Cell type‐specific Alzheimer's disease polygenic risk scores are associated with distinct disease processes in preclinical Alzheimer's disease. (1st February 2022)

• Record Type: Journal Article
• Authors: Yang, Hyun‐Sik; Kang, Daniel; Menon, Vilas; Klein, Hans‐Ulrich; Ge, Tian; Finucane, Hilary; Chibnik, Lori B.; Hohman, Timothy J.; Mayeux, Richard; De Jager, Philip L; Sperling, Reisa A.
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 3
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-02-01
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.055304
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Conventional polygenic risk score (PRS) summarizes genetic risk throughout the genome and lacks biological specificity. Here, leveraging human brain single nucleus RNA sequencing (sNuc‐Seq) data, we introduce a new method to capture cell‐type specific Alzheimer's disease (AD) PRS, and implicate different cell types in distinct disease processes in preclinical AD. Method: Cognitively unimpaired (CU) older adults of European descent who participated in the Anti‐Amyloid Treatment in Asymptomatic Alzheimer's (A4) study screening PET were included in this study. Cortical amyloid‐β burden (Aβ) was quantified as cortical composite SUVR from the florbetapir PET. Screening cognition was measured with Preclinical Alzheimer Cognitive Composite (PACC). We defined cell type‐specific genes as genes expressed higher in a given cell type (FDR<0.05) compared to other cell types in a published human neocortex sNucSeq study, for eight major cell types identified from the same study. Published AD genome‐wide association study results were processed with PRS‐CS to infer posterior effect size of each genetic variant, and we derived cell type‐specific ADPRS by weighted summation and standardization of the genetic risk of AD localized to cell type‐specific genes. We excluded the APOE region from ADPRS, and adjusted for APOE ε4, ε2, age, sex, and the first three genotype principal components in all analyses. Result: In CU older adults from the A4 screen data (n=2961; 71.3±4.8 years old), microglia‐specific ADPRS had the strongest association with higher Aβ (beta=0.018, p=2.0×10 ‐8 ), while neuron‐specific ADPRSs showed weaker associations (Figure 1A). By contrast, excitatory neuron‐specific ADPRS had the strongest association with lower PACC (additionally adjusted for years of education; beta=‐0.16, p=1.0×10 ‐4 ; Figure 1B), and this association remained significant even after adjusting for Aβ (beta=‐0.14, p=6.1×10 ‐4 ). The excitatory neuron‐specific ADPRS was negatively associated with years of education (beta=‐0.12, p=0.018), suggesting a shared heritability between neuron‐specific AD risk and lower cognitive reserve. Conclusion: These results further implicate microglia in fibrillar Aβ accumulation and support the neuronal role in cognitive reserve. Cell type‐specific ADPRSs derived using our novel method demonstrate specific relationships with AD pathology and cognition, and may improve power to predict progression during the preclinical stages of AD.