High glucose promotes cisplatin chemoresistance in MDA-MB-231 breast cancer derived cells through changes in gene expression and multiple signaling pathways.

Epidemiological and experimental evidence indicate that hyperglycemia, transient hyperinsulinemia and insulin-like growth factor-1 promote an aggressive phenotype of breast tumor cells during diabetes mellitus. The present study evaluated the effect of high glucose (HG) concentration (30 mmol/l) on cisplatin sensitivity and gene expression in MDA-MB-231 triple-negative breast cancer (TNBC)-derived cells. Cisplatin cytotoxicity was assessed by the MTT assay and was attenuated by HG. This effect was accompanied by reduced caspase-3 activation, indicating impaired apoptosis. Differentially expressed genes between HG and normal glucose (5.6 mmol/l) conditions were analyzed by microarray. The upregulation of tetraspanin 1 (TSPAN1) and frizzled 3 (FZD3) was validated by reverse transcription-quantitative PCR. HG promotes cisplatin chemoresistance by preventing apoptosis. Additionally, HG altered the expression of genes involved in glucose, glycerophospholipid, purine and pyrimidine metabolism. Regulatory pathways affected included Wnt/β-catenin, p53 network, NF-κB survival and DNA damage response. Candidate genes potentially associated with HG-induced chemoresistance included SLC6A6, TSPAN1, TNFSF7, PDE4B, RPS13, COX4I1, RAD17, LRP8, SMARCA2, EHD1, LEF1 and EIF4G2. Survival analysis using the KM-Plotter platform revealed that YBX3 and amphiregulin expression were significantly associated with poor prognosis. By contrast, amyloid precursor protein (APP) and N-cadherin (CDH2) showed non-significant trends. Coexpression analyses demonstrated that APP and CDH2 were positively correlated with YBX3. This suggests they may contribute to tumor aggressiveness as part of YBX3-driven regulatory networks, rather than as independent prognostic markers. Collectively, the present findings demonstrated that HG promotes cisplatin resistance and aggressive features in TNBC cells through multiple metabolic and signaling pathways. Notably, these include the Wnt/β-catenin axis. Candidate biomarkers with potential prognostic and therapeutic relevance in diabetic breast cancer (BC) were also identified.