The Mycobacterium smegmatis HesB Protein, MSMEG_4272, Is Required for In Vitro Growth and Iron Homeostasis

A-type carrier (ATC) proteins are proposed to function in the biogenesis of Fe-S clusters, although their exact role remains controversial. The genome of Mycobacterium smegmatis encodes a single ATC protein, MSMEG_4272, which belongs to the HesB/YadR/YfhF family of proteins. Attempts to generate an MSMEG_4272 deletion mutant by two-step allelic exchange were unsuccessful, suggesting that the gene is essential for in vitro growth. CRISPRi-mediated transcriptional knock-down of MSMEG_4272 resulted in a growth defect under standard culture conditions, which was exacerbated in mineral-defined media. The knockdown strain displayed reduced intracellular iron levels under iron-replete conditions and increased susceptibility to clofazimine, 2,3-dimethoxy-1,4-naphthoquinone (DMNQ), and isoniazid, while the activity of the Fe-S containing enzymes, succinate dehydrogenase, and aconitase were not affected. This study suggests that MSMEG_4272 plays a role in the regulation of intracellular iron levels and is required for in vitro growth of M. smegmatis, particularly during exponential growth.


Introduction
Iron-sulphur (Fe-S) clusters are ubiquitous protein co-factors involved in diverse biological processes [1][2][3][4][5]. In vivo, Fe-S cluster assembly and transfer to apo-proteins require multi-protein systems, and three such systems have been identified in prokaryotes, namely the nitrogen fixation associated (NIF), iron-sulphur cluster (ISC), and sulphur mobilization (SUF) systems [6][7][8]. Fe-S cluster assembly occurs by highly conserved steps. The process is initiated by sulphur mobilisation from cysteine by a cysteine desulphurase (IscS, SufS, NifS) and persulphide transfer to the scaffold protein (IscU, SufB, NifU) [9]. Fe-S cluster assembly on the scaffold protein is followed by transfer to an apo-protein, either directly or via a transfer protein [10]. Although the process has been studied extensively, questions remain about the identity of the iron chaperone and the process of Fe-S cluster transfer. In eukaryotes, frataxin was proposed as the iron chaperone, while characterisation of the bacterial frataxin orthologue, CyaY, suggests that it functions as an iron-dependent regulator of Fe-S cluster assembly [11,12]. IscA, SufA, and nif IscA (HesB) are A-type carrier (ATC) proteins associated with the ISC, SUF, and NIF systems, respectively. However, their role in Fe-S cluster biogenesis remains controversial as they were observed to bind both iron [13][14][15] and Fe-S clusters [14][15][16] in vitro. E. coli has two additional ATC proteins, ErpA and NfuA, which are encoded elsewhere in the genome [7,[16][17][18][19].

Generation of Mutant Strains of M. smegmatis
The allelic exchange substrate for deletion of MSMEG_4272 in M. smegmatis was generated by cloning upstream (1500 bp) and downstream (1578 bp) homologous regions into the suicide plasmid p2NIL. The primers used for amplification are indicated in Table S2. The selection markers were cloned from pGOAL17 to generate the plasmid p2NIL4272LS. A MSMEG_4272 protein depletion strain was generated by creating an allelic exchange substrate to introduce the ID-tag at the C-terminus of the MSMEG_4272 gene in the M. smegmatis chromosome. The inducible degradation (ID) tag was amplified from pdacB_SsrA-tag plasmid, while the MSMEG_4272 gene, with 300 bp upstream, was amplified from M. smegmatis mc 2 155 genomic DNA [37]. A fusion of these two fragments was generated by overlap PCR, using the primers indicated in Table S2, and cloned into pJET1.2 to generate pJET_4272ssr. The MSMEG_4272 downstream (881 bp) region was amplified by PCR and fused to the ID-tagged MSMEG_4272 by overlap PCR (pJET_4272ID). The upstream/gene/ID-tag/downstream fusion was generated such that the MSMEG_4272 stop codon was removed and introduced at the end of the ID-tag. To ensure efficient homologous recombination, the MSMEG_4272 upstream region was extended from 300 bp to 1500 bp by subcloning a 1200 bp region from a previously prepared plasmid, pJET_4272_up (Table S1). Following cloning into p2NIL (pN_4272ID), selectable markers from pGOAL 17 were added, resulting in pNG4272ID.
Mutants were generated using the two-step allelic exchange approach as previously described by Parish and Stoker [38]. Briefly, the first homologous recombination event was selected for on 7H10 AD agar, containing the appropriate antibiotics and x-gal, rendering colonies blue, antibiotic resistant, and sucrose sensitive. The second homologous recombination event was selected for on 7H10 AD agar with 2% sucrose and x-gal, rendering colonies white, antibiotic sensitive, and sucrose resistant. Following the selection of the second cross-over event, colonies were screened for the desired modification by PCR. The primers used for PCR screening are indicated in Table S2. Mutant genotypes were confirmed by southern blotting analysis ( Figure S1). This mutant, 4272ID, was transformed with the pMC1s::HIV2Pr plasmid to generate the final strain in which the MSMEG_4272 protein levels could be modulated (4272ID::HIV2Pr).

CRISPRi Mediated MSMEG_4272 Gene Silencing in M. smegmatis
The CRISPR interference (CRISPRi)-based system was used for MSMEG_4272 gene silencing in M. smegmatis as previously described [39]. A sgRNA (5 -GTCCCCGTCGAGGGT-GCGGTCGTC-3 ) targeting MSMEG_4272 was selected based on a previous study that ranked possible sgRNA sequences on the PAM score, mismatch score, GC content, and the homopolymer repeat count [39]. The sgRNA oligos were commercially synthesized and cloned into the pLJR962 vector backbone, using the BsmBI restriction site to generate the pSG4272 plasmid. The pLJR962 and pSG4272 plasmids were individually transformed into electro-competent M. smegmatis mc 2 155 and 4272ID strains, resulting in WT::empty, WT::sgRNA, 4272ID::empty, and 4272ID::sgRNA.

Real-Time Quantitative PCR
M. smegmatis cells (10 mL) were harvested after 21 h and 33 h of growth by centrifugation and resuspended in 1 mL RNAProBlue solution. Cells were disrupted by bead beating (FastPrep) for three cycles of 40 s at 4 watts, with 1 min cooling on ice between cycles. Cellular debris was removed by centrifugation at 12,000× g for 15 min, and a chloroform ex-traction was performed before purification using the Nucleospin RNA II kit. An on-column DNase treatment was performed according to the manufacturer's instructions. The RNA integrity was assessed using the Aligent 2100 Bioanalyzer, and only samples with an RIN above 8 were analysed. The Transcriptor First Stand cDNA synthesis kit (Roche, Mannheim, Germany) was used to reverse transcribe 90 ng of RNA as set out in the manufacturer's instructions. Gene-specific primers used for cDNA synthesis of MSMEG_4272 (RT_4272) and sigA (rtsmSiga-r1) transcripts are indicated in Table S2. Quantitative PCR was performed with the SsoAdvanced™ Universal SYBR Green Supermix and the QuantStudio 5 system. qPCR primers are indicated in Table S2. Genomic DNA was used to generate a standard curve for each gene, and MSMEG_4272 expression was determined relative to sigA for each sample.

Intracellular Iron Determination
The intracellular iron levels were measured using the ferrozine assay as previously described [40,41]. Intracellular iron was determined after M. smegmatis growth in M-Media, with and without the supplementation of 2 µM iron. Cells were harvested by centrifugation (3220× g) and washed once with PBS containing 0.05% (v/v) Tween-80. Cells were resuspended in 1 mL of NaOH (50 mM) and ribolysed for three cycles (30 s at 4.5 Watts). The iron content was standardized using the total protein content, as determined by the Bradford protein assay.
The activity of the Fe-S cluster containing enzymes, succinate dehydrogenase, and aconitase was evaluated to indirectly assess the effect on Fe-S cluster biogenesis. The M. smegmatis strains, in which MSMEG_4272 expression could be transcriptionally silenced, were initially cultured for 15 h, without Atc, whereafter the culture was split in two, and Atc was added to one of the cultures to induce transcriptional silencing. Cells were harvested from 5 mL of each culture at an OD 600~0 .6-0.8 and washed once with 5 mL PBS-T buffer (1× PBS, 0.05% (v/v) Tween80). Pellets were stored at −80 • C until needed. The succinate dehydrogenase assay was performed as previously described [41], and absorbance was monitored at 500 nm every minute for 30 min using the Multiskan SkyHigh Microplate Spectrophotometer (Thermo Scientific, Vilnius, Lithuania). The aconitase assay was performed under anaerobic conditions in a Bactron anaerobic glove box and developed by combining previously published protocols and the Abcam Aconitase Enzyme Activity Microplate Assay Kit [44]. In an anaerobic cuvette, 0.75 mM MnCl 2 was added to 100 µL cell lysate. Isocitrate (20 mM) was added to the reaction anaerobically, and the spectrum was recorded every 30 s at 240 nm using UV spectroscopy. Readings were plotted to determine the linear range of the curve. The extinction coefficient for isocitrate at 240 nm (3.6 mM −1 cm −1 ) was used to calculate the isocitrate concentration change per minute. Enzyme activity for both the succinate dehydrogenase and aconitase assays were analysed as previously described [45].

Statistical Analysis
All statistical analyses were done using GraphPad Prism ® version 8.0 software package. Growth curve data were evaluated as separate replicates using non-linear or linear regression analysis. Non-linear regressions were fitted to a sigmoidal (dose-responsive) curve. To compare the LogEC50 values, an unpaired t-test was used without assuming a consistent standard deviation. The LogEC50 values represent the point at which 50% of the maximum cell density has been reached. Similarly, p-values were obtained by comparing the Hillslope of each growth curve using an unpaired t-test, without assuming a consistent standard deviation. Phenotype characterization data for all strains were analysed by using an unpaired t-test, analysing each row individually, without assuming a consistent standard deviation.

MSMEG_4272 Function Is Sensitive to the Modulation of Protein Levels
We sought to investigate the function of the HesB/YadR/YfhF family protein, MSMEG_4272, in M. smegmatis by generating a strain lacking this protein. Initial attempts to delete the MSMEG_4272 gene in M. smegmatis by two-step allelic exchange failed to yield mutants, as only wild-type or single-cross overs were recovered after the second selection step (Table S3). In an alternate approach, we generated a strain in which the chromosomal copy of MSMEG_4272 was fused to a C-terminal ID-tag which is comprised of a myc-tag, the TB SsrA tag, an HIV-2 protease cleavage site, and a flag-tag. In this system, induction of HIV-2 protease expression by the addition of Atc exposes the modified SsrA sequence, targeting the tagged MSMEG_4272 protein for degradation [37]. FLAG-and myc-tags allow monitoring of cleavage and degradation respectively ( Figure 1a). Despite two rounds of culturing in the presence of the Atc to induce protease expression, adequate depletion of the MSMEG_4272 protein could not be achieved (Figure 1b,c). Although the presence of the protease-encoding plasmid in the 4272ID::HIV2Pr strain decreased the amount of mycand FLAG-tagged protein detected (panel 2), the decrease was seen both in the presence and absence of Atc, suggesting that protease expression may be leaky in the absence of the inducer. This is further supported by the observation that the addition of Atc had no impact on growth, while 4272ID::HIV2Pr exhibited a growth defect relative to 4272ID (Figure 1d).
Next, we employed the mycobacterial CRISPRi system [39,46] to generate strains in which the MSMEG_4272 gene expression could be silenced. This approach allowed silencing of MSMEG_4272 expression on a transcriptional level, whereas the previous approach degraded MSMEG_4272 after its translation. MSMEG_4272 does not appear to be part of an operon and this approach should therefore not alter the expression of any genes located downstream of MSMEG_4272. The plasmid encoding the MSMEG_4272-directed sgRNA and inducible dCas9 nuclease (pSG4272) was therefore introduced into both the wild-type M. smegmatis and the 4272ID strains. We reasoned that the C-terminal tag introduced for protein depletion presented a convenient way of monitoring MSMEG_4272 protein levels. The control strains, WT::empty and 4272ID::empty, which contain the vector without a sgRNA, displayed comparable growth in the presence and absence of Atc, indicating that dCas9 expression did not affect growth ( Figure S2). Induction of the MSMEG_4272-directed sgRNA and dCas9 protease in the 4272ID::sgRNA strain completely inhibited growth ( Figure 2), while in the absence of Atc, growth of the strain was comparable to that of 4272ID::empty ( Figure S2b).
Western blot analysis confirmed depletion of the MSMEG_4272-ID fusion protein in the 4272ID::sgRNA strain in the presence, but not in the absence, of Atc ( Figure 2b). Surprisingly, no growth defect was observed for the WT::pSG4272 in the presence of Atc ( Figure S2a). We hypothesize that this strain may have sufficient levels of functional MSMEG_4272 protein to allow normal growth, and that introduction of the tag on the C-terminus of MSMEG_4272 either alters protein stability or activity/function, thereby unmasking the essentiality of this protein when combined with transcriptional knock-down.  To further characterize these strains, we grew them in the absence of Atc for 15 h and then induced transcriptional silencing. At 9 h following Atc addition, the 4272ID::sgRNA strain displayed a reduced growth rate relative to the other strains, confirming that the protein is required for normal growth under standard conditions ( Figure 3). When comparing the Hill-slope (i.e., change of OD over time during exponential growth) of strains cultured in the presence and absence of Atc, a significant difference was observed for both WT::sgRNA (p = 0.005) and 4272ID::sgRNA (p = 0.001), suggesting a change in the growth rate, although the difference in OD values at each time point for the WT::sgRNA strain was marginal. bodies to determine protein degradation. Aliquots from (1) 4272ID and (2) 4272ID::HIV2Pr, with and without Atc, were taken after 16 h, 25 h, and 40 h of growth (dot-dash lines). Protein quantity for western blotting analysis was standardized by protein concentration using the Bradford assay.   Western blotting with anti-FLAG antibody was used to determine MSMEG_4272 protein levels for 4272ID::sgRNA after 45 h. Bradford assays were used to standardise the total protein loaded for western blotting. Higher molecular weight bands represent multimers of the protein due to incomplete reduction. Western blot analysis confirmed depletion of the MSMEG_4272-ID fusion protein in the 4272ID::sgRNA strain in the presence, but not in the absence, of Atc ( Figure 2b). Surprisingly, no growth defect was observed for the WT::pSG4272 in the presence of Atc (Figure S2a). We hypothesize that this strain may have sufficient levels of functional MSMEG_4272 protein to allow normal growth, and that introduction of the tag on the Cterminus of MSMEG_4272 either alters protein stability or activity/function, thereby unmasking the essentiality of this protein when combined with transcriptional knock-down.
To further characterize these strains, we grew them in the absence of Atc for 15 h and then induced transcriptional silencing. At 9 h following Atc addition, the 4272ID::sgRNA strain displayed a reduced growth rate relative to the other strains, confirming that the protein is required for normal growth under standard conditions ( Figure 3). When comparing the Hill-slope (i.e., change of OD over time during exponential growth) of strains cultured in the presence and absence of Atc, a significant difference was observed for both WT::sgRNA (p = 0.005) and 4272ID::sgRNA (p = 0.001), suggesting a change in the growth rate, although the difference in OD values at each time point for the WT::sgRNA strain was marginal. To verify transcriptional silencing of MSMEG_4272, transcript levels were determined at 6 (exponential phase) and 18 h (stationary phase) after the addition of Atc ( Figure  4). The WT::sgRNA and 4272ID::sgRNA strains grown in the presence of Atc showed a reduction in MSMEG_4272 expression relative to no Atc controls log2 fold change of 2.85 (±0.09) in WT::sgRNA and 2.53 (±0.03) in 4272ID::sgRNA). This data also confirms that the phenotypic differences observed between WT::sgRNA and 4272ID::sgRNA are not due to differences in transcriptional silencing. To verify transcriptional silencing of MSMEG_4272, transcript levels were determined at 6 (exponential phase) and 18 h (stationary phase) after the addition of Atc (Figure 4). The WT::sgRNA and 4272ID::sgRNA strains grown in the presence of Atc showed a reduction in MSMEG_4272 expression relative to no Atc controls log 2 fold change of 2.85 (±0.09) in WT::sgRNA and 2.53 (±0.03) in 4272ID::sgRNA). This data also confirms that the phenotypic differences observed between WT::sgRNA and 4272ID::sgRNA are not due to differences in transcriptional silencing.

Phenotypic Impact of MSMEG_4272 Transcriptional Silencing on M. smegmatis
The impact of MSMEG_4272 silencing on M. smegmatis was characterized by assessing drug susceptibility and the activity of Fe-S cluster-dependent enzymes. Silencing of MSMEG_4272 did not affect the activity of succinate dehydrogenase and aconitase (Figure S3). Since mutants with defects in Fe-S cluster biogenesis often show increased susceptibility to oxidative stress, the susceptibility of the strains to compounds that generate reactive oxygen species was investigated [26,47]. An increase in the susceptibility to clofazimine and DMNQ was observed in the presence of Atc in WT::sgRNA (two-fold) and 4272ID:: sgRNA (four-fold). We observed a large degree of variability between replicates for isoniazid susceptibility, represented by the broad MIC ranges in Table S4. In general, MSMEG_4272 depletion in WT::sgRNA and 4272ID::sgRNA increased susceptibility to isoniazid. In addition, the susceptibility of the control strain, 4272ID::empty, was lower than that of WT::empty, which may be due to the C-terminal ID-tag altering protein function or stability.
ATC proteins are proposed to function as iron chaperones or to contribute to the intracellular iron pool. We, therefore, investigated the impact of transcriptionally silencing MSMEG_4272 on growth in Chelex-treated MM media, with (iron-replete conditions) or without (iron limiting conditions) supplemental iron ( Figure 5). Consistent with previous reports, irrespective of Atc addition, a significant reduction in growth (OD600) was observed for all strains under iron limiting conditions, when compared to iron-replete conditions after 57 h of culturing (WT::sgRNA (without Atc, p = 0.028) and 4272ID::sgRNA (without Atc, p = 0.029)) [45]. For 4272ID::empty with Atc, WT::sgRNA with Atc, and 4272ID::sgRNA without Atc, no significant difference in their growth was observed when compared under either iron limiting or iron-replete conditions. RT-qPCR used to determine the total number of MSMEG_4272 transcript copies, at 6 (grey) and 18 h (black) after the addition of Atc, divided by the total number of SigA transcript copies. Plot displays the relative expression of three independent experiments (n = 3) performed as biological replicates. Where appropriate, statistical significance is indicated by p < 0.05 (*) and p < 0.01 (**). GraphPad Prism 8 software was used to determine the p values using an unpaired t test.

Phenotypic Impact of MSMEG_4272 Transcriptional Silencing on M. smegmatis
The impact of MSMEG_4272 silencing on M. smegmatis was characterized by assessing drug susceptibility and the activity of Fe-S cluster-dependent enzymes. Silencing of MSMEG_4272 did not affect the activity of succinate dehydrogenase and aconitase ( Figure S3). Since mutants with defects in Fe-S cluster biogenesis often show increased susceptibility to oxidative stress, the susceptibility of the strains to compounds that generate reactive oxygen species was investigated [26,47]. An increase in the susceptibility to clofazimine and DMNQ was observed in the presence of Atc in WT::sgRNA (two-fold) and 4272ID:: sgRNA (four-fold). We observed a large degree of variability between replicates for isoniazid susceptibility, represented by the broad MIC ranges in Table S4. In general, MSMEG_4272 depletion in WT::sgRNA and 4272ID::sgRNA increased susceptibility to isoniazid. In addition, the susceptibility of the control strain, 4272ID::empty, was lower than that of WT::empty, which may be due to the C-terminal ID-tag altering protein function or stability.
ATC proteins are proposed to function as iron chaperones or to contribute to the intracellular iron pool. We, therefore, investigated the impact of transcriptionally silencing MSMEG_4272 on growth in Chelex-treated MM media, with (iron-replete conditions) or without (iron limiting conditions) supplemental iron ( Figure 5). Consistent with previous reports, irrespective of Atc addition, a significant reduction in growth (OD 600 ) was observed for all strains under iron limiting conditions, when compared to iron-replete conditions after 57 h of culturing (WT::sgRNA (without Atc, p = 0.028) and 4272ID::sgRNA (without Atc, p = 0.029)) [45]. For 4272ID::empty with Atc, WT::sgRNA with Atc, and 4272ID::sgRNA without Atc, no significant difference in their growth was observed when compared under either iron limiting or iron-replete conditions. For the 4272ID::sgRNA strain, the addition of Atc at 15 h resulted in a more severe growth defect in MM media with supplemental iron than in 7H9 (Figure 5d). One of the differences between MM media and 7H9 is the source of nitrogen, which is asparagine in MM media and glutamic acid and ammonia in 7H9. To investigate if the growth defect observed in MM media was related to the nitrogen source, the asparagine in MM media was replaced with glutamic acid. This failed to rescue the growth defect of the 4272ID::sgRNA strain, suggesting that it was not due to inability to utilize asparagine ( Figure S4). Addition of Atc to MM media, without supplemental iron, at 15 h also resulted in a severe growth defect for the 4272ID::sgRNA strain (Figure 5c). However, the impact of iron limitation on growth was difficult to evaluate given the severity of the phenotype under iron replete conditions in this media. For the WT::sgRNA strain cultured under iron For the 4272ID::sgRNA strain, the addition of Atc at 15 h resulted in a more severe growth defect in MM media with supplemental iron than in 7H9 (Figure 5d). One of the differences between MM media and 7H9 is the source of nitrogen, which is asparagine in MM media and glutamic acid and ammonia in 7H9. To investigate if the growth defect observed in MM media was related to the nitrogen source, the asparagine in MM media was replaced with glutamic acid. This failed to rescue the growth defect of the 4272ID::sgRNA strain, suggesting that it was not due to inability to utilize asparagine ( Figure S4). Addition of Atc to MM media, without supplemental iron, at 15 h also resulted in a severe growth defect for the 4272ID::sgRNA strain (Figure 5c). However, the impact of iron limitation on growth was difficult to evaluate given the severity of the phenotype under iron replete conditions in this media. For the WT::sgRNA strain cultured under iron limiting conditions, no significant difference was observed when comparing growth in the presence and absence of Atc.
Comparison of the intracellular iron levels of the strains grown under iron limiting and iron-replete conditions revealed a reduction in iron levels for all strains grown under iron limiting conditions ( Figure 6). In addition, the 4272ID::sgRNA showed significantly reduced intracellular iron levels under iron replete conditions in the presence of Atc. This suggests that depleting MSMEG_4272 levels alters iron acquisition or storage in M. smegmatis under iron replete conditions, and this may partly explain the severe growth defect of the 4272ID::sgRNA strain when grown in MM containing 2 µM supplemental iron.
Microorganisms 2023, 11, x FOR PEER REVIEW 10 of 14 limiting conditions, no significant difference was observed when comparing growth in the presence and absence of Atc. Comparison of the intracellular iron levels of the strains grown under iron limiting and iron-replete conditions revealed a reduction in iron levels for all strains grown under iron limiting conditions ( Figure 6). In addition, the 4272ID::sgRNA showed significantly reduced intracellular iron levels under iron replete conditions in the presence of Atc. This suggests that depleting MSMEG_4272 levels alters iron acquisition or storage in M. smegmatis under iron replete conditions, and this may partly explain the severe growth defect of the 4272ID::sgRNA strain when grown in MM containing 2 μM supplemental iron.

Discussion
The process of Fe-S cluster biogenesis in mycobacteria has not been studied extensively, and as such, the role of A-type carriers in this genus is unclear. We sought to address this by investigating the role of the HesB/YadR/YfhF family protein, MSMEG_4272, in M. smegmatis by generating an MSMEG-4272-deficient strain. Attempts to construct an MSMEG_4272 knock-out mutant strain by two-step allelic exchange were unsuccessful, suggesting that the gene may be essential for in vitro growth. This is consistent with the prediction by forward genetic screens that MSMEG_4272 is essential under specific conditions [39]. The severe growth defect observed in the 4272ID::sgRNA strain in the presence of Atc ( Figure 2) further supports this prediction. In E. coli, where multiple A-type carriers are present, iscA or sufA single mutants are viable, while double mutants lacking either iscA-sufA or iscA-erpA presented a null-growth phenotype under aerobic and anaerobic conditions [13,18]. In Synechococcus sp., both the iscA and sufA single and the iscA-sufA double mutant was viable under aerobic conditions, suggesting that additional redundancy exists in this organism [48]. In M. smegmatis, the essentiality of MSMSEG_4272 is consistent with the absence of other A-type carrier homologues in this organism.
When attempting to generate a MSMEG_4272 knock-down strain using a previously described ssrA-tag-based protein depletion system, we encountered two problems. Firstly, reduced growth was observed in the presence and absence of Atc, suggesting that expression of the HIV-2 protease was leaky, and this was supported by the western

Discussion
The process of Fe-S cluster biogenesis in mycobacteria has not been studied extensively, and as such, the role of A-type carriers in this genus is unclear. We sought to address this by investigating the role of the HesB/YadR/YfhF family protein, MSMEG_4272, in M. smegmatis by generating an MSMEG-4272-deficient strain. Attempts to construct an MSMEG_4272 knock-out mutant strain by two-step allelic exchange were unsuccessful, suggesting that the gene may be essential for in vitro growth. This is consistent with the prediction by forward genetic screens that MSMEG_4272 is essential under specific conditions [39]. The severe growth defect observed in the 4272ID::sgRNA strain in the presence of Atc (Figure 2) further supports this prediction. In E. coli, where multiple Atype carriers are present, iscA or sufA single mutants are viable, while double mutants lacking either iscA-sufA or iscA-erpA presented a null-growth phenotype under aerobic and anaerobic conditions [13,18]. In Synechococcus sp., both the iscA and sufA single and the iscA-sufA double mutant was viable under aerobic conditions, suggesting that additional redundancy exists in this organism [48]. In M. smegmatis, the essentiality of MSMSEG_4272 is consistent with the absence of other A-type carrier homologues in this organism.
When attempting to generate a MSMEG_4272 knock-down strain using a previously described ssrA-tag-based protein depletion system, we encountered two problems. Firstly, reduced growth was observed in the presence and absence of Atc, suggesting that expression of the HIV-2 protease was leaky, and this was supported by the western blotting results (Figure 1). Secondly, although a lower level of the MSMEG_4272 protein was observed in the presence of the HIV-2 protease, the protein levels increased over time, even in the presence of Atc. Reduced MSMEG_4272 protein levels resulted in an extended lag phase. However, after 16 h, the growth rate increased significantly, suggesting that sufficient protein levels were achieved after this time point.
Generating a MSMEG_4272-deficient strain using the CRISPRi system was more successful than the protein depletion approach. RT-qPCR revealed that the CRISPRi system achieved a log 2 fold-change of 2.85 in WT::sgRNA and 2.53 in 4272ID::sgRNA after 6 h of sgRNA induction (Figure 4), with similar levels measured after 18 h of induction. Previous reports, investigating the relationship between transcriptional repression and the resulting phenotypic effect, used the same sgRNA sequence and observed a log 2 fold-change of 2.97, consistent with our findings [39,46]. Furthermore, a vulnerability index of −1.33 was observed for MSMSEG_4272, suggesting that a high level of inhibition is required before a growth defect is observed [46]. We hypothesize that the phenotypic differences between the 4272ID::sgRNA strain and the WT::sgRNA strain are due to either decreased functionality or decreased stability of the ID-tag fusion protein, which further decreases the amount of functional MSMEG_4272 protein in the 4272ID::sgRNA strain. Furthermore, the growth defect of the 4272ID::sgRNA strain was most prominent between 25 and 35 h, with the OD reaching the same level as the uninduced strain by 38 h. This may point to a higher demand for MSMEG_4272 during exponential growth, which cannot be achieved in the 4272ID::sgRNA strain. The observation that transcriptional repression in the 4272ID::sgRNA strain is comparable after 21 and 33 h of growth suggests this recovery is not due to increased expression but rather due to decreased demand for the protein as the culture enters stationary phase. Previous studies have shown that certain A-type carrier proteins are only required under specific conditions [15], such as oxidative stress. Therefore, MSMEG_4272 may be important during exponential growth when Fe-S cluster demand is high.
Significantly lower intracellular iron levels were observed when MSMEG_4272 was silenced in the 4272ID::sgRNA strain under iron replete conditions. In the cyanobacteria Synechococcus, transcription of isiA, a marker for iron limitation, was upregulated in the ∆iscA mutant under iron replete conditions, suggesting that it incorrectly senses iron levels [48]. Furthermore, increased growth of the E.coli iscA − /sufA − double mutant under anaerobic conditions compared to aerobic conditions, led to the hypothesis that these proteins are required to maintain iron in an accessible form only when oxygen is present [13], while nif IscA from A. vinelandii was proposed to function as a non-specific iron donor which contributes to the iron pool. These results implicate A-type carrier proteins in maintaining iron homeostasis and our findings suggest that MSMEG _4272 may have a similar role in M. smegmatis. Analysis of the activity of the Fe-S cluster-dependent enzymes succinate dehydrogenase and aconitase under iron replete conditions revealed no difference in activity when MSEMG_4272 was silenced. This was surprising given the proposed role for ATC proteins in Fe-S cluster assembly. However, this could suggest that MSMEG_4272 is functioning independently of the SUF system. An increase in the sensitivity to isoniazid was observed and possibly linked to altered iron homeostasis given that KatG is a hemecontaining enzyme and that FurA, a regulator of katG expression, is iron-binding [49].
Growth of the 4272ID::sgRNA strain was impaired in MM media in the presence and absence of iron when transcriptional silencing was induced, and unlike the growth defect in 7H9 AD, no recovery was observed at later time points. M. smegmatis can utilize asparagine, glutamic acid, and ammonia as nitrogen sources [50], with asparagine being the main nitrogen source in MM media and glutamic acid and ammonia the nitrogen sources in 7H9. Replacing asparagine with glutamic acid failed to restore growth of the 4272ID::sgRNA strain, suggesting that the phenotype was not related to asparagine utilization. Recently, deletion of sufT, the last gene in the sufR-sufB-sufD-sufC-csd-nifU-sufT operon, was found to alter levels of TCA cycle intermediates, amino acids, and sulphur-containing metabolites in M. tuberculosis; changes are thought to be mediated by defects in lipoic acid synthesis [51]. Transferring these observations to M. smegmatis should be done cautiously, though, as significant differences in the nitrogen assimilation and amino acid metabolic pathways in M. smegmatis have been reported [52]. The underlying mechanism of the growth defect of the MSMEG_4272-deficient strain in MM media therefore requires further investigation and should include more extensive metabolite profiling.

Conclusions
This study suggests that MSMEG_4272 plays a role in iron homeostasis in M. smegmatis and that modulating MSMEG_4272 levels alters isoniazid susceptibility. Further investigation is required to understand the underlying mechanism of these phenotypes and to elucidate the molecular function of this HesB/YadR/YfhF family protein in mycobacteria.