Increased Resistance to Staphylococcus aureus Endophthalmitis in BALB/c Mice: Fas Ligand Is Required for Resolution of Inflammation but Not for Bacterial Clearance

Abstract

FasL was recently shown be required for bacterial clearance in C57BL/6 mice that express the FasL.1 allotype. The FasL.2 allotype is expressed in BALB/c mice and exhibits increased binding affinity to and increased cytotoxic activity against Fas⁺ target cells. Therefore, we hypothesized that BALB/c mice would be more resistant to Staphylococcus aureus-induced endophthalmitis. To test this hypothesis, C57BL/6, BALB/c, and BALB(gld) mice received intravitreal injections of 2,500 CFU of S. aureus (RN6390). Clinical examinations, electroretinography (ERG), histology, and bacterial quantification were performed at 24, 48, 72, and 96 h postinjection. The myeloperoxidase (MPO) assay was used to quantitate neutrophil infiltration. At 96 h postinfection, 86% of C57BL/6 mice presented with complete destruction of the eye, compared to only 29% of BALB/c mice with complete destruction. To our surprise, in the absence of Fas ligand, BALB(gld) mice showed no difference in bacterial clearance compared to BALB/c mice. However, histology and ERG analysis revealed increased retinal damage and significant loss of retinal function. MPO analysis revealed equal numbers of neutrophils in BALB(gld) and BALB/c mice at 24 h postinfection. However, at 48 h, the neutrophil numbers remained significantly elevated in BALB(gld) mice, correlating with the increased retinal damage observed in BALB(gld) mice. We conclude that the increased resistance to S. aureus induced endophthalmitis in BALB/c mice is not dependent upon the FasL. However, in contrast to C57BL/6 mice, FasL is required for resolution of inflammation and protecting host tissue from nonspecific damage in BALB/c mice.

INTRODUCTION

Bacterial endophthalmitis is a rare but serious infection of the posterior segment of the eye that can occur as a complication following ocular surgery or globe-penetrating injury, often resulting in significant retinal damage and loss of vision (1–4). The retinal damage occurs directly, as a consequence of bacterial toxins and cytolytic factors that lyse cells and damage host tissue, as well as indirectly, as a consequence of host inflammation that results in bystander damage of normal tissue (5–7). The extent of damage is dependent upon variables that are interconnected, the virulence of the infecting organism, and the severity of the host inflammatory response. Consequently, in spite of aggressive therapy with broad-spectrum antibiotics aimed at the infecting pathogen, endophthalmitis often leads to partial or complete loss of vision due to the damage of photoreceptor cells and/or retinal detachment (1, 2). Thus, more successful treatment of bacterial endophthalmitis will require not only the elimination of the infecting bacteria but also proper regulation of the host immune response to minimize nonspecific bystander destruction of normal tissue.

The host immune response to intraocular infection is multifactorial, with the clearance of a bacterial infection dependent upon the early recruitment and activation of neutrophils (8). Several studies have demonstrated the importance of cytokines and chemokines in a timely recruitment and activation of neutrophils (9). More recently, however, Fas ligand has also been shown to be critical in the clearance of bacterial endophthalmitis and may play a direct role in neutrophil recruitment and activation of neutrophils (10), supporting earlier reports that identified FasL as a potent chemoattractant for neutrophils (11, 12). Fas ligand is constitutively expressed in tissues of the normal eye and has been shown to play a central role in the maintenance of the immune-privileged environment by inducing apoptosis in infiltrating inflammatory cells (13). On the other hand, in a model of Staphylococcus aureus-induced endophthalmitis, it has been shown that while normal mice readily cleared an infection of S. aureus (500 CFU), mice deficient in Fas ligand were unable to clear the same-size inoculum (10). In the absence of FasL, bacteria grew more rapidly and fewer neutrophils were recruited to the site of infection, suggesting that FasL is important for early activation of innate immunity within the eye (10).

Findings from our laboratory support the concept that FasL is important in triggering the early phase of innate immunity. We demonstrated that FasL engagement of Fas⁺ macrophages not only induces apoptosis but also induces the production of neutrophil chemotactic factors (14). Moreover, using an ocular tumor model, we demonstrated that overexpression of the membrane form of FasL within the immune-privileged eye induced a potent neutrophil-mediated inflammatory response, leading to the termination of ocular immune privilege (15, 16). Together, these studies suggest that a more efficient and/or rapid engagement of the Fas receptor on cells resident within the eye will trigger a quicker innate-mediated inflammatory response, accelerate pathogen clearance, and ultimately increase resistance to bacterial endophthalmitis.

In order to test whether increased activation of the Fas receptor by FasL leads to more efficient bacterial clearance, we examined whether a polymorphism of murine Fas ligand, which increases the specific activity of FasL, affects resistance to S. aureus-induced endophthalmitis. Sequence analysis of FasL cDNA from several strains of mice determined that inbred mice segregate into two FasL allotypes: mFasL.1 and mFasl.2 (17). The mFasL.2 allotype exhibits 9-fold-higher binding affinity to the Fas receptor and cytotoxic activity against Fas⁺ target cells than mFasL.1. Polymorphisms in the FasL gene have also been found in humans and have been associated with increased risk of some cancers (18–20). However, it is not known how these FasL polymorphisms may influence the immune response and susceptibility to infection in humans. Moreover, the C57BL/6 mice used in previous reports express the mFasL.1 allotype (weaker binding affinity), whereas BALB/c mice express the mFasL.2 allotype (stronger binding affinity). Motivated by our data showing that FasL engagement of the Fas receptor induces not only apoptosis but the production of proinflammatory mediators and neutrophil chemotactic factors, we asked whether BALB/c mice that express the mFasL.2 allotype develop an innate-mediated inflammatory response that is more robust than mice expressing the mFasL.1 variant and if this would result in increased resistance to bacterial endophthalmitis. The data presented herein indicate that BALB/c mice are indeed more efficient at bacterial clearance, resulting in increased resistance to endophthalmitis compared to that of C57BL/6 mice. However, the increased resistance in BALB/c mice was not dependent upon the Fas-FasL pathway. Our results indicate that in contrast to C57BL/6 mice, the Fas-FasL pathway is not required for bacterial clearance in BALB/c mice. Moreover, neutrophil infiltration is not dependent upon FasL expression, as shown previously in C57BL/6 mice. Rather in BALB/c mice, the Fas-FasL pathway is critical in resolution of inflammation and protecting the retina from nonspecific bystander tissue damage.

MATERIALS AND METHODS

Animals.

Female and male C57BL/6J mice, BALB/cJ mice, and BALB(gld) (Cpt.C3-Tnfsf6gld) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All animals were treated according to the Association for Research in Vision and Ophthalmology Resolution on the Use of Animals in Research. All procedures involving mice were approved by the Schepens IACUC.

Induction of endophthalmitis.

Mice received an intravitreal injection of 0.5 μl of sterile physiological saline (Hospira, Lake Forest, IL), in which 500, 2,500, 5,000, or 10,000 CFU of S. aureus (strain RN6390) was suspended, into the left eye, according to a previously described protocol (10). Briefly, an intravitreal injection of S. aureus, just posterior to the limbus-parallel conjunctival vessels, was performed using a borosilicate microcapillary pipette pulled to a tip size of 50 μm. Dilation of the pupil with 1% tropicamide (Alcon Laboratories, Inc., Fort Worth, TX) allowed direct visualization of the needle during the injection. During injection, if the lens or retina was damaged, the mouse was removed from the study. In addition, any mice that developed hemorrhage at the time of injection were also removed from the study. The right eye of each mouse was untreated and served as an internal control for electroretinography studies. Before injection, S. aureus cells were grown in standard Bacto brain heart infusion broth (BD Bioscience, Sparks, MD) and diluted in phosphate-buffered saline (PBS) during log phase to the appropriate concentration (500, 2,500, 5,000, or 10,000 CFU/0.5 μl).

Clinical scoring.

At 24, 48, 72, and 96 h following injection, clinical exams were performed by slit lamp microscopy. Mice were given a clinical score of 0 to 3, as previously described (21). A score of 0 was equivalent to a clear anterior chamber, clear vitreous, and clear view of the retina; a score of 1 demonstrated mild aqueous flare, mild vitreal haze, and a slightly obscured view of the retina. A score of 2 showed moderate aqueous flare, dense vitreal haze, poor pupil dilation, and a moderately obscured view of the retina; and a score of 3 had intense aqueous flare, opaque vitreous, and a completely obscured view of the retina.

Histological analysis and quantification of retinal folding.

Mice were sacrificed at 24, 48, 72, and 96 h following injection, and eyes were enucleated using Stevens curved sharp-tip scissors. Eyes were fixed in 10% buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. To quantitate the extent of retinal folding in BALB/c and BALB(gld) mice, each section (5 or 6 sections per eye) was examined by a blinded observer who counted the number of retinal folds per 10-μm section (5 or 6 sections per eye) (22, 23). These sections were cut through or immediately adjacent to the optic nerve.

Quantification of intraocular bacteria.

Mice were sacrificed at 24, 48, 72, and 96 h following injection, and eyes were enucleated using Stevens curved sharp-tip scissors. Eyes were placed in 1.0 ml sterile PBS on ice and homogenized by bead beating (FastPrep FP120; Thermo Fisher Scientific, Waltham MA) with 1.0-mm glass beads for 45 s at a maximum speed. Homogenates were serially diluted and plated on brain heart infusion agar plates as previously described (21).

Electroretinography.

Electroretinography was performed on mice at 24, 48, 72, and 96 h postinjection using a slight modification of a previous protocol (10). Briefly, mice were dark adapted for at least 4 h and were anesthetized, and the pupils were dilated using 1% tropicamide ophthalmic solution (Bausch & Lomb, Tampa, FL). Following anesthesia, body temperature was maintained at 37°C using a microwave heat pad. Gold wire electrodes (0.25 mm) (Alfa Aesar, Ward Hill, MA) were placed on the cornea after application of a hypromellose ophthalmic prism solution (Akorn, Inc.) and connected to a visual electrodiagnostic system (UTAS-E 3000; LKC Technologies, Gaithersburg, MD). Needle electrodes placed in the anterior scalp and the tail served as reference and ground leads, respectively. The b-wave amplitude (trough of a-wave to peak of b-wave) in response to a bright flash in a Ganzfeld illumination sphere was assessed in the injected left eye and contralateral (internal control) right eye simultaneously. A total of 30 readings taken at 0.6 cd-s/m² flash intensity with a 1-s interval between flashes were taken and averaged. The retinal function was defined as the ratio of the b-wave amplitudes of the experimental infected eye to the contralateral nontreated eye.

Myeloperoxidase assay.

Mice were sacrificed at 24 and 48 h following injection, and eyes were enucleated using Stevens curved sharp-tip scissors. Eyes were snap-frozen in liquid nitrogen and were homogenized in 1.0 ml of an assay buffer provided by an MPO fluorometric detection kit (Stressgen Assay Designs, Ann Arbor, MI). The homogenates were read using a multidetection microplate reader (Synergy 2; BioTEK, Highland Park, VT) and the concentration was derived from an MPO standard curve, prepared as directed in the detection kit.

Statistical analysis.

Where appropriate, the data were analyzed with an unpaired t test with P values of <0.05 as the basis for rejection of the null hypothesis. The Gehan-Breslow-Wilcoxon and log rank (Mantel-Cox) tests were used to analyze the percentage of eyes destroyed following infection. Statistical analysis and graphing were performed on a computer (Prism, Gen5, and Microsoft Excel).

RESULTS

Endophthalmitis in C57BL/6 and BALB/c mice.

To determine whether BALB/c mice are more resistant to S. aureus-induced endophthalmitis than C57BL/6 mice, wild-type BALB/c and C57BL/6 mice were injected intravitreally with an increasing concentration of S. aureus (500, 2,500, 5,000, and 10,000 CFU of S. aureus) (see Fig. S1 in the supplemental material). Eyes were examined by slit lamp microscopy and given a score of 0, 1, 2, or 3 (0, cleared; 3, destroyed). From these scores, the percentage of eyes destroyed (score of 3) was calculated (see Fig. S1). Both strains of mice efficiently cleared 500 CFU, while 10,000 CFU destroyed all eyes in both strains. However, a strain difference in resistance to endophthalmitis became apparent at the 2,500- and 5,000-CFU dose. At 2,500 CFU, 80% of C57BL/6 eyes were destroyed, compared to only 28% in the BALB/c group. Even at 5,000 CFU, the BALB/c mice exhibited increased resistance, where 90% of the C57BL/6 eyes were destroyed, compared to only 70% in the BALB/c group.

The 2,500-CFU dose, where the most significant difference in resistance was observed between strains, was chosen to further elucidate the strain difference in resistance to endophthalmitis. BALB/c and C57BL/6 mice were injected intravitreally with 2,500 CFU of S. aureus and eyes were examined by slit lamp microscopy and given a score of 0, 1, 2, or 3 (0, cleared; 3, destroyed) (Fig. 1C). From these scores, the percentage of eyes destroyed (score of 3) was calculated (Fig. 1D). In C57BL/6 mice, endophthalmitis developed rapidly, with 28% of the eyes destroyed as early as 48 h postinfection, compared to 0% in the BALB/c group. By 96 h postinfection, 86% of the C57BL/6 eyes infected with 2,500 CFU of S. aureus were destroyed, compared to only 29% of the BALB/c eyes. A representative example of disease progression in C57BL/6 versus BALB/c mice is shown macroscopically and histologically at 24, 48, 72, and 96 h postinfection (Fig. 1A and B). In BALB/c mice, slit lamp examination revealed mild inflammation that peaked at 48 h postinfection and subsequently resolved fully by 96 h postinfection. In contrast, C57BL/6 mice displayed severe inflammation, characterized by significant vitreal and aqueous haze and poor pupil dilation at 48 h postinfection. The marked difference in the percentage of eyes destroyed in C57BL/6 mice (86%) compared to that in BALB/c mice (29%) demonstrates increased resistance to bacterial endophthalmitis in BALB/c mice.

Fig 1.

Clinical examination and retinal histology of C57BL/6J and BALB/cJ mice following intravitreal injections of 2,500 CFU of S. aureus. At 24, 48, 72, and 96 h postinoculation, the eyes were examined by slit lamp microscopy. Representative eyes of BALB/c (A) and C57BL/6 (B) mice were photographed, enucleated, sectioned, and stained with H&E. (C) Each eye was given a score of 0, 1, 2, or 3 (0, cleared; 1, mild aqueous flare and vitreal haze; 2, moderate aqueous flare and dense vitreal haze; 3, destroyed). (D) At 96 h, 86% of eyes from C57BL/6 mice and 29% of eyes from BALB/c mice were destroyed (score, 3). H&E pictures were taken of whole eyes at ×4 magnification. n = 14 C57BL/6 mice, n = 15 BALB/c mice. *, P < 0.001 (Gehan-Breslow-Wilcoxon and log rank [Mantel-Cox] tests).

Bacterial quantification.

We performed studies to determine whether C57BL/6 mice contained a higher bacterial load, indicating an inability to effectively clear the infection. Bacterial quantification was performed on whole eyes enucleated at 24, 48, and 96 h postinfection (Fig. 2). No significant difference in bacterial load was detected between C57BL/6 and BALB/c mice at 24 h postinfection. At 48 h postinfection, the number of viable bacteria recovered from C57BL/6 eyes was approximately 3 logs greater than that from BALB/c mice. By 96 h, bacterial load in C57BL/6 mice remained approximately 3 logs greater (maximum, 2.4 × 10⁶ CFU/eye) than that in BALB/c (maximum, 6 × 10³ CFU/eye). These data indicate that C57BL/6 mice are unable to effectively clear a 2,500-CFU S. aureus inoculum and are therefore more susceptible to S. aureus-induced endophthalmitis.

Fig 2.

Bacterial quantification of C57BL/6 and BALB/c mice following intravitreal injections of 2,500 CFU of S. aureus. Bacterial quantification was performed at 24, 48, and 96 h postinoculum, and data are presented as CFU ± standard errors of the means (SEM). The bacterial quantification revealed a significantly increased bacterial load in C57BL/6 mice (n = 10) compared to that of BALB/c mice (n = 10) at 48 and 96 h postinoculum. **, P < 0.001.

Retinal function and histopathology.

Retinal function, as determined by the b-wave amplitude measured by ERG, corresponded with the histological findings. Control mice that received sterile-saline intravitreal injections displayed a transient loss of retinal function at 24 h postinjection, returning to 80 to 90% of normal by 96 h (Fig. 3A). BALB/c mice displayed a loss in retinal function that was approximately 40% of normal at 24 h postinfection and recovered to approximately 50% of normal by 96 h postinfection (Fig. 3A). In contrast, C57BL/6 displayed complete loss of retinal function at 24 h postinfection, with no recovery at 96 h postinfection. Histological analysis at 24 h postinfection revealed a significant neutrophil infiltrate in both BALB/c and C57BL/6 mice (Fig. 3B), and myeloperoxidase analysis revealed no significant difference in the number of infiltrating neutrophils (7.1 ± 0.2 [BALB/c] versus 7.4 ± 0.2 [C57BL/6] relative units of MPO; n = 10/group). At higher magnification, hematoxylin and eosin (H&E) staining of retinas from C57BL/6 mice reveals significant neutrophil infiltration throughout the inner and outer plexiform layers, correlating with substantial retinal damage (Fig. 3B). In contrast, in BALB/c mice, the neutrophils were primarily located in the vitreous and lining the retinal ganglion cell layer, with very few neutrophils infiltrating only the inner plexiform layer (Fig. 3B). By 96 h postinfection, the inflammation subsided significantly in BALB/c mice (Fig. 3C). However, the prolonged loss of retinal function (50% of normal) in BALB/c mice corresponded with the presence of retinal folding that was not observed in BALB/c mice injected with saline only (Fig. 3C). In contrast, histological analysis of eyes from C57BL/6 mice at 96 h postinfection demonstrated significant inflammation and complete destruction of the neural retina (Fig. 3C). Taken together, these data support the conclusion that C57BL/6 mice are more susceptible to S. aureus-induced endophthalmitis than BALB/c mice.

Fig 3.

Retinal histology and ERG assessment of retinal function in C57BL/6 and BALB/c mice following intravitreal injections of sterile saline or 2,500 CFU of S. aureus. (A) The b-wave amplitude was determined, and the data are presented as the ratio of injected eye to contralateral internal control eye ± SEM. The saline-injected control groups displayed a transient loss of retinal function. In contrast, C57BL/6 mice lost complete retinal function in the injected eye, while BALB/c mice displayed a loss in retinal function of 48% of normal. At 24 h (B) and 96 h (C), representative eyes were enucleated, sectioned, and stained with H&E. White arrows (neutrophils in the retina), black arrows (neutrophils along the RGC layer), black arrowheads (folds in RGC layer) (L, lens;V, vitreous; R, retina; RGC, retina ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer). H&E pictures were taken at ×4 and ×60 magnification at 24 h and ×10 magnification at 96 h. Saline, n = 3 BALB/c mice and n = 3 C57BL/6 mice; 2,500 CFU of S. aureus, n = 5 BALB/c mice and n = 5 C57BL/6 mice. **, P < 0001.

Endophthalmitis in BALB(gld) mice.

FasL is required for bacterial clearance in C57BL/6 mice (10). BALB/c mice express a polymorphism of FasL that exhibits 9-fold-higher binding affinity to the Fas receptor and cytotoxic activity against Fas⁺ target cells than the FasL expressed in C57BL/6 mice (17). The following series of studies was performed to determine whether the relative resistance to S. aureus-induced endophthalmitis in BALB/c mice is dependent upon Fas ligand. BALB(gld) mice expressing a mutant/nonfunctional form of FasL were challenged with an intravitreal injection of 2,500 CFU of S. aureus. No significant difference was observed in the percentage of eyes destroyed, with 40% destruction in BALB/c and 35% in BALB(gld) mice (Fig. 4A). Moreover, there was no significant difference in bacterial load at 96 h postinfection in BALB/c and BALB(gld) mice (Fig. 4B). Taken together, these data demonstrate that the increased resistance to endophthalmitis in BALB/c mice does not depend on FasL and, further, that FasL is not required for bacterial clearance in BALB/c mice.

Fig 4.

Clinical examination and bacterial quantification of BALB/c and BALB(gld) mice following intravitreal injections of 2,500 CFU of S. aureus. At 24, 48, 72, and 96 h postinoculum, the eyes were examined by slit lamp microscopy. (A) Each eye was given a score of 0, 1, 2, or 3 (0, cleared; 1, mild aqueous flare and vitreal haze; 2, moderate aqueous flare and dense vitreal haze; 3, destroyed). There was no significant difference in percentage of eyes destroyed. At 96 h postinoculum, 40% of the eyes were destroyed (score, 3) in BALB/c and 35% in BALB(gld) mice. BALB/c mice, n = 10; BALB(gld) mice, n = 10. (B) Bacterial quantification revealed no significant difference in bacterial load at 96 h postinoculum between BALB/c mice (n = 15) and BALB(gld) mice (n = 15).

Retinal function in BALB/c and BALB(gld) mice.

Interestingly, while there was no significant difference in bacterial clearance between BALB/c and BALB(gld) mice, there was a significant difference in retinal function retained at 96 h postinfection. In order to assess retinal function in mice that cleared the infection, ERG was performed only on eyes with a clinical score of <3. In mice that successfully cleared the infection, BALB/c mice retained nearly 58% of normal retinal function at 96 h postinfection, compared to only 36% of normal retinal function retained in BALB(gld) mice (Fig. 5A). Histological analysis at 96 h postinfection revealed extensive retinal folding in BALB(gld) retinas compared to BALB/c retinas (Fig. 5C and D). Quantification of retinal folds in histological sections taken from eyes at 96 h postinfection further corroborates the ERG results, demonstrating a significant increase in retinal folding in BALB(gld) compared to BALB/c mice (Fig. 5B). These data imply that while FasL is not required for bacterial clearance, it is critical for protecting host tissue from inflammation-induced damage.

Fig 5.

Retinal histology and ERG assessment of retinal function of BALB/c and BALB(gld) mice following intravitreal injections of 2,500 CFU of S. aureus. (A) The b-wave amplitude was determined, and the ratio of injected eye to contralateral internal control eye was calculated. The BALB/c mice lost 42% of normal retinal function in the infected eye, while BALB(gld) mice lost nearly 64%. (B) At 96 h, representative eyes were enucleated, sectioned, and stained with H&E. Mild inflammation was observed in BALB/c mice, while BALB(gld) mice presented with significant retinal damage. INL, inner nuclear layer; ONL, outer nuclear layer. (C and D) H&E pictures were taken at ×4 magnification and ×20 magnification. BALB/c mice, n = 10; BALB(gld) mice, n = 10. *, P < 0.05; **, P < 0.01.

Neutrophil infiltration in BALB/c and BALB(gld) mice.

Neutrophils are the predominant cell type present in the early innate immune-mediated inflammatory response to ocular infections (8, 9, 24). A delay in neutrophil infiltration will allow the bacteria to multiply, leading to the development of destructive endophthalmitis. On the other hand, a prolonged neutrophil presence can lead to nonspecific host tissue destruction. The constitutive expression of Fas ligand in the eye is thought to protect ocular tissue from the damaging effects of inflammation by inducing apoptosis of Fas⁺-infiltrating inflammatory cells and preventing inflammation (13, 25). Therefore, we performed a myeloperoxidase assay to quantify the neutrophil infiltrate and to determine if the increased retinal damage observed in the BALB(gld) mice was due to a prolonged neutrophil infiltrate (Fig. 6). The myeloperoxidase analysis revealed no significant difference in the number of neutrophils at 24 h postinfection, indicating no delay in neutrophil infiltration in BALB(gld) mice. However, while the number of neutrophils significantly decreased in WT mice at 48 h postinfection, the numbers of neutrophils remained elevated and actually increased in the absence of functional FasL in BALB(gld) mice at 48 h postinfection. This prolonged neutrophil infiltration correlates well with the increased retinal damage observed in BALB(gld) mice (Fig. 5A). These data indicate that in the absence of functional FasL, there is prolonged neutrophil presence that leads to increased nonspecific tissue damage in BALB(gld) mice.

Fig 6.

PMN infiltration into infected eyes of BALB/c and BALB(gld) mice following intravitreal injections of 2,500 CFU of S. aureus. At 24 and 48 h postinoculum, the eyes were enucleated and processed for myeloperoxidase (MPO). Similar MPO activity was observed in BALB/c and BALB(gld) mice at 24 h postinoculum. At 48 h, persistent PMN infiltration was observed in BALB(gld) compared to the significant decrease in PMN infiltration observed in BALB/c mice. BALB/c mice, n = 5; BALB(gld) mice, n = 5. *, P < 0.001.

DISCUSSION

It was demonstrated previously that FasL was required for bacterial clearance from the posterior segment of C57BL/6 mice challenged with S. aureus (10). Herein, we demonstrate that BALB/c mice that express an allotypic polymorphism of FasL resulting in a 9-fold increase in cytotoxic activity are more resistant than C57BL/6 mice to S. aureus-induced endophthalmitis. An intravitreal injection of 2,500 CFU of S. aureus resulted in the destruction of 86% of the eyes in C57BL/6 mice, compared to only 29% destruction in BALB/c mice. To our surprise, though, the increased resistance in BALB/c mice was not dependent upon FasL. BALB(gld) mice were able to clear an S. aureus infection as efficiently as BALB/c mice. However, while the absence of FasL did not impede bacterial clearance in BALB(gld) mice, increased retinal damage and loss of retinal function was observed following bacterial clearance in BALB(gld) mice compared to in BALB/c mice. Moreover, the increased retinal damage in BALB(gld) mice coincided with a prolonged neutrophil infiltration. Taken together, these data demonstrate that in BALB/c mice, FasL is not required for bacterial clearance, but, rather, FasL is critical in resolution of the inflammatory response and protecting the retina from nonspecific damage.

FasL is constitutively expressed on ocular tissue, where it is thought to be critical in the maintenance of immune privilege by inducing apoptosis of Fas⁺-infiltrating inflammatory cells (13, 25). While FasL was first identified as an apoptosis-inducing ligand, more recent studies revealed nonapoptotic functions of FasL, where FasL triggered the Fas receptor to promote cell proliferation, migration, and release of proinflammatory cytokines (26–31). While it is not completely understood how the different pathways are regulated in vivo, it is thought that the Fas receptor engages these pathways in a cell- and tissue-dependent manner. Moreover, Gao and colleagues suggested a link between FasL-mediated apoptosis and immune deviation in the eye that is dependent upon the microenvironment, in particular, the production of Th2 versus Th1 cytokines (32). Promotion of Th2-type cytokines (interleukin 10 [IL-10]) inhibits inflammation and minimizes the inflammatory-induced damage to the ocular tissue. Many studies have documented that C57BL/6 mice favor a Th1 response to infection, while, BALB/c mice favor a Th2-like response, and these responses are characterized by differences in the cytokines produced: Th1-like responses are characterized by the production of IL-2 and gamma interferon (IFN-y), and Th2-like responses are characterized by the production of IL-3, IL-4, IL-5, IL-6, and IL-10 (33–35). Therefore, these strain-specific differences in the microenvironment during an infection may alter FasL function and ultimately susceptibility to S. aureus endophthalmitis.

In 2005, Engelbert and Gilmore demonstrated that FasL was required for the early recruitment of granulocytes that are critical mediators of bacterial clearance in S. aureus endophthalmitis in C57BL/6 mice (10). In C57BL/6(gld) mice (that express a nonfunctional FasL), the inability to clear an S. aureus infection corresponded with a significant decrease in the number of infiltrating granulocytes at 24 h postinfection. This implied that increased susceptibility to infection coincided with decreased neutrophil infiltration, while increased resistance to bacterial infection, as we observed in BALB/c mice, would coincide with a quicker and/or more robust neutrophil infiltration. To our surprise, we observed no difference in the rate or quantity of neutrophil infiltration between the more resistant BALB/c and more susceptible C57BL/6 strains of mice at 24 h postinfection (7.1 ± 0.2 versus 7.4 ± 0.2 relative units of MPO; n = 10/group). In addition, there was no reduction in the number of infiltrating neutrophils in BALB(gld) mice that express a nonfunctional FasL compared to BALB/c mice. Taken together, these data indicate that in contrast to C57BL/6 mice, FasL is not required for the infiltration of neutrophils in BALB/c mice. Moreover, the increased resistance of BALB/c mice to S. aureus infection (only 29% of BALB/c eyes infected with 2,500 CFU of S. aureus were destroyed compared to 86% of the C57BL/6 eyes) is not dependent upon the rate and/or number of neutrophils that infiltrate the eye.

Many studies have documented several differences in the immune response of C57BL/6 and BALB/c mice. Studies comparing the host immune response between resistant and susceptible strains of mice have been performed extensively to dissect the mechanisms of innate and adaptive immunity critical to resistance in several mouse models of infection, including bacterial keratitis (34–38). In regard to differences in immune mechanisms, Hazlett and colleagues established an important role for adaptive immunity in bacterial keratitis, demonstrating that CD4⁺ T cells contribute to the pathogenesis of P. aeruginosa keratitis in C57BL/6 mice (34, 39). Moreover, mouse strains favoring the development of a Th1-type response (C57BL/6) are susceptible, while mice favoring a Th2-type response (BALB/c) are resistant. Interestingly, though, while an important role for adaptive immunity has been established in bacterial keratitis, such a role for adaptive immunity in endophthalmitis has not yet been demonstrated. While it is well documented that the anterior segment displays significantly increased resistance to infection compared to the posterior segment (40, 41), no studies have yet been performed to examine the role of Th1- and Th2-type responses in resistance to infection in the posterior segment. While our data demonstrate increased resistance to S. aureus endophthalmitis in BALB/c mice that favor a Th2-type response, additional studies will be required to fully elucidate the mechanisms critical to resistance to infection in the posterior segment.

Unlike the adaptive immune response, the innate immune response is immediate and has been shown to be critical in bacterial clearance during endophthalmitis (10). Recent studies have also begun to investigate the effect the microenvironment has on the development of the innate immune response, in particular neutrophil activation (42). Therefore, it is possible that while our data show no difference in the rate of neutrophil infiltration between resistant and susceptible mouse strains, the neutrophil function and ability to clear the bacteria may be different between strains of mice. This possibility is supported by a study of resistance and susceptibility to methicillin-resistant S. aureus (MRSA), where two different subsets of neutrophils were identified in mice with different susceptibilities to infection by MRSA (43). One subset of neutrophils produced IL-12 and CCL3 and was associated with resistance to MRSA, while a second subset of neutrophils produced IL-10 and CCL2 and was associated with susceptibility to MRSA (43). In addition, a recent study using a subcutaneous Staphylococcus aureus infection model demonstrates an association between increased resistance with neutrophil infiltration and a Th2-like response (44). However, an alternate possibility is that neutrophils are not the critical factor in the increased resistance to S. aureus endophthalmitis observed in BALB/c mice, but rather another component of innate immunity, such as antimicrobial factors, is the critical factor. Future studies will be focused on further elucidating the immune mechanism(s) responsible for the increased resistance to S. aureus endophthalmitis in BALB/c mice.

In BALB/c mice, the function of FasL appears to be more closely associated with the prevention of nonspecific tissue damage secondary to host-mediated inflammation. This is further supported by the prolonged neutrophil infiltration, increased infiltration into the retina, and increased retinal damage observed in BALB(gld) mice that express a nonfunctional form of FasL. Previous studies by our laboratory and others demonstrated that within the retina, FasL is primarily expressed on retinal microglia, astrocytes, the inner nuclear layer, and the retinal pigment epithelium (45–47). Therefore, we would predict that microglia within the retina ganglion cell (RGC) layer and astrocytes that line the retina are most likely to play a critical role in inducing apoptosis of infiltrating neutrophils, resulting in reduced infiltration of neutrophil into the retina and shutting down the innate-mediated inflammation. A previous study using a corneal keratitis model demonstrates that FasL is required for the clearance of infiltrating neutrophils and resolution of inflammation (48). In the absence of FasL, polymorphonuclear leukocyte (PMNs) persisted, resulting in increased nonspecific corneal damage. Moreover, others demonstrate that FasL.2 (expressed in BALB/c mice) is much more efficient at killing Fas⁺ targets than FasL.1 (expressed in C57BL/6 mice) (17). Taken together, these data suggest that the sharp decline in PMNs we observe at 48 h postinfection in BALB/c mice is due to the expression of FasL, in particular FasL.2, which is more cytotoxic for Fas⁺ targets and thus more efficient at killing Fas⁺ neutrophils. However, additional studies must be performed to specifically identify the FasL⁺ cells critical in inducing apoptosis of Fas⁺ inflammatory cells.

In conclusion, our data demonstrate increased resistance to S. aureus endophthalmitis in BALB/c versus C57BL/6 mice via an FasL-independent mechanism. In addition, the fact that (i) FasL is required for bacterial clearance in C57BL/6 mice but not BALB/c mice and (ii) FasL is critical in protecting the retina from nonspecific damage in BALB/c but not C57BL/6 mice reveals the complexity of the Fas-FasL system and how the function of FasL in ocular innate immunity may be differentially regulated by genetic background. Specifically, differences in genetic background may alter FasL function through changes in the immune response phenotype (Th1 versus Th2), leading to changes in (i) the cytokine milieu, (ii) the form of FasL expressed, and (iii) the susceptibility of Fas⁺ targets to apoptosis. Therefore, the diverse activities of FasL and the complexity of the Fas-FasL system must be taken into account when elucidating the function of FasL in a particular disease model.