Systematic Identification of Circular RNA in Alzheimer’s Disease

Abstract

Circular (circ)RNAs are an extensive family of RNAs in which the 3′ and 5′ ends are covalently linked. In humans, circRNAs have been implicated in a range of age‐associated pathologies, including Alzheimer’s Disease (AD), Parkinson’s Disease, and cancer. However, only a few of them have been functionally characterized. Given that circRNAs lack 5− and 3− ends and are refractory to exoribonuclease digestion, they have long half‐lives and can serve as effective ‘sponges’, binding cellular factors (RNAs and proteins) in a stable manner. One particular circRNA that is highly abundant in the human brain, ciRS‐7, was found to associate with and ‘sponge’ a specific microRNA, miR‐7, which is an important regulator of post‐transcriptional gene expression in the human brain. We hypothesize that many other circular RNAs may influence the gene expression programs in the AD brain through binding microRNAs or other molecules.Here, our primary goal was to identify all circRNAs relevant to AD using publicly available deep‐sequenced datasets and specifically designed bioinformatic and genomic tools and software. We analyzed four different publicly available data sets (n = 102 total brains) for the presence of circRNAs. We aligned the RNA sequencing data to the human genome (hg19 Ensembl v82) with aligners such as STAR and TopHat2, used CIRCexplorer2 to parse and annotate circRNA junctions in each sample, and processed these junctions through pipelines such as edgeR and DESeq2 (Fig. 1). After analyzing the junctions, we identified the body of each circRNA using the de novo assembly functions of CIRCexplorer2, which we then ran through the same edgeR and DESeq2 pipelines.The results reveal substantial differences in the abundance of circRNAs in AD brains relative to normal brains (Table 1). Moreover, these dysregulated circRNAs do not appear to be confined to one section of the brain, as they appear in the frontal lobe, hippocampus, and lateral temporal lobe. We propose that these circRNAs may represent possible diagnostic or prognostic factors. Assessment of their possible implication in AD pathogenesis is underway through the identification of circRNA‐interacting partners – RNA‐binding proteins and RNAs.differences in RNA content in control (age‐matched) and AD brain.Volcano plot showing differences in circular RNAs (also more and less abundant) in AD relative to control brain samples.Figure 1 Table showing the genes encoding the most commonly and differently expressed circRNAs associated with AD in published studies. Parent Gene Full circRNA Name Relation to Alzheimer’s DOCK1 hsa_circ_chr10_128768965_128780237_F Associated with axon function, neuroinflammation, spine morphogenesis, and neurite growth NTRK2 hsa_circ_chr9_87317073_87325706_F May be a genetic susceptibility gene contributing to AD pathology 1 KIF1B hsa_circ_chr1_10292307_10338186_F Transports synaptic vesicles necessary for nerve impulses DLG1 hsa_circ_chr3_196786759_196796131_R Increased expression shown to cause increased neuronal strength, affecting memory TRAPPC9 hsa_circ_chr8_141407718_141415797_R Associated with cognitive disability in multiple diseases (including autism, HD, etc.) APC hsa_circ_chr5_112154662_112163703_F Involved in synapse assembly between neurons 1Genetic association of neurotrophic tyrosine kinase receptor type 2 (NTRK2) With Alzheimer’s disease. ( PMID: 17918233) Chen Z… Go RC American journal of medical genetics. Part B, Neuropsychiatric genetics: the official publication of the International Society of Psychiatric Genetics 2008.