Overexpression of HMGB3 and its prognostic value in breast cancer

Abstract

Background: High mobility group protein B3 (HMGB3) is abundantly expressed in a number of malignancies, contributing to tumor cell growth and predicting poor outcomes. More research on the connection between HMGB3 and breast cancer is needed. The prognostic significance of HMGB3 in breast cancer was examined and validated in this study.

Methods: Using The Cancer Genome Atlas (TCGA) database RNA sequencing and clinical data, we investigated the associations between HMGB3 expression and tumor mutations, prognosis, and immune infiltration in breast cancer. The Gene Expression Profiling Interactive Analysis (GEPIA), Tumor Immune Estimation Resource (TIMER), breast cancer gene-expression miner (bc-GenExMiner), UALCAN, OncoLnc, cBio Cancer Genomics Portal (cBioPortal), and LinkedOmics databases were applied to examine the levels of expression, mutation, coexpression, and immune correlation of HMGB3 in breast cancer. cBioPortal and the Database for Annotation, Visualization, and Integrated Discovery (DAVID) were used for coexpression and enrichment analyses, respectively. Experimental tests and a separate cohort of breast cancer patients in our center were used for validation. To determine independent risk factors affecting breast carcinoma prognosis, multivariate Cox regression analysis was performed. The Kaplan-Meier method was applied to analyze the connection between HMGB3 expression and overall survival time in breast cancer.

Results: Pan-cancer investigation using the GEPIA and UALCAN databases revealed a high level of HMGB3 expression in different malignancies, including breast cancer. HMGB3 might be a potential diagnostic biomarker, according to the receiver operating characteristic (ROC) curve (AUC=0.932). And immunohistochemistry confirmed higher HMGB3 protein expression in breast cancer tissues in clinical samples. Experimental tests also showed that breast cancer cells have higher expression of HMGB3, and knockdown of HMGB3 can promote the proliferation of breast cancer cells and increase sensitivity to chemotherapy. Human epidermal growth factor receptor 2 (HER2), Nottingham Prognostic Index (NPI), basal-like status, nodal status (N+), triple-negative status, and Scarff-Bloom-Richardson (SBR) grade all showed positive correlations with HMGB3 expression. Conversely, HMGB3 expression was negatively associated with the expression of estrogen receptor (ER) and progesterone receptor (PR) in breast cancer. Breast cancer patients with high HMGB3 expression had poor overall survival, which was validated by an analysis of a separate cohort of breast cancer patients in our center. Cox regression analysis identified high HMGB3 expression as an independently associated risk factor for breast carcinoma. The amount of immunological infiltration was substantially linked with the high expression of HMGB3. The chromosome centromeric region, ATPase activity, and the cell cycle are critical areas where HMGB3 is involved, according to enrichment analysis. Therefore, we suspected that HMGB3 might be a potential biomarker for detecting and treating breast carcinoma.

Conclusion: Breast cancer tissues had higher HMGB3 expression than normal breast tissues. HMGB3 overexpression may serve as an indicator for poor breast cancer outcomes.

Keywords: HMGB3; bioinformatic analysis; biomarker; breast cancer; prognosis.