A Novel Mechanism of Growth Phase-dependent Tolerance to Isoniazid in Mycobacteria

Background: The mechanism underlying mycobacterial phenotypic tolerance to isoniazid is unknown.

Results: MDP1, a mycobacterial histone-like protein, down-regulates KatG expression.

Conclusion: Down-regulation of KatG by MDP1 causes growth phase-dependent phenotypic tolerance to isoniazid in mycobacteria.

Significance: Understanding the mechanism by which mycobacteria acquire tolerance to isoniazid is important for developing novel therapies.

Abstract

Tuberculosis remains one of the most deadly infectious diseases worldwide and is a leading public health problem. Although isoniazid (INH) is a key drug for the treatment of tuberculosis, tolerance to INH necessitates prolonged treatment, which is a concern for effective tuberculosis chemotherapy. INH is a prodrug that is activated by the mycobacterial enzyme, KatG. Here, we show that mycobacterial DNA-binding protein 1 (MDP1), which is a histone-like protein conserved in mycobacteria, negatively regulates katG transcription and leads to phenotypic tolerance to INH in mycobacteria. Mycobacterium smegmatis deficient for MDP1 exhibited increased expression of KatG and showed enhanced INH activation compared with the wild-type strain. Expression of MDP1 was increased in the stationary phase and conferred growth phase-dependent tolerance to INH in M. smegmatis. Regulation of KatG expression is conserved between M. smegmatis and Mycobacterium tuberculosis complex. Artificial reduction of MDP1 in Mycobacterium bovis BCG was shown to lead to increased KatG expression and susceptibility to INH. These data suggest a mechanism by which phenotypic tolerance to INH is acquired in mycobacteria.

Introduction

Tuberculosis is a disease caused by infection with Mycobacterium tuberculosis complex and remains a serious threat to health around the world. Approximately one-third of the world's population is infected with M. tuberculosis. The current World Health Organization report shows that 8.8 million new tuberculosis cases arose, and 1.4 million people died from tuberculosis in 2010. Although medications are indispensable for treating infectious diseases, one of the predominant problems in tuberculosis chemotherapy is the prolonged treatment duration. The current treatment of tuberculosis with first-line antitubercular agents including isoniazid (isonicotinic acid hydrazide, INH)³, rifampin, pyrazinamide, streptomycin, and ethambutol requires at least six months to cure the acute disease, yet there is still a relapse rate of 2 to 3% (1). For latent tuberculosis, the standard treatment takes six to nine months using INH alone.

The relatively long duration of tuberculosis chemotherapy is not only due to reduced metabolism based on the slow growth rates of the pathogens but also the emergence of drug-resistant cells. There are two possible mechanisms by which M. tuberculosis acquires drug resistance. First, spontaneous chromosomal mutations in genes related to drug resistance can result from irregular drug supply, inappropriate drug prescriptions, and poor patient adherence to treatment (2). Secondly, M. tuberculosis can acquire phenotypic drug resistance in the absence of genotypic alterations in drug-target genes (3). In particular, it is well known that INH tolerance is acquired by M. tuberculosis during the stationary phase, and requires prolonged tuberculosis chemotherapy.

INH is one of the key drugs used to control tuberculosis (4). It is critical in tuberculosis therapy because of its potent bactericidal activity against organisms growing actively in the pulmonary cavities, whereas the sterilizing activity of INH is reduced nearly 1,000-fold in M. tuberculosis cells during the stationary phase of growth (5).

The mode of action of this drug is complicated. INH is a prodrug that is converted into the active form by the mycobacterial catalase-peroxidase, KatG (6). The expression of KatG is regulated by an iron-containing transcription factor, furA, which is situated immediately upstream of katG (7). In its active form, INH inhibits both InhA, which is a primary target of INH (8), and an enoyl-acyl carrier protein reductase of the fatty acid synthase II (9, 10).

DNA sequencing of INH-resistant clinical isolates has revealed several mutations associated with resistance to INH. In addition to mutations in katG and inhA, ahpC (coding for alkyl hydroperoxide) (11) and ndh (coding for NADH dehydrogenase) (12) have also been reported to be associated with INH resistance. Mutations in kasA, which encodes ketoacyl acyl carrier protein synthase, are involved in INH resistance (13). However, later studies showed that overexpression of KasA did not lead to INH resistance in M. tuberculosis (14), and mutations in kasA are not likely to participate in INH resistance (4). Among these mechanisms, INH resistance in M. tuberculosis is most commonly associated with mutations in katG. In addition, a recent study by Ando et al. (15) revealed that mutations in the intergenic region of furA and katG also affected katG expression and conferred INH resistance. Although the reason why mycobacteria in the stationary phase acquire INH tolerance has not been fully elucidated, it is likely that the regulated expression of genes involved in the action of this drug is responsible for INH tolerance.

Transcriptional regulators are thought to play important roles in growth phase-dependent bacterial adaptive responses, including drug tolerance. Histone-like proteins are possible candidate transcriptional regulators in such responses. Recently, it was reported that Lsr2, a histone-like protein highly conserved in mycobacteria, inhibits a wide variety of DNA-interacting enzymes to regulate genes induced by antibiotics and those associated with inducible multidrug tolerance (16). Mycobacterial DNA-binding protein 1 (MDP1) is another histone-like protein in mycobacteria that binds to genomic DNA at guanine and cytosine residues. MDP1 is generally a negative regulator of gene expression (17) that participates in the slow growth rate of mycobacteria (18). Expression of this protein is enhanced in both stationary and dormant mycobacteria (19, 20).

In this study, we describe for the first time that MDP1 negatively regulates KatG expression, which, in turn, causes phenotypic tolerance to INH. The current study describes a novel molecular mechanism by which phenotypic drug tolerance is acquired in mycobacteria. This mechanism may strongly impact our understanding of phenotypic tolerance to INH in M. tuberculosis.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Antimicrobial Agents

Mycobacterium smegmatis mc²155 (WT) and its histone-like protein/MDP1-deficient strain (MDP1-KO) were kindly provided by Dr. Thomas Dick (Novartis Institute for Tropical Diseases). An MDP1-complemented strain (MDP1-Comp) was generated previously (21). All M. smegmatis strains were cultured in Luria-Bertani (LB) broth or on LB agar plates (Sigma) aerobically at 37 °C on a magnetic stirrer set to rotate at 130 rpm. Mycobacterium bovis BCG strains were cultured in 7H9 broth base (Becton Dickinson and Company) supplemented with glycerol, 10% albumin-dextrose-catalase (ADC), and 0.05% Tween 80 (7H9-ADC-Tween), or on 7H11 agar plates supplemented with oleic acid, albumin, dextrose, and catalase (OADC). Ethambutol, rifampin, levofloxacin, and INH were purchased from Sigma. Rifampin was dissolved in ethanol, whereas levofloxacin, INH, and ethambutol were dissolved in distilled water. Stock solutions of each drug were filter-sterilized through 0.2-μm pore-size polyethersulfone membrane filters (Iwaki) except for rifampin, which was dissolved in ethanol.

Broth Microdilution

For the estimation of the minimum inhibitory concentrations (MICs) of each antibiotic, we used the broth microdilution method as previously described (22). Briefly, serial 2-fold dilutions of compounds were added to LB broth (for M. smegmatis) or 7H9-ADC-Tween (for BCG) to achieve final concentrations ranging from 128–0.125 μg/ml. The diluted antibiotic was then dispensed into the wells of microdilution plates at 0.1 ml per well. Aliquots of mycobacterial cells were then inoculated to a final concentration of ∼10⁴ CFU/well. After incubation at 37 °C for 4 days (M. smegmatis) or 14 days (BCG), the MICs were determined as the lowest concentrations of compound that prevented visible growth.

Cell Viability against INH

M. smegmatis and BCG cells were grown aerobically at 37 °C in liquid medium under appropriate conditions for 2–6 days. At each time point, an aliquot of each culture was withdrawn and adjusted to an optical density at 600 nm (A₆₀₀) of 0.1 and subsequently diluted 1:100 in fresh medium. After addition of INH solution to a final concentration of 6.25 μg/ml (for M. smegmatis) or 0.125 μg/ml (for BCG), cells were grown aerobically at 37 °C for 24 h (M. smegmatis) or for 48 h (BCG). Serial 10-fold dilutions of the cell suspensions were plated on agar plates to estimate the number of viable bacteria in the inoculum. After incubation for 4 days (M. smegmatis) or 14 days (BCG) at 37 °C, colonies were counted, and the proportion growing in the presence of various drug concentrations was compared with the total number of viable bacteria in the inoculum.

RNA Extraction

Cells were suspended in 1 ml of TRIzol reagent (Invitrogen) and disrupted using a Mini-BeadBeater. After incubation for 5 min at room temperature, 0.2 ml of chloroform was added, and the samples were shaken vigorously for 15 s. Cell lysates were centrifuged at 12,000 × g for 10 min at 4 °C, and the colorless upper aqueous phases were transferred to fresh tubes. Total RNA in the aqueous phase was precipitated by mixing samples with isopropyl alcohol followed by centrifugation. The pellets were washed with 75% ethanol, dried, and resuspended in 100 μl of diethylpyrocarbonate-treated dH₂O. To remove the chromosomal DNA, samples were processed with a TURBO DNA-free kit (Applied Biosystems) according to the manufacturer's instructions. RNA quantification was performed by spectrophotometry using a NanoDrop 3300 (Thermo Scientific).

Microarray Analysis

A customized high-density oligonucleotide whole genome expression array (NimbleGen Systems) was designed for M. smegmatis mc²155 using the genome sequence and ORF predictions available from the J. Craig Venter Institute. Total RNA was extracted from M. smegmatis WT and MDP1-KO cells in the exponential phase (A₆₀₀ ∼ 0.8). The cDNA synthesis, hybridization, and scanning were performed by NimbleGen Systems. Microarray data analysis was performed using GeneSpring GX (Agilent Technologies). The data presented are the results from one experiment.

Real-time Quantitative RT-PCR (qRT-PCR)

To confirm the results of microarray analysis, qRT-PCR was performed using Power SYBR green (Applied Biosystems). To generate cDNA samples, equal amounts of RNA were reverse transcribed using the High-capacity reverse transcription kit (Applied Biosystems) according to the manufacturer's instructions. Primers were designed using Primer Express software (version 2.0; Applied Biosystems), and the sequences of the primers used were as follows: sigA (forward), 5′-CGTTCCTCGACCTCATCCA-3′; sigA (reverse), 5′-GCCCTTGGTGTAGTCGAACTTC-3′; katG (forward), 5′-GACCGCGAATGACCTTGTGT-3′; and katG (reverse), 5′-TGTCGGACTGGGCATAAACC-3′. A standard curve was generated for relative quantification of the PCR products, and a control reaction lacking reverse transcriptase was performed for every RNA sample. The major housekeeping σ factor gene, sigA, was used to normalize mRNA levels. Analysis was performed on triplicate biological samples that were each assayed in triplicate.

Western Blotting

Mycobacterial cells were grown at 37 °C and harvested at the indicated time points. To obtain whole-cell protein extracts, cells were washed three times with PBS and disrupted using a Mini-BeadBeater as described previously (23). Quantities of protein were determined by the Bradford assay (24) using the Bio-Rad protein assay kit (Bio-Rad). The same amount of total protein from each strain (2 μg per well) was separated by SDS-PAGE. Immunoblotting was carried out after SDS-PAGE and transfer of proteins to a PVDF membrane by incubation with anti-MDP1 and anti-KatG antibodies, as described previously (25). Anti-heat shock protein Hsp65 (GroEL2) antibody was used as a loading control (26, 27). The expression levels of different proteins were analyzed using the public domain software Image J (a Java image processing program developed by the National Institutes of Health Image for Macintosh).

Reduction of Nitroblue Tetrazolium (NBT)

This test detects INH activation by KatG and is dependent on the reduction of NBT by superoxide-free radical in the presence of INH to form an insoluble formazan. Reduction was monitored qualitatively following electrophoresis of whole-cell extracts on native gels, as described previously (28). Before samples were subjected to native PAGE, protein concentrations were quantified by Bradford assay. The gel was soaked in 50 ml of 50 mm sodium phosphate (pH 7.0) containing 68 mg of INH, 12.5 mg of NBT, and 15 ml of 30% H₂O₂. Color development was complete after 30 min. The gel was then rinsed with distilled water and soaked in 7% acetic acid, 1% glycerol in distilled water before visualization. The levels of KatG activity in each strain observed in the native PAGE gel were also analyzed using ImageJ software.

Transformation and Isolation of Recombinant M. bovis BCG Strains

Preparation of M. bovis BCG competent cells and electroporation procedures were performed according to standard procedures (29). Briefly, M. bovis BCG strain Tokyo was grown in 7H9-OADC-Tween until an A₆₀₀ of 0.6 was reached. Cells were harvested by centrifugation, washed several times, and resuspended in one-tenth of their original volume of 10% glycerol to obtain competent cells. The MDP1-antisense plasmid (18) as well as the empty vector, pMV261, were introduced into BCG-competent cells using a Gene pulser II (Bio-Rad) with the following settings: 2.5 kV, 129 ohms, 50 μF. After electroporation, cells were resuspended in 4 ml of 7H9-ADC-Tween and incubated for 20 h at 37 °C. Following incubation, appropriate dilutions were added to 7H11 agar containing 10 μg/ml kanamycin. One clone of empty vector transformant (pMV261) and two clones of transformants carrying MDP1-antisense plasmid (pAS-MDP1-1 and pAS-MDP1-2) were chosen for further analysis.

Expression and Purification of Recombinant MDP1

Recombinant MDP1 was obtained as described previously (30). Briefly, the oligonucleotide primers (forward, 5′-CCCCATATGAACAAAGCAGAGCTCATTGAC-3′; reverse, 5′-CCCAAGCTTTTTGCGACCCCGCCGAGCGG-3′) were synthesized based on the nucleotide sequences of BCG and M. tuberculosis mdp1. The amplified DNA fragment were digested with NdeI and HindIII, cloned into the same sites of pET22b+ (Novagen, Darmstadt, Germany), and introduced into Escherichia coli BL21(DE3). Expression of recombinant protein was induced by addition of 0.1 mm isopropyl-1-thio-β-d-galactopyranoside, and the cells were then incubated for 10 h at 22 °C. After incubation, cells were disrupted, and the supernatant was collected by centrifugation 8,000 × g for 30 min. After filtering the supernatant through a 0.22-μm membrane, it was applied to a 1-ml Hi-Trap chelating column charged previously with 100 mm NiSO₄ and equilibrated with 20 mm sodium phosphate, pH 7.4, 10 mm imidazole, and 0.5 m NaCl. After washing out unbound proteins, the protein was eluted with the same buffer containing 300 mm imidazole. The fractions containing MDP1 were collected and dialyzed against PBS. The purity of the protein was confirmed by staining as a single band with Coomassie Brilliant Blue R-250 following separation by SDS-PAGE.

Electrophoretic Mobility Shift Assay (EMSA)

Primers (forward, 5′-CTCTGACAGGCGCCAATGCG-3′; reverse, 5′-GACCAGAAGGCTACTGCTTT-3′) were used to amplify a 80-bp fragment containing the furA promoter region, as described by Milano et al. (31). The amplified DNA fragment was labeled with digoxigenin with a digoxigenin gel shift kit, second generation (Roche Diagnostics), according to the manufacturer's protocol. EMSA was also performed using the same kit according to the manufacturer's instructions. Briefly, purified recombinant MDP1 protein was incubated with 40 fg of digoxigenin-labeled double-stranded DNA fragments in a final volume of 25 μl. Incubations were carried out at 4 °C for 2 h in a solution of 10 mm Tris-HCl (pH 7.6), 50 mm NaCl, 5% (v/v) glycerol, 1 mm EDTA, and 1 mm DTT. For the competition assay, unlabeled oligonucleotide was added to a 100-fold excess. The reaction mixtures were loaded onto 6% polyacrylamide gels that were pre-electrophoresed at 100 V for 1 h in 1× Tris borate-EDTA buffer consisting of 89 mm Tris, 89 mm borate, and 2 mm EDTA (pH 8.3). Polyacrylamide gels were electrophoresed at 100 V at ambient temperature until the bromphenol blue dye front reached the bottom. The probes were transferred onto a positively charged nylon membrane (Roche Diagnostics) and detected according to the kit protocol.

RESULTS

MDP1-deficient M. smegmatis Showed Increased Susceptibility to INH

To determine whether MDP1 is involved in the acquisition of drug sensitivity in mycobacteria, we analyzed the susceptibility of M. smegmatis WT, MDP1-KO, and MDP1-Comp cells to antibiotics. As shown in Fig. 1A, a marked increase in susceptibility to INH was observed in MDP1-KO cells. We determined the MICs of INH for each strain using the broth microdilution method described previously (32). As presented in Table 1, the MIC of MDP1-KO cells was 8.0 μg/ml, whereas the MICs for the other strains were 4-fold higher. In contrast, M. smegmatis WT, MDP1-KO, and MDP1-Comp were equally susceptible to the other drugs tested, including ethambutol, rifampin, and levofloxacin (Table 1).

FIGURE 1.

Comparison of the INH susceptibilities of M. smegmatis WT, MDP1-KO and MDP1-Comp. A, broth microdilution method. INH susceptibilities were analyzed by inoculating each bacterial strain in 7H9-ADC-Tween culture medium containing serially diluted INH. B, comparison of CFU count. Each strain was exposed to INH at a concentration below the MIC. After incubation for 48 h, cell suspensions were inoculated onto LB agar to estimate the number of viable bacteria in the inoculum. MDP1-KO CFUs (solid bar) decreased markedly after exposure to INH compared with WT (open bar) and MDP1-Comp (hatched bar) CFUs. Mean values and S.E. for three experiments are shown. conc., concentration.

TABLE 1.

MICs against various antibiotics by the broth microdilution method

EB, ethambutol; RFP, rifampin; LVFX, levofloxacin.

	EB	RFP	LVFX	INH
	μg/ml	μg/ml	μg/ml	μg/ml
WT	1.0	8	0.5	32
KO	1.0	8	0.5	8
Comp	1.0	8	0.5	32

To clarify the effects of culture medium components on susceptibility to antibiotics, we performed broth microdilution assays using 7H9-ADC-Tween as the culture medium. The MICs of each strain in 7H9-ADC-Tween were the same as those in LB broth (data not shown). To confirm the differential susceptibility to INH between strains, we counted the CFUs in each strain after exposure to INH at a concentration below the MIC for MDP1-KO cells (Fig. 1B). We found that MDP1-KO cells showed decreased viability following treatment with INH compared with WT and MDP1-Comp cells. These results suggested that susceptibility to INH is affected by the presence of MDP1 in M. smegmatis cells. Our results are consistent with the data recently reported by Dahl et al. (33).

Increased Transcription of katG in MDP1-KO

To investigate the effect of MDP1 on the expression of genes related to INH resistance, we performed DNA microarray analysis to compare gene expression profiles between M. smegmatis WT and MDP1-KO strains. The expression levels of the genes related to INH resistance, such as inhA and ahpC, were similar in MDP1-KO and WT cells (Table 2). However, a significant increase in katG expression was observed in MDP1-KO cells.

TABLE 2.

Expression of genes related to INH resistance analyzed by microarray

	WT	MDP1-KO
MSMEG_3461 (katG)	2.910772	6.552576
MSMEG_3151 (inhA)	3.056271	2.7222855
MSMEG_4891 (ahpC)	164.84808	168.63435

To confirm the increased katG transcription in MDP1-KO cells, total RNA from M. smegmatis WT, MDP1-KO, and MDP1-Comp cells was isolated and used for qRT-PCR analysis of katG expression. The expression levels of katG mRNA were determined by the comparative threshold cycle method and then normalized to sigA expression. As expected, MDP1-KO cells possessed a 2-fold higher expression level of katG mRNA compared with WT cells (Fig. 2). In contrast, MDP1-Comp cells exhibited 2-fold lower katG expression than the WT strain, whereas the viability of MDP1-Comp cells treated with INH was similar to that of WT cells.

FIGURE 2.

Analysis of katG expression by qRT-PCR. Transcription of katG was quantified by qRT-PCR. The relative expression levels of katG in MDP1-KO (solid bar) and MDP1-Comp (hatched bar) cells were compared with that in WT cells (open bar) after normalization to the housekeeping gene, sigA. The cDNA level is expressed as the average of three replicates. For statistical analysis, one-way ANOVA test was used to obtain the p value. *, p < 0.01.

Increased KatG Protein Expression in MDP1-KO

To support the results of DNA microarrays and qRT-PCR, Western blotting was used to analyze the expression level of KatG protein. As a loading control, anti-Hsp65 antibody was used to confirm that the protein concentrations were equal in all samples. The results demonstrated that M. smegmatis WT and MDP1-Comp cells produced similar amounts of KatG, whereas KatG expression was increased in MDP1-KO cells (Fig. 3A). These data suggest that MDP1 regulates the KatG protein level in M. smegmatis cells.

FIGURE 3.

Influence of MDP1-deficiency on KatG protein expression and INH activation. M. smegmatis wild-type (WT), MDP1-KO (KO), and MDP1-Comp (Comp) cells were grown at 37 °C for 48 h. A, KatG expression was analyzed by Western blotting using rabbit anti-KatG antibodies. B, the level of activation of INH by KatG in each strain was estimated by NBT reduction assay. INH-dependent reduction of NBT to formazan was monitored on a native PAGE gel.

Enhanced Activation of INH in MDP1-KO Cells

Next, we investigated whether INH activation is increased in MDP1-KO cells, which produce larger amounts of KatG than WT cells (Fig. 3A). The INH-dependent NBT reduction assay is commonly used to estimate the level of INH activation by KatG because the KatG enzyme produces free radicals upon activation of INH (34, 35).

The NBT reduction assay was performed using cell lysates derived from M. smegmatis WT, MDP1-KO, and MDP1-Comp cells. A single band of activity was detected in each lane at the same distance on the gel as a purple formazan product. As shown in Fig. 3B, whole-cell lysates of MDP1-KO cells exhibited a higher capacity to reduce NBT to formazan in the presence of INH than other strains. To confirm that the different NBT-reducing activities between strains were due to different KatG expression levels in each strain and not to amino acid substitutions, we compared their katG nucleotide sequences and found no amino acid substitutions (data not shown). These data suggest that increased expression of KatG results in the activation of INH, and consequently, MDP1-KO cells are less viable following treatment with INH.

MDP1 Affects INH Susceptibility in Growth Phase-dependent Manner

Our previous study revealed that the expression of MDP1 is up-regulated in the stationary phase of mycobacterial culture (21). Therefore, we investigated whether the different expression levels of MDP1 during different growth states produced the INH-resistant phenotype in M. smegmatis. The expression of MDP1 was shown to increase in both WT and MDP1-Comp cells during the stationary growth phase (day 6 in Fig. 4). In contrast, expression of KatG was observed during the logarithmic growth phase in both WT and MDP1-Comp cells but diminished during the stationary growth phase (Fig. 4). MDP1-KO cells, however, showed enhanced KatG expression during the logarithmic growth phase (Fig. 2), whereas KatG was still expressed even in the stationary growth phase (Fig. 4).

FIGURE 4.

Expression of MDP1 and KatG proteins during the logarithmic and stationary growth phases in M. smegmatis. Analysis of the expression of MDP1 and KatG during the logarithmic (day 2) and stationary (day 6) growth phases was performed by Western blotting. WT, M. smegmatis; KO, WT MDP1-KO; Comp, MDP1-Comp.

A time course analysis was performed investigating the expression levels of MDP1 and KatG and the INH susceptibility of M. smegmatis WT cells. Western blotting revealed a time-dependent up-regulation of MDP1 expression, whereas the expression of KatG was inversely decreased (Fig. 5A). Cell viability in the presence of INH was also found to vary at each time point. On day 2 (exponential phase), a significant reduction in the number of CFUs was observed after treatment with INH at a concentration of 6.25 μg/ml, whereas most of the cells survived exposure to INH at the same concentration on day 6 (stationary phase) (Fig. 5B). These results revealed that the acquisition of INH tolerance occurred in a time-dependent manner and was positively correlated with increased MDP1 expression.

FIGURE 5.

Growth phase-dependent expression of MDP1 and KatG and susceptibility to INH in M. smegmatis WT. A, the expression levels of MDP1 and KatG were monitored by Western blotting on days 2, 3, and 6. B, INH susceptibility was analyzed by CFU assay. Initially, M. smegmatis WT cells were grown at 37 °C in LB broth without INH. On days 2, 3, and 6, cells were collected and exposed to INH at a concentration of 0.0 μg/ml or 6.25 μg/ml for 2 days. After INH treatment, the drug was removed from the culture media, and cells were diluted and plated onto INH-free LB agar plates. The number of viable cells obtained following treatment with INH (open bar) was compared with the number obtained without INH treatment (solid bar) at each time point. Mean values and S.E. for three experiments are shown. The statistical significance of the differences in CFUs between the different time points was determined by one-way ANOVA. *, p < 0.01.

MDP1 Affects INH Susceptibility of M. bovis BCG

To clarify whether MDP1 influences the susceptibility of other mycobacterial species to INH, drug sensitivity was measured in two M. bovis BCG clones carrying the MDP1-antisense plasmid (pAS-MDP1-1 and pAS-MDP1-2) and the reference strain (pMV261). M. bovis BCG is an attenuated strain of M. bovis that belongs to the M. tuberculosis complex. The architecture of the katG gene and its upstream regulatory region are the same in M. bovis BCG and M. tuberculosis. Western blotting showed that MDP1 expression was decreased by half, and KatG expression was increased >2-fold, in both pAS-MDP1-1 and pAS-MDP1-2 compared with pMV261 (Fig. 6A). We then determined the MICs of these strains in the presence of INH and found that mutants with reduced MDP1 expression showed increased susceptibility to INH, consistent with data obtained for M. smegmatis MDP1-KO (Figs. 6 and 7). These results suggest that MDP1 influences the susceptibility of not only the avirulent rapidly growing M. tuberculosis complex but also the slow growing M. tuberculosis complex to INH.

FIGURE 6.

MDP1 affects the expression of KatG protein and INH susceptibility in M. bovis BCG. A, comparison of MDP1 and KatG expression in two M. bovis BCG clones carrying the MDP1-antisense plasmid and a reference strain. BCG cells were grown at 37 °C for 10 days. MDP1-antisense plasmids (pAS-MDP1−1 and −2) reduced the expression of MDP1 compared with the control plasmid (pMV261). B, comparison of the INH susceptibility of the two M. bovis BCG clones carrying MDP1-antisense plasmids and the reference strain. The effect of INH was monitored by the broth microdilution method. conc., concentration.

FIGURE 7.

CFU assay to estimate the INH susceptibility of recombinant BCG harboring the MDP1 antisense plasmid. Knockdown of MDP1 expression by antisense production (solid and hatched bars) in BCG revealed increased susceptibility of the bacterial cells to INH compared with recombinant BCG harboring the control plasmid (open bar) in the presence of 0.125 μg/ml INH for 48 h. Mean values and S.E. for three experiments are shown. The statistical significance of the differences in CFUs between the different time points was determined by one-way ANOVA. *, p < 0.01.

MDP1 Binds to DNA Sequence in FurA-KatG Promoter Region

A previous report revealed that katG is negatively regulated by FurA, a homologue of the ferric uptake regulator Fur, by binding to its promoter region (36). It was also shown that furA and katG are co-transcribed from a common regulatory region located immediately upstream of the furA gene (31). Furthermore, inactivation of furA resulted in katG up-regulation, which increased the sensitivity of mycobacterial cells to INH (7). We hypothesized that MDP1 may directly bind to the promoter region of the FurA-KatG operon and performed gel-shift assays to test the hypothesis. We observed that co-incubation of MDP1 with a digoxigenin-labeled DNA fragment containing the furA-katG promoter region resulted in the disappearance of a DNA band. This was due to altered electrophoretic mobility of the MDP1-DNA complex because of the high basic charge of MDP1 (pI = 12.4). We confirmed the specificity of this binding by showing the absence of a band shift in the presence of non-digoxigenin-labeled competitor (Fig. 8). These data show that MDP1 has the ability to bind to the furA-katG promoter region.

FIGURE 8.

Gel-shift assay to analyze the interaction between MDP1 and the furA-katG DNA promoter region. Recombinant MDP1 was incubated with either digoxigenin (DIG)-labeled DNA probes including the sequence of the furA-katG promoter region alone, or together with a 100-fold weight excess of non-digoxigenin labeled probes. Left lane, digoxigenin-labeled probes alone. Middle lane, digoxigenin-labeled probes co-incubated with recombinant MDP1 protein. Right lane, digoxigenin-labeled probes co-incubated with recombinant MDP1 in the presence of an excess of the non-digoxigenin-labeled probes.

DISCUSSION

MDPI is a conserved histone-like DNA binding protein found in mycobacteria including M. tuberculosis, Mycobacterium leprae, and M. smegmatis. Because chromatin-associated proteins are thought to organize the bacterial chromosome and exert a regulatory influence on transcription, recombination and DNA replication (37), it has been postulated that MDP1 may contribute to the regulation of gene expression. In this study, we found that MDP1 down-regulates the expression of KatG in mycobacteria.

KatG is an antioxidant enzyme that converts hydrogen peroxide (H₂O₂) into water and oxygen via its catalase activity (38, 39). Because mycobacteria encounter oxidative stress inside host cells, antioxidant proteins such as KatG play a significant role in intracellular survival. Sherman et al. (40) reported that INH-resistant katG mutants acquired a second mutation resulting in hyperexpression of alkyl hydroperoxidase (AhpC) to compensate for the loss of KatG activity in the detoxification of organic peroxides. However, overexpression of AhpC did not result in increased M. tuberculosis survival, and deletion of ahpC did not alter the expression of katG. Thus, AhpC does not compensate for the loss of KatG activity. In addition to DNA-binding activity, we discovered recently that MDP1 enzymatically converts H₂O₂ into water and oxygen in the presence of Fe²⁺ (41). Therefore, it is conceivable that both KatG and MDP1 play a significant role in H₂O₂ detoxification in mycobacterial cells. MDP1 rather than AhpC may compensate for reduced KatG activity, and therefore their expression levels are reciprocally regulated.

It is also known that katG is regulated negatively by FurA, a homologue of the ferric uptake regulator Fur, which is encoded by a gene located immediately upstream of katG (36). Similar to FurA, MDP1 has affinity to iron (41), and the expression levels of both FurA and MDP1 are influenced by the environmental iron concentration (42). EMSA assay revealed that MDP1 has an actual capacity to bind to the promoter region of furA-katG, suggesting that MDP1 is a trans-acting factor that regulates the transcription of the FurA-KatG operon by directly binding to its promoter region. Future studies, including chromatin immunoprecipitation assays, will clarify the effect of MDP1 on the furA-katG genomic region.

Drug tolerance is a phenomenon seen in many microorganisms during their growth in vivo (3, 43, 44). It is known that phenotypic tolerance occurs when the environmental or physiological status of the bacteria change. For example, environmental factors such as low pH (45), depletion of certain nutrients (46), and high magnesium or calcium concentration (47) induce phenotypic tolerance. The inhibition of growth occurs in the stationary phase and is the most common cause of reduced drug susceptibility in all bacteria (48).

It has been reported that the expression of histone-like proteins varies with the age of the culture. For example, Fis is a cofactor in a site-specific recombination system that is expressed at a high level in the early exponential phase of E. coli growth (49). In contrast, integration host factor is found to be most abundant when the culture enters the stationary phase (50). Because the expression of MDP1 is enhanced in the stationary growth phase (20), this molecule may affect the gene profile of cells in the stationary phase. The current study shows that alteration of MDP1 expression induces growth phase-dependent tolerance to INH by controlling KatG expression in M. smegmatis. A schematic diagram showing a hypothetical mechanism of INH tolerance regulated by MDP1 is provided in Fig. 9. To our knowledge, this is the first description of a molecular mechanism underlying growth phase-dependent tolerance to INH in mycobacteria.

FIGURE 9.

Schematic diagram showing the hypothetical mechanism of INH tolerance regulated by MDP1. INH is a prodrug that is converted to an activated, bactericidal form by the enzymatic activity of KatG. In the logarithmic growth phase, cells expressing KatG show susceptibility to INH. In contrast, up-regulation of MDP1 expression occurs in cells during the stationary phase, which leads to suppression of KatG expression probably by direct binding of MDP1 to the furA-katG promoter region. Thus, cells with reduced KatG activity become phenotypically tolerant to INH.

Regulation of KatG expression appears to be conserved in several mycobacteria, including M. tuberculosis and M. smegmatis (31). We show here that reduction of MDP1 expression by antisense knockdown increases the sensitivity of M. bovis BCG to INH (Fig. 7). M. bovis BCG and M. tuberculosis show similar susceptibilities to INH, and the DNA sequences of their katG genes and upstream regulatory regions are identical. Taken together, MDP1 may affect the expression level of KatG and the susceptibility of pathogenic M. tuberculosis to INH. Our data suggest that the artificial reduction of MDP1 expression may enhance the efficacy of INH and shorten the length of tuberculosis chemotherapy against both active and latent diseases.