Background/Purpose: Identifying molecular signatures associated with anti-citrullinated protein antibody (ACPA) positive individuals who are ‘at-risk’ for future RA (At-Risk) and early RA (ERA) requires comprehensive datasets to integrate multi-omics platforms to define personalized mechanisms. In this study, we measured single cell epigenomic and transcriptomic profiles from peripheral blood mononuclear cells (PBMCs) of At-Risk individuals, ERA patients and controls (CON). We performed a cross-sectional analysis using a novel bioinformatic tool to integrate the large-scale multi-omics single cell data.

Methods: Three cohorts were studied including 26 At-Risk with elevated ACPA, 6 ERA, and 35 controls (CON). scRNA-seq and scATAC-seq data were paired for each participant (pt). Seurat and ArchR packages identified co-embedded clusters between scRNA and scATAC cells for each sample, which were annotated to identify cell types. Each cluster was processed through the Taiji pipeline (Nat Commun 2022;13:6221) to create regulatory networks and PageRanks for transcription factors (TFs). Groups were identified using K-means clustering with Pearson correlation as the distance metric. Chi-square test was applied and K-means group-specific TFs were identified by Wilcoxon test.

Results: 1613 clusters from 67 samples spanning 21 cell types were identified and grouped into 5 K-means groups displaying distinct TF regulatory patterns. 4 groups were enriched for individual cell lineages, such as CD4 and CD8 T cells, B cells and monocytes. Group 2 (G2) was multilineage and was significantly enriched in the At-Risk and ERA (8.2 clusters/pt) compared to CON (5.5 clusters/pt) (p< 0.001). 344 TFs were G2-specific, such as embryonic development TFs SP7, FOXL2, and TFAP2C (p< 0.001 each). Reactome pathway analysis in G2 showed enrichment in RUNX2, SUMOylation, and YAP1 pathways, each of which have been previously implicated in RA. Remarkably, the G2 signatures were shared by multiple lineages in individual At-Risk and ERA participants. CD4 T naïve, CD4 TCM, and CD8 T naïve cells had greatest enrichment (47% vs 24% of total clusters in G2, p< 0.0001; 38% vs 22%, p< 0.01; 55% vs 27%, p< 0.05, respectively for At-Risk/ERA vs CON). In many cases, more than one cell lineage displayed the signature in an individual participant. Of interest, MAIT cells with the TF signature were only found in CON (49% vs. 0% in At-Risk/ERA, p=0.01). Figure 1 shows a heatmap illustrating the cell lineages with the At-Risk/ERA signature.

Conclusion: Distinctive TF profiles are enriched in PBMCs from ACPA+ At-Risk and ERA compared to controls, especially in CD8 and CD4 T cells. These TFs involve pathogenic pathways that have been identified in RA. Furthermore, PageRanks and pathways were similar between At-Risk and ERA suggesting that these mechanisms antedate classifiable disease. Moreover, the cell lineages with these TFs and pathways varied between individual participants, suggesting common mechanisms can occur across multiple cell types. These findings could explain the diversity of clinical responses in RA to targeted therapies because each patient can have an unique combination of pathogenic cell types displaying the RA TF signature.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/integrated-single-cell-multi-omics-analysis-in-at-risk-for-future-rheumatoid-arthritis-ra-and-early-ra-reveals-shared-transcription-factor-profiles-in-multiple-cell-lineages/