Abstract
Background
Genetic risk factors for Alzheimer disease (AD) demonstrate distinct effects across diverse ancestral populations. The ancestral heterogeneity (admixture) of Caribbean Hispanics from Puerto Rico (PR), makes studies of the PR population important in the discovery of ancestry‐specific factors in AD. To expand ongoing genomic investigations of AD in PR individuals, it is necessary to characterize functional downstream effects by studying gene expression and regulation. Here we characterized the differences in gene expression, splicing, and RNA editing of the protein coding transcriptome from peripheral blood in PR individuals. to identify case vs control differential expression, splicing, and RNA editing in this diverse population.
Method
Poly‐A selected RNA was from peripheral whole blood of 76 PR individuals over the age of 65 (39 AD, 37 cognitively normal controls) was sequenced and analyzed with a standard bioinformatics pipeline. Differential expression between PR cases and controls was calculated using DESeq2, alternative splicing using LeafCutter software, and RNA editing with REDITools and linear models. All analyses were adjusted for sex, age, and sequencing coverage. For each analysis, pathway enrichment analysis of Gene Ontology Biological Processes and KEGG gene sets were used to identify underlying biological pathways.
Result
A total of 761 genes (518 up‐regulated, 243 down‐regulated) were differentially expressed between PR AD and controls (adjusted p ≤ 0.05). At the transcript level, 561 genes had a significant (FDR ≤ 0.05) differential splicing event. We also identified 35,246 total RNA editing sites. While there was no significant difference globally, 422 sites in 159 genes showed nominally significant editing difference (p ≤ 0.05). These genes, isoforms, and RNA editing sites overlap little with previous investigations of the transcriptomes of Non‐Hispanic Whites and African‐American AD with only a few genes in common. However, pathway enrichment across all three PR transcriptomic analyses consistently reveals differences in both the adaptive and innate immune response pathways, consistent with other ancestries.
Conclusion
Transcriptomic analyses of diverse populations in AD, shows stark divergence at the single gene level. However, the convergence on immune molecular pathways suggest shared underlying disease etiology and the possibility of broad therapeutic options.