# Distinct Regulatory Genomic Architectures Distinguish Early-Onset from Late-Onset Alzheimer’s Disease

## TL;DR

This study finds that early-onset and late-onset Alzheimer’s disease have different genetic regulatory patterns, with early-onset being more focused and late-onset more widespread.

## Contribution

The study reveals distinct regulatory genomic architectures for EOAD and LOAD using a multi-tissue integrative approach.

## Key findings

- EOAD variants show a focal regulatory profile, affecting only GSE1 in the dorsolateral prefrontal cortex.
- LOAD variants have broad, multi-tissue effects involving genes like APOE and APH1B.
- LOAD-related genes are linked to amyloid processing and Notch signaling, while EOAD genes relate to chromatin and epigenetic regulation.

## Abstract

Background/Objectives: Alzheimer’s disease (AD) exhibits marked genetic heterogeneity between early-onset (EOAD) and late-onset (LOAD) forms. EOAD is typically associated with highly penetrant variants, whereas LOAD follows a polygenic architecture dominated by non-coding variation. However, the tissue-specific regulatory consequences of these variants remain insufficiently characterized. This study aimed to compare the regulatory genomic architectures underlying EOAD and LOAD using a multi-tissue integrative approach. Methods: GWAS-associated variants for EOAD and LOAD were retrieved from the GWAS Catalog using a relaxed significance threshold (p < 1 × 10−5). Variants were functionally annotated and integrated with GTEx v8 eQTL data across 13 neurologically relevant tissues and peripheral blood. Regulatory effects were evaluated using eQTL slope estimates. Basal gene expression patterns were assessed using GTEx RNA-seq data, and protein–protein interaction and functional enrichment analyses were performed using the STRING database. Results: A total of 287 variants were analyzed (32 EOAD, 255 LOAD), with minimal overlap. EOAD exhibited a highly focal regulatory profile, identifying GSE1 as the sole eQTL-regulated gene, restricted to the dorsolateral prefrontal cortex (BA9). In contrast, LOAD displayed a broad multi-tissue regulatory architecture involving APH1B, APOE, CEP63, and HAVCR2, with heterogeneous tissue-specific effects. LOAD-regulated genes converged on pathways related to γ-secretase activity, amyloid precursor protein processing, and Notch signaling, whereas GSE1-associated interactions were enriched for chromatin organization and epigenetic repression. Conclusions: EOAD and LOAD exhibit distinct regulatory genomic architectures, with EOAD characterized by focal, region-specific regulation and LOAD by widespread, tissue-dependent effects, highlighting stage-specific molecular mechanisms contributing to AD heterogeneity.