Genetic studies of AD had identified mutations in the amyloid-beta precursor protein (APP) presenilin (PSEN1) and presenilin 2 (PSEN2) genes cause Mendelian forms of AD. These mutations have been found in a small number of people, but the identification of such mutations and genes provided a better understanding of the biology of AD. More recent studies using genome-wide association studies (GWAS) approaches in late onset AD have identified more than thirty loci that influence risk for AD. For some of these loci genetic variants in specific genes have been identified, but for the majority of these loci the disease gene remains to be identified. The goal of this project is to use genetic and genomic approaches to identify functional networks enriched for AD risk and protective loci, and to use this information to determine how myeloid cellular function is impacted by these genetic factors in disease-relevant cell types. The ultimate goal is to use this information to identify drugs that target these networks and to test the drugs in appropriate cellular models. During the last three years we have revealed substantial evidence to support the hypothesis that many AD risk genes operate within an inter-connected network that modulates myeloid cell function. We plan to extend this work to more fully characterize the affected network(s) using bioinformatics and to validate our findings through experimental approaches in BV2 cells and human induced pluripotent stem cells (hiPSC)-derived microglia.

A major obstacle for the translation of Late onset AD (LOAD) GWAS data into an actionable (from a drug discovery perspective) understanding of disease etiology is not only the increased complexity of the genetic architecture of LOAD compared to autosomal-dominant AD, but also the fact that (with a few exceptions) the causal genes are much more difficult to identify unequivocally. Indeed, most genetic variants associated with LOAD don’t occur within genes but instead fall into non-coding regions of the genome that modulate the expression of one or more genes in specific cell types. By applying novel computational and statistical methods for integrated analyses of human genetics and genomics datasets, we have identified strong candidate genes, biochemical pathways and cellular processes. In particular, we have discovered that variants in most of the LOAD GWAS loci function to modulate the expression of genes that play important roles in cells of the myeloid lineage, such as monocytes and macrophages (including microglia), the professional phagocytes of the innate immune system. Several of these candidate genes (together with well-validated LOAD genes like APOE, TREM2 and ABCA7) cluster around the phagocytic clearance of lipid-rich cellular debris (hereafter referred to as efferocytosis), a core and evolutionarily conserved function of myeloid cells that is essential for the maintenance of tissue homeostasis, immune tolerance, and the resolution of inflammation. A recent publication indicated that members of the MS4A family are lipid sensors in the olfactory system.  This observation led us to proposed that the MS4A genes linked to AD could be lipid sensors for microglia, potentially unifying them with the efferocytosis hypothesis. The goal of this research is to identify candidate the causal gene(s) in the MS4A cluster (an AD-associated locus) and to investigate their role in AD pathogenesis for the development of therapeutic agents.

The Dominantly Inherited Alzheimer Network (DIAN) established an international, multicenter registry of individuals (gene carriers and noncarriers; asymptomatic and symptomatic) who are at risk of carrying a known causative mutation for AD in the amyloid precursor protein (APP), presenilin 1 (PSEN1), or presenilin 2 (PSEN2) genes. Individuals are evaluated upon enrollment in DIAN and longitudinally thereafter with standard instruments including the Uniform Data Set of the Alzheimer’s Disease Centers and protocols developed by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) for structural, functional, amyloid imaging , biological fluids (blood; CSF), and histopathological examination of cerebral tissue in individuals who come to autopsy. We have expanded our network of centers and begun longitudinal characterization of a large series of autosomal dominant AD (ADAD) kindreds with known disease-causing mutations and will continue longitudinal follow up of these kindreds to identify the earliest detectable changes associated with development of disease and to characterize the temporal series of events that occurs up to and including the diagnosis of symptomatic AD. The goal of the Genetics Core of the DIAN initiative is to provide genetic information and useful biological and genomic materials to the research community for the study of AD. We have already collected genomic samples from 531 individuals and generated fibroblasts from 99 individuals. We anticipate collection of an additional 125 new individuals, including participants from NIH and self-funded sites. We will expand the fibroblast and induced pluripotent stem cell collection. The Core will maintain and curate a list of pathogenic mutations and confirm that new DIAN families carry an ADAD mutation. The Core will also generate GWAS and APOE genotype data on all individuals and obtain biological materials (fibroblasts, induced pluripotent stem cells, white blood cells) to perform cell-based functional studies. All data will be placed in the DIAN database. We will support all projects in DIAN and perform analyses with other Cores to identify novel factors modulating age at onset in these families.

Family-based approaches led to the identification of disease-causing Alzheimer’s Disease (AD) variants in the genes encoding amyloid-beta precursor protein (APP), presenilin 1 (PSEN1) and presenilin 2 (PSEN2). Subsequently, the identification of these genes led to the Aβ-cascade hypothesis and recently to the development of drugs that target that pathway. In this proposal, we will identify rare risk and protective alleles. In a recent study, we identified a rare coding variant in TREM2 with large effect size for risk for AD, confirming that rare coding variants play a role in the etiology of AD. We will use sequence data from families densely affected by AD, because we hypothesize that these families are enriched for genetic risk factors. We already have access to sequence data from 695 families (2,462 individuals), that combined with the ADSP data will lead to a very large family-based dataset: more than 805 families and 4,512 participants. Our preliminary results support the flexibility of this approach and strongly suggest that protective and risk variants with large effect size will be found. The identification of those variants and genes will lead to a better understanding of the biology of the disease.

To date, a total of 1,454 multiplex late onset AD (LOAD) families have been recruited with 8,543 family members clinically assessed and DNA sampled. We have also recruited 1,030 controls. Genome-wide SNP arrays have been generated on 5,428 individuals, exome chip genotyping on 1,278 individuals, whole exome sequencing in 1,484 and whole genome in 928 family members and controls. The NIA-LOAD FBS provides an excellent opportunity to improve our understanding of the clinical and biological impact of genetic variation in the elderly. Phenotypic information is continually updated in these families by regular cognitive evaluations and autopsy at the time of death to confirm the diagnosis of LOAD. We have begun to recruit additional family members with a particular emphasis on the offspring generation. We have been able to bank brain tissue from family members creating one of the largest collections of brain tissues for familial LOAD. We will now expand biological sampling to include RNA and peripheral blood mononuclear cells in selected families. As additional genes and variants are identified, the members of the NIA LOAD Family Study will again play a central role as we explore: What is the impact of these risk and protective variants on disease risk? Are the genetic variants highly penetrant? What is the risk of developing LOAD in offspring? Can the presence of variants be used for stratification of patients into specific subtypes for clinical trials? Can the family data be used to identify novel biomarkers of disease risk, age at onset onset or progression? The NIA-LOAD FBS dataset is uniquely poised to address these clinical and biological questions because of its large size, rigorous ascertainment criteria, standardized clinical assessment and lack of restriction to specific mutations. This is by far the largest collection of LOAD families available in the world and virtually every major genetic study of Alzheimer’s disease has included patients and controls from the NIA-LOAD FBS dataset.