Characterization of Genetic Landscape and Novel Inflammatory Biomarkers in Patients With Adult-Onset Still's Disease

Abstract

Objective: Adult-onset Still disease (AOSD) is a systemic autoinflammatory disorder (AID) of unknown etiology. Genetic studies have been limited. Here, we conducted detailed genetic and inflammatory biomarker analysis of a large cohort with AOSD to investigate the underlying pathology and identify novel targets for potential treatment.

Methods: We investigated AOSD cases (n = 60) for rare germline and somatic variants using whole exome sequencing with virtual gene panels. Transcriptome profiles were investigated by bulk RNA sequencing whole blood. Cytokine profiling was performed on an extended patient cohort (n = 106) alongside measurements of NLRP3 inflammasome activation using a custom assay and type I interferon (IFN) score using a novel method.

Results: We observed higher than expected frequencies of rare germline variants associated with monogenic AIDs in AOSD cases (AOSD 38.4% vs healthy controls [HCs] 20.4%) and earlier onset of putative somatic variants associated with clonal hematopoiesis of indeterminate potential. Transcriptome profiling revealed a positive correlation between Still Activity Score and gene expression associated with the innate immune system. ASC/NLRP3 specks levels and type I IFN scores were significantly elevated in AOSD cases compared with HCs (P = 0.0001 and 0.0015, respectively), in addition to several cytokines: interleukin (IL)-6 (P < 0.0001), IL-10 (P < 0.0075), IL-12p70 (P = 0.0005), IL-18 (P < 0.0001), IL-23 (P < 0.0001), IFN-α2 (P = 0.0009), and IFNγ (P = 0.0002).

Conclusion: Our study shows considerable genetic complexity within AOSD and demonstrates the potential utility of the ASC/NLRP3 specks assay for disease stratification and targeted treatment. The enriched genetic variants identified may not by themselves be sufficient to cause disease, but may contribute to a polygenic model for AOSD.

Figure 1

Rare germline variant analysis. (A) The number of variants identified in each panel (CHIP‐associated, autoinflammation, and type I interferonopathy) was normalized to their corresponding gene panel size (gene count). Bar charts for each segment illustrate the proportion of AOSD and HC cohorts (%) carrying a rare germline variant in each gene panel. (B) The frequency of cases carrying rare variants across genes in the autoinflammation gene panel in each cohort. (C) Summary of the rare germline variants shared by more than one AOSD case, which are color‐coded to reflect the associated gene panel: CHIP (brown), autoinflammation (orange), type I interferonopathy (blue). (D) The proportion of each cohort (%) carrying multiple variants across all gene panels. AOSD, adult‐onset Still disease; CHIP, chromatin immunoprecipitation; HC, healthy control.

Figure 2

Somatic variant analysis in AOSD. Variants are color coded according to the variant calling tool used for identification (black: HaplotypeCaller [putative germline]; red: Mutect2 [putative somatic]; purple: found by both HaplotypeCaller and Mutect2 [uncertain origin]). (A) An overview of VAF across all putative somatic variants identified by Mutect2. Genes located on an X chromosome are highlighted *; biological males with high VAF % on these genes are annotated with #. Somatic variants previously reported in COSMIC are presented as stars annotated with their COSV numbers. (B) Mutational landscape summary of all putative somatic and germline variants identified in AOSD cases within the preselected gene panels (y‐axis gene panel: CHIP = dark red; autoinflammation = orange; type I interferonopathy = blue) in age order (x‐axis). (C) The distribution of putative somatic variants identified in CHIP‐associated genes across each age group. AOSD, adult‐onset Still disease; CHIP, clonal hematopoiesis of indeterminant potential; COSMIC, Catalog of Somatic Mutations in Cancer; COSV, genomic mutation identifier; VAF, variant allele fractions.

Figure 3

Transcriptome analysis of AOSD cases: (A) Principal component analysis plot illustrating the relationship between AOSD cases and HCs. (B) Volcano plot highlighting the log2 fold change and ‐log10 adjusted P value of 2,830 differentially expressed genes in AOSD cases. (C) Pathway enrichment analysis performed through Enrichr using terms from the Reactome pathway database, enriched pathway terms are arranged in order of P value. (D) The relative expression of 170 genes shared among the top three pathways are displayed in a heatmap. The gene expression profiles of AOSD (purple) and healthy control (green) samples are grouped into hierarchical clusters and the average SAS of each cluster was recorded. Linear regression analysis showed significant positive correlation between SAS scores and increasing gene expression. AOSD, adult‐onset Still disease; HC, healthy control; SAS, Still Activity Score.

Figure 4

ASC/NLRP3 specks and IFN activity levels in AOSD and disease control cohorts: (A) ASC/NLRP3 specks in healthy controls (HC, n= 30) compared with AOSD cohort subsets (AOSD#1 (n = 39), AOSD#2 (n = 30) and AOSD#3 (n = 34), sJIA (n = 12), SchS (n = 10), CAPS (n = 11) and FMF (n = 31) cohorts. The weighted bars represent median values. (B) Scatterplot comparing ASC/NLRP3 specks to CRP (mg/L) taken from all disease cohorts. (C) ASC/NLRP3 specks comparisons within the AOSD cases, categorized by mild (0–2), moderate (3–4) and severe (5–7) Still Activity Score (SAS). (D) Change in ASC/NLRP3 specks in AOSD cases following treatment with canakinumab or placebo. (E) Boxplots comparing IFN scores (Sera) between HCs and cases from each disease cohort. (F) Comparison of IFN scores calculated from sera against IFN scores calculated from the gene expression of interferon response genes (geomeans/10), analyzed by linear regression. Statistically significant pairwise comparisons are annotated with * =P < 0.05; ** =P < 0.01; *** =P < 0.001; **** =P < 0.0001. AOSD, adult‐onset Still disease; ASC, antibody‐secreting cell; CAPS, cryopyrin‐associated periodic syndrome; CRP, C‐reactive protein; FMF, familial Mediterranean fever; HC, healthy control; IFN, interferon; SchS, Schnitzler syndrome; sJIA, systemic juvenile idiopathic arthritis.

Figure 5

Cytokine profile in patients with AOSD, disease, and healthy controls: The serum cytokine levels for 13 targets are displayed on dot plots, comparing healthy controls to AOSD, sJIA, SchS, CAPS, and FMF cases. AOSD cases are highlighted in red, weighted bars represent median values, and statistically significant pairwise comparisons are annotated with asterisks. * =P < 0.05; ** =P < 0.01; *** =P < 0.001; **** =P < 0.0001. AOSD, adult‐onset Still disease; CAPS, cryopyrin‐associated periodic syndrome; FMF, familial Mediterranean fever; HC, healthy control; IFN, type I interferon; IL, interleukin; SchS, Schnitzler syndrome; sJIA, systemic juvenile idiopathic arthritis.

Figure 6

Correlation between rare genetic variant counts against biomarker levels: Bar charts (median, IQR) illustrating (A) ASC/NLRP3 specks; (B) IL‐1β; (C) IL‐6; and (D) IL‐18 levels (pg/mL) in AOSD cases carrying 0, 1, 2, and 3+ variants (germline and/or putative somatic) within our preselected gene panels. (E) IFN scores in AOSD cases with and without a variant in the type I interferonopathy gene panel. AOSD, adult‐onset Still disease; ASC, antibody‐secreting cell; IFN, interferon; IL, interleukin; IQR, interquartile range.