Abstract
Uveal melanoma (UM) is a deadly intraocular neoplasm in the adult population and harbors limited therapeutic effects from the current treatment. Here, we aimed to investigate the role of hypoxia in UM progress. We adopted the Cancer Genome Atlas data set as a training cohort and Gene Expression Omnibus data sets as validating cohorts. We first used consensus clustering to identify hypoxia-related subtypes, and the C1 subtype predicted an unfavorable prognosis and exhibited high infiltration of immunocytes and globally elevated immune checkpoint expression. Besides this, the patients with the C1 subtype were predicted to respond to the PD-1 treatment. By the least absolute shrinkage and selection operator algorithm, we constructed a hypoxia risk score based on the hypoxia genes and identified 10 genes. The risk score predicted patient survival with high performance, and the high-risk group also harbored high immunocyte infiltration and immune checkpoint expression. Furthermore, we confirmed that the risk genes were upregulated under hypoxia, and knockdown of CA12 inhibited the epithelial-mesenchymal transition process, clone formation ability, and G1/S phase transformation of the UM cells. The CD276 was also downregulated when CA12 knockdown was performed. These results validate the prognostic role of the hypoxia signature in UM and demonstrate that CA12 is a critical factor for UM cell progression as well as a target to improve immunotherapeutic effects. We believe our study contributes to the understanding of hypoxia's roles in UM and provides a novel target that will benefit future therapeutic strategy development.