Clinical Cancer Research Vol. 10, 5857-5861, September 1, 2004
© 2004 American Association for Cancer Research
Molecular Oncology, Markers, Clinical Correlates |
SEL1L and Squamous Cell Carcinoma of the Esophagus
Paola Granelli1,
Monica Cattaneo2,
Stefano Ferrero3,
Luca Bottiglieri3,
Silvano Bosari3,
Giuseppe Fichera1 and
Ida Biunno2,4
1 Dipartimento di Scienze Chirurgiche e Trapianti, Ospedale Maggiore Policlinico Milano Istituto di Ricovero e Cura a Carattere Scientifico, Milan; 2
Istituto Tecnologie Biomediche, Consiglio Nazionale delle Ricerche, Segrate-Milan; 3
Dipartimento di Medicina Chirurgica e Odontoiatria, Ospedale S. Paolo e Istituto di Ricovero e Cura a Carattere Scientifico Ospedale Maggiore, Università degli Studi di Milano, Milan; 4
Centro Interdisciplinare Studi Bio-molecolari e Applicazioni Industriali (CISI) Università di Milano, Milan, Italy
 |
ABSTRACT
|
|---|
The gene SEL1L is involved both in human breast and pancreatic cancer progression. It is located on 14q24.3–31, a region known to be lost in invasive cancer of the esophagus. We aimed to assess whether SEL1L could become a useful biomarker for this cancer. We assessed SEL1L mRNA and protein expression in 35 patients and found it to be weak in low-grade and strong in high-grade dysplasia. Although the majority of cancer patients showed differential expression (mRNA and protein) of SEL1L, in five cases it was completely absent; these patients had the worst outcomes. SEL1L immunoreactivity was negative in normal tissue samples from five patients with mild esophagitis as well as in normal mucosa adjacent to the tumor. We hypothesize that SEL1L could influence those cellular changes that mediate the transition from a normal mucosa to a neoplastic lesion and may help in the identification of those patients at higher risk of developing this cancer. The specific impact of SEL1L in esophageal cancer needs further investigation.
|
INTRODUCTION
|
|---|
Several lines of evidence indicate that the gene SEL1L is involved in cancer progression. Indeed, SEL1L plays a role in breast and pancreatic tumor growth and aggressiveness, possibly involving cell-matrix interactions. It has been reported that SEL1L expression decreases breast cancer cell aggressiveness in vivo and in vitro and that down-regulation correlates with poor prognosis (1)
. Generally, SEL1L is abundantly expressed only in the pancreas (2
, 3)
and more specifically in the acini and islets of Langerhans (4
, 5)
. However, 36% of primary pancreatic cancers do not express the gene, and a significant correlation was found with Dpc4. The ectopic and inducible expression of SEL1L in a pancreatic cancer cell line delays tumor growth in nude mice (6)
. SEL1L resides on human chromosome 14q24.3–31 (2
, 4)
, and molecular cytogenetic evaluation of gastric and malignant esophagus revealed loss of chromosome 14q in over 25% of the cardia tumors analyzed. In invasive cancers of the esophagus, loss of this chromosomal band occurs at high rate (7, 8, 9, 10, 11, 12)
.
We aimed to assess whether assessing the pattern of SEL1L expression could improve the efficacy of endoscopic surveillance and differentiate between benign mucosa and premalignant conditions of the esophagus. It is well known that squamous cell carcinoma of the esophagus arises from a series of preinvasive lesions or intraepithelial neoplasia. Dysplasias vary from an abnormal epithelium that can closely resemble the normal cells to an epithelium that is so disorganized from the architectural and cytologic point of view that malignancy is clearly evident.
We analyzed 35 patients with squamous cell carcinoma of the esophagus and found that, contrary to the surrounding normal epithelia, dysplastic and neoplastic cells expressed appreciable levels of the gene. We hypothesize that the presence of SEL1L in esophageal epithelium could mediate changes of the normal mucosa toward a dysplastic and neoplastic condition.
|
MATERIALS AND METHODS
|
|---|
Patient Population.
We studied 35 patients with esophageal squamous cell carcinoma, including 28 males (median age, 64.7 years; range, 40–81 years) and 7 females (median age, 61.3 years; range, 33–76 years). In 9 cases tumors were localized in the upper third of the thoracic esophagus, in 15 cases they were localized in the middle third, and in the remaining 11 patients, the lesion was located in the lower third. In all cases a biopsy of esophageal mucosa was also taken at a remote location (>3 cm) from the tumors. All of the patients underwent diagnostic upper gastrointestinal endoscopy before treatment, either surgical or palliative. Our investigation concerned the evaluation of 18 moderately (grade 2) and 16 poorly differentiated cancers (grade 3). Only 1 patient had an early tumor (T1N0M0), whereas 13 had positive lymph nodes (stages IIB and III). In six patients, distant metastasis occurred (stage IV). Sixteen patients were suitable for surgery (esophageal resection). Nineteen underwent palliation by either adjuvant therapy (chemoradiotherapy) or endoscopic treatments (stents or laser therapy for infiltrating and bulging lesions, respectively). Mortality at 36 months was 43.7% and 89.4%, respectively, in the surgical and palliation groups.
As nonneoplastic controls, five patients (three males and two females) with mild esophagitis were investigated.
Immunoperoxidase Staining.
Paired bioptic specimens were taken with routine forceps. One fragment was immediately placed in a sterile vial and frozen at –80°C for RNA extraction; the second was formalin-fixed and paraffin-embedded. From each paraffin block 4-µm slices were cut from different sections; one section was stained with hematoxylin-eosin for routine light microscopy examination, and the other was used for SEL1L immunohistochemical detection by a specific mouse monoclonal antibody using the immunoperoxidase technique. Briefly, tissue sections were dewaxed, rehydrated in xylene/alcohol, and subjected to antigen retrieval by three 6-min microwave cycles in sodium citrate (pH 6.0). Endogenous peroxidase was blocked by incubation for 10 minutes with 3% hydrogen peroxide in deionized water. Sections were then incubated overnight at 4°C with a mouse monoclonal antibody specific for SEL1L. Antigen-antibody detection was performed with an antimouse peroxidase-conjugated secondary antibody (Envision-HRP mouse; DAKO, Carpinteria, CA). Sections were stained with 3,3'-diaminobenzidine substrate and quickly counterstained with hematoxylin. Control sections were obtained by omitting the primary antibody or by use of an unrelated mouse monoclonal antibody.
SEL1L immunoreactivity was evaluated in at least 200 cells in representative high-power fields and scored as follows: 0, no immunoreactivity; +, <10% immunoreactive dysplastic or neoplastic cells; ++, 10 to 50% immunoreactive cells; +++, >50% immunoreactive cells.
RNA Extraction and Expression Analysis.
Total RNA was extracted from minced bioptic samples by use of the kit from Talent (Trieste, Italy) according to the manufacturer’s instructions.
Total RNA (1 µg) was used in each reverse transcription reaction containing 5 µmol/L MgCl2, 1x reaction buffer [50 mmol/L Tris-HCl (pH 8.8), 8 mmol/L MgCl2, 30 mmol/L KCl, 1 mmol/L dithiothreitol], 1 µmol/L deoxynucleotide triphosphates, 5 units of RNase inhibitor (RNasin), 0.8 µg of oligo-p(dT)15 primer, 1.6 µg of random primer, and 15 units of Avian Myeloblastosis Virus Reverse Transcriptase (Amersham, Cologno Monzese, Italy). The reaction mixture was incubated for 10 minutes at 25°C and for 60 minutes at 42°C. The enzyme was denatured at 99°C for 5 minutes, and the mixture was chilled on ice. PCR amplifications were performed in Perkin-Elmer Corp. (Fremont, CA) thermal cycler using 2 µL of reverse transcription product per reaction. The PCR conditions used, which allowed specific detection of the entire SEL1L cDNA, consisted of 3 minutes at 94°C, followed by 30 cycles at 94°C for 1 minute, annealing temperature for 1 minute at 60°C, and extension at 72°C for 3 minutes, followed by a final extension of 5 minutes at 72°C.
All PCR products were electrophoresed on 1.5% agarose gel and stained with ethidium bromide.
Oligonucleotide primer sequences were as follows: IBD7, 5'-ccggccccgagaggaggatgcgggtc-3'; IBD8, 5'-gctggacccagtgcctattactgtgg-3'; and HPRT, 5'-aattatggacaggactgaacgtc-3' and 5'-cgtggggtccttttcaccagcaag-3'.
|
RESULTS
|
|---|
Immunohistochemical Results.
The histologically normal squamous epithelium present either adjacent to tumors (21 cases) or distant from tumors (35 biopsies) was consistently devoid of SEL1L immunoreactivity. Similarly, the five nonneoplastic controls lacked SEL1L immunoreactivity. SEL1L immunoreactivity was found in 30 squamous cell carcinomas (86%). The immunoreactivity was consistently present in a diffuse pattern in the cytoplasm of neoplastic cells. In 4 cases, a minority of tumor cells were SEL1L immunoreactive (score, +), whereas 10 tumors were ++ (28%) and 16 carcinomas were +++ (46%). Dysplastic foci bordering neoplastic tissue were seen in eight cases. The five cases displaying low-grade dysplasias were all SEL1L immunoreactive, and the staining was restricted to the lower third of the epithelium. Four of these cases were scored +, whereas the remaining one was scored ++. The three cases displaying high-grade dysplasia were SEL1L immunoreactive, with staining scattered through the full thickness of the epithelium. One case was scored ++, and one case was scored +. The results are detailed in Table 1
, and Table 2
lists the tumor/patient data and summarizes the RNA and protein SEL1L values obtained by reverse transcription-PCR and immunohistochemistry. Five patients completely lacked the SEL1L expression both at the RNA and protein levels. Good levels of SEL1L expression was observed in 31% of the patients, who to date are still surviving; the majority of them had stage IIA disease. The remaining 19 (deceased) patients had discrete and heterogeneous SEL1L expression. No apparent and immediate clinicopathologic associations can be drawn from our data.
SEL1L RNA and Protein Analysis.
RNA expression scanning of the 35 squamous cell carcinoma cases analyzed showed an ubiquitous pattern of transcription (Fig. 1)
. Thirty of the samples (Fig. 1
, Lanes 1–4 and 10–22; a representative number of squamous cell carcinoma bioptic fragments) displayed appreciable levels of SEL1L gene; however 5 samples (Lanes 5–9) completely lacked the transcript. The quality of the neosynthesized cDNA was evaluated by the amplification of the HPRT gene (Fig. 1
, bottom gel). The presence or absence of SEL1L was confirmed by immunohistochemical analysis of the same samples. In all of the cases, a small bioptic specimen of normal-appearing epithelium was identifiable and was systematically unreactive (Fig. 2A)
. Neoplastic cells showed a differential pattern of cytoplasmic reactivity. Thirty of the squamous cell carcinoma samples investigated were positive for SEL1L immunoreactivity and confirmed the RNA expression results (Tables 2
and 3
). In these cases, the scattered plasma cells were strongly positive. Typically, low-grade dysplastic cells showed a weak SEL1L cytoplasmic immunoreactivity, generally limited to the deeper epithelial layers, whereas, high-grade dysplasia bordering infiltrative squamous cell carcinomas were strongly immunoreactive (Fig. 2B and C)
. Five of the squamous cell carcinomas analyzed were negative for SEL1L immunoreactivity and confirmed the RNA expression results. Three of the five cases in which the protein was completely absent were stage III, one was stage IIB, and one was stage IV; these patients had the worst prognosis.
|
DISCUSSION
|
|---|
Most benign and malignant neoplasms of the esophagus are epithelial in origin and arise through the progression of premalignant conditions: Barrett’s esophagus for adenocarcinoma and squamous dysplasia for squamous cell carcinoma (13, 14, 15)
. The clinical outcome is very heterogeneous and cannot be predicted satisfactorily by existing clinical or molecular prognostic factors; early detection and treatment is the most important factor in patient survival. Thus, a better understanding of the genetic pathways involved in the progression of normal epithelium to squamous cell carcinoma may aid clinicians in devising more effective management, classifying patient’s malignant risks, and developing better follow-up strategies.
Here we report on the potential use of SEL1L expression as a biomarker of squamous cell carcinoma of the esophagus. SEL1L resides on human chromosome 14q24.3–31, and the loss of chromosome 14q was reported previously in over 25% of cardia tumors. In invasive cancers of the esophagus, loss of this chromosomal band occurs at a high rate (7, 8, 9, 10, 11, 12)
. Immunohistochemical data and reverse transcription-PCR analysis showed loss of SEL1L protein and transcript in 14% of the esophageal carcinomas analyzed; the events behind this silencing are unknown at present. Lack of SEL1L activity can occur through genetic events, including homozygous deletion, intragenic mutation, rearrangement of chromosomal material, or loss of heterozygosity, or through epigenetic changes, such as aberrant methylation of the promoter region. In human cancers, this heritable but nongenetic modification is a powerful mechanism by which tumor suppressor gene activity can be inhibited. In esophageal cancer, the reduction or loss of gene expression by CpG hypermethylation has been reported for the p16, p15, and p14 genes (16)
. The rarity of mutations in the promoter and coding regions of SEL1L together with the presence of a CpG island in the 5' region of the gene makes of SEL1L a methylation-sensitive candidate gene.
We previously reported that a possible role of SEL1L is to modulate the aggressive behavior of human breast and pancreatic cancer. Indeed, relatively moderate levels of the protein decrease anchorage-independent growth of both breast and pancreatic cancer, and this may be mediated through the transforming growth factor-ß/Smad signaling pathway and cell-matrix interactions (1
, 6)
. For squamous cell carcinomas, neoplastic progression is usually characterized by well-defined premalignant stages that progress from low- to high-grade dysplasia. The biological markers detected in association with these events may be used in conjunction with histopathology to help identify individuals who are at an earlier stage of disease and therefore at higher risk to develop the neoplasia. We focused our attention on these premalignant lesions. Normal esophageal mucosa lacks SEL1L expression, as is the case for normal pancreatic ductal cells (6)
and not for normal breast ductal cells (1)
. The presence of the protein in early neoplastic events (low- and high-grade dysplasia) of the esophagus and its persistence in neoplastic cells indicate that the protein is required for mediation of the changes in squamous cell differentiation and suggest that it may be a useful tool to detect dysplasia and to diagnose carcinoma at an early, and possibly treatable, stage. Further studies are necessary, however, to determine whether the effects are unique to esophageal cells or represent a general mechanism initiating the conversion of normal epithelial cell types during mammalian tumorigenesis.
The data presented here suggest that SEL1L in squamous cell carcinoma could play a role in the early phases of tumor formation. Further investigation on the possible role along the pathway from benign to preneoplastic to neoplastic conditions in the gastrointestinal tract are being planned mainly in two different areas. The first is Barrett’s esophagus, a well-known condition involving the appearance of intestinal metaplasia at the esophagogastric junction in patients with gastroesophageal disease. In fact, in the last decade, a dramatically increased incidence of adenocarcinoma of the cardia has been described, which is 30 to 125 times greater among patients with Barrett’s esophagus. In addition, it is known that after surgical correction of gastroesophageal reflux, dysplasia may develop in Barrett’s esophagus, progressing from low- to high-grade dysplasia to adenocarcinoma. The second area of interest is the well-known sequence of adenoma to carcinoma of the colon. Different genetic abnormalities occur along this pathway leading to cancer. A simple immunohistochemical test, sensitive and specific, might help identify patients at higher risk of developing adenocarcinoma. Studies in both of these areas are already in progress.
|
FOOTNOTES
|
|---|
Grant support: M. Cattaneo received a fellowship from Fondazione Italiana Ricerca sul Cancro (FIRC).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Ida Biunno, Via Fratelli Cervi 93, 20090 Segrate, Milan, Italy. Phone: 39-02-26422712; Fax: 39-02-26422770; E-mail: ida.biunno{at}itb.cnr.it
Received 1/14/04;
revised 4/28/04;
accepted 5/25/04.
|
REFERENCES
|
|---|
- Orlandi R, Cattaneo M, Troglio F, et al SEL1L expression decreases breast tumor cell aggressiveness in vivo and in vitro. Cancer Res, 62: 567-74, 2002.[Abstract/Free Full Text]
- Biunno I, Appierto V, Cattaneo M, et al Isolation of a pancreas-specific gene located on human chromosome 14q31: expression analysis in human pancreatic ductal carcinomas. Genomics, 46: 284-6, 1997.[CrossRef][Medline]
- Cattaneo M, Orlandi R, Ronchini C, et al The expression of SEL1L and TAN-1 in normal and neoplastic cells. Int J Biol Markers, 15: 26-32, 2000.[Medline]
- Harada Y, Ozaki K, Suzuki M, et al Complete cDNA sequence and genomic organization of a human pancreas-specific gene homologous to Caenorhabditis elegans sel-1. J Hum Genet, 44: 330-6, 1999.[CrossRef][Medline]
- Donoviel DB, Donoviel MS, Fan E, Hadjantonakis A, Bernstein A. Cloning and characterization of Sel-1l, a murine homolog of the C. elegans sel-1 gene. Mech Dev, 78: 203-7, 1998.[CrossRef][Medline]
- Cattaneo M, Orlandini S, Beghelli S, et al SEL1L expression in pancreatic adenocarcinoma parallels SMAD4 expression and delays tumor growth in vitro and in vivo. Oncogene, 22: 6359-68, 2003.[Medline]
- van Dekken H, Geelen E, Dinjens WN, et al Comparative genomic hybridization of cancer of the gastroesophageal junction: deletion of 14Q31–32.1 discriminates between esophageal (Barrett’s) and gastric cardia adenocarcinomas. Cancer Res, 59: 748-52, 1999.[Abstract/Free Full Text]
- van Dekken H, Alers JC, Riegman PH, Rosenberg C, Tilanus HW, Vissers K. Molecular cytogenetic evaluation of gastric cardia adenocarcinoma and precursor lesions. Am J Pathol, 158: 1961-7, 2001.[Abstract/Free Full Text]
- Riegman PH, Vissers KJ, Alers JC, et al Genomic alterations in malignant transformation of Barrett’s esophagus. Cancer Res, 61: 3164-70, 2001.[Abstract/Free Full Text]
- Doak SH, Jenkins GJ, Parry EM, et al Chromosome 4 hyperploidy represents an early genetic aberration in premalignant Barrett’s oesophagus. Gut, 52: 623-8, 2003.[Abstract/Free Full Text]
- Tselepis C, Morris CD, Wakelin D, et al Upregulation of the oncogene c-myc in Barrett’s adenocarcinoma: induction of c-myc by acidified bile acid in vitro. Gut, 52: 174-80, 2003.[Abstract/Free Full Text]
- Walch AK, Zitzelsberger HF, Bruch J, et al Chromosomal imbalances in Barrett’s adenocarcinoma and the metaplasia-dysplasia-carcinoma sequence. Am J Pathol, 156: 555-66, 2000.[Abstract/Free Full Text]
- Glickman JN. Section II: pathology and pathologic staging of esophageal cancer. Semin Thorac Cardiovasc Surg, 15: 167-79, 2003.[CrossRef][Medline]
- Dolan K, Walker SJ, Gosney J, Field JK, Sutton R. TP53 mutations in malignant and premalignant Barrett’s esophagus. Dis Esophagus, 16: 83-9, 2003.[Medline]
- Spinelli P, Falsitta M. Barrett’s esophagus: ablative endoscopic techniques. Tumori, 89: 117-21, 2003.[Medline]
- Xing EP, Nie Y, Song Y, et al Mechanisms of inactivation of p14ARF, p15INK4b, and p16INK4a genes in human esophageal squamous cell carcinoma. Clin Cancer Res, 5: 2704-13, 1999.[Abstract/Free Full Text]
Top
ABSTRACT
INTRODUCTION
