TET2-mutant clonal hematopoiesis and risk of gout

Abstract

Gout is a common inflammatory arthritis caused by precipitation of monosodium urate (MSU) crystals in individuals with hyperuricemia. Acute flares are accompanied by secretion of proinflammatory cytokines, including interleukin-1β (IL-1β). Clonal hematopoiesis of indeterminate potential (CHIP) is an age-related condition predisposing to hematologic cancers and cardiovascular disease. CHIP is associated with elevated IL-1β, thus we investigated CHIP as a risk factor for gout. To test the clinical association between CHIP and gout, we analyzed whole exome sequencing data from 177 824 individuals in the MGB Biobank (MGBB) and UK Biobank (UKB). In both cohorts, the frequency of gout was higher among individuals with CHIP than without CHIP (MGBB, CHIP with variant allele fraction [VAF] ≥2%: odds ratio [OR], 1.69; 95% CI, 1.09-2.61; P = .0189; UKB, CHIP with VAF ≥10%: OR, 1.25; 95% CI, 1.05-1.50; P = .0133). Moreover, individuals with CHIP and a VAF ≥10% had an increased risk of incident gout (UKB: hazard ratio [HR], 1.28; 95% CI, 1.06-1.55; P = .0107). In murine models of gout pathogenesis, animals with Tet2 knockout hematopoietic cells had exaggerated IL-1β secretion and paw edema upon administration of MSU crystals. Tet2 knockout macrophages elaborated higher levels of IL-1β in response to MSU crystals in vitro, which was ameliorated through genetic and pharmacologic Nlrp3 inflammasome inhibition. These studies show that TET2-mutant CHIP is associated with an increased risk of gout in humans and that MSU crystals lead to elevated IL-1β levels in Tet2 knockout murine models. We identify CHIP as an amplifier of NLRP3-dependent inflammatory responses to MSU crystals in patients with gout.

Graphical abstract

Figure 1.

CHIP is associated with increased risk of gout. (A) Association between CHIP and gout in the MGBB. Generalized linear model adjusting for age deciles and sex was used to calculate OR and 95% CI. Individuals without CHIP were used as reference group. (B) Association between CHIP and gout in the UKB. Generalized linear model adjusting for age deciles, sex, WBC, eGFR, and BMI was used to calculate OR and 95% CI. Individuals without CHIP were used as reference group. (C) Cumulative incidence of gout in individuals with and without CHIP in UKB. Individuals were identified as events at the time of gout diagnosis and censored at the end of follow-up, at the time of death, or at the time of hematologic malignancy diagnosis. (D) Cumulative incidence of gout among individuals without CHIP and CHIP with VAF <10% in UKB. Individuals were identified as events at the time of gout diagnosis and censored at the end of follow-up, at the time of death, or at the time of hematologic malignancy diagnosis. (E) Cumulative incidence of gout among individuals without CHIP and CHIP with VAF ≥10% in UKB. Individuals were identified as events at the time of gout diagnosis and censored at the end of follow-up, at the time of death, or at the time of hematologic malignancy diagnosis. (F) Forest plot of HR for the association between CHIP and gout in the UKB. Cox proportional hazards model adjusting for age deciles, sex, WBC, eGFR, and BMI was used to calculate HR and 95% CI. Individuals without CHIP were used as reference group.

Figure 2.

CHIP increases risk of gout in the presence of elevated serum urate levels. (A) Association between serum urate levels and gout diagnosis in males and females. P values were computed using Wilcoxon rank-sum test. (B) Forest plot of HR for the association between CHIP and gout. This analysis included individuals with serum urate levels above the median of UKB. Cox proportional hazards model adjusting for age deciles, sex, WBC, eGFR, and BMI was used to calculate HR and 95% confidence intervals (95% CI). Individuals without CHIP were used as reference group.

Figure 3.

Loss of Tet2 exacerbates MSU crystal-induced inflammatory phenotype in murine models. (A) Serum cytokine array of Tet2-deficient (Tet2 KO, n = 5) vs WT (n = 5) mice after 6 hours of treatment with PBS (left plot) and MSU crystals (right plot). Blue dots highlight cytokines with P < .05 using 2-sample Student t test; red dots highlight cytokines with adjustment of false discovery rate (FDR < 0.1). (B) Flow cytometry analysis of peripheral blood and peritoneal fluid from WT and Tet2 KO mice after 6 hours of treatment with PBS and MSU crystals. Two-sample Student t test was used to compare the fraction of CD11b⁺ cells between MSU-treated WT and Tet2 KO animals. (C) Paw edema elicited by subcutaneous injection of MSU crystals into the foot pads of WT (n = 10) and Tet2 KO (n = 10) mice. P values were calculated using 2-way analysis of variance. (D) Representative hematoxylin and eosin stained paw cross sections of WT and Tet2 KO mice treated with PBS and MSU crystals shows increased cellular infiltrate (40× magnification).

Figure 4.

Loss of Tet2 augments MSU crystal-induced secretion of IL-1β in macrophages. (A) Time-course analysis of IL-1β levels in supernatant of WT and Tet2 KO BMDMs treated with an MSU crystal dose of 100 µg/mL. P values were obtained using 2-sample Student t test. (B) Cytokine array in supernatant of Tet2 KO vs WT BMDMs after 6 hours of treatment with MSU crystals. Red dots highlight cytokines with FDR < 0.1. (C) KEGG pathway enrichment analysis of differentially expressed genes in Tet2 KO vs WT BMDMs treated with MSU crystals. (D) Heatmap of selected differentially expressed genes in inflammatory pathways from Tet2 KO vs WT BMDMs after administration of MSU crystals. (E) Dose-response analysis of IL-1β levels in supernatant of MSU crystal-treated BMDMs obtained from WT, Tet2 KO, Nlrp3 KO, and Tet2 KO+Nlrp3 KO mice. P values were obtained using 2-sample Student t test. (F) Dose-response analysis of IL-1β levels in supernatant of MSU crystal-treated BMDMs incubated with MCC950. BMDMs were obtained from WT and Tet2 KO mice. P values were obtained using 2-sample Student t test.