Fig 1.
Absence of fsrA and gelE enhances intracellular survival of E. faecalis in RAW264.7 macrophages at 20 hpi.
(A-B) Intracellular CFU in RAW264.7 macrophages infected with OG1RF WT and derived mutants of genes implicated in intracellular persistence (blue), virulence (orange), or GelE proteolytic targets (pink). (n = 18 for WT and n = 5-6 for mutants) (C) Schematic diagram of the genomic locus comprising fsrABDC/gelE/sprE, represented in different shaded arrows for each transcriptional unit. Promoters are represented by bent arrows (grey: constitutive promoter of fsrA, black: fsrA-regulated promoter). The respective gene size in base pairs (bp) is indicated below each gene. (D-E) Intracellular CFU in RAW264.7 macrophages infected with genetic deletion mutants of fsrABDC and gelE (n = 9), as well as a transposon insertion mutant of sprE (n = 5). (F-J) Analysis of gelatinase activity phenotype. (F, I) Gelatinase activity of E. faecalis strains on Todd-Hewitt agar + 3% gelatin at 24 h. Halo formation (white dashed line) indicates gelatinase activity. Representative images of n = 3 are shown. (G, J) Detection of GelE secretion in E. faecalis culture supernatants at 4, 12 and 24 h by Western blotting. CTD = C-terminal domain. Representative images of n = 2 are shown. (H) Schematic diagram of GelE domains: a putative signal peptide (S.P., 1-30 a.a.), a propeptide (30-192 a.a.), the mature GelE protease (192-495 a.a.) and a C-terminal domain (CTD, 495-509 a.a.). (K-L) Intracellular CFU in RAW264.7 macrophages infected with WT, ΔgelE and ΔgelE-complemented OG1RF strains. (n = 4); All bar graphs shown represent mean ± SD of biological replicates. Statistical significance of each strain against WT was assessed using one-way ANOVA with Dunnett’s multiple comparisons test. Only comparisons with p < 0.05 are annotated. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
Fig 2.
Increased intramacrophage survival in the absence of fsrA and gelE is conserved across wound isolates of E. faecalis.
(A) Characterization of 49 E. faecalis wound isolates for fsrABDC/gelE genotypes and gelatinase activity phenotypes. (B-D) Intracellular CFU in RAW264.7 macrophages infected with four selected gentamicin-susceptible fsrABDC+gelE+ (grey bars), fsrABDC-gelE+ (green bars) and fsrABDC-gelE- (orange bars) E. faecalis wound isolates and laboratory strains (JH2-2, OG1RF WT, ΔfsrABDC and ΔgelE) at (B) 2 hpi, (C) 6 hpi and (D) 20 hpi. Bars represent mean ± SD from n = 4. Statistical significance of each strain against WT was assessed using one-way ANOVA with Dunnett’s multiple comparisons test. Only comparisons with p < 0.05 are annotated. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
Fig 3.
Lack of GelE promotes the accumulation of replicating intracellular bacterial reservoirs.
(A) Quantification of intracellular replicating (eFluor-) and non-replicating (eFluor+) E. faecalis strains from macrophage lysates at 2, 6, and 20 hpi using flow cytometry. WT OG1RF fixed with 4% PFA prior to infection (Fixed WT) was included as non-proliferating controls. Representative histograms from n = 2-4 are shown. (B) Proportion (%) of intracellular replicating E. faecalis (eFluor-) population from WT and mutant E. faecalis-infected macrophage lysates. Bars represent mean ± SD of n = 4, except fixed WT (n = 2). Statistical significance of each strain against WT was assessed using one-way ANOVA with Dunnett’s multiple comparisons test. Only comparisons with p < 0.05 are annotated. ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001. (C) Representative confocal microscopy images of RAW264.7 macrophages infected by E. faecalis strains pre-stained with eFluor 670 (magenta) at 20 hpi from n = 3. Samples were fixed and post-stained for Enterococcus-specific Group D antigen (green), double-stranded DNA (dsDNA; blue) and F-actin (white).
Fig 4.
Intracellularly replicating WT OG1RF induce Fsr quorum sensing at 6 hpi.
(A) Gating of eFluor+GFPhi (GFPhi) and eFluor+GFPlo (GFPlo) infected macrophages used for cell sorting at 2, 6 and 20 hpi. Representative dotplots from n = 4 are shown. (B) Relative quantification (%) of GFPhi and GFPlo macrophages as gated in (A). Bars represent mean ± SD of n = 4. Statistical analysis was assessed using one-way ANOVA with Tukey’s multiple comparisons test. (C) Confocal fluorescence images of sorted GFPhi and GFPlo macrophages, with intracellular GFP+eFluor670lo bacteria indicating proliferating E. faecalis (n = 1). Scale bar = 10 μm. (D-E) Relative gene expression of fsrABDC and its associated regulon in intracellular replicating (GFPhi) and non-replicating (GFPlo) E. faecalis populations at (D) 6 hpi and (E) 20 hpi, normalised by the -ΔΔCt method to the housekeeping gene recA and to the gene expression of the baseline 2 hpi GFPlo population. Bars represent mean ± SEM of n = 4. Statistical significance between GFPhi and GFPlo populations at each timepoint was assessed using unpaired T-test. For all graphs, only comparisons with p < 0.05 are annotated. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
Fig 5.
Loss of GelE proteolytic activity delays egress from macrophages and promotes intracellular accumulation of E. faecalis in macrophages.
(A) LDH cytotoxicity quantification of culture supernatants from E. faecalis-infected macrophages at 2, 6 and 20 hpi using the antibiotic protection assay. Bars represent mean ± SD from n = 5. Statistical significance against WT was assessed using one-way ANOVA with Dunnett’s multiple comparisons test. A subset of this data comparing macrophage cytotoxicity from WT infection at 2, 6 and 20 hpi is shown in S1B Fig. (B-C) Quantification of extracellular egressed bacteria from culture supernatants of RAW264.7 macrophages infected with (B) fsrABDC/gelE deletion mutants and (C) gelE-complemented OG1RF strains at 6 to 12 hpi. Each data point represents mean ± SD of n = 3-4. At each timepoint, statistical significance against WT was assessed using two-way ANOVA with Dunnett’s multiple comparisons test. LOD = limit of detection. (D) Representative Z-projections of WT- and ΔgelE-infected RAW264.7 macrophages at 6, 9 and 12 hpi from (B), captured with confocal microscopy and stained for Enterococcus specific Group D antigen (yellow), dsDNA (cyan), and F-actin (magenta) (n = 2). Orthogonal Z-axis projections along the blue and red lines are shown in the adjacent colored boxes. White and yellow arrows indicate matched top view and Z-axis projections of selected macrophages with dense intracellular E. faecalis. For all graphs, only comparisons with p < 0.05 are annotated. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
Fig 6.
Absence of GelE promotes intracellular survival of E. faecalis in murine wounds.
(A-B) Total CFU from wound homogenates of (A) mono-infection or (B) competitive infection at 1, 3, and 5 dpi. Median of 3-5 animals per group from one independent experiment is shown. Dotted lines show inoculum CFU for mono-infection. (C) Flow cytometry analysis of intracellular E. faecalis (stained for Group D antigen) in various immune cells at 5 dpi. Median from 6-9 animals per group from 2 independent experiments is shown. (D-E) Quantification of (D) intracellular CFU at 1, 3 and 5 dpi or (E) total CFU at 5 dpi from enzymatically dissociated wounds. Median from 4-5 animals per group from one independent experiment is shown. For A-D, only comparisons with p < 0.05 are annotated. At each timepoint, statistical significance between infection groups was assessed using Mann-Whitney test, except for C, which was assessed using Kruskal-Wallis test with Dunn's multiple comparisons test. * = p < 0.05, ** = p < 0.01, *** = p < 0.001. n.s. = not significant.
Fig 7.
Increased expression of QS-regulated GelE in replicating populations drives egress from macrophages and wound cells.
Replicating intracellular E. faecalis exhibit Fsr QS-mediated induction of gelE expression compared to non-replicating populations, promoting egress from macrophages and from wound infection-associated cells via its proteolytic action on an unknown host substrate. By contrast, absence of GelE resulted in egress-defective, dense intracellular populations. Created in BioRender. Tanoto, F. (2026) https://BioRender.com/eo3dbrt.