Malignant gliomas can be induced in mice through the combined expression of activated forms of both KRas and Akt in glial progenitor cells. To determine the reliance of these tumors on continued KRas signaling in vivo, we generated a viral vector that allows the expression of KRas to be controlled post-delivery. Tumor-free survival rates were compared between those animals with continued KRas expression and animals in which KRas expression was suppressed. KRas signaling was found to be required for the maintenance of these tumors in vivo; inhibition of KRas expression resulted in apoptotic tumor regression and increased survival. Subsequent reexpression of KRas reinitiated tumor growth, indicating that a percentage of the progenitor cells survived and retained tumorigenic properties.
Glioblastoma multiforme (GBM) is the most common and aggressive primary brain tumor. It is also the most fatal; mean survival is <1 year from the time of diagnosis, with <10% survival after 2 years. Despite major improvements in imaging, radiation, and surgery, the prognosis for patients with this disease has not changed in the last 20 years. Treatment options for patients with GBM include surgery, radiation, and chemotherapy (1). However, surgery and radiation are extremely difficult due to the location and invasive nature of the tumor, and chemotherapy remains controversial, as many studies have failed to show prolonged median survival in treated patients (2). In the past several years, genes that are altered in tumor tissue relative to normal brain tissue have been found (3). However, the effects of these genetic alterations have not been fully characterized in vivo; those that can be productively targeted for therapeutic intervention in patients remain to be identified.
Glioblastomas are classified as either primary or secondary GBM (4). Primary GBM develops de novo, whereas secondary GBM progresses from a low-grade glioma to a high-grade glioma through the acquisition of additional genetic alterations (5). Primary glioblastomas comprise 80% of all GBM (6). Although these two types cannot be distinguished pathologically, they seem to have distinct genetic alterations (reviewed in ref. 6). Epidermal growth factor receptor amplification and overexpression, PTEN loss, and INK4a deletion are hallmarks of primary GBM. In contrast, platelet-derived growth factor receptor overexpression, p53 mutation, cyclin-dependent kinase 4 overexpression, or RB loss, and PTEN loss characterize secondary GBM (6). In both cases, activated receptor tyrosine kinases (i.e., epidermal growth factor receptor and platelet-derived growth factor receptor) activate common downstream signaling pathways including the Ras pathway and the phosphatidylinositol 3-kinase (PI3K)/Akt pathway (1, 7). Loss of PTEN expression is also common in both primary and secondary GBM (6). PTEN is a phosphatase that functions to inhibit the PI3K/Akt pathway (7). In the absence of PTEN, Akt activity is elevated leading to increased proliferation and inhibition of apoptosis (8). Akt activation has also been documented in GBM as a result of increased PI3K activity due to mutation within the regulatory subunit of PI3K (9).
Recently, a mouse model of human GBM based on the avian RCAS/TVA system was developed (10). In this model, the retroviral receptor, TVA, is expressed under the control of the nestin promoter, which is active in neural and glial progenitors. This promoter was chosen because GBM is thought to arise from the supporting glial cells of the brain (11). Replicating mammalian cells that express TVA are susceptible to infection by the subgroup A avian leukosis virus–derived RCAS vectors (Fig. 1A and B; reviewed in ref. 10). Although replication-competent in avian cells, these viruses are replication-defective in mammalian cells; therefore, the viral vectors cannot spread in the target animals. In addition, because little viral envelope protein is produced, there is no interference to superinfection. Thus, there is theoretically no limit to the number of experimental genes that can be introduced into a TVA-expressing mammalian cell. The ability of these cells to be infected by multiple viruses allows efficient modeling of GBM, because multiple oncogenic alterations can be easily introduced into the same cell or animal.
In this study, we examined the role of Ras signaling in glioblastoma maintenance in the Kras- and Akt-induced glioblastomas in the RCAS/TVA model. Inhibition of KRas expression resulted in apoptotic tumor regression and increased survival of tumor-bearing mice, strongly suggesting that KRas signaling is required for tumor maintenance in vivo. Subsequent reexpression of KRas reinitiated tumor growth, indicating that some progenitor cells survived and retained tumorigenic properties.
Materials and Methods
Transgenic mice. Nestin-TVA mice have been previously described (12). The mice were maintained on standard food or doxycycline-containing food pellets (Harlan-Teklad, Madison, WI). All experiments were done in compliance with the guiding principles of the “Care and Use of Animals” (available at http://www.nap.edu/books/0309053773/html/) and were approved by the Van Andel Research Institute, Institutional Animal Care and Use Committee prior to experimentation.
Genotype analysis. DNA was prepared from tail biopsies using an AutoGenprep 960 automated DNA isolation system. PCR of the TVA allele was carried out in 25 μL total volume containing 2.5 μL 10× PCR buffer (Invitrogen, Carlsbad, CA), 1.0 μL DMSO, 1.25 μL 50 mmol/L MgCl2, 0.625 mmol/L of each nucleotide, 1 unit of Taq polymerase (Invitrogen), 6.25 μg/mL of primers TVA-386 sense and TVA-786 antisense. The following primer sequences were used: TVA-386 sense 5′-AGCTGGTGAGATGGGACTGAAC-3′; TVA-786 antisense 5′-CGAACATTCAAAGCCTCCAG-3′ (to detect a 400 bp fragment). Samples were amplified for 30 cycles (94°C for 30 seconds, 56°C for 30 seconds, and 72°C for 45 seconds).
Isolation of primary astrocytes. Primary astrocytes were isolated from the brains of Nestin-TVA-positive newborn mice. Single cells were obtained by mincing the tissue followed by digestion in 0.05% trypsin for 20 minutes at 37°C. The cells were resuspended in DMEM with 10% fetal bovine serum (FBS) and 1× penicillin/streptomycin. The cells were split 1:3 when confluent.
Vector constructs. The retroviral vectors used in this study were either replication-competent avian leukosis virus long terminal repeats, splice acceptor, and Bryan polymerase-containing vectors of envelope subgroup A (designated RCASBP(A); Fig. 1A and B), or no splice acceptor vectors (designated RCANBP(A); Fig. 1C and D). RCASBP(A)Akt has been previously described (8). The RCASBP(A) destination vector has also been described (13). To generate RCASBP(A)Tet-off (Fig. 1B), the Tet-off element was PCR-amplified from genomic DNA isolated from the human breast adenocarcinoma MCF7 Tet-off stable cell line (Clontech, Palo Alto, CA). The forward primer (5′-GGATCCGCCACCATGTCTAGATTAGATAAAAG-3′) introduced a BamHI restriction site and the reverse primer (5′-CTCGAGCTACCCACCGTACTCGTCAATTC-3′) introduced an XhoI restriction site and stop codon. The amplified product was cloned into TOPO Blunt (Invitrogen) to generate TOPO-Tet-off and was sequence-verified. TOPO-Tet-off was then digested with BamHI and XhoI and the Tet-off fragment was subcloned into pENTR3C (Invitrogen) to generate pENTR3C-Tet-off. RCASBP(A)Tet-off was generated by mixing 300 ng of pENTR3C-Tet-off with 300 ng of the RCASBP(A) destination vector in the presence of the LR Clonase Enzyme Mix as described (14). LR reactions were done for 2 hours at room temperature and 1 μL was transformed into 50 μL TOP10-One Shot Competent Cells (Invitrogen). Transformations were plated on agar plates containing 100 μg/mL ampicillin.
The RCANBP(A) destination vector has been previously described (14). To generate pENTR3C-TRE, pTRE (Clontech) was digested with XhoI and EcoRI. The TRE fragment was cloned into pENTR3C digested with SalI and EcoRI. To generate pENTR3C-TRE-KRas, RCASBP(A)FLAG-KRas (a gift from Eric Holland, Departments of Surgery (Neurosurgery), Neurology, and Cell Biology, Memorial Sloan Kettering Cancer Center, New York, NY) was digested with XhoI and XbaI and the FLAG-KRas fragment was cloned into pENTR3C-TRE also digested with XhoI and XbaI. RCANBP(A)TRE-KRas (Fig. 1D) was generated by mixing 300 ng of pENTR3C-TRE-FLAG-KRas with 300 ng of the RCANBP(A) destination vector in the presence of the LR Clonase Enzyme Mix as described above.
Cell culture. DF-1 cells were grown in DMEM-high glucose supplemented with 10% FBS (Invitrogen), 1× penicillin/streptomycin, and maintained at 39°C (15, 16). The DF-1 cultures were passaged 1:3 when confluent.
Virus propagation. Virus infection was initiated by calcium phosphate transfection of plasmid DNA that contained the retroviral vector in proviral form (17). In standard transfections, DF-1 cells were plated at 30% confluency, allowed to attach (2-3 hours), and 5 μg of purified plasmid DNA was introduced by the calcium phosphate precipitation method previously described (18), followed by a 5-minute glycerol shock at 39°C (15% glycerol in the medium). Viral spread was monitored by assaying culture supernatants for avian leukosis virus capsid protein by ELISA as previously described (19). Virus stocks were generated from the cell supernatants. The supernatants were cleared of cellular debris by centrifugation at 2,000 × g for 10 minutes at 4°C, filtered through a 0.45-μm filter, and stored in aliquots at −80°C. Virus was determined to be replication-competent by a reading of 0.200 or greater on ELISA for the viral capsid protein (19).
Viral infections in vitro. Astrocytes were seeded in six-well plates at a density of 5 × 104 cells/well and were maintained in DMEM with 10% FBS, 1× penicillin/streptomycin at 37°C. After the cells attached, the medium was removed and replaced with 1 mL of filtered virus-containing medium in the presence of 8 μg/mL polybrene (Sigma, St. Louis, MO) for 2 hours at 37°C. The virus was removed and replaced with fresh medium, and the cells were incubated at 37°C for 1 hour. Cells were infected again for another 2 hours, after which the virus-containing medium was replaced with fresh medium and the cells were incubated overnight at 37°C.
Western blotting. Infected astrocytes were washed with PBS and 150 μL SDS-lysis buffer was added to each well of a six-well dish. The cell lysates were boiled for 10 minutes and passed through a 26-gauge needle five times. Following centrifugation, the proteins were separated on a 14% Tris-glycine polyacrylamide gel, transferred to nitrocellulose, and incubated for 1 hour at room temperature in blocking solution (0.05% Tween 20 in TBS with 5% nonfat dry milk). Blots were immunostained for FLAG-KRas using an anti-FLAG monoclonal antibody (Sigma) at a 1:500 dilution and anti-α-tubulin 1:1,000 (Oncogene, San Diego, CA). All antibodies were diluted in blocking solution. Western blots were incubated in the primary antibody for 1 hour at room temperature with constant shaking and then washed thrice in TBS-T wash buffer (0.05% Tween 20 in TBS). The blots were then incubated with an anti-mouse IgG-horseradish peroxidase secondary antibody diluted 1:3,000 (Sigma) for 1 hour at room temperature. The blots were washed thrice in TBS-T wash buffer, incubated with enhanced chemiluminescence solutions according to the manufacturer's specifications (Amersham, Piscataway, NJ), and exposed to film.
Viral infections in vivo. Infected DF-1 cells from a confluent culture in a 10 cm dish were trypsinized, pelleted, resuspended in 100 μL medium, and placed on ice. Mice were injected intracranially 4 mm ventral from the bregma (intersection of the coronal and sagittal sutures) with 5 μL of cells using a gastight Hamilton syringe. Mice were monitored daily for tumor development. Censored survival data was analyzed using a log-rank test of the Kaplan-Meier estimate of survival.
Histology and histochemical staining. Mice were euthanized when they displayed obvious signs of distress or as indicated. Brain tissues were fixed in formalin overnight, cut into three sections, and then dehydrated through a graded alcohol series in a Ventana Renaissance processor (Ventana Medical Systems, Tucson, AZ). Tissues were paraffin-embedded and 5 μm sections were adhered to glass slides. Sections were stained with H&E or left unstained for immunohistochemistry.
Immunohistochemistry. The sections were deparaffinized and treated with 3% hydrogen peroxide for 30 minutes to quench endogenous peroxidase activity. Sections were blocked for 30 minutes at room temperature with 1% bovine serum albumin and 5% normal goat serum in PBS. KRas expression was detected using a monoclonal antibody to the FLAG epitope (Sigma) at a 1:250 dilution, and Akt expression was detected using a monoclonal antibody to the HA epitope (Covance, Berkeley, CA) at a 1:1,000 dilution. The sections were incubated with the primary antibodies for 1 hour at room temperature and were then washed in PBS-T. The signal was detected with the Vectastain ABC kit (Vector Labs, Burlingame, CA) and visualized with 3,3′-diaminobenzidine tetrahydrochloride (Vector Labs). Sections were counterstained with hematoxylin. Apoptosis was detected by terminal nucleotidyl transferase–mediated nick end labeling (TUNEL) staining using the in situ cell death detection kit (Roche, Indianapolis, IN) according to the manufacturer's specifications.
Results and Discussion
Intracranial infection of Nestin-TVA mice with either RCASBP(A)Akt or RCASBP(A)KRas alone is insufficient to form tumors, but the combination of activated forms of both Akt and KRas induces glioblastomas that are histologically similar to human GBM (8). To determine the reliance of these tumors on continued KRas signaling, we generated a viral vector in which expression of the inserted gene could be regulated post-delivery using the tetracycline (tet)-inducible system. We used the RCANBP(A) virus (14), which lacks the splice acceptor at the 3′ end of the envelope gene (Fig. 1C), so that sequences inserted into this region are transcribed from an internal promoter and not the viral long terminal repeat. A tet-responsive element (TRE) was inserted upstream of the KRas gene to generate RCANBP(A)TRE-KRas (Fig. 1D). The virus was propagated in DF-1 cells, an immortalized chicken embryo fibroblast cell line (15, 16). No KRas expression was observed in these cells, because expression from the TRE requires the presence of a tetracycline transcriptional activator such as Tet-off or a reverse tetracycline transcriptional activator such as Tet-on. In the Tet-on system, the Tet-responsive gene is only expressed in the presence of doxycycline; in the Tet-off system, the Tet-responsive gene is repressed in the presence of doxycycline (20).
Tet-regulated expression of KRas in primary astrocytes in vitro. To test the ability of KRas expression to be induced from the TRE-KRas virus, Nestin-TVA-positive astrocytes were infected in culture with the TRE-KRas virus alone or with both the TRE-KRas and Tet-off viruses. The infected cells were cultured in the presence or absence of doxycycline and KRas expression was visualized by Western blot (Fig. 1E). KRas expression was observed only in those cells infected with both the TRE-KRas and Tet-off viruses. This expression is tightly regulated, because no KRas expression was detected in cells infected with the TRE-KRas virus alone or in cells infected with both the TRE-KRas and Tet-off viruses in the presence of doxycycline (Fig. 1E).
Tumor formation in Nestin-TVA mice infected with Akt, Tet-off, and TRE-KRas viruses. Intracranial infection of Nestin-TVA mice with RCASBP(A)Akt and RCASBP(A)KRas induces GBM that are histologically similar to human GBM (8). To test the ability of the TRE-KRas virus to induce tumors in vivo, newborn Nestin-TVA mice were injected intracranially with Akt, Tet-off, and TRE-KRas viruses. As early as 3 weeks of age, several mice began showing signs of tumor formation (e.g., macrocephaly, lethargy, or cachexia). Histologic examination revealed tumors that were similar to those generated by injection of both RCASBP(A)Akt and RCASBP(A)KRas (8), and closely resembled human GBM (Fig. 2A). TUNEL staining revealed few apoptotic cells (Fig. 2B). Immunohistochemical staining of the tumors with antibodies to the FLAG epitope tag on KRas (Fig. 2C) and the HA epitope tag on Akt (Fig. 2D) showed that both genes were expressed in the tumors, but not in the normal surrounding tissue. Whereas 46% of the Nestin-TVA-positive animals developed tumors, no tumors were detected in any of the TVA-negative animals injected with the same viruses (P < 0.00004; Fig. 3A). This shows the specificity of the infection to only those cells expressing TVA.
KRas expression is required for the maintenance of glioblastoma in vivo. To determine the reliance of these tumors on continued KRas signaling, tumors were induced by injection of Nestin-TVA mice with Akt, Tet-off, and TRE-KRas viruses at birth. Because ∼50% of the injected Nestin-TVA-positive mice develop tumors (Fig. 3A), only tumor-bearing mice were selected for doxycycline treatment. Doxycycline was given continuously through the food to 14 tumor-bearing mice from the time of weaning (∼21 days of age) until the experimental end point (4 months of age) to suppress KRas expression. Four of the treated mice expired shortly after the start of treatment as a result of tumor progression. The remaining 10 mice became asymptomatic. Tumor recurrence was not observed in any of the animals maintained on doxycycline. A separate cohort of 14 tumor-bearing mice was given standard feed (without doxycycline), and all eventually died of GBM. Tumor-free survival rates were compared between untreated mice and mice given doxycycline to determine whether the administration of doxycycline increased survival. The resulting Kaplan-Meier curve shows a significant increase in tumor-free survival for the mice treated with doxycycline (P < 0.0006; Fig. 3B).
Inhibition of KRas expression results in apoptotic tumor regression. To determine if the increased survival in the doxycycline-treated group was a result of reduced KRas expression, brain tissue from mice treated with doxycycline for 3 or 14 days was analyzed for the level of expression of both KRas and Akt by immunohistochemistry for the FLAG and HA epitopes on KRas and Akt, respectively. After 3 days of treatment, the level of KRas expression was significantly decreased relative to that in untreated mice, whereas the level of Akt expression remained unchanged (Fig. 4). After 14 days of doxycycline treatment, no tumor was visible and no KRas or Akt expression was detected (Fig. 5). A significant number of vacuolated cells were observed in sections from treated mice relative to untreated mice. Because the tumors from doxycycline-treated mice showed signs of regression, TUNEL staining was done to determine if apoptotic cells could be detected within the tumor. After 3 days of doxycycline treatment, a significant amount of apoptosis was detected within the tumor (Fig. 4B, compared with untreated mice Fig. 2B). After 14 days of doxycycline treatment, no tumor cells or apoptotic cells were visible (Fig. 5).
Reexpression of KRas following doxycycline treatment reinitiates tumor growth. To determine if reexpression of KRas could reinitiate tumor growth, doxycycline was withdrawn from animals treated for either 25 or 45 days. The mice were observed for signs of tumor growth, and tumor-free survival rates were compared between the two groups (Fig. 6). Interestingly, all seven mice in each group had recurrent disease. Upon reactivation of KRas, tumorigenesis quickly resumed in the mice that were treated for only 25 days. The majority of those animals (57%) displayed recurrent disease within 3 weeks of doxycycline withdrawal. In contrast, in mice that had been treated for 45 days, tumor formation was not observed until 3 months after doxycycline withdrawal. The tumors that formed after doxycycline withdrawal reexpressed Kras, as detected by immunohistochemistry for the FLAG epitope (data not shown). These data suggest that a percentage of the progenitor cells capable of reestablishing the tumor survived apoptotic cell death following the loss of KRas expression upon doxycycline treatment, and that this percentage decreases the longer KRas expression is suppressed. Future work will focus on determining the length of time required to eliminate all progenitor cells capable of reestablishing disease.
This model shows the critical importance of the Ras pathway in glioblastoma maintenance and indicates that continuous suppression of Ras signaling is necessary and sufficient to suppress the tumorigenic potential of the glial progenitor cells. In addition, this regulated expression system will allow the evaluation of the role of other genes and pathways in this context. This has important clinical implications for pharmacologic agents targeting these pathways in GBM patients.
Grant support: Michigan Economic Development Corporation, and the Michigan Technology Tri-Corridor (Michigan Animal Models Consortium grant 085P1000815 and Michigan Center for Biological Information grant 085P1000819).
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.
We thank Eric Holland for the Nestin-TVA mice, the RCASBP(A)Akt and FLAG-KRas viral vectors, and helpful advice; Brendan Looyenga, Scott Robertson, Marleah Russo, Arianne Folkema, Cassie Zylstra, and Robert Sigler for technical assistance and the vivarium staff for excellent animal husbandry; Bree Berghuis and James Resau of the Analytical Cellular and Molecular Microscopy Laboratory; Kyle Furge of the Bioinformatic Special Program for technical assistance. The authors gratefully acknowledge the Van Andel Research Institute, the Michigan Economic Development Corporation, and the Michigan Technology Tri-Corridor for support (Michigan Animal Models Consortium grant 085P1000815 and Michigan Center for Biological Information grant 085P1000819).