The inflammasome NLRP3 plays a protective role against a viral immunopathological lesion

NLRP3 inflammasome in SK pathogenesis.

Abstract

Herpes simplex 1 infection of the eye can cause blindness with lesions in the corneal stroma largely attributable to inflammatory events that include components of both adaptive and innate immunity. Several innate immune responses are triggered by herpes simplex 1, but it is unclear how such innate events relate to the subsequent development of stromal keratitis. In this study, we compared the outcome of herpes simplex 1 ocular infection in mice unable to express NLRP3 because of gene knockout (NLRP3^−/−) to that of wild-type mice. The NLRP3^−/− mice developed more-severe and earlier stromal keratitis lesions and had higher angiogenesis scores than did infected wild-type animals. In addition, NLRP3^−/− mice generated an increased early immune response with heightened chemokines and cytokines, including interleukin-1β and interleukin-18, and elevated recruitment of neutrophils. Increased numbers of CD4⁺ T cells were seen at later stages of the disease in NLRP3^−/− animals. Reduction in neutrophils prevented early onset of the disease in NLRP3^−/− animals and lowered levels of bioactive interleukin-1β but did not lower bioactive interleukin-18. In conclusion, our results indicate that NLRP3 has a regulatory and beneficial role in herpetic stromal keratitis pathogenesis.

Introduction

The outcome of a viral infection depends on many host factors as well as properties of the virus itself. In some instances, damage to tissues is largely the consequence of immunoinflammatory events set off by an infection. An example is herpetic SK, which is characterized by a blinding ocular lesion [1, 2]. Studies in animal models indicate that SK represents an immune inflammatory event orchestrated mainly by CD4⁺ T cells, but the early events induced by the infection that result in the development of overt SK are not fully understood [1–3]. Prominent events of the early response to HSV ocular infection include invasion by innate immune cells, expression of proinflammatory cytokines, particularly IL-1β and IL-6, as well as the involvement of several chemokines, angiogenic factors, and neuropeptides [4–8]. Control of early events is considered important because their modulation represents a potentially useful approach for therapy. One unresolved issue is how the infection triggers innate recognition events. Although a critical role for TLR activation by the virus has been demonstrated [9–11], other mechanisms are likely involved and have not been fully investigated.

One likely mechanism, so far, poorly explored, is that the infection may serve to activate one or more members of the inflammasome family. There is evidence that some viruses activate inflammasomes, such as NLRP3, in immunologic cells that respond to the infection, such as macrophages and dendritic cells [12–16]. The activation of NLRP3 occurs through a wide range of molecules, including self-derived, environmentally-derived, and pathogen-derived activators. Therefore, as various stimuli with divergent structures and biochemical properties activate the NLRP3 inflammasome, a common cellular event may be elicited by the different stimuli and serves as the activating signal for the NLRP3 inflammasome [17–24]. This results in the formation of the biochemical complex known as the NLRP3 inflammasome, which recruits and activates procaspase 1, which acts to process IL-1β and IL-18 into their bioactive forms and concurrently acts to initiate the process of pyroptosis [25, 26].This situation has been advocated for influenza virus infection, where the viral M2 ion channel induces an H⁺ imbalance in infected cells resulting in NLRP3 inflammasome activation [27]. In addition, adenovirus activates NLRP3 through the disruption of lysosomal membranes and the release of cathepsin B into the cytoplasm [28, 29]. For HSV, in vitro results have shown that infection of human foreskin fibroblasts results in the activation of IFI16/204 and NLRP3 inflammasomes [30], but it is not clear whether similar events occur in vivo. Many groups have shown that mice lacking NLRP3 have defective inflammatory responses to a range of situations, which include autoimmunity, metabolic diseases, and immunoinflammatory lesions [24, 31–34]. With the exception of influenza, few studies have evaluated the role of NLRP3 during in vivo responses to viral infections. Moreover, with influenza, some groups showed that, compared with WT animals, NLRP3^−/−mice were more susceptible [19, 35], but others have reported contradictory results [36].

It is not known whether NLRP3 is activated and serves to influence the outcome of infection with HSV. We investigated this issue in an SK model system in which lesions were largely the consequence of an immunoinflammatory process [19, 29]. In this study, we compared the responses to HSV ocular infection in mice that lack expression of NLRP3 because of gene knockout with WT animals. Surprisingly, NLRP3^−/− mice manifested an early onset, more-severe SK lesions, and angiogenesis when compared with WT animals. This was associated with significantly increased early neutrophil infiltration into the corneas and heightened cytokines and chemokines, including cleaved IL-1β, IL-18, and elevated T cell numbers compared with WT mice. Reduction of the early neutrophil infiltration prevented the early onset of the disease and reduced the levels of cleaved IL-1β in NLRP3^−/− animals. Our study indicates that NLRP3 has an immunoregulatory function in SK pathogenesis modulating the early immune response after HSV infection.

MATERIALS AND METHODS

Mice

Female, 6–8-wk-old, C57BL/6 mice were purchased from Harlan Sprague Dawley Inc. (Indianapolis, IN, USA). NLRP3 knockout (NLRP3^−/−) mice were the kind gift of Dr. Gabriel Nuñez (University of Michigan). The animals were housed in American Association of Laboratory Animal Care–approved facilities at the University of Tennessee, Knoxville. All investigations followed guidelines of the institutional animal care and use committee.

Virus

HSV strain RE Tumpey was propagated in Vero cell monolayers (American Type Culture Collection CCL81; Manassas, VA, USA). Virus was grown in Vero cell monolayers, titrated, and stored in aliquots at −80°C until used. Corneal infections of mice were performed under deep anesthesia. The mice were lightly scarified on their corneas with a 27-gauge needle, and a 3-µl drop containing 10⁴ PFU of HSV RE was applied to 1 eye. Scratched animals were used as controls. These mice were monitored for the development of SK lesions. The SK lesion severity and angiogenesis in the eyes of mice were examined by slit-lamp biomicroscopy (Kowa Company, Nagoya, Japan). The scoring system was as follows: 0, normal cornea; +1, mild corneal haze; +2, moderate corneal opacity or scarring; +3, severe corneal opacity but iris visible; +4, opaque cornea and corneal ulcer; and +5, corneal rupture and necrotizing keratitis [37]. The severity of the angiogenesis was recorded as described previously [38]. According to this system, a grade of +4 for a given quadrant of the circle represented a centripetal growth of 1.5 mm toward the corneal center. The score of the 4 quadrants of the eye were then summed to derive the neovessel index (range, 0–16) for each eye at a given time point.

Histopathology

For histopathologic analysis, eyeballs from different groups were extirpated on d 15 PI and stored in 10% formalin. In brief, the samples were put into a Tissue-Tek processor (Sakura Finetek USA, Torrance, CA, USA) overnight, which removes all the moisture content from the samples and embeds it in paraffin. Tissue-Tek was automatically programmed, treating the samples sequentially with 100% alcohol, 100% xylene, and paraffin. Six-micrometer sections were then cut using a microtome and stained with H&E (Richard Allen Scientific, Kalamazoo, MI, USA).

ELISA

Corneal samples were pooled group-wise (4–5 corneas per sample) and collected in PBS with antiprotease cocktail. After homogenization of the sample with a tissue homogenizer (Kontes Pellet Pestle mortar; Kimble Chase, Vineland, NJ, USA), the concentrations of bioactive IL-1β, IL-18, and VEGF were measured by sandwich ELISA, according to the manufacturers’ instructions (Mouse IL-1β ELISA Ready-SET-Go!; Mouse IL-18 Platinum ELISA; Mouse VEGF-A Platinum ELISA; eBioscience, San Diego, CA, USA).

Quantification of mRNA expression levels by qRT-PCR

Total mRNA was isolated from corneal and lymph node cells using mirVana microRNA Isolation Kit (Ambion, Austin, TX, USA). For RNA, cDNA was made with 500 ng of RNA using oligo (dT) primer and ImProm-II Reverse Transcription System (Promega, Fitchburg, WI, USA). TaqMan gene expression assays (IL-6, IL-12, TNF-α, IL-17, MIP2, KC, IL-10, proteinase 3, elastase, and cathepsin) were purchased from Applied Biosystems (Foster City, CA, USA) and were used to quantify mRNAs using a 7500 Fast Real-Time PCR System (Applied Biosystems). The expression levels of the target genes were normalized to β-actin and with the comparative cycle threshold (ΔCT) method, and relative quantification between control and infected mice was performed using the 2^−ΔΔCT × 1000 formula.

Virus recovery and titration

Corneas were extracted on d 2, 4, 7, and 10 PI and placed on sterile-ice, 2-ml, straight-wall, ground-glass tissue homogenizers (Wheaton, Millville, NJ, USA) with media and were homogenized. Homogenates were centrifuged (2250 g at 4°C) for 5 min, placed on ice, and immediately plated titrations were performed by a standard plaque assay, as described previously [39]. Titers were calculated as log¹⁰ PFU/ml per a standard protocol [40].

Flow cytometry

Cell preparation.

Single-cell suspensions were prepared from the corneas and cervical DLN of mice at different time points PI. Corneas were excised, pooled group-wise, and digested with 60 U/ml Liberase (Roche Diagnostics, Indianapolis, IN, USA) for 35 min at 37°C in a humidified atmosphere of 5% CO₂. After incubation, the corneas were disrupted by grinding with a syringe plunger on a cell strainer, and a single-cell suspension was made in complete RPMI 1640 medium.

Staining for flow cytometry.

The single-cell suspensions obtained from corneas and DLN were stained for different cell surface molecules for FACS. All steps were performed at 4°C. A total of 1 × 10⁶ cells were stained with the respective antibodies for 30 min on ice. Finally, the cells were washed 3 times and resuspended in 1% paraformaldehyde. The stained samples were acquired with FACS LSR (BD Biosciences, Franklin Lakes, NJ, USA), and the data were analyzed using the FlowJo software (Tree Star, Ashland, OR, USA). For corneas, total cell numbers were calculated by acquiring the totality of the sample and taking into consideration the total number of corneas in the sample. The antibodies used included anti-CD4 Percp, anti-CD45 APC cy7, anti-CD11b Percp, anti-Ly6G Pacific blue, anti-CD4 APC, anti-CD45 Percp, anti- CD11b PE, anti- Ly6G FITC, for 30 min on ice.

To enumerate the number of IFN-γ– and IL-17–producing CD4⁺ T cells, intracellular cytokine staining was performed, as previously described [41]. In brief, 10⁶ freshly isolated cells from lymph node and corneas were cultured in U-bottom 96-well plates. Cells were left unstimulated or stimulated with PMA (50 ng) and ionomycin (500 ng) for 4 h in the presence of brefeldin A (10 µg/ml). Subsequently, cell-surface staining was performed, followed by intracellular-cytokine staining using a Cytofix/Cytoperm kit (BD Pharmingen, San Jose, CA, USA) in accordance with the manufacturer’s recommendations. The antibodies used were anti-IFN-γ APC and anti-IL-17 PE. The fixed cells were resuspended in 1% paraformaldehyde. The stained samples were acquired with a FACSCalibur (BD Biosciences), and the data were analyzed using the FlowJo software. For DLN approximately 200,000 events were recorded. For corneal samples, depending on the number of corneas pooled, approximately 300,000 to 1.5 × 10⁶ events were recorded.

FLICA assay.

Active caspase 1 was detected using a fluorescent inhibitor of caspases (FLICA assay; ImmunoChemistry Technologies, Bloomington, MN, USA), according to the manufacturer’s instructions. Briefly, single-cell suspensions were prepared from pools of infected corneas previously collagen digested and stained for anti-CD45. Then 10 µl of a 30× FLICA solution was added. The culture plates were covered with aluminum foil and incubated 1 h at 37°C in 5% CO₂. Following incubation, the cells were washed with wash buffer. At the end, the samples were labeled with cell propidium iodide (Molecular Probes, Eugene, OR, USA) and acquired with a FACS LSR (BD Biosciences). The data were analyzed with FlowJo software.

Murine treatment with soluble anti-Ly6G.

WT and 2 groups of NLRP3^−/− mice were ocularly infected with HSV RE Tumpey. Only one of the groups of NLRP3^−/− mice were treated with 50 μg/kg of anti-mouse Ly6G mAbs (clone 1A8; Bio X Cell, West Lebanon, NH, USA) intraperitoneally from d −1 to d 6 PI. Animals in the control group, NLRP3^−/− and WT mice, were given an isotype control (IgG2b) Ab (LTF-2; Bio X Cell) following the same regimen. All experiments were repeated 2 times.

Statistics.

The statistical significance between the 2 groups was determined using unpaired, 1-tailed Student’s t test. When data did not show a normal distribution, the Mann-Whitney test was used, and 1-way ANOVA with Tukey’s multiple comparison tests was used to calculate the level of significance of the experiments with >2 groups to compare. P ≤ 0.001 (***), P ≤ 0.01 (**), P ≤ 0.05 (*) were considered significant, and results were expressed as means ± sem. For all statistical analysis, GraphPad Prism software (GraphPad Software, La Jolla, CA, USA) was used.

RESULTS

NLRP3 deficiency induces an early onset and more-severe disease

To investigate the potential role of NLRP3 in the pathogenesis of SK, the outcome of HSV ocular infection was compared in WT animals and mice lacking NLRP3 because of gene knockout. In addition, at d 7 and 15 PI, eyes were collected and histologic sections examined for the level of inflammation. Because herpetic SK is an immunoinflammatory lesion and research on some other immunoinflammatory diseases indicate that NLRP3^−/− mice express higher resistance than WT animals do [34, 42, 43], we anticipated that NLRP3^−/− mice would be more refractory to SK. Unexpectedly, NLRP3^−/− mice developed more-severe SK lesions than WT mice did (Fig. 1A). Usually, SK lesions in WT animals start to manifest on about d 9 PI, but lesions were evident in NLRP3^−/− mice as early as d 6 PI (P < 0.05) (Fig. 1A and B). Consistently, HSV-infected NLRP3^−/− mice exhibited increased cellular infiltration, as could be seen in tissue sections compared with HSV-infected WT mice (Fig. 1B). At d 15 PI, when both NLRP3^−/− and WT mice showed clinically and histologically evident corneal lesions, both parameters were exacerbated in NLRP3^−/− animals (P = 0.001) (Fig. 1B). Alongside infected and naive animals, scratched control samples were also included in the analyses. None showed lesions and were more or less identical to naïve controls (Fig. 1B). These experiments demonstrated that, in the absence of NLRP3, HSV infection generates an earlier onset and more-severe manifestations of ocular lesions.

Figure 1. NLRP3^−/− animals have an early onset and more-severe disease.

C57BL/6 (WT) and NLRP3^−/− animals were scarified and infected with HSV (HSV infected) and scarified but uninfected (scratched control). The disease progression was analyzed throughout time in a blinded manner using a scale described in the “Materials and Methods.” (A) The progression of SK and angiogenesis lesion severity was significantly increased in the NLRP3^−/− compared with the WT mice. (B) Representative eye photos show increased SK lesions and angiogenesis development in NLRP3^−/− mice compared with WT mice on d 7 and 15 PI. Those eyes were processed for cryosections, and H&E staining was carried out on 6-μm sections. Histopathology pictures were taken at original magnification ×40 microscope augmentation. Data are representative of 3 independent experiments and show means ± sem (n = 8 mice/group). Inf., infected. ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

HSV-induced IL-1β and IL-18 levels are NLRP3 independent

One of the consequences of inflammasome signaling involves the inflammasome-dependent secretion of IL-1β and IL-18 [17]. These cytokines are known to be synthesized as proproteins without significant biologic activity until caspase 1 cleaves them into their bioactive forms for subsequent release [24, 44]. To examine the presence of IL-1β and IL-18 in our model, WT and NLRP3^−/− animals were ocularly infected with HSV, and on d 2, 7, and 15 PI, corneas were collected for the measurement of such cytokines by an ELISA assay that detects the bioactive form of both cytokines. Surprisingly, IL-1β and IL-18 were expressed in NLRP3^−/− corneas (Fig. 2A and B). Comparing both groups of animals, IL-18 levels were similar at all time points, but IL-1β levels showed some differences. IL-1β levels were increased at all time points and were significantly different at d 7 PI. Next, to explore the levels of bioactive caspase 1 in both groups of animals, HSV-infected WT and NLRP3^−/− corneas were collected at d 2 PI, and caspase 1 was measured using a fluorescent inhibitor of caspases (FLICA assay). As shown in Fig. 2C NLRP3^−/− and WT corneas expressed similar high levels of bioactive caspase 1 in CD45⁺ corneal cells. Taken together these results indicate that, upon HSV infection, maturation of caspase 1 and subsequent increased levels of IL-1β and IL-18 can occur independently of NLRP3.

Figure 2. Mice lacking NLRP3 present bioactive levels of IL-1β and IL-18 following ocular challenge with HSV.

C57BL/6 (WT) and NLRP3^−/− mice were infected with HSV, and corneal samples were processed to measure IL-1β and IL-18 by ELISA and activated caspase 1 by FLICA assay. (A) Quantification of mature IL-1β was performed at different time points. On d 7 PI, NLRP3^−/− mice had significantly increased levels of matured IL-1β. (B) Quantification of matured IL-18 protein was performed at different time points. Matured IL-18 was similar between NLRP3^−/− and WT animals. (C) Representative histogram of FLICA⁺ cells gated on total CD45⁺ PI cells infiltrated in the corneas of WT and NLRP3^−/− animals. Spleen of naïve mice was used as an isotype control. Data are representative of 2 independent experiments and show means ± sem (n = 3; each sample is representative of 5 corneas). *P ≤ 0.05.

NLRP3 absence amplifies the proinflammatory cytokine and chemokine responses

To assess the effect of NLRP3 on the production of proinflammatory cytokine and chemokine production in SK, mRNA was prepared from HSV-infected NLRP3^−/− and WT corneal extracts on d 2 and d 15 PI. Subsequently, IL-6, MIP2, IL-17, IL-12, TNF-α, and KC were measured by TaqMan qRT-PCR. As shown in Fig. 3A, by d 2 PI, almost all cytokine levels were significantly increased in NLRP3^−/− mice compared with WT mice (1.5–5-fold higher) (P ≤ 0.05). However, on d 15 PI, although IL-6, IL-12, MIP2, and TNF-α were increased in the NLRP3^−/−, compared with the WT animals (around 2 to 5 fold higher) (P ≤ 0.05), there were no significant differences in IL-17 and KC levels between the 2 groups. We also quantified the levels of the anti-inflammatory cytokine IL-10 in both groups and observed that on d 15 PI, it was around 2-fold up-regulated in NLRP3^−/− compared with WT animals. In addition, pools of corneas were collected on d 2 and 15 PI for measurement of VEGF by ELISA. As shown in Fig. 3B, VEGF was significantly increased at both time points by around 2- and 4-fold in NLRP3^−/− compared with WT animals. These data indicate that the deficiency of NLRP3 results in early increase of proinflammatory and proangiogenic cytokines and chemokines and IL-10 in response to HSV infection.

Figure 3. NLRP3 deficiency increases proinflammatory cytokines and chemokines C57BL/6 (WT) and NLRP3^−/− mice corneas were scarified and infected with HSV.

On d 2 and 15 PI, corneas were collected. (A) Relative fold change in mRNA expression of IL-6, IL-12, TNF-α, MIP2, KC, IL-17, and IL-10 was examined by RT-qPCR and compared between both groups. (B) Quantification of VEGF was measured by ELISA on d 2 and 15 PI. NLRP3^−/− mice had significantly increased levels of VEGF at both time points. Data represent means ± sem from 3 different independent experiments (n = 3; each sample is representative of 5 corneas). Inf., infected ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Effect of NLRP3 on viral clearance

To compare viral titers between NLRP3^−/− and WT animals corneas were extracted on d 2, 4, 7 and 10 PI, and viral titers were detected by plaque assay. The results revealed that NLRP3^−/− animals presented 1-fold higher viral titers on d 2 (Fig. 4A). By d 7 PI, both groups of animals cleared the virus. These data suggest that animals unable to respond with NLRP3 have slightly increased viral titers at early stages of the disease. However, although the viral titers were statistically significant, such differences were not considered to be biologically meaningful.

Figure 4. Corneal viral titers and neutrophil infiltration in HSV infected eyes of NLRP3^−/− and WT animals.

C57BL/6 (WT) and NLRP3^−/− mice were scarified and infected with HSV. (A) Corneal tissue was collected on d 0, 2, 7 and 10 PI, and titration was performed by standard plaque assay as described in the “Materials and Methods” section. Titers were calculated as log¹⁰ PFU/ml. Data are representative of 3 independent experiments and show means ± sem (n = 8 mice/group). (B and C) Corneas were collected at different time points to analyze the neutrophil infiltration throughout the disease. Numbers of total neutrophil infiltration (B) and representative FACS plots and percentages (C) are shown for d 2, 7, and 15 PI. At all time points, neutrophil infiltration was significantly increased in NLRP3^−/− compared with WT mice. Data are representative of 3 independent experiments and show means ± sem (n = 6; each sample is representative of 2 corneas). KO, knock out. **P ≤ 0.01, *P ≤ 0.05.

NLRP3 deficiency increases the influx of neutrophils at early stages

To evaluate the extent of neutrophil infiltration, pools of corneas from NLRP3^−/− and WT animals were collected on d 2, 7, and 15 PI. After collagen digestion, the pools of corneas were processed to quantify neutrophils by FACS. On d 2 and 7 PI, neutrophil infiltration was around 2- and 10-fold higher in NLRP3^−/− compared with WT animals (P ≤ 0.05), respectively (Fig. 4B and C). On d 15 PI, even though NLRP3^−/− animals had higher numbers of neutrophils infiltrating the cornea, there were no significant differences when compared with WT mice (Fig. 4B and C). The results indicate that, in the absence of NLRP3, the infiltration of neutrophils continues to increase from early time points until the peak of the disease.

Loss of NLRP3 leads to increased Th1 and Th17 cell responses

Because NLRP3^−/− mice developed clinical SK lesions as soon as 7 d PI, the level of CD4⁺, CD4⁺Th1⁺, and CD4⁺Th17⁺ corneal and DLN infiltration was examined on d 7 and 15 PI. The numbers of cells of several phenotypes were measured on pools of 4 corneas and the DLN by FACS. On d 7 PI, NLRP3^−/− mice showed around 2- to 4-fold increased CD4⁺ and CD4⁺Th1⁺ corneal cell numbers compared with WT animals (P ≤ 0.05) (Fig. 5A). Additionally, on d 15 PI, CD4⁺ T cells and Th1 cells increased around 1.5-fold in NLRP3^−/− corneas compared with WT corneas (Fig. 6A). Examination of DLN cell number revealed that on d 7 PI, NLRP3^−/− animals showed increased CD4⁺, Th1⁺, and Th17 cells compared with WT mice (Fig. 5B) (around 1.5-fold increase). Finally, on d 15 PI, the numbers of CD4⁺, CD4⁺Th1⁺, and CD4⁺Th17⁺ cells were increased in NLRP3^−/− DLN compared with WT DLN (around 1.5- to 2.5-fold increase) (Fig. 6B). In accordance with the severe disease seen in NLRP3^−/− mice, these results provide evidence that, as early as 7 and 15 d PI, the major orchestrators of this disease, CD4⁺ T cells, were significantly increased in NLRP3^−/− compared with WT mice. Additionally, T_regs were also quantified in corneas and DLN on d 15 PI. The ratio of T_reg to Th1 cells was calculated, and there were no significant differences between both groups in either cornea or DLN (Fig. 6C). These experiments indicate that mice unable to activate NLRP3 after HSV ocular infection generate a more-intense adaptive immune response compared with WT animals. This is probably the consequence of the amplified immune response that occurs during early stages of the disease in NLRP3^−/− mice.

Figure 5. NLRP3^−/−mice exhibit increased corneal and lymph node cellular infiltrates at d 7 PI.

C57BL/6 (WT) and NLRP3^−/− animals were infected with HSV. On d 7 PI, corneas and DLN were collected and stimulated with PMA/ionomycin for 4 h. (A) Representative FACS plots and percentages (left) and numbers of CD4⁺ T cells and CD4⁺ IFN-γ secreting cells from pooled corneas. (B) Representative FACS plots and percentages (left) and numbers of total CD4⁺T cells (right), CD4⁺ IFN-γ and IL-17 from lymph nodes. Data are representative of 3 independent experiments and show means ± sem (n = 8). In the case of corneas each sample is representative of 3 corneas. SSC, side scatter. **P ≤ 0.01, *P ≤ 0.05.

Figure 6. NLRP3^−/− mice exhibit increased corneal and lymph node cellular infiltrates on d 15 PI.

C57BL/6 (WT) and NLRP3^−/− animals were infected with HSV. On d 15 PI, corneas and lymph nodes were collected and processed for stimulation with PMA/ionomycin for 4 h. (A) Representative FACS plots and percentages and numbers of CD4⁺ T cells and CD4⁺ IFN-γ secreting cells from corneas taken on d 15 PI. (B) Representative FACS plots and percentages and numbers of total CD4⁺ T cells. CD4⁺ IFN-γ and IL-17 from lymph nodes were determined on 15 d PI. Cell ratios are for total numbers of T_regs per Th1 in lymph nodes (C) and corneas (D) on d 15 PI. Data are representative of 3 independent experiments and show means ± sem (n = 8). In the case of corneas, each sample is representative of 3 corneas. SSC, side scatter. ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Neutrophil reduction prevents the early onset of SK

Previously we showed that NLRP3^−/− animals presented high levels of IL-1β and IL-18 despite the absence of NLRP3. However, it has been demonstrated that in neutrophil-mediated inflammatory responses, neutrophil-derived serine proteases can also cleave pro-IL-1β and pro-IL-18. To explore the neutrophil-derived serine proteases expression, pools of 5 corneas were collected on d 2 PI from NLRP3^−/− and WT animals. Corneas from both groups were processed for the extraction of mRNA and quantification of the aforementioned proteases. Exposure to HSV significantly increased the neutrophil proteinase 3 and cathepsin b by around 3-fold in NLRP3^−/−compared with WT animals (Fig. 7). Even though elastase levels were increased in NLRP3^−/− animals, the differences between groups were not significant.

Figure 7. NLRP3^−/− animals have high levels of neutrophil-derived proteases C57BL/6 (WT) and NLRP3^−/− mice corneas were scarified and infected with HSV.

(A–C) On d 2 PI, corneas were collected and relative fold change in mRNA expression of neutrophil elastase (A), proteinase-3 (B), and cathepsin (C) was examined by qRT-PCR and compared between both groups. Data represent means ± sem from 3 different independent experiments (n = 3; each sample is representative of 5 corneas). Inf., infected., **P ≤ 0.01, *P ≤ 0.05.

Neutrophils together with CD4⁺ T cells are known to be the main cells driving the immunopathogenesis of SK [45]. However, the early infiltration of neutrophils is usually not clinically evident. To evaluate whether the early infiltration of neutrophils was the main cause of the early onset of disease in knockout mice, NLRP3^−/− animals were depleted of neutrophils using anti-Ly6G (clone 1A8; Bio X Cell), and NLRP3^−/− and WT animals received isotype control (IgG2b) Ab (LTF-2; Bio X Cell) from d −1 to d 6 PI. (Fig. 8A). To confirm neutrophil reduction, histopathology and neutrophil count by FACS were used. The results showed that the treatment effectively reduced neutrophil infiltration (Fig. 8B and C). Comparing NLRP3^−/− treated animals with NLRP3^−/− control animals, the SK scores were reduced by approximately 40%. Accordingly, anti-Ly6G treatment in NLRP3^−/− animals did not lead to an early onset of the disease (Fig. 8D). In addition, administration of anti-Ly6G decreased IL-1β levels by approximately 4-fold in NLRP3^−/− treated compared with NLRP3^−/− control animals (P ≤ 0.05), but IL-18 did not change significantly (P ≥ 0.05) (Fig. 8E and F). These data suggest that the infiltration of neutrophils has an important role driving the early manifestation of the disease in NLRP3^−/− animals. In addition, the increased levels of IL-1β in NLRP3^−/− animals could be neutrophil dependent.

Figure 8. Reduction of the early neutrophilic infiltrate prevents the early onset of the disease in NLRP3^−/− animals.

C57BL/6 (WT) and 2 groups of NLRP3^−/− animals were scarified in the eye and infected with HSV. (A) Using mAbs against Ly6G from d −1 to d 6 PI, neutrophils were reduced in 1 group of NLRP3^−/− animals (NLRP3^−/− TRT). The other 2 groups, including NLRP3^−/− and WT mice, were used as controls (NLRP3^−/− control and WT control) and treated with isotype control (IgG2b) Abs from d −1 to d 6 PI. (B) Representative FACS plots and percentages of corneas collected on d 7 PI show that mouse anti-Ly6G treatment effectively reduced the infiltration of neutrophils in corneas. (C) Representative histopathologic pictures taken at original magnification ×40 microscope augmentation show that NLRP3^−/− TRT animals had less cellular infiltration and fewer neutrophils than NLRP3^−/− control mice. (D) The progression of SK and angiogenesis was evaluated throughout the disease, and both were significantly increased in NLRP3^−/− control compared with NLRP3^−/− TRT and WT control animals. (E and F) Quantification of bioactive IL-18 and IL-1β protein on d 7 PI by ELISA. NLRP3^−/− mice treated with anti-Ly6G had similar levels of bioactive IL-18 to NLRP3^−/− and WT control. However, bioactive concentrations of IL-1β were reduced in NLRP3^−/− TRT compared with NLRP3^−/− control WT and NLRP3^−/− animals. Data are representative of 2 independent experiments and show means ± sem (n = 8 mice/group). TRT, treated. **P ≤ 0.01, *P ≤ 0.05.

DISCUSSION

Ocular infection with HSV sets off an array of events that succeeded in clearing virus from the cornea, but the tissue was damaged by a CD4⁺ T cell–orchestrated chronic inflammatory lesion that impairs vision [1]. An unresolved issue is how early events during infection relate to the subsequent immunoinflammatory SK lesions. In this report, we determined whether the NLRP3 inflammasome participates in the early response to HSV by comparing the outcome of infection in WT mice with animals that lack the NLRP3 function because of gene knockout. Unexpectedly, we observed that without NLRP3 mice developed more-severe lesions than did intact animals. Along with more-severe clinical lesions, infiltrates of CD4⁺ T cells, and neutrophils that were higher, the levels of proinflammatory chemokines and cytokines, including IL-1β and IL-18, were also increased. The heightened lesions in NLRP3^−/− mice appeared dependent on neutrophils because removal of such cells lessened the lesion severity. Our results indicate that NLRP3 may have a modulatory role, acting in some way to diminish the severity of lesions in ocularly HSV-infected animals.

Few studies have evaluated the role of NLRP3 in the pathogenesis of a viral infection [19, 35, 36]. Most past reports have focused on immunoinflammatory [34, 42, 43], autoimmune diseases [31–34, 46], and some types of cancers [47]. The studies showed that NLRP3 helped mediate inflammatory events, which markedly reduced if NLRP3 was absent or if its function was inhibited. Previous reports on the role of NLRP3 in influenza have revealed a confused phenotype. Although some advocate that NLRP3 has a proinflammatory role that is beneficial for viral clearance and animal survival [19, 35], others show that the presence or absence of NLRP3 does not influence the final outcome of the disease [36]. However, in diseases such as colitis and fungal infection, NLRP3 was shown to have beneficial roles [18, 48–52]. In colitis, the absence of NLRP3 reduced the levels of IL-18, which was shown to be critically involved in the maintenance of intestinal homeostasis [50].With the fungal infection, the explanation for more-severe lesions in the absence of NLRP3 was attributed to the impaired IL-1β response [18, 48–53]. Our study is the first, to our knowledge, to evaluate any role for NLRP3 during an in vivo infection with HSV. However, studies in vitro with HSV had indicated that activation of NLRP3 and IFI16/204 inflammasomes occurs with the subsequent maturation of the IL-1β response [29, 30]. The component of the virus that activates NLRP3 is not known, but Johnson et al. [30] suggested that NLRP3 could be acting indirectly following activation by causing the production of reactive oxygen species during HSV infection and that another inflammasome IFI16/204 could also be involved in the HSV response. Our observation that, in the absence of NLRP3, an inflammatory response followed by SK occurs means that NLRP3 inflammasome formation is not an essential event responsible for the early inflammatory response in the eye to HSV. Conceivably, other recognition systems are primarily involved, such as the known TLR ligand activity of the virus itself [10, 11, 54] or the involvement of additional inflammasomes. In support of the latter possibility, we found, in preliminary studies, that mice lacking ASC, the adaptor molecule for the assembly of many inflammasomes, had a less-severe response to HSV than did WT animals. Further studies are needed to evaluate the involvement of additional inflammasomes, such as the aforementioned IFI16/204.

Although our studies argue that NLRP3 may not be critical for driving the early inflammatory response to HSV, they do make the point that some form of recognition must be occurring because lesions were more severe in the absence of NLRP3. We also observed that the more-severe phenotype depended on the presence of neutrophils because when those cells were depleted, the more-severe phenotype in NLRP3^−/− was eliminated. Although it was not clear what generated the heightened IL-1β responses, this molecule could have been acting in a positive feedback loop with neutrophils. Thus, it is known that IL-1 signaling through one of its receptors results in the activation of the transcription factors NF-κB and AP-1 [55, 56]. These, in turn, cause the expression of vascular adhesion molecules and induction of chemokines, such as KC and MIP2, known to increase neutrophil infiltration [56]. Although caspase 1 is the master regulator of IL-1β, inflammatory mediators, such as neutrophil-derived proteases can also cause IL-1β to mature [57–59]. In support of this, administration of neutrophil-derived protease inhibitors was shown to reduce levels of IL-β in a mouse model of arthritis [60]. Because the neutrophil-derived proteases, proteinase, and cathepsin-G were shown to increase in NLRP3^−/− animals, it is conceivable that neutrophils also contributed in the maturation of IL-1β. Another explanation for increased IL-1β production in NLRP3^−/− animals could be due to the activation of another inflammasome, which leads to caspase 1 activation, and consequently, mature IL-1β production.

The most fascinating and yet unexplained aspect of our study was the observation that SK lesions in NLRP3^−/− animals were more severe than in animals with functional NLRP3. In fact, our results imply that NLRP3 could act to regulate the inflammatory reaction to HSV in normal animals. Initially, we suspected that the apparent regulatory effect might be the consequence of less ability of NLRP3^−/− animals to control viral infection with the virus itself driving the more severe inflammatory reaction. However, the difference in viral titers in NLRP3^−/− and WT animals were minimal, and both WT and NLRP3^−/− animals cleared virus in the same time frame. Another explanation for the regulatory influence could relate to differences in the balance of the T cell responses with less T_reg cell induction in NLRP3^−/− animals than in WT animals. However, we could not detect significant difference in the effector to T_reg cell relationship between NLRP3^−/− and WT infected animals. Conceivably, the more-severe outcome in NLRP3^−/− mice could be explained by lesser levels of regulatory IL-10 made by NLRP3^−/− animals, but our results did not support this scenario.

Another explanation for the apparent regulatory effect of NLRP3 is that it interacts in some way with other inflammasomes that are activated by the virus. Our preliminary data that animals unable to activate multiple inflammasomes because of ASC^−/− showed minimal SK lesions would support such an explanation, and this issue is being further investigated in our laboratory. Lastly, NLRP3 could have an inflammasome-independent role in HSK. Thus, NLRP3 was shown to have an inflammasome-independent role, acting as a key transcription factor in Th2 differentiation [61]. NLRP3 was shown to traffic to the nucleus together with IRF4 in Th2 cells, where it bound to regulatory DNA sites and promotes the transcription of genes encoding IL-4, IL-5, and IL-13. Therefore, it is conceivable that, after HSV infection, NLRP3 could traffic into the nucleus and promote the transcription of some anti-inflammatory genes. Thus in the absence of NLRP3 there would be less transcription of anti-inflammatory genes, which might explain the increased SK lesion severity in NLRP3^−/− animals. These ideas are currently being further investigated.

In conclusion, our studies reveal an unexpected role for the NLRP3 inflammasome in an ocular inflammatory response to HSV infection. Our observation that NLRP3 could have a regulatory effect, although still unexplained at a mechanistic level, might be exploitable therapeutically.

AUTHORSHIP

F.G. and B.T.R. conceived and designed the experiments; F.G., S.B., P.D., L.H., S.K.B., and U.J. performed the experiments; F.G. and B.T.R. analyzed the data and wrote the paper.

Glossary

DLN

draining lymph node

KC (CXCL1)

chemokine ligand 1

MIP2

macrophage inflammatory protein 2

NLRP3

NLR family, pyrin domain containing 3

PI

postinfection

qRT-PCR

quantitative real-time polymerase chain reaction

SK

stromal keratitis

VEGF

vascular endothelial factor

WT

wild type