| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Human Cancer Biology |
Authors' Affiliations: 1 Department of Clinical Pharmacology, Oxford University, Oxford, United Kingdom; 2 Digestive Diseases Centre, University Hospitals of Leicester, Leicester, United Kingdom; 3 University of Birmingham; 4 Birmingham Heartlands Hospital, Birmingham, United Kingdom; and 5 London Research Institute, Cancer Research UK, London, United Kingdom
Requests for reprints: Mark R. Anderson, Department of Gastroenterology, City Hospital, Dudley Road, Birmingham, B18 7QH, United Kingdom. Phone: 44-121-507-4589; Fax: 44-121-507-5665; E-mail: markanderson55{at}hotmail.com.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Experimental Design: Met and HGF expression in esophageal tissues were assessed using immunohistochemistry and ELISA. Met-positive cell lines (OE33 and SEG1) and a Met-negative cell line (TE7) were incubated with HGF. Real-time reverse transcription-PCR and Western blotting were used to assess levels of E-cadherin expression. Nuclear TCF/ß-catenin signaling was assessed following reporter construct transfection. Agar colony formation was used to assess anchorage-independent growth. A panel of 72 resected esophageal adenocarcinomas were assessed for Met expression by immunohistochemistry and correlated to survival data.
Results: An increased expression of Met was seen along the metaplasia- adenocarcinoma sequence. Met-positive cells showed reductions in E-cadherin mRNA (37% and 69%) and protein expression following stimulation with HGF (P < 0.01). OE33 and SEG-1 showed up to a 2-fold increase in the levels of ß-catenin nuclear signaling (P < 0.01). TE7 only responded when transfected to express Met; E-cadherin expression decreased by 64% (P < 0.01). HGF stimulation led to increased agar colony formation (P < 0.01). Patients with Met-positive tumors showed lower 6-month survival rates after surgical resection than those with Met-negative tumors (P < 0.05).
Conclusions: Met activation induces changes consistent with early invasion, such as down-regulation of E-cadherin, increased nuclear TCF/ß-catenin signaling, and anchorage-independent growth. This is supported by ex vivo data associating Met with reduced short-term survival. Inhibitors of Met may be effective treatment for esophageal adenocarcinoma.
Esophageal adenocarcinoma mostly arises from Barrett's metaplasia (3). The progression from Barrett's metaplasia involves a sequence of changes from the development of low-grade to high-grade dysplasia and adenocarcinoma (4). This sequence is characterized by alterations in the expression of adhesion molecules, growth factors, and tyrosine kinase receptors (5).
The Met receptor is a tyrosine kinase receptor, the product of a proto-oncogene (6). It acts as a receptor for hepatocyte growth factor (HGF), a potent mitogen and pro-motility agent for epithelial cells (7, 8). HGF is primarily produced by mesenchymal cells to act on Met-expressing epithelial cells in a paracrine fashion (9).
Met has roles in gastrointestinal development (10), mucosal healing (11), and epithelial cell migration (12), processes that require disassembly of cell-cell adhesions.
The predominant adhesion protein of epithelial tissue is E-cadherin (13), and this is down-regulated in esophageal cancer (14). E-cadherin binds to ß-catenin at the cell membrane and is linked to the control of ß-catenin–regulated transcription (15, 16). The ß-catenin protein is found in three cellular pools: membranous, cytoplasmic, and nuclear. The translocation among these is tightly regulated (17), and the dynamic equilibrium determines the signaling role (18).
Nuclear ß-catenin is seen in esophageal tumorigenesis (19), and many catenin target genes show increased expression (20–22).
Studies have shown an association between HGF/Met stimulation and increased phosphorylation of ß-catenin in cell lines (23–25).
Studies of the expression of Met in esophageal malignancy showed increased expression in tumors compared with normal mucosa (26–28). In breast cancer, high Met expression has been correlated with lower survival rates (29, 30). The situation in esophageal adenocarcinomas remains to be addressed.
Here, we describe the effect of HGF/Met signaling in esophageal adenocarcinoma cell lines and show a down-regulation of E-cadherin and an increase in ß-catenin nuclear signaling. We assess the effect of HGF/Met activation on the growth of esophageal cells in an anchorage-independent medium. Finally, we investigate the survival rates of patients with esophageal adenocarcinoma when stratified according to levels of different tyrosine kinase (Met, Erb-B2, and Src) expression.
Our data illustrate that the Met receptor is associated with poorer prognosis and as such is a potential key therapeutic target.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Quantitative real-time PCR. RNA isolation was done using Trizol reagent (Invitrogen, Paisley, United Kingdom). cDNA was synthesized from RNA (1 mg/mL) using a cDNA reverse transcription kit (Promega, Southampton, United Kingdom).
All reactions were done on a real-time PCR PE7700 cycler (ABI prism). 18S rRNA was assessed as an internal standard control using 18S-specific VIC-labeled probes (PE Biosystems, Warrington, United Kingdom). The gene-specific primer and probe sequences were as follows: E-cadherin probe, FAM-AAATTCACTCTGCCCAGGACGCGG-TAMRA; E-cadherin forward primer, 5'-GGCGCCACCTGGAGAGA-3'; E-cadherin reverse primer, 5'-TGTCGACCGGTGCAATCTT-3'.
The real-time PCR reaction mix was made up in 1x Mastermix (PE Biosystems).
Each reaction was replicated thrice. Cycle threshold values for E-cadherin and 18S were obtained graphically. Gene expression was normalized to the 18S value. Relative gene expression was then normalized to a value of 1.0 (100%) for the unstimulated control group. Control reactions without cDNA were done as a negative control.
Western blot. Cell lysates were prepared from culture dishes. Proteins were separated on an 8% SDS-polyacrylamide electrophoresis gel and transferred onto Hybond polyvinylidene difluoride nitrocellulose membranes (Amersham Pharmacia, Amersham, United Kingdom) as previously described (34).
Proteins were detected using primary antibodies at the following concentrations: 1 µg/mL (cytokeratin-19 mouse monoclonal IF-15; Santa Cruz Biotechnology, Santa Cruz, CA), 1 µg/mL (Met rabbit polyclonal SC-161; Santa Cruz Biotechnology), 0.1 µg/mL (E-cadherin mouse monoclonal C20820; Transduction Laboratories, Exeter, United Kingdom). Protein bands were visualized using enhanced chemiluminescence reagent (Amersham) on Kodak film.
Equal protein loading was ensured by normalizing to the housekeeping protein cytokeratin-19. Kaleidoscope standards were used to determine protein molecular weights. A homogenate of adipose tissue was used as a negative control for Met. A lysate of cell line MDA-MB-231 was used as a positive control for Met (35).
Transfection assays. TOPFLASH is a reporter construct in which a luciferase gene is located downstream of c-fos promoter containing ß-catenin/TCF–binding sites (36). The amount of ß-catenin–mediated nuclear transcription is inferred by measuring luciferase activity. TOPFLASH was a kind gift from Hans Clevers (Utrecht, The Netherlands).
To control for the level of successful transfection between wells, a ß-galactosidase construct under the control of a constitutively active cytomegalovirus promoter (37) was transfected simultaneously (kind gift from Chris Dawson, University of Birmingham, United Kingdom). Wild-type APC construct was also transfected to see if this altered the level of signaling (kind gift from B. Vogelstein, Baltimore, MD).
The Met receptor plasmid was transfected transiently into receptor-negative cells (TE-7) to see if they would then respond to HGF. The plasmid was a kind gift from Paulo Amati (University of Rome).
Transfection reactions were done using LipofectAMINE 2000 reagent and Opti-MEM (Invitrogen). Cells were cultured for 24 hours after transfection to allow expression of reporter enzymes before assays of luciferase and galactosidase were done as previously described (38).
Negative control groups included cells transfected but not stimulated with HGF and cells treated with transfection reagents but lacking the plasmid. Transfection experiments and stimulations were carried out in triplicate. The luciferase/galactosidase value for each sample was tabulated and expressed as a ratio of the value for unstimulated cells (given a value of 1).
Agar colony assay. To assess the anchorage-independent growth of the cell lines, the ability to form colonies in agar was measured, using methods as previously described (39).
In brief, trypsinized cells (5 x 103/mL) were seeded on the top layer of an agar coated six-well insert and were cultured in DMEM, with or without HGF. HGF was supplemented to the bottom agar layer so that when this had diffused into both layers it would give a final concentration of 100 ng/mL.
The number of colonies per well were counted at day 10 after seeding. Results were expressed as mean number ± SD of colonies per well of triplicates. Wells supplemented with HGF were compared with unsupplemented control wells.
ELISA assay. A commercially available kit (Human HGF Immunoassay DHG00, Quantikine, R&D Systems, Abingdon, United Kingdom) was used according to manufacturer's instructions. HGF concentration was measured in duplicate tissue biopsies from matched fundic, Barrett's, and normal esophageal mucosa from 10 different subjects. Briefly, biopsies were collected and immediately snap-frozen in ornithine carbamyl transferase embedding compound (R.A. Lamb, Eastbourne, United Kingdom) and isopentane (Fisher Scientific, Leicester, United Kingdom). Each biopsy was then sectioned using a cryostat into 100 µL of total cell lysis buffer and assayed for HGF according to the manufacturer's instructions. The final HGF concentration was normalized to the total cellular protein content for each lysed sample.
Human tissues. To assess the expression of tyrosine kinases along the metaplasia-dysplasia-adenocarcinoma sequence, paraffin-embedded biopsies were identified from a patient database at University Hospital Birmingham NHS Trust. In all patients, the samples of Barrett's metaplasia showed intestinal-type metaplasia. The samples of normal gastric mucosa, normal esophageal mucosa, dysplastic Barrett's (confirmed as high grade by two pathologists), and esophageal adenocarcinoma had been confirmed histologically. Ten different samples of each tissue type were used for immunohistochemistry.
To assess the prognostic implication of different kinase expression profiles, samples were obtained from a case series of 72 patients who underwent curative resection of adenocarcinoma of the esophagus at Birmingham Heartlands Hospital between December 1990 and July 1998. No patient received preoperative chemotherapy or radiotherapy. Thoracic esophaegectomy was done with standard two-field resection of regional lymph nodes. Macroscopic tumor was judged to be completely resected in all patients. All specimens were fixed in formalin and processed into paraffin. Three-micrometer sections of the primary cancer were cut for use in immunohistochemistry. Tumours were staged according to the 1987 Unio Internationale Contra Cancrum Classification based on pTNM subsets. Survival times were calculated as date of operation to date of death or date of censor if still alive. All dying patients died of their primary cancer. Causes of death were obtained from hospital records or from the West Midlands Cancer Intelligence Unit. Permission to use the series has been passed by the East Birmingham Local Research Ethics Committee.
Immunohistochemistry. The streptavidin-biotin indirect immunoperoxidase method was used for immunohistochemistry as previously described (34). Slides were incubated with the following primary antibodies: ErbB-2 rabbit polyclonal A00485 (10 µg/mL; DAKO, Ely, United Kingdom), Met rabbit polyclonal SC-161 (67 ng/mL; Santa Cruz Biotechnology), and Src mouse monoclonal SC-5266 (6.7 µg/mL; Santa Cruz Biotechnology).
A slide of human breast adenocarcinoma was simultaneously processed as a positive control. Negative controls consisted of the same tissue but for which the primary antibody was omitted.
To further assess the specificity of the primary antibodies, peptide absorption controls were done. The primary antibody was incubated first with a 5-fold excess solution of the specific blocking peptide (Blocking peptides sc-161 P for Met and sc-5266 P for Src; Autogen, Calne, United Kingdom). For the ErbB-2 antibody, absorption blocking was carried out using a lysate of cell line MCF-7 known to express ErbB-2 (40). The specificity of the antibody was determined by the absence of immunoreactivity on sections previously known to be positive.
All sections were scored by three independent observers (R. Harrison, S. Chowdary, and M. Anderson). Each slide was classed as positive for expression only if strong staining was seen in more than two thirds of cells.
All observers were blind to the clinical data from the prognostic series. A concordance rate of 86% was reached (percentage of the slides on which all three opinions were the same).
Statistics and analysis. Statistical analysis was done using InStat (GraphPad Software, San Diego, CA). Significance was taken at the 5% level. Advice on statistical tests was obtained from R. Joshi and Dr. H. Ross (University of Birmingham).
Unpaired t test was used to determine if statistical significance existed between the means of two data sets. For dose response curves, an initial ANOVA test was done to insure against false positives from multiple t tests. The log-rank test was used to compare overall survival distribution data, and Fisher's exact test was used to assess the proportions of patients surviving to set time points.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
The cell lines were assessed for the expression of the Met receptor by Western blotting, the results of which are shown in Fig. 2 compared with a housekeeping gene cytokeratin-19. Two of the cell lines (OE33 and SEG-1) expressed the Met receptor, but the TE-7 cell line did not. This resulted in the TE-7 cell line acting as an additional negative control group to show that any observed effects of HGF in the other cells were dependent on Met receptor expression.
|
Effect of Met activation on E-cadherin expression: assessment by real-time reverse transcription-PCR. To investigate the functional effects of Met signaling, the effect of HGF stimulation on E-cadherin expression was explored using quantitative real-time reverse transcription-PCR. Cells were stimulated with HGF at 100 ng/mL. RNA was harvested after 30 minutes, 2 hours, and 24 hours. Results were compared with control groups of unstimulated cells for which the media had been changed at the same time points. Levels of E-cadherin mRNA were calculated relative to those for 18S which served as an internal control.
The results are shown in Fig. 3 .
|
SEG-1 cells showed a mean 68% reduction (95% CI, 55-81% reduction; P = 0.008) in mRNA levels after 30 minutes of stimulation, and levels were returning to normal by 2 and 24 hours after stimulation. This dramatic decrease and quick recovery was consistent with the fact that they were also the fastest growing cell line.
TE-7 cells (which lack the Met receptor) showed no significant response to HGF stimulation.
These results show that Met-expressing cells show a reduction in E-cadherin mRNA in response to stimulation with HGF at 100 ng/mL.
In all cell lines, any response to HGF was lost by 24 hours after stimulation. To ensure that this was not simply due to breakdown of the recombinant HGF, a stimulation was done using HGF-supplemented media that had been made up 24 hours before and incubated at 37°C. This still produced a down-regulation and therefore had a half-life of >24 hours; thus, deterioration of the HGF was not a confounding factor.
To assess the effect of different concentrations of HGF, a dose response curve was produced by stimulating OE33 cells with different doses of HGF (1, 10, 50, 100, and 500 ng/mL) and measuring E-cadherin mRNA levels after 2 hours. The minimum effective dose to cause down-regulation was found to be between 50 and 100 ng/mL.
However, these were cells that were routinely cultured in 10% FCS. To investigate whether the amount of calf serum influenced the dose response, the experiment was repeated with OE33 cells grown in only 1% FCS. This showed that reducing the amount of calf serum altered the dose response curve by shifting it to the left, and that 50 ng/mL of HGF was now enough to cause a significant down-regulation of E-cadherin mRNA levels (P = 0.006).
No significant alterations in E-cadherin mRNA levels were seen at any dose used on TE-7 cells, showing that these were truly unresponsive and not simply dose dependent.
To see if the TE-7 cells lack of response was due to lack of the receptor, we induced Met expression using transient transfection with a Met-encoding plasmid and then stimulated the cells with HGF. The baseline level of E-cadherin mRNA was decreased by 64% in Met-TE-7 cells compared with wild-type TE-7 cells (P < 0.0001). A similar decrease was still seen after adding HGF at 100 ng/mL (P < 0.0001).
Effect of met activation on E-cadherin expression: assessment by Western blotting. To assess whether the reductions in mRNA levels led to actual changes in E-cadherin protein expression, the technique of Western blotting was used.
OE33 and TE-7 cell lines were stimulated with HGF at 100 ng/mL, and the cells were then lysed at 12, 24, and 48 hours after stimulation to allow time for any effect on protein expression to be detected. Control cells were given fresh media only and were lysed at the same time points. These represented a positive control group for normal E-cadherin expression in the cell line at each particular time point.
The results for OE33 are shown in Fig. 4 . Only the principal bands are displayed in this diagram. A reduction in E-cadherin expression was evident at 12 hours, but the difference between unstimulated cells and cells receiving HGF was most pronounced at 24 hours. By 48 hours, the cells had recovered the normal level of expression, and no difference was seen between control cells and stimulated cells.
|
Effect of Met activation on ß-catenin signaling. To investigate whether the down-regulation of E-cadherin was linked to a subsequent increase in nuclear signaling via ß-catenin–mediated transcription, we used the transient transfection of reporter construct.
The construct termed TOPFLASH is a reporter in which luciferase expression is controlled by a c-fos promoter containing multiple Tcf-binding sites. Only the binding of ß-catenin/Tcf complexes activates luciferase transcription; thus, an assay of luciferase activity is a quantitative report of the level of transcriptionally active nuclear ß-catenin/Tcf complexes.
Cells with TOPFLASH were stimulated with HGF at 100 ng/mL for 2 hours and compared with control cells given only fresh normal media. A separate set of cells were cotransfected with wild-type APC in addition to the TOPFLASH plasmid and underwent the same stimulation and control experiments. All experiments were done thrice, and the results are displayed as the mean ± 95% CI.
Results for OE33 cells are shown in Fig. 5A . A significant 2-fold increase in ß-catenin–mediated transcription was seen in cells stimulated with HGF for 2 hours. This increase could be partially abrogated by cotransfection with wild-type APC, which increases ß-catenin breakdown in the cytoplasm.
|
Results for TE-7 cells are shown in Fig. 5C. Stimulation with HGF did not cause a significant change in the level of ß-catenin–mediated transcription in any TE-7 group. However, it was noted that the addition of wild-type APC reduced the basal level of ß-catenin–mediated transcription by 50% but was unaffected by the presence of HGF.
Effect of Met activation on anchorage-independent growth. To investigate whether Met activation could influence cell invasion or early metastasis, we assessed the ability of the cell lines to form colonies in an anchorage-independent medium. We used an assay of colony numbers in agar gels with and without HGF supplementation. The concentration of HGF in supplemented agar was 100 ng/mL. The concentration of cells added to the agar when first plated was 5 x 10 3/mL. Colonies were counted after 10 to 14 days. The means of three experiments are given.
The TE-7 cell line that lacked the Met receptor was unable to form any colonies in the agar gels, with or without the presence of HGF.
The OE33 cells and the SEG-1 cells were able to form colonies in agar. An increase in colony formation was seen in response to HGF supplementation in both cell lines. OE33 showed a 37% increase in mean colony numbers per well from 68 (range, 54-78) to 93 (range, 79-105; P < 0.05). SEG-1 showed a 20% increase in mean colony numbers per well from 275 (range, 240-324) to 331 (range, 318-342; P < 0.05).
Therefore, HGF provides an additional survival signal in anchorage-independent media to cells that express the Met receptor.
Association between Met expression and clinical variables. To investigate whether this in vitro data may hold clinical relevance, we assessed the relationship between expression of three different tyrosine kinases and the survival of patients with esophageal adenocarcinoma.
The expression of three tyrosine kinases (ErbB-2, Met, and Src), known to be present in a proportion of esophageal adenocarcinomas and previously linked to ß-catenin signaling, were assessed using immunohistochemistry on 72 cases of esophageal adenocarcinoma for which tumor-node-metastasis staging data and long-term survival figures after surgical resection were known.
There were 60 cases (83%) that were positive for Met expression. The median survival was 12 months in those that were positive and 16 months in those that were negative (not significant). The number of patients alive by the 6-month time point was 38 of 60 positive cases (63%) compared with 11 of 12 negative cases (92%). This difference in short-term survival was statistically significant (Fisher's exact test, P < 0.05).
The proportion of patients that had lymph node spread in the resection specimen (stage N1) was not significantly different in those that were positive for Met (62%) compared with those that were negative (42%). The survival distributions of the two groups are shown in Fig. 6 .
|
There were 61 of the 72 cases (85%) that were strongly positive for Src expression. The median survival and the survival rates at different time points were not significantly different. The proportion of patients that had lymph node spread was not significantly different in those that were positive for Src (59%) compared with those that were negative (55%). Thus, neither ErbB-2 nor Src expression was shown to correlate significantly with prognosis.
Thus, Met expression was the only tyrosine kinase shown to be significantly associated with poorer short-term survival.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
For a protein not only to be expressed early in a pathogenesis but to also have an influence on later survival may seem paradoxical. One reason behind this may be that although the Met receptor is present, the ligands to activate it are expressed at higher levels further along the sequence and may thus exert an influence on later metastasis. The conventional ligand for Met is HGF, and a recent study using reverse transcription-PCR has shown higher levels of HGF expression in Barrett's epithelium compared with squamous mucosa (41), consistent with our findings here. In addition, another study has shown that Met may also be activated by the presence of the transforming growth factor-
and the epidermal growth factor receptor (42).
Data presented here have shown that Met receptor activation in esophageal cell lines is associated with increases in ß-catenin nuclear signaling, which precede a reduced expression of E-cadherin. Activation of this receptor also imparts a pro-proliferative signal to cells growing in an anchorage-independent medium. The dose of HGF needed to observe such effects in our cancer cell model system was between 50 and 100 ng/mL, varying according to experimental conditions. It is not known what physiologic dose range is found in the microenvironment of an esophageal adenocarcinoma, but HGF is also known to be present in saliva (43), which constantly reaches the esophagus. A study on the serum level of HGF in patients with gastric carcinoma found that levels were higher in patients with metastases (44). However, it is local and intratumoral concentrations that are likely to be relevant in vivo as these growth factors act by autocrine and paracrine signaling. An important finding of our report is that we have shown that intramucosal HGF levels in Barrett's metaplasia are raised.
The mechanism of E-cadherin down-regulation is not known. However, the response is time dependent, and this suggests that a repressor protein may be used and ultimately used up. Snail and Sip-1 are two proteins that are known to bind to repressor elements in the cadherin gene promoter region (45, 46), and it is possible that one of these is triggered by Met activation. The results from the real-time reverse transcription-PCR suggest that there is a compensatory increase in E-cadherin expression 24 hours after HGF stimulation. This is partly explained by the time-dependent utilization of the HGF and the metabolic breakdown of this growth factor once the receptor-ligand complex is internalized by the cells. Another explanation would be a gradual decrease in repressor element levels following prolonged receptor activation.
Previously, work on lung cancer cell lines has shown a similar effect on colony formation in agar to the one shown here, with an increased colony number following HGF/Met activation (47). It is known that Met can phosphorylate Akt, which provides resistance to anoikis upon loss of matrix contact (48), and this may be of relevance in promoting anchorage-independent growth.
Finally, we have shown that Met shows increasingly frequent expression in Barrett's dysplasia and adenocarcinoma. It seems to influence patient survival. Those that overexpress Met display worse short-term survival at 6 months after surgery than those that do not express the receptor. It has no demonstrable influence on nodal spread in these patients who underwent resection. It would have been of interest to investigate a link between expression profile and T and M stage, but as all these cases were obtained from operable cases, T1/T2 and M0 was the oncologic grading in the majority of cases at the time of resection. The data presented here need further studies to confirm whether the expression of Met is a robust, independent variable. However, no significant association was seen with the other tyrosine kinases that we assessed, suggesting that this is specific to the Met receptor.
In conclusion, this body of work has increased our understanding of the importance of Met receptor signaling in esophageal adenocarcinoma and has now set the stage for the testing of specific Met receptor inhibitors as potential treatments. Several pharmaceutical companies are currently developing such agents that have selectivity for Met and have been used to reduce cell proliferation or increase apoptosis in vitro (49, 50). It would be timely now to test such drugs in the context of esophageal cell model systems and ultimately in human patients.
We have shown that HGF/Met activity leads to decreased E-cadherin expression, increased ß-catenin signaling, and an increased colony formation of esophageal cells in vitro. The expression of Met in esophageal adenocarcinoma is associated with a poorer prognosis in vivo, and this is now a prime area for targeting therapies.
|
Footnotes |
|---|
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.
Received 5/18/06; revised 7/ 5/06; accepted 8/ 3/06.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
B pathway involving TNF receptor associated factors. Oncogene 1997;14:2899–916.[CrossRef][Medline] in Barrett's oesophagus: a potential novel mechanism of action. Oncogene 2002;21:6071–81.[CrossRef][Medline]B represents the central event in the neoplastic progression associated with Barrett's esophagus: a possible link to the inflammation and overexpression of COX-2, PPAR
and growth factors. Dig Dis Sci 2004;49:1075–83.[CrossRef][Medline]B. J Biol Chem 2002;277:25203–8.This article has been cited by other articles:
|
|
 |
|
  F. Ginty, S. Adak, A. Can, M. Gerdes, M. Larsen, H. Cline, R. Filkins, Z. Pang, Q. Li, and M. C. Montalto The Relative Distribution of Membranous and Cytoplasmic Met Is a Prognostic Indicator in Stage I and II Colon Cancer Clin. Cancer Res., June 15, 2008; 14(12): 3814 - 3822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Ferry, M. Anderson, K. Beddard, S. Tomlinson, P. Atherfold, J. Obszynska, R. Harrison, and J. Jankowski A Phase II Study of Gefitinib Monotherapy in Advanced Esophageal Adenocarcinoma: Evidence of Gene Expression, Cellular, and Clinical Response Clin. Cancer Res., October 1, 2007; 13(19): 5869 - 5875. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
