Background/Purpose: African-American (AA) ethnicity is associated with a 3-fold higher risk of developing systemic lupus erythematosus (SLE). In addition, there is an increased risk of lupus nephritis (2-fold), high-risk histological features, and resistance to treatment. This may account for the increased mortality rate compared to Caucasian patients, especially in women. In Phase One of the Accelerating Medicines Partnership (AMP) study, we used single-cell RNA sequencing (ssRNA-Seq) on kidney biopsies from patients with active lupus nephritis to identify pathways that were differentially expressed in AA patients.

Methods: Single cells from renal biopsies obtained for clinical purpose for active nephritis were processed using CEL-Seq2. The bioinformatic pipeline to generate the expression levels of unique molecular identifiers (UMIs) from RNA reads has been previously described. We used canonical correlation analysis to identify common sources of variation between AAs and Caucasians. Cell clusters were identified using t-distributed stochastic neighbor embedding (t-SNE). Cluster identity was defined based on the presence of known cell lineage markers, enrichment of known gene set or using publicly available gene expression atlases. Next, we identified differentially expressed genes within each cluster with > 1.5-fold change in expression (p <0.05, Bonferroni). Finally, we applied Ingenuity Pathway Analysis (IPA) (QIAGEN Bioinformatics) to identify pathways of interest.

Results: Samples from 16 AA and 13 Caucasian patients were obtained. Of the 3829 sequenced cell libraries, we used 2358 which passed our quality filter for a total of 30155 UMIs. We identified 12 cell clusters: CD4, CD8, B cells, NK, monocytes, myeloid, NKT, distal tubule, proximal tubule, plasma cells, fibroblasts, and cell cycle. We identified 42 unique genes differentially expressed between AAs and Caucasians (Table 1). IPA identified a stronger type 1 interferon signature in AA patients, especially in CD4+ lymphocytes. In Caucasians, we identified selective expression of genes related to macrophage activation and that FKB5, potential mediator of the immunosuppressant effect of mycophenolate and rapamycin analogues, is preferentially expressed in fibroblasts.

Conclusion: AA lupus nephritis patients have a stronger Type I interferon gene signature in immune and renal cells. In Caucasian patients, infiltrating macrophages have an activated profile and fibroblasts express FK506 binding protein 5, suggesting a potential mechanism for their better response to immunosuppression. These results indicate that ethnicity may predict a response to both current and upcoming treatments, paving the way for a more personalized approach to treatment in lupus nephritis. Further work in Phase 2 of AMP will confirm and extend these findings.

Table 1.