MAML1 regulates cell viability via the NF-κB pathway in cervical cancer cell lines

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

available at www.sciencedirect.com

www.elsevier.com/locate/yexcr

Research Article

MAML1 regulates cell viability via the NF-κB pathway in cervical cancer cell lines Yanin Kuncharina , Naunpun Sangphechb , Patipark Kueanjindaa , Parvapan Bhattarakosola, c , Tanapat Palagad,⁎ a

Medical Microbiology Interdisciplinary Program, Graduate School, Chulalongkorn University, Payathai Road, Pathumwan, Bangkok 10330, Thailand b Biotechnology Program, Faculty of Science, Chulalongkorn University, Payathai Road, Pathumwan, Bangkok 10330, Thailand c Department of Microbiology, Faculty of Medicine, Chulalongkorn University, Payathai Road, Pathumwan, Bangkok 10330, Thailand d Department of Microbiology, Faculty of Science, Chulalongkorn University, Payathai Road, Pathumwan, Bangkok 10330, Thailand

A R T I C L E I N F O R M A T I O N

A B S T R A C T

Article Chronology:

The Notch signaling pathway plays important roles in tumorigenesis in a context-dependent

Received 25 November 2010

manner. In human cervical cancer, alterations in Notch signaling have been reported, and both

Revised version received 7 May 2011

tumor-suppressing and tumor-promoting roles of Notch signaling have been proposed; however,

Accepted 9 May 2011

the precise molecular mechanisms governing these roles in cervical cancer remain controversial.

Available on line 24 May 2011

MAML is a transcriptional co-activator originally identified by its role in Notch signaling. Recent evidence suggests that it also plays a role in other signaling pathways, such as the p53 and β-

Keywords: Notch signaling MAML Dominant-negative Cervical cancer NF-κB

catenin pathways. MAML is required for stable formation of Notch transcriptional complexes at the promoters of Notch target genes. Chromosomal translocations affecting MAML have been shown to promote tumorigenesis. In this study, we used a truncated dominant-negative MAML1 (DNMAML) to investigate the role of MAML in HPV-positive cervical cancer cell lines. Three human cervical cancer cell lines (HeLa, SiHa and CaSki) expressed all Notch receptors and the Notch target genes Hes1 and MAML1. Among these 3 cell lines, constitutive appearance of cleaved Notch1 was found only in CaSki cells, which suggests that Notch1 is constitutively activated in this cell line. Gamma secretase inhibitor (GSI) treatment, which suppresses Notch receptor activation, completely abrogated this form of Notch1 but had no effect on cell viability. Overexpression of DN-MAML by retroviral transduction in CaSki cells resulted in significant decreases in the mRNA levels of Hes1 and Notch1 but had no effects on the levels of MAML1, p53 or HPV E6/E7. DN-MAML expression induced increased viability of CaSki cells without any effect on cell cycle progression or cell proliferation. In addition, clonogenic assay experiments revealed that overexpression of DNMAML resulted in increased colony formation compared to the overexpression of the control vector. When the status of the NF-κB pathway was investigated, CaSki cells overexpressing DNMAML exhibited loss of phospho-IκBα, decreased total IκBα and nuclear localization of NF-κB p65, which suggests that the NF-κB pathway is hyperactivated. Furthermore, increased level of cleaved Notch1 was detected when DN-MAML was expressed. When DN-MAML-overexpressing cells

⁎ Corresponding author at: Department of Microbiology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand. Fax: + 66 2 2527576. E-mail address: [email protected] (T. Palaga). Abbreviations: MAML, Mastermind-like; DN, dominant-negative; GSI, gamma secretase inhibitor; HPV, Human papilloma virus.

0014-4827/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2011.05.005

1831

E XP E RI ME N T AL C E L L R E S EA RC H 31 7 ( 20 1 1) 1 8 3 0– 1 84 0

were treated with GSI, significantly decreased cell viability was observed, indicating that inhibition of Notch signaling using GSI treatment and DN-MAML expression negatively affects cell viability. Taken together, targeting Notch signaling using DN-MAML and GSI treatment may present a novel method to control cell viability in cervical cancer cells. © 2011 Elsevier Inc. All rights reserved.

Introduction The evolutionarily conserved Notch signaling pathway plays important roles in cell fate determination, cell proliferation and apoptosis during the development of vertebrates [1]. Aberrations in the activation of this signaling pathway have been shown to be involved in tumor formation of tissues from various origins [2]. Notch signaling is initiated by specific interactions between Notch receptors and ligands, resulting in enzymatic cleavage of the Notch receptors. Gamma secretase is responsible for the cleavage that releases intracellular Notch receptor, thereby permitting it to enter the nucleus. In the nucleus, intracellular Notch receptor forms a complex with a DNA binding protein, CSL, at the promoters of its target genes. This complex, together with other recruited transcriptional coactivators, drives the transcription of target genes. This signaling cascade is called the “canonical” Notch signaling pathway and is dependent on CSL [3]. Depending on the context, Notch signaling can either promote or inhibit tumor formation [2]. In human cervical cancer, alterations in Notch signaling have been shown to be associated with abnormal cell fates in epithelial neoplasia [4]. Previous studies have extensively demonstrated that the components of the Notch signaling pathway can act as promoters or suppressors of HPVmediated cervical cancer [5–10]. As a tumor suppressor, Notch signaling is involved in the cross-regulation of HPV E6/E7 because down-regulation of Notch1 is important for sustained expression of viral E6/E7 [5]. Furthermore, overexpression of constitutively active Notch1 has been shown to result in growth arrest of HPV + cervical cancer cells and HeLa cells [11,12]. In contrast, Notch signaling cooperates with other signaling pathways, such as the NF-κB, MAPK and PI3K/AKT pathways, to promote malignancy in cervical cancer cell lines [7,10,13]. Therefore, the roles that Notch signaling plays in cervical cancer remain controversial. The MAML proteins (MAML1-3) comprise a family of transcriptional co-activators and were first identified by their association with the Notch signaling pathway [14,15]. The N-terminal basic domain of MAML is responsible for interacting with the ankyrin repeats in the intracellular domain of Notch receptors. MAML contains two transactivation domains that are involved in the recruitment of transcriptional co-activators, such as CBP/p300 [16]. Studies of the structure of MAML upon interacting with Notch/CSL on the promoters of target genes have revealed that MAML is required for the formation of stable ternary Notch transcription complexes [17,18]. A truncated form of MAML1 containing only the N-terminal region of MAML1 (amino acids 12–74) was used as a dominantnegative (DN) form of MAML that effectively suppressed expression of the target genes of Notch signaling [19]. Previous studies have shown that DN-MAML overexpression often phenocopies Notchdeficient phenotypes in immune cells and tumors [15,19,20]. Besides acting as a binding partner with the Notch/CSL transcriptional complex, a recent study also pointed toward a broader role for MAML because it interacts with p53, MEF2c and β-catenin [16]. MAML1 has been shown to enhance p53-mediated gene expression by increas-

ing the stability of the p53 protein and by enhancing phosphorylation and acetylation of p53, independent of Notch signaling [21]. Interestingly, a chromosomal translocation involving MAML2, which disrupted Notch signaling, was reported in mucoepidermoid carcinoma [22]. Therefore, MAML family proteins have emerged as critical factors in tumorigenesis through Notch-dependent and Notch-independent manners [15]. The role of MAML in cervical cancer, however, has not been directly addressed. In this study, we used DN-MAML to investigate the effects of disrupting MAML function in cervical cancer cell lines. We report that MAML may have a protective role in the progression of cervical cancer, partly by regulating the crosstalk between the Notch and NF-κB pathways. In addition, we demonstrated that targeting Notch signaling using GSI and DN-MAML is effective at decreasing the viability of cervical cancer cell lines. These results indicate the potential for the use of similar agents as novel therapies for cervical cancer.

Materials and methods Cell lines and reagents HPV-positive cervical cancer cell lines, HeLa and SiHa, were maintained in Eagle's modified essential medium, and CaSki cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum (v/v) (HyClone, USA), 100 U/ml penicillin (General Drugs House Co., Ltd., Thailand), 0.4 mg/ml streptomycin (M & H Manufacturing Co., Ltd., Thailand), 1% sodium pyruvate and 1% HEPES. All cell lines were obtained from ATCC (Manassas, VA, USA). A gamma secretase inhibitor (GSI), DAPT (Sigma-Aldrich, MO, USA), was dissolved in DMSO and used as previously described [23]. DMSO was used as a mock vehicle control when DAPT was used to treat cells.

Transfection and retroviral transduction The retroviral plasmid vector for expression of DN-MAML [MSCVMam(12–74)-EGFP] was a kind gift from Dr. Warren Pear (University of Pennsylvania, USA). A control empty vector, MSCV-IRES-GFP (plasmid 20672), was obtained from Addgene (USA). The retroviral vectors and packaging construct pCL-Ampho (Imagenex) were cotransfected into 293 T cells using the FuGene® HD transfection reagent (Roche, USA) according to the manufacturer's instructions. Culture supernatants containing retroviruses were harvested twice at 48 and 72 h after transfection and were used to transduce CaSki cells, as described elsewhere. Transduction efficiency was confirmed by fluorescent microscopy and flow cytometry.

Western blot To monitor protein levels of Notch1 and cleaved Notch1, cells were treated as indicated, and cell lysates were prepared using lysis

1832

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

buffer, as previously described [23]. For detecting phosphoproteins in the NF-κB pathway, RIPA buffer with the addition of a phophatase inhibitor cocktail was used to prepare cell lysates (SigmaAldrich, USA). Notch1, Notch2 and cleaved Notch1 (Val1744) were detected using rabbit antibodies for Notch1 (C20; Santa Cruz Biotechnology Inc., USA), Notch2 and cleaved Notch1 (Cell Signaling Technology, USA), respectively. Hes1 was detected by rabbit monoclonal antibody for Hes1 (AB15470; Millipore). Antibodies for NF-κB p65, IKKα/β and IκBα were all obtained from Cell Signaling Technology. Secondary antibodies against rabbit IgG conjugated to horseradish peroxidase were obtained from GE Healthcare (UK). Signals were detected by the chemiluminescent method.

RT-PCR and quantitative RT-PCR Total RNA was isolated from cells (treated as indicated) using TriZol reagent (Invitrogen, USA). cDNA was prepared using reverse transcriptase (Fermentas, USA) and random hexamers (Invitrogen). PCR reactions were conducted using primers specific for human Notch1, Notch2, Notch3, Notch4, Hes1, MAML1 and TP53. β-actin was used as a loading control. The forward and reverse primers used for PCR amplification are as follows: human Notch1 (5′-CAGCCTGCACAACCAGACAGA-3′ and 5′-TGAGTTGATGAGG TCCTCCA G-3′); Notch2 (5′-TGAGTAGGCTCCATCCAGTC-3′ and 5′-TGGTGTCAG GTAGGGAT GCT-3′); Notch3 (5′-TCTTGCTGCTGGTCATTCTC-3′ and 5′-TGCCTCAT CCTCTTCAGTTG-3′); Notch4 (5’-CACTGAGCCAAGGCATAGAC-3′ and 5′-ATCTCCA CCTCACACCACTG-3′); Hes1 (5′-ACGACACCGGATAAACCAAA-3′ and 5′-CGGAGG TGCTTCACTGTCAT-3′); MAML1 (5′-CAGCATCAGTTGCTTTTGGA-3′ and 5′-CTG CTCTGAGGCATGTTTTG-3′); and β-actin (5′-ACCAACTGGGACGACATGGAGAA3′ and 5′-GTGG TGGTGAAGCTGTAGCC-3′). PCR reactions were carried out using Bioer Life Express®. PCR products were analyzed by electrophoresis on 2% agarose gels and visualized with a Gel Documentation System with Quantity One 4.4.1 software (Bio-Rad, USA). For quantitative RT-PCR (qPCR), cDNA was prepared as described above. The qPCR amplifications were performed with the 1× Maxima™ SYBR Green/ROX qPCR Master Mix (Fermentas, USA) according to the manufacturer's protocol. The forward and reverse primers used for qPCR amplification are as follows: human Notch1 (5′-CAGCCTGCACAACCAGACAGA-3′ and 5′-TGAGTTGATGAGGTCCTCCAG-3′); Hes1 (5′-ACGACACCGGATAAACCAAA-3′ and 5′-CGGAGGTGCTTCACTGTCAT-3′); MAML1 (5′-CAGCATCAGTTGCTTTTGGA-3′ and 5′-CTGCTCTGAGGCATGTTTTG-3′); TP53 (5′-TCCACTACAACTACATGTGTAAC-3′ and 5′-GTGAAATATTCTCCATC CAGTG-3′); HPV16 E6 (5′-TCAAAAGCCACTGTGTCCTGA-3′ and 5′-CGTGTTCTT GATGATCTGCAA-3′); HPV16 E7 (5′-ATGACAGCTCAGAGGAG GAG-3′ and 5′-TCCTAGTGTGCCCATTAACAG-3′); and β-actin (5′-ACCAACTGGGACGACATGGA GAA-3′ and 5′-GTGGTGGT GAAGCTGTAGCC-3′). β-actin was used as a reference gene. qPCRs were carried out using the MJ Mini Personal Thermal cycler (BioRad, USA). The relative expression levels were calculated and analyzed by 2-ΔΔCP.

Immunofluorescent staining CaSki cells were transduced with retroviruses as described above, were fixed in 4% paraformaldehyde for 10 min at room temperature and permeabilized in PBS with 0.2% Triton X-100

for 2 min. After washing with PBS, cells were stained with rabbit anti-NF-κB p65 (C22B4) and anti-rabbit IgG (H + L) F(ab')2 fragments conjugated to Alexa Fluor® 555 (Cell Signaling Technology, USA). After mounting in confocal microscope mounting media, the cells were observed under an inverted fluorescent microscope.

MTT assays, cell cycle analyses, clonogenic assays and cell proliferation assays For cell viability assays, cells were treated as indicated and then subjected to MTT assays, as previously described [23]. For cell cycle analyses, treated cells were fixed in 70% cold ethanol at 4 °C for 4 h and then were treated with RNaseA (10 mg/ml) at 37 °C for 30 min. After washing, cells were stained with propidium iodide (1 mg/ml; Sigma-Aldrich, USA) for 30 min in the dark and subjected to FACS analysis (BD Biosciences, USA). Acquired data were analyzed with the Summit 5.0 software program (Dako Colorado Inc., USA). For clonogenic assays, cells were treated as indicated and plated at cell densities of 300 or 400 cells/well, in triplicate, in 12-well plates. After allowing the cells to form colonies by incubation in drug-free medium for 10 days, the resulting colonies were stained with 2% crystal violet in methanol and were counted. For cell proliferation assay, retrovirally transduced cells were labeled with Cell Proliferation Dye eFluor® 670 (eBioscience) according to manufacturer's instruction. After 4 days of incubation, cells were fixed in 4% paraformaldehyde and analyzed by FACS analysis. Acquired data were analyzed by FlowJo flow cytometry analysis software. Proliferation index is the average number of divisions undergone in all cells.

Statistical analyses All data were analyzed by independent t-tests using SPSS software. A p value of less than 0.05 was considered statistical significance.

Results Expression of Notch, MAML1 and Hes-1 in human cervical cancer cell lines We first examined the expression profiles of Notch receptors, MAML1 and Hes1 in three human cervical cancer cell lines that are HPV-positive. As shown in Fig. 1A, all cell lines tested in this study expressed Notch1, 2 and 3; MAML1; and Hes1 but at different levels. Notch4 mRNA was not detected in any of these cell lines (data not shown). We next examined Notch1 protein levels in these cell lines and found that Notch1 was present in all of them. Interestingly, CaSki cell lysates displayed a pattern of two distinct bands with sizes that corresponded to the size of Notch1 (~ 118 and 114 kDa) (Fig. 1B). When we used an antibody specific for cleaved Notch1 (Val1744), we could detect this active form of Notch1 only in CaSki cells (Fig. 1C). These results strongly suggest that in all cervical cell lines that we tested, Notch receptors are expressed and might be activated because one of the target genes of the Notch signaling pathway, Hes1, is detectable. In addition, MAML1 was also expressed in all three cell lines. More importantly, Notch1

1833

-ve

SiHa

CaSki

A

HeLa

Jurkat

E XP E RI ME N T AL C E L L R E S EA RC H 31 7 ( 20 1 1) 1 8 3 0– 1 84 0

B

HeLa

SiHa

CaSki Notch1 118 kDa Notch1 114 kDa β-actin

C

HeLa

SiHa

CaSki cleaved Notch1 (Val1744) β-actin

Fig. 1 – Expression profiles of Notch1-3, MAML1 and Hes1 mRNA and protein Notch1 in human cervical cancer cell lines. (A) RNA from the indicated cell lines was analyzed for mRNA expression by RT-PCR. Jurkat mRNA was used as a positive control. (B–C) Cell lysates from HeLa, SiHa and CaSki cells were analyzed for the presence of Notch1 (B) and cleaved Notch1 (Val1744) (C) by Western blot.

A

B

HeLa

CaSki

Notch1

Notch2

β-actin

β-actin

CaSki

SiHa Cleaved Notch1 (Val1744) β-actin

C

Notch1 (118 kDa) expression

D

Notch1 (114 kDa) expression

Fig. 2 – GSI treatment and activation of Notch receptors. (A). Human cervical cancer cell lines were treated with DAPT (50 μM) or vehicle control (DMSO) for 4 days. Cell lysates were used to detect Notch1 and cleaved Notch1 (Val1744) by Western blot. (B) CaSki cells were treated in a manner similar to the method described in panel (A). Cell lysates were analyzed for Notch2 expression by Western blot. (C–D) Intensities of the bands corresponding to ~118 kDa (C) and ~114 kDa (D) from the blots obtained in (A) were quantified by Quantity One Software.

1834

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

was constitutively active in CaSki cells but not in the other two cell lines.

Notch1 in CaSki cells completely disappeared upon GSI treatment. This result indicates that DAPT effectively inhibited gamma secretase enzymatic activity, and cleavage of Notch1 in CaSki cells is dependent on gamma secretase activity. Interestingly, DAPT treatment did not affect cell viability in all cell lines tested, from the MTT assays (Fig. 6). Because CaSki cells display constitutively active Notch1 and GSI treatment removed cleaved Notch1, these cells were used to further explore the role of Notch signaling and MAML.

Effects of GSI on human cervical cancer cell lines Gamma secretase, a multi-subunit enzyme, plays a crucial role in initiating Notch signaling by mediating the cleavage of Notch receptors upon receptor–ligand engagement. All Notch receptors share a common enzyme complex for their processing, and treatment with inhibitors specific for gamma secretase can circumvent the problem of functional redundancy among Notch receptors. We used a GSI, DAPT, to target this enzyme, and we examined its effects on cervical cancer cell lines. Upon DAPT treatment, the levels of Notch1 in HeLa and SiHa cells remained the same, while in CaSki cells, the levels of Notch1 (114 kDa) slightly increased (Figs. 2A and C). An increase in Notch1 levels upon GSI treatment is probably the result of the accumulation of uncleaved Notch1 on the cell surface. When the intensity of the bands at 118 and 114 kDa were measured, we found that GSI treatment did not affect the intensity of band at 118 kDa in all cell lines but did cause an increase in the 114 kDa band in CaSki cells only (Figs. 2C and D). This trend was also observed when Notch2 was detected in CaSki cells upon treatment with DAPT (Fig. 2B). As expected, cleaved

A

Effects of DN-MAML overexpression on gene expression in CaSki cells To interfere with the function of endogenous MAML, DN-MAML was transiently overexpressed in CaSki cells by retroviral transduction. As shown in Fig. 3A, more than 95% of the cells were GFP + when they were transduced with either the control vector or the DN-MAML-containing vector. We next examined the gene expression profiles of CaSki cells overexpressing DN-MAML. As expected, Hes1 expression levels were significantly decreased, indicating that Notch signaling-mediated gene expression is efficiently suppressed by DN-MAML expression (Fig. 3B). Consistent with the transcript levels, the protein levels of Hes1 also

MSCV-IRES-GFP GFP+ 96% GFP+ 0.26%

MSCV-Mam(12-74)-EGFP Control untransduced cells

GFP+ 98%

Relative expression level normalized to β actin

B 2.0

1.6

MSCV-IRES-GFP MSCV-Mam(12-74)-EGFP

C

1.2

Hes1

0.8

β-actin

0.4

CaSki 0 Hes1 Notch1

MAML1

TP53 HPV E6 HPV E7

Fig. 3 – Overexpression of DN-MAML and its effects on gene expression in CaSki cells. (A) CaSki cells were retrovirally transduced with the control empty vector MSCV-IRES-GFP (upper) or MSCV-Mam(12–74)-EGFP (lower) for 4 days. Cells were analyzed for GFP fluorescence by FACS. Results are representative of three independent experiments. (B) Total RNA from retrovirally transduced CaSki cells was analyzed for mRNA expression by quantitative RT-PCR. Results are presented as means ± SD of triplicate determinations. * indicates statistical significance where p < 0.05. (C) Cell lysates from CaSki cells retrovirally transduced with the control empty vector or MSCV-Mam (12–74)-EGFP were analyzed for Hes1 by Western blot.

1835

E XP E RI ME N T AL C E L L R E S EA RC H 31 7 ( 20 1 1) 1 8 3 0– 1 84 0

decreased in DN-MAML-overexpressing cells (Fig. 3C). In addition, expression of Notch1 was also decreased by DN-MAML, implying that Notch1 itself may be under the regulation of MAML-mediated transcription control in CaSki cells. The levels of expression of MAML1, p53 and HPV E6 did not change after DN-MAML expression. The expression level of HPV E7 showed the trend of increasing upon DN-MAML expression but it did not reach statistical significance.

DN-MAML expression enhanced cell viability and NF-κB activation Previous reports have indicated that hyperactivation of Notch signaling affects cell survival and cell cycle progression in cervical cancer cells [11,12]. Therefore, we tested whether interfering with MAML function could affect cell viability in CaSki cells. As shown in Fig. 4A, DN-MAML expression resulted in significantly increased cell viability 4 days after transduction. This increase in cell viability was not the result of increased cell cycle progression because similar cell cycle profiles were observed in cells overexpressing DN-MAML and in control cells (Figs. 5A and B). Furthermore, the

A

percentages of apoptotic cells as indicated by the sub-G1 population in the cell cycle were similar between vector control and DN-MAML transducing cells, suggesting that decreasing apoptosis is not the primary cause of increasing cell viability in DN-MAML overexpressing CaSki cells (Fig. 5A). When clonogenic assays were carried out, a significant increase in the number of colonies that were formed was observed in DN-MAML-overexpressing cells (Figs. 4B and C). More than 90% of cells that formed colonies were GFP + (data not shown). Interestingly, colonies derived from the DN-MAML-overexpressing cells were smaller in size than those derived from the control cells (Fig. 4B). Indeed, increase in percentages of cells with smaller in size, as observed in FACS analysis as forward scatter of cells, in DN-MAML overexpressing CaSki cells were evident (Fig. 5C). We next investigated whether DN-MAML overexpression induced higher rate of proliferation in CaSki cells by cell proliferation dye. As shown in Fig. 4D, the patterns of cell proliferation between the control vector transducing cells and DN-MAML overexpressing cells were similar and there was no difference in the proliferation indices. This result implies that overexpressing DN-MAML did not affect the rate of proliferation.

C

200

Percent colony formation (%)

Percent cell viability (%)

140 120 100

MSCV-IRES-GFP

80

MSCV-Mam(12-74)-EGFP

60 40 20

180 160 120

CaSki

B

MSCV-Mam (12-74)EGFP

100 80 60 40 20 0

0

MSCV-IRES-GFP

140

CaSki

D Proliferation index = 2.11 Proliferation index = 2.12

MSCV-IRES-GFP

MSCV-Mam(12-74)-EGFP

MSCV-Mam(12-74)-EGFP

Cell Count

MSCV-IRES-GFP

Cell Proliferation Dye eFluoro® 670

Fig. 4 – Effects of DN-MAML expression on cell viability of CaSki cells. (A) Retrovirally transduced CaSki cells were subjected to MTT assays for cell viability. Results are presented as means ± SD of triplicate determinations. * indicates statistical significance where p < 0.05. (B) Colony formation by clonogenic assay of CaSki cells transduced with control empty vector or MSCV-Mam(12–74)-EGFP. (C) Clonogenic assays were carried out using retrovirally transfected cells. After 10 days, the colonies were counted. Results are presented as means ± SD of triplicate determinations. * indicates statistical significance where p < 0.05. (D) Cell proliferation assays were carried out by labeling CaSki cells transduced with the control empty vector or MSCV-Mam (12–74)-EGFP with cell proliferation Dye eFluor® 670. Cells were incubated for 4 days before analysis by FACS. The proliferation indices were calculated by FlowJo flow cytometry analysis software.

1836

B

A 12.92% 65.17%

15.03% 65.16%

19.82%

Untransduced CaSki

MSCV-IRES-GFP

0.57%

16.11% 62.66%

S

G1

Sub G1

MSCV-IRES-GFP

MSCVMam(12-74) -EGFP

C

21.12%

MSCV-IRES-GFP

MSCV-Mam(12-74)-EGFP

MSCV-Mam(12-74) -EGFP

Fig. 5 – Effects of DN-MAML expression on cell cycle progression. (A) Retrovirally transduced CaSki cells were subjected to cell cycle analysis using propidium iodine staining by FACS. Results are representative of two independent experiments. (B) Average percentages of cells in each phase of the cell cycle in (A) are plotted. (C) Forward and side scatter dot plots obtained from transduced CaSki cells were shown.

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

Untransduced CaSki

0.48%

G2/M

21.78%

Percent Cell Cycle (%)

0.41%

1837

E XP E RI ME N T AL C E L L R E S EA RC H 31 7 ( 20 1 1) 1 8 3 0– 1 84 0

We further explored which signaling pathways are responsible for the increasing cell viability of DN-MAML-overexpressing cells. The NF-κB pathway has been shown to play a vital role in tumor progression, including in cervical cancers; in addition, crosstalk with Notch signaling components has been previously reported [9,24]. Therefore, we examined the status of components of the NF-κB pathway in DN-MAML-overexpressing cells and found decreases in both the levels of phospho-IκBα and total IκBα (Fig. 6A). Furthermore, immunostaining of NF-κB p65 revealed that the protein mainly located in the cytoplasm of the control vector-transduced cells, while the signals were equally detected in both the nuclei and cytoplasms of DN-MAMLoverexpressing cells (Fig. 7). These results suggest that interfering with MAML function using DN-MAML results in hyperactivation of NF-κB, which may drive increased cell viability in CaSki cells. To explore the possible mechanism leading to NF-κB activation in DN-MAML overexpressing cells, we detected the level of

A

cleaved Notch1 (Val1744) which is an indicator of the ligandinduced cleavage of Notch1. As shown in Fig. 6B, overexpression of DN-MAML led to elevated level of cleaved Notch1. Since in DNMAML overexpressing cells, the transcriptions of Notch target genes (such as Hes1) decreased, increased level of cleaved Notch1 may suggest that it functions independent on the CSL/MAML transcription complex, possibly by forming a complex with IKK signalosome and activates NF-κB pathway [10]. Because DN-MAML overexpression and GSI treatment did not yield the exact same phenotype in terms of cell viability, we next determined whether targeting Notch signaling using two different approaches has a synergistic effect on cell viability. To accomplish this, we used GSI treatment to target the processing of the Notch receptor and DN-MAML overexpression to target the “canonical” CSL-mediated gene transcription. Surprisingly, we found that GSI treatment alone at a concentration of 25 μM in vector control transduced CaSki cells did not significantly alter cell viability, while the combination of GSI treatment and DN-MAML overexpression

C 105

106

101

96

113

Phospho-IκB-α

IκB-α

D β-actin

Percentage of viability

120

B

Cleaved Notch1

*

100 80 60

0.25% DMSO 25 μM DAPT

40 20 0

Notch 1 β-actin

Fig. 6 – Effects of DN-MAML expression on NF-κB activation and DN-MAML overexpression and GSI treatment on cell viability. (A) CaSki cells were retrovirally transduced with control empty vector or MSCV-Mam(12–74)-EGFP for 4 days. Cell lysates were analyzed for the presence of phospho-IκBα, total IκBα by Western blot. (B) CaSki cells were retrovirally transduced with control empty vector or MSCV-Mam(12–74)-EGFP for 4 days. Cell lysates were analyzed for the presence of cleaved Notch1 and total Notch1 by Western blot. (C) CaSki cells were treated with various concentrations of DAPT for 4 days, and cell viability was measured with MTT assay. Results are presented as means ± SD of triplicate determinations. (D) CaSki cells transduced with MSCV-IRES-GFP or MSCV-Mam(12–74)-EGFP were treated with DAPT (25 μM; light gray) or vehicle control (dark gray) for 2 days before analysis with MTT assay. Results are presented as means ± SD of triplicate determinations. The viability of untransduced cells not treated with DAPT was set at 100%. * indicates statistical significance where p < 0.05.

1838

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

NF-κB p65

GFP

Overlay

A

B

C

D

E

F

MSCV-IRESGFP

MSCV-Mam (12-74)-EGFP

Fig. 7 – Localization of NF-κB p65 in CaSki cells overexpressing DN-MAML. CaSki cells were retrovirally transduced with MSCV-IRES-GFP (A–C) or MSCV-Mam(12–74)-EGFP (D–F), as described above. Cells were fixed and stained with rabbit anti-NF-κB p65 antibody and anti-rabbit IgG conjugated to Alexa Fluor® 555 before observation by fluorescent microscope.

resulted in significantly reduced cell vitality (Fig. 6C). Similar, but not better, results were obtained with higher dose (50 μM) of GSI treatment (data not shown).

Discussion Notch signaling has long been associated with cervical cancer, but its precise roles in this type of cancer remain controversial. Aberrant Notch signaling cooperates with other signaling pathways, such as the MAPK, NF-κB and PI3K pathways, to give rise to invasive cervical cancer, which suggests that Notch signaling plays an oncogenic role in cervical cancer [9,25,26]. In contrast, Notch signaling during a certain stage of cervical cancer induces the down-modulation of HPV E6/E7 expression via decreased AP-1 activation, suggesting that it has a protective role in HPV-driven cervical cancer [11]. Because MAML is a major binding partner of Notch/CSL during activation of Notch signaling, we used DN-MAML to interfere with endogenous MAML function and to elucidate the relationship between MAML, Notch

GSI

DN-MAML

?

DN-MAML

Canonical Notch Signaling

Notch Receptor-Ligand Cleaved Notch

signaling and cervical cancer. DN-MAML used in this study was shown previously to suppress activation of CSL-mediated canonical Notch signaling [19]. Indeed, in CaSki cells, in which constitutively active Notch1 was observed, overexpression of DN-MAML significantly reduced Hes-1 mRNA expression and the protein levels. However, we did not test whether DN-MAML affects other signaling pathways that have been shown to use MAML as a co-activator. Among 3 cell lines tested, CaSki cells were the only cell line which showed two bands with the size corresponding to 114 and 118 kDa when Notch1 was detected. Furthermore, the appearance of cleaved Notch1 was also detected in this cell line, which was dependent upon the activity of gamma secretase. Interestingly, when Notch1 was detected in CaSki cells upon GSI treatment, the intensity of bands corresponding to 114 kDa increased while those corresponding to 118 kDa remained the same. This result led us to propose that the proteins with the size of 114 kDa detected in CaSki cells may represent Notch1 which is cleaved by TNFα converting enzyme (TACE). The resulting piece of Notch receptor is sequentially processed by gamma secretase. In the absence of

GSI

?

Net Effect

Cell Survival Or Cell Death

Other Signalings such as NF-κB Notch independent-MAML DN-MAML

Fig. 8 – Model of the relationship between the Notch signaling pathway, MAML and other signaling pathways as they relate to the survival of cervical cancer cell lines. See Discussion for details.

E XP E RI ME N T AL C E L L R E S EA RC H 31 7 ( 20 1 1) 1 8 3 0– 1 84 0

GSI, gamma secretase continuously cleave Notch1 which result in the appearance of cleaved Notch1. When GSI is present, however, the piece of Notch1 accumulates post cleavage by TACE. In CaSki cells, expression of DN-MAML resulted in higher cell viability than expression of the control vector, and increased colony formation was observed in clonogenic assays in cells expressing DN-MAML. Together with previous reports, these results imply that MAML, probably via Notch signaling, may have a protective role in cervical cancer progression. We did not find that DN-MAML expression had any effect on HPV E6/E7 or p53 mRNA levels. The increased cell viability in DN-MAML-expressing cells appears to be partially due to activation of the NF-κB pathway. We found decreases in total IκBα in DN-MAML-overexpressing CaSki cells, while phosphorylation of NF-κB p65 did not change in this setting. Increased nuclear localization of p65, however, was detected in DN-MAML-overexpressing cells. This result is consistent with previous reports suggesting that activation of Notch signaling in HeLa cells results in cell cycle arrest and apoptosis through down-regulation of NF-κB activation via decreased nuclear NF-κB p50 and increased cytosolic IκBα [8]. Interestingly, Song et al. demonstrated that crosstalk between Notch signaling and NF-κB is mediated by direct regulation of IKKα and IKK activity by Notch1 [10]. This report implied that Notch1 is associated with the IKK signalosome, and down-modulation of IKK affects CSL-mediated reporter gene activity and NF-κB reporter activity. Whether MAML is required for this Notch/IKKα transcriptional complex is currently not known. When DN-MAML is overexpressed in CaSki, we detected increased in the level of cleaved Notch1, despite decreased expression of Hes1. This observation, together with the report by Song et al., led us to propose that increased cleaved Notch1 in DN-MAML expressing CaSki activates NF-κB pathway which result in increased cell viability and colony formations. How expression of DN-MAML leading to more cleaved Notch1 requires further investigation. Recently, MAML was shown to also function as a transcriptional coactivator for other signaling pathways, such as the p53 and βcatenin pathways, suggesting that it may play a key role in the crosstalk among different signaling pathways [16]. Even though we did not detect any changes in p53 mRNA expression, DNMAML expression may affect the post-translational modifications of protein p53 which greatly affect the protein functions. Some of these novel functions of MAML are independent of Notch signaling. Thus, it is of great interest to further investigate whether the phenotypes obtained in our study are due to the inhibitory effects on Notch signaling or on other pathways. In this study, GSI treatment of CaSki cells did not result in changes in cell viability. Therefore, DN-MAML overexpression and GSI treatment in our studies did not yield similar results in terms of cell viability. This disparity might be caused by the ability of MAML to also function in other pathways besides Notch in this cell line; in addition, DN-MAML expression might interfere with those other functions that are independent of Notch signaling (Fig. 8). To this end, we did not detect any effects of DN-MAML expression on the mRNA expression levels of p53 and viral E6/E7. Interestingly, GSI treatment and DN-MAML overexpression together did reduce cell viability in CaSki cells. This result led us to propose the model depicted in Fig. 8. When GSI is used to inhibit Notch signaling, cleaved Notch receptors are not generated and the “canonical” Notch signaling pathway is suppressed, while

1839

other Notch-independent, MAML-dependent pathways may operate to keep cell viability in check. When DN-MAML is overexpressed, canonical Notch signaling and other Notchindependent MAML functions are suppressed, but increased cleaved Notch stimulates NF-κB activation without relying on the transcriptional activity of the Notch/CSL complex. The net effect of interfering with MAML function leads to activation of the NF-κB pathway, possibly by increased cleaved Notch, which results in increased cell viability. When both GSI treatment and DN-MAML are used, all the signaling (canonical Notch signaling/cleaved Notch/Notch-independent MAML-dependent signaling) positively regulating cell survival is suppressed, resulting in decreased cell viability. This model highlights the importance of “canonical” Notch signaling, which requires MAML, and the signaling pathways that might not involve MAML. In conclusion, interference with MAML function in human cervical cancer cell lines affected cell viability, partly via activation of the NF-κB pathway. Therefore, MAML may have a suppressive role in cervical cancers, possibly through crosstalk between the Notch and NF-κB pathways.

Acknowledgments This work (AS613A) was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission, the Rachadapiseksomphot Endowment Funds from Chulalongkorn University and the Thai Government Stimulus Package 2 (TKK2555) under the Project for Establishment of Comprehensive Center for Innovative Food, Health Products and Agriculture. The authors are grateful to Drs. Warren Pear, Barbara Osborne and Tannishtha Reya for sharing reagents and to Noppadol Sa-ard-iam and Supranee Buranapraditkul for their helps with the FACS analysis.

REFERENCES

[1] S. Artavanis-Tsakonas, M.D. Rand, R.J. Lake, Notch signaling: cell fate control and signal integration in development, Science 284 (1999) 770–776. [2] F. Radtke, K. Raj, The role of Notch in tumorigenesis: oncogene or tumour suppressor? Nat. Rev. Cancer 3 (2003) 756–767. [3] F. Radtke, N. Fasnacht, H.R. Macdonald, Notch signaling in the immune system, Immunity 32 (2010) 14–27. [4] P. Zagouras, S. Stifani, C.M. Blaumueller, M.L. Carcangiu, S. Artavanis-Tsakonas, Alterations in Notch signaling in neoplastic lesions of the human cervix, Proc. Natl Acad. Sci. U. S. A. 92 (1995) 6414–6418. [5] C. Talora, D.C. Sgroi, C.P. Crum, G.P. Dotto, Specific down-modulation of Notch1 signaling in cervical cancer cells is required for sustained HPV-E6/E7 expression and late steps of malignant transformation, Genes Dev. 16 (2002) 2252–2263. [6] A. Rangarajan, R. Syal, S. Selvarajah, O. Chakrabarti, A. Sarin, S. Krishna, Activated Notch1 signaling cooperates with papillomavirus oncogenes in transformation and generates resistance to apoptosis on matrix withdrawal through PKB/Akt, Virology 286 (2001) 23–30. [7] S. Weijzen, A. Zlobin, M. Braid, L. Miele, W.M. Kast, HPV16 E6 and E7 oncoproteins regulate Notch-1 expression and cooperate to induce transformation, J. Cell. Physiol. 194 (2003) 356–362.

1840

E XP E RI ME N T AL C E L L R E SE A RC H 31 7 ( 20 1 1) 1 8 30 – 1 84 0

[8] J. Yao, L. Duan, M. Fan, J. Yuan, X. Wu, Notch1 induces cell cycle arrest and apoptosis in human cervical cancer cells: involvement of nuclear factor kappa B inhibition, Int. J. Gynecol. Cancer 17 (2007) 502–510. [9] B. Ramdass, T.T. Maliekal, S. Lakshmi, M. Rehman, P. Rema, P. Nair, G. Mukherjee, B.K. Reddy, S. Krishna, M. Radhakrishna Pillai, Coexpression of Notch1 and NF-kappaB signaling pathway components in human cervical cancer progression, Gynecol. Oncol. 104 (2007) 352–361. [10] L.L. Song, Y. Peng, J. Yun, P. Rizzo, V. Chaturvedi, S. Weijzen, W.M. Kast, P.J. Stone, L. Santos, A. Loredo, U. Lendahl, G. Sonenshein, B. Osborne, J.Z. Qin, A. Pannuti, B.J. Nickoloff, L. Miele, Notch-1 associates with IKKalpha and regulates IKK activity in cervical cancer cells, Oncogene 27 (2008) 5833–5844. [11] C. Talora, S. Cialfi, O. Segatto, S. Morrone, J. Kim Choi, L. Frati, G. Paolo Dotto, A. Gulino, I. Screpanti, Constitutively active Notch1 induces growth arrest of HPV-positive cervical cancer cells via separate signaling pathways, Exp. Cell Res. 305 (2005) 343–354. [12] L. Wang, H. Qin, B. Chen, X. Xin, J. Li, H. Han, Overexpressed active Notch1 induces cell growth arrest of HeLa cervical carcinoma cells, Int. J. Gynecol. Cancer 17 (2007) 1283–1292. [13] T.T. Maliekal, J. Bajaj, V. Giri, D. Subramanyam, S. Krishna, The role of Notch signaling in human cervical cancer: implications for solid tumors, Oncogene 27 (2008) 5110–5114. [14] L. Wu, T. Sun, K. Kobayashi, P. Gao, J.D. Griffin, Identification of a family of mastermind-like transcriptional coactivators for mammalian notch receptors, Mol. Cell. Biol. 22 (2002) 7688–7700. [15] L. Wu, J.D. Griffin, Modulation of Notch signaling by mastermind-like (MAML) transcriptional co-activators and their involvement in tumorigenesis, Semin. Cancer Biol. 14 (2004) 348–356. [16] A.S. McElhinny, J.L. Li, L. Wu, Mastermind-like transcriptional co-activators: emerging roles in regulating cross talk among multiple signaling pathways, Oncogene 27 (2008) 5138–5147. [17] Y. Nam, P. Sliz, L. Song, J.C. Aster, S.C. Blacklow, Structural basis for cooperativity in recruitment of MAML coactivators to Notch transcription complexes, Cell 124 (2006) 973–983. [18] Y. Nam, A.P. Weng, J.C. Aster, S.C. Blacklow, Structural requirements for assembly of the CSL.intracellular Notch1.Mastermind-like 1

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

transcriptional activation complex, J. Biol. Chem. 278 (2003) 21232–21239. I. Maillard, A.P. Weng, A.C. Carpenter, C.G. Rodriguez, H. Sai, L. Xu, D. Allman, J.C. Aster, W.S. Pear, Mastermind critically regulates Notch-mediated lymphoid cell fate decisions, Blood 104 (2004) 1696–1702. I. Maillard, L. Tu, A. Sambandam, Y. Yashiro-Ohtani, J. Millholland, K. Keeshan, O. Shestova, L. Xu, A. Bhandoola, W.S. Pear, The requirement for Notch signaling at the beta-selection checkpoint in vivo is absolute and independent of the pre-T cell receptor, J. Exp. Med. 203 (2006) 2239–2245. Y. Zhao, R.B. Katzman, L.M. Delmolino, I. Bhat, Y. Zhang, C.B. Gurumurthy, A. Germaniuk-Kurowska, H.V. Reddi, A. Solomon, M.S. Zeng, A. Kung, H. Ma, Q. Gao, G. Dimri, A. Stanculescu, L. Miele, L. Wu, J.D. Griffin, D.E. Wazer, H. Band, V. Band, The notch regulator MAML1 interacts with p53 and functions as a coactivator, J. Biol. Chem. 282 (2007) 11969–11981. K. Roser, K.T. Jakel, J. Bullerdiek, T. Loning, Significance of molecular-cytogenetic findings in mucoepidermoid carcinoma as an example of salivary gland tumors, Pathologe 26 (2005) 359–366. S. Suwanjunee, W. Wongchana, T. Palaga, Inhibition of gamma-secretase affects proliferation of leukemia and hepatoma cell lines through Notch signaling, Anticancer Drugs 19 (2008) 477–486. C. Osipo, T.E. Golde, B.A. Osborne, L.A. Miele, Off the beaten pathway: the complex cross talk between Notch and NF-kappaB, Lab. Invest. 88 (2008) 11–17. O. Chakrabarti, K. Veeraraghavalu, V. Tergaonkar, Y. Liu, E.J. Androphy, M.A. Stanley, S. Krishna, Human papillomavirus type 16 E6 amino acid 83 variants enhance E6-mediated MAPK signaling and differentially regulate tumorigenesis by notch signaling and oncogenic Ras, J. Virol. 78 (2004) 5934–5945. K. Veeraraghavalu, V.K. Subbaiah, S. Srivastava, O. Chakrabarti, R. Syal, S. Krishna, Complementation of human papillomavirus type 16 E6 and E7 by Jagged1-specific Notch1-phosphatidylinositol 3-kinase signaling involves pleiotropic oncogenic functions independent of CBF1;Su(H);Lag-1 activation, J. Virol. 79 (2005) 7889–7898.