Caspase 1-independent activation of interleukin-1beta in neutrophil-predominant inflammation

Abstract

Objective: Interleukin-1beta (IL-1beta) is a key cytokine linked to the pathogenesis of acute arthritis. Caspase 1, neutrophil elastase, and chymase all process proIL-1beta to its biologically active form. This study was undertaken to examine the potential contributions of each of these proteases in experimental models of inflammatory arthritis.

Methods: Caspase 1-deficient (Casp1-/-) and wild-type (WT) mice were tested for their response to arthritogenic K/BxN serum transfer for induction of arthritis or injection of monosodium urate monohydrate (MSU) crystals for induction of peritonitis. All mice were prophylactically treated with inhibitors of neutrophil elastase or chymase. Arthritic paws were tested for the presence of IL-1beta protein by enzyme-linked immunosorbent assay and Western blotting. Neutrophils and mast cells from WT and mutant mice were tested for their ability to secrete IL-1beta after in vitro stimulation, in the presence of protease inhibitors.

Results: Casp1-/- and WT mice developed paw swelling to the same extent in the K/BxN serum transfer-induced arthritis model. MSU crystal injection into Casp1-/- mice also resulted in neutrophil influx and production of measurable peritoneal IL-1beta protein. Both of these responses were attenuated with neutrophil elastase inhibitors. K/BxN serum transfer-induced arthritis was also reduced by treatment with a chymase inhibitor. Casp1-/- neutrophils and mast cells, when exposed to MSU crystals, secreted similar amounts of IL-1beta protein upon in vitro stimulation with lipopolysaccharide, albeit at lower levels than that secreted by WT cells. Elastase and chymase inhibitors reduced the amount of IL-1beta released by these cells.

Conclusion: The production of IL-1beta by neutrophils and mast cells is not exclusively dependent on caspase 1, and other proteases can compensate for the loss of caspase 1 in vivo. These pathways might therefore compromise the caspase 1-targeted therapies in neutrophil-predominant arthritis.

Figure 1. IL-1β is processed and secreted in K/BxN serum transferred arthritis even in the absence of caspase-1

(A) Clinical scores and joint swelling were similar in Casp1^−/− and WT mice. Casp1^−/− (squares, N=19) and C57BL/6 (circles, n=20) mice were injected with 150 μl of pooled K/BxN sera on day 0. Swelling and clinical scores were serially monitored, and averages ± SEM are shown. No significant differences were observed (p>0.05 by ANOVA). Data are pooled from five independent experiments. (B) Joint inflammation and destruction are similar in C57BL/6 and Casp1^−/− mice. Mice were sacrificed on day 7 at the peak of arthritis and rear paws were removed, and prepared for histology. Representative ankles joints stained with H&E and safranin O are shown (original magnification x200). (C) IL-1β and IL-18 were assayed by ELISA in disrupted joint tissue from C57BL/6 (solid) and Casp1^−/− (striped) mice on day 0 and day 7. Shown are means ± SEM. Processed IL-1β was detectable by immunoblotting lysates of paws from C57BL/6 and Casp1^−/− mice on day 7.

Figure 2. Mast cells and neutrophils use several proteases to process IL-1β

(A) C57BL/6 and Casp1^−/− mast cells were stimulated for 4 hr with LPS (1μg/ml) and IL-1β mRNA transcript levels were analyzed by qPCR and normalized to18S mRNA. (B) Mast cells from C57BL/6 (B6, dots) and Casp1^−/− (C1, stripes) mice were stimulated for 16 hrs with LPS (1 μg/ml). The levels of secreted IL-1β were assayed by capture ELISA and cell lysates were probed for the presence of IL-1β. (C) Thioglycolate induced neutrophils from C57BL/6 micewere stimulated with LPS (100 ng/ml) in the presence (dark) or absence (light) of the MeOSuc-Ala-Ala-Pro-Val-cmk (500 μM) inhibitor. After 16 hrs levels of IL-1β secretion were measured by ELISA. (D) WT mast cells were stimulated for 16 hrs with LPS (1 μg/ml) in the presence of the Suc- Val-Pro-Phe^p(OPh)₂ inhibitor (dark) or vehicle (light). Additional wells received a calcium ionophore (A23187,0.5 μg/ml) for the last hr of incubation with LPS to enhance IL-1β secretion. Mast cells were also stimulated for 16 hrs with anti-FcγRII/III crosslinking in the presence of Suc-Val-Pro-Phe^p(OPh)₂ (dark) or vehicle (light). IL-1β secretion was measured by ELISA. Results are averages of three independent experiments ± SEM. * denotes p<0.05 versus control.

Figure 3. Serum transferred arthritis is attenuated by chymase and elastase inhibitors

Casp1^−/− mice were injected with 150 μl pooled K/BxN sera on day 0. Groups of mice were treated with vehicle (square, n=7), chymase inhibitor (Suc-Val- Pro-Phe^p(OPh)₂)(circle, n=7) or elastase inhibitor ONO-5046 (triangle, n=8) daily i.p. (A) Ankle thickness was assessed daily. Data shown were pooled from two independent experiments. * designates p<0.05 by ANOVA and post hoc Bonferonni comparisons. (B) On day 5 after serum transfer, mice were sacrificed and their paws were snap frozen. Pulverized paws were lysed and assayed for IL-1β and IL-18 y ELISA (n=4/group). * designates p<0.05 by ANOVA and significantly different to the WT vehicle treated mice by Dunnet’s test. (C) Lysates of pooled paws were immunoblotted for the presence of IL-1β and actin.

Figure 4. MSU stimulated IL-1β secretion by neutrophils depends on caspase-1 and other proteases

(A) Neutrophils from WT (solid) and Casp1^−/− (striped) were stimulated in culture for 4 hrs with MSU crystals (0.5 mg/ml), lysed and IL-1β mRNA transcripts were assessed by qPCR. The fold induction is relative to PBS treated control cultures. (B) Thioglycolate elicited neutrophils from WT (solid) and Casp1^−/− (striped) mice were harvested and stimulated with the indicated doses of MSU crystals. The levels of IL-1β released into the supernatant were measured after 16 hrs by ELISA. Thioglycolate stimulated peritoneal neutrophils from C57BL/6 (C) and Casp1^−/− (D) mice were stimulated in vitro with the indicated doses of MSU crystals in the presence of the MeOSuc-Ala-Ala-Pro-Val-cmk (500 μM, dark) inhibitor or vehicle (light). After 16 hrs IL-1β secretion to the supernatant was measured by ELISA. Results are averages of three independent experiments ± SEM.

Figure 5. Redundant roles for caspase-1 and elastase in murine MSU induced peritonitis

WT, Casp1^−/− or Il1r1^−/− mice (n=4/group) were i.p. injected with 1ml PBS or 3 mg/ml MSU crystals as indicated. Some groups of mice were i.p. injected with 1 mg MeOSuc-Ala-Ala-Pro-Val-cmk (AAPV) 1 hr before MSU crystal injection. The peritoneum was lavaged after 6 hrs. (A) Neutrophil influx was determined by cell count and flow cytometry. (B) IL-1β levels in lavage fluids were assayed by ELISA. Shown were means ± SEM. Significance was assessed by ANOVA with post hoc Bonferroni pair wise comparisons.