Pharmacologic Abrogation of the Mitotic Spindle Checkpoint by an Indolocarbazole Discovered by Cellular Screening Efficiently Kills Cancer Cells

Published OnlineFirst April 14, 2009; DOI: 10.1158/0008-5472.CAN-08-3597

Research Article

Ailine Stolz,1 Celia Vogel,1 Verena Schneider,1 Norman Ertych,1 Anne Kienitz,1 Hongtao Yu,2 and Holger Bastians1

1Institute for Molecular Biology and Tumor Research, Philipps University Marburg, Marburg, Germany and 2Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, Texas

Abstract separate pathways, which control the activation of caspases and The mitotic spindle checkpoint represents a signal transduc- a slow but continuous degradation of cyclin B during mitosis, tion pathway that prevents the onset of anaphase until all respectively (5, 6). The drug-induced activation of caspases during chromosomes are properly aligned on a metaphase plate. mitosis might be a key event required for efficient tumor cell killing Partial inactivation of this checkpoint allows premature (6, 7), but the mechanisms leading to caspase activation remain separation of sister chromatids and results in aneuploidy, elusive. Interestingly, the mitotic spindle checkpoint is required for which might contribute to tumorigenesis. Unlike other cell the efficient induction of apoptosis on antimitotic drug treatment, cycle checkpoints, the spindle checkpoint is essential for cell regardless of whether a cell dies in mitosis or in the subsequent G1 viability, giving rise to the idea that the spindle checkpoint phase. Thus, compromising the mitotic spindle checkpoint results itself might represent a valuable target for anticancer therapy. in a significant decrease in sensitivity of tumor cells toward We used a cell-based screen and identified the indolocarbazole antimitotic drugs (8–12). These findings were recently supported compound Go¨6976 as a pharmacologic inhibitor of the spindle by an unbiased small interfering RNA (siRNA) screen that identified checkpoint. Go¨6976 potently overrides a spindle checkpoint– spindle checkpoint proteins as major determinants for the mediated mitotic arrest by abrogating the phosphorylation sensitivity to paclitaxel (13). The molecular basis for a requirement and kinetochore localization of several spindle checkpoint of the spindle checkpoint in the activation of the apoptotic proteins. We identified the Aurora-A and Aurora-B kinases, machinery is unclear, but it might be related to inhibition of Akt which have been previously implicated in proper mitotic during a mitotic arrest (13) or to a dephosphorylation and progression and spindle checkpoint function, as targets for activation of caspase-9 that requires a prolonged duration of Go¨6976. Accordingly, Go¨6976treatment causes severe mitotic mitosis (14). abnormalities and chromosome alignment defects, which are The spindle checkpoint pathway requires the function of various not properly detected by the drug-inactivated spindle check- proteins, including the kinases Bub1, BubR1, Mps1, Mad1, Mad2, point. This results in an aberrant progression of mitosis, Bub3, and several others, which are recruited to kinetochores on leading to apoptosis in various human cancer cell lines, activation of the checkpoint (2, 3). Furthermore, inhibition of the including spindle checkpoint–compromised cancer cells. centromeric Aurora-B kinase or overexpression of the centrosomal Thus, our work describes a novel and promising strategy for Aurora-A kinase led to spindle checkpoint malfunction (15–17). anticancer treatment that targets the mitotic spindle check- Activation of the spindle checkpoint in response to misaligned point. [Cancer Res 2009;69(9):3874–83] chromosomes either in the early phases of a regular mitosis or chronically on antimitotic drug treatment causes the inhibition of the ubiquitin ligase activity within the anaphase-promoting Introduction complex/cyclosome (APC/C), which is required for the degradation of various mitotic substrates, including securin and cyclin B. Antimitotic drugs, including various taxanes and Vinca alkaloids, Accordingly, failure of the spindle checkpoint results in a are frequently used for anticancer treatment (1). It is well premature separation of sister chromatids and an unscheduled established that these drugs activate the spindle assembly exit from mitosis, which leads to the generation of aneuploid checkpoint, a signal transduction pathway that monitors the progenitors (2, 3). proper alignment of chromosomes during mitosis (2, 3). The drug- In human cancer, a partial dysfunction of the spindle check- induced and spindle checkpoint–mediated mitotic arrest is point (15, 18–22) is a frequent event and might contribute to followed by the induction of apoptosis, which occurs either in tumorigenesis (18, 23–28). In fact, mice heterozygous and mitosis or in the postmitotic G phase on escape from a prolonged 1 hypomorphic for spindle checkpoint genes show gross aneuploidy mitotic arrest (4). The latter decision might depend on two and carcinogen-induced or spontaneous tumorigenesis (20, 29). Importantly, the analyses of homozygous knockout mice as well as Note: Supplementary data for this article are available at Cancer Research Online siRNA studies in human cells revealed that the spindle checkpoint (http://cancerres.aacrjournals.org/). is essential for cell viability (30–33). The essential nature of the A. Stolz and C. Vogel contributed equally to this work. Requests for reprints: Holger Bastians, Institute for Molecular Biology and Tumor spindle checkpoint raises the interesting idea to use the spindle Research, Philipps University Marburg, Emil-Mannkopff-Strasse 2, D-35037 Marburg, checkpoint as a novel target for anticancer therapy. Germany. Phone: 49-6421-2863113; Fax: 49-6421-2865932; E-mail: [email protected] To test this concept, we performed a cell-based screen and marburg.de. I2009 American Association for Cancer Research. identified the indolocarbazole Go¨6976 that inhibits the mitotic doi:10.1158/0008-5472.CAN-08-3597 spindle checkpoint by restraining the phosphorylation and

Cancer Res 2009; 69: (9). May 1, 2009 3874 www.aacrjournals.org

Spindle Checkpoint Inhibition Induces Apoptosis

Figure 1. Go¨6976 overrides a spindle damage–induced mitotic arrest. A, a cell-based screen performed to identify pharmacologic spindle checkpoint inhibitors. B, Go¨6976 is an inhibitor of spindle damage–induced mitotic arrest. HCT116 cells were arrested in mitosis with 300 nmol/L nocodazole for 14 h and subsequently treated with various protein kinase inhibitors. After 2 h of incubation, the mitotic index was determined. C, representative examples of fluorescence-activated cell sorting (FACS) profiles of HCT116 cells treated with nocodazole alone or followed by Go¨6976 treatment as in B. D, dose-response curves for Go¨6976 and roscovitine after treatment with nocodazole or Taxol according to B. Points, mean of at least three independent experiments; bars, SD.

kinetochore localization of several spindle checkpoint proteins. 200 Amol/L ALLN (Calbiochem), 20 Amol/L MG132 (Calbiochem), 5 to Interestingly, we identified the mitotic Aurora-A and Aurora-B 100 Amol/L roscovitine (Sigma), 0.2 to 20 Amol/L Go¨6976 (Calbiochem), 0.05 A A A kinases as targets for Go¨6976 in vitro and in vivo and show that the to 2 mol/L ZM447439 (Tocris), 50 mol/L H89 (Calbiochem), 100 mol/L A pharmacologic inactivation of the spindle checkpoint efficiently protein kinase A (PKA) inhibitor (Calbiochem), 1 mol/L Ro31-8220 (Calbiochem), 1 Amol/L Ro32-0432 (Calbiochem), 40 Amol/L rottlerin kills human cancer cells. (Calbiochem), 50 Amol/L PD98059 (Calbiochem), 10 Amol/L U0126 (Calbiochem), 30 Amol/L SB202190 (Calbiochem), 30 Amol/L SB203580 (Calbiochem), 50 nmol/L staurosporine (Sigma), 20 Amol/L LY294002 Materials and Methods (Sigma), 0.1 to 2 Amol/L UCN-01 (a gift from the Developmental Cell culture and treatments. All cell lines were maintained under Therapeutics Program of the National Cancer Institute), and 10 Amol/L standard culture conditions (11) and treated with 300 nmol/L nocodazole Raf inhibitor (BAY 43-9006). Cells were synchronized in G1-S by a double- (Sigma), 100 nmol/L Taxol (Sigma), or 68 Amol/L monastrol (Biomol), thymidine block using 2 mmol/L thymidine (Sigma). www.aacrjournals.org 3875 Cancer Res 2009; 69: (9). May 1, 2009

Cancer Research

Flow cytometry and determination of the mitotic index. The DNA Microscopy. Cells were fixed/permeabilized in 2% paraformaldehyde/ content and the mitotic index were determined as described (11). methanol and incubation with antibodies was carried out for 2 h. Images were Western blotting. Cell lysates were prepared, and SDS-PAGE, semidry taken with a Leica DM6000B microscope and a charge-coupled device Western blotting, and detection were performed as described (11) using camera (Orca-ER, Hamamatsu) with a Z-optical spacing of 0.2 Am. Images the following antibodies: Bub1 (Bethyl; Stephen Taylor, University of were deconvolved using the Leica LAS-AF software and maximum projection Manchester, Manchester, United Kingdom), BubR1 (Chemicon; Stephen images are shown. Pixel quantitations of CREST and Bub1/BubR1 signals Taylor), Bub3 (BDBiosciences), Mps1 (Santa Cruz Biotechnology), Plk1 on kinetochores were performed using the Leica LAS-AF software. (Santa Cruz Biotechnology), Aurora-A (Santa Cruz Biotechnology; pT288: Kinase assays. In vitro kinase assays using recombinant Aurora kinases Cell Signaling), Aurora-B (BDBiosciences; pT232: Cell Signaling), cyclin B (34) were performed for 30 min at 30jC in kinase buffer [50 mmol/L Tris-

(Santa Cruz Biotechnology), tubulin (Sigma), actin (Sigma), securin HCl (pH 7.7), 100 mmol/L KCl, 5 mmol/L MgCl2, 1 mmol/L DTT] in the (NeoMarkers), histone H3 (pS-10; Cell Signaling), and poly(ADP-ribose) presence of 50 to 2,000 Amol/L of ATP (Sigma), 1 ACi [32P]ATP (6,000 Ci/ polymerase (PARP; Pharmingen). Signals were quantitated using the mmol; Amersham), and 7 Ag MBP (Sigma). Substrate phosphorylation was ImageJ software (NIH). visualized by autoradiography and quantitated using a phosphoimager Chromosome spread analysis. Premature sister chromatid separation (Fuji). For immunoprecipitation kinase assays, Aurora-A or Aurora-B was and karyotype analysis were performed as described (11). immunoprecipitated from lysates and subjected to in vitro kinase assays.

Figure 2. Go¨6976 is an inhibitor of the mitotic spindle checkpoint. A, Go¨6976 induces mitotic exit and proteasome-dependent proteolysis of cyclin B and securin. HCT116 cells were treated with nocodazole (N) or Taxol (T) and incubated with DMSO (D)or2Amol/L Go¨6976 (G) in the presence or absence of 200 Amol/L ALLN (A). The mitotic index was determined (left) and the protein levels of cyclin B and securin were determined on Western blot (right). B, Go¨6976 induces premature sister chromatid separation. Nocodazole-arrested HCT116 cells were treated with DMSO or Go¨6976 for 10 and 15 min and metaphase spreads were prepared. Examples of typical metaphase spreads and the quantitation of cells showing premature sister chromatid separations (n = 400) are shown. C, Go¨6976 reduces kinetochore localization of Bub1 and BubR1. HCT116 cells were treated with nocodazole to arrest cells in mitosis followed by incubation with 20 Amol/L MG132 and 2 Amol/L Go¨6976 for 2 h. Cells were fixed and immunofluorescence staining was performed for Bub1, BubR1, and CRESTand chromosomes were stained with Hoechst dye. Pixel intensities for Bub1 and BubR1 at kinetochores were quantified relative to the CRESTsignals. The graphs represent quantitat ions from 80 kinetochores from six cells. D, Go¨6976 inhibits phosphorylation of Bub1, Mps1, Aurora-A, and Aurora-B. Cells were treated as in C in the presence or absence of ALLN and the indicated mitotic proteins were detected on Western blots. For each cell population, the mitotic index was determined.

Cancer Res 2009; 69: (9). May 1, 2009 3876 www.aacrjournals.org

Spindle Checkpoint Inhibition Induces Apoptosis

Figure 3. Go¨6976 causes chromosome alignment defects and premature and accelerated exit from mitosis. A, asynchronously growing HCT116 cells were treated with 2 Amol/L Go¨6976 for the indicated times and the DNA content and the mitotic index were determined. B, cells were treated with 2 Amol/L Go¨6976 for 6 h and fixed cells were stained for a-tubulin, CREST, and DNA. Typical examples of maximal projections of deconvolved Z-stacks are shown. C, HCT116 cells were synchronized in mitosis by a nocodazole block and released into medium with or without 2 Amol/L Go¨6976. At the indicated time points, cells were fixed and stained for a-tubulin and DNA. D, cells were treated as in C and the DNA content and the mitotic index were determined. Representative FACS profiles are shown. Points, mean of at least three independent experiments; bars, SD.

Apoptosis assays. Caspase activity assays, detection of PARP cleavage, activated protein/extracellular signal-regulated kinase kinase, and chromosomal DNA fragmentation assays as well as determination of p38, Raf, and phosphatidylinositol 3-kinase (Fig. 1B). Among the sub-G1 DNA content were performed as described previously (11). For the small-molecule inhibitors tested, we identified the indolo- caspase inhibition, the pan-caspase inhibitor Boc-D-FMK (100 Amol/L; carbazole Go¨6976 (Supplementary Fig. S1A) as a compound that Calbiochem) was used. potently abrogated the induced mitotic arrest and forced cells to exit mitosis with a 4N DNA content (Fig. 1B and C). In HCT116 Results and HeLa cells, Go¨6976 showed a half-maximal effect at 0.8 to A Identification of the indolocarbazole compound Go¨6976 1.2 mol/L in the presence of nocodazole and Taxol, which is f as a potent abrogator of a spindle damage–induced mitotic 25 times more potent than the CDK1 inhibitor roscovitine in arrest. We performed a cell-based screen, in which HCT116 cells abrogating the mitotic arrest (Fig. 1D). Very similar results were were first treated with nocodazole to arrest cells in mitosis in a obtained after treatment with the KSP/Eg5 inhibitor monastrol spindle checkpoint–dependent manner. Subsequently, cells were and in various human cell lines (Supplementary Fig. S1B and C). treated with small-molecule inhibitors followed by determination We also tested established structurally related and similarly of the mitotic index to monitor the mitotic arrest (Fig. 1A). potent PKC inhibitors (Ro31-8220 and Ro32-0432) as well as the Because several kinases are involved in spindle checkpoint structurally unrelated PKC inhibitor rottlerin and found no effect signaling, we screened commercially available protein kinase on the spindle damage–induced mitotic arrest, suggesting that 1inhibitors, including established inhibitors for PKA, protein Go¨6976 acts independently of PKC inhibition. In addition, other kinase C (PKC), cyclin-dependent kinase (CDK), Chk1, mitogen- indolocarbazole compounds, including UCN-01 or staurosporine, www.aacrjournals.org 3877 Cancer Res 2009; 69: (9). May 1, 2009

Cancer Research had no or little effects on the spindle damage–induced mitotic DNA content was reformed, whereas CDK inhibitors caused a high arrest (Fig. 1B). No significant inhibition of CDK1 by Go¨6976 was rate of multinucleation (Supplementary Fig. S2C), supporting the observed (Supplementary Fig. S1D). notion that Go¨6976 acts independently from CDK1. Go¨6976 inhibits the mitotic spindle checkpoint. In the Consistent with a spindle checkpoint inhibition by Go¨6976, we presence of nocodazole or Taxol, treatment with 2 Amol/L found a high rate of premature chromatid separation on Go¨6976 Go¨6976 induced a rapid exit from mitosis, which was associated treatment (Fig. 2B). Prolonged treatment led to a significant increase with the rapid degradation of cyclin B and securin that could be in aneuploid cells, whereas the overall cell cycle progression was not blocked by the proteasome inhibitor ALLN, indicating that Go¨6976 significantly disturbed (Supplementary Fig. S2D and E), although relieves the inhibition of the APC/C (Fig. 2A). Go¨6976 allowed the dead cells became increasingly apparent. entry into mitosis but suppressed the accumulation of mitotic Go¨6976 inhibits the proper phosphorylation and localiza- cells in the presence of nocodazole, whereas roscovitine treatment tion of spindle checkpoint proteins. We investigated whether prevented the entry into mitosis (Supplementary Fig. S2A). Go¨6976 interferes with the localization or amount and Furthermore, premature exit from mitosis led to endoreduplication phosphorylation of spindle checkpoint proteins. Significantly, (Supplementary Fig. S2B), which is associated with spindle Go¨6976 treatment reduced the kinetochore localization of Bub1 checkpoint impairment as shown previously (35). Go¨6976 treat- and BubR1 by f80% (Fig. 2C). Aurora-B was also found to be ment forced cells out of mitosis and a single nucleus with a 4N mislocalized from the centromere region to chromosome arms

Figure 4. Go¨6976 inhibits the Aurora-A and Aurora-B kinases in vitro and in vivo. A, in vitro kinase assays using the indicated purified recombinant kinases in the presence of different concentrations of Go¨6976. Representative autoradiographs are shown. B, the in vitro activity of purified recombinant Aurora-A and Aurora-B was determined in the presence of different concentrations of Go¨6976 or ZM447439 and the IC50 values were calculated in the presence of 50 Amol/L ATP. C, detection of phosphorylated Aurora-A and Aurora-B kinases on Western blots using lysates from mitotic HCT116 cells treated with nocodazole for 14 h followed by incubation with MG132 in the presence or absence of 2 Amol/L Go¨6976 or 2 Amol/L ZM447439 for 2 h. D, Aurora-A and Aurora-B kinases were immunopurified from mitotic HeLa cells treated with nocodazole, ALLN, and 1 or 2 Amol/L of Go¨6976 and their kinase activities were determined. Columns, mean of at least three independent experiments; bars, SD.

Cancer Res 2009; 69: (9). May 1, 2009 3878 www.aacrjournals.org

Spindle Checkpoint Inhibition Induces Apoptosis

Figure 5. The aberrant progression through mitosis induced by Go¨6976 causes apoptosis in human cancer cells. A, Go¨6976-induced apoptosis requires progression through mitosis. HCT116 cells were synchronized at the G1-S transition by a double-thymidine block and released into fresh medium. Seven hours after release, cells were treated with various concentrations of Go¨6976, roscovitine, or roscovitine plus Go¨6976, and another 7 h later, caspase-3 activity was determined. B, Go¨6976-induced apoptosis is dependent on caspase activity. HCT116 cells were treated with 20 Amol/L Go¨6976 in the presence or absence of 100 Amol/L Boc-D-FMK for 24 and 48 h and the DNA content and proportion of apoptotic cells were determined. Representative FACS profiles are shown. C, Go¨6976 induces caspase activity in various human cancer cell lines. The indicated human cancer cell lines and two nontransformed cell lines were treated with 15 Amol/L Go¨6976 for 24 h and caspase-3 activity was measured in intact cells. D, synergistic effect of Taxol and Go¨6976. HCT116 cells were treated with 100 nmol/L Taxol for 14 h followed by the addition of DMSO or 2 Amol/L Go¨6976 for additional 2 h. The DNA content was determined by FACS analyses and representative DNA profiles and quantitation of apoptotic cells are shown. E, quantitative analysis of cell death on combined treatment with 2 Amol/L Go¨6976 and 100 nmol/L Taxol for 38 h. Columns, mean of at least three independent experiments; bars, SD.

on Go¨6976 treatment (Supplementary Fig. S3), whereas Aurora-A were frequently found near the spindle poles, suggesting that retained its centrosomal localization (data not shown). Further- those chromosomes are not properly attached to microtubules. more, on Western blots, we found that the hyperphosphorylation Similar chromosome alignment defects were observed when cells of Bub1, Mps1, Aurora-A, and Aurora-B, which is closely were released from a monastrol block into Go¨6976 (Supplemen- associated with spindle checkpoint activation, was abolished tary Fig. S4A). on Go¨6976 treatment both on forced exit from mitosis and in Remarkably, the severe alignment defects induced by Go¨6976 did mitotic cells (Fig. 2D). not cause a robust mitotic arrest as would be expected in the Go¨6976 causes aberrant spindle morphology and chromo- presence of a functional spindle checkpoint. Instead, the accumu- some alignment defects. Next, we investigated the mitotic lation of mitotic cells was only weak and transient (Fig. 3A) but progression in the presence of Go¨6976. Treatment of asynchro- dependent on the spindle checkpoint (Supplementary Fig. S4B), nously growing HCT116 cells with 2 Amol/L Go¨6976 resulted in suggesting that the mitotic arrest cannot be maintained, possibly a transient accumulation of mitotic cells (Fig. 3A) and this was mediated by Go¨6976-induced spindle checkpoint inhibition. associated with profound alterations in the spindle structure and Go¨6976 accelerates exit from an aberrant mitosis. To test severely misaligned chromosomes (Fig. 3B). The spindles often whether Go¨6976 treatment indeed allows an exit from mitosis in seemed cylindrical and flattened and groups of chromosomes the presence of unaligned chromosomes, we released cells from a www.aacrjournals.org 3879 Cancer Res 2009; 69: (9). May 1, 2009

Cancer Research nocodazole block in the presence or absence of Go¨6976 and inhibition of Bub1, Mps1, Plk1, and p38 by Go¨6976 was found. followed the progression through mitosis qualitatively by immu- However, Aurora-B kinase activity was significantly inhibited by nofluorescence (Fig. 3C) and quantitatively by monitoring the Go¨6976 in these assays (Fig. 4A). We determined the IC50 values for mitotic index (Fig. 3D). Whereas control cells progressed normally Go¨6976 and ZM447439, which is a potent and well-characterized through mitosis and exited mitosis properly, Go¨6976-treated cells inhibitor for Aurora-A and Aurora-B kinases in vitro (36), and exhibited a significant acceleration of mitotic progression and found that Go¨6976 inhibited Aurora-B with similar potency as exited from mitosis in the presence of misaligned chromosomes. ZM447439 (100 versus 80 nmol/L at 50 Amol/L ATP; Fig. 4B). Most Go¨6976-treated cells underwent cytokinesis despite the However, our analyses revealed that Go¨6976 was nearly five times presence of unaligned and missegregated chromosomes, producing more potent than ZM447439 in inhibiting the Aurora-A kinase (120 daughter cells with highly unequal DNA content (Fig. 3D, top). By versus 580 nmol/L at 50 Amol/L ATP; Fig. 4B). As expected for the end of mitosis, apoptotic cells became readily apparent, an ATP-competitive inhibitor, the inhibitory activity of suggesting that the severe missegregation observed causes cell Go¨6976 decreased with higher ATP concentrations (Supplementary death (Fig. 3C). Fig. S5A; Supplementary Text). To test whether Go¨6976 also inhibits The mitotic Aurora-A and Aurora-B kinases are inhibited by the Aurora kinases in vivo, we evaluated the activation loop– Go¨6976 in vitro and in vivo. To gain insights into the mitotic specific phosphorylation of Aurora-A and Aurora-B kinases (pT288 target kinase of Go¨6976, we evaluated candidate mitotic kinases for Aurora-A and pT232 for Aurora-B), which is indicative of their known to be involved in spindle checkpoint function. Because we activation. In agreement with previous data (36), treatment of observed a lack of phosphorylation of Bub1, Mps1, Aurora-A, mitotic cells with ZM447439 abolished selectively the activation of and Aurora-B (Fig. 2D), which might relate to their functional cellular Aurora-B (75% inhibition) but left the activity of Aurora-A inactivation, we evaluated these kinases in particular. In vitro,no unchanged. However, in agreement with our in vitro data, Go¨6976

Figure 6. The role of the spindle checkpoint for the efficacy of Go¨6976. A, a spindle damage–induced mitotic arrest is not overridden by Go¨6976 in HT29 cells. HCT116 and HT29 cells were arrested in mitosis by treatment with 150 nmol/L nocodazole for 16 h followed by the addition of different concentrations of Go¨6976. After 2 h of incubation, the mitotic index was determined. B, reduced caspase activation in HT29 cells on Go¨6976 treatment. HCT116 and HT29 cells were synchronized at G1-S and released into the cell cycle. Seven hours after release, cells were treated with various concentrations of Go¨6976 and the activity of caspase-3 was determined after 7 h of treatment. C, spindle checkpoint–compromised cancer cells show reduced caspase activation on spindle damage. HCT116 and isogenic derivatives expressing lower levels of Mad2 (HCT116-MAD2 +/À) or Mad1 (HCT-MAD1-kd) were treated with 150 nmol/L nocodazole for 38 h and caspase-3 activity was determined in intact cells. D, Go¨6976 reduces the already lowered mitotic arrest in spindle checkpoint–compromised cancer cells on spindle damage. Spindle checkpoint–compromised and parental HCT116 cells were treated with 150 nmol/L nocodazole in the presence or absence of 2 Amol/L Go¨6976 and the mitotic index was determined at various time points. E, activation of caspase-3 in spindle checkpoint–compromised cancer cells in response to Go¨6976 treatment. The indicated HCT116 cell lines were treated with different Go¨6976 concentrations and the induction of caspase-3 activity was determined. Columns, mean of at least three independent experiments; bars, SD.

Cancer Res 2009; 69: (9). May 1, 2009 3880 www.aacrjournals.org

Spindle Checkpoint Inhibition Induces Apoptosis treatment inhibited the activity of both Aurora-A and Aurora-B, strong and prolonged mitotic arrest in response to spindle damage with similar potency in vivo (60% and 58% inhibition, respectively; compared with most cancer cell lines, including HCT116 cells Fig. 4C; Supplementary Fig. S5B). (Supplementary Fig. S8). Accordingly, Go¨6976 could only weakly We also determined the activity of Aurora-A and Aurora-B override the spindle checkpoint in HT29 cells (Fig. 6A), which was immunoprecipitated from drug-treated cells. Again, Go¨6976 directly associated with resistance toward Go¨6976 (Fig. 6B), treatment caused significant inhibition of Aurora-A and Aurora-B suggesting that the spindle checkpoint override is directly (Fig. 4D), whereas immunoprecipitated CDK1-cyclin B complexes associated with the sensitivity toward Go¨6976. It is important to were not inhibited by Go¨6976 (Supplementary Fig. S5C). These note that HT29 cells are not resistant per se because they are results provide strong evidence that Go¨6976 is a potent inhibitor of capable of inducing apoptosis on chemotherapeutic treatment Aurora-A and Aurora-B in vitro and in vivo, although we cannot (data not shown). exclude the possibility that Go¨6976 targets additional kinases Targeting of spindle checkpoint–compromised cancer cells during mitosis. by Go¨6976. Several reports have indicated that the spindle In addition to its crucial function in chromosome alignment checkpoint is required for the efficient induction of cell death after and spindle checkpoint function, Aurora-B is also required for treatment with antimitotic drugs (8, 9, 11, 12, 39). Accordingly, spindle cytokinesis (16). Indeed, treatment of asynchronously growing cells checkpoint–compromised HCT116 cells with lowered expression of with ZM447439 prevents cytokinesis and results in profound MAD1 or MAD2 showed a lowered mitotic arrest and reduced polyploidization. Surprisingly, although Go¨6976 inhibits Aurora-B caspase-3 activity in response to nocodazole treatment (Fig. 6C; in vivo, neither profound defects in cytokinesis nor polyploidization ref. 11). In both spindle checkpoint–impaired cell lines, Go¨6976 was detected (Supplementary Fig. S6). treatment further lowered the mitotic arrest in response to spindle The aberrant progression through mitosis induced by damage (Fig. 6D) and induced caspase-3 activity similar to parental Go¨6976 causes apoptosis in tumor cells. To test whether the cells (Fig. 6E), suggesting that spindle checkpoint–compromised Go¨6976-induced aberrant progression of mitosis causes apoptosis, cancer cells are sensitive toward Go¨6976 treatment. we synchronized HCT116 cells at G1-S, treated the cells with Go¨6976 before entry into mitosis, and evaluated the activation of caspase-3 in intact cells. Indeed, Go¨6976 treatment induced Discussion caspase-3 activity, PARP cleavage, and DNA fragmentation after Due to its essential nature, the spindle checkpoint has been progression through a single mitosis and this was significantly suggested as a potential novel anticancer drug target (32, 33). We suppressed when entry into mitosis was prevented by inhibiting wished to prove this conceptual idea and have identified the CDK1 (Fig. 5A; Supplementary Fig. S7A). Further, Go¨6976-induced indolocarbazole Go¨6976 as an inhibitor of the spindle checkpoint. apoptosis was caspase dependent (Fig. 5B). Importantly, we show that this inhibitor induces a highly aberrant We treated 15 human cell lines with Go¨6976 and found a and accelerated mitosis that results in efficient killing of human significant but varying induction of caspase activity in most cancer cancer cells. Thus, we propose to add the spindle checkpoint to the cell lines (Fig. 5C), whereas nontransformed human cells (human growing list of attractive mitotic drug targets. BJ-hTert fibroblasts and primary human umbilical vein endothelial The ‘‘classic’’ mitotic drug target is the mitotic spindle, whose cells) exhibited low caspase-3 activity. The reason for the different function is inhibited by microtubule-targeting drugs including sensitivities observed is not known at present but could be various taxanes or Vinca alkaloids (40). However, due to the recapitulated in colony formation assays comparing the sensitivity ability to inhibit microtubule function in interphase or even in of HeLa and BJ-hTert fibroblasts toward Go¨6976 (Supplementary differentiated cells, a plethora of unwanted side effects, including Fig. S7B; Supplementary Text). Accordingly, we found that HeLa neutropenia and neuropathy, is observed after treatment with and HCT116 cancer cells showed a higher rate of cell killing than these drugs. Therefore, novel mitotic drug targets that function BJ-hTert fibroblasts after treatment with Go¨6976 (Supplementary exclusively in mitosis, including the Plk1 and the Aurora kinases, Fig. S7C and D). We also compared the efficacy of Go¨6976 with but also mitotic Eg5/KSP or Cenp-E kinesins, have gained ZM447439 and Taxol and found a higher efficacy of Taxol but a considerable interest (1, 41). Similarly, no function for the mitotic lower efficacy of ZM447439 toward HeLa and HCT116 cells spindle checkpoint pathway outside of mitosis has been described (Supplementary Fig. S7D and E). thus far. Thus, inhibition of the spindle checkpoint is expected to Go¨6976 acts synergistically with Taxol. Recent data showed have no effect on interphase or differentiated cells. However, it is that the efficacy of antimitotic drugs could be greatly enhanced important to note that targeting mitosis per se might harbor the by subsequent inactivation of the spindle checkpoint (7, 37). To risk of chromosome missegregation in surviving cells that could investigate whether Taxol and a Go¨6976-mediated inactivation of contribute to de novo tumorigenesis. the spindle checkpoint act synergistically, cells were treated for At present, is not clear how antimitotic drugs induce tumor 14 hours with Taxol to arrest cells in mitosis followed by treatment cell death. Most recent results have indicated that cells activate with Go¨6976 to induce exit from mitosis. In fact, a synergistic effect caspase-dependent pathways during mitosis and die either in of Go¨6976 and Taxol was clearly observed (Fig. 5D and E). It is mitosis or on an unscheduled exit from mitosis (5–7). However, important to note that a reversed order of drug treatment caused it is unknown how caspases are activated on drug treatment, an inhibition of the spindle checkpoint before Taxol treatment and but the activation of the spindle checkpoint associated with a reduced sensitivity toward Taxol (data not shown; Supplementary transient mitotic delay seems to be necessary for the efficient Fig. S2B; refs. 11, 35, 38). induction of cell death (8–13). Interestingly, during mitosis, A spindle checkpoint override is required for efficacy of caspase-9 is inhibited by CDK1/cyclin B–mediated phosphoryla- Go¨6976. To investigate if the efficacy of Go¨6976 requires the tion. Prolonged mitotic arrest, which requires the spindle override of the mitotic spindle checkpoint, we took advantage of a checkpoint, may lead to a progressive dephosphorylation, human colon carcinoma cell line, HT29, which exhibits a very activation of caspase-9, and the induction of apoptosis, thereby www.aacrjournals.org 3881 Cancer Res 2009; 69: (9). May 1, 2009

Cancer Research providing a timer mechanism monitoring the length of mitosis selective Aurora-B inhibitors, including ZM447439 or hesperadin (14). In contrast to this mechanism, we show here when (36, 43). Whereas ZM447439 and hesperadin inhibit selectively a targeting the spindle checkpoint pathway that the induction of Taxol-induced spindle checkpoint, Go¨6976 overrides the spindle apoptosis is not dependent on a functional spindle checkpoint. checkpoint on Taxol or nocodazole treatment. Moreover, in Similarly, because Aurora-B is required for the spindle check- contrast to ZM447439 or hesperadin, Go¨6976 did not cause defects point, ZM447439 also induces spindle checkpoint–independent in cytokinesis or induce polyploidization. At present, we cannot apoptosis.3 Thus, the therapeutic concept of spindle checkpoint explain these interesting differences, but they might be related to inhibition is expected to be efficient in tumor cells harboring a the concomitant inhibition of Aurora-A by Go¨6976 or due to compromised spindle checkpoint. different inhibitor selectivity for distinct Aurora-B–containing We show that Go¨6976 targets the mitotic Aurora kinases, which protein complexes. might explain the phenotypes observed after treatment with Previous work has identified the c-Jun NH2-terminal kinase Go¨6976. Go¨6976 treatment causes spindle formation defects and inhibitor SP600125 as an inhibitor of the spindle checkpoint kinase severe chromosome misalignment, phenotypes attributable to Mps1 (44). Remarkably, SP600125-mediated inhibition of Mps1 led Aurora-A inhibition (42). In addition, inhibition of Aurora-B might to an inactivation of the spindle checkpoint in cancer cells but not explain the checkpoint override in the presence of misaligned in nontransformed fibroblasts. Although we have not investigated chromosomes because Aurora-B has been implicated in spindle the spindle checkpoint activity in nontransformed cells, we found checkpoint function (36, 43). Thus, the efficacy of Go¨6976 might that Go¨6976 killed many cancer cells more efficiently than depend on a two-step mechanism: first, the induction of mitotic nontransformed human cells. The molecular basis for this damage associated with a transient mitotic delay, and second, the selectivity is not clear at present but deserves further detailed override of the spindle checkpoint allowing exit from an aberrant investigation in future studies. mitosis. The latter might be particularly supportive for the induction of apoptosis. In fact, when tumor cells are first treated Disclosure of Potential Conflicts of Interest with Taxol followed by the inactivation of the spindle checkpoint, No potential conflicts of interest were disclosed. the induction of cell death is greatly enhanced (Fig. 5; refs. 7, 37). Interestingly, the reverse order of events, Taxol treatment of cells in which the spindle checkpoint is already inhibited, led to Acknowledgments endoreduplication and a reduced sensitivity toward Taxol3 (13, 35, Received 9/17/08; revised 2/9/09; accepted 2/19/09; published OnlineFirst 4/14/09. Grant support: University Medical Center Giessen and Marburg and 38), suggesting that a therapeutic combination of Taxol and spindle Deutsche Forschungsgemeinschaft Heisenberg fellowship (H. Bastians) and Behring- checkpoint inhibitors strictly depends on the order of drug Ro¨ntgen-Stiftung. treatment. The costs of publication of this article were defrayed in part by the payment of page in vitro charges. This article must therefore be hereby marked advertisement in accordance Although we show that Go¨6976 inhibits Aurora-B and with 18 U.S.C. Section 1734 solely to indicate this fact. in vivo, we found some remarkable differences to the inhibition by We thank Stephen Taylor, Paolo Sassone-Corsi, Bert Vogelstein, Robert Benezra, Qi Min Zhan, Stefan Dimitrov, Stefan Gaubatz, Martin Eilers, Matthias Dobbelstein, and Mathias Schmidt for providing antibodies, cell lines, and plasmids; Rolf Mu¨ller for support; Katja Scheffler for experimental help; the Developmental Therapeutics Program of the National Cancer Institute for the gift of UCN-01; Gary Gorbski for discussions; and 3 A. Stolz, C. Vogel, and H. Bastians, unpublished results. Heike Krebber and all members of the Bastians lab for comments on the manuscript.

References activation of the spindle assembly checkpoint and assembly checkpoint, inducing resistance to Taxol. mitotic slippage. Cancer Cell 2005;8:49–59. Cancer Cell 2003;3:51–62. 1. Schmidt M, Bastians H. Mitotic drug targets and the 9. Sudo T, Nitta M, Saya H, Ueno NT. Dependence of 16. Vader G, Maia AF, Lens SM. The chromosomal development of novel anti-mitotic anticancer drugs. paclitaxel sensitivity on a functional spindle assembly passenger complex and the spindle assembly check- Drug Resist Updat 2007;10:162–81. checkpoint. Cancer Res 2004;64:2502–8. point: kinetochore-microtubule error correction and 2. Bharadwaj R, Yu H. The spindle checkpoint, aneuploi- 10. Masuda A, Maeno K, Nakagawa T, Saito H, Takahashi beyond. Cell Div 2008;3:10. dy, and cancer. Oncogene 2004;23:2016–27. T. Association between mitotic spindle checkpoint 17. Ruchaud S, Carmena M, Earnshaw WC. Chromo- 3. Musacchio A, Salmon ED. The spindle-assembly impairment and susceptibility to the induction of somal passengers: conducting cell division. Nat Rev Mol checkpoint in space and time. Nat Rev Mol Cell Biol apoptosis by anti-microtubule agents in human lung Cell Biol 2007;8:798–812. 2007;8:379–93. cancers. Am J Pathol 2003;163:1109–16. 18. Wang X, Jin DY, Ng RW, et al. Significance of MAD2 4. Rieder CL, Maiato H. Stuck in division or passing 11. Kienitz A, Vogel C, Morales I, Muller R, Bastians H. expression to mitotic checkpoint control in ovarian through: what happens when cells cannot satisfy the Partial downregulation of MAD1 causes spindle check- cancer cells. Cancer Res 2002;62:1662–8. spindle assembly checkpoint. Dev Cell 2004;7:637–51. point inactivation and aneuploidy, but does not confer 19. Hernando E, Nahle Z, Juan G, et al. Rb inactivation 5. Brito DA, Rieder CL. Mitotic checkpoint slippage in resistance towards taxol. Oncogene 2005;24:4301–10. promotes genomic instability by uncoupling cell cycle humans occurs via cyclin B destruction in the 12. Kasai T, Iwanaga Y, Iha H, Jeang KT. Prevalent loss of progression from mitotic control. Nature 2004;430: presence of an active checkpoint. Curr Biol 2006;16: mitotic spindle checkpoint in adult T-cell leukemia 797–802. 1194–200. confers resistance to microtubule inhibitors. J Biol 20. Sotillo R, Hernando E, Diaz-Rodriguez E, et al. Mad2 6. Gascoigne KE, Taylor SS. Cancer cells display profound Chem 2002;277:5187–93. overexpression promotes aneuploidy and tumorigenesis intra- and interline variation following prolonged 13. Swanton C, Marani M, Pardo O, et al. Regulators of in mice. Cancer Cell 2007;11:9–23. exposure to antimitotic drugs. Cancer Cell 2008;14: mitotic arrest and ceramide metabolism are determi- 21. Guardavaccaro D, Frescas D, Dorrello NV, et al. 111–22. nants of sensitivity to paclitaxel and other chemother- Control of chromosome stability by the h-TrCP-REST- 7. Kim M, Liao J, Dowling ML, et al. TRAIL inactivates apeutic drugs. Cancer Cell 2007;11:498–512. Mad2 axis. Nature 2008;452:365–9. the mitotic checkpoint and potentiates death induced 14. Allan LA, Clarke PR. Phosphorylation of caspase-9 by 22. Perez de Castro I, de Carcer G, Malumbres M. A by microtubule-targeting agents in human cancer cells. CDK1/cyclin B1 protects mitotic cells against apoptosis. census of mitotic cancer genes: new insights into tumor Cancer Res 2008;68:3440–9. Mol Cell 2007;26:301–10. cell biology and cancer therapy. Carcinogenesis 2007;28: 8. Tao W, South VJ, Zhang Y, et al. Induction of apoptosis 15. AnandS,Penrhyn-LoweS,VenkitaramanAR. 899–912. by an inhibitor of the mitotic kinesin KSP requires both AURORA-A amplification overrides the mitotic spindle 23. Cahill DP, Lengauer C, Yu J, et al. Mutations of

Cancer Res 2009; 69: (9). May 1, 2009 3882 www.aacrjournals.org

Spindle Checkpoint Inhibition Induces Apoptosis

mitotic checkpoint genes in human cancers. Nature 31. Kalitsis P, Earle E, Fowler KJ, Choo KH. Bub3 gene G, Kao GD. Inactivation of the mitotic checkpoint as a 1998;392:300–3. disruption in mice reveals essential mitotic spindle determinant of the efficacy of microtubule-targeted 24. Saeki A, Tamura S, Ito N, et al. Frequent impairment checkpoint function during early embryogenesis. Genes drugs in killing human cancer cells. Mol Cancer Ther of the spindle assembly checkpoint in hepatocellular Dev 2000;14:2277–82. 2004;3:661–9. carcinoma. Cancer 2002;94:2047–54. 32. Michel L, Diaz-Rodriguez E, Narayan G, Hernando E, 38. Vogel C, Kienitz A, Muller R, Bastians H. The mitotic 25. Yoon DS, Wersto RP, Zhou W, et al. Variable levels of Murty VV, Benezra R. Complete loss of the tumor spindle checkpoint is a critical determinant for top- chromosomal instability and mitotic spindle checkpoint suppressor MAD2 causes premature cyclin B degrada- oisomerase-based chemotherapy. J Biol Chem 2005;280: defects in breast cancer. Am J Pathol 2002;161:391–7. tion and mitotic failure in human somatic cells. Proc 4025–8. 26. Takahashi T, Haruki N, Nomoto S, et al. Identification Natl Acad Sci U S A 2004;101:4459–64. 39. Vogel C, Hager C, Bastians H. Mechanisms of mitotic of frequent impairment of the mitotic checkpoint and 33. Kops GJ, Foltz DR, Cleveland DW. Lethality to human cell death induced by chemotherapy-mediated G2 molecular analysis of the mitotic checkpoint genes, cancer cells through massive chromosome loss by checkpoint abrogation. Cancer Res 2007;67:339–45. hsMAD2 and p55CDC, in human lung cancers. Onco- inhibition of the mitotic checkpoint. Proc Natl Acad 40. Jordan MA, Wilson L. Microtubules as a target for gene 1999;18:4295–300. Sci U S A 2004;101:8699–704. anticancer drugs. Nat Rev Cancer 2004;4:253–65. 27. Weaver BA, Cleveland DW. Does aneuploidy cause 34. Tang Z, Shu H, Oncel D, Chen S, Yu H. Phosphory- 41. Keen N, Taylor SS. Aurora-kinase inhibitors as cancer? Curr Opin Cell Biol 2006;18:658–67. lation of Cdc20 by Bub1 provides a catalytic mechanism anticancer agents. Nat Rev Cancer 2004;4:927–36. 28. Shichiri M, Yoshinaga K, Hisatomi H, Sugihara K, for APC/C inhibition by the spindle checkpoint. Mol Cell 42. Marumoto T, Zhang D, Saya H. Aurora-A—a guardian Hirata Y. Genetic and epigenetic inactivation of mitotic 2004;16:387–97. of poles. Nat Rev Cancer 2005;5:42–50. checkpoint genes hBUB1 and hBUBR1 and their 35. Vogel C, Kienitz A, Hofmann I, Muller R, Bastians 43. Hauf S, Cole RW, LaTerra S, et al. The small molecule relationship to survival. Cancer Res 2002;62:13–7. H. Crosstalk of the mitotic spindle assembly check- Hesperadin reveals a role for Aurora B in correcting 29. Ricke RM, van Ree JH, van Deursen JM. Whole point with p53 to prevent polyploidy. Oncogene 2004; kinetochore-microtubule attachment and in maintain- chromosome instability and cancer: a complex relation- 23:6845–53. ing the spindle assembly checkpoint. J Cell Biol 2003; ship. Trends Genet 2008;24:457–66. 36. Ditchfield C, Johnson VL, Tighe A, et al. Aurora B 161:281–94. 30. Dobles M, Liberal V, Scott ML, Benezra R, Sorger PK. couples chromosome alignment with anaphase by 44. Schmidt M, Budirahardja Y, Klompmaker R, Chromosome missegregation and apoptosis in mice targeting BubR1, Mad2, and Cenp-E to kinetochores. Medema RH. Ablation of the spindle assembly lacking the mitotic checkpoint protein Mad2. Cell 2000; J Cell Biol 2003;161:267–80. checkpoint by a compound targeting Mps1. EMBO 101:635–45. 37. Lee EA, Keutmann MK, Dowling ML, Harris E, Chan Rep 2005;6:866–72.

www.aacrjournals.org 3883 Cancer Res 2009; 69: (9). May 1, 2009

Pharmacologic Abrogation of the Mitotic Spindle Checkpoint by an Indolocarbazole Discovered by Cellular Screening Efficiently Kills Cancer Cells

Ailine Stolz, Celia Vogel, Verena Schneider, et al.

Cancer Res 2009;69:3874-3883. Published OnlineFirst April 14, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-08-3597

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2009/04/13/0008-5472.CAN-08-3597.DC1

Cited articles This article cites 44 articles, 13 of which you can access for free at: http://cancerres.aacrjournals.org/content/69/9/3874.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/69/9/3874.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/69/9/3874. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.