A SAR study has been carried out around a modified scaffold of the natural product psammaplin A obtained by replacing the o-bromophenol unit by an indole ring. A series of indole psammaplin A constructs were generated in a short synthetic sequence that starts with the functionalization of the C3 indole position with in situ generated nitrosoacrylate, and this is followed by protection of the β-indole-α-oximinoesters, saponification, condensation with symmetrical diamines, and deprotection. Biochemical and cellular characterization using U937 and MCF-7 cells confirmed that many of these analogues displayed more potent actitivies than the parent natural product. Moreover, in addition to the reported HDAC and DNMT dual epigenetic inhibitory profile of the parent compound, some analogues, notably 4a (UVI5008), also inhibited the NAD(+)-dependent SIRT deacetylase enzymes. The SAR study provides structural insights into the mechanism of action of these multiple epigenetic ligands and paves the way for additional structural exploration to optimize their pharmacological profiles. Because of their multi(epi)target features and their action in ex vivo samples, the indole-based psammaplin A derivatives are attractive molecules for the modulation of epigenetic disorders.

Retinoid-related molecules with an adamantyl group (adamantyl arotinoids) have been described with selective activities towards the retinoid receptors as agonists for NR1B2 and NR1B3 (RARβ,γ) (CD437, MX3350-1) or RAR antagonists (MX781) that induce growth arrest and apoptosis in cancer cells. Since these molecules induce apoptosis independently of RAR transactivation, we set up to synthesize novel analogs with impaired RAR binding. Here we describe adamantyl arotinoids with 2,2'-disubstituted biaryl rings prepared using the Suzuki coupling of the corresponding fragments. Those with cinnamic and naphthoic acid end groups showed significant antiproliferative activity in several cancer cell lines, and this effect correlated with the induction of apoptosis as measured by caspase activity. Strikingly, some of these compounds, whereas devoid of RAR binding capacity, were able to activate RXR.

The synthesis of dioxepine bastadin 3, a tyrosine-tyramine derivative with a dibenzo-1,3-dioxepine scaffold that is rarely present among natural products, is described. The dibenzo-1,3-dioxepine ring was formed early in the sequence and the (E)-2-(hydroxyimino)-N-alkylamide was generated in the last step by oxidation of the 2-amino-N-alkylamide precursor. The presumably biogenetic late-stage ring formation starting from congener bastadin 3 failed. A new synthesis of this alkaloid was also developed. This new route requires a minimal use of protecting groups and the order of the two key steps was reversed relative to the route to dioxepine bastadin 3.

We recently reported on a series of retinoid-related molecules containing an adamantyl group, a.k.a. adamantyl arotinoids (AdArs), that showed significant cancer cell growth inhibitory activity and activated RXRα (NR2B1) in transient transfection assays while devoid of RAR transactivation capacity. We have now explored whether these AdArs could also bind and inhibit IKKβ, a known target that mediates the induction of apoptosis and cancer cell growth inhibition by related AdArs containing a chalcone functional group. In addition, we have prepared and evaluated novel AdArs that incorporate a central heterocyclic ring connecting the adamantyl-phenol and the carboxylic acid at the polar termini. Our results indicate that the majority of the RXRα activating compounds lacked IKKβ inhibitory activity. In contrast, the novel heterocyclic AdArs containing a thiazole or pyrazine ring linked to a benzoic acid motif were potent inhibitors of both IKKα and IKKβ, which in most cases paralleled significant growth inhibitory and apoptosis inducing activities.