Critical reappraisal of mechanistic links of copy number variants to dimensional constructs of neuropsychiatric disorders in mouse models

Abstract

Copy number variants are deletions and duplications of a few thousand to million base pairs and are associated with extraordinarily high levels of autism spectrum disorder, schizophrenia, intellectual disability, or attention-deficit hyperactivity disorder. The unprecedented levels of robust and reproducible penetrance of copy number variants make them one of the most promising and reliable entry points to delve into the mechanistic bases of many mental disorders. However, the precise mechanistic bases of these associations still remain elusive in humans due to the many genes encoded in each copy number variant and the diverse associated phenotypic features. Genetically engineered mice have provided a technical means to ascertain precise genetic mechanisms of association between copy number variants and dimensional aspects of mental illnesses. Molecular, cellular, and neuronal phenotypes can be detected as potential mechanistic substrates for various behavioral constructs of mental illnesses. However, mouse models come with many technical pitfalls. Genetic background is not well controlled in many mouse models, leading to rather obvious interpretative issues. Dose alterations of many copy number variants and single genes within copy number variants result in some molecular, cellular, and neuronal phenotypes without a behavioral phenotype or with a behavioral phenotype opposite to what is seen in humans. In this review, I discuss technical and interpretative pitfalls of mouse models of copy number variants and highlight well-controlled studies to suggest potential neuronal mechanisms of dimensional aspects of mental illnesses. Mouse models of copy number variants represent toeholds to achieve a better understanding of the mechanistic bases of dimensions of neuropsychiatric disorders and thus for development of mechanism-based therapeutic options in humans.

Keywords: Research Domain Criteria (RDoC); attention-deficit hyperactivity disorder; autism spectrum disorder; copy number variants; intellectual disability; schizophrenia.

Figure 1

Genetic background of different breeding strategies. The targeted gene (white band) is shown with background alleles originating from the embryonic stem (ES) cell donor strain (red), a breeder (green), and randomly mixed alleles of both parents (yellow) in non-congenic mice, congenic mice, co-isogenic mice and F1 hybrid mice. Non-congenic wild-type and mutant mice systematically differ in alleles flanking the targeted gene at the F2 generation due to recombination of alleles. By backcrossing such a mouse to the breeder for more than 10 generations (>10N), the genetic background of a congenic mouse is saturated with alleles of the breeder, thereby minimizing the systematic difference in the flanking regions between wild-type and mutant mice. A co-isogenic mouse is developed in ES cells of a mouse strain and bred with the same mouse line. The F1 hybrid is made by crossing a co-isogenic mutant mouse is crossed with another inbred mouse and the identical genetic background is present between wild-type and mutant mice at the F1 generation.

Figure 2

Impact of deletions of individual 22q11.2 gene on various behavioral constructs relevant to mental disorders. Bold gene names indicate that congenic or co-isogenic mice have been developed. When mice are tested for the selected phenotypes, they are left blank. Deletions of genes cause phenotypes consistent with (red) and opposite to (blue) what is seen in humans with 22q11.2 hemizygous deletions. Deletions cause no detectible effect on a phenotype (black). Phenotypes have not been examined (blank). Non-congenic mice are not included in analysis. Non-congenic mice that are backcrossed for less than 10 generations or carry alleles of both C57BL/6N and C57BL/6J are not included, as these C57BL/6 substrains differ in many behavioral measures due to allelic differences^–. Abbreviations: PPI, prepulse inhibition; WM, working memory; SO, affiliative social interaction, sociability, neonatal and adult vocalization and adult aggression; Anx, anxiety-like behaviors. Dgcr2; Prodh^,; Rtn4r; Ranbp1; Dgcr8^,; Arvcf; Comt^,–;Tbx1^,; Gp1bb; Sept5^,,; Hira.

Figure 3

Hypothetical links between genes, neuronal phenotypes and behavioral phenotypes of a CNV. Hypothetical genes and neuronal phenotypes are indicated as letters in upper and lower cases, respectively. Behavioral phenotypes are indicated as numbers. The causal link is indicated by red arrows.