Cyclic-di-GMP and cyclic-di-AMP activate the NLRP3 inflammasome

Abstract

The cyclic dinucleotides 3'-5'diadenylate (c-diAMP) and 3'-5' diguanylate (c-diGMP) are important bacterial second messengers that have recently been shown to stimulate the secretion of type I Interferons (IFN-Is) through the c-diGMP-binding protein MPYS/STING. Here, we show that physiologically relevant levels of cyclic dinucleotides also stimulate a robust secretion of IL-1β through the NLRP3 inflammasome. Intriguingly, this response is independent of MPYS/STING. Consistent with most NLRP3 inflammasome activators, the response to c-diGMP is dependent on the mobilization of potassium and calcium ions. However, in contrast to other NLRP3 inflammasome activators, this response is not associated with significant changes in mitochondrial potential or the generation of mitochondrial reactive oxygen species. Thus, cyclic dinucleotides activate the NLRP3 inflammasome through a unique pathway that could have evolved to detect pervasive bacterial pathogen-associated molecular patterns associated with intracellular infections.

Figure 1

Cyclic dinucleotides activate the inflammasome in THP-1 cells and murine BMMs. (A) PMA-differentiated THP-1 cells (left panel) were treated with ATP (5 mM; 6 h) or transfected (Lipofectamine 2000; lpftm) with c-diGMP (cdG; 5 nmol; 6 h) or poly(dA:dT) (dA:dT; 5 μg/ml; 6 h). LPS-primed BMMs (right panel) were treated with ATP (5 mM; 6 h) or transfected (X-tremegene HP; Xtrm) with c-diGMP (10 nmol; 6 h), c-diAMP (10 nmol; 6 h), cGMP (10 nmol; 6 h) or GTP (10 nmol; 6 h). Supernatants were immunoblotted for immature IL-1β (pro-IL1β), active IL-1β (p17), immature caspase-1 (pro-Cas1) or active caspase-1 (Cas1, p20 and p10, respectively). (B) IL-1β levels were measured by ELISA (BD Biosciences) in 6 h supernatants of LPS-primed BMMs after the indicated treatment with c-diGMP or c-diAMP, as in A. (C) IL-1β levels were measured by ELISA in BMMs infected with IPTG-induced L. pneumophila strains (MOI 10; 6 h), ectopically expressing cdgS3, cdgS3B or cdgS4 [35]. Graphs present means±standard error from three independent experiments. BMMs, bone marrow-derived macrophages; c-diAMP, 3′-5′ diadenylate; c-diGMP, 3′-5′ diguanylate; ELISA, enzyme-linked immunosorbent assay; MOI, multiplicity of infection.

Figure 2

Cation mobilization during c-diGMP NLRP3 inflammasome activation.(A) IL-1β levels were measured by ELISA in LPS-primed BMMs treated with KCl (130 mM), Glibenclamide (Gli; 50 μM), 2-APB (100 μM) or U73122 (10 μM) 30 min before stimulation with ATP (5 mM; 30 min), MSU (100 μg/ml; 4 h) and silica (500 μg/ml, 6 h); or transfection with c-diGMP (10 nmol; 6 h), c-diAMP (10 nmol; 6 h) and poly(dA:dT) (5 μg/ml; 6 h); or infected with S. typhimurium (MOI 25; 6 h). (B) IFN-β expression was evaluated by Q–PCR in LPS-primed BMMs treated as in panel A. Results were normalized to GAPDH and reported as relative fold change. Graphs present means±standard error from three independent experiments. BMMs, bone marrow-derived macrophages; c-diAMP, 3′-5′ diadenylate; c-diGMP, 3′-5′ diguanylate; ELISA, enzyme-linked immunosorbent assay; GAPDH, Glyceraldehyde-3-phosphate dehydrogenase; IFN-β, interferon beta; MOI, multiplicity of infection; Q–PCR, quantitative polymerase chain reaction.

Figure 3

c-diGMP-stimulated inflammasome activation is dependent on NLRP3. (A) Secreted IL-1β and Cas1 levels were evaluated by immunoblotting in LPS-primed WT, NLRP3^−/−, NLRC4^−/− or ASC^−/− BMMs, as in Fig 1. (B) IFN-β and GAPDH expression were evaluated by PCR, in BMMs treated as in panel A. (C) Secreted IL-1β and caspase-1 levels were evaluated in PMA-differentiated THP-1 cells stably expressing control shRNA (shCtrl), shNLRP3, shASC or the re-expression of a knockdown resistant ASC cDNA (shASC^mut), as above. Data are representative of more than three independent experiments. ASC, apoptosis speck-like protein with CARD; BMMs, bone marrow-derived macrophages; Cas1, caspase-1; c-diGMP, 3′-5′ diguanylate; IFN-β, interferon beta; IL-1β, interleukin 1 beta; GAPDH, Glyceraldehyde-3-phosphate dehydrogenase; PCR, polymerase chain reaction; shctrl, control short hairpin RNA; WT, wild type.

Figure 4

c-diGMP-stimulated inflammasome activation is independent of MPYS/STING. (A) Secreted IL-1β and Cas1 were evaluated in LPS-primed WT* or MPYS/STING[−/−] BMMs, as in Fig 1. *(Note, the WT data (right panel) is a replicate of WT data presented in Fig 3A.) (B) IFN-β expression was evaluated by Q–PCR in WT or MPYS/STING[−/−] BMMs 6 h after transfection with c-diGMP (cdG; 10 nmol), poly(I:C) (I:C; 1 μg/ml) or poly(dA:dT) (dA:dT; 1 μg/ml) as in Fig 2. Results are reported as relative log fold change with respect to untreated controls and graphed as means±standard error from three independent experiments. (C) Peritoneal leukocytes were enumerated by FACS (see supplementary Fig S5A,B online) 15 hrs after IP injection of PBS or c-diGMP (cdG; 160 nmols). The scatter plot represents data from two independent experiments (n=6 mice for c-di-GMP and n=4 mice for PBS treatment). Statistical significance was evaluated by student’s t-test (Graphpad). BMMs, bone marrow-derived macrophages; Cas1, caspase-1; c-diGMP, 3′-5′ diguanylate; FACS, fluorescence-activated cell sorting; IP, intraperitoneal; IFN-β, interferon beta; IL-1β, interleukin 1 beta; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; WT, wild type.