Screening from TCM Database@Taiwan and QSAR model for identifying HER2 inhibitors

Abstract

Human epidermal growth factor receptor type 2 (HER2) overexpression, which has been reported to increase the malignancy of human ovarian cancer cells and the metastatic potential of human breast cancer cells, is an important factor in tumour formation and tumour growth. However, currently available HER2 inhibitors, such as Lapatinib, for cancer therapy cause adverse side effects including diarrhoea, rash and possible liver toxicity. We hoped to find novel agents that cause less adverse side effects by performing virtual screening process on the world's largest traditional Chinese medicine compound database. The results thus obtained were then validated using 3D quantitative structure–activity relationship model. Top three candidates were selected from the docking results. The top three candidates and the control both formed a hydrogen bond with the key residue, Lys724. This showed that the candidates and the control have similar binding effects to HER2. These candidates were investigated using comparative molecular field analysis and comparative molecular similarity indices analysis models. The results from these models showed high correlation coefficients (r 2) of 0.9547 and 0.9226, respectively. All top three candidates had high docking scores, favourable pharmacophores and functional groups forming stable hydrogen bonds with HER2. These properties suggested stable binding affinities and favourable interaction with HER2. We concluded that these candidates may be further investigated as potential HER2 inhibitors.