IL-1β processing during Pseudomonas aeruginosa infection is mediated by neutrophil serine proteases and is independent of NLRC4 and Caspase-1

Abstract

To examine the role of caspase-1 and the NLRC4 inflammasome during bacterial infection, C57BL/6, IL-1β^−/−, Caspase-1^−/−, and NLRC4^−/− mouse corneas were infected with ExoS/T or ExoU expressing Pseudomonas aeruginosa. We found that IL-1β was essential for neutrophil recruitment and bacterial clearance, and was produced by myeloid rather than resident cells. In addition, neutrophils were found to be the primary source of mature IL-1β during infection, and that there was no significant difference in IL-1β processing between C57BL/6 and Caspase-1^−/− or NLRC4^−/− infected corneas. IL-1β cleavage by human and mouse neutrophils was blocked by serine protease inhibitors, and was impaired in infected neutrophil elastase (NE)^−/− corneas. NE^−/− mice also had an impaired ability to clear the infection. Together, these results demonstrate that during P. aeruginosa infection, neutrophils are the primary source of mature IL-1β, and that IL-1β processing is dependent on serine proteases and not NLRC4 or caspase-1.

IL-1β is a critical mediator of the host response to microbial infections. Production of functional IL-1β is a two-step process requiring an initial signal to induce transcription of the 31kDa pro-form, and a second signal for post-translational cleavage to the 17kDa mature form (1, 2). IL-1β processing by macrophages in vitro requires inflammasome and caspase-1 activation following exposure to Gram-negative bacteria (3–7), Gram-positive bacteria (8, 9), or yeast (10, 11). However, neutrophils rather than macrophages are generally the first cells to infiltrate the site of bacterial infection.

In the current study, we demonstrate that in a well characterized model of P. aeruginosa corneal infection (12, 13) neutrophils are the primary source of IL-1β in vivo, and that IL-1β cleavage during infection is independent of NLRC4 and caspase-1, but is dependent on neutrophil elastase. Given the ubiquitous presence of neutrophils in acute bacterial infections, it is likely that this alternate mechanism of IL-1β processing contributes to the host response in other microbial infections.

Materials and Methods

Source of mice

C57BL/6 mice were from The Jackson Laboratory (Jax, Bar Harbor, ME), IL-1β^−/−mice were obtained from Dr. Iwakura (University of Tokyo, Japan), and caspase1^−/− and NLRC4^−/− mice were generated by R. Flavell (Yale University, CT) and Millenium Pharmaceuticals, respectively. Neutrophil elastase gene knockout (NE^−/−) mice were generated by Shapiro (14), and purchased from Jax. All gene knockout mice are fully backcrossed to C57BL/6.

Bacterial strains and culture conditions

P. aeruginosa ExoS expressing strain PAO1 (15), and ExoU expressing strain 19660 (from ATCC) were examined in our recent studies (12, 13). Bacteria were grown in BHI (BD Diagnostics) to mid log phase (~1×10⁸ bacteria/ml), washed and diluted in sterile PBS to 1 × 10⁵ bacteria/2.5μl.

Murine model of P. aeruginosa corneal infection

The corneal epithelium was abraded, and infected topically with 1 × 10⁵ P. aeruginosa in 2.5μl PBS as described (12, 13). For systemic neutrophil depletion, 400 μg anti-neutrophil NIMP-R14 antibody was injected into the peritoneal cavity one day prior to infection. We, and others showed that NIMP-R14 is specific for murine neutrophils (12, 13, 16, 17).

Detection of cytokines in the cornea

Infected corneas were dissected and homogenized in PBS using a Mixer Mill MM300 (Retsch) for 4 min at 33Hz. Cytokines were measured by ELISA (R&D systems).

Colony forming units (CFU) quantification from infected cornea

At 24 or 48h post infection, mice were euthanized by CO₂ asphyxiation, eyes were homogenized in 1ml PBS, and serial log dilutions of bacterial homogenate were plated on BHI. CFU were counted manually after 18hr.

Bone marrow chimeras

Bone marrow cells were isolated from C57BL/6-GFP and IL-1β^−/− mice as described (18, 19). Recipient C57BL/6 mice received 2 × 600 Gy doses of full body irradiation 3h apart, then reconstituted with 5×10⁶ total bone marrow cells by tail vein injection. Chimeric mice were used 4 weeks later. This method results in up to 75% reconstitution of the myeloid cell population in recipient corneas (19).

Western blot analysis

Corneas were homogenized in cell lysis buffer. For peritoneal neutrophils, treated cells were washed in ice cold PBS and lysed. 30μg protein was fractionated on 12% SDS-PAGE, transferred to a nitrocellulose membrane and incubated with goat primary antibodies to IL-1β (R&D systems) or β-actin (Cell Signaling Technology). Reactivity was determined using HRP-conjugated secondary antibodies and developed with Supersignal West Femto Maximum Sensitivity Substrate (Pierce).

Flow Cytometry

Corneas were incubated in type I collagenase (Sigma) at 82U/cornea for 2h at 37°C. Fc receptors were blocked for 20 min with anti-mouse CD16/32 antibody (eBiosciences), and incubated with Alexa488-NIMP-R14 (in-house) and PeCy5-F4/80 (eBiosciences) to detect neutrophils and macrophages, respectively. Cells were then washed in 2mL of FACS buffer (1% FBS in PBS) and fixed in 0.5% PFA for analysis by flow cytometry.

For intracellular staining, cells were incubated at 4°C overnight with Protein Transport Inhibitor Cocktail (eBioscience), washed and further incubated in 20 min in permeabilization buffer (eBioscience). APC- conjugated anti-mouse IL-1β antibody (eBiosciences) was used, and stained cells were washed in FACS buffer and fixed in PFA for analysis.

Confocal Microscopy

Images were collected using an UltraVIEW VoX spinning disk confocal system (PerkinElmer) mounted on a Leica DMI6000B microscope equipped with a HCX PL APO 100×/1.4 oil immersion objective using a 0.2-micron step size. Images were then imported into Metamorph Image Analysis Software (Molecular Devices Corp) where maximum projections were generated from the original stacks and visualized following 2D deconvolution.

Peritoneal macrophage and neutrophil isolation and stimulation

For macrophages, mice were injected i.p with 3% thioglycolate, cells were recovered by peritoneal lavage after 3 days. Neutrophils were obtained following i.p injection of 3% thioglycolate 18h and 3h before lavage, and separated on a 90% Percoll (GE Healthcare) column. Cell purity was >97% for both populations.

Isolation of Human neutrophils from peripheral blood

Human neutrophils were isolated from the peripheral blood of healthy volunteers following informed consent as approved by the Institutional Review Board of University Hospitals of Cleveland. Heparinized blood was incubated with 3% dextran in PBS, and separated on 10 ml Ficoll- Paque Plus (GE Healthcare). Erythrocytes were lysed, and a >97% neutrophil population was obtained.

Neutrophil Protease activity assay and inhibition

Stimulated human neutrophils were lysed in lysis buffer (50mM Tris, 1% Triton-X100, 0.25% deoxycholate, 150mM NaCl, 1mM EGTA) and hydrolysis of colorimetric elastase substrate MeOSuc-AAPV-pNA (Calbiochem) was quantified in assay buffer (0.1M HEPES, 0.5M NaCl, 10% DMSO) at 405nm. To inhibit protease activity, rhSLPI (R&D systems), neutrophil elastase inhibitor III and IV (Calbiochem) and serine protease inhibitor 3,4 DCIC (Sigma Aldrich) was used.

Statistical Analysis

Statistical analysis was performed using ANOVA with Tukey post-test analysis (Prism; GraphPad Software). P values less than 0.05 were considered significant.

RESULTS

IL-1β from bone marrow derived cells regulates neutrophil recruitment to the cornea and bacterial clearance

To determine the role of IL-1β in chemokine production and neutrophil recruitment, C57BL/6 and IL-1β^−/− corneas were dissected 6h or 24h after infection with P. aeruginosa strain PAO1, and CXCL2/MIP-2 in homogenates was measured by ELISA. Neutrophils and macrophages were quantified by flow cytometry. Infected IL-1β^−/− corneas had significantly less CXCL2 (Figure 1A) and fewer neutrophils (Figure 1B) compared with C57BL/6 corneas, whereas there were fewer macrophages overall.

Figure 1. The role of IL-1β in P. aeruginosa corneal infection.

C57BL/6 and IL-1β^−/− corneas were infected with 1×10⁵ PAO1. A. After 6h, corneas were homogenized and cytokines were measured by ELISA. B. Total F4/80+ and NIMP-R14+ cells 24h post infection. (A,B: mean ± sd for 5 mice per group). C. CFU of C57BL/6, IL-1β^−/−, C57BL/6/IL-1β^−/− (Recipient/Donor) and C57BL/6/GFP-B6 (Recipient/Donor) at 24h and 48h after infection. Data points represent individual eyes. Results are representative of two independent experiments with five mice per group.

Consistent with impaired neutrophil recruitment, CFU was significantly higher in IL-1β^−/− compared with C57BL/6 corneas 48h post infection (Figure 1C). CFU was also elevated in chimeric mice given donor IL-1β^−/− bone marrow cells compared with those given C57BL/6 cells, indicating that myeloid cells rather than resident corneal epithelial cells or fibroblasts are the major source of IL-1β. Corneal opacity relating to cellular infiltration was lower in IL-1β^−/− and IL-1β^−/−/C57BL/6 chimeras compared with C57BL/6 mice (Supplemental Figure S1A). Our previous studies showed that impaired neutrophil infiltration resulted in increased corneal opacification at later time points due to higher bacterial load (12, 13). Consistent with these studies, we found that at 48h, corneal opacity was higher in the absence of IL-1β (Supplemental Figure S1A). Together, these data demonstrate that myeloid cell production of IL-1β regulates P. aeruginosa survival and corneal disease.

Neutrophils are the major source of IL-1β in infected corneas

To identify the IL-1β secreting myeloid cells in infected corneas, intracellular IL-1β was examined by flow cytometry. We found that > 80% NIMP-R14 positive neutrophils in infected corneas were IL-1β+ve (Figure 2A). Intracellular staining was confirmed by confocal microscopy (Figure 2B). F4/80+ve macrophages comprised ~6% total IL-1β positive cells in the cornea at 24h p.i., although the percent increased over 72h (Supplemental Figure S1B).

Figure 2. The role of neutrophils in IL-1β production in vivo.

A–C: Intracellular IL-1β in cells from C57BL/6 mice 24h after infection. A: flow cytometry scatter plot and histogram showing double positive cells. B: Representative cells visualized by confocal microscopy (original magnification is x100). C–E: Neutrophil depletion by intraperitoneal injection of NIMP- R14 18h prior to corneal infection. Total bone marrow cells stained with F4/80 or NIMP-R14 antibody, and analyzed by flow cytometry. Note selective depletion of neutrophils (C). D: Secreted and processed IL-1β. ELISA data are mean ± SD of five mice per group; western blot shows three representative corneas from each group. E. Bacterial CFU (Data points represent individual eyes). This experiment was repeated twice with similar results.

To determine if neutrophils are required for IL-1β production, we selectively depleted neutrophils by intraperitoneal injection of NIMP-R14 as shown in Figure 2C. Following corneal infection, neutrophil depleted mice had significantly lower secreted and mature IL-1β than control mice (Figure 2D). CFU was also significantly elevated in neutrophil depleted mice (Figure 2E), and corneal opacity as an indicator of neutrophil infiltration was significantly less (Figure 2E, Supplemental Figure S2A). Together, these findings demonstrate that neutrophils are the major source of IL-1β in early stage P. aeruginosa corneal infection.

Neutrophil mediated IL-1β processing in vivo is caspase-1 and NLRC4 independent

To ascertain the role of NLRC4 and caspase-1 in IL-1β processing by neutrophils, peritoneal neutrophils and macrophages were isolated from C57BL/6, caspase-1^−/− and NLRC4^−/− mice, incubated with 50:1 P. aeruginosa, and IL-1β was quantified by ELISA. IL-1β production by C57BL/6 macrophages was increased in response to P. aeruginosa (Figure 3A), and as reported (5–7), P. aeruginosa - induced IL-1β was ablated in caspase-1^−/− and reduced over five-fold in NLRC4^−/− compared with C57BL/6 macrophages. IL-1β secretion by caspase-1^−/− and NLRC4^−/− neutrophils was also significantly lower than from C57BL/6 neutrophils, although inhibition was partial (Figure 3B, C).

Figure 3. The role of caspase-1 and NLRC4in IL-1β processing.

A–B. in vitro: Secreted IL-1β from peritoneal macrophages (A) and neutrophils (B) from C57BL/6, caspase-1^−/− and NLRC4^−/− mice after 3h incubation with PAO1. C. Western blot of PAO1-stimulated neutrophils cleaved IL-1β. D, E: IL-1β production in C57BL/6, Caspase-1^−/− and NLRC4^−/− corneas 24h after infection with PAO1. IL-1β was measured by ELISA (D) and western blot was performed from infected corneas at 24h (E). ELISA data are mean ±sd of 5 mice per group, and western blot shows two representative corneas from each group. F–H: PAO1 infected corneas. F: Representative flow cytometry from infected caspase-1^−/− corneas. These experiments were repeated twice with similar results.

In contrast to in vitro findings, total and mature IL-1β in Caspase-1^−/− and NLRC4^−/−corneas were not significantly different from C57BL/6 mice after infection with P. aeruginosa strain PAO1 (Figure 3D, E). Also, NIMP-R14+ neutrophils were the predominant IL-1β producing cells in infected Caspase-1^−/− corneas (Figure 3F). Consistent with these findings, there were no significant differences in corneal opacification or bacterial CFU between C57BL/6 mice and either caspase-1^−/− or NLRC4^−/− mice, or in caspase-1^−/− mice infected with ExoU expressing P. aeruginosa 19660 (Supplemental Figure S2).

Taken together, these data indicate that during P. aeruginosa infection, caspase-1 and NLRC4 have either no role or a redundant role in IL-1β processing in the cornea, thereby implying that an alternate mechanism for IL-1β processing occurs in vivo.

IL-1β processing by murine and human neutrophils is mediated by serine proteases

As serine proteases also degrade mature IL-1β (20–22), we examined their role in neutrophil IL-1β secretion. Serine proteases in human neutrophils were incubated 30 min with the elastase inhibitor NEI III, a broad serine protease inhibitor 3,4 DCIC, or the Cathepsin G inhibitor rhSLPI. Live PAO1 or flagellin was added, and elastase activity and secreted IL-1β were measured after 3h incubation.

As shown in Figure 4A, elastase activity was not increased following incubation with PAO1 or flagellin, indicating constitutive activity. However, elastase activity was significantly lower in the presence of NEI III and 3,4 DCIC, whereas there was no difference after incubation with rhSLPI. IL-1β secretion was elevated after stimulation with PAO1 or flagellin, and was ablated in the presence of NEI III or 3,4 DCIC. IL-1β production was significantly lower after incubation with the rhSLPI- cathepsin G inhibitor, although the effect was partial (Figure 4B). Similarly, IL-1β secretion by P. aeruginosa - stimulated mouse neutrophils was significantly lower in the presence of murine elastase inhibitor NEIIV or 3,4 DCIC (Figure 4C). There were no cytotoxic effects even at the highest concentration of the inhibitors (Supplemental Figure S2D).

Figure 4. The role of serine proteases in IL-1β processing.

A. Intracellular neutrophil elastase activity of human neutrophils after incubation with SLPI (1μg/mL), NEI III (500μM) or 3,4 DCIC (100μM), and 3h incubation with live PAO1 or purified flagellin. B. IL-1β production by human neutrophils under the same conditions. C. PAO1 or flagellin induced IL-1β secretion by C57BL/6 peritoneal neutrophils in the presence of serine protease inhibitors. D: IL-1β expression and E: CFU in infected neutrophil elastase (NE)^−/− mice. Data points represent individual corneas.

To determine the role of elastase in IL-1β processing during P. aeruginosa keratitis, C57BL/6 and neutrophil elastase (NE)^−/− corneas were infected with 1×10⁵ PAO1, and mature IL-1β was examined after 24h. As shown in Figure 4D, mature IL-1β was lower in NE^−/− compared with C57BL/6 corneas, indicating a requirement for neutrophil elastase in IL-1β processing. NE ^−/− mice also had significantly higher CFU, indicating impaired bacterial clearance (Figure 4E). Corneal opacification was also higher in NE^−/− (Supplemental Figure S2F).

Taken together, these findings demonstrate that neutrophil elastase mediates IL-1β processing and susceptibility, consistent with impaired bacterial clearance during P.aeruginosa infection.

DISCUSSION

Neutrophils rather than macrophages are generally the first cells recruited to the site of bacterial infection. In the current study, we demonstrate that neutrophils are also the predominant source of pro- and mature forms of IL-1β in P. aeruginosa keratitis. We show that in addition to macrophages which are completely dependent on caspase-1 and NLRC4, IL-1β processing was impaired in NLRC4^−/− and caspase-1^−/− neutrophils in vitro. This observation is consistent with a recent study showing RNA expression of NLRP3, NLRC4 and caspase-1 in neutrophils, and also NLRP3 mediated IL-1β processing (23).

In contrast to in vitro studies, infected caspase-1^−/− and NLRC4^−/− mice were able to process IL-1β. Instead, IL-1β cleavage was blocked in NE^−/− mice, demonstrating an essential role for neutrophil elastase. Caspase-1 and elastase cleave the 31kD pro-IL-1β at different sites; however, both enzymes generate the 17–18kD mature form of IL-1β, which we also detect (20–22). Neutrophil elastase was also shown to mediate IL-1β processing in a model of sterile inflammation (24). In contrast to data presented here, earlier reports did find a role for caspase-1 in P. aeruginosa keratitis (25, 26); however, those studies focused on later time points in infection when macrophages are also present.

In conclusion, results from the current study demonstrate an essential role for neutrophil derived, inflammasome and caspase-1 independent process of IL-1β processing during bacterial infection that is likely to occur in other causes of acute microbial disease where neutrophils are present.