Altered plakoglobin expression at mRNA and protein levels correlates with clinical outcome in patients with oropharynx squamous carcinomas

Altered Plakoglobin Expression at mRNA and Protein Levels Correlates With Clinical Outcome in Patients With Oropharynx Squamous Carcinomas SILVANA PAPAGERAKIS, MD, MS, PHD, AL-HASSAN SHABANA, BDS, PHD, JOE¨L DEPONDT, MD, MS, PHD, LAURENCE PIBOUIN, MS, PHD, CLAUDINE BLIN-WAKKACH, MS, PHD, AND ARIANE BERDAL, BDS, MS, PHD Previous studies have established that expression of plakoglobin is down-regulated during malignant transformation. The aim of this study was to evaluate for the first time the expression of plakoglobin at the mRNA and protein levels in primary oropharyngeal squamous cell carcinomas (SCCs) and determine the extent to which the patterns of expression correlated with clinical parameters. Plakoglobin expression was evaluated in 37 new tumor cases and normal oral epithelium using immunofluorescence, reverse transcriptase–polymerase chain reaction (RT-PCR), and Northern and Western blotting analysis. The results indicated that the steady-state levels of plakoglobin protein were down-regulated in all tumors compared with normal epithelium. Furthermore, in 87.1% of the tumors, plakoglobin immunoreactivity displayed an abnormal cytoplasmic localization that was inversely correlated with tumor size and directly correlated with a poor clinical outcome for the patient. Northern blotting analysis revealed that down-regulation of mRNA expression occurred in only 65.6% of the tumors, with plakoglobin mRNA levels similar to normal epithelium in the remaining cases. In the tumors expressing mRNA levels similar to those of normal tissue, a 3.7-kb transcript was detected in addition to the expected 3.4-kb transcript observed in normal epithelium. RT-PCR analysis of the 3' untranslated region of

the 3.7-kb plakoglobin mRNA transcript identified a 297-base insertion from ⴙ2369 to ⴙ2666 that had been previously reported only in transformed cell lines (GenBank M23410). Interestingly, the prognosis was poor for patients with tumors expressing both RNA transcripts. These results are consistent with the concept that complex regulation of plakoglobin expression and intracellular routing may contribute to malignant transformation. The study also shows evidence that the level of expression and intracellular localization of plakoglobin may be useful in predicting the course of disease in patients with oropharyngeal SCC. HUM PATHOL 35:75-85. © 2004 Elsevier Inc. All rights reserved. Key words: plakoglobin, catenin, adhesion, RNA, squamous cell carcinoma, oropharynx, prognosis. Abbreviations: APC, adenomatous polyposis coli; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HRP, horseradish peroxidase; PBS, phosphate-buffered saline; PKP, plakophilin; LEF/TCF, lymphoid enhancer factor/T cell factor; RT-PCR, reverse transcriptase–polymerase chain reaction; SCC, squamous cell carcinoma; SDS, sodium dodecyl sulfate; TBS, Tris-buffered saline; UTR, untranslated region.

Squamous cell carcinoma (SCC) of the oropharynx is an aggressive epithelial malignancy that is the sixth most frequent human tumor diagnosed worldwide.1 The survival rate for individuals afflicted with SCC is among the lowest of the major cancers.2 Because it is well recognized that changes in intercellular adhesion accompany tumor dedifferentiation and invasion,3 the mechanisms underlying intercellular adhesion have been extensively studied in tumor cells both in vivo and in vitro. However, molecular mechanisms that contribute to tumor progression remain largely unknown. Plakoglobin, also known as ␥-catenin, is a structural and functional homologue of ␤-catenin and armadillo (arm, the product of the Drosophila segment polarity gene) that has both adhesive and signaling roles.4

Structurally, plakoglobin consists of a central core of 13 arm repeats flanked by an N-terminal domain and a C-terminal domain. The central core repeats of plakoglobin function as binding sites for various cellular proteins, including cadherins, transcription factors of the LEF/TCF1 family, and the tumor-suppressor protein adenomatous polyposis coli (APC).5 The N-terminal domain regulates the stability of plakoglobin, whereas the C-terminal domain has an apparent transactivation function.6 The interaction of plakoglobin and ␤-catenin with cadherins is necessary for the maintenance of intercellular adhesion and function of epithelia.4 More precisely, ␤-catenin regulates the interaction between classical cadherins and the cytoskeleton at adherens junctions, whereas plakoglobin interacts with both classical and desmosomal cadherins.4,7 The interaction of plakoglobin with desmosomal cadherins is essential for desmosome assembly, whereas complexes between plakoglobin and classical cadherins appear to have a regulatory function.8 In normal tissues, plakoglobin is localized almost exclusively in the cell membrane after immunohistochemical procedures. Staining is detected rarely in the cytoplasm and never in the nucleus.9 In contrast, pronounced cytoplasmic staining is observed for both cad-

From the Laboratory of Orofacial Biology and Pathology, INSERM EMI-U 0110, University of Paris 7, IFR 58, Institut Biomedical des Cordeliers, Paris, France, and the Department of Otolaryngology and Head and Neck Surgery, Richat Hospital, Paris, France. Accepted for publication August 4, 2003. Address correspondence and reprint requests to Dr. Silvana Papagerakis, Department of Pediatric Dentistry, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, Mail Code 7888, San Antonio, TX 78284. 0046-8177/$—see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2003.08.018

75

HUMAN PATHOLOGY

Volume 35, No. 1 (January 2004)

following reasons: (1) No carcinoma cells were identified in the frozen sections (n ⫽ 4); (2) different pathologies, such as tuberculosis and adenocarcinomas, were identified in the frozen sections (n ⫽ 5); (3) the sample was too small to cut (n ⫽ 3); and (4) the tumor was localized in additional areas, such as the hypopharynx and larynx (n ⫽ 8). The patients remaining in the study ranged in age from 42 to 81 years (average, 59 years), and the male-to-female ratio was 35:2. The patients’ clinicopathologic data were retrieved from the original medical and pathological reports and recorded (Table 1). No patients in the study presented with contemporary multicentric lesions. The study was double-blinded, and the investigators did not have access to the patients’ clinical data. Serial sections (6 ␮m thick) were obtained from each tumor sample using a cryomicrotome (Bright, Huntington, UK), air-dried, and stored at ⫺80°C until use. Tumor grading was done using every tenth section after staining with toluidine blue. The remaining tissue was subjected to immunofluorescence, biochemical, and molecular evaluation for plakoglobin. All experiments were performed in duplicate.

herins and catenins, including plakoglobin, in primary oral SCC.10-12 Recent literature suggests that cytoplasmic and nuclear plakoglobin complexes integrate signals from Wnt/wingless (wg) proto-oncogenes and the tumor-suppressor protein APC that direct cell fate and regulate cell proliferation.6,13-15 In adults, inappropriate activation of elements of the Wnt cascade is associated with a number of cancers, and several studies implicate ␤-catenin in this process.16 Although the role of plakoglobin in proliferation and tumor progression is less well documented, several facts suggest that plakoglobin may act as a tumor suppressor. First, the plakoglobin gene demonstrates loss of heterozygosity in certain human tumors.17 Second, plakoglobin overexpression can decrease the tumorigenicity of transformed cells.18 Finally, even a modest level of plakoglobin expression restricts the proliferative potential of normal cells in vivo.19 Down-regulation of cadherins and catenins is frequently observed in many types of human cancers, including oral SCC,20 but plakoglobin expression at the level of transcription has not yet been examined in oropharyngeal SCC. The present prospective study investigated the expression of plakoglobin at the mRNA and protein levels of primary SCCs of the oral and pharyngeal mucosa. Expression of the plakophilins (PKP1, PKP2, PKP3, and p0071), other members of the armadillo multigene familly, has been previously characterized in the same tumor samples.21 Data presented here are consistent with the concept that alterations of the complex regulation of plakoglobin expression and intracellular routing may contribute to malignant transformation. The study also suggests that the level of expression and intracellular localization of plakoglobin may be useful in predicting the course of disease in patients with SCC.

Immunofluorescence Sections were stained for plakoglobin using a standard indirect immunofluorescence method as described previously.20 Immunofluorescence staining of tumors was performed with the primary antibody anti-plakoglobin monoclonal antibody (clone PG-15, IgG1 class, 250 ␮g/mL; Transduction Laboratories, Lexington, KY), diluted at 1:500 in phosphatebuffered saline (PBS). The sections were incubated with the primary antibody at 4°C for 16 hours in a moist chamber. The samples were then washed several times in PBS and treated for 1 hour at 37°C with the second antibody, fluoresceinisothiocyanate– conjugated goat anti-mouse antibody (1.0 mg/mL, 1:40 dilution; Southern Biotechnology Associates, Birmingham, AL). The samples received a final wash in PBS before mounting using an aqueous medium (Fluoprep; Biome´ rieux, Marcy-l’Etoile, France). No staining was obtained when nonimmune serum or PBS was used instead of the primary antibody, confirming the specificity of the primary antibody. Biopsy specimens containing surface epithelium were used as positive controls. All specimens were primary tumors that had undergone no earlier treatment that may have affected the immunostaining results. Unfixed material was used to minimize antigen masking or modification that could have altered labeling. Epithelial cells were identified in different sections of each sample using a mouse monoclonal anti-cytokeratin antibody with broad specificity (KL1, ready to use; Immunotech, Marseille, France). The various compartments within the epithelium, confirmed by KL1 staining, are described as basal, suprabasal, and superficial cell layers. Cell labeling was examined using a Leitz DMRB microscope (Leitz, Westlar, Germany) equipped with epifluorescence illumination, oil-immersion PL Fluotar objectives, and automatic photographic capability. Two investigators independently evaluated the immunostaining results, assigning a score based on the extent and intensity of immunoreactivity. Agreement between investigators was ⬎90%. Differences were resolved by consensus. Plakoglobin-positive cells were reported as a percentage of total cytokeratin-positive carcinoma cells. Based on the extent of cell labeling, the immunostaining reaction was classified as homogeneous (⬎50% to 100%), focal (⬍50% to 10%), or negative (⬍10%). Using an intensity scale ranging from 0 to 6, homogeneous staining was scored as 6 (strong), 5 (moderate), or 4 (weak) and focal staining was scored as 3 (strong),

MATERIALS AND METHODS Tumor Specimens Samples of human oropharyngeal cancers were obtained with the signed consent of all patients. The study was conducted in accordance with French bioethical rules. The following criteria were used to select 57 new cases of primary SCC from patients at the Department of Otolaryngology and Head and Neck Surgery, Bichat Hospital (Paris, France) between September 1996 and September 1999: (1) clinical and histological diagnosis of oral and pharyngeal SCC; (2) no previous treatment of the tumor; (3) no evidence of distant metastasis; and (4) a Karnofsky score ⱖ70%.22 Tumor biopsy specimens were obtained from the 57 patients during endoscopic examination or surgical dissection under general anesthesia. Fragments of ⬍5 mm were cut from the lateral side of each tumor and snap-frozen in liquid nitrogen. These fragments usually included both surface and invasive fronts of the tumor. The remainder of the biopsy material was fixed in normal saline solution and sent for routine histopathologic examination and grading by the Department of Pathology, Bichat Hospital. The diagnosis reported by the pathologist was confirmed in our laboratory by analysis of the first and last frozen sections cut from each fragment used in the study. As a result of this process, 20 of the 57 original carcinoma patients were excluded from the study for one or more of the

76

PLAKOGLOBIN EXPRESSION IN SQUAMOUS CELL CARCINOMA (Papagerakis et al)

TABLE 1. Clinicopathological Data for the 37 Patients With Oropharynx SCC and Plakoglobin Protein (Immunofluorescence). Score/Pattern and Western Blot Analysis) and RNA Status

No

Age (years)

Sex

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

51 52 70 51 75 72 42 66 75 55 66 56 55 45 60 51 66 48 61 50 45 44 51 73 72 56 50 54 64 48 81 50 59 61 57 74 68

M M M M M M M F M M M M M M M M M M M M M M M M M M M M M M F M M M M M M

Site

Histology (WHO Classification)

T/N/M Staging

Follow-up (Months)

Survival Status

Oncologic Status

IF Score/Pattern

WB Protein Level

Number of Transcripts/RNA Level

T SP PW T PW PW PW T SP T PW PW PW PW T PW PW T T T SP SP T T PW PW PW PW TA T TA PW PW PW PW T SP

I III II II III III I II II II I III III II II III II I III I III I I III III II I II II II II I II III II III II

4/1/0 1/0/0 1/1/0 4/1/0 1/0/0 2/2/0 1/0/0 2/1/0 4/2/0 4/3/0 2/1/0 2/3/0 1/0/0 4/3/0 2/1/0 2/1/0 1/0/0 2/0/0 4/2/0 2/0/0 3/3/0 3/0/0 4/3/0 3/3/0 2/0/0 2/3/0 1/3/0 1/0/0 1/1/0 1/0/0 4/3/0 4/0/0 4/3/0 4/3/0 4/3/0 1/1/0 3/3/0

29 29 28 28 28 28 27 22 22 22 20 20 10 10 7 29 28 28 27 24 24 24 21 5 12 5 27 29 27 27 27 22 27 12 27 23 22

A A A A A A A A A A A A A A A A A A A A A A A A A D D D A A A A A D A D D

NED NED NED NED NED NED NED NED NED NED NED NED NED NED NED LR LR LR LR LR LR LR LR LR LR LR LR Met/LR Met C Met C LR,Bo M Bo M LR,Bo/Br M Bo/Li M Lu/Li M LR,Lu/Li M, LR,Li M

3m/c 3c 5c 0 1m 2c 1c 0 2m/c 0 1c 3c 2c 5m 2c 1m/c 3m/c 2c 5m 2c 1c 3m/c 0 1m 0 2m/c 2m/c 5m/c 3m/c 5c 4m/c 1m/c 5m/c 2c 0 5m/c 2c

N/A ⫹ N/A 0 ⫹ ⫹ ⫹ 0 ⫹ 0 ⫹ ⫹ ⫹ ⫹⫹ ⫹ ⫹ ⫹ ⫹ N/A ⫹ N/A ⫹ 0 ⫹ 0 ⫹ ⫹ ⫹⫹ ⫹ ⫹⫹ ⫹⫹ ⫹ ⫹⫹ ⫹ 0 N/A ⫹

N/A 1/lower N/A 1/lower 1/lower 1/lower 1/lower 1/lower 1/lower 2/normal 1/lower 1/lower 1/lower 1/lower 1/lower 2/normal 2/normal 2/normal N/A 1/lower N/A 1/lower 0 1/lower 0 1/lower 2/normal 2/normal 2/normal 2/normal 2/normal 1/lower 2/normal 2/normal 1/lower N/A 1/lower

Abbreviations: F, female; M, male; TA, tonsilar area; PW, pharyngeal wall; SP, soft palate; T, tongue; WHO, World Health Organization; NED, no evidence of disease; LR, local recurrence; M, distant metastasis; Br, brain; Lu, lung; Li, liver; Bo, bone; Met, metachronous primary; C, controlled; D, died of disease; A, alive; m, membranous; c, cytoplasmic (in the presence or absence of membrane labeling). IF, immunofluorescence; WB, Western blot; N/A, the tumor sample size did not permit RNA and protein extraction.

Assay Reagent (Pierce, Rockford, IL). All samples were stored at ⫺80°C until use.

2 (moderate), or 1 (weak). A score of 0 indicated no staining. The topographical staining pattern was categorized as typical (membrane staining) or atypical (cytoplasmic staining, with or without cell surface expression) and scored separately.

Western Blotting Equal amounts of protein (40 ␮g/lane) from each sample were electrophoresed in 7.5% sodium dodecyl sulfate (SDS)-polyacrylamide gel under reducing conditions and transferred onto nitrocellulose membranes (Hybond ECL; Amersham, Les Ulis, France). Blots were stained with Ponceau S red (SigmaAldrich, Lyon, France). Nonspecific binding sites were blocked by treating the membranes with 10% non-fat dry milk in Tris-buffered saline (TBS) (137 mmol NaCl, 20 mmol Tris-HCl, pH 7.6) containing 0.1% Tween-20 for 16 hours at 4°C. After blocking, the membranes were incubated with monoclonal anti-plakoglobin antibody (clone PG-15, IgG1 class; Transduction Laboratories, Lexington KY) diluted 1:2,000 in TBS with 0.1% Tween-20 blocking buffer for 1 hour at 37°C. Excess antibody was removed by washing the membranes 3 times for 10 minutes in TBS with 01% Tween-20, followed by 3 times for 10 minutes in TBS. After

RNA and Protein Extraction Because the remaining biopsy material was insufficient for 5 specimens, total RNA and protein were extracted from only 32 of the 37 SCC biopsy specimens previously analyzed by immunohistochemistry (20 with focal heterogeneous staining, 6 with homogeneous staining, and 6 with negative staining). Briefly, frozen tumor sections as well as histologically normal epithelium surrounding the tumor, were homogenized with a Polytron (Paris, France). Total RNA and proteins were isolated using Tri Reagent (Euromedex, Souffelweyershen, France) according to the manufacturer’s instructions. RNA concentration was measured by optical density. RNA integrity was assessed using ethidium bromide staining of 18S and 28S ribosomal RNA after agarose gel electrophoresis. Protein concentration was quantitated using BCA Protein

77

HUMAN PATHOLOGY

Volume 35, No. 1 (January 2004)

linked to the membrane by baking at 80°C for 2 hours. The membranes were prehybridized at 65°C for 30 minutes in 0.2 mol phosphate buffer containing 7% SDS, 1% bovine serum albumin, and 1 mmol EDTA, then hybridized with the radiolabeled plakoglobin probe overnight under the same conditions. Stringency washes were performed with 0.5⫻ SSC (1⫻ ⫽ 1.5 mmol sodium citrate and 15 mmol sodium chloride, pH 7) and 0.1% SDS at 65°C, 3 times for 15 minutes. After the washes, blots were exposed to X-ray film (Eastman Kodak, Rochester, NY) for 1 to 5 days at ⫺80°C. Then all membranes were stripped by washing in 50% formamide and 10 mmol phosphate buffer, pH 6.5, for 30 minutes at 65°C and rehybridized with a GAPDH cDNA probe. Autoradiographs of the hybridized blots were prepared using x-ray film (Eastman Kodak, Rochester, NY). The optical density of the bands was quantitated by computer-assisted video densitometry (Genomic, Visio-Mic I, Collonges sous Sale`ves, France). To compare the steady-state levels of RNA expression between tumor and control tissue, the optical density of the plakoglobin band was normalized to the optical density of the GAPDH band as a control for intersample variation in RNA content.

incubation with HRP-coupled anti-mouse IgG (Amersham), diluted 1:4,000 in blocking buffer for 1 hour at 37°C, and washing as described earlier, immunoreactive proteins were detected by enhanced chemiluminescence (ECL System; Amersham). When necessary, blots were stripped in 62.5 mmol Tris-HCl, pH 6.8, containing 100 mmol 2-␤ mercaptoethanol and 2% SDS, for 30 minutes at 56°C, washed in TBS with 0.1% Tween-20, and reequilibrated overnight in TBS before reprobing. Control blots were probed directly with anti-mouse IgG HRP conjugate without previous incubation with the primary anti-plakoglobin antibody.

Messenger RNA Amplification by Reverse Transcriptase-Polymerase Chain Reaction and Sequencing of the Amplified Products Plakoglobin expression was analyzed by reverse transcriptase-polymerase chain reaction (RT-PCR) amplification. Total RNA (1 ␮g) was reverse-transcribed using a cDNA kit (Invitrogen, Groningen, The Netherlands), according to the manufacturer’s instructions. A 1-␮L aliquot of the resultant complementary DNA was amplified in 50 ␮L of 1⫻ PCR buffer (Invitrogen) containing 100 ␮mol/L deoxyribonucleotide triphosphate (Invitrogen), 1.5 mmol/L MgCl2 (Invitrogen), 0.025U/␮L AmpliTaq polymerase (Invitrogen), 0.1 ␮mol/L forward primer, and 0.1 ␮mol/L reverse primer. The thermocycler was programmed for 35 cycles of 30 seconds at 94°C, 1 minute at 60°C (for F1/R1) or 50 seconds at 55°C (for F2/R2), and 1 minute at 72°C. RNA without previous reverse transcription served as a control for absence of genomic DNA contamination. Primers were designed based on the published sequence for the plakoglobin gene (GenBank M23410-7) and synthesized by Genset SA (Paris, France) and Oligo Express (Montreuil, France). Two different pairs of primers were used to amplify different regions of the plakoglobin gene. For the 1061 to 1915 region, the forward primer (F1) 5'-CGTGCAGATCATGCGTAACTACAG-3' and reverse primer (R1) 5'-ATGTTCTCCACCGACGAGTACAGG-3' were used. For the 2118 to 2804 region, the forward primer (F2) 5'-CTCACCAACTCCCTCTTCAA-3' and reverse primer (R2) 5'-AAAGCAGGAGCAGAACACTA-3' were used. These primers were designed to amplify 854- and 686-bp fragments, respectively, of plakoglobin-complementary DNA. In addition, amplification of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) from the same samples served as a control. All PCR reactions were performed with a PerkinElmer GeneAmp PCR System 2400 (PerkinElmer, Boston, MA). The PCR products were recovered using standard agarose gel (2% wt/vol) electrophoresis, stained with ethidium bromide, gel-purified (QIAquick Gel Extraction kit; Qiagen, Courtaboeuf, France) and directly sequenced by Oligo Express. The obtained sequences were compared with sequences available in databases using the BLAST algorithm23 and aligned with the known plakoglobin sequence using the Clustal-W algorithm.24

Correlation With Clinicopathologic Parameters and Statistical Analysis The immunohistochemical, biochemical, and molecular results were correlated with SCC histological grade and stage, presence of satellite nodules, distant metastasis formation, and patient survival. SCCs were classified as highly differentiated (I), moderately differentiated (II), or poorly differentiated (III) according to the World Health Organization classification system.25 SCCs were staged using the TNM system of the International Union Against Cancer.26 Patient follow-up ranged between 5 and 29 months, with a mean of 21.88 ⫾ 7.44 months. Survival curves were computed by the KaplanMeier method using the StatView computer program (StatView Software, Cary, NC). The association between plakoglobin immunoreactivity and the clinicopathologic parameters of the tumors was analyzed using the ␹2 test. P values ⬍0.05 were considered significant.

RESULTS Clinical Data Thirty-seven patients were followed for a mean of 22 months (Table 1). Local failure occurred in 19 cases, local recurrences developed in 17 cases, and metachronous lesions developed in 3 cases. (One patient developed both a recurrence and a metachronous lesion.) Seven patients developed distant metastases to bone (2 cases), liver (1 case), bone and brain (1 case), bone and liver (1 case), or lung and liver (2 cases). Four patients developed both recurrence and either 1 metastasis (2 patients) or 2 metastases (2 patients). Of the 31 patients alive at the end of the study, 15 were disease-free (5 to 29 months), 10 had recurrent tumors (5 to 29 months), 2 had metachronous lesions (27 months), and 4 had distant metastases (22 to 27 months). The 6 patients who died during the study had 1 or more tumors; 1 patient died with distant metastases (12 months), 2 patients died with recurrences (5 and 27 months), 1 patient died with both a local recurrence and a metachronous lesion (29 months), and 2 patients

Northern Blotting The amplified products obtained from RT-PCR of control samples were separated on 1.2% Tris-acetate-ethylenediamine-tetraacetic acid agarose gel, purified, and used to generate cDNA probes for Northern blotting analyses of the tumor samples. The probes were labeled with (P32) deoxycytidine triphosphate using a random primer radiolabeling kit (Boehringer Mannheim, Meylan, France). Total RNA (10 ␮g/lane) was fractionated on a 1% agarose gel containing 2.2 mol formaldehyde using electrophoresis and transferred onto a Hybond N⫹ membrane (Amersham). The RNA was cross-

78

PLAKOGLOBIN EXPRESSION IN SQUAMOUS CELL CARCINOMA (Papagerakis et al)

FIGURE 1. Immunolocalization of plakoglobin in tumor and normal tissue. Plakoglobin was detected in serial sections of tumor and normal tissues using a standard immunofluorescence method. (A) Plakoglobin exhibited strong immunoreactivity in both spinous and basal cells of the nonneoplastic surface epithelia, as a network pattern at the cell membrane. (B) Dysplastic epithelium exhibited a similar pattern with slightly reduced plakoglobin expression. In contrast, a heterogeneous pattern of varying intensity was observed in (C) severe dysplastic and (E) invasive epithelia, (D) carcinoma in situ, and (F) frank carcinoma. (F) A focal (heterogeneous) pattern was observed in certain tumor tissues; in this frame, only a small island of tumor cells was stained. In C, D, E, and F, note the lack of defined membranous staining in the tumor tissues compared with normal tissue, as well as the cytoplasmic staining. (Original magnification ⫻ 640.)

along with a lack of defined membranous staining were observed in invasive and tumor epithelia (Fig 1D, E, and F). In total, positive immunoreactivity for plakoglobin was observed in 31 SCC samples (Table 1). Of these, 87.1% displayed an atypical subcellular distribution for plakoglobin with positive immunoreactivity in the cytoplasm whether membrane immunoreactivity was present (48% of cases) or not (52% of cases) (Fig 1D and F). Complete absence of immunoreactivity for plakoglobin was seen in 6 cases.

died with distant metastases and a local recurrence (22 and 23 months). Immunofluorescent Analysis of Plakoglobin The immunoreactivity of plakoglobin was examined in SCCs and in normal, dysplastic, and invasive epithelia of the oropharynx. Strong immunoreactivity for plakoglobin was observed in the basal and spinous layers of the normal surface epithelium, with a network pattern seen in the cell membrane (Fig 1A). In contrast, focal areas of immunoreactivity varying in intensity and/or subcellular distribution were observed in mild (Fig 1B) and severe dysplasia (Fig 1C), invasive epithelia (Fig 1E), carcinoma in situ (Fig 1D), and frank SCC carcinoma (Fig 1F). Cytoplasmic staining, indicating an abnormal distribution of plakoglobin,

mRNA and Protein Analysis of Plakoglobin Plakoglobin is an 82-kDa protein encoded by a 3.4-kb mRNA (Fig 2A). Some human tumor cell lines express an additional 3.7-kb mRNA1 that contains an 79

HUMAN PATHOLOGY

Volume 35, No. 1 (January 2004)

80

PLAKOGLOBIN EXPRESSION IN SQUAMOUS CELL CARCINOMA (Papagerakis et al)

insertion of 297 nucleotides in the 3' noncoding region (indicated by “Y” in Fig 2A). Our data indicate that plakoglobin expression at the protein and mRNA levels is particularly complex. Immunohistochemistry revealed important variations in plakoglobin protein levels and distribution in these tissues (Table 1). Western blotting analysis using the same antibody confirmed the specificity of the antibody, because plakoglobin protein migrated identically with an apparent molecular weight of about 82 kDa in both tumoral and normal control epithelium (Fig 2B). Similar to the results of our immunofluorescence analysis. Plakoglobin protein levels were down-regulated in all tumor samples analyzed compared with control samples. Plakoglobin was not detectable by Western blotting in the 6 samples with negative immunoreactivity. Northern blotting and RT-PCR analyses provided the first evidence that plakoglobin mRNA is expressed in oral SCC (Fig 2C, D, and E). Tumor cDNA was amplified using the primer set F1/R1 that overlapped the 3' coding region of plakoglobin mRNA. As expected, these primers amplify a DNA product of about 850 bp. GAPDH was amplified as an internal standard. Interestingly, plakoglobin mRNA expression was detected in all protein-positive tumors and also in 4 of the 6 tumors (tumors 4, 8, 10 and 35) in which protein was not detected by immunofluorescence and Western blotting. No mRNA was detected in the 2 remaining plakoglobin protein–negative tumors (data not shown). Northern blotting analyses detected a single plakoglobin transcript of approximately M(r) 3.4 kb in 19 cases of the tumor samples tested at the RNA level (Table 1). A longer transcript of approximately M(r) 3.7 kb was detected along with the 3.4-kb transcript in the remaining 34.5% (11 cases) of tumor samples. Only the 3.4-kb transcript was detected in normal control epithelium. Densitometric analysis revealed that expression of the 3.4-kb transcript (normalized to GAPDH) was down-regulated in the tumors expressing this transcript exclusively when compared with control epithelium (the 65.5%, Fig 2D). In contrast, in tumor samples expressing both transcripts (the 34.5%), the steady-state levels of the 3.4-kb transcript were similar to the expression levels in normal epithelia. No mRNA was detected by Northern blotting of the 2 tumors that were also negative by RT-PCR. Previous studies using cell lines suggest that the

3.7-kb transcript could result from the modification of the 3' untranslated region (UTR) of plakoglobin mRNA.7 We investigated this possibility using another primer set (F2/R2) that overlaps the 3' UTR of the plakoglobin mRNA. Based on the published sequence of 3.4-kb and 3.7-kb plakoglobin mRNA,7 these primers amplify a 400-bp region of the 3.4-kb transcript and a 700-bp region of the 3.7-kb transcript (Fig 2E). The insertion in the 700-bp amplicon was assessed by sequence analysis, revealing an insertion of 297 nucleotides (position ⫹2369 to ⫹2666; GenBank M23410). All samples examined displayed the smaller band (Fig 2E). In contrast, 34.5% (11 cases) expressed both the 700-bp and the 400-bp cDNA. These samples corresponded exactly to the samples that displayed both the 3.4-kb and 3.7-kb transcripts by Northern blotting. The remaining cases that expressed only the 400-bp band were the same cases that displayed only the 3.4-kb transcript by Northern blotting. In summary, our data indicate that down-regulation of plakoglobin at the protein level in SCC is associated with two different patterns of mRNA expression. On the one hand, there was a decrease (compared with the normal epithelia) in the steady-state mRNA level of the 3.4-kb transcript in tumors expressing a single transcript. On the other hand, in tumors expressing both the 3.4-kb and 3.7-kb transcripts, levels of mRNA for the 3.4-kb transcript were similar to those for normal epithelia. In addition, plakoglobin mRNA expression was detected in 4 of the 6 tumors in which plakoglobin protein expression was not detected by immunofluorescence and Western blotting. Statistical Analysis Clinicopathologic Parameters and Plakoglobin Protein Expression

Interestingly, different correlations between the plakoglobin immunoreactivity score and tumor growth were observed, demonstrating an inverse correlation with tumor size (␹2 ⫽ 4.430; 1 degree of freedom; P ⬍0.03). Immunoreactivity was strongest in the small tumors (T1) and lowest in large tumors (T3 to T4). In fact, 64% of T1 tumors exhibited obvious immunoreactivity for plakoglobin (score 3 to 6), whereas most (73%) of the large tumors (T2 to T4) displayed only focal and weak or negative immunoreactivity (score 0 to

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ FIGURE 2. Plakoglobin protein and mRNA expression in SCC and normal tissue. Normal epithelial tissue (control C) and 5 representative tumor samples (1 to 5) are shown. (A) Plakoglobin is encoded by a 3.487-kb mRNA containing a 3' UTR region of 1133 bp (2354 to 3487). In some tumors, a 297-base insertion (“Y”) is found inside the untranslated region (2369 to 2666). Specific primers used for this study amplify regions within the coding region (F1 to R1) or the UTR (F2 to R2). (B) Plakoglobin appeared more abundant in normal epithelium (C) than in tumor samples (1 to 5). Six of the tumors were negative. One representative negative tumor sample is shown (5). (C) RT-PCR amplification was performed using specific plakoglobin (F1 to R1) and GAPDH primers. The resultant products were resolved on a 2% agarose gel. All the tumors express different amounts of plakoglobin mRNA (1 to 5). Note that tumor sample 5 was positive for mRNA expression but negative for plakoglobin protein (2B). (D) Evaluation of plakoglobin mRNA expression (normalized to GAPDH) revealed that expression levels were lower than in normal tissue (C) in 65.5% of tumors (tumors 3 to 5) and similar to that of normal tissue in the remaining 34.5% of tumors (samples 1 and 2). In addition, 2 different transcripts (3.4 and 3.7 kb) were found in 34.5% of tumors (samples 1 and 2). Normal epithelium (C) expressed only the smaller transcript (3.4 kb). (E) RT-PCR analysis of the 3' noncoding region of the plakoglobin gene that 34.5% of tumors presented both 400and 700-bp products (tumors 1 and 2). Direct sequencing of the amplified products revealed a 297-base sequence (2369 to 2666) inserted with the 3’UTR in the 700-bp amplicon (2A).

81

HUMAN PATHOLOGY

Volume 35, No. 1 (January 2004)

TABLE 2. Correlation Between Plakoglobin Immunofluorescence Score and Patient Survival Immunohistochemistry Score

Alive n (%)

Dead n (%)

Total

0-1 2-6 Total

13 (100%) 18 (75 %) 31

0 (0%) 6 (25%) 6

13 (100%) 24 (100%) 37

ing the follow-up period; 4 patients developed local recurrences, 2 patients developed metachronous lesions, 1 patient developed both a recurrence and a metachronous lesion, 1 patient developed double distant metastasis, and 2 patients developed both a recurrence and either 1 metastasis (1 case) or 2 metastases (1 case). Moreover, the tumor samples obtained from patients belonging to the first group exhibited only an atypical subcellular distribution for plakoglobin. Of these tumors, 73% (8 cases) displayed only focal immunoreactivity or were completely negative for plakoglobin (score 0 to 4). In contrast, 63.16% (12 cases) of the patients from the second group (with tumor samples that displayed no insertion in the plakoglobin 3' UTR and thus had a single band detected by RT-PCR) had very good clinical outcomes during the follow-up period, being alive and disease-free at the end of the study. The 2 patients whose tumor samples were negative at both the mRNA and protein levels (cases 23 and 25) developed local recurrences during the follow-up period.

NOTE: ␹2 ⫽ 3.879; 1 degree of freedom; P ⬍ 0.04.

2). Among large tumors (T2 to T4) staining positive for plakoglobin, 85% displayed an atypical subcellular distribution. These immunohistochemical data indicated that down-regulation of plakoglobin expression in SCC was associated with tumor growth. Patient survival was positively correlated with the plakoglobin immunoreactivity score (Table 2)(␹2 ⫽ 3.879; 1 degree of freedom; P ⬍0.04). Oropharyngeal SCC patients with tumors demonstrating weak immunoreactivity for plakoglobin had poor survival. Analysis of tumor samples from deceased patients revealed that all (100%) displayed an atypical subcellular distribution, lying predominantly in the cytoplasm whether or not membrane immunoreactivity was present. Furthermore, 66.7% of the tumors obtained from patients who died during the follow-up period exhibited only a moderate focal immunoreactivity (score of 2). Of the metastatic tumors, 85.7% exhibited an atypical subcellular distribution for plakoglobin; the remaining tumors (14.3%) were negative. Moreover, 57% of the metastatic tumors displayed only focal immunoreactivity or were completely negative (score 0 to 3). However, log-rank (Mantel-Cox) statistics for metastasis curves indicated that these differences were not statistically significant. The plakoglobin immunoreactivity score was not significantly correlated with other clinical parameters, such as histological differentiation, recurrences, and lymph node invasion. No significant correlations with patient age or sex were noted.

DISCUSSION Plakoglobin is a structural and regulatory constituent of all known plaque-bearing cell– cell junctions, and is the only protein common to desmosomes and all other types of adhesive junctions.4 Recent data suggests that another armadillo protein, p0071, is expressed in both desmosomes and adherens junctions in epithelial cells.27 Cell– cell adhesion plays a critical role in the development and maintenance of epithelial tissue integrity and function. Lack of plakoglobin leads to severe cell adhesion defects in developing heart muscle in mice.28,29 There is evidence that plakoglobin plays an important role in human pathology as well. Naxos disease involves a 2-bp deletion at the 3' end of the plakoglobin gene, coding for a truncated protein (75 kDa).30 Patients with Naxos disease exhibit heart and skin abnormalities similar to those described in mice lacking plakoglobin.28 During cancer progression, disturbances of intercellular adhesion contribute to malignant progression, invasion, and metastasis.3 In a previous study,20 we reported that desmoplakin expression and plakoglobin expression are down-regulated in human oropharyngeal SCC. Here, we showed that plakoglobin normally localized to the cell membrane, is re-

Clinicopathologic Parameters and Plakoglobin mRNA Expression

Using RT-PCR with the F2/R2 primer set that overlaps the 3' UTR of the plakoglobin mRNA, we divided the tumors into 2 groups: (1) tumors displaying both 700- and 400-bp bands (11 cases) and (2) tumors exhibiting only a single, smaller band of 400 bp (19 cases) (Tables 1 and 3).The 700-bp band resulted from the 297 base insertion in the plakoglobin 3' UTR. Patient clinical outcome was positively correlated with the presence of the 297 base insertion in the 3' UTR of the plakoglobin mRNA (␹2 ⫽ 8.320; 1 degree of freedom; P ⬍0.03) (Table 3). Oropharyngeal SCC patients with tumors demonstrating this insertion exhibited poor clinical outcome over the follow-up period. Interestingly, 91% (10 cases) of the tumor samples displaying the 2 bands by RT-PCR were obtained from patients who had an undesirable clinical outcome dur-

TABLE 3. Correlation Between Patient’s Clinical Outcome and Status of the 297 Nucleotides Insertion in the 3' UTR of the Plakoglobin mRNA Insertion of 297 Nucleotides

NED n (%)

LR/M/Met n (%)

Total

Absent Present Total

12 (63.16%) 1 (9%) 13

7 (36.84%) 10 (91%) 17

19 (100%) 11 (100%) 30

NOTE: ␹2 ⫽ 8.320; 1 degree of freedom; P ⬍ 0.03. For abbreviations, see Table 1.

82

PLAKOGLOBIN EXPRESSION IN SQUAMOUS CELL CARCINOMA (Papagerakis et al)

bin may play a role in regulating cell proliferation that is independent of its role in mediating cell– cell adhesion.15 Recent data from studies of human colon cancer indicate that wild-type plakoglobin functions as an oncogene when expression is not regulated.40 In the present study, decreased plakoglobin expression in the membrane and prominent cytoplasmic staining were directly related to tumor size and decreased patient survival. Taken together, these data suggest different functions for plakoglobin that are related to the subcellular distribution of the protein and its tumor specificity. Thus reduced and/or abnormal plakoglobin immunoreactivity in oropharyngeal SCC is indicative of an unfavorable prognosis.20 Although alterations of immunohistochemical expression of plakoglobin are evident in many types of human carcinomas,38,41 the molecular changes leading to down-regulation of the protein in oral SCC remain largely unknown. In this study, down-regulation at the protein level was observed for all tumor samples. This reduced expression at the protein level was correlated with down-regulation of mRNA expression in 65.6% of tumors. Indeed, comparison of our mRNA analysis with the pattern of protein staining clearly shows that downregulation of plakoglobin mRNA is the mechanism most frequently associated with reduced immunoreactivity in these tumors. This conclusion is consistent with recently published reports of plakoglobin mRNA analysis in human cell lines.42,43 These studies reported that down-regulation of plakoglobin expression occurs at the mRNA level in cervical42 and human bladder43 carcinoma cell lines. In both bladder and cervical lines, alterations in plakoglobin gene expression were associated with absent or reduced E-cadherin gene expression.42-45 These results suggest that the alteration of plakoglobin expression occurs at the RNA level, and that E-cadherin may play an important role in this regulation. E-cadherin down-regulation at the protein and mRNA level has been reported in head and neck SCC,41,46 suggesting that E-cadherin down-regulation may accompany plakoglobin reduction in these tumors. Interestingly, a mutation in plakoglobin has been described in a gastric cancer cell line,47 suggesting that abnormal expression of plakoglobin may be due to genetic mutation. However, no mutation or deletion has been reported for genes encoding for any cadherin– catenin complex components in oral SCC.41 In the present study, plakoglobin protein downregulation was not accompanied by mRNA reduction in 34.5% of tumors examined. This suggests that protein down-regulation occurs through at least two different mechanisms, and that posttranscriptional regulation may occur in these tumors. Analysis of these same tumors without RNA reduction (the 34.5%) by Northern blotting and RT-PCR, revealed transcripts of 3.4 and 3.7 kb. Direct sequencing of the corresponding PCR products (Fig 2) indicated that the longer transcript corresponds to an insertion of 297 bases (from ⫹2369 to ⫹2666; GenBank M23410) in the 3' UTR of plakoglobin mRNA. The sequence of this transcript is identical with the sequence previously published by

duced and/or absent in the cell membranes of oropharyngeal SCC, and that this reduction correlates with an undesirable clinical outcome for patients with this disease. A new tumor-suppressor gene has been identified by allelic loss on chromosome 17.31 Abnormalities in chromosome 17 are frequently associated with head and neck SCC.31 Because plakoglobin is located on chromosome 17, band q12-q21,17,32 it is a good candidate for this tumor-suppressor gene. Moreover, loss of heterozygosity on chromosome 17 is observed in breast and ovarian cancers,17 suggesting that plakoglobin may function as a generalized tumor-suppressor gene in man. In experimental animals, plakoglobin overexpression plays a role in tumor suppression.18,33 Interestingly, our analysis reveals that plakoglobin immunoreactivity is inversely correlated with tumor growth. Indeed, plakoglobin immunoreactivity was stronger in smaller tumors, whereas a reduction or complete absence of the protein was seen in larger tumors. Moreover, reduced plakoglobin immunoreactivity was correlated with an unfavorable prognosis. Similar data have been reported for other human carcinomas, including bronchial,34 hepatocellular,35 renal,36 bladder,37 and lung34 cancers. In skin tumors, abnormal expression of plakoglobin (ie, membranous immunostaining partially lost with cytoplasmic and nuclear localization) was previously shown to be the sole predictor of high proliferation, and also was correlated with tumor grade.38 Our data provide the first evidence that plakoglobin absence and/or abnormal distribution is correlated with decreased survival in patients with oropharyngeal SCC. Our immunofluorescence analysis indicated that most of the tumors (87.1%) exhibited an abnormal cytoplasmic distribution of plakoglobin that correlated with a poor outcome for the patients. Interestingly, this abnormal cytoplasmic distribution was observed preferentially in large tumors. Also, it has been reported that reduced plakoglobin expression correlates with metastasis formation in human carcinoma.39 Although our study failed to find a significant correlation between plakoglobin protein expression level and metastasis formation, immunostaining was abnormal (in 85.7%) or absent (in 14.3%) in tumors that metastasized. A pronounced abnormal cytoplasmic distribution of plakoglobin staining was seen in 43% of the tumors that developed distant metastasis. Interestingly, in human breast carcinoma, most of the adhesion molecules down-regulated in the primary tumor were reestablished in the metastases from the same patient. This was true for E-cadherin and for ␣- and ␤-catenin, but never for plakoglobin. The reduction in plakoglobin expression was greater in metastatic lesions than in primary tumors.39 This suggests that the downregulation of plakoglobin expression in human breast carcinomas is not a transient event, although the underlying mechanism remains unknown.39 These reports suggest that investigation of plakoglobin expression in tumors that metastasize from squamous cell carcinoma may be of great interest as well. Recent in vitro observations suggest that plakoglo83

HUMAN PATHOLOGY

Volume 35, No. 1 (January 2004)

Franke et al,7 whose study was the first to report these 2 transcripts in vitro. More precisely, only 2 human carcinoma cell lines were known to express both transcripts. Our study has provided the first evidence of these 2 transcripts in vivo. The 297 base insertion is not present in mRNA from normal human tissue (present study7) or the 3' UTR of rat mRNA (GenBank RNU58858). Computer analysis has determined that the inserted sequence contains a complete open reading frame, suggesting that this sequence could be translated. However, consistent with previous reports, no evidence for plakoglobin protein isoforms was observed in this study. Alternatively, as proposed previously by Franke et al,7 this sequence may represent an intronic region inserted during aberrant alternative splicing. Interestingly, retention of the entire noncoding region of intron 7 of the desmoplakin gene (another desmosomal plaque component) by aberrant splicing has been described in patients with genodermatosis.48 The 3' UTR region of plakoglobin mRNA contains more than 1000 nucleotides. This is also true for almost all known adhesion molecules, including ␣- and ␤-catenin. The size of this region is variable (ranging from 1500 bases for PKP2 to fewer than 100 bases for PKP4). The biological significance of the 3'UTR noncoding region of the adhesion molecules is not presently known. However, evidence is accumulating that strongly implicates the 3' UTR region of mRNA in the regulation of gene expression.49,50 It is now apparent that the 3'UTR region can specifically control the nuclear export, polyadenylation status, subcellular targeting, and rates of translation and degradation of mRNA.49 Our data indicate that tumoral oral epithelium expresses 2 sets of transcripts that differ by the length of their 3' UTR region, and their steady-state levels (transcriptional efficiency and/or mRNA stabilization). Indeed, the steady-state mRNA levels were higher in all tumor samples in which the longer 3' UTR region was observed. A poor clinical outcome was observed for patients with tumors expressing both transcripts. Several other diseases, including neuroblastoma and acute myelogeneous leukemia, are caused by alterations in the 3' UTR that affect the expression of several genes.49-51 Taken together, these findings suggest that the observed perturbations in the 3' UTR region may be responsible, at least in part, for the altered mRNA expression and/or protein distribution in these tumors. In summary, we have demonstrated alterations of plakoglobin at the mRNA and protein levels in oropharyngeal SCC. We also have demonstrated changes in intracellular distribution of this protein. These alterations may affect the intercellular adhesion and signal transduction pathways involved in oral carcinogenesis and cancer progression. Our results imply that regulation at multiple levels controls the amount and distribution of plakoglobin mRNA and protein in epithelium. Aberrant expression of plakoglobin mRNA and protein expression was correlated with a poor outcome for the patients who presented with oropharyngeal SCC, suggesting a key role for plakoglobin in the development of oral cancer.

Acknowledgment. The authors sincerely thank Dr. David Carnes (Department of Periodontology, University of Texas Health Science Center at San Antonio) for the English editing.

REFERENCES 1. Parkin DM, Pisani P, Ferlay J: Estimates of the worldwide incidence of 25 major cancers in 1990. Int J Cancer 80:827-841, 1999 2. Lacy PD, Piccirillo JF, Merritt MG, et al: Head and neck squamous cell carcinoma: Better to be young. Otolaryngol Head Neck Surg 122:253-258, 2000 3. Behrens J: Cadherins and catenins: Role in signal transduction and tumor progression. Cancer Metastasis Rev 18:15-30, 1999 4. Hatzfeld M: The armadillo family of structural proteins. Int Rev Cytol 186:179-225, 1999 5. Ben-Ze’ev A: The dual role of cytoskeletal anchor proteins in cell adhesion and signal transduction. Ann NY Acad Sci 886:37-47, 1999 6. Simcha I, Shtutman M, Salomon D, et al: Differential nuclear translocation and transactivation potential of beta-catenin and plakoglobin. J Cell Biol 141:1433-1448, 1998 7. Franke WW, Goldschmidt MD, Zimbelmann R, et al: Molecular cloning and amino acid sequence of human plakoglobin, the common junctional plaque protein. Proc Natl Acad Sci U S A 86: 4027-4031, 1989 8. Levis JE, Wahl JK III, Sass KM, et al: Cross-talk between adherens junctions and desmosomes depends on plakoglobin. J Cell Biol 136:919-934, 1997 9. Pannone G, Lo Muzio L, Bucci P, et al: Psysiopathology of beta and gamma catenin in the oral epithelium. Minerva Stomatol 47:583-588, 1998 10. Williams HK, Sanders DSA, Jankowski JAZ, et al: Expression of cadherins and catenins in oral epithelial dysplasia and squamous cell carcinoma. J Oral Pathol Med 27:308-317, 1998 11. Bagutti C, Speight PM, Watt FM: Comparison of integrin, cadherin, and catenin expression in squamous cell carcinomas of the oral cavity. J Pathol 186:8-16, 1998 12. Lo Muzio L, Staibano S, Pannone G, et al: Beta- and gammacatenin expression in oral squamous cell carcinomas. Anticancer Res 19:3817-3826, 1999 13. Klymkowsky MW, Williams BO, Barish GD, et al: Membraneanchored plakoglobins have multiple mechanisms of action in wnt signaling. Mol Biol Cell 10:3151-3169, 1999 14. Zhurinsky J, Shtutman M, Ben-Ze’ev A: Plakoglobin and (␤)-catenin: Protein interactions, regulation and biological roles. J Cell Sci 113:3127-3139, 2000 15. Hakimelahi S, Parker HR, Gilchrist AJ, et al: Plakoglobin regulates the expression of the anti-apoptotic protein BCL-2. J Biol Chem 275:10905-10911, 2000 16. Lustig B, Behrens J: The Wnt signaling pathway and its role in tumor development. J Cancer Res Clin Oncol 129:199-221, 2003 17. Aberle H, Bierkamp C, Torchard D, et al: The human plakoglobin gene is localizes on chromosome 17q21 and is subjected to loss of heterozygosity in breast and ovarian cancers. Proc Natl Acad Sci U S A 92:6384-6388, 1995 18. Simcha I, Geiger B, Yehuda-Levenberg S, et al: Suppression of tumorigenicity by plakoglobin: An augmenting effect of N-cadherin. J Cell Biol 133:199-209, 1996 19. Charpentier E, Lavker RM, Acquista E, et al: Plakoglobin suppresses epithelial proliferation and hair growth in vivo. J Cell Biol 149:503-520, 2000 20. Depondt J, Shabana AH, Florescu-Zorila S, et al: Downregulation of desmosomal molecules in oral and pharyngeal squamous cell carcinomas as a marker for tumour growth and distant metastasis. Eur J Oral Sci 107:183-193, 1999 21. Papagerakis S, Shabana AH, Depondt J, et al: Immunohistochemical localization of plakophilins (PKP1, PKP2, PKP3, and p0071) in primary oropharyngeal tumors: Correlation with clinical parameters. HUM PATHOL 34:565-572, 2003 22. Karnofsky DA: Determining the extent of the cancer and clinical planning for cure. Cancer 22:730-734, 1968

84

PLAKOGLOBIN EXPRESSION IN SQUAMOUS CELL CARCINOMA (Papagerakis et al) 23. Altschul SF, Gish W, Miller W, et al: Basic local alignment search tool. Mol Biol 215:403-410, 1990 24. Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673-4680, 1994 25. Pindborg JJ, Reichart PA, Smith CJ, et al: Histological typing of cancer and precancer of the oral mucosa, in Sobin LH (ed): World Health Organization Classification of Tumors. Berlin, Springer-Verlag, 1997, pp 11-13 26. Sobin LH, Wittekind CH (eds): International Union Against Cancer (UICC) TNM Classification of Malignant Tumors. New York, Wiley, 1997 27. Hatzfeld M, Green KJ, Sauter H: Targeting of p0071 to desmosomes and adherens junctions is mediated by different protein domains. J Cell Sci 116:1219-1233, 2003 28. Bierkamp C, McLaughlin KJ, Schwarz H, et al: Embryonic heart and skin defects in mice lacking plakoglobin. Dev Biol 180:780785, 1996 29. Ruiz P, Brinkmann V, Ledermann B, et al: Targeted mutation of plakoglobin in mice reveals essential functions of desmosomes in the embryonic heart. J Cell Biol 135:215-225, 1996 30. McKoy G, Protonotarios N, Crosby A, et al: Identification of a deletion in plakoglobin in arrhythmogenic right ventricular cardiomyopathy with palmoplantar keratoderma and woolly hair (Naxos disease). Lancet 355:2119-2124, 2000 31. Nunn J, Scholes AG, Liloglou T, et al: Fractional allele loss indicates distinct genetic populations in the development of squamous cell carcinoma of the head and neck (SCCHN). Carcinogenesis 20:2219-2228, 1999 32. Cowley CM, Simrak D, Spurr NK, et al: The plakophilin 1 (PKP1) and plakoglobin (JUP) genes map to human chromosomes 1q and 17, respectively. Hum Genet 100:486-488, 1997 33. Ben-Ze’ev A: Cytoskeletal and adhesion proteins as tumor suppressors. Curr Opin Cell Biol 9:99-108, 1997 34. Pantel K, Passlick B, Vogt J, et al: Reduced expression of plakoglobin indicates an unfavorable prognosis in subsets of patients with non–small-cell lung cancer. J Clin Oncol 16:1407-1413, 1998 35. Endo K, Ueda T, Ueyama J, et al: Immunoreactive E-cadherin, alpha-catenin, beta-catenin, and gamma-catenin proteins in hepatocellular carcinoma: Relationships with tumor grade, clinicopathologic parameters, and patients’ survival. HUM PATHOL 31:558565, 2000 36. Buchner A, Oberneder R, Riesenberg R, et al: Expression of plakoglobin in renal cell carcinoma. Anticancer Res 18:4231-4235, 1998 37. Syrigos KN, Harrington K, Waxman J, et al: Altered gamma-

catenin expression correlates with poor survival in patients with bladder cancer. J Urol 160:1889-1893, 1998 38. Papadavid E, Pignatelli M, Zakynthinos S, et al: Abnormal immunoreactivity of the E-cadherin/catenin (␣-, ␤-, and ␥-) complex in premalignant and malignant non-melanocytic skin tumours. J Pathol 196:154-162, 2002 39. Ilyas M: Adhesion molecule expression in breast cancer: The phoenix in tumour metastasis? J Pathol 190:3-5, 2000 40. Kolligs FT, Kolligs B, Hajra KM, et al: Gamma-catenin is regulated by the APC tumor suppressor and its oncogenic activity is distinct from that of beta-catenin. Genes Dev 14:1319-1331, 2000 41. Wijnhoven BP, Dinjens WN, Pignatelli M: E-cadherin-catenin cell-cell adhesion complex and human cancer. Br J Surg 87:9921005, 2000 42. Denk C, Hulsken J, Schwarz E: Reduced gene expression of E-cadherin and associated catenins in human cervical carcinoma cell lines. Cancer Lett 120:185-193, 1997 43. Giroldi LA, Bringuier PP, Shimazui T, et al: Changes in cadherin-catenin complexes in the progression of human bladder carcinoma. Int J Cancer 82:70-76, 1999 44. Bringuier PP, Giroldi LA, Umbas R, et al: Mechanisms associated with abnormal E-cadherin immunoreactivity in human bladder tumors. Int J Cancer 83:591-595, 1999 45. Popov Z, Gil-Diez de Medina S, Lefrere-Belda MA, et al: Low E-cadherin expression in bladder cancer at the transcriptional and protein level provides prognostic information. Br J Cancer 83:209214, 2000 46. Sakaki T, Wato M, Tamura I, et al: Correlation of E-cadherin and alpha-catenin expression with differentiation of oral squamous cell carcinoma. J Osaka Dent Univ 33:75-81, 1999 47. Caca K, Kolligs FT, Ji X, et al: Beta- and gamma-catenin mutations, but not E-cadherin inactivation, underlie T-cell factor/ lymphoid enhancer factor transcriptional deregulation in gastric and pancreatic cancer. Cell Growth Differ 10:369-376, 1999 48. Whittock NV, Ashton GH, Dopping-Hepenstal PJ, et al: Striate palmoplantar keratoderma resulting from desmoplakin haploinsufficiency. J Invest Dermatol 113:940-946, 1999 49. Conne B, Stutz A, Vassalli JD: The 3' untranslated region of messenger RNA: A molecular “hotspot” for pathology? Nat Med 6:637-641, 2000 50. Mignone F, Gissi C, Liuni S, et al: Untranslated region of mRNAs. Genome Biol 3:0004, 2002 51. Mbongolo Mbella EG, Bertrand S, Huez G, et al: A GG nucleotide sequence of the 3' untranslated region of amyloid precursor protein mRNA plays a key role in the regulation of translation and the binding of proteins. Mol Cell Biol 13:4572-4579, 2000

85