Mitochondrial (MT) dysfunction has been associated with several neurodegenerative diseases including Alzheimer’s disease (AD). While MT-copy number differences have been implicated in AD, the effect of MT heteroplasmy on AD has not been well characterized. Here, we analyzed over 1800 whole genome sequencing data from four AD cohorts in seven different tissue types to determine the extent of MT heteroplasmy present. While MT heteroplasmy was present throughout the entire MT genome for blood samples, we detected MT heteroplasmy only within the MT control region for brain samples. We observed that an MT variant 10398A>G (rs2853826) was significantly associated with overall MT heteroplasmy in brain tissue while also being linked with the largest number of distinct disease phenotypes of all annotated MT variants in MitoMap. Using gene-expression data from our brain samples, our modeling discovered several gene networks involved in mitochondrial respiratory chain and Complex I function associated with 10398A>G. The variant was also found to be an expression quantitative trait loci (eQTL) for the gene MT-ND3. We further characterized the effect of 10398A>G by phenotyping a population of lymphoblastoid cell-lines (LCLs) with and without the variant allele. Examination of RNA sequence data from these LCLs reveal that 10398A>G was an eQTL for MT-ND4. We also observed in LCLs that 10398A>G was significantly associated with overall MT heteroplasmy within the MT control region, confirming the initial findings observed in post-mortem brain tissue. These results provide novel evidence linking MT SNPs with MT heteroplasmy and open novel avenues for the investigation of pathomechanisms that are driven by this pleiotropic disease associated loci.

Cerebral organoids comprise of diverse cell types found in the developing human brain, and can be leveraged for the identification of critical cell types perturbed by genetic risk variants in common, neuropsychiatric disorders. There is great interest in developing high-throughput, multiplexed technologies to associate genetic variants with cell types. Here, we describe a multiplexed, quantitative approach (oFlowSeq) by utilizing CRISPR-Cas9, FACS sorting and nextgeneration sequencing. Using oFlowSeq, we found that deleterious mutations in autismassociated gene KCTD13 resulted in increased proportions of Nestin + cells and decreased proportions of TRA-1-60 + cells within mosaic cerebral organoids. We further identified that a locuswide CRISPR-Cas9 survey of another 19 genes in the 16p11.2 locus resulted in >2.5% maximum editing efficiencies for short and long indels, suggesting a high feasibility for an unbiased, locuswide experiment using oFlowSeq. Our approach presents a novel method to identify genotypeto-cell type imbalances in an unbiased, multiplexed, quantitative manner.

Mitochondrial (MT) dysfunction has been associated with several neurodegenerative diseases including Alzheimers disease (AD). While MT-copy number differences have been implicated in AD, the effect of MT heteroplasmy on AD has not been well characterized. Here, we analyzed over 1,800 whole genome sequencing data from four AD cohorts in seven different tissue types to determine the extent of MT heteroplasmy present. While MT-heteroplasmy was present throughout the entire MT genome for blood samples, we detected MT-heteroplasmy only within the MT control region for brain samples. We observed that an MT variant 10398A>G (rs2853826) was significantly associated with overall MT-heteroplasmy in brain tissue while also being linked with the largest number of distinct disease phenotypes of all annotated MT variants in MitoMap. Using gene-expression data from our brain samples, our modeling discovered several gene networks involved in mitochondrial respiratory chain and Complex I function associated with 10398A>G. The variant was also found to be an expression quantitative trait loci (eQTL) for the gene MT-ND3. We further characterized the effect of 10398A>G by phenotyping a population of lymphoblastoid cell-lines (LCLs) with and without the variant allele. Examination of RNA sequence data from these LCLs reveal that 10398A>G was an eQTL for MT-ND4. We also observed in LCLs that 10398A>G was significantly associated with overall MT-heteroplasmy within the MT control region, confirming the initial findings observed in post- mortem brain tissue. These results provide novel evidence linking MT SNPs with MT heteroplasmy and open novel avenues for the investigation of pathomechanisms that are driven by this pleiotropic disease associated loci.