Brain regional alternative splicing and clinical traits in Alzheimer’s disease

Abstract

AbstractBackgroundAlzheimer’s disease (AD) is one of the most complex diseases and is characterized by neuropathological features, including accumulation of beta‐amyloid plaques and tau neurofibrillary tangles. These neuropathological features exhibit heterogeneity across brain regions in different patients, leading to different symptoms. Dysregulation of alternative splicing (AS) has been observed to contribute to neuropathological features in AD; however, their molecular mechanisms via AS remain unclear. Therefore, we aimed to systemically investigate the impact of AS (i.e., exon skipping) on AD‐related neuropathological features (i.e., Braak stage and plaque density) and the functional connections of alternately spliced genes across multiple brain regions in AD patients.MethodWe analyzed RNA‐seq data generated by the MSBB project representing two brain regions: the frontal region (frontal pole; FP, n = 175, inferior frontal gyrus; IF, n = 87) and the temporal (parahippocampal gyrus; PH, n = 93 and superior temporal gyrus; ST, n = 101). The data were mapped to the hg19 reference genome with the STAR aligner and exon skipping rates were estimated as percent spliced in (PSI) by rMATs. Then, we performed a linear regression to evaluate clinical or pathogenic associations between exon PSI levels and the Braak stage and plaque density mean for each brain region, considering sex as a covariate.ResultWe identified 192 and 158 exon skipping (ES) associated with Braak stages in the frontal and temporal regions, respectively. In addition, 863 and 845 events were associated with plaque density. Notably, while the two regions shared only 20% (60 events) of Braak‐associated ES, they coincided in 90% (808 events) of plaque‐associated ES. Enrichment analysis of functional annotations found that frontal‐specific ES associated with Braak stage was enriched in interleukin signaling and the neurodegenerative‐related CDK5 pathway, while temporal‐specific ES was enriched in lipoprotein clearance functional pathways. For plaque density‐associated ES shared between regions, the most significantly enriched pathway was Notch4 signaling, represented for example by SOD2, in which skipping of the functional domain‐encoding exon 4 was associated with plaque mean in both regions.ConclusionThis study provides a landscape of brain region‐dependent AS events and of AD risk genes that contribute to neuropathological features in AD patients.