Host and bacterial proteases influence biofilm formation and virulence in a murine model of enterococcal catheter-associated urinary tract infection

Abstract

Enterococcus faecalis is a leading causative agent of catheter-associated urinary tract infection (CAUTI), the most common hospital-acquired infection. Its ability to grow and form catheter biofilm is dependent upon host fibrinogen (Fg). Examined here are how bacterial and host proteases interact with Fg and contribute to virulence. Analysis of mutants affecting the two major secreted proteases of E. faecalis OG1RF (GelE, SprE) revealed that while the loss of either had no effect on virulence in a murine CAUTI model or for formation of Fg-dependent biofilm in urine, the loss of both resulted in CAUTI attenuation and defective biofilm formation. GelE^-, but not SprE^- mutants, lost the ability to degrade Fg in medium, while paradoxically, both could degrade Fg in urine. The finding that SprE was activated independently of GelE in urine by a host trypsin-like protease resolved this paradox. Treatment of catheter-implanted mice with inhibitors of both host-derived and bacterial-derived proteases dramatically reduced catheter-induced inflammation, significantly inhibited dissemination from bladder to kidney and revealed an essential role for a host cysteine protease in promoting pathogenesis. These data show that both bacterial and host proteases contribute to CAUTI, that host proteases promote dissemination and suggest new strategies for therapeutic intervention.

Fig. 1

GelE and SprE are functionally redundant in CAUTI. Catheter-implanted mice were infected with ~2 × 10⁷ CFU of the indicated strains (see Table S1). Following 24 hrs of infection, total numbers of CFU recovered were determined for a Catheter, b Bladder and c Kidney. Each symbol represents an individual mouse and symbols touching the dashed lines indicate values below the limit of detection (LOD, 40 CFU). Data shown are pooled from two independent experiments. For each strain, the mean value is shown as the horizontal line. *p < 0.05, **p < 0.005, ***p < 0.001

Fig. 2

∆GelE has protease activity in urine, but not BHI. Expression of protease activity against a FITC-casein substrate a and relative transcription of the genes encoding the E. faecalis secreted proteases from the WT strain, as determined by real time RT-PCR b were assessed following 24 h of culture in the indicated medium. Protease activity of various strains (see Table S1) on protease indicator plates c and in supernatants from strains grown in BHI d or urine e following overnight culture. Values are normalized vs. culture density (OD₆₀₀) and are compared to BHI a, b or compared to uninoculated media (Medium; d, e). On indicator plates, protease activity is apparent as a zone of clearing around the bacterial growth. Data are derived from a minimum of three independent experiments are shown as the mean ± SEM. For the indicated pair-wise comparisons, ***p < 0.001

Fig. 3

GelE-independent activation of SprE in urine. a GelE⁻ and GelE⁺ strains expressing SprE were constructed by complementing the indicated mutants with a plasmid expressing HA-tagged SprE (pSprE) or GelE and HA-tagged SprE (pGS). Supernatants (super) and cell lysates (cell) were then prepared following overnight culture in the indicated medium and subjected to a Western blot analysis to detect HA-tagged SprE. Arrows at the right indicate pro-SprE (closed) and mature SprE (open). Migration of several molecular weight standards are shown on the left. All blots are derived from the same experiment and processed in parallel. b Morphology of WT and ΔGS strains following overnight growth in the indicated medium was determined by phase contrast microscopy following fixation

Fig. 4

In urine GelE and SprE are functionally redundant for processing Fg and can enhance Fg-dependent biofilm formation. The ability of supernatants derived from the indicated strains (see Table S1) to process Fg is shown following overnight culture in BHI a or urine b. Cell-free supernatants were prepared, Fg was added and after an additional 6 or 48 h of incubation, the resulting cleavage patterns were resolved by SDS-PAGE and staining with Coomasie Brilliant blue for comparison to Fg incubated in uninoculated BHI (BHI) or urine (Urine). Migration of the Aα, Bβ, and γ chains of Fg and of several molecular weight standards are shown the right and left, respectively. All blots are derived from the same experiment and processed in parallel. The ability of alexa-fluor Fg to associate with the cell surface of the indicated strains following growth in optimized M9 medium c or urine d was analyzed by FACS. Enterococcal cells were counter stained with SYTO59. A PBS-washed WT and a SYTO59-only WT were used as control. Numbers in each panel indicates the percentage of the total number of cells contained within the indicated quadrant. The ability of the supplements indicated at the right to enhance growth in optimized M9 medium e or urine f was assessed by growth following overnight incubation by determination of CFUs. The ability of secreted proteases to enhance biofilm formation in the absence (unsupp.) or presence of Fg (+Fg) in TSBG g or urine h is shown. Biofilm formation by the indicated strains was analyzed following 48 h of incubation in 96-well plates by absorbance (A595) following staining with Crystal Violet. Biofilm was normalized with growth density measured at OD600. Data shown are from a minimum of three independent experiments and are shown as the mean ± SEM. For the indicated pair-wise comparisons, *p < 0.05, **p < 0.005, ***p < 0.001, ns not significant

Fig. 5

Protease inhibitor treatment reduced bacterial burden in CAUTI. a The ability of a protease inhibitor cocktail (cOmplete™ Mini) to inhibit E. faecalis protease activity was assessed following overnight growth of the indicated strains in urine. Protease activity in supernatants against a FITC-casein substrate was then determined. b The growth of the WT strain in urine unsupplemented (unsupp.) or supplemented as indicated, in the presence or absence of the inhibitor cocktail was assessed by determination of CFUs following overnight growth. c Experimental timeline for treatment of murine CAUTI. The inhibitor cocktail was administered by intraperitoneal injection (dose) in four groups of mice at 0, 20, 100, and 500 mg/kg. Infection: time of catheter implantation followed immediately by infection by the WT strain. Sac: time of mouse sacrifice. d Urines collected at the indicated time points were tested for inhibition of endogenous protease activity by the addition of Fg followed by SDS–PAGE as described above (Fig. 4). The migration of several molecular weight standards is shown (M) and that of the Fg chains shown on the left. All blots are derived from the same experiment and processed in parallel. Following sacrifice, bacterial burdens were determined in catheter e, bladder f and kidney g. Total CFU recovered is shown. Each symbol represents an individual mouse. Symbols touching the dashed line indicate that recovery was below the limit of detection (LOD, 40 CFU). Infection data are pooled from two independent experiments. For each dose, the mean value is shown as the horizontal line. Other data are shown as the mean ± SEM derived from at least three independent experiments. *p < 0.05, **p < 0.005, ***p < 0.001

Fig. 6

A cysteine protease inhibitor reduces dissemination. Groups of mice were treated with the protease inhibitors indicated or mock-treated (PBS), infected and analyzed as described in Fig. 5 for determination of bacterial burdens in catheter a, bladder b and kidney c. Inhibitors were used at the following concentrations: Pefabloc, 50 mg/kg/day; EDTA, 30 mg/kg/day; E64, 2 mg/kg/day; Mixed: Pefabloc + EDTA + E64 (50 mg/kg/day, 30 mg/kg/day, 2 mg/kg/day respectively). Histology of bladders in H&E-stained section is shown for mice d uninfected (naïve), e catheter-implanted and infected (mock) and f E64-treated, catheter-implanted and infected by the WT strain. Labels are: L lumen; lp, lamina propria, ue urothelium, m muscularis. Scale bars: 200 µm. *p < 0.05, **p < 0.005