Vellucci, Licia (2025) MicroRNA profilation from serum and total and neuronal extracellular vesicles in patients affected by schizophrenia and treatment-resistant schizophrenia. [Tesi di dottorato]

Documento PDF vellucci_licia_38.pdf Download (5MB)

Abstract

Background: Treatment-resistant schizophrenia (TRS), defined as the persistence of symptoms despite adequate antipsychotic treatment, remains frequently underrecognized, with clozapine, the gold standard, introduced only after an average delay of 4–9 years. In this study, we aimed to identify which Positive and Negative Syndrome Scale (PANSS) factors, executive function measures, and selected microRNAs (miRNAs) are most strongly associated with TRS, and to investigate the molecular underpinnings targeted by these miRNAs, thereby hypothesizing aberrant mechanisms contributing to the TRS condition. Methods: We examined 60 patients with schizophrenia (25 treatment-responsive, 35 TRS) following a structured retrospective-prospective evaluation of antipsychotic response. Seventeen serum miRNAs were assessed. Principal Component Analysis (PCA) identified clinical variables and miRNAs with loading scores ≥ 0.4, which were used to train three machine learning models. Random Forest (RF) provided the best balance of accuracy, sensitivity, specificity, and generalizability. Based on RF with nested cross-validation, selected miRNAs underwent Gene Ontology (GO) and KEGG enrichment analyses with Benjamini–Hochberg correction. Results: PCA highlighted PANSS emotional (EMO), disorganization (DIS), and excitement (EXC) factors, processing speed, and miRNAs (miR-132-3p, miR-181-5p, miR-203a-3p, miR-199a-5p, miR-128-3p). RF classified TRS patients based on higher EMO, DIS, EXC, miR-132-3p, miR-181- 5p, and miR-203a-3p, with 62.9% sensitivity, 60% specificity, and 0.71 AUC. GO analysis indicated that miR-132, miR-181b, and miR-203a regulate stress response, apoptosis, cell cycle, differentiation, and protein homeostasis, converging on synaptic signaling and neuroplasticity pathways (ErbB, glutamatergic, dopaminergic). Predicted targets included key neurotransmission genes (DRD1, HOMER1, MAPK1/3, GRIN2A/D), also modulated by antipsychotics. Discussion: TRS is characterized by a combined clinical-molecular signature involving PANSS domains, and selected miRNAs converging on synaptic and neuroplasticity networks. Dopaminergic, glutamatergic and ErbB pathways emerged as central hubs, suggesting that antipsychotic effects may be mediated, at least in part, through miRNA-driven mechanisms underlying TRS.

Downloads

Actions (login required)

Modifica documento