Resistance to kinase inhibitors may also be effected by MLN4924 aberrant activation of redundant signalling pathways to that of the target, an example being MET amplification inresistancetoEGFR kinase inhibitors. As CEM/ AKB16 cells were highly resistant to Aurora B inhibition it appears that sustained Aurora B activity in the presence of ZM447439 may still be driving resistance in these cells rather than activation 
of an alternative pathway. Previous work from our laboratory on drug resistance mediated by tubulin mutations showed that CEM cells acquire additional point mutations in tubulin at higher levels of resistance. Both CEM/AKB8 and CEM/AKB16 cells expressed the Aurora B G160E mutation described for CEM/ AKB4 cells, however no additional mutations in Aurora B were observed, further demonstrating theimportance of the160 residue in drug binding and high-level resistance. Our study of phosphorylated Histone H3 levels showed that CEM/AKB4 cells maintain resistance to Aurora B inhibition at 16 mM ZM, despite this drug concentration being sufficient to induce apoptosis and cell death. This is consistent with off-target kinase inhibition of ZM447439, where at high drug concentrations the contribution of targeting additional cytotoxic pathways to Aurora B inhibition becomes significant. Therefore the resistant phenotype in CEM/AKB16 cells may potentially be mediated LEE011 through alterations in these other targets of ZM447439. ZM447439 has been shown to potently inhibit Aurora A as well as Aurora B in biochemical assaysand we analysed CEM/AKB16 cells for alterations in Aurora A. We found no changes in gene or protein expression of Aurora A in CEM/AKB16 cells and no mutations in the Aurora A gene. Additionally, CEM/AKB16 cells were as equally sensitive as CEM cells to a selective Aurora A inhibitor MLN8237, suggesting that ZM447439 resistance in these cells is not mediated through an Aurora A dependent pathway. It is possible that alterations in other unknown targets of ZM447439 may be responsible, and ultimately, an understanding of the precise mechanisms underpinning resistance in the more highly resistant CEM/AKB8 and CEM/AKB16 cells will shed further light on the mode of action of this drug. Aurora B inhibitors remain a promising area for targeted anticancer therapy, yet a fuller understanding of drug response and resistance mechanisms will aid their clinical implementation. Our findings have confirmed that resistance to these agents is likely across a variety of malignancies and that point mutations in Aurora B, particularly of the 160 residue, may be highly significant markers of treatment outcome. Moreover, our analysis of highly resistant cells suggests that sustained or high-level drug treatment may give rise to an evolution of multiple mechanisms of resistance in patients. Accordingly, our models provide a basis for designing and testing alternative Aurora B inhibitors, and for screening agents that may be employed in combination therapeutic approaches. The two highly homologous human tankyrase isoforms, TNKS1 and TNKS2, are members of the poly ADP-ribose polymerasefamily of 17 proteins that share a catalytic PARP domain. These PARP proteins cleave NAD+into ADP-ribose and nicotinamide and transfer the ADP-ribose units onto their substrates, resulting in a post-translational modification referred to as PARsylation. Cellular functions of many PARP proteins remain unknown. PARP1 and PARP2, the two best characterized family members, are key players in homologous recombination DNA damage response and have been pursued as cancer targets for over a decade. Structural studies of PARP inhibitor complexes reveal that these compounds are anchored in the nicotinamide pocket in a very similar manner.