Expression of Activated Epidermal Growth Factor Receptors, Ras-Guanosine Triphosphate, and Mitogen-activated Protein Kinase in Human Glioblastoma Multiforme Specimens

Expression of Activated Epidermal Growth Factor Receptors, Ras-Guanosine Triphosphate, and Mitogen-activated Protein Kinase in Human Glioblastoma Multiforme Specimens
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  1 of 16 Copyright © by the Congress of Neurological Surgeons Volume 45(6), December 1999, p 1442 Expression of Activated Epidermal Growth Factor Receptors, Ras-Guanosine Triphosphate, and Mitogen-activated Protein Kinase in Human Glioblastoma Multiforme Specimens [Clinicopathological Studies]Feldkamp, Matthias M. M.D.; Lala, Prateek B.Sc.; Lau, Nelson B.Sc.; Roncari, Luba M.Sc.; Guha, Abhijit M.D. Samuel Lunenfeld Research Institute (MMF, PL, NL, LR, AG), Mount Sinai Hospital; Division of Neurosurgery (MMF, AG), Toronto Western Hospital, University Health Network, and University of Toronto; and Department of Surgical Oncology (AG), Ontario Cancer Institute/Princess Margaret Hospital, Toronto, Ontario, CanadaReceived, October 20, 1998.Accepted, August 2, 1999. Abstract OBJECTIVE: Amplification of the epidermal growth factor receptor (EGFR) is a common event in the molecularpathogenesis of high-grade astrocytic tumors, occurring in 50% of glioblastoma multiforme (GBM) cases. A subset of GBMs also express a constitutively phosphorylated truncated receptor (EGFRvIII). Expression of transfected EGFRvIII in cells has been reported to activate the Ras-mitogen-activated protein kinase pathway andto provide a growth advantage. Novel therapeutic agents targeting signal transduction pathways are entering early clinical trials; determination of which GBMs express EGFRvIII might help identify patients who might benefitfrom these biological agents. METHODS : A cohort of 15 flash-frozen surgical specimens (12 GBMs, 2 gliosarcomas, and 1 adult low-grade glioma) were evaluated for EGFR and EGFRvIII expression and for EGFR activation status using immunohistochemical (IHC) analysis, Western blotting, and reverse transcription-polymerase chain reaction assays. Levels of activated Ras-guanosine triphosphate were measured using a nonradioactive luciferase-based technique. Mitogen-activated protein kinase activation was determined using a myelin basic protein assay. IHC analysis was performed on paraffin-embedded, formalin-fixed, pathological specimens. Normal control samples included white matter specimens distal to tumors (n = 5), a sample obtained during a lobectomy for treatment of epilepsy (n = 1), and cultured fetal human astrocytes (n = 1). RESULTS : We demonstrated higher levels of activated Ras and mitogen-activated protein kinase in GBM specimens, compared with normal brain tissue or the low-grade glioma. There was a very good correlation between results obtained using specialized molecular techniques and those obtained using routine IHC techniques. Screening for EGFRvIII expression may be of prognostic importance, because patients with EGFRvIII-positive tumors exhibited shorter life expectancies (mean survival time for patients with EGFRvIII-positive tumors, 4.5 ± 0.6 mo; mean survival time for patients with EGFRvIII-negative tumors, 11.2 ±0.9 mo).CONCLUSION: We demonstrated that routine IHC techniques using commercially available antibodies are capable of identifying which GBM specimens express EGFRvIII and whether the EGFRs are activated. Such a molecular classification of GBMs might allow us to determine which patients might benefit from biologically targeted therapies. In addition, characterization of specimens with respect to their EGFRvIII status seems to be of prognostic value.Astrocytomas are the most common primary malignant neoplasms of the central nervous system, with prognoses being inversely related to their histological grades ( 18 ). The transformation of a low-grade astrocytoma(LGA) (World Health Organization Grade II) to a glioblastoma multiforme (GBM) (World Health Organization Grade IV) is associated with the stepwise accumulation of genetic mutations ( 22 ). Amplification of the epidermal growth factor receptor ( EGFR ) gene on chromosome 7p has been identified as one of the final events in this progression, occurring in 50% of GBMs ( 3, 20 ). Amplification of the EGFR  gene is associated with overexpression of the EGFR ( 4 ), a M  r  170,000, 1186-amino acid, surface-localized, receptor protein tyrosine kinase. Activation of the receptor by its ligands (epidermal growth factor, transforming growth factor-[alpha], vaccinia virus growth factor, and amphiregulin) through phosphorylation results in the activation of downstream intracellular signaling pathways, including the Ras-mitogen-activated protein kinase (MAPK) signaling pathway ( 23 ). We previously demonstrated that the expression of wild-type EGFRs, together with the expression of platelet-derived growth  2 of 16 factor receptors, in these tumors results in the activation of Ras and the Ras-MAPK signaling pathway in astrocytoma cells ( 11 ).The process of EGFR  gene amplification is associated with intragenic deletions and rearrangements, frequently resulting in large deletions of exons 2 to 7 of the EGFR  gene ( 7, 48 ). Subsequent coercive splicing results in an identical truncated messenger ribonucleic acid (mRNA) transcript, lacking the 801 bases encoding amino acids 6 to 273 of the extracellular domain of the receptor ( 50 ). This coercive splicing results in the creation of a novel glycine splice site in the translated protein, whereas the carboxyl portion of the protein undergoes in-frame translation. The resulting M  r  140,000 truncated protein (EGFRvIII, also known as p140 EGFR ,[DELTA]EGFR, or EGFRde2–7) is constitutively activated ( 7, 31 ) and confers an in vivo growth advantage to cells implanted into nude mice ( 31 ). Approximately 17 to 23% of GBMs express EGFRvIII because of EGFR  gene amplification ( 56 ). An additional 40% of GBMs (and some ovarian cancers, non-small cell lung cancers, and breastcancers, as well as precancerous Barrett’s epithelium) do not exhibit EGFR  amplification but do express EGFRvIII (as detected by Western blotting, immunohistochemical [IHC] analysis, or reverse transcription [RT]-polymerase chain reaction [PCR]) ( 9, 26, 39, 55 ). Therefore, it has been reported that as many as 62% of all GBMs express EGFRvIII at the protein level ( 55 ). The constitutive phosphorylation of EGFRvIII results in enhanced activation of various signaling pathways, including the Ras-MAPK pathway ( 25, 38 ).In the present study, we examined a cohort of 15 adult glial tumor specimens, including 12 GBMs, 2 gliosarcomas, and 1 LGA. Various specimens of non-neoplastic white matter were used as control samples. We performed extensive molecular investigations on this small cohort of tumors, including Western blot analysis, RT-PCR, and IHC analysis to assess EGFR and EGFRvIII expression and the presence of activated forms of the EGFR. We also measured the levels of activated Ras-guanosine triphosphate (GTP) and the kinase activity of MAPK in these tumors. In this article, we demonstrate a good correlation between the results of molecular techniques (Western blotting and RT-PCR) and IHC analyses, as well as a relationship between Ras activation andMAPK activity. These results extend our previous observations that suggested a role for Ras activation in GBM pathogenesis and demonstrate the feasibility of using readily available IHC techniques for the rapid molecular characterization of GBM specimens. Finally, we demonstrate for the first time that EGFRvIII expression has potential clinical significance, because Kaplan-Meier plots indicate substantial reductions in mean survival times for patients whose tumors express EGFRvIII. MATERIALS AND METHODS Tumor specimens All tumor specimens were obtained from the University of Toronto Nervous System Tumor Bank at The Toronto Hospital. All tumor samples were obtained from patients undergoing craniotomies for tumor resection, and samples were placed in cryovials and immediately flash-frozen in liquid nitrogen in the operating room. A random cohort of 15 specimens was selected and included 12 GBM specimens, 2 gliosarcomas, and 1 LGA. Patient characteristics are presented in Table 1 . Up to seven specimens of non-neoplastic white matter were used as control samples. Five of these were lobectomy specimens collected far from a tumor, one was obtained from a patient undergoing a lobectomy for treatment of intractable seizures, and one was a culture of normal fetal human astrocytes (Clonetics, San Diego, CA). All normal control specimens underwent analysis to confirm the absence of neoplastic cells within the specimen, with frozen sections from both ends of each specimen being independently analyzed using hematoxylin/eosin staining. The characteristics of these normal control specimens are presented in Table 2 .  3 of 16 Table 1. Patient Characteristics and Pathological Classification of Tumors aa  Pathological diagnoses were obtained from surgical pathological records and were confirmed by a review of hematoxylin/eosin-stained sections. Survival times were calculated as the times from the initial onset of symptoms to death; patients still alive at the time of this study are indicated as censored observations. R, right; L, left; WHO, World Health Organization.Table 2. Characteristics of Normal Specimens aa  Ras-guanosine triphosphate (GTP) levels could be measured for seven specimens; mitogen-activated protein kinase (MAPK) levels were measured for three of these. H&E, hematoxylin and eosin. Western blot analysis Tumor specimens were lysed in ice-cold phosphorylation lysis buffer (50 mmol/L 4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid, pH 7.0, 150 mmol/L NaCl, 10% glycerol, 1% Triton X-100, 1.5 mmol/L MgCl 2 , 1 mmol/L ethylene glycol bis[[beta]-aminoethyl ether]- N  , N  , N  ', N  '-tetra-acetic acid, 100 mmol/L NaF, 10 mmol/L sodium pyrophosphate, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 0.1 mmol/Lphenylmethylsulfonyl fluoride, 0.1 mmol/L sodium orthovanadate). The protein concentration of each lysate was determined using the bicinchoninic acid method (Pierce Chemical Co., Rockford, IL). Aliquots of 100 µg of protein(for all Western blots except those using the activation-specific antibody) or 300 µg (for blots using theactivation-specific antibody) were combined with 0.5 volume of 3× sample buffer (187.5 mmol/L Tris·HCl, pH 6.8,6%, w/v, sodium dodecyl sulfate, 30% glycerol, 150 mmol/L dithiothreitol, 0.1%, w/v, bromphenol blue), and proteins were separated on 10% sodium dodecyl sulfate-polyacrylamide gels. Proteins were transferred to PolyScreen polyvinylidene difluoride membranes (NEN Research Products/Du Pont, Boston, MA), using a semi-drytransfer apparatus (Bio-Rad, Hercules, CA). Membranes were probed for the expression and activation status of the wild-type EGFR and EGFRvIII using the following three antibodies: 1) a polyclonal sheep anti-EGFR antibody that recognizes the intracellular domains of both the wild-type EGFR and EGFRvIII (anti-EGFR IC ; Upstate Biotechnology, Lake Placid, NY), 2) a rabbit polyclonal antibody that specifically recognizes EGFRvIII (a gift from Albert Wong, Kimmel Cancer Center, Philadelphia, PA), and 3) a mouse monoclonal antibody that recognizes activated/phosphorylated EGFR (Transduction Laboratories, Lexington, KY). Proteins were detected using alkalinephosphatase-conjugated secondary antibodies and the CDP-Star chemiluminescence system (NEN Research Products, Boston, MA), and membranes were exposed to Kodak X-OMAT film (Kodak, Rochester, NY). RT-PCR assay for EGFRvIII expression RT-PCR was performed on each tumor sample in three separate experiments. Ribonucleic acid (RNA) was extracted using RNeasy RNA extraction columns (Qiagen, Valencia, CA). Two micrograms of RNA were subjected  4 of 16 to first-strand synthesis using random polyhexamer primers and SuperScript II reverse transcriptase (Life Technologies, Inc., Gaithersburg, MD). Two microliters of first-strand cDNA were subjected to PCR using Qiagen Taq  deoxyribonucleic acid polymerase Master Mix buffer, containing 1.5 mmol/L MgCl 2 . The primers used were 5'-GGGGAATTCGCGATGCGACCCTCCGGG-3' (sense) and 5'-GGGAAGCTTTCCGTTACACACTTTGCG-3' (antisense), amplifying the wild-type EGFR as a 1037-base pair fragment and EGFRvIII as a 236-base pair fragment ( 55 ). PCRconditions were as follows: initial denaturation for 4 minutes at 94°C, 40 cycles of PCR (94°C for 80 s, 56°C for60 s, and 72°C for 135 s), and final extension at 72°C for 10 minutes. PCR products were resolved on vertical 8%polyacrylamide gels, and deoxyribonucleic acid fragments were identified using ethidium bromide staining.Tumors were characterized as either expressing or not expressing EGFRvIII mRNA. Ras·GTP assay We previously developed, in collaboration with Gerry Boss (University of California, San Diego), a luciferase-based enzymatic technique for measuring activated Ras·GTP levels in flash-frozen specimens ( 11, 13, 44 ). This assay was modified to be entirely nonradioactive. Briefly, approximately one-half of a cryovial of tissue was homogenized at 4°C, in 2.5 ml of lysis buffer (50 mmol/L 4-[2-hydroxyethyl]-1-piperazineethanesulfonicacid, pH 7.4, 10 mmol/L MgCl 2 , 150 mmol/L NaCl, 1% Nonidet P-40, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 0.5mmol/L phenylmethylsulfonyl fluoride), and centrifuged. The supernatant was split into two equal aliquots and immunoprecipitated overnight with the neutralizing rat anti-Ras monoclonal antibody Y13-259 (which binds and immunoprecipitates Ras-bound guanosine diphosphate [GDP] or GTP; Oncogene Science) or rat immunoglobulin G (negative control; Jackson Immunoresearch Laboratories, West Grove, PA), linked to protein G-agarose with a secondary goat anti-rat Fc immunoglobulin G antibody. After multiple washings, the GTP and GDP nucleotides were eluted from the immunoprecipitated Ras-bound complex by heating. GTP was used to phosphorylate adenosine diphosphate, in the presence of nucleotide diphosphate kinase. The generated adenosine triphosphate (ATP) activated luciferin in the presence of luciferase; a luminometer was used for detection. GDP levels were assayed by converting GDP to GTP using pyruvate kinase, in the presence of phosphoenolpyruvate; the GTP was subsequently converted to ATP as in the GTP assay described above. The amounts of GTP and GDP in the samplescould be quantified, because known amounts of GTP and GDP standards were also assayed. Background levels measured for the rat immunoglobulin G precipitates were subtracted to yield the absolute amounts of GTP and GDP in the tumor specimens. The absolute amount of activated Ras·GTP extracted from each specimen wasexpressed per milligram of protein lysate; reported values are means of two or three experiments performed for each sample. Myelin basic protein (MBP) kinase assay MAPK activity was quantified using a MBP kinase assay. Tumor specimens were lysed in ice-cold kinase lysis buffer (25 mmol/L 4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid, pH 7.5, 0.3 mol/L NaCl, 1.5 mmol/L MgCl 2 , 0.2 mmol/L ethylenediamine tetra-acetic acid, 20 mmol/L [beta]-glycerophosphate, 1% Triton X-100, 0.5 mmol/L dithiothreitol, 20 µg/ml aprotinin, 20 µg/ml leupeptin, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/Lsodium orthovanadate). Three hundred micrograms of protein were immunoprecipitated overnight with protein A-Sepharose beads and 35 µl of anti-extracellular signal-regulated kinase 2 (ERK2) antibody (Santa CruzBiotechnology, Santa Cruz, CA), which preferentially recognizes the ERK2 isoform of MAPK but also immunoprecipitates the ERK1 MAPK isoform. Immunoprecipitates were washed once with kinase lysis buffer, three times with 1% Nonidet P-40 in phosphate-buffered saline with 2 mmol/L Na 3 VO 4 , and once with 100mmol/L Tris·HCl, pH 7.5, and 0.5 mol/L LiCl in water. Lysates were finally washed once with kinase reactionbuffer (12.5 mmol/L 3-[ N  -morpholino]-propanesulfonic acid, pH 7.5, 12.5 mmol/L [beta]-glycerophosphate, 7.5 mmol/L MgCl 2 , 0.5 mmol/L ethylene glycol bis[[beta]-aminoethyl ether]- N  , N  , N  ', N  '-tetra-acetic acid, 0.5 mmol/L NaF, 0.5 mol/L Na 3 VO 4 ). After this final wash, immunoprecipitates were incubated for 1 hour at 30°C with 50 µl of kinase reaction buffer supplemented with 1.5 mg/ml bovine MBP (Sigma Chemical Co., St. Louis, MO), 2 mmol/L unlabeled ATP, and 1 µCi of [[gamma]- 32 P]ATP (specific activity, 3000 Ci/mmol; NEN Research Products). The kinase reaction was terminated by addition of 25 µl of 3× sample buffer; samples were boiled for 5 minutes at100°C and loaded onto 15% sodium dodecyl sulfate-polyacrylamide gels. MBP was identified as a M  r  17,000 protein by Coomassie blue staining, and kinase activity was quantified using a STORM 860 PhosphorImager (Molecular Dynamics, Sunnyvale, CA). Kinase activity was normalized to MBP kinase activity in quiescent, serum-starved, HER14 cells (NIH 3T3 fibroblasts transfected to express the wild-type human EGFR). HER14 cells stimulated with epidermal growth factor for 5 minutes provided a positive control sample for each experiment. IHC analysis  5 of 16 Formalin-fixed, paraffin-embedded, tissue sections were probed with the following four antibodies: 1) a mouse monoclonal antibody that recognizes both the wild-type EGFR and EGFRvIII (EGFR IC , used at 1:100; Transduction Laboratories), 2) a mouse monoclonal antibody that recognizes only the extracellular domain of the wild-type EGFR (EGFR EC , used at 1:80; Novocastra, Newcastle, UK), 3) a mouse monoclonal antibody that is directed against the novel glycine splice site and that recognizes only EGFRvIII (EGFRvIII clone DH8.3, used at 1:20; Novocastra), and 4) a mouse monoclonal antibody that specifically recognizes activated/phosphorylated EGFR (EGFR Act , used at 1:50; Transduction Laboratories). All slides were subjected to antigen retrieval by microwave boiling in citrate buffer. The secondary antibody was a goat anti-mouse antibody (used at 1:200; Zymed Laboratories, South San Francisco, CA); antigens were detected using the avidin-biotin complex method (Vector Laboratories, Burlingame, CA), with diaminobenzidine as the substrate. Slides were counterstained with hematoxylin. All sections were independently evaluated by two of the authors (MMF and AG), and staining was classified as follows: absent, 0; trace, 1; positive, 2 to 4, based on the number of stained cells and the intensity of staining. These independent observations, although subjective, exhibited substantial interobserver correlation, and all nonidentical scores were resolved by consensus. Survival analysis Clinical data were obtained from patient charts. Survival times were calculated as the time from the onset of symptoms to the time of death. Kaplan-Meier curves were constructed using JMP version 3.2 software (SAS Institute, Cary, NC). Group differences were analyzed using the log-rank test. RESULTS Expression of the wild-type EGFR Expression of the wild-type EGFR was assessed by Western blotting ( M  r  170,000 band, assayed using anti-EGFR IC ; Upstate Biotechnology) ( Fig. 1 ), RT-PCR ( Fig. 2 ), and IHC analysis (anti-EGFR EC ; Novocastra) ( Fig. 3 A ). Because of the competitive nature of PCR, our PCR method preferentially amplifies the shorter EGFRvIII product, compared with the longer wild-type EGFR, when both mRNA species are present (positive control) ( Fig. 2 ). Therefore, although tumors that expressed EGFRvIII invariably expressed the wild-type EGFR as well (according to Western blotting and IHC analysis), PCR analysis revealed only the EGFRvIII band in these cases. The anti-EGFR IC  antibody that was used to detect both wild-type and truncated EGFRs frequently revealed multiple bands even when only the wild-type receptor was expressed (e.g., low-grade tumors); these bands represent partially glycosylated forms of the receptor ( M  r  150,000–160,000). Therefore, to differentiate partiallyglycosylated wild-type EGFR from EGFRvIII, it was necessary to use a specific antibody against EGFRvIII (seebelow).FIGURE 1. Western blots of 15 tumor samples and control specimens. Samples 1 to 14 represent World Health Organization Grade IV tumor samples; Cases 2 and 13 contain sarcomatous elements. LG , Grade II LGA. Lysates from normal brain ( NB ) tissue and cultured normal human astrocytes ( NHA ) (Clonetics) were used as negative control samples. U118 cell lines expressing only the wild-type EGFR (negative control) or both the wild-type EGFRand EGFRvIII (positive control) were also probed. Upper  , 300 µg of lysate, separated by sodium dodecylsulfate-polyacrylamide gel electrophoresis and probed with an activation-specific mouse monoclonal antibody (Transduction Laboratories). Middle  and lower  , 100 µg of lysate, separated by sodium dodecylsulfate-polyacrylamide gel electrophoresis and probed with a polyclonal rabbit antibody specific for EGFRvIII ( middle ) or with a polyclonal pan-EGFR antibody that recognizes both the wild-type and truncated receptors ( lower  ). 170  and 140 , positions of the wild-type and truncated receptors, respectively.  Asterisks , locations of an aberrant EGFR of approximately M  r  190,000, which seems to be constitutively activated in Case 6 ( upper  ).
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks