Identify the PANoptosis signature and prognostic model via a multimachine-learning computational framework for bladder urothelial carcinoma

Objective: As accumulating evidence suggests that pan-opoptosis plays a significant role in tumor progression, it is essential to elucidate its implications for tumor prognosis and treatment. This study aims to characterize the pan-apoptotic features of patients with bladder urothelial carcinoma (BLCA) and to develop a novel model to guide clinical diagnosis and treatment, while further investigating the associated molecular mechanisms underlying tumor progression. Methods: Firstly, samples with BLCA were divided into two clusters based on the expression of PANoptosis genes. Subsequently, 369 PANoptosis-associated genes were identified through differential expression gene analysis. A novel model was then developed by integrating Cox regression analysis with four machine learning algorithms. Furthermore, Immunohistochemistry, EdU assays, Quantitative real-time PCR, and Western blotting experiments were employed to validate the model. Results: The PANoptosis model was developed and has demonstrated robust performance in prognostic prediction. And the high PANS group had higher TIDE scores than low PANS group, which suggested that the low PANS group obtained more benefit from ICB treatment than high PANS group. Moreover, our study revealed that GNLY was highly expressed in NK cells, while VSIG2 exhibits elevated expression levels in tumor cells. Notably, the expression of VSIG2 was found to be positively correlated with the degree of malignancy in BLCA. Meanwhile, the function of VSIG2 in BLCA has been explored and the results showed the proliferation capacity of BLCA cells diminished following the knockdown of VSIG2. Finally, our research identified compounds or drugs targeting VSIG2 through molecular docking techniques. And the small molecule compound quercetin could target the VSIG2 protein, effectively reversing the enhanced proliferative capacity of BLCA induced by VSIG2 overexpression. Conclusion: The PANoptosis model could accurately predict the prognosis of BLCA patients and guide the treatment of BLCA. Meanwhile, the treatment of patients with high expression of VSIG2 by the small molecule compound quercetin has opened up a new direction for clinical treatment.