Background/Purpose: Systemic lupus erythematosus (SLE) is a multi-organ autoimmune disorder with a prominent genetic component. Evidence has shown that individuals of African-Ancestry (AA) experience the disease more severely and with an increased co-morbidity burden compared to European-Ancestry (EA) populations. However, the relationship between genetics, molecular pathways, and disease severity has not been fully delineated. Here, we examined AA and EA SLE-associated single nucleotide polymorphisms (SNPs) and linked them via expression quantitative trait loci (eQTL) across multiple tissues to genes with altered expression (E-Genes). Putative EA and AA E-Gene signatures were coupled with SLE differential expression (DE) datasets and upstream regulators to map candidate molecular pathways. Together, these genetic and gene expression analyses enable a better understanding of how the identified SNPs could contribute to aberrant immune function as well as the influence of ancestry on the genetic basis of SLE.

Methods: A previous SLE Immunochip study (Langefeld et al., 2017) identified SNPs significantly associated with SLE in AA (2,970 cases; 2,452 controls) and EA (6,748 cases; 11,516 controls) cohorts. eQTL mapping identified E-Genes from SLE SNPs and their ancestry-specific SNP proxies (based on linkage disequilibrium) via the GTEx database. For both ancestral groups, E-Gene lists were examined for the significant enrichment of gene ontogeny (GO) terms, canonical IPA^® (Qiagen) pathways and BIG-C™ categories. Next, we analyzed the gene expression profiles of predicted E-Genes across multiple SLE DE datasets, including those from blood and multiple tissues. Differential expressed genes (DEGs) were identified and subjected to pathway analysis with IPA^®, clustering using MCODE and visualization in Cytoscape with the ClusterMaker2 plugin. Drug candidates targeting E-Genes, DEGs and upstream regulators (UPRs) were identified using CLUE, IPA^® and STITCH.

Results: A total of 908 Immunochip SNPs were mapped to 252 eQTLs and coupled to 760 E-Genes (207 in EAs, 30 in AAs, 523 shared; Figure 1A). Shared E-Genes were highly enriched in interferon signaling, whereas EA E-Genes were associated with nucleotide degradation and AA E-Genes were linked to multiple biosynthesis and intracellular signaling pathways (e.g., retinol biosynthesis and AMPK signaling). Protein-protein interaction (PPI) networks of clustered EA, AA and shared E-Genes illustrate the high degree of ancestral overlap evident within each E-Gene set (Figure 1B). Clustering analysis of all DE E-Genes and IPA-predicted UPRs highlight disease-associated pathways that are both shared and ancestry-specific. Drug candidate comparison identified a total of 115 drugs targeting EA, AA and shared E-Genes and their molecular pathways.

Conclusion: Using a bioinformatics-based approach that utilizes pathway analysis and gene expression data, we were able to discover novel ancestry-dependent and ancestry-agnostic candidate causal targets in SLE and couple those with drug discovery tools to identify new therapies with the potential to impact disease processes within and across specific populations.  

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-integration-of-genetic-data-molecular-pathway-analysis-and-differential-expression-to-delineate-the-impact-of-ancestral-differences-on-lupus/