New Insights into the Antimicrobial Effect of Mast Cells against Enterococcus faecalis

Abstract

Enterococcus faecalis has emerged as an important cause of life-threatening multidrug-resistant bacterial infections in the hospital setting. The pathogenesis of enterococcal infections has remained a relatively neglected field despite their obvious clinical relevance. The objective of this study was to characterize the interactions between mast cells (MCs), an innate immune cell population abundant in the intestinal lamina propria, and E. faecalis. This study was conducted with primary bone marrow-derived murine MCs. The results demonstrated that MCs exerted an antimicrobial effect against E. faecalis that was mediated both by degranulation, with the concomitant discharge of the antimicrobial effectors contained in the granules, and by the release of extracellular traps, in which E. faecalis was snared and killed. In particular, the cathelicidin LL-37 released by the MCs had potent antimicrobial effect against E. faecalis. We also investigated the specific receptors involved in the recognition of E. faecalis by MCs. We found that Toll-like receptors (TLRs) are critically involved in the MC recognition of E. faecalis, since MCs deficient in the expression of MyD88, an adaptor molecule required for signaling by most TLRs, were significantly impaired in their capacity to degranulate, to reduce E. faecalis growth as well as to release tumor necrosis factor alpha (TNF-α) and interleukin 6 (IL-6) after encountering this pathogen. Furthermore, TLR2 was identified as the most prominent TLR involved in the recognition of E. faecalis by MCs. The results of this study indicate that MCs may be important contributors to the host innate immune defenses against E. faecalis.

INTRODUCTION

Enterococcus species are part of the normal intestinal microbiota. From here, they can disseminate and cause a variety of infections in the immunocompromised host, including urinary tract infections, bacteremia, endocarditis, and sepsis (1). In recent years, Enterococcus faecalis has emerged as an important cause of nosocomial infection, with mortality rates exceeding 50% in critically ill patients, cancer patients, and some transplant patients (2, 3). The intrinsic resistance of E. faecalis to several of the most commonly used antibiotics, including penicillinase-resistant penicillins, cephalosporins, lincosamides, nalidixic acid, low-level aminoglycoside, and low-level clindamycin as well as its extreme capacity to acquire resistance to the remaining antibiotic classes, such as penicillin, chloramphenicol, tetracyclines, rifampin, fluoroquinolones, aminoglycosides (high levels), and vancomycin (4) by the acquisition of mobile genetic elements (5) is of major concern for most hospitals and health care facilities. In particular, the increasing resistance of enterococci to vancomycin and teicoplanin is a challenging and serious public health concern (6).

Asymptomatic E. faecalis colonization frequently precedes clinical infection, and colonized individuals represent a potential source for the spread of the microorganism. Unfortunately, there is no known effective therapy to decolonize patients with antibiotic-resistant enterococci. The increasing health care problem posed by E. faecalis highlights the urgent need for new avenues of therapeutic treatment and/or decolonization. In this regard, greater insight into the host-pathogen interactions during enterococcal infection and a better understanding of the mechanisms of host defense against this pathogen will be the foundation for advances in treatment and prevention modalities.

An important front of host defense in the intestinal mucosa consists of various innate immune cells, including mast cells (MCs) and macrophages, which sense entry of foreign elements into the mucosa and orchestrate an appropriate inflammatory response (7, 8). MCs, in particular, are multifunctional and highly effective tissue-dwelling cells, which are considered important components of the immune system. MCs have an important immunoregulatory function, particularly at the mucosal border between the body and the environment (9). They are located in close proximity to blood vessels, where they can efficiently regulate vascular permeability (10). In addition, MCs are able to modulate the activities of neighboring effector cells through the release of a broad array of prestored or newly synthesized mediators (11). MCs produce four main classes of mediators, including a wide variety of cytokines and chemokines, preformed granule-associated mediators, newly generated lipid mediators, and endogenous antimicrobial agents such as antimicrobial peptides or reactive oxygen species (ROS), which can be released from the granules upon activation (12). These mediators have been shown to exhibit various roles in tissue remodelling, angiogenesis, cellular recruitment, and/or change of vascular permeability and host defense, such as the recruitment of neutrophils to the site of infection (11). Despite their known role in the initiation of allergic reactions, chronic inflammatory processes, and activation during certain types of parasitic infections (13, 14), there is now clear evidence that MCs also play a prominent role in the early immune response to invading pathogens (10, 15). MCs can also contribute to host defense by exerting a direct antimicrobial effect against pathogens either through phagocytosis (16) or by releasing extracellular traps (MCETs), which are structures composed of DNA, histones, and granule proteins where the pathogens are snared and killed (17, 18). After translocation through the intestinal barrier, E. faecalis comes most probably into contact with the MCs present beneath the intestinal epithelium. However, relatively little is known about the MCs' interactions with E. faecalis. Because MCs play such a prominent role in regulating the immune response, we can predict that these interactions may be of high relevance for fine-tuning the immune response to this invading pathogen. In the current study, we have investigated the dynamic interplay between E. faecalis and MCs. Our results show that MCs exerted an antimicrobial activity against E. faecalis that was mediated by the release of MCETs and by the discharge of granular antimicrobial compounds. We also show that recognition of E. faecalis by MCs was mediated by Toll-like receptors (TLRs), since the production of cytokines and degranulation of MCs in response to E. faecalis was impaired in MCs lacking the common adaptor molecule MyD88, which is required for almost all TLR activation cascades. Furthermore, we could narrow down the recognition of E. faecalis by MCs to TLR2. The results obtained in this study provide new insight into the interplay between E. faecalis and components of the innate immune system.

MATERIALS AND METHODS

Bacteria.

E. faecalis strains DSMZ 20478 and OG1RF as well as different clinical isolates kindly provided by M. Probst-Kepper (Städtisches Klinikum Braunschweig, Germany) were used in this study. Stock cultures were maintained at −80°C and cultured at 37°C in brain heart infusion (BHI) medium for 6 h. Bacteria were harvested at mid-log phase (optical density at 600 nm [OD₆₀₀], ∼0.5), centrifuged for 10 min at 4,000 rpm, washed with sterile phosphate-buffered saline (PBS), and adjusted to 10% transmission at 600 nm. Bacteria suspensions were further diluted in cell culture medium to achieve the required concentrations. For some experiments, E. faecalis was cocultivated with LL-37 (AnaSpec, San Jose, CA) added for 90 min.

Mice.

Inbred, pathogen-free, 8- to 12-week-old C57BL/6 mice were purchased from Harlan-Winkelmann (Borchen, Germany). C57BL/6 TLR2^−/− mice were a gift from S. Weiss (Helmholtz Center for Infection Research, Braunschweig, Germany), and C57BL/6 MyD88^−/− mice were kindly provided by T. Sparwasser (TWINCORE, Hannover, Germany). Mice were maintained under standard conditions and according to institutional guidelines. All experiments were approved by the ethics board Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, Oldenburg, Germany.

Generation of bone marrow-derived MCs (BMMCs).

MCs were isolated and differentiated as previously described (18). Briefly, mice were sacrificed by CO₂ asphyxiation and the femurs and tibias were removed. The bone marrow was harvested by repeated flushing with Iscove's modified Dulbecco's medium (IMDM). Bone marrow cells were incubated in IMDM supplemented with 10% fetal calf serum (FCS), 2 mM l-glutamine, 1 mM pyruvate, 100 U/ml of penicillin, 100 μg/ml of streptomycin, and 20 U/ml of recombinant murine interleukin 3 (IL-3; BioLegend, San Diego, CA) for 21 days. Nonadherent cells were transferred to fresh culture plates every 2 to 3 days to remove adherent macrophages and fibroblasts. The resulting cell population consisted of 98% MCs as determined by flow cytometry analysis using anti-mouse CD117 antibody (Caltag Laboratories, Hamburg, Germany) and a FACSCalibur TM flow cytometer (Becton Dickinson, San Jose, CA).

MC infection assay.

BMMCs were harvested, washed twice with IMDM without antibiotics, and seeded in 48-well tissue culture plates at a density of 2 × 10⁶/ml. E. faecalis was added to wells containing BMMCs at a multiplicity of infection (MOI) of 1 bacterium per cell (1:1) or to wells containing medium alone. The MOI of 1:1 was chosen based on a previous report (18) to ensure optimal conditions for the investigation of extracellular killing mechanisms elicited by MCs. Kinetics of bacterial growth in the presence or absence of MCs was monitored at various time points of infection by determination of CFU in the cell cultures. The total amount of bacteria in the wells was determined by plating serial dilutions of 20 μl volume of each well. The percentage of bacterial growth inhibition in the presence of MCs was calculated using the bacterial growth based on CFU after 90 min with the following formula: [(CFU in medium alone − CFU in medium with MCs)/CFU in medium alone] × 100.

In some experiments, BMMCs were treated with 1 μM Ca²⁺ ionophore ionomycin (Sigma, Deisenhofen, Germany), 100 μM cromolyn (Sigma, Deisenhofen, Germany), and/or 50 mU of micrococcal nuclease (New England BioLabs, Frankfurt am Main, Germany).

Toluidine stain.

A total of 5 × 10⁵ MCs per well were infected with E. faecalis at an MOI of 10:1 and incubated for 2 h at 37°C and 5% CO₂. MCs were then fixed with 3% paraformaldehyde (PFA) for 10 min at room temperature, and 100 μl of the MC suspension was centrifuged on slides at 150 × g for 5 min and dried on air at room temperature. Five microliters of 0.5% toluidine blue in PBS was placed on the slides, which were incubated further for 5 min at room temperature. After the incubation step, slides were rinsed with water until the blue color disappeared and analyzed by light microscopy (Zeiss Axiophot microscope with an attached Zeiss Axiocam HRc digital camera and Axiovision software 4.8; Carl Zeiss, Oberkochen, Germany).

β-Hexosaminidase assay.

MCs were centrifuged at progressing times after stimulation, and degranulation was determined by assessing the percentage of β-hexosaminidase released into the culture supernatant according to the method of Kuehn et al. (19). Briefly, the cell pellets were lysed with distilled H₂O and the extracts were analyzed for total β-hexosaminidase activities. A 50-μl volume of cell culture lysates and 100 μl of 2 mM p-nitrophenyl-N-acetyl-β-d-glucosaminidase in 0.04 M sodium citrate buffer (pH 4.5) were added to each well of a 96-well plate, and color was allowed to develop for 30 min at 37°C. The enzyme reaction was terminated by adding 50 μl of 0.4 M glycine-NaOH (pH 10.7). The absorbance at 405 nm of each sample was measured with a 96-well Tecan Sunrise reader (Tecan Group Ltd., Männedorf, Switzerland). Total release of β-hexosaminidase was calculated as the percentage of the maximum release of C48/80 degranulated cells; untreated MCs served as positive controls.

Cytokine determination.

The determination of IL-6 and tumor necrosis factor alpha (TNF-α) levels was performed by specific enzyme-linked immunosorbent assay (ELISA), using matched antibody pairs and recombinant cytokines as standards. Briefly, 96-well microtiter plates were coated with the corresponding purified anti-human capture monoclonal anti-IL-6 or TNF-α antibody (Pharmingen, San Jose, CA) at a concentration of 2 μg/ml in sodium bicarbonate buffer overnight at 4°C. The wells were washed and then blocked with 1% bovine serum albumin-PBS before the serum samples and the appropriate standard were added to each well. Biotinylated rat monoclonal anti-IL-6 or -TNF-α antibody (BD Pharmingen) at 2 μg/ml was added as the second antibody. Detection was performed with streptavidin-peroxidase, and the plates were developed by use of TMB (3,3′,5,5′-tetramethylbiphenyl-4,4′-diamine).

Quantitative RT-PCR (qRT-PCR).

Total RNA was prepared using the GeneJET RNA purification kit (Fisher Scientific, Schwerte, Germany). RNA was reverse transcribed with reverse transcriptase (Hoffmann-La Roche, Basel, Switzerland), and cDNA synthesis was performed with a Gibco reverse transcription-PCR (RT-PCR) kit according to the manufacturer's instructions. The single-stranded cDNA was then subjected to PCR amplification under standard reaction conditions. The PCR primer sequences for the TLR2 gene were as follows: forward, 5′-GCT CCA GGT CTT TCA CCT CTA TTC-3′; reverse: 5′-TCC AGC AGG AAA GCA GAC TCG CTT A-3′. The sequences for the housekeeping gene β-actin are as follows: forward, 5′-TGG AAT CCT GTG GCA TCC ATG AAA C-3′, and reverse, 5′-TAA AAC GCA GCT CAG TAA CAG TCC G-3′. The resultant PCR products were electrophoresed on a 2% agarose gel, stained with ethidium bromide, and photographed (Gel Doc XR system; Bio-Rad, Hercules, CA).

Generation of skin air pouches.

Mice were anesthetized with isoflurane (Isoba; Essex Tierarznei, Munich, Germany) and infected subcutaneously with 5 × 10⁷ CFU of green fluorescent protein (GFP)-expressing E. faecalis. Mice were killed by CO₂ inhalation at 2 h after bacterial inoculation, and the infiltrating inflammatory cells were isolated from the side of infection by extensively rinsing with warm IMDM. Inflammatory cells were incubated for 5 min at 4°C with anti-CD16/CD32 antibodies to block the Fc receptor (FcR), followed by phycoerythrin (PE)-conjugated anti-CD117 antibodies. After incubation for 30 min at 4°C, cells were washed and flow cytometry analysis was performed using a FACSCalibur TM (Becton Dickinson). MCs were gated according to their expression of CD117 antigen (FL2). MCs associated with green-labeled enterococci were identified by the expression of green fluorescence (FL1).

LIVE/DEAD staining.

MCs were seeded on glass coverslips, incubated for 5 min, and then infected with E. faecalis at an MOI of 10 bacteria per mast cell. After 2 h of incubation at 37°C in a 5% CO₂ atmosphere, gentamicin (100 μg/ml) was added to kill nonphagocytosed bacteria, and MCs were further incubated for 2 h at 37°C and 5% CO₂. Discrimination between intact and damaged MCs was performed using the LIVE/DEAD viability/cytotoxicity kit for animal cells (Molecular Probes, Leiden, The Netherlands) according to the manufacturer's instructions. Briefly, infected MCs were washed with PBS and incubated with 250 μl of PBS containing 2 μM calcein AM and 8 μM ethidium homodimer 1. Cells were incubated at room temperature for 15 min and photographed under fluorescein isothiocyanate illumination in an epifluorescence microscope (Zeiss Axiophot, Zeiss, Germany) coupled to a camera (model AxioCam HRc; Zeiss) and analyzed by AxioVision 4.8 software (Zeiss).

Visualization of MC extracellular traps.

To examine the release of extracellular traps by fluorescence microscopy, BMMCs were seeded on poly-l-lysine-covered glass coverslips, infected with E. faecalis at an MOI of 1:1, and fixed with 4% paraformaldehyde at 4 h of infection. BMMCs were then stained using the LIVE/DEAD cell viability kit for mammalian cells (Invitrogen, Karlsruhe, Germany) by following the manufacturer's recommendations and examined using a Zeiss Axiophot microscope with an attached Zeiss Axiocam HRc digital camera and Axiovision 4.8 software (Zeiss). The LIVE/DEAD BacLight bacterial viability kit (Invitrogen) was used in some experiments to determine the viability of E. faecalis according to the recommendations of the manufacturer.

Double-immunofluorescence microscopy.

A total of 5 × 10⁵ MCs were seeded on poly-l-lysine-treated coverslips and infected with E. faecalis at an MOI of 10 to 1. After 90 min of incubation, the coverslips were rinsed and cells were fixed with 3.7% formaldehyde. For double-immunofluorescence staining, extracellular bacteria were stained with polyclonal rabbit anti-E. faecalis antibodies, followed by Alexa green-conjugated goat anti-rabbit antibodies (Sigma, Deisenhofen, Germany). After several washes, the cells were permeabilized by 0.01% Triton X-100 in PBS and washed again, and intracellular bacteria were stained by anti-E. faecalis antibodies, followed by Alexa red-conjugated goat anti-rabbit antibodies (Sigma). The fluorescence images were obtained using a Zeiss Axiophot microscope with an attached Zeiss Axiocam HRc digital camera and Axiovision 4.8 software (Zeiss).

Electron microscopy.

For scanning electron microscopy, samples were fixed with 5% formaldehyde and 2% glutaraldehyde in cacodylate buffer for 1 h on ice and washed in Tris-EDTA (TE) buffer. Dehydration was performed with a graded series of acetone, and samples were critical-point dried with CO₂ and sputter coated with gold-palladium before examination in a Zeiss field emission scanning electron microscope (Merlin) at 5 kV using the Everhart-Thornley secondary electron (SE) detector and the in-lens SE detector in a 25:75 ratio.

Statistics.

Data were analyzed by using Excel 2007 (Microsoft Office) or GraphPad Prism 5.0 (GraphPad sofware). All data are presented as means ± standard deviations (SDs). Comparison between groups was made by use of t test or one-way analysis of variance (ANOVA) or nonparametric Mann-Whitney test. P values of ≤0.05 were considered significant.

RESULTS

E. faecalis interacts with MCs in vitro and in vivo.

Scanning electron microscopy was used to visualize the interactions of E. faecalis with MCs in in vitro cultures. As shown in Fig. 1A and B, E. faecalis attached to the surface of MCs. To demonstrate that this interaction also occurs in an in vivo environment, GFP-expressing E. faecalis was injected subcutaneously in the backs of mice and the inflammatory cells recruited to the site of infection were collected at 2 h after bacterial inoculation. Infiltrated cells were stained with PE-conjugated anti-CD117 antibodies to identify the MC population by flow cytometry. Anti-CD117 antibody was specific for MCs and did not bind E. faecalis (see Fig. S1 in the supplemental material). The results showed that approximately 28% ± 2% of MCs isolated from the infected skin contained GFP-expressing E. faecalis (Fig. 1C). As MCs have been previously shown to induce growth inhibition of several pathogens (17, 18), we determined if MCs were also capable to influence the growth of E. faecalis in in vitro cultures. For this purpose, the growth of E. faecalis was determined over time in the presence and absence of MCs. Results in Fig. 1D show that MCs were also very efficient at inhibiting the growth of E. faecalis. MCs were capable of inhibiting the growth of a wide range of clinical isolates (Fig. 1E).

FIG 1.

Association of E. faecalis with MCs. (A) Scanning microscopy picture of E. faecalis attached to the surface of MCs (bacteria are indicated by white arrows). (B) High magnification of an MC harboring E. faecalis (white arrow). Bars represent 3 μm in (A) and 2 μm in (B). (C) Histogram analysis showing MCs associated with E. faecalis during in vivo infection. Mice were subcutaneously inoculated with 5 ×10⁷ GFP-expressing E. faecalis organisms, infiltrated cells were isolated from the site of bacterial inoculation at 2 h of infection, and the amount of MCs (CD117⁺) containing E. faecalis was determined by flow cytometry. The percentage of MCs containing E. faecalis was determined by the increase in green fluorescence (GFP) within the gated CD117⁺ cell population. The results of one representative experiment out of three are shown. (D) Growth kinetics of E. faecalis in the presence (squares) or absence (circles) of MCs (MOI, 1:1). A compilation of the results of three independent experiments is shown. *, P < 0.05. (E) Growth inhibition of E. faecalis clinical isolates in the presence of MCs determined at 90 min of coculture. The percentage of bacterial growth inhibition in the presence of MCs was calculated using net bacterial growth based on CFU after 90 min with the following formula: [(CFU in medium alone − CFU in medium with MCs)/CFU in medium alone] × 100. Each bar represents the mean ± SD of triplicates from three independent experiments.

TLR2 signaling is crucially involved in MC recognition of E. faecalis.

As TLR2 is among the most important pattern recognition receptors for Gram-positive microorganisms, we examined the potential contribution of TLR2 to recognition of E. faecalis by MCs. For this purpose, we characterized the response of MCs derived from mice deficient in MyD88, an adaptor molecule required for signaling by most TLRs, to in vitro challenge with E. faecalis. MyD88^−/− MCs were significantly impaired in their capacity to degranulate in response to E. faecalis, as measured by the release of β-hexasominidase (Fig. 2A). MyD88^−/− MCs were also less effective than wild-type MCs at inhibiting E. faecalis growth (Fig. 2B).

FIG 2.

(A and B) MCs derived from wild-type TLR2^−/− or MyD88^−/− mice were impaired in their capacity to degranulate (A) and to inhibit growth of E. faecalis (120 min of infection) (B). Each bar represents the mean ± SD of quadruplicates from two independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (C to E) Scanning microscopy picture of E. faecalis cocultivated with either wild-type MCs (C), TLR2^−/− MCs (D), or medium alone (E) (white arrows indicate membrane vesicle formation in the surface of E. faecalis). Bars represent 1 μm.

MCs deficient in the expression of TLR2 also exhibited significantly lower levels of degranulation (Fig. 2A) and growth-inhibitory effects against E. faecalis (Fig. 2B) than wild-type MCs. Furthermore, E. faecalis cocultured with wild-type MCs showed vesicle-like structures on the bacterial surface (Fig. 2C) which were less evident on bacteria cultivated with TLR2^−/− MCs (Fig. 2D) or untreated control bacteria (Fig. 2E). The number of enterococci displaying vesicle-like structures after cocultivation with either wild-type MCs, TLR2^−/− MCs, or cell culture medium was quantified by scanning electron microscopy after counting more than 300 individual organisms in 10 selected chains. A significantly higher percentage of E. faecalis bacteria producing vesicles was found after cocultivation with wild-type MCs (89.9%) than with TLR2^−/− MCs (12.2%) or control bacteria (6.4%) (P < 0.005 for E. faecalis cocultured with wild-type MCs versus TLR2^−/− MCs and P < 0.005 for E. faecalis cocultured with MCs versus medium alone).

The release of inflammatory cytokines such as IL-6 (Fig. 3A) and TNF-α (Fig. 3B) by MCs in response to E. faecalis was also mediated by the TLR2 signaling pathway, since MCs deficient in the expression of MyD88 or TLR2 released significantly lower levels of these cytokines after exposure to E. faecalis. Furthermore, the pathogen induced the upregulation of TLR2 in MCs, as demonstrated by the increased level of tlr2 mRNA in MCs after exposure to E. faecalis (Fig. 3C).

FIG 3.

(A and B) MCs derived from either wild-type, TLR2^−/−, or MyD88^−/− mice were altered in their capacity to release IL-6 (A) and TNF-α (B) in response to E. faecalis. Each bar represents the mean ± SD of three independent experiments. **, P < 0.01. (C) Expression of tlr2 mRNA on MCs after infection with E. faecalis. MCs were exposed to E. faecalis for 2 h, washed, and further incubated in the presence of gentamicin for 6 h and 24 h. Total RNA was isolated, followed by quantitative measurement of tlr2 mRNA expression in MCs after infection with E. faecalis by qRT-PCR. β-actin served as an internal control, and the results of the qRT-PCR were calculated according to the threshold cycle (ΔΔC_T) method (38). One representative result out of three is shown.

MCs release antimicrobial extracellular traps (MCETs) in response to E. faecalis.

We next investigated the potential mechanism used by MCs to inhibit growth of E. faecalis. To determine if killing of E. faecalis by MCs was dependent on phagocytic uptake, MCs were infected with E. faecalis for 90 min and the amount of internalized bacteria was determined by double-immunofluorescence microscopy. The fluorescence microscopy photographs in Fig. 4A show that E. faecalis is scarcely internalized by MCs. These results suggest that the antimicrobial effect of MCs against E. faecalis was essentially mediated by extracellular mechanisms.

FIG 4.

Double-immunofluorescence staining of E. faecalis-infected MCs (A) and MC released extracellular traps after exposure to E. faecalis. MCs were infected with E. faecalis at an MOI of 10 to 1 for 90 min and stained for double immunofluorescence. Extracellular bacteria were stained by anti-E. faecalis antibodies, followed by Alexa green-conjugated goat anti-rabbit antibodies (ii). MCs were then permeabilized by 0.01% Triton X-100 in PBS, and intracellular bacteria were stained by anti-E. faecalis antibodies, followed by Alexa red-conjugated goat anti-rabbit antibodies (iii). The DNA is stained in blue (i). The lack of red bacteria in the merged multifluorescence picture (iv) shows that E. faecalis is located mainly extracellularly. MCs were seeded on poly-l-lysine-coated glass slides and left untreated (B) or infected with E. faecalis for 3 h (C). MCs were then fixed with 4% paraformaldehyde and examined by immunofluorescence microscopy after histone (red) and DAPI (blue) staining. Bars represent 10 μm. (D) Microscopic quantification of MCETs released per field of view (40× magnification) in untreated control, E. faecalis-infected (MOI 10:1), and PMA (200 μM)-treated MCs. Each bar represents the mean ± SD of three independent experiments. **, P < 0.01; ***, P < 0.001.

As it has been previously reported that MCs release MCETs with an antimicrobial effect (17), we evaluated whether MCs released ETs after encountering E. faecalis. MCs were infected for 3 h with E. faecalis or medium alone, fixed, and processed for immunofluorescence microscopy. Histone staining reveals formation of MCETs by MCs in coculture with E. faecalis (Fig. 4C) but not by MCs cultured in medium alone (Fig. 4B). Quantification of MCETs released after exposure to E. faecalis is shown in Fig. 4D. MCs stimulated for 3 h with phorbol myristate acetate (PMA) served as a positive control.

We next determined the potential contribution of MCETs to the E. faecalis growth inhibition exerted by the MCs. For this purpose, we analyzed the viability of E. faecalis cultured in the presence of MCs for 3 h using LIVE/DEAD staining. The confocal microscopy image in Fig. 5A shows that E. faecalis microorganisms entrapped in the MCETs were dead (arrows). However, we observed that a proportion of free E. faecalis organisms were also dead, suggesting that MCs also killed E. faecalis in a MCET-independent manner (Fig. 5B). To determine the proportional contribution of MCETs to the growth inhibition of E. faecalis, MCs were cocultivated with E. faecalis at an MOI of 1:1 for 90 min in the presence or absence of 50 mU of endonuclease, which has been shown to dismantle MCETs by digesting the nuclear backbone (17). Destruction of MCETs resulted in only partial growth inhibition of E. faecalis by MCs, suggesting that an additional mechanism was also involved (Fig. 5C).

FIG 5.

LIVE/DEAD staining of MCs infected with E. faecalis. (A and B) MCs were seeded on poly-l-lysine coated glass slides, then infected with E. faecalis for 3 h at an MOI of 10:1, fixed with 4% paraformaldehyde, and examined by immunofluorescence microscopy after staining with LIVE/DEAD reagents. Dead bacteria appear red, and viable bacteria appear green. Arrows indicate dead bacteria. Bars represent 10 μm (A) and 5 μm (B). (C) Growth kinetics of E. faecalis in medium alone (circles), in medium with nuclease (triangles), or in coculture with either untreated (squares) or nuclease-treated (inverted triangles) MCs. Each point represents the mean ± SD of triplicates from three independent experiments. *, P < 0.05.

Release of antimicrobial compounds by MCs contributes to the growth inhibition of E. faecalis.

We next determined whether the additional mechanism used by MCs to inhibit the growth of E. faecalis was mediated by the release of antimicrobial compounds contained in the MCs granules. Indeed, toluidine blue staining indicated a massive degranulation of MCs after encountering E. faecalis (Fig. 6B). Toluidine blue-stained uninfected MCs are shown in Fig. 6A for comparison. The degranulation of MCs was also evidenced by scanning electron microscopy (Fig. 6C to E). We then determined the contribution of antimicrobial compounds released by MCs after degranulation to the growth inhibition of E. faecalis using MCs which had been treated with cromolyn, a stabilizing agent that inhibits MC degranulation. Treatment with cromolyn strongly reduces the capacity of MCs to inhibit the growth of E. faecalis. Thus, E. faecalis inhibition by MCs was reduced to 27.4-fold when MCs were treated with cromolyn to block degranulation (Fig. 6F).

FIG 6.

Contribution of granule components released by MCs to the antimicrobial effect against E. faecalis. (A and B) Uninfected MCs (A) or MCs infected for 3 h with E. faecalis (B) were fixed and stained with toluidine blue. (C to E) Scanning electron photographs showing an uninfected MC (C) and a MC degranulating after infection with E. faecalis at an MOI of 10:1 (bacteria on the MC surface are indicated by white arrows) (D). (E) Enlarged view of E. faecalis on the MC surface. Bars represent 10 μm (A) and (B), 5 μm (C and D), and 1 μm (E). (F) Growth kinetics of E. faecalis in medium alone (circles), in medium with cromolyn (triangles), or in coculture with either untreated (squares) or cromolyn-treated (inverted triangles) MCs. Each point represents the mean ± SD of triplicates from three independent experiments. (G) Growth of E. faecalis in medium alone (white bar) or cocultured with MCs either left untreated (dark gray bar) or treated with cromolyn (light gray bar), nuclease (hatched bar), or both (black bar) for 90 min. Each point represents the mean ± SD of triplicates of three independent experiments. *, P < 0.05; ***, P < 0.005.

Together, these results indicate that the antimicrobial effect exerted by MCs against E. faecalis was mediated in part by the release of MCETs and in part by the release of antimicrobial compounds after degranulation. To confirm this postulation, we determined the levels of growth inhibition of E. faecalis by degranulated MCs and in the presence of nuclease to dismantle the MCETs. MCs treated with cromolyn and nuclease were completely impaired in their capacity to inhibit the growth of E. faecalis (Fig. 6G). These results confirmed the contributions of both MCETs and release of granule compounds to the antimicrobial effect of MCs against E. faecalis.

The cathlecidin LL-37 exhibits strong antimicrobial effects against E. faecalis.

One important antimicrobial factor released by MCs is the cathlecidin LL-37 (20, 21). Therefore, we examined the effect of LL-37 on E. faecalis. As shown in Fig. 7A, LL-37 exerted a strong antimicrobial effect against E. faecalis at a concentration of 10 μg/ml. This effect was also observed when different clinical isolates of E. faecalis were tested (data not shown). E. faecalis incubated in the presence of LL-37 (10 μg/ml) displayed the same membrane vesicle formation as was already observed on bacteria cocultivated with MCs; nontreated E. faecalis bacteria served as controls (Fig. 7B and C). The number of enterococci displaying vesicle-like structures after cultivation in the absence or in the presence of LL-37 (10 μg/ml) was determined by scanning electron microscopy after counting 100 individual bacteria (data not shown). A significantly higher percentage of E. faecalis bacteria produced vesicles after treatment with LL-37 (94.4%) than did bacteria grown in medium alone (5.7%) (P < 0.0005).

FIG 7.

Antimicrobial effect of LL-37 against E. faecalis. (A) Growth of E. faecalis in the presence of different concentrations of cathelicidin LL-37 for 120 min. Each bar represents the mean ± SD of quadruplicates from two independent experiments. *, P < 0.05. Scanning electron microscopy photographs of untreated (B) or LL-37-treated (10 μg/ml) (C) E. faecalis. Arrows indicate membrane vesicle formation on the surface of E. faecalis. Bars represent 1 μm. (Inset) Enlarged scanning electron micrograph. Arrows indicated vesicle formation on the bacterial surface. Bar represents 1 μm.

DISCUSSION

The nosocomial human pathogens E. faecalis and E. faecium are classically considered the main cause of enterococcal bacteremia. Enterococci are the third most common pathogens isolated from human bloodstream infections (22–25), and it has been previously reported that up to 90% of enterococcal infections in humans are caused by E. faecalis (26). In addition to bloodstream infections, they are also important causes of urinary tract infections, endocarditis, and intra-abdominal and pelvic infections not only in the hospital environment but also, even more concerning, in the community environment (27). The high propensity of these pathogens to acquire and express new antibiotic resistance determinants further increases their ability to sustain antibiotic selection, impeding bacterial clearance and thus promoting gastrointestinal colonization (28). Furthermore, transfer of transposable mobile genetic elements carrying antibiotic resistance determinants from E. faecalis to other bacterial genera by broad-host-range conjugative elements aggravates the problems associated with the spreading of multidrug-resistant pathogens (5).

Despite the clinical importance of E. faecalis infections, little is known about the interactions of E. faecalis with the host immune system. Such knowledge could lead to new strategies to improve the natural host resistance to this emerging pathogen. Investigations to unravel the interaction of enterococci with host immune cells have been strongly focused on the role of professional phagocytotic cells. In this regard, several reports have highlighted the crucial role of polymorphonuclear neutrophils in the early control of E. faecalis (29). However, little is known about the interactions of E. faecalis with other important immune cell types with high immune modulatory functions, like MCs. MCs are inflammatory cells which are typically located immediately beneath the epithelial surfaces exposed to the outer environment such as the skin and the mucosa but also the respiratory, genitourinary, and gastrointestinal tracts (10). Because many of these sites are also common entry ports for pathogens, MCs represent one of the first types of immune cells encountered by an invading pathogen and therefore initiating the inflammatory immune response against these microbes (30). Recent evidence has suggested that MCs have a beneficial contribution to both innate and adaptive immunity during infection (31). In this study, we performed an in-depth characterization of the interactions of E. faecalis with MCs. The results demonstrated that MCs exhibited a remarkable antimicrobial effect against E. faecalis.

MCs have been shown to be able to kill bacteria by two different mechanisms (16): (i) phagocytosis (mainly Gram-negative bacteria) and/or (ii) extracellularly either by the release of antimicrobial peptides such as LL-37 by degranulation or by the formation of MCETs (17) or by a combination of both (18). We found no evidence that E. faecalis could actively invade MCs as was previously demonstrated for other Gram-positive pathogens, like Staphylococcus aureus (18), or Gram-negative bacteria, like Escherichia coli (32). These pathogens are able to persist for long periods intracellularly within MCs without losing viability by gaining access into MCs by a route distinct from the classical endosome-lysosome pathway.

Our study shows that MCs became activated after encountering E. faecalis and exerted a direct extracellular antimicrobial activity against this pathogen that was mediated by various extracellular mechanisms. One of those mechanisms involves the formation of MCETs, where E. faecalis microorganisms are trapped and killed. However, the level of MCET formations was not as pronounced as has been reported for other pathogens, like Streptococcus pyogenes (17), and therefore, killing of E. faecalis by MCETs cannot account fully for the antimicrobial effect of MCs observed in this study. This was further confirmed by the diminished but still significant antimicrobial effect of MCs after dismantling of the MCETs by nuclease treatment. The reason why only a percentage of MCs in the cultures released MCETs after encountering E. faecalis is not yet clear; however, this is likely to reflect a heterogeneity in the physiological status of the MCs in the culture. A similar phenomenon has been reported for extracellular trap formation by neutrophils (33).

The second extracellular mechanism used by MCs to control E. faecalis consisted of the discharge of granule antimicrobial compounds. Several studies have shown that activation of MC degranulation leads to an effective killing of various Gram-positive as well as Gram-negative bacteria (34). In the present study, blocking of MC degranulation by cromolyn strongly impaired the antimicrobial effect exerted by MCs against E. faecalis, which argues for a predominant contribution of MC degranulation to the antimicrobial response of MCs to E. faecalis. MC degranulation is generally accompanied by the release of antimicrobial peptides such as cathepsin G or cathelicidins (20, 35), which are known to inhibit growth of various Gram-positive bacteria, including S. pyogenes or S. aureus (20, 36, 37). Our results show that E. faecalis was very sensitive to the antimicrobial effect of cathelicidin LL-37, indicating a potential major role for this antimicrobial peptide in the antimicrobial activity of MCs against E. faecalis, which has been suggested in other studies (38–41). Scanning electron microscopy examination revealed that E. faecalis microorganisms exposed to either MCs or LL-37 developed vesicle-like structures on the bacterial surface. Such vesicles have been described mainly for Gram-negative microorganisms, like E. coli (42), Salmonella enterica (43), or Pseudomonas aeruginosa (44), and some Gram-positive microorganisms, like S. aureus (45, 46), Streptococcus pneumoniae (47), mycobacteria (48), and Bacillus anthracis (49). The function of these vesicles is not yet clear, but they may be related to bacterial responses to environmental stress, which in our study might have been produced by the antimicrobial compounds of MCs acting in the bacterial cell wall.

There are multiple direct and indirect pathways by which MCs can be selectively activated by pathogens. These include TLRs, coreceptors and complement receptors, among which the TLR family occupies a central position. Signaling via these receptors guides the immune system to produce an effective immune response to invading pathogens. We also investigated in this study the receptors involved in the recognition of E. faecalis by MCs. Using MCs deficient in MyD88, a critical adaptor molecule for most TLRs, we demonstrated that recognition of E. faecalis by MCs was largely mediated by TLRs. Among the different TLRs, we selectively evaluated the role of TLR2 because of its important role in the recognition of Gram-positive microorganisms. TLR2 signaling has been reported to be critically involved in the control of other pathogens, like Francisella tularensis, by MCs (50). Furthermore, it has been demonstrated that lipoteichoic acid of E. faecalis induces the expression of chemokines in macrophages, mainly via the TLR2/PAFR signaling pathway (51). Also, the involvement of other TLRs in the immune response to enterococci has been demonstrated in recent studies involving the recognition of enterococcal nucleic acids by the endosomal TLR7 and TLR9 in macrophages (52). Our results reveal an important role for TLR2 signaling for activation of MCs in response to E. faecalis. Thus, MCs deficient in the expression of TLR2 were significantly impaired in their capacity to release inflammatory cytokines such as IL-6 and TNF-α. IL-6 release by MCs was recently shown to be important to regulate the selective influx of dendritic cells (DCs) into inflamed lymph nodes, thus enhancing opportunities for effective T cell-DC interaction and therefore for mounting an effective adaptive immune response to bacterial pathogens (53). In particular, the release of TNF-α by local MCs can be seen as an important protective mechanism elicited by MCs during enterococcal infection, because TNF-α serves as an important chemoattractant for phagocytic cells like polymorphonuclear leukocytes (PMNs) and macrophages, which are involved in the elimination of this pathogen. Furthermore, the antimicrobial effect of MCs against E. faecalis was strongly reduced in TLR2^−/− MCs in comparison with wild-type MCs. We also found that TLR2^−/− MCs exhibited a capacity to release MCETs similar to that of wild-type MCs (data not shown); we speculated that MC degranulation and release of antimicrobial compounds may be driven by TLR2 signaling. Indeed, TLR2^−/− MCs were strongly impaired in their capacity to degranulate after encountering E. faecalis. Interestingly, our results also showed that MCs upregulated TLR2 expression after encountering E. faecalis. Increasing the availability of TLR2 by the MCs can serve to amplify the magnitude of the inflammatory response mediated by this receptor.

In summary, this study provides for the first time experimental evidence that MCs exert antimicrobial activity against E. faecalis and thereby supports a protective effect of this immune cell type against this important pathogen. Further studies are, however, needed to elucidate the further facets of MC-E. faecalis interactions as well as the bacterial determinants involved in MC activation. Understanding the mechanisms and scope of the contribution of MCs to host defense will be crucial to regulating their activity therapeutically during bacterial infections.