UCSF

We used a robust imputation method on two case–control cohorts (a small UCSF cohort and a large cohort from the Alzheimer’s Disease Genetics Consortium [ADGC]) to identify HLA haplotypes associated with Alzheimer’s disease and followed up these studies with direct sequencing of the HLA region in AD cases and controls, including both typical amnestic and atypical clinical forms of disease. In our imputed study, we found the haplotype A*03:01~B*07:02~DRB1*15:01~DQA1*01:02~DQB1*06:02 (p = 9.6 x 10-4, odds ratio [OR] [95% confidence interval] = 1.21 [1.08–1.37]) was associated with increased risk of AD in the combined UCSF + ADGC cohort (n = 11,690). Secondary analysis suggested that this effect may be driven primarily by individuals who are negative for APOE-ε4. Separate analyses of class I and II haplotypes further supported the role of class I haplotype A*03:01~B*07:02 (p = 0.03, OR = 1.11 [1.01–1.23]) and class II haplotype DRB1*15:01~DQA1*01:02~DQB1*06:02 (DR15) (p = 0.03, OR = 1.08 [1.01–1.15]) as risk factors for AD. We followed up these genetic associations in a separate clinical dataset representing the spectrum of cognitively normal controls, individuals with mild cognitive impairment, and individuals with AD to assess their relevance to disease. Carrying A*03:01~B*07:02 was associated with higher CSF amyloid levels. We also found a dose-dependent association between the DR15 haplotype and greater rates of cognitive decline on two different assessments.

We also directly sequenced the HLA region in an expanded cohort of AD cases, controls, and atypical AD cases seen at UCSF. We corroborated the accuracy of HLA imputation via direct sequencing of 308 overlapping samples and confirmed the association of the haplotype previously associated with AD risk in the UCSF cohort. We also found that the A*03:01~B*07:02~C*07:02~DRB1*15:01~DQB1*06:02~DPB1*04:01 haplotype was associated with decreased risk of atypical AD in our cohort (p = 0.01, OR = 0.18 [0.02-0.74]). Taken together, our findings corroborate a role of the HLA in AD risk and suggest a differential role of HLA variation in amnestic versus atypical AD.