Reversion of a RND transporter pseudogene uncovers latent stress resistance in Brucella ovis

Abstract

Small-molecule screens can advance therapeutic discovery while yielding new insights into pathogen biology. Through a luminescence-based screen, we identified clinically approved dihydropyridines that reduced the fitness of the intracellular pathogen Brucella ovis within mammalian phagocytes. Given the established role of dihydropyridines as inhibitors of mammalian L-type calcium channels and our observation that drug treatment perturbed calcium and manganese levels in host phagocytes, we initially hypothesized a host-directed mechanism of action. However, dose-response assays in axenic medium revealed that these drugs can directly inhibit B. ovis growth. To investigate the genetic basis of B. ovis susceptibility to dihydropyridine treatment, we selected for mutants capable of growing in the presence of cilnidipine. Cilnidipine-resistant isolates carried single-base deletions in the bepE pseudogene that restored an open reading frame encoding an RND-family transporter subunit. B. ovis is an ovine venereal pathogen that has experienced significant pseudogenization in its recent evolutionary history. Frameshift mutations that restored bepE function in B. ovis increased its resistance not only to dihydropyridines but also to a broad range of membrane-disrupting agents, including bile acid. Conversely, deleting bepE in Brucella abortus, a related zoonotic species that retains an intact version of the gene, increased its sensitivity to bile acid in vitro and to cilnidipine in the intracellular niche. We conclude that bepE is a key determinant of chemical stress resistance in Brucella spp., and that its pseudogenization in B. ovis contributes to the documented hypersensitivity of this host-restricted lineage to chemical stressors.

Figure 1.. The dihydropyridine Ca²⁺ blockers nicardipine and cilnidipine inhibit intracellular B. ovis growth and cilnidipine inhibits B.ovis axenic growth.

(A) Intracellular inhibitory activities of nicardipine and cilnidipine during B. ovis THP-1 macrophage infection. B. ovis luminescence was measured 48 h post-treatment and normalized to signal from infected untreated cells. IC50 values were calculated using GraphPad Prism software. (B) Axenic inhibitory activity of nicardipine and cilnidipine during B. ovis growth in liquid medium. Optical densities at 600 nm were measured after 48 h of growth and normalized to untreated cultures. (C) Chemical structures of nicardipine and cilnidipine. In A and B, data represent the mean and standard deviation of 3 biological replicates.

Figure 2.. Nicardipine treatment alters the THP-1 macrophage metallome.

Elemental content of untreated THP-1 cells or cells treated with 25 μM nicardipine for 48 hours was determined using triple quadrupole inductively coupled plasma mass spectrometry (ICP-QQQ). Each element was normalized by the total phosphorus in each sample (M/P). (A) Calcium and (B) Manganese levels were significantly elevated upon nicardipine treatment. Other elements are presented in Fig. S2. Bars represent the mean ± standard deviation of 7 biological replicates measured over 2 independent experiments. The full metallome data set was analyzed using multiple unpaired t-tests and Bonferroni-Dunn method to adjust for multiple comparisons (***, adjusted P < 0.001).

Figure 3.. B. ovis cilnidipine-resistant mutants harbor single-nucleotide deletions that resurrect the open reading frame of the pseudogene bepE.

(A) Diagram of the bepDE locus (top) and alignment of bepE nucleotide sequences from wild-type B. ovis, four cilnidipine-resistant mutants, and B. abortus (bottom). The 2 bp deletion that led to the pseudogenization of B. ovis bepE, and the single base deletions in the cilnidipine-resistant mutants are highlighted. (B) Schematic of the bepE open reading frame B. abortus 2308, B. ovis ATCC 25840, and a B. ovis cilnidipine-resistant mutant. The site of the 2 bp frameshift in wild-type B. ovis compared to the other Brucella species, and site of the restoring frameshift mutations are indicated. (C) Alignment of the partial BepE amino acid from wild-type B. ovis ATCC 25840, four cilnidipine-resistant mutants and B. abortus. Differences from the consensus of the genus, exemplified by B. abortus, are highlighted in red. (D) Alignment of the entire BepE protein sequence from representative Brucella species lineages and one B. ovis cilnidipine-resistant mutant. Differences from the consensus are indicated in black.

Figure 4.. Restoration of B. ovis BepE function confers detergent resistance and enhanced tolerance of dihydropyridines during macrophage infection.

(A) Growth of the wild-type (WT) B. ovis and cilnidipine-resistant mutants on TSA blood plates with or without detergents. Cultures of each genotype were normalized by optical densitiy, 10-fold serially diluted, and spotted onto TSA blood plates that were untreated or contained 0.005% SDS, 0.009% deoxycholate, or 0.005% CHAPS. (B) Bacteria recovered from THP-1 macrophage-like cells infected with WT B.ovis or a B. ovis bepE restored mutant. (C, D) Bacteria recovered from infected THP-1 cells treated with 25 μM nicardipine or 25 μM cilnidipine at (C) 24- or (D) 48-h post infection. Infections were performed three times; values are means ± SD from three independent trials. Statistical significance was calculated at 24 hpi and 48 hpi with an unpaired t-test (*, P < 0.05; **, P < 0.01).

Figure 5.. Loss of BepE function renders B. abortus sensitive to deoxycholate.

(A) Colony-forming units of B. abortus WT and ΔbepE (BAB_RS17475) enumerated after growth on TSA blood medium alone (Untreated) or containing 25 μM cilnidipine, 25 μM nicardipine, 1% deoxycholate (w/v), or 1% CHAPS (w/v). Dotted line indicates limit of detection. (B) Deoxycholate dose-response curves with IC50 values. Growth of B. abortus WT, ΔbepE, and the complementation strain ΔbepE glmS::bepDE in liquid medium with deoxycholate was assessed by optical density after 48 h. Values represent mean ± SD of three independent trials, each normalized to the untreated control. (C) Growth of serially-diluted B. abortus strains spotted onto TSA blood plates with or without 0.1% deoxycholate (w/v).

Figure 6.. B. abortus bepE does not contribute to survival in THP-1 macrophages but confers enhanced resistance to cilnidipine in the intracellular niche.

(A) B. abortus CFU recovered from THP-1 macrophages after infection with WT, ΔbepE, and the complementation strain ΔbepE glmS::bepDE. (B) WT B. abortus cells recovered from infected THP-1 macrophages untreated or treated with 25 μM nicardipine or 25 μM cilnidipine. (C, D) CFUs of WT, ΔbepE, and ΔbepE glmS::bepDE recovered from THP-1 macrophages at 24 (C) or 48 hpi (D). Infected macrophages were treated with 25 μM nicardipine or 25 μM cilnidipine. Infections were performed three times; values are means ± SD from three independent trials. Statistical significance was assessed at 48 hpi using two-way ANOVA on log10 transformed CFU values, followed by Tukey’s multiple comparison’s test compared with the untreated WT group (*, P < 0.01; **, P < 0.001).