Cytosolic double-stranded RNA activates the NLRP3 inflammasome via MAVS-induced membrane permeabilization and K+ efflux

Abstract

The nucleotide-binding oligomerization domain-like receptor pyrin domain-containing 3 (Nlrp3) inflammasome plays an important role in inflammation by controlling the maturation and secretion of the cytokines IL-1β and IL-18 in response to multiple stimuli including pore-forming toxins, particulate matter, and ATP. Although the pathways activated by the latter stimuli lead to a decrease in intracellular K(+) concentration, which is required for inflammasome activation, the mechanism by which microbial RNA activates Nlrp3, remains poorly understood. In this study, we found that cytosolic poly(I:C), but not total RNA from healthy macrophages, macrophages undergoing pyroptosis, or mitochondrial RNA, induces caspase-1 activation and IL-1β release through the Nlrp3 inflammasome. Experiments with macrophages deficient in Tlr3, Myd88, or Trif, indicate that poly(I:C) induces Nlrp3 activation independently of TLR signaling. Further analyses revealed that the cytosolic sensors Rig-I and melanoma differentiation-associated gene 5 act redundantly via the common adaptor mitochondrial antiviral signaling (Mavs) to induce Nlrp3 activation in response to poly(I:C), but not ATP or nigericin. Mechanistically, Mavs triggered membrane permeabilization and K(+) efflux independently of the inflammasome which were required for poly(I:C)-induced Nlrp3 activation. We conclude that poly (I:C) activates the inflammasome through an Mavs-dependent surveillance pathway that converges into a common K(+) lowering step in the cytosol that is essential for the induction of Nlrp3 activation.

Figure 1

Cytosolic poly(I:C) induces Nlrp3 activation independent of TLR signaling. (A) BMDC derived from WT or Myd88^−/−/Ticam^−/− DKO mice were primed with poly(I:C) (10μg/ml) or TNF (100ng/ml) for 6 hours and then stimulated with ATP (5 mM) for 30 minutes or stimulated directly with poly(I:C) conjugated with lipofectamine for 6 hours. Extracts were prepared from cells plus culture supernatants and immunoblotted with an antibody detecting active caspase-1 (p20). (B) WT or Myd88^−/−/Ticam^−/− DKO BMDC were primed with LPS or TNF and stimulated with poly(I:C) conjugated, or not, with lipofectamine for 6 hours. IL-1β was measured in cell-free supernatants by ELISA. Values represent the mean of triplicate wells ± SD. (**) p<0.01. Results are representative of three independent experiments.

Figure 2

Poly(I:C) induces inflammasome activation via Mavs. (A) LPS-primed BMDM derived from WT, Ddx58^−/−, Ifih1^−/−, Pkr^−/−, Ticam^−/−, Ips1^−/− or Ticam^−/−/Ips1^−/− DKO mice were stimulated with high (pIC H) or low (pIC L) molecular weight poly(I:C) conjugated with lipofectamine for 6 hours or ATP or nigericin for 30 min. (A) IL-1β secretion was assessed by ELISA in cell free supernatant. Values represent the mean of triplicate wells ± SD. (*) p<0.01. (B-C), extracts were prepared from cells plus culture supernatants and immunoblotted with an antibody detecting active caspase-1 (p20). All results are representative of three independent experiments.

Figure 3

VSV and E. coli, but not silica or ATP, induce inflammasome activation via Mavs. (A C) LPS-primed BMDM or (D-E) DC derived from WT or Mavs−/− mice were stimulated with ATP (A) (1,25mM; 2,5mM; 5mM) for 30 minutes, or Silica (B) (100μg/ml; 200μg/ml, 400 μg/ml) for 6 hours, or poly(I:C) conjugated with lipofectamine for 6 hours or infected with E. coli or VSV for 24 hours (C-D). IL-1β secretion was assessed by ELISA in cell free supernatant. Values represent the mean of triplicate wells ± SD. (*) p<0.01. (A-D) Results are representative of three independent experiments.

Figure 4

dsRNA induces membrane permeabilization via Mavs. LPS-primed BMDM derived from WT, Ddx58^−/−, Ifih1^−/−, Pkr^−/−, Ticam^−/−, Mavs^−/− or Ticam^−/−/Mavs^−/− DKO (T/M DKO) mice were stimulated with poly(I:C) (10 μg/ml) conjugated with lipofectamine or infected with Salmonella. Membrane permeabilization was evaluated after 6 hours by analyzing the release of cytosolic LDH in the supernatant. Values represent the mean of triplicate wells ± SD. (*) p<0.01 Results are representative of three independent experiments.

Figure 5

dsRNA-induced membrane permeabilization is required for K+ ellux and inflammasome activation. Unprimed and LPS-primed BMDM were pretreated with NAC (30 mM) and stimulated with poly(I:C) (10 μg/ml) conjugated with lipofectamine for 6 hours (A-E) or with ATP (C-D) (5 mM) for 30 minutes. (A and E) Membrane permeabilization was evaluated by analyzing the release of cytosolic LDH in the supernatant. (B-C) intracellular K+ content was evaluated by ICP/MS. (D) Caspase-1 activation was evaluated by immunblotting with an antibody detecting active caspase-1 (p20). (E) IL-1β was evaluated by ELISA in cell free supernatant. (A-C and E) Values represent the mean of triplicate wells ± SD. (*) p<0.01 (A-E) Results are representative of three independent experiments.

Figure 6

Activation of the Nlrp3-inflammasome by poly(I:C) requires K+ efflux. A, LPS- primed BMDM were stimulated with high (pIC H) or low (pIC L) molecular weight poly(I:C) conjugated with lipofectamine for 6 hours (A-D) or ATP for 30 min (B-C), in buffer containing low (5mM) or high (70mM) K+ concentration (B-D). (A) Intracellular K+ was evaluated by ICP/MS, (B) IL-1β secretion was evaluated by ELISA in cell free supernatant. (C), caspase-1 activation was evaluated by immunblotting with an antibody detecting active caspase-1 (p20) and (D) membrane permeabiliziation was evaluated by analyzing the release of cytosolic LDH in the supernatant. (A, B and D). Values represent the mean of triplicate wells ± SD. (*) p<0.01. (A-D) Results are representative of three independent experiments.

Figure 7

Rig-I and Mda5 act redundantly via the common adaptor Mavs to induce NLRP3 activation. (A-D) LPS-primed BMDM derived from WT, Ddx58^−/−Ifih1^−/− DKO or Mavs^−/− mice were stimulated poly(I:C) conjugated, or not, with lipofectamine for 6 hours (A-D) (A) membrane permeabilization was evaluated by analyzing the release of cytosolic LDH in the supernatant (B) Intracellular K+ was evaluated by ICP/MS. (C) Caspase-1 activation was evaluated by immunoblotting with an antibody detecting active caspase-1 (p20). (D) IL-1β secretion was evaluated by ELISA in cell free supernatant. (A, B and D). Values represent the mean of triplicate wells ± SD. (*) p<0.01. (A-D) Results are representative of three independent experiments.