| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cancer Therapy: Preclinical |
Authors' Affiliations: Departments of 1 Cancer Biology, 2 Urologic Oncology, and 3 Pathology, 4 Brain Tumor Center and Department of Neurosurgery, Departments of 5 Pathology and 6 Urology, The University of Texas M.D. Anderson Cancer Center, Houston, Texas
Requests for reprints: Menashe Bar-Eli, Department of Cancer Biology, Unit 173, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: 713-794-4004; Fax: 713-792-8747; E-mail: mbareli{at}mdanderson.org.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Experimental Design: The kinase domain found within exons 18 to 21 of the EGFR from 11 bladder cancer cell lines and 75 patient tumors were subjected to automated sequencing. EGFRvIII expression was determined by immunohistochemistry using a urothelial carcinoma tissue microarray, and its expression was subsequently verified by reverse transcription PCR, real-time PCR, and Western blot analysis, using an EGFRvIII-transfected glioblastoma cell line and glioblastoma tumors as positive controls.
Results: Our analysis failed to detect mutations within the tyrosine kinase domain of EGFR in the 11 cell lines and 75 patients tested. The initial analysis of EGFRvIII expression by immunohistochemistry revealed that at least 50% of the patient tumors expressed EGFRvIII in a urothelial carcinoma tissue microarray. Conflicting reports exist, however, regarding the extent of EGFRvIII expression in tissues owing to the specificity of the antibodies and the methodologies used. Therefore, we sought to validate this observation by reverse transcription PCR, real-time PCR, and Western blot analysis. In these assays, none of the samples were positive for EGFRvIII except for control transfectants and glioblastomas.
Conclusions: When our results are taken together, we conclude that alterations within the tyrosine kinase domain and expression of EGFRvIII are rare events in bladder cancer. The present study has clinical implications in selecting tyrosine kinase inhibitors for the therapy of urothelial carcinoma.
Recently, point mutations or short deletions within the kinase domain of EGFR have been reported in lung, head and neck, and colorectal cancers (4, 5, 10–13). All of these alterations occur in exons 18 to 21 of EGFR. These mutations within the kinase domain of EGFR raised considerable interest because they have been reported to correlate, in some patients with NSCLC, with clinical responses to the tyrosine kinase inhibitor gefitinib (ZD1839, Iressa; refs. 4, 5, 14). Clinical trials with gefitinib in patients with NSCLC revealed a low response rate in most studies carried out in the U.S. and Europe, whereas a higher response rate was reported in Japan, especially among nonsmoking women with EGFR mutations.
In addition to mutations within the kinase domain of EGFR, the most frequent alteration, the EGFR variant III (EGFRvIII), is characterized by a consistent and tumor-specific in-frame deletion of 801 bp from the extracellular domain. This truncation splits a codon and results in a novel glycine at the fusion junction (15, 16). The resulting protein has a constitutively active tyrosine kinase (17) that greatly enhances the tumorigenicity of these cells (18, 19). The EGFRvIII form is in 40% of glioblastomas, but not in any normal tissues, making it an ideal target for tumor-specific immunotherapy for glioblastomas (15, 20–22). Indeed, the expression of EGFRvIII has been clinically correlated with enhanced tumor cell growth, invasion, and metastasis (23, 24), due to its constitutive kinase activity and its ability to deregulate signal transduction and cell cycle checkpoints (24–27). Furthermore, it has recently been reported that the neoplastic phenotype of EGFRvIII is relatively resistant to gefitinib and that higher doses, repeated dosing, and longer exposures are required to decrease receptor phosphorylation (28).
Several studies have shown that gefitinib can inhibit bladder cancer cell proliferation and promote apoptosis in vitro (29–32), supporting the idea that a subset of patients might benefit from treatment with this drug. In this study, we sought to examine the status of EGFR mutations/truncations in gefitinib-responding and nonresponding urothelial carcinoma cell lines and in the tumor specimens of patients with bladder cancer. We report that EGFR mutations were very rare within the kinase domain, and thus, must not account for the response of urothelial carcinoma cells to gefitinib. Moreover, the expression of EGFRvIII in bladder cancer was extremely rare, and therefore, does not contribute to the malignant phenotype of this tumor.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Human tissue samples. Mutational analysis of EGFR was done on DNA extracted from 75 frozen human bladder tumor samples. The immunohistochemical expression patterns of EGFR were analyzed in a tissue microarray containing 251 samples of bladder tumors. The expression levels of full-length and truncated forms of EGFR were verified by Western blotting, quantitative real-time PCR, and reverse transcription PCR in 24, 12, and 9 samples, respectively.
The cohort of patients selected for the different experiments (i.e., mutational analysis, immunohistochemistry, PCR, real-time PCR, and Western blot) was chosen at random and was based on the availability of DNA, RNA, and protein. Among the 75 cases chosen for sequencing, 21 samples were also included in the tissue microarray slide. These 21 samples were also used for reverse transcription PCR and real-time PCR analysis. Eight of the 24 samples analyzed by Western blot were included in the tissue microarray. There were no patients treated with gefitinib in this cohort.
For DNA extraction, tumor tissue samples from transurethral resections were grossly dissected and the presence of tumor tissue in a sample selected for DNA extraction was confirmed on parallel frozen section. The portion of the tissue containing predominantly tumor cells (
90%) was dissected, cut into small pieces, and transferred to a conical tube containing 5 mL of PBS.
The 12 paired samples of adjacent urothelium and tumor were prepared from cystectomy specimens as previously described (36). In brief, one representative section from the central area of grossly identified tumor was removed, cut into small pieces, and transferred to a conical tube containing 5 mL of PBS. The tube was mechanically agitated for 5 minutes using a vortex mixer to release the tumor cells. The presence of the tumor in the tissue was confirmed by microscopic analysis of a parallel frozen section stained with H&E. Cell suspensions of urothelium adjacent to tumors were prepared by mechanical scraping of the mucosal surface with a razor blade. First, a selected adjacent mucosal area was marked with India ink and its surface urothelium was scraped with a razor blade. Multiple histologic sections of the scraped area were submitted for microscopic examinations to rule out contamination with invasive tumor or with grossly occult microscopic papillary lesions. The samples were then transferred to conical tubes containing PBS, and their purity was determined by cytologic examinations of cytospin preparations. The presence of cells of interest in all samples was confirmed by microscopic analysis of cytospin preparations. The baseline expression levels of EGFR were tested on cell suspensions of normal urothelial cells prepared from the ureters of nephrectomy specimens done for other medical reasons.
Only those tissue samples that yielded >90% microscopically recognizable intact normal, dysplastic, or tumor cells were used for DNA and RNA extraction. Six transitional cell carcinomas were classified according to the three-tier WHO histologic grading system (grades 1-2, low grade; grade 3, high grade) and growth pattern (papillary versus nonpapillary; ref. 37). The levels of invasion were defined according to the tumor-node-metastasis staging system (38). Stage T1 (lamina propria invasion) was divided into T1a (no muscularis mucosae invasion) and T1b (muscularis mucosae invasion), which have a significantly higher risk or progression. So, as in our previous publications, the tumors were dichotomized into superficial (Ta-T1a) and invasive (T1b and higher) groups (39).
Mutational analysis. DNA was isolated from tumor specimens using the QIAmp DNA kit, whereas the DNA from the cell lines was isolated by a genomic DNA extraction kit (Qiagen, Valencia, CA). PCR was used to amplify the four exons comprising the kinase domain of the EGFR gene. The primer pairs used were previously described by Lynch et al. (4) and generated by Integrated DNA Technologies, Inc. (Coralville, IA): exon 18, 5'-CAA ATG AGC TGG CAA GTG CCG TGT C-3' (sense) and 5'-GAG TTT CCC AAA CAC TCA GTG AAA C-3' (antisense); exon 19, 5'-CCT TAG GTG CGG CTC CAC AGC-3' (sense) and 5'-CAT TTA GGA TGT GGA GAT GAG C-3' (antisense); exon 20, 5'-CCA TGA GTA CGT ATT TTG AAA CTC-3' (sense) and 5'-CAT ATC CCC ATG GCA AAC TCT TGC-3' (antisense); and exon 21, 5'-CTA ACG TTC GCC AGC CAT AAG TCC-3' (sense) and 5'-GCT GCG AGC TCA CCC AGA ATG TCT GG-3' (antisense). PCR was done using Taq DNA polymerase (Promega, Madison, WI) with the following cycle conditions for exons 18, 20, and 21: 5 minutes of denaturation at 94°C, followed by 35 cycles of 94°C for 30 seconds, 58°C for 30 seconds, and 72°C for 1 minute, followed by an additional extension for 5 minutes at 72°C. The PCR reaction for exon 19 was done with Titanium Taq DNA polymerase (Clontech, Mountain View, CA) with the following cycle conditions: 3 minutes of denaturation at 95°C, followed by 30 cycles of 95°C for 30 seconds, 70°C for 30 seconds, and 72°C for 1 minute then followed by an additional extension for 5 minutes at 72°C. Nested PCR primers were used for sequencing for exons 18, 20, and 21: exon 18, 5'-CAA GTG CCG TGT CCT GGC ACC CAA GC-3' (sense) and 5'-CCA AAC ACT CAG TGA AAC AAA GAG-3' (antisense); exon 20, 5'-GAA ACT CAA GAT CGC ATT CAT GC-3' (sense) and 5'-GCA AAC TCT TGC TAT CCC AGG AG-3' (antisense); exon 21, 5'-CAG CCA TAA GTC CTC GAC GTG G-3' (sense) and 5'-CAT CCT CCC CTG CAT GTG TTA AAC-3' (antisense). For sequencing, exon 19 primers were the same as amplification primers. Prior to sequencing, PCR products were purified using QIAquick PCR Purification Kit (Qiagen). The purified PCR product was then subjected to automated sequencing using 3730xl DNA Analyzer (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions.
Immunohistochemistry. The tissue microarray was assembled as previously described (36). In brief, histologic slides from 251 bladder cancers and 4 normal urotheliums were reviewed, and the most representative, well-preserved areas of the tumor tissue were selected and marked. The donor paraffin blocks were punched in areas of interest using a microarray instrument (Beecher Instruments, Inc., Sun Prairie, WI) and 0.6 mm cores of the tumor tissue were transferred to a recipient block containing 85 noninvasive low-grade (grade 1-2) papillary carcinomas, 26 noninvasive high-grade (grade 3) papillary carcinoma, and 140 invasive high-grade (grade 3) nonpapillary carcinomas. Overall, the microarray contained 111 superficial (Ta-T1a) and 140 invasive (T1b and higher) transitional cell carcinomas of the bladder. The tissue microarrays were deparaffinized in xylene and rehydrated in graded ethanol. Antigen retrieval was done using a 0.01 mol/L sodium citrate buffer (pH 6.0) and microwaving on high for 30 minutes. Endogenous peroxidase activity was quenched with 0.6% hydrogen peroxide. Specimens were blocked with 5% normal goat serum. The samples were incubated for 2 hours at 4°C with anti-EGFRvIII antibody Ab-18 (Neomarker, Union City, CA) in 1:20 dilution with PBS. The primary antibody was washed with PBS and incubated with a sheep anti-mouse biotinylated secondary antibody (Amersham Biosciences, Buckinghamshire, England) in a 1:250 dilution with PBS. Streptavidin-horseradish peroxidase and then 3-amino-9-ethylcarbazole were added. The samples were counterstained with Gill's no. 3 hematoxylin and then washed with water and mounted with universal mount.
Real-time PCR. RNA (2 µg/50 µL) from urothelial cell lines (UM-UC-1, UM-UC-2, UM-UC-3, UM-UC-6, UM-UC-9, UM-UC-10, UM-UC-12, UM-UC-13, UM-UC-14, UM-UC-15, UM-UC-16, and UM-UC-17) as well as tumor and adjacent specimens were made into cDNA using TaqMan reverse transcriptase reagents (Applied Biosystems). The cDNA was then diluted to a 1:10 concentration for subsequent reverse transcription PCR. RNA was extracted from the glioblastoma tumor samples (4A2, 5A2, A7, being positive for EGFRvIII expression and B1, C2, E1, being negative for EGFRvIII)7 using High Pure RNA Paraffin Kit (Roche, Indianapolis, IN). The primers and fluorescence probes were designed according to the Assays-by-Design Service provided by Applied Biosystems using primers specific to exons 2 to 3 and exons 13 to 14 of EGFR (assay numbers, Hs01076087_m1 and Hs01076071_m1, respectively). Reaction components for reverse transcription PCR included TaqMan Universal PCR Master Mix (Applied Biosystems), 20x Assay Mix (Applied Biosystems), RNA-free water, and the diluted cDNA (1:10 and 1:1,000). The amplifications were carried out in an Applied Biosystems 7700 Prism reverse transcription PCR device using the following temperature profile: denaturation at 95°C for 15 seconds, annealing and extension at 60°C for 1 minute. Amplifications were done in triplicate for each gene and compared with three housekeeping genes (18S rRNA, cyclophilin A, and human acidic ribosomal phosphoprotein PO), which served as internal controls. The relative expression of EGFR wild-type and total forms were calculated and quantified relative to 18S rRNA. In our experience, 18S provides very consistent results in the real-time PCR assay, and therefore, we used it as a standard in making relative expression calculations. The relative expression of the truncated form of EGFR was calculated by subtracting the wild-type form of EGFR from the total EGFR-relative expression in the samples.
Reverse transcription PCR analysis. RNA was extracted from the tumors and cell lines using TRIzol reagent protocol (Invitrogen) and was converted to cDNA using high-capacity cDNA archive kit (Applied Biosystems). Using 150 ng of cDNA in a 25 µL reaction AmpliTaq Gold PCR Mix (Applied Biosystems), PCR was done with the following cycle conditions: 3 minutes of denaturation at 94°C, followed by 35 cycles of 94°C for 45 seconds, 60°C for 45 seconds, and 72°C for 2 minutes followed by an additional extension for 7 minutes at 72°C. EGFR primers used were as previously described (40) and purchased from Integrated DNA Technologies: the forward primer sequence was 5'-ATG CGA CCC TCC GGG ACG-3' and the reverse sequence was 5'-GAG TAT GTG TGA AGG AGT-3'. For loading control, the glyceraldehyde-3-phosphate dehydrogenase forward primer was 5'-GAG CCA CAT CGC TCA-3' and the reverse primer was 5'-CTT CTC ATG GTT CAC ACC C-3'. The amplified fragments were analyzed on 2% agarose gel stained with ethidium bromide and visualized using UV light.
Western blot. Whole-cell lysate was prepared using Triton X-100 lysis buffer at 4°C. Lysates were cleared by a 10-minute centrifugation at 14,000 x g, and protein concentration was determined by the Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA). Each sample contained 100 µg, which was boiled for 5 minutes at 100°C in buffer containing SDS, and was separated by 6% and 10% SDS-PAGE. The proteins were electrophoretically transferred to a polyvinylidene difluoride membrane (Millipore Corp., Bedford, MA). The membrane was blocked overnight in blocking solution [5% nonfat dry milk/TBS plus 0.1% Tween 20 (TTBS)], and then incubated with two primary EGFR antibodies (2232 and Ab-15) targeting the cytoplasmic domain at 1:1,000 dilution (Cell Signaling, Inc., Beverly, MA and Neomarker, respectively). The primary antibody was washed in TTBS and then incubated in secondary antibody peroxidase-conjugated donkey anti-rabbit or sheep anti-mouse (Amersham Biosciences) at a 1:2,000 dilution in 0.5% nonfat dry milk/TTBS for 2 hours and then washed with TTBS. Proteins were visualized using enhanced chemiluminescence reagent (Amersham Biosciences) according to the manufacturer's instructions, followed by exposure to X-ray film. Equal loading of protein samples was confirmed by incubating membranes with a primary antibody for pan-actin (Ab-5) at 1:1,000 dilution (Neomarker).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
50% of the tumor samples stained positive for EGFRvIII (Fig. 2A and B
). The four normal urothelium on the array also stained positive (Fig. 2C), a contradiction with previous reports that normal tissue did not express EGFRvIII (21). This immunohistochemistry staining raised concerns about the specificity of the antibody used to detect EGFRvIII expression and prompted us to validate its expression by real-time PCR, reverse transcription PCR, and Western blot analysis.
|
30% of the samples. To confirm the lack of EGFRvIII expression in the tested samples, reverse transcription PCR of wild-type and truncated EGFR as well as Western blot analysis was done. For reverse transcription PCR, the U87MG-EGFRvIII cell line was again used as a positive control. The 352 bp band corresponded with the expression of EGFRvIII and was only seen in the positive control, U87MG-EGFRvIII (Fig. 4
). Wild-type EGFR is detected in the 253J B-V cell line (Fig. 4). Glyceraldehyde-3-phosphate dehydrogenase was used as an internal loading control, as represented by the 420 bp band. Reverse transcription PCR verified the absence of expression of EGFRvIII at the mRNA level (Fig. 4). A total of 21 samples were analyzed by real-time PCR and reverse transcription PCR. Ten of these samples were negative for EGFRvIII expression and 11 were positive for EGFRvIII as determined by tissue microarray staining, indicating that the observed lack of EGFRvIII expression was not due to the a priori selection of samples that were negative for EGFRvIII expression by immunohistochemistry. Western blot analysis of normal urothelium, 24 patient tumors, and 4 cell lines using EGFR antibodies targeting the cytoplasmic domain revealed no EGFRvIII protein expression. Of these patients' samples, five specimens were EGFRvIII-negative and three were EGFRvIII-positive based on immunohistochemistry. Representative samples are shown in Fig. 5A
. Protein expression of EGFRvIII was seen only in the positive control, U87MG-EGFRvIII cell line (Fig. 5A and B). Taken together, our results show that the expression of the truncated variant of EGFR in urothelial carcinoma is a rare event.
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Interestingly enough, four of the cell lines tested were found to be sensitive to gefitinib in vitro, whereas the other seven cell lines were resistant to the cytotoxic effects of the drug (34). Resistance to gefitinib was associated with uncoupling between the EGFR and mitogen-activated protein kinase, and could be predicted by analyzing the activation of Gsk-3ß and cyclin D1. In certain cell lines, this uncoupling was due to a dependence on platelet-derived growth factor-ß signaling for cell cycle progression (34). Our data indicate that the response to gefitinib by bladder cancer cells cannot be attributed solely to the status of EGFR mutations within the kinase domain and that other variables and cell properties must affect the response to gefitinib.
Indeed, recent work, based on the retrospective analysis of a group of patients with NSCLC treated with gefitinib, suggests that increased EGFR gene copy number and EGFR protein overexpression may be associated with improved survival (42). In bladder cancer, it was shown that EGFR is overexpressed in 31% to 48% of tumors, mainly in association with muscle invasion (43–45). Thus, it will be important to correlate EGFR gene amplification and protein expression with the response to gefitinib or other tyrosine kinase inhibitor drugs targeting EGFR in the ongoing clinical trials in patients with bladder cancer (46).
In this study, we were also unable to detect measurable levels of EGFRvIII expression in any tumor cell line or patient specimens. Although our initial immunostaining for EGFRvIII in urothelial carcinoma tissue microarrays detected EGFRvIII expression in 50% of the samples, subsequent analysis by real-time PCR, reverse transcription PCR, and Western blot failed to detect its expression in the tumor samples. These conflicting results between immunohistochemistry staining and reverse transcription PCR and real-time PCR for the expression of EGFRvIII were also documented for breast cancer (41). We now stress the point of validating any data that emerge from using antibody and tissue microarrays to measure the expression of EGFRvIII, which may produce false-positive results.
Although we could not detect EGFRvIII mRNA or protein expression in any of the investigated bladder cancer cell lines, or in the large number of tumor specimens, in contrast, many of these samples do express wild-type EGFR at various levels. Given that there was expression of wild-type EGFR, that it has several potential ligands, and that it can be activated in several ways, it is doubtful that EGFRvIII is a major drive in the EGFR pathway in bladder cancer. Moreover, our study shows that EGFRvIII does not play a role in bladder cancer and therefore does not represent a viable target for bladder cancer therapy.
These results have clinical implications in the sense that when considering treatment with tyrosine kinase inhibitors for bladder cancer, there is no need to consider EGFRvIII-preferential inhibitors such as AG1478 (47). However, the challenge of predicting which patients will respond to gefitinib and other tyrosine kinase inhibitors remains to be elucidated.
|
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 2/20/06; revised 5/ 3/06; accepted 5/15/06.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
transforming growth factor receptor binding and transmembrane signaling. Proc Natl Acad Sci U S A 1994;91:10217–22.This article has been cited by other articles:
|
|
 |
|
  P. C. Black, G. A. Brown, T. Inamoto, M. Shrader, A. Arora, A. O. Siefker-Radtke, L. Adam, D. Theodorescu, X. Wu, M. F. Munsell, et al. Sensitivity to Epidermal Growth Factor Receptor Inhibitor Requires E-Cadherin Expression in Urothelial Carcinoma Cells Clin. Cancer Res., March 1, 2008; 14(5): 1478 - 1486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Franovic, L. Gunaratnam, K. Smith, I. Robert, D. Patten, and S. Lee Translational up-regulation of the EGFR by tumor hypoxia provides a nonmutational explanation for its overexpression in human cancer PNAS, August 7, 2007; 104(32): 13092 - 13097. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Shrader, M. S. Pino, L. Lashinger, M. Bar-Eli, L. Adam, C. P.N. Dinney, and D. J. McConkey Gefitinib Reverses TRAIL Resistance in Human Bladder Cancer Cell Lines via Inhibition of AKT-Mediated X-Linked Inhibitor of Apoptosis Protein Expression Cancer Res., February 15, 2007; 67(4): 1430 - 1435. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Shrader, M. S. Pino, G. Brown, P. Black, L. Adam, M. Bar-Eli, C. P.N. Dinney, and D. J. McConkey Molecular correlates of gefitinib responsiveness in human bladder cancer cells Mol. Cancer Ther., January 1, 2007; 6(1): 277 - 285. [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 |
