A3A-ligand evolution against anti-hypermutational human metastasis

Many human tumor-cells during carcinogenesis acquire expanding metastasis by heterogeneous-cell hypermutations. In humans, most hypermutations are induced by APOBEC3A enzymes (A3A). To reduce hypermutation levels, drug-like inhibitors are one of the searched A3A-targeted metastatic therapies. However, development of anti-A3A inhibitors have been limited because side-effect risks to other related human A3 enzymes (B to H and AID) with healthy functions, such as antiviral defenses and/or antibody diversification. To computationally predict large numbers of highly specific inhibitor candidates, the ssDNA-binding and dimer-interface of the human A3A have been targeted here by mimicking natural evolution as accelerated by co-evolutionary docking. After randomly generating tens of thousands of raw-children anti-A3A, hundreds of candidates were selected by their low toxicity, best-fitting and consensus affinities. For each A3A cavity, co-evolution 2-4 iterations were capable of predicting children within low nanomolar affinity ranges. Those anti-A3A candidates predicting significant cross-docking to A3B, A3C, A3D, A3F, A3G or AID, were eliminated. The final anti-A3A list of candidates may be tested for anti-hypermutation activities in cell culture or animal tumor models