The polyadenylase PAPI is required for virulence plasmid maintenance in pathogenic bacteria

Abstract

Many species of pathogenic bacteria harbor critical plasmid-encoded virulence factors, and yet the regulation of plasmid replication is often poorly understood despite playing a critical role in plasmid-encoded gene expression. Human pathogenic Yersinia, including the plague agent Y. pestis and its close relative Y. pseudotuberculosis, require the type III secretion system (T3SS) virulence factor to subvert host defense mechanisms and colonize host tissues. The Yersinia T3SS is encoded on the IncFII plasmid for Y ersinia virulence (pYV). Several layers of gene regulation enables a large increase in expression of Yersinia T3SS genes at mammalian body temperature. Surprisingly, T3SS expression is also controlled at the level of gene dosage. The number of pYV molecules relative to the number of chromosomes per cell, referred to as plasmid copy number, increases with temperature. The ability to increase and maintain elevated pYV plasmid copy number, and therefore T3SS gene dosage, at 37°C is important for Yersinia virulence. In addition, pYV is highly stable in Yersinia at all temperatures, despite being dispensable for growth outside the host. Yet how Yersinia reinforces elevated plasmid replication and plasmid stability remains unclear. In this study, we show that the chromosomal gene pcnB encoding the polyadenylase PAP I is required for regulation of pYV plasmid copy number (PCN), maintenance of pYV in the bacterial population outside the host, robust T3SS activity, and Yersinia virulence in a mouse infection model. Likewise, pcnB/PAP I is also required for robust expression of the Shigella flexneri virulence plasmid-encoded T3SS. Furthermore, Yersinia and Shigella pcnB/PAP I is required for maintaining normal PCN of model antimicrobial resistance (AMR) plasmids whose replication is regulated by sRNA, thereby increasing antibiotic resistance by ten-fold. These data suggest that pcnB/PAP I contributes to the spread and stabilization of virulence and AMR plasmids in bacterial pathogens, and is essential in maintaining the gene dosage required to mediate plasmid-encoded traits. Importantly pcnB/PAP I has been bioinformatically identified in many species of bacteria despite being studied in only a few species to date. Our work highlights the potential importance of pcnB/PAP I in antibiotic resistance, and shows for the first time that pcnB/PAP I reinforces PCN and virulence plasmid stability in natural pathogenic hosts with a direct impact on bacterial virulence.

Figure 1.. Y. pseudotuberculosis pYV plasmid copy number (PCN) dynamically changes in response to temperature.

(A) pYV plasmid copy number per chromosome was determined for wildtype Y. pseudotuberculosis IP2666pIB1 grown at 26°C or 37°C/low calcium using droplet digital PCR (dd)PCR. (B) Relative pYV plasmid copy number of wildtype Y. pseudotuberculosis YPIII/pIBX was determined following growth at 26°C or 37°C/low calcium using a luciferase reporter stably incorporated in the pIBX plasmid, normalized to bacterial cell density (OD₆₀₀). (A-B) Averages of three independent experiments are shown ± standard deviation (Student t-test, *** p<0.001, ** p<0.01, * p<0.05). (C) Representative phase contrast (left) and fluorescence (right) images of Y. pseudotuberculosis YPIII/pIBX cells where the pYV plasmid is visualized using a ParB-msfGFP fluorescent marker. Images are shown for cells grown at 26°C or under the 37°C/low calcium condition. (D) Histogram depicting the number of ParB-msfGFP-labeled puncta of pYV quantified from data illustrated in panel C. Three biological replicates per condition were imaged and analyzed; for each strain and condition, between 790 to 4428 cells were analyzed per replicate.

Figure 2.. Disruption of the pcnB gene encoding poly(A) polymerase PAP I decreases pYV PCN as well as T3SS expression and activity.

(A) PAP I from Y. pseudotuberculosis IP2666pIB1, E. coli K12, and S. flexneri M90T are highly conserved. Known PAP I domains include the head domain (blue), the neck domain (green), the body domain (purple), and the leg domain (yellow) [62]. Residues in boxes represent known catalytic residues from EcPAP and the residue indicated with an arrow represents the residue that was mutated in the L291R pcnB suppressor mutants. (B) pYV plasmid copy number per chromosome was determined for Y. pseudotuberculosis IP2666pIB1 strains grown at 26°C or 37°C/low calcium using ddPCR. Averages of three independent experiments are shown ± standard error of the mean. Statistical significance was calculated using a one-way ANOVA with Tukey’s multiple comparisons test (**** p<0.0001, *** p<0.001, ** p<0.01, * p<0.05). (C) Y. pseudotuberculosis IP2666pIB1 strains were grown in low calcium media at 37°C before preparing samples for western blotting. Bacterial pellets were collected and secreted proteins (Yops) were precipitated from the supernatant (sup). Images shown represent a single blot that was independently probed for YopE and RpoA, which serves as a loading control.

Figure 3.. The pcnB gene is required for maintenance of the pYV plasmid.

(A) Representative phase contrast (left) and fluorescence (right) images of Y. pseudotuberculosis YPIII/pIBX ΔpcnB cells where the pYV plasmid is visualized using a ParB-msfGFP fluorescent marker. Images are shown for cells grown at 26°C or under the 37°C/low calcium condition. White arrowheads indicate plasmidless cells. (B) Histograms comparing wildtype vs. ΔpcnB cells under 26°C or 37°C/low calcium conditions. Mean fluorescence intensity values are plotted on the x-axis (log₁₀). Shaded error bars represent the mean ± standard error of the mean from bootstrapping the data 100 times. Both histograms were randomly sampled without replacement to match the lowest n value of either condition (7,003 and 10,033 for the 26°C and 37°C/low calcium conditions, respectively). Both distributions are significantly different (p < 10⁻¹⁰⁰) by Kolmogorov-Smirnov tests. A.U. indicates arbitrary units. (C) Histograms depicting the number of ParB-msfGFP-labeled puncta of pYV quantified from data illustrated in panel A. Only cells with at least one detected pYV spot were considered. Three biological replicates were imaged and analyzed per condition. For each strain and condition, between 2063 to 14,479 cells were analyzed per replicate. (D) Same as (C) except that the comparison is between wildtype and ΔpcnB cells grown at 26°C. For each strain and condition, between 760 to 4179 cells were analyzed per replicate. (E) Same as (D) except for the 37°C/low calcium condition. For each strain and condition, between 2046 to 14,479 cells were analyzed per replicate. Distributions in panels D and E are significantly different (p < 10⁻⁸⁰) by Kolmogorov-Smirnov tests.

Figure 4.. PAP I point mutations lead to a decrease in pYV PCN, plasmid stability, and T3SS activity.

(A) Y. pseudotuberculosis IP2666pIB1 strains harboring various PAP I^FLAG alleles grown at 26°C or 37°C/low calcium were pelleted and prepared for western blot analyses. Images shown represent a single blot that was cut after transferring and independently probed for expression of PAP I^FLAG (top) and RpoA (loading control, bottom). Data shown is representative of three biological replicates. (B) Y. pseudotuberculosis IP2666pIB1 strains were grown at 26°C before diluting and spotting onto either LB or low calcium plates containing Congo red dye and incubated at 26°C or 37°C, respectively, prior to imaging. (C) Relative pYV plasmid copy number was estimated for strains in the Y. pseudotuberculosis pIBX/YPIII background. For each timepoint, luminescence was measured and normalized to cell density (OD₆₀₀). Averages of three independent experiments are shown ± standard error of the mean. Each strain per condition was compared to the wildtype (WT) strain in a one-way ANOVA with Dunnett’s multiple comparisons test (**** P<0.0001, *** P<0.001, ** P<0.01, * P<0.05). (D) Y. pseudotuberculosis IP2666pIB1 strains were grown at 37°C/low calcium and secreted proteins in the supernatant fractions were precipitated and visualized after separation on as SDS-PAGE gel by Coomassie blue staining. Bovine serum albumin (BSA) was used as a loading and protein precipitation control. The YopE T3SS effector protein band from four independent experiments was visualized, quantified, and normalized to BSA from the same samples. Bars represent the mean fold change in YopE/BSA relative to wildtype and error bars represent the standard error of the mean between different experiments. Statistical significance was determined using a one-way ANOVA with Dunnett’s multiple comparisons test (**** P<0.0001, *** P<0.001, ** P<0.01, * P<0.05).

Figure 5.. The pcnB gene is required for plasmid-mediated antibiotic resistance in Y. pseudotuberculosis.

(A) Wildtype and ΔpcnB Y. pseudotuberculosis IP2666pIB1 harboring pGFP-uv were spotted simultaneously onto either plain LB plates or LB plates containing carbenicillin to assess retention of pGFP-uv. Plates were incubated at 26°C for ~16 hours before imaging spots under a GFP filter. (B) Retention of the pGFP-uv plasmid by wildtype and ΔpcnB Y. pseudotuberculosis IP2666pIB1 was assessed at 26°C over 40 hours following the removal of carbenicillin. (C) Wildtype and ΔpcnB Y. pseudotuberculosis IP2666pIB1 harboring either no vector, a pTrc99::empty vector control, or a pTrc99::PAP I recombinant plasmid were spotted onto either plain LB plates or LB plates containing carbenicillin to select for pTrc99 and 0.2 mM IPTG to induce PAP I expression. Plates were incubated at 26°C for ~16 hours prior to imaging. (D) Wildtype and ΔpcnB Y. pseudotuberculosis IP2666pIB1 harboring either no vector or a pNF06 mini-F plasmid with a ccdAB TA system (pNF06::ccdAB) were spotted onto either plain LB plates or LB plates containing 25 μg/mL kanamycin to select for pNF06. Plates were incubated at 26°C for ~16 hours prior to imaging.

Figure 6.. The pcnB gene is required for normal Y. pseudotuberculosis virulence in a mouse IP infection model.

Six- to eight-week-old C57Bl/6 mice were infected with 1–3×10³ Y. pseudotuberculosis IP2666pIB1 via intraperitoneal (IP) injection. Tissues were collected, homogenized, and plated to determine CFU per gram (CFU/g) tissue 4 days post-inoculation. The log₁₀ CFU/g tissue from the spleens (A) and livers (B) of infected mice are shown. Each circle represents one mouse organ; lines represent geometric means.

Figure 7.. The pcnB gene is required for plasmid-encoded expression of T3SS and antibiotic resistance genes in Shigella flexneri.

(A) S. flexneri M90T ΔipaH2.5 and ΔipaH2.5ΔpcnB, as well as a virulence plasmid cured (VP⁻) BS103 strain, were incubated on Congo red media at 37°C. (B) S. flexneri M90T ΔipaH2.5 and ΔipaH2.5ΔpcnB carrying pTRC99 with and without IPTG-inducible Yersinia PAP I were incubated at 37°C, and secreted T3SS effector proteins were visualized after separation on as SDS-PAGE gel using anti-IpaB, anti-IpaC, and anti-IpaD antibodies. (C) S. flexneri M90T ΔipaH2.5 and ΔipaH2.5ΔpcnB carrying pTRC99 with and without IPTG-inducible Yersinia PAP I were grown on either plain TCS media or TCS containing ampicillin and IPTG. Shown is one of two independent experiments.