Reinvestigation of Passerini and Ugi scaffolds as multistep apoptotic inducers via dual modulation of caspase 3/7 and P53-MDM2 signaling for halting breast cancer

Abstract

Selective induction of breast cancer apoptosis is viewed as the mainstay of various ongoing oncology drug discovery programs. Passerini scaffolds have been recently exploited as selective apoptosis inducers via a caspase 3/7 dependent pathway. Herein, the optimized Passerini caspase activators were manipulated to synergistically induce P53-dependent apoptosis via modulating the closely related P53-MDM2 signaling axis. The adopted design rationale and synthetic routes relied on mimicking the general thematic features of lead MDM2 inhibitors incorporating multiple aromatic rings. Accordingly, the cyclization of representative Passerini derivatives and related Ugi compounds into the corresponding diphenylimidazolidine and spiro derivative was performed, resembling the nutlin-based and spiro MDM-2 inhibitors, respectively. The study was also extended to explore the apoptotic induction capacity of the scaffold after simplification and modifications. MTT assay on MCF-7 and MDA-MB231 breast cancer cells compared to normal fibroblasts (WI-38) revealed their promising cytotoxic activities. The flexible Ugi derivatives 3 and 4, cyclic analog 8, Passerini adduct 12, and the thiosemicarbazide derivative 17 were identified as the study hits regarding cytotoxic potency and selectivity, being over 10-folds more potent (IC₅₀ = 0.065–0.096 μM) and safer (SI = 4.4–18.7) than doxorubicin (IC₅₀ = 0.478 μM, SI = 0.569) on MCF-7 cells. They promoted apoptosis induction via caspase 3/7 activation (3.1–4.1 folds) and P53 induction (up to 4 folds). Further apoptosis studies revealed that these compounds enhanced gene expression of BAX by 2 folds and suppressed Bcl-2 expression by 4.29–7.75 folds in the treated MCF-7 cells. Docking simulations displayed their plausible binding modes with the molecular targets and highlighted their structural determinants of activities for further optimization studies. Finally, in silico prediction of the entire library was computationally performed, showing that most of them could be envisioned as drug-like candidates.