Synergistic Targeting of Vemurafenib-Resistant Melanoma via Network-Guided Drug Combination and Biophysical Validation

Vemurafenib, a selective inhibitor targeting the BRAF V600E mutation, has improved outcomes in advanced melanoma. However, resistance frequently emerges, limiting its long-term effectiveness and highlighting the need for combination therapies. This study aimed to identify synergistic drug combinations to overcome vemurafenib resistance in BRAF-mutant melanoma using an integrative computational approach. We applied a modified version of SynGeNet, which combines gene expression data with protein–protein interaction networks, to predict effective drug combinations in the vemurafenib-resistant A375 melanoma cell line. The analysis identified sorafenib and pioglitazone as the most promising candidates. Sorafenib is a multi-kinase inhibitor targeting signaling pathways involved in proliferation and angiogenesis, while pioglitazone activates PPARγ and modulates stress responses. Transcriptomic profiling of resistant cells revealed enrichment in nucleotide metabolism, protein trafficking, and corticosteroid response pathways, alongside suppression of mitotic and cell cycle processes. In vitro validation confirmed that the sorafenib/pioglitazone combination reduces cell viability with a strong synergistic effect. We further applied a novel biophysical platform integrating QCM-D and lectin–glycan interaction analysis to assess glycosylation dynamics. The combination treatment reduced the glycan viscoelastic index, suggesting a shift toward a less metastatic phenotype. The experimental design integrates computational drug synergy prediction, transcriptomic profiling, and biophysical glycan analysis. A375 human melanoma cells, either vemurafenib-sensitive or -resistant, were treated with sorafenib and pioglitazone individually and in combination. Resistance was induced by gradual vemurafenib exposure. RNA-Seq was performed on biological replicates from each condition. Libraries were prepared from total RNA and sequenced on an Illumina platform in paired-end mode. Gene expression data were used to evaluate transcriptional responses and pathway enrichment. Additionally, whole-cell glycomic profiling was conducted using a quartz crystal microbalance with dissipation monitoring (QCM-D) and lectin-binding analysis. Changes in the glycan viscoelastic index (gVI) served as a surrogate for metastatic potential and phenotypic plasticity under treatment. The approach supports the identification of therapeutic synergies and biomarkers associated with resistance and EMT-like transitions.