We have established a line of transgenic rats expressing v-erbB, the viral form of epidermal growth factor receptor (EGFR), under transcriptional regulation of the S100β promoter. Reverse transcriptase-polymerase chain reaction revealed highest transgene expression in the cerebellum followed by the cerebrum, ovary, and testis. Other organs, including the lung, heart, salivary gland, colon, liver, kidney, and spleen, did not show detectable transgene expression. Of 23 homozygous rats that died or were killed because they became moribund between 25 and 91 weeks of age, 15 (65%) showed the presence of brain tumors (mean age, 59 weeks). Of the 10 heterozygous rats killed between 61 and 91 weeks of age, 4 (40%) showed the presence of brain tumors (mean, 77 weeks). With 3 exceptions, all tumors were located within or near the cerebellum (83%). There were 2 major histologic types; one type displayed a solid growth pattern with predominantly perivascular infiltration of adjacent central nervous system tissue and the meninges. Tumors showed histologic features of malignancy with occasional lung metastases. There was a consistent, strong immunoreactivity for S100 protein but no significant expression of glial, neuronal, or meningothelial markers. These tumors were classified as malignant gliomas. A second tumor type was less invasive and characterized by isomorphic cells with round to ovoid nuclei and clear perinuclear halos expressing S100 but no neuronal or glial marker proteins. They were diagnosed as oligodendrogliomas. This is the first transgenic rat model that spontaneously develops brain tumors. Because v-erbB is structurally and functionally similar to the truncated form of EGFR amplified and overexpressed in human glioblastomas, S100β-v-erbB transgenic rats may serve as a useful animal model for the identification of EGFR-related molecular targets and as a tool for the assessment of novel therapeutic approaches.
There are several transgenic and knockout models of gliomas in mice with abnormal expression of genes that are involved in the pathogenesis of human brain neoplasms (1). Genetically modified genes include H-ras (2, 3), EGFR (3, 4), Nf1 (5), and p53 (5). Ding et al reported that 95% of transgenic mice expressing v(12)Ha-ras under the transcriptional regulation of the glial fibrillary acidic protein (GFAP) promoter developed low- and high-grade astrocytomas within 2 to 6 months (2). GFAP-v(12)Ha-ras/GFAP-EGFRvIII double transgenic mice had decreased survival with 50% of mice dead at 2 to 4 weeks from oligodendrogliomas and oligoastrocytomas (3). Reilly et al established a mouse model involving Nf1 and p53 that shows development of low-grade astrocytoma and glioblastoma (5). Weiss et al reported that transgenic mice expressing v-erbB under the regulation of the S100β promoter developed low-grade oligodendrogliomas (50% penetrance by 6 months of age), whereas transgenic mice heterozygous for ink4a/arf or p53 developed anaplastic oligodendrogliomas (4).
These transgenic and knockout mice have contributed significantly to the understanding of the development of gliomas and the cooperation of transformation-associated genes. However, as a result of their small size, mice have limited suitability for experimental therapeutic studies. Available rat models are typically based on intracranial injection of established brain tumor cell lines and are thus highly artificial. To obtain a long-needed animal model suitable for use in the testing of novel approaches to targeted brain tumor therapy, we have established a line of transgenic rats carrying an S100β-v-erbB transgene (4). v-erbB is the viral homolog of EGFR (6, 7), which is structurally and functionally similar to the truncated form of EGFR, frequently observed in human glioblastomas (8, 9).
Materials and Methods
Generation and Screening of Transgenic Rats
The transgene S100β-v-erbB, which has been previously used to establish transgenic mice that develop oligodendrogliomas (4), was microinjected into pronuclei of Wistar Hannover rat one-cell embryos by YS New Technology Institute (Tochigi, Japan); newborn rats were first maintained at CLEA Japan, Inc. (Shizuoka, Japan) and then transferred to the International Agency for Research on Cancer (IARC) animal house in Lyon, France. Of 78 potential transgenic rats screened by polymerase chain reaction (PCR), 8 carried the S100β-v-erbB transgene. One founder rat (no. 59) died 5 days after birth. Another founder rat (no. 45) died of a brain tumor in the cerebellum, histologically resembling those of a low-grade human oligodendroglioma (World Health Organization grade II) at 8 weeks of age. Six remaining founder rats (nos. 13, 23, 32, 52, 53, and 55) were first mated with nontransgenic Wistar Hannover rats to increase the number of heterozygous rats and then between heterozygous rats to establish a homozygous line.
DNA was extracted from rat tails at 1 to 1.5 weeks of age. Tails were incubated in 500 μL tail buffer (50 mM Tris-HCl pH 8.0; 100 mM NaCl; 1% SDS) and 25 μL proteinase K solution (15.1 mg/mL) overnight at 55°C. DNA was precipitated with saturated NaCl (6 M) and isopropanol (500 μL). Precipitated DNA was washed with 500 μL 70% ethanol and then dissolved in 100 μL TE buffer.
PCR was carried out in a total volume of 10 μL consisting of 20 mM Tris-HCl, pH 8.4, 50 mM KCl, and 2 mM MgCl2, 0.2 mM of each dNTP, 0.2 μM of sense and antisense primers, 0.5 U Platinum Taq DNA polymerase (Invitrogen, Paisley, U.K.), and 1 μL DNA solution. After heating at 95°C for 5 minutes, 36 cycles of denaturation at 95°C for 30 seconds, annealing at 62°C for 30 seconds, and extension at 72°C for 30 seconds were performed followed by a final extension at 72°C for 5 minutes in gradient Thermocycler (Biometra, Gottingen, Germany). Primers used to detect the S100β-v-erbB fragment (210 bp) were 5′-CTC ACA GCA ATC TCA AAG CTC CCC-3′ (sense) and 5′-AGC CTC CAA AGT CAG GTT GAT GAG C-3′ (antisense).
Southern blot was carried out to distinguish between homozygous and heterozygous transgenic rats. Genomic DNA (3.5 μg) extracted from tail tissues was digested with 14 units of Bts1 at 37°C for 4 hours and then resolved in a 1% agarose gel in 1× TAE for 2.5 to 3 hours. The gel was treated with 0.25 N HCl for 20 minutes and 0.5 M NaOH/1.5 M NaCl for 30 minutes, respectively. DNA was then transferred onto a nylon membrane overnight in 10× SSC. After transfer, the membrane was washed in 50 mM NaPi for 10 minutes on the desk and 10 minutes with shaking, baked at 80°C for 1 hour, and then crosslinked with UV. Prehybridization was performed in 5 mL of Church's buffer for 2 hours. A 527-bp segment of v-erbB was labeled with α-32P dCTP (ICN 3000 Ci/mmol) using a Prime-It RMT Random Primer Labeling Kit (Stratagene, Amsterdam, The Netherlands) and was used as a probe for hybridization. Hybridization was performed in 7 mL Church's buffer containing the previous 32P-labeled probe at 60°C overnight. After washing, the membrane was exposed to a phosphorimaging screen for 3 days. Quantification of bands was carried out using the Imagequant software. The membrane was tripped and rehybridyzed with a probe synthesized from a 750-bp DNA segment (Xba/EcoR) of PTC. Data obtained were used for calibration of DNA input.
Reverse Transcriptase-Polymerase Chain Reaction
Reverse transcriptase-polymerase chain reaction (RT-PCR) was carried out to assess tissue-specific expression of the transgene. First-strand cDNA was synthesized as follows: 2 μg of DNase-treated total RNA and 0.5 μg of oligo dT12-18 (Pharmacia, Uppsala, Sweden) in a total volume of 11 μL were heated to 70°C for 10 minutes and then chilled on ice. A mix consisting of 4 μL of 5× first-strand cDNA buffer (Gibco BRL, Gaithersburg, MD), 2 μL of 100 mM DTT, 1 μL of 10 mM dNTPs, and 1 μL of Rnase block (40 units/μL; Stratagene, La Jolla, CA) was added into the tube and allowed to heat at 42°C for 2 minutes. Then 200 units of SuperScript II Rnase H-reverse transcriptase (Gibco-BRL, Gaithersburg, MD) were added, and the reaction was continued at 42°C for 50 minutes. After inactivation at 70°C for 15 minutes, the cDNA was stored at −20°C until use. A reaction without RT was performed in parallel to control genomic DNA contamination.
PCR was performed in a total volume of 10 μL containing 1 μL cDNA template and v-erbB primers (the same as those used for genotyping), 0.5 U Taq DNA polymerase (Sigma, St. Louis, MO), 2 mM MgCl2, 2 mM of each dNTP, 2 μM of sense and antisense primers, 10 mM Tris-HCl (pH 8.3), and 50 mM KCl in a Robot Thermal Cycler (Stratagene) as follows: initial denaturation for 5 minutes at 95°C, 36 cycles with denaturation at 95°C for 30 seconds, annealing at 62°C for 30 seconds, and extension at 72°C for 30 seconds followed by a final extension for 5 minutes at 72°C. After PCR, 7 μL of products were run on an agarose gel and stained with ethidium bromide. A PCR using β-actin primers was used in parallel to control RNA/cDNA quantity.
Total RNA (30 μg in 6.5 μL) was denatured in 22.1 μL of denaturation buffer (consisting of 1× MOPS, 2.5% formaldehyde, and 65% formamide) at 65°C for 5 minutes and then chilled on ice for 5 minutes. Four microliters of loading buffer (consisting of 50% glycerol, 1 mM EDTA pH 8.0, 0.25% bromphenol bleu, 0.25% xylene cyanol, and 0.5 mg/mL ethidium bromide) was then added. The mix was loaded onto a 1% agarose gel containing 2.28% formaldehyde and run in 1× MOPS buffer at 70 V for 30 minutes and then 100 V for 2 hours. After being washed with 10× SSC for 30 minutes, RNA was transferred onto a nylon membrane overnight in 10× SSC. After the overnight transfer, membrane was washed in 50 mM NaPi for 10 minutes on the desk and 10 minutes with shaking, baked at 80°C for 1 hour, and then was crosslinked with UV. Prehybridization was performed in 5 mL of Church's buffer for 2 hours. DNA fragment of v-erbB (527 bp) was labeled with α-32P dCTP (ICN 3000 Ci/mmol) using a Prime-It RMT Random Primer Labelling Kit (Stratagene, Amsterdam, The Netherlands) and used as a probe for hybridization. Hybridization was performed in 7 mL Church's buffer containing the previous 32P-labeled probe at 60°C overnight. After washing, the membrane was exposed to a phosphorimaging screen for 3 to 14 days.
Differential Polymerase Chain Reaction to Detect Amplification of S100β-v-erbB Transgene
To assess amplification of the S100β-v-erbB gene in brain tumors compared with normal brain tissues in transgenic rats, we carried out differential PCR for S100β-v-erbB gene using rat β-actin gene as a reference. The primer sequences were as follows: 5′-CTC ACA GCA ATC TCA AAG CTC CCC-3′ (sense) and 5′-AGC CTC CAA AGT CAG GTT GAT GAG C-3′ (antisense) for S100β-v-erbB (product 210 bp), and 5′- CTG TGG CAT CCA TGA AAC TA -3′ (sense) and 5′- GCT AAG GCT GCT TAC CTT GA -3′ (antisense) for rat β-actin gene (PCR product 187 bp). DNA was extracted from brain tumors and adjacent normal brain tissues from the same rats as previously described (10). PCR was carried out in a total volume of 10 μL containing 100 ng DNA, 1 μL of 10× PCR buffer without MgCl2 (Invitrogen, Carlsbad, CA), 2 mM of MgCl2, 0.2 mM of each dNTP, 0.1 μL of Platinum Taq DNA Polymerase (Invitrogen), 0.2 M of each β-actin primer, and 0.1 M of each S100β-v-erbB primer. PCR amplification was performed in a T3 Thermocycler (Biometra) as follows: initial denaturation for 5 minutes at 95°C, 26 to 32 cycles with denaturation at 95°C for 30 seconds, annealing at 62°C for 30 seconds, and extension at 72°C for 30 seconds. The PCR product was analyzed on a 2% agarose gel containing ethidium bromide. Densitometry of the PCR fragments was measured using a DC120 Zooming Digital Camera and Kodak Digital Science ID Image analysis Software (Kodak, Rochester, NY).
Histology and Immunohistochemistry
Autopsy was carried out on all rats presenting with clinical signs of disease. Brains and other macroscopically abnormal organs were fixed in buffered formalin and stained with hematoxylin & eosin. To classify brain tumors, immunohistochemical stainings were carried out on an automated Nexus staining apparatus (Ventana Medical Systems, Strasbourg, France) following the manufacturer's guidelines. Antibodies used were GFAP (Dako, Baar, Switzerland; dilution 1:300), synaptophysin (Zymed, San Francisco, CA; dilution 1:50), microtubule-associated protein-2 (MAP-2C 11; 1:1,000), neurofilament protein (NF, 200 kD subunits; Sigma; dilution 1:200), S100 (Dako; 1:2,000), neuronal nuclei (NeuN; Chemicon International, Temecula, CA; 1:4,000), and epithelial membrane antigen (EMA; Dako; 1:20).
Characteristics of the antisera to v-erbB have been previously reported (7, 12). For immunohistochemistry of v-erbB, endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide in methanol for 30 minutes at room temperature. With no unmasking treatment, incubation with 5% skimmed milk for 1 hour was performed. Sections were incubated overnight at 4°C with the primary monoclonal antibody to v-erbB (dilution 1:50). Negative controls were served as the brain tumor sections without primary antibody incubation.
Tissue-Specific Transgene Expression in S100β-v-erbB Transgenic Rats
RT-PCR revealed that the transgene was expressed in the brain of offspring from all 6 founder rats (numbers 13, 23, 32, 52, 53, and 55) with the highest expression levels in offspring from the founder rat 55 (data not shown). Offspring from founder rat 55 were therefore chosen to establish a homozygous line and carry out further characterization. RT-PCR revealed a highest expression of the transgene in the cerebellum with somewhat lower expression in the cerebrum, testis, and ovary, and no detectable expression in other tissues, including the lung, kidney, colon, liver, heart, spleen, and salivary gland (Fig. 1). Transgene expression was not detectable by Northern blot in any of the tissues analyzed in homozygous rats of line 55, including the cerebellum and cerebrum (data not shown).
Development of Brain Tumors in S100β-v-erbB Transgenic Rats
Lethal brain tumors developed in 15 homozygous rats (7 males and 8 females) at 25 to 91 weeks of age (mean, 59 weeks) and in 4 male heterozygous rats at 61 to 91 weeks (mean, 77 weeks; Fig. 2). All but 3 tumors were located within or near the cerebellum (Fig. 3A). Of 23 homozygous rats that died or were killed because they became moribund between 25 and 91 weeks of age, 15 (65%) showed the presence of brain tumors. Of the 10 heterozygous rats killed between 61 and 91 weeks of age, 4 (40%) showed the presence of brain tumors.
Histologic Features of Brain Tumors
There were histologically 2 types of brain tumors. One type displayed a solid growth pattern composed of fibrillary tumor cells growing in a fascicular pattern and of epithelioid tumor cells with eosinophilic cytoplasm (Fig. 3B-G). Most of these tumors, particularly in the cerebellum, infiltrated through the Virchow-Robin-spaces resulting in perivascular tumor clusters (Fig. 3B) that often extended into the adjacent central nervous system tissue and meninges and were usually large, whereas in 4 tumors, diffuse single-cell invasion of surrounding brain tissues was noted. Tumors showed signs of malignancy such as brisk mitotic activity (Fig. 3C), geographic tumor necroses (Fig. 3D), invasion into the skull bones (2 tumors), and lung metastases (3 tumors; Fig. 3E). There was a consistent and diffuse immunoreactivity for S100 protein (Fig. 3F) without significant immunoreactivity for GFAP, synaptophysin, NeuN, microtubule-associated protein-2 (MAP2C), neurofilament (NF), and epithelial membrane antigen (EMA). The diagnoses of glioblastoma and malignant meningioma were considered, but the lack of meningeal markers and lack of immunoreactivity for GFAP as well as for MAP2C prompted us to preliminary classify these tumors as malignant gliomas.
The second type of brain tumor in S100β-v-erbB transgenic rats was typically located within or near the cerebellar granular cell layer (Fig. 3H-L). It was observed at 33 to 47 weeks of age (mean, 38 weeks) in homozygous rats. Small focal neoplastic proliferations were observed within or near the granular layer of the cerebellum (Fig. 3H). These tumors were histologically composed of isomorphic cells with round to ovoid nuclei and clear perinuclear halos (Fig. 3J). They displayed some mitotic activity (Fig. 3J) and, at least focally, prominent endothelial proliferation (Fig. 3I). The diagnosis of neurocytoma was considered. However, because these tumors lacked immunoreactivity for neuronal markers but were S100-positive in at least some fraction of neoplastic cells, they were classified as oligodendrogliomas.
Transgene in Brain Tumors in Transgenic Rats
Differential PCR revealed that S100β-v-erbB/β-actin ratios were similar between pairs of brain tumors and normal brain tissues from 13 transgenic (10 homozygous and 3 heterozygous) rats analyzed, suggesting the absence of amplification of S100β-v-erbB transgene in brain tumors developed in transgenic rats (data not shown).
Immunohistochemistry revealed that higher v-erbB expression was observed focally or diffusely in brain tumors compared with surrounding normal brain tissues in 11 of 15 (73%) of brain tumors analyzed (Fig. 3G, L).
One homozygous rat developed a malignant neuroectodermal tumor in the subcutaneous soft tissue near the left eye at 70 weeks of age. One heterozygous rat developed a pituitary tumor (86 weeks), and another heterozygous rat developed invasive ductal carcinoma of the breast (91 weeks); both of these were considered to be sporadic, because these are common tumors in aged Wistar rats (13). No abnormalities or neoplastic or preneoplastic lesions were observed in any other organs of transgenic rats.
In contrast to transgenic mice, only a few lines of transgenic rats have been established as models of human neoplasms. Transgenic rats overexpressing SV40 T antigen under control of the phosphoenolpyruvate carboxykinase promoter developed pancreatic islet cell carcinomas with regional lymph node metastases (14). Male MMTV-neu transgenic rats developed mammary cancer in an androgen-dependent fashion (15). Development of mammary fibroadenomas and papillary adenomas was reported in rats overexpressing c-erbB-2 under the control of MMTV long terminal repeat promoter (16), whereas rats carrying the HTLV-pX gene under control of the mouse H-2Kd promoter developed mammary carcinomas (17). Transgenic rats carrying the HTLV-pX gene under the control of a rat lymphocyte-specific protein tyrosine kinase proximal promoter developed thymic tumors (18). Male transgenic rats expressing SV40 T antigen under probasin promoter developed prostate carcinomas at 100% incidence before reaching 15 weeks of age (19). Transgenic rats expressing SV40 large T antigen under control of albumin promoter exhibited a 100% incidence of liver tumors by 24 to 36 weeks of age (20). Transgenic rats carrying the human c-Ha-ras were highly susceptible to mammary carcinogenesis induced by N-methyl-N-nitrosourea (MNU) (21) and 7,12-dimethylabenz[a]anthracene (DMBA) (21) and esophageal carcinogenesis induced by N-nitrosomethylbenzylamine (NMBA) (22).
Reasons for availability of only a few transgenic rats may be the result of the difficulty of microinjection in rat embryos, cost of production and maintenance, and limited opportunity to cross with other transgenic rats. However, since techniques for generation of transgenic rats by the advent of lentiviral vectors have been established (23, 24), more genetically modified rat models are expected to be established in the future. The larger size of rats is clearly advantageous, providing better access for microsurgery and tissue sampling, better imaging, and easy manual handling for the purpose of therapy (25). Furthermore, mice and rats are significantly different with respect to genes involved in immunity, metabolic detoxification, reproduction, chemosensation, and conservation of genes involved in human diseases (25).
The work presented here represents the first transgenic rat model that develops brain tumors. Sporadic brain tumors are generally very rare in rats, including the Wistar strain (13, 26, 27). In a combined analysis of 11 studies (total 60,000 rats), only 0.3% of rats developed brain tumors (gliomas, meningiomas, and granular cell tumors) (26); in another study, 11 of 891 Wistar rats (1%) that were observed up to 120 weeks developed brain tumors spontaneously (13). However, in contrast to mice, rats are known to be highly susceptible to the development of brain tumors induced by chemical carcinogens such as N-methyl-N-nitrosourea (MNU) and N-ethyl-N-nitrosourea (ENU) (28). Transplacental exposure to ENU causes a high incidence of brain tumors in the offspring, particularly oligodendrogliomas (29-31). The genes involved in neurocarcinogenesis by MNU, ENU, and related alkylating agents have not yet been identified with the exception of neu mutations in ENU-induced malignant schwannomas (32).
The aim of the present study was to establish a line of transgenic rats that develop malignant gliomas. We chose the viral form of EGFR, which is structurally and functionally similar to the truncated form of EGFR (8) that is frequently amplified and overexpressed in human glioblastomas (33, 34). In mice, expression of v-erbB causes the development of oligodendrogliomas (4). In agreement with our observation of higher expression of transgene in the cerebellum than in the cerebrum, most tumors were located in the cerebellum in S100β-v-erbB transgenic rats. This is consistent with the findings that in S100β-v-erbB transgenic mice, v-erbB expression was observed in subcortical white matter and in the granular and Purkinje layers of the cerebellum (4), and that oligodendrogliomas frequently developed in the cerebellum (K. Aldape, personal communication, 2006).
The relatively long latency period of brain tumor development in the line 55 of S100β-v-erbB transgenic rats (mean 59 weeks in homozygous rats) may be the result of low expression of the transgene, which was detectable only with RT-PCR, but not in Northern blots. However, one founder rat died of oligodendroglioma at only 8 weeks of age (see "Materials and Methods"), possibly as a result of higher expression of v-erbB. Line 55 of S100β-v-erbB transgenic rats may be useful for the elucidation of genetic alterations required for the development of brain tumors in addition to v-erbB expression.
Two distinct histologic types of brain tumors developed in S100β-v-erbB transgenic rats. One histologic type was diagnosed as oligodendroglioma. Early focal lesions of these tumors were observed within the granular layer of the cerebellum, also suggesting the possibility of a neuronal origin. However, immunohistochemistry showed no evidence of immunoreactivity to neuronal markers such as synaptophysin, NeuN, MAP2C, or NF. This corroborates the observation that transgenic mice carrying the identical transgene typically develop oligodendrogliomas (4). Furthermore, GFAP-EGFRvIII and RAS double transgenic mice have been reported to develop oligodendrogliomas and mixed oligoastrocytomas (3). This suggests that the alteration of the EGFR signaling pathway resulting from v-erbB or EGFRvIII expression is associated with the development of oligodendrogliomas. In human oligodendrogliomas, EGFR overexpression in the absence of gene amplification is frequently observed (35) in the absence of EGFRvIII expression (36).
Another histologic type was diagnosed as malignant glioma, which displayed clear evidence of malignant behavior, that is, the presence of geographic necrosis, brisk mitotic activities, and invading capacity such as spreading to the surface of the cerebellum, cerebrum, skull bone, and metastasizing to the lung. Diagnosis of "glioblastoma" was considered, but because of the lack of convincing GFAP and MAP2C immunoreactivity, and the presence of consistent and diffuse S100 immunoreactivity, we were rather cautious in using the term "glioblastoma." In humans, complete lack of GFAP immunoreactivity is a rare event, only occurring in rather advanced, dedifferentiated glioblastomas.
Histologic features of these tumors, their preferential cerebellar localization, and the presence of lung metastases suggest that these brain tumors are unique and considerably different from human gliomas. This may be a potential disadvantage when these tumors are used as a model for human gliomas. However, development of 2 different histologic tumor types in S100β-v-erbB transgenic rats may be potentially interesting, because these tumors may develop from common stem cells (i.e. neuroectodermal origin) and therefore present with either oligodendroglial or glioblastoma-like phenotypes. This hypothesis is not unrealistic, because v-erbB is expressed under transcriptional regulation of the S100β promoter in this transgenic model. S100β is a small acidic Ca2+-binding protein that is expressed at high levels in the brain (37) and plays important roles in mediating glial-neuronal interactions and in modulating the maturation of oligodendroglia from oligodendrocyte progenitor cells (38). It is expressed particularly in astrocytes, but also oligodendroglial cells, Schwann cells, and ependymal cells (37, 39, 40). The S100β promoter element functions as a promoter with a magnitude of activity similar to SV40 enhancer (37). Alternatively, malignant gliomas in S100β-v-erbB transgenic rats may be a more malignant form of oligodendrogliomas, which develop through progression from the less invasive oligodendrogliomas.
EGFR amplification is a frequent genetic alteration in human primary (de novo) glioblastomas (41, 42) and is often associated with structural alterations (8). The most common variant, present in 20% to 50% of glioblastomas with EGFR amplification, is variant III (EGFRvIII), which results from a nonrandom 801-bp in-frame deletion of exons 2-7 of the EGFR gene (9). This results in loss of the extracellular portion (domains I and II) of EGFR, which therefore becomes unable to bind to its ligands such as EGF and TGFα (43, 44). However, it activates constitutively the tyrosine kinase domain in a ligand-independent manner, leading to cell proliferation through the RAS and mitogen-activated protein kinase signaling pathways (45). EGFRvIII is thus structurally and functionally similar to the product of the retroviral oncogene v-erbB (7, 8). EGFRvIII has been considered a potential target for a selective, tumor-specific therapy of glioblastomas (45-48), because EGFRvIII expression is restricted to neoplastic cells (49, 50). A recent clinical study suggests that EGFR kinase inhibitors are significantly effective in glioblastomas coexpressing EGFRvIII and intact PTEN (51). These S100β-v-erbB transgenic rats may serve as models to study therapies targeting EGFRvIII and related downstream molecules.
The authors thank Mrs. Anne-Marie Camus, Mrs. Marie Pierre Cros, Mrs. Dominique Galendo, and Mrs. Nicole Lyandrat for technical assistance. The authors also thank YS New Technology Institute and CLEA Japan, Inc. for microinjection and maintenance of transgenic rats.