Gene profiling and the cyclin-dependent kinase inhibitor flavopiridol: what's in a name?

Flavopiridol represents one in an ever-expanding group of compounds currently called molecularly targeted agents. Such compounds, in contrast to conventional cytotoxic drugs, which exert a more general disruptive effect on key cellular structures or processes (e.g., DNA replication, formation of the mitotic spindle, etc.), target pathways directly implicated in neoplastic transformation. The prototype of this type of compound is imatinib mesylate, or Gleevec, an inhibitor of the Bcr/Abl kinase that has enjoyed considerable success in the treatment of chronic myelogenous leukemia and related disorders (1). Flavopiridol is a semi-synthetic rohitukine alkaloid that functions as a broad inhibitor of cyclin-dependent kinases (CDKs), enzymes critically involved in cell cycle progression and the disordered proliferation of tumor cells (2). In addition to its cell cycle inhibitory effects, preclinical studies have shown that flavopiridol is a highly potent inducer of apoptosis in malignant cells, particularly those of hematopoietic origin (3). Flavopiridol has also been shown to potentiate the lethal effects of both conventional cytotoxic as well as more novel agents in epithelial and leukemic cells (4, 5). On the basis of its very promising activity in preclinical studies, flavopiridol was the first CDK inhibitor to enter clinical trials in humans (6). Currently, flavopiridol continues to undergo extensive clinical evaluation, either as a single agent or in combination with other cytotoxic drugs. Although flavopiridol has demonstrated some, albeit limited, activity in early trials (7), to date it has not fulfilled the promise predicted for it based on encouraging preclinical results. However, evidence is now emerging that the activity of flavopiridol in humans may have been limited by pharmacokinetic factors, including suboptimal administration schedules as well as binding of the drug to plasma proteins. Consequently, attention is now focusing on newer hybrid schedules (e.g., bolus in conjunction with a prolonged infusion) to circumvent these problems. In the intervening period, interest in CDK inhibitors has persisted, and several other agents of this class, including UCN-01 (8), CYC202 (9), and BMS 387032 (10) have now entered clinical trials. Efforts to elucidate the mechanism of action of CDK inhibitors have provided the following insights: such agents may have multiple targets in addition to CDKs, and it is unclear whether CDK inhibition is primarily responsible for lethality. For example, UCN-01 was originally developed as a selective PKC inhibitor (11) but was subsequently found to act as a CDK inhibitor (12). Since then, UCN-01 has been shown to exert multiple other actions that may be as or possibly more directly related to lethality, including inhibition of Chk1 (13) and Akt/PDK1 (14). Flavopiridol, in particular, has been found to exert an exceptionally broad spectrum of actions in addition to CDK inhibition, all of which might plausibly contribute to cytotoxicity. These include anti-angiogenic activity (15), transcriptional repression via inhibition of the positive transcription elongation factor-B (PTEF-b), CDK9/cyclin T complex (16), downregulation of anti-apoptotic proteins (e.g., XIAP, Mcl-1, BclxL; refs. 17, 18), repression of p21 (19), DNA duplex formation (20), inhibition of survivin phosphorylation (21), down-regulation of cyclin D1 (22), and most recently, inhibition of IKK and NF-nB (23). It is important to note that the pleiotropic mode of actions of such agents is not restricted to CDK inhibitors; in fact, the putatively specific Bcr/Abl kinase inhibitor imatinib mesylate is known to inhibit other tyrosine kinases, including c-Kit, a finding that has significant clinical implications, for example, in gastrointestinal stromal tumors (24). It is clear that in the case of novel agents, such as flavopiridol, a better understanding of which pathways are being targeted will be critical to the development of optimal therapeutic strategies. However, until recently, identification of such targets has been a largely empirical endeavor. With the development of gene profiling strategies, this situation seems likely to change. For example, in a study reported in this issue by Lü et al. (25), a DNA array approach was employed to characterize changes in gene expression in four epithelial tumor types (prostate and gliomas) exposed to cytotoxic concentrations of flavopiridol. A number of findings, including some that were quite unexpected, emerged from this study. Lü et al. identified a diverse group of 220 genes that were either significantly upor down-regulated following exposure of Mol Cancer Ther 2004;3(7):873–5