[Laryngeal squamous cell carcinoma-derived exosomes promote neuronal axonal growth by remodeling the neural...

Objective: Perineural invasion (PNI) is a critical determinant of poor prognosis in laryngeal squamous cell carcinoma (LSCC), but its underlying mechanisms remain unclear. This study aimed to investigate whether LSCC-derived exosomes induce axonal growth by delivering neuroactive molecules, thereby contributing to tumor perineural invasion. Methods: Clinical data from the laryngeal cancer cohort of The Cancer Genome Atlas Head and Neck Squamous Cell Carcinoma (TCGA-HNSC) dataset were analyzed. Propensity score matching (PSM) and Cox regression were used to evaluate the prognostic value of nerve density, and these findings were validated using 35 pairs of laryngeal cancer and adjacent normal tissue specimens collected at Yantai Yuhuangding Hospital between 2022 and 2026 to assess neural morphological changes. Exosomes were isolated from the human LSCC cell line AMC-HN-8, characterized by quality-control assays, and co-cultured with PC12 cells. A rescue experiment using GW4869, a specific inhibitor of neutral sphingomyelinase, was performed to confirm the exosome-dependent effect. Neurite outgrowth was evaluated by immunofluorescence, and the expression of axonal growth-related genes was measured by RT-qPCR. Targeted metabolomics was employed for the absolute quantification of neuroactive metabolites within the vesicles and for pathway enrichment analysis. Results: After PSM adjustment, high nerve density was identified as an independent poor prognostic factor in LSCC patients (HR=2.10, P=0.035), with particularly pronounced prognostic value in the early-stage node-negative (N0) subgroup (HR=4.07, P=0.001). Pathological sections showed high expression of the neural markers βIII-tubulin and PGP9.5 in LSCC tissues (βIII-tubulin: t=2.234, P<0.05; PGP9.5: t=2.575, P<0.05). Exosomes were successfully isolated from AMC-HN-8 cells and passed quality control. In vitro assays showed that LSCC-derived exosomes significantly promoted neurite extension and branching in PC12 cells (t=4.147, P<0.000 1) and upregulated core axonal growth genes, including GAP-43, NEFL, and NEFM (GAP-43: t=3.698, P<0.05; NEFL: t=5.113, P<0.01; NEFM: t=5.263, P<0.01); this effect was completely reversed by the exosome-release inhibitor GW4869 (t=3.535, P<0.001). Targeted metabolomics revealed a specific enrichment of 12 neurotransmitters and metabolites within LSCC exosomes, centered on glutamine (83.411 μmol/L, FC=1.88) and glutamate (18.461 μmol/L, FC=1.21), which were significantly enriched in signaling pathways such as "central carbon metabolism in cancer" and "glutamatergic synapse". Conclusion: Nerve density is a potential adverse prognostic factor in patients with LSCC. LSCC-derived exosomes can directly induce axonal growth in neuron-like cells, suggesting that tumor cells actively remodel the neural microenvironment and drive axonal growth through exosome-mediated long-range signaling.