Single Cell RNASeq Analysis of AD and Control Post Mortem Brain Samples

Abstract

Alzheimer’s disease (AD) is a devastative late life disease, molecularly characterized by biochemical lesions of amyloid β plaques and Tau tangles. the 2nd most frequent one worldwide with no early detection methods/cure. I have applied single cell sequencing and analysed post mortem brain samples (using a Chromium machine) from 2 AD and 2 control frontal cortex samples. All the samples had ethics approval, and the RNA was tested for quality. The samples were obtained from the Edenborough MRC brain bank. Very few genes have been linked to far to AD (mainly through GWAS studies [1]). Overall, I quantified about 24,000 nuclei. Using statistical, bioinformatics and computational analyses I have detected significant differential expression changes in astrocyte, neuronal and microglial marker genes. I also conducted network analyses. The percentage of reads passing filter was slightly low (68%). Overall output was a bit lower than optimal. There was still over 1.6Bn clusters generated though, which is within the specification for an S1 flow-cell, and base call quality (quality control) was excellent: 97% >Q30. The read counts mapping to genome occupancy rate was also fairy low (63-81%). The number of barcodes is as follows: sample 1 (AD, Braak I-II) - 6,223 sample 2 (control) 6,262 sample 3 (AD Braak I-II) 5,332 sample 4 (AD, Braak I-II) 5,332 as well. The data is available for download under the GEO (accession number GSE175814). to assess global and cell-type specific gene expression impacts in Alzheimer’s versus control brain samples, we performed a holistic comparison of matching cell populations in AD and controls with the software cell Harmony. To identify possible gene regulatory network underlying transcriptional regulation in these distinct populations, we further explored prior evidenced transcription factor targets relationships from cell Harmony. Importantly, we observe both common and cell-type specific core transcription factors regulators, with both enriched downstream targets and regulated transcription factors. Upregulated networks in Astrocytes were associated with broad regulation by HIF1A, SOX2, NRF1, and RB1. the proportion of endothelial cells was higher in the Alzheimer’s disease samples than the neurological control samples. results reveal that the cell type-specific transcriptomic changes in Alzheimer’s disease are associated with 4 molecular pathways: angiogenesis in endothelial cells, immune response in endothelial cells and microglia, myelination in oligodendrocytes, and synaptic signalling in astrocytes and neurons. I also compared the results with microarray data from large cohort studies that examined samples from the prefrontal cortex (Alzheimer’s disease: n = 310; Neurological controls: n = 157) or temporal cortex (Alzheimer’s disease: n = 106; Neurological controls: n = 135). Among the DEGs identified in our snRNA-seq analysis, 1,113 and 764 genes were significantly differentially expressed in the microarray data from the prefrontal cortex and temporal cortex, respectively. subcluster analysis of microglia identified 13 subpopulations; VGF signalling was also found as changed. In the future, imaging of post mortem Alzheimer’s patients may verify brain cellular morphology/cell volume of vascular changes in the Alzheimer’s compared to control samples. My study findings [2] may open new venues to our understanding of the disease underlying molecular processes and cellular signaling. Regarding the glial cell changes, our main conclusion is that the glial cells may be reactive in the disease and not causal.