SGM Journals
PHYSIOLOGY

The alternative sigma factor SigF of Mycobacterium smegmatis is required for survival of heat shock, acidic pH and oxidative stress

Susanne Gebhard, Anja Hümpel, Alexander D. McLellan, Gregory M. Cook

  • Department of Microbiology and Immunology, Otago School of Medical Sciences, University of Otago, PO Box 56, Dunedin, New Zealand
  • Correspondence
    Gregory M. Cook
    greg.cook{at}stonebow.otago.ac.nz

    • Microbiology 2008; 154(9):2786–2795 · https://doi.org/10.1099/mic.0.2008/018044-0

    View at publisher PubMed

    Abstract

    The alternative sigma factor SigF of Mycobacterium tuberculosis has been characterized in detail as a general-stress, stationary-phase sigma factor involved in the virulence of the bacterium. While a homologous gene has been annotated in the genome of the fast-growing Mycobacterium smegmatis, little experimental evidence is available on the function of this gene. Here, we demonstrate that SigF of M. smegmatis is required for resistance to hydrogen peroxide, heat shock and acidic pH, but not for survival in human neutrophils. No difference in sensitivity to isoniazid was observed between the wild-type strain and the ΔsigF mutant, suggesting that SigF-mediated resistance to hydrogen peroxide was via a pathway independent of KatG or AhpC. RT-PCR and 5′-RACE (rapid amplification of cDNA ends) analyses showed that sigF of M. smegmatis was co-transcribed with rsbW (thought to encode an anti-sigma factor for SigF) and MSMEG_1802 (unknown function) and was expressed from two promoters, one upstream of MSMEG_1802 and the second upstream of rsbW. Analysis of transcriptional lacZ fusion constructs in the sigF-deletion background revealed that the MSMEG_1802 promoter was dependent on SigF for expression. Moreover, MSMEG_1802-lacZ was induced twofold upon entry into stationary phase, while exposure of exponentially growing cultures to various stress conditions (e.g. heat, cold, ethanol, hydrogen peroxide or different pH values) did not lead to induction of MSMEG_1802-lacZ. Expression of rsbW-lacZ was independent of SigF and remained constant throughout the growth cycle and under various stress conditions unless the bacteria were challenged with d-cycloserine.

    Edited by: G. R. Stewart

    INTRODUCTION

    In the course of investigations into the molecular mechanisms underlying the pathogenesis and persistence of Mycobacterium tuberculosis, several studies have focused on the role of global regulators, such as sigma factors. The genome of M. tuberculosis encodes 13 sigma factors, two primary factors (SigA and SigB) and 11 alternative factors (SigC to SigM in alphabetical order) (Cole et al., 1998; Manganelli et al., 2004). One of these sigma factors that has been studied in detail is the alternative sigma factor SigF, which is similar to the general stress sigma factor, SigB, of Bacillus subtilis (DeMaio et al., 1997). The gene encoding this sigma factor, sigF, is part of a gene cluster comprising usfY (unknown function), usfX (encoding a SigF-specific anti-sigma factor) and sigF, where usfX and sigF appear to be co-transcribed (Beaucher et al., 2002; DeMaio et al., 1997).

    Deletion mutants of M. tuberculosis in sigF have been constructed and characterized. While none of the mutants was impaired in growth in broth culture or human macrophages in vitro, they showed reduced virulence in mouse and guinea pig infection models (Chen et al., 2000; Geiman et al., 2004; Karls et al., 2006). Furthermore, it was shown that SigF was required for survival of M. tuberculosis in the lung, but not in the spleen of guinea pigs (Karls et al., 2006). SigF did not appear to be required for survival of stress conditions such as heat and cold shock or long-term stationary-phase survival (Chen et al., 2000). In fact, a recent study using both deletion as well as overexpression of sigF in M. tuberculosis suggested that SigF has a role in the regulation of the structure and function of the mycobacterial cell wall, rather than being a general stress-response sigma factor (Williams et al., 2007).

    The activity of SigF from M. tuberculosis is post-translationally regulated by the anti-sigma factor UsfX, which in turn is negatively regulated by two anti-anti-sigma factors, RsfA and RsfB, in response to different physiological stress conditions (Beaucher et al., 2002). In addition, several other proteins have been shown to interact with SigF or UsfX of M. tuberculosis and these are thought to exert some control on the activity of SigF (Parida et al., 2005). There appears to be no consensus regarding the expression of sigF from M. tuberculosis in response to different stress conditions such as infection of macrophages (Graham & Clark-Curtiss, 1999; Mariani et al., 2000), nutrient starvation (Betts et al., 2002; DeMaio et al., 1996; Michele et al., 1999), stationary-phase growth (DeMaio et al., 1996; Graham & Clark-Curtiss, 1999; Manganelli et al., 1999), or exposure to oxidative stress, alcohol, antibiotics and temperature shock (DeMaio et al., 1996; Manganelli et al., 1999; Michele et al., 1999).

    When the gene encoding SigF was first identified in M. tuberculosis, the authors were able to show the presence of homologous genes in other slow-growing mycobacteria, but not in any of the fast-growers tested, including Mycobacterium smegmatis (DeMaio et al., 1996, 1997). Inspection of the genome of M. smegmatis () revealed two open reading frames, MSMEG_1803 and MSMEG_1804, which have been annotated as rsbW (or usfX) and sigF, respectively, and recent literature has included SigF from M. smegmatis in discussions of mycobacterial sigma factors (Chowdhury et al., 2007; Manganelli et al., 2004; Waagmeester et al., 2005). However, no experimental evidence is available regarding the role of SigF as an alternative sigma factor of M. smegmatis. The genome of M. smegmatis contains twice as many sigma factors as that of M. tuberculosis (26 versus 13), suggesting significant differences in the regulatory repertoire between these two species (Waagmeester et al., 2005). However, while some sigma factors are unique to either M. tuberculosis (e.g. SigC, SigK and SigI) or M. smegmatis (e.g. multiple copies of SigH) and thus are likely to represent specific adaptations to the respective environments of these two species, several sigma factors, including SigF, are common to both (Waagmeester et al., 2005).

    In this paper, we report on the construction and characterization of a sigF deletion mutant of M. smegmatis. Furthermore, we used RT-PCR and transcriptional fusions to lacZ to study the expression of sigF in response to various stress conditions to determine whether sigF of M. smegmatis plays a similar role to the gene in M. tuberculosis.

    METHODS

    Bacterial strains and growth conditions.

    All strains and plasmids used in this study are listed in Table 1⇓. Escherichia coli strains were grown in Luria–Bertani (LB) medium at 37 °C with agitation (200 r.p.m.). Mycobacterium smegmatis strain mc2155 (Snapper et al., 1990) and derived strains were routinely grown at 37 °C, 200 r.p.m., in LB containing 0.05 % (w/v) Tween 80 (LBT), in Middlebrook 7H9 medium (Difco) supplemented with 10 % albumin-glucose-catalase enrichment (ADC; Becton Dickinson) and 0.05 % (w/v) Tween 80, or in modified Sauton's medium (ST). The composition of this medium per litre was as follows: 0.5 g MgSO4 heptahydrate, 2 g citric acid monohydrate, 23 g K2HPO4 trihydrate, 0.3 g KCl, 1 g l-asparagine, 2 g glycerol, 0.5 g Tween 80, 320 μl 0.5 M FeCl3 and 100 μl 1 M NH4Cl. For carbon-limited growth, glycerol was reduced to 1 g l−1; for nitrogen-limited growth, l-asparagine was reduced to 0.5 g l−1. M. smegmatis transformants were grown at 28 °C for propagation of temperature-sensitive vectors and at 40 °C for allelic-exchange mutagenesis. Selective media contained kanamycin (50 μg ml−1 for E. coli; 20 μg ml−1 for M. smegmatis), gentamicin (20 μg ml−1 for E. coli; 5 μg ml−1 for M. smegmatis) or hygromycin B (200 μg ml−1 for E. coli; 50 μg ml−1 for M. smegmatis). Solid media contained 1.5 % agar.

    Table 1.

    Bacterial strains, plasmids and primers used in this study

    Optical density was measured at 600 nm (OD600) using culture samples diluted in saline to bring OD600 to below 0.5 when measured in cuvettes of 1 cm light path length in a Jenway 6300 spectrophotometer.

    DNA manipulation and cloning of constructs.

    All molecular biology techniques were carried out according to standard procedures (Sambrook et al., 1989). Restriction or DNA-modifying enzymes and other molecular biology reagents were obtained from Roche Diagnostics or New England Biolabs. All primer sequences used are listed in Table 1⇑.

    Genomic DNA of M. smegmatis was isolated by a modification of the method of Gonzalez-y-Merchand et al. (1996) as described previously (Gebhard et al., 2006). To create a construct for the deletion of sigF, the gentamicin resistance cassette (Gmr), aacC-1, from pBSL141 (Alexeyev et al., 1995) was excised using EcoRI and SacI. PCR products of approximately 820 bp flanking the sigF gene of M. smegmatis on either side were amplified using the primers sigFKO1 with sigFKO2, and sigFKO3 with sigFKO4, respectively. The left-flank PCR product was digested with SpeI and EcoRI, the right-flank product with SacI and SpeI. Both flanks and the gentamicin resistance cassette were ligated into the SpeI site of pBluescript II KS (Stratagene). The resulting assembly, left flank/Gmr/right flank, was subcloned as a SpeI fragment into pPR23 (Pelicic et al., 1997), creating plasmid pSG45. The expected double-crossover event would result in a deletion–insertion at the sigF locus, eliminating 88 % of the sigF coding sequence in exchange for the gentamicin resistance marker (Fig. 1⇓).

    Figure image not available in archive
    Fig. 1.

    The sigF gene locus and construction of a sigF deletion mutant of M. smegmatis. (a) Schematic of the sigF gene loci from M. tuberculosis (Mtb) and M. smegmatis (Msm). (b) Schematic of allelic replacement of M. smegmatis sigF with aacC-1. SacI restriction sites and fragment sizes as detected in (c) are indicated. The black bar shows the fragment used as a probe. (c) Southern hybridization analysis of replacement of sigF with aacC-1. SacI-digested genomic DNA from the wild-type (WT) and sigF deletion mutant SG128 (ΔsigF) were probed with radiolabelled left-flank PCR product of the deletion construct.

    Allelic replacement of sigF was carried out essentially as described by Pelicic et al. (1997). In brief, a culture of M. smegmatis carrying pSG45 was grown at 28 °C with agitation (200 r.p.m.) to an OD600 of approximately 0.6–0.8, followed by plating onto low-salt LBT plates (2 g NaCl l−1) containing gentamicin and 10 % sucrose at 40 °C, selecting for double-crossover events. To confirm the correct deletion of sigF, candidate colonies were screened by Southern hybridization analysis of SacI-digested genomic DNA. Replacement of sigF with the gentamicin resistance marker created strain SG128 (ΔsigF : : aacC-1).

    For complementation of the sigF mutation, a 1.8 kb PCR product encompassing both the rsbW and sigF genes plus 570 bp upstream DNA and 37 bp downstream DNA was amplified by PCR using primers csigFF and csigFR and cloned into the HindIII site of the integrative E. coli/mycobacteria shuttle vector pUHA267 (Lee et al., 1991), creating plasmid pSG57. Transformation of strain SG128 with pSG57 resulted in strain SG158.

    To create a transcriptional fusion of the rsbWsigF operon to lacZ, a 655 bp PCR product, encompassing 577 bp of DNA upstream of rsbW plus 78 bp of its coding region, was amplified using primers PsigFF and PsigFR. The product was cloned into the ApaI and Asp718 sites of pJEM15 (Timm et al., 1994), creating plasmid pAH55. Similarly, a transcriptional MSMEG_1802 fusion was created by amplification of a 426 bp PCR product encompassing 211 bp DNA upstream of MSMEG_1802 plus 215 bp of its coding region using primers P1802F and P1802R and cloning into the ApaI and Asp718 sites of pJEM15, creating plasmid pAH59.

    Stress conditions.

    Stress experiments were performed with mid-exponential-phase (OD600 0.7–1.0) cultures grown in LBT medium. For heat-shock (50 °C for 4 h) and cold-shock experiments (28 °C for 8 h), the cultures were placed at the indicated temperatures without further manipulation. For acid and alkaline stress, the cells were collected by centrifugation (16 000 g, 1 min) and resuspended to an OD600 of 0.7 in citrate/phosphate buffer (100 mM, pH 4) or sodium borate buffer (13 mM, pH 9), respectively. For hyperosmotic and hypoosmotic stress, cells were collected as described above and resuspended to an OD600 of 0.7 in 0.5 M NaCl or distilled water, respectively. For oxidative, alcohol and antibiotic stress, the respective compounds were added directly to the mid-exponential-phase culture. Hydrogen peroxide was tested at concentrations between 0.1 and 10 mM, rifampicin and isoniazid from 1 to 16 μg ml−1 and d-cycloserine from 4 to 64 μg ml−1. DMSO was included as a solvent control, and had no effect on cell growth at all concentrations tested (up to 0.13 %). Cultures were incubated under stress conditions at 37 °C with agitation (200 r.p.m.) for 2 h or 4 h as stated. All stress experiments were carried out in triplicate. Survival was determined by viable cell counts from eight replicate spots of 5 μl per dilution. β-Galactosidase activities of strain AH56 (harbouring pAH55) were determined as described previously (Gebhard et al., 2006) and were expressed as Miller units (Miller, 1972), calculated as the increase in A420 per minute per 1 ml of cell suspension (normalized to an OD600 of 1.0) used, and multiplied by a factor of 1000.

    Statistical analyses were performed using paired or unpaired two-tailed t-tests for survival and β-galactosidase experiments, respectively. P-values are only reported where significant differences were found.

    Determination of in vitro killing by human neutrophils.

    Human neutrophils were isolated from heparinized peripheral blood from healthy adult donors using dextran sedimentation and centrifugation over a Ficoll-Hypaque gradient as described previously (Faldt et al., 2002). Cells were washed with PBS buffer and remaining erythrocytes removed by hypotonic lysis with water. The isolated neutrophils were resuspended to a concentration of 4×107 cells ml−1 in RPMI 1640 medium supplemented with 10 % autologous human serum or plasma. M. smegmatis strains were grown to an OD600 between 0.7 and 1.0 in LBT medium and diluted to 1.5×108 c.f.u. ml−1 in the same medium. Equal volumes of prepared mycobacterial cells and neutrophils (m.o.i. of 3.75) were diluted 10-fold into 1 ml prewarmed RPMI 1640 medium in 24-well plates and incubated at 37 °C and 5 % CO2 for 10 min, 20 min or 120 min. The neutrophils were lysed on ice for 20 min using 0.1 % saponin, and survival of M. smegmatis was determined by viable cell count. As a control, the mycobacterial cells were incubated in RPMI 1640 medium without neutrophils.

    Isolation of RNA and reverse transcriptase (RT)-PCR.

    RNA was extracted from 5 ml of broth culture grown to an OD600 of 0.5 in LBT medium. Cells were harvested by centrifugation (16 000 g, 1 min), washed in 1 ml GTC buffer [5 M guanidine thiocyanate, 0.5 % (w/v) Sarkosyl, 0.5 % (w/v) Tween 80, 100 mM 2-mercaptoethanol, 25 mM sodium citrate (pH 7)] and resuspended in 1 ml TRIzol reagent (Invitrogen). Cells were ruptured by two cycles of bead-beating with 0.5 ml zirconia beads (0.1 mm diameter) in a Mini-Beadbeater (Biospec) at 5000 r.p.m. for 1 min. RNA was isolated according to the manufacturer's instructions for the TRIzol reagent. The resulting RNA was resuspended in 50 μl DNase buffer [20 mM Tris/HCl (pH 7.5), 10 mM MgCl2] and 20 U RNase-free DNaseI (Roche) were added, followed by incubation at 37 °C for 1 h. Final purification of RNA was performed using RNeasy kit spin columns (Qiagen) according to the manufacturer's instructions. RNA was quantified using a NanoDrop ND-1000 spectrophotometer. RT-PCR reactions were performed with the primers rsbWRTF and sigFRTR (Table 1⇑) to determine whether rsbW and sigF were co-transcribed, and with primers 1802RTF and sigFRTR (Table 1⇑) to detect a potential MSMEG_1802rsbWsigF transcript, using the Titan One Tube RT-PCR system (Roche) according to the manufacturer's instructions. The RT step was carried out at 55 °C for 30 min, followed by 94 °C for 2 min and 30 cycles of PCR consisting of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s and elongation at 72 °C for 45 s for the rsbWsigF transcript or at 68 °C for 1 min for the MSMEG_1802rsbWsigF transcript. Final elongation was carried out at 72 °C for 5 min. As a template, 150 ng (for the rsbWsigF transcript) or 220 ng (for the MSMEG_1802rsbWsigF transcript) of RNA was added per reaction. Control reactions with 150 ng RNA or 25 ng genomic DNA of M. smegmatis as template were performed using Taq DNA polymerase (Roche) according to the manufacturer's instructions.

    Determination of transcriptional start sites (TSSs).

    The TSSs of sigF were mapped by 5′-RACE (rapid amplification of cDNA ends) using the components of the 3′/5′-RACE kit (Roche) according to the manufacturer's instruction. First-strand cDNA was synthesized from 5 μg total RNA of M. smegmatis with the sigF-specific primer sigF-RACE1. The resulting cDNA was purified and dA-tailed following the kit instructions. Purified dA-tailed cDNA was then used as a template for PCR using the Oligo dT-anchor primer and sigF-specific primer sigF-RACE2. The two resulting PCR products (550 bp and 1.1 kb) were gel purified and used as template for a second PCR using the PCR anchor primer and sigF-specific nested primer sigFKO2. These PCR products were cloned into pGEM-T Easy (Promega) according to the manufacturer's instructions. Three clones containing the correct-sized insert for both products were sequenced using primer sigFKO2 (500 bp product) or P1802R (1 kb product), respectively, and the last nucleotide before the poly-A tail to align with the genome sequence was chosen as the most likely TSS.

    DNA and protein sequence analysis.

    Preliminary sequence data for M. smegmatis mc2155 were obtained from the Institute for Genomic Research website at . Sequence data for M. tuberculosis were obtained from the Institut Pasteur website at .

    RESULTS AND DISCUSSION

    Construction of a sigF deletion mutant of M. smegmatis

    Sequence analysis of the region surrounding sigF of M. smegmatis showed that the gene is found in the same genetic locus as the homologous gene of M. tuberculosis (DeMaio et al., 1997), downstream of its putative anti-sigma factor (usfX or rsbW) and adjacent to a gene encoding a probable bifunctional protein involved in long-chain fatty acid synthesis (bcc) (Fig. 1a⇑). However, M. smegmatis does not appear to contain a homologue to usfY, a gene with unknown function in M. tuberculosis, and neither MSMEG_1802 (annotated as ChaB protein of unknown function) nor MSMEG_1806 (conserved hypothetical protein) has sequence similarity to UsfY. blastp analysis (Altschul et al., 1990) showed that SigF of M. smegmatis and M. tuberculosis are 78 % identical (data not shown).

    To determine whether sigF of M. smegmatis encoded an alternative sigma factor with a similar function to SigF of M. tuberculosis, a sigF deletion mutant of M. smegmatis was created by allelic-exchange mutagenesis. A construct for the replacement of sigF with the gentamicin resistance marker (aacC-1) was cloned into pPR23 (Pelicic et al., 1997) as described in Methods and transformed into M. smegmatis mc2155. Knockout mutants were selected as described previously (Pelicic et al., 1997). The left flank of the deletion construct is located on a 2 kb SacI fragment (Fig. 1b⇑). The coding region of sigF contains two SacI recognition sites. Replacement of sigF with aacC-1 removes those SacI sites, while introducing a new site at the end of the gentamicin resistance cassette, resulting in a band shift from 2 kb in the wild-type to 2.9 kb in the deletion mutant (strain SG128) in Southern hybridization analysis of SacI-digested genomic DNA probed with a radiolabelled left-flank PCR product (Fig. 1b, c⇑). While the fragments of both the wild-type and deletion mutant migrated slightly larger than expected, the difference between the two bands was consistent with the expected 900 bp shift (Fig. 1c⇑).

    In vitro phenotype of the sigF deletion strain

    To study the general growth characteristics of the sigF deletion strain, SG128, in comparison to the wild-type, we monitored the growth of both strains in a number of different media. No differences in growth rate, length of lag phase or final optical density reached were seen between the wild-type and SG128 ΔsigF when grown in LBT, 7H9/ADC or ST minimal medium (limited for either nitrogen or carbon) (data not shown). A previously described sigF deletion mutant of M. tuberculosis H37Rv also showed similar growth to its parent strain in vitro (Karls et al., 2006). In contrast, a sigF deletion mutant of M. tuberculosis CDC1551, while showing the same exponential growth rate, had a markedly decreased lag phase and grew to a threefold higher density as compared to the wild-type (Chen et al., 2000). SigF therefore does not appear to play a role in exponentially growing cultures of M. smegmatis, consistent with the phenotype of the M. tuberculosis H37Rv sigF mutant (Karls et al., 2006).

    SigF has been described as a stress-response sigma factor of slow-growing mycobacteria (DeMaio et al., 1996). We therefore wanted to investigate whether SigF was required for survival of the fast-growing M. smegmatis when exposed to various stress conditions. No significant differences between the sigF deletion strain and the wild-type were detectable after 4 h exposure to alkaline pH (pH 9), hyperosmotic (0.5 M NaCl) or hypoosmotic (H2O) shock or exposure to ethanol [5 % (v/v)] as determined by viable cell counts (Table 2⇓). However, the sigF deletion mutant was significantly impaired in survival after 4 h of heat shock (50 °C) (P<0.05) or exposure to acidic pH (pH 4) (P<0.01), with a 10-fold and 20-fold reduction in viable cells, respectively, as compared to the wild-type (Table 2⇓). Although no difference was seen between the wild-type and sigF-deletion mutant after 2 h exposure to hydrogen peroxide concentrations up to 5 mM, exposure to 7.5 mM hydrogen peroxide caused an approximately 100-fold decrease (P<0.005) in viability of SG128 ΔsigF as compared to the wild-type (Table 2⇓). Neither of the strains survived 2 h exposure to 10 mM hydrogen peroxide (data not shown). Complementation of the mutant, by supplying a single copy of the putative rsbWsigF operon to the cell (strain SG158), completely restored survival of both heat and acid stress as well as hydrogen peroxide stress (Table 2⇓), thus confirming that the observed phenotype was solely due to the loss of SigF activity. For a previously described sigF deletion mutant of M. tuberculosis, no defect in survival of heat shock was observed, while survival at acidic pH or exposure to oxidative stress-generating agents had not been tested (Chen et al., 2000). A similar phenotype of sensitivity to heat shock (52 °C) and oxidative stress (diamide or plumbagin) as seen here has been described for a M. tuberculosis mutant in a different alternative sigma factor, SigH (Raman et al., 2001). A sigH mutant of M. smegmatis also displayed increased sensitivity to oxidative stress by cumene hydroperoxide exposure, but no defect in survival of heat shock at 42 °C or 53 °C, or oxidative stress due to hydrogen peroxide exposure (Fernandes et al., 1999). However, a M. smegmatis double mutant in SigH and another stress-response sigma factor, SigE, was impaired in survival of both heat shock and oxidative stress (cumene hydroperoxide) (Fernandes et al., 1999). Raman et al. (2001) were able to identify a regulatory network between several stress-response sigma factors of M. tuberculosis, where SigH is responsible for stress-inducible expression of SigE, and both SigH and SigE are required for full induction of SigB. Considering the similar phenotype observed in the present study, it is conceivable that SigF of M. smegmatis either belongs to such a network of cross-regulation, or provides a separate, additional mechanism to respond to these stress conditions. This apparent redundancy of regulators for overlapping stress responses is further supported by the isolation of a M. tuberculosis mutant in yet another sigma factor, SigJ, which is also sensitive to hydrogen peroxide stress (Hu et al., 2004). The requirement of SigF and SigH for resistance of M. smegmatis to oxidative stress is in accordance with the recent observation that the promoter of dps, a gene expressed under oxidative stress and starvation, can only be recognized by RNA polymerase containing one of these sigma factors (Chowdhury et al., 2007).

    Table 2.

    Survival of stress conditions by wild-type M. smegmatis, the sigF deletion strain (SG128) and the complemented strain (SG158)

    Cells were exposed to stress conditions as detailed in Methods. Viable cell counts were performed immediately after exposure to the stress condition (t0) and after 2 h (t2) or 4 h exposure (t4). Data are shown as the means±sd of two or three independent experiments. nd, Not determined.

    Exposure to antibiotics is another condition that has been shown to induce expression of sigF in M. tuberculosis (Michele et al., 1999), and a sigF deletion mutant had an increased susceptibility to rifamycin drugs (Chen et al., 2000). We therefore determined the sensitivity of the M. smegmatis sigF mutant, SG128, to several antibiotics. No differences were seen in the MICs for rifampicin (4–8 μg ml−1), isoniazid (4 μg ml−1) or d-cycloserine (16–32 μg ml−1) between strain SG128 ΔsigF and the wild-type. The MIC for rifampicin determined here for M. smegmatis was considerably higher than results obtained previously for M. tuberculosis (0.25 μg ml−1) (Chen et al., 2000). It is therefore possible that the relatively higher resistance of M. smegmatis to rifamycin antibiotics is responsible for the differences in phenotype of the respective sigF mutants. As discussed above, strain SG128 was more sensitive to killing by hydrogen peroxide, yet we observed no differences in its susceptibility to isoniazid as compared to the wild-type. Detoxification of hydrogen peroxide in mycobacteria involves two pathways, requiring the catalase-peroxidase KatG and the alkyl hydroperoxide reductase AhpC, respectively (Pagan-Ramos et al., 1998). Both of these have been linked to altered susceptibility of the bacterium to isoniazid (Dhandayuthapani et al., 1996; Heym et al., 1993; Zhang et al., 1992). It therefore appears that SigF of M. smegmatis regulates the response to hydrogen peroxide via a pathway independent of KatG or AhpC. Supporting this notion, expression of ahpC was not altered in the sigF deletion strain SG128 as compared to the wild-type (A. Hümpel & G. M. Cook, unpublished observations).

    SigF is not required for survival of M. smegmatis in human neutrophils

    As described above, the sigF deletion strain SG128 displayed a pronounced sensitivity to hydrogen peroxide exposure. While M. smegmatis is generally considered a non-pathogenic saprophytic soil bacterium, it has been shown to cause opportunistic skin infections and soft tissues lesions (Brown-Elliott & Wallace, 2002; Wallace et al., 1988). We were therefore interested to determine whether strain SG128 ΔsigF displayed a reduced ability to survive in human neutrophils, which produce a wide range of oxidants, including hydrogen peroxide, when stimulated (Winterbourn et al., 2006). No difference in survival was seen between either the wild-type and strain SG128 ΔsigF or the complemented strain SG158 when incubated with human neutrophils at an m.o.i. of 3.75 (data not shown), suggesting that SigF was not involved in survival of M. smegmatis in human neutrophils. This is in agreement with results obtained for a M. tuberculosis sigF mutant, which was not impaired in growth within human monocytes or in its susceptibility to lymphocyte-mediated growth inhibition (Chen et al., 2000).

    rsbW and sigF are co-transcribed in M. smegmatis

    In M. tuberculosis, usfX and sigF are co-transcribed from a SigF-dependent promoter upstream of usfX (Beaucher et al., 2002). To determine whether rsbW and sigF are also co-transcribed in M. smegmatis, we carried out RT-PCRs using primers that anneal to rsbW (rsbWRTF) and sigF (sigFRTR), respectively (Fig. 2a⇓, Table 1⇑). Thus, a product can only be obtained if rsbW and sigF are co-transcribed. The expected size for a product was 1.02 kb. Using RNA isolated from an exponentially growing culture of M. smegmatis as template for RT-PCR, a product of the expected sizes was obtained (Fig. 2b⇓, lane 2). Control PCRs performed without the reverse transcriptase step yielded no products (data not shown). These results show that rsbW and sigF of M. smegmatis, like those of M. tuberculosis, are co-transcribed.

    Figure image not available in archive
    Fig. 2.

    RT-PCR analysis of rsbWsigF transcription. (a) Schematic of RT-PCR reactions performed. Open triangles indicate the annealing position of the primers used within the sigF locus of M. smegmatis. Sizes of expected products for a rsbWsigF and MSMEG_1802rsbWsigF transcript are indicated. (b) RT-PCR for a potential rsbWsigF transcript. A 5 μl volume of the reaction was analysed by agarose gel electrophoresis and DNA was visualized by ethidium bromide staining. Lane 1, DNA marker (band sizes in kb are indicated on the left); lane 2, RT-PCR product from 150 ng RNA. (c) RT-PCR for a potential MSMEG_1802rsbWsigF transcript. A 10 μl volume of the reaction was analysed by agarose gel electrophoresis and DNA was visualized by ethidium bromide staining. Lane 1, DNA marker (band sizes in kb are indicated on the left); lane 2, RT-PCR product from 220 ng RNA.

    As shown in Fig 1(a)⇑ and Fig. 2(a)⇑, an open reading frame (MSMEG_1802), annotated as chaB (unknown function) is present upstream of rsbW and transcribed in the same direction. To determine if this gene was also co-transcribed with rsbWsigF, a further RT-PCR was carried out using primers annealing within MSMEG_1802 (1802RTF) and sigF (sigFRTR), respectively, which would yield a 1.5 kb product from a potential MSMEG_1802rsbWsigF transcript (Fig. 2a⇑, Table 1⇑). As shown in Fig. 2(c)⇑, a faint product of 1.5 kb was obtained, showing that there was indeed a transcript present containing all three cistrons.

    To further characterize the sigF transcript, the TSS was determined by 5′-RACE analyses using a set of nested primers annealing within sigF (Table 1⇑). Two products were obtained, ∼500 bp and 1 kb in length, and the corresponding TSSs were mapped to an adenyl residue 17 bp upstream of the rsbW start codon and a guanyl residue 41 bp upstream of the MSMEG_1802 start codon, respectively (Fig. 3⇓). It therefore appears that sigF of M. smegmatis is transcribed from two promoters, one upstream of rsbW (PrsbW), and one upstream of MSMEG_1802 (PMSMEG_1802). Sequence analysis of PrsbW revealed regions with similarity to the −10 and −35 elements of mycobacterial promoters (Agarwal & Tyagi, 2006), but no similarity to known SigF-dependent promoters from mycobacteria (Rodrigue et al., 2007) (Fig. 3b⇓). In contrast, PMSMEG_1802 contained −10 and −35 elements with strong similarity to the consensus sequence of SigF-dependent promoters of M. tuberculosis (GGWWT-N16-17-GGGTAY) (Rodrigue et al., 2007) (Fig. 3a⇓).

    Figure image not available in archive
    Fig. 3.

    Analysis of promoter regions upstream of sigF. (a) PMSMEG_1802 region; (b) PrsbW region. Transcriptional start sites were determined by 5′-RACE (traces are given as reverse sequence), and the most likely +1 nucleotides are indicated in bold. The start codons of MSMEG_1802 and rsbW, as well as putative −10 and −35 elements of the promoters, are shown in bold; putative ribosome-binding sites (rbs) are underlined.

    Expression studies with transcriptional fusions to lacZ

    In order to determine the activity of the two promoters identified above, transcriptional fusions of both regions to lacZ were created. The rsbW-lacZ construct (pAH55) contained 570 bp upstream of the start codon of rsbW, and the MSMEG_1802-lacZ construct (pAH59) contained 210 bp upstream of the start codon of MSMEG_1802. Cells of wild-type M. smegmatis harbouring pAH55 had β-galactosidase activities of around 55 Miller units (MU) when grown to mid-exponential phase in LBT medium (data not shown). Cells of wild-type M. smegmatis harbouring pAH59 displayed significantly lower levels of β-galactosidase activity (P<0.001) of around 12 MU when grown to mid-exponential phase in LBT medium (Fig. 4⇓).

    Figure image not available in archive
    Fig. 4.

    Expression of MSMEG_1802-lacZ. (a) Effect of sigF deletion on expression. Cells of wild-type M. smegmatis (WT), the sigF deletion mutant SG128 (ΔsigF) or the sigF-complemented strain SG158 (sigF+) carrying the MSMEG_1802-lacZ construct pAH59 were grown to OD600 0.7–1.0 in LBT medium and β-galactosidase (β-Gal) activities, expressed in Miller units (MU), determined. Results are given as means±sd of three technical replicates. Significant differences from the wild-type (P<0.05) are indicated by two asterisks. (b) Expression throughout the growth cycle in LBT medium. Cells of M. smegmatis carrying pAH59 were grown in LBT medium, and OD600 (▪) and β-galactosidase activities, expressed as Miller units (grey bars) were monitored over time. β-Galactosidase activities are given as means±sd of duplicate assays performed on each sample. Representative results from two independent experiments are shown.

    As mentioned above, PMSMEG_1802 contained a sequence with similarity to SigF-dependent promoters of M. tuberculosis, while PrsbW did not. To determine if any of these promoters were dependent on SigF in M. smegmatis, both transcriptional fusion constructs were introduced into the sigF deletion strain SG128. Strikingly, expression of MSMEG_1802-lacZ was reduced sixfold in strain SG128 as compared to the wild-type (Fig. 4a⇑), with β-galactosidase activities of the mutant strain (2 MU) being similar to that of cells harbouring the empty vector pJEM15 (<2 MU, data not shown). These data strongly suggest SigF-dependence of PMSMEG_1802. Complementation of the sigF deletion in strain SG158 restored expression to wild-type levels (Fig. 4a⇑). In contrast, expression of rsbW-lacZ was unchanged in strain SG128 (data not shown), and PrsbW therefore appears to be expressed independently of SigF.

    To determine whether transcription from the PMSMEG_1802 or PrsbW promoters was induced upon entry into stationary phase, M. smegmatis cells harbouring pAH55 and pAH59 were grown in LBT medium, and β-galactosidase activities were monitored throughout growth. For pAH55 (rsbW-lacZ) similar levels of activity (around 60 MU) were observed throughout all growth phases (data not shown). In contrast, expression from PMSMEG_1802 (pAH59) was induced approximately twofold (from ∼12 MU to 23 MU) upon entry into stationary phase (Fig. 4b⇑). These results are comparable to data of M. tuberculosis sigF in Mycobacterium bovis BCG, where 3.6-fold induction upon entry into stationary phase as compared to exponentially growing cells was observed (DeMaio et al., 1996). However, the same study reported further induction (9.8-fold relative to exponential phase) in late stationary phase (DeMaio et al., 1996), whereas β-galactosidase activities of AH60 in the present study remained at a similar level over 40 h after entry into stationary phase.

    To determine whether the transcriptional fusion constructs of pAH55 and pAH59 were inducible in exponentially growing cultures by certain stress conditions, M. smegmatis cells harbouring these constructs were subjected to various stresses as described above. No change in rsbW-lacZ expression was seen after 2 h exposure to cold, mild heat shock (42 °C), ethanol, hydrogen peroxide, acidic or alkaline pH, hypoosmotic or hyperosmotic shock or exposure to rifampicin or isoniazid compared to unexposed control cells (data not shown). The only condition tested which led to a change in rsbW-lacZ expression was exposure to 1 MIC of d-cycloserine (16 μg ml−1) causing a twofold induction (P<0.001) from 53 MU to 100 MU. As a control, cells were also exposed to d-cycloserine in the presence of 200 μg chloramphenicol ml−1 to prevent any further synthesis of β-galactosidase. No induction compared to unexposed control cells was observed under these conditions (data not shown), confirming that the observed induction of rsbW-lacZ was indeed due to exposure to d-cycloserine. In M. tuberculosis, SigF appears to have a role in the regulation of genes involved in cell envelope biosynthesis (Geiman et al., 2004; Williams et al., 2007). It may therefore be that cell wall stress, e.g. due to exposure to d-cycloserine, causes an increase in expression of sigF, as observed here for M. smegmatis. The MSMEG_1802-lacZ construct did not show differential expression in response to any of the conditions tested here (data not shown). The general lack of induction of both constructs in response to most stress conditions tested is in accordance with previous results obtained after 2 h of stress exposure of M. tuberculosis (Manganelli et al., 1999). However, taking into account the different growth rates of M. smegmatis and M. tuberculosis, the conditions chosen in the present study were thought to be more comparable to the 24 h exposure that DeMaio et al. (1996) used for their study of M. bovis BCG, where significant induction of sigF was observed in response to stress conditions such as oxidative stress, exposure to alcohol or cold shock. In the present and previous studies of the activity of SigF in mycobacteria, little if any regulation was observed at the transcriptional level (Beaucher et al., 2002; Betts et al., 2002; DeMaio et al., 1996; Graham & Clark-Curtiss, 1999; Manganelli et al., 1999; Mariani et al., 2000), while more pronounced differences were seen at the translational and post-translational level (Beaucher et al., 2002; Michele et al., 1999). These findings support the argument that SigF activity may be regulated more strongly through the action of anti-sigma factors than through induction of sigF transcription (Beaucher et al., 2002).

    During the review process of this manuscript, a paper by Singh & Singh (2008) was published on the expression of sigF in M. smegmatis. The authors show that usfXsigF (synonymous with rsbWsigF) was expressed constitutively throughout the growth cycle and was not induced by the following stressors: rifampicin, streptomycin, NaCl, hydrogen peroxide and heat shock (45 °C), thus supporting the data presented here. However, usfXsigF was upregulated (<twofold) after 4–8 h exposure of exponentially growing cells to the following stressors: isoniazid, detergent (SDS), cold shock (15 °C), nutrient depletion (PBS), and ethambutanol (Singh & Singh, 2008). We did not test ethambutanol, PBS or SDS as inducers of rsbWsigF expression and the concentration of isoniazid used by Singh & Singh (2008) was 12-fold higher than that used in our study. Moreover, we routinely measured the transcriptional response of rsbWsigF after 2–4 h exposure to the test stressor and not the 4–8 h used in the Singh study.

    In conclusion, our findings demonstrate that sigF of M. smegmatis encodes an alternative stress-response sigma factor, similar to its homologue from M. tuberculosis. Allelic-exchange mutagenesis showed that this sigma factor is required by M. smegmatis for the survival of heat shock, acid stress and oxidative stress. SigF-mediated resistance to oxidative stress most likely involves a KatG- and AhpC-independent pathway, because the sigF deletion strain displays unaltered sensitivity to isoniazid, and ahpC expression was unaltered in the sigF deletion mutant as compared to the wild-type (unpublished results). Despite its role in resistance to oxidative stress, SigF does not appear to be required for the survival of M. smegmatis in human neutrophils. RT-PCR and 5′-RACE analyses showed that the sigF locus of M. smegmatis had two promoters, one located upstream of rsbW and one upstream of MSMEG_1802. Transcription of MSMEG_1802-lacZ was SigF-dependent and induced twofold upon entry into stationary phase, but could not be induced in exponential phase after exposure to various stressors. Expression of rsbW-lacZ was independent of SigF and remained constant throughout the growth cycle and under various stress conditions unless cells were challenged with d-cycloserine.

    Acknowledgments

    This work was funded by a New Zealand Lottery Health Grant. A. H. was supported by a University of Otago Postgraduate Scholarship.

    References

    1. Agarwal, N. & Tyagi, A. K. (2006). Mycobacterial transcriptional signals: requirements for recognition by RNA polymerase and optimal transcriptional activity. Nucleic Acids Res 34, 4245–4257.
    2. Alexeyev, M. F., Shokolenko, I. N. & Croughan, T. P. (1995). Improved antibiotic-resistance gene cassettes and omega elements for Escherichia coli vector construction and in vitro deletion/insertion mutagenesis. Gene 160, 63–67.
    3. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403–410.
    4. Beaucher, J., Rodrigue, S., Jacques, P. E., Smith, I., Brzezinski, R. & Gaudreau, L. (2002). Novel Mycobacterium tuberculosis anti-sigma factor antagonists control σF activity by distinct mechanisms. Mol Microbiol 45, 1527–1540.
    5. Betts, J. C., Lukey, P. T., Robb, L. C., McAdam, R. A. & Duncan, K. (2002). Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol 43, 717–731.
    6. Brown-Elliott, B. A. & Wallace, R. J., Jr (2002). Clinical and taxonomic status of pathogenic nonpigmented or late-pigmenting rapidly growing mycobacteria. Clin Microbiol Rev 15, 716–746.
    7. Chen, P., Ruiz, R. E., Li, Q., Silver, R. F. & Bishai, W. R. (2000). Construction and characterization of a Mycobacterium tuberculosis mutant lacking the alternate sigma factor gene, sigF. Infect Immun 68, 5575–5580.
    8. Chowdhury, R. P., Gupta, S. & Chatterji, D. (2007). Identification and characterization of dps promoter of Mycobacterium smegmatis: promoter recognition by stress specific ECF sigma factors σH and σF. J Bacteriol 189, 8973–8981.
    9. Cole, S. T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., Gordon, S. V., Eiglmeier, K., Gas, S. & other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544.
    10. DeMaio, J., Zhang, Y., Ko, C., Young, D. B. & Bishai, W. R. (1996). A stationary-phase stress-response sigma factor from Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 93, 2790–2794.
    11. DeMaio, J., Zhang, Y., Ko, C. & Bishai, W. R. (1997). Mycobacterium tuberculosis sigF is part of a gene cluster with similarities to the Bacillus subtilis sigF and sigB operons. Tuber Lung Dis 78, 3–12.
    12. Dhandayuthapani, S., Zhang, Y., Mudd, M. H. & Deretic, V. (1996). Oxidative stress response and its role in sensitivity to isoniazid in mycobacteria: characterization and inducibility of ahpC by peroxides in Mycobacterium smegmatis and lack of expression in M. aurum and M. tuberculosis. J Bacteriol 178, 3641–3649.
    13. Faldt, J., Dahlgren, C. & Ridell, M. (2002). Difference in neutrophil cytokine production induced by pathogenic and non-pathogenic mycobacteria. APMIS 110, 593–600.
    14. Fernandes, N. D., Wu, Q. L., Kong, D., Puyang, X., Garg, S. & Husson, R. N. (1999). A mycobacterial extracytoplasmic sigma factor involved in survival following heat shock and oxidative stress. J Bacteriol 181, 4266–4274.
    15. Gebhard, S., Tran, S. L. & Cook, G. M. (2006). The Phn system of Mycobacterium smegmatis: a second high-affinity ABC-transporter for phosphate. Microbiology 152, 3453–3465.
    16. Geiman, D. E., Kaushal, D., Ko, C., Tyagi, S., Manabe, Y. C., Schroeder, B. G., Fleischmann, R. D., Morrison, N. E., Converse, P. J. & other authors (2004). Attenuation of late-stage disease in mice infected by the Mycobacterium tuberculosis mutant lacking the SigF alternate sigma factor and identification of SigF-dependent genes by microarray analysis. Infect Immun 72, 1733–1745.
    17. Gonzalez-y-Merchand, J. A., Estrada-Garcia, I., Colston, M. J. & Cox, R. A. (1996). A novel method for the isolation of mycobacterial DNA. FEMS Microbiol Lett 135, 71–77.
    18. Graham, J. E. & Clark-Curtiss, J. E. (1999). Identification of Mycobacterium tuberculosis RNAs synthesized in response to phagocytosis by human macrophages by selective capture of transcribed sequences (SCOTS). Proc Natl Acad Sci U S A 96, 11554–11559.
    19. Hanahan, D., Jessee, J. & Bloom, F. R. (1991). Plasmid transformation of Escherichia coli and other bacteria. Methods Enzymol 204, 63–113.
    20. Heym, B., Zhang, Y., Poulet, S., Young, D. & Cole, S. T. (1993). Characterization of the katG gene encoding a catalase-peroxidase required for the isoniazid susceptibility of Mycobacterium tuberculosis. J Bacteriol 175, 4255–4259.
    21. Hu, Y., Kendall, S., Stoker, N. G. & Coates, A. R. (2004). The Mycobacterium tuberculosis sigJ gene controls sensitivity of the bacterium to hydrogen peroxide. FEMS Microbiol Lett 237, 415–423.
    22. Karls, R. K., Guarner, J., McMurray, D. N., Birkness, K. A. & Quinn, F. D. (2006). Examination of Mycobacterium tuberculosis sigma factor mutants using low-dose aerosol infection of guinea pigs suggests a role for SigC in pathogenesis. Microbiology 152, 1591–1600.
    23. Lee, M. H., Pascopella, L., Jacobs, W. R., Jr & Hatfull, G. F. (1991). Site-specific integration of mycobacteriophage L5: integration-proficient vectors for Mycobacterium smegmatis, Mycobacterium tuberculosis, and bacille Calmette-Guérin. Proc Natl Acad Sci U S A 88, 3111–3115.
    24. Manganelli, R., Dubnau, E., Tyagi, S., Kramer, F. R. & Smith, I. (1999). Differential expression of 10 sigma factor genes in Mycobacterium tuberculosis. Mol Microbiol 31, 715–724.
    25. Manganelli, R., Proveddi, R., Rodrigue, S., Beaucher, J., Gaudreau, L. & Smith, I. (2004). Sigma factors and global gene regulation in Mycobacterium tuberculosis. J Bacteriol 186, 895–902.
    26. Mariani, F., Cappelli, G., Riccardi, G. & Colizzi, V. (2000). Mycobacterium tuberculosis H37Rv comparative gene-expression analysis in synthetic medium and human macrophage. Gene 253, 281–291.
    27. Michele, T. M., Ko, C. & Bishai, W. R. (1999). Exposure to antibiotics induces expression of the Mycobacterium tuberculosis sigF gene: implications for chemotherapy against mycobacterial persistors. Antimicrob Agents Chemother 43, 218–225.
    28. Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
    29. Pagan-Ramos, E., Song, J., McFalone, M., Mudd, M. H. & Deretic, V. (1998). Oxidative stress response and characterization of the oxyR-ahpC and furA-katG loci in Mycobacterium marinum. J Bacteriol 180, 4856–4864.
    30. Parida, B. K., Douglas, T., Nino, C. & Dhandayuthapani, S. (2005). Interactions of anti-sigma factor antagonists of Mycobacterium tuberculosis in the yeast two-hybrid system. Tuberculosis (Edinb) 85, 347–355.
    31. Pelicic, V., Jackson, M., Reyrat, J. M., Jacobs, W. R., Jr, Gicquel, B. & Guilhot, C. (1997). Efficient allelic exchange and transposon mutagenesis in Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 94, 10955–10960.
    32. Raman, S., Song, T., Puyang, X., Bardarov, S., Jacobs, W. R., Jr & Husson, R. N. (2001). The alternative sigma factor SigH regulates major components of oxidative and heat stress responses in Mycobacterium tuberculosis. J Bacteriol 183, 6119–6125.
    33. Rodrigue, S., Brodeur, J., Jacques, P. E., Gervais, A. L., Brzezinski, R. & Gaudreau, L. (2007). Identification of mycobacterial sigma factor binding sites by chromatin immunoprecipitation assays. J Bacteriol 189, 1505–1513.
    34. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
    35. Singh, A. K. & Singh, B. N. (2008). Conservation of sigma F in mycobacteria and its expression in Mycobacterium smegmatis. Curr Microbiol 56, 574–580.
    36. Snapper, S. B., Melton, R. E., Mustafa, S., Kieser, T. & Jacobs, W. R., Jr (1990). Isolation and characterization of efficient plasmid transformation mutants of Mycobacterium smegmatis. Mol Microbiol 4, 1911–1919.
    37. Timm, J., Lim, E. M. & Gicquel, B. (1994). Escherichia coli–mycobacteria shuttle vectors for operon and gene fusions to lacZ: the pJEM series. J Bacteriol 176, 6749–6753.
    38. Waagmeester, A., Thompson, J. & Reyrat, J. M. (2005). Identifying sigma factors in Mycobacterium smegmatis by comparative genomic analysis. Trends Microbiol 13, 505–509.
    39. Wallace, R. J., Jr, Nash, D. R., Tsukamura, M., Blacklock, Z. M. & Silcox, V. A. (1988). Human disease due to Mycobacterium smegmatis. J Infect Dis 158, 52–59.
    40. Williams, E. P., Lee, J. H., Bishai, W. R., Colantuoni, C. & Karakousis, P. C. (2007). Mycobacterium tuberculosis SigF regulates genes encoding cell wall-associated proteins and directly regulates the transcriptional regulatory gene phoY1. J Bacteriol 189, 4234–4242.
    41. Winterbourn, C. C., Hampton, M. B., Livesey, J. H. & Kettle, A. J. (2006). Modeling the reactions of superoxide and myeloperoxidase in the neutrophil phagosome: implications for microbial killing. J Biol Chem 281, 39860–39869.
    42. Zhang, Y., Heym, B., Allen, B., Young, D. & Cole, S. (1992). The catalase-peroxidase gene and isoniazid resistance of Mycobacterium tuberculosis. Nature 358, 591–593.