Abstract IA24: Targeting Chk1

Abstract

Abstract The majority of traditional anticancer drugs inhibit DNA synthesis either by directly damaging DNA or by inhibiting synthesis of deoxyribonucleotide precursors. DNA damage induces cell cycle arrest through activation of cell cycle checkpoints whose goal is to prevent further DNA synthesis or mitosis until the damage is repaired. These checkpoints have undergone intense investigation as potential therapeutic targets, and Chk1 inhibitors (Chk1i) have emerged as promising novel therapeutic agents (1). Chk1 was initially recognized as a regulator of the DNA damage-induced S and G2 checkpoints, and its inhibition forced S phase progression followed by mitotic catastrophe of arrested cells. Subsequently, it was found that the combination of Chk1i with antimetabolites (particularly gemcitabine and cytarabine) more effectively enhanced cell killing (2, 3), and these combinations entered clinical trials. Little clinical benefit has been reported to date, and development of several Chk1i has been terminated for toxicity; whether the toxicity was due to on-target effects remains to be established. It is worth noting one outlier response in which a patient receiving irinotecan and the Chk1i AZD7762 attained a durable response; the sensitivity was tracked to a mutation in RAD50. This suggests that subsets of patients may have therapeutic benefit from this combination (4). Intriguingly, a few cell lines are also highly sensitive to Chk1i as a single agent suggesting subsets of tumors may exist that are uniquely sensitive to these drugs (5). Development of AZD7762 and another Chk1i, MK-8776, were terminated, but clinical trials are continuing with two other Chk1i: LY2606368 and GDC-0425; and a Phase I trial with V158411 is planned. The sensitivity to Chk1i as a single agent results from inappropriate activation of CDK2 in S phase, which in turn depends on the differential activation of the upstream phosphatase CDC25A (5, 6). This results in rapid appearance of DNA double-strand breaks that are dependent on Mre11 and Mus81 nucleases (7). The ability of Chk1i to abrogate DNA damage-induced S phase arrest is commonly attributed to activation of CDC25A/CDK2, yet we find that abrogation of arrest occurs even when CDK2 is inhibited. In contrast, inhibition of CDC7 prevents S phase progression. Similarly, the sensitivity to ribonucleotide reductase inhibitors such as gemcitabine is independent of CDK2 but dependent on CDC7. In this case, the continued absence of dNTPs prevents S phase progression, but DNA helicases are activated by Chk1i creating excess single-strand DNA that exhausts the protective single-strand binding protein RPA. The unprotected DNA is then susceptible to nuclease activity. The mechanistic link between Chk1 and CDC7 remains to be established. Administering a combination of two drugs to a patient, for example gemcitabine and a Chk1i, requires knowledge of how both drugs work in a human. One critical problem with preclinical experiments is that they are too often performed with continuous incubation over many days, yet in a patient, the drugs are usually administered over a short period of time. In the case of gemcitabine, a brief incubation in cell culture, or a bolus administration to a mouse or patient, leads to rapid and irreversible inhibition of ribonucleotide reductase and persistent S phase arrest. To assess the optimum time of addition of Chk1i, we added MK-8776 for 6 h periods following a 6 h treatment with gemcitabine. The greatest growth inhibition in vitro occurred when the addition of MK-8776 was delayed to 18 h, and this correlated with the maximal accumulation of cells in S phase. This delayed schedule was also shown to be more effective than concurrent administration in a xenograft model (3). Despite the apparent efficacy of MK-8776 in combination with gemcitabine, experiments to assess the single agent activity of MK-8776 have not detected tumor growth inhibition. We propose several explanations: when used as a single agent, a higher concentration of MK-8776 is required to inhibit cell growth than when combined with gemcitabine; cells also appear to recover better following a 6-h incubation with MK-8776 as a single agent than a 6-h treatment in combination suggesting that longer target inhibition is required for single agent activity. These 6-h treatment periods were selected because they reflect the in vivo concentration and half-life of MK-8776 in patients. These disappointing effects may be circumvented with alternate Chk1i that achieve greater bioavailability for a longer period of time. In this regard, both GDC-0425 and LY2606368 exhibit longer plasma half-lives, and Vernalis has a lead compound that has exhibited single agent activity in a xenograft model (Andrew Massey, personal communication). In summary, while initial clinical trials have generally resulted in marginal therapeutic activity, there may be subsets of patients who may respond well, and Chk1i with prolonged bioavailability may have enhanced in vivo efficacy either alone or in combination.