Gadolinium-based compounds induce NLRP3-dependent IL-1β production and peritoneal inflammation

Abstract

Objective: Nephrogenic systemic fibrosis (NSF) is a progressive fibrosing disorder that may develop in patients with chronic kidney disease after administration of gadolinium (Gd)-based contrast agents (GBCAs). In the setting of impaired renal clearance of GBCAs, Gd deposits in various tissues and fibrosis subsequently develops. However, the precise mechanism by which fibrosis occurs in NSF is incompletely understood. Because other profibrotic agents, such as silica or asbestos, activate the nucleotide-binding oligomerisation domain (NOD)-like receptor protein 3 (NLRP3) inflammasome and initiate interleukin (IL)-1β release with the subsequent development of fibrosis, we evaluated the effects of GBCAs on inflammasome activation.

Methods: Bone marrow derived macrophages from C57BL/6, Nlrp3(-/-) and Asc(-/-) mice were incubated with three Gd-containing compounds and IL-1β activation and secretion was detected by ELISA and western blot analysis. Inflammasome activation and regulation was investigated in IL-4- and interferon (IFN)γ-polarised macrophages by ELISA, quantitative real time (qRT)-PCR and NanoString nCounter analysis. Furthermore, C57BL/6 and Nlrp3(-/-)mice were intraperitoneally injected with GBCA and recruitment of inflammatory cells to the peritoneum was analysed by fluorescence-activated cell sorting (FACS).

Results: Free Gd and GBCAs activate the NLRP3 inflammasome and induce IL-1β secretion in vitro. Gd-diethylenetriaminepentaacetic acid also induces the recruitment of neutrophils and inflammatory monocytes to the peritoneum in vivo. Gd activated IL-4-polarised macrophages more effectively than IFNγ-polarised macrophages, which preferentially expressed genes known to downregulate inflammasome activity.

Conclusions: These data suggest that Gd released from GBCAs triggers a NLRP3 inflammasome-dependent inflammatory response that leads to fibrosis in an appropriate clinical setting. The preferential activation of IL-4-differentiated macrophages is consistent with the predominantly fibrotic presentation of NSF.

Keywords: Cytokines; Inflammation; Magnetic Resonance Imaging.

Figure 1

Induction of processed IL-1β by GBCA is dependent on the NLRP3 inflammasome. (A) IL-1β secretion by BMDM obtained from WT mice after LPS priming (20 ng/ml) and stimulation with the indicated concentrations of Gd containing compounds, compared to stimulation by nigericin (1hr) and poly(dA:dT) (6 hrs), as detected by ELISA. (B) Western blot analysis of supernatants (SN) and cell lysates (CL) from BMDM described above, analyzed for IL-1β and for β-actin, respectively. (C) IL-1β secretion of BMDM obtained from Nlrp3^−/− and Asc^−/− mice. Omniscan^®, Gd-DTPA and Gd-Cl₃ were used at 10mM, 500 μM and 2.5 μM, respectively. Values are mean ± SEM of 4 independent experiments, each with duplicates of BMDM obtained from one mouse.

Figure 2

Effect of macrophage polarization on GBCA-induced IL-1β production. (A) Expression of inducible nitric oxide synthase (iNos) and arginase 1 (arg1) in IFNγ- and IL-4-polarized macrophages after LPS priming (20 ng/ml for 4 hrs) expressed relative to the housekeeping gene guanine nucleotide-binding protein subunit beta-2-like 1 (GNB2L1), as measured by qRT-PCR. Values are mean ± SEM of 6 independent experiments. (B–C) IL-1β secretion of IFNγ-polarized (B) and IL-4-polarized (C) cells after LPS priming (20 ng/ml) and incubation with Gd containing compounds, poly(dA:dT) or silica (200 μg/ml) for 6 hrs or nigericin for 1 hr. Values are mean ± SEM of 3 independent experiments, each with duplicates of BMDM obtained from one mouse. Statistical significance: ★= p<0.05; ★★= p<0.01; ★★★=p<0.001.

Figure 3

Gene expression profiles of IFNγ- and IL-4-polarized BMDM. mRNA expression of inflammasome regulating genes detected with the NanoString nCounter in untreated and LPS primed (20 ng/ml) macrophages as well as IFNγ- and IL-4-polarized BMDM obtained as described in the methods. Values are mean ± SEM of 3 independent experiments, each using BMDM obtained from different mice. Statistical significance: ★= p<0.05; ★★= p<0.01; ★★★=p<0.001.

Figure 4

Gd-DTPA induced peritoneal inflammation. WT and Nlrp3^−/− mice were injected with either PBS or 500 μM Gd-DPTA in PBS and peritoneal exudate cells were collected 24 hrs later. CD11b⁺ cells were further stained for Ly6C and Ly6G expression to identify inflammatory monocytes (Ly6C⁺ Ly6G^int) and granulocytes (Ly6C⁺ Ly6G⁺). (A) Representative FACS plots. (B) Bar graphs displaying mean ± SEM for compiled data from 3 experiments involving WT (n=8) and Nlrp3^−/−(n=7) mice. Statistical significance: ★= p<0.05; ★★= p<0.01; ★★★=p<0.001.