Decision letter: Biosensor-integrated transposon mutagenesis reveals rv0158 as a coordinator of redox homeostasis in Mycobacterium tuberculosis

Full text Figures and data Side by side Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Mycobacterium tuberculosis (Mtb) is evolutionarily equipped to resist exogenous reactive oxygen species (ROS) but shows vulnerability to an increase in endogenous ROS (eROS). Since eROS is an unavoidable consequence of aerobic metabolism, understanding how Mtb manages eROS levels is essential yet needs to be characterized. By combining the Mrx1-roGFP2 redox biosensor with transposon mutagenesis, we identified 368 genes (redoxosome) responsible for maintaining homeostatic levels of eROS in Mtb. Integrating redoxosome with a global network of transcriptional regulators revealed a hypothetical protein (Rv0158) as a critical node managing eROS in Mtb. Disruption of rv0158 (rv0158 KO) impaired growth, redox balance, respiration, and metabolism of Mtb on glucose but not on fatty acids. Importantly, rv0158 KO exhibited enhanced growth on propionate, and the Rv0158 protein directly binds to methylmalonyl-CoA, a key intermediate in propionate catabolism. Metabolite profiling, ChIP-Seq, and gene-expression analyses indicate that Rv0158 manages metabolic neutralization of propionate toxicity by regulating the methylcitrate cycle. Disruption of rv0158 enhanced the sensitivity of Mtb to oxidative stress, nitric oxide, and anti-TB drugs. Lastly, rv0158 KO showed poor survival in macrophages and persistence defect in mice. Our results suggest that Rv0158 is a metabolic integrator for carbon metabolism and redox balance in Mtb. Editor's evaluation This manuscript is important as it describes the powerful combination of TnSeq approaches with a reporter for mycothiol redox potential to identify genes in Mycobacterium tuberculosis that are required to maintain the reducing environment of the cell. Through this study, the authors provide compelling data that Rv0158 functions as a critical regulator of bacterial responses to endogenous reactive oxygen species. This study will be important to investigators interested in bacterial physiology and how this can impact pathogenesis. https://doi.org/10.7554/eLife.80218.sa0 Decision letter eLife's review process Introduction Mycobacterium tuberculosis (Mtb) is the etiological agent of tuberculosis (TB), a disease that results in approximately 1.6 million deaths annually (Global tuberculosis report, 2022). Two major concerns: the emergence of drug-resistant strains and few anti-TB drugs in the pipeline, necessitate the discovery of novel drug targets. Mtb requires oxygen (O2) for growth (Wayne and Hayes, 1996; Boshoff and Barry, 2005). Low O2 levels, as encountered in hypoxic granulomas, trigger a transition from actively growing to a non-replicating persistent state in Mtb (Boshoff and Barry, 2005). While this metabolic adaptation is an important contributor toward tolerance against host-induced stressors and anti-TB drugs, it obstructs the completion of the Mtb life cycle (Ernst, 2012). Therefore, for successful progression and subsequent disease transmission, Mtb requires O2. The immediate consequence of an aerobic lifestyle is the generation of endogenous reactive oxygen species (eROS) as metabolic by-products when O2 adventitiously interacts with and extracts an electron from quinones, flavins, and transition metal centres distributed throughout the cell (Imlay, 2013; Imlay, 2009; Imlay, 2003). eROS can cause oxidative damage to DNA, proteins, and lipids, driving the bacteria to deploy detoxification and repair machinery to keep eROS-induced deleterious effects below a threshold level (Imlay, 2009). While multiple reports investigate this cellular requirement (Ayer et al., 2012; Brynildsen et al., 2013; Radzinski et al., 2018), an understanding of global pathways that keep Mtb- eROS levels in check, is absent. The primary niche of Mtb is inside phagocytic cells, where the bacteria are exposed to superoxide (O2.-), nitric oxide (NO), their reaction product peroxynitrite (ONOO.), and hypochlorite stress (Ehrt and Schnappinger, 2009; Philips and Ernst, 2012). In response to host-induced redox stress, Mtb deploys antioxidant enzymes catalase-peroxidase (KatG), superoxide dismutase (Fe-SodA), thioredoxin reductase (TrxB2), alkyl hydroperoxide reductase (AhpC), and redox buffers- thioredoxin (trxC), and ergothioneine (Table 1). In addition to this, Mtb maintains a reduced cytosol by producing mycothiol (MSH; a low-molecular-weight sugar thiol) in millimolar amounts (Kumar et al., 2011). Interestingly, the cytoplasmic MSH-mediated redox buffering is intricately linked to the radical detoxification system by a membrane-associated oxidoreductase complex (MRC; SodA, DoxX, and SseA) in Mtb (Nambi et al., 2015). Disruption of any MRC-component impedes mycothiol recycling, increases oxidative damages, and enhances susceptibility to thiol stress (Nambi et al., 2015). Furthermore, Mtb expresses virulence factors that modulate host immune signalling and ROS- generating machinery to reduce the concentrations of ROS encountered by the bacteria inside the host (Chai et al., 2020; Köster et al., 2017). Additionally, the thick cell envelope consisting of cyclopropanated mycolic acids, phthiocerol dimycocerosates (PDIM), lipoarabinomannan, and phenolic glycolipid I (oxygen radical scavengers) confers an excellent anatomical barrier to and detoxification of exogenous oxidants (Hatfull and Jacobs, 2014; Tyagi et al., 2015). Also, sulfur metabolism pathways (Hatzios and Bertozzi, 2011; Kunota et al., 2021), RHOCS (a redox homeostatic system; PknG, ribosomal protein L13, and RenU) (Wolff et al., 2015), DNA repair, and redox sensors (two-component systems such as DosR/S/T, WhiB family of transcriptional factors Hatfull and Jacobs, 2014; Kumar et al., 2011, and SigH-AosR/RshA Khan et al., 2021) allow Mtb to adapt and survive in response to exogenous redox stressors. While Mtb has evolutionarily adapted to counteract the host-generated exogenous redox stress, the pathogen is markedly sensitive to cell-permeable molecules that artificially elevate ROS and RNI inside the bacteria (Singh et al., 2008; Tyagi et al., 2015). For example, Mtb retains viability under high millimolar (mM) concentrations of exogenous ROS and RNI (Voskuil et al., 2011). However, it displays exceptional sensitivity to low micromolar (μM) concentrations of endogenous superoxide (2, 3-Dihydro-1,4, naphthoquinone) (Tyagi et al., 2015) and NO (PA-824) donors (Singh et al., 2008). In this context, PA-824 or Pretomanid, which releases NO inside Mtb, is lethal against both replicating and non-replicating Mtb (Singh et al., 2008) and has been approved to treat Multi-drug resistant (MDR) tuberculosis as BPaLM regimen (World Health Organization, 2022). Taken together, these data indicate that disrupting intracellular redox homeostasis by enhancing eROS levels can be a promising strategy for identifying novel targets to kill Mtb. Table 1 List of redox-related proteins in Mtb which are either membrane bound or secreted out into periplasm/ extracellular space. Protein/small MW thiolsRv IDReferences1KatGRv1908cAntioxidant enzymeBraunstein et al., 2003; Tucci et al., 20202SodARv3846Braunstein et al., 2003, Vargas-Romero et al., 2016, Tucci et al., 20203AhpCRv2428Nieto R et al., 2016, Tucci et al., 20204TpxRv1932Probable thiol peroxidaseTucci et al., 20205TrxB2Rv3913Thioredoxin reductaseWong et al., 20186TrxCRv3914ThioredoxinWong et al., 2018, Tucci et al., 20207Rv0526Possible thioredoxinKe et al., 20188ThiXRv0816Probable thioredoxinKe et al., 20189ErgothioneineSao Emani et al., 2013 Genetic determinants of redox homeostasis can coordinate the response to anti-TB drugs inside macrophages (Mishra et al., 2019; Mishra et al., 2017). Using a genetic biosensor (Mrx1-roGFP2) of MSH redox potential (EMSH), we found that heterogeneity in EMSH of Mtb drives multidrug tolerance inside macrophages (Bhaskar et al., 2014; Mishra et al., 2019). Interestingly, several central metabolic enzymes (e.g., isocitrate lyase; Icl) and redox regulators are required to maintain EMSH, ergothioneine biogenesis, NAD+/NADH poise, and to counter the lethal action of antibiotics in Mtb (Chawla et al., 2012; Mishra et al., 2017; Nandakumar et al., 2014; Saini et al., 2016; Singh et al., 2009). Indeed, the link between metabolism and antibiotic-mediated killing is further substantiated by studies showing that sensitivity to anti-TB drugs (bedaquiline, Q203, moxifloxacin, and clofazimine) is influenced by mycobacterial respiration and carbon catabolism (Lamprecht et al., 2016; Mackenzie et al., 2020; Shee et al., 2022). It appears that Mtb repurposes pathways involved in metabolism and redox homeostasis to mitigate immune and antibiotic pressures. These findings underscore the need for a comprehensive understanding of redox homeostasis in Mtb. In this work, we repurposed transposon mutagenesis- deep sequencing (TnSeq) by combining it with the Mrx1-roGFP2 redox biosensor to discover novel pathways controlling basal EMSH of Mtb: –280±5 mV (Figure 1A). We built a Himar1-based saturated Tn mutant library (Sassetti et al., 2003; Sassetti et al., 2001; Sassetti and Rubin, 2003) in Mtb expressing Mrx1-roGFP2 (Mtb-roGFP2) and exploited FACS to reliably separate mutants displaying a shift from basal to oxidative EMSH during aerobic growth. We then developed a next-generation sequencing (NGS) and analysis pipeline for constructing a system-level network of genetic factors essential for maintaining redox homeostasis under aerobic growth in Mtb. We established the effectiveness of our strategy by discovering the role of an unknown transcriptional regulator encoded by rv0158 in integrating the metabolism of carbon source with redox balance and bioenergetics in Mtb. Figure 1 with 3 supplements see all Download asset Open asset Detection and enrichment of EMSH-oxidized Mtb-Tn mutants under in vitro growing conditions using FACS. (A) A saturated transposon mutant library in Mtb -roGFP2 was generated using a genetically engineered, temperature-sensitive mycobacteriophage pHAE180. Transposon mutants of Mtb were selected on 7H11 agar containing hygromycin and kanamycin. Genomic DNA was isolated and TraDIS (TnSeq) protocol was utilized to identify the transposon insertion site within the Mtb genome in the library of Mtb-transposon mutants. Mtb and Mtb Tn- Library expressing Mrx1-roGFP2 were grown in 7H9 supplemented with OADC and analyzed by flow-cytometry by exciting with 405 and 488 nm lasers at a constant emission (510 nm). The program BD FACS-Suite software was used to analyze the population distribution of bacteria, and a unique colour represented each population. EMSH-basal, EMSH-reduced, and EMSH-oxidized subpopulations are shown in red, blue, and yellow, respectively. FACS Dot plot of (B) untreated Mtb expressing Mrx1-roGFP2 grown in vitro is shown in red. (C) Mtb cells treated with an oxidant- 10 mM cumene hydroperoxide (CHP) (shown in yellow). (D) Tn-Library (Tn Lib). The subpopulation indicated in blue is the EMSH-reduced population as determined by Mtb cells treated with 20 mM dithiothreitol. (E) The EMSH-oxidized subpopulation (≈ 4%; EMSH OXD) was isolated by sorting, regrown on 7H11 agar, and resorted for three cycles to obtain EMSH-oxidized Tn mutants. 10,000 events per sample were analyzed. (F) The calculated EMSH of WT Mtb, input Tn Lib, and flow-sorted EMSH-oxidized Tn mutants. The data are means ± SEM of two independent experiments (n=6). (p>0.05: ns, p<0.0001: **, one-way ANOVA with Tukey’s multiple comparisons test). Figure 1—source data 1 FACS dot plots and EMSH numerical values of WT Mtb, input Tn Lib, and flow-sorted EMSH-oxidized Tn mutants. https://cdn.elifesciences.org/articles/80218/elife-80218-fig1-data1-v2.zip Download elife-80218-fig1-data1-v2.zip Results Himar1-based transposon mutagenesis in Mtb-roGFP2 A library of transposon mutants (Tn-Library) comprising >105 independent insertion events was generated in Mtb-roGFP2 (Figure 1A). The pool was allowed to grow to mid-log-phase (OD600 nm = 0.6–0.8), and genomic DNA was isolated for deep sequencing of transposon insertion (TnSeq) by Illumina technology. Using Transposon-directed insertion-site sequencing (TraDIS; Barquist et al., 2016), we identified the location of Tn insertion sites in the Tn-Library and detected insertion at ≈ 66% (49623 out of 74604) of TA sites (Supplementary file 1). Next, we quantified the frequency of Tn insertions at each open-reading frame (ORF) of Mtb (Supplementary file 1). Eighty-five percent of coding-DNA sequences incorporated at least 10 Tn insertion events in our Tn-Library. Consistent with previous studies, we observed a lack of Tn insertion in the TA sites of essential ORFs (≈11%) in the Mtb genome (DeJesus et al., 2017). Therefore, we conclude that the density of Tn insertion reached near saturation in the Tn-Library. Flow-cytometry-based selection of Mtb-Tn mutants with altered basal EMSH We defined basal EMSH as the EMSH of mid-log phase Mtb grown in 7H9+ADS under standard aerobic culture conditions (180 RPM, 37 °C) (Figure 1B and F, Figure 1—figure supplement 1b–f). Mtb exhibited a basal EMSH of -278.2±4.7 mV (Figure 1F). We aimed to identify genes that, when mutated by transposon insertion, increase eROS and induce a shift in the basal EMSH of Mtb towards oxidative. The Tn-mutants displaying oxidative-EMSH were detected by quantifying the Mrx1-roGFP2 fluorescence ratio (excitation: 405 nm /488 nm; emission: 510 nm) of the library using flow-cytometry (Bhaskar et al., 2014). To sort out EMSH-oxidized Tn-mutants, we utilized Fluorescence-activated Cell Sorting (FACS). To set the gates for sorting EMSH-oxidized Tn-mutants by flow-cytometry, we treated wild-type (WT) Mtb H37Rv expressing Mrx1-roGFP2 with an oxidant, cumene hydroperoxide (CHP). Treatment with 10 mM CHP for 5 min maximally increased the 405/488 ratio (Bhaskar et al., 2014; Shee et al., 2022), indicating 100% oxidation of biosensor in WT Mtb (EMSH-oxidized; –240 mV; Figure 1C, Figure 1—figure supplement 1g). Next, we subjected the Tn-mutant library (without CHP treatment) to FACS and selected the fraction of mutants displaying an increase in the biosensor ratio (Figure 1D). The biosensor ratio of this untreated Tn-mutant fraction overlapped with the ratiometric increase induced in WT Mtb by CHP, confirming the reliability of the sorting strategy (Figure 1D). Therefore, our strategy allowed the gating and isolation of Tn-mutants that were basally oxidized without exposure to oxidants such as CHP. Using this approach, we consistently identified ≈ 4% of Tn-Library displaying oxidative-EMSH under aerobic culture conditions without exposure to exogenous oxidants (Figure 1D, Figure 1—figure supplement 1i). This mutant fraction was purified using FACS (Figure 1—figure supplement 1j). To enhance selectivity, sorted mutants from the first round were recultured on 7H11 agar plates, colonies were pooled, and re-sorted by flow-cytometry. This cycle was repeated thrice, improving the sensitivity and reliability of our selection by gradually enriching the minor fraction of EMSH-oxidized Tn-mutants with each passage to >90% (Figure 1E, Figure 1—figure supplement 2). As expected, the enriched fraction of Tn-mutants showed a higher 405/488 ratio (0.7–0.9), which corresponded to an oxidative EMSH of –249±6.2 mV (Figure 1F). Consistent with this, measurement using the ROS-sensitive dye CellROX Deep Red showed that EMSH-oxidized Tn-mutants accumulated ≈10-fold higher eROS than WT Mtb and the input pool of the Tn-Library (Figure 1—figure supplement 3a). In addition to this, co-treatment of EMSH-oxidized Tn mutants with antioxidant molecules (Shee et al., 2022; combination of catalase (17.5 U/mL)+thiourea (1 mM)+bipyridyl (250 μM)) significantly decreased eROS levels (Figure 1—figure supplement 3b). Thus, by combining Tn mutagenesis with a redox biosensor, we reliably isolated redox-altered mutants of Mtb at a genomic scale. The EMSH-oxidized Tn-mutant pool exhibited hallmarks of oxidative damage The oxidative EMSH and increased eROS of Tn-mutants imply that this mutant pool experiences oxidative stress despite favorable aerobic growth conditions. Moreover, while each mutant is likely to get oxidized to a different extent, we expect that there will be an overall increase in the hallmarks of oxidative stress in the EMSH-oxidized Tn-mutants fraction. To examine this, we assessed various oxidative stress markers, such as DNA damage, lipid peroxidation, and expression of antioxidant genes in EMSH-oxidized Tn-mutants. The 8-hydroxy deoxyguanosine (an oxidative product of deoxyguanosine David et al., 2007; 8-OH dG) levels and lipid peroxidative products were ≈two-fold higher in EMSH-oxidized Tn-mutants as compared to the Tn-Library and WT Mtb, indicating increased DNA and lipid damage (Figure 2A and B). As a positive control, we treated WT Mtb with CHP and found a similar increase in DNA and lipid damage (Figure 2A and B). Another signature of oxidative stress inside bacteria is the upregulation of oxidative stress-responsive genes. We found that antioxidant enzymes, such as ahpC, ahpD, trxB1, DNA repair protein alkA, a regulator of iron-sulfur cluster biogenesis sufR, and redox-stress responsive sigma factor sigH were ≈two- to eight-fold upregulated in EMSH-oxidized Tn-mutants (Figure 2C). Figure 2 Download asset Open asset EMSH-oxidized Tn mutants displayed signature of oxidative damage under in vitro growth conditions. Bacterial cultures were grown to log phase (OD600 nm ≈ 0.6) (A) 8-hydroxy 2-deoxy guanosine, a major product of oxidative DNA damage was quantified using DNA/RNA Oxidative Damage ELISA Kit (Cayman chemicals). (B) Cellular lipid-hydroperoxides generation was measured by FOX2 assay. WT Mtb treated with 500 μM of cumene hydroperoxide (CHP) for 2 hr is used as a positive control. The data are means ± SEM of two independent experiments (n=4). (p>0.05: ns, p<0.01: , p<0.001: *, p<0.0001: , one-way ANOVA with Tukey’s multiple comparisons test). (C) Realtime-quantitative PCR analysis showing increased basal expression of oxidative stress response genes. 16 S expression was used as a control. The dotted line indicates a 1.5-fold- change (cut-off value). (D) Mtb, Tn-Library (Tn Lib), and EMSH-oxidized Tn-mutants (EMSH-OXD Tn) were treated with a non-bactericidal concentration of CHP and the ratiometric response was measured Data of two independent experiments in comparisons are with to the WT Mtb. , (E) cultures were exposed to oxidants with indicated concentration for hr and was The data and are means ± SEM of two independent (p>0.05: ns, p<0.01: , p<0.0001: **, ANOVA with multiple comparisons test). Figure data 1 values used to plot the in Figure Download We assessed the response of EMSH-oxidized Tn-mutants to an exogenous We exposed WT Mtb expressing the and EMSH-oxidized Tn-mutants to a non-bactericidal concentration of CHP et al., to CHP in a but increase in the 405/488 ratio to 0.6) in WT Mtb and the of antioxidant machinery (Figure In EMSH-oxidized Tn-mutants showed a significantly and increase in the 405/488 ratio to Figure The in biosensor ratio indicated that EMSH-oxidized Tn-mutants an impaired to an and antioxidant The of intracellular redox homeostasis enhances susceptibility toward exogenous stress et al., Consistent with this, the EMSH-oxidized Tn-mutant pool was to CHP, (a and (an = hr of (Figure data indicate that EMSH-oxidized Tn-mutants of oxidative TraDIS analysis revealed genetic determinants of redox homeostasis under aerobic growth conditions in Mtb To the site of transposon insertions in redox-altered EMSH-oxidized Tn-mutants, the TraDIS protocol et al., was we used a analysis pipeline expression analysis on the et al., and of Data in et al., to identify 368 genes containing significantly Tn insertions in EMSH-oxidized Tn-mutants to the input Tn-Library change or Figure supplement file 2). We this set of genetic factors genes to be with redox homeostasis were of the (Figure our for of redox For example, genes involved in metabolism metabolism Kunota et al., 2021), thiol buffering antioxidant enzymes homeostasis sufR, et al., and respiration were in the EMSH-oxidized Tn-mutant pool (Figure Additionally, we that several genes to functions such as biogenesis oxidized DNA and lipid repair membrane and metabolism were of Mtb redoxosome (Figure file we that of genes in the redoxosome been directly or with redox in Mtb file Figure 3 with 2 supplements see all Download asset Open asset TraDIS analysis determinants of redox homeostasis in Mtb under standard aerobic culture conditions. (A) showing genetic factors regulating EMSH of Mtb on TraDIS These factors are on the and The of the indicates the enrichment of genes in the EMSH-oxidized Tn pool compared to the Tn The for the change is at the of the are to of to colour indicates change is (B) The EMSH of Mtb mutants transposon in the genes identified as of Mtb The mutants were grown to log phase nm ≈ 0.6) under standard aerobic growth and EMSH was The basal The data are means ± SEM of two independent experiments (n=4). (p>0.05: ns, p<0.01: , p<0.001: *, p<0.0001: , one-way ANOVA with multiple comparisons test). (C) of the network of genes and the are shown as and the genes are shown as The of the is to their The was generated using The Rv0158 in the has the a in the Figure data 1 EMSH numerical values of Mtb mutants transposon in the genes identified as of Mtb Download While redox to 368 genes of the redoxosome requires independent we our data by 10 genes that are of the redoxosome and EMSH of Mtb strains in these genes. These 10 mutants were generated by transposon mutagenesis in the Mtb from Mtb and Mtb H37Rv displayed basal EMSH to mV; Figure As expected, the EMSH of Mtb mutants displayed an oxidative shift of mV to mV and exhibited higher eROS and DNA damage as compared to WT Mtb (Figure Figure supplement 2). Therefore, we that the set of 368 genes a to the of Mtb factors in maintaining redox balance in Mtb. analysis identified rv0158 as an important regulator of redox balance in Mtb To the of our the of that Mtb in maintaining redox balance during aerobic growth is not similar studies been on et al., 2013; et al., 2021) and (Ayer et al., 2012; Radzinski et al., a to analyze the redoxosome and to discover a network of regulators responsible for maintaining basal EMSH in growing Mtb. To this we generated a network by integrating 368 redoxosome genes with the Mtb transcriptional network et al., Figure A similar has been used to identify transcriptional regulators drug tolerance et al., 2016), adaptation to or et al., 2008) and to et al., in Mtb. In our network of and a node a or a and a indicates the transcriptional of a by a the of genes and in the network can be into three levels with in the in the and genes in the (Figure the in the are responsive to (Kumar et al., 2011; et al., 2003; et al., 2022), regulator et al., 2003; et al., 2009; et al., et al., 2021), and protein that as to Mtb et al., 2019; et al., In of the of genes or has the by in the network (Figure We measured which the fraction of in the network that a A high indicates that the node is in maintaining the We found that the in the and compared to in the In we that rv0158 has the in the network and the in the (Figure These indicate that rv0158 is both a and a in the Furthermore, rv0158 is a of 2 out of the 3 and 2003) in the In rv0158 to in the of the with the to redox signalling in Mtb. indicates that rv0158 lipid metabolism and oxidative stress response in Mtb on our in we aimed to the of how rv0158 redox homeostasis in Mtb. We first the rv0158 in Mtb by an strategy and the by (Figure supplement 1). to findings with the Tn-mutant EMSH of rv0158 KO displayed oxidative shift mV; Figure accumulated eROS (Figure and superoxide radical (Figure exhibited increased DNA (Figure and lipid compared to WT Mtb and rv0158 (Figure The mutant showed a ≈ increase in NAD+/NADH ratio to WT Mtb and rv0158 (Figure and Figure with 1 supplement see all Download asset Open asset rv0158 KO experiences enhanced oxidative damage to and altered NAD+/NADH during aerobic growth. Bacterial cultures were grown to log phase nm ≈ 0.6) (A) The EMSH of WT Mtb rv0158 and rv0158 The basal The data are means ± SEM of two independent experiments (n=4). (p>0.05: ns, p<0.001: *, one-way ANOVA with multiple comparisons test). eROS and superoxide levels were quantified using FACS by the fluorescence of CellROX Deep Red dye