Skip Navigation


Carcinogenesis Advance Access originally published online on July 9, 2007
Carcinogenesis 2007 28(11):2305-2312; doi:10.1093/carcin/bgm158
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
28/11/2305    most recent
bgm158v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Fauquette, V.
Right arrow Articles by Pigny, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fauquette, V.
Right arrow Articles by Pigny, P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Transcription factor AP-2{alpha} represses both the mucin MUC4 expression and pancreatic cancer cell proliferation

Valérie Fauquette1,2, Sébastien Aubert1,2,4, Sophie Groux-Degroote3, Brigitte Hemon1,2, Nicole Porchet1,2,4, Isabelle Van Seuningen1,2 and Pascal Pigny1,2,4,*

1 INSERM, U837, Place de Verdun, 59045 Lille cedex, France
2 Faculté de médecine, Centre de Recherche Jean-Pierre Aubert, Université de Lille 2, Place de Verdun, 59045 Lille cedex, France
3 UMR CNRS 8576, Université des Sciences et Technologies de Lille 1, 59655 Villeneuve d'Ascq cedex, France
4 Centre Hospitalier Régional et Universitaire, 59037 Lille cedex, France

* To whom correspondence should be addressed. Tel: +33 3 202 988 50; Fax: +33 3 205 385 62; Email: pigny{at}lille.inserm.fr


   Abstract
Top
 Abstract
Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
MUC4 is a transmembrane mucin expressed in pancreatic ductal adenocarcinoma (DAC) in contrast to normal pancreas, and is an independent predictor of poor prognosis in patients with invasive DAC. Our aim was therefore to investigate the mechanisms that control MUC4 expression in pancreatic cancer cells. We focused our study on activator protein (AP)-2{alpha} transcription factor that acts as a tumour suppressor gene in several cancers. In a series of 18 human DAC, using immunohistochemistry, we confirmed that MUC4 was exclusively expressed in cancerous or preneoplastic lesions in 83% of the samples. On the contrary, AP-2 was mainly expressed by non-tumoural ductal cells (61%) or endocrine cells (67%). Moreover, MUC4 and AP-2 were never found co-expressed suggesting an inhibitory role of AP-2{alpha} in normal ductal cells. In CAPAN-1 and CAPAN-2 cells, transient AP-2{alpha} over-expression decreased both MUC4 mRNA and apomucin levels by 20–40% by a mechanism involving inhibition of MUC4 promoter. By chromatin immunoprecipitation and gel-shift assays, we demonstrated that this inhibition involved two AP-2 cis-elements located in the –475/–238 region of the promoter. CAPAN-1 clones, which stably over-expressed AP-2{alpha}, displayed a strong MUC4 down-regulation (–38 to –100%), a significant decrease of both cell proliferation and invasion concomitant to the up-regulation of p27 cyclin-dependent kinase inhibitor. In conclusion, our data provide evidence that AP-2{alpha} is an important in vivo negative regulator of MUC4 expression in human pancreatic tissue and that AP-2{alpha} may play a tumour-suppressive role in pancreatic DAC.

Abbreviations: AP, activator protein; BSA, bovine serum albumin; cdk, cyclin-dependent kinase; ChIP, chromatin immunoprecipitation; DAC, ductal adenocarcinoma; PCR, polymerase chain reaction; siRNA, small interfering RNA; WD, well-differentiated

Pancreatic cancer is an especially devastating form of cancer, with only 3% of patients found to survive 5 years after diagnosis. Pathologically, most pancreatic cancers correspond to ductal adenocarcinoma (DAC), which develops from epithelium in a multistep process (1). Pancreatic intraepithelial neoplasia is considered as a precursor of DAC. At the time of diagnosis, >80% of patients have locally advanced or metastatic disease. The inability to detect cancer at onset, the aggressiveness of DAC and the lack of effective therapies are responsible for the low patient survival rate. Also, to improve patient prognosis, further understanding of the molecular mechanisms of cancer progression is mandatory. Development of new therapeutic tools based on these mechanisms is also required. Transcription factors that have been implicated in the pathogenesis of malignancy could thus serve as novel therapeutic targets (2).

The activator protein (AP)-2 family comprises five isoforms of 52 kDa proteins encoded by independent genes (3), the most studied being AP-2{alpha}, AP-2ß and AP-2{gamma}. They share a common structure, possessing a proline/glutamine-rich transactivation domain in the N-terminal region, and a helix-span-helix domain in the C-terminal region, which mediates dimerization and site-specific DNA binding. AP-2 transcriptional activity is negatively controlled by sumolation, i.e. conjugation with SUMO peptides (4). Depending on the cellular context, AP-2 transcription factors are individually associated either with cell differentiation and development (3,5) or with cancer progression/regression (6). For example, loss of AP-2 expression results in the transition of melanoma cells to the metastatic phenotype (7), indicating that AP-2 may have a tumour-suppressive role. Recent data where AP-2{alpha} was over-expressed in ovarian cancer cells confirmed such a role both in vitro and in vivo (8). On the contrary, in breast cancer AP-2 induced ErbB-2 and ErbB-3 proto-oncogene expression (9) leading to an increase of the malignant potential of cancer cells.

The membrane-bound mucin MUC4 is a high molecular weight glycoprotein expressed by epithelial cells. Its large extracellular domain comprises a tandem repeat domain rich in Ser/Thr and two epidermal growth factor-like domains (10). MUC4 is the human homologue of the rat sialo–mucin complex, a heterodimeric glycoprotein acting as an intramembrane ligand for p185neu/ErbB-2 via one of its epidermal growth factor-like domains (11). Whereas MUC4 is not expressed in normal pancreas, up to 75% of pancreatic DAC show de novo expression of MUC4 at the mRNA level (12,13). Swartz et al. (14) demonstrated a gradual expression of MUC4 in pancreatic carcinogenesis from pancreatic intraepithelial neoplasia type 1 (17%) to DAC (89%). Moreover, Saitou et al. (15) recently demonstrated that a high MUC4 expression was an independent predictor of poor outcome in patients with invasive DAC.

In this report, our aim was to characterize the molecular mechanisms governing MUC4 expression in pancreatic cancer. We previously characterized MUC4 promoter, which is composed of a proximal and a distal regulatory region, and identified several consensus binding sites for AP-2 (16). We thus decided to study the relationships between AP-2 and MUC4. To check the relevance of our hypothesis, we first studied MUC4 and AP-2{alpha} expression in a series of human DAC and demonstrated that their expression was mutually exclusive in the tumoural tissue. We thus undertook to investigate the effects of AP-2{alpha} on MUC4 transcriptional regulation and expression in CAPAN-1 and CAPAN-2 pancreatic cancer cells. Lastly, we also analysed cell proliferation and invasion in vitro in CAPAN-1 clones that stably over-expressed AP-2{alpha}. Our results demonstrate that AP-2{alpha} efficiently represses MUC4 expression both in vitro and in vivo, as well as pancreatic cancer cell proliferation and invasion in vitro.


   Materials and methods
Top
 Abstract
 Materials and methods
Results
 Discussion
 Supplementary material
 Funding
 References
 
Pancreatic tumour specimens and immunohistochemistry
Pancreatic tissues were obtained from 18 patients who underwent pancreatic surgery for primary DAC in the Surgical Department of the Lille University Hospital. No patient received either chemotherapy or radiotherapy before the surgical resection. An informed consent was obtained from each patient before inclusion in the study. The histological slides for each case were reviewed by two pathologists (S.A., E.L.) and one representative slide was selected for immunohistochemical analysis (S.A.). Immunohistochemistry studies were conducted on formalin-fixed, paraffin-embedded tissues using an automated immunostainer (Nex ES, Ventana Medical System, Strasbourg, France) as described previously (17). The mouse monoclonal anti-MUC4 antibody (m8G7), directed against the tandem repeat region, was used at 1:2000 dilution. For AP-2{alpha}, a pre-treatment step (inactivation of peroxidase activity and antigen retrieval) was realized as described (18) before performing incubation with the rabbit polyclonal C18 antibody (Santa Cruz Laboratory, Santa Cruz, CA, USA) at 1:750 dilution in phosphate-buffered saline containing 1% bovine serum albumin (BSA) and 0.1% Triton X-100. The secondary antibody is a mix of anti-rabbit and anti-mouse IgG (Ventana). In each series, a negative control (omission of the primary antibody) and positive control (melanoma specimen for AP-2{alpha} and bronchial tissue for MUC4) were included (see Supplementary data 1, available at Carcinogenesis Online).

Immunofluorescence
Cells were grown on a sterile glass coverslip (Lab-Tek® chamber slide, Nalge Nunc, Dominique Dutcher SAS, Brumath, France) 48 h at 37°C and washed three times with TBS [Tris–HCl 50 mM (pH 7.5), NaCl 150 mM]. Cells were fixed with 4% paraformaldehyde for 10 min, and then washed three times for 10 min with TBST (TBS, 0.2% Triton X-100). Slides were incubated in TBST containing 1% BSA for 30 min before adding the primary antibody (anti-AP-2{alpha} C-18) at a final concentration of 20 µg/ml for 1 h (dilution 1:100). Slides were washed three times with TBST containing 1% BSA. Fluorescein-conjugated antibody (Santa Cruz Laboratory) was added at a 1:250 dilution in TBST containing 1% BSA and left for 45 min in the dark. Slides were washed three times and mounted with Vectashield with 4',6-diamidino-2-phenylindole (H-1200, Vector Laboratories, Biovalley, France). Fluorescence was read on a FluoArc HBO100 (Zeiss, Le Pecq, France). Negative controls were run by omission of the primary antibody (data not shown).

Transient transfections
Pancreatic cancer cells CAPAN-1 or CAPAN-2 were transfected with 1 µg of the pGL3 basic vector (Promega, Charbonnières, France), carrying promoter elements of MUC4 upstream of the luciferase gene as described previously (16), and various amounts of an expression vector encoding AP-2{alpha} (pRSV-AP-2 gift from Dr H. Hurst, ICRF Molecular Oncology Unit, London, UK). Results were expressed as fold induction relative to the co-transfection performed in the presence of the corresponding empty expression vector, as described previously (19).

Stable transfections
CAPAN-1 cells were transfected with either pcDNA 3.1 AP-2{alpha} expression vector (gift from Dr R.H. Dashwood, Oregon State University, Corvallis, OR) or pcDNA 3.1His-C control vector (Invitrogen, France) using Effectene® reagent according to the manufacturer's instructions. Stable transfectants were selected by growth in the media supplemented with 300 µg/ml of G418 (Geneticin G418, Invitrogen) and clones resistant to G418 were obtained. Their pattern of AP-2{alpha} and MUC4 expression was determined by western blotting. Eight control clones and 18 clones stably transfected with AP-2{alpha} expression vector were obtained.

Small interfering RNA assays
Cells (2 x 105) were transfected with 100 nM of ON-TARGETplus SMARTpool® directed against AP-2{alpha} (Dharmacon, Perbio, France) using Dharmafect 4 reagent as described previously (20). Controls included mock transfected cells, and cells transfected with siCONTROL non-targeting small interfering RNA (siRNA) pool.

Semi-quantitative Reverse transcription-polymerase chain reaction
Total RNA was prepared 48 h after transient transfection using the Rneasy mini-kit (Qiagen, Courtaboeut, France) as described previously (16). MUC4 mRNA levels were evaluated by semi-quantitative Reverse transcription-polymerase chain reaction (PCR) using 28S as an internal control as described previously (16). For AP-2{alpha}, we used the Rt-PCR protocol described by Wajapeyee et al. (21). A densitometry analysis of the bands was performed using the Clara Vision Gel Smart-Gel analysis software, as described previously (19).

Site-directed mutagenesis
Three AP-2-binding sites from the –461/–1 region of MUC4 promoter were mutated separately using the QuickChange Site-Directed Mutagenesis kit (Stratagene, Saint Quentin en Yvelines, France) as described in (19). The –173 site was mutated from 5'-GGGGCCCC-3' (wild-type) to 5'-AAGGAATA-3', the –382 site from 5'-CCCCTGGGG-3' (wild-type) to 5'-CAATGGAA-3' and the –429 site from 5'-TCCCCA-3' (wild-type) to 5'-TACCAA-3'. The mutated plasmid DNA was sequenced on both strands before being used in cell transfection experiments.

Nuclear extract preparation and electrophoretic mobility-shift assay
Nuclear and cytoplasmic extracts from CAPAN-1 or CAPAN-2 cells were prepared as described previously (22). Binding studies were performed using nuclear extracts (8 µg) and oligonucleotides containing the –173 AP-2 site (see Figure 3), the –382, the –429 or the –401 binding sites (5'-AAGCTCTGGTGGGACAGGGGC-3') present in MUC4 promoter or their mutated versions (see above) as described previously (19). Supershift experiments were performed by adding 4 µl of an anti-AP-2{alpha} (C-18, sc-184) or anti-Sp1 (sc-59) antibody to the reaction mixture overnight at 4°C. Samples were run as described in (19).

Total cellular extract preparation and western blotting
After scraping, the cells were pelleted, washed with 1x phosphate-buffered saline and incubated in lysis buffer [50 mM Tris–HCl (pH 7.5), containing 150 mM NaCl, 1% NP40, 5 mM sodium fluoride, 5 mM sodium orthovanadate, 0.25% sodium deoxycholate and a cocktail of protease inhibitors] for 30 min at 4°C. After 10 min of centrifugation at 13 210g (4°C), the supernatant (total extract) was stored at –80°C. Western blot was performed as described previously (19) using specific primary antibodies: C-18 for AP-2{alpha}, sc-528 for p27kip1, sc-397 for p21waf1, sc-764 for c-myc, all from Santa Cruz Biotechnology (Tebu, France), and A5441 for ß-actin (from Sigma, Lyon, France). The C-18 antibody also recognized, but to a lesser extent, AP-2ß and AP-2{gamma}. The anti-MUC4 mouse monoclonal antibody m8G7 was a gift from Dr S.K. Batra, University of Nebraska Medical Center (Omaha, NE, USA). Peroxidase-conjugated secondary antibodies were used and immunoreactive bands were visualized as described previously (19). The western blot of MUC4 was performed as described previously (23).

Cell proliferation and invasion
Cell proliferation was evaluated by seeding 105 cells into six-well plates on day 0. Three representative wells were counted every day during 5 days using a Malassez haemocytometer. Three independent experiments were carried out. Cell invasion was evaluated using 24-well Matrigel invasion chambers (BD Biosciences, Le Pont de Claix, France) as described previously (23) with the following modifications: 105 cells were plated in the top chamber, the incubation time was extended to 48 h, 15% fetal calf serum was used as a chemoattractant in the lower chamber. After staining, the total number of cells on the lower surface of the filters was counted under microscopy at x100 magnification.

Chromatin immunoprecipitation assay
The chromatin immunoprecipitation (ChIP) assay protocol described by Piessen et al. (20) was used with slight modifications. Briefly, 15 x 106 CAPAN-1 cells were cross-linked by 1% formaldehyde for 10 min at room temperature. Cross-linking was stopped, cells were washed and nuclei were pelleted before being sonicated with the Bioruptor system (Diagenode, Belgium). The chromatin solution was precleared as described (20). After centrifugation, a supernatant aliquot was removed as a control (input) and the remainder was precipitated overnight with 6 µg of anti-AP-2{alpha} antibody or with normal rabbit IgG (Upstate Biotechnology, Chandlers Ford, Hampshire, UK.) as negative control. Protein G-agarose gel slurry was then added (vol/vol) and the incubation was continued for 2 h. After reversal of the cross-linking, the DNA was purified and 100 ng of input control or ChIP samples were used as a template for PCR using the primer set designed to amplify the –475/–238 region of MUC4 promoter (5'-TCTTTCCCCCATTCATAC-3' and 5'-GAAAACACCGATACACCC-3'). PCR was performed in a 25 µl volume containing 1 U AmpliTaq (Roche, Meylan, France) and 5 pmoles of each primer using the following protocol: 2 min at 94°C followed by 32 cycles of 45 s at 94°C, 1 min of annealing at 50°C and 1 min of extension at 72°C followed by a 10 min final extension. PCR products (20 µl) were analysed on a 2% agarose gel.

Statistical analysis
Data are presented as mean ± standard deviation. Differences in the mean of two samples were analysed by Student's t-test with differences <0.05 considered significant.


   Results
Top
 Abstract
 Materials and methods
 Results
Discussion
 Supplementary material
 Funding
 References
 
Expression levels of AP-2{alpha} and MUC4 in human pancreatic adenocarcinoma
Tissues from 18 patients with pancreatic DAC (14 men and 4 women, age range: 50–78 years) were immunostained with an anti-MUC4 or anti-AP-2{alpha} antibody. As already demonstrated, no MUC4 protein was observed in normal pancreatic tissue (Figure 1B) whereas 83% (15/18) of the samples expressed MUC4 protein at the cytoplasmic or membrane level, both in areas with severe dysplasia or poorly differentiated tumours (Figure 1D). Tumoural tissue areas that expressed MUC4 never showed co-expression of AP-2 (compare panels C and D). On the contrary, AP-2 was frequently observed in the nuclei and cytoplasm of endocrine islets [67% of the patients (12/18)] and in non-tumoural ductal cells (in 61% of the cases [11/18)] mainly at the cytoplasmic level (Figure 1A). Only one well-differentiated (WD) tumour (5.5% of the samples) was immunostained for AP-2 (Figure 1E) but not for MUC4 (Figure 1F). In conclusion, AP-2 was mainly present in non-cancerous pancreatic tissue and was never found co-expressed with MUC4.


Figure 1
View larger version (85K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Immunohistochemistry on pancreatic human tissues was performed as described in the Materials and Methods with an anti-AP-2 antibody (C-18) directed against the C-terminal end of human AP-2{alpha} (panels A, C and E) or an anti-MUC4 antibody (panels B, D and F). Magnification x200. (A, B) Non-tumoural ductal cells. (C, D) poorly differentiated ductal adenocarcinoma. (E, F) WD ductal adenocarcinoma.

 
AP-2{alpha} regulates MUC4 mRNA levels and apomucin expression in pancreatic cancer cells
Having shown a negative relationship between AP-2 and MUC4 expression in non-tumoural and tumoural tissues, we studied the mechanisms of MUC4 regulation by AP-2{alpha} in vitro. As an experimental model, we used two WD pancreatic cancer cell lines CAPAN-1 and CAPAN-2 that both expressed AP-2{alpha} and MUC4 (Figure 2A and B). Immunofluorescence analysis showed that AP-2{alpha} is localized both in the nucleus and cytoplasm of CAPAN-1 cells and mainly in the nucleus of CAPAN-2 cells (Figure 2A). This was further supported by immunoblot analysis of AP-2{alpha} demonstrating that nuclear extracts mainly expressed the monomer form of AP-2{alpha} (~50 kDa), whereas slow migrating forms (~72, 95 and 120 kDa) were the major species in cytosolic extracts of both cell lines (data not shown). We next evaluated whether a transient over-expression of AP-2{alpha} could affect the level of endogenous MUC4 mRNA and protein levels in the cell lines. Rt-PCR and the densitometry analysis of the bands showed that transient over-expression of AP-2{alpha} led to a 40% decrease of MUC4 mRNA level for 1 µg expression vector (P < 0.01, Figure 2B). This negative regulation of MUC4 by AP-2{alpha} was confirmed at the protein level by immunocytochemistry (Figure 2C). To check the specificity of the negative effect, a transient silencing of AP-2{alpha} was performed using a pool of siRNA against this transcription factor, or several controls. As shown in Figure 2D, AP-2{alpha} mRNA level, assessed by the AP-2 to 28S ratio, was reduced by 40–50% by the AP-2{alpha} siRNA in both cell lines. This silencing was accompanied by a slight up-regulation of MUC4 mRNA level [+86% in CAPAN-1 cells (P < 0.05), +30% in CAPAN-2 cells (not significant)]. In conclusion, AP-2{alpha} is an inhibitor of MUC4 endogenous expression at the transcriptional level in two pancreatic cancer cell lines.


Figure 2
View larger version (43K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. (A) As an experimental model, we used two WD pancreatic cancer cell lines CAPAN-1 and CAPAN-2. Expression of AP-2{alpha} and MUC4 was evaluated by immunofluorescence. (B) Effect of AP-2 transient over-expression on MUC4 mRNA levels in CAPAN-1 and CAPAN-2 cells. Cells were cultured in 40 mm dishes at 40% confluence before being transfected with 1 µg of AP-2{alpha} expression vector or empty expression vector (C, control). RNA was prepared 48 h later using the Rneasy mini-kit (Qiagen). The mRNA levels of MUC4 and AP-2{alpha} were checked by Rt-PCR. 28S was used as an internal control. The MUC4 to 28S ratio obtained after densitometry analysis of the bands is shown. The ratio obtained with the empty vector was set at 1. The data shown are mean ± standard deviation of three independent experiments (**P < 0.01). (C) Effect of AP-2{alpha} over-expression on MUC4 protein level. Immunocytochemistry was performed on transiently transfected cells as described in (19). The percentage of MUC4-positive cells is indicated at the bottom of each figure. (D) Effect of transient silencing of AP-2{alpha} on MUC4 expression. siRNA experiments were carried out as described in the Materials and Methods. MUC4, AP-2{alpha} and 28S mRNA level were assessed by Rt-PCR. Results are expressed as MUC4 to 28S ratio (MUC4) or AP-2{alpha} to 28S ratio (AP-2{alpha}) obtained after densitometry analysis of the bands. Control (NT) corresponds to the cells transfected with non-targeting siRNA. The data are mean ± standard deviation of two independent experiments run in triplicate.

 
AP-2 selectively inhibits MUC4 promoter
The MUC4 promoter contains several AP-2-binding sites (5'-GCCNNNGGC-3') located both in the proximal and distal regions (Figure 3A). By transient co-transfections, we showed that AP-2{alpha} repressed the transcriptional activity of the –461/–1 region by 40% in CAPAN-1 and by 80% in CAPAN-2 cells (Figure 3B). This repression was lost on the shorter fragment and not observed on the distal promoter. Individual contribution of the AP-2-binding sites was studied by site-directed mutagenesis in both cell lines. When the –173 or the –429 AP-2-binding sites on the –461/–1 region was mutated (Figure 3C), the repressive effect of AP-2{alpha} was deeply reduced (P < 0.01 in CAPAN-1 cells). The reduction was slightly lower when the –382 site was mutated (P < 0.05 in both cell lines). Since no interaction was noticed with the –401 site by EMSA, this site was not studied by mutagenesis. These results suggest that three sites located at –173, –382 and –429 are implicated in the repression of MUC4 promoter by AP-2{alpha}.


Figure 3
View larger version (37K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. (A) Schematic representation of MUC4 promoter. The numbering refers to the A of the first ATG; the bent arrows represent the position of the transcriptional start sites. The main cis-elements for Sp-1, AP-1, AP-2 and PEA3 are shown. (B) AP-2{alpha} represses the transcriptional activity of MUC4 promoter in CAPAN-1 and CAPAN-2 cells. The MUC4-pGL3 promoter constructs were transiently co-transfected with 0.2 µg of pRSV-AP-2 or corresponding empty vector. Results are expressed as fold induction of luciferase activity relative to the empty expression vector (value set at 1.0). Values are means ± standard deviation for three independent experiments, where co-transfections were run in triplicate. (C) Effect of mutation of the AP-2-binding sites on the responsiveness of the –461/–1 region of MUC4 promoter to AP-2{alpha}. Three putative AP-2-binding sites located at –173, –382 and –429 were mutated separately before performing transient co-transfections. Results are expressed as mean ± standard deviation for three independent experiments. *P < 0.05, **P < 0.01 (Student's t-test).

 
AP-2{alpha} directly interacts with its cognate elements in the MUC4 proximal promoter
The AP-2 binding to its putative binding sites was studied in vitro by EMSA. The four sites located at –173, –382, –401 and –429 were studied in both cell lines (Figure 4A–C). Two retarded complexes (Figure 4A–C, arrow # 1) were obtained when incubating nuclear extracts with the radiolabelled probes containing the –173, –382 or the –429 site. No interaction was observed with the –401 site (data not shown). Competition experiments performed with an excess of unlabelled probe led to a substantial decrease of the shifted bands whereas a mutated probe did not modify the pattern. A supershift was observed when the anti-AP-2{alpha} antibody was added to the binding reaction (Figure 4A–C, arrow # 2 for CAPAN-1 cells; Figure 4B and C, arrow # 2 for CAPAN-2 cells). The use of an irrelevant anti-Sp1 antibody did not modify the retarded band formation, thus indicating that Sp1 did not compete with AP-2 to bind to those GC-rich sites. In conclusion, AP-2{alpha} is able to interact directly in vitro with its cognate elements in MUC4 promoter, respectively, located at –173 and –382 in both cell lines and at –429 only in CAPAN-1 cells, suggesting that the repressive effect of AP-2{alpha} on MUC4 transcription could be a direct one.


Figure 4
View larger version (42K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. DNA binding to the –429 site (A), to the –382 site (B) or to the –173 site (C) of MUC4 promoter was evaluated by EMSA. Lane 1: labelled probe only. Two specific retarded complexes (arrow # 1) were seen in CAPAN-1 (lane 2) and CAPAN-2 cells (lane 7). Specificity was assessed by addition of a 500-fold excess of an unlabelled wild-type probe (lanes 3 and 8) or of an unlabelled mutated probe (lanes 4 and 9). Supershift experiments were performed by adding an anti-AP-2{alpha} (lanes 5 and 10) or an anti-Sp1 antibody (lanes 6 and 11) in the reaction mixture. Supershifted bands are shown by arrow # 2. The nucleotide sequences of the three cis-elements are shown at the bottom of the figure.

 
Effects of AP-2{alpha} stable over-expression in CAPAN-1 cells
To further investigate the roles of AP-2{alpha}, stable transfection with an AP-2{alpha} expression vector was performed in CAPAN-1 cells. Among transfectants, four independent clones ({alpha}27, {alpha}38, {alpha}42 and {alpha}45) showing 1.4- to 3.7-fold higher levels of AP-2{alpha} than control clone C5 (transfected with the empty vector) were selected (Figure 5A). All clones expressed reduced levels of MUC4 apomucin by western blotting (from –38 to nearly –100%) and deeply reduced levels of MUC4 mRNA by real-time PCR (at least –80%, data not shown).


Figure 5
View larger version (38K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5. Characterization of the CAPAN-1 clones stably transfected with an empty vector (C5) or with an AP-2{alpha} expression vector ({alpha}27, {alpha}38, {alpha}42 and {alpha}45). (A) Western blot analysis of the expression of AP-2{alpha} and MUC4. Bands were quantified by densitometry and a ratio (specific protein to ß-actin) was calculated to evaluate differences between C5 and selected clones. Corresponding ratio values are shown under each figure. nc: not calculated. (B) Western blot analysis of different cell cycle regulators in CAPAN-1 clones over-expressing AP-2{alpha}. Total cellular extracts were analysed by western blot. ß-actin immunostaining was performed to check the protein loading. Bands were quantified and a ratio (specific protein to ß-actin) was calculated to evaluate differences between mocked and transfected cells. Corresponding ratio values are shown under each figure. (C) ChIP was realized as described in the Materials and Methods. Ca1: CAPAN-1 cells. Inp: input fraction. AP-2 represents the fraction precipitated with the anti-AP-2{alpha} antibody and IgG the fraction precipitated with a rabbit IgG (negative control). An antibody to Inp ratio was calculated for each case after densitometry analysis of the bands. Values are as follows: Ca1: 0.08, C5: not calculated, {alpha}27: 0.24 and {alpha}38: 0.15.

 
To document the molecular consequences of AP-2{alpha} over-expression, we studied the protein levels of different key players of the cell cycle in the stable clones (Figure 6B). AP-2{alpha} over-expression was accompanied by an important up-regulation of the cyclin-dependent kinases (cdk) inhibitor p27 (from 5- to 12-fold compared with C5). There was a trend towards a down-regulation of c-myc, especially in {alpha}42, whereas variations of p21 remained weaker. In conclusion, AP-2{alpha} over-expression is associated with a substantial up-regulation of p27.


Figure 6
View larger version (21K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6. A) Cell proliferation was evaluated during 120 h. Cells were counted every day and values are expressed in number of cells per microlitre. Values are mean ± standard deviation for three independent experiments run in triplicate, *P < 0.05, **P < 0.01. (B) Cell invasion was evaluated using 24-well Matrigel invasion chambers with 15% fetal calf serum as chemoattractant. The graph shows the total number of invaded cells counted 48 h after seeding. Values are mean ± standard deviation derived from two independent experiments run in triplicate, *P < 0.05, **P < 0.01.

 
ChIP experiments were performed to check whether AP-2{alpha} could directly act in vivo on MUC4 promoter to down-regulate its expression. As shown in Figure 5C, AP-2{alpha} binding to the –475/–238 region of MUC4 promoter was very weak or nearly undetectable in parental CAPAN-1 cells and control clone C5, respectively, in favour of a moderate or null effect of AP-2{alpha} in these cells. On the contrary, the binding to this region of the promoter was increased from 2- to 3-fold in clones {alpha}27 and {alpha}38 compared with parental CAPAN-1 cells. The –253/–76 region containing the –173 AP-2 cis-element and the distal –2596/–2377 region were also tested but no AP-2{alpha} binding was observed (Supplementary data 2, available at Carcinogenesis Online). Thus, MUC4 down-regulation by AP-2{alpha} in AP-2{alpha} over-expressing clones may be due to its recruitment to the –475/–238 region of MUC4 promoter that contains the –382 and –429 AP-2-binding sites.

Finally, the effect of AP-2{alpha} over-expression on cell proliferation and invasion was assessed in vitro. Concerning cell proliferation (Figure 6A), at 96 h all clones, except {alpha}42, demonstrated a significant growth inhibition compared with control clone C5 (P < 0.05). At 120 h, the proliferation rate of the four clones was significantly decreased (P < 0.01). Cell invasion was evaluated using the Matrigel invasion chambers. As shown in Figure 6B, the number of invading cells was significantly lower for all clones compared with control clone C5 indicating a reduced invasiveness (P < 0.05 to P < 0.01), whatever their residual level of MUC4 expression was. These results demonstrate that AP-2{alpha} over-expression results in an in vitro inhibition of pancreatic cancer cell growth and invasion.


   Discussion
Top
 Abstract
 Materials and methods
 Results
 Discussion
Supplementary material
 Funding
 References
 
AP-2{alpha} is considered as a tumour suppressor gene that regulates important factors of cancer development and progression such as the cell adhesion protein MUC18 (7), the thrombin receptor PAR-1 (24) or the cdk inhibitor p21waf1 (25) in melanoma, ovarian or colon carcinoma cells. In this last model, recent data showed that loss of AP-2{alpha} expression resulted in an increase of the in vivo tumourigenicity of the cells associated with a down-regulation of E-cadherin (26). Similarly, reintroduction of AP-2{alpha} in ovarian cancer cells greatly reduced their metastatic potential (8). In this report, we show for the first time that AP-2{alpha} is a repressor of MUC4 mucin tumour marker. We found a negative correlation between AP-2{alpha} and MUC4 expression in vivo in human pancreatic DAC. As already shown by others (12–14), MUC4 is expressed in 83% of pancreatic DAC samples, either poorly differentiated or WD. No expression was found in normal acini or ductal epithelial cells. On the contrary, AP-2{alpha} was found at the cytoplasmic level in non-tumoural ductal cells, as already shown by Motonaga et al. (27). In our series, only one WD DAC (5.5%) expressed AP-2{alpha} but was not labelled by the anti-MUC4 antibody. Whether AP-2{alpha} is associated with an epithelial differentiation of DAC is currently not known. However, in normal prostate epithelium, AP-2{alpha} expression is associated with luminal differentiation and is lost early in the development of prostate adenocarcinoma (28). Our results suggest that loss of AP-2{alpha} occurs during pancreatic carcinogenesis, probably as a late event, and contributes to the neo-expression of MUC4. Moreover, our transcriptional studies demonstrated that AP-2{alpha} is a potent repressor of MUC4 transcription in vitro in pancreatic cancer cells, further emphasizing the negative link between AP-2 and MUC4 expression. ChIP experiments indicated that the repressive effect of AP-2{alpha} was in part direct since AP-2{alpha} binding to the –475/–238 region of MUC4 promoter was selectively reinforced in clones where MUC4 expression was down-regulated. More accurately, this effect involves two AP-2 cis-elements of the proximal promoter located at –429 and –382, as shown by EMSA and site-directed mutagenesis. Unexpectedly, we did observe an expression of AP-2{alpha} in parental CAPAN-1 and CAPAN-2 cells that expressed high levels of MUC4. This could be explained by the presence of AP-2{alpha} in the cytoplasm of these cells that prevented it to interact with MUC4 promoter as demonstrated by ChIP experiments. Moreover, the nuclear fraction and especially the cytoplasmic fraction contained high molecular weight species (~72 kDa) that could correspond to less active sumolated forms of AP-2{alpha} (4). On the other hand, in stable clones that over-expressed AP-2{alpha}, the nuclear content of AP-2{alpha} was higher (data not shown) and a direct interaction with MUC4 promoter occurred (as shown by ChIP) allowing its down-regulation. Collectively, these in vitro and in vivo results suggest that AP-2{alpha} is involved in the repression of MUC4 in pancreatic tissues. Loss of AP-2{alpha} during pancreatic carcinogenesis could thus contribute to MUC4 expression.

To study the biological properties of pancreatic cancer cells that expressed various amounts of MUC4, we selected four CAPAN-1 clones that stably over-expressed AP-2{alpha}. All clones exhibited a reduced proliferation rate in vitro compared with control clones, regardless the level of reduction of MUC4 expression (–38% to nearly –100%). This was accompanied by a significant reduction of invasiveness by Matrigel invasion assay. One can hypothesize that a moderate decrease of MUC4 is sufficient to alter mitogenic intracellular pathways in pancreatic cancer cells. This expands the results of a previous study (23) in which a CD/HPAF cell clone stably transfected by a MUC4 antisense RNA led to a 80–90% decrease of MUC4 expression and a 30% decrease of in vitro cell growth. However, this antisense MUC4 clone expressed 2-fold lower levels of total and phosphorylated ErbB-2, a fact that could contribute at least in part to the suppression of tumour growth and metastasis.

In addition, AP-2{alpha} over-expression could modify the expression of other target genes than MUC4 that play key roles in cell proliferation. The first candidate is MUC1 since Adriance et al. (29) previously demonstrated that the inhibition of Muc1 transcription by activated ErbB-2 depends on the –247/–108 region of its promoter that contains a conserved AP-2 cis-element. Indeed, our preliminary data showed that AP-2{alpha} also repressed the expression of the membrane-bound mucin MUC1 in pancreatic cancer cells (data not shown). Another potential target gene of AP-2{alpha} is the cdk inhibitor p21waf1, which is a negative regulator of the cell cycle. Indeed, a previous report showed that AP-2{alpha} transactivated the p21 promoter, and induced its expression in SW480 colon carcinoma cells (25). Such a transactivation depends on p53 that targets AP-2{alpha} to the p53 cis-binding elements of p21 promoter (30). Interestingly, p53 is mutated (A159V) in CAPAN-1 cells (31) that are considered as p53 deficient (32), a fact that could explain the absence of p21 induction by AP-2{alpha} in these cells. Since p27 belongs to the same group of cdk inhibitor and since its loss of expression predicts a poor prognosis for patients with resectable DAC (33), we decided to look for variations of its expression in response to AP-2{alpha}. Whereas p27 was faintly expressed in parental CAPAN-1 cells, all AP-2{alpha} over-expressing clones exhibited a strong increase of p27. As already shown (34), up-regulation of p27 induced a cell cycle arrest in the G1 phase, leading to an inhibition of pancreatic cancer cell proliferation. This questions the role of AP-2{alpha} in modulating p27 level in pancreatic cancer cells. Since p27 expression is down-regulated by activation of the Ras–MEK–ERK pathway (34), one possibility is that AP-2{alpha} over-expression may elicit an inhibition of this pathway and consequently a p27 up-regulation. Alternatively, AP-2{alpha} could act directly on p27 promoter that contains two active GC boxes in the –549/–511 region (35), an issue that needs to be addressed in the next future. Previously published data using p27–/– mice demonstrated that p27 is an inducer of intestinal goblet cell differentiation as assessed by alcian blue staining or MUC2 expression (36). These data suggested that the expression of the secreted mucin MUC2, which plays a tumour suppressor function in the intestine (37), depends on p27 in a positive way. On the contrary, that of the membrane-bound mucin MUC4, which favours tumour development and progression, is negatively related to p27 levels in pancreatic tumours.

In conclusion, we showed that MUC4, recently defined as a strong indicator of poor prognosis in pancreatic cancer patients (15), is a new target gene of AP-2{alpha}, which down-regulates its expression in pancreatic cell lines leading to growth inhibition. We also showed that AP-2{alpha} expression is lost in MUC4 expressing DAC cells. Therefore, experimental strategies leading to a stable AP-2{alpha} over-expression could be a new and valuable therapeutic option in pancreatic cancer.


   Supplementary material
Top
 Abstract
 Materials and methods
 Results
 Discussion
 Supplementary material
Funding
 References
 
Supplementary data 1 and 2 can be found at http://carcin.oxfordjournals.org/.


   Funding
Top
 Abstract
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
References
 
Ministère de la Recherche et de la Technologie, France (to V.F.).


   Acknowledgments
 
This work is dedicated to the memory of Dr Jean Pierre Aubert, former director of the INSERM Unit 560, who has made major scientific contributions to the field of mucins. We thank Dr E.Leteurtre for her help in selecting patients with DAC and Dr N.Moniaux for his technical advice.

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 

  1. Takaori K, et al. Pancreatic intraepithelial neoplasia. Pancreas (2004) 28:257–262.[CrossRef][Web of Science][Medline]
  2. Libermann TA, et al. Targeting transcription factors for cancer gene therapy. Curr. Gene Ther. (2006) 6:17–33.[CrossRef][Web of Science][Medline]
  3. Eckert D, et al. The AP-2 family of transcription factors. Genome Biol. (2005) 6:246.[CrossRef][Medline]
  4. Eloranda JJ, et al. Transcription factor AP-2 interacts with the SUMO-conjugating enzyme UBC9 and is sumolated in vivo. J. Biol. Chem. (2002) 277:30798–30804.[Abstract/Free Full Text]
  5. Hilger-Eversheim K, et al. Regulatory roles of AP-2 transcription factors in vertebrate development, apoptosis and cell-cycle control. Gene (2000) 260:1–12.[CrossRef][Web of Science][Medline]
  6. Gershenwald JE, et al. Dominant-negative transcription factor AP-2 augments SB-2 melanoma tumor growth in vivo. Oncogene (2001) 20:3363–3375.[CrossRef][Web of Science][Medline]
  7. Jean D, et al. Loss of AP-2 results in up-regulation of MCAM/MUC18 and an increase in tumour growth and metastasis of human melanoma cells. J. Biol. Chem. (1998) 273:16501–16508.[Abstract/Free Full Text]
  8. Sumigama S, et al. Suppression of invasion and peritoneal carcinomatosis of ovarian cancer cells by overexpression of AP-2alpha. Oncogene (2004) 23:5496–5504.[CrossRef][Web of Science][Medline]
  9. Vernimmen D, et al. Identification of HTF (HER2 transcription factor) as an AP-2 (activator protein-2) transcription factor and contribution of the HTF binding site to ERBB2 gene overexpression. Biochem. J. (2003) 370:323–329.[CrossRef][Web of Science][Medline]
  10. Moniaux N, et al. Complete sequence of the human mucin MUC4: a putative cell membrane-associated mucin. Biochem. J. (1999) 338:325–333.[CrossRef][Web of Science][Medline]
  11. Carraway KL 3rd, et al. An intramembrane modulator of the ErbB2 receptor tyrosine kinase that potentiates neuregulin signaling. J. Biol. Chem. (1999) 274:5263–5266.[Abstract/Free Full Text]
  12. Balagué C, et al. In situ hybridization shows distinct patterns of mucin gene expression in normal, benign, and malignant pancreatic tissues. Gastroenterology (1995) 109:953–964.[CrossRef][Web of Science][Medline]
  13. Andrianafahanana M, et al. Mucin (MUC) gene expression in human pancreatic adenocarcinoma and chronic pancreatitis: a potential role of MUC4 as a tumour marker of diagnostic significance. Clin. Cancer Res. (2001) 7:4033–4040.[Abstract/Free Full Text]
  14. Swartz MJ, et al. MUC4 expression increases progressively in pancreatic intraepithelial neoplasia. Am. J. Clin. Pathol. (2002) 117:791–796.[Abstract/Free Full Text]
  15. Saitou M, et al. MUC4 expression is a novel prognostic factor in patients with invasive ductal carcinoma of the pancreas. J. Clin. Pathol. (2005) 58:845–852.[Abstract/Free Full Text]
  16. Perrais M, et al. Characterisation of human mucin gene MUC4 promoter: importance of growth factors and proinflammatory cytokines for its regulation in pancreatic cancer cells. J. Biol. Chem. (2001) 276:30923–30933.[Abstract/Free Full Text]
  17. Mariette C, et al. Transcriptional regulation of human mucin MUC4 by bile acids in oesophagus cancer cells is promoter dependent and involves activation of phosphatidylinositol 3-kinase signalling pathway. Biochem. J. (2004) 377:1–8.[CrossRef][Web of Science][Medline]
  18. Van der Sluis M, et al. The murine Muc2 mucin gene is transcriptionally regulated by the zinc-finger GATA-4 transcription factor in intestinal cells. Biochem. Biophys. Res. Commun. (2004) 325:952–960.[CrossRef][Web of Science][Medline]
  19. Fauquette V, et al. The antagonist regulation of the human MUC4 and ErbB-2 genes by the Ets protein PEA3 in pancreatic cancer cells: implications for the proliferation/differentiation balance in the cells. Biochem. J. (2005) 386:35–45.[CrossRef][Web of Science][Medline]
  20. Piessen G, et al. Regulation of the human mucin MUC4 by taurodeoxycholic and taurochenodeoxycholic bile acids in oesophageal cancer cells is mediated by hepatocyte nuclear factor 1alpha. Biochem. J. (2007) 402:81–91.[CrossRef][Web of Science][Medline]
  21. Wajapeyee N, et al. Activator protein 2alpha status determines the chemosensitivity of cancer cells: implications in cancer chemotherapy. Cancer Res. (2005) 65:8628–8634.[Abstract/Free Full Text]
  22. Van Seuningen I, et al. The K protein domain that recruits the IL-1 responsive K protein kinase lies adjacent to a cluster of c-Src and Vav SH3-binding sites. Implications that K protein acts as a docking platform. J. Biol. Chem. (1995) 270:26976–26985.[Abstract/Free Full Text]
  23. Singh AP, et al. Inhibition of MUC4 expression suppresses pancreatic tumour cell growth and metastasis. Cancer Res. (2004) 64:622–630.[Abstract/Free Full Text]
  24. Tellez C, et al. Role and regulation of the thrombin receptor (PAR-1) in human melanoma. Oncogene (2003) 22:3130–3137.[CrossRef][Web of Science][Medline]
  25. Zeng YX, et al. AP-2 inhibits cancer cell growth and activates p21WAF1/CIP1 expression. Nat. Genet (1997) 15:78–82.[CrossRef][Web of Science][Medline]
  26. Schwartz B, et al. Loss of AP-2{alpha} results in deregulation of E-cadherin and MMP-9 and an increase in tumorigenicity of colon cancer cells in vivo. Oncogene (2007) 26:4049–4058.[CrossRef][Web of Science][Medline]
  27. Motonaga K, et al. Age related expression of werner's syndrome protein in selected tissues and coexpression of transcription factors. J. Clin. Pathol. (2002) 55:195–199.[Abstract/Free Full Text]
  28. Ruiz M, et al. AP-2{alpha} transcription factor expression is associated with luminal differentiation and is lost in prostate cancer. Clin. Cancer Res. (2001) 7:4086–4095.[Abstract/Free Full Text]
  29. Adriance MC, et al. Downregulation of Muc1 in MMTV-c-Neu tumours. Oncogene (2004) 23:697–705.[CrossRef][Web of Science][Medline]
  30. Mc Pherson LA, et al. Tumour suppressor activity of AP-2{alpha} mediated through a direct interaction with p53. J. Biol. Chem. (2002) 277:45028–45033.[Abstract/Free Full Text]
  31. Loukopoulos P, et al. Orthotopic transplantation models of pancreatic adenocarcinoma derived from cell lines and primary tumours and displaying varying metastatic activity. Pancreas (2004) 29:193–203.[CrossRef][Web of Science][Medline]
  32. Eisold S, et al. The effect of adenovirus expressing wild type p53 on 5 fluorouracil chemosensitivity is related to p53 status in pancreatic cancer cell lines. World J. Gastroenterol. (2004) 10:3583–3589.[Medline]
  33. Lu CD, et al. Loss of p27Kip1 expression independently predicts poor prognosis for patients with resectable pancreatic adenocarcinoma. Cancer (1999) 85:1250–1260.[CrossRef][Web of Science][Medline]
  34. Gysin S, et al. Pharmacologic inhibition of RAF–>MEK–>ERK signaling elicits pancreatic cancer cell cycle arrest through induced expression of p27Kip1. Cancer Res. (2005) 65:4870–4880.[Abstract/Free Full Text]
  35. Minami S, et al. Molecular cloning and characterization of the human p27Kip1 gene promoter. FEBS Lett. (1997) 411:1–6.[CrossRef][Web of Science][Medline]
  36. Yang WC, et al. Targeted inactivation of p27kip1 is sufficient for large and small intestinal tumorigenesis in the mouse, which can be augmented by a western-style high-risk diet. Cancer Res. (2003) 63:4990–4996.[Abstract/Free Full Text]
  37. Velcich A, et al. Colorectal cancer in mice genetically deficent in the mucin Muc2. Science (2002) 295:1726–1729.[Abstract/Free Full Text]
Received February 12, 2007; revised June 4, 2007; accepted July 3, 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Cancer Res.Home page
S. Aubert, V. Fauquette, B. Hemon, R. Lepoivre, N. Briez, D. Bernard, I. Van Seuningen, X. Leroy, and M. Perrais
MUC1, a New Hypoxia Inducible Factor Target Gene, Is an Actor in Clear Renal Cell Carcinoma Tumor Progression
Cancer Res., July 15, 2009; 69(14): 5707 - 5715.
[Abstract] [Full Text] [PDF]

Home page
 Mol Cancer ResHome page
S. Han, X. Sun, J. D. Ritzenthaler, and J. Roman
Fish Oil Inhibits Human Lung Carcinoma Cell Growth by Suppressing Integrin-Linked Kinase
Mol. Cancer Res., January 1, 2009; 7(1): 108 - 117.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
28/11/2305    most recent
bgm158v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Fauquette, V.
Right arrow Articles by Pigny, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fauquette, V.
Right arrow Articles by Pigny, P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?