Antagonistic interaction between two key endodontic pathogens Enterococcus faecalis and Fusobacterium nucleatum

ABSTRACT

Background

Endodontic infections are known to be caused by pathogenic bacteria. Numerous previous studies found that both Fusobacterium nucleatum and Enterococcus faecalis are associated with endodontic infections, with Fusobacterium nucleatum more abundant in primary infection while Enterococcus faecalis more abundant in secondary infection. Little is known about the potential interactions between different endodontic pathogens.

Objective

This study aims to investigate the potential interaction between F. nucleatum and E. faecalis via phenotypical and genetic approaches.

Methods

Physical and physiological interactions of F. nucleatum and E. faecalis under both planktonic and biofilm conditions were measured with co-aggregation and competition assays. The mechanisms behind these interactions were revealed with genetic screening and biochemical measurements.

Results

E. faecalis was found to physically bind to F. nucleatum under both in vitro planktonic and biofilm conditions, and this interaction requires F. nucleatum fap2, a galactose-inhibitable adhesin-encoding gene. Under our experimental conditions, E. faecalis exhibits a strong killing ability against F. nucleatum by generating an acidic micro-environment and producing hydrogen peroxide (H₂O₂). Finally, the binding and killing capacities of E. faecalis were found to be necessary to invade and dominate a pre-established in vitro F. nucleatum biofilm.

Conclusions

This study reveals multifaceted mechanisms underlying the physical binding and antagonistic interaction between F. nucleatum and E. faecalis, which could play a potential role in the shift of microbial composition in primary and secondary endodontic infections.

Introduction

Endodontic infections are categorized as primary and secondary infection according to the time when microbial infection occurs [1]. Primary infection is caused by invasion and colonization of the pulp by oral microbes, while secondary infection typically results from a persistent microbial root canal infection after initial endodontic treatment due to retention or introduction of microbes during treatment of primary infection [2].

Many studies have characterized microbes associated with primary and secondary infections and established that they harbor distinct microbial communities [1–3]. Based on current knowledge, Fusobacterium nucleatum is mainly associated with primary endodontic infection, and often detected in high prevalence and high abundance [3–9]. While the prevalence of F. nucleatum in secondary infection remains relatively high, its abundance drops markedly [8,10,11]. In contrast, Enterococcus faecalis is rarely detected in primary infection, but during secondary infection, its prevalence increases to 20–77% in general while its abundance could be as high as 90% [8,11–15]. These intriguing findings promoted us to investigate the potential interaction between F. nucleatum and E. faecalis in the current study.

Both E. faecalis and F. nucleatum have been extensively studied as individual pathogens [15–19]. E. faecalis is a Gram-positive facultative anaerobe that naturally inhabits the human gastrointestinal tract [20,21]. Known virulence functions of E. faecalis include its ability to penetrate dentin tubules to establish biofilms and to survive for a prolonged period in harsh environments such as low pH, low nutrition, and low oxygen [15,16,22]. F. nucleatum is an oral Gram-negative strict anaerobe with stringent requirements for growth and survival [23,24]. F. nucleatum plays a key role as a ‘bridging’ organism to promote oral biofilm formation [25]. F. nucleatum’s connection with endodontic infection also includes virulence factors, such as dysregulation of inflammasomes in dental pulp cells [26]. Except for an early report on the possible in vitro co-aggregation between F. nucleatum and E. faecalis [27], the interaction between these two important oral pathogens remains largely unexplored. In this study, we used physiological and genetic approaches to characterize the interaction between E. faecalis and F. nucleatum.

Methods

Bacterial strains and culture conditions

F. nucleatum wild type strain ATCC 23726 and its mutant derivatives defective in outer membrane autotransporter proteins, including Fn1449(fap2), Fn1526(radD), Fn2058(aim1), Fn2047, Fn0254, Fn1554, Fn1253(aid1), and Fn1893 [28], were maintained under anaerobic conditions (10% H₂, 10% CO₂, 80% N₂) at 37°C on either Columbia agar supplemented with 5% sheep blood or in Columbia Broth (CB; BD Difco, Detroit, MI, USA). Thiamphenicol (MP Biomedicals, Irvine, CA) at 5 μg per mL was used for selection and maintenance of F. nucleatum mutant strains possessing the catP determinant. E. faecalis wild type strain OG1RF (ATCC 47077) and clinical strain DX1 (isolated by Dr. Xuesong He’s laboratory) were grown aerobically at 37°C on brain-heart infusion (BHI; BD Difco, Detroit, MI, USA) agar plates or broth. For two-species cocultures, we used JVN (Josamycin, Vancomycin and Norfloxacin) selective agar plates in anerobic conditions to grow and count F. nucleatum cells [29] and used BHI plates in aerobic conditions to grow and count E. faecalis cells.

Coaggregation assay

Coaggregation visual assays were performed according to the following protocol [28]: exponential phase cultures of E. faecalis and F. nucleatum cells were washed with and re-suspended in Coaggregation buffer (CAB, 150 mM NaCl, 1 mM Tris, 0.1 mM CaCl₂, 0.1 mM MgCl₂·H₂O; pH 7.5) to a final concentration of 2 × 10⁹ cells per mL. Suspensions of strains to be examined for coaggregation were combined with an equal volume of a test strain adjusted to the same cellular concentration in CAB to a total volume of 1 mL in a reaction tube. The reaction mixtures were immediately vortexed for 10s and incubated for 120 min. The evaluation was performed using a visual scoring system ranging from 0 to 4. A score of 0 was assigned for no visible coaggregation and a score of 4 described complete sedimentation of strains with a clear supernatant [30].

For the coaggregation inhibition assay, either D-galactose, D-Mannose, L-arginine, L-leucine, L-glutamic acid was added to the reaction tube containing only F. nucleatum cells to a final concentration of 100 mM. The suspension was then vortexed and incubated for 5 min prior to the addition of the coaggregation test partner. Once the partner strain was added, the reaction mixture was vortexed again and the assay was evaluated by the quantitative coaggregation assay. The final concentration of each inhibitor per coaggregation reaction was 50 mM [31].

The quantitative (spectrophotometric) coaggregation assays were performed similar as the visual assay described above with the following additional steps: optical density measured at 600 nm (OD600) of reaction mixtures were obtained spectrophotometrically. After 120 min of incubation, reaction mixtures were centrifuged at low speed (100 g for 1 min) to pellet the co-aggregated cells while leaving non-aggregated bacteria in suspension. OD600 of the supernatants were measured after the 120 min incubation. Relative coaggregation of species A and B was determined by dividing the difference between the total turbidity of each partner strain and the coaggregation supernatant turbidity by the total turbidity of each partner strain using the formula: {[OD600(A)+OD600(B)]−OD600(A + B)}/[OD600(A)+OD600(B)] [32].

Competition assay

The competition assays on agar plates between E. faecalis and F. nucleatum were performed using a previously established protocol [33]. For competition assays under planktonic conditions, exponential phase cultures of E. faecalis and F. nucleatum were resuspended in pre-reduced fresh CB to a final OD600 of 0.1, equivalent to approximately 1 × 10⁸ cells per mL. An equal volume of E. faecalis and F. nucleatum were mixed, and cultures were incubated at 37°C anaerobically for 24 h and 48 h. CFU was then measured at each timepoint using the growth conditions described above.

pH measurements and pH buffered media

pH measurements of bacterial supernatants were performed using a pH meter. Buffered pH medium was made by mixing the medium with filter sterilized buffer K₂HPO₄/KH₂PO₄ (pH 7.4) to a 22 mM/14 mM final concentration.

Hydrogen peroxide (H₂O₂) measurement

For Amplex Red (Invitrogen Thermo Fisher Scientific, Eugene, OR, USA) assay, all testing reagent solutions were prepared according to manufacturer’s instruction. 1X Reaction Buffer without H₂O₂ working solution was used as the negative control and 10 μM H₂O₂ was used as the positive control. Diluted test solutions in 1X Reaction Buffer and a volume of 50 µL was used for each reaction. 50 µL of the Amplex Red reagent solution was added to each black/clear bottom microplate well containing the standards, controls, and samples. The plates were incubated at room temperature in the dark for 30 min before the color change was observed. Fluorescence was then measured with a fluorescence microplate reader using excitation at 530 ± 12.5 nm and fluorescence detection at 590 ± 17.5 nm. Background fluorescence, determined for a non-H₂O₂ control reaction, is subtracted from each value.

Effect of H₂O₂ on F. nucleatum viability

Two concentrations (100 µM and 200 µM) of the H₂O₂ were added to the Columbia broth to evaluate their effects on growth of F. nucleatum ATCC23726. F. nucleatum viability was measured by CFU counting at each timepoint (30 min, 60 min and 120 min). F. nucleatum cells on Columbia broth (CB; BD Difco, Detroit, MI, USA) without H₂O₂ was used as the control.

Preparation of E. faecalis-incubated PBS solution and catalase rescue assay

To prepare E. faecalis-incubated PBS solutions, E. faecalis cells were pelleted from an exponential phase culture, washed three times with buffered PBS and resuspended in buffered PBS to an OD600 of 1. The E. faecalis-containing PBS solution was then incubated at 37°C under micro-aerobic conditions (N₂ 90%, CO₂ 5%, O₂ 5%) for 2 h and centrifuged (10 min at 17,000 x g, 4°C). The supernatant was collected by filtration (0.22 μm) and used to measure H₂O₂ production and F. nucleatum killing capacity. For the catalase rescue assay, 2.5 mg of catalase (Thermo Scientific, USA) was added to 1 ml E. faecalis-incubated PBS solution or fresh unbuffered PBS solution and incubated at 37°C anaerobically for 12 h. Then, exponential phase F. nucleatum cells were added to the solutions, incubated under micro-aerobic conditions for 2 h, and assayed for viability.

Biofilm based binding and invasion assays

The 35 mm glass bottom dish (Cellvis, CA, USA) was treated with poly-lysine (Electron microscopy sciences, PA, USA) [34]and air dried thoroughly prior to inoculation. To grow F. nucleatum biofilm, F. nucleatum cells (concentration~ 1 × 10⁸ cells per mL) were seeded into each dish, which contains 1 mL of pre-reduced CB supplemented with hemin at 5 μg per mL and menadione at 1 μg per mL. Each dish was allowed to grow biofilm under anaerobic conditions (5% H₂, 5% CO₂, 90% N₂) at 37°C for 48 h prior before E. faecalis cells (concentration ~ 2 × 10⁹ cells per mL) were added for binding and invasion studies. For binding assays, 2 mL of E. faecalis (concentration ~ 2 × 10⁹ cells per mL) were added to the F. nucleatum biofilm and incubated for 30 min. The unbound E. faecalis cells were washed away with pre-reduced CB (three times). The resulting biofilms were scraped off from the dishes, vortexed extensively to be dispersed into single cells, and counted for both F. nucleatum and E. faecalis respectively within the mixture using the same assays described above. For invasion assays, 2 mL of pre-reduced CB was added to the washed F. nucleatum biofilms that have been pre-incubated with E. faecalis for 30 min, and incubated anaerobically for another 4 h, 10 h and 24 h before being counted for F. nucleatum and E. faecalis respectively.

Results

E. faecalis and F. nucleatum binding is mediated by the fusobacterial adhesin Fap2

We explored whether E. faecalis and F. nucleatum could physically bind to each other. Employing a coaggregation assay [28], we showed that F. nucleatum wildtype ATCC 23726 coaggregated with E. faecalis ATCC type strain OG1RF (Figure 1(a)) under planktonic conditions. Using a quantitative coaggregation assay [32], we further demonstrated that, in addition to E. faecalis OG1RF, a clinical isolate, E. faecalis DX1, also showed strong coaggregation with F. nucleatum ATCC 23726 (Figure 1(d)).

Figure 1.

The qualitative and quantitative coaggregation assay for E. faecalis (Ef) and F. nucleatum (Fn). (a) Coaggregation assay to screen eight different F. nucleatum ATCC23726 outer-membrane protein mutants for defective coaggregation phenotype with E. faecalis OG1RF. The arrow points to the fap2 mutant with binding deficiency. (b) Coaggregation assay to screen sugars and amino acids for their inhibitory efforts on interaction between E. faecalis and F. nucleatum. The arrow points to galactose with inhibitory effort. (c) Microscopic observation of coaggregation between E. faecalis and F. nucleatum under different conditions: a. E. faecalis alone; b. F. nucleatum alone; c. E. faecalis and F. nucleatum; d. E. faecalis and F. nucleatum fap2. For each interacting pair, the pictures of 10 random views were taken and 1 representative image was shown. 100X magnification. The scale bar is 10 µm. (d) Spectrophotometric coaggregation assay to quantitatively measure coaggregation between wild type E. faecalis (Ef WT)/ E. faecalis clinical isolate (Ef DX1) and wild type F. nucleatum (Fn WT) and its outer-membrane protein mutants. (e) Spectrophotometric coaggregation assay to quantitatively measure coaggregation between E. faecalis and F. nucleatum in the presence of selected sugars and amino acids, which are D-galactose (Gal), D-Mannose (Man), L-arginine (Arg), L-leucine (Leu) and L-glutamic acid (Glu). All co-aggregation assays have 120 min incubation time. The data presented are based on experiments conducted in triplicate.

F. nucleatum is known to have a set of autotransporter-like outer membrane proteins that interact with other microbes, and our laboratory has previously generated mutants for genes encoding these proteins [28]. To explore the potential mechanism underlying E. faecalis-F. nucleatum binding, the entire panel of these outer membrane protein mutants was tested for their coaggregation ability with E. faecalis. The qualitative coaggregation measurements shown in Figure 1(a) and Table 1 indicate that only the fap2 mutant of F. nucleatum was defective in binding to E. faecalis. Quantitative measurement (Figure 1(d)) confirmed that the fap2 mutant displayed significantly reduced coaggregation capability with E. faecalis compared to wild-type F. nucleatum. When observed under the microscope, as shown in Figure 1(c), while F. nucleatum and E. faecalis monocultures remained suspended as individual cells, large aggregates were present when both species were co-incubated. This coaggregation phenotype was absent when the F. nucleatum fap2 mutant was co-incubated with E. faecalis. These microscopic visual observations further validated the conclusion that F. nucleatum binds to E. faecalis via its Fap2 adhesin.

Table 1.

Scoring coaggregation between E. faecalis (Ef) with F. nucleatum (Fn) wildtype (WT) and its mutants. The scores were determined after 120 min of incubation. Scores were assigned as follows: 0 – no visible co-aggregation; 1 – small aggregates that stay suspended; 2 – larger aggregates that settle slowly and leave the supernatant turbid; 3 – large aggregates that settle quickly but leave the supernatant slightly turbid; 4 – complete sedimentation with a clear supernatant. Experiments were repeated independently at least three times.

	Fn WT	Δfap2	Δrad D	Δaim1	Δ2047	Δ0254	Δ1554	Δaid1	Δ1893
Auto-aggregation	0	0	0	0	0	0	0	0	0
Ef OG1RF	4	0	4	4	4	4	4	4	4

F. nucleatum fap2 encodes a galactose-inhibitable adhesin that has been previously shown to be required for hemagglutination, coaggregation, and adherence to mammalian cells [35]. The data above suggest that Fap2 plays a role in mediating the interaction with E. faecalis as well. To further confirm this finding, we tested if the observed binding between E. faecalis and F. nucleatum might be galactose inhibitable. As shown in Figure 1(b) and Table 2, we confirmed that 50 mM galactose indeed resulted in a drastic reduction in the coaggregation between E. faecalis and F. nucleatum, while other selected carbohydrates or amino acids at the same concentration did not have any significant inhibitory effects. The strong inhibitory effect of galactose on the coaggregation was further supported by a quantitative coaggregation assay (Figure 1(e)).

Table 2.

Scoring inhibitory effect of sugars or amino acids on co-aggregation between E. faecalis and F. nucleatum. The scores were determined after 120 min of incubation. Scores were assigned as follows: 0 – no visible coaggregation; 1 – small aggregates that stay suspended; 2 – larger aggregates that settle slowly and leave the supernatant turbid; 3 – large aggregates that settle quickly but leave the supernatant slightly turbid; 4 – complete sedimentation with a clear supernatant. Experiments were repeated independently at least three times.

	CAB	D-Galactose	D-Mannose	L-Arginine	L-Leucine	L-Glutamic acid
Ef OG1RF/ Fn WT	4	0	3	4	4	4

E. faecalis exerts an inhibitory effect on F. nucleatum growth

After demonstrating physical binding between E. faecalis and F. nucleatum, we explored possible physiological interactions between these two bacteria. A plate-based competition assay showed that, when E. faecalis and F. nucleatum cells were spotted simultaneously in proximity (less than 1 mm apart) on an agar plate, E. faecalis exhibited an inhibitory effort against F. nucleatum (Figures 2(a,b)). Since there was no overlap of the two bacterial species spotted on the agar surface, the result indicates that the antagonistic effect is mediated through E. faecalis production of diffusible molecule(s). Quantitative assays were developed to measure the outcome when E. faecalis and F. nucleatum were co-aggregated and co-incubated under planktonic conditions. Consistent with the result of plate-based competition assay, growth of F. nucleatum in planktonic conditions was significantly inhibited by E. faecalis (Figures 2(c,d)). The CFU data shown in Figures 2(c,d) indicate that the inhibitory effort is bactericidal instead of bacteriostatic.

Figure 2.

The qualitative and quantitative analysis of E. faecalis’ (Ef) impact on F. nucleatum (Fn) growth under different conditions. (a) Colony competition assay on agar plate to measure the interaction between E. faecalis wildtype OG1RF and F. nucleatum wildtype ATCC23726 when two bacteria were inoculated more than 10 mm apart. (b) Colony competition assay on plate to measure the interaction between E. faecalis and F. nucleatum when two bacteria were inoculated right next to each other (less than 1 mm). (c) Competition assay under planktonic condition when E. faecalis and F. nucleatum were co-incubated for 24 h and 48 h, then CFU of E. faecalis in the mixture were counted with selected agar plate in comparison with E. faecalis grown in monoculture. (d) Competition assay under planktonic condition when E. faecalis and F. nucleatum were co-incubated for 24 h and 48 h, then CFU of F. nucleatum in the mixture were counted with selected culture in comparison with F. nucleatum grown in monoculture. Graph shows means and SD of readings from two individual experiments performed in triplicates. *P < 0.05; ** P < 0.01; *** P < 0.001.

E. faecalis inhibits F. nucleatum by generating an acidic environment

To further explore the killing factor(s) of E. faecalis against F. nucleatum, we tested the pH of monocultures and co-cultures of E. faecalis and F. nucleatum during exponential growth after starting at pH 7. The pH of E. faecalis overnight cultures was around 5.09, the pH of F. nucleatum was about 6.39, and the co-culture pH was around 5.33. We examined the growth of F. nucleatum at pH 5 and pH 7, and found that the growth of F. nucleatum in the pH 5 medium was significantly inhibited when compared to growth in neutral pH medium (Figure 3(a)), suggesting that the low pH environment generated by E. faecalis could be one of the inhibitory factors.

Figure 3.

The acidic environment generated by E. faecalis inhibits F. nucleatum growth. (a) pH challenging assay. F. nucleatum ATCC 23726 viability was tested under different pH conditions in Columbia broth. The data presented are based on experiments conducted in triplicates. WT, wild type. (b) Competition assay under buffered pH condition. Quantitative analysis of the competition effects between E. faecalis (Ef) and F. nucleatum (Fn) under buffered pH condition. Data are expressed as relative CFU (%) compared with the Fn alone CFU at each timepoint as 100%. Graph shows means and SD of readings from two individual experiments performed in triplicate. Data represent the means and standard deviation of at least three independent experiments. *P < 0.05; ** P < 0.01; *** P < 0.001. WT, wild type.

To further investigate if low pH may act as an inhibitory factor, we repeated the planktonic competition assay using the same setup described above in a medium buffered by high concentrations of K₂HPO₄/ KH₂PO₄ to maintain the pH at 7. While F. nucleatum survived better in this buffered medium, it still suffered a significant reduction in CFU when co-cultured with E. faecalis in comparison with its own monoculture (P < 0.05) (Figure 3(b)). The data indicate that even though the acidic environment generated by E. faecalis can play an important role in inhibiting F. nucleatum under the conditions tested, there are other killing factor(s) at play.

E. faecalis inhibits F. nucleatum by producing H₂O₂ under starvation stress conditions

It has been well-documented that E. faecalis is able to produce H₂O₂ [36], which may also contribute to the observed growth inhibition of F. nucleatum. When co-cultured with E. faecalis, we demonstrated, using the Amplex Red kit, that E. faecalis is indeed able to generate H₂O₂ (Figure 4(a)). By varying the experimental conditions using different growth media (such as Brain-Heart Infusion, Columbia Broth or PBS) and atomic environments (such as N₂, CO₂ or O₂), we found that E. faecalis was able to generate significant amounts of H₂O₂ (as high as 167 µM extracellular H₂O₂) when incubated in PBS with a brief oxygen exposure or in a micro-aerobic environment (Figure 4(a)).

Figure 4.

E. faecalis (Ef) H₂O₂ production and its impact on F. nucleatum (Fn) viability under starvation condition. (a) AmplexRed assay to measure extracellular H₂O₂ production by E. faecalis in PBS solution under starvation condition. (b) F. nucleatum viability in the presence of H₂O₂. Bacterial viability of F. nucleatum (CFU) after 30 min, 60 min and 120 min of incubation with 100 µM H₂O₂ and 200 µM H₂O₂ under starvation condition. (c) F. nucleatum viability (CFU) in filtered E. faecalis-incubated PBS (EIP) solution or fresh PBS with or without catalase treatment. Graph shows means and SD of readings from two individual experiments performed in triplicate. Columns with the symbol * are statistically different from the control group. *P < 0.05; ** P < 0.01; *** P < 0.001.

We then tested F. nucleatum’s sensitivity to H₂O₂. As shown in Figure 4(b), 100 µM H₂O₂ is sufficient to kill F. nucleatum, which suggests that H₂O₂ generated by E. faecalis (167 µM extracellular H₂O₂ detected in PBS solution) would be able to kill F. nucleatum. For experimental validation, we incubated E. faecalis in buffered PBS solution for two hours, collected the supernatant via centrifugation and filtration, then added fresh F. nucleatum cells into the filtered supernatants. As expected, Figure 4(c) shows that the E. faecalis-incubated PBS solutions exhibited strong killing against F. nucleatum in comparison to the fresh PBS solution without E. faecalis incubation. Most interestingly, when we treated the E. faecalis-incubated PBS solutions with catalase, the resulting solution was no longer able to kill F. nucleatum (Figure 4(c)), suggesting H₂O₂ is a major killing factor under the tested experimental conditions.

It is worthwhile to mention that pH and H₂O₂ based E. faecalis killing against F. nucleatum under planktonic condition (as described above) seems to be binding independent because F. nucleatum wildtype and the fap2 mutant were found to be killed in the same rate (data not shown). However, the outcome is very different under the biofilm condition as described below.

E. faecalis cells bind and dominate pre-established F. nucleatum biofilm

The biofilms of F. nucleatum wildtype and the fap2 mutant were pre-established and subjected to the addition of E. faecalis wildtype cells. As shown in Figure 5(a), after 30 min incubation followed by extensive washing, a high number of E. faecalis cells were still bound to the F. nucleatum wildtype biofilm, while significantly lower numbers of E. faecalis cells were able to integrate into the preformed F. nucleatum fap2 biofilm. The data indicate that Fap2-mediated E. faecalis-F. nucleatum binding is critical in mediating the integration of E. faecalis into the pre-existing F. nucleatum biofilm. Meanwhile, when the same F. nucleatum biofilms bound with E. faecalis were further incubated in growth media for longer periods of time (4, 10 and 24 h), E. faecalis exhibited its ability to kill F. nucleatum and dominated the biofilm with extensive growth. Compared to wildtype, F. nucleatum fap2 mutant biofilm cells suffered less viability loss during the first hours (4 and 10 h) of incubation (Figures 5(a,b)), likely due to less bound E. faecalis cells to the fap2 mutant biofilm. However, at 24 h, E. faecalis within both F. nucleatum wildtype and fap2 mutant biofilms grew to same biomass, as reflected by the similar CFU, which was accompanied by the almost total loss of viability of both the F. nucleatum wildtype and the fap2 mutant.

Figure 5.

E. faecalis (Ef) binds and dominates pre-establishedF. nucleatum (Fn) biofilms. (a) The viability of E. faecalis wildtype OG1RF (CFU) after binding with the pre-established F. nucleatum ATCC23726 biofilms (both wildtype, WT, and the fap2 mutant) for 30 min then being washed extensively. The resulting biofilms were then incubated in the growth medium for another 4 h, 10 h and 24 h and counted for E. faecalis CFU. (b) The viability of F. nucleatum after the resulting biofilms were incubated in the growth medium for another 4 h, 10 h and 24 h. Time 0 refers to the timepoint right after the washing step and added fresh medium before the following incubation. Data are expressed as relative percentage CFU via comparing F. nucleatum cells from the E. faecalis invasion group vs F. nucleatum alone as 100% for both WT and the fap2 mutant, respectively. Graph shows means and SD of readings from experiments performed in duplicates. The black asterisk * denotes statistical difference between the experimental data at 4 h, 10 h and 24 h in comparison with the data at time 0 of the same co-culture (F. nucleatum WT/ E. faecalis or F. nucleatum fap2/ E. faecalis). The red asterisk * denotes statistical difference between different cocultures of F. nucleatum WT/ E. faecalis and F. nucleatum fap2/ E. faecalis groups at the same timepoint. *P < 0.05; ** P < 0.01; *** P < 0.001.

Discussion

There has been extensive research looking at interspecies interactions and investigating their impact on bacterial physiology and disease association within the oral microbiome. However, the studies on the interactions between critical individual microorganisms found at the endodontic infection lesion and their impact on pathogenicity is very limited. Here, we used various in vitro assays to demonstrate a strong binding between E. faecalis and F. nucleatum under both planktonic and biofilm conditions, which is mediated by the F. nucleatum outer-membrane protein Fap2. Furthermore, we showed that E. faecalis exhibited a strong killing ability against F. nucleatum by generating an acidic micro-environment and producing hydrogen peroxide. These interactions may contribute to the observation that E. faecalis was able to invade and dominate pre-established F. nucleatum biofilms.

Bacteria do not exist as isolated individuals, but rather as members of multispecies communities. To adapt and cope with the ever-changing microenvironment, they need to engage in constant interaction with neighboring residents. Thus, inter-species interactions are the most important processes influencing bacterial physiology, ecology, disease initiation, and disease progression. Our study presents in vitro evidence emphasizing the importance of microbial interspecies interaction in the context of endodontic infection.

It has been well documented under clinical conditions that there is often a shift in E. faecalis abundance, from being seldom detected to one of the most dominant species within the lesion during the transition from primary to secondary endodontic infection [3,5,8,9,11–13]. Yet it remains to be fully determined how this shift happens. Being able to penetrate dentin tubules, cope with harsh environments such as extensive starvation and resist H₂O₂ could contribute to E. faecalis’s survival after failed root canal treatments for primary infection and allow E. faecalis to gain a competitive advantage in secondary infection [22]. Our study shows that E. faecalis displays a strong physical binding to and inhibitory effect against F. nucleatum, one of the most dominant pathogens in primary endodontic infection, which may facilitate its colonization and contribute to its increased prevalence in secondary infection.

There are many follow-up questions that need be addressed to obtain a more comprehensive mechanistic understanding of F. nucleatum-E. faecalis interaction and its role in endodontic infection progression. This could be partly achieved through genetic and molecular studies to address questions such as: what is the counterpart of F. nucleatum Fap2 in E. faecalis and how does the physical interaction impact gene expression in E. faecalis and F. nucleatum.

We fully recognized that the real disease process within the infected root canal is very complex and dynamic, the in vitro data presented in this study may not fully recapitulate the in vivo situation. Therefore, additional in vivo studies are being planned to further validate the hypothesis. We have collected many more clinical samples from the root canals of primary and secondary endodontic infection patients. These samples will be subjected to DNA extraction and deep-sequencing, which will provide additional data for us to further explore the endodontic microbiome shift during the progression of primary to secondary infection. The special attention will be paid to the shift from F. nucleatum to E. faecalis among these samples. Forsyth has an existing mouse model system where endodontic infection could be induced by injecting endodontic pathogens into mice’s root canals [37]. We plan to use this animal model to conduct in vivo studies to explore whether E. faecalis is able to attach, invade and dominate a pre-established F. nucleatum biofilm within a mouse root canal.

Mouth is the gateway to the GI tract. It is interesting to note that the natural habitat of E. faecalis is the gut rather than the oral cavity, yet it manages to colonize inside the tooth root canal and become one of the most dominant endodontic pathogens in secondary infections, through the interaction with F. nucleatum. This shows the deep connection between oral and gut microbiome especially under disease conditions. The inhibitory function of E. faecalis over F. nucleatum may also allow E. faecalis to keep F. nucleatum in check in the gut.

In conclusion, our data reveal the antagonistic interaction between two key endodontic pathogens, which may help to shed light on microbial shifts from primary to secondary infection. The study also provides future directions to further understand endodontic microbial infections at molecular and genetic levels, which may lead to the development of diagnostic and therapeutic tools against endodontic infections.

Funding Statement

This work was supported by the National Natural Science Foundation of China; Forsyth institute internal fund

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author Contributions

D. Xiang, contributed to design the experiments, data acquisition, analysis, and interpretation, drafted and critically revised the manuscript; P. Dong, L. Cen, contributed to data interpretation, and critically revised the manuscript; B. Bor and R. Lux, contributed to conception, data interpretation, drafted and critically revised the manuscript; W. Shi, Q. Yu, X. He, and T. Wu, contributed to conception, design, data analysis, interpretation, drafted and critically revised the manuscript. All authors gave final approval and agreed to be accountable for all aspects of the work.