Abstract
Numerous modern scientific studies have demonstrated that animal venoms harbor a wealth of diverse anticancer active components, serving as a valuable resource for the development of natural antitumor drugs. AI-based computation and prediction models enable rapid screening of extensive active peptides. In this study, the anticancer activity of seven peptides was predicted using our previous deep learning model. Further verification experiments confirmed that Lpep3 can selectively and efficiently inhibit the growth of leukemia cells. Electron microscopy observations revealed cell shrinkage in morphology and honeycomb-like perforations on the cell membrane in the treated group. It is hypothesized that high-concentration peptides disrupt the cell membrane and increase cell permeability, which was confirmed by trypan blue staining and Calcein-AM/PI double-staining assays. Lpep3 induces the release of lactate dehydrogenase (LDH) and ATP in a concentration-dependent manner, further suggesting that this peptide disrupts the cell membrane. In addition, although Lpep3 does not affect the cell cycle of MV-4-11, it can induce cell apoptosis. Western blotting and RT-qPCR results showed that compared with the control group, the expression levels of Bax were upregulated, while the expression level of Bcl-2 protein was downregulated in the Lpep3 group. In vivo experiments demonstrated that Lpep3 has good biological safety, and compared with the control group, the Lpep3 group could inhibit the growth of tumor cells in mice. Collectively, Lpep3 is characterized by high potency and specificity and may serve as a promising lead compound for the development of anti-leukemia drugs.