Dysregulation of Heat Shock Proteins in Auditory Hallucinations of Schizophrenia: Insights from Molecular, Neuroimaging, and Machine Learning PerspectivesReceived
نویسندگان
1 Assistant Professor, Department of Psychiatry, Arak University of Medical Sciences, Arak, Iran.
doi
10.22111/jep.2025.53542.1098چکیده
Schizophrenia (SCZ) is a complex psychiatric disorder affecting ~1% of the global population, marked by hallucinations, delusions, and cognitive deficits that impair daily life. Genetic, environmental, and neurodevelopmental factors underpin its etiology, yet molecular mechanisms are unclear. Heat shock proteins (HSPs)—molecular chaperones like HSP70/HSPA8 (HSP1), HSP90/HSP90AA1 (HSP3), and HSP40/DNAJ (HSP4)—maintain proteostasis, aiding protein folding and stress responses. In the brain, they protect neurons from oxidative stress and inflammation, key in SCZ pathogenesis. Proteostatic failures may drive neuronal misfiring linked to hallucinations.The role of HSPs in SCZ hallucinations remains underexplored. These symptoms stem from dysregulated dopamine and sensory processing in limbic-cortical networks. Systematic reviews link elevated HSPs to brain changes: prefrontal cortex (PFC) volume loss, with HSP-driven gliosis and synaptic pruning deficits disrupting executive function and reality testing. Hippocampal atrophy, tied to memory distortions fueling hallucinations, involves HSP dysregulation in clearing amyloid-like proteins, sparking neuroinflammation.fMRI shows HSPs affecting functional connectivity, like weakened default mode network (DMN) integrity, blurring self-external boundaries. Machine learning integrates this: models using TCGA/ICGC transcriptomics predict SCZ outcomes at ~85-90% accuracy (initial training accuracy of 92% tempered via rigorous 10-fold cross-validation with nested hyperparameter tuning; overfitting mitigated by L2 regularization, early stopping, and recursive feature elimination; independent validation on Psychiatric Genomics Consortium [PGC] data [n=500] confirmed 82% accuracy, identifying genes like PDE4D (cAMP modulation), PDP1 (mitochondrial stress), and RORA (circadian disruption).HSPs, along with potential epigenetic regulators, promise as biomarkers for early diagnosis via assays or imaging, enabling personalized therapies like HSP inducers (e.g., geranylgeranylacetone) to restore balance. Longitudinal studies are needed to track dynamics.This review merges molecular, neuroimaging, and ML data to clarify HSPs in hallucinations, proposing targeted therapies for precision psychiatry and reducing SCZ's burden.