Abstract
Neuroblastoma (NB) is the most common solid tumor apart from central nervous system malignancies in children aged 0-14 years, and the outcomes of high-risk patients are dismal. High mobility group box 3 (HMGB3) plays an oncogenic role in many cancers; however, its biological role in NB is still unclear. Using data mining, we found that HMGB3 expression was markedly elevated in NB patients with unfavorable prognoses. When HMGB3 expression in NB cell lines was inhibited, cell proliferation, migration, and invasion were suppressed, and HMGB3 knockdown inhibited NB tumor development in mice. RT-PCR was employed to detect mRNA expression of nine coexpressed genes in response to HMGB3 knockdown, and TPX2 was identified. Furthermore, overexpression of TPX2 reversed the cell proliferation effect of HMGB3 silencing. Multivariate Cox regression analysis indicated that HMGB3 and TPX2 might be independent prognostic factors for overall survival and event-free survival, which showed the highest significance (p < 0.001). According to the nomogram predictor constructed, the integration of gene expression and clinicopathological features exhibited better prognostic prediction power. Furthermore, the random forest algorithm and receiver operating characteristic curves also showed that HMGB3 and TPX2 played important roles in discriminating the vital status (alive/dead) of patients in the NB datasets. Our informatics analysis and biological experiments suggested that HMGB3 is correlated with the unfavorable clinical outcomes of NB, and plays an important role in promoting cell growth, proliferation, and invasion in NB, potentially representing a new therapeutic target for tumor progression.