Essential Role of NLRP3 Inflammasome in Mediating IL-1β Production and the Pathobiology of Staphylococcus aureus Endophthalmitis

ABSTRACT

Staphylococcal endophthalmitis is one of the leading causes of blindness following ocular surgery and trauma. Dysregulated inflammation during bacterial endophthalmitis causes host-induced inflammatory damage and vision loss if it remains unchecked. Emerging evidence indicates that inflammasome plays a critical role in regulating innate immunity in various infectious and inflammatory diseases. However, the role of the inflammasome in endophthalmitis remains elusive. Here, using a mouse model of Staphylococcus (S) aureus endophthalmitis, we show that NLRP3/ASC/Caspase-1 signaling regulates IL-1β production in endophthalmitis. We also show that S. aureus and its cell wall components and toxins induce the activation of the NLRP3 inflammasome complex in mouse eyes. Moreover, we found that both infiltrating neutrophils and retinal microglia contribute toward NLRP3 activation and IL-1β production in S. aureus-infected eyes. Furthermore, our data using NLRP3^−/− and IL-1β^−/− mice revealed that NLRP3 and IL-1β deficiency leads to increased intraocular bacterial burden and retinal tissue damage. Altogether, our study demonstrated an essential role of NLRP3 inflammasome activation in regulating innate immune responses in bacterial endophthalmitis.

INTRODUCTION

Bacterial endophthalmitis is one of the most devastating ocular complications of eye surgery and trauma. With increasing life expectancy and demographic aging, there is an increased demand for ocular surgeries such as cataracts, age-related macular degeneration (AMD), and glaucoma, in addition to the frequent intravitreal injections of anti-VEGF drugs to treat AMD and diabetic retinopathy (1, 2). These therapeutic interventions have been linked to a steady rise in the incidence of bacterial endophthalmitis. A wide variety of Gram-positive and Gram-negative pathogens has been shown to cause endophthalmitis, where Staphylococcus aureus is the most common causative agent of severe endophthalmitis and accounts for 10% of all endophthalmitis cases (3–6). After entering into the vitreous cavity, S. aureus proliferates rapidly and induces severe intraocular inflammation. Being an immune-privileged site, the retina possesses a strong innate defense mechanism in response to inflammation-mediated tissue damage following S. aureus invasion to reduce the destructive effects caused by pathogen- as well as host-induced inflammatory response (7–12). Over the past several years, our laboratory has investigated and reported pathological mechanisms and regulation of innate responses during S. aureus endophthalmitis (1, 7, 8, 11–16). In addition to our study, various independent studies have confirmed that patients with bacterial endophthalmitis, as well as experimental models, exhibit increased levels of inflammatory cytokines, including IL-1β (1, 7, 8, 10–12, 17–20).

To mount an appropriate reparative response to any insult or infection, the host cell triggers innate immune responses through various cellular receptors. Toll-like receptors (TLRs) are an extensively studied member of the pattern recognition receptors (PRRs) family, which contain an extracellular domain to recognize bacterial components like lipoteichoic acid, lipoproteins, LPS, flagellin, and unmethylated CpG DNA (7, 10, 11). Nod-like receptors (NLRs) are another large family of intracellular PRRs comprising 22 and 34 distinct members in humans and mice, respectively (21–23). Among the NLRs family, the nucleotide-binding domain and leucine-rich repeat-containing family, pyrin containing 3 (NLRP3) is the most thoroughly characterized. It is an important component of the innate immune system that is known to assemble various proteins such as apoptosis-associated speck-like proteins (ASC) and caspase-1 to form a complex called inflammasome (23–25). A variety of pathogens and other noxious stimuli have been demonstrated to activate the inflammasomes resulting in the activation and splicing of pro-caspase-1 into an active caspase-1 (caspase p10 and p20) and subsequent proteolytic cleavage of pro-IL-1β into the active IL-1β, which orchestrates the innate immune response (24, 26–28).

Both Gram-positive and Gram-negative bacterial species have been shown to activate the NLRP3 inflammasome and induce IL-1β production to mount an innate immune response (29, 30). Innate immune cells such as macrophages, neutrophils, and microglial cells, have been shown to process cytokines via the NLRP3 inflammasome in response to bacterial toxins and S. aureus (31–35). More recently, the expression of NLRP3 and subsequent activation of IL-1β following S. aureus infection in human and rat conjunctivital goblet cells has been demonstrated (36). Over the past few years, we have demonstrated the involvement of retinal residential cells such as microglia and Müller glial cells in pathogen recognition and initiation of the innate immune response by producing inflammatory cytokines, and antimicrobial peptides (8, 12). We showed that bacterial cell wall components and live S. aureus induce IL-1β secretion in mouse eyes (1, 6, 13). Although the role of NLRP3 in IL-1β production has been well studied, its role in bacterial endophthalmitis has not been investigated. In this study, we demonstrated the essential role of NLRP3 inflammasome in S. aureus-induced IL-1β production and in the innate defense using IL-1β-dSRed, NLRP3^−/−, and IL-1β^−/− mouse models of endophthalmitis.

RESULTS

S. aureus infection induces the production of IL-1β in the eye.

Bacterial infections trigger the activation of the innate immune system resulting in the production of inflammatory chemokines and cytokines. Previously, we have reported the production of multiple chemokines and cytokines in bacterial endophthalmitis (1, 2, 8, 13). Among S. aureus-induced intraocular cytokines, IL-1β is one of the predominantly produced cytokines. Here in this study, we investigated the molecular mechanism of IL-1β production in S. aureus endophthalmitis. To gain a better understanding of the kinetics of IL-1β production, we performed a time course study by infecting mice eyes with S. aureus. Our results show that S. aureus induced the mRNA expression of IL-1β as early as 6 h postinfection with increasing levels at 12 and 24 h (Fig. 1A). The production of IL-1β at protein levels followed a similar pattern with increased accumulation at 12, 24, and 48 h postinfection (Fig. 1B). To further confirm these findings, we used DsRed transgenic mice, where the DsRed gene is under the control of the IL-1β promoter, allowing the visualization of IL-1β production by red fluorescence (37, 38). The histological analysis revealed that S. aureus infection induces time-dependent expression of DsRed, mostly in infiltrating cells, indicating the in vivo production of IL-1β during endophthalmitis (Fig. 1C).

FIG 1.

S. aureus infection induces IL-1β production in the mouse retina. C57BL/6 and DsRed mouse (n = 6) eyes were intravitreally injected with PBS (control, C) or 5,000 CFU of S. aureus (SA), strain RN6390. At indicated time point postinfection, eyes were enucleated, and WT mouse retinal tissue was subjected to qPCR (A) and ELISA (B) for IL-1β expression. The data are expressed as relative fold change by normalizing the expression of genes with respect to control. The retinal cryosections from DsRed mouse were imaged for red fluorescent protein expression (C). Results are representative of at least two independent experiments. Statistical analysis was performed using one-way ANOVA. ***, P < 0.0001. Data are shown as the mean ± SD.

S. aureus toxins and cell wall components induce NLRP3 inflammasome in the retina.

IL-1β is produced in an immature form by transcriptional regulation that is cleaved and activated by a proteolytic enzyme, caspase-1, following inflammasome activation. Previously, our transcriptomic analysis indicated the activation of inflammasomes in pathways that regulate inflammation in endophthalmitis (39). Similarly, our recent RNAseq analysis of S. aureus-infected mouse retina showed the differential dysregulation of inflammasome-related genes (Fig. 2A). As NLRP3 is one of the best-studied inflammasomes in regulating IL-1β, we sought to determine its role in our model. Our data show that S. aureus significantly induced the expression of NLRP3, ASC, and caspase-1 mRNA transcripts in a time-dependent manner (Fig. 2B). We further confirmed the induction of NLRP3 and ASC at the protein levels in the S. aureus-infected mouse retina (Fig. 2C and D). To visualize the colocalization of the NLRP3 inflammasome with IL-1β expression, we used DsRed mice and observed the red fluorescent protein expression along with NLRP3 in infiltrating cells (Fig. 2E). NLRP3 expression was also detected in other residential retinal cell types.

FIG 2.

S. aureus infection induces the activation of the NLRP3/ASC/Caspase-1 pathway in the mouse retina. C57BL/6 mouse (n = 6) eyes were intravitreally injected either with 5,000 CFU of S. aureus (SA), strain RN6390, or PBS (control, C). At indicated time point postinfection, eyes were enucleated, mouse retinal tissue was subjected to RNAseq, and a heat-map was generated for genes modulating inflammasome pathways (A). The expression of NLRP3, ASC, and Caspase-1 was confirmed by qPCR (B). The protein levels of NLRP3 and ASC were detected by immunoblotting (C). Densitometry analysis was performed using ImageJ (D). The data are expressed as relative fold change by normalizing the expression of genes/proteins with respect to housekeeping control (GAPDH/β-actin respectively). The retinal cryosections from DsRed mouse were immunostained for NLRP3 (green) and imaged with red fluorescent protein (IL-1β-red); a few representative cells with NLRP3 and DsRed colocalization are indicated with white arrows, and some cells only expressing NLRP3 are indicated by cyan arrows (E). R, retina; ONH, optic nerve head; VC, vitreous chamber. Results are representative of at least two independent experiments. Statistical analysis was performed using one-way ANOVA. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. Data are shown as the mean ± SD.

Since S. aureus cell wall components and toxins contribute toward the pathogenesis of endophthalmitis (13, 14), we assessed their role in NLRP3 inflammasome activation. Our results show bacterial culture media supernatant (referred to as media), heat-killed S. aureus (HKSA), alpha-toxin, and Staphylococcal peptidoglycan (PGN) and lipoteichoic acid (LTA) induced the expression of NLRP3, ASC, caspase-1, and IL-1β mRNA transcripts in mouse retina (Fig. 3), similar to live S. aureus, indicating their ability to differentially regulate NLRP3 inflammasome activation.

FIG 3.

S. aureus cell wall components and toxins differentially induce NLRP3 signaling in the mouse retina. C57BL/6 mouse (n = 6) eyes were intravitreally injected with either PBS (control, C) or PGN (100 ng/eye), LTA (100 ng/eye), alpha-toxin (100 ng/eye), overnight SA grown culture media (referred as media) (2 μL/eye), and HKSA (5000CFU/eye) for 24h. Following the challenge, the mouse retina was subjected to qPCR for NLRP3, Caspase-1, and IL-1β mRNA transcripts. The data are expressed as relative fold change by normalizing the expression of genes with respect to housekeeping control (GAPDH). Results are representative of at least two independent experiments. Statistical analysis was performed using one-way ANOVA. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. Data are shown as the mean ± SD.

NLRP3/Caspase-1 mediates IL-1β production during S. aureus endophthalmitis.

Since IL-1β is activated by proteolytic cleavage by caspase-1 following inflammasome activation, we sought to determine the role of NLRP3 and caspase-1 in bacterial endophthalmitis. To start, we utilized a pharmacological inhibition approach via a lysosomal acidification inhibitor, bafilomycin, and the pan-caspase inhibitor Q-VD-OPH, as these are known to inhibit inflammasome activation. Our results show that bafilomycin reduced NLRP3 whereas both bafilomycin and Q-VD-OPH attenuated caspase-1 mRNA expression in mouse retinal tissue (Fig. 4A). Similarly, bafilomycin and Q-VD-OPH treated eyes showed a significant reduction in S. aureus-induced IL-1β levels (Fig. 4B). The inhibitory effects of bafilomycin on NLRP3 and ASC expression were confirmed at protein levels (Fig. 4C and D). This coincided with reduced cleavage of caspase-1 and active IL-1β levels in eyes treated with bafilomycin or Q-VD-OPH (Fig. 4C, D).

FIG 4.

S. aureus-induced IL-1β secretion is NLRP3/Caspase-1 dependent. C57BL/6 mouse (n = 6) eyes were intravitreally injected with bafilomycin A1 (Bafilo, 200 nM) or pan-caspase inhibitors Q-VD-OPH (QVD, 100 nM) 24 h before induction of endophthalmitis. PBS injected eyes were used as a control (C). At indicated time point postinfection, eyes were enucleated, and retinal tissue was subjected to qPCR for NLRP3 and caspase-1 mRNA expression (A). The physiological consequence of NLRP3 and caspase-1 inhibition was measured by measuring the production of IL-1β by ELISA (B). The expression of NLRP3, ASC, caspase-1(p20), and an active form of IL-1β was assessed by immunoblotting (C). Densitometry for immunoblots was performed using ImageJ (D). The data are expressed as relative fold change by normalizing the expression of genes/proteins with respect to housekeeping control (GAPDH/β-actin respectively). Results are representative of at least two independent experiments. Statistical analysis was performed using one-way ANOVA. **, P < 0.001; ***, P < 0.0001; ns, not significant. Data are shown as the mean ± SD.

NLRP3 and IL-1β deficiency exacerbate S. aureus endophthalmitis.

To further investigate the role of NLRP3 and IL-1β in bacterial endophthalmitis, we used NLRP3^−/− and IL-1β^−/− mice. Our results show that NLRP3 and IL-1β deficiency lead to increased intraocular bacterial burden compared to WT mice (Fig. 5A). We also observed a significant reduction in IL-1β cytokine production in both NLRP3^−/− and IL-1β^−/− mice (Fig. 5B). Since IL-1β levels were significantly decreased in both of these knockout mice, we conclude that NLRP3 is required to produce IL-1β during ocular bacterial infection. The histopathological analysis (Fig. 5C) revealed that both NLRP3 and IL-1β KO mouse eyes had increased retinal tissue damage, quantified by retinal folding (Fig. 5D). The gross pathological changes also show a slight increase in cellular infiltration and pus formation within the vitreous cavity of these KO mice (Fig. 5C). Collectively, these results indicate that the lack of NLRP3 and IL-1β is detrimental during bacterial endophthalmitis.

FIG 5.

NLRP3 and IL-1β deficiency increases bacterial burden and retinal tissue damage in bacterial endophthalmitis. C57BL/6, NLRP3^−/−, and IL-1β^−/− mice (both on B6 background) (n = 6) eyes were injected intravitreally either with 5,000 CFU of S. aureus (SA), strain RN6390, or PBS (control, C). Twenty-four hours postinfection, eye lysates were subjected to intraocular bacterial burden estimation by plate count method (A) and ELISA for measurements of IL-1β levels (B). From a different set of experiments, the eyes were enucleated and subjected to histopathological analysis by H&E staining (C). The blue rectangle represents the higher magnification image of the selected tissue area. Red arrows indicate retinal detachments and folds; asterisks (*) represent cellular infiltration in the vitreous cavity. The retinal folds per retina were counted and presented as a bar graph (D). R, retina; ONH, optic nerve head; VC, vitreous chamber; L, lens; C, cornea. Results are representative of at least two independent experiments. Statistical analysis was performed using one-way ANOVA. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. Data are shown as the mean ± SD.

Infiltrating PMNs are the predominant source of IL-1β in bacterial endophthalmitis.

Since we observed increased intraocular levels of IL-1β in experimental models with similar observation in human bacterial endophthalmitis (40), we sought to determine its cellular source. Previously, we have shown that polymorphonuclear leukocytes (PMNs) are one of the predominant infiltrating immune cells and contribute to the pathobiology of bacterial endophthalmitis (8). To determine the role of PMNs as potential source of IL-1β, we made DsRed mice neutropenic and compared it with their wild-type counterparts following S. aureus infection. Our results show that only WT DsRed mice exhibited the induced expression of red fluorescent protein following S. aureus infection, whereas neutropenic mice lack this response (Fig. 6A). To further confirm this finding, we performed flow cytometry analysis, and found an increased population of Ly6G⁺DsRed⁺ cells, indicating PMNs (Fig. 6B). These results indicate infiltrated PMNs are producers of IL-1β in the eye during endophthalmitis.

FIG 6.

S. aureus induced NLRP3 activation and IL-1β expression in infiltrating PMNs during endophthalmitis. WT and neutropenic (PMN^−/−) DsRed mouse (n = 6) eyes were injected intravitreally either with 5,000 CFU of S. aureus (SA), strain RN6390, or PBS (control, C). Twenty-four hours postinfection, eyes were enucleated and immunostained with anti-CD45 antibodies (green) and imaged with DsRed fluorescence (red). White arrows indicate a few representative cells showing colocalization of CD45 with DsRed (A). In another set of experiments, retinal single-cell suspension was prepared and stained for anti-Ly6G antibody. Flow cytometric analysis was performed for DsRed-Ly6G double-positive cells (B).

Activated retinal microglia show NLRP3 induction and IL-1β production.

Among retinal residential cells, microglia and Müller glia were shown to initiate innate immune responses in bacterial endophthalmitis (6, 12). We demonstrated that these cells express TLRs and produce various cytokines in response to pathological challenges (6, 12, 14). To investigate the role of retinal residential cells in NLRP3 activation and IL-1β production, first we confirmed the expression of NLRP3 and DsRed in S. aureus infected retinal flat-mount cryosections. Our results show the colocalization of microglial marker Iba-1 with both NLRP3 and DsRed, which indicates the activation of NLRP3 inflammasome in retinal microglia during bacterial endophthalmitis (Fig. 7A and B). To further investigate the role of microglia in S. aureus-mediated inflammasome induction and IL-1β secretion, we used the mouse BV2 microglia cell line. Our data show that S. aureus challenge significantly induced the NLRP3 and IL-1β mRNA expression in BV2 cells (Fig. 8A). The BV2 cells also exhibited increased IL-1β secretion in culture supernatant following S. aureus infection (Fig. 8B). Our immunostaining data also revealed the induction of NLRP3, caspase-1, and IL-1β in S. aureus-infected BV2 cells (Fig. 8C). Thus, both in vivo and in vitro data indicate the NLRP3 activation and IL-1β production in microglia during S. aureus endophthalmitis.

FIG 7.

Retinal microglia exhibit S. aureus-induced NLRP3 and IL-1β expression. C57BL/6 (n = 6) mouse eyes were injected intravitreally with 5,000 CFU of S. aureus (SA), strain RN6390. Twenty-four hours postinfection, eyes were enucleated and retinal flat-mount (A), as well as cryosections (B), were immunostained for anti-NLRP3 and anti-IL-1β antibodies along with the microglial marker (Iba1).

FIG 8.

S. aureus challenge induced NLRP3 signaling in cultured mouse microglia. Mouse microglial cell line BV2 (1 × 10⁶/mL in a 35 mm dish, n = 3/condition) was challenged with S. aureus (SA) at MOI 10:1 for indicated time points. Uninfected cells were used as control. Following incubation, the cells were subjected to RNA isolation and qPCR for measuring IL-1β and NLRP3 mRNA transcripts (A). The data are expressed as relative fold change by normalizing the expression of genes with respect to housekeeping control (GAPDH). The conditioned media were subjected to quantification of secreted IL-1β by ELISA (B). In another set of experiments, the infected and uninfected control cells were subjected to immunostaining for anti-NLRP3, anti-caspase-1, and anti-IL-1β antibodies (C). Statistical analysis was performed using one-way ANOVA. *, P < 0.05; **, P < 0.001; ***, P < 0.0001. Data are shown as the mean ± SD.

DISCUSSION

Pathogenic microorganisms employ various strategies to invade and establish an infection inside the host; therefore, the immune system exhibits diverse mechanisms to sense and eradicate them. Different pathogen-associated molecular patterns (PAMPs) can be detected by the immune system to elicit protective responses (41, 42). Infectious agents, such as bacteria and viruses, are recognized by PRRs such as TLRs, NLRs, and RIG-I-like receptors resulting in the activation of innate responses. Over the last decade, our laboratory has been investigating inflammatory signaling during intraocular infections, with an emphasis on the TLR2-mediated innate response in S. aureus endophthalmitis (3, 8, 15, 43). In contrast, the role of non-TLR mediated innate immune signaling in this disease remains elusive. However, our prior retinal transcriptomics (39) and RNAseq data have shed light on the non-TLR mediated innate immune response, with an upregulation of multiple inflammasome-related genes. Therefore, in this study, we sought to determine the role of NLRP3 inflammasome in bacterial endophthalmitis. Our study demonstrates NLRP3 and IL-1β deficiencies are detrimental for the eye and indicate their protective roles in bacterial endophthalmitis.

Secretion of IL-1β in bacterial endophthalmitis has been reported in patients (40) and animal models by both our laboratory as well as others (1, 2, 5, 8, 17, 44). Unlike other cytokines, IL-1β secretion needs two signals, a PAMP-mediated TLR activation to induce pro-IL-1β and a danger signal for the assembly and activation of the NLRP3 inflammasome (45). Our data show an induced expression of IL-1β and NLRP3 by heat-killed S. aureus, PGN, and LTA, indicating that these bacterial cell wall components trigger the priming signal for inflammasome activation. We previously reported that inside the vitreous chamber, S. aureus is recognized by TLR2, which are abundantly expressed by multiple retinal cell types, including glial and photoreceptor cells (8, 11, 15). Thus, we postulate that during the early phase of infection, Staphylococcal cell wall components activate TLR2 signaling to provide the initial stimulus for inflammasome activation. However, as the bacterium proliferates and reaches the stationary growth phase, it starts secreting toxins (46). These bacterial toxins (e.g., alpha-toxin) directly damage the retinal tissue (13), but also provide a secondary stimulus to activate NLRP3 and Caspase-1 mediated cleavage of pro-IL-1β. S. aureus toxins, or lipoproteins alone, do not activate NLRP3 (47); however, a recent study revealed bacterial extracellular vesicles (EVs) containing both first (toxins) and second (lipoproteins) stimuli trigger NLRP3 signaling in macrophages (48). S. aureus strains deficient in toxin production show reduced inflammation during endophthalmitis (49, 50). In conjunction, a recent study has shown the efficacy of nanosponges in neutralizing the effects of toxins and improving the disease outcome in bacterial endophthalmitis (51). Together, our data and prior studies (32–35, 52–54) indicate that Staphylococcal toxins trigger an inflammatory response, in part, by activating the NLRP3 inflammasome.

Given the emerging role of inflammasomes in various infectious and inflammatory diseases, they have been implicated in the pathobiology of ocular conditions, such as sterile corneal inflammation (55), corneal wound healing (56), dry eye disease (57), S. pneumoniae (58), and HSV keratitis (59). However, the role of the inflammasome in the pathogenesis of endophthalmitis is not known. Our interest in investigating the inflammasome role emerged from our transcriptomics data showing the induced expression of NLRP3, and the activation of inflammasomes are known to induce inflammation, a hallmark of endophthalmitis. We found that S. aureus induced NLRP3 expression, and both lysosomal acidification inhibitor (bafilomycin A1) and a pan-caspase inhibitor (Q-VD-OPH) attenuated IL-1β secretion, indicating the activation of the classical NLRP3-Casp1-IL-1β signaling pathway (28). Due to an increase in inflammation during endophthalmitis, retinal tissue damage is inevitable (16) and we have reported the protective effects of several anti-inflammatory molecules (1, 3, 15, 60). Based on these studies, we hypothesized that the lack of NLRP3 or IL-1β could be protective in ocular infections. In contrast, we observed that NLRP3^−/− and IL-1β^−/− mouse eyes had a relatively higher bacterial burden and exhibited more tissue damage (i.e., retinal folding, tissue infiltrates, and pus exudates). Thus, the complete deficiency of the NLRP3-IL-1β axis is detrimental during infections and indicates the protective role of inflammasome activation in bacterial endophthalmitis. This is consistent with the notion that an early inflammatory response is necessary to protect the host (45, 48).

Our study shows NLRP3 is essential in evoking an effective innate immune response to limit S. aureus growth inside the eye. Uncontrolled inflammasome activation, however, may result in an excessive inflammation leading to retinal tissue damage in endophthalmitis (46). Therefore, a fine balance is required in the initiation and the termination of inflammatory responses (1, 3, 61). One of the potential mechanisms of inflammasome-mediated tissue damage is pyroptosis, a type of programmed cell death characterized by gasdermin D (GSDM)-mediated membrane rupture, resulting in the release of cytosolic contents into the extracellular spaces. Upon inflammasome activation, caspase-1 activates GSDM, which subsequently forms pores on the cell membrane. These pores facilitate the secretion of IL-1β and IL-18 (61). Our RNAseq data showing induced expression of GSDM indicates the potential involvement of pyroptotic cell death in endophthalmitis. Pyroptotic cell death has also been implicated in CMV retinitis (62, 63). Although our current study focused on elucidating the role of NLRP-3/IL-1β activation, we observed induced expression of IL-18 in S. aureus-infected eyes, and Q-VD-OPH treatment attenuated IL-18 expression (data not shown). As our RNAseq data showed induced expression of IL-18R, further studies are warranted to elucidate the functional role of IL-18 in our disease model. Similarly, in addition to NLRP3, S. aureus infected retina exhibited induced expression of NLRP-12, NLRP1-1a, and NLRC5, indicating their involvement in the pathogenesis of endophthalmitis. Although our data did not show induced expression of AIM2, its role has been implicated in activating inflammasome via sensing bacterial DNA during infections (64), including those caused by S. aureus (65–67). While the focus of this study is on NLRP3, our data indicates several other inflammasomes may also be involved in orchestrating retinal innate responses during bacterial endophthalmitis.

NLRP3 is ubiquitously expressed in the retina; additionally, IL-1β can be produced by various retinal cell types including residential and infiltrating immune cells. In an attempt to identify the sources of IL-1β during endophthalmitis, we used DsRed IL-1β reporter mice. Our immunostaining and flow cytometry analysis revealed neutrophils as the major cell type producing IL-1β. As we reported earlier, neutropenic mice are more susceptible to endophthalmitis (8, 68); we postulate that the lack of IL-1β production in NLRP3 KO mouse neutrophils may be responsible for impaired bacterial clearance and increased tissue damage. In the brain, microglia and astrocytes express PRRs and inflammasome has been shown to play an important role in the pathogenesis of neuroinflammatory and neurodegenerative diseases (69, 70). We also reported the activation of microglia in endophthalmitis (6). Herein, NLRP3 expression is induced in retinal microglia, and it co-localizes with IL-1β in S. aureus-infected mouse retina. Moreover, these findings were confirmed in cultured mouse microglial cells challenged with S. aureus showing induced expression of NLRP3 and Casp-1, and IL-1β production. Here, we demonstrated how the NLRP3 inflammasome plays an essential role in orchestrating retinal innate immune response in endophthalmitis by regulating IL-1β production Moreover, using the IL-1β reporter mice, we observed that infiltrating PMNs are the cellular source of IL-1β production and NLRP3 activation.

In conclusion, we show that S. aureus induces IL-1β secretion in the retina through inflammasome activation, which plays a protective role in bacterial endophthalmitis. Both retinal residential microglia and infiltrating PMNs are the cellular source of NLRP3 mediated IL-1β production in bacterial endophthalmitis. Collectively, our results indicate that the NLRP3 inflammasome plays an essential role in orchestrating retinal innate immune response in endophthalmitis by regulating IL-1β production. The mechanistic insights into the regulation of NLRP3 and IL-1β during endophthalmitis could aid in the development of immunomodulatory therapeutic moieties for the treatment of intraocular infections.

MATERIALS AND METHODS

Animals and ethics statement.

C57BL/6 wild-type (WT) female mice (6–8 weeks of age) were obtained from the Jackson Laboratory (Bar Harbor, ME). NLRP3^−/− mice were provided by Dr. Gabriel Nuñez (University of Michigan, Ann Arbor, MI), and IL-1β^−/− mice were procured from Jackon Labs (C57BL/6J-Il1b^em2Lutzy/Mmjax, stock# 068082-JAX). DsRed mice were received from Dr. Akira Takashima (University of Toledo, Toledo, OH). These mice were bred in-house and maintained in a pathogen-free, restricted-access Division of Laboratory Animal Resources (DLAR) facility at Kresge Eye Institute. All animals were maintained on a 12-h light 12-h dark cycle and provided LabDiet rodent chow (PicoLab; LabDiet, St. Louis, MO) and free access to tap water. All the procedures were conducted in compliance with the Association for Research in Vision and Ophthalmology (ARVO) statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee (IACUC) of Wayne State University.

Induction of S. aureus endophthalmitis.

Endophthalmitis was induced in mice by intravitreal injection of S. aureus strain RN6390 (5000 CFU/eye) using a 34G needle attached to a 10 μL syringe as described previously (1–3). In the treatment groups, mice were intravitreally injected with Bafilomycin A1 (200 nM/eye) and pan-caspase inhibitor Q-VD-OPH (100 nM/eye; Sigma-Aldrich, St. Louis, MO) diluted in sterile phosphate-buffered saline (PBS). Contralateral eyes injected with sterile PBS served as control.

Bacterial strain and cell culture.

The S. aureus strain RN6390 was grown on Tryptic Soy Agar and Broth (TSA/TSB), and the bacterial count was adjusted to desired CFU by centrifugation followed by dilution in PBS. As reported previously, for the preparation of heat-killed S. aureus (HKSA), bacterial culture suspension (10⁵ CFU/mL) was kept in a 100-degree water bath for 5–10 min. Any possible cell viability was then verified by bacterial plating (13). Similarly, for cell-free media supernatant, the culture media from S. aureus that was grown overnight was centrifuged at 5,000 × g for 5 min. The resulting clear culture supernatant was then passed through a 0.22 μm syringe filter to remove any possible live bacteria. The S. aureus cell wall components, PGN (cat # tlrl-pgns2) and LTA (tlrl-pslta), were purchased from InvivoGen (InvivoGen, San Diego, CA), and alpha-toxin (H9395-1MG) was procured from Sigma (Sigma-Aldrich, St. Louis, MO).

For in vitro studies, an immortalized mouse microglial cell line, BV2, was maintained in low-glucose Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 5% FBS and 100U/mL penicillin-streptomycin cocktail (Invitrogen, Carlsbad, CA) in a humidified 5% CO₂ incubator at 37°C. Before experiments, cells were adapted to antibiotic-free and serum-free DMEM for 18 h (growth factor starvation).

Neutrophil depletion.

Mice were made neutropenic by intraperitoneal injection of Anti-Ly6G-1A8 antibody (200 μg/100 μL/mice) (R&D Systems, Minneapolis, MN, USA) as described previously (68, 71). Following neutropenia, endophthalmitis was induced in these mice as indicated above.

RNA extraction and quantitative real-time PCR (qPCR).

Total RNA was extracted from mouse neural retina and BV2 cells using TRIzol per manufacturer’s instructions (Invitrogen, Carlsbad, CA). Total RNA was reverse transcribed using Maxima first-strand cDNA synthesis kit per manufacturer’s protocol (Thermo Scientific, Rockford, IL). cDNA was subjected to qPCR for a gene of interest using mouse-specific TaqMan primers and probes set (Integrated DNA Technologies, Coralville, IA). The data were analyzed using the comparative ΔΔC_T method. The target gene expression was normalized using endogenous GAPDH gene controls.

Enzyme-linked immunosorbent assay (ELISA).

ELISA was performed to quantify the levels of cytokines in mice ocular tissue as well as the BV2 cell-conditioned media. Mouse eyes were enucleated and homogenized in sterile PBS as described previously (2, 3, 15). Total protein was estimated using the Micro BCA protein estimation kit (Thermo Scientific, Rockford, IL, USA) per the manufacturer’s instructions. ELISA was performed for interleukin-1-beta (IL-1β) per the manufacturer’s protocol (R&D Systems, Minneapolis, MN).

Immunoblotting.

For immunoblotting, mice neural retinas were lysed by sonication in PBS containing a protease and phosphatase inhibitor cocktail (Thermo Scientific, Rockford, IL). For in vitro experiments, cells were lysed using radioimmunoprecipitation (RIPA) lysis buffer containing a protease and phosphatase inhibitor cocktail. Total protein concentration was estimated, and 30–40 μg protein was used for blotting. Denatured proteins were resolved in a 12–16% SDS-polyacrylamide gel and transferred onto a nitrocellulose membrane (0.2–0.4 μm) (Bio-Rad Laboratories). Following blocking, the blots were incubated with anti-NLRP3 (1:1,000), anti-ASC (1:1,000), anti-cleaved caspase-1 (Asp297) (1:1,000), anti-cleaved-IL-1β (Asp116) (1:1,000) (Cell signaling Technology, Boston, MA), and anti-β-actin (1:5,000) (Sigma-Aldrich, St. Louis, MO) antibodies overnight at 4°C. Following washing, blots were incubated with goat anti-rabbit/anti-mice-IgG-HRP conjugate (Bio-Rad, Hercules, CA). Protein bands were developed using SuperSignal West Femto Chemiluminescent Substrate and visualized using iBright FL1500 Imaging Systems (ThermoFisher Scientific, Rockford, IL). β-actin was used as a control for protein loading. Densitometry for Western blot images was performed using ImageJ software (Rasband, W.S., ImageJ, NIH, Bethesda, Maryland, http://rsb.info.nih.gov/ij/, 1997–2009).

Histopathology and immunofluorescence staining.

Following euthanasia, eyes were enucleated and fixed in 4% paraformaldehyde (PFA) for histopathological examination. The embedding, sectioning, and hematoxylin and eosin (H&E) staining was performed by Excalibur Pathology, Inc. (Oklahoma City, OK, USA), and Pathscan Enabler IV (Meyer Instruments, Inc.) was used to scan H&E stained slides.

Immunostaining for retinal cryosections as well as flat mounts were performed as described previously (16, 72). For in vivo experiments, eyes were enucleated and fixed in 4% PFA, dehydrated, and embedded in paraffin. For immunostaining, retinal cryosections were fixed in 4% PFA for 20 min at RT followed by washes with PBS (4 × 10 min each). Eye sections were permeabilized and blocked with 10% normal goat serum with 0.5% Triton X-100 for 2 h at RT and incubated overnight with primary antibody (1:100). The next day, sections were washed with PBS (4 × 10 min each) and incubated with anti-mouse/rabbit Alexa Fluor 485/594-conjugated secondary antibody (1:200) for 2h at RT. The eye sections were extensively washed with PBS (4 × 10 min each), and the slides were mounted in Vectashield anti-fade mounting medium (Vector Laboratories, Burlingame, CA) and visualized using Eclipse 90i fluorescence microscope (Nikon, Melville, NY).

For in vitro experiments, BV2 cells were cultured in four-well chamber slides (Fisher Scientific, Rochester, NY). Following stimulation, cells were washed three times with PBS and fixed in 4% paraformaldehyde for 15 min. The fixed cells were permeabilized and blocked simultaneously in 1% (wt/vol) BSA containing 0.4% Triton X-100 for 1h at room temperature followed by incubation with primary antibodies (1:100 dilution) overnight at 4°C. Cells were washed with PBS and incubated with specific fluorescein isothiocyanate (FITC)-conjugated secondary antibodies (1:200 dilutions) for 1 h at room temperature. Following incubation, cells were washed with PBS and mounted in Vectashield anti-fade mounting medium with DAPI (Vector Laboratories). Slides were visualized using an Eclipse 90i fluorescence microscope (Nikon, Melville, NY).

Statistical analysis.

Statistical analyses were performed using GraphPad Prism V8.1.2 (Graph Pad, San Diego, CA). All data were expressed as mean ± SD unless indicated otherwise. Statistical significance was determined using either two-way ANOVA or the unpaired t tests as indicated in the figure legends. A P < 0.05 was considered statistically significant.