CHARACTERIZATION OF PSYCHOSIS BASED ON IMMUNO-METABOLIC MARKERS: EX VIVO AND IN VIVO ANALYTICAL APPROACHES

Background Systemic immune dysregulation has been implicated in psychosis pathophysiology, potentially contributing to white matter (WM) alterations and neuronal dysfunction. While immune-brain interactions have been observed in neuropsychiatric disorders, the underlying mechanisms, including potential lipidic changes in brain cells, remain poorly understood. This study integrates in vivo and post-mortem analyses to explore immune-FA interactions and lipidomic profiles in psychosis, addressing both diagnostic-specific immune contributions to WM integrity and molecular alterations in the dorsolateral prefrontal cortex (DLPFC). Methods In the in vivo analysis, we examined 39 first-episode psychosis (FEP) patients and 48 healthy controls (HC) from the GET UP cohort. Gene expression levels of immune markers were quantified in peripheral blood using qPCR. Fractional anisotropy (FA) was assessed via diffusion tensor imaging (DTI) and tract-based spatial statistics (TBSS). Interaction models between immune markers and FA were tested across WM tracts. In the ex vivo analysis, MALDI imaging mass spectrometry (MALDI-IMS) was performed on DLPFC brain tissue from 5 individuals (aged 21–39 years) to map lipid species at subregional and layer-specific levels. Results Significant gene-FA interactions were observed for five immune markers (CXCR3, CD4, CD27, CD38, and IL12a), with p-values < 0.001 (TFCE-corrected). CXCR3 and CD4 exhibited diagnostic group-dependent discordant trends, highlighting differential immune contributions to WM microstructure in FEP and HC. For instance, CXCR3 correlated positively with FA in FEP and negatively in HC, suggesting a shift toward maladaptive neuroinflammatory responses in psychosis. CD27 displayed opposing correlations, reflecting divergent immune-WM remodeling roles between groups. Regions of significant FA alteration included the corpus callosum, cingulum, and superior longitudinal fasciculus. MALDI-IMS revealed distinct lipid distributions between WM and gray matter (GM), with sulfatides enriched in WM and phosphatidylinositols and gangliosides predominant in GM. Supra- and infra-granular GM layers demonstrated specific lipid patterns, reflecting metabolic and neuronal functions. These findings validated the methodological rigor of MALDI-IMS and highlighted its utility for capturing spatial lipidomic heterogeneity. Conclusions and Perspectives This study underscores the diagnostic specificity of immune-FA interactions in early psychosis and establishes MALDI-IMS as a robust methodological platform for spatial lipidomics. The observed immune-mediated WM alterations and lipid profiles offer novel insights into psychosis pathophysiology. Future research will focus on integrating lipidomic data with in vivo imaging, validating findings in external cohorts, and expanding ex vivo analyses to early psychosis brain samples. These approaches aim to refine diagnostic tools and identify therapeutic targets bridging molecular and systemic changes in psychosis.