Prostate cancer (PCa) exhibits marked metabolic heterogeneity, yet the prognostic implications of amino acid metabolism remain insufficiently characterized. In this study, we developed a 9-gene amino acid metabolic risk signature through an integrative analysis of single-cell and bulk transcriptomic datasets, leveraging machine learning to stratify patients into distinct prognostic subgroups. The model demonstrated robust predictive accuracy in both TCGA and independent GEO cohorts, with significant associations to immune microenvironment remodeling and therapeutic vulnerabilities. Mechanistically, multi-omics analyses (SCENIC, CellChat, pseudotime trajectory) delineated regulatory networks underlying amino acid metabolic dysregulation, highlighting FUS as a potential oncogenic regulator. Experimental validation across cellular, murine, and human models supported a role for FUS in promoting tumor aggressiveness. Through bioinformatic analysis, we identified potential signaling pathways underlying FUS involvement in prostate cancer progression. Our study establishes a clinically actionable amino acid metabolic signature and nomogram for PCa risk stratification, while suggesting FUS as a candidate therapeutic target. These findings bridge computational discovery with mechanistic validation, providing novel insights into the amino acid metabolic dependencies that govern prostate cancer (PCa) progression.

SLC6A14 is an exosomal protein derived from bronchial epithelial cells. This study aims to investigate how exosomal SLC6A14 contributes to airway inflammation and mucus hypersecretion in chronic obstructive pulmonary disease (COPD).An in vitro COPD model was established by treating normal human bronchial epithelial cells (BEAS-2B) with 8% cigarette smoke extract (CSE) for 48 h. Interventions included SLC6A14 overexpression, ZFP36L1 overexpression, and exosome treatment. For the in vivo experiments, a COPD mouse model was induced by long-term cigarette smoke (CS) exposure for 6 months, and exosomes were administered during the final 2 weeks. Levels of inflammatory cytokines and mucin secretion were measured by RT-qPCR, ELISA, and Western blot. Lung tissue pathology, fibrosis, goblet cell hyperplasia, and mucus expression were analyzed using HE staining, Masson staining, immunohistochemistry, and multiplex immunofluorescence. An in vitro COPD model was established by treating normal human bronchial epithelial cells (BEAS-2B) with 8% cigarette smoke extract (CSE) for 48 h. Interventions included SLC6A14 overexpression, ZFP36L1 overexpression, and exosome treatment. For the in vivo experiments, a COPD mouse model was induced by long-term cigarette smoke (CS) exposure for 6 months, and exosomes were administered during the final 2 weeks. Levels of inflammatory cytokines and mucin secretion were measured by RT-qPCR, ELISA, and Western blot. Lung tissue pathology, fibrosis, goblet cell hyperplasia, and mucus expression were analyzed using HE staining, Masson staining, immunohistochemistry, and multiplex immunofluorescence. SLC6A14 expression was significantly upregulated in the lungs of CS-exposed mice and in CSE-treated BEAS-2B cells. Its overexpression triggered inflammatory responses and mucin expression in BEAS-2B cells irrespective of CSE treatment. Mechanistically, SLC6A14 could enhance IL-8 mRNA stability by suppressing the RNA-binding protein ZFP36L1. Comparatively, ZFP36L1 overexpression reduced CSE-induced inflammation and mucin expression in BEAS-2B cells by decreasing IL-8 mRNA. Exosomes released from CSE-treated BEAS-2B cells, which were enriched in SLC6A14, promoted inflammation and mucin secretion in recipient BEAS-2B cells. Importantly, exosomes isolated from the SLC6A14-silenced and CSE-treated BEAS-2B cells significantly alleviated airway inflammation, goblet cell hyperplasia, fibrosis, and mucus hypersecretion in CS-exposed mice. CSE stimulated bronchial epithelial cells to release exosomes enriched with SLC6A14, and these exosomes aggravated COPD-related airway inflammation and mucus hypersecretion by disrupting the ZFP36L1/IL-8 signaling axis.