Ribonucleotide reduction in Mycobacterium tuberculosis: function and expression of genes encoding class Ib and class II ribonucleotide reductases

Abstract

Mycobacterium tuberculosis, the causative agent of tuberculosis, possesses a class Ib ribonucleotide reductase (RNR), encoded by the nrdE and nrdF2 genes, in addition to a putative class II RNR, encoded by nrdZ. In this study we probed the relative contributions of these RNRs to the growth and persistence of M. tuberculosis. We found that targeted knockout of the nrdF2 gene could be achieved only in the presence of a complementing allele, confirming that this gene is essential under normal, in vitro growth conditions. This observation also implied that the alternate class Ib small subunit encoded by the nrdF1 gene is unable to substitute for nrdF2 and that the class II RNR, NrdZ, cannot substitute for the class Ib enzyme, NrdEF2. Conversely, a DeltanrdZ null mutant of M. tuberculosis was readily obtained by allelic exchange mutagenesis. Quantification of levels of nrdE, nrdF2, nrdF1, and nrdZ gene expression by real-time, quantitative reverse transcription-PCR with molecular beacons by using mRNA from aerobic and O(2)-limited cultures showed that nrdZ was significantly induced under microaerophilic conditions, in contrast to the other genes, whose expression was reduced by O(2) restriction. However, survival of the DeltanrdZ mutant strain was not impaired under hypoxic conditions in vitro. Moreover, the lungs of B6D2/F(1) mice infected with the DeltanrdZ mutant had bacterial loads comparable to those of lungs infected with the parental wild-type strain, which argues against the hypothesis that nrdZ plays a significant role in the virulence of M. tuberculosis in this mouse model.

FIG. 1.

Genomic organization of RNR-encoding genes in M. tuberculosis H37Rv. The gene notation is that of TubercuList (http://genolist.pasteur.fr/TubercuList/).

FIG. 2.

Targeted knockout of RNR-encoding genes in M. tuberculosis. (A) Schematic representation of the inactivated nrdF2 allele showing the site of insertion of the hygromycin cassette, the position of the 1,126-bp BamHI-BglII probe (striped box) used for Southern blot analysis, and the extent of homologous DNA used in the knockout construct. (B) Southern blot analysis of nrdF2 recombinant strains. Chromosomal DNA from wild-type M. tuberculosis strain H37Rv (WT), a single-crossover recombinant carrying tandem copies of the inserted and wild-type alleles (SC), a single-crossover recombinant carrying a functional nrdF2 gene integrated at the attB locus (SCc), and a double-crossover recombinant with a crossover at the chromosomal locus of the nrdF2 allele arising from an SCc background (DCc) were digested with BamHI and probed with the BamHI-BglII probe containing the 5′ portion of the nrdF2 gene. (C) Schematic representation of the inactivated ΔnrdZ allele showing the 814-bp SalI deletion and the positions of the 2,011-bp SalI-PvuII probe (striped box) and the MluI sites used for Southern blot analysis. (D) Southern blot analysis of five ΔnrdZ allele replacement strains. Chromosomal DNA from wild-type M. tuberculosis H37Rv (WT) and five ΔnrdZ isolates were digested with MluI and probed with the SalI-PvuII probe shown in panel C. Abbreviations: Bg, BglII; H, HindIII; S, SalI; M, MluI; Pv, PvuII.

FIG. 3.

Inhibition of M. tuberculosis strains H37Rv and ΔnrdZ by HU, as determined by BACTEC susceptibility testing. Cultures were inoculated into Middlebrook 7H12 test medium supplemented with cyanocobalamin containing different concentrations of HU. Growth (which was directly proportional to the growth index) was monitored for 10 days. Symbols: ▪, H37Rv; □, ΔnrdZ.

FIG. 4.

Quantitative analysis of nrd gene expression in M. tuberculosis cultured under the Wayne model conditions (44, 45). (A) Growth of M. tuberculosis H37Rv in slowly stirred sealed tubes, showing a gradual cessation of growth as available oxygen becomes limiting. Three independent tubes were harvested at most of the times indicated by arrows; the exception was the 202-h time point, when two tubes were harvested. (B) Relative abundance of message as determined by quantitative RT-PCR by using molecular beacons in aerated (stirred) Dubos medium (open bars) and under Wayne model conditions. Copies of mRNA were normalized to 1,000 copies of sigA mRNA and the values shown are the means and standard errors. Log-phase samples were taken at 83 h (grey bars), NRPI samples were taken at 151 h (striped bars), NRPI/II samples were taken at 202 h (cross-hatched bars), and NRPII samples were taken at 302 h (solid bars).

FIG. 5.

Growth of the M. tuberculosis H37Rv and ΔnrdZ strains in B6D2/F₁ mice. Mice were infected with M. tuberculosis by using an aerosol, and bacillary loads in the lungs were determined over a 14-month infection period, as described in Materials and Methods. Symbols: ▪, mice infected with the parental strain; □, mice infected with the ΔnrdZ strain. Each point represents the mean for four mice, and the error bars indicate the standard deviations.