The copper-responsive RicR regulon contributes to Mycobacterium tuberculosis virulence

Abstract

As with most life on Earth, the transition metal copper (Cu) is essential for the viability of the human pathogen Mycobacterium tuberculosis. However, infected hosts can also use Cu to control microbial growth. Several Cu-responsive pathways are present in M. tuberculosis, including the regulated in copper repressor (RicR) regulon, which is unique to pathogenic mycobacteria. In this work, we describe the contribution of each RicR-regulated gene to Cu resistance in vitro and to virulence in animals. We found that the deletion or disruption of individual RicR-regulated genes had no impact on virulence in mice, although several mutants had Cu hypersensitivity. In contrast, a mutant unable to activate the RicR regulon was not only highly susceptible to Cu but also attenuated in mice. Thus, these data suggest that several genes of the RicR regulon are required simultaneously to combat Cu toxicity in vivo or that this regulon is also important for resistance against Cu-independent mechanisms of host defense.

Importance: Mycobacterium tuberculosis is the causative agent of tuberculosis, killing millions of people every year. Therefore, understanding the biology of M. tuberculosis is crucial for the development of new therapies to treat this devastating disease. Our studies reveal that although host-supplied Cu can suppress bacterial growth, M. tuberculosis has a unique pathway, the RicR regulon, to defend against Cu toxicity. These findings suggest that Cu homeostasis pathways in both the host and the pathogen could be exploited for the treatment of tuberculosis.

FIG 1

Contribution of RicR-regulated genes to Cu resistance. (A) Model of the RicR regulon in M. tuberculosis. Cytoplasmic MymT can bind with up to six Cu⁺ ions (black circles). LpqS and Rv2963 are predicted to be membrane-associated proteins. RicR is autoregulated and also represses socAB under low-Cu conditions. MmcO is an MCO. (B) Cu sensitivity assays assessing the ability of RicR regulon mutants to survive at the indicated concentrations of CuSO₄ after 10 days. CFU were enumerated after 14 to 21 days of incubation on solid medium with trace amounts of Cu. Data are representative of at least two experiments, each done in triplicate. Abbreviations: comp., complemented; <L.O.D., below the limit of detection (which was 100 CFU). (C) Agar plate assay evaluating the Cu susceptibility of RicR regulon mutants. Serial dilutions of M. tuberculosis cultures were spotted onto 7H11-OADC agar plates with the CuSO₄ concentrations indicated. The contrast was adjusted to make the images clearer here and in Fig. 5 and 6. Data are representative of two independent experiments.

FIG 2

Contribution of RicR-regulated genes to virulence in mice. (A) CFU from lungs and spleens harvested on days 1 (n = 3) and 21 (n = 4) and at ~8 weeks (n = 4) from WT C57BL/6 mice infected with M. tuberculosis H37Rv RicR regulon single mutants. The initial dose of infection was ~500 to 1,000 CFU/mouse. Each datum point represents the average number of CFU from organs of three or four mice and the standard deviation. (B) Infection of WT C57BL/6 mice with a moderately large dose (~2,000 CFU/mouse) of WT or lpqS::ΦMycoMarT7 (MHD131) M. tuberculosis. Because of the slow movement and labored breathing of the mice infected with the lpqS mutant, they were sacrificed at day 26. N.D., not determined.

FIG 3

The transposon insertion in MHD131 increased the expression of mmcO. (A) Map of the lpqS locus in strain MHD131. The ΦMycoMarT7 transposon was inserted at nucleotide (nt) 130 of lpqS. lpqS is 393 nt long. (B) Immunoblot analysis showed that MmcO increased upon 4 h of Cu treatment or in a ricR mutant (left). Data are representative of three biological replicates. In another experiment, robust induction of MmcO was observed after 24 h of treatment with 50 µM CuSO₄ (right). (C) MmcO levels were higher in the lpqS mutant than in WT M. tuberculosis. Immunoblotting for dihydrolipoamide acetyltransferase (DlaT) was used as a loading control for all experiments.

FIG 4

Overexpression of mmcO resulted in high Cu resistance. (A) Deletion of mmcO from the lpqS mutant resulted in WT Cu resistance. A Cu sensitivity assay (top) was performed with the WT strain and the lpqS and lpqS mmcO mutant strains. Data are representative of two experiments each done in triplicate. <L.O.D., below the limit of detection (which was 100 CFU). Immunoblot analysis (bottom) shows that disruption of mmcO in the lpqS mmcO strain eliminates MmcO protein from this strain. Data are representative of three biological replicates. (B) Deletion of mmcO from the lpqS mutant did not alleviate the hypervirulence phenotype. The CFU counts in the lungs and spleens of C57BL/6 mice infected with the WT, lpqS, and lpqS mmcO strains are plotted. The initial dose of infection was ~500 CFU/mouse. Measurements were made on days 1 (n = 3), 21 (n = 4), and 56 (n = 4). Two-way analysis of variance and a Bonferroni posttest showed that at day 21, both the lpqS and lpqS mmcO mutant strains were statistically significantly different from the WT strain (***, P < 0.001) and that at day 56, the lpqS mutant strain was statistically significantly different from the WT strain (**, P < 0.01). The data represent the mean ± standard deviation of a typical experiment that was done twice.

FIG 5

Deletion of mmcO had no effect on virulence. (A) Cu sensitivity assay (top) of WT and mmcO mutant strains in the H37Rv and CDC1551 backgrounds. We also complemented the mmcO mutation in the H37Rv background. This is representative of two independent experiments, each done in triplicate. At 150 µM, no CFU of the H37Rv strains were detected. Immunoblot analysis (bottom) of the same strains with polyclonal antibodies to MmcO. Antibodies to dihydrolipoamide acetyltransferase (DlaT) were used to show even loading of cell lysates. (B) Agar plate assay assessing the Cu sensitivity of the M. tuberculosis strains in panel A. Serial dilutions of M. tuberculosis cultures were spotted onto agar with the indicated CuSO₄ concentrations. Data are representative of two independent experiments. (C) CFU counts in the lungs and spleens of C57BL/6 mice infected with WT, mmcO, and mmcO-complemented (comp.) M. tuberculosis strain H37Rv. The results shown are for days 1 (n = 3), 21 (n = 4), and 56 (n = 4). ns, not significant. The data represent the mean ± SD of a typical experiment that was done twice.

FIG 6

Deletion of both mmcO and mymT was not sufficient to attenuate M. tuberculosis in vivo. (A) Immunoblot analysis shows that MmcO protein was absent from an mmcO mymT double mutant. The WT, mmcO, mymT, mmcO mymT, and complemented (comp.) double mutant strains were used. Whole-cell lysates were analyzed with antibodies to MmcO. Dihydrolipoamide acetyltransferase (DlaT) was used as a loading control. (B) Agar plate assay (top) determining the Cu susceptibility of the WT, mmcO, mymT, mmcO mymT, and complemented double mutant strains. Data are representative of two independent experiments. The results of a liquid-based Cu sensitivity assay of the aforementioned M. tuberculosis strains are also shown (bottom). Data are representative of two experiments done in triplicate. (C) CFU counts in the lungs and spleens of C57BL/6 mice infected with the WT and mmcO; mymT, and mmcO mymT mutant M. tuberculosis strains. The results shown are for days 1 (n = 3), 21 (n = 4), and 56 (n = 4). The data represent means ± standard deviations. ns, not significant.

FIG 7

Repression of the entire RicR regulon sensitized M. tuberculosis to Cu and attenuated M. tuberculosis in mice. (A) DNA affinity chromatography shows that RicR dissociates from the lpqS promoter in the presence of Cu as previously described (top) while RicR_C38A constitutively bound to DNA regardless of the presence of Cu (bottom). Protein was eluted from the DNA with sequential increasing amounts of CuSO₄ or a high salt concentration (last). (B) A Cu sensitivity assay revealed that the p^c-ricR_C38A strain was hypersensitive to Cu. The WT CDC1551/pMV306 (empty vector), ricR/pMV306, ricR/pMV-ricR⁺, ricR/pMV-p^c-ricR⁺, and ricR/pMV-p^c-ricR_C38A strains were tested for Cu sensitivity. Data represent the mean ± standard deviation of one experiment that was done three times. (C) Constitutive repression of the ricR regulon attenuated M. tuberculosis growth in mice. CFU counts in the lungs of C57BL/6 mice infected with the WT/pMV306, ricR/pMV-p^c-ricR⁺, ricR/pMV-p^c-ricR_C38A, ricR/pMV306, or ricR/pMV-ricR⁺ (ricR comp.) strain are shown. The data were separated into two graphs for clarity but represent infections done within the same week. The WT infection in both panels represents the same experiment. The initial dose of infection was ~500 CFU/mouse. The data represent the means ± standard deviations at days 1 (n = 3), 21 (n = 4), and 63 (n = 4) postinfection. The WT and p^c-ricR_C38A data are representative of two independent infections. ****, P < 0.0001; ***, P < 0.001 (two-way analysis of variance with a Bonferroni posttest). (D) Proteasomal-degradation-defective strains were hypersensitive to Cu. The WT/pMV306 and mpa/pMV306, mpa/pMV-mpa⁺, pafA/pMV306, and pafA/pMV-pafA⁺ mutant H37Rv strains were exposed to Cu for 10 days. The data represent the means ± standard deviations of one typical experiment that was done three times. <L.O.D., below the limit of detection.