MAP3K10 Promotes the Proliferation and Decreases the Sensitivity of Pancreatic Cancer Cells to Gemcitabine by Upregulating Gli-1 and Gli-2

Cancer Letters 329 (2013) 228–235

Contents lists available at SciVerse ScienceDirect

Cancer Letters

journal homepage: www.elsevier.com/locate/canlet

MAP3K10 promotes the proliferation and decreases the sensitivity of pancreatic cancer cells to gemcitabine by upregulating Gli-1 and Gli-2

Yong An a, Baobao Cai a, Jianmin Chen b, Nan Lv a, Jie Yao c, Xiaofeng Xue d, Min Tu a, Dong Tang c, Jishu Wei b, ⇑ Kuirong Jiang b, Junli Wu b, Qiang Li b, Wentao Gao b, Yi Miao a,b, a Laboratory of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China b Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China c Department of General Surgery, The First Affiliated Hospital of Yangzhou University, Yangzhou, China d Department of General Surgery, The First Affiliated Hospital of Suzhou University, Suzhou, China article info abstract

Article history: Pancreatic ductal adenocarcinoma (PDAC) is among the most lethal human malignancies and is regulated Received 8 August 2012 by Sonic Hedgehog (Shh) signaling. Recently, MAP3K10 has been shown to regulate Shh signaling, sug- Received in revised form 18 October 2012 gesting a role for MAP3K10 in the tumorigenesis of PDAC. We determined the expression status of Accepted 2 November 2012 MAP3K10 in PDAC tissues and cell lines, and analyzed the viability and cell proliferation of PDAC cells with an overexpression or knockdown of MAP3K10 in vitro. MAP3K10 was upregulated in PDAC tissues and cell lines. Overexpression of MAP3K10 promoted the proliferation and decreased the gemcitabine Keywords: sensitivity of pancreatic cancer cells. In contrast, knockdown of MAP3K10 signiﬁcantly decreased cell MAP3K10 proliferation and sensitized cells to gemcitabine. However, neither overexpression nor knockdown of Sonic Hedgehog pathway Pancreatic cancer MAP3K10 affected cell migration. Moreover, overexpression of MAP3K10 resulted in upregulation of Chemoresistance Gli-1 and Gli-2 in PDAC cells. Our results indicate a novel and important role for MAP3K10 in the proliferation and chemoresistance of PDAC. Our study suggests that targeting MAP3K10 is a potential strategy for the development of alternative therapies for pancreatic cancers. Ó 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction strated that the Shh pathway promotes the tumorigenesis of pancreatic cancers via enhancing cell proliferation [8,10–12], Pancreatic cancer is one of the most lethal human malignancies. increasing invasion and metastasis [10,13–17], and protecting Great improvements have been made in surgical and chemothera- against apoptosis [14,18,19]. In a phase I trial testing an inhibitor peutic approaches during the past decades; however, the overall of smoothend (SMO), vismodegib (GDC-0449), exciting responses 5-year survival of pancreatic cancer is still below 5% [1]. Thus, it were observed in patients with advanced basal cell carcinoma is urgent to develop alternative therapeutic strategies for pancre- (BCC) or medulloblastoma. However, only one in eight patients atic cancers, such as molecular targeted therapies. with pancreatic carcinoma experienced a stable disease as the best Increasing evidences suggest that molecular signaling pathways response [20]. Therefore, to develop efﬁcient therapies targeting regulating cell development are subverted in the aberrant morpho- the hedgehog pathway, it is essential to further investigate the reg- genetic process of cancers [2,3]. One such pathway is the Hedgehog ulation of hedgehog signaling in pancreatic cancer. signaling pathway, an essential pathway during embryonic devel- A recent study showed that the Shh pathway was regulated by opment [4]. Sonic Hedgehog (Shh) is a secreted mammalian ortho- MAP3K10 [21]. MAP3K10 is a member of the mitogen-activated log of the Hedgehog family and plays multiple roles during protein kinase kinase kinase (MAP3K) family. MAP3K10 activates embryonic development [5]. Activation of the Shh pathway trig- C-Jun N-terminal kinase (JNK) signaling and the p38 mitogen-acti- gers a series of intracellular events through the GLI transcriptional vated-protein kinase (MAPK) pathway, as well as regulates apopto- effectors Gli-1, Gli-2 and Gli-3 [6,7]. sis in many neurodegenerative diseases [22,23]. The regulation of Deregulation of Shh signaling has been found in many cancers, Shh signaling by MAP3K10 suggests that MAP3K10 may play an including pancreatic cancer [8,9]. Amounting data have demon- important role in the progression and maintenance of pancreatic cancers. Therefore, we investigated the functions of MAP3K10 in the proliferation, cell cycle progression, migration and apoptosis ⇑ Corresponding author. Address: Department of General Surgery, The First Afﬁliated Hospital of Nanjing Medical University, 300 Guangzhou Rd., Nanjing of pancreatic cancer cells. We found that MAP3K10 increased cell 210029, China. Tel.: +86 25 68136590; fax: +86 25 83781992. proliferation, promoted cell cycle progression and suppressed E-mail address: [email protected] (Y. Miao).

0304-3835/$ - see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.canlet.2012.11.005 Y. An et al. / Cancer Letters 329 (2013) 228–235 229 gemcitabine-induced apoptosis of pancreatic cancer cells by containing 20 lL viral supernatant (1 107 UT/ml) and 5 lg/ml polybrene. After 6 h upregulating Gli-1 and Gli-2. of incubation, the virus-containing medium was replaced with fresh medium. 48 h later, transfected cells were screened by puromycin. Then, successfully transfected cells were subcultured in 96-well plates with one cell per well. Two weeks later, 2. Materials and methods stably transfected cells from monoclony were passaged, cultured and verified by PCR and western-blot. For the knockdown experiments, construct sh2, with a 2.1. Cell cultures and tissue collections knockdown efficiency P80%, was used for further studies. The shRNA sequences used for further studies were as follows: shMAP3K10, 50-CCGGGCACATGTTTGAT- Human pancreatic cancer cell lines were purchased from the Shanghai Cell Bank GACCTTCGTCAAGAGCGAAGGTCATCAAACATGTGCTTTTTG-30 and shcontrol, 50- (Shanghai, China) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) CCGGCCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGGTTTTTG-30. (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Sigma, St Louis, MO, USA). Human pancreatic cancer specimens and their adjacent normal pancreatic tissues (26 pairs) and six normal pancreas tissues were collected 2.4. Reverse transcription-PCR (RT-PCR) from patients who underwent surgery according to an approved human protocol at the First Affiliated Hospital with Nanjing Medical University. Written informed con- Total RNA from cells and tissues was extracted using Trizol reagent (Invitrogen, sent was obtained from every patient. Carlsbad, CA, USA). cDNA was synthesized using the PrimeScript RT Kit (Takara, Da- lian, China). Quantitative RT-PCR was performed with FastStart Universal SYBR Green Master (Rox) (Roche, USA) with an ABI 7500 (Applied Biosystems, Life Tech- 2.2. Immunohistochemistry nologies Corporation, California, USA). Briefly, the optimized reaction was performed in duplicates at a final reaction volume of 25 lL containing 0.25 lL After tissue collection, 4 lm thick paraffin-embedded tissue sections were (0.3 lM) of each forward and reverse primer, 2.5 lL of cDNA (50 ng), 12.5 lL SYBR deparaffinized in xylene, rehydrated in graded alcohol and blocked in methanol Green Master (ROX) (1X) and 9.5 lL of nuclease free water. The qPCR cycling was containing 3% hydrogen peroxide. The slides were covered with a blocking solution performed as follows: initial denaturation at 95 °C for 10 min followed by 40 cycles for 1 h at room temperature and incubated with a rabbit-anti-human MAP3K10 of denaturation at 95 °C for 10 s, annealing for 60 s at 56 °C (MAP3K10), 58 °C antibody (Abcam, Cambridge, MA, USA) for 2 h at 37 °C. After rinsing with phos- (GSK3b, Gli-1) or 60 °C (DYRK2, Gli-2) and finally a melting curve profile was set phate-buffered saline (PBS; pH 7.4) solution, sections were treated with a goat- at 95 °C (15 s), 60 °C (15 s), 95 °C (15 s). The relative expression of mRNA was anti-rabbit secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) examined as the inverse log of the DDCT and normalized to the reference gene, for 1 h at 37 °C. Next, the slides were incubated with a 3-diaminobenzidine solution GAPDH. Primers for qRT-PCR were synthesized by Invitrogen (shanghai, china) for 10 min, and then counterstained with hematoxylin. and the sequences were as follows: MAP3K10 sense: 50-AGAACCACAACCTCGCA- GAC-30, antisense: 50-TTCATAGCCACGCCATACGC-30; DYRK2 sense: 50-CGCCAGAAG- 2.3. Plasmid construction, lentivirus packaging and stable transfection cell lines TAGCAGCAGGAC-30, antisense: 50-CCTGGATCTGTCCGTGAGCGT-30; GSK3b sense: generation 50-GACATTTCACCTCAGGAGTGC-30; antisense: 50-GTTAGTCGGGCAGTTGGTGT-30; GLI1 sense: 50-AGCCAAGCACCAGAATCGGAC-30; antisense: 50-GTTTGGTCACAT A human MAP3K10 complementary CDS sequence was amplified from total GGGCGTCAG-30; GLI2 sense: 50-GGGTCTGGGGTCAGCCTTTGGA-30; antisense: 50- mRNA of BxPC-3 cells using PrimeSTAR HS DNA Polymerase (Takara, DR010A, Da- AATGGCGACAGGGTTGACGGT-30; GAPDH sense: 50-TCACCCACACTGTGCCCATCTAC- lian, China) with the forward primer 5-GCCCGCTAGCCATGGAGGAGGAGGAGGG-3 GA-30, antisense:50-CAGCGGAACCGCTCATTGCCAATGG-30. and the reverse primer 5-CGGTGGCGCGCCGGCCTTAGTGGGAGCC-3. The sequence was inserted into a Lv-CMV-EGFP vector to construct a MAP3K10 overexpression lentivirus. Three shRNA plasmids (sh1, sh2 and sh3) for human MAP3K10 were de- 2.5. Western-blotting signed against different MAP3K10 targets and constructed in pLKO.1-puro vectors. A scrambled sequence was designed as a negative control. All plasmids were veri- Cells were lysed using RIPA buffer with 1% PMSF on ice. The concentration of to- fied by sequencing (Invitrogen). To generate stable MAP3K10 overexpression or tal protein was determined using a BCA kit (Keygen, Nanjing, China). Equal amounts knockdown cell lines, BxPC-3 and PANC-1 cells were cultured in six-well plates of protein (30 lg) were resolved with 10% SDS-PAGE and transferred to polyvinyli- to 40% confluence. The medium was then replaced with 2 ml fresh culture medium dene difluoride (PVDF) membranes (Millipore, Bedford, USA) using a mini trans-blot

Fig. 1. MAP3K10 was overexpressed in pancreatic cancers. (a) qRT-PCR analysis of MAP3K10 expression in 26 pairs of pancreatic cancer tissues and adjacent tissues (p < 0.05). (b) A scatter plot showing the MAP3K10 expression in the PDAC tissues and adjacent tissues (p < 0.05). (c) qRT-PCR and western blot analysis of MAP3K10 expression in four pancreatic cancer cell lines (p < 0.05). BxPC-3 showed a relative high expression, while PANC-1 demonstrated a low level. (d) Immunohistochemistry showing the intracellular location of MAP3K10 in PDAC sample. Magniﬁcation, 20X. 230 Y. An et al. / Cancer Letters 329 (2013) 228–235

Fig. 2. Generation of stable transfected BxPC-3 and PANC-1 cells. (a) Analysis of MAP3K10 expression in stable transfected BxPC-3 cells using qRT-PCR. BxPC-3 cells were transduced with three shRNAs and one control shRNA (designated as ‘BxPC-3-sh1’, ‘BxPC-3-sh2’, ‘BxPC-3-sh3’ and ‘BxPC-3-shcontrol’, respectively). (b) Analysis of MAP3K10 expression in stable transfected PANC-1 cells using qRT-PCR. PANC-1-MAP3K10, MAP3K10-overexpressing PANC-1 cells; PANC-1-EGFP, PANC-1 cells transfected with a vector-expressing EGFP. NT indicates non-transduced cells. (c) Western blots were performed to conﬁrm the efﬁciency of transfected overexpression and knockdown.

apparatus (Bio-Rad laboratories, Hercules, CA, USA). Membranes were probed with 2.8. Cell migration assay primary antibodies for 12 h at 4 °C, and then incubated with secondary antibodies for 2 h at room temperature. MAP3K10, Dual-speciﬁcity tyrosine-phosphorylation Cell migration analyses were performed with a chamber 6.5 mm in diameter regulated kinases 2 (DYRK2) and GLI2 antibodies were purchased from Abcam and a pore size of 8.0 lm (Corning, Corning, NY, USA). First, cells were starved for (Cambridge, MA, USA). Glycogen synthase kinase 3b (GSK3b) and GLI1 antibodies 24 h in 2% FBS-DMEM media. Then, 1 105 cells were seeded into the upper were from Cell Signaling Technology (Danvers, MA, USA). The goat-anti-rabbit sec- chambers in 200 lL serum-free DMEM supplemented with 0.1% BSA and 600 lL ondary antibody was purchased from Beyotime (Nantong, China). GAPDH (Beyo- 10% FBS-DMEM medium was added into the lower wells. Cells were incubated time, Nantong, China) was used as an internal control. Electrochemiluminescence for 24 h at 37 °C. Cells that invaded the bottom of the Matrigel-coated membrane was performed with a Chemilmager 5500 imaging system (Alpha Innotech Co., were stained with 0.1% crystal violet in methanol. Stained cells were soaked in San Leandro, CA, USA). 33% ice-cold acetic acid and oscillated for 10 min. Then the absorbance of 33% ice-cold acetic acid containing crystal violet was assessed using a microplate reader (Tecan, Shanghai, China) at a 570 nm wavelength. OD (optical density) value reﬂects the number of penetrated cells indirectly. Cell migration was assessed by OD value.

2.6. Cell viability and proliferation assays 2.9. Statistical analysis Aliquots of 2 103 cells were seeded in 96-well plates (Costar, Cambridge). Cell viability was determined using an MTT assay 72 h later. Cell proliferation was All experiments were repeated in triplicate. All values were mean ± standard determined using an EdU (5-ethynyl-20-deoxyuridine) assay. After 48 h of cultur- deviation (SD). Statistical significance was determined with a Student’s t-test using ing, cells were incubated with 100 lL of EdU for 2 h at 37 °C using a Cell-Light™ SPSS 15.0. P-values < 0.05 were considered as statistically significant. EdU ApolloÒ643 In vitro Imaging Kit (RIBO, Guangzhou, China). After fixation with 4% polyoxymethylene, cells were treated with Alexa 594 azide, and then counterstained with Hoechst 33342 for detection and imaging with a fluorescence micro- 3. Results scope (Olympus). 3.1. Increased MAP3K10 expression in PDAC tissues and cells

To quantitate tissues’ mRNA relative expression of MAP3K10, 2.7. Cell cycle and apoptosis analysis cDNAs from six normal pancreases were mixed and MAP3K10 mRNA level was defined as control. MAP3K10 was overexpressed Cell cycle and apoptosis were assessed by flow cytometry (Becton Dickinson, in the pancreatic cancers. Of the 26 paired samples, 21 showed sig- San Jose, CA, USA). For cell cycle analysis, cells were cultured in 6-well plates for 24 h, harvested and fixed in ethanol at 20 °C overnight. Cells were collected by nificantly higher levels of MAP3K10 mRNA in the cancer tissues centrifugation, washed with PBS, and re-suspended in 500 lL of PBS with 0.2% than in the adjacent tissues (Fig. 1a, p < 0.05). The mean MAP3K10 Triton X-100, 10 mM EDTA, 100 lg/ml RNase A and 50 lg/ml propidium iodide mRNA in tumors and tumor-adjacent normal tissues was 6.46 and (PI) at room temperature for 30 min. For apoptosis analyses, cells were treated with 2.15, respectively (Fig. 1b). To quantitate MAP3K10 mRNA and pro- gemcitabine (5 lMor25lM) for 48 h. The cell pellets were collected, washed with tein level of cell lines, total cDNAs and total proteins of 26 tumor- PBS, and suspended in 100 lLof1 binding buffer, stained with 5 lL of Phycoery- thrin (PE)–Annexin-V and 5 lL of 7-AAD at room temperature for 15 min in the adjacent tissues were mixed and defined as control. In comparison dark. The stained cells were immediately analyzed by flow cytometry. with tumor-adjacent normal tissue, all four pancreatic cancer cell Y. An et al. / Cancer Letters 329 (2013) 228–235 231

Fig. 3. MAP3K10 promoted the viability and proliferation of pancreatic cancer cells in vitro. (a) The viability of MAP3K10 knockdown BxPC-3 cells over 72 h was measured using MTT assays (p < 0.05). (b) The viability of MAP3K10 overexpression PANC-1 cells over 72 h was measured using MTT assays (p < 0.05). (c) Cell proliferation was determined with an EdU (5-ethynyl-20-deoxyuridine) assay. An EdU positive cell indicates cell proliferation. (d) Quantitation of the data from Fig. 3c(p < 0.05). (e) Cell cycle progression was analyzed using flow cytometry. S + G2 phase reflects cell proliferation. (f) Quantitation of the data from Fig. 3e(p < 0.05). lines displayed elevated MAP3K10 mRNA levels and 3/4 of the cell 80% lower expression when compared with non-transduced lines had increased MAP3K10 protein levels (Fig. 1c, p < 0.05). In BxPC-3 cells (p < 0.05). addition, immunohistochemistry showed that MAP3K10 was overexpressed in PDAC samples compared with that of the adjacent tis- 3.3. MAP3K10 promoted the viability and proliferation of pancreatic sues (Fig. 1d). Fig. 1d shows that MAP3K10 was mainly located in cancer cells in vitro the cytoplasm of tumor cells. MAP3K10 knockdown significantly suppressed the viability of BxPC-3 cells (BxPC-3-shMAP3K10 vs. BxPC-3shcontrol: 3.2. Generation of stable transfected cells 23.31 ± 1.89% vs. 52.19 ± 4.32%, p < 0.05) (Fig. 3a). In contrast, MAP3K10 overexpression promoted the viability of PANC-1 cells To investigate the roles of MAP3K10 in PDAC, we constructed (PANC-1-MAP3K10 vs. PANC-1-EGFP: 50.78 ± 4.57% vs. MAP3K10 overexpression and knockdown lentiviral constructs. 40.46 ± 2.48%, p < 0.05) (Fig. 3b). Cell proliferation assays with BxPC-3 cells were infected with MAP3K10 knockdown lentiviral EdU (5-ethynyl-20-deoxyuridine) showed that MAP3K10 overex- constructs (Fig. 2a). PANC-1 cells were infected with MAP3K10 pression enhanced the proliferation of PANC-1 cells, while overexpression lentiviral constructs (Fig. 2b). Stable infected cells MAP3K10 knockdown decreased the proliferation of BxPC-3 cells were selected for further experiments. The expression levels of (Fig. 3c and d, p < 0.05). To investigate the mechanism of MAP3K10 were confirmed by qRT-PCR (Fig. 2a and b) and western MAP3K10-modulated cell growth, cell cycle progression was ana- blots (Fig. 2c). MAP3K10-overexpressing PANC-1 cells showed a lyzed by flow cytometry. MAP3K10 knockdown BxPC-3 cells 116-fold higher MAP3K10 expression than non-transduced showed a reduced proportion of cells in the S + G2 phase (BxPC- PANC-1 cells, whereas MAP3K10-knockdown BxPC-3 cells had an 3-shMAP3K10 vs. BxPC-3-shcontrol: 33.77 ± 6.16% vs. 232 Y. An et al. / Cancer Letters 329 (2013) 228–235

Fig. 4. MAP3K10 did not affect the migration of pancreatic cancer cells. (a) Cell migration was performed using a Matrigel migration assay. The upper chambers were seeded with various cell lines. The membranes of the chambers were stained with 0.1% crystal violet 24 h later. Magniﬁcation, 20. (b) Quantiﬁcation of the data of Fig. 4a. OD values from three independent experiments were assessed.

Fig. 5. MAP3K10 decreased the gemcitabine sensitivity of pancreatic cancer cells. (a) After treatment with 5 or 25 lM gemcitabine for 48 h, the apoptosis rate was analyzed with ﬂow cytometry. UR + LR is indicative of apoptosis. (b and c) Quantiﬁcation of the data from Fig. 5a(p < 0.05).

60.75 ± 6.05%, p < 0.05), while MAP3K10 overexpression PANC-1 3.4. MAP3K10 did not affect the migration of pancreatic cancer cells cells exhibited an elevated percentage of cells in the S + G2 phase (PANC-1-MAP3K10 vs. PANC-1-EGFP: 61.72 ± 8.98% vs. We next determined if MAP3K10 promoted the cell migration of 43.18 ± 3.94%, p < 0.05). These results indicate that MAP3K10 pos- pancreatic cancer cells using a matrigel migration assay. MAP3K10 itively regulated the growth of pancreatic cancer cells via modulat- knockdown cells did not signiﬁcantly effect BxPC-3 cell migration ing cell cycle progression. compared with that of control groups (BxPC-3-shMAP3K10 vs. Y. An et al. / Cancer Letters 329 (2013) 228–235 233

Fig. 6. MAP3K10 up-regulates Gli-1 and Gli-2. (a–d) qRT-PCR analysis of MAP3K10 on the transcriptional regulation of DYRK2 (a), GSK3b (b), GLI1 (c) and GLI2 (d). mRNA levels were determined in BxPC-3-shMAP3K10, PANC-1-MAP3K10, and their control group cells. The relative levels were normalized to GAPDH (p < 0.05). (e) Parallel samples from Fig. 6a–d were immunoblotted for the proteins of DYRK2, GSK3b, GLI1 and GLI2. GAPDH served as loading control (p < 0.05). Each group was analyzed in triplicate, and the data are presented as the average ± SD.

BxPC-3 shcontrol: 0.61 ± 0.08 vs. 0.58 ± 0.06, p = 0.564). A similar BxPC-3 cells was higher than that of control groups (p < 0.05), result was observed in PANC-1 cells (PANC-1-MAP3K10 vs. while MAP3K10 overexpression reduced the apoptosis rate of PANC-1-EGFP: 0.38 ± 0.04 vs. 0.4 ± 0.04, p = 0.529) (Fig. 4). PANC-1 cells compared with the control groups (p < 0.05). In addition, MAP3K10 knockdown increased the frequency of apoptosis of 3.5. MAP3K10 decreased the sensitivity of pancreatic cancer cells to BxPC-3-shMAP3K10 cells treated with 5 lM gemcitabine gemcitabine (16.14% ± 0.93%) when compared with BxPC-3-shcontrol cells (11.32% ± 2.12%) (p < 0.05). The frequency of apoptosis after To assess the effect of MAP3K10 on the sensitivity of pancreatic treatment with 25 lM gemcitabine was 42.19% ± 2.42% and cancer cells to chemotherapy, we examined the cell apoptosis rate 30.84% ± 3.57% in BxPC-3-shMAP3K10 and BxPC-3-shcontrol cells, of PANC-1 and BxPC-3 cells after treatment with 5 or 25 lM gem- respectively. After treatment with 5 lM gemcitabine, MAP3K10 citabine for 48 h. The apoptosis rate of MAP3K10 knockdown overexpression decreased the apoptosis frequency of PANC-1- 234 Y. An et al. / Cancer Letters 329 (2013) 228–235

MAP3K10 cells (14.98% ± 1.63%) when compared with PANC-1- mediating the Shh pathway via MAP3K10 in pancreatic cancer EGFP cells (21.92% ± 2.29%) (p < 0.05). The frequency of apoptosis cells. Indeed, priming phosphorylation by DYRK2 is a prerequisite after treatment with 25 lM gemcitabine was 23% ± 2.06% and for GSK3b phosphorylation of eIF2B [45], NFAT [46] and c-Jun/c- 37.31% ± 2.56% in PANC-1-MAP3K10 cells and PANC-1-EGFP cells, Myc [26], indicating a more general role for DYRK2 in the priming respectively. These results suggest that MAP3K10 decreased the phosphorylation of GSK3b substrates. Perhaps the phosphorylation gemcitabine sensitivity of pancreatic cancer cells (Fig. 5, p < 0.05). of GLI by GSK3b may require the priming phosphorylation by DYRK2. In addition, DYRK2 and GSK3b both have many substrates and overlap with the other signaling pathways. The role of 3.6. MAP3K10 upregulated Gli-1 and Gli-2 in pancreatic cancer cells MAP3K10 in pancreatic cancer cells cannot be explained solely by its effect on the Shh pathway, as it is involved in C-JNK and To elucidate the underlying mechanisms by which MAP3K10 p38-MAPK signaling [22]. The mechanisms of MAP3K10 in tumor- promoted the proliferation and viability of pancreatic cancer cells, igenesis are likely to be complex and needs further investigation. we investigated the expression of several key proteins in the Shh In summary, we showed that MAP3K10 played a key role in the pathway in MAP3K10 knockdown BxPC-3 cells and MAP3K10 over- proliferation and gemcitabine resistance, but not in the migration, expression PANC-1 cells. MAP3K10 overexpression suppressed the of pancreatic cancer cells. A possible mechanism of these effects mRNA levels of DYRK2 and GSK3b, but increased Gli-1 and Gli-2 may be through the upregulation of Gli-1 and Gli-2 by MAP3K10. expression. In contrast, knockdown of MAP3K10 resulted in en- Our data suggest that MAP3K10 is a potential target for the devel- hanced mRNA levels of DYRK2 and GSK3b, but decreased Gli-1 opment of alternative therapeutic strategy for pancreatic cancers. and G-li-2 mRNA levels (Fig. 6a–d, p < 0.05). Similar changes were observed for the proteins of these genes (Fig. 6e). These data indi- cated that MAP3K10 maybe promoted the Shh pathway in pancre- Acknowledgement atic cancer cells. This work was supported by grants from the National Nature Science Foundation of China (no. 81170336 and 30972912) and 4. Discussion the Postgraduate Innovation Project of Jiangsu Province, China (no. JX 22013186). In this study, we demonstrated that MAP3K10 was generally overexpressed in PDAC tissues and cell lines. In addition, our in vitro data showed that MAP3K10 promoted the viability and References proliferation, as well as decreased the gemcitabine sensitivity, of [1] D. Li, K. Xie, R. Wolff, J.L. Abbruzzese, Pancreatic cancer, Lancet 363 (2004) pancreatic cancer cells. These results suggest that MAP3K10 plays 1049–1057. an important role in the tumorigenesis and survival of pancreatic [2] P.A. Beachy, S.S. Karhadkar, D.M. Berman, Tissue repair and stem cell renewal cancers. in carcinogenesis, Nature 432 (2004) 324–331. [3] M.P. Fogarty, J.D. Kessler, R.J. Wechsler-Reya, Morphing into cancer: the role of The survival of human pancreatic cancer depends on the Shh developmental signaling pathways in brain tumor formation, J. Neurobiol. 64 pathway [24]. MAP3K10 activates the Shh pathway by increasing (2005) 458–475. the level of GLI proteins [21]. In our study, we found that MAP3K10 [4] M. Hammerschmidt, A. Brook, A.P. McMahon, The world according to hedgehog, Trends Genet. 13 (1997) 14–21. increased GLI1 and GLI2 mRNA and protein levels, indicating that [5] M. Varjosalo, J. Taipale, Hedgehog: functions and mechanisms, Genes Dev. 22 MAP3K10 promotes pancreatic cancer cell malignant transforma- (2008) 2454–2472. tion maybe by activating the Shh pathway. It was reported that [6] D. Kalderon, Transducing the hedgehog signal, Cell 103 (2000) 371–374. [7] A.P. McMahon, More surprises in the Hedgehog signaling pathway, Cell 100 MAP3K10 modulated the activity of the Shh pathway via regula- (2000) 185–188. tion of DYRK2 and GSK3b [21]. In accordance, we found that [8] S.P. Thayer, M.P. di Magliano, P.W. Heiser, C.M. Nielsen, D.J. Roberts, G.Y. MAP3K10 decreased both the mRNA and protein levels of DYRK2 Lauwers, et al., Hedgehog is an early and late mediator of pancreatic cancer and GSK3b. tumorigenesis, Nature 425 (2003) 851–856. [9] D.M. Berman, S.S. Karhadkar, A. Maitra, R. Montes De Oca, M.R. Gerstenblith, K. DYRK2 is a member of an evolutionarily conserved family of Briggs, et al., Widespread requirement for Hedgehog ligand stimulation in DYRKs, a scaffold for an E3 ubiquitin ligase complex, and controls growth of digestive tract tumours, Nature 425 (2003) 846–851. mitotic transition [25]. A study revealed that DYRK2 directly phos- [10] G. Feldmann, N. Habbe, S. Dhara, S. Bisht, H. Alvarez, V. Fendrich, et al., Hedgehog inhibition prolongs survival in a genetically engineered mouse phorylated and induced degradation of GLI2 [21]. Furthermore, re- model of pancreatic cancer, Gut 57 (2008) 1420–1430. cent studies demonstrated that DYRK2 regulated tumor [11] R.L. Yauch, G.J. Dijkgraaf, B. Alicke, T. Januario, C.P. Ahn, T. Holcomb, et al., progression through modulation of c-Jun and c-Myc [26] and in- Smoothened mutation confers resistance to a Hedgehog pathway inhibitor in medulloblastoma, Science 326 (2009) 572–574. duced apoptosis in response to DNA damage by regulating p53 [12] J.P. Morton, M.E. Mongeau, D.S. Klimstra, J.P. Morris, Y.C. Lee, Y. Kawaguchi, [27]. DYRK2 was proposed to be a favorable prognostic marker in et al., Sonic hedgehog acts at multiple stages during pancreatic tumorigenesis, pulmonary adenocarcinoma [28], as well as a predictive marker Proc. Natl. Acad. Sci. USA 104 (2007) 5103–5108. [13] J.M. Bailey, A.M. Mohr, M.A. Hollingsworth, Sonic hedgehog paracrine for chemotherapy in non-small cell lung cancer [29]. These results signaling regulates metastasis and lymphangiogenesis in pancreatic cancer, collectively suggest that DYRK2 act as a tumor suppressor, which is Oncogene 28 (2009) 3513–3525. in accordance with our ﬁndings. We found that MAP3K10 overex- [14] S.G. Chun, W. Zhou, N.S. Yee, Combined targeting of histone deacetylases and hedgehog signaling enhances cytoxicity in pancreatic cancer, Cancer Biol. Ther. pression enhanced the proliferation and reduced the gemcitabine 8 (2009) 1328–1339. sensitization of pancreatic cancer cells, accompanied by decreased [15] G. Feldmann, S. Dhara, V. Fendrich, D. Bedja, R. Beaty, M. Mullendore, et al., expression of DYRK2. Blockade of hedgehog signaling inhibits pancreatic cancer invasion and GSK3b is a member of GSK3 family [30]. Accumulating evi- metastases: a new paradigm for combination therapy in solid cancers, Cancer Res. 67 (2007) 2187–2196. dences have shown that GSK3b is a convergence of multiple signal [16] S. Nagai, M. Nakamura, K. Yanai, J. Wada, T. Akiyoshi, H. Nakashima, et al., Gli1 transduction pathways deregulated in cancers, such as the Wnt contributes to the invasiveness of pancreatic cancer through matrix pathway [31], the PI3K pathway [32–35], the hedgehog pathway metalloproteinase-9 activation, Cancer Sci. 99 (2008) 1377–1384. [17] J. Dai, K. Ai, Y. Du, G. Chen, Sonic hedgehog expression correlates with distant [36] and the Notch pathway [37–39]. It is accepted that GSK3b is metastasis in pancreatic adenocarcinoma, Pancreas 40 (2011) 233–236. a primary regulator in pancreatic cancers [40–44]. In our study [18] K.P. Olive, M.A. Jacobetz, C.J. Davidson, A. Gopinathan, D. McIntyre, D. Honess, we found that MAP3K10 overexpression increased cell prolifera- et al., Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer, Science 324 (2009) 1457–1461. tion, reduced gemcitabine sensitization and decreased GSK3b pro- [19] E. Lonardo, P.C. Hermann, M.T. Mueller, S. Huber, A. Balic, I. Miranda-Lorenzo, tein levels, suggesting that DYRK2 may play a major role in et al., Nodal/activin signaling drives self-renewal and tumorigenicity of Y. An et al. / Cancer Letters 329 (2013) 228–235 235

pancreatic cancer stem cells and provides a target for combined drug therapy, [34] B.W. Doble, J.R. Woodgett, Exploring pluripotency with chemical genetics, Cell Cell Stem Cell 9 (2011) 433–446. Stem Cell 4 (2009) 98–100. [20] P.M. LoRusso, C.M. Rudin, J.C. Reddy, R. Tibes, G.J. Weiss, M.J. Borad, et al., [35] T. Otto, S. Horn, M. Brockmann, U. Eilers, L. Schuttrumpf, N. Popov, et al., Phase I trial of hedgehog pathway inhibitor vismodegib (GDC-0449) in Stabilization of N-Myc is a critical function of Aurora A in human patients with refractory, locally advanced or metastatic solid tumors, Clin. neuroblastoma, Cancer Cell 15 (2009) 67–78. Cancer Res. 17 (2011) 2502–2511. [36] J. Jia, K. Amanai, G. Wang, J. Tang, B. Wang, J. Jiang, Shaggy/GSK3 antagonizes [21] M. Varjosalo, M. Bjorklund, F. Cheng, H. Syvanen, T. Kivioja, S. Kilpinen, et al., Hedgehog signalling by regulating cubitus interruptus, Nature 416 (2002) Application of active and kinase-deficient kinome collection for identification 548–552. of kinases regulating hedgehog signaling, Cell 133 (2008) 537–548. [37] L. Espinosa, J. Ingles-Esteve, C. Aguilera, A. Bigas, Phosphorylation by glycogen [22] K.A. Gallo, G.L. Johnson, Mixed-lineage kinase control of JNK and p38 MAPK synthase kinase-3 beta down-regulates notch activity, a link for Notch and pathways, Nat. Rev. Mol. Cell Biol. 3 (2002) 663–672. Wnt pathways, J. Biol. Chem. 278 (2003) 32227–32235. [23] Z. Xu, A.C. Maroney, P. Dobrzanski, N.V. Kukekov, L.A. Greene, The MLK family [38] K. Uemura, A. Kuzuya, Y. Shimozono, N. Aoyagi, K. Ando, S. Shimohama, et al., mediates c-Jun N-terminal kinase activation in neuronal apoptosis, Mol. Cell GSK3beta activity modifies the localization and function of presenilin 1, J. Biol. Biol. 21 (2001) 4713–4724. Chem. 282 (2007) 15823–15832. [24] S. Jones, X. Zhang, D.W. Parsons, J.C. Lin, R.J. Leary, P. Angenendt, et al., Core [39] Y.H. Jin, H. Kim, M. Oh, H. Ki, K. Kim, Regulation of Notch1/NICD and Hes1 signaling pathways in human pancreatic cancers revealed by global genomic expressions by GSK-3alpha/beta, Mol. Cells 27 (2009) 15–19. analyses, Science 321 (2008) 1801–1806. [40] A.V. Ougolkov, M.E. Fernandez-Zapico, D.N. Savoy, R.A. Urrutia, D.D. Billadeau, [25] S. Maddika, J. Chen, Protein kinase DYRK2 is a scaffold that facilitates assembly Glycogen synthase kinase-3beta participates in nuclear factor kappaB- of an E3 ligase, Nat. Cell Biol. 11 (2009) 409–419. mediated gene transcription and cell survival in pancreatic cancer cells, [26] N. Taira, R. Mimoto, M. Kurata, T. Yamaguchi, M. Kitagawa, Y. Miki, et al., Cancer Res. 65 (2005) 2076–2081. DYRK2 priming phosphorylation of c-Jun and c-Myc modulates cell cycle [41] A.V. Ougolkov, M.E. Fernandez-Zapico, V.N. Bilim, T.C. Smyrk, S.T. Chari, D.D. progression in human cancer cells, J. Clin. Invest. 122 (2012) 859–872. Billadeau, Aberrant nuclear accumulation of glycogen synthase kinase-3beta [27] N. Taira, K. Nihira, T. Yamaguchi, Y. Miki, K. Yoshida, DYRK2 is targeted to the in human pancreatic cancer: association with kinase activity and tumor nucleus and controls p53 via Ser46 phosphorylation in the apoptotic response dedifferentiation, Clin. Cancer Res. 12 (2006) 5074–5081. to DNA damage, Mol. Cell. 25 (2007) 725–738. [42] W. Wilson 3rd, A.S. Baldwin, Maintenance of constitutive IkappaB kinase [28] S. Yamashita, M. Chujo, K. Tokuishi, K. Anami, M. Miyawaki, S. Yamamoto, activity by glycogen synthase kinase-3alpha/beta in pancreatic cancer, Cancer et al., Expression of dual-specificity tyrosine-(Y)-phosphorylation-regulated Res. 68 (2008) 8156–8163. kinase 2 (DYRK2) can be a favorable prognostic marker in pulmonary [43] W. Mai, K. Kawakami, A. Shakoori, S. Kyo, K. Miyashita, K. Yokoi, et al., adenocarcinoma, J. Thorac. Cardiovasc. Surg. 138 (2009) 1303–1308. Deregulated GSK3{beta} sustains gastrointestinal cancer cells survival by [29] S. Yamashita, M. Chujo, T. Moroga, K. Anami, K. Tokuishi, M. Miyawaki, et al., modulating human telomerase reverse transcriptase and telomerase, Clin. DYRK2 expression may be a predictive marker for chemotherapy in non-small Cancer Res. 15 (2009) 6810–6819. cell lung cancer, Anticancer Res. 29 (2009) 2753–2757. [44] B. Marchand, I. Tremblay, S. Cagnol, M.J. Boucher, Inhibition of glycogen [30] N. Embi, D.B. Rylatt, P. Cohen, Glycogen synthase kinase-3 from rabbit skeletal synthase kinase-3 activity triggers an apoptotic response in pancreatic cancer muscle. Separation from cyclic-AMP-dependent protein kinase and cells through JNK-dependent mechanisms, Carcinogenesis 33 (2012) 529–537. phosphorylase kinase, Eur. J. Biochem. 107 (1980) 519–527. [45] Y.L. Woods, P. Cohen, W. Becker, R. Jakes, M. Goedert, X. Wang, et al., The [31] B.W. Doble, J.R. Woodgett, GSK-3: tricks of the trade for a multi-tasking kinase, kinase DYRK phosphorylates protein-synthesis initiation factor eIF2Bepsilon at J. Cell Sci. 116 (2003) 1175–1186. Ser539 and the microtubule-associated protein tau at Thr212: potential role [32] P.S. Knoepfler, A.M. Kenney, Neural precursor cycling at sonic speed: N-Myc for DYRK as a glycogen synthase kinase 3-priming kinase, Biochem. J. 355 pedals, GSK-3 brakes, Cell Cycle 5 (2006) 47–52. (2001) 609–615. [33] M. Bechard, S. Dalton, Subcellular localization of glycogen synthase kinase [46] Y. Gwack, S. Sharma, J. Nardone, B. Tanasa, A. Iuga, S. Srikanth, et al., A genome- 3beta controls embryonic stem cell self-renewal, Mol. Cell Biol. 29 (2009) wide Drosophila RNAi screen identifies DYRK-family kinases as regulators of 2092–2104. NFAT, Nature 441 (2006) 646–650.