National Institutes of Health Washington, United States
Disclosure information not submitted.
Background/Purpose: Systemic lupus erythematosus (SLE) is an autoimmune disease that disproportionately affects individuals of Amerindian ancestry and those who self-identify as Hispanic. The term “Hispanic” encompasses diverse populations, genetic backgrounds, and environments, making it challenging to understand the drivers of disease disparities. Since DNA methylation (DNAm) is shaped by genetic and environmental factors, it may help disentangle these influences. We aim to assess whether DNAm differs between genetically similar SLE individuals living in Peru and the U.S. Methods: We utilized whole blood genotype and methylation data from SLE cohorts in the U.S. (n=1,671) and Peru (n=74). We used principal component analysis (PCA), including genetic data from the 1000 Genomes and Peruvian Genome Project, to identify genetically similar individuals.We selected individuals within 1.75 standard deviations of the Peruvian population centroid based on the first three principal components (PCs). DNAm was profiled using the Illumina EPIC v2.0 array. After quality control and normalization, 801,370 CpG sites were retained. We linearly modeled differential methylation by country (U.S. vs. Peru) using limma, adjusting for age, disease duration, estimated blood cell proportions, and genetic PCs 1–3. Pathway analyses was performed using missMethyl. Results: We identified genetically similar Peruvian (n=64) and U.S. individuals (n=86). Results showed 13,736 CpG sites differentially methylated by country at a false discovery rate (q) < 0.05 (Figure 1). Of these, 870 CpGs had an average absolute methylation difference ≥3% between the two groups. The sites with the largest differences in methylation between U.S and Peru SLE cases were located in genes involved in immune function (Table 1). In particular, a CpG site within PSCA had an average of 19% lower methylation in U.S. compared to Peruvian samples. This gene is linked to immune cell infiltration and may reflect cohort-specific differences in inflammatory regulation. Another CpG site within NMI had an average of 11% less methylation in U.S. samples. NMI enhances STAT-mediated transcription in response to cytokines such as IL-2 and IFN-γ, both critical to SLE pathogenesis. Additionally, a Cpg within TXNIP was 7% more methylated in Peruvian samples. TXNIP is a vitamin D–responsive gene involved in oxidative stress and immune signaling via JAK-STAT and PI3K/Akt pathways. Conclusion: As an epigenetic mechanism shaped by genetic and environmental factors, DNAm offers insights into population-level disease variation. Our findings suggest that geographic context is associated with epigenetic variation in SLE when accounting for similar genetic ancestry. The most differentially methylated CpGs were in genes involved in immune regulation, pointing to geographically distinct immune responses potentially shaped by environmental exposures. Although the precise drivers are uncertain, these differences may reflect an interplay of social, cultural, and environmental factors. Understanding how these shape epigenetic patterns may offer new insights into SLE pathogenesis across populations.
