Bridging the Scales via Personalized Cellular Modeling and Deep Phenotyping in Schizophrenia

TL;DR

This study links synaptic deficits in patient-derived neurons to brain structure and cognitive issues in schizophrenia, offering a new way to understand and treat the disorder.

Contribution

The study establishes a patient-specific bridge from synapse biology to clinical phenotypes in schizophrenia using a multiscale framework.

Findings

Reduced excitatory synapse density and transcriptomic signatures predict individual alterations in brain structure and cognition in schizophrenia.

Genetically driven neuronal gene expression patterns and synapse density correlate with macro-scale brain and cognitive phenotypes in vivo.

The multiscale framework provides a foundation for mechanism-based stratification and precision target identification for cognitive impairment in schizophrenia.

Abstract

Do synaptic deficits in patient-derived neurons predict individual differences in brain circuitry and cognitive impairment in schizophrenia (SCZ)?

In a multiscale framework integrating magnetic resonance imaging, electroencephalography, and cognitive data from more than 500 participants with in vitro phenotyping of donor-matched induced pluripotent stem cell (iPSC)–derived neurons, reduced excitatory synapse density and transcriptomic signatures predicted the individual alterations in brain structure, electrophysiology, and cognition in vivo, providing a mechanistic link from synapse deficits to cognitive impairments in SCZ.

This study establishes a patient-specific bridge from synapse biology to the individual clinical phenotype, offering a road map for mechanism-based target and drug discovery.

While growing evidence implicates synaptic dysfunction as a key pathophysiological mechanism in cognitive impairments in schizophrenia (SCZ), it remains unknown how individual alterations in synaptic connectivity translate into corresponding neural circuit dysfunction and cognitive deficits.

To test whether genetically driven variability in excitatory neurons’ transcriptome and synapse density in patient-derived neurons in vitro explain individual changes in cortical morphology, electrophysiology, and cognitive impairments in vivo.

This multimodal case-control study integrated deep clinical phenotyping (magnetic resonance imaging, electroencephalography, and cognitive assessments) across 2 independent cohorts with schizophrenia and healthy controls (N = 461) with donor-matched induced pluripotent stem cell (iPSC)–derived neurons (n = 80). Machine learning, transcriptome imputation, and reverse dynamic causal modeling were applied to link cellular and systems-level phenotypes. Data were collected between September 16, 2014, and November 10, 2023, and analyzed from January 2022 to January 2026.

The primary outcome was associations between cellular phenotypes (gene expression, synapse density) and individual-level brain structure, electrophysiology, and cognition.

This multiscale translational framework was implemented in 461 individuals with SCZ and healthy controls across 2 independent cohorts. In both cohorts (cohort 1 [C1]: mean [SD] age: 35.1 [11.6] years; 46 female participants [31.1%]; cohort 2 [C2]: mean [SD] age, 36.9 [11.7] years; 140 female participants [44.57%]), cognitive impairments in SCZ were associated with specific gray matter volume reductions across multiple brain regions, in particular the right dorsolateral prefrontal cortex, as well as disturbed electrophysiological activity in the gamma band. Importantly, the individual-level differences in the genetically driven neuronal gene expression patterns and synapse density in vitro predicted the macro-scale alterations of brain structural (C1: r = 0.39; 95% CI, 0.21-0.55; P < .001; iPSC: r = 0.31; 95% CI, −0.07 to 0.60; P = .049; C2: r = 0.23; 95% CI, 0.07-0.37; P = .003), electrophysiological (theta: r = 0.19; 95% CI, 0.04-0.32; P = .05; gamma1: r = 0.17; 95% CI, 0.028-0.31; P = .005; gamma2: r = 0.22; 95% CI, 0.07-0.35; P < .001), and cognitive (C1: r = 0.76; 95% CI, 0.66-0.83; P < .001; iPSC: r = 0.77; 95% CI, 0.57-0.89; P < .001; C2: r = 0.17; 95% CI, 0.02-0.32; P = .02) phenotypes in vivo, providing a mechanistic link from synapse deficits to cognitive impairments in SCZ.

These findings establish a patient-specific link between genetically driven alterations in gene expression, synaptic dysfunction, and large-scale brain and cognitive phenotypes in SCZ. This multiscale framework provides a foundation for mechanism-based stratification and precision target identification for cognitive impairment.

This genetic association study tests whether genetically driven variability in excitatory neurons’ transcriptome and synapse density in patient-derived neurons in vitro explain individual changes in cortical morphology, electrophysiology, and cognitive impairments in vivo.

Figures