Role of sortase-assembled Ebp pili in Enterococcus faecalis adhesion to iron oxides and its impact on extracellular electron transfer

ABSTRACT

Enterococcus faecalis sortase-assembled endocarditis and biofilm-associated pili (Ebp) are virulence factors implicated in enterococcal biofilm-associated infections and gastrointestinal colonization. We previously showed that E. faecalis biofilm metabolism is influenced by extracellular electron transfer (EET) under iron-rich conditions, raising the question of whether Ebp pili also play a role in EET. Here, we report a novel role of Ebp pili in E. faecalis adhesion to the iron oxides magnetite, goethite, and hematite, where the EbpA tip adhesin contributes to this interaction. Adhesion by Ebp pili is conditionally important for EET to iron oxides, as pilus mutants are attenuated in EET under non-static growth conditions. In alignment with the established role of EET in redox homeostasis, we find that EET to ferricyanide supports E. faecalis anaerobic growth on glycerol. Furthermore, in an antibiotic-treated mouse gastrointestinal colonization model, we show that E. faecalis mutants deficient in EET poorly colonize the intestinal niche. Taken together, our findings suggest that Ebp pili can influence E. faecalis metabolic fitness by promoting EET to iron oxides, raising new questions about how Ebp pili shape E. faecalis interactions with environmental ecosystems. Additionally, the important role of EET in E. faecalis colonization of the dysbiotic gastrointestinal environment highlights the need for further inquiry into how EET contributes to E. faecalis microbial pathogenesis.

IMPORTANCE

In this study, we explored the interplay between extracellular electron transfer (EET) and an Enterococcus faecalis biofilm factor, the endocarditis and biofilm-associated pili (Ebp). We demonstrate that Ebp pili have a novel role in adhesion to iron oxides, which consequently promotes EET to iron oxides under non-static conditions. Along with our findings that E. faecalis EET can be coupled to anaerobic cell growth, our results point to a potential ecological role of Ebp pili in natural environments, outside of its established function in adhesion to host ligands. We provide the first evidence of the contribution of EET to E. faecalis colonization of the antibiotic-treated murine intestinal niche, which adds to the limited experimental evidence linking EET and microbial pathogenesis, as well as highlights the need for further studies of EET in bacterial pathogens.

OBSERVATION

Extracellular electron transfer (EET) is a process in which microorganisms reduce extracellular substrates, such as iron, and is frequently connected to anaerobic energy metabolism (1, 2). Although EET has been extensively reviewed in the gram-negative model species Geobacter sulfurreducens and Shewanella oneidensis (3–5), the discovery of a flavin-based EET genetic locus in gram-positive bacteria, including pathogens and members of the gut microbiota, raises the question of how microbial EET can influence human health or disease (6, 7). The gut commensal and opportunistic pathogen E. faecalis also performs EET (6, 8–11), but the physiological significance of EET in E. faecalis is not well-characterized. As our previous observations showed a relationship between E. faecalis biofilm metabolism and EET (8), this study focuses on investigating whether endocarditis and biofilm-associated pili (Ebp), an essential factor in E. faecalis biofilm formation, play a role in mediating EET. Given that EET has a prominent impact on the energy metabolism of another lactic acid bacterium Lactiplantibacillus plantarum (12, 13), we also sought to explore the importance of EET to E. faecalis using both in vitro and animal models.

Preliminary microscopy observations of E. faecalis grown in culture media autoclaved with ferric chloride showed some cells tightly surrounded by a dense material (Fig. S1A). This material might have been insoluble iron precipitates, inadvertently produced by autoclaving, as it was reminiscent of iron deposits previously observed in the extracellular matrix of E. faecalis biofilms grown under similar conditions (8). Immunofluorescence staining showed colocalization of the dense material with Ebp pili in the parent strain, whereas the dense material did not associate with the pilus-null ΔebpABC mutant (Fig. S1A). As Ebp pili are associated with adhesion to fibrinogen, collagen, and platelets (14, 15), as well as attachment to abiotic surfaces such as polystyrene (16), we hypothesized that there could be a similar interaction between Ebp pili and insoluble iron forms. Hence, we assessed the interaction between Ebp pili and iron by co-incubating E. faecalis with iron oxide magnetite, enabling the convenient separation of iron-adherent and iron-non-adherent bacteria using a magnet (17) (Fig. 1B).

Fig 1.

Ebp pilus expression is important for E. faecalis adhesion and EET to iron oxides. (A) Flow cytometry and immunofluorescence microscopy images of E. faecalis cells after magnetic separation using magnetite. Bacterial suspensions of E. faecalis OG1RF (left panels) were mixed with magnetite and then separated into magnetite-bound (middle panels) and non-bound fractions (right panels) using a magnetic rack. For flow cytometric analysis, the magnetite-bound bacterial fraction was released from magnetite by resuspension in 1× PBS containing 100 mM EDTA prior to staining. E. faecalis cells were stained with SYTO 9 DNA stain (green) and labeled for EbpC (red). Histograms show representative data from three independent experiments. Scale bars represent 20 µm. (B) Schematic diagram (left panel) illustrating how magnetic separation was used to isolate adherent and non-adherent fractions of bacteria. Culture turbidity (right panel) of non-adherent bacterial subpopulations after assaying with increasing concentrations of magnetite. Data points reflect optical density values at 600 nm (OD₆₀₀) of the non-adherent fraction after a magnet was used to remove magnetite and magnetite-bound bacteria. Horizontal lines represent the mean measurement of three independent experiments and statistical comparisons represent differences between WT and ΔebpABC at each magnetite concentration. *P < 0.05, **P < 0.01 by unpaired Student’s t-test. (C) Iron oxide reduction by Ebp pilus mutants as assessed by the ferrozine assay under non-static (left panel) or static (right panel) growth conditions. Horizontal lines represent the mean measurement of three independent experiments. Statistical comparisons represent differences between WT and the corresponding pilus mutant. **P < 0.01, ***P < 0.001 by one-way ANOVA with Tukey’s multiple comparisons.

Heterogeneous expression of Ebp pili in E. faecalis OG1RF produces subpopulations of piliated and non-piliated cells (14–16, 18). Through immunofluorescence microscopy and flow cytometry, we observed that the pilus-expressing cell population selectively adheres to magnetite (Fig. 1A; Fig. S1B). The ΔebpABC mutant showed significantly less adherence to magnetite compared with the parental wild-type strain (Fig. 1B; Fig. S2). Using the EbpA^AWAGA mutant, which has a mutated metal ion-dependent adhesion site (MIDAS) motif to disrupt EbpA function without perturbing pilus biogenesis (19), we found that the tip adhesin EbpA contributes to magnetite binding (Fig. S2C). Magnetite binding was also attenuated in the single deletion mutants of EbpA and major fiber pilin EbpC, whereas no significant differences were observed for the deletion mutant in cell wall anchor pilin EbpB (Fig. S2C). Findings from single deletion mutants overall correlate with previously published biofilm formation phenotypes due to defective pilus biogenesis (20). We were also interested in investigating whether the adhesion phenotype was similar for goethite and hematite, both of which are among the most common iron oxides found in soils (21). To test for adhesion to other iron oxides, we separated non-adherent bacteria from mineral-bound bacteria using differential density centrifugation with a sucrose solution and similarly observed adhesion to goethite and hematite in a pilus-dependent manner (Fig. S2D).

To determine if the adhesin function of Ebp pili influences EET to iron oxides, bacteria were grown in the presence of iron oxides and the amount of ferrous iron produced and released into the supernatant was quantified by the ferrozine assay. As the control, we used an EET-deficient mutant containing an insertional inactivation of the ndh3 gene. E. faecalis Ndh3 is a NADH dehydrogenase that had previously been characterized to be a component of the electron transport chain involved in EET, where it oxidizes NADH and transfers electrons to demethylmenaquinone (6, 9). Ndh3 is distinct from Ndh2, which is likely the NADH dehydrogenase specific for aerobic respiration (9). Significant differences in iron reduction were observed between parental and pilus mutant strains when bacterial growth was conducted under non-static conditions, whereas these differences were absent or less distinct under static growth conditions (Fig. 1C). Due to the heme auxotrophy of E. faecalis and the absence of heme in the culture media, observations made here were likely not influenced by the effects of aerobic respiration. Collectively, our findings indicate that Ebp pili play an accessory role in enhancing EET through the aggregation of E. faecalis cells with iron oxides. Although we did not differentiate between direct and indirect electron transfer, the results suggest that close contact between bacteria and iron oxides is important for efficient electron transfer. It is also tempting to speculate that the binding of Ebp pili to iron oxides is an evolutionary trait that augments bacterial EET in environments, such as soils and sediments, where iron oxides are commonly present. Ferrous iron produced by EET might additionally modulate pilus function by competition with or displacement of the native divalent cation in the MIDAS motif of EbpA (Fig. S3).

Previous work in Listeria monocytogenes demonstrated that the electron transport chain involved in EET can route electrons not only to extracellular iron but also through extracellular reductases such as fumarate reductase (22). As anaerobic glycerol dissimilation in E. faecalis is dependent on fumarate reductase (23), we predicted that EET could provide an alternative route to external electron acceptors in place of fumarate. To test if the membrane-impermeable ferricyanide could act as the electron acceptor for EET during glycerol dissimilation, we grew E. faecalis macrocolonies anaerobically on an agar base supplemented with ferricyanide. To limit excess fermentation of amino acids, the growth medium was excluded from the agar base, and instead, the bacterial inoculum was resuspended in the growth medium prior to spotting on agar plates. We find that the presence of either fumarate or ferricyanide promotes E. faecalis growth on glycerol, whereas growth was deficient in the absence of an electron acceptor or when the EET-deficient ndh3::tn mutant was used (Fig. 2A). Our findings confirmed that E. faecalis EET can be coupled to anaerobic cell growth, which is consistent with metabolic and growth phenotypes implicated in EET for L. monocytogenes and L. plantarum (6, 13). It would be particularly interesting to make further comparisons against L. plantarum, as it is a lactic acid bacterium similar to E. faecalis and has been shown to combine features of fermentation and respiration in a hybrid form of metabolism (13). As EET also promoted L. monocytogenes colonization of the mouse gastrointestinal tract (6), we investigated whether E. faecalis EET plays a similar role. As anticipated, lower numbers of ndh3::tn mutant were recovered in an antibiotic-treated mouse gastrointestinal model of colonization when compared with the parent strain (Fig. 2B).

Fig 2.

E. faecalis EET contributes to optimal fitness in vitro and in vivo. (A) Anaerobic macrocolony growth with glycerol as the carbon source in the presence of either 2.5 mM fumarate or 2.5 mM ferricyanide as the electron acceptor. After 24 h growth in an anaerobic jar, macrocolonies were excised from the agar and CFU enumerated. Data shown are from five independent experiments, and each dot represents the average of three technical replicates. Horizontal lines represent the median CFU. Statistical comparisons were assessed by one-way ANOVA with Tukey’s multiple comparisons. **P < 0.01, ***P < 0.001, ns non-significant. (B) E. faecalis colonization of lower gastrointestinal tract using an antibiotic-treated mice model of gut colonization. Data points represent individual mice, and the dotted line indicates the limit of detection at CFU < 40. Data shown are from four independent experiments with five mice per group in each experiment. Horizontal lines represent the median CFU. Statistical comparisons were assessed by the Mann-Whitney test. **P < 0.01, ***P < 0.001, ****P < 0.0001, ns non-significant. (C) Proposed model for the accessory role of Ebp pili in promoting adhesion and EET to iron oxides. Efficient EET to iron oxides under non-static conditions is mediated by Ebp pili, which has a novel role in adhesion to magnetite, goethite, and hematite. Ebp pili are not essential for EET, but adhesion promotes close contact with iron oxides for more efficient EET. The tip pilin EbpA, which is responsible for adhesion to host collagen and fibrinogen, contributes to the adhesion to iron oxides, but it is not known if the binding mechanism is similar for these substrates. EET contributes to the maintenance of redox homeostasis (NAD⁺/NADH balance), which is known to be important for various metabolic processes such as the anaerobic glycerol dissimilation pathway in E. faecalis. Although lactate fermentation is the primary mode of NAD⁺ regeneration in E. faecalis, experiments in another lactic acid bacterium L. plantarum have demonstrated that EET supplements fermentative pathways to increase metabolic flux and yield. Finally, the physiological consequence of EET to iron oxides would be highly dependent on the thermodynamic favorability of electron transfer in these environments, but a possibility is that EET serves as a strategy for E. faecalis to maintain competitive fitness by expanding the range of electron acceptors that it is able to utilize. Components are not drawn to scale. The figure was created using BioRender.

In summary, we describe a novel role of E. faecalis Ebp pili in adherence to iron oxides, which may promote EET and enhance metabolic fitness in environmental ecosystems where iron oxides are present (Fig. 2C). Although we show that EET can be coupled to anaerobic growth of E. faecalis, it is possible that EET may play additional roles, such as in iron uptake or detoxification (24, 25). Iron availability is a strong driver for colonization of the gastrointestinal tract by pathogens (26), but it is still unclear how Ebp pili will interact with different iron forms present in the gastrointestinal tract. Nevertheless, we demonstrate that mutants deficient in EET poorly colonize the mouse gastrointestinal tract, indicating that EET contributes to E. faecalis outgrowth during gut dysbiosis.

DATA AVAILABILITY

All data generated or analysed during this study are included in this published article and its supplemental material.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/spectrum.02337-24.

Supplemental Material. spectrum.02337-24-s0001.pdf.

Supplemental text, Figures S1 to S3, and Table 1.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.