SUMMARY New variants, especially in non-coding regions, are expected to be discovered through the ongoing AD Sequencing Project (ADSP). This proposal will investigate circular RNAs (circRNAs) and RNA binding proteins (RBPs) that regulate or are regulated by these circRNAs. Recent genomic studies have discovered thousands of circRNAs produced from both protein-coding genes and non-coding regions of the genome via a process known as back-splicing. CircRNAs are more enriched in neuronal tissues and are often derived from genes specific for neuronal and synaptic function. The discovery of these circRNAs demands a coordinated investigation of RBPs that interact with the circRNAs. Mutations in and dysfunction of RBPs are known to be major mechanisms contributing to the pathophysiology in frontotemporal dementia, ALS and AD. However, the contributions of the circRNA:RBP network to these disease mechanisms are largely unknown. The novel biology of circRNAs opens an entirely new window into mechanisms of neurodegeneration in ADRD. CircRNAs could contribute to neurodegeneration by acting as sponges that sequester miRNA/RBPs away from normal mRNA targets, altering splicing or expression. RBPs also regulate circRNA production by binding to the flanking intronic sequences of circRNAs which contain many conserved binding sites of splicing factors/RBPs. Thus, sequestration of RBPs in protein aggregates could cause dysfunctional regulation of circRNAs. The history of genomics indicate that discovery of each new nucleotide species expands our understanding of disease mechanisms. The discovery of circRNA presents a major unexplored avenue of RNA metabolism that demands investigation. We hypothesize that changes in the levels of circRNAs contributes to the pathophysiology of ADRD, and that discovery of key circRNAs or circRNA-RBP interactions in aging human brains could uncover novel biomarkers, disease mechanisms or therapeutic targets. In this proposal, by leveraging large public and our own RNA-seq data (rRNA-depleted), we will apply several methods to detect and characterize AD-related circRNAs from multiple human brain regions, and integrate them with ADSP genetic findings (Aim 1). In Aim 2, aside from discovering AD-related RBPs from human brain RNA-seq, proteomics and ADSP WES/WGS data, we will leverage the ENCODE CLIP-seq data for RBP binding to identify putative RBP-circRNA interactions with AD, i.e. AD-related functional RNA elements. Finally, in Aim 3, we will select the top 10% of the circRNAs (~200) and RBPs (~150) for further high-throughput functional evaluation with a novel, powerful 3D human organoid model of ADRD, termed AstAD that exhibits the full range of tau pathology and neurodegeneration. We anticipate that our integrative analyses of ADSP genetics, circRNA, mRNA, RBP and the high-throughput AstAD functional screen readouts can help generate testable hypothesis for future molecular mechanisms experimental design.