Unraveling the genetic architecture of copy number variants associated with schizophrenia and other neuropsychiatric disorders

Abstract

Recent studies show that the complex genetic architecture of schizophrenia (SZ) is driven in part by polygenic components, or the cumulative effect of variants of small effect in many genes, as well as rare single-locus variants with large effect sizes. Here we discuss genetic aberrations known as copy number variants (CNVs), which fall in the latter category and are associated with a high risk for SZ and other neuropsychiatric disorders. We briefly review recurrent CNVs associated with SZ, and then highlight one CNV in particular, a recurrent 1.6-Mb deletion on chromosome 3q29, which is estimated to confer a 40-fold increased risk for SZ. Additionally, we describe the use of genetic mouse models, behavioral tools, and patient-derived induced pluripotent stem cells as a means to study CNVs in the hope of gaining mechanistic insight into their respective disorders. Taken together, the genomic data connecting CNVs with a multitude of human neuropsychiatric disease, our current technical ability to model such chromosomal anomalies in mouse, and the existence of precise behavioral measures of endophenotypes argue that the time is ripe for systematic dissection of the genetic mechanisms underlying such disease. © 2016 Wiley Periodicals, Inc.

Keywords: CNVs; CRISPR; behavioral assays; genetics; schizophrenia.

Figure 1. Risk for SZ, ID, and ASD associated with recurrent CNVs

Odds ratios (ORs) for SZ were obtained from (Rees et al., 2014b, Mulle, 2015). ORs for ID were estimated using ID data from (Kaminsky et al., 2011), when zero CNV were observed in controls the OR was estimated by Haldane's correction (adding 0.5 to each cell in the 2×2 matrix). ORs for ASD were estimated from ASD case data (Sanders et al., 2015), n = 4687 cases) and controls (Cooper et al., 2011), n = 8329 controls) using Fisher's exact test in R; when zero CNV were observed in controls the OR was estimated by Haldane's correction.

Figure 2. Generating CNVs using CRISPR

A. To generate an inversion or deletion, CRISPRs are designed at the desired breakpoints to create DSBs. The resulting inversion or deletion is generated only if the DSBs are created on the same homologous chromosome (cis). B. If the DSBs are generated on different homologous chromosomes (trans), this may result in a duplication. Generation of DSBs in cis or trans is random and cannot be controlled. C. As an example of how to generate a translocation, two CRISPRs are needed. The first CRISPR is designed to create a DSB outside the first gene of interest on chromosome 3, and the second CRISPR is designed to create a DSB outside the other gene of interest on chromosome 10.

Figure 3. Generating subdeletions using CRISPR

Designing subdeletions of a CNV is relatively simple. Briefly, the recurrent deletion is divided into three (or more) segments (A, B, C). Each of these segments is separately deleted using the corresponding CRISPR. Functional studies can then be done on each subdeletion to determine which segment contains the gene(s) that confer risk for a disorder, such as schizophrenia (SZ). Other rearrangements are possible with this design, as outlined in Figure 2.