Towfique Raj - US grants

DESCRIPTION (provided by applicant): In this project, we will investigate the transcriptomes of a human brain tissue to resolve RNA transcripts and functional variants that contribute to the process of cognitive decline in the aging human brain. Our approach using next generation RNA sequencing to create these transcriptomes will allow an unparalleled perspective on the complexity of transcriptional regulation of the brain, which is a tissue with one of the most diverse content of alternatively spliced transcripts and non-coding RNA molecules, many of which may not have been previously discovered. Further, the availability of genome-wide genotype data on the same subjects gives us a unique opportunity to explore the genetic effects on RNA expression. Thus, we may gain insight not only into the transcriptome diversity of the aging human brain but also into the manner in which genetic variation associated with clinically impaired cognition exert a functional consequence on the human brain. ! ! PUBLIC HEALTH RELEVANCE: The goal of this proposal is to understand which molecules of a brain tissue are involved in the cognitive decline in older individuals. In understanding how these molecules relate to genetic risk for aging and human cognitive function will allow investigators to better understand the onset of cognitive decline and diseases of cognition, such as Alzheimer's disease (AD). This may help to develop clinical tests that may be able to predict who is at risk of developing AD and to develop new treatments to delay or prevent onset of cognitive decline and AD.

Alzheimer's disease (AD) is an age-related neurodegenerative disease characterized by progressive cognitive decline and dementia. Recent genome scans have identified over twenty novel AD susceptibility loci, and several of these loci implicate the immune system. Our recent gene expression analyses of data from healthy young individuals have implicated 12 AD susceptibility genes in myeloid cell function, whose expression, relative to each risk allele, is altered in the primary monocytes. We have also recently identified AD risk-increasing alleles of the myeloid cell surface receptor CD33 that are associated with diminished A? uptake by human monocytes. Furthermore, AD GWAS signals, as well as the most strongly validated coding variants associated with AD (in APOE, TREM2 and ABCA7), coalesce around genes that are necessary for efficient phagocytic clearance of cellular debris by myeloid cells. Therefore, these loci represent excellent candidates as the first step in the cascade of molecular events that link genetic risk factors to the altered innate immune function that contributes to AD pathology. Indeed there is conflicting evidence as to the relative importance in AD pathogenesis of peripheral myeloid cells that subsequently enter the brain versus tissue resident myeloid cells such as microglia. Because of their shared ontology regulation of expression of many myeloid specific genes is likely to be shared between monocytes and microglia. Given the ease of access to blood monocytes throughout the disease process, compared to microglia that are only accessible at autopsy we propose to explore the functional consequences of commonly- occurring genetic variation on the transcriptome in peripheral monocytes from AD patients. Our central hypothesis is that peripheral monocyte-derived cells, such as macrophages and monocytes will manifest changes in gene expression of these AD susceptibility genes and other genes in the same molecular pathways that reflect the stages of AD pathophysiology. To test this hypothesis, we will characterize the transcriptome/methylome profiles from peripheral monocytes of 200 AD cases and 200 age-matched controls. This will be followed up with profiling of monocytes stimulated with anti- (myelin) and pro-inflammatory (LPS and A?) stimuli. Through innovative computational approaches, we will integrate various datasets from monocytes in order to identify causal drivers and molecular networks underlying AD pathogenesis such as A? clearance and neuroinflammation. We will also incorporate data from over 500 AD brains to assess if the monocyte-specific transcriptional networks recapitulate changes seen in the AD brains. Finally, we will perform highly multiplexed mass cytometry-based immune phenotype profiling to investigate the activation state and phagocytic capacity of monocytes. We will validate our most promising candidate genes from the functional studies in human microglia using immunohistochemistry. The proposed research is innovative because it will not only identify genetic mechanisms in peripheral monocytes which may contribute to interindividual risk for AD but may lead to the discovery of novel immune biomarkers.

Alzheimer's disease (AD) is an age-related neurodegenerative disease characterized by progressive cognitive decline and dementia. Our long-term goal is to understand the pathogenesis of AD and pave the way towards novel diagnostics and therapeutics. The objective here is to investigate the mechanistic link between dysregulation of mRNA splicing and neurodegeneration. Our central hypothesis is that AD-associated genetic variants influence mRNA splicing, which result in changes to pathways/networks and gene expression downstream contributing to cellular events leading to protein aggregation in AD. We will test our hypothesis by pursuing the following specific aims: In aim 1, we will perform a genome wide analysis using publically available multimodal expression data from post-mortem autopsy brain tissue (n=1,300) to discover aberrant RNA splicing events in the AD transcriptome. We will identify common and rare genetic variants regulating alternative splicing events in the brain (splicing QTLs). We will use novel long-read small molecule real time (SMRT) sequencing (IsoSeq) to generate full-length transcripts from 100 postmortem brain tissues to profile the full complexity of the AD transcriptome. We will validate these changes using proteomic, biochemical and cellular approaches. In aim 2, we will validate the RNA targets of disease-specific RNA binding proteins (RBPs) using enhanced crosslinking-immunoprecipitation (eCLIP) in control and AD brains. In aim 3, we will validate RNA targets of RBPs via lentivirally delivered shRNA knockdown and RBP binding sites via antisense oligonucleotide (ASO) inhibition in human pluripotent stem cell-derived neuronal and glial cell populations. This project will have a large overall impact by providing a comprehensive survey of RNA regulation in AD brain and adding functional and mechanistic interpretation of genetic variants associated with AD susceptibility. This contribution will be significant because it will provide a foundation for further mechanistic studies that will elucidate the drivers of disease and pave the way for novel therapeutic avenues, likely based on ASO- mediated approaches aimed at counteracting splicing changes by blocking proximal splicing enhancers or suppressors.

With ageing populations world-wide, neurodegenerative diseases are placing an ever increasing burden on long- term well-being, healthcare costs and family life. Despite decades of research and enormous investment, no disease-modifying treatment is available for the most common of these diseases: Alzheimer?s (AD). The majority of these, to-date unsuccessful, efforts have focused on one potential cause of AD: amyloid-? aggregation. Combining population-scale data collection, human genetics and machine learning provides a way forward to uncover and characterize new causal cellular processes involved in AD. This would provide an array of potential therapeutic targets, increasing the chance that one will be more easily modulated than the amyloid-? pathway. AD-specific genomic datasets of unprecedented scale are being actively collected: whole genome sequencing (WGS) from ~20k individuals, gene expression (RNA-seq) and epigenomics (ATAC-seq, histone ChIP-seq) from >1000 post-mortem AD brains, single-cell transcriptomes and similar modalities in peripheral and brain-resident innate immune cells (which we and others have shown to be AD-relevant). Effectively integrating these diverse data to better understand AD represents a substantial computational challenge, both in terms of data scale and analysis complexity. This proposal leverages state-of-the-art deep learning (DL) and machine learning (ML), combined with human genetic analyses, to address this challenge. We will train DL models to predict epigenomic signals and RNA splicing from genomic sequence, enabling in silico mutagenesis to estimate the functional impact (a ?delta score?) of any genetic variant. The delta scores will be used in genetic analyses that distinguish causal associations: cellular changes that drive AD pathogenesis rather than downstream/side effects of disease. Delta scores will aid in associating both rare and common variants to AD. To achieve sufficient power, rare variants must be aggregated (e.g. for a gene): delta scores will allow filtering out many likely non-functional (particularly non-coding) variants. Most common variants from AD Genome Wide Association Studies (GWAS) are simply correlated with the causal variant due to linkage disequilibrium (LD). Delta scores, combined with trans-ethnic GWAS, will enable estimation of the likely causal variant(s). These analyses will highlight variants and genes involved in AD. However, genes do not operate in a vacuum so robust probabilistic ML will be used to learn cell-type and disease-specific gene regulatory networks from sorted bulk and single-cell RNA-seq. The detected networks will be integrated with our genetic findings to discover network neighborhoods/pathways especially enriched in AD variants. Such pathways will be prime candidates for future functional and therapeutic studies of AD.

PROJECT SUMMARY Alzheimer's disease (AD) is an age-related neurodegenerative disease characterized by progressive cognitive decline and dementia. Although the mechanisms underlying AD are still largely unknown, it is clear that aging of the brain is a key risk factor for this disease. Several lines of evidence indicate that microglia, the myeloid immune cells of the brain, play a crucial role in the processes involved in normal aging of the central nervous system and development of AD. Recent genetic studies have identified over twenty novel AD risk loci, and network analysis have shown that a major part of these loci play a role in the myeloid immune system. In addition, gene expression changes have been found in microglia of aged individuals, subjects with AD and in microglia in mice models for AD. How these microglia changes contribute to AD is not yet clear. Answering this question is an important step towards understanding the mechanisms involved in AD. In addition, this could lead to unravelling novel targets for treatment of AD and related neurodegenerative disorders. The long-term goal of this project is to deepen our insight into the role of microglia cells in AD and to identify microglia-related targets for treatment of age-related disorders of the central nervous system. An important first step towards reaching this goal is to understand what the changes that have been found microglia in AD tell us in terms of changes in gene and protein expression and functions. The overall objective of this study is therefore to identify the AD-associated common genetic variants that alters microglia gene expression at baseline and in response to inflammatory stimuli. By combining the unique expertise of the two PIs in the isolation and culture of human microglia, as well as advanced computational genomics analysis of human myeloid immune cells. In Aim 1, we will use 264 existing microglia samples that we have previously isolated of different regions of 103 brain donors to generate genotype and transcriptome profiles. By combining these data with existing microglia transcriptomic datasets, we will be able to generate a map of how AD-associated genetic loci influence gene expression (or expression quantitative trait loci, eQTL) and splicing (sQTL). In aim 2, we will use interferon (IFN) stimulated microglial samples that we have previously collected to characterize how AD-associated risk variants alter IFN-stimulated transcriptome changes. These profiles will be validated in new microglia transcriptomes from an independent cohort of 15 donors to determine the response of these samples to interferon and A? and sort subsets with different phagocytic capacity. The transcriptome profiles and expression and splicing QTL that will be generated in these two aims will be made publically available and we will apply these profiles to very large available gene datasets on aged and AD brain tissue and peripheral monocytes to investigate how gene expression changes relate to changes in microglia function. Together, we expect that this study will provide key information bridging AD genetics to molecular mechanisms in microglia, setting the stage for future mechanistic studies in model systems.

Alzheimer's disease (AD) is an age-related neurodegenerative disease characterized by progressive cognitive decline and dementia. Despite more than fifty years of research, no cures exist and the standard of treatment remains unsatisfactory. Genome wide association studies (GWAS) indicate that AD risk reflects both highly penetrant rare variants as well as common single nucleotide polymorphisms with small effect sizes. By overlapping GWAS and post-mortem and myeloid cell expression analyses, we have identified common variants with expression and splicing quantitative trait loci that may contribute to altered gene expression or alternative splicing; however, demonstrating which risk loci are the causal contributors to disease risk remains an intractable problem. Here, we will apply statistical approaches to prioritize putative causal variants in AD-associated loci by incorporating expanded AD GWAS functional annotations, single cell chromatin profiles and histone modification and microglia gene expression datasets. We will then apply a human induced pluripotent stem cell (hiPSC)-based approach to manipulate the genotype of prioritized putative causal AD risk variants that alter gene expression or alternative splicing of mRNAs, focusing largely on genes implicated in phagocytosis and lysosomal-autophagy function. Through isogenic comparisons of neurons, microglia, and astrocytes, we propose to examine the molecular and functional effects of perturbing five putative causal SNPs separately testing their cell autonomous and non-cell autonomous impact on cellular function. Our hope is that this work may identify novel therapeutic points of intervention in order to prevent or slow disease course in individuals with AD. This project will have a large overall impact by providing a mechanistic interpretation of genetic variants associated with AD susceptibility.