Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Mechanism of action and therapeutic efficacy of Aurora kinase B inhibition in MYC overexpressing medulloblastoma

Mechanism of action and therapeutic efficacy of Aurora kinase B inhibition in MYC overexpressing... www.impactjournals.com/oncotarget/ Oncotarget, Vol. 6, No.5 Mechanism of action and therapeutic efficacy of Aurora kinase B inhibition in MYC overexpressing medulloblastoma 1,2,3,* 1,* 1 1 Roberto Jose Diaz , Brian Golbourn , Claudia Faria , Daniel Picard , David 1 4 4 1 1 Shih , Denis Raynaud , Michael Leadly , Danielle MacKenzie , Melissa Bryant , 1 1 1,2,3 1 Matthew Bebenek , Christian A. Smith , Michael D. Taylor , Annie Huang and 1,2,3 James T. Rutka The Hospital for Sick Children. Arthur and Sonia Labatt Brain Tumour Research Centre, Toronto, Ontario, Canada Department of Laboratory Medicine & Pathobiology, University of Toronto, Toronto, Ontario, Canada Division of Neurosurgery, Department of Surgery, University of Toronto, Toronto, Ontario, Canada Analytical Facility for Bioactive Molecules, The Hospital for Sick Children, Toronto, Ontario, Canada These authors contributed equally to this work Correspondence to: James T. Rutka, email: james.rutka@sickkids.ca Keywords: Aurora kinase, medulloblastoma, tumor biology, molecular therapy,cell-cycle Received: June 29, 2014 Accepted: December 24, 2014 Published: December 31, 2014 This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT Medulloblastoma comprises four molecular subgroups of which Group 3 medulloblastoma is characterized by MYC amplification and MYC overexpression. Lymphoma cells expressing high levels of MYC are susceptible to apoptosis following treatment with inhibitors of mitosis. One of the key regulatory kinases involved in multiple stages of mitosis is Aurora kinase B. We hypothesized that medulloblastoma cells that overexpress MYC would be uniquely sensitized to the apoptotic effects of Aurora B inhibition. The specific inhibition of Aurora kinase B was achieved in MYC- overexpressing medulloblastoma cells with AZD1152-HQPA. MYC overexpression sensitized medulloblastoma cells to cell death upon Aurora B inhibition. This process was found to be independent of endoreplication. Using both flank and intracranial cerebellar xenografts we demonstrate that tumors formed from MYC-overexpressing medulloblastoma cells show a response to Aurora B inhibition including growth impairment and apoptosis induction. Lastly, we show the distribution of AZD1152- HQPA within the mouse brain and the ability to inhibit intracranial tumor growth and prolong survival in mice bearing tumors formed from MYC-overexpressing medulloblastoma cells. Our results suggest the potential for therapeutic application of Aurora kinase B inhibitors in the treatment of Group 3 medulloblastoma. options. We have recently reported on the in-vitro pro- INTRODUCTION apoptotic and in-vivo cytostatic effects of Aurora B inhibition in glioblastoma using the inhibitor AZD1152- Recent genetic analysis of medulloblastoma (MB) HQPA.[4] Aurora kinase B (Aurora B) is an essential has revealed four genetically and epidemiologically serine/threonine kinase responsible for regulation distinct subgroups.[1, 2] One of these subgroups of multiple events in mitosis including chromosome designated as Group 3 medulloblastoma (G3MB) affects condensation, centromere-kinetochore dynamics, primarily infants and children and carries an overall chromatid segregation, and cytokinesis.[5] One of the survival of 45% for infants and 58% for children at five hallmarks of Aurora kinase B inhibition is G2/M arrest years as reported in a retrospective international meta- and subsequent escape into endoreplication cycles.[6] analysis.[3] Given the poor clinical outcomes observed Multiple regulatory or structural components essential to in patients with G3MB tumors with current treatment mitotic progression have been previously targeted to block regimens, there is a need to explore novel therapeutic www.impactjournals.com/oncotarget 3359 Oncotarget tumor cell proliferation including microtubules,[7] cyclin expression and Aurora B expression in G3MB (R=0.57, dependent kinase 1,[8] Aurora kinases,[9] and Polo-like P=0.002, N=27, Fig. 1C). Although WNT tumors express kinase 1.[10-12] When tumor cells overexpress c-Myc high levels of AURKB mRNA we did not observe a (MYC) either endogenously or by genetic modification, correlation to MYC mRNA expression in this small subset the induction of a G2/M arrest has been shown to trigger of tumor samples (R=0.42, P=0.3, N=8). Aurora kinase tumor cell death in-vitro and to block tumor growth in- gene expression is increased in fetal cerebellum and in all vivo.[8, 13, 14] This observation was also made recently subgroups of MB compared to adult cerebellum, reflecting in B-cell and T-cell lymphoma cells overexpressing the proliferative capacity of fetal and tumor tissue. MYC and treated with the Aurora B inhibitor AZD1152- To further evaluate the expression of Aurora kinase HQPA.[15, 16] MYC expression has also been reported to A and B in relation to MYC, protein expression in a confer sensitivity to radiation and DNA damaging agents number of unsynchronized MB cell lines was evaluated (etoposide, cisplatin) in medulloblastoma cells.[17] (Fig. 1D). The D425, D458 and MED8A cells, all of Amplification of MYC, as well as MYC which have known amplification of MYC,[18, 21, 22] overexpression, is a negative prognostic factor for overall showed concurrent marked expression of both Aurora B survival in MB.[18, 19] Approximately 11% of G3MB and MYC protein. Using a NanoString assay with a probe tumors demonstrate MYC amplification.[20] Furthermore, set for MB subgroup signature genes,[23] the D425 cell all G3MB tumors express MYC at high levels and line was demonstrated to share a similar gene expression express genes associated with elevated MYC levels.[20] signature as G3MB tumors (Supplementary Fig. S1). The We hypothesized that MB cells overexpressing MYC D425 cell line was derived from primary tumor tissue,[21] would be uniquely sensitized to the effects of Aurora B whereas the D458 line was derived from cerebrospinal inhibition and that this property could be harnessed for fluid within the same patient after failure of radiotherapy the in-vivo treatment of MYC-overexpressing MB tumors. and chemotherapy with cyclophosphamide, cisplatin, and The goal of our study was not only to determine if MYC vincristine.[22, 24] As such, our results suggest that the overexpression in human MB cells sensitized the cells to concordance between Aurora B and MYC expression is the apoptotic effects of Aurora B inhibition, but also to maintained in the setting of recurrent G3MB. further define the mechanism triggering this response. We demonstrate that Aurora B inhibition triggers cell death Myc overexpression sensitizes medulloblastoma independent of DNA replication and that transient Aurora cells to cell death induced by Aurora B inhibition B inhibition results in a unique impaired growth response in MYC-overexpressing cells. Having defined the response The UW228 and UW426 MB cell lines show time-course we proceeded to optimize in-vivo therapy with low expression of Aurora B and MYC (Supplementary AZD-1152 HQPA, achieving a prolongation in survival Fig. S2A). However, the cells can be modified to stably of mice bearing cerebellar xenografts of MB cells having overexpress MYC by retroviral transduction[25, 26] – MYC amplification and endogenously overexpressing the stable cell lines are identified here as UW228-Myc MYC. and UW426-Myc. Comparison of protein expression in MYC-overexpressing cells compared to isogenic RESULTS controls showed that increased expression of MYC was associated with an increase in Aurora A and B protein levels in unsynchronized cell culture (Supplementary Fig. Co-expression of Aurora B and MYC in Group 3 S2A). The cell-cycle distribution of UW228 cells did not medulloblastoma change with MYC overexpression while a reduction in G1/G0 cells with associated increase in polyploid cells was observed in UW426-Myc cells compared to control MYC has been shown to directly regulate the (Supplementary Fig. S2B). expression of Aurora A and indirectly the expression of The co-expression of Aurora B and MYC in MB Aurora B in B-cell lymphoma.[15] Therefore, we sought cells suggested that Aurora B activity could be important to determine if Aurora kinase gene expression correlates for cell survival in the presence of excess MYC, as is the with MYC expression in human MB. AURKA and AURKB case for B-cell lymphoma cells.[15] Therefore, we tested mRNA expression showed a positive correlation with MYC the ability of Aurora B inhibition to elicit cell death in mRNA expression (AURKA vs MYC: R=0.32, P=0.001, MYC overexpressing MB cells versus isogenic controls. N=103; AURKB vs MYC: R=0.37, P = 0.0001, N=103) We selected the Aurora B inhibitor AZD1152-HQPA while no correlation exists between AURKC and MYC for our studies as we had previously demonstrated that expression (Fig. 1A). The highest MYC expression was it had activity in brain and flank xenografts of human observed in WNT and G3MB relative to other subgroups, glioblastoma cells.[4] A series of drug concentrations normal fetal cerebellum, and adult cerebellum (Fig. 1B). was used to assess the lowest concentration of drug that Furthermore, there was a modest correlation between MYC www.impactjournals.com/oncotarget 3360 Oncotarget could inhibit MB cell proliferation in UW228 and UW426 sensitized cells to induction of cell death upon Aurora cells (Fig. 2A). Cell proliferation was inhibited in a B inhibition (Fig. 2A). Inhibition of proliferation was concentration-dependent manner both in isogenic controls achieved with 25 nM AZD1152-HQPA in UW426 and and MYC overexpressing cells (Fig. 2A). UW426 cells UW426-Myc as well as UW228 and UW228-Myc showed almost complete inhibition of proliferation when cells indicating a potent block of mitosis by Aurora exposed to 100 nM AZD1152-HQPA for 72 hrs, while B inhibition. The 72-hour time point was selected for UW426-Myc cells had cell counts that were almost half of determination of the concentration effects of AZD1152- those at baseline, indicating that cell loss had occurred (P< HQPA based on previous observations that induction of 0.001, N=3). A similar effect is observed for UW228 and cell death in response to Aurora B inhibition is a time- UW228-Myc cells, indicating that MYC overexpression dependent phenomenon.[4] Figure 1: Aurora kinase mRNA and protein expression in relation to Myc expression in medulloblastoma. A) mRNA expression of AURKA, AURKB, and AURKC in relation to MYC mRNA level in 103 medulloblastoma tumor samples. B) MYC mRNA expression in fetal cerebellum (fCb), adult cerebellum (aCb), and medulloblastoma tumors subgrouped according to RNA expression profile, ANOVA P<0.0001. C) Correlation between AURKB mRNA expression and MYC mRNA expression in medulloblastoma tumors subgrouped as Group 3. D) Western blot showing protein expression of Aurora A, Aurora B, and MYC in multiple medulloblastoma cell lines. Cell lines harboring MYC amplification are indicated by a star. The loading control was β-Actin. Total protein loaded was 30 μg. www.impactjournals.com/oncotarget 3361 Oncotarget In order to demonstrate the specificity of inhibition B on T232 is a key requirement for kinase activity.[30, of MB cells at 100 nM AZD1152-HQPA, cell lysates 31] Inhibition of Aurora B autophosphorylation without were probed for reduction in phosphorylation of Histone a change in total Aurora B levels was achieved upon H3 Serine 10, a specific substrate for Aurora B.[27, exposure of control and MYC-overexpressing cells to 100 28] Histone H3 Serine 10 phosphorylation is reduced nM AZD1152-HQPA for 24 hr (Fig. 2B). We note a weak with exposure to 100 nM AZD1152-HQPA over 24 phosphorylation signal of Aurora C (35 kDa) in UW425 hours without a change in total Histone H3 or β-actin and UW228 cells, which is abolished with AZD1152- protein levels (Fig. 2B). To further assess if AZD1152- HQPA (Fig 2B). The autophosphorylation of Aurora A HQPA could be interacting with Aurora A, a previously was observed to increase slightly in both wild-type and reported therapeutic target for MB,[29] we examined MYC-overexpressing cells with no change in total Aurora the inhibition of Aurora A and B autophosphorylation. A levels (Fig. 2B). Autophosphorylation of Aurora A on T288 and Aurora The onset of cell loss in MYC-overexpressing Figure 2: Cell viability effects and specificity of Aurora B inhibition with AZD1152-HQPA in medulloblastoma. A) Cell counts obtained by Vi-Cell XR counter (Beckman-Coulter, Mississauga, Ontario) in wild-type versus MYC overexpressing medulloblastoma cells exposed for 72 hours to varying concentrations of AZD1152-HQPA. *P <0.05 for comparison between wild-type and MYC overexpressing cells. Data represent mean for three independent experiments. Error bars are standard error of the mean. A total of 1 x 10 cells were cultured for 24 hours prior to exposure to AZD1152-HQPA. B) Western blots showing the inhibition of Histone H3 (Ser 10) phosphorylation and Aurora B autophosphorylation by AZD1152-HQPA at 100 nM in MYC overexpressing or wild-type cells when the drug is solubilized in water or DMSO. Labels as follows: phosphohistone H3 Serine 10 (pH3 Ser 10), histone H3 (H3), aurora kinase A (AurA), aurora kinase B (AurB), Aurora A phosphothreonine 288 (pAurA), Aurora B phosphothreonine 232 (pAurB), Aurora C phosphothreonine 198 (pAurC), β-Actin (Actin). Whole cell lysate from untreated U251 GBM cells was used as a reference antibody positive control (+). Total protein loaded was 30 μg. www.impactjournals.com/oncotarget 3362 Oncotarget MB cells compared to isogenic controls took place after Downregulation of the serine/threonine kinase 48 hours of continuous exposure to 100 nM AZD1152- Ark5/Nuak1 has been shown to promote cell death in HQPA in our time-course experiments (Fig. 3A). This MYC expressing cells.[33] Furthermore, LATS1 levels cell loss was associated with PARP-1 cleavage indicating are regulated by Ark5/Nuak1 phosphorylation and the activation of apoptosis pathways[32] preferentially LATS1 is needed for activation of Aurora B.[34, 35] in Myc overexpressing cells (Fig. 3B). While UW228 In order to confirm the specificity of AZD1152-HQPA and UW426 cells showed recovery of proliferation after action on Aurora B independent of Ark5/Nuak1, we release from 48 hr of Aurora B inhibition, proliferation probed for changes in LATS1 levels upon continuous was impaired in UW426-Myc and UW228-Myc cells after medulloblastoma cell exposure to AZD1152-HQPA for up release, suggesting that Aurora B inhibition has unique to 96 hours. We observed that LATS1 levels are preserved anti-proliferative effects on Myc overexpressing cells (Fig. in the presence of 100 nM AZD1152-HQPA over the 3C). first 24-48 hours. A slight reduction in LATS1 protein Figure 3: Sensitization of Myc overexpressing medulloblastoma cells to cell death and impaired cell proliferation in response to Aurora B inhibition. A) Cell counts for wild-type and MYC overexpressing medulloblastoma cells treated with continuous 100 nM AZD1152-HQPA exposure for up to 96 hours. *P <0.05 for comparison between vehicle and AZD1152-HQPA treated Myc overexpressing cells. Data represent mean for three independent experiments. Error bars are standard error of the mean. B) Western blots for PARP cleavage in response to 48 hours of sustained Aurora B inhibition in Myc overexpressing medulloblastoma cells versus control. 30 μg total protein loaded. C) Cell counts in MYC overexpressing or wild-type cells exposed to 100 nM AZD1152-HQPA for 48 hours followed by drug withdrawal. Data represent mean for three independent experiments. Error bars are standard error of the mean. *P<0.05 for comparison of MYC versus wild-type cells. www.impactjournals.com/oncotarget 3363 Oncotarget is seen at 72 -96 hours in both the wild-type and Myc- Aurora C phosphosignal. Aurora A and B levels remained overexpressing cell lines. (Supplementary Fig. S3) These stable after 48 hr of Aurora B inhibitor exposure (Fig. 4B). results indicate that the reduction of Aurora B kinase Also, no change in MYC protein level was observed as a activity and tumor cell polyploidy observed in the first result of Aurora B inhibition (Fig. 4C). The block in cell 24-48 hours of Aurora B inhibition with AZD1152-HQPA proliferation was associated with Caspase 3 cleavage upon is independent of Ark5/Nuak1 kinase regulated LATS1 continuous exposure to 100 nM AZD1152-HQPA for 48 levels. hr (Fig. 4D). In addition to UW228-Myc and UW426-Myc cells, In addition to the demonstration of a biochemical we also examined effects of Aurora B inhibition in MB cell response that is concordant with specific Aurora B lines with endogeneous Myc overexpression. We observed inhibition, we observed effects on cellular morphology a block in cell proliferation using an MTS assay over a and DNA content characteristic of Aurora B inhibition. period of 96 hr in D425 and D458 cells at AZD1152- Both wild-type and MYC overexpressing cells showed HQPA concentrations between 25 nM and 1000 nM (Fig. the development of a large cell, multinucleated phenotype 4A) Specific inhibition of Aurora B was also observed after 48 hr exposure to 100 nM AZD1152-HQPA in D458 and D425 cells with 100 nM AZD1152-HQPA (Supplementary Fig. S4A). Cells that overexpressed (Fig. 4B). These D458 and D425 cells did not show an MYC had greater DNA content compared to wild-type Figure 4: Anti-proliferative and pro-apoptotic effects of Aurora B inhibition in Group 3 medulloblastoma cells. A) MTS cell viability assay of D425 and D458 cells exposed to varying concentration of AZD1152-HQPA over a 96 hour period. Error bars represent standard error of the mean of 8 replicates per group per time point. B) Western blots showing the inhibition of Histone H3 (Ser 10) phosphorylation and Aurora B autophosphorylation by AZD1152-HQPA at 100 nM in D425 and D458 endogenous MYC overexpressing cells. Labels as follows: phosphohistone H3 Serine 10 (pH3 Ser 10), histone H3 (H3), aurora kinase A (AurA), aurora kinase B (AurB), Aurora A phosphothreonine 288 (pAurA), Aurora B phosphothreonine 232 (pAurB). Total protein loaded 30 μg. Note the lower band observed in the AurA blot is non-specific binding after re-probing of the membrane. C) Western blot for MYC in D425 and D458 cells exposed to 0.01% DMSO or 100 nM AZD1152-HQPA for 48 hours. Total protein loaded 30 μg. D) Western blots for cleaved caspase-3 (CC3) in D425 and D458 cells exposed to 0.01% DMSO or 100 nM AZD1152-HQPA for 48 hours. Positive control (+ CON) was 10 μL of etoposide treated Jurkat cell lysate (Cell Signaling). www.impactjournals.com/oncotarget 3364 Oncotarget isogenic controls when cytokinesis was blocked by Aurora Cell death effect of Aurora B inhibition in B inhibition (Supplementary Fig. S4B and S5). Induction medulloblastoma does not require endoreplication of endoreplication was associated with an increase in the proportion of sub-G (apoptotic) cells only in MYC Multiple studies have suggested that endoreplication overexpressing cells after 48 hr of Aurora B inhibition is essential for the apoptosis response triggered by Aurora (Supplementary Fig. S5). B inhibition.[16, 36-38] If this were to be the case for MB cells, this would limit the concurrent use of Aurora B inhibitors along with DNA synthesis inhibitors currently used in combination chemotherapy regimens for MB. Figure 5: Effect of DNA polymerase inhibition on cell viability in Myc overexpressing cells subjected to Aurora B inhibition. A) Percentage of cells with greater than 4N DNA content determined by FACS in MYC overexpressing cells after 24 hr of exposure to 0.01% DMSO (DMSO), 0.5 μM aphidicolin (APH), 100 nM AZD1152-HQPA (AZD1152), or 100 nM AZD1152-HQPA and 0.5 μM aphidicolin (AZD1152 + APH). *P<0.05. B) Cell counts in MYC ovexpressing cells continuously treated for 96 hr with 100 nM AZD1152-HQPA with or without exposure to 0.5 μM aphidicolin (APH) for the first 24 hr. Cells not treated with AZD1152-HQPA but exposed to aphidicolin show return of proliferation upon removal of aphidicolin after 24 hr. *P<0.05 for comparison between cells with and without inhibition of DNA replication by aphidicolin over the first 24 hr. C) Percentage of cells with subG DNA content determined by FACS in MYC overexpressing cells after 48 hr of continuous exposure 100 nM AZD1152-HQPA in the absence (AZD1152) or presence of 0.5 μM aphidicolin (AZD1152 + APH) for the first 24 hr. The graph also shows the proportion of subG cells when MYC overexpressing cells are exposed for 48 hr to control media containing 0.01% DMSO in the absence (DMSO) or presence of 0.5 μM aphidicolin (APH) for the first 24 hr. www.impactjournals.com/oncotarget 3365 Oncotarget Therefore we sought to determine if endoreplication to aphidicolin for the first 24 hours resulted in lower cell was a requirement for cell loss in MYC overexpressing viability in both UW426-Myc and UW228-Myc cells in medulloblastoma. In order to address this question, the which Aurora B inhibition was maintained for 72 hr (Fig. alpha-DNA polymerase inhibitor aphidicolin[39] was used 5B). In the absence of Aurora B inhibition, aphidicolin to block DNA replication for 24 hr. If endoreplication is blocked cell proliferation. The release from the aphidicolin a requirement for triggering apoptosis, the expected result block did not affect cell viability (Fig. 5B). Interestingly, would be the protection of MYC overexpressing cells from endoreplication occurred in UW426 and UW228 cells the cell death induced by Aurora B inhibition. Incubation which did not demonstrate an apoptotic response to of medulloblastoma cells with 0.5 μM aphidicolin and 100 Aurora B inhibition (Fig. 3, Supplementary Fig. S5). nM AZD1152-HQPA for 24 hr resulted in a significant These findings are in keeping with a DNA replication- reduction in the number of cells having > 4N DNA content independent trigger for apoptosis in MYC overexpressing (Fig. 5A). However, we did not see a protective effect of medulloblastoma cells when subjected to Aurora B blocking endoreplication on cell viability in cells exposed inhibition. to the Aurora B inhibitor (Fig. 5B,C). In fact, exposure Figure 6: Aurora B inhibition in a Myc overexpressing medulloblastoma flank xenograft model. A) Flank tumors were established in female nude mice by subcutaneous injection of UW426-Myc cells in a matrigel matrix. At 24 days post tumor cell implantation, mice were randomized to receive vehicle or 25 mg/kg/day AZD1152-HQPA x 4 days. Tumors were excised and frozen in th liquid nitrogen on the 5 day for Western blot analysis. Western blot shows inhibition of Aurora kinase B target Histone H3 Serine 10 (pH3 Ser10) and inhibition of autophosphorylation of Aurora B at threonine 232 (pAurB) in tumors from animals that received AZD1152-HQPA. Total protein loaded was 30 µg. B) H&E stain of flank UW426-Myc tumors in mice that received either vehicle or AZD1152-HQPA 25 mg/kg/day x 4 days. Scale bar 100 μm. C) Immunohistochemistry for Histone H3 Serine 10 phosphorylation and cleaved caspase-3 in UW426-Myc flank tumors from mice treated with vehicle (n=7) or AZD1152-HQPA (n=5). Graphs represent cell counts in more than 1000 cells per tumor per condition, *P<0.05. Scale bars 100 μm. D) Flank tumor mass on day 29 after tumor cell implantation in animals that received 50 mg/kg twice daily for 2 days starting on day 14 after cell implantation. Median and interquartile range depicted (Vehicle, n=7; AZD1152, n=10). E) Flank tumor volume in mice treated with vehicle (circles, n=6) or AZD1152-HQPA (squares, n=10) at 50 mg/kg day twice daily for 2 days starting on day 14 after tumor cell implantation. Tumor volumes are significantly different (P<0.05) from 4 days after treatment onwards. www.impactjournals.com/oncotarget 3366 Oncotarget the cerebellum and subarachnoid spaces from mice that Growth of MYC medulloblastoma flank received AZD1152-HQPA showed large multinucleated xenografts is impaired by Aurora B inhibition cells on H&E histology at day 7 of therapy (Fig. 7C). Furthermore, the proportion of phosphohistone H3 To test the results of Aurora B inhibition in-vivo, (Ser10) positive cells decreased and the proportion of we studied the drug-target effects using whole tumor cleaved-caspase 3 positive tumor cells increased with lysates probed for Histone H3 Serine 10 phosphorylation drug treatment (Fig. 7C,D). In addition to a reduction in and Aurora B autophosphorylation after administering tumor growth we observed an increase in mean survival of 25 mg/kg/day of inhibitor subcutaneously for 5 days mice bearing intracranial D458 tumors from 18 ± 0 days in (Fig. 6A). The cells in tissue sections from UW426-Myc the vehicle treated group to 34 ± 3 days in the AZD1152- tumors exposed to AZD1152-HQPA were multinucleated HQPA group (Log Rank P= 0.003, Fig. 7E). All of the and had lost the nodular pattern of cell organization drug treated animals died after withdrawal of therapy. BLI observed in untreated tumors (Fig. 6B). The proportion of demonstrated tumor growth (5/5) and metastasis (3/5) phosphohistone H3 (Ser10) positive cells was lower in the after cessation of drug therapy. Analysis of H&E spinal flank tumors from drug treated mice compared to vehicle cord sections revealed micrometastasis in 3/4 vehicle controls (Fig. 6C). Furthermore, the proportion of cleaved treated mice and 4/5 AZD1152-HQPA treated mice on day Caspase-3 positive cells was higher in mice with UW426- 7 of treatment, suggesting that Aurora B inhibition does Myc flank tumors that were treated with AZD1152-HQPA not alter the propensity for metastasis in MYC amplified compared to vehicle controls (Fig. 6C). Having confirmed medulloblastoma cells. target activity of AZD1152-HQPA for medulloblastoma flank tumors, the growth rate of the tumors was studied DISCUSSION by measuring tumor volume. A significant reduction in UW426-Myc tumor volume and weight at 28 days after Aurora kinase family proteins A and B have unique tumor cell implantation was achieved with subcutaneous functions in mitosis despite sharing a high similarity. administration of 50 mg/kg twice daily for 2 days (Fig. In this study, we have inhibited the kinase activity of 6D). This high dose, short course therapy was tested Aurora B in MB cells with both transgene mediated based on the observations that at least 48 hr of Aurora B and endogenous overexpression of MYC. As previously inhibition was required to trigger cell death in UW426- reported AZD1152-HQPA inhibits both Aurora B Myc cells. and Aurora C autophosphorylation[4] and this was demonstrated in the UW425 and UW228 wild-type and Aurora B inhibition impairs cerebellar MYC MYC-overexpressing cells within this study. However, medulloblastoma xenograft growth and prolongs medulloblastoma cells D425 and D458 lack an Aurora C survival phosphosignal suggesting that Aurora C function is not required for mitosis in at least some medulloblastoma cells. It is important to note that Aurora C has overlapping The pharmacokinetic parameters for subcutaneous function with Aurora B in mitosis.[40] administration of AZD1152-HQPA in nude mice using -1 We show in this study that medulloblastoma cells a one-component model were as follows: k = 0.24 hr ; overexpressing MYC are sensitized to apoptosis induced V = 190 μL; C = 13.3 ng/μL; t = 2.9 hours; AUC 0 1/2 linear upon Aurora B inhibition. The induction of apoptosis in = 68 ng • hours/μL (Fig. 7A). The calculated effective these cells is time dependent, with cell loss occurring only therapeutic plasma concentration time was 11 hr for after 48 hours of continuous inhibitor exposure. While a dose of 2.5 mg (equivalent to 50 mg/kg for a 25 gm endoreplication takes place upon Aurora B inhibition as mouse). The biodistribution of AZD1152-HQPA in the a result of cytokinesis block and G2/M escape and this brain was confirmed using LC/MS/MS after subcutaneous process is driven by MYC overexpression, we found administration of the drug in a phosphate buffered saline that inhibition of DNA replication does not protect solution. The peak brain content of AZD1152-HQPA against cell loss. Furthermore, endoreplication occurred was 0.7 ± 0.2 ng/mg brain tissue (n=4) at 2 hr after in cells that did not overexpress MYC and these cells administration. did not show an apoptotic response. These observations The D458 cell line was modified to express suggest that DNA synthesis inhibitors should not Luciferase by lentiviral transduction (D458-Luc/ affect the therapeutic effects of Aurora B inhibition in GFP). We observed the formation of D458-Luc/GFP medulloblastoma cells overexpressing MYC. This is an tumor cell grafts in the cerebellum and monitored important consideration since medulloblastoma is often growth by bioluminescence imaging (BLI) over time. treated with multiple chemotherapeutic drugs, including Daily administration of AZD1152-HQPA 50 mg/kg inhibitors of DNA synthesis such as methotrexate and subcutaneously for a 21-day period resulted in impaired topoisomerase inhibitors.[41] A synergistic effect on tumor growth as measured by percent change in photon blocking cell proliferation has been shown by Aurora B flux on BLI on day 7 of therapy (Fig. 7B). Tumor cells in www.impactjournals.com/oncotarget 3367 Oncotarget Figure 7: Aurora B inhibition in D458 human medulloblastoma intracranial xenograft model. A) Plasma concentration of AZD1152-HQPA in mice measured by LC/MS/MS at different time points after administration of 100 mg/kg subcutaneously in the dorsal skin fold. Error bars represent S.E.M. B) Whole brain content of AZD1152-HQPA in non-tumor bearing mice measured by LC/MS/MS at different time points after subcutaneous administration of 100 mg/kg (~3 mg per animal). Error bars represent S.E.M. C) Bioluminescence measurements represented as a change in photon flux from intracranial tumor at 1 week compared to start of treatment in mice treated with vehicle (n=12) or AZD1152-HQPA 50 mg/kg/day (n=12), P<0.01. D) H&E stain and immunohistochemistry for Histone H3 Serine 10 phosphorylation (pH3 Ser10) and cleaved caspase-3 (CC3) on cerebellar D458 tumors in mice that received either vehicle (n=6) or AZD1152-HQPA 50 mg/kg/day x 7 days (n=7). Scale bars 100 μm. E) Percentage of cells positive for pH3Ser10 or cleaved caspase-3 in cerebellar D458 tumors from mice that received either vehicle (n=6) or AZD1152-HQPA 50 mg/kg/day x 7 days (n=7), *P<0.05. F) Survival curve for mice bearing D458 cerebellar tumors treated with vehicle (n=5) or AZD1152-HQPA (n=5) 50 mg/kg/day for 21 days beginning 7 days after tumor cell implantation, Log-Rank P=0.003. www.impactjournals.com/oncotarget 3368 Oncotarget inhibition and concurrent inhibition of DNA synthesis appropriate downstream effects of Aurora B inhibition. with Topoisomerase II inhibitor daunorubicin in human The orthotopic cerebellar xenograft model demonstrated leukemia.[42] that AZD1152-HQPA can cross the blood brain barrier at Contrary to our results and those of Yang et al.,[42] effective concentrations to inhibit tumor cell growth and retinal pigment epithelium cells overexpressing MYC prolong survival. in which DNA replication is blocked by aphidicolin are AZD1152-HQPA has a clear cytostatic effect in- protected from induction of apoptosis by VX-680, an vivo, with regrowth of tumor upon withdrawal of drug inhibitor of both Aurora A and B.[16] It is important to therapy. We have demonstrated that if cells are exposed to note that DNA synthesis was inhibited prior to Aurora persistent Aurora B inhibition in-vitro for at least 48 hours kinase inhibition in that study, which means that the cells that the rate of cell proliferation is reduced following drug did not enter a 4N state. We propose that the trigger for the withdrawal. These effects are not mirrored in our in-vivo apoptotic response to Aurora B inhibition is initiated prior models and may be explained by sub-therapeutic levels to endoreplication, but requires entry into G2/M. This is of Aurora B inhibitor at the tumor site, combined with in keeping with the recently reported mechanism of ATM- persistence of tumor cells that are not actively cycling. dependent DNA damage response activated upon telomere Pharmacokinetic analysis indicates that during treatment deprotection that results from prolongation of mitosis by there were periods in which the plasma concentration of mitotic inhibitors, including Aurora kinase B inhibitor the drug was below the in-vitro therapeutic target of 100 Hesparadin.[43] nM. Given the significant suppression of tumor growth Prior reports of Aurora B inhibition in colon with a single 48-hour period of high-dose therapy in the carcinoma cells with ZM447439 in which post-mitotic UW426-Myc flank model, it is possible that sustained and G1 arrest was initiated by forced expression of p21 efficient tumor suppression could be achieved with a short or p27, showed that endoreplication was required for (48-72 hr) course of AZD1152-HQPA administered at a apoptosis in these cells.[37] We did not observe apoptosis dose of 50 mg/kg every 8 hours in mice. The presence of in UW426 and UW228 cells despite the induction of AZD1152-HQPA in brain has been recently documented endoreplication in these cells. In order to explain this and we also confirm this finding.[47] AZD1152 which is cell-type dependent response to Aurora B inhibition and the pro-drug for AZD1152-HQPA is the only selective the delayed apoptotic response to Aurora B inhibition Aurora B inhibitor that has been assessed in phase I in UW426-Myc and UW228-Myc cells we propose and II clinical trials. In a clinical trial AZD1152 was that transcriptional changes take place that are unique administered as a 1200 mg infusion over 7 days every 21 to MYC-overexpressing cells and that these changes days in 32 AML patients with completion of ≥ 2 cycles modify the cellular response to Aurora B inhibition. in 50%.[48] The geometric mean plasma half-life of Previous groups have reported upregulation of pro- AZD1152-HQPA was 77.5 hr with steady-state plasma apoptotic genes and down-regulation of anti-apoptotic concentration being 245.2 ng/mL (0.483 M).[48] Thus, genes with Aurora B inhibition in glioblastoma and colon a plasma concentration greater than 100 nM, which was cancer cells.[44, 45] Medulloblastoma cells engineered the therapeutically effective concentration in our tumor to overexpress MYC by retroviral transduction showed model, is achievable in humans. an upregulation of ribosomal biosynthesis genes, which Our study demonstrates that intracranial is in keeping with the role of MYC as a transcriptional medulloblastoma tumor growth can be impaired with amplifier driving pathways involved in RNA synthesis AZD1152-HQPA. These findings provide evidence to and protein production.[46] The role of transcriptional support the investigation of Aurora B inhibitors for the response in driving sensitization to Aurora B inhibition treatment of primary CNS malignancy and specifically and its potential for therapeutic application remains to be their use as a targeted therapeutic agent for G3M tumors. addressed. Changes in the gene expression state of MYC- The use of Aurora B inhibitors for the treatment of overexpressing cells could explain the observation that G3M tumors may be further evaluated by examining the these cells proliferate at a slower rate than untreated cells effectiveness and optimal dosing of of AZD1152-HQPA in when released from Aurora B inhibition after a period of murine models of G3M tumors such as those described by 48 hours. Pei et al.[49] and Kawauchi et al.[50] We have demonstrated the therapeutic potential of Aurora B inhibition in MYC overexpressing MATERIALS AND METHODS medulloblastoma in both a flank and an intracranial xenograft model. The flank xenograft model allowed us to assess the efficacy of AZD1152-HQPA in a location where Cell culture blood brain barrier penetration does not have a bearing on the treatment of tumor cells. This model further facilitated Medulloblastoma cells were cultured in standard the accurate biochemical readouts that confirm AZD1152- media at 37°C in a 5% CO atmosphere. All cells were HQPA could effectively reach its target and initiate the www.impactjournals.com/oncotarget 3369 Oncotarget confirmed to be mycoplasma negative by PCR techniques paraformaldehyde at room temperature for 10 minutes. at the Hospital for Sick Children microbiology laboratory. The cell membranes were permeabilized with 0.5% v/v Adherent cells were passaged using 0.05% Trypsin/0.53 Triton-X100 (Sigma) in PBS. Antigens were blocked with mM EDTA. A list of all cell lines and culture media is 1% bovine serum albumin fraction V (Merck) in PBS for provided in Supplementary Methods S1. 1 hr at room temperature followed by immunolabeling. The primary and secondary antibodies used, incubation protocol, and DNA staining and are described in Bioluminescent cell line Supplementary Methods S1. An Olympus 1X81 spinning disc confocal microscope (Olympus Canada Inc., D458 cells were transfected with lentivirus carrying Richmond Hill, ON, Canada) with Yokogawa scan head an expression cassette for Luciferase and GFP (D458-Luc/ (Yokogawa Corporation of America, Sugar Land, TX, GFP) as described previously.[4] USA) was used to visualize the cells with 20X air or 40X water immersion lenses (Carl Zeiss AG, Germany). Images Western Blots were processed using Volocity 5.5 imaging software. FACS cell cycle analysis Cell lysates were derived by using RIPA lysis buffer for non-phosphoproteins, or modified RIPA buffer as described previously.[4] Lysates from flank tumors were Cells were grown to confluence, dissociated, and obtained by grinding the liquid nitrogen frozen whole pelleted for resuspension in 50 µL of staining media (1M tumor with a mortar and pestle and suspending the powder HEPES (pH 7.2), 1M NaN , 2% fetal bovine serum in in modified RIPA lysis buffer. Proteins were separated by Hank’s Buffered Salt Solution). The cell suspension was SDS-PAGE on 10 to 12.5% gels and transferred to PVDF gently mixed with 1 mL 80% ethanol. The cells were membranes using a semi-dry transfer apparatus (Bio-Rad, centrifuged at 200 x g for 5 min at 4°C for pelleting and Hercules, CA). The membranes were washed in TBS-T resuspended in Hank’s Buffered Salt Solution (HBSS, (100 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.1% Tween-20). Wisent Inc.) containing 2 mg/mL RNAse A (Qiagen Inc., Incubation with primary antibody in TBS-T with 5% Toronto, ON, Canada) for 5 minutes at room temperature. Bovine Serum Albumin (Sigma) or 5% non-fat milk was Cells were then pelleted and resuspended in HBSS performed at 4°C overnight followed by incubation with containing 0.1 mg/mL propidium iodide and 0.6% (w/v) horseradish peroxidase conjugated secondary antibody NP40 at room temperature for 30 minutes. Subsequently, for 1 hr and enhanced chemiluminescence detection the cell pellet was resuspended in 500 µL of staining (Supplementary Methods S1). media and filtered through a 100 µm nylon cell strainer (BD Biosciences Discovery Labware, Bedford, MA, Cell viability assays USA). A total of 10,000 cells were sorted for each cell line and experimental condition and each experiment was performed in triplicate. An LSR II analyzer (Becton Cell viability was assessed over 96 hr Dickinson, Franklin Lakes, NJ, USA) was used with using MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3- FACSDiva software. The excitation wavelength was 523 carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- nm and the emission filter was LP600, BP610/20. Data tetrazolium) absorbance (Promega, Madison, WI, USA) at analysis was conducted with FlowJo (Tree Star Inc., 490 nm with at least 8 repeats per treatment condition per Ashland, OR, USA). Doublets were excluded by gating time point in each experiment. For cell count experiments, using the PI-intensity versus FSC-W graph as described 1 x 10 cells were cultured on a 3-cm Petri dish for 24 hr previously.[4] prior to exposure to drug inhibitor and grown for a total period of 96 hr with varying concentrations of AZD1152- Xenograft tumor model HQPA (Selleck Chemicals LLC, Houston, TX, USA) dissolved in dimethyl sulfoxide (DMSO, Sigma, St. Louis, MO, USA). The final DMSO concentration was Animal experiments were approved by the Hospital 0.01% v/v in media. The live cell number was determined for Sick Children Animal Care Committee (protocol by trypan-blue exclusion using a Vi-Cell XR counter 1000010458) and conducted in accordance with the (Beckman-Coulter, Mississauga, Ontario). Experiments Ontario Animals for Research Act and the Canadian were independently repeated three times. Council for Anima Care guidelines. Athymic nude 8-10 week old female mice Foxn1nu/Foxn1nu (Charles River, Immunofluorescent labeling and imaging Sherbrook, QC, Canada) were used for flank xenografts and intracranial xenografts. Isofluorane gas anaesthetic was used to induce an insensate state for inoculation Cells were washed with warm (37°C) phosphate of cells into the subcutaneous space at the right flank. buffered saline (PBS, Wisent Inc.) and fixed with 4% www.impactjournals.com/oncotarget 3370 Oncotarget 6 For implantation, 2 x 10 cells were suspended in a 1:1 after drug administration, blood was drawn by cardiac mixture of PBS and Matrigel (Becton Dickinson). A total puncture and collected in EDTA tubes. The mice were volume of 200 µL cell suspension was injected into the perfused with heparinized saline (5 units heparin/mL subcutaneous space using a 30 Ga needle. Fourteen days 0.9% NaCl). Subsequently the brain was excised and after cell inoculation palpable tumors were measured frozen in liquid nitrogen. Blood samples were centrifuged with a digital caliper to estimate the tumor volume (in at 800 x g for 10 min and the supernatant plasma was mm ) using the formula: tumor volume = length (mm) x frozen in liquid nitrogen. Frozen brain and blood samples 2 2 width (mm )/2. Animals were weighed for drug dosage were submitted to the Analytical Facility for Bioactive calculation and randomized to a vehicle or drug treatment Molecules, The Hospital for Sick Children for liquid group. Subcutaneous bolus injection in the dorsal skin fold chromatography-tandem mass spectrometry (LC/MS/MS) of each animal was performed daily for 4 days. AZD1152- analysis to determine AZD1152-HQPA concentration. HQPA was administered at a weight-determined volume Pharmacokinetic parameters were calculated from linear to achieve 25 or 50 mg/kg/day using a 2 or 4 mg per mL regression analysis of Log (plasma concentration) versus of 3.94% v/v DMSO in PBS solution. The vehicle group time graph assuming a one-component model. received a weight-determined volume of PBS with 3.94% DMSO vehicle solution. Tumor volumes were measured Immunohistochemistry every 2 days until day 30 after implantation of tumor cells. Animals were euthanized by carbon dioxide chamber Flank tumors and whole brains were fixed and either at day 5 or day 30 post-implantation and tumors processed for immunohistochemistry as described excised and immediately immersed in 3.7% formaldehyde previously.[4] Tissues were incubated with primary for histologic analysis or frozen in liquid nitrogen for antibody at room temperature for 1 hr. Rabbit polyclonal protein analysis. Formalin-fixed whole tumors were anti-cleaved Caspase 3 antibody (Cell Signaling) 1:500, weighed prior to sectioning and anti-phospho Histone H3 (Ser 10) (Cell Signaling) Intracranial implantation of 2.5 x 10 cells suspended 1:200 were used. TBS-T was used for washes. Detection in 3 µL of PBS was performed using sterile technique was performed with avidin-biotin-horseradish peroxidase under isofluorane gas anaesthetic. Cells were placed into complex (ABC; Vector Laboratories) followed by the right cerebellum at a depth of 2.5 mm via a single diaminobenzidine as the chromogen. Nuclei were burr hole 2 mm lateral to the midline and 1 mm posterior counterstained with hematoxylin. Slides were visualized to the junction of the sagittal and lamdoid sutures. Mice on an Olympus 1X37 light microscope. Cells were counted received 5 mg/kg ketoprofen analgesic and a bolus of in 20X fields to greater than 1000 cells per specimen. 0.5 mL 0.9% saline subcutaneously in the immediate post-operative period. At day 7, after implantation all mRNA expression analysis in medulloblastoma mice were imaged using Luciferin bioluminescence as tumors described previously.[4] Mice with defined posterior fossa tumors were randomized to AZD1152-HQPA or vehicle treatment. AZD1152-HQPA was given at a dose The expression profiles of medulloblastoma tumors, of 50 mg/kg/day using a 4 mg AZD1152-HQPA per mL cell lines, and normal samples were retrieved from GEO of 3.94% v/v DMSO in PBS administered in the dorsal accession GSE21140. The samples were expression- skin fold. Treatment was continued for 21-days and profiled and assigned subgroups as previously described Luciferin bioluminescence imaging was repeated at 7, 15, [20, 51] Pearson correlation was used to determine and 28 days after initiation of therapy. The vehicle group the association between MYC mRNA expression and received a weight-determined volume of PBS with 3.94% AURKA, AURKB, and AURKC mRNA expression in v/v DMSO vehicle solution. The survival endpoints were medulloblastoma tumors. death, weight loss greater than 20%, inability to mobilize, continuous seizure, moribund state. Tumors were excised mRNA expression profiling in cell lines for histological analysis 7 days after initiation of drug therapy and at the time of animal death. Subgrouping of D283, D425, Daoy, Med 8A, ONS-76, RES262, UW228 cells in comparison to 103 Pharmacokinetic experiment and AZD1152- medulloblastomas was performed by NanoString analysis HQPA quantification for selected signature genes differentiating Wnt, SHH, Group 3, and Group 4 tumors as previously described.[23] Mice were injected subcutaneously in the dorsal skin fold with 100 mg/kg AZD1152-HQPA using a solution of 4 mg AZD1152-HQPA per mL of 3.94% v/v DMSO in PBS. At 2 hr, 4 hr, 6 hr, 12 hr, and 24 hr www.impactjournals.com/oncotarget 3371 Oncotarget Rutka JT. Aurora kinase B/C inhibition impairs malignant Statistics glioma growth in vivo. J Neurooncol. 2012; 108(3):349- The PASW Statistics 18 software (SPSS Inc., 5. Lens SM, Voest EE and Medema RH. Shared and separate Chicago, IL) or Prism 6 (GraphPad Software, Inc., functions of polo-like kinases and aurora kinases in cancer. CA, USA) were used for statistical analysis. Data were Nat Rev Cancer. 2010; 10(12):825-841. assessed for normality by the Kolmogorov-Smirnov test. 6. Ditchfield C, Johnson VL, Tighe A, Ellston R, Haworth C, All measures are reported as mean +/- standard error of Johnson T, Mortlock A, Keen N and Taylor SS. Aurora B the mean. Means were compared by independent samples couples chromosome alignment with anaphase by targeting Student’s t-test. One-way ANOVA was used to compare BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol. mean absorbance at varying concentrations of inhibitor for 2003; 161(2):267-280. the MTS cell viability assay with post-hoc LSD testing 7. Tseng SH, Bobola MS, Berger MS and Silber JR. if significant differences were found between groups. For Characterization of paclitaxel (Taxol) sensitivity in human post-hoc LSD testing after ANOVA a P-value of less than glioma- and medulloblastoma-derived cell lines. Neuro- 0.025 was selected as significant. In all other analysis oncology. 1999; 1(2):101-108. a P-value less than 0.05 was selected for significance. Survival analysis was performed using a Kaplan-Meier 8. Goga A, Yang D, Tward AD, Morgan DO and Bishop JM. plot. The Breslow (Generalized Wilcoxon) test was used Inhibition of CDK1 as a potential therapy for tumors over- for inferential analysis with a P-value less than 0.05 expressing MYC. Nature medicine. 2007; 13(7):820-827. selected as significant. 9. Wilkinson RW, Odedra R, Heaton SP, Wedge SR, Keen NJ, Crafter C, Foster JR, Brady MC, Bigley A, Brown E, ACKNOWLEDGEMENTS Byth KF, Barrass NC, Mundt KE, Foote KM, Heron NM, Jung FH, et al. AZD1152, a selective inhibitor of Aurora B Dr. Diaz completed this work as a Canada Vanier kinase, inhibits human tumor xenograft growth by inducing Graduate Scholar at the University of Toronto. apoptosis. Clin Cancer Res. 2007; 13(12):3682-3688. 10. Lee C, Fotovati A, Triscott J, Chen J, Venugopal C, Singhal GRANT SUPPORT A, Dunham C, Kerr JM, Verreault M, Yip S, Wakimoto H, Jones C, Jayanthan A, Narendran A, Singh SK and Dunn SE. Polo-like kinase 1 inhibition kills glioblastoma Canadian Institutes for Health Research (CIHR), multiforme brain tumor cells in part through loss of SOX2 MOP-74610. b.r.a.i.n.child and the Laurie Berman Fund and delays tumor progression in mice. Stem Cells. 2012; for b.r.a.i.n.child and the Laurie Berman Fund for Brain 30(6):1064-1075. Tumour Research. 11. Harris PS, Venkataraman S, Alimova I, Birks DK, Donson AM, Knipstein J, Dubuc A, Taylor MD, Handler MH, REFERENCES Foreman NK and Vibhakar R. Polo-like kinase 1 (PLK1) inhibition suppresses cell growth and enhances radiation 1. Taylor MD, Northcott PA, Korshunov A, Remke M, Cho sensitivity in medulloblastoma cells. BMC Cancer. 2012; YJ, Clifford SC, Eberhart CG, Parsons DW, Rutkowski 12:80. S, Gajjar A, Ellison DW, Lichter P, Gilbertson RJ, 12. Nihal M, Stutz N, Schmit T, Ahmad N and Wood GS. Pomeroy SL, Kool M and Pfister SM. Molecular subgroups Polo-like kinase 1 (Plk1) is expressed by cutaneous T-cell of medulloblastoma: the current consensus. Acta lymphomas (CTCLs), and its downregulation promotes cell neuropathologica. 2012; 123(4):465-472. cycle arrest and apoptosis. Cell Cycle. 2011; 10(8):1303- 2. Northcott PA, Jones DT, Kool M, Robinson GW, Gilbertson RJ, Cho YJ, Pomeroy SL, Korshunov A, Lichter P, Taylor 13. Li Q and Dang CV. c-Myc overexpression uncouples DNA MD and Pfister SM. Medulloblastomics: the end of the replication from mitosis. Molecular and cellular biology. beginning. Nature reviews Cancer. 2012; 12(12):818-834. 1999; 19(8):5339-5351. 3. Kool M, Korshunov A, Remke M, Jones DT, Schlanstein 14. Horiuchi D, Kusdra L, Huskey NE, Chandriani S, Lenburg M, Northcott PA, Cho YJ, Koster J, Schouten-van ME, Gonzalez-Angulo AM, Creasman KJ, Bazarov AV, Meeteren A, van Vuurden D, Clifford SC, Pietsch T, von Smyth JW, Davis SE, Yaswen P, Mills GB, Esserman LJ Bueren AO, Rutkowski S, McCabe M, Collins VP, et al. and Goga A. MYC pathway activation in triple-negative Molecular subgroups of medulloblastoma: an international breast cancer is synthetic lethal with CDK inhibition. J Exp meta-analysis of transcriptome, genetic aberrations, Med. 2012; 209(4):679-696. and clinical data of WNT, SHH, Group 3, and Group 15. den Hollander J, Rimpi S, Doherty JR, Rudelius M, Buck 4 medulloblastomas. Acta neuropathologica. 2012; A, Hoellein A, Kremer M, Graf N, Scheerer M, Hall MA, 123(4):473-484. Goga A, von Bubnoff N, Duyster J, Peschel C, Cleveland 4. Diaz RJ, Golbourn B, Shekarforoush M, Smith CA and www.impactjournals.com/oncotarget 3372 Oncotarget JL, Nilsson JA, et al. Aurora kinases A and B are up- 25. Huang A, Ho CS, Ponzielli R, Barsyte-Lovejoy D, Bouffet regulated by Myc and are essential for maintenance of the E, Picard D, Hawkins CE and Penn LZ. Identification of malignant state. Blood. 2010; 116(9):1498-1505. a novel c-Myc protein interactor, JPO2, with transforming activity in medulloblastoma cells. Cancer research. 2005; 16. Yang D, Liu H, Goga A, Kim S, Yuneva M and Bishop 65(13):5607-5619. JM. Therapeutic potential of a synthetic lethal interaction between the MYC proto-oncogene and inhibition of 26. Zhou L, Picard D, Ra YS, Li M, Northcott PA, Hu Y, aurora-B kinase. Proc Natl Acad Sci U S A. 2010; Stearns D, Hawkins C, Taylor MD, Rutka J, Der SD and 107(31):13836-13841. Huang A. Silencing of thrombospondin-1 is critical for myc-induced metastatic phenotypes in medulloblastoma. 17. von Bueren AO, Shalaby T, Oehler-Janne C, Arnold L, Cancer research. 2010; 70(20):8199-8210. Stearns D, Eberhart CG, Arcaro A, Pruschy M and Grotzer MA. RNA interference-mediated c-MYC inhibition 27. Hsu JY, Sun ZW, Li X, Reuben M, Tatchell K, Bishop prevents cell growth and decreases sensitivity to radio- and DK, Grushcow JM, Brame CJ, Caldwell JA, Hunt DF, Lin chemotherapy in childhood medulloblastoma cells. BMC R, Smith MM and Allis CD. Mitotic phosphorylation of Cancer. 2009; 9:10. histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell. 2000; 18. Northcott PA, Shih DJ, Peacock J, Garzia L, Morrissy 102(3):279-291. AS, Zichner T, Stutz AM, Korshunov A, Reimand J, Schumacher SE, Beroukhim R, Ellison DW, Marshall CR, 28. Giet R and Glover DM. Drosophila aurora B kinase is Lionel AC, Mack S, Dubuc A, et al. Subgroup-specific required for histone H3 phosphorylation and condensin structural variation across 1,000 medulloblastoma genomes. recruitment during chromosome condensation and to Nature. 2012; 488(7409):49-56. organize the central spindle during cytokinesis. J Cell Biol. 2001; 152(4):669-682. 19. Park AK, Lee SJ, Phi JH, Wang KC, Kim DG, Cho BK, Haberler C, Fattet S, Dufour C, Puget S, Sainte-Rose C, 29. El-Sheikh A, Fan R, Birks D, Donson A, Foreman NK Bourdeaut F, Grill J, Delattre O, Kim SK and Park WY. and Vibhakar R. Inhibition of Aurora Kinase A enhances Prognostic classification of pediatric medulloblastoma chemosensitivity of medulloblastoma cell lines. Pediatr based on chromosome 17p loss, expression of MYCC and Blood Cancer. 2010; 55(1):35-41. MYCN, and Wnt pathway activation. Neuro Oncol. 2012; 30. Yasui Y, Urano T, Kawajiri A, Nagata K, Tatsuka M, 14(2):203-214. Saya H, Furukawa K, Takahashi T, Izawa I and Inagaki M. 20. Northcott PA, Korshunov A, Witt H, Hielscher T, Eberhart Autophosphorylation of a newly identified site of Aurora-B CG, Mack S, Bouffet E, Clifford SC, Hawkins CE, French is indispensable for cytokinesis. J Biol Chem. 2004; P, Rutka JT, Pfister S and Taylor MD. Medulloblastoma 279(13):12997-13003. comprises four distinct molecular variants. Journal of 31. Ohashi S, Sakashita G, Ban R, Nagasawa M, Matsuzaki H, clinical oncology : official journal of the American Society Murata Y, Taniguchi H, Shima H, Furukawa K and Urano of Clinical Oncology. 2011; 29(11):1408-1414. T. Phospho-regulation of human protein kinase Aurora-A: 21. Bigner SH, Friedman HS, Vogelstein B, Oakes WJ and analysis using anti-phospho-Thr288 monoclonal antibodies. Bigner DD. Amplification of the c-myc gene in human Oncogene. 2006; 25(59):7691-7702. medulloblastoma cell lines and xenografts. Cancer research. 32. Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE 1990; 50(8):2347-2350. and Poirier GG. Specific proteolytic cleavage of poly(ADP- 22. Friedman HS, Colvin OM, Kaufmann SH, Ludeman SM, ribose) polymerase: an early marker of chemotherapy- Bullock N, Bigner DD and Griffith OW. Cyclophosphamide induced apoptosis. Cancer research. 1993; 53(17):3976- resistance in medulloblastoma. Cancer research. 1992; 3985. 52(19):5373-5378. 33. Liu L, Ulbrich J, Muller J, Wustefeld T, Aeberhard L, Kress 23. Northcott PA, Shih DJ, Remke M, Cho YJ, Kool M, TR, Muthalagu N, Rycak L, Rudalska R, Moll R, Kempa Hawkins C, Eberhart CG, Dubuc A, Guettouche T, S, Zender L, Eilers M and Murphy DJ. Deregulated MYC Cardentey Y, Bouffet E, Pomeroy SL, Marra M, Malkin expression induces dependence upon AMPK-related kinase D, Rutka JT, Korshunov A, et al. Rapid, reliable, 5. Nature. 2012; 483(7391):608-612. and reproducible molecular sub-grouping of clinical 34. Humbert N, Navaratnam N, Augert A, Da Costa M, medulloblastoma samples. Acta neuropathologica. 2012; Martien S, Wang J, Martinez D, Abbadie C, Carling D, de 123(4):615-626. Launoit Y, Gil J and Bernard D. Regulation of ploidy and 24. He XM, Wikstrand CJ, Friedman HS, Bigner SH, Pleasure senescence by the AMPK-related kinase NUAK1. EMBO S, Trojanowski JQ and Bigner DD. Differentiation J. 2010; 29(2):376-386. characteristics of newly established medulloblastoma 35. Yabuta N, Mukai S, Okada N, Aylon Y and Nojima H. The cell lines (D384 Med, D425 Med, and D458 Med) and tumor suppressor Lats2 is pivotal in Aurora A and Aurora their transplantable xenografts. Laboratory investigation; B signaling during mitosis. Cell Cycle. 2011; 10(16):2724- a journal of technical methods and pathology. 1991; 64(6):833-843. www.impactjournals.com/oncotarget 3373 Oncotarget 36. Oke A, Pearce D, Wilkinson RW, Crafter C, Odedra R, 47. Marchetti S, Pluim D, van Eijndhoven M, van Tellingen Cavenagh J, Fitzgibbon J, Lister AT, Joel S and Bonnet O, Mazzanti R, Beijnen JH and Schellens JH. Effect of the D. AZD1152 rapidly and negatively affects the growth and drug transporters ABCG2, Abcg2, ABCB1 and ABCC2 on survival of human acute myeloid leukemia cells in vitro and the disposition, brain accumulation and myelotoxicity of the in vivo. Cancer research. 2009; 69(10):4150-4158. aurora kinase B inhibitor barasertib and its more active form barasertib-hydroxy-QPA. Invest New Drugs. 2013. 37. Kaestner P, Stolz A and Bastians H. Determinants for the efficiency of anticancer drugs targeting either Aurora-A 48. Lowenberg B, Muus P, Ossenkoppele G, Rousselot P, Cahn or Aurora-B kinases in human colon carcinoma cells. JY, Ifrah N, Martinelli G, Amadori S, Berman E, Sonneveld Molecular cancer therapeutics. 2009; 8(7):2046-2056. P, Jongen-Lavrencic M, Rigaudeau S, Stockman P, Goudie A, Faderl S, Jabbour E, et al. Phase 1/2 study to assess 38. Li M, Jung A, Ganswindt U, Marini P, Friedl A, Daniel the safety, efficacy, and pharmacokinetics of barasertib PT, Lauber K, Jendrossek V and Belka C. Aurora kinase (AZD1152) in patients with advanced acute myeloid inhibitor ZM447439 induces apoptosis via mitochondrial leukemia. Blood. 2011; 118(23):6030-6036. pathways. Biochem Pharmacol. 2010; 79(2):122-129. 49. Pei Y, Moore CE, Wang J, Tewari AK, Eroshkin A, Cho 39. Ikegami S, Taguchi T, Ohashi M, Oguro M, Nagano H YJ, Witt H, Korshunov A, Read TA, Sun JL, Schmitt and Mano Y. Aphidicolin prevents mitotic cell division EM, Miller CR, Buckley AF, McLendon RE, Westbrook by interfering with the activity of DNA polymerase-alpha. TF, Northcott PA, et al. An animal model of MYC-driven Nature. 1978; 275(5679):458-460. medulloblastoma. Cancer Cell. 2012; 21(2):155-167. 40. Sasai K, Katayama H, Stenoien DL, Fujii S, Honda 50. Kawauchi D, Robinson G, Uziel T, Gibson P, Rehg J, Gao R, Kimura M, Okano Y, Tatsuka M, Suzuki F, Nigg C, Finkelstein D, Qu C, Pounds S, Ellison DW, Gilbertson EA, Earnshaw WC, Brinkley WR and Sen S. Aurora-C RJ and Roussel MF. A mouse model of the most aggressive kinase is a novel chromosomal passenger protein that can subgroup of human medulloblastoma. Cancer Cell. 2012; complement Aurora-B kinase function in mitotic cells. Cell 21(2):168-180. Motil Cytoskeleton. 2004; 59(4):249-263. 51. Northcott PA, Fernandez LA, Hagan JP, Ellison DW, 41. Klesse LJ and Bowers DC. Childhood medulloblastoma: Grajkowska W, Gillespie Y, Grundy R, Van Meter T, current status of biology and treatment. CNS Drugs. 2010; Rutka JT, Croce CM, Kenney AM and Taylor MD. The 24(4):285-301. miR-17/92 polycistron is up-regulated in sonic hedgehog- 42. Yang J, Ikezoe T, Nishioka C, Tasaka T, Taniguchi A, driven medulloblastomas and induced by N-myc in sonic Kuwayama Y, Komatsu N, Bandobashi K, Togitani K, hedgehog-treated cerebellar neural precursors. Cancer Koeffler HP, Taguchi H and Yokoyama A. AZD1152, a research. 2009; 69(8):3249-3255. novel and selective aurora B kinase inhibitor, induces growth arrest, apoptosis, and sensitization for tubulin depolymerizing agent or topoisomerase II inhibitor in human acute leukemia cells in vitro and in vivo. Blood. 2007; 110(6):2034-2040. 43. Hayashi MT, Cesare AJ, Fitzpatrick JA, Lazzerini-Denchi E and Karlseder J. A telomere-dependent DNA damage checkpoint induced by prolonged mitotic arrest. Nature structural & molecular biology. 2012; 19(4):387-394. 44. Li J, Anderson MG, Tucker LA, Shen Y, Glaser KB and Shah OJ. Inhibition of Aurora B kinase sensitizes a subset of human glioma cells to TRAIL concomitant with induction of TRAIL-R2. Cell Death Differ. 2009; 16(3):498-511. 45. Shah OJ, Lin X, Li L, Huang X, Li J, Anderson MG, Tang H, Rodriguez LE, Warder SE, McLoughlin S, Chen J, Palma J, Glaser KB, Donawho CK, Fesik SW and Shen Y. Bcl-XL represents a druggable molecular vulnerability during aurora B inhibitor-mediated polyploidization. Proceedings of the National Academy of Sciences of the United States of America. 2010; 107(28):12634-12639. 46. Nie Z, Hu G, Wei G, Cui K, Yamane A, Resch W, Wang R, Green DR, Tessarollo L, Casellas R, Zhao K and Levens D. c-Myc is a universal amplifier of expressed genes in lymphocytes and embryonic stem cells. Cell. 2012; 151(1):68-79. www.impactjournals.com/oncotarget 3374 Oncotarget http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Oncotarget Pubmed Central

Mechanism of action and therapeutic efficacy of Aurora kinase B inhibition in MYC overexpressing medulloblastoma

Loading next page...
 
/lp/pubmed-central/mechanism-of-action-and-therapeutic-efficacy-of-aurora-kinase-b-DfW1yVPLkS

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
Copyright
Copyright: © 2015 Diaz et al.
eISSN
1949-2553
Publisher site
See Article on Publisher Site

Abstract

www.impactjournals.com/oncotarget/ Oncotarget, Vol. 6, No.5 Mechanism of action and therapeutic efficacy of Aurora kinase B inhibition in MYC overexpressing medulloblastoma 1,2,3,* 1,* 1 1 Roberto Jose Diaz , Brian Golbourn , Claudia Faria , Daniel Picard , David 1 4 4 1 1 Shih , Denis Raynaud , Michael Leadly , Danielle MacKenzie , Melissa Bryant , 1 1 1,2,3 1 Matthew Bebenek , Christian A. Smith , Michael D. Taylor , Annie Huang and 1,2,3 James T. Rutka The Hospital for Sick Children. Arthur and Sonia Labatt Brain Tumour Research Centre, Toronto, Ontario, Canada Department of Laboratory Medicine & Pathobiology, University of Toronto, Toronto, Ontario, Canada Division of Neurosurgery, Department of Surgery, University of Toronto, Toronto, Ontario, Canada Analytical Facility for Bioactive Molecules, The Hospital for Sick Children, Toronto, Ontario, Canada These authors contributed equally to this work Correspondence to: James T. Rutka, email: james.rutka@sickkids.ca Keywords: Aurora kinase, medulloblastoma, tumor biology, molecular therapy,cell-cycle Received: June 29, 2014 Accepted: December 24, 2014 Published: December 31, 2014 This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT Medulloblastoma comprises four molecular subgroups of which Group 3 medulloblastoma is characterized by MYC amplification and MYC overexpression. Lymphoma cells expressing high levels of MYC are susceptible to apoptosis following treatment with inhibitors of mitosis. One of the key regulatory kinases involved in multiple stages of mitosis is Aurora kinase B. We hypothesized that medulloblastoma cells that overexpress MYC would be uniquely sensitized to the apoptotic effects of Aurora B inhibition. The specific inhibition of Aurora kinase B was achieved in MYC- overexpressing medulloblastoma cells with AZD1152-HQPA. MYC overexpression sensitized medulloblastoma cells to cell death upon Aurora B inhibition. This process was found to be independent of endoreplication. Using both flank and intracranial cerebellar xenografts we demonstrate that tumors formed from MYC-overexpressing medulloblastoma cells show a response to Aurora B inhibition including growth impairment and apoptosis induction. Lastly, we show the distribution of AZD1152- HQPA within the mouse brain and the ability to inhibit intracranial tumor growth and prolong survival in mice bearing tumors formed from MYC-overexpressing medulloblastoma cells. Our results suggest the potential for therapeutic application of Aurora kinase B inhibitors in the treatment of Group 3 medulloblastoma. options. We have recently reported on the in-vitro pro- INTRODUCTION apoptotic and in-vivo cytostatic effects of Aurora B inhibition in glioblastoma using the inhibitor AZD1152- Recent genetic analysis of medulloblastoma (MB) HQPA.[4] Aurora kinase B (Aurora B) is an essential has revealed four genetically and epidemiologically serine/threonine kinase responsible for regulation distinct subgroups.[1, 2] One of these subgroups of multiple events in mitosis including chromosome designated as Group 3 medulloblastoma (G3MB) affects condensation, centromere-kinetochore dynamics, primarily infants and children and carries an overall chromatid segregation, and cytokinesis.[5] One of the survival of 45% for infants and 58% for children at five hallmarks of Aurora kinase B inhibition is G2/M arrest years as reported in a retrospective international meta- and subsequent escape into endoreplication cycles.[6] analysis.[3] Given the poor clinical outcomes observed Multiple regulatory or structural components essential to in patients with G3MB tumors with current treatment mitotic progression have been previously targeted to block regimens, there is a need to explore novel therapeutic www.impactjournals.com/oncotarget 3359 Oncotarget tumor cell proliferation including microtubules,[7] cyclin expression and Aurora B expression in G3MB (R=0.57, dependent kinase 1,[8] Aurora kinases,[9] and Polo-like P=0.002, N=27, Fig. 1C). Although WNT tumors express kinase 1.[10-12] When tumor cells overexpress c-Myc high levels of AURKB mRNA we did not observe a (MYC) either endogenously or by genetic modification, correlation to MYC mRNA expression in this small subset the induction of a G2/M arrest has been shown to trigger of tumor samples (R=0.42, P=0.3, N=8). Aurora kinase tumor cell death in-vitro and to block tumor growth in- gene expression is increased in fetal cerebellum and in all vivo.[8, 13, 14] This observation was also made recently subgroups of MB compared to adult cerebellum, reflecting in B-cell and T-cell lymphoma cells overexpressing the proliferative capacity of fetal and tumor tissue. MYC and treated with the Aurora B inhibitor AZD1152- To further evaluate the expression of Aurora kinase HQPA.[15, 16] MYC expression has also been reported to A and B in relation to MYC, protein expression in a confer sensitivity to radiation and DNA damaging agents number of unsynchronized MB cell lines was evaluated (etoposide, cisplatin) in medulloblastoma cells.[17] (Fig. 1D). The D425, D458 and MED8A cells, all of Amplification of MYC, as well as MYC which have known amplification of MYC,[18, 21, 22] overexpression, is a negative prognostic factor for overall showed concurrent marked expression of both Aurora B survival in MB.[18, 19] Approximately 11% of G3MB and MYC protein. Using a NanoString assay with a probe tumors demonstrate MYC amplification.[20] Furthermore, set for MB subgroup signature genes,[23] the D425 cell all G3MB tumors express MYC at high levels and line was demonstrated to share a similar gene expression express genes associated with elevated MYC levels.[20] signature as G3MB tumors (Supplementary Fig. S1). The We hypothesized that MB cells overexpressing MYC D425 cell line was derived from primary tumor tissue,[21] would be uniquely sensitized to the effects of Aurora B whereas the D458 line was derived from cerebrospinal inhibition and that this property could be harnessed for fluid within the same patient after failure of radiotherapy the in-vivo treatment of MYC-overexpressing MB tumors. and chemotherapy with cyclophosphamide, cisplatin, and The goal of our study was not only to determine if MYC vincristine.[22, 24] As such, our results suggest that the overexpression in human MB cells sensitized the cells to concordance between Aurora B and MYC expression is the apoptotic effects of Aurora B inhibition, but also to maintained in the setting of recurrent G3MB. further define the mechanism triggering this response. We demonstrate that Aurora B inhibition triggers cell death Myc overexpression sensitizes medulloblastoma independent of DNA replication and that transient Aurora cells to cell death induced by Aurora B inhibition B inhibition results in a unique impaired growth response in MYC-overexpressing cells. Having defined the response The UW228 and UW426 MB cell lines show time-course we proceeded to optimize in-vivo therapy with low expression of Aurora B and MYC (Supplementary AZD-1152 HQPA, achieving a prolongation in survival Fig. S2A). However, the cells can be modified to stably of mice bearing cerebellar xenografts of MB cells having overexpress MYC by retroviral transduction[25, 26] – MYC amplification and endogenously overexpressing the stable cell lines are identified here as UW228-Myc MYC. and UW426-Myc. Comparison of protein expression in MYC-overexpressing cells compared to isogenic RESULTS controls showed that increased expression of MYC was associated with an increase in Aurora A and B protein levels in unsynchronized cell culture (Supplementary Fig. Co-expression of Aurora B and MYC in Group 3 S2A). The cell-cycle distribution of UW228 cells did not medulloblastoma change with MYC overexpression while a reduction in G1/G0 cells with associated increase in polyploid cells was observed in UW426-Myc cells compared to control MYC has been shown to directly regulate the (Supplementary Fig. S2B). expression of Aurora A and indirectly the expression of The co-expression of Aurora B and MYC in MB Aurora B in B-cell lymphoma.[15] Therefore, we sought cells suggested that Aurora B activity could be important to determine if Aurora kinase gene expression correlates for cell survival in the presence of excess MYC, as is the with MYC expression in human MB. AURKA and AURKB case for B-cell lymphoma cells.[15] Therefore, we tested mRNA expression showed a positive correlation with MYC the ability of Aurora B inhibition to elicit cell death in mRNA expression (AURKA vs MYC: R=0.32, P=0.001, MYC overexpressing MB cells versus isogenic controls. N=103; AURKB vs MYC: R=0.37, P = 0.0001, N=103) We selected the Aurora B inhibitor AZD1152-HQPA while no correlation exists between AURKC and MYC for our studies as we had previously demonstrated that expression (Fig. 1A). The highest MYC expression was it had activity in brain and flank xenografts of human observed in WNT and G3MB relative to other subgroups, glioblastoma cells.[4] A series of drug concentrations normal fetal cerebellum, and adult cerebellum (Fig. 1B). was used to assess the lowest concentration of drug that Furthermore, there was a modest correlation between MYC www.impactjournals.com/oncotarget 3360 Oncotarget could inhibit MB cell proliferation in UW228 and UW426 sensitized cells to induction of cell death upon Aurora cells (Fig. 2A). Cell proliferation was inhibited in a B inhibition (Fig. 2A). Inhibition of proliferation was concentration-dependent manner both in isogenic controls achieved with 25 nM AZD1152-HQPA in UW426 and and MYC overexpressing cells (Fig. 2A). UW426 cells UW426-Myc as well as UW228 and UW228-Myc showed almost complete inhibition of proliferation when cells indicating a potent block of mitosis by Aurora exposed to 100 nM AZD1152-HQPA for 72 hrs, while B inhibition. The 72-hour time point was selected for UW426-Myc cells had cell counts that were almost half of determination of the concentration effects of AZD1152- those at baseline, indicating that cell loss had occurred (P< HQPA based on previous observations that induction of 0.001, N=3). A similar effect is observed for UW228 and cell death in response to Aurora B inhibition is a time- UW228-Myc cells, indicating that MYC overexpression dependent phenomenon.[4] Figure 1: Aurora kinase mRNA and protein expression in relation to Myc expression in medulloblastoma. A) mRNA expression of AURKA, AURKB, and AURKC in relation to MYC mRNA level in 103 medulloblastoma tumor samples. B) MYC mRNA expression in fetal cerebellum (fCb), adult cerebellum (aCb), and medulloblastoma tumors subgrouped according to RNA expression profile, ANOVA P<0.0001. C) Correlation between AURKB mRNA expression and MYC mRNA expression in medulloblastoma tumors subgrouped as Group 3. D) Western blot showing protein expression of Aurora A, Aurora B, and MYC in multiple medulloblastoma cell lines. Cell lines harboring MYC amplification are indicated by a star. The loading control was β-Actin. Total protein loaded was 30 μg. www.impactjournals.com/oncotarget 3361 Oncotarget In order to demonstrate the specificity of inhibition B on T232 is a key requirement for kinase activity.[30, of MB cells at 100 nM AZD1152-HQPA, cell lysates 31] Inhibition of Aurora B autophosphorylation without were probed for reduction in phosphorylation of Histone a change in total Aurora B levels was achieved upon H3 Serine 10, a specific substrate for Aurora B.[27, exposure of control and MYC-overexpressing cells to 100 28] Histone H3 Serine 10 phosphorylation is reduced nM AZD1152-HQPA for 24 hr (Fig. 2B). We note a weak with exposure to 100 nM AZD1152-HQPA over 24 phosphorylation signal of Aurora C (35 kDa) in UW425 hours without a change in total Histone H3 or β-actin and UW228 cells, which is abolished with AZD1152- protein levels (Fig. 2B). To further assess if AZD1152- HQPA (Fig 2B). The autophosphorylation of Aurora A HQPA could be interacting with Aurora A, a previously was observed to increase slightly in both wild-type and reported therapeutic target for MB,[29] we examined MYC-overexpressing cells with no change in total Aurora the inhibition of Aurora A and B autophosphorylation. A levels (Fig. 2B). Autophosphorylation of Aurora A on T288 and Aurora The onset of cell loss in MYC-overexpressing Figure 2: Cell viability effects and specificity of Aurora B inhibition with AZD1152-HQPA in medulloblastoma. A) Cell counts obtained by Vi-Cell XR counter (Beckman-Coulter, Mississauga, Ontario) in wild-type versus MYC overexpressing medulloblastoma cells exposed for 72 hours to varying concentrations of AZD1152-HQPA. *P <0.05 for comparison between wild-type and MYC overexpressing cells. Data represent mean for three independent experiments. Error bars are standard error of the mean. A total of 1 x 10 cells were cultured for 24 hours prior to exposure to AZD1152-HQPA. B) Western blots showing the inhibition of Histone H3 (Ser 10) phosphorylation and Aurora B autophosphorylation by AZD1152-HQPA at 100 nM in MYC overexpressing or wild-type cells when the drug is solubilized in water or DMSO. Labels as follows: phosphohistone H3 Serine 10 (pH3 Ser 10), histone H3 (H3), aurora kinase A (AurA), aurora kinase B (AurB), Aurora A phosphothreonine 288 (pAurA), Aurora B phosphothreonine 232 (pAurB), Aurora C phosphothreonine 198 (pAurC), β-Actin (Actin). Whole cell lysate from untreated U251 GBM cells was used as a reference antibody positive control (+). Total protein loaded was 30 μg. www.impactjournals.com/oncotarget 3362 Oncotarget MB cells compared to isogenic controls took place after Downregulation of the serine/threonine kinase 48 hours of continuous exposure to 100 nM AZD1152- Ark5/Nuak1 has been shown to promote cell death in HQPA in our time-course experiments (Fig. 3A). This MYC expressing cells.[33] Furthermore, LATS1 levels cell loss was associated with PARP-1 cleavage indicating are regulated by Ark5/Nuak1 phosphorylation and the activation of apoptosis pathways[32] preferentially LATS1 is needed for activation of Aurora B.[34, 35] in Myc overexpressing cells (Fig. 3B). While UW228 In order to confirm the specificity of AZD1152-HQPA and UW426 cells showed recovery of proliferation after action on Aurora B independent of Ark5/Nuak1, we release from 48 hr of Aurora B inhibition, proliferation probed for changes in LATS1 levels upon continuous was impaired in UW426-Myc and UW228-Myc cells after medulloblastoma cell exposure to AZD1152-HQPA for up release, suggesting that Aurora B inhibition has unique to 96 hours. We observed that LATS1 levels are preserved anti-proliferative effects on Myc overexpressing cells (Fig. in the presence of 100 nM AZD1152-HQPA over the 3C). first 24-48 hours. A slight reduction in LATS1 protein Figure 3: Sensitization of Myc overexpressing medulloblastoma cells to cell death and impaired cell proliferation in response to Aurora B inhibition. A) Cell counts for wild-type and MYC overexpressing medulloblastoma cells treated with continuous 100 nM AZD1152-HQPA exposure for up to 96 hours. *P <0.05 for comparison between vehicle and AZD1152-HQPA treated Myc overexpressing cells. Data represent mean for three independent experiments. Error bars are standard error of the mean. B) Western blots for PARP cleavage in response to 48 hours of sustained Aurora B inhibition in Myc overexpressing medulloblastoma cells versus control. 30 μg total protein loaded. C) Cell counts in MYC overexpressing or wild-type cells exposed to 100 nM AZD1152-HQPA for 48 hours followed by drug withdrawal. Data represent mean for three independent experiments. Error bars are standard error of the mean. *P<0.05 for comparison of MYC versus wild-type cells. www.impactjournals.com/oncotarget 3363 Oncotarget is seen at 72 -96 hours in both the wild-type and Myc- Aurora C phosphosignal. Aurora A and B levels remained overexpressing cell lines. (Supplementary Fig. S3) These stable after 48 hr of Aurora B inhibitor exposure (Fig. 4B). results indicate that the reduction of Aurora B kinase Also, no change in MYC protein level was observed as a activity and tumor cell polyploidy observed in the first result of Aurora B inhibition (Fig. 4C). The block in cell 24-48 hours of Aurora B inhibition with AZD1152-HQPA proliferation was associated with Caspase 3 cleavage upon is independent of Ark5/Nuak1 kinase regulated LATS1 continuous exposure to 100 nM AZD1152-HQPA for 48 levels. hr (Fig. 4D). In addition to UW228-Myc and UW426-Myc cells, In addition to the demonstration of a biochemical we also examined effects of Aurora B inhibition in MB cell response that is concordant with specific Aurora B lines with endogeneous Myc overexpression. We observed inhibition, we observed effects on cellular morphology a block in cell proliferation using an MTS assay over a and DNA content characteristic of Aurora B inhibition. period of 96 hr in D425 and D458 cells at AZD1152- Both wild-type and MYC overexpressing cells showed HQPA concentrations between 25 nM and 1000 nM (Fig. the development of a large cell, multinucleated phenotype 4A) Specific inhibition of Aurora B was also observed after 48 hr exposure to 100 nM AZD1152-HQPA in D458 and D425 cells with 100 nM AZD1152-HQPA (Supplementary Fig. S4A). Cells that overexpressed (Fig. 4B). These D458 and D425 cells did not show an MYC had greater DNA content compared to wild-type Figure 4: Anti-proliferative and pro-apoptotic effects of Aurora B inhibition in Group 3 medulloblastoma cells. A) MTS cell viability assay of D425 and D458 cells exposed to varying concentration of AZD1152-HQPA over a 96 hour period. Error bars represent standard error of the mean of 8 replicates per group per time point. B) Western blots showing the inhibition of Histone H3 (Ser 10) phosphorylation and Aurora B autophosphorylation by AZD1152-HQPA at 100 nM in D425 and D458 endogenous MYC overexpressing cells. Labels as follows: phosphohistone H3 Serine 10 (pH3 Ser 10), histone H3 (H3), aurora kinase A (AurA), aurora kinase B (AurB), Aurora A phosphothreonine 288 (pAurA), Aurora B phosphothreonine 232 (pAurB). Total protein loaded 30 μg. Note the lower band observed in the AurA blot is non-specific binding after re-probing of the membrane. C) Western blot for MYC in D425 and D458 cells exposed to 0.01% DMSO or 100 nM AZD1152-HQPA for 48 hours. Total protein loaded 30 μg. D) Western blots for cleaved caspase-3 (CC3) in D425 and D458 cells exposed to 0.01% DMSO or 100 nM AZD1152-HQPA for 48 hours. Positive control (+ CON) was 10 μL of etoposide treated Jurkat cell lysate (Cell Signaling). www.impactjournals.com/oncotarget 3364 Oncotarget isogenic controls when cytokinesis was blocked by Aurora Cell death effect of Aurora B inhibition in B inhibition (Supplementary Fig. S4B and S5). Induction medulloblastoma does not require endoreplication of endoreplication was associated with an increase in the proportion of sub-G (apoptotic) cells only in MYC Multiple studies have suggested that endoreplication overexpressing cells after 48 hr of Aurora B inhibition is essential for the apoptosis response triggered by Aurora (Supplementary Fig. S5). B inhibition.[16, 36-38] If this were to be the case for MB cells, this would limit the concurrent use of Aurora B inhibitors along with DNA synthesis inhibitors currently used in combination chemotherapy regimens for MB. Figure 5: Effect of DNA polymerase inhibition on cell viability in Myc overexpressing cells subjected to Aurora B inhibition. A) Percentage of cells with greater than 4N DNA content determined by FACS in MYC overexpressing cells after 24 hr of exposure to 0.01% DMSO (DMSO), 0.5 μM aphidicolin (APH), 100 nM AZD1152-HQPA (AZD1152), or 100 nM AZD1152-HQPA and 0.5 μM aphidicolin (AZD1152 + APH). *P<0.05. B) Cell counts in MYC ovexpressing cells continuously treated for 96 hr with 100 nM AZD1152-HQPA with or without exposure to 0.5 μM aphidicolin (APH) for the first 24 hr. Cells not treated with AZD1152-HQPA but exposed to aphidicolin show return of proliferation upon removal of aphidicolin after 24 hr. *P<0.05 for comparison between cells with and without inhibition of DNA replication by aphidicolin over the first 24 hr. C) Percentage of cells with subG DNA content determined by FACS in MYC overexpressing cells after 48 hr of continuous exposure 100 nM AZD1152-HQPA in the absence (AZD1152) or presence of 0.5 μM aphidicolin (AZD1152 + APH) for the first 24 hr. The graph also shows the proportion of subG cells when MYC overexpressing cells are exposed for 48 hr to control media containing 0.01% DMSO in the absence (DMSO) or presence of 0.5 μM aphidicolin (APH) for the first 24 hr. www.impactjournals.com/oncotarget 3365 Oncotarget Therefore we sought to determine if endoreplication to aphidicolin for the first 24 hours resulted in lower cell was a requirement for cell loss in MYC overexpressing viability in both UW426-Myc and UW228-Myc cells in medulloblastoma. In order to address this question, the which Aurora B inhibition was maintained for 72 hr (Fig. alpha-DNA polymerase inhibitor aphidicolin[39] was used 5B). In the absence of Aurora B inhibition, aphidicolin to block DNA replication for 24 hr. If endoreplication is blocked cell proliferation. The release from the aphidicolin a requirement for triggering apoptosis, the expected result block did not affect cell viability (Fig. 5B). Interestingly, would be the protection of MYC overexpressing cells from endoreplication occurred in UW426 and UW228 cells the cell death induced by Aurora B inhibition. Incubation which did not demonstrate an apoptotic response to of medulloblastoma cells with 0.5 μM aphidicolin and 100 Aurora B inhibition (Fig. 3, Supplementary Fig. S5). nM AZD1152-HQPA for 24 hr resulted in a significant These findings are in keeping with a DNA replication- reduction in the number of cells having > 4N DNA content independent trigger for apoptosis in MYC overexpressing (Fig. 5A). However, we did not see a protective effect of medulloblastoma cells when subjected to Aurora B blocking endoreplication on cell viability in cells exposed inhibition. to the Aurora B inhibitor (Fig. 5B,C). In fact, exposure Figure 6: Aurora B inhibition in a Myc overexpressing medulloblastoma flank xenograft model. A) Flank tumors were established in female nude mice by subcutaneous injection of UW426-Myc cells in a matrigel matrix. At 24 days post tumor cell implantation, mice were randomized to receive vehicle or 25 mg/kg/day AZD1152-HQPA x 4 days. Tumors were excised and frozen in th liquid nitrogen on the 5 day for Western blot analysis. Western blot shows inhibition of Aurora kinase B target Histone H3 Serine 10 (pH3 Ser10) and inhibition of autophosphorylation of Aurora B at threonine 232 (pAurB) in tumors from animals that received AZD1152-HQPA. Total protein loaded was 30 µg. B) H&E stain of flank UW426-Myc tumors in mice that received either vehicle or AZD1152-HQPA 25 mg/kg/day x 4 days. Scale bar 100 μm. C) Immunohistochemistry for Histone H3 Serine 10 phosphorylation and cleaved caspase-3 in UW426-Myc flank tumors from mice treated with vehicle (n=7) or AZD1152-HQPA (n=5). Graphs represent cell counts in more than 1000 cells per tumor per condition, *P<0.05. Scale bars 100 μm. D) Flank tumor mass on day 29 after tumor cell implantation in animals that received 50 mg/kg twice daily for 2 days starting on day 14 after cell implantation. Median and interquartile range depicted (Vehicle, n=7; AZD1152, n=10). E) Flank tumor volume in mice treated with vehicle (circles, n=6) or AZD1152-HQPA (squares, n=10) at 50 mg/kg day twice daily for 2 days starting on day 14 after tumor cell implantation. Tumor volumes are significantly different (P<0.05) from 4 days after treatment onwards. www.impactjournals.com/oncotarget 3366 Oncotarget the cerebellum and subarachnoid spaces from mice that Growth of MYC medulloblastoma flank received AZD1152-HQPA showed large multinucleated xenografts is impaired by Aurora B inhibition cells on H&E histology at day 7 of therapy (Fig. 7C). Furthermore, the proportion of phosphohistone H3 To test the results of Aurora B inhibition in-vivo, (Ser10) positive cells decreased and the proportion of we studied the drug-target effects using whole tumor cleaved-caspase 3 positive tumor cells increased with lysates probed for Histone H3 Serine 10 phosphorylation drug treatment (Fig. 7C,D). In addition to a reduction in and Aurora B autophosphorylation after administering tumor growth we observed an increase in mean survival of 25 mg/kg/day of inhibitor subcutaneously for 5 days mice bearing intracranial D458 tumors from 18 ± 0 days in (Fig. 6A). The cells in tissue sections from UW426-Myc the vehicle treated group to 34 ± 3 days in the AZD1152- tumors exposed to AZD1152-HQPA were multinucleated HQPA group (Log Rank P= 0.003, Fig. 7E). All of the and had lost the nodular pattern of cell organization drug treated animals died after withdrawal of therapy. BLI observed in untreated tumors (Fig. 6B). The proportion of demonstrated tumor growth (5/5) and metastasis (3/5) phosphohistone H3 (Ser10) positive cells was lower in the after cessation of drug therapy. Analysis of H&E spinal flank tumors from drug treated mice compared to vehicle cord sections revealed micrometastasis in 3/4 vehicle controls (Fig. 6C). Furthermore, the proportion of cleaved treated mice and 4/5 AZD1152-HQPA treated mice on day Caspase-3 positive cells was higher in mice with UW426- 7 of treatment, suggesting that Aurora B inhibition does Myc flank tumors that were treated with AZD1152-HQPA not alter the propensity for metastasis in MYC amplified compared to vehicle controls (Fig. 6C). Having confirmed medulloblastoma cells. target activity of AZD1152-HQPA for medulloblastoma flank tumors, the growth rate of the tumors was studied DISCUSSION by measuring tumor volume. A significant reduction in UW426-Myc tumor volume and weight at 28 days after Aurora kinase family proteins A and B have unique tumor cell implantation was achieved with subcutaneous functions in mitosis despite sharing a high similarity. administration of 50 mg/kg twice daily for 2 days (Fig. In this study, we have inhibited the kinase activity of 6D). This high dose, short course therapy was tested Aurora B in MB cells with both transgene mediated based on the observations that at least 48 hr of Aurora B and endogenous overexpression of MYC. As previously inhibition was required to trigger cell death in UW426- reported AZD1152-HQPA inhibits both Aurora B Myc cells. and Aurora C autophosphorylation[4] and this was demonstrated in the UW425 and UW228 wild-type and Aurora B inhibition impairs cerebellar MYC MYC-overexpressing cells within this study. However, medulloblastoma xenograft growth and prolongs medulloblastoma cells D425 and D458 lack an Aurora C survival phosphosignal suggesting that Aurora C function is not required for mitosis in at least some medulloblastoma cells. It is important to note that Aurora C has overlapping The pharmacokinetic parameters for subcutaneous function with Aurora B in mitosis.[40] administration of AZD1152-HQPA in nude mice using -1 We show in this study that medulloblastoma cells a one-component model were as follows: k = 0.24 hr ; overexpressing MYC are sensitized to apoptosis induced V = 190 μL; C = 13.3 ng/μL; t = 2.9 hours; AUC 0 1/2 linear upon Aurora B inhibition. The induction of apoptosis in = 68 ng • hours/μL (Fig. 7A). The calculated effective these cells is time dependent, with cell loss occurring only therapeutic plasma concentration time was 11 hr for after 48 hours of continuous inhibitor exposure. While a dose of 2.5 mg (equivalent to 50 mg/kg for a 25 gm endoreplication takes place upon Aurora B inhibition as mouse). The biodistribution of AZD1152-HQPA in the a result of cytokinesis block and G2/M escape and this brain was confirmed using LC/MS/MS after subcutaneous process is driven by MYC overexpression, we found administration of the drug in a phosphate buffered saline that inhibition of DNA replication does not protect solution. The peak brain content of AZD1152-HQPA against cell loss. Furthermore, endoreplication occurred was 0.7 ± 0.2 ng/mg brain tissue (n=4) at 2 hr after in cells that did not overexpress MYC and these cells administration. did not show an apoptotic response. These observations The D458 cell line was modified to express suggest that DNA synthesis inhibitors should not Luciferase by lentiviral transduction (D458-Luc/ affect the therapeutic effects of Aurora B inhibition in GFP). We observed the formation of D458-Luc/GFP medulloblastoma cells overexpressing MYC. This is an tumor cell grafts in the cerebellum and monitored important consideration since medulloblastoma is often growth by bioluminescence imaging (BLI) over time. treated with multiple chemotherapeutic drugs, including Daily administration of AZD1152-HQPA 50 mg/kg inhibitors of DNA synthesis such as methotrexate and subcutaneously for a 21-day period resulted in impaired topoisomerase inhibitors.[41] A synergistic effect on tumor growth as measured by percent change in photon blocking cell proliferation has been shown by Aurora B flux on BLI on day 7 of therapy (Fig. 7B). Tumor cells in www.impactjournals.com/oncotarget 3367 Oncotarget Figure 7: Aurora B inhibition in D458 human medulloblastoma intracranial xenograft model. A) Plasma concentration of AZD1152-HQPA in mice measured by LC/MS/MS at different time points after administration of 100 mg/kg subcutaneously in the dorsal skin fold. Error bars represent S.E.M. B) Whole brain content of AZD1152-HQPA in non-tumor bearing mice measured by LC/MS/MS at different time points after subcutaneous administration of 100 mg/kg (~3 mg per animal). Error bars represent S.E.M. C) Bioluminescence measurements represented as a change in photon flux from intracranial tumor at 1 week compared to start of treatment in mice treated with vehicle (n=12) or AZD1152-HQPA 50 mg/kg/day (n=12), P<0.01. D) H&E stain and immunohistochemistry for Histone H3 Serine 10 phosphorylation (pH3 Ser10) and cleaved caspase-3 (CC3) on cerebellar D458 tumors in mice that received either vehicle (n=6) or AZD1152-HQPA 50 mg/kg/day x 7 days (n=7). Scale bars 100 μm. E) Percentage of cells positive for pH3Ser10 or cleaved caspase-3 in cerebellar D458 tumors from mice that received either vehicle (n=6) or AZD1152-HQPA 50 mg/kg/day x 7 days (n=7), *P<0.05. F) Survival curve for mice bearing D458 cerebellar tumors treated with vehicle (n=5) or AZD1152-HQPA (n=5) 50 mg/kg/day for 21 days beginning 7 days after tumor cell implantation, Log-Rank P=0.003. www.impactjournals.com/oncotarget 3368 Oncotarget inhibition and concurrent inhibition of DNA synthesis appropriate downstream effects of Aurora B inhibition. with Topoisomerase II inhibitor daunorubicin in human The orthotopic cerebellar xenograft model demonstrated leukemia.[42] that AZD1152-HQPA can cross the blood brain barrier at Contrary to our results and those of Yang et al.,[42] effective concentrations to inhibit tumor cell growth and retinal pigment epithelium cells overexpressing MYC prolong survival. in which DNA replication is blocked by aphidicolin are AZD1152-HQPA has a clear cytostatic effect in- protected from induction of apoptosis by VX-680, an vivo, with regrowth of tumor upon withdrawal of drug inhibitor of both Aurora A and B.[16] It is important to therapy. We have demonstrated that if cells are exposed to note that DNA synthesis was inhibited prior to Aurora persistent Aurora B inhibition in-vitro for at least 48 hours kinase inhibition in that study, which means that the cells that the rate of cell proliferation is reduced following drug did not enter a 4N state. We propose that the trigger for the withdrawal. These effects are not mirrored in our in-vivo apoptotic response to Aurora B inhibition is initiated prior models and may be explained by sub-therapeutic levels to endoreplication, but requires entry into G2/M. This is of Aurora B inhibitor at the tumor site, combined with in keeping with the recently reported mechanism of ATM- persistence of tumor cells that are not actively cycling. dependent DNA damage response activated upon telomere Pharmacokinetic analysis indicates that during treatment deprotection that results from prolongation of mitosis by there were periods in which the plasma concentration of mitotic inhibitors, including Aurora kinase B inhibitor the drug was below the in-vitro therapeutic target of 100 Hesparadin.[43] nM. Given the significant suppression of tumor growth Prior reports of Aurora B inhibition in colon with a single 48-hour period of high-dose therapy in the carcinoma cells with ZM447439 in which post-mitotic UW426-Myc flank model, it is possible that sustained and G1 arrest was initiated by forced expression of p21 efficient tumor suppression could be achieved with a short or p27, showed that endoreplication was required for (48-72 hr) course of AZD1152-HQPA administered at a apoptosis in these cells.[37] We did not observe apoptosis dose of 50 mg/kg every 8 hours in mice. The presence of in UW426 and UW228 cells despite the induction of AZD1152-HQPA in brain has been recently documented endoreplication in these cells. In order to explain this and we also confirm this finding.[47] AZD1152 which is cell-type dependent response to Aurora B inhibition and the pro-drug for AZD1152-HQPA is the only selective the delayed apoptotic response to Aurora B inhibition Aurora B inhibitor that has been assessed in phase I in UW426-Myc and UW228-Myc cells we propose and II clinical trials. In a clinical trial AZD1152 was that transcriptional changes take place that are unique administered as a 1200 mg infusion over 7 days every 21 to MYC-overexpressing cells and that these changes days in 32 AML patients with completion of ≥ 2 cycles modify the cellular response to Aurora B inhibition. in 50%.[48] The geometric mean plasma half-life of Previous groups have reported upregulation of pro- AZD1152-HQPA was 77.5 hr with steady-state plasma apoptotic genes and down-regulation of anti-apoptotic concentration being 245.2 ng/mL (0.483 M).[48] Thus, genes with Aurora B inhibition in glioblastoma and colon a plasma concentration greater than 100 nM, which was cancer cells.[44, 45] Medulloblastoma cells engineered the therapeutically effective concentration in our tumor to overexpress MYC by retroviral transduction showed model, is achievable in humans. an upregulation of ribosomal biosynthesis genes, which Our study demonstrates that intracranial is in keeping with the role of MYC as a transcriptional medulloblastoma tumor growth can be impaired with amplifier driving pathways involved in RNA synthesis AZD1152-HQPA. These findings provide evidence to and protein production.[46] The role of transcriptional support the investigation of Aurora B inhibitors for the response in driving sensitization to Aurora B inhibition treatment of primary CNS malignancy and specifically and its potential for therapeutic application remains to be their use as a targeted therapeutic agent for G3M tumors. addressed. Changes in the gene expression state of MYC- The use of Aurora B inhibitors for the treatment of overexpressing cells could explain the observation that G3M tumors may be further evaluated by examining the these cells proliferate at a slower rate than untreated cells effectiveness and optimal dosing of of AZD1152-HQPA in when released from Aurora B inhibition after a period of murine models of G3M tumors such as those described by 48 hours. Pei et al.[49] and Kawauchi et al.[50] We have demonstrated the therapeutic potential of Aurora B inhibition in MYC overexpressing MATERIALS AND METHODS medulloblastoma in both a flank and an intracranial xenograft model. The flank xenograft model allowed us to assess the efficacy of AZD1152-HQPA in a location where Cell culture blood brain barrier penetration does not have a bearing on the treatment of tumor cells. This model further facilitated Medulloblastoma cells were cultured in standard the accurate biochemical readouts that confirm AZD1152- media at 37°C in a 5% CO atmosphere. All cells were HQPA could effectively reach its target and initiate the www.impactjournals.com/oncotarget 3369 Oncotarget confirmed to be mycoplasma negative by PCR techniques paraformaldehyde at room temperature for 10 minutes. at the Hospital for Sick Children microbiology laboratory. The cell membranes were permeabilized with 0.5% v/v Adherent cells were passaged using 0.05% Trypsin/0.53 Triton-X100 (Sigma) in PBS. Antigens were blocked with mM EDTA. A list of all cell lines and culture media is 1% bovine serum albumin fraction V (Merck) in PBS for provided in Supplementary Methods S1. 1 hr at room temperature followed by immunolabeling. The primary and secondary antibodies used, incubation protocol, and DNA staining and are described in Bioluminescent cell line Supplementary Methods S1. An Olympus 1X81 spinning disc confocal microscope (Olympus Canada Inc., D458 cells were transfected with lentivirus carrying Richmond Hill, ON, Canada) with Yokogawa scan head an expression cassette for Luciferase and GFP (D458-Luc/ (Yokogawa Corporation of America, Sugar Land, TX, GFP) as described previously.[4] USA) was used to visualize the cells with 20X air or 40X water immersion lenses (Carl Zeiss AG, Germany). Images Western Blots were processed using Volocity 5.5 imaging software. FACS cell cycle analysis Cell lysates were derived by using RIPA lysis buffer for non-phosphoproteins, or modified RIPA buffer as described previously.[4] Lysates from flank tumors were Cells were grown to confluence, dissociated, and obtained by grinding the liquid nitrogen frozen whole pelleted for resuspension in 50 µL of staining media (1M tumor with a mortar and pestle and suspending the powder HEPES (pH 7.2), 1M NaN , 2% fetal bovine serum in in modified RIPA lysis buffer. Proteins were separated by Hank’s Buffered Salt Solution). The cell suspension was SDS-PAGE on 10 to 12.5% gels and transferred to PVDF gently mixed with 1 mL 80% ethanol. The cells were membranes using a semi-dry transfer apparatus (Bio-Rad, centrifuged at 200 x g for 5 min at 4°C for pelleting and Hercules, CA). The membranes were washed in TBS-T resuspended in Hank’s Buffered Salt Solution (HBSS, (100 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.1% Tween-20). Wisent Inc.) containing 2 mg/mL RNAse A (Qiagen Inc., Incubation with primary antibody in TBS-T with 5% Toronto, ON, Canada) for 5 minutes at room temperature. Bovine Serum Albumin (Sigma) or 5% non-fat milk was Cells were then pelleted and resuspended in HBSS performed at 4°C overnight followed by incubation with containing 0.1 mg/mL propidium iodide and 0.6% (w/v) horseradish peroxidase conjugated secondary antibody NP40 at room temperature for 30 minutes. Subsequently, for 1 hr and enhanced chemiluminescence detection the cell pellet was resuspended in 500 µL of staining (Supplementary Methods S1). media and filtered through a 100 µm nylon cell strainer (BD Biosciences Discovery Labware, Bedford, MA, Cell viability assays USA). A total of 10,000 cells were sorted for each cell line and experimental condition and each experiment was performed in triplicate. An LSR II analyzer (Becton Cell viability was assessed over 96 hr Dickinson, Franklin Lakes, NJ, USA) was used with using MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3- FACSDiva software. The excitation wavelength was 523 carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- nm and the emission filter was LP600, BP610/20. Data tetrazolium) absorbance (Promega, Madison, WI, USA) at analysis was conducted with FlowJo (Tree Star Inc., 490 nm with at least 8 repeats per treatment condition per Ashland, OR, USA). Doublets were excluded by gating time point in each experiment. For cell count experiments, using the PI-intensity versus FSC-W graph as described 1 x 10 cells were cultured on a 3-cm Petri dish for 24 hr previously.[4] prior to exposure to drug inhibitor and grown for a total period of 96 hr with varying concentrations of AZD1152- Xenograft tumor model HQPA (Selleck Chemicals LLC, Houston, TX, USA) dissolved in dimethyl sulfoxide (DMSO, Sigma, St. Louis, MO, USA). The final DMSO concentration was Animal experiments were approved by the Hospital 0.01% v/v in media. The live cell number was determined for Sick Children Animal Care Committee (protocol by trypan-blue exclusion using a Vi-Cell XR counter 1000010458) and conducted in accordance with the (Beckman-Coulter, Mississauga, Ontario). Experiments Ontario Animals for Research Act and the Canadian were independently repeated three times. Council for Anima Care guidelines. Athymic nude 8-10 week old female mice Foxn1nu/Foxn1nu (Charles River, Immunofluorescent labeling and imaging Sherbrook, QC, Canada) were used for flank xenografts and intracranial xenografts. Isofluorane gas anaesthetic was used to induce an insensate state for inoculation Cells were washed with warm (37°C) phosphate of cells into the subcutaneous space at the right flank. buffered saline (PBS, Wisent Inc.) and fixed with 4% www.impactjournals.com/oncotarget 3370 Oncotarget 6 For implantation, 2 x 10 cells were suspended in a 1:1 after drug administration, blood was drawn by cardiac mixture of PBS and Matrigel (Becton Dickinson). A total puncture and collected in EDTA tubes. The mice were volume of 200 µL cell suspension was injected into the perfused with heparinized saline (5 units heparin/mL subcutaneous space using a 30 Ga needle. Fourteen days 0.9% NaCl). Subsequently the brain was excised and after cell inoculation palpable tumors were measured frozen in liquid nitrogen. Blood samples were centrifuged with a digital caliper to estimate the tumor volume (in at 800 x g for 10 min and the supernatant plasma was mm ) using the formula: tumor volume = length (mm) x frozen in liquid nitrogen. Frozen brain and blood samples 2 2 width (mm )/2. Animals were weighed for drug dosage were submitted to the Analytical Facility for Bioactive calculation and randomized to a vehicle or drug treatment Molecules, The Hospital for Sick Children for liquid group. Subcutaneous bolus injection in the dorsal skin fold chromatography-tandem mass spectrometry (LC/MS/MS) of each animal was performed daily for 4 days. AZD1152- analysis to determine AZD1152-HQPA concentration. HQPA was administered at a weight-determined volume Pharmacokinetic parameters were calculated from linear to achieve 25 or 50 mg/kg/day using a 2 or 4 mg per mL regression analysis of Log (plasma concentration) versus of 3.94% v/v DMSO in PBS solution. The vehicle group time graph assuming a one-component model. received a weight-determined volume of PBS with 3.94% DMSO vehicle solution. Tumor volumes were measured Immunohistochemistry every 2 days until day 30 after implantation of tumor cells. Animals were euthanized by carbon dioxide chamber Flank tumors and whole brains were fixed and either at day 5 or day 30 post-implantation and tumors processed for immunohistochemistry as described excised and immediately immersed in 3.7% formaldehyde previously.[4] Tissues were incubated with primary for histologic analysis or frozen in liquid nitrogen for antibody at room temperature for 1 hr. Rabbit polyclonal protein analysis. Formalin-fixed whole tumors were anti-cleaved Caspase 3 antibody (Cell Signaling) 1:500, weighed prior to sectioning and anti-phospho Histone H3 (Ser 10) (Cell Signaling) Intracranial implantation of 2.5 x 10 cells suspended 1:200 were used. TBS-T was used for washes. Detection in 3 µL of PBS was performed using sterile technique was performed with avidin-biotin-horseradish peroxidase under isofluorane gas anaesthetic. Cells were placed into complex (ABC; Vector Laboratories) followed by the right cerebellum at a depth of 2.5 mm via a single diaminobenzidine as the chromogen. Nuclei were burr hole 2 mm lateral to the midline and 1 mm posterior counterstained with hematoxylin. Slides were visualized to the junction of the sagittal and lamdoid sutures. Mice on an Olympus 1X37 light microscope. Cells were counted received 5 mg/kg ketoprofen analgesic and a bolus of in 20X fields to greater than 1000 cells per specimen. 0.5 mL 0.9% saline subcutaneously in the immediate post-operative period. At day 7, after implantation all mRNA expression analysis in medulloblastoma mice were imaged using Luciferin bioluminescence as tumors described previously.[4] Mice with defined posterior fossa tumors were randomized to AZD1152-HQPA or vehicle treatment. AZD1152-HQPA was given at a dose The expression profiles of medulloblastoma tumors, of 50 mg/kg/day using a 4 mg AZD1152-HQPA per mL cell lines, and normal samples were retrieved from GEO of 3.94% v/v DMSO in PBS administered in the dorsal accession GSE21140. The samples were expression- skin fold. Treatment was continued for 21-days and profiled and assigned subgroups as previously described Luciferin bioluminescence imaging was repeated at 7, 15, [20, 51] Pearson correlation was used to determine and 28 days after initiation of therapy. The vehicle group the association between MYC mRNA expression and received a weight-determined volume of PBS with 3.94% AURKA, AURKB, and AURKC mRNA expression in v/v DMSO vehicle solution. The survival endpoints were medulloblastoma tumors. death, weight loss greater than 20%, inability to mobilize, continuous seizure, moribund state. Tumors were excised mRNA expression profiling in cell lines for histological analysis 7 days after initiation of drug therapy and at the time of animal death. Subgrouping of D283, D425, Daoy, Med 8A, ONS-76, RES262, UW228 cells in comparison to 103 Pharmacokinetic experiment and AZD1152- medulloblastomas was performed by NanoString analysis HQPA quantification for selected signature genes differentiating Wnt, SHH, Group 3, and Group 4 tumors as previously described.[23] Mice were injected subcutaneously in the dorsal skin fold with 100 mg/kg AZD1152-HQPA using a solution of 4 mg AZD1152-HQPA per mL of 3.94% v/v DMSO in PBS. At 2 hr, 4 hr, 6 hr, 12 hr, and 24 hr www.impactjournals.com/oncotarget 3371 Oncotarget Rutka JT. Aurora kinase B/C inhibition impairs malignant Statistics glioma growth in vivo. J Neurooncol. 2012; 108(3):349- The PASW Statistics 18 software (SPSS Inc., 5. Lens SM, Voest EE and Medema RH. Shared and separate Chicago, IL) or Prism 6 (GraphPad Software, Inc., functions of polo-like kinases and aurora kinases in cancer. CA, USA) were used for statistical analysis. Data were Nat Rev Cancer. 2010; 10(12):825-841. assessed for normality by the Kolmogorov-Smirnov test. 6. Ditchfield C, Johnson VL, Tighe A, Ellston R, Haworth C, All measures are reported as mean +/- standard error of Johnson T, Mortlock A, Keen N and Taylor SS. Aurora B the mean. Means were compared by independent samples couples chromosome alignment with anaphase by targeting Student’s t-test. One-way ANOVA was used to compare BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol. mean absorbance at varying concentrations of inhibitor for 2003; 161(2):267-280. the MTS cell viability assay with post-hoc LSD testing 7. Tseng SH, Bobola MS, Berger MS and Silber JR. if significant differences were found between groups. For Characterization of paclitaxel (Taxol) sensitivity in human post-hoc LSD testing after ANOVA a P-value of less than glioma- and medulloblastoma-derived cell lines. Neuro- 0.025 was selected as significant. In all other analysis oncology. 1999; 1(2):101-108. a P-value less than 0.05 was selected for significance. Survival analysis was performed using a Kaplan-Meier 8. Goga A, Yang D, Tward AD, Morgan DO and Bishop JM. plot. The Breslow (Generalized Wilcoxon) test was used Inhibition of CDK1 as a potential therapy for tumors over- for inferential analysis with a P-value less than 0.05 expressing MYC. Nature medicine. 2007; 13(7):820-827. selected as significant. 9. Wilkinson RW, Odedra R, Heaton SP, Wedge SR, Keen NJ, Crafter C, Foster JR, Brady MC, Bigley A, Brown E, ACKNOWLEDGEMENTS Byth KF, Barrass NC, Mundt KE, Foote KM, Heron NM, Jung FH, et al. AZD1152, a selective inhibitor of Aurora B Dr. Diaz completed this work as a Canada Vanier kinase, inhibits human tumor xenograft growth by inducing Graduate Scholar at the University of Toronto. apoptosis. Clin Cancer Res. 2007; 13(12):3682-3688. 10. Lee C, Fotovati A, Triscott J, Chen J, Venugopal C, Singhal GRANT SUPPORT A, Dunham C, Kerr JM, Verreault M, Yip S, Wakimoto H, Jones C, Jayanthan A, Narendran A, Singh SK and Dunn SE. Polo-like kinase 1 inhibition kills glioblastoma Canadian Institutes for Health Research (CIHR), multiforme brain tumor cells in part through loss of SOX2 MOP-74610. b.r.a.i.n.child and the Laurie Berman Fund and delays tumor progression in mice. Stem Cells. 2012; for b.r.a.i.n.child and the Laurie Berman Fund for Brain 30(6):1064-1075. Tumour Research. 11. Harris PS, Venkataraman S, Alimova I, Birks DK, Donson AM, Knipstein J, Dubuc A, Taylor MD, Handler MH, REFERENCES Foreman NK and Vibhakar R. Polo-like kinase 1 (PLK1) inhibition suppresses cell growth and enhances radiation 1. Taylor MD, Northcott PA, Korshunov A, Remke M, Cho sensitivity in medulloblastoma cells. BMC Cancer. 2012; YJ, Clifford SC, Eberhart CG, Parsons DW, Rutkowski 12:80. S, Gajjar A, Ellison DW, Lichter P, Gilbertson RJ, 12. Nihal M, Stutz N, Schmit T, Ahmad N and Wood GS. Pomeroy SL, Kool M and Pfister SM. Molecular subgroups Polo-like kinase 1 (Plk1) is expressed by cutaneous T-cell of medulloblastoma: the current consensus. Acta lymphomas (CTCLs), and its downregulation promotes cell neuropathologica. 2012; 123(4):465-472. cycle arrest and apoptosis. Cell Cycle. 2011; 10(8):1303- 2. Northcott PA, Jones DT, Kool M, Robinson GW, Gilbertson RJ, Cho YJ, Pomeroy SL, Korshunov A, Lichter P, Taylor 13. Li Q and Dang CV. c-Myc overexpression uncouples DNA MD and Pfister SM. Medulloblastomics: the end of the replication from mitosis. Molecular and cellular biology. beginning. Nature reviews Cancer. 2012; 12(12):818-834. 1999; 19(8):5339-5351. 3. Kool M, Korshunov A, Remke M, Jones DT, Schlanstein 14. Horiuchi D, Kusdra L, Huskey NE, Chandriani S, Lenburg M, Northcott PA, Cho YJ, Koster J, Schouten-van ME, Gonzalez-Angulo AM, Creasman KJ, Bazarov AV, Meeteren A, van Vuurden D, Clifford SC, Pietsch T, von Smyth JW, Davis SE, Yaswen P, Mills GB, Esserman LJ Bueren AO, Rutkowski S, McCabe M, Collins VP, et al. and Goga A. MYC pathway activation in triple-negative Molecular subgroups of medulloblastoma: an international breast cancer is synthetic lethal with CDK inhibition. J Exp meta-analysis of transcriptome, genetic aberrations, Med. 2012; 209(4):679-696. and clinical data of WNT, SHH, Group 3, and Group 15. den Hollander J, Rimpi S, Doherty JR, Rudelius M, Buck 4 medulloblastomas. Acta neuropathologica. 2012; A, Hoellein A, Kremer M, Graf N, Scheerer M, Hall MA, 123(4):473-484. Goga A, von Bubnoff N, Duyster J, Peschel C, Cleveland 4. Diaz RJ, Golbourn B, Shekarforoush M, Smith CA and www.impactjournals.com/oncotarget 3372 Oncotarget JL, Nilsson JA, et al. Aurora kinases A and B are up- 25. Huang A, Ho CS, Ponzielli R, Barsyte-Lovejoy D, Bouffet regulated by Myc and are essential for maintenance of the E, Picard D, Hawkins CE and Penn LZ. Identification of malignant state. Blood. 2010; 116(9):1498-1505. a novel c-Myc protein interactor, JPO2, with transforming activity in medulloblastoma cells. Cancer research. 2005; 16. Yang D, Liu H, Goga A, Kim S, Yuneva M and Bishop 65(13):5607-5619. JM. Therapeutic potential of a synthetic lethal interaction between the MYC proto-oncogene and inhibition of 26. Zhou L, Picard D, Ra YS, Li M, Northcott PA, Hu Y, aurora-B kinase. Proc Natl Acad Sci U S A. 2010; Stearns D, Hawkins C, Taylor MD, Rutka J, Der SD and 107(31):13836-13841. Huang A. Silencing of thrombospondin-1 is critical for myc-induced metastatic phenotypes in medulloblastoma. 17. von Bueren AO, Shalaby T, Oehler-Janne C, Arnold L, Cancer research. 2010; 70(20):8199-8210. Stearns D, Eberhart CG, Arcaro A, Pruschy M and Grotzer MA. RNA interference-mediated c-MYC inhibition 27. Hsu JY, Sun ZW, Li X, Reuben M, Tatchell K, Bishop prevents cell growth and decreases sensitivity to radio- and DK, Grushcow JM, Brame CJ, Caldwell JA, Hunt DF, Lin chemotherapy in childhood medulloblastoma cells. BMC R, Smith MM and Allis CD. Mitotic phosphorylation of Cancer. 2009; 9:10. histone H3 is governed by Ipl1/aurora kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell. 2000; 18. Northcott PA, Shih DJ, Peacock J, Garzia L, Morrissy 102(3):279-291. AS, Zichner T, Stutz AM, Korshunov A, Reimand J, Schumacher SE, Beroukhim R, Ellison DW, Marshall CR, 28. Giet R and Glover DM. Drosophila aurora B kinase is Lionel AC, Mack S, Dubuc A, et al. Subgroup-specific required for histone H3 phosphorylation and condensin structural variation across 1,000 medulloblastoma genomes. recruitment during chromosome condensation and to Nature. 2012; 488(7409):49-56. organize the central spindle during cytokinesis. J Cell Biol. 2001; 152(4):669-682. 19. Park AK, Lee SJ, Phi JH, Wang KC, Kim DG, Cho BK, Haberler C, Fattet S, Dufour C, Puget S, Sainte-Rose C, 29. El-Sheikh A, Fan R, Birks D, Donson A, Foreman NK Bourdeaut F, Grill J, Delattre O, Kim SK and Park WY. and Vibhakar R. Inhibition of Aurora Kinase A enhances Prognostic classification of pediatric medulloblastoma chemosensitivity of medulloblastoma cell lines. Pediatr based on chromosome 17p loss, expression of MYCC and Blood Cancer. 2010; 55(1):35-41. MYCN, and Wnt pathway activation. Neuro Oncol. 2012; 30. Yasui Y, Urano T, Kawajiri A, Nagata K, Tatsuka M, 14(2):203-214. Saya H, Furukawa K, Takahashi T, Izawa I and Inagaki M. 20. Northcott PA, Korshunov A, Witt H, Hielscher T, Eberhart Autophosphorylation of a newly identified site of Aurora-B CG, Mack S, Bouffet E, Clifford SC, Hawkins CE, French is indispensable for cytokinesis. J Biol Chem. 2004; P, Rutka JT, Pfister S and Taylor MD. Medulloblastoma 279(13):12997-13003. comprises four distinct molecular variants. Journal of 31. Ohashi S, Sakashita G, Ban R, Nagasawa M, Matsuzaki H, clinical oncology : official journal of the American Society Murata Y, Taniguchi H, Shima H, Furukawa K and Urano of Clinical Oncology. 2011; 29(11):1408-1414. T. Phospho-regulation of human protein kinase Aurora-A: 21. Bigner SH, Friedman HS, Vogelstein B, Oakes WJ and analysis using anti-phospho-Thr288 monoclonal antibodies. Bigner DD. Amplification of the c-myc gene in human Oncogene. 2006; 25(59):7691-7702. medulloblastoma cell lines and xenografts. Cancer research. 32. Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE 1990; 50(8):2347-2350. and Poirier GG. Specific proteolytic cleavage of poly(ADP- 22. Friedman HS, Colvin OM, Kaufmann SH, Ludeman SM, ribose) polymerase: an early marker of chemotherapy- Bullock N, Bigner DD and Griffith OW. Cyclophosphamide induced apoptosis. Cancer research. 1993; 53(17):3976- resistance in medulloblastoma. Cancer research. 1992; 3985. 52(19):5373-5378. 33. Liu L, Ulbrich J, Muller J, Wustefeld T, Aeberhard L, Kress 23. Northcott PA, Shih DJ, Remke M, Cho YJ, Kool M, TR, Muthalagu N, Rycak L, Rudalska R, Moll R, Kempa Hawkins C, Eberhart CG, Dubuc A, Guettouche T, S, Zender L, Eilers M and Murphy DJ. Deregulated MYC Cardentey Y, Bouffet E, Pomeroy SL, Marra M, Malkin expression induces dependence upon AMPK-related kinase D, Rutka JT, Korshunov A, et al. Rapid, reliable, 5. Nature. 2012; 483(7391):608-612. and reproducible molecular sub-grouping of clinical 34. Humbert N, Navaratnam N, Augert A, Da Costa M, medulloblastoma samples. Acta neuropathologica. 2012; Martien S, Wang J, Martinez D, Abbadie C, Carling D, de 123(4):615-626. Launoit Y, Gil J and Bernard D. Regulation of ploidy and 24. He XM, Wikstrand CJ, Friedman HS, Bigner SH, Pleasure senescence by the AMPK-related kinase NUAK1. EMBO S, Trojanowski JQ and Bigner DD. Differentiation J. 2010; 29(2):376-386. characteristics of newly established medulloblastoma 35. Yabuta N, Mukai S, Okada N, Aylon Y and Nojima H. The cell lines (D384 Med, D425 Med, and D458 Med) and tumor suppressor Lats2 is pivotal in Aurora A and Aurora their transplantable xenografts. Laboratory investigation; B signaling during mitosis. Cell Cycle. 2011; 10(16):2724- a journal of technical methods and pathology. 1991; 64(6):833-843. www.impactjournals.com/oncotarget 3373 Oncotarget 36. Oke A, Pearce D, Wilkinson RW, Crafter C, Odedra R, 47. Marchetti S, Pluim D, van Eijndhoven M, van Tellingen Cavenagh J, Fitzgibbon J, Lister AT, Joel S and Bonnet O, Mazzanti R, Beijnen JH and Schellens JH. Effect of the D. AZD1152 rapidly and negatively affects the growth and drug transporters ABCG2, Abcg2, ABCB1 and ABCC2 on survival of human acute myeloid leukemia cells in vitro and the disposition, brain accumulation and myelotoxicity of the in vivo. Cancer research. 2009; 69(10):4150-4158. aurora kinase B inhibitor barasertib and its more active form barasertib-hydroxy-QPA. Invest New Drugs. 2013. 37. Kaestner P, Stolz A and Bastians H. Determinants for the efficiency of anticancer drugs targeting either Aurora-A 48. Lowenberg B, Muus P, Ossenkoppele G, Rousselot P, Cahn or Aurora-B kinases in human colon carcinoma cells. JY, Ifrah N, Martinelli G, Amadori S, Berman E, Sonneveld Molecular cancer therapeutics. 2009; 8(7):2046-2056. P, Jongen-Lavrencic M, Rigaudeau S, Stockman P, Goudie A, Faderl S, Jabbour E, et al. Phase 1/2 study to assess 38. Li M, Jung A, Ganswindt U, Marini P, Friedl A, Daniel the safety, efficacy, and pharmacokinetics of barasertib PT, Lauber K, Jendrossek V and Belka C. Aurora kinase (AZD1152) in patients with advanced acute myeloid inhibitor ZM447439 induces apoptosis via mitochondrial leukemia. Blood. 2011; 118(23):6030-6036. pathways. Biochem Pharmacol. 2010; 79(2):122-129. 49. Pei Y, Moore CE, Wang J, Tewari AK, Eroshkin A, Cho 39. Ikegami S, Taguchi T, Ohashi M, Oguro M, Nagano H YJ, Witt H, Korshunov A, Read TA, Sun JL, Schmitt and Mano Y. Aphidicolin prevents mitotic cell division EM, Miller CR, Buckley AF, McLendon RE, Westbrook by interfering with the activity of DNA polymerase-alpha. TF, Northcott PA, et al. An animal model of MYC-driven Nature. 1978; 275(5679):458-460. medulloblastoma. Cancer Cell. 2012; 21(2):155-167. 40. Sasai K, Katayama H, Stenoien DL, Fujii S, Honda 50. Kawauchi D, Robinson G, Uziel T, Gibson P, Rehg J, Gao R, Kimura M, Okano Y, Tatsuka M, Suzuki F, Nigg C, Finkelstein D, Qu C, Pounds S, Ellison DW, Gilbertson EA, Earnshaw WC, Brinkley WR and Sen S. Aurora-C RJ and Roussel MF. A mouse model of the most aggressive kinase is a novel chromosomal passenger protein that can subgroup of human medulloblastoma. Cancer Cell. 2012; complement Aurora-B kinase function in mitotic cells. Cell 21(2):168-180. Motil Cytoskeleton. 2004; 59(4):249-263. 51. Northcott PA, Fernandez LA, Hagan JP, Ellison DW, 41. Klesse LJ and Bowers DC. Childhood medulloblastoma: Grajkowska W, Gillespie Y, Grundy R, Van Meter T, current status of biology and treatment. CNS Drugs. 2010; Rutka JT, Croce CM, Kenney AM and Taylor MD. The 24(4):285-301. miR-17/92 polycistron is up-regulated in sonic hedgehog- 42. Yang J, Ikezoe T, Nishioka C, Tasaka T, Taniguchi A, driven medulloblastomas and induced by N-myc in sonic Kuwayama Y, Komatsu N, Bandobashi K, Togitani K, hedgehog-treated cerebellar neural precursors. Cancer Koeffler HP, Taguchi H and Yokoyama A. AZD1152, a research. 2009; 69(8):3249-3255. novel and selective aurora B kinase inhibitor, induces growth arrest, apoptosis, and sensitization for tubulin depolymerizing agent or topoisomerase II inhibitor in human acute leukemia cells in vitro and in vivo. Blood. 2007; 110(6):2034-2040. 43. Hayashi MT, Cesare AJ, Fitzpatrick JA, Lazzerini-Denchi E and Karlseder J. A telomere-dependent DNA damage checkpoint induced by prolonged mitotic arrest. Nature structural & molecular biology. 2012; 19(4):387-394. 44. Li J, Anderson MG, Tucker LA, Shen Y, Glaser KB and Shah OJ. Inhibition of Aurora B kinase sensitizes a subset of human glioma cells to TRAIL concomitant with induction of TRAIL-R2. Cell Death Differ. 2009; 16(3):498-511. 45. Shah OJ, Lin X, Li L, Huang X, Li J, Anderson MG, Tang H, Rodriguez LE, Warder SE, McLoughlin S, Chen J, Palma J, Glaser KB, Donawho CK, Fesik SW and Shen Y. Bcl-XL represents a druggable molecular vulnerability during aurora B inhibitor-mediated polyploidization. Proceedings of the National Academy of Sciences of the United States of America. 2010; 107(28):12634-12639. 46. Nie Z, Hu G, Wei G, Cui K, Yamane A, Resch W, Wang R, Green DR, Tessarollo L, Casellas R, Zhao K and Levens D. c-Myc is a universal amplifier of expressed genes in lymphocytes and embryonic stem cells. Cell. 2012; 151(1):68-79. www.impactjournals.com/oncotarget 3374 Oncotarget

Journal

OncotargetPubmed Central

Published: Dec 31, 2014

References