RESUMEN
A collection of 9050 natural products, their derivatives, and mimetics, was virtually screened against the human Atg3-Atg8 (Atg - autophagy) binding scaffold. By blocking this interaction, the lipidation of Atg8 does not occur and the formation of autophagosomes is inhibited. Forty-three (43) potential ligands were tested using enhanced Green Fluorescent Protein (eGFP) tagged LC3, the human ortholog of Atg8, in MCF7 breast cancer cells. Three hits showed single digit µM IC50 values with AT110, an isoflavone derivative, being the best at 1.2 ± 0.6 µM. Molecular modelling against Atg8 in conjunction with structural activity relationship (SAR) strongly supports the binding to this target. Testing in a panel of cancer cell lines showed little cytotoxic effect as compared to chloroquine. However, same concentration of AT110 was shown to be toxic to young zebrafish embryos. This can be explained in terms of the autophagy process being very active in the zebrafish embryos rendering them susceptible to AT110 whereas in the cancer cells tested the autophagy is not usually active. Nevertheless, AT110 blocks autophagy flux in the zebrafish confirming that the ligand is modulating autophagy. A small molecule non-cytotoxic autophagy inhibitor would open the door for adjunct therapies to bolster many established anticancer drugs, reducing their efficacious concentration thus limiting undesirable site effects. In addition, since many cancer types rely on the autophagy mechanism to survive a therapeutic regime, recurrence can potentially be reduced. The discovery of AT110 is an important step in establishing such an adjunct therapy.