Send Orders for Reprints to
[email protected] Current Cancer Drug Targets, 2014, 14, 000-000
1
The Role of E-Cadherin Down-Regulation in Oral Cancer: CDH1 Gene Expression and Epigenetic Blockage G. Pannone1,#, A. Santoro1,#, A. Feola2, P. Bufo1,12, P. Papagerakis3, L. Lo Muzio4, S. Staibano5, F. Ionna6, F. Longo6, R. Franco7, G. Aquino7, M. Contaldo8, S. De Maria9, R. Serpico8, A. De Rosa8, C. Rubini10, S. Papagerakis3, A. Giovane2, V. Tombolini11,12, A. Giordano13, M. Caraglia2,13,* and M. Di Domenico2,13,* 1
Department of Clinical and Experimental Medicine - Section of Anatomic Pathology and Cytopathology - University of Foggia – Foggia – Italy; 2Department of Biochemistry, Biophysics and General Pathology, Second University of Napoli – Napoli- Italy; 3Orthodontics and Paedriatic Dentistry, School of Dentistry, University of Michigan- MI- USA; 4 Department of Clinical and Experimental Medicine - Section of Oral Pathology - University of Foggia – Foggia – Italy; 5Section of Pathological Anatomy – Department of Biomorphological and Functional Sciences – University ‘Federico II’– Napoli – Italy; 6Department of Maxillofacial Surgery, Istituto Nazionale per lo Studio e la Cura dei Tumori – Fondazione ‘G. Pascale’ – Napoli – Italy; 7Section of Anatomic Pathology and Cytopathology, Istituto Nazionale per lo Studio e la Cura dei Tumori – Fondazione ‘G. Pascale’ – Napoli – Italy; 8Multidisciplinary Department of Medical-Surgical and Odontostomatological Specialties - Second University of Napoli – Napoli- Italy; 9 Department of Experimental Medicine, Second University of Napoli – Napoli- Italy; 10Section of Anatomic Pathology, Università Politecnica delle Marche, Ancona, Italy; 11Department of Radiology, Oncology and Pathological Anatomy Sciences,La Sapienza University of Rome, Rome, Italy; 12Spencer-Lorillard Foundation Rome, Italy; 13Sbarro Institute for Cancer Research and Molecular Medicine, Center for Biotechnology, Temple University, Philadelphia, PA, USA Abstract: Background: The prognosis of the oral squamous cell carcinoma (OSCC) patients remains very poor, mainly due to their high propensity to invade and metastasize. E-cadherin reduced expression occurs in the primary step of oral tumour progression and gene methylation is a mode by which the expression of this protein is regulated in cancers. In this perspective, we investigated E-cadherin gene (CDH1) promoter methylation status in OSCC and its correlation with Ecadherin protein expression, clinicopathological characteristics and patient outcome. Methods: Histologically proven OSCC and paired normal mucosa were analyzed for CDH1 promoter methylation status and E-cadherin protein expression by methylation-specific polymerase chain reaction and immunohistochemistry. Colocalization of E-cadherin with epidermal growth factor (EGF) receptor (EGFR) was evidenced by confocal microscopy and by immunoprecipitation analyses. Results: This study indicated E-cadherin protein down-regulation in OSCC associated with protein delocalization from membrane to cytoplasm. Low E-cadherin expression correlated to aggressive, poorly differentiated, high grade carcinomas and low patient survival. Moreover, protein down-regulation appeared to be due to E-cadherin mRNA downregulation and CDH1 promoter hypermethylation. In an in vitro model of OSCC the treatment with EGF caused internalization and co-localization of E-cadherin with EGFR and the addition of demethylating agents increased E-cadherin expression. Conclusion: Low E–Cadherin expression is a negative prognostic factor of OSCC and is likely due to the hypermethylation of CDH1 promoter. The delocalization of E-cadherin from membrane to cytoplasm could be also due to the increased expression of EGFR in OSCC and the consequent increase of E-cadherin co-internalization with EGFR.
Keywords: CDH1 methylation, clinical outcome, E-cadherin, EGFR, Epithelial Mesenchymal Transition, Methylation Specific PCR, Oral squamous cell carcinoma, Real-Time PCR. INTRODUCTION Development of malignant tumours is partially characterized by the ability of the tumour cells to overcome *Address correspondence to this author at the Department of Biochemistry, Biophisics and General Patology, Second University of Naples, Via L. De Crecchio, 7, 80138 Naples, Italy; Tel: +390815665871; Fax: +39081 450169; E-mails:
[email protected],
[email protected] # These Authors have equally contributed to the work 1568-0096/14 $58.00+.00
cell-cell adhesion and to invade surrounding tissues. Ecadherin, the main epithelial adhesion molecule [1], has been involved in carcinogenesis because it is frequently lost in human epithelial cancer, including OSCC [2]. Efforts to identify novel molecular predictors of tumour behaviour and therapeutic targets for OSCC have reported the association of several genetic and epigenetic alterations associated with their early and progressive development [3]. Aberrant promoter methylation and the associated loss of gene expression is a common finding in human cancers [4,5]. © 2014 Bentham Science Publishers
2
Current Cancer Drug Targets, 2014, Vol. 14, No. 2
Reduced expression of E-cadherin is regarded as one of the main molecular events involved in dysfunction of the cellcell adhesion system, triggering cancer invasion and metastasis. CDH1 is a tumour suppressor gene located on chromosome 16q22.1 [6]. E-cadherin, the prototypic member of the cadherin transmembrane protein family, is a homodimeric protein, mediating homophilic Ca++-dependent intercellular adhesion in the adherens junctions of the epithelial cell surface [7]. It interacts with similar E-cadherin molecules on opposing cell membrane and, through its cytoplasmic domain, with cytoskeleton actin and catenin proteins [8]. By providing tissue integrity and promoting cell polarity, E-cadherin is involved in the maintenance of the epithelial phenotype. E-cadherin is expressed by most normal epithelial tissue, while numerous clinical studies have proved that the reduction or the loss of E-cadherin expression in many epithelial invasive or in situ tumours is important events in the acquisition of an invasive and metastatic potential [9]. Loss of E-cadherin expression and, hence, cadherin switching is involved in Epithelial-Mesenchymal Transition (EMT) phenomenon, that is a process in which epithelial cells lose their polarity and become more invasive and migratory, promoting tumour progression and metastasis development [10]. EMT is associated with increased invasion and metastases. Down-regulation of E-cadherin expression includes mechanisms like DNA mutation, transcriptional control and promoter methylation [11,12]. Epigenetic alterations comprise heritable changes in gene expression that are not caused by changes in the primary DNA sequence, are increasingly being recognized for their roles in oral carcinogenesis. These epigenetic alterations may change the methylation status of cytosine bases (C) in the context of CpG dinucleotides within the DNA [13,14]. Methylation of clusters of CpGs called “CpG-islands” in the promoters of CDH1 gene has been associated with inherited E-cadherin gene silencing [15]. Saito et al. (1998) have observed that CpG methylation of E-cadherin gene promoter occurred in 17% of OSCC and causes reduction of Ecadherin expression in the tumous, resulting in acquisition of the invasive phenotype [15]. Moreover, in oral cancer Ecadherin protein membrane expression is frequently reduced or lost [13]. An increasing expression of the EGFR and its modified transduction EGFR pathway are represented in a multitude of tumours, indicating a more aggressive phenotipe of these cancers [16,17]. There is a considerable number of evidence that links the regulation of cadherins and the tyrosine kinase receptors pathways. Fluctuating level of transforming growth factor (TGF), EGF, platelet-derived growth factor (PDGF) in tumour microenvironment has a key player in the development of an aggressive phenotype in cancer cells and in EMT. Ligands together with -catenin, integrins 4 and 5 induce the dissolution of cell-cell junctions and the tumour motility which is supported by a network mediated by SNAIL, SNAIL2 and E2 transcription factors [18]. These key downstream effectors include membrane ruffling in cytoskeleton that drive cell polarity and migration [19]. In the present manuscript we discuss about the hypermethylation of the CDH1 promoter and E-cadherin protein
Pannone et al.
down-regulation/delocalization as well as prognostic factors in OSCC. We describe the effects of the demethylating agents that can increase E-cadherin expression, in addition we provide data suggesting that the E-cadherin downregulation could be due to the its interaction with EGFR and successive co-internalization. MATERIALS AND METHODS Study Population and Clinicopathological Data In the present study, we analysed 186 OSCC cases for whom both paraffin-embedded tumour blocks and frozen tissues were available. In details, 186 paraffin-embedded tumour blocks (134 males and 52 females) with age ranging from 28 to 92 years underwent immune-histo-chemistry (IHC) examination; 69 paraffin-embedded tumour blocks underwent methylation-specific PCR (MSP) analysis; 13 frozen tissues were used for molecular investigation. The OSCC tumour sites included tongue (n = 76), lip (n = 35), cheek oral mucosa (n = 40), floor of the mouth (n = 11), gingival (n = 12), palate (n = 6) and retromolar trigone (n = 6). Upon approval by the Ethical Committee of the all Institutions, we identified patients from the Oral Surgery Registry of three Italian University Hospitals. The patients came from various geographical areas of Italy such as north, middle and south, and underwent oral maxillo-facial surgery without previous treatment. All patients received surgical or radio/chemo-therapeutic treatment only with curative intention between 1990 and 2006. The limitation of the material we had available, allowed us to analyze 164 out of the 186 cases by immunohistochemistry (IHC), and 69 by both IHC and MSP to correlate protein expression with promoter methylation. Finally, 13 representative OSCCs were selected for quantitative analysis of E-cadherin mRNA using Real Time RT-PCR. The study patient population consisted of 134 males and 52 females, with a mean age of 70.5 years (range: 28-92 years), who had a complete and long-term follow-up information available. The mean follow-up time of the studied cases was 39.41 months. All patients or their relatives gave their informed written consent. Demographical and clinical data together with follow-up status were extracted from clinical records. The histopathological diagnosis, reporting about grade and stage of all OSCCs, were made in the different three Italian University Hospitals. Tumour extent determined from clinical records, computed tomography and magnetic resonance imaging data, was revised and classified according to the 2002 TNM classification [20]. Special care was taken in assessing tumour nuclear grade on paraffin-fixed, haematoxylin and eosin (H&E)-stained sections and in defining it by appropriate grading system [2]. Immunohistochemistry IHC analysis in different types of OSCC was performed as previously described [21]. Primary monoclonal anti-Ecadherin antibody (BD, Lexington, KY) has been diluted 1: 250 with 0.05 M Tris-HCl buffer pH 7.4 containing 1% bovine serum albumin and incubated overnight. The
CDH1 Gene Expression and Epigenetic Blockage
specificity of the anti-E-Cadherin antibody (Ab) and its IHC utilization technique has been previously described in the literature [4]. The proliferative index of neoplastic epithelial cells was identified in different sections of each sample using a mouse monoclonal anti-Ki67 antibody (anti-Ki67, Cell Marque Corporation, Rocklin, CA). Negative control slides without primary Ab were included for each staining. Two observers evaluated the results of the IHC staining separately. Immunostained cells were counted in at least 10 high power field (HPF) analyzed at optical microscope (OLYMPUS BX41, at x40). The topographical staining pattern was also evaluated and recorded as membranous (M), cytoplasmic (C), or mixed, with prevalent membranous (M/C) or cytoplasmic (C/M) staining. Tumour inflammatory infiltrate has been evaluated according to the criteria of Wada et al. [22]. For each case, the cumulative percentage of positive cells among all sections examined was determined. Inter-rate reliability between the two investigators blindly and independently examining the immunostained sections was assessed by the Cohen’s K test, yielding K values higher than 0.70 in almost all instances. RNA Extraction 13 representative OSCCs were selected for quantitative analysis of E-cadherin mRNA using Real Time RT-PCR. For this molecular investigation we have used frozen tissues. Twenty/forty mg of deep-frozen tissue specimen were used for total RNA extraction. Total RNA was isolated by RNeasy minikit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions. The structural integrity of all tested total RNA samples was verified by agarose gel electrophoresis. Real-Time PCR Real-Time PCR analysis of CDH1 gene was performed by using the iCycler® apparatus (BioRad, Hercules, CA) with sequence-specific primer pairs for the gene tested: respectively, CDH1 forward, 5'-GGC GCC ACC TGG AGA GA-3' and reverse, 5'-TGT CGA CCG GTG CAA TCT T-3'. Amplification of the housekeeping gene glyceraldehydephosphate-dehydrogenase (GAPDH) from the same samples was used as internal control. The primers used were the following: GAPDH forward, 5'-TGG TAT CGT GGA AGG ACT CAT GAC-3' and reverse, 5'-ATG CCA GTG AGC TTC CCG TTC AGC-3'. These primers were designed to amplify 456 bp and 257 bp fragments of CDH1 and GAPDH, respectively, complementary DNA (cDNA). The Real-Time PCR products were electrophoresed on 2% agarose gel containing TAE (standard Tris-Acetate-EDTA electrophoretic buffer). The amplicons of expected size were extracted, purified and controlled for sequence. Results were evaluated by means of ICYCLER IQ Real-Time Detection System Software® (BioRad). Data were calculated on threshold cycle (Ct). We assumed Ct = CDH1 Ct – GAPDH Ct, Ct = pathological Ct – normal tissue Ct, and the fold change of expression as 2-Ct. The last value represents the pathological versus normal mucosa gene expression. Mean values from three independent experiments were taken as results.
Current Cancer Drug Targets, 2014, Vol. 14, No. 2 3
DNA Extraction, Sodium Bisulfite Modification, and Methylation Specific PCR Following careful examination of Haematoxylin-eosin stained slides, we selected tissue sections with the greatest proportion of malignant tissue. Paraffin blocks with corresponding normal epithelium far-away from tumour site were also selected. Five 10 M sections were cut from each formalin-fixed, paraffin-embedded tumour sample and transferred into microcentrifuge tubes. DNA extraction and sodium bisulfite conversion were performed as previously described [4]. All methylation-specific PCRs were optimized to detect >5% methylated substrate in each sample. The primers used for Nested-PCR to flank CDH1 and methylated/unmethylated (M/U) CDH1 internal sequences were the following: CDH1 external primers: forward, 5’GTG TTT TYG GGG TTT ATT TGG TTGT-3’, and reverse, 5’- TAC RAC TCC AAA AAC CCA TAA CTA ACC- 3’; internal primers for methylated CDH1: forward, 5’- TGT AGT TAC GTA TTT ATT TTT AGT GGC GTC3’, and reverse, 5’-CGA ATA CGA TCG AAT CGA ACCG-3’; internal primers for un-methylated CDH1: forward, 5’-TGG TTG TAG TTA TGT ATT TAT TTT TAG TGG TGTT-3’, and, reverse, 5’-ACA CCA AAT ACA ATC AAA TCA AAC CAAA-3’. Cell Culture and Treatment Oral squamous carcinoma cells (KB) were obtained from American Type Culture Collection (ATCC) and grown in Dulbecco minimum essential medium (DMEM) supplemented with 1 mM sodium pyruvate, 10% fetal bovine serum (FBS), 2 mM glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin in a humidified incubator containing 5% CO2 at 37°C. All experiments were performed in quiescent cell growing for 24 h in phenol red-free DMEM (Lonza; Verviers, Belgium) and 5% charcoal-stripped serum (Biowest; Nuaillé, France). Western Blotting and Immunoprecipitation Western-blotting analysis was performed as previously described [23]. Subconfluent KB cells were starved for 24 h and then treated with EGF for 15, 45 and 60 min or with SAM and Azacitidine-S-adenisylmethionine (10-100 nM) for 72h. Cells were lysed and cell extracts were obtained with a Lysis buffer (Tris HCl pH 50mM, EDTA 1mM, Triton 1%, NaCl 150mM. MgCl2 5mM, EGTA 1mM) with phosphatase and protease inhibitors. 50g/ml of protein extract were separated by sodium dodecyl-sulfate polyacrilamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. Filters were stained with a 10% Ponceau S solution to verify equal loading and transfer efficiency before being blocked with 5% milk at room temperature for one hour. Then the filters were incubated overnight with EGFR (Cell Signalling) and E-cadherin (Santa Cruz, CA, USA) for 2 hours. Finally, after incubation with secondary antibodies for one hour and washed throughly, the signal was detected with the aid of a chemiluminescence kit (ECL, Amersham Pharmacia Inc). For immunoprecipitation experiments, lysates were immunoprecipitated with antiEGFR antibodies (Cell Signalling) and pre-immune serum.
4
Current Cancer Drug Targets, 2014, Vol. 14, No. 2
Pannone et al.
Immunocomplexes were incubated for 12h and then supplemented with protein G-Sepharose beads. After 2h, the beads were collected and washed four times with lysis buffer and boiled in Laemmly buffer for immunoblotting analysis [24]. The results were derived from 3 different experiments and the p value 60%
Demographic and clinicopathological features of OSCC patients. Age (Years) 28-92 Gender Male 134
Female 52 Site
Lip
35
Tongue
76
Cheek oral mucosa
40
Oral floor (FOM)
11
Gingiva
12
Palate
6
Retromolar trigone
6
Total
186 TNM Staging
T
N0
T1
75
T2
36
T3
4
T4
N1
N2
Total
9
5
89
15
16
67
1
3
8
18
2
2
22
Total
133
27
26
186
St1
St2
St3
St4
81
35
27
43
Stages
Histological Grade G1
G2
G3
64
87
35
CDH1 Gene Expression and Epigenetic Blockage
Current Cancer Drug Targets, 2014, Vol. 14, No. 2 5
Immunohistochemistry
whereas in 22 cases, death was due to other causes. In 16 cases, information about patients’ follow-up status was not available. We observed that normal epithelium surrounding the lesions showed strong expression of E-cadherin placed in all histological layers with an almost exclusively membrane localization (Fig. 1A). On the other hand, heterogeneous areas of immunoreactivity varying in intensity and/or subcellular distribution were observed in tumour tissues (Fig. 1B-F). E-cadherin localization in tumour cells was prevalently at cytoplasm level, especially in poorly differentiated, aggressive and proliferative areas, detected through evaluation of mitotic rate by Ki-67 immunostaining in corresponding serial sections (data not shown). Well differentiated tumours retained the membrane staining of Ecadherin while in the remaining cases, a mixed membrane and cytoplasm expression was detected. In the whole, Ecadherin positivity has been observed in 152/164 cases of OSCC with frequent cytoplasm localization (73 cases) that was statistically significant compared to the membrane localization of corresponding normal peritumoural oral epithelium (p
