Resistance to Tyrosine Kinase Inhibition by Mutant Epidermal Growth Factor Receptor Variant III Contributes to the Neoplastic Phenotype of Glioblastoma Multiforme

Cell Lines, Tissue Culture, and Inhibitor Studies.

All of the tissue culture was performed in antibiotic-free Zinc Option medium (Life Technologies, Inc., Grand Island, NY) containing 10% heat-inactivated FCS, and all of the cell lines were shown to be free of Mycoplasma contamination as described previously (3) . For the purposes of these investigations, the following cell lines were used as described previously, A431, human epidermoid carcinoma cells which express high levels of wild-type EGFR (26) , and NR6, Swiss 3T3 murine fibroblasts mock-transfected (NR6 Null, shown previously to be devoid of either wild-type EGFR or mutant EGFRvIII expression), or transfected with either wild-type EGFR (NR6W) or mutant EGFRvIII (NR6M) cDNA (12) . NR6M is fully tumorigenic in vivo and is dependent on mutant EGFRvIII expression for its neoplastic phenotype (12) . The U87MG.wtEGFR and U87MG.ΔEGFR malignant glioma cell lines, derived from transfection of the parental U87 malignant glioma (U87MG) cell line stably transfected with and expressing human wild-type EGFR or mutant EGFRvIII (19) , respectively, were provided by Dr. Webster Cavenee (Ludwig Institute, University of California San Diego, San Diego, CA). Stock solutions of gefitinib (courtesy of Dr. Alan Wakeling, AstraZeneca, Cheshire, United Kingdom) were prepared in DMSO. Cell lines were cultured in humidified incubators at 37°C/5%CO₂ and treated with 0.1, 1, or 10 μm gefitinib unless otherwise indicated, for the reported times. Control samples were mock-treated with an equal amount of DMSO. Growth inhibition and viability were assessed by trypan blue dye exclusion using light microscopy and a standard hemocytometer for cell counting.

Determination of Receptor Density by Quantitative Fluorescence Activated Cell Sorting Analysis.

As described previously by Wikstrand et al. (4) , we have standardized the Quantitative Fluorescence Activated Cell Sorting methodology as marketed by Banks Laboratory (Fisher, IN) for the determination of wild-type EGFR and mutant EGFRvIII receptor density by comparing to standard Scatchard analysis with iodinated epidermal growth factor and iodinated monoclonal antibodies. Results presented in Table 1⇓ for A431, NR6 Null, NR6W, and NR6M cells were provided from that analysis.

Western Blot Analysis.

All cell lines were mock-treated or treated with gefitinib and were cultured for up to 7 days, as indicated in the figure legends. Briefly, at the time of harvest, cells were washed once in 1× Dulbecco’s PBS, centrifuged, lysed, equivalent amounts of protein were separated by SDS-PAGE (4–20% acrylamide), and were transferred to polyvinylidene difluoride (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom) for enhanced chemiluminescence detection, as described previously (26) . The polyvinylidene difluoride membranes were blocked for 1 h with 5% nonfat milk in 1× Tris-buffered saline/0.1% Tween 20 and were then probed with rabbit polyclonal IgG phospho-Tyr-1173 EGFR (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit polyclonal IgG pan EGFR (Cell Signaling Technology, Beverly, MA), rabbit polyclonal IgG phospho-Ser-473 Akt (Cell Signaling Technology); rabbit polyclonal IgG pan Akt (Cell Signaling Technology); and monoclonal IgG1 anti-β-actin clone AC-15 (Sigma-Aldrich, Inc., St. Louis, MO). The membranes were washed three times in 1× Tris-buffered saline/0.1% Tween 20 and incubated for 2 h with antirabbit IgG (Cell Signaling) or antimouse IgG (Cell Signaling) conjugated to horseradish peroxidase. The membranes were washed six times in 1× Tris-buffered saline/0.1% Tween 20, and immunoblotted proteins were visualized using the Phototope-horseradish peroxidase Western Blot Detection System (Cell Signaling), according to the manufacturer’s instructions. For Western analysis, a mean constitutive activity or fold induction was determined for each experiment, with expression levels normalized to β-actin for all of the blots. To compare results between lanes, densitometry data were normalized to the 45.5 kDa β-actin band, which was constitutively expressed to similar levels in all of the lanes. Values were calculated using scanning densitometry software (Kodak Digital Science 1D Image Analytical Software, Rochester, NY).

Flow Cytometric Cell Cycle Analysis.

In vitro mock-treated and gefitinib-treated cells were harvested with 0.125% trypsin, washed twice in 1× Dulbecco’s PBS with 0.1% glucose, counted, and fixed overnight with 70% (vol/vol) ethanol. Cells were then centrifuged at 1333 × g for 10 min, resuspended at a concentration of 1 × 10⁶ cells/ml in 1× Dulbecco’s PBS with 0.1% glucose, 50 μg/ml propidium iodide (Sigma), and 100 units/ml of RNase A (Sigma), and analyzed by flow cytometry.

In Vitro Invasion Assay.

The BD Biocoat Matrigel Invasion Chamber assay was performed according to manufac-turer’s instructions (Becton Dickinson Labware, Bedford, MA). Briefly, 1 × 10⁶ cells were mock-treated or treated with gefitinib (1 or 10 μm) in equivalent amounts of DMSO for 24 h, at which time 5 × 10⁴ cells were added to control or invasion-chamber wells for an additional 24 h with fresh gefitinib. Twenty-four h later, the cells on the upper side of the chamber were scraped, and the ones on the lower side of the chamber were stained with Diff Quik (Dade Behring, Newark, DE) and counted using light microscopy with a standardized grid. All of the experimental treatment and control samples were performed in triplicate.

Statistical Analysis.

The Student’s t test was used to determine significant changes in activities and assess statistical significance for cell cycle analysis. Data are presented as the mean ± SE, and were analyzed using Microsoft Excel XP software (Microsoft, Seattle, WA).

Inhibition of Mutant EGFRvIII Receptor Phosphorylation Requires Higher Concentration and Longer Exposure to Gefitinib Than Wild-Type EGFR.

To determine the effectiveness of inhibition of phosphorylation by gefitinib, we tested multiple cell lines expressing no or low levels of EGFR and high levels of wild-type EGFR or mutant EGFRvIII (Table 1)⇓ . For no or low-level EGFR expression, we assayed NR6 Swiss murine fibroblasts expressing neither EGFR nor EGFRvIII (NR6 Null), as well as the human glioblastoma multiforme cell line U87 (U87MG), endogenously expressing low amounts of wild-type EGFR. A431 human epidermoid carcinoma cells, NR6 cells expressing wild-type EGFR (NR6W), and U87 cells expressing wild-type EGFR (U87MG.wtEGFR) were tested because of their high level of stably expressed wild-type EGFR. Similarly, NR6 cells expressing mutant EGFRvIII (NR6M), as well as U87 cells (U87MG.ΔEGFR), which express the mutant receptor, we assayed because of their stable expression of EGFRvIII. All of the cells were mock-treated (DMSO only) or treated daily for 7 days with fresh medium and gefitinib at a final concentration of 0.1, 1, or 10 μm. Cells were harvested, lysed, and total cell lysates were subjected to SDS-PAGE for Western blot analysis.

After in vitro daily treatment with gefitinib at various concentrations (0.1, 1, or 10 μm), phosphorylation of EGFR tyrosine 1173, a key regulatory tyrosine residue involved in EGFR (34 , 35) , as well as EGFRvIII signal transduction (18 , 19) , was reduced significantly in high EGFR-expressing A431, NR6W, and U87MG.wtEGFR cells, but was not as decreased in mutant EGFRvIII-expressing cells NR6M and U87MG.ΔEGFR at 24 h (Fig. 1, A and B)⇓ . However, after 3 days, phosphorylated EGFR expression in mutant EGFR-expressing cell lines was reduced with higher concentrations of gefitinib and was beginning to decline at lower treatment concentrations with prolonged exposure. Inasmuch, reductions in both wild-type and mutant receptor phosphorylation were sustained after day 3, continuing to decrease at days 5 and 7 for mutant EGFRvIII (Fig. 1, A and B)⇓ . As expected, neither form of EGFR was detected in NR6 Null cells (Fig. 1A)⇓ . Parental U87MG cells, which endogenously express low levels of wild-type EGFR (18) , did not display receptor phosphorylation, regardless of concentration or time tested (Fig. 1A)⇓ . Thus, mutant EGFRvIII does not appear to be as sensitive to TK inhibition as wild-type EGFR when treated with lower concentrations of gefitinib for shorter periods of time, but displays a dose-dependent inhibition with higher concentrations (1 and 10 μm) at day 5 and later (Fig. 1, A and B)⇓ . From this, EGFRvIII-expressing tumors may require increased concentration and repeated dosing to inhibit mutant receptor phosphorylation.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

A, phosphorylated epidermal growth factor receptor (EGFR) expression in wild-type EGFR and mutant EGFRvIII cells treated daily with various concentrations of gefitinib. After in vitro daily treatment of A431, NR6 Null, NR6W, NR6M, U87MG, U87MG.wtEGFR, and U87MG.ΔEGFR cells with vehicle-only or gefitinib at various concentrations (0.1, 1, or 10 μm), phospho-tyrosine 1173 EGFR, pan EGFR, and β-actin expression were analyzed by Western blot analysis for days 1, 3, 5, and 7. Days 1 and 3 are depicted for all cell lines, whereas days 5 and 7 are shown for illustrative purposes for mutant EGFRvIII cells. At each daily harvest, cells were washed and lysed at which time Western blot analysis was performed, as described in “Materials and Methods.” This figure is representative of three independent experiments. B, phosphorylated EGFR expression in wild-type EGFR and mutant EGFRvIII cells treated with various concentrations of gefitinib at days 1, 3, 5, and 7. Graphical representation of relative percentage of expression of phospho-tyrosine 1173 EGFR by Western blot in 0.1–10 μm gefitinib-treated NR6W and NR6M cells. This figure is representative of three independent experiments.

Phosphorylation of Akt Is Decreased with Increasing Concentration of Gefitinib in Cells Expressing Wild-Type EGFR but Is Resistant to Inhibition in Mutant EGFRvIII-Expressing Cells.

To determine what effect decreased phosphorylation of EGFR has on downstream signaling, we investigated the expression of the phosphatidylinositol-3′kinase downstream effector Akt, considered to be a promitotic kinase central to EGFR-promoted growth, neoplasia, and invasion (21 , 36 , 37) . Whereas the tumor suppressor PTEN, a lipid phosphatase that is a key negative regulator of Akt activity, is commonly mutated in glioblastoma multiforme tumors and cell lines such as U87MG (38 , 39) , it was not known if PTEN is functional in NR6 cells. Because mutant EGFRvIII expression and mutation of PTEN often occur concomitantly in glioblastoma multiformes, we were interested in the expression of Akt and PTEN in cells expressing wild-type EGFR or mutant EGFRvIII.

For the purposes of determining Akt expression in our studies, we investigated the phosphorylation of Akt serine 473, a residue that has been shown to be constitutively phosphorylated in some HER-2 overexpressing malignancies (40) and head and neck cancers (41) . We found that phosphorylated Akt was down-regulated after treatment with gefitinib in a dose-dependent manner in cells expressing EGFR, including U87MG.wtEGFR (Fig. 2, A and B)⇓ . However, phosphorylated Akt expression was unchanged in all of the mutant EGFRvIII-expressing cells assayed at all of the concentrations and times tested after treatment with gefitinib (Fig. 2, A and B)⇓ . It is interesting to note that Akt phosphorylation was not reduced in U87MG.ΔEGFR, which expresses low levels of wild-type EGFR and high levels of mutant EGFRvIII. These data would suggest that cells expressing mutant EGFRvIII, which is significantly inhibited in phosphorylation with repeated dosing and prolonged exposure, still signal intracellularly via Akt after treatment with gefitinib. As expected, immunoblot detection of PTEN in U87MG.wtEGFR and U87MG.ΔEGFR was unsuccessful (data not shown), as PTEN is mutated in this malignant glioma line (42) . However, PTEN was detected and constitutively phosphorylated in both NR6W and NR6M cells (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

A, phosphorylated Akt expression in wild-type epidermal growth factor receptor (EGFR) and mutant EGFRvIII cells treated daily with various concentrations of gefitinib. After in vitro daily treatment of A431, NR6 Null, NR6W, NR6M, U87MG, U87MG.wtEGFR, and U87MG.ΔEGFR cells with vehicle only or gefitinib at various concentrations (0.1, 1, or 10 μm), phospho-serine 473 Akt, pan Akt, and β-actin expression were analyzed by Western blot analysis for days 1, 3, 5, and 7. Day 3 is depicted for all cell lines, whereas days 5 and 7 are shown for illustrative purposes for mutant EGFRvIII cells. At each daily harvest, cells were washed and lysed at which time Western blot analysis was performed, as described in “Materials and Methods.” This figure is representative of three independent experiments. B, phosphorylated Akt expression in wild-type and mutant EGFRvIII cells treated with various concentrations of gefitinib at day 3. Graphical representation of relative percentage of expression of phospho-serine 473 Akt by Western blot in 0.1–10 μm gefitinib-treated A431, NR6 Null, NR6W, NR6M, U87MG, U87MG.wtEGFR, and U87MG.ΔEGFR cells. This figure is representative of three independent experiments.

Growth of Cells Expressing Wild-Type EGFR Are Inhibited in a Dose-Dependent Manner by Gefitinib, but Cells Expressing EGFRvIII Are Resistant to Inhibition.

On the basis of the observation that phosphorylation of Akt was differentially affected in cells expressing wild-type EGFR versus mutant EGFRvIII after treatment with gefitinib, we investigated the growth of these cells under the same conditions. Growth of NR6 Null cells was unaffected at all of the concentrations of gefitinib tested (Fig. 3)⇓ . NR6W cells demonstrated a dose-dependent reduction in growth (Fig. 3)⇓ and viability (data not shown) after treatment with gefitinib. In contrast, NR6M cells displayed an enhanced relative growth rate (Fig. 3)⇓ and unaffected viability (data not shown) with increasing gefitinib concentration and length of exposure when compared with cells expressing wild-type EGFR after treatment. These data were recapitulated in the U87MG glioblastoma cells, in which U87MG.ΔEGFR demonstrated a significantly greater growth rate than U87MG.wtEGFR, even at the highest concentrations of gefitinib tested (data not shown). We did not observe any nonspecific cytotoxicity for any of the cell lines tested at the concentration of 10 μm. However, we have determined gefitinib to be nonspecifically toxic at concentrations of 20 μm and higher in vitro for all of the cell lines tested here (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Growth of cells expressing no epidermal growth factor receptor (EGFR), wild-type EGFR, or mutant EGFRvIII when treated daily with gefitinib at various concentrations. Graphical representation of cell counting as determined by hemacytometry for NR6 Null, NR6W, and NR6M when treated with 0.1–10 μm gefitinib for 5 days. This figure is representative of three independent experiments; bars, ±SE.

To additionally investigate the effect that gefitinib has on cell cycle progression after treatment, propidium iodide staining and flow cytometric cell cycle analysis were performed. As described above, wild-type and mutant EGFRvIII-expressing cells were mock-treated or treated with 0.1, 1, or 10 μm gefitinib for 3 days. At the end of 3 days, cells were harvested, treated with RNase A, stained with propidium iodide, and analyzed by flow cytometry. G₀/G₁ gates were set to characterize whole cells with normal DNA content, i.e., indicative of cells in the resting/nondividing state, whereas S + G₂-M gates were set to characterize whole cells undergoing DNA synthesis or having twice the baseline DNA content, i.e., indicative of cells in the mitotic state. Treatment with gefitinib had no effect on NR6 Null at any concentration tested (Fig. 4)⇓ . However, gefitinib reduced DNA synthesis and cell cycle progression of wild-type EGFRexpressing cells in a dose-dependent manner, such that there were statistically significant increases in the resting G₀/G₁ phases and decreases in the mitotic S + G₂-M phases of A431, U87MG.wtEGFR (data not shown), and NR6W cells (Fig. 4)⇓ treated with increasing concentrations of gefitinib. There were statistically significant increases in G₀/G₁ for 1 μm and 10 μm gefitinib-treated NR6W cells compared with mock-treated NR6W cells (P = 0.036 and P = 0.032, respectively), whereas there were statistically significant decreases in S + G₂-M for 1 μm and 10 μm gefitinib-treated NR6W cells compared with mock-treated NR6W cells (P = 0.045 and P = 0.040, respectively). Collectively, these data are consistent with the observation that a dose-dependent growth reduction was observed for wild-type EGFR-expressing cells, as determined by cell counting (Fig. 3)⇓ .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

A, cell cycle analysis of S + G₂/M phases for cells expressing no epidermal growth factor receptor (EGFR), wild-type EGFR, or mutant EGFRvIII when treated daily with gefitinib at various concentrations. Propidium iodide staining assay for in vitro mock-treated and 0.1, 1, and 10 μm gefitinib-treated NR6 Null, NR6W, and NR6M cells, as described in “Materials and Methods.” P = 0.045 for NR6W (0) versus NR6W (1); P = 0.04 for NR6W (0) versus NR6W (10); P = 0.076 for NR6M (0) versus NR6M (10). This figure is representative of three independent experiments; bars, ±SE.

In contrast, gefitinib was not effective in reducing DNA synthesis in mutant EGFRvIII-expressing NR6M (Fig. 4)⇓ and U87MG.ΔEGFR cells (data not shown). Decreases in DNA synthesis for gefitinib-treated NR6M cells compared with mock-treated cells did not reach statistical significance (P > 0.05) at any concentration tested. These findings support the observation that mutant EGFRvIII-expressing cells survived treatment with gefitinib better (i.e., higher cell counts and higher viability) than cells expressing wild-type EGFR (Fig. 3)⇓ .

Invasion by Wild-Type EGFR-Expressing Cells Is Inhibited by Gefitinib, Whereas Invasion by Mutant EGFRvIII-Expressing Cells Is Resistant to Inhibition.

We have shown previously that expression of mutant EGFRvIII is correlated with the up-regulation of genes and molecular effectors intrinsically involved in tumor invasion (43) , as determined by quantitation of specific mRNA transcripts. In addition, we have demonstrated that mutant EGFRvIII confers an invasive phenotype when expressed in different cellular backgrounds (3 , 12) . To assess if the invasive effects of mutant EGFRvIII were inhibited by gefitinib, we determined how treatment with gefitinib affected invasion of wild-type EGFR and mutant EGFRvIII cells in vitro. NR6 Null, NR6W, and NR6M cells were mock-treated (DMSO only) or treated with gefitinib at 1 or 10 μm (in equivalent volumes of DMSO) for an initial 24 h and were then placed into control (migration) or Matrigel invasion assay chambers with fresh medium and drug for an additional 24 h, for a total culture time of 48 h. In comparison to mock-treated cells, invasion of NR6W cells was reduced by 48% and 69% after treatment with 1 μm and 10 μm gefitinib, respectively (Fig. 5)⇓ . In contrast, invasion by NR6M was reduced by 33% and 40% after treatment with 1 μm and 10 μm gefitinib, respectively (Fig. 5)⇓ . Thus, it appears that unlike wild-type EGFR, invasive characteristics conferred by mutant EGFRvIII are more resistant to inhibition by gefitinib, even with increasing drug concentration (Fig. 5)⇓ . We have obtained similar results that demonstrate there is a dose-dependent inhibition of invasion by A431, U87MG, and U87MG.wtEGFR cells, whereas U87MG.ΔEGFR cells are resistant to this inhibitor (data not shown). These data are consistent with previous observations we have made, as well as those made by others that mutant EGFRvIII expression strongly correlates with enhanced growth rate and invasion (43 , 44) , which help contribute to is resistance to inhibition even at the maximally tolerated dose in vitro as well as in vivo (26) .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Cell counts for migration and invasion of cells expressing no epidermal growth factor receptor (EGFR), wild-type EGFR, or mutant EGFRvIII when treated daily with gefitinib at various concentrations. Graphical representation of raw cell counts for NR6 Null, NR6W, and NR6M cells treated with 0.1–10 μm gefitinib for 48 h. This figure is representative of three independent experiments; bars, ±SE.