br Synthetic lethal approaches In addition to the potential
Synthetic lethal approaches In addition to the potential utility of ATM and ATR as chemo- or radiosensitisers, recent studies suggest that such compounds may have single agent activity in certain subsets of patients through induction of synthetic lethality. Two genes are considered synthetically lethal, if mutation or inactivation of either gene or gene product alone has no effect on cellular viability, whereas simultaneous defects in both genes/gene products lead to cell death (Kaelin, 2005). The observed single-agent anti-tumour activity of AZD6738 in ATM-deficient but not ATM-proficient xenograft models suggests that such a synthetic lethal interaction could exist between the ATM and ATR signalling pathways (Guichard et al., 2013, Jones et al., 2013), potentially due to the overlapping and cooperating roles of these pathways in coordinating cancer progression with DNA repair. In our own studies we were able to show that non-small cell lung cancer cells deficient in both ATM and p53 are particularly sensitive to ATR inhibition in vitro, suggesting that the functional status of p53 may be important in this setting as well (Weber et al., 2013). Similarly, PARP inhibitors possess single agent activity in ATM-deficient tumour cells in vitro and in vivo (Aguilar-Quesada et al., 2007, Weston et al., 2010, Williamson et al., 2010). However, the cytotoxicity of PARP inhibitors was shown to be enhanced when the function of both ATM and p53 was lost (Williamson et al., 2012, Kubota et al., 2014). Several putative synthetic lethal interactions between the ATM or ATR signalling pathways and other DDR components have been reported. Inhibition or knockdown of ATM has been shown to be synthetically lethal in cells with defects in the Fanconi anaemia pathway, components of which are commonly mutated or lost in cancer (Kennedy and D'Andrea, 2006, Kennedy et al., 2007). ATM inhibition or -deficiency has also been shown to be synthetically lethal in combination with APE1 inhibitors (Sultana et al., 2012) or functional loss of XRCC1 (Sultana et al., 2013a), which is frequently deregulated in breast and ovarian cancers (Abdel-Fatah et al., 2013, Sultana, Abdel-Fatah, Abbotts, et al., 2013). ATR inhibition has been shown to be synthetically lethal with XRCC1 or ERCC1 loss (Sultana, Abdel-Fatah, Abbotts, et al., 2013, Mohni et al., 2014). Furthermore, the observed increased cytotoxicity of the ATR inhibitors AZ20 and AZD6738 in MRE11A mutated LoVo cells in vitro and in vivo (Jacq et al., 2012, Foote et al., 2013, Guichard et al., 2013) points towards a potential synthetic lethal interaction between ATR inhibition and functional loss of MRE11. This synthetic lethality may include other components of the MRN complex as well (Al-Ahmadie et al., 2014). Hence, pharmaceutical inhibition of ATM and/or ATR may provide the basis for the selective treatment of DDR pathway-deficient cancers.
Biomarkers and patient selection To increase the success rate for drug development it has been proposed that biomarkers to identify patient subgroups likely to get the greatest benefit from new drugs should be identified early, alongside biomarkers of target inhibition (pharmacodynamic; PD), pathway modulation, and anticipated biological effect(s) (Yap et al., 2010). Although patient selection biomarkers for ATM and ATR inhibitors have not yet been established, published in vitro and in vivo data suggest a number of potential strategies (Table 3). Increased tumour DDR capacity has been associated with resistance to radio- and chemotherapies, for example by upregulating DNA repair capacity in response to platinum-based therapy (Martin et al., 2008, Oliver et al., 2010). Although resistance to therapy is likely to be multifactorial and highly dependent on the particular therapy under consideration (e.g. cisplatin, taxanes, radiation), conceptually, a test for increased tumour DDR capacity might identify a group of patients who would gain particular benefit from combination therapy with either an ATM or ATR inhibitor. Interestingly, elevated levels of phospho-ATM (Ser1981) prior to radiotherapy has been associated with radioresistance and poor prognosis in cervical cancer (Roossink et al., 2012). This suggests that levels of ATM autophosphorylation in tumour tissues could serve as a potential biomarker to identify patients that might get the greatest benefit from a combination of ATM inhibitor treatment and radiotherapy. As outlined above, many human tumours acquire defects in the DDR in order to tolerate DNA replication stress and genomic instability that are characteristic of oncogene activation during cancer development. These DDR defects offer the potential to use ATM or ATR inhibition as a synthetic lethal approach, in an analogous manner to the use of PARP inhibitors in BRCA1/2 defective tumours. Functional defects in several components of the DDR pathway have been reported to confer sensitivity to ATM and/or ATR inhibition, including p53 (Reaper et al., 2011), XRCC1 (Sultana, Abdel-Fatah, Abbotts, et al., 2013, Sultana, Abdel-Fatah, Perry, et al., 2013), FANCD2 (Kennedy et al., 2007), MRE11A (Jacq et al., 2012, Foote et al., 2013, Guichard et al., 2013), RAD50 (Al-Ahmadie et al., 2014), BRCA1 (Albarakati et al., 2014) and ATM (Reaper et al., 2011). Whole exome/genome sequencing offers a potential method to identify patient subgroups or particular cancer types with DDR defects (Kandoth et al., 2013), although the relationships between reported mutations and loss of protein function for many of the DDR proteins have yet to be determined. Of the DDR genes, p53 is the most widely studied with an extensive literature on the functional consequences of a wide range of mutations (Petitjean et al., 2007). Therefore, p53 mutation status may provide an enrichment strategy for selecting patients for treatment with ATM or ATR inhibitors, but p53 status remains a complex biomarker to interpret (Olivier et al., 2010).