Expression of a novel mycobacterial phosphodiesterase successfully lowers cAMP levels resulting in reduced tolerance to cell wall-targeting antimicrobials

Abstract

cAMP and antimicrobial susceptibility in mycobacteriaAntimicrobial tolerance, the ability to survive exposure to antimicrobials via transient nonspecific means, promotes the development of antimicrobial resistance (AMR). The study of the molecular mechanisms that result in antimicrobial tolerance is therefore essential for the understanding of AMR. In gram-negative bacteria, the second messenger molecule 3'',5''-cAMP has been previously shown to be involved in AMR. In mycobacteria, however, the role of cAMP in antimicrobial tolerance has been difficult to probe due to its particular complexity. In order to address this difficulty, here, through unbiased biochemical approaches consisting in the fractionation of clear protein lysate from a mycobacterial strain deleted for the known cAMP phosphodiesterase (Rv0805c) combined with mass spectrometry techniques, we identified a novel cyclic nucleotide-degrading phosphodiesterase enzyme (Rv1339) and developed a system to significantly decrease intracellular cAMP levels through plasmid expression of Rv1339 using the constitutive expression system, pVV16. In Mycobacterium smegmatis mc²155, we demonstrate that recombinant expression of Rv1339 reduced cAMP levels threefold and resulted in altered gene expression, impaired bioenergetics, and a disruption in peptidoglycan biosynthesis leading to decreased tolerance to antimicrobials that target cell wall synthesis such as ethambutol, D-cycloserine, and vancomycin. This work increases our understanding of the role of cAMP in mycobacterial antimicrobial tolerance, and our observations suggest that nucleotide signaling may represent a new target for the development of antimicrobial therapies.

Keywords: 3′,5′-cAMP; antimicrobials tolerance; bioenergetics; mycobacteria; peptidoglycan; phosphodiesterases.

Figure 1

Phylogenetic analysis, structure prediction, recombinant expression, and cell-free extract in vitro activity of Rv1339 expressed in M. smegmatis mc²155.A, maximum-likelihood phylogenetic consensus tree of selected members of typical (PF02112) and atypical (PF12706) phosphodiesterase (PDE) class-II families. B, ribbon representation of a homology model of Rv1339 established using Modeller 9.23 (115) with the crystal structure of the metallo-β-lactamase fold protein YhfI from Bacillus subtilis (PDB: 6KNS) as the template. YhfI was found to be a homodimer in solution by size-exclusion chromatography (29). The structure exhibits the alpha-beta-sandwich configuration characteristic of the metallo-β-lactamase fold, which consists of two β sheets of seven β strands each (purple) and α helices (green) capping the β sheet. The two zinc cations in the active sites are depicted in black. The figure was created using UCSF Chimaera (116). C, zoomed view of one of the active sites of the homology model of Rv1339 showing the coordinating histidine and aspartate residues. Asp180 was chosen for mutagenesis because it appears to be involved in the coordination of both zinc cations. D, Western blot of the pVV16 (empty vector) control, Rv1339-expressing, and Rv1339 D180-expressing M. smegmatis mc²155 strains. The proteins were probed with an α-His antibody, and each lane was loaded with 50 μg of clear soluble lysate. E, relative PDE activity of the clear soluble lysate after 30 min of incubation compared to time 0 for the pVV16 (empty vector) control (black), Rv1339-expressing (red) and Rv1339 D180-expressing (gray) M. smegmatis mc²155 strains. The data are presented as the means ± SDs of two biological replicates and three technical replicates. Unpaired two-tailed Student’s t tests were used to compare the data, and p < 0.05 was considered significant. ∗∗∗∗p < 0.0001 as analyzed by Student's t test. PDB, Protein Data Bank.

Figure 2

Expression of Rv1339 in M. smegmatis mc²155 compromises growth, depletes cAMP, and redirects bioenergetics toward ATP synthesis.A, growth curves in 7H9 medium of the pVV16 (empty vector) control (black), Rv1339-expressing (red) and Rv1339 D180-expressing (gray) M. smegmatis mc²155 strains. B, intracellular cAMP and c-di-AMP levels in these strains measured by LC-MS. Oxygen consumption rates (C) and extracellular acidification rates (D) of the strains measured using Seahorse XFP analysis. The first three measurement cycles were obtained in the absence of a carbon source after which a mixture of glucose and glycerol was injected to a final concentration of 0.2% each. E, fractions of total AXP species in the mid-log phase of bacterial growth of the strains measured by LC-MS. F, calculated adenylate energy charge (AEC; (F)) and fold change in the NADH/NAD⁺ ratio measured using the alcohol dehydrogenase method (G). Increased AEC and NADH/NAD⁺ ratios in the Rv1339-expressing strain indicate a shift toward synthesis of ATP. The statistical analysis was performed with a two-tailed Student’s t test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 and ∗∗∗∗p < 0.0001. The data are presented as the means ± SDs of three biological replicates and three replicates.

Figure 3

Expression of Rv1339 alters the transcriptome, compromises cell wall integrity, and disrupts peptidoglycan biosynthesis.A, RNA-Seq of the pVV16 (empty vector) control and Rv1339-expressing M. smegmatis mc²155 strains in the mid-log phase of growth. Principal component analysis (PCA) assessing the quality of data produced using normalized counts subjected to variance stabilizing transformation (VST). Technical replicates appear to cluster in accordance with their respective experimental group. B, volcano plot illustrating log2 fold changes and adjusted p-values of each gene in the differential gene expression analysis of Rv1339-expressing strain compared with the empty control vector strain. Horizontal dotted line shows adjusted p-value cut-off at 0.05 and vertical dotted lines show log2 fold change cut-off at a magnitude of 1.5. Nonsignificantly differentially expressed genes with log2 fold change under 1.5 magnitude are indicated by a black circle; nonsignificantly differentially expressed genes with log2 fold change over 1.5 magnitude are indicated by a black cross; significantly differentially expressed genes with log2 fold change under 1.5 magnitude are indicated by a black diamond; and significantly differentially expressed genes with log2 fold change over 1.5 magnitude are indicated by red triangles and labeled with their old locus tags (where this tag was unavailable, respective new locus tags were used). C, membrane potential, expressed as fluorescence ratio (Red/Green) of DiOC2, of intact cells and cells treated with the depolarizing agent CCCP, comparing the pVV16 (empty vector) control (black), Rv1339-expressing (red) and Rv1339 D180-expressing (gray) M. smegmatis mc²155 strains. D, spot dilution assay of the strains grown for 3 days in 7H10 and 7H10 containing 5 μg/ml of malachite green or 0.01% SDS. E, volcano plot for the untargeted metabolomic analysis displaying the differential abundance of metabolites in the empty control vector and Rv1339-expressing strains, cut-off log2 onefold change p < 0.01 (green lines). F, abundances in the different strains of selected metabolites involved in lysine and/or peptidoglycan biosynthesis. The experiments were performed in biological duplicates and technical quadruplicates. Unpaired two-tailed Student’s t tests were used to compare the data. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 and ∗∗∗∗p < 0.0001. The data are presented as the means ± SDs from two biological replicates and three replicates. CCCP, carbonyl cyanide m-chlorophenyl hydrazine; DiOC₂, 3,3′ diethyloxicarbocianide chloride.

Figure 4

Expression of Rv1339 reduces tolerance to antibiotics targeting cell wall biosynthesis. Time-kill assay experiments of the pVV16 (empty vector) control, Rv1339-expressing and Rv1339 D180A-expressing M. smegmatis mc²155 strains in the presence of three cell wall-targeting antibiotics: D-cycloserine (A), vancomycin (B), and ethambutol (C) compared to control antibiotics with other targets: levofloxacin (D), ciprofloxacin (E), and rifampin (F). Circles and solid lines indicate vehicle alone and triangles and dashed lines 1 × MIC₅₀. The data are presented as the means ± SDs from two biological replicates and three technical replicates.