Integration of genetic, transcriptomic, and clinical data provides insight into 16p11.2 and 22q11.2 CNV genes

Abstract

Deletions and duplications of the 16p11.2 and 22q11.2 copy number variant (CNV) regions are associated with brain-related disorders including autism spectrum disorder, schizophrenia, bipolar disorder, obesity, and intellectual disability. The contribution of individual genes in each CNV to each phenotype is unknown. We propose a novel in silico approach to systematically interrogate each gene's effect on brain-related disorders. As CNVs affect RNA copy number, we attempted to fine-map each CNV region by asking whether natural gene expression variation in any individual gene would increase risk for the same disorder(s). We first used large genotyped cohorts for five CNV-associated traits: autism spectrum disorder, schizophrenia, bipolar disorder, BMI, and IQ. Imputing the transcriptome from the genome, we performed association testing for CNV genes with the genotype datasets. We found individual genes within 16p11.2 associated with schizophrenia (TMEM219, INO80E, YPEL3), BMI (TMEM219, SPN, TAOK2, INO80E), and IQ (SPN), with independent replication. Using a conditional analysis, we showed that INO80E explains the schizophrenia association. Second, we used a biobank containing electronic health data and genotypes for 23,000 individuals to perform phenome-wide association studies with predicted expressions of 16p11.2 and 22q11.2 genes and over 1,500 health traits, finding eight significant gene-trait pairs. Third, we compared the medical phenome of CNV carriers to controls within an independent group of 3 million individuals. We identified CNV traits may have indicated genetic testing, other traits previously observed in CNV carriers, and novel traits. Our results provide new insight into the contribution of individual CNV genes to CNV-associated traits. Limitations

Data integration of bulk and single-cell transcriptomics from cerebral organoids and post-mortem brains to identify cell types and cell type specific driver genes in autism

“…Moreover, in addition to KCTD13, Orgo-Seq prioritizes 2 other candidate driver genes in the 16p11.2 locus (YPEL3 and INO80E) in neuroepithelial cells. A recent report showed evidence for these 2 genes as candidate schizophrenia-associated genes in the locus using a large biobank 48 , which further reaffirms our finding using Orgo-Seq that there is unlikely to be a single driver gene in the 16p11.2 locus.…”

“…Among our top 3 candidate genes in the 16p11.2 locus, two of the genes (YPEL3 and INO80E) were recently reported to be associated with schizophrenia through an association study of imputed gene expression with 5 medical traits identified from electronic health records for over 3 million individuals 48 . In Ypel3 -/mice, an association with absence of startle reflex (P=2.2×10 -5 ) and an association with short tibia (P=5.1×10 -6 ) had been reported 49 .…”

Combinations of genes at the 16p11.2 and 22q11.2 CNVs contribute to neurobehavioral traits

“…With this enhancement, we find one 22q11.2 gene ( PPIL2 ) significantly associated with schizophrenia at a permutation-based threshold (Table 1, Supplementary table 5). We note that the permutation-based threshold is less conservative than the experiment-wide thresholds used in previous analysis [22]. However, we can identify five top genes at 22q11.2 associated with BMI ( YDJC, CCDC116, PPIL2, THAP7, UBE2L3 ), primarily located outside the canonical CNV region (LCR D-E), three with bipolar disorder ( TMEM191B, TUBA8, PPIL2 ), six with ASD ( CLTCL1, AC004471.10, UFD1L, DGCR14, CCDC188, DGCR9 ), and two with IQ ( SEPT5, LINC00896 ) (Table 1, Supplementary table 5).…”

“…We included noncoding genes, as they have not received significant attention in studies of these regions, despite some evidence of miRNA contribution to 22q11.2 phenotypes. In addition, we considered flanking genes within 200kb of the region, as there is suggestive evidence of broader transcriptional effects in CNV carriers, and because we previously found evidence of flanking gene involvement in psychosis [22,27]. Supplementary tables 1 and 2 contain single and pairwise CNV genes used in analysis.…”

Tissue and cell-type specific molecular and functional signatures of 16p11.2 reciprocal genomic disorder across mouse brain and human neuronal models

“…These include, for example, SEZ6L2 (synapse numbers, dendritic morphology, and neuritogenesis) (Qiu et al, 2021; Yaguchi et al, 2017), CORO1A (filopodia formation, required for initial neurite formation) (Alvarez Juliá et al, 2016; Dent et al, 2007), DOC2A (spontaneous neurotransmission associated with its calcium-dependent translocation) (Courtney et al, 2018; Groffen et al, 2006), TLCD3B (altered composition and levels of sphingolipids and glycerolipids associated with cellular membranes leading to synaptic protein mislocalization) (Tomasello et al, 2021), TAOK2 (brain size, neural connectivity and excitatory transmission) (Richter et al, 2019); PRRT2 (neuronal excitability) (Fruscione et al, 2018); and MAPK3 (dendritic alterations of cortical pyramidal neurons) (Blizinsky et al, 2016). Recently, integration of genetic regulation of gene expression with genome-wide association data from human cohorts pointed to INO80E as the potential driver of schizophrenia due to 16p11.2 duplication and to both SPN and INO80E as contributors to the increased body mass index due to 16p11.2 deletion (Vysotskiy et al, 2021). Other model studies have also pointed to interaction between 16p11.2 genes as potentially being responsible for abnormal phenotypes (Iyer et al, 2018; McCammon et al, 2017).…”