Mycobacterium tuberculosis requires SufT for Fe-S cluster maturation, metabolism, and survival in vivo

Iron-sulfur (Fe-S) cluster proteins carry out essential cellular functions in diverse organisms, including the human pathogen Mycobacterium tuberculosis (Mtb). The mechanisms underlying Fe-S cluster biogenesis are poorly defined in Mtb. Here, we show that Mtb SufT (Rv1466), a DUF59 domain-containing essential protein, is required for the Fe-S cluster maturation. Mtb SufT homodimerizes and interacts with Fe-S cluster biogenesis proteins; SufS and SufU. SufT also interacts with the 4Fe-4S cluster containing proteins; aconitase and SufR. Importantly, a hyperactive cysteine in the DUF59 domain mediates interaction of SufT with SufS, SufU, aconitase, and SufR. We efficiently repressed the expression of SufT to generate a SufT knock-down strain in Mtb (SufT-KD) using CRISPR interference. Depleting SufT reduces aconitase’s enzymatic activity under standard growth conditions and in response to oxidative stress and iron limitation. The SufT-KD strain exhibited defective growth and an altered pool of tricarboxylic acid cycle intermediates, amino acids, and sulfur metabolites. Using Seahorse Extracellular Flux analyzer, we demonstrated that SufT depletion diminishes glycolytic rate and oxidative phosphorylation in Mtb. The SufT-KD strain showed defective survival upon exposure to oxidative stress and nitric oxide. Lastly, SufT depletion reduced the survival of Mtb in macrophages and attenuated the ability of Mtb to persist in mice. Altogether, SufT assists in Fe-S cluster maturation and couples this process to bioenergetics of Mtb for survival under low and high demand for Fe-S clusters.

Introduction heightened requirement for Fe-S cluster biogenesis [24]. S. aureus SufT is dispensable for Fe-S maturation under standard growth conditions or repair of damaged Fe-S clusters [24]. Similar to this, an initial report on SufT of Mycobacterium smegmatis (Msm-non-pathogenic and fast grower) was dispensable under standard growth conditions, but essential during Fe-limitation [25]. In contrast, Mtb SufT is predicted to be essential for viability under standard growth conditions [17], suggesting that the function of SufT is distinct in Mtb as compared to S. aureus or Msm.
In this study, we performed an in silico examination of Mtb SufT and identified the role of a conserved cysteine residue in Fe-S cluster maturation. SufT participated in Fe-S cluster maturation and repair by physically associating with other Suf proteins and Fe-S cluster acceptor proteins in Mtb. We systematically depleted SufT and examined bacillary growth under ambient conditions, upon exposure to redox stress, and inside macrophages and mice. Lastly, we assessed the effect of SufT depletion on TCA cycle metabolites, amino acid metabolism, and bioenergetics of Mtb. Our findings provide functional insights on how SufT contributes to Fe-S cluster maturation and cellular homeostasis to promote survival of Mtb in vitro and in vivo.

In silico analysis indicates the role of SufT in Fe-S cluster maturation
The function of a large majority of DUF59-containing proteins remains unknown [26] except for a few DUF59 proteins involved in phenylacetic acid degradation and the maturation of iron-sulfur (Fe-S) cluster-containing proteins [24,[27][28][29]. The DUF59 domain-containing proteins display 9 modular structures (S1 to S9 modules in S1A Fig). The Mtb SufT DUF59 domain contains an extra N-terminal motif (S2 module in S1A Fig) that did not exhibit any homology to previously defined domains (S1A and S1B Fig). The DUF59 domain contains a conserved motif containing a hyperactive cysteine with heightened nucleophilicity (DPE-X26-31-T-X2/3-C) [30]. Both of these features are preserved in Mtb SufT (S1B Fig). The Mtb SufT NMR structure (PDB ID: 5IRD) shows a α/β-topology with five helices, and three-stranded, mixed parallel/anti-parallel β-sheets (Fig 1). The sequence arrangement of the regular secondary structures is α1-α2-β1-β2-α3-β3-α4-α5. We mapped the NMR structure of Mtb SufT on to the structure of proteins containing DUF59 domain in Bacillus subtilis YitW, (PDB ID: 3LNO), Thermotoga maritima TM0487 (PDB ID: 1UWD), and Homo sapiens FAM96a (PDB ID: 2M5H), and confirmed that the conserved D, E, T, and C residues are located commonly on solvent-exposed loops (Fig 1). However, while DET-associated hydroxyl side chains position towards one another to form a pocket with an average distance of~5Å between them in SufT of B. anthracis, T. maritima, and H. sapiens, the same pocket in Mtb SufT exhibits an average distance of 12-17Å (Fig 1). Similar to B. anthracis and T. maritima DUF59, but not the H. sapiens DUF59, the cysteine thiol (C62) clusters with the DET-associated hydroxyl group in Mtb SufT (Fig 1). Moreover, the strict conservation of proline in position 63 is likely to have a role in properly arranging C62 in the active site of Mtb SufT (Fig 1). The overall arrangement of D, E, T, C, and P residues can therefore create an environment to bind a Fe ion or a Fe-S cluster (Fig 1) [31], although, a well-recognized Fe-S cluster binding motif could not be identified. The structural feature of Mtb SufT, along with its organization within the suf operon, suggests that SufT might have a role in Fe-S cluster maturation.

Mtb SufT interacts with Fe-S cluster biogenesis proteins; SufS and SufU
Our in silico analysis suggests the importance of C62 and D, E, and T residues in creating a pocket for coordinating Fe or a Fe-S cluster in SufT. To examine this possibility, we purified histidine-tagged wild type (WT) SufT, or alanine mutant of the hyperactive cysteine residue (SufT C62A ) from E. coli cultured in growth conditions optimized for the maximum insertion of Fe or Fe-S clusters in vivo [16]. Purified SufT did not display the characteristic straw brown color associated with Fe or Fe-S cluster binding proteins. Agreeing to this, UV-visible spectroscopy did not reveal absorbance features associated with Fe (~420 nm) or Fe-S clusters (~340 nm The DUF59 containing protein (YHR112W) in yeast mediates transfer of Fe-S clusters to acceptor/client proteins by physically interacting with the multiple proteins involved in Fe-S cluster assembly [30]. Importantly, the highly conserved cysteine residue in YHR112W likely regulates the formation of biologically active protein complexes with the Fe-S cluster biogenesis factors [30]. On this basis, we examined the interaction of SufT WT and SufT C62A with other Suf proteins involved in Fe-S cluster biogenesis in Mtb. Previous studies have shown that Mtb . The side chains of the conserved D, E, T, and C (red encircled) residues are highlighted. Oxygen, sulfur, and nitrogen atoms are represented with red, yellow, and blue color, respectively. Distance in Angstrom (Å) is represented between the sidechain hydroxyl of the T, D, and E residues. A proline (P25 in Mtb) present in the motif as DPE (blue encircled) and the consecutive proline (P63 in Mtb) next to the C residue are represented in the Mtb SufT structure. https://doi.org/10.1371/journal.ppat.1010475.g001

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb SufT does not interact with other Suf proteins in the yeast-two-hybrid assays [11,32]. While surprising, these studies are consistent in that certain Mtb protein: protein interactions require a natural host environment rather than a non-native host such as yeast [33]. Therefore, we used a mycobacterial two-hybrid system (mycobacterial protein fragment complementation assay [M-PFC]) to examine interacting partners of SufT [34]. The M-PFC system has been successfully used to demonstrate interactions between Mtb proteins using Msm as a surrogate. The M-PFC assay is based on the proximity-based reconstitution of murine dihydrofolate reductase (mDHFR) activity due to interaction between two proteins fused to mDHFR fragments F [1,2] and F [3]. The reconstitution of mDHFR activity can be scored by detecting growth on a medium containing trimethoprim (TRIM), a drug that preferentially targets bacterial DHFR (Fig 2A). We tested the interaction of Mtb SufT with SufU and SufS, both of which are known to catalyze Fe-S cluster transfer to the client protein; the final step towards Fe-S cluster maturation. We co-expressed SufT WT or SufT C62A with SufU or SufS fused to the C-terminus of F [3] and F [1,2] mDHFR fragments in Msm (designated as SufT [F3] or SufT C62A[F3] and SufS [F1,2] or SufU [F1,2] ), respectively. Additionally, the previously reported yeast GCN4 interacting pair (GCN4 [F3] /GCN4 [F1,2] ) was taken as a positive control [34]. Msm co-expressing SufT [F3]/ SufS [F1,2] , SufT [F3] /SufU [F1,2] , SufT [F3] /SufT [F1,2] , and GCN4 [F1,2] / GCN4 [F3] combinations grew on 7H11 plates containing HYG, KAN and TRIM, indicating homodimerization of SufT and its interaction with SufS and SufU. In contrast, Msm coexpressing SufT [F3] /GCN4 [F1,2] did not show any growth on HYG/KAN/TRIM-containing medium. Interestingly, the substitution of Cys by alanine completely disrupted SufT homodimerization and association with SufS and SufU ( Fig 2B). The expression of SufT [F3] and SufT C62A[F3] was comparable in Msm (S2D Fig), suggesting that mutation of the cysteine residue resulted in the loss of interaction.

SufT physically associates with Fe-S cluster proteins aconitase and SufR
Having shown the interaction of SufT with SufS and SufU, we next asked if SufT directly associates with the proteins requiring Fe-S clusters for their function. To test this, we co-expressed SufT with 4Fe-4S cluster containing Mtb proteins Acn and SufR in Msm using the M-PFC vectors. We also co-expressed ESAT6 and CFP10 as these proteins associate strongly in mycobacteria [35]. The growth on HYG/KAN/TRIM containing medium was readily detected in the case of SufT [F3] /Acn [F1,2] , SufT [F3] /SufR [F1,2] , and ESAT6 [F1,2] /CFP10 [F3] , indicating interaction of SufT with Acn and SufR. Msm co-expressing SufT C62A with Acn or SufR did not grow on HYG/KAN/TRIM plates ( Fig 3A).
We further validated M-PFC findings by re-examining the interaction between SufT and Acn in Mtb using in vivo pull-down assays. To do this, we utilized SufT-proficient (wtMtb H37Rv) and a SufT-deficient strain of Mtb. Since Mtb SufT is proposed to be essential for the viability of Mtb [17], we generated a SufT-deficient strain using CRISPR-based RNA interference (CRISPRi) [36]. In CRISPRi, dCas9 and small guide RNA (sgRNA) expression was induced using anhydrotetracycline [ATc] from TetR-regulated promoter system. We first confirmed  Fig 3B). Using an anti-SufT antibody, we confirmed an~95% reduction in SufT levels at 48 h post-ATc treatment (Fig 3C).
We next performed immunoprecipitation of Acn from the cell-free lysate of the SufT-proficient control strain or the SufT-deficient (SufT-KD) strain. We observed comparable levels of Acn in control and the SufT-KD strain (Fig 3D, upper panel). The proteins pulled down from the lysate using the anti-SufT antibody were separated and probed with the Anti-Acn antibody by immunoblotting. A strong signal was obtained in the case of the control strain, whereas the signal intensity was markedly reduced (~4 fold) in the SufT-KD strain (Fig 3D lower panel). A similar Two independent plasmid constructs for candidate proteins X and Y are shown in fusion to complementary mDHFR fragments F [1,2] and F [3], respectively. Co-transformation of XF [1,2] and YF [3] fusions in Msm results in the functional reconstitution of mDHFR activity and subsequent growth on TRIM (0 and 25 μg/mL; concentration) plates, whereas proteins that do not interact will not reconstitute F [1,2] and F [3] and consequently no growth on TRIM plates. Components of vectors, aph confers resistance to KAN; hyg confers resistance to HYG; hsp60p is the hsp60 promoter; oriM is the origin of replication for propagation in mycobacteria; oriE is the origin of replication for propagation in E. coli; and int and attP are the integrase and phage attachment sites, respectively, from mycobacteriophage L5. (B) M-PFC based demonstration of C62 dependent specific protein-protein interaction of SufT with SufT, SufU and SufS. Msm as a host was co-transformed with plasmid construct expressing 1) SufT [

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb pull-down using a rabbit Isotype IgG antibody (IgG:Iso), yielded a low background signal of Acn ( Fig 3D lower panel), affirming our findings that SufT specifically interacts with Acn in Mtb. In sum, our data suggest that SufT-mediated Fe-S cluster maturation requires direct interaction with proteins involved in Fe-S cluster biogenesis and Fe-S cluster requiring client proteins.

SufT depletion reduces the activity of Fe-S cluster containing proteins
Next, we directly tested the role of Mtb SufT in Fe-S cluster maturation by measuring the activity of Acn, a known 4Fe-4S cluster-containing enzyme, in the SufT-KD strain. Acn activity was reduced to~50% in the SufT-KD strain upon treatment with various concentrations of ATc for 72 h (Fig 4A). Under these conditions, the Acn levels remain unaltered; indicating that the loss of activity may be related to reduced 4Fe-4S cluster assembly of Acn ( Fig 4A). We

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb have recently shown that Mtb SufR, in a 4Fe-4S cluster bound form, represses the expression of suf operon [16]. Loss of 4Fe-4S cluster from SufR de-represses the suf operon [16]. Since SufT interacts with SufR, we expect that SufT is required for maturation of 4Fe-4S cluster on SufR to maintain its repressor activity. To examine this, we measured the expression of suf operon in the SufT-KD strain. As expected, the expression of suf operon was induced in the SufT-KD strain (Fig 4B), indicating that abnormal Fe-S cluster assembly on SufR upon SufT depletion likely resulted in the de-repression of the suf operon. Mtb encodes a homolog of cysteine desulfurase IscS (Rv3025c) that can insert the Fe-S cluster in Mtb proteins independent of the suf system [2,37]. One possibility is that the depletion of SufT results in the reduced transcription of IscS system, which thereby diminishes the occupancy of the 4Fe-4S cluster on Acn and SufR. However, the transcription of iscS remained unaltered upon SufT depletion

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb ( Fig 4B). Lastly, the 4Fe-4S clusters of Acn and SufR are sensitive to ROS [38,39]. However, ROS in the SufT-KD strain remained comparable to the control strain (S4 Fig). Also, the expression of known antioxidant genes (sodA, katG, ahpC, and ahpD) was not affected in the SufT-KD strain (Fig 4B), clearly suggesting that defective Acn and SufR activities are not a consequence of Fe-S cluster damage due to heightened ROS. Based on the above results, we conclude that Acn and SufR activities are likely reduced due to decreased occupancy of the 4Fe-4S cluster upon SufT depletion.

SufT repairs damaged Fe-S clusters in Mtb
Since respiration-linked ROS damages Fe-S clusters of Acn during aerobic growth [3,24,39], the steady-state activity of Acn reflects the equilibrium between the inactive (Fe-S cluster damaged) and active fraction of the enzyme [39]. The addition of ferrous ions (Fe[II]) shifts the equilibrium towards active fraction, whereas Fe(II) chelator dipyridyl had an opposite effect [40]. We asked if SufT is required to maintain the balance between the active and inactive fraction of Acn in Mtb. Acn activity was monitored in the cell-free extract of aerobically growing SufT-KD strain in the presence or absence of Fe(II). The addition of Fe(II) increased the activity of Acn in the control strain by 2-fold ( Fig 4C), indicating that the enzyme is partially operational (50%) under steady-state conditions. Inclusion of Fe(II) increased the Acn activity by 3-fold in the SufT-KD strain, which amounted to Acn activity of only 30% upon under steadystate conditions ( Fig 4C). Furthermore, the addition of 500 μM of dipyridyl (DIP) further decreased the steady-state activity of Acn by 22% and 40% in the control and the SufT-KD strains as compared to untreated cells, respectively ( Fig 4C). This is consistent with an equilibrium shift towards inactive Acn fraction (lesser 4Fe-4S cluster) in response to SufT depletion under aerobic growing condition.
Transient exposure to H 2 O 2 converts the active [4Fe-4S] 2+ form of Acn to inactive [3Fe-4S] 1+ form [39]. This can be repaired to the [4Fe-4S] 2+ active state by the cellular Fe-S cluster biogenesis machinery upon mitigation of H 2 O 2 stress [24]. We exposed the cell-free lysate of 48 h ATc treated control and SufT-KD strains to 450 μM of H 2 O 2 for 1 min and Acn activity was measured. In both cases, Acn activity was reduced by 50% in comparison to untreated cell lysates ( Fig 4D). However, removal of H 2 O 2 by catalase fully restored the Acn activity in the cell free lysates of control strain, but not in the lysate of SufT-KD strain ( Fig 4D). Above findings suggest that SufT is required to maintain the steady-state activity of Acn and repair H 2 O 2 damaged Fe-S clusters in Mtb.

SufT depletion results in widespread metabolic realignment in Mtb
Since Fe-S cluster proteins are central to carbon metabolism, amino acid, and nucleotide biosynthesis [41], we assessed the influence of SufT depletion on these metabolites in Mtb. We utilized targeted, quantitative liquid chromatography-mass spectrometry (LC-MS/MS) based approaches (as described earlier [42] to analyze the steady-state amounts of amino acids, nucleotides and tricarboxylic acid cycle intermediates, in SufT-KD strain with and without ATc. Consistent with the reduced activity of Acn, the quantity of the Acn substrate (citrate) was elevated, and the product (α-ketoglutarate) was decreased in the SufT-KD strain ( Fig 5A). Similarly, the substrate (fumarate) and the product (malate) of another 4Fe-4S cluster containing fumarase were increased and decreased, respectively (Fig 5A).
Accumulation of succinate indicates reduced activity of succinate dehydrogenase (SDH), which could be due to impaired maturation of Fe-S cluster on SDH in the SufT-KD strain ( Fig  5A). A Fe-S cluster enzyme, lipoate synthase, provides lipoic acid to activate the E2 subunits of pyruvate dehydrogenase complex (PDHC) and H protein of the glycine cleavage system (GCS)

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb [43]. Accumulation of glycine and pyruvate, which are the substrates of GCS and PDHC, respectively, indicates impaired lipoate synthase activity upon SufT depletion (Fig 5A). Similarly, valine and isoleucine that are catabolized by another lipoic-acid dependent enzyme (branched-chain keto-acid dehydrogenase complex [BCKDC]) [44] were accumulated in the SufT-KD strain (Fig 5B). The depletion of TCA cycle intermediates, along with the abundance of some amino acids (e.g., Arg, Lys, Ser, Thr, Prol, and Phe) ( Fig 5B) suggest a compensatory mechanism to balance cataplerosis (removal of TCA cycle intermediates) by anaplerosis in the SufT-KD. Likewise, the sulfur-containing metabolites [e.g., cystathionine (CTH), SAH, and SAM] were reduced, indicating disturbed sulfur metabolism in the SufT-KD. Due to relatively unstable nature of cysteine and methionine, we only measured CTH, SAM, and SAH as representative of sulfur amino acids. Further experiments are needed to fully understand the role of

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb SufT in sulfur metabolism. Lastly, we observed glutamine accumulation along with a decrease in the purine nucleotides in the SufT-KD strain (Fig 5A and 5B). This could be due to the reduced activity of a 4Fe-4S cluster enzyme glutamine phosphoribosylpyrophosphateamidotransferase (GPATase), which uses glutamine for de novo purine nucleotide biosynthesis [43,45,46]. Overall, data indicate that SufT is required for maintaining Fe-S homeostasis in Mtb in order to support carbon, amino acid, and nucleotide metabolism.

SufT-depleted Mtb is bioenergetically deficient
A major realignment of central carbon metabolism upon SufT-depletion suggests altered bioenergetics. We utilized extracellular flux (XF) analyzer to noninvasively track the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) of the SufT-depleted strain. ECAR is an indication of proton (H + ) translocation linked to glycolysis and TCA cycle, while OCR is a reliable indicator of respiration due to oxidative phosphorylation (OXPHOS).
To quantify the basal and maximum rates of OCR and ECAR, we cultured Mtb in the XF microchamber, exposed to glucose, and then to the uncoupler carbonyl cyanide m-chlorophenyl hydrazine (CCCP) (Fig 6A). The addition of CCCP stimulates respiration to the maximal capacity manageable by Mtb. The difference between basal and CCCP induced OCR provides an estimate of the bioenergetics reserve, a capacity crucial for survival under redox stress [47]. Control and SufT-KD strains were grown in standard 7H9 media and treated with or without ATc for 72 h to deplete SufT. For measuring OCR and ECAR, cultures were diluted, and an equal number of SufT-KD and control cells (2x10 6 cells/well) adhered to XF cell culture plates as described in Materials and Methods. In the presence of glucose, the SufT-KD strain had a significantly lower OCR (57%) and ECAR (59%) than the control strain (Fig 6A and 6B). The addition of CCCP (5 μM) stimulated OCR and ECAR in both the strains however, the increase was significantly lower in the SufT-depleted strain (Fig 6A and 6B). The SufT-KD strain showed a 31% reduction in bioenergetics reserves (Fig 6C), exposing a likely vulnerability to redox stresses. By plotting the OCR and ECAR data, we generated a metabolic phenogram, which clearly showed that the SufT-depleted strain is metabolically derailed (Fig 6D). In sum, these data collectively implicate SufT in maintaining the bioenergetic homeostasis of Mtb.

Mtb SufT is required for growth and promotes survival under oxidative stress, nitric oxide exposure, and inside macrophages
We first examined the phenotypic consequence of SufT depletion on growth of Mtb in liquid broth cultures exposed to various concentrations of ATc over time. As shown in (Figs 7A and S5A), increasing the concentration of ATc uniformly reduces the growth of the SufT-KD strain. Interestingly, on the solid agar medium, addition of 200 ng/mL of ATc nearly eliminated the growth of the SufT-KD colonies (Fig 7B). These results indicate the requirement of SufT for promoting growth under aerobic growing conditions.
Oxidative and nitrosative stress damage Fe-S clusters and increase the demand for Fe-S biogenesis and repair. Importantly, the suf operon showed heightened expression in response to hydrogen peroxide (H 2 O 2 ) and nitric oxide (NO) exposure (19,20). On this basis, we tested the requirement of SufT for adaptation of Mtb under H 2 O 2 and NO stress. Exposure to 10 mM H 2 O 2 resulted in~3-fold (p = 0.0017) decreased survival of the SufT-KD strain relative to unstressed bacteria (Fig 7C). However, a significant effect of NO on bacterial survival (2 .5-fold) (p = 0.0002) was only evident at 1mM of DETA/NO (Fig 7D). We next assessed the survival of SufT-KD inside macrophages. We infected PMA-differentiated THP-1 macrophages and resting RAW264.7 murine macrophages with the control and SufT-KD strains at multiplicity of infection (MOI) 2 and monitored the survival over time. In these models,

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb control and SufT-KD strains without the ATc inducer showed unrestrictive growth over time (Fig 7E and 7F). In contrast, ATc pretreatment for 72 h, prevented the proliferation of SufT-KD in THP-1 and RAW264.7 macrophages. Lastly, we treated RAW264.7 with the iNOS inhibitor 1400W [48] and examined the growth phenotype of the SufT-KD strain. The growth inhibition exhibited by the SufT-KD strain was partially reversed upon treatment of RAW264.7 with 1400W (S5B Fig). Taken together, these data suggest that SufT provides resistance to oxidative and NO stress and is required for survival of Mtb inside macrophages.

SufT is required for growth and persistence of Mtb in mice
We next sought to determine the impact of SufT depletion on the growth and persistence of Mtb in mice. The Tet-based regulatory system used in this study to deplete SufT has been

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb successfully utilized to analyze the requirement of essential genes in Mtb during acute and chronic phase of infection [49]. We infected BALB/c mice with the SufT-KD strain by aerosol. As previously reported [49], we initiated SufT depletion at day 7 post-infection to assess acute phase phenotype and for the chronic phase at day 21 post-infection by feeding doxycycline (DOX), a tetracycline derivative frequently used for gene silencing in vivo [50]. Wt Mtb proliferated for 21 days in mice followed by the onset of adaptive immunity resulting in the

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb stabilization of bacterial load in the lungs. The SufT-KD strain multiplied at a rate comparable to wt Mtb in the absence of DOX [49]. However, bacterial counts were reduced to 56-fold (p = 0.0073) at 21 days and 21-fold (p = 0.00079) at 42 days post-infection when SufT depletion was initiated during the acute (day 7) and chronic (day 21) stage, respectively (Fig 8A). The gross and histopathological changes in the lungs of the SufT-KD strain after 42 days of infection were comparable with the bacillary load observed (Fig 8B and 8C). The extent of pulmonary tissue damage was significantly greater in the SufT-KD strain in DOX untreated animals (score 5.6) relative to DOX treated animals (score 0.6) (Fig 8D). Since ATc/DOX addition sustains dCas9 expression only for a few days [36], we feel that a relatively smaller effect of SufT depletion on the survival and persistence of Mtb is likely due to decrease in dCas9 induction and SufT depletion over time. Altogether, these results confirm that SufT is necessary for Mtb's survival during all stages of infection in mice.

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb

Discussion
The presence of a DUF59 domain on Mtb SufT and its physical linkage with other suf genes strongly suggest a possible role in Fe-S cluster biogenesis. Here, we have presented exhaustive genetic, biochemical, and animal studies to show that DUF59 containing Mtb SufT physically interacts with Fe-S cluster biogenesis proteins and Fe-S cluster proteins to coordinate metabolism, bioenergetics, growth, and stress survival in vitro and in vivo. The physiological deficiencies manifested by the SufT-depleted strain are indicative of defects in Fe-S cluster maturation, as described for other DUF59 containing SufT homologs [24,27,51,52]. For example, DUF59 domain-containing proteins from Arabidopsis thaliana (e.g., HCF101, AE7), humans (Fam96a, Fam96c), S. aureus (SufT), and Saccharomyces cerevisiae (CIA2A and CIA2B) are involved in the maturation of Fe-S clusters to control central metabolism, respiration, photosynthesis, and growth.
However in contrast to the major requirement of other SufT homologs for de novo Fe-S cluster assembly upon apo-protein [24,27,29,51,52], Mtb SufT seems to also protect Fe-S clusters from leaching (e.g., dipyridyl) and reactivation of H 2 O 2 -damaged Fe-S clusters. While the mechanism of SufT mediated protection, maturation, and regeneration of Fe-S clusters needs further experimentation, data showing a direct interaction between SufT, SufU and SufS suggest its function in the carriage and transfer of Fe-S clusters to the client proteins during assembly. In line with this, DUF59 containing A. thaliana HCF101 binds 4Fe-4S cluster and transfers the cluster to an acceptor apo-protein akin to an Fe-S cluster carrier [51]. However, Mtb SufT contains only one strictly conserved cysteine, whereas Fe-S carriers generally contain two or more cysteine residues for Fe-S cluster coordination [31]. Consistent with this, we found that Mtb SufT does not bind Fe or Fe-S cluster, rather the conserved cysteine residue is essential for establishing interaction of SufT with SufS, SufU, Acn, and SufR. A likely possibility is that the association of SufT with Acn and SufR might be required for the physical elimination of Fe-S cluster damaging agents (e.g., H 2 O 2 , dipyridyl, NO) from the Acn and SufR active site.
In addition to the DUF59 domain-containing proteins, organisms encode multiple auxiliary factors for Fe-S cluster maturation under prevailing environmental conditions [53][54][55][56]. For example, both SufT and Nfu perform Fe-S maturation of Acn in S. aureus [24]. However, Nfu is preferred under normal growth conditions, whereas SufT prevails under stress conditions [24]. Homologs of accessory proteins performing Fe-S maturation have not been identified in Mtb, indicating a broader role for Mtb SufT in Fe-S cluster homeostasis under ambient and stress conditions. Consistent with this premise, the SufT-depleted strain not only displayed defective growth in standard culture media but also showed poor survival under highly oxidative or nitrosative conditions. In the absence of multiple auxiliary factors, Mtb likely responds to a graded demand for Fe-S cluster biogenesis by regulating the expression of sufT via the transcriptional regulator SufR [16]. Under standard growth conditions, SufR coordinates a 4Fe-4S cluster and maintains the basal expression of the suf operon [15]. Stress conditions (e.g., NO, ROS) damage the SufR Fe-S cluster resulting in a loss of repressor function and an induction of the suf operon, including sufT [19].
The physiological role of SufT in Fe-S cluster maturation is supported by the metabolite analysis, which shows impaired pool of TCA cycle intermediates, and accumulation of substrates utilized by the lipoamide-requiring enzymes (PDH, BKC, GCS). The synthesis of lipoic acid requires Mtb LipA, which utilizes two [4Fe-4S] clusters for catalysis [57]. Our data suggest that multiple phenotypes associated with SufT-depletion could be a consequence of decreased activities of the lipoate-requiring enzymes. Importantly, S. aureus SufT contributes to Fe-S

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb maturation under conditions of high demand for lipoic acid [58]. Further experiments are needed to confirm if SufT is involved in the maturation of 4Fe-4S clusters of LipA in Mtb.
Cysteine is the source of sulfur required for the Fe-S cluster formation. Mtb preferentially utilizes methionine to generate cysteine via methionine cycle intermediates (SAM, SAH, homocysteine, and cystathionine) using the reverse transsulfuration pathway [59]. Since the accumulation of cysteine is toxic in Mtb [60], it is utilized to form mycothiol [61], and also consumed by cysteine desulfurases (IscS and SufS) to release sulfane sulfur for Fe-S biogenesis [37]. Since methionine and SAM are essential metabolites [62], they are regenerated from cysteine and homoserine via the transsulfuration and homoserine transacetylase pathways, respectively [59,62]. Other than methionine, inorganic sulfate can be assimilated to produce cysteine by the action of a Fe-S cluster-dependent enzyme APS reductase (APSR) encoded by cysH [60]. The balance between these pathways likely ensures the balance of methylation, antioxidant, and Fe-S cluster homeostasis in Mtb [60,62,63]. Our data provide new insight into this important metabolic branch point. Expectedly, the levels of TCA cycle metabolites and amino acids linked to the activity of Fe-S cluster enzymes were altered upon SufT-depletion. However, an unexpected finding was the diminished amounts of SAM, SAH, and cystathionine in the SufT-depleted strain. This could well be a consequence of the unregulated induction of the suf operon upon SufT-depletion. Due to defective Fe-S maturation of the repressor SufR, the SufT-KD strain constitutively overexpresses a cysteine desulfurase (sufS), which release sulfur from cysteine for Fe-S biogenesis. This increase in sufS likely results in increased utilization of cysteine for repairing Fe-S clusters upon SufT-depletion. A consequence of this would be the net loss of SAM, SAH, and cystathionine, which also depend on the methionine and cysteine pool via the methionine cycle and transsulfuration pathways. Moreover, defective Fe-S maturation would have rendered the Fe-S cluster-dependent APSR pathway an unlikely route to generate cysteine, further raising the dependency of the SufT-KD strain on the reverse transsulfuration for cysteine formation. All of this could lead to an exhaustion of SAM, SAH, and cystathionine in the SufT-KD strain. Because methionine and SAM are critical for host infection and the virulence of Mtb [62], the malfunctioning of the SufT-KD strain in vitro and in vivo is likely due to the collapse of multiple mechanisms associated with Fe-S cluster and methionine metabolism. This observation is consistent with the fact that the activity of several Fe-S cluster dependent enzymes are also contingent on SAM (e.g., lipA, methyltransferase) [57,64].
In summary, in this study, we used in vivo and in vitro approaches to define the role of DUF59 containing SufT in the physiology and virulence of Mtb. We used a SufT-depleted strain as a model system to examine the consequence of defective Fe-S metabolism on growth, stress response, metabolome, bioenergetics, and pathogenesis. In doing so, we discovered a previously unknown link between the Fe-S cluster and methionine metabolism in Mtb. The study presented provides a framework for future studies examining the link between Fe-S homeostasis and pathogenesis of human pathogens such as Mtb.

Ethics statement
Animal experimentation. This study was carried out in strict accordance with the guidelines provided by the Committee for the Purpose of Control and Supervision on Experiments on Animals (CPCSEA), Government of India. The protocol of animal experiment was approved by animal ethical committee on the Ethics of Animal Experiments, Indian Institute of Science (IISc), Bangalore, India (Approval number: CAF/Ethics/544/2017). All efforts were made to minimize the suffering.

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb

Strain and plasmid construction
For construction of SufT-KD strain, CRISPR interference (CRISPRi) technology was used as described [36]. An inactive version of Streptococcus pyogenes cas9 having mutations D10A and H820A (dcas9) was expressed by pRH2502 from a TetR-regulated uvtetO promoter. Gene specific small guide RNA (sgRNA) was expressed in pRH2521 vector under the control of a TetRregulated smyc promoter (P myc1 tetO). To deplete sufT, gene specific sgRNAs were designed for two regions between 98-127 bp and 171-191 bp of sufT ORF and cloned in pRH2521. The SufT-KD secondary liquid culture in 7H9 was divided equally when OD 600nm reached to 0.1-0.2 and cultured in the presence or absence of ATc (200 ng/mL). Depletion of sufT was verified by RT-qPCR. Based on the significant repression of sufT, we chose sgRNA targeting sufT region between 98-127 bp for further study. All the experiments were performed with 72 h ATc treated SufT-KD culture, unless otherwise mentioned.

Bioinformatic analysis
The DUF59 family protein sequence alignment of amino acid residues from the H. sapiens, S. meliloti, S. aureus, B. anthracis, M. smegmatis, M. tuberculosis, and M. leprae were performed with ClustalW and identity of Rv1466 with homologous sequence distribution was achieved by BLASTp. The three-dimensional structural model for Mycobacterium tuberculosis SufT (PDB ID: 5IRD), Bacillus anthracis YitW (PDB ID: 3LNO), Thermotoga maritima TM0487 (PDB ID: 1UWD), and Homo sapiens FAM96a (PDB ID: 2M5H) were downloaded from RCSB PDB, a protein structure database (www.rcsb.org) [66]. The structures were visualized and analyzed in Pymol [The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC]. The DPE TC motif residues were highlighted and the distances between the selected residues were calculated.

RNA isolation and RT-qPCR experiments
Total RNA extraction was conducted using the FastRNA Pro Blue Kit (MP Biomedicals, USA) in accordance with the manufacturer's instruction and further purified using RNeasy spin columns (Qiagen, USA) as described in [67]. For RT-qPCR analysis, total RNA was extracted from SufT-KD and control strains after 24 h and 48 h of ATc treatment. After DNase treatment, total 600 ng of RNA was used for cDNA synthesis by using random hexamer oligonucleotide primer (iScript Select cDNA Synthesis Kit, BioRad, USA). Gene specific primers (S1 Table) and iQ SYBER Green Supermix (BioRad, USA) were used for RT-qPCR (StepOne Plus, Thermo, USA). In order to obtain meticulous expression levels, PCR expression was

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb normalized with Mtb 16S rRNA expression as endogenous control and StepOne software v2.3 (Thermo) was utilized for data analysis.

Aconitase assay
The Acn activity was measured by monitoring the disappearance of cis-aconitate at wavelength (λ) 240 nm in a UV spectrophotometer (Thermo Scientific Biomat 3S, USA) as described [68][69][70]. One unit (U) of Acn activity is defined as 1 μmol cis-aconitate formed or converted per minute. Reaction mixtures for Acn assay contained 25 mM Tris-HCl (pH 8.0), 100 mM NaCl, and 50 μg Mtb cell lysates in 1 mL of reaction volume. Reaction was initiated by adding 0.15 mM cis-aconitate and monitored by following the disappearance of cis-aconitate at λ 240 nm after every 15 sec for 30 min [68,69]. Absorbance at λ 240 nm was plotted against time. Acn activity was calculated from linear portion of the curve in initial 5 min. An extinction coefficient of 3,500 M -1 cm -1 was used to calculate the rates.

ROS measurement
Exponentially growing SufT-KD culture at an initial OD 600 nm of 0.15 was divided in two and was either treated with 200 ng/mL of ATc for 72 h or left untreated. The cells were harvested by centrifugation at 4500xg for 5 min and re-suspended in 200 μL of growth medium. As per manufacturer instructions, CellRox Deep Red (Invitrogen, Waltham, MA) was added to a final concentration of 5 μg/mL and cells were agitated on a rocker (Biobee Tech, Bangalore, India) for 30 min. After incubation, cells were washed to remove residual dye by centrifugation at 4500xg for 5 min. Cells were re-suspended in 300 μL phosphate-buffered saline (PBS), (pH 7.4) and then fixed by addition of 4% paraformaldehyde (PFA) for 1 h at room temperature. Fluorescence was measured at a fixed emission (670 nm) after excitation with a red laser (640 nm) using a BD FACSVerse Flow cytometer (BD Biosciences, San Jose, CA).

Western blot and immune-precipitation
For western blot, whole cells lysate (30 μg) was separated on 12% SDS-PAGE and then transferred onto a PVDF membrane (GE Healthcare, Piscataway, NJ, USA). Membrane was blocked using 5% (w/v) non-fat dry milk and incubated for 3 h at room temperature with primary antibody (Acn and SufT at 1:10000 dilutions). After washing with 1xTBST (Tris-buffered saline pH 7.5 with 0.05% Tween-20), membranes were incubated in goat anti-rabbit IgG HRPconjugated secondary antibody (1:10000 dilution) for 1 h. The autoradiography signals were visualized using ECL advance Western blotting detection kit (BioRad, USA).
Immunoprecipitation of Acn and SufT was performed by using protein A/G Mix magnetic bead (Cat. No.-LSKMAGAG02 Sigma Aldrich, USA) as per manufacturer's protocol. Equal amounts of cell free lysates (200 μg lysate of SufT-KD strain with and without ATc) were incubated with anti-SufT IgG antibody at 4˚C for 1 h. Further 25 μL of protein A/G Mix magnetic beads were added and incubated with the antibody-lysates mix at 4˚C for 4 h with continuous mixing. The beads were washed thrice with PBST (PBS with 0.01% Tween-20) and eluted with acidic glycine (pH 2.5) or boiled with gel loading 1X laemmli buffer directly as per requirement of experiments. The eluates were separated on 12% SDS-PAGE gel and probed with anti-Acn IgG antibody.

Mycobacterial protein fragment complementation (M-PFC) assay
The M-PFC assay is based on the functional reconstitution of murine dihydrofolate reductase (mDHFR) enzyme activity mediated by interaction between two candidate proteins fused to

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb mDHFR fragments F [1,2] and F [3]. The reconstitution of mDHFR activity can be scored by detecting growth on 7H11 plate containing trimethoprim (TRIM) 0, 25 and 50 μg/mL, a drug that preferentially targets bacterial DHFR. Two independent constructs were prepared: 1) SufT WT or SufT C62A were expressed in fusion with C-terminus of F [3] fragment mDHFR and 2) Acn, SufR, SufS, SufU, SufT, and GCN4 fused to the C-terminus of F [1,2]

Purification of SufT and UV-vis spectroscopy
Mtb sufT (Rv1466) ORF was PCR-amplified using gene-specific oligonucleotides (Rv1466_Nde1_FP and Rv1466_Xho1; S1 Table), digested with NdeI-Xho1, and ligated into similarly digested His-tag-based expression vector, pET28a (TAKARA BIO, Clontech Laboratories, CA, USA) to generate pET28a-SufT construct. N-terminal histidine-tagged SufT was overexpressed in E. coli BL21 λDE3 by 0.6 mM IPTG [Isopropyl β-D-1-thiogalactopyranoside; MP Biomedicals, USA (16 h, 18˚C)]. To facilitate Fe-S cluster formation, cultures were incubated on ice for 18 min prior to induction and were supplemented with 300 μM ferric ammonium citrate and 75 μM L-methionine (Amresco, USA) and purified as described [16]. Single primer method of site directed mutagenesis approach was used to create cysteine to alanine substitutions (C62A). After the PCR, DpnI was used to digest the wild-type methylated template plasmid. The reaction mixture containing the mutated sufT gene was used to transform E. coli BL21λDE3. Primer sequences are enlisted in S1 Table. Resulting clones were verified by sequencing, and the mutant SufT C62A variants of the SufT WT were purified as described earlier.
The UV-visible absorption spectroscopy of purified SufT protein was carried out in Thermo scientific spectrophotometer (Thermo scientific, USA) at 25˚C. Absorption spectra of SufT WT were recorded between λ 200 nm-800 nm in sodium phosphate buffer (20 mM and pH 7.4).

Iron estimation
For Fe estimation, freshly purified SufT was saturated with FeSO 4 (300 μM) and desalted using PD10 desalting column. Total Fe was quantified as described [71]. Total 100 μM and 0.1 mL of purified desalted SufT protein was heated at 95˚C for 30 min after treatment with 22% HNO 3 (0.1 mL). Samples were cooled to ambient temperature followed by addition of 0.6 mL ammonium acetate (7.5% w/v), 0.1 mL freshly prepared ascorbic acid (12.5% w/v) and 0.1 mL ferene (10 mM). The concentration of Fe present in the protein was determined by measuring the absorbance of the product, Fe-ferene complex at 593 nm, which was compared with a standard curve prepared from dilutions of freshly prepared Fe(III) solution in the range of 0-200 μM.
Total Fe was also estimated by atomic absorption spectroscopy (AAS) performed on atomic absorption spectrophotometer system (Shimadzu AA6880, Japan). AAS is a spectro-analytical procedure for the quantitative determination of ionizing elements using the absorption spectra. Total 200 μM of purified proteins of SufT WT, SufT C62A, BSA (negative control) and catalase (positive control) were saturated with 300 μM of FeSO 4 followed by desalting by PD10 column. Protein samples were estimate as described earlier [72] with little modifications. Briefly, equal volume of protein samples (100 μL) was cold digested with 10 mL of nitric acid overnight.

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb Next day 10 mL of di-acid (mixture of nitric acid and perchloric acid in the ratio of 9:3 ratios) was added and heat digested on sand hot plate. Digestion was confirmed by the sequential raising of red and white fumes. Precaution was taken to avoid charring of samples to get a final colorless solution. The digested samples were adjusted to 25 mL, using double distilled water and used for further iron analysis. Air compressor (4 psi) and gas (acetylene, 0.9 psi) were adjusted for smooth flame. Once AAS is ready, standard curve from 0, 1, 2, 3, 4 and 5 ppm of Fe (prepared from iron standard solution of 1000 mg/L of AAS grade, SRL) were generated followed by measurement of Fe in protein samples.

Metabolite extraction and LC MS/MS based analysis
SufT-KD strain with and without ATc were grown in the standard 7H9 medium and harvested after 72 h of ATc treatment during the log phase (OD 600nm of 0.8-1.4). Cells were quenched for 5 min in 4 volumes of 60% methanol (maintained at -45˚C) and then centrifuged at 4800xg (-5˚C). The pellet was re-suspended in 700 μL of 60% methanol (maintained at -45˚C) and then centrifuged at 4800xg (-5˚C). The pellet obtained was re-suspended in 1 mL of 75% ethanol and kept at 80˚C for 3 min with intermittent vortexing at 1.5 min interval, immediately followed by incubation on ice for 5 min and centrifugation at 17000xg for 15 min. The final supernatant was dried on a vacuum concentrator for 3-4 h and then stored at -80˚C till further analysis. Metabolites were resolved and analyzed, with appropriate multiple reaction monitoring (MRM) methods, as described earlier [42]. At the time of metabolite measurement, the extract was dissolved in a suitable solvent as described [42] and injected into the Synergi Fusion-RP column on Agilent's 1290 infinity series UHPLC system coupled to a triple-quadrupole type mass spectrometer (Sciex QTRAP 6500).

OCR and ECAR measurements
The SufT-KD strain grown for 72 h with or without ATc in standard 7H9 medium and equal number of cells (2x10 6 CFU/well) were adhered to the bottom of a XF cell culture microplate (Agilent technologies, USA), using Cell-Tak as a cell and tissue adhesive (Corning Cell-Tak). OCR and ECAR were measured using Agilent XF Extracellular Flux Analyzer (Agilent, US). Assays were carried out in unbuffered 7H9 media (pH 7.35) without glucose and ATc. Basal OCR and ECAR were measured for initial 21 min. Further, freshly prepared 2 mg/mL glucose as carbon source was added automatically in 7H9 unbuffered media, through port A and port B of the cartridge plate. To achieve maximum rate of respiration three measurements points were taken followed by 5 μM CCCP (Sigma-Aldrich, USA) treatment. Spare respiratory capacity, was calculated from % OCR value, by subtracting third basal reading, before glucose addition (normalized as 100%) from first point after CCCP addition.

H 2 O 2 and NO stress
Control and SufT-KD strains were grown in standard 7H9 media and treated with or without ATc for 72 h to deplete SufT. Cultures were diluted to an OD 600nm of 0.15 and exposed to 0.5 mM, 1mM DETA-NO and 5mM and 10 mM H 2 O 2 . After 12 h cells were diluted appropriately and plated on ADS-7H11 plates. Colonies were counted after 3-4 weeks of incubation at 37˚C.

Cell line experiments
RAW264.7 murine macrophage cell line and PMA differentiated THP-1 cells were infected with Control strain with ATc and SufT-KD strains with and without ATc. Infection was performed at MOI 2 for 4 h, followed by washing thoroughly to remove extracellular bacteria with

PLOS PATHOGENS
Fe-S cluster maturation is essential for maintaining redox homeostasis and bioenergetics in Mtb warm PBS and suspended in the DMEM/RPMI-1640 media, containing 10% FBS. For CFU determination macrophages were lysed using 0.06% SDS and further serially diluted in 7H9 medium (0.05% Tween-80) and plated on OADC-7H11 at indicated time points. Colonies were counted after 3-4 weeks of incubation at 37˚C. For experiments with the iNOS inhibitor, 25 μM of 1400W was added to the media containing infected RAW264.7 after the 4 h of phagocytosis and remained throughout incubation.

Animal experiments
For the chronic model of infection, 5-6 week old female BALB/c, mice (n = 6 per group) were infected by aerosol with approximately 200 bacilli per mouse with the SufT-KD strain using a Madison chamber by aerosol generation. Mice were divided in three groups, SufT-KD-DOX (without doxycycline), SufT-KD+DOX Acute phase (doxycycline started after 7 days of infection) and SufT-KD+DOX chronic phase (doxycycline started after 21 days of infection). Doxycycline was given in drinking water 1 mg/mL in 5% sucrose solution. Water bottles were light protected and replaced twice a week. At indicated times post infection, mice were euthanized, and the lungs were harvested for bacillary load, tissue histopathology analysis, and pathological scoring as described [46]. The remaining tissue samples from each mouse were homogenized and bacillary load was quantified by plating serial dilutions of tissue homogenates onto Middlebrook 7H11-OADC agar plates supplemented with lyophilized BBL MGIT PANTA antibiotic mixture (polymyxin B, amphotericin B, nalidixic acid, trimethoprim, and azlocillin, as supplied by BD; USA). Colonies were observed and counted after 4 weeks of incubation at 37˚C.

Statistical analysis
All data were graphed and analyzed with Prism v8.0 (GraphPad) unless otherwise stated. Statistical analyses were performed using Student's t-tests (two-tailed). Where comparison of multiple groups was made either multiple t test or two-way ANOVA with multiple comparisons was performed. Differences with a p value of < 0.05 were considered significant.