Abstract / Summary
Abstract Background The blood-brain barrier (BBB), the brain’s obligate energy gateway, has to our knowledge not been examined by spatially resolved profiling in any psychiatric disorder. Metabolic dysregulation is implicated in schizophrenia spectrum disorders (SSD), yet prior post-mortem evidence derives largely from bulk dorsolateral prefrontal cortex (DLPFC) tissue, in which vascular and parenchymal signals are mixed. The insular cortex, a salience-network hub implicated in SSD, has received little molecular attention. Methods We used the GeoMx Digital Spatial Profiler with the Human Whole Transcriptome Atlas to profile post-mortem insular cortex from 8 donors with schizophrenia or schizoaffective disorder and 8 matched controls (Netherlands Brain Bank-Psychiatry). Under neuropathological guidance, three CD31-positive (BBB) and three CD31-negative (parenchymal) areas of interest (AOIs) were delineated per donor in cortical layers III to VI, yielding 96 profiles of which 95 passed quality control. Results Endothelial enrichment of the BBB AOIs was confirmed by five canonical markers. Unsupervised clustering separated BBB from parenchyma, and within the BBB SSD samples formed a distinct subcluster. Gene set enrichment analysis identified three glucose metabolism pathways enriched in SSD BBB and seven mitochondrial energy metabolism pathways enriched in SSD parenchyma. At gene level, reported as exploratory because no gene survived correction for multiple testing, the endothelial glycolytic regulator PFKFB3 was increased in SSD BBB despite reduced expression of the transporters GLUT1 and MCT1 , suggesting a compensatory downregulation of glucose and lactate transporters associated with heightened glycolytic flux. Additionally, the mitochondrial gene COX7A1 was increased in parenchyma. Conclusions The BBB may contribute to the metabolic pathophysiology of SSD. Expression of energy metabolism pathways was increased in both compartments, opposite in direction to the reductions reported in DLPFC, suggesting that metabolic changes in SSD may be region-specific rather than generalised, and that bulk tissue analysis averages out a distinct BBB signal.