Nonclassical transpeptidases of Mycobacterium tuberculosis alter cell size, morphology, the cytosolic matrix, protein localization, virulence, and resistance to β-lactams

Abstract

Virtually all bacteria possess a peptidoglycan layer that is essential for their growth and survival. The β-lactams, the most widely used class of antibiotics in human history, inhibit D,D-transpeptidases, which catalyze the final step in peptidoglycan biosynthesis. The existence of a second class of transpeptidases, the L,D-transpeptidases, was recently reported. Mycobacterium tuberculosis, an infectious pathogen that causes tuberculosis (TB), is known to possess as many as five proteins with L,D-transpeptidase activity. Here, for the first time, we demonstrate that loss of L,D-transpeptidases 1 and 2 of M. tuberculosis (LdtMt1 and LdtMt2) alters cell surface morphology, shape, size, organization of the intracellular matrix, sorting of some low-molecular-weight proteins that are targeted to the membrane or secreted, cellular physiology, growth, virulence, and resistance of M. tuberculosis to amoxicillin-clavulanate and vancomycin.

FIG 1

Field emission scanning electron microscopy to assess cell morphology and length of M. tuberculosis. FESEM images of late-exponential-phase wild-type (blue) (A), M1 (red) (B), M1c1 (pink) (C), M2 (green) (E), M2c2 (maroon) (F), M12 (purple) (H), M12c12 (lime green) (I), M12c1 (yellow) (J), and M12c2 (orange) (K) are shown. Magnification, ×20,000 (bar, 200 nm). (D, G, and L) Statistical analysis of cell lengths for wild-type, mutant, and complemented strains. Values are means and standard deviations (SD); n ≥ 100 for each sample. One-way analysis of variance (ANOVA) was performed, with all strains being compared to wild-type M. tuberculosis (*, P < 0.05). M12, M12c1, and M2 were found to be significantly shorter than wild-type M. tuberculosis.

FIG 2

Scanning and transmission electron microscopy to assess intracellular morphology. Images of wild-type M. tuberculosis (A, E, and I), M1 (B, F, and J), M2 (C, G, and K), and M12 (D, H, and L) were obtained from culture at exponential phase of in vitro growth by FESEM (A to D) and TEM (E to L). (D) White arrows indicate cell surface depressions in M12 cells. (I, J, K, and L) Black arrows indicate the cell wall of M. tuberculosis. (A to D) Magnification, ×35,000 (bar, 200 nm); (E to H) magnification, ×80,000 (bar, 100 nm); (I to L) magnification, ×100,000 (bar, 20 nm).

FIG 3

Growth of wild-type, M1, M2, and M12 strains in both in vitro and in vivo settings. In vitro growth (A, E, and I) of wild-type (blue), M1 (red), M1c1 (pink), M2 (green), M2c2 (maroon), M12 (purple), M12c12 (lime green), M12c1 (yellow), and M12c2 (orange) in 7H9 broth. In vivo growth in the lungs (B, F, and J) and spleens (C, G, and K) of mice infected with the strains listed above was assessed. Bacterial burdens in the lungs and spleen were determined by enumerating CFU. Values are means ± SD (n = 4/group). Kaplan-Meier survival analysis (D, H, and L) of mice infected with all strains (n = 15/group) is also shown.

FIG 4

Proteomic analysis of cytosolic proteins of wild-type M. tuberculosis versus M12. (A) Distribution of proteins whose targeting was altered in M12. Proteins that are otherwise membrane bound or secreted by wild-type M. tuberculosis are sequestered in the cytoplasm in M12. (B) Gaussian curve of the frequency distribution of the molecular masses of proteins whose localization has been altered. Proteins in the wild-type strain have a geometric mean molecular mass of 36.17 ± 0.4928 kDa, with a median of 30.6 kDa. Proteins in M12 that show an altered distribution have a geometric mean molecular mass of 21.63 ± 2.291 kDa, with a median of 17.0 kDa. Two-tailed Student's t test was performed to assess significance (*, P < 0.05).

FIG 5

Activity of amoxicillin-clavulanate and vancomycin against M. tuberculosis strains. Bacterial burden in the lungs and spleens of mice infected with the wild type (circles), M12 (squares), or M12c12 (triangles) and treated with isoniazid plus ethambutol (A and B), amoxicillin-clavulanate (C and D), vancomycin (E and F), or amoxicillin-clavulanate plus vancomycin (G and H) or not treated (I and J). Because of the different levels of infection at the start of treatment, CFU counts were normalized to those at week 0 and are plotted as a percentage, where the CFU count at week 0 is designated 100%. Normalized means ± SD are shown for all treatments (n = 4/time point/group). Two-tailed Student's t test (*, P < 0.05) was performed at week 12 for each treatment and infecting strain to assess the significance of the treatment relative to no treatment.