Gout Risk Allele Regulating IRF5 Expression Is Associated with Enhanced IL-1β Production in Response to Palmitate and Monosodium Urate Crystals

Abstract

Interferon Regulatory Factor 5 plays an important role in the regulation of innate immune responses by amplifying the Nuclear Factor κB response, which is critical in gout inflammation. Furthermore, the rs4728141 polymorphism C allele was associated with both increased IRF5 expression and susceptibility to gout. We examine the association between rs4728141 and cytokine production in response to various Toll-Like Receptor ligands and describe the transcriptomic and proteomic changes observed in patients with gout and controls in relation to this polymorphism. We examine the transcriptome of freshly isolated peripheral blood mononuclear cells (PBMCs) from 93 normouricemic donors and 63 gout patients as well as serum inflammatory proteome in 197 control and 195 gout samples. Stimulation experiments of freshly isolated human PBMCs were performed over 24 h, followed by RNA-sequencing in gout patients and cytokine production measurement by ELISA in normouricemic donors and gout patients. The rs4728141 C allele was associated with increased IL-1β expression in unstimulated PBMCs of controls, but not in gout. No association between the polymorphism and serum inflammatory proteome was found. As expected, an increased IRF5 expression was observed in stimulated PBMCs of rs4728141 C allele carriers in response to several stimulations. Interestingly, IL-1β production was specifically enhanced in association to the rs4728141 C allele when cells were stimulated with palmitate with or without monosodium urate crystals. This pattern of cytokine production shows a functional impact of rs4728141 in gout through altered IL-1β production.

Keywords: gout; inflammation; interferon regulatory factor 5; interleukin-1.

Figure 1

Gene expression of key inflammatory cytokines in unstimulated PBMCs in relation to rs4728141. Expression of IRF5 (a), IL-1B (b), TNF (c), and IL1RN (d) in freshly isolated PBMCs from normouricemic controls (n = 93) and gout patients (n = 63). The values represent log2 normalized gene counts. The p value is obtained after performing a linear regression, and individual groups were compared using the ANOVA and Tukey test. Statistical significance: * p <= 0.05; NS. p > 0.05.

Figure 2

Changes in inflammatory proteome associated with rs4728141. A total of 73 inflammatory proteins in normouricemic control (a) and gout (b) groups were examined (n = 197 and 195). The genotypes were assigned 0, 1, and 2 values, and linear regressions were performed for each protein. The X-axis represents the coefficient estimates and the Y-axis –log₁₀ of the nominal p values, with the line drawn at p = 0.05. No statistically significant changes were observed after multiple testing correction.

Figure 3

IRF5 and IL-1β expression after in vitro 24 h PBMC stimulation. Cells were stimulated with palmitate (50 μM), with and without the presence of monosodium urate crystals (300 μg/mL), LPS (10 ng/mL), E. coli (10⁶/mL), S. aureus (10⁶/mL), M. tuberculosis lysate (5 μg/mL), B. burgdorferi (10⁶/mL), C. albicans (10⁶/mL), Poly(I:C) (10 μg/mL), and CpG (1 μg/mL), followed by transcriptomic analysis. (a) Row scaled heatmap of IRF5 expression colored by means. Linear regressions were performed to identify if the risk allele is associated with changes in the cytokine production, and the result is indicated next to the stimuli names. In statistically significant associations, heterozygous and homozygous risk allele groups were compared to the homozygous wild-type group (ANOVA and Tukey correction). The strongest correlation was found in the stimulations with palmitate (b), and no changes in IL-1β expression were found (c). Statistical significance: * p <= 0.05; ** p <= 0.01; NS. p > 0.05.

Figure 4

IL-1β cytokine production after in vitro 24 h PBMC stimulation. Cells were stimulated with palmitate (50 μM), with and without the presence of monosodium urate crystals (300 μg/mL), LPS (10 ng/mL), E. coli (10⁶/mL), S. aureus (10⁶/mL), M. tuberculosis lysate (5 μg/mL), B. burgdorferi (10⁶/mL), and C. albicans (10⁶/mL), followed by cytokine measurements by ELISA in the supernatants. Row scaled heatmaps colored by median. Samples were split into 2 groups—normouricemic controls (a) and gout (b). Linear regressions were performed to identify if the risk allele is associated with changes in the cytokine production, and the result is indicated next to the stimuli names. In statistically significant associations, heterozygous and homozygous risk allele groups were compared to the homozygous wild-type group (ANOVA and Tukey correction). Boxplots with the key changes represented by palmitate stimulations, with and without MSU, are included separately (c,d). A ratio of IL-1β production between palmitate, with and without MSU, was calculated for every patient followed by a similar comparison (e). Statistical significance: * p <= 0.05; ** p <= 0.01; NS. p > 0.05.