathog. 2024 Apr 672 18;20(4):e1012124. 673 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 23 67. Tan S, Sukumar N, Abramovitch RB, Parish T, Russell DG. Mycobacterium tuberculosis 674 Responds to Chloride and pH as Synergistic Cues to the Immune Status of its Host Cell. Bishai 675 WR, editor. PLoS Pathog. 2013 Apr 4;9(4):e1003282. 676 68. Sørensen AL, Nagai S, Houen G, Andersen P, Andersen AB. Purification and characterization of 677 a low-molecular-mass T-cell antigen secreted by Mycobacterium tuberculosis. Infect Immun. 678 1995 May;63(5):1710–7. 679 69. Brodin P, Majlessi L, Marsollier L, De Jonge MI, Bottai D, Demangel C, et al. Dissection of 680 ESAT-6 System 1 of Mycobacterium tuberculosis and Impact on Immunogenicity and Virulence. 681 Infect Immun. 2006 Jan;74(1):88–98. 682 70. Hsu T, Hingley-Wilson SM, Chen B, Chen M, Dai AZ, Morin PM, et al. The primary 683 mechanism of attenuation of bacillus Calmette–Guérin is a loss of secreted lytic function 684 required for invasion of lung interstitial tissue. Proc Natl Acad Sci USA. 2003 Oct 685 14;100(21):12420–5. 686 71. Stanley SA, Raghavan S, Hwang WW, Cox JS. Acute infection and macrophage subversion by 687 Mycobacterium tuberculosis require a specialized secretion system. Proc Natl Acad Sci USA. 688 2003 Oct 28;100(22):13001–6. 689 72. Fortune SM, Jaeger A, Sarracino DA, Chase MR, Sassetti CM, Sherman DR, et al. Mutually 690 dependent secretion of proteins required for mycobacterial virulence. Proc Natl Acad Sci USA. 691 2005 Jul 26;102(30):10676–81. 692 73. Guo Q, Bi J, Wang H, Zhang X. Mycobacterium tuberculosis ESX-1-secreted substrate protein 693 EspC promotes mycobacterial survival through endoplasmic reticulum stress-mediated 694 apoptosis. Emerging Microbes & Infections. 2021 Jan;10(1):19–36. 695 74. Chen JM, Boy-Röttger S, Dhar N, Sweeney N, Buxton RS, Pojer F, et al. EspD Is Critical for the 696 Virulence-Mediating ESX-1 Secretion System in Mycobacterium tuberculosis. J Bacteriol. 2012 697 Feb 15;194(4):884–93. 698 75. Kruh NA, Troudt J, Izzo A, Prenni J, Dobos KM. Portrait of a Pathogen: The Mycobacterium 699 tuberculosis Proteome In Vivo. Aziz RK, editor. PLoS ONE. 2010 Nov 11;5(11):e13938. 700 76. Wang Y, Li Z, Wu S, Fleming J, Li C, Zhu G, et al. Systematic Evaluation of Mycobacterium 701 tuberculosis Proteins for Antigenic Properties Identifies Rv1485 and Rv1705c as Potential 702 Protective Subunit Vaccine Candidates. Ehrt S, editor. Infect Immun. 2021 Feb 16;89(3):e00585-703 20. 704 77. Kana BD, Gordhan BG, Downing KJ, Sung N, Vostroktunova G, Machowski EE, et al. The 705 resuscitation/i5 promoting factors of Mycobacterium tuberculosis are required for virulence and 706 resuscitation from dormancy but are collectively dispensable for growth in vitro. Molecular 707 Microbiology. 2008 Feb;67(3):672–84. 708 78. Becker K, Sander P. Mycobacterium tuberculosis lipoproteins in virulence and immunity - 709 fighting with a double-edged sword. FEBS Lett. 2016 Nov;590(21):3800–19. 710 79. Schwebach JR, Glatman-Freedman A, Gunther-Cummins L, Dai Z, Robbins JB, Schneerson R, 711 et al. Glucan Is a Component of the Mycobacterium tuberculosis Surface That Is Expressed In 712 Vitro and In Vivo. Infect Immun. 2002 May;70(5):2566–75. 713 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 24 80. Leisching G, Pietersen RD, Wiid I, Baker B. Virulence, biochemistry, morphology and host-714 interacting properties of detergent-free cultured mycobacteria: An update. Tuberculosis. 2016 715 Sep;100:53–60. 716 81. Sala A, Calderon V, Bordes P, Genevaux P. TAC from Mycobacterium tuberculosis: a paradigm 717 for stress-responsive toxin–antitoxin systems controlled by SecB-like chaperones. Cell Stress 718 and Chaperones. 2013 Mar;18(2):129–35. 719 82. Lee MH, Kim HL, Seo H, Jung S, Kim BJ. A secreted form of chorismate mutase (Rv1885c) in 720 Mycobacterium bovis BCG contributes to pathogenesis by inhibiting mitochondria-mediated 721 apoptotic cell death of macrophages. J Biomed Sci. 2023 Dec 18;30(1):95. 722 83. Perez-Riverol Y, Csordas A, Bai J, Bernal-Llinares M, Hewapathirana S, Kundu DJ, et al. The 723 PRIDE database and related tools and resources in 2019: improving support for quantification 724 data. Nucleic Acids Research. 2019 Jan 8;47(D1):D442–50. 725 726 727 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 25 Figure 1. Acid stress promotes the formation of viable but non-replicating M. tuberculosis. A. 728 M. tuberculosis ::pTiGc will constitutively express a GFP (cell viability marker) and express 729 TurboFP635 when cultured with theophylline (inducer). Upon removal of theophylline, VBNR or 730 slowly replicating M. tuberculosis can be identified through retention of the red fluorescent signal. B. 731 M. tuberculosis S169::pTiGc was cultured to an OD 600 ~ 1 prior to transferring into either pH6,5 or 732 pH4,5 media for 48 hours. Aliquots of cultures were collected before and after stress for imaging 733 flow cytometry. Cell pellets were collected for cell lysate proteomics and culture supernatants were 734 collected for culture filtrate proteomics. C. M. tuberculosis S169::pTiGc was cultured with 735 theophylline and a high intensity of red fluorescence was detected (orange). Following the removal 736 of the inducer, actively replicating (pH6,5 cultures) bacteria had a reduction in red fluorescence 737 (green), however, pH stressed cultures retained a high red fluorescence intensity (red). Created with 738 BioRender.com. 739 Figure 2. Proteins identified in the cell lysates of actively replicating and VBNR enriched cell 740 lysates. A. Volcano plot representing differential protein abundance for M. tuberculosis S169 cell 741 lysate proteins. Volcano plot generated by FragPipe-Analyst using protein abundance data for 742 proteins identified in the cell lysates of actively replicating and acid stressed M. tuberculosis S169. 743 Following the removal of contaminant proteins and proteins only identified in one biological 744 replicate, 77 proteins were more abundant, and 269 proteins were less abundant in pH 4.5 cell lysates 745 when compared to pH 6.5 cell lysates (adjusted p-value 2). B. Gene ontology 746 enrichment of proteins significantly less abundant in the cell lysates of acid stressed M. tuberculosis 747 S169. GO enrichment analysis revealed that GO terms associated with universal stress proteins were 748 enriched in the 269 proteins found to be significantly less abundant in the cell lysates of acid stressed 749 M. tuberculosis S169 when compared to that of actively replicating M. tuberculosis S169. 750 Figure 3. Identification of culture filtrate proteins from pH 6.5 and pH 4.5 cultures. The 751 diagram demonstrates the data analysis of DDA mass spectrometry data for M. tuberculosis S169 752 culture filtrates. Following automated database searching, 461 protein groups were identified in at 753 least two of the three biological replicate experiments in culture filtrate recovered from pH 6.5 and 754 pH 4.5 cultures. Of these, 192 protein groups were only identified in pH 6.5 cultures filtrates and 45 755 proteins were only identified in pH 4.5 culture filtrates. For proteins identified under both test 756 conditions, a paired student t-test with a Benjamini-Hochberg correction of 0,05 was used to identify 757 83 differentially abundant proteins (log2 FC >/<0,05). In total we identified 275 protein groups in the 758 culture filtrates of pH 6.5 culture and 128 protein groups in the culture filtrates of pH 4.5 cultures. 759 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint 26 Figure 4. Enriched pathways of proteins identified in the culture filtrates of actively replicating 760 and acid stressed M. tuberculosis S169. GO enrichment analysis confirmed the enrichment of 761 proteins from the extracellular region through enrichment of pathways secreted and extracellular 762 region. A. GO enrichment analysis of proteins from actively replicating M. tuberculosis S169 763 revealed the enrichment of pathways for carbon metabolism, carboxylic acid metabolism, and 764 oxoacid metabolic process. B. Proteins identified in the extracellular fraction of acid-stressed M. 765 tuberculosis S169 revealed the enrichment of pathways for protein folding, lipoprotein, zymogen and 766 enzyme binding, and protease. 767 768 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2025. ; https://doi.org/10.1101/2025.10.03.680234doi: bioRxiv preprint