Resources

Increasing quantitative and biological understanding of the role of TREM2 in Alzheimer’s disease

One of the strongest risk factors for late onset AD is the loss-of-function (lof) variants in the TREM2 gene2.  Understanding the impact of the TREM2 lof variants on gene expression in both disease and non-disease contexts will enable deeper understanding of its biology.   

The TREM2 project is underway whilst prospective brain tissue is being collected.  Our  transcriptomic methods are being used to generate datasets from retrospectively acquired brain tissue from AD and control donors and on a genetically stratified group of donors (TREM2 variant).  This study aims to increase our quantitative and biological understanding of the role of TREM2 in AD.  It will combine transcriptomic and lipidomic data with histological and genotyping data from AD and healthy control donors with and without the TREM2 variants.  This work is in collaboration with Biogen and will be completed at the beginning of 2021. 

Fancy et al., 2022 - Mechanisms contributing to differential genetic risks for TREM2 R47H and R62H variants in Alzheimer’s Disease

Single nuclear transcriptional signatures of dysfunctional brain vascular homeostasis in Alzheimer’s disease

Brain perfusion and normal blood brain barrier integrity are reduced early in Alzheimer’s disease (AD). We performed single nucleus RNA sequencing of vascular cells isolated from AD and control brains to characterise pathological transcriptional signatures. We found that endothelial cells (EC) are enriched for expression of genes associated with susceptibility to AD. EC transcriptional signatures identified mechanisms for impaired β-amyloid clearance. Evidence for immune activation was found with upregulation of interferon signalling genes in EC and in pericytes (PC). Transcriptional signatures suggested dysregulation of vascular homeostasis and angiogenesis with upregulation of pro-angiogenic signals (HIF1A) and metabolism in EC, but downregulation of homeostatic growth factor pathways (VEGF, EGF, insulin) in EC and PC and of extracellular matrix genes in fibroblasts (FB). Our genomic dissection of vascular cell risk gene enrichment suggests a potentially causal role for EC and defines transcriptional signatures associated with microvascular dysfunction in AD.

Tsartsalis et al. 2021 - Single nuclear transcriptional signatures of dysfunctional brain vascular homeostasis in Alzheimer’s disease

Enrichment method to enable the study of large numbers of glia from the post mortem human brain

Amy Smith, supported by Karen Davey, has developed an enrichment method to enable the study of large numbers of glia from the post-mortem human brain by combining nuclei-FACS with single-nuclei RNA-sequencing.  This results in a 5-fold increase in the number of glial transcriptomes obtained for the study of microglia-astrocyte interactions in AD.  This initial work has supported optimisation of methods and validation of a comprehensive analysis pipeline developed as part of the project. 

Smith et al., 2022 - Diverse human astrocyte and microglial transcriptional responses to Alzheimer’s pathology

View gene expression data

View and download raw data

The effects of experimental delay on freshly resected tissue for single nuclei transcriptomics

Using freshly-resected brain tissue from McGill University’s ‘The Neuro’ state-of-the art biobanking platform, this project aims to systematically assess how differing processing techniques (time and temperature) cause artifacts in transcriptional data relevant to interpreting brain autopsy studies.  The collaboration between McGill and Imperial DRI will include advanced bioinformatics and create a user-friendly open science website where researchers can deposit their human brain transcriptional datasets and assess to what degree their data may be subject to processing artifacts associated with time and temperature.  Tissue collection is currently underway with snRNAseq to start in early 2021. 

MAP: Prospective brain collection

Brain tissue is being prospectively acquired from several brain banks in a standardised manner and sampled from 8 brain regions ranging from those with high to those with low levels of pathology.  The aim is to have 12 AD and six controls brains which are well phenotyped and without co-morbidities.   To achieve this, approximately 40 brains need to be collected and histopathological analysis conducted to ensure they fit the inclusion/exclusion criteria.  Currently we have collected 14 AD cases and 6 controls and they are undergoing histopathology to determine their suitability.