Endocrine-Related (2008) 15 559–568

Effects of the Aurora inhibitor VX-680 on anaplastic thyroid cancer-derived cell lines

Yannick Arlot-Bonnemains, Enke Baldini1, Benedicte Martin, Jean-Guy Delcros, Matteo Toller 2, Francesco Curcio2, Francesco S Ambesi-Impiombato2, Massimino D’Armiento1 and Salvatore Ulisse1

CNRS-UMR 6061 ‘Ge´ne´tique et De´veloppement’, IFR 140 GFAS, Faculte´ de Me´decine, Universite´ Rennes 1, 35043 Rennes Cedex, France 1Department of Experimental Medicine, ‘Sapienza’ University of Rome, Viale del Policlinico 155, 00161 Rome, Italy 2Department of Pathology and Experimental Medicine and Clinic, University of Udine, 33100 Udine, Italy (Correspondence should be addressed to S Ulisse; Email: [email protected])

Abstract Anaplastic thyroid (ATC) are aggressive tumors, which exhibit misregulations leading to uncontrolled cellular proliferation and genomic instability. They fail to respond to chemotherapeutic agents and radiation therapy, and most patients die within a few months of diagnosis. In the present study, we evaluated the in vitro effects on ATC cells of VX-680, an inhibitor of the Aurora serine/threonine involved in the regulation of multiple aspects of chromosome segregation and . The effects of VX-680 on proliferation, apoptosis, soft agar colony formation, cell cycle, and ploidy were tested on the ATC-derived cell lines CAL-62, 8305C, 8505C, and BHT-101. Treatment of the different ATC cells with VX-680 inhibited proliferation in a time- and dose-dependent manner, with the IC50 between 25 and 150 nM. The VX-680 significantly impaired the ability of the different cell lines to form colonies in soft agar. Analysis of caspase-3 activity showed that VX-680 induced apoptosis in the different cell lines. CAL-62 cells exposed for 12 h to VX-680 showed an accumulation of cells with R4N DNA content. Time-lapse analysis demonstrated that VX- 680-treated CAL-62 cells exit metaphase without dividing. Moreover, H3 phosphorylation was abrogated following VX-680 treatment. In conclusion, our data demonstrated that VX-680 is effective in reducing cell growth of different ATC-derived cell lines and warrant further investigation to exploit its potential therapeutic value for ATC treatment. Endocrine-Related Cancer (2008) 15 559–568

Introduction The genetic instability represents a hallmark of solid Although derived from the same cell type, different cancers including thyroid carcinomas, in which it has thyroid show specific histological features, been suggested to underlie the progression to more biological behavior, and degree of differentiation, as a aggressive phenotypes (Shahedian et al.2001, Fagin consequence of different genetic alterations (Fagin 2002, 2002, Wreesmann et al.2002, Are & Shaha 2006). Over Sherman 2003). The majority of thyroid cancers are the last decade, our knowledge regarding the molecular represented by the differentiated papillary and follicular processes controlling the mitotic phase of the cell cycle thyroid carcinomas that, following dedifferentiation, are has increased considerably. As a result, different mitotic thought to give rise to the anaplastic thyroid carcinoma proteins, expression or function of which has been found (ATC; Fagin 2002, Wreesmann et al.2002, Are & Shaha to be altered in human cancer tissues, are now thought to 2006). The latter is one of the most aggressive and lethal play a role in tumor genetic instability. These include the tumors in humans, which fails to respond to all available three family members, Aurora-A, Aurora- chemotherapeutic agents and radiotherapy. Therefore, B, and Aurora-C, implicated in the regulation of multiple most patients die within a year of the diagnosis (Fagin aspects of chromosome segregation and cytokinesis 2002, Wreesmann et al.2002, Are & Shaha 2006). (Katayama et al.1999, Takahashi et al.2000, Tanner

Endocrine-Related Cancer (2008) 15 559–568 Downloaded from Bioscientifica.comDOI: 10.1677/ERC-08-0021 at 10/02/2021 09:39:17AM 1351–0088/08/015–559 q 2008 Society for Endocrinology Printed in Great Britain Online version via http://www.endocrinology-journals.orgvia free access Y Arlot-Bonnemains et al.: Effects of VX-680 on ATC cells et al.2000, Miyoshi et al.2001, Nigg 2001, Sakakura including the ATC (Harrington et al.2004, Hata et al. et al.2001, Ota et al.2002, Carmena & Earnshaw 2003, 2005, Matthews et al.2006, Manfredi et al.2007). Vagnarelli & Earnshaw 2004). During the cell cycle, Among these, VX-680 is a potentAurora kinaseinhibitor, their expression is tightly regulated, being maximal in the shown to suppress tumor growth in xenograft models, G2/M phase, while their rapid degradation at the end of leading to regression of , colon, and pancreatic by the ubiquitin–proteasome pathway is required tumors at well-tolerated doses (Harrington et al.2004). for the cell entry into a new cell cycle (Arlot-Bonnemains In the present study, we evaluated the effects of VX- et al.2001, Nigg 2001, Carmena & Earnshaw 2003, 680 on cell cycle progression, proliferation, apoptosis, Vagnarelli & Earnshaw 2004). Aurora-A localizes onto and soft agar colony formation of different ATC- the duplicated centrosomes and is involved in their derived cell lines. positioning, the recruitment of components at the formation of mitotic spindle and its stability (Nigg 2001, Carmena & Earnshaw 2003, Vagnarelli & Earn- Materials and methods shaw 2004). Aurora-B is a chromosomal passenger Cell line and materials protein, which at the beginning of mitosis associates with The anaplastic carcinoma-derived cell lines CAL-62, chromatin where it forms a complex with proteins as 8305C, 8505C, and BHT-101 were purchased from inner centromere protein (INCENP), survivin, and German Collection of Microorganisms and Cell Cultures borealin, leading to the phosphorylation of histone H3 (DMSZ, Braunschweig, Germany). Agar, monoclonal (Nigg 2001, Carmena & Earnshaw 2003, Vagnarelli & anti-b-tubulin, polyclonal anti-g-tubulin, and anti-actin Earnshaw 2004). Moreover, during the transition from antibodies were from Sigma Chemical Co. Nonidet P-40 anaphase to telophase, Aurora-B plays a role in the (NP-40) and Protease Inhibitor Cocktail Set III were from mitotic spindle dynamics and cleavage furrow, and can Calbiochem (San Diego, CA, USA). Bradford protein be observed in the midbody of cytokinetic cells. Aurora- assay kit and electrophoresis reagents were from Bio-Rad C is also a chromosomal passenger protein, shown to Laboratories. The Aurora kinase inhibitor VX-680 was co-localize and form complexes with Aurora-B, from KAVA Technology (San Diego, CA, USA). The INCENP, and survivin in mitotic cells (Yan et al.2005). goat polyclonal antibody against TACC-3 was from Recently, we reported that all three Aurora kinases are Santa Cruz Biotechnology (Heidelberg, Germany). The overexpressed in human thyroid cancer-derived cell lines polyclonal anti-Aurora-C antibody was generated and tissues (Ulisse et al.2006). These observations may against a 16 amino acid peptide of the C-terminal part help to clarify the molecular mechanisms involved in the of Aurora-C (aa 259–275) by Eurogentec (Seraing, pathogenesis and the genetic instability of thyroid Belgium). The monoclonal antibodies against Aurora-A cancers. In fact, the overexpression of Aurora-A has (31C1) and Aurora-B (AIM-1), and the polyclonal been shown to induce centrosome amplification and to antibody against P-histone H3 were from Abcam potentiate the oncogenic action of Ras. Both these events (Cambridge, UK). The secondary antibodies (tetra- are involved in human thyrocytes transformation and methyl rhodamine isothiocynate (TRITC)-conjugated chromosome instability (Miyoshi et al.2001, Tatsuka anti-goat or anti-rabbit and fluorescein isothiocynate et al.2005). Furthermore, it has been documented that (FITC)-conjugated anti-mouse IgG) were from Jackson p53 is phosphorylated by Aurora-A on Ser215, inducing Laboratories (West Grove, PA, USA). Vectashield was the inhibition of the p53 transactivating action on several from Vector Laboratories (Burlingame, KS, USA). genes (Liu et al.2004). On the other hand, p53 binding has been shown to inactivate Aurora-A in certain cell types (Chen et al.2002). Thus, alterations in the crosstalk Cell cultures between Aurora kinase-A and p53 could be relevant in The normal strain of human thyrocytes (HTU5) was thyroid cancer progression by compromising the cultured as previously described (Curcio et al. 1994). faithfulness of chromosome segregation (Parry et al. These diploid and non-tumorogenic cells retain in 1998, Shahedian et al.2001). culture the functional features of normal human The finding that the expression of all three Aurora thyrocytes. The different ATC-derived cells, CAL-62, kinases is altered in human thyroid cancers may also have 8305C, 8505C, and BHT-101 have been shown to potential therapeutic implications. In fact, over the last possess an abnormal karyotype and predicted loss- few years, specific inhibitors of Aurora kinases have been of-function mutations of the p53 gene (Gioanni et al. identified, which may open a new scenario in cancer 1991, Olivier et al. 2002). The different cell lines were therapy, especially against those cancers that do not cultured in the appropriate media at 37 8Cin5%CO2 respond to the available chemotherapeutic agents, humidified atmosphere.

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access 560 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 559–568

Western blot policeman and fixed in ice-cold ethanol. The cell The cells were lysed in RIPA buffer (50 mM Tris–HCl samples were analyzed for DNA content (propidium (pH 7.4), 1% NP-40, 0.5% sodium deoxycholate, iodide) and/or BrdU content (FITC) as previously 150 mM sodium chloride, 1 mM EDTA, 1! Protease described (30) using an EPICS Elite Flow cytometer Inhibitor Cocktail Set III), sonicated, and then centri- (Coultronics, Hialeah, FL, USA). fuged at 15 000 g for 20 min. Protein concentrations were determined by the Bradford assay. Aliquots of Colony formation in soft agar 50 mg cell protein extracts were electrophoresed on a Petri dishes of 3.5 cm in diameter were first prepared 12.5% polyacrylamide gel and transferred onto nitro- by adding 3 ml complete media with 0.4% soft agar. cellulose membranes. The membranes were then washed Adherent ATC cell cultures were trypsinized and with TBST (50 mM Tris–HCl (pH 7.4), 150 mM NaCl, centrifuged to obtain a single-cell suspension of 0.05% Tween-20), saturated with 5% low-fat milk in 150 000 viable cells/ml. The suspension was mixed TBST and then incubated at 4 8C overnight with with complete medium containing 0.4% soft agar at a antibodies against Aurora-A (1:500), Aurora-B (1:500), ratio 1:2 and then divided into two aliquots, one of Aurora-C (1:500), or actin (1:1000) in TBS-T. After which was supplemented with VX-680 250 nM and the washing, the membranes were incubated with appro- other with the vehicle (DMSO). These suspensions priate horseradish peroxidase-conjugated secondary were plated onto the Petri dishes, 1 ml/dish, and antibodies against mouse or rabbit IgG (1:20 000) in incubated at 37 8C and 5% CO2. Treated and non- TBST and developed using the chemiluminescence treated plates were photographed, few hours after Super Signal kit (Pierce, Rockford, IL, USA). Aurora plating to exclude the presence of cell aggregates (data kinases and actin immunoreactive bands were quantified not shown) and again after 2 weeks. The size of the by scanning densitometry, using Molecular Analyst PC colonies was measured by MetaVue software (Uni- software for the Bio-Rad model 670 scanning densi- versal Imaging Corp., Downingtown, PA, USA) and tometer. The different Aurora kinases/actin ratios were those larger than 50 mm in diameter were scored. calculated and the values obtained for VX-680-treated cells were normalized against those found in control cells Caspase-3 assay and reported as a fold of variation. The different ATC cells were cultured in the absence (DMSO) or presence of 250 nM VX-680 for 72 h. Proliferation assay Following treatment, the cells were rinsed with PBS The CAL-62 cells were cultured in the absence and collected by scraping in PBS. The cells were then (dimethyl sulfoxide, DMSO) or the presence of used to evaluate caspase-3 activity using the caspase- 500 nM VX-680 for different periods of time (1– 3/CPP32 fluorimetric assay kit (Biovision, Mountain 5 days). The dose-dependent effects of VX-680 on cell View, CA, USA). proliferation were evaluated by treating the different ATC cells for 4 days with different concentrations of Time-lapse analysis the Aurora inhibitor (5–500 nM). The cells were pulse The CAL-62 cells were cultured in the absence labeled with 30 mM BrdU for 2 h before the end of the (DMSO) or presence of 250 nM VX-680 and observed incubation time. The BrdU incorporation was analyzed for 24 h under a microscope (Leica DM-IRBE) by means of a colorimetric immunoassay using the cell equipped with an incubation chamber at 37 8C and proliferation ELISA kit (Roche Applied Science), 5% CO2. The cell pictures were performed every 5 min according to the manufacturer’s instructions. The using the MetaVue software. results from VX-680-treated cells were compared with those observed in control cells and expressed as Immunofluorescence (IF) a fold of variation versus control. The CAL-62 cells cultured on glass cover slips were treated with 250 nM VX-680 or vehicle (DMSO) for 6 h. FACS analysis The cells were fixed in cold methanol for 2 min, washed, The CAL-62 cells were cultured in the absence and preincubated with 3% BSA in PBS for 1 h at room (DMSO) or the presence of 250 nM VX-680 for 6, temperature. After three washes with PBS, the cells were 12, or 72 h. In some experiments, the cells were treated incubated with the antibodies anti-TACC3 (1:100), anti- with 30 mM BrdU 20 min before harvesting. The cells Aurora-A (1:200), anti-Aurora-B (1:200), anti-Aurora-C were collected in PBS by scraping with a rubber (1:200), anti-P-histone H3 (1:1000), anti-g-tubulin

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access www.endocrinology-journals.org 561 Y Arlot-Bonnemains et al.: Effects of VX-680 on ATC cells

(1:200), or anti-b-tubulin (1:200) for 2 h at room temperature. After washing, the cover slips were incubated with TRITC-conjugated anti-goat or anti- rabbit (1:100) and FITC-conjugated anti-mouse (1:100) antibodies for 1 h at room temperature and then mounted in Vectashield containing 1 mg/ml DAPI. The cover slips were observed with a microscope Leica DMRXA.

Statistical analysis

The results were expressed as the meanGS.E.M. of three independent experiments. Statistical significance of data was evaluated by the Student t-test using the Epistat computer program. The results were considered significantly different if P values were lower than 0.05.

Results Effects of VX-680 on CAL-62, 8305C, 8505C, and BHT-101 cell proliferation The expression of the Aurora kinase family members (Aurora-A, Aurora-B, and Aurora-C) in the different ATC-derived cell lines was evaluated by western blot analysis in comparison with the normal strain of human thyrocyte HTU5. Similarly to that previously observed in Figure 1 Effects of the VX-680 on different anaplastic thyroid the 8305C cells (Ulisse et al. 2006), a significant carcinoma (ATC)-derived cell lines proliferation. (A) The CAL- augmentation of all three proteins in the cancer cells 62 cells were cultured in the absence (DMSO) or the presence was observed (data not shown). The effect of the Aurora of 500 nM VX-680 for different periods of time. (B) The different ATC cells were cultured in the absence or the presence of kinase inhibitor VX-680 on the CAL-62 cell proliferation 500 nM VX-680 or (C) with different concentrations of VX-680 was evaluated on cells cultured from 1 to 5 days, in the (5–500 nM) for 4 days. The cells were pulse labeled with BrdU presence of 500 nM VX-680 or the vehicle (DMSO) as for 2 h before the end of the incubation time, and then analyzed by a colorimetric immunoassay. In B and C, the results from VX- control. The dose of 500 nM was adopted in these initial 680-treated cells were compared with those observed in control experiments since it was shown to elicit maximal cells and expressed as a fold of variation versus control. Data G response on different tumor cell types in vitro reported represent the mean S.E.M. of three independent experiments. Statistical significances of data were assessed by (Harrington et al.2004). Our results demonstrated that the Student t-test. *P!0.01. VX-680 inhibits the CAL-62 proliferation in a time- dependent manner (Fig. 1A). In particular, significant Effects of VX-680 on ATC cells colony formation (P!0.01) inhibition of BrdU incorporation was detected in soft agar after only 24 h of treatment. The anti-proliferative effect We evaluated the effects of the Aurora kinase inhibitor on became maximal at 4–5 days of treatment, when BrdU the ability of the different ATC cell lines to form colonies incorporation was significantly reduced by more than in soft agar. In these experiments, the cells were cultured 90%. Similar inhibition of proliferation was also in the absence or presence of 250 nM VX-680 for observed on the 8305C, 8505C, and BHT-101 cells 2 weeks. Control cells started to form noticeable colonies following VX-680 (500 nM) treatment for 4 days after 6–8 days of culture. As reported in Fig. 2 and (Fig. 1B). We then evaluated the dose-dependent effects Table 1, VX-680 treatment for 14 days significantly of VX-680 on the different ATC-derived cell lines reduced the number and size of colonies by w70% in the proliferation by treating the cells for 4 days in the 8305C and 90% in the CAL-62, 8505C, and BHT-101. presence of increasing concentrations of the inhibitor (5–500 nM). The results of these experiments, reported in Fig. 1C, showed that the inhibition of the different Effects of VX-680 on CAL-62 cell ploidy and cell ATC cells proliferation was dose dependent with half- cycle progression maximal inhibitory concentrations (IC50) comprised Inhibition of Aurora kinase activity by VX-680 has between 25 and 150 nM. been demonstrated to generate polyploid cells as a

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access 562 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 559–568

Table 1 Effects of the VX-680 on soft agar colony formation of different anaplastic thyroid carcinoma (ATC)-derived cell lines

Number of colonies

ATC cells Control VX-680 P

CAL-62 34.8G1.7 2.2G0.9 !0.001 8305C 37.8G2.2 12.7G2.3 !0.001 8505C 11.7G2.0 0.33G0.3 !0.01 BHT-101 23.6G3.6 2.1G0.4 !0.001

The different ATC cells were cultured in soft agar in 3.5 cm Petri dishes in the absence (DMSO) or presence of 250 nM VX-680 for 2 weeks. Treated and non-treated plates were then photographed and number and size the colonies was measured by MetaVue software, scoring those larger than 50 mm.

control cells completed their mitosis in about 200 min (Fig. 3C). By contrast, twice this time was necessary for VX-680-treated cells to process mitosis without any cytokinesis.

Effects of the VX-680 on Aurora kinases sub- cellular localization, TACC3 localization, and histone H3 phosphorylation in CAL-62 cells We next investigated the effects of VX-680 on CAL-62 mitotic structures and proteins. To ascertain that VX- 680 effects were due to the inhibition of Aurora kinase activities and not to changes in their protein levels, we performed western blot experiments on cell protein Figure 2 Effects of the VX-680 on different anaplastic thyroid carcinoma (ATC)-derived cell lines colony formation in soft extracts from control and the CAL-62 cells treated with agar. The different cell lines were plated in soft agar onto 3.5 cm 250 nM VX-680 for 6 h (Fig. 4A). Densitometric Petri dishes in the absence or presence of VX-680 (250 nM). analysis of western blot results of three independent Treated and non-treated plates were photographed after plating to exclude the presence of cell aggregates (data not shown) and experiments showed no significant variations between after 14 days of incubation. The colony size was determined control and treated cells in Aurora-A (1.17G0.2-fold), using the MetaVue software and those larger than 50 mm were G G scored. Photographs reported in the figure are representative of Aurora-B (0.94 0.11-fold), or Aurora-C (0.97 0.09- one of three similar experiments. fold) proteins level. The IF experiments showed that centrosomal Aurora-A localization was maintained in result of multiple rounds of DNA synthesis in the cells exposed to VX-680 (250 nM) for 6 h (Fig. 4B). absence of cytokinesis. We therefore evaluated cell However, the mitotic cells showed aberrant spindles ploidy by FACS analysis and cell cycle by time-lapse characterized by shorter microtubules. In addition, the experiments in control and VX-680-treated CAL-62 localization on spindle microtubules of the transform- cells. The cell cultures exposed for 6 h to 250 nM VX- ing acidic coiled-coil 3 (TACC3) protein, a substrate of 680 showed an accumulation of cells with 4N DNA Aurora-A which plays a major role in spindle content. Longer exposure to VX-680 (12 h) resulted in microtubules growth and stability, was completely a population of cells with R4N DNA, and a clear abrogated in VX-680-treated cells (Fig. 4B). Aurora-B accumulation of cells with 4N/8N DNA content localization onto the condensing chromatin during (Fig. 3A). IF experiments showed that control prophase was also maintained in treated cells, but, as CAL-62 cells are characterized by large nuclei and a shown in Fig. 4C, histone H3 phosphorylation was no small amount of cytoplasm, while after treatment with longer detectable. In control cells, Aurora-C was solely 250 nM VX-680 for 24 h the cells displayed a observed localized onto the midbody of cytokinetic remarkable increase in cell size and contained multiple cells (Fig. 4D), but following VX-680 treatment no nuclei (Fig. 3B). Time-lapse analysis showed that cells in telophase could be identified.

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access www.endocrinology-journals.org 563 Y Arlot-Bonnemains et al.: Effects of VX-680 on ATC cells

Figure 3 Effects of the VX-680 on the CAL-62 cell ploidy and cell cycle time span. (A) The CAL-62 cells were incubated for 6 or 12 h with 250 nM VX-680 or the vehicle (DMSO). The cells were pulse labeled with BrdU for 2 h before the end of the incubation time, fixed, and analyzed by FACS. (B) The CAL-62 cells were exposed for 24 h to 250 nM VX-680 or vehicle alone (DMSO) as control. Following treatment, control and treated cells were fixed and stained with DAPI and b-tubulin. Scale bar, 10 mm. (C) Time-lapse analysis of the CAL-62 cells cultured in the absence or presence of 250 nM VX-680. The cell pictures were performed every 5 min during 24 h using the MetaVue software. Data reported are representative of one of three similar experiments.

Effects of the VX-680 in inducing ATC cell of the mitotic process (Nigg 2001, Carmena & Earnshaw apoptosis 2003, Vagnarelli & Earnshaw 2004). The findings that In order to assess whether the reduced proliferation and their overexpression may cause cell transformation and the profound alterations in mitosis were associated with that they are overexpressed in several types of human an induction of apoptosis, the accumulation of sub-G0 malignancies, including thyroid cancers, have lead to nuclei in the CAL-62 cells following a treatment with their evaluation as a potential target for anticancer VX-680 (250 nM) for 72 h was investigated. As reported therapy (Katayama et al.1999, Takahashi et al.2000, in Fig. 5A, the percentage of sub-G0 nuclei in control Tanner et al.2000, Miyoshi et al.2001, Sakakura et al. cells was 0.6G0.2 and significantly (P!0.01) increased 2001, Ota et al.2002, Mortlock et al.2005, Sorrentino to 25.3G2.6 in VX-680-treated polyploid cells. More- et al.2005, Ulisse et al.2006). over, a significant (P!0.01) increase of caspase-3 In the present study, we evaluated the in vitro effects of activity was found in the different ATC-derived cells VX-680 on different well-characterized ATC-derived following VX-680 (250 nM) treatment for 72 h with cell lines. We first demonstrated that treatment of these respect to control cells, as reported in Fig. 5B. cells with VX-680 lead to a time- and dose-dependent inhibition of proliferation, with IC50 between 25 and 150 nM, in agreement with the IC50 reported for this Discussion inhibitor on other cancer cell types (Harrington et al. Over the last decade, our knowledge regarding the 2004). Furthermore, VX-680 induced apoptosis in all the molecular players controlling the mitotic process has ATC-derived cell lines tested. Finally, we showed that considerably increased leading to the identification of the VX-680 treatment significantly reduced the ability of new potential oncogenes. These include the three the different ATC cells to form colonies in soft agar. In members of the Aurora kinase family, Aurora-A, this context, it is worthwhile to note that the selective Aurora-B, and Aurora-C, which regulate many aspects inhibition of Aurora-B expression in the ATC-derived

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access 564 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 559–568

cell line ARO (by RNA interference or inhibition of the kinase activity by the quinazoline derivative N-[4-(6,7- dimethoxy-quinazolin-4-ylamino)-phenyl]-benzamide) has been shown to reduce cell proliferation and colony formation in soft agar by only 50%, while no activation of the apoptotic cascade was observed (Sorrentino et al. 2005). This would suggest that the sole inhibition of Aurora-B may not have the expected therapeutic value in ATC treatment (Nikiforow 2005). Moreover, in a recent report Aurora-A, but not Aurora-B or Aurora-C, was found among the most frequently and most strongly overexpressed genes in ATC (Wiseman et al.2007). This is also consistent with the finding that gain of chromosome 20q, where the Aurora-A gene is located (20q13.2–q13.3), is frequently reported in ATC and suggested to play a role in committing differentiated thyroid cancer cells to anaplasia (Rodrigues et al.2004). The VX-680 has been demonstrated to inhibit the three Aurora kinases with inhibition constant (Ki) ranging between 0.6 and 18 nM, showing more than 100-fold selectivity with respect to other kinases tested (Harrington et al. 2004). The only exceptions are the Fms-related -3 (FLT-3) and the c-ABL tyrosine kinase, which are inhibited by the VX-680 with Ki of 30 and 20 nM respectively (Harrington et al. 2004, Carter et al. 2005). Both tyrosine kinases have been detected in thyroid cancer cells and proposed to play a role in malignant transformation (Kung et al.2006, Kurebayashi et al. 2006). However, the inhibition of c-ABL by Imatinib mesylate had negligible effects in suppressing human ATC cell growth (Dziba & Ain 2004, Kurebayashi et al. 2006). Similarly, the SU5416 (3-substituted indolin-2- one), reported to inhibit FLT-3, RET (rearranged upon transformation), vascular endothelial growth factor receptor, and c-kit with similar potency, causes prolifer- ation arrest only in papillary thyroid cancer cells expressing the RET/PTC1 oncogene (Mologni et al. 2006). Therefore, the inhibitory effects of VX-680 on ATC cell growth reported here are likely to be consequent to the specific inhibition of the Aurora kinases. Consistent

Figure 4 Effects of the VX-680 on Aurora kinase protein levels and subcellular localization, centrosome maturation, TACC3 localization, and histone H3 phosphorylation in the CAL-62 cells. The CAL-62 cells were incubated for 6 h with 250 nM VX-680 or DMSO as control. Following treatment, the cells were used for the subsequent experiments. (A) Western blot analysis of Aurora kinase protein levels. (B) Subcellular localization of Aurora-A and TACC3. Following treatment, the cells were fixed and stained for Aurora-A and g-tubulin or TACC3 and b-tubulin. (C) Subcellular localization of Aurora-B and phosphorylated histone H3. Following treatment, the cells were fixed and stained for Aurora-B and P-histone H3. (D) Subcellular localization of Aurora-C in control CAL-62 cells. The cells were fixed and stained for Aurora-C and b-tubulin as a marker for microtubules. Scale bar, 10 mm.

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access www.endocrinology-journals.org 565 Y Arlot-Bonnemains et al.: Effects of VX-680 on ATC cells

with this, we found that the treatment of CAL-62 cells with VX-680 causes a major alteration in centrosome functions with abnormal spindle formation characterized by the presence of short microtubules. We and others demon- strated that Aurora-A kinase activity is required for the phosphorylation and localization of the TACC3 protein on the spindle microtubules. TACC3, in complex with the Ch-Tog protein, is essential in spindle microtubule growth and stability (Mori et al.2007, Ulisse et al. 2007); hence, the complete abrogation of TACC3 localization following the VX-680 treatment could explain the aberrant spindle formationintheCAL-62cells. Histone H3 is also a well-recognized target of Aurora-B kinase and its phosphorylation is thought to mediate chromosome condensation during prophase (Crosio et al. 2002). In agreement with a previous study (Harrington et al. 2004), we showed that the VX- 680 treatment of the CAL-62 cells inhibits histone H3 phosphorylation. Finally, the inhibition of Aurora kinase activity has been demonstrated to generate polyploid cells as a result of multiple rounds of DNA synthesis in the absence of cytokinesis (Kawasaki et al.2001). Recently, endor- eduplication and apoptosis in response to the VX-680 have been shown to be conditioned by the p53–p21- dependent post-mitotic checkpoint. In fact, the cells with intact checkpoint function arrest with 4N DNA content, while those with compromised p53-dependent pathway undergo endoreduplication and apoptosis (Gizatullin et al.2006). All the ATC cell lines employed in the present study are characterized by predicted loss-of- function mutations of the p53 gene (Gioanni et al.1991, Olivier et al.2002). In agreement with the above report, we observed that VX-680 induced endoreduplication in the CAL-62 cells and apoptosis in all the ATC cell lines tested. Both IF and time-lapse experiments suggest the ability of treated cells to enter the mitotic process. However, IF experiments showed that in treated cells all mitoses were arrested in prophase. On the other hand, time-lapse analysis demonstrated that VX-680-treated cells escape mitosis without dividing. Over the last few years,a number ofdifferent inhibitors of the Aurora kinases have been developed and some of them were reported to enter in Phase I clinical trial Figure 5 Effects of the VX-680 on different anaplastic thyroid (Mortlock et al.2005, Matthews et al.2006). These carcinoma (ATC)-derived cell lines apoptosis. (A) The CAL-62 include VX-680, which efficiently inhibits tumor growth cells were incubated for 72 h with 250 nM VX-680 or the vehicle in a variety of in vivo xenograft models, inducing (DMSO). The cells were then scraped in PBS, fixed, and analyzed by FACS. The M1 bar identified the sub-G0/apoptotic regression of leukemia, colon, and pancreatic tumors at cell population scored. (B) Caspase-3 activity was measured in well-tolerated doses (Harrington et al. 2004). This extracts prepared from the different ATC-derived cell lines inhibitor recently entered in Phase II clinical trial on exposed for 72 h to 250 nM VX-680 or to the vehicle alone (DMSO). Data reported represent the meanGS.E.M. of three patients with chronic myelogenous leukemia and independent experiments. Statistical significance of data was Philadelphia chromosome-positive acute lymphocytic assessed by the Student t-test. *P!0.01, **P!0.05. leukemia (Vertex Pharmaceuticals Reports 2007

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access 566 www.endocrinology-journals.org Endocrine-Related Cancer (2008) 15 559–568

Pipeline Progress). It has to be mentioned, however, that Curcio F, Ambesi-Impiombato FS, Perrella G & Coon HG Merck has recently suspended the enrolment in clinical 1994 Long-term culture and functional characterization of trials of the Aurora kinase inhibitor, MK-0457 (VX-680), follicular cells from adult normal human thyroids. PNAS pending a full analysis of all safety data for the drug. The 91 9004–9008. decision was based on preliminary safety data, in which a Dziba JM & Ain KB 2004 Imatinib mesylate (gleevec; STI571) monotherapy is ineffective in suppressing human QTc prolongation was observed in one patient. Patients anaplastic thyroid carcinoma cell growth in vitro. currently enroled in these trials may continue to be Journal of Clinical Endocrinology and Metabolism 89 treated with MK-0457, with additional monitoring for 2127–2135. QTc prolongation. Fagin JA 2002 Branded from the start-distinct oncogenic In conclusion, we demonstrated that VX-680 initiating events may determine tumor fate in the thyroid. inhibits proliferation and induces apoptosis in different Molecular Endocrinology 16 903–911. ATC-derived cell lines. In addition, the inhibitor Gioanni J, Zanghellini E, Mazeau C, Zhang D, Courdi A, strongly reduced the ability of these cell lines to form Farges M, Lambert JC, Duplay H & Schneider M colonies in soft agar. These findings warrant further 1991 Characterization of a human cell line from investigations to exploit the potential therapeutic value anaplastic thyroid carcinoma. Bulletin du Cancer 78 1053–1062. of Aurora kinase inhibition in the treatment of ATC. Gizatullin F, Yao Y, Kung V, Harding MW, Loda M & Shapiro GI 2006 The Aurora kinase inhibitor VX-680 induces endoreduplication and apoptosis preferentially in Acknowledgements cells with compromised p53-dependent postmitotic checkpoint function. Cancer Research 66 7668–7677. This study was supported by the Agenzia Spaziale Harrington EA, Bebbington D, Moore J, Rasmussen RK, Italiana (ASI), the Ligue Nationale Contre le Cancer, Ajose-Adeogun AO, Nakayama T, Graham JA, Demur C, and the Association pour la Recherche sur le Cancer Hercend T, Diu-Hercend A et al. 2004 VX-680, a potent (ARC). The authors declare that there is no conflict of and selective small-molecule inhibitor of the Aurora interest that would prejudice the impartiality of this kinases, suppresses tumor growth in vivo. Nature scientific work. Medicine 10 262–267. Hata T, Furukawa T, Sunamura M, Egawa S, Motoi F, Ohmura N, Marumoto T, Saya H & Horii A 2005 RNA interference targeting aurora kinase a suppresses tumor growth and References enhances the taxane chemosensitivity in human pancreatic cancer cells. Cancer Research 65 2899–2905. Are C & Shaha AR 2006 Anaplastic thyroid carcinoma: Katayama H, Ota T, Jisaki F, Ueda Y, Tanaka T, Odashima S, biology, pathogenesis, prognostic factors, and treatment Suzuki F, Terada Y & Tatsuka M 1999 Mitotic kinase approaches. Annals of Surgical Oncology 13 453–464. expression and progression. Journal of Arlot-Bonnemains Y, Klotzbucher A, Giet R, Uzbekov R, the National Cancer Institute 91 1160–1162. Bihan R & Prigent C 2001 Identification of a functional Kawasaki A, Matsumura I, Miyagawa JI, Ezoe S, Tanaka H, destruction box in the Xenopus laevis aurora-A kinase Terada Y, Tatsuka M, Machii T, Miyazaki H, Furukawa Y pEg2. FEBS Letters 508 149–152. et al. 2001 Downregulation of an AIM-1 kinase couples Carmena M & Earnshaw WC 2003 The cellular geography of with megakaryocytic polyploidization of human hemato- aurora kinases. Nature Reviews. Molecular Cell Biology 4 poietic cells. Journal of Cell Biology 22 275–287. 842–854. Kung SP, Lee CH, Yang AH, Chi CW, Tseng LM & Wu CW Carter TA, Carter TA, Wodicka LM, Shah NP, Velasco AM, 2006 Expression of c-kit, Flk-1, and Flk-2 receptors in Fabian MA, Treiber DK, Milanov ZV, Atteridge CE, benign and malignant tumors of follicular epithelial origin. Biggs WH III et al. 2005 Inhibition of drug-resistant Journal of the Chinese Medical Association 69 74–79. mutants of ABL, KIT, and EGF receptor kinases. PNAS Kurebayashi J, Okubo S, Yamamoto Y, Ikeda M, Tanaka 102 11011–11016. K, Otsuki T & Sonoo H 2006 Additive antitumor Chen SS, Chang PC, Cheng YW, Tang FM & Lyn YS 2002 effects of gefitinib and imatinib on anaplastic thyroid Suppression of the STK15 oncogenic activity requires a cancer cells. Cancer Chemotherapy and Pharma- transactivation-independent p53 function. EMBO Journal cology 58 460–470. 21 4491–4499. Liu Q, Kaneko S, Yang L, Feldman RI, Nicosia SV, Chen J & Crosio C, Fimia GM, Loury R, Kimura M, Okano M, Zhou M, Cheng JQ 2004 Aurora-A abrogation of p53 DNA binding Sen S, Allis CD & Sassone-Corsi P 2002 Mitotic and transactivation activity by phosphorylation of serine phosphorylation of histone H3: spatio-temporal regulation 215. Journal of Biological Chemistry 279 52175–52182. by mammalian aurora kinases. Molecular and Cellular Manfredi MG, Ecsedy JA, Meetze KA, Balani SK, Biology 22 874–885. Burenkova O, Chen W, Galvin KM, Hoar KM, Huck JJ,

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access www.endocrinology-journals.org 567 Y Arlot-Bonnemains et al.: Effects of VX-680 on ATC cells

LeRoy PJ et al. 2007 Antitumor activity of MLN8054, an Shahedian B, Shi Y, Zou M & Farid NR 2001 Thyroid orally active small-molecule inhibitor of . carcinoma is characterized by genomic instability: PNAS 104 4106–4111. evidence from p53 mutations. Molecular Genetics and Matthews N, Visintin C, Hartzoulakis B, Jarvis A & Metabolism 72 155–163. Selwood DL 2006 Aurora A and B kinases as targets for Sherman SI 2003 Thyroid carcinoma. Lancet 361 501–511. cancer: will they be selective for tumors? Expert Review Sorrentino R, Libertini S, Pallante PL, Troncone G, of Anticancer Therapy 6 109–120. Palombini L, Bavetsias V, Spalletti-Cernia D, Laccetti P, Miyoshi Y, Iwao K, Egawa C & Noguchi S 2001 Association Linardopoulos S, Chieffi P et al. 2005 Aurora B of centrosomal kinase STK15/BTAK mRNA expression overexpression associates with the thyroid carcinoma with chromosomal instability in human breast cancers. undifferentiated phenotype and is required for thyroid International Journal of Cancer 92 370–373. carcinoma cell proliferation. Journal of Clinical Mologni L, Sala E, Cazzaniga S, Rostagno R, Kuoni T, Endocrinology and Metabolism 90 928–935. Puttini M, Bain J, Cleris L, Redaelli S, Riva B et al. 2006 Takahashi T, Futamura M, Yoshimi N, Sano J, Katada M, Inhibition of RET tyrosine kinase by SU5416. Journal of Takagi Y, Kimura M, Yoshioka T, Okano Y & Saji S Molecular Endocrinology 37 199–212. 2000 Centrosomal kinases, HsAIRK1 and HsAIRK3, are Mori D, Yano Y, Toyo-oka K, Yoshida N, Yamada M, overexpressed in primary colorectal cancers. Maramatsu M, Zhang D, Saya H, Toyoshima YY, Kinoshita Japanese Journal of Cancer Research 91 1007–1014. K et al. 2007 NDEL1 phosphorylation by Aurora-A kinase is Tanner MM, Grenman S, Koul A, Johannsson O, Meltzer P, essential for centrosomal maturation, separation and TACC3 Pejovic T, Borg A & Isola JJ 2000 Frequent amplification recruitment. Molecular and Cellular Biology 27 352–367. of chromosomal region 20q12-q13 in ovarian cancer. Mortlock AA, Keen NJ, Jung FH, Heron NM, Foote KM, Clinical Cancer Research 6 1833–1839. Wilkinson RW & Green S 2005 Progress in the Tatsuka M, Sato S, Kitajima S, Suto S, Kawai H, Miyauchi M, Ogawa I, Maeda M, Ota T & Takata T 2005 Overexpression development of selective inhibitors of Aurora kinase. of Aurora-A potentiates HRAS-mediated oncogenic Current Topics in Medicinal Chemistry 5 807–821. transformation and is implicated in oral . Nigg EA 2001 Mitotic kinases as regulators of cell division Oncogene 24 1122–1127. and its checkpoints. Nature Reviews. Molecular Cell Ulisse S, Delcros JG, Baldini E, Toller M, Curcio F, Giacomelli L, Biology 2 21–32. Prigent C, Ambesi-Impiombato FS, D’Armiento M & Arlot- Nikiforow YE 2005 Anaplastic carcinoma of the thyroid-will Bonnemains Y 2006 Expression of Aurora kinases in human Aurora B light a path for treatment? Journal of Clinical thyroid carcinoma cell lines and tissues. International Endocrinology and Metabolism 90 1243–1245. Journal of Cancer 119 275–282. Olivier M, Eeles R, Hollstein M, Khan MA, Harris CC & Ulisse S, Baldini E, Toller M, Delcros J-G, Gue`ho A, Curcio F, Hainaut P 2002 The IARCT P53 Database: new online De Antoni E, Giacomelli L, Ambesi-Impiombato FS, mutation analysis and recommendations to users. Human Bocchini S et al. 2007 Transforming acidic coiled-coil 3 and Mutation 19 607–614. Aurora-A interact in human thyrocytes and their expression Ota T, Suto S, Katayama H, Han ZB, Suzuki F, Maeda M, is deregulated in thyroid cancer tissues. Endocrine-Related Tanino M, Terada Y & Tatsuka M 2002 Increased mitotic Cancer 14 827–837. phosphorylation of histone H3 attributable to AIM- Vagnarelli P & Earnshaw WC 2004 Chromosomal passen- 1/Aurora-B overexpression contributes to chromosome gers: the four-dimensional regulation of mitotic events. number instability. Cancer Research 62 5168–5177. Chromosoma 113 211–222. Parry EM, Ulucan H, Wyllie FS, Wynford-Thomas D & Wiseman SM, Masoudi H, Niblock P, Turbin D, Rajput A, Parry JM 1998 Segregational fidelity of chromosomes in Hay J, Bugis S, Filipenko D, Huntsman D & Gilks B 2007 human thyroid tumour cells. Chromosoma 107 491–497. Anaplastic thyroid carcinoma: expression profile of Rodrigues RF, Roque L, Rosa-Santos J, Cid O & Soares J targets for therapy offers new insights for disease 2004 Chromosomal imbalances associated with anaplas- treatment. Annals of Surgical Oncology 14 719–729. tic transformation of follicular thyroid carcinomas. British Wreesmann VB, Ghossein RA, Patel SG, Harris CP, Schnaser Journal of Cancer 90 492–496. EA, Shaha AR, Tuttle RM, Shah JP, Rao PH & Singh B 2002 Sakakura C, Hagiwara A, Yasuoka R, Fujita Y, Nakanishi M, Genome-wide appraisal of thyroid cancer progression. Masuda K, Shimomura K, Nakamura Y, Inazawa J, Abe T American Journal of Pathology 161 1549–1556. et al. 2001 Tumour-amplified kinase BTAK is amplified Yan X, Cao L, Li Q, Wu Y, Zhang H, Saiyin H, Liu X, Zhang X, and overexpressed in gastric cancers with possible Shi Q & Yu L 2005 Aurora C is directly associated with involvement in aneuploid formation. British Journal of Survivin and required for cytokinesis. Genes to Cells: Cancer 84 824–831. Devoted to Molecular and Cellular Mechanisms 10 617–626.

Downloaded from Bioscientifica.com at 10/02/2021 09:39:17AM via free access 568 www.endocrinology-journals.org