Mycobacterium tuberculosis gene Rv2136c is dispensable for acid resistance and virulence in mice

Summary

The gene Rv2136c is annotated to encode the Mycobacterium tuberculosis (Mtb) homologue of Escherichia coli’s undecaprenyl pyrophosphate phosphatase. In previous work, a genetic screen of 10,100 Mtb transposon mutants identified Rv2136c as being involved in acid resistance in Mtb. The Rv2136c:Tn strain was also sensitive to sodium dodecyl sulfate, lipophilic antibiotics, elevated temperature and reactive oxygen and nitrogen intermediates and was attenuated for growth and persistence in mice. However, none of these phenotypes could be genetically complemented, leading us to generate an Rv2136c knockout strain to test its role in Mtb pathogenicity. Genetic deletion revealed that Rv2136c is not responsible for any of the phenotypes observed in the transposon mutant strain. An independent genomic mutation is likely to have accounted for the extreme attenuation of this strain. Identification of the mutated gene will further our understanding of acid resistance mechanisms in Mtb and may offer a target for anti-tuberculosis chemotherapy.

1. Introduction

Mycobacterium tuberculosis (Mtb) is the most common cause of death worldwide due to a single bacterial pathogen, killing approximately two million people each year.¹ In the host, Mtb is found primarily within phagosomes of macrophages, where it encounters a relatively hostile environment. During the initial stages of infection, Mtb is capable of arresting phagosome fusion with acidic and hydrolytic lysosomes, allowing Mtb to reside in a mildly acidic environment (pH ~6.5) within non-activated macrophages.² However, this block in fusion is overcome upon macrophage activation by interferon gamma, resulting in acidification of the phagolysosome to pH ~4.5.^3,4 The acidic, hydrolytic environment of the phagolysosome may eradicate some microorganisms. However, Mtb is not sterilized in macrophages and can persist in a non-replicative state indefinitely.

Although many acid resistance mechanisms have been elucidated for enteric pathogens that encounter extremely acidic environments, such as the stomach, little is known about Mtb and its mechanisms to resist phagolysosomal acidity. Few acid sensitive Mtb mutants have been identified. Mtb deficient in MgtC, a putative magnesium transporter, is attenuated for growth in vitro under mildly acidic conditions with low Mg²⁺ concentrations.⁵ Mtb mutants lacking OmpAtb, an acid-dependent porin, are sensitive to low pH conditions in vitro and are attenuated in macrophages and mice.⁶ In a screen of 10,100 Mtb transposon mutants for impaired ability to recover from acid stress, 21 mutants were found to be hypersusceptible at pH 4.5.⁷ Killing of Mtb at pH 4.5 was largely dependent on the medium used. Most of the mutants only displayed enhanced susceptibility in 7H9 medium containing albumin and the dispersal agent Tween 80. At low pH, both Tween 80 and albumin can release fatty acids, which are toxic to Mtb.⁸ When a sodium phosphate-citrate buffer was used, albumin was omitted and Tween 80 was replaced with Tyloxapol, wild-type Mtb exhibited prolonged survival at pH 4.5. Of the 21 mutants, only two, Rv3671c:Tn and Rv2136c:Tn, remained acid sensitive under these conditions. Both strains were unable to maintain intrabacterial pH (pH_IB) homeostasis in vitro and within activated macrophages. Moreover, both mutants were also severely attenuated in mice, raising the possibility that pH_IB homeostasis is essential for Mtb’s growth and persistence in vivo. Studies are ongoing on the role of Rv3671c, a transmembrane serine protease, in acid resistance. In contrast, studies of Rv2136c have been delayed by prolonged and unsuccessful efforts to complement the phenotype of the transposon mutant with a wild-type copy of Rv2136c and/or members of the operon of which it is a part (O. Vandal, C. Darby, S. Ehrt, unpublished observations).⁷

Rv2136c is annotated as the Mtb homolog of the Escherichia coli uppP (also known as bacA), an undecaprenyl pyrophosphate phosphatase involved in synthesis of cell wall polymers including peptidoglycan, lipopolysaccharides, and teichoic acids. UppP dephosphorylates undecaprenyl pyrophosphate to generate undecaprenyl phosphate, the lipid carrier of biosynthetic intermediates for cell wall components. E. coli expresses four UppPs and a quadruple deletion mutant exhibits growth defects and abnormal cell morphology.⁹ Staphylococcus aureus and Streptococcus pneumoniae mutants lacking uppP are more susceptible than wild-type to bacitracin, an antibiotic that functions through sequestration of undecaprenyl pyrophosphate, and show reduced virulence in a mouse model of infection.¹⁰ Deletion of the E. coli uppP homolog in the non-pathogenic Mycobacterium smegmatis results in impaired biofilm formation as well as attenuated smegma development in an in vivo model.¹¹ The role of UppP in pathogenic mycobacteria has apparently not been explored.

If indeed the gene product of Rv2136c functions as a UppP, its disruption by transposon insertion could potentially result in altered membrane integrity, thereby rationalizing this strain’s sensitivity to acid as well as to sodium dodecyl sulfate (SDS), lipophilic antibiotics, elevated temperatures, and reactive oxygen and nitrogen intermediates.¹² A Streptococcus mutans strain disrupted in dgk, an undecaprenyl kinase, was sensitive to acidic conditions, providing evidence that membrane integrity is important for acid resistance.¹³ Membrane modifications can also result in enhanced susceptibility to other host stresses. A known Mtb cell wall mutant disrupted in lysX, which exhibits altered lysinylated phospholipid production, is attenuated in mice, with a 1 log₁₀ reduction in lung colony forming units (CFUs) at day 28.¹⁴ The Rv2136c:Tn mutant mouse phenotype was more severe than that of the lysX mutant, with 2 log₁₀ reduction in lung CFUs by day 21 and no CFUs recovered at day 56 post infection, even though the Rv2136c:Tn mutant grows normally under replicating conditions in vitro.¹² Because none of the observed phenotypes of Rv2136c:Tn could be complemented by a wild-type copy of Rv2136c in trans, we sought to determine whether disruption of Rv2136c is responsible for the pleiotropic phenotypes and extreme attenuation seen in this strain. If not, it would be appropriate to launch a search for another Mtb gene that is presumably disrupted in this strain and that is conditionally essential during infection of the mouse.

2. Materials and methods

2.1. Strains and media

All strains were cultivated in Middlebrook broth 7H9 at pH 6.6 with 0.2% glycerol, 0.2% dextrose, 0.5% bovine serum albumin, 0.085% NaCl and 0.05% Tween-80 (7H9) or on Middlebrook 7H11 agar containing 0.5% glycerol and 10% OADC (oleic acid, albumin, dextrose, catalase supplement) (Difco). For pH_IB and acid sensitivity measurements, bacteria were washed and resuspended in sodium phosphate citrate buffer pH 4.5 with 0.02% Tyloxapol (Pcit-Tyl-4.5). Hygromycin 50 μg/mL and streptomycin 25 μg/mL were included when required for selection. The complemented strain contains a plasmid with a constitutively active hsp60 promoter driving a wild-type copy of the Rv2136c gene, an attB site for chromosomal integration, and a streptomycin resistance cassette.

2.2. Mutant construction

ΔRv2136c was constructed via allelic exchange using a specialized transducing phage phAE87.¹⁵ Briefly, ~500 bp fragments containing the upstream and downstream regions of the Rv2136c locus were amplified by PCR and cloned into pJSC284-loxP to flank the hygromycin resistance cassette. The plasmid was digested with PacI and packaged into the temperature-sensitive phage phAE87. The phage was amplified in M. smegmatis at 30 °C and used to infect Mtb as described.¹⁶

2.3. Southern blot analysis

To confirm deletion of Rv2136c, genomic DNA from wild-type H37Rv and knockout candidates was digested with AflIII, separated by agarose gel electrophoresis, and transferred to a nylon membrane that was probed with a ~ 500 bp fragment containing the upstream gene Rv2137c. Upon exposure with the ECL Direct Nucleic Acid Labeling and Detection System (Amersham) a 3.5 kb band for wild-type Mtb and a 5.3 kb band for Rv2136c-deficient Mtb would result.

2.4. Real-time PCR

Mtb cultures were grown in 7H9 medium to an optical density (OD5_{80 nm}) of 2 and an equal volume of buffer containing guanidinium thiocyanate (4 M), sodium lauryl sulfate (0.5%), trisodium citrate (25 mM), and 2-mercaptoethanol (0.1 M) was added. Cultures were pelleted, resuspended in TRIzol, and bead beaten three times. RNA was extracted and qRT-PCR was performed using gene-specific Taqman probes and primers for sigA and Rv2136c (Biosearch Technologies).

2.5. pH_IB measurements

All Mtb strains were transformed with a pH-sensitive ratiometric GFP plasmid to monitor pH_IB and selected using hygromycin. Cultures were grown to mid log phase in 7H9, washed twice, resuspended in Pcit-Tyl-4.5 to an OD_{580 nm} = 0.1 (~5 × 10⁷ CFU/mL). At 2 and 6 days, the fluorescence of each well was read on a Molecular Dynamics SpectraMax M5 spectrofluorometer at excitation 395 nm, emission 510 nm and at excitation 475 nm, emission 510 nm. pH_IB was measured using the ratio of reading 1 (excitation 395 nm) to reading 2 (excitation 475 nm) and converting to pH_IB in reference to a calibration curve.

2.6. Measurement of acid sensitivity

Mid-log phase cultures were washed with Pcit-Tyl-4.5 buffer and centrifuged at 120 g for 10 minutes. Single cell suspension were adjusted to an OD₅₈₀ of 0.1 and incubated at 37 °C. After 2 and 6 days, samples were serially diluted and plated on 7H11 agar. CFUs were enumerated after 2 weeks.

2.7. Measurement of sensitivity to SDS

Mtb strains were grown to mid-log phase, a 10-fold dilution series was made from OD_{580 nm} of 0.1 to 0.001, and 5 μL was spotted onto 7H11 agar plates containing OADC with or without 0.0085% SDS.

2.8. Mouse infection

C57BL/6 8-week-old female mice (Jackson Laboratories) were infected by aerosol using a Middlebrook inhalation exposure system (Glas-Col). Early-log-phase Mtb cultures were aerosolized as single-cell suspensions in PBS to deliver ~100-200 CFU per mouse. Serial dilutions of lung homogenates from four mice per data point were plated onto 7H11 agar plates to quantify CFUs..

3. Results

3.1. Generation of an Mtb strain lacking Rv2136c

Our inability to complement the phenotypes observed in the Rv2136c:Tn mutant led us to generate an Mtb knockout strain, ΔRv2136c, in which the Rv2136c ORF was deleted and replaced with a hygromycin resistance cassette (Figure 1A). The first 20 base pairs and the last 48 base pairs of Rv2136c were retained in the knock out construct in an effort to maintain expression of the downstream gene Rv2135c, which is in a putative operon with Rv2136c. Deletion of Rv2136c was confirmed by Southern blot (Figure 1B) and quantitative real-time PCR (Figure 1C). Generation of this mutant confirms that Rv2136c is non-essential in vitro, as predicted by TraSH analysis.¹⁷

Figure 1. Generation of Rv2136c-deficient Mtb.

(A) Genomic region of wild-type Rv2136c and design of disrupted allele. Dashed lines indicate AflIII restriction products.

(B) Southern blot of AflIII restricted DNA from wild-type (WT) and ΔRv2136c (KO) strains.

(C) Level of Rv2136c transcripts assessed by quantitative Real-time PCR in WT and KO strains. Results show means ± SEM of one sample from three individual experiments. Samples in which transcript was not detected were labeled n. d.

3.2. Rv2136c is not responsible for acid or SDS sensitivity phenotype

We initially examined whether deletion of Rv2136c resulted in sensitivity to acid and SDS, as was observed in the Rv2136c:Tn mutant. We transformed ΔRv2136c and Rv2136c:Tn with a plasmid expressing a pH-sensitive ratiometric GFP. This pHGFP differentially fluorescences at two excitation wavelengths depending on the proton milieu, thus allowing for the non-invasive measurement of pH_IB over a dynamic range of pH 5.5 to 8.5. ΔRv2136c maintained pH_IB and viability after 2 and 6 days at pH 4.5, and was indistinguishable from wild-type Mtb in these respects. In contrast, Rv2136c:Tn was unable to maintain pH_IB and exhibited a 1 log₁₀ reduction in CFUs after 2 days at pH 4.5 (Figure 2A and B). Also, ΔRv2136c was not hypersensitive to SDS, while the Rv2136c:Tn mutant was attenuated for growth in the presence of 0.0085% SDS (Figure 2C). Additional phenotypes of Rv2136c:Tn, including sensitivity to heat (45 °C) and to various lipophilic antibiotics, were also not observed in ΔRv2136c (data not shown).

Figure 2. Rv2136c is dispensable for acid and SDS resistance and for virulence in mice.

(A) Intrabacterial pH (pH_IB) measurements of wild-type (H37Rv) and mutants after 2 and 6 days post exposure to phosphate citrate buffer plus tyloxapol pH 4.5. Means ± SEM of triplicate experimental samples are shown from one experiment, representative of two.

(B) Survival of wild-type and mutant strains after 2 and 6 days post exposure to pH 4.5. Means ± SEM of triplicate experimental samples are shown from one experiment, representative of two.

(C) Sensitivity of wild-type and mutant bacteria to 0.0085% SDS. Data are representative of results from two independent experiments.

(D) Growth and persistence of wild-type, ΔRv2136c, and ΔRv2136c Complemented (COMP) strains in mice. Data are means ± SEM from four mice per time point and group and are the results of one experiment.

3.3. Rv2136c is dispensable for Mtb’s virulence in mice

Although Rv2136c does not appear to be involved in acid resistance mechanisms, we sought to determine if it contributed to Mtb’s virulence. After infection of mice, ΔRv2136c exhibited the same logarithmic growth and subsequent persistence as wild-type (H37Rv) Mtb (Figure 2D). This is in contrast to the Rv2136c:Tn strain, which was highly attenuated for growth and persistence in the mouse model of infection.¹² Therefore, Rv2136c is not essential for establishment or persistence of infection in mice.

4. Discussion

In this study, construction and characterization of an Rv2136c-deficient Mtb strain allowed us to establish that Rv2136c does not contribute to acid resistance in Mtb and is not required for virulence in mice.

The previous report that initially characterized Rv2136 in Mtb described a transposon mutant strain wherein the transposon inserted within the last 80 nucleotides of the Rv2136c gene. This proximity to the 3’ end of the gene may lead to the production of a truncated protein with different functions, a dominant negative dysfunctional protein, or no protein at all. In an attempt to implicate Rv2136c in the phenotypes observed, the authors tried to complement the defects, but were unsuccessful in doing so.⁷ The inability to complement the Rv2136c:Tn phenotypes by supplying a wild-type copy of Rv2136c in trans indicated that Rv2136c may not be involved in acid resistance. By constructing a genetic knockout strain, we have confirmed this. The absence of parallel phenotypes among Rv2136c:Tn and ΔRv2136c suggests that a genomic mutation independent of the transposon insertion is responsible for the severe attenuation of this strain.

In addition to its dispensability towards acid resistance, Rv2136c is also not required for the virulence of Mtb. Although Rv2136c is the only Mtb gene with amino acid similarity to the E coli UppP, additional UppP(s) may exist that can compensate for the loss of Rv2136c. The majority of E. coli’s UppP activity is attributed to BacA; however, three additional UppPs (YbjG, PgpB, and YeiU) have been identified, each having unique substrate specificities.^18,19 Similar redundancy in UppP activity may exist in Mtb and the importance of each enzyme may depend on the environmental conditions encountered by the bacterium. Enzymatic studies are required to determine if Rv2136c does in fact function as a UppP.

There could be several explanations for the defects in the Rv2136c:Tn strain. Polar effects of the transposon on other genes in the Rv2136c operon are unlikely to have caused the phenotype, as the transposon mutant was not complemented even with the entire operon. Quantitative real-time PCR was performed on the Rv2136c:Tn strain to evaluate levels of Rv2135c transcript. We did not detect expression of Rv2135c in the Rv2136c:Tn strain, although it was readily detectable in wild type Mtb (data not shown). Complementation of Rv2136c:Tn strain with a construct containing Rv2136c, Rv2135c, Rv2134c, and Rv2133c resulted in over-expression of Rv2135c. Given that the phenotypes of these two Rv2136c:Tn strains were identical whether Rv2135c was undetectable or over-expressed, it appears that Rv2135c is not involved in the observed phenotypes. Generation of a dominant negative protein was another unlikely explanation, as transformation of wild-type Mtb with a plasmid encoding the putative truncated protein conferred no phenotype (data not shown). Most likely, the phenotype is due to an independent mutation at another locus. We have confirmed that there is no mutation within the Rv3671c gene, eliminating the possibility that a mutation in at least one gene known to control acid sensitivity could produce the observed phenotypes in the Rv2136c:Tn strain (data not shown). We are currently sequencing the Rv2136c:Tn genome to identify genetic mutations that may account for the acid sensitivity and the extreme attenuation in mice.

In conclusion, we do not think Rv2136c itself warrants further attention as a potential target for TB drug discovery. However, the product of the gene responsible for acid sensitivity could prove to be an attractive target for chemotherapy.