Abstract. Corruption of the Rb and p53 pathways occurs in virtually all human cancers. This could be because it lends oncogene-bearing cells a surfeit of Cdk activity and growth, enabling them to elaborate strategies to evade tumorsuppressive mechanisms and divide inappropriately. Targeting both Cdk activities and the PI3K pathway might be therefore a potentially universal means to palliate their deficiency in cancer cells. We showed that the killing efficacy of roscovitine and 16 other purines and potentiation of roscovitine-induced apoptosis by the PI3K inhibitor, LY294002, decreased with increasing corruption of the Rb and p53 pathways. Further, we showed that purines differing by a single substitution, which exerted little lethal effect on distant cell types in rich medium, could display widelydiffering cytotoxicity profiles toward the same cell types in poor medium. Thus, closely-related compounds targeting similar Cdks may interact with different targets that could compete for their interaction with therapeutically-relevant Cdk targets. In the perspective of clinical development in association with the PI3K pathway inhibitors, it might thus be advisable to select tumor cell type-specific Cdk inhibitors on the basis of their toxicity in cell-culture-based assays performed at a limiting serum concentration sufficient to suppress their interaction with undesirable crossreacting targets whose range and concentration would depend on the cell genotype.
IntroductionCancers arise from mutations that enable somatic cells to break tissue homeostasis and proliferate in settings where they would normally be sentenced to either exit the cell cycle or die (1). The retinoblastoma (Rb) and p53 proteins appear as key components in homeostatic pathways (2). Rb is a powerful repressor of cell cycle-regulating gene transcription, whose sequential phosphorylation and inactivation by cyclin D-and cyclin E-Cdks is required for cells to depart from quiescence (3). p53 is a transcription factor that induces cell cycle exit or apoptosis upon activation by many stresses, including oncogenic insults (2,4). Human cancers invariably harbor a functional alteration of one, or more, component(s) of the Rb and p53 pathways and typically overexpress cyclins D or E and/or certain Cdks (2).Targeting the cell cycle has thus emerged as a promising avenue for anticancer therapy, prompting the search for potent and selective Cdk inhibitors intended to stop tumor growth (5-7). Several ATP competitive Cdk inhibitors have now reached the stage of clinical trials (flavopiridol, UCN-01, CYC-202 and BMS-387032). Intriguingly, these compounds not only arrest cells in G1 and G2 but also trigger apoptosis in certain contexts. Such an unforeseen lethal effect could result from the fortuitous interaction of the drugs with a Cdkunrelated target. A more tantalizing hypothesis, however, is that Cdk inhibition is able to reactivate proapoptotic pathways silenced, and/or extinguish prosurvival pathways triggered by hyperactivated Cdks. Indeed, inhibition of Cdk7...