miRNA centered regulatory networks identify FN1 and miR27b as metastatic drivers in HPV negative head and neck cancer.

The metastasis of human papilloma virus (HPV)-negative head and neck squamous cell carcinoma (HNSC) is associated with poor prognosis, yet the regulatory mechanisms remain unclear. In this study, we developed a metastasis prediction model using the transcriptomes of 333 HPV-negative HNSC tumor samples, representing one of the largest HPV-negative HNSC sample sets to date. Our model, which leverages miRNA-centered regulatory networks extended by regulome data, showed superior predictive performance with an AUC of 0.805 compared to existing metastasis-related biomarkers. We identified three key miRNAs in our model with significant differential expression in metastatic patients. Notably, miR-27b was downregulated and its target gene FN1 was overexpressed, leading to activation of the cell adhesion pathway. Single-cell RNA-seq analysis confirmed FN1 overexpression specifically in the malignant cells of metastasis patients with HPV-negative HNSC. In vitro, miR-27b was found to reduce metastatic potential by negatively regulating FN1 expression at the translational level. These findings suggest FN1, driven by miR-27b, as a potential driver of metastasis in HPV-negative HNSC, offering new avenues for biomarker development and targeted therapeutic strategies.