Brain metabolism perturbation can contribute to traits and diseases. We conducted the first large-scale CSF and brain genome-wide association studies, which identified 219 independent associations (59.8% novel) for 144 CSF metabolites and 36 independent associations (55.6% novel) for 34 brain metabolites. Most of the novel signals (97.7% and 70.0% in CSF and brain) were tissue specific. We also integrated MWAS-FUSION approaches with Mendelian Randomization and colocalization to identify causal metabolites for 27 brain and human wellness phenotypes and identified eight metabolites to be causal for eight traits (11 relationships). Low mannose level was causal to bipolar disorder and as dietary supplement it may provide therapeutic benefits. Low galactosylglycerol level was found causal to Parkinson's Disease (PD). Our study expanded the knowledge of MQTL in central nervous system, provided insights into human wellness, and successfully demonstrates the utility of combined statistical approaches to inform interventions.
Triggering receptor expressed on myeloid cells 2 (TREM2) plays a critical role in microglial activation, survival, and apoptosis, as well as in Alzheimer’s disease (AD) pathogenesis. We previously reported the MS4A locus as a key modulator for soluble TREM2 (sTREM2) in cerebrospinal fluid (CSF). Here, we pursued the largest CSF study to identify additional novel genetic modifiers of sTREM2. We performed the genome-wide association study (GWAS) of CSF sTREM2 in 3350 individuals of European ancestry from eight cohorts. To identify functional genes, we performed multi-ethnic fine mapping with 250 non-European individuals. Functional validation using peripheral blood mononuclear cell-derived macrophages was performed. Colocalization and Mendelian randomization (MR) was performed for the causality of sTREM2 in AD. The known MS4A locus and three novel loci were identified at genome-wide significance. In the MS4A locus, we identified two independent missense variants (p.M178V in MS4A4A and p.A112T in MS4A6A) that drove the association through multi-ethnic fine mapping. They showed an epistatic effect for sTREM2 and AD risk. The novel TREM2 locus on chromosome 6 contained two rare missense variants (rs142232675 p.D87N, P = 2.71×10-10; rs75932628 p.R47H, P = 7.16×10-19) associated with sTREM2 and AD risk. Another novel locus in the intergenic region between TGFBR2 and RBMS3 (rs73823326, P = 3.86×10-9) included a regulatory variant with a microglia-specific chromatin loop for promoters of TGFBR2. Using cell- based assays we demonstrate that TGFBR2, but not RBMS3, modify sTREM2 levels. The last novel locus NECTIN2 on chromosome 19 (rs11666329, P = 2.52×10-8) was independent of APOE genotype and colocalized with cis-eQTL of NECTIN2 in the brain cortex and cis-pQTL of NECTIN2 in CSF (PP.H4 = 0.83). This variant was also associated with AD risk (P = 1.52×10-66) and age at onset (P = 3.26×10-5). Our MR results confirmed the significant protective effect of sTREM2 on reducing AD risk. To our knowledge, this is the largest study to date aiming at identifying genetic modifiers of CSF sTREM2. We provided novel insights for MS4A and TREM2, the two well- known gens for AD risk. We also identified TGFBR2 and NECTIN2 as additional modulators involved in TREM2 biology. Our findings provide new insight to unravel sTREM2 modulators and their role in AD.
Hippocampal sclerosis of aging (HS) is one of the most predominant causes of dementia in elder populations and it is often misdiagnosed as Alzheimer’s disease (AD). HS has recently been found to show a hippocampal atrophy pattern that might be key for the differentiation of this pathology from AD in the early stages of dementia. In this study, we aimed to find SNPs associated with AD diagnosis that might be related to HS. To that end, we calculated a polygenic risk score (PRS) combining SNPs associated with AD by Bellenguez et al. (2022) and assessed their relation with HS-by-proxy based on magnetic resonance imaging (MRI) data of hippocampal subfield volumes from 1,130 individuals without dementia syndrome. Our results show a significant association of AD-PRS with fimbria and hippocampal body and head volumes. Association with these hippocampal subfields was also observed on the individual effects of the AD-related variants in the SHARPIN (rs34173062) and TNIP1 (rs871269) genes. Besides, these loci remained consistently associated to HS-aging after increasing the sample size by including an independent cohort of healthy individuals (N = 729). The results presented in this study suggest that some SNPs that have recently been identified as AD variants might be closely related to atrophy in hippocampal subfields instead of AD, and hence, could be instrumental to refine AD and HS pathways, their interrelationships, and to elaborate specific treatment for each condition.
Progressive supranuclear palsy (PSP) is a rare neurodegenerative tauopathy. We genotyped all the available histopatologically confirmed PSP cases from the Spanish National Biobank Network and patients with a probable PSP diagnosis recruited from research groups linked to the Dementia Genetics Spanish Consortium (DEGESCO). Here, we present the results of the genome-wide association study (GWAS) performed after the first stage of this project. DNA samples were genotyped by the Centro Nacional de Genotipado (Santiago de Compostela, Spain) using the Affymetrix Axiom 815K Spanish biobank array. After quality control, we kept 156 histopathological (discovery) and 197 clinical (validation) PSP samples. Then, we selected 650 and 800 sex-matched controls from GR@ACE/DEGESCO for stratified genome-wide association of the discovery and validation sets, respectively (Table 1). We ran PC1-4 adjusted logistic regressions with plink v2.00 and meta-analyzed the discovery and validation sets using METAL. Finally, we meta-analyzed our results with 36 previously reported SNPs 1 reaching the largest PSP sample size to date (N = 3053). We found a strong genome-wide significant (GWS) signal associated to the MAPT H1/H2 haplotype (rs8070723 G: OR = 3.40; 95%CI = 2.64-4.39; p = 2.60·10 −20 , Figure 1) and identified a previously unreported signal in chromosome 7p13 (rs76827509 T: OR = 0.45; 95%CI = 0.35-0.58; p = 3.09·10 −10 ). We replicated, with nominal significance, 5 GWS SNPs reported in previous GWAS (Table 2) and found significant effects for NFASC and TOMM40. Because APOE and TOMM40 are in strong linkage disequilibrium, we also checked the effect of APOE variants. We found a protective effect (OR = 0.66; 95%CI = 0.48-0.90; p = 9.75·10 −3 ) of the rs429358 C allele (APOE ε4) in our dataset. The strong GWS signal of the H1 haplotype confirms the validity of our GWAS results 2 . We were able to replicate several PSP loci (Table 1). The NFASC variant reached borderline GWS significance after meta-analysis and is a good PSP risk factor candidate. Finally, the APOE ε4 allele was protective for PSP in our cohort. The second stage of PSP/DEGESCO genotyped 218 additional PSP samples. 1 Sanchez-Contreras, M. Y. et al. Mol. Neurodegener. (2018) 2 Höglinger, G. U. et al. in Nature Genetics (2011)