Multi-gene anoikis signature combined with tumor microenvironment stratifies prognosis in cutaneous melanoma with experimental validation of PTK6.

BACKGROUND

Cutaneous melanoma (CM) exhibits aggressive behavior driven by anoikis resistance and dysregulated tumor microenvironment (TME). The combined prognostic value of anoikis susceptibility and immune landscape remains uncharacterized.

METHODS

Using transcriptomic data from 471 TCGA-CM patients, we developed an 11-gene anoikis risk score (RS) through LASSO-Cox regression, with external validation in the GSE65904 cohort (n = 214). This risk score was integrated with ESTIMATE-quantified immune/stromal features to construct the RS-TME classifier. Functional validation focused on Protein Tyrosine Kinase 6 (PTK6), a key prognostic driver, employing in vitro and in vivo models.

RESULTS

We established an 11-gene RS with superior predictive capacity compared to conventional clinical indicators. Consensus clustering identified two molecular subtypes where Cluster B showed significantly improved survival compared to Cluster A (p < 0.001). Single-cell RNA sequencing confirmed immune cells, particularly macrophages, possessed substantially higher anoikis scores than malignant cells. This microenvironment foundation guided development of the RS-TME classifier, which stratified patients into three prognostic groups. Low-risk/high-TME patients exhibited significantly superior survival outcomes (p < 0.001), with external validation in the GSE65904 cohort (p = 0.002). Crucially, PTK6 knockdown suppressed tumor progression through increased apoptosis, reduced cell viability, impaired migration, and decreased tumor volume, mediated by Hippo pathway activation.

CONCLUSIONS

The RS-TME classifier is the first model integrating anoikis resistance and immune microenvironment to predict survival and immunotherapy response. Simultaneously, PTK6 targeting suppresses CM progression via Hippo-mediated anoikis resistance, providing a foundation for personalized treatment strategies.