In Rainey et al published the results of a targeted

In 2008, Rainey et al. published the results of a targeted compound library screen for potential inhibitors of the ATM kinase. In this study, the compound CP466722 was identified as a highly selective and rapidly reversible ATM inhibitor, which did not inhibit PI3K or related PIKK family members. Most importantly, the authors were able to not only validate the radiosensitising effect mediated by ATM inhibition, but also showed that transient inhibition of ATM is sufficient to achieve a significant increase in radiation-induced cytotoxicity (Rainey et al., 2008). This raises the possibility that clinically relevant radiosensitisation may not require prolonged ATM inhibition which could help to reduce side effects in a clinical setting. Starting out from the chemical structure of the PI3K inhibitor LY294002, KuDOS Pharmaceuticals (in 2005 acquired by AstraZeneca) developed the first potent and selective ATM inhibitor: KU-55933. In a study published in 2004, Hickson et al. demonstrated, that KU-55933 confers marked sensitisation to ionising radiation and DNA DSB-inducing chemotherapeutics, such as the topoisomerase II inhibitors etoposide and doxorubicin, in cancer cells. Importantly, in cells derived from A–T patients, which express no functional ATM, no radiosensitisation was observed, further confirming the selectivity of the compound for ATM (Hickson et al., 2004). To improve the pharmacokinetics and bio-availability, the core structure of KU-55933 was further optimised which led to the development of KU-60019, which, like KU-55933, is an ATP-competitive ATM inhibitor. It was shown to be more effective at blocking radiation-induced phosphorylation of ATM downstream targets than KU-55933 and to possess greater potency as radiosensitiser (Golding et al., 2009). Again, A–T fibroblasts were not radiosensitised by the compound, arguing for ATM being specifically targeted. This and further in vitro studies demonstrated potent radiosensitisation of several glioblastoma cell lines by KU-60019 – a promising observation given that glioblastoma cells are usually very resistant to radiotherapy. Notably, viability of cultured human astrocytes, which are terminally differentiated and not actively dividing, was not affected by short-term exposure to KU-60019. This observation suggests that ATM inhibition alone is not toxic for normal Nimodipine tissues outside the radiation field (Golding et al., 2012). Even though KU-60019 showed better solubility in aqueous solutions than KU-55933, bio-availability was still poor. When administered intraperitoneally or orally, systemic plasma levels only reach low micromolar concentrations (Biddlestone-Thorpe et al., 2013), limiting utility for in vivo studies. Biddlestone-Thorpe et al. bypassed this limitation by directly injecting KU-60019 into orthotopic gliomas grown in mice. They showed that ATM inhibition by KU-60019 markedly radiosensitised the glioma xenografts in vivo, leading to a significant increase in survival time of the mice. Importantly, KU-60019 showed an even greater radiosensitising effect in p53-mutant glioma, resulting in extended survival and, in some cases, apparent cure of the treated mice (Biddlestone-Thorpe et al., 2013). Though KU-60019 may not be of clinical utility, this proof-of-principle study provided evidence that pharmacological ATM inhibition can induce a potent radiosensitisation of cancer cells in vivo. Recently, KU-59403, another ATM inhibitor from this class of compounds, has been described. This compound not only possesses improved potency over KU-55933, it also exhibits improved solubility and bio-availability, allowing for the study of effects of pharmacological ATM inhibition in animal models of human cancer. Batey et al. demonstrated that following administration to mice, KU-59403 showed good tissue distribution. In subcutaneous tumour xenografts, concentrations above those required for in vitro activity were reached and maintained for at least 4h (Batey et al., 2013), making this the first reported in vivo active ATM inhibitor. In this study, it was shown that while KU-59403 alone had no impact on tumour growth rate, it greatly enhanced the anti-tumour activity of the topoisomerase inhibitors etoposide and irinotecan. Notably, the authors found the presence of KU-59403 at the time of etoposide dosing necessary to observe the chemosensitising effect. In contrast, delaying administration of KU-59403 by only 4h completely abolished chemosensitisation. Contrary to Biddlestone-Thorpe et al., Batey and colleagues found that chemo- and radiosensitisation by ATM inhibition was not p53-dependent. Possible causes for this discrepancy include the use of different inhibitors and administration routes or the use of different tumour models (orthotopic glioma vs. subcutaneous colon cancer models). A contributing factor could also be the nature of the p53 mutations that the cancer cells carry as loss of function, dominant negative or gain of function mutations in p53 may affect therapeutic responses of cancer cells in different ways. Also, the increased sensitivity of p53-deficient glioma cells to ATM inhibition in combination with ionising radiation observed by Biddlestone-Thorpe was only seen in vivo and not in vitro. The radiosensitising effect of KU-59403, however, was only studied in vitro. Further studies will be required to address the question if the functional status of p53 plays a role in the radiosensitising potency of ATM inhibitors.