Co-expression of prepulse inhibition and Schizophrenia genes in the mouse and human brain

Abstract

Schizophrenia is a complex psychiatric disorder with genetic and phenotypic heterogeneity. Accumulating rare and genome-wide association study (GWAS) common risk variant information has yet to yield robust mechanistic insight. Leveraging large-scale gene deletion mouse phenomic data thus has potential to functionally interrogate and prioritize human disease genes. To this end, we applied a cross-species network-based approach to parse an extensive mouse gene set (188 genes) associated with disrupted prepulse inhibition (PPI), a Schizophrenia endophenotype. Integrating PPI genes with high-resolution mouse and human brain transcriptomic data, we identified functional and disease coherent co-expression modules through hierarchical clustering and weighted gene co-expression network analysis (WGCNA). In two modules, Schizophrenia risk and mouse PPI genes converged based on telencephalic patterning. The associated neuronal genes were highly expressed in cingulate cortex and hippocampus; implicated in synaptic function and neurotransmission and overlapped with the greatest proportion of rare variants. Concordant neuroanatomical patterning revealed novel core Schizophrenia-relevant genes consistent with the Omnigenic hypothesis of complex traits. Among other genes discussed, the developmental and post-synaptic scaffold TANC2 (Tetratricopeptide repeat, ankyrin repeat and coiled-coil containing 2) emerged from both networks as a novel core genetic driver of Schizophrenia altering PPI. Aspects of psychiatric disease comorbidity and phenotypic heterogeneity are also explored. Overall, this study provides a framework and galvanizes the value of mouse preclinical genetics and PPI to prioritize both existing and novel human Schizophrenia candidate genes as druggable targets.

Keywords: Cross-species; Endophenotype; Mouse models; Prepulse inhibition; Schizophrenia.

Fig. 1

Study overview using mouse prepulse inhibition (PPI) genes and brain expression to stratify human Schizophrenia common and ultra-rare variants and identify potentially novel ultra-rare variant genes with „core”-like properties. Created with BioRender.com.

Fig. 2

Prepulse inhibition (PPI) phenotypes associated with disruption of selected mouse genes. (A) Gria1 (Glutamate Ionotropic Receptor AMPA Type Subunit 1), (B)Gad2 (Glutamic acid decarboxylase, 2), (C)Cpe (Carboxypeptidase E), (D) Tdo2 (Tryptophan 2,3-dioxygenase), (E) Mphosph9 (M-phase phosphoprotein 9), (F) Aspa (Aspartoacylase), (G) Olig1 (Oligodendrocyte transcription factor 1), (H)Lrrtm1 (Leucine rich repeat transmembrane neuronal 1), (I)Ntf5 (Neurotrophin 5). Comparisons were made using the soft windowing approach of the International Mouse Phenotyping Consortium (IMPC) web resource at www.mousephenotype.org and hence, control group numbers vary accordingly. WT = wildtype control mice, MUT = mutant mice.

Fig. 3

(A) Unsupervised hierarchical clustering analysis of prepulse inhibition (PPI) gene expression in brain. Brain regions (BR) were clustered according to in situ hybridization (ISH) expression values (Allen Brain Atlas) of PPI genes and revealed two clusters (BR-A and BR-B). PPI genes gathered in three clusters PPI-1, PPI-2 and PPI-3. Color-coded lines represent brain regions of interest for PPI. (B) Visualization of the color-coded brain regions from the heatmap indicated in mouse brain coronal brain sections with corresponding Bregma levels for each section. Enrichment analyses for each of the PPI gene modules PPI-1 (C), PPI-2 (D) and PPI-3 (E). (F) % overlap of PPI genes (188) with Schizophrenia genome wide associated study (GWAS) common variants from GWAS catalog. Created with Biorender. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

(A) Unsupervised hierarchical clustering of transcriptomic data from entire genome based on BR-A. Brain regions (BR) were clustered according to in situ hybridization (ISH) expression values (Allen Brain Atlas) and revealed three brain region clusters. Genes gathered in three extended clusters (I, II, III). (B) The overlap of established schizophrenia common variants with each of the extended clusters. (C, D, E). Enrichment analyses for each of the Schizophrenia ultra-rare and common variant gene overlaps with each of the extended gene modules are shown.

Fig. 5

Schizophrenia-relevant „core” genes enriched in Midnight Blue module. (A) Module enrichments for established Schizophrenia common and rare variants and mouse prepulse inhibition (PPI) genes. The Midnight Blue module is significantly enriched in common and ultra-rare variants and tends to be enriched in mouse PPI genes. The red line in each plot indicates the significance threshold of padj <0.05. (B) Midnight Blue module functional and disease annotations. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)