Mycobacterium tuberculosis universal stress protein Rv2623 regulates bacillary growth by ATP-Binding: requirement for establishing chronic persistent infection

Abstract

Tuberculous latency and reactivation play a significant role in the pathogenesis of tuberculosis, yet the mechanisms that regulate these processes remain unclear. The Mycobacterium tuberculosisuniversal stress protein (USP) homolog, rv2623, is among the most highly induced genes when the tubercle bacillus is subjected to hypoxia and nitrosative stress, conditions thought to promote latency. Induction of rv2623 also occurs when M. tuberculosis encounters conditions associated with growth arrest, such as the intracellular milieu of macrophages and in the lungs of mice with chronic tuberculosis. Therefore, we tested the hypothesis that Rv2623 regulates tuberculosis latency. We observed that an Rv2623-deficient mutant fails to establish chronic tuberculous infection in guinea pigs and mice, exhibiting a hypervirulence phenotype associated with increased bacterial burden and mortality. Consistent with this in vivo growth-regulatory role, constitutive overexpression of rv2623 attenuates mycobacterial growth in vitro. Biochemical analysis of purified Rv2623 suggested that this mycobacterial USP binds ATP, and the 2.9-A-resolution crystal structure revealed that Rv2623 engages ATP in a novel nucleotide-binding pocket. Structure-guided mutagenesis yielded Rv2623 mutants with reduced ATP-binding capacity. Analysis of mycobacteria overexpressing these mutants revealed that the in vitro growth-inhibitory property of Rv2623 correlates with its ability to bind ATP. Together, the results indicate that i) M. tuberculosis Rv2623 regulates mycobacterial growth in vitro and in vivo, and ii) Rv2623 is required for the entry of the tubercle bacillus into the chronic phase of infection in the host; in addition, iii) Rv2623 binds ATP; and iv) the growth-regulatory attribute of this USP is dependent on its ATP-binding activity. We propose that Rv2623 may function as an ATP-dependent signaling intermediate in a pathway that promotes persistent infection.

Figure 1. Generation of Δrv2623 M. tuberculosis strain.

(A) Genomic organization of the rv2623 gene locus. Genes appear as large arrows in their native orientation. Small arrows represent forward and reverse primers used for long range PCR. Sizes of the rv2623 deletion and hyg-insertion are indicated. The location of the radiolabeled probe (black bar) as well as relevant Sph1 sites (S) is indicated as they pertain to the Southern blot. (B) An autoradiograph of the Southern blot shows radiolabeled Sph1 fragments from the wild type (lane 1) and Δrv2623 strain (lane 2). The sizes indicated represent those expected for wild type and the deletion mutant.

Figure 2. In vitro characterization of Δrv2623.

(A) Growth curves of cultures inoculated (10⁶ CFU/ml) into 7H9+10% OADC+0.05% Tween 80 (top) and in minimal Sauton's media (bottom); Erdman (closed boxes, solid line) and Δrv2623 (open boxes, dashed line) cultures. Error bars represent the standard error of the means; each curve is a combination of at least three independent experiments. (B) Overexpression of Rv2623 in M. smegmatis. The top panel represents serial dilutions (1∶10) of the empty vector pMV261-containing negative control strain and Rv2623-overexpressing strain harboring pMV261::rv2623. Diluted stationary phase M. smegmatis culture was spotted (5 µl) onto solid 7H10 media supplemented with 10% OADC and kanamycin (40 µg/mL). Photographs were taken after three days incubation at 37°C. Growth of corresponding strains in liquid medium was assessed based on the time to detection determined using a BD BACTEC 9000MB system (bottom). The various strains were inoculated at 10⁴ CFU/ml in triplicates. Data shown are representative of several independent experiments. ***p<0.001.

Figure 3. In Vivo growth of and pathology caused by Δrv2623 in guinea pigs.

Outbred Hartley guinea pigs given an aerosol challenge of ∼30 CFU were assessed for pulmonic bacterial burden (A,D) and the severity of lung pathology (B,E). Closed box, open box, and triangle represent guinea pigs infected with Erdman, in (A,B,D,E), Δrv2623, in (A,B,D,E), and Δrv2623 attB::P_hsp60 Rv2623, in (A,B), or Δrv2623 attB::P_rv2623 Rv2623, in (D,E). Comparing the wildtype Erdman and the Δrv2623 strains: *p<0.05; **p<0.01; ***p<0.001. Comparing the Δrv2623 and the Δrv2623::complemented strains (Δrv2623 attB::P_hsp60 Rv2623 or Δrv2623 attB::P_rv2623 Rv2623): ⁺⁺⁺p<0.001; ⁺p<0.05. (C) Hematoxylin & Eosin-stained lung sections (40 days post infection) from guinea pigs infected with Erdman (top), Δrv2623 (middle), and Δrv2623 attB::P_hsp60 Rv2623 (bottom) M. tuberculosis. Error bars represent the standard error of the mean.

Figure 4. In Vivo growth of Δrv2623 in mice.

(A) C57BL/6 mice infected with various strains of M. tuberculosis via the aerosol route with a low dose (∼100 CFU) were assessed for lung bacteria burden. Wild type Erdman (closed box, solid line); Δrv2623 (open box, dashed line) and the complemented strain Δrv2623 P_hsp60::Rv2623 (triangle, dashed line). (B) Survival curve of C3H/HeJ mice infected via aerosol with 750–1000 CFU. Erdman and Δrv2623-infected mice are represented by solid and dashed lines, respectively. (C) Kinetics of infection, established via aerolization (inoculum: ∼1,000 CFU) of wildtype Erdman (dark), Δrv2623 (dark grey), and the complemented strain Δrv2623 P_hsp60::Rv2623 (light grey), as assessed by lung bacterial burden. **p<0.01; ***p<0.001.

Figure 5. M. tuberculosis Rv2623 is a nucleotide-binding USP.

High Performance Liquid Chromatography (HPLC) analysis of endogenously bound nucleotides from purified His₆-Rv2623. Nucleotides species were identified based on their specific retention times on the Mono Q HR 5/5 column, represented by peaks in absorbance at 260 nm, which correspond to that of nucleotide standards (not shown). Bound nucleotides were extracted by boiling, and separated and quantified from a standard curve that relates absorbance peak area to the known amount of ATP injected onto the column (inset).

Figure 6. Structure and phylogeny of Rv2623 from M. tuberculosis.

(A,B) A ribbon representation of the Rv2623 monomer (A) and dimer (B) with bound ATP (sticks) and Mg²⁺ (chocolate spheres). The three, mutually perpendicular pseudo-two-fold axes of the dimer are represented by lines with double arrows (along x, y) and a central ellipse (along z). The atoms of the bound ATP are colored cyan (carbon), red (oxygen), blue (nitrogen), and orange (phosphorus) in (A) and (B). (C) A structure-based sequence alignment of Rv2623, the N631 subfamily consensus, Methanococcus jannaschii protein 0577 (MJ0577), and domains 1 and 2 of Rv2623. Invariant residues in the alignment (>85% conserved in N631) are shaded in bold red and similarities are boxed in blue but left unshaded. Regions with consensus ATP binding motifs comprising L1/L2 (domain 1) and L3/L4 (domain 2) are colored dark violet and smudge, respectively. The positions of the mutated amino acids (D15, G117) are indicated in green. The structure-based sequence alignment was produced using ESPript and the structural representations were produced using PyMOL.

Figure 7. Design and stability of ATP-binding–deficient Rv2623 mutants.

(A) A ribbon and stick representation of the mutation sites within the ATP binding pocket of domain 1. Mg²⁺ is shown as a green sphere; dotted lines indicate hydrogen-bonding contacts; atoms that constitute ATP are colored cyan (carbon), red (oxygen), blue (nitrogen), and orange (phosphorus). (B) The ATP-binding capacity of mutant Rv2623 was compared to that of wild type protein following nucleotide extraction and HPLC. Data presented are derived from analysis of three independent protein preparations. ATP binding capacity is expressed as: [(the amount of ATP bound in mutant)/(the amount of ATP bound by wild type Rv2623) * 100]. (C) Thermal denaturation curves of two individual protein preparations of Rv2623_WT (WT-1, WT-2) as compared to Rv2623_G117A (G117A) and Rv2623_D15E (D15E). The data is expressed as the negative first-derivative of the fluorescence intensity as a function of temperature.

Figure 8. ATP binding by Rv2623 is required for its ability to attenuate growth.

Growth of M. smegmatis (A) or M. tuberculosis (B) overexpressing wild type or mutant Rv2623 inoculated at 10⁴, 10⁵ CFU/ml into a BD BACTEC 9000MB system; the time to detection of triplicate cultures is indicated. While overexpression of wildtype Rv2623 attenuates mycobacterial growth, overexpression of either ATP-binding-deficient Rv2623 mutant (Rv2623_G117A or Rv2623_D15E) significantly shifts the time to detection back towards that of the pMV261 vector control strain. ***p<0.001.