Cortical patterning of abnormal morphometric similarity in psychosis is associated with brain expression of schizophrenia-related genes

Abstract

Schizophrenia has been conceived as a disorder of brain connectivity, but it is unclear how this network phenotype is related to the underlying genetics. We used morphometric similarity analysis of MRI data as a marker of interareal cortical connectivity in three prior case-control studies of psychosis: in total, n = 185 cases and n = 227 controls. Psychosis was associated with globally reduced morphometric similarity in all three studies. There was also a replicable pattern of case-control differences in regional morphometric similarity, which was significantly reduced in patients in frontal and temporal cortical areas but increased in parietal cortex. Using prior brain-wide gene expression data, we found that the cortical map of case-control differences in morphometric similarity was spatially correlated with cortical expression of a weighted combination of genes enriched for neurobiologically relevant ontology terms and pathways. In addition, genes that were normally overexpressed in cortical areas with reduced morphometric similarity were significantly up-regulated in three prior post mortem studies of schizophrenia. We propose that this combined analysis of neuroimaging and transcriptional data provides insight into how previously implicated genes and proteins as well as a number of unreported genes in their topological vicinity on the protein interaction network may drive structural brain network changes mediating the genetic risk of schizophrenia.

Keywords: Allen Human Brain Atlas; dysconnectivity; morphometric similarity; network neuroscience; psychosis.

Fig. 1.

Case–control differences in regional morphometric similarity. (A) Case and control distributions of regional morphometric similarity strength, i.e., the average similarity of each region to all other regions, pooling data from all three primary studies. (B) Distributions of morphometric similarity strength for a region with significantly reduced morphometric similarity in cases, namely left hemisphere caudal middle frontal part 1. (C) Regional morphometric similarity averaged over controls from all three datasets. (D) $t$ statistics and Hedge’s $g$ effect sizes for the case–control differences in regional morphometric similarity in each dataset. (E) $t$ statistics for regional case–control differences averaged across datasets in all regions and in the 18 cortical areas where the difference was statistically significant across datasets (FDR = 0.05). (F) Scatterplot of mean control regional morphometric similarity ($x$ axis) vs. case–control $t$ statistic ($y$ axis). Control morphometric similarity (from C) is strongly negatively correlated with case–control morphometric similarity differences (from D; Pearson’s $r$ = −0.76, $P<0.001$). Most cortical regions have positive morphometric similarity in controls, which decreases in cases (47% of regions), or negative morphometric similarity in controls, which increases in cases (36% of regions). Statistically significant regions are circled in red or blue according to whether their mean $t$ statistic increases or decreases, respectively, in patients. MS, morphometric similarity.

Fig. 2.

Gene expression profiles related to case–control differences in morphometric similarity. (A) Scatterplot of regional PLS1 scores (weighted sum of 20,647 gene expression scores) vs. case–control differences in regional morphometric similarity (Cobre dataset). (B) Cortical map of regional PLS1 scores. (C) Cortical map of mean case–control morphometric similarity differences averaged across all datasets. Here, we include intrahemispheric left hemisphere edges only (Materials and Methods). (D) Genes that are strongly positively weighted on PLS1 (e.g., LYSMD4) correlate positively with case–control differences in regional morphometric similarity ($r=0.44$, $P<0.001$), whereas genes that are strongly negatively weighted on PLS1 (e.g., C1orf95) correlate negatively with case–control differences in morphometric similarity ($r=-0.37$, $P<0.001$).

Fig. 3.

Enrichment analysis of genes transcriptionally related to morphometric similarity. (A) PPI network for PLS− genes ($Z<-3$) highlighted with some of the significantly GO-enriched biological processes: nervous system development in red and adenylate cyclase-modulating GPCR signaling pathway in blue. The most interconnected set of proteins was coded by several genes previously implicated in schizophrenia (highlighted in B; details are in the text and SI Appendix, section S8.8).