The NLRP12 inflammasome recognizes Yersinia pestis

Abstract

Yersinia pestis, the causative agent of plague, is able to suppress production of inflammatory cytokines IL-18 and IL-1β, which are generated through caspase-1-activating nucleotide-binding domain and leucine-rich repeat (NLR)-containing inflammasomes. Here, we sought to elucidate the role of NLRs and IL-18 during plague. Lack of IL-18 signaling led to increased susceptibility to Y. pestis, producing tetra-acylated lipid A, and an attenuated strain producing a Y. pseudotuberculosis-like hexa-acylated lipid A. We found that the NLRP12 inflammasome was an important regulator controlling IL-18 and IL-1β production after Y. pestis infection, and NLRP12-deficient mice were more susceptible to bacterial challenge. NLRP12 also directed interferon-γ production via induction of IL-18, but had minimal effect on signaling to the transcription factor NF-κB. These studies reveal a role for NLRP12 in host resistance against pathogens. Minimizing NLRP12 inflammasome activation may have been a central factor in evolution of the high virulence of Y. pestis.

Figure 1. Infection of mice with Y. pestis-pYtbLpxL is controlled by IL-18

(A) Survival of mice infected s.c. with 500 CFU of Yersinia pestis KIM1001 (C57Bl/6: n=8) or KIM1001-pYtbLpxL (C57Bl/6 and Tlr4^−/− [TLR4 KO]: n=10) bacteria. (B) Mortality of animals infected s.c. with 500 CFU of KIM1001-pYtbLpxL (n= 7 for Ifnar^−/− [IFNαβR KO], 8 for Il12b^−/− [IL-12p40 KO], 10 for C57Bl/6, Il1r1^−/− [IL-1R1 KO], Il18r1^−/− [IL-18R KO],Il18^−/− [IL-18 KO] and TLR4 KO). Statistical differences: TLR4, IL-18 or IL-18R vs IL-1R and other strains: p<0.002. IL-12p40, IL-18, IL-18R, TLR4 vs C57Bl/6: p<0.001. (C,D) Survival of mice deficient in (C) IL-1β (II1b^−/−, IL1β KO) and IL-1R and (D) IL-18 infected s.c. with 500 CFU of KIM1001-pYtbLpxL (n=10 of each genotype). (E,F) Concentrations of spleen IL-1β and IL-18 from C57BL/6 mice infected i.v. with 500 CFU of KIM1001 or KIM1001-pYtbLpxL for 44 hours. (G) IL-1β in supernatants from BMDM stimulated with 10 m.o.i. (multiplicity of infection) of KIM5 or KIM5-pYtbLpxL for 6 hours, 50µg/ml of gentamicin was added to wells after 3 h.p.i. (H) Immunoblot of IL-1β in the combined lysates and supernatants of BMDM stimulated with 10 m.o.i. of Y. pestis KIM5, KIM5-pEcLpxL, KIM5-pYtbLpxL and 1 m.o.i. of Salmonella typhimurium. Both pro-IL-1β (upper band) and mature IL-1β (lower band) are shown. Representative experiments out of 3–5 performed are shown. P values = *<0.05, **<0.01, ***<0.001. Also see Figure S1 and Table S1.

Figure 2. NLRP12 is involved in host resistance to attenuated Y. pestis

(A,B) Survival of C57Bl/6 (circles) and (A) Nlrp3^−/− (NLRP3 KO, squares) and (B) Nlrp12^−/− (NLRP12 KO, squares) mice infected s.c. with 500 CFU of KIM1001-pYtbLpxL. (C)_Survival of C57BL/6 (circles), NLRP12 KO (squares) or TLR4 KO (triangle) mice (n=10 of each) infected i.p. with 500 CFU of S. typhimurium, p<0.003 (NLRP12 KO or WT vs TLR4 KO). Also see Figure S2.

Figure 3. NLRP12 mediates Y. pestis-induced release IL-1β and IL-18

(A) IL-1β released from neutrophil-enriched peritoneal cells from C57BL/6 (black bars), NLRP12 KO (gray bars) and Casp1^−/− (Caspase-1 KO, white bars) mice. (B–I) IL-18 (B), IL-1β (C,E,F,G,H,I) and TNF (D,G) released from C57BL/6, NLRP12 KO, NLRP3 KO, ASC KO and caspase-1 KO BMDM (B–D); C57BL/6 and NLRP12 KO BMDM (E); C57BL/6 BMDM (F,G); or C57BL/6 and TLR4 KO BMDM (I). Infection with Yersinia strains occurred for 6 hours, with an addition of 50 µg/ml gentamicin to limit bacterial growth after 3 hours. Yersinia strains were added at 10 m.o.i., S. typhimurium at 1 m.o.i. Alum (130 µg/ml) stimulations (C,I) lasted 6 hours after priming for 3 hours with 10 ng/ml KIM5-YtbLpxL LPS. Shown are mean for triplicate cultures (with S.D.) in representative experiments out of 3 to 10 performed. P values = *<0.05, **<0.01, ***<0.001.Statistical comparisons are between wild type cells and multiple mutant cells (A-E,I), or between unstimulated and multiple bacterial strains (F). Also see Figure S3.

Figure 4. NLRP12 is necessary for optimal maturation of IL-1β and caspase-1 after infection with Y. pestis

(A-B,G) Q-PCR of (A) Nlrp12, (B) Nlrp3, or (G) Il1b from BMDM infected with 10 m.o.i. of Y. pestis KIM5 or KIM5-pYtbLpxL for (G) 4 hours or (A–B) 6 hours, with gentamicin addition after 3 hours. (C,E) FACS histograms showing active caspase-1 after FLICA staining with corresponding (D,F) percent positive cells of (C–D) bone marrow cells after 6 hour of challenge with 10 m.o.i. of Y. pestis strains (gentamicin added after 3 hrs) or (E–F) Ly6G or F4/80 positive splenocytes from mice infected with 500 CFU of KIM1001 or KIM1001-pYtbLpxL for 24 hours. Values from unstimulated cells are subtracted in D. LPS primed cells treated with nigericin (10 µM) served as a control (C,D). (H) Immunoblot for caspase-1 p10 or IL-1β p17 in supernatant or cell lysate from BMDM exposed for 10 hours to poly(dA:dT) (LPS primed as in Figure 3), KIM5 or KIM5-pYtbLpxL. Shown is a representative of 2 (E–G) or 3–5 (A–D, H, I) experiments. P values: *<0.05, **<0.01, ***<0.001. Also see Figure S4.

Figure 5. NLRP12 and IL-18 control infection with Y. pestis in vivo

(A) IL-18 in serum (ng/ml) of C57Bl/6 or NLRP12 KO mice infected i.v. with 500 CFU Y. pestis KIM1001 for 44 hours, normalized by bacterial loads (for each animal). n=5. (B–D) IL-18 or IL-1β in (B–C) spleen or (D) serum of WT or NLRP12 KO mice infected i.v. with 500 CFU KIM1001-pYtbLpxL for 44 hours. (B,C): uninfected mice: n=3, infected animals: n=8. (D): Uninfected mice: n=4), infected mice: n=5). (E) Spleen CFU of mice infected i.v. with either KIM1001 or KIM1001-pYtbLpxL (n=5). Horizontal lines indicate median values. (F) Histology of fixed H&E stained liver sections from (top) WT, (middle) NLRP12 KO, or (bottom) IL-18R KO mice infected i.v. with (left) KIM1001 or (right) KIM1001-pYtbLpxL for 44 hours. Stars represent bacterial clusters, arrows represent foci of inflammatory cells, primarily neutrophils. (G) Survival of C57Bl/6 (n=10), IL-18 KO and NLRP12 KO (n=8) mice with 10 CFU s.c. of KIM1001. (H) Spleen CFU or (I) spleen IL-1β of C57Bl/6 and NLRP3 KO infected for 44 hrs with 500 CFU i.v. of KIM1001-pYtbLpxL. Representative experiments out of 3 performed are shown. P values = *<0.05, **<0.001. Also see Figure S5.

Figure 6. NLRP12 induces IFNγ via IL-18 signaling

(A,B) Survival of mice, (A): C57Bl/6: n=10 and Ifnar1^−/− × Ifngr1^−/− (IFNαβR × IFNγR DKO): triangles, n=8, and (B) C57Bl/6: squares, n=10, IFNαβR KO: n=8, IFNγR KO: n=7, infected s.c. with 500 CFU of KIM1001-pYtbLpxL. (C–D) IFNγ in spleen homogenates from C57BL/6 and NLRP12 KO mice infected i.v. with 500 CFU of KIM1001-pYtbLpxL. Uninfected mice: n=3, infected mice C: n=8, D: n=5) Samples were harvested 46 hrs after infection. Horizontal lines indicate median values. Experiments shown are representative out of 3 performed. P values: *<0.05 **<0.001.