Cancer Cell Biology
Establishment of a human non-small cell lung cancer cell line resistant to gefitinib
Article first published online: 10 MAR 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 116, Issue 1, pages 36–44, 10 August 2005
How to Cite
Koizumi, F., Shimoyama, T., Taguchi, F., Saijo, N. and Nishio, K. (2005), Establishment of a human non-small cell lung cancer cell line resistant to gefitinib. Int. J. Cancer, 116: 36–44. doi: 10.1002/ijc.20985
- Issue published online: 12 MAY 2005
- Article first published online: 10 MAR 2005
- Manuscript Accepted: 21 DEC 2004
- Manuscript Received: 1 JUL 2004
- non-small cell lung cancer
The epidermal growth factor receptor (EGFR) tyrosine-kinase inhibitor gefitinib (Iressa®, ZD1839) has shown promising activity preclinically and clinically. Because comparative investigations of drug-resistant sublines with their parental cells are useful approaches to identifying the mechanism of gefitinib resistance and select factors that determine sensitivity to gefitinib, we established a human non-small cell lung carcinoma subline (PC-9/ZD) that is resistant to gefitinib. PC-9/ZD cells are ∼180-fold more resistant to gefitinib than their parental PC-9 cells and PC-9/ZD cells do not exhibit cross-resistance to conventional anticancer agents or other tyrosine kinase inhibitors, except AG-1478, a specific inhibitor of EGFR. PC-9/ZD cells also display significant resistance to gefitinib in a tumor-bearing animal model. To elucidate the mechanism of resistance, we characterized PC-9/ZD cells. The basal level of EGFR in PC-9 and PC-9/ZD cells was comparable. A deletion mutation was identified within the kinase domain of EGFR in both PC-9 and PC-9/ZD, but no difference in the sequence of EGFR cDNA was detected in either cell line. Increased EGFR/HER2 (and EGFR/HER3) heterodimer formations were demonstrated in PC-9/ZD cells by chemical cross-linking and immunoprecipitation analysis in cells unexposed to gefitinib. Exposure to gefitinib increased heterodimer formation in PC-9 cells, but not in PC-9/ZD cells. Gefitinib inhibits EGFR autophosphorylation in a dose-dependent manner in PC-9 cells but not in PC-9/ZD cells. A marked difference in inhibition of site-specific phosphorylation of EGFR was observed at Tyr1068 compared to other tyrosine residues (Tyr845, 992 and 1045). To elucidate the downstream signaling in the PC9/ZD cellular machinery, complex formation between EGFR and its adaptor proteins GRB2, SOS, and Shc was examined. A marked reduction in the GRB2-EGFR complex and absence of SOS-EGFR were observed in PC-9/ZD cells, even though the protein levels of GRB2 and SOS in PC-9 and PC-9/ZD cells were comparable. Expression of phosphorylated AKT was increased in PC-9 cells and inhibited by 0.02 μM gefitinib. But the inhibition was not significant in PC-9/ZD cells. These results suggest that alterations of adaptor-protein-mediated signal transduction from EGFR to AKT is a possible mechanism of the resistance to gefitinib in PC-9/ZD cells. These phenotypes including EGFR–SOS complex and heterodimer formation of HER family members are potential biomarkers for predicting resistance to gefitinib. © 2005 Wiley-Liss, Inc.
Chemotherapy has played a central role in the treatment of patients with inoperable NSCLC for over 30 years, although its efficacy seems to be of very limited value.1, 2 Human solid tumors, including lung cancer, glioblastoma, breast cancer, prostate cancer, gastric cancer, ovarian cancer, cervical cancer and head and neck cancer, express epidermal growth factor receptor (EGFR) frequently, and elevated EGFR levels are related to disease progression, survival, stage and response to therapy.2, 3, 4, 5, 6, 7, 8, 9, 10 The therapies directed at blocking EGFR function are attractive.
Interest in target-based therapy has been growing ever since the clinical efficacy of STI-571 was first demonstrated,11, 12, 13 and small molecules and monoclonal antibodies that block activation of the EGFR and HER2 have been developed over the past few decades. The leading small-molecule EGFR tyrosine-kinase inhibitor, gefitinib (Iressa®, ZD1839), has shown excellent antitumor activity in a series of Phase I and II studies,14, 15 and Phase II international multicenter trials (Iressa Dose Evaluation in Advanced Lung Cancer (IDEAL) 1 and 2) yield an overall RR of 11.8–18.4% and overall disease control rate of 42.2–54.4% (gefitinib 250 mg/day) in patients with advanced non-small cell lung cancer (NSCLC) who had undergo at least 2 previous treatments with chemotherapy. INTACT 1 and 2 (‘Iressa’ NSCLC Trials Assessing Combination Therapy) have demonstrated that gefitinib does not provide improvement in survival when added to standard first line platinum-based chemotherapy vs. chemotherapy alone in advanced NSCLC.16, 17 Two small retrospective studies reported recently that activating mutation of EGFR correlate with sensitivity and clinical response to gefitinib and erlotinib.18, 19, 20 Although information of EGFR mutation may enable to identify the subgroup of patients with NSCLC who will respond to gefitinib and erlotinib, it would be expected that acquired resistance would develop in such patients after treatment. The problem of acquired resistance to gefitinib might be growing, but there has been no preclinical research about the mechanism of developing resistance to gefitinib. We established resistant subline using PC-9 that is highly sensitive to gefitinib.
Establishment of drug-resistant sublines and comparative investigations with their parental cells to identify their molecular, biological and biochemical properties are useful approaches to elucidating the mechanism of the drug's action. Our study describes the establishment of a gefitinib-resistant cell line and its characterization at the cellular and subcellular levels. The PC-9/ZD cell line is the first human NSCLC cell line resistant to gefitinib ever reported. PC-9 is a lung adenocarcinoma cell line that is highly sensitive to gefitinib at its IC50-value of 0.039 μM, but the PC-9/ZD subline, which has a level of EGFR expression comparable to that of PC-9 cells, is specifically resistant to gefitinib. Thus, PC-9 and PC-9/ZD cells will provide useful information about the mechanism of developing resistance to gefitinib and molecules as surrogate markers for predicting chemosensitivity to gefitinib.
Material and methods
Drugs and cells
Gefitinib(N-(3-chloro-4-fluorophenyl)-7-methoxy-6-[3-(morpholin-4-yl)propoxy] quinazolin-4-amine) was supplied by AstraZeneca Pharmaceuticals (Cheshire, UK). AG-1478, AG-825, K252a, staurosporin, genistein, RG-14620 and Lavendustin A were purchased from Funakoshi Co. Ltd (Tokyo, Japan).
NSCLC cell line PC-9 (derived from a patient with adenocarcinoma untreated previously) was provided by Prof. Hayata of Tokyo Medical University (Tokyo, Japan).21 PC-9 and PC-9/ZD cells were cultured in RPMI-1640 medium (Sigma, St. Louis, MO) supplemented with 10% FBS (GIBCO-BRL, Grand Island, NY), penicillin and streptomycin (100 U/ml and 100 μg/ml, respectively; GIBCO-BRL) in a humidified atmosphere of 5% CO2 at 37°C. Gefitinib-resistant PC-9/ZD cells were selected from a subculture that had acquired resistance to gefitinib using the following procedure. Cultured PC-9 cells were exposed to 2.5 μg/ml N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) for 24 hr and then washed and cultured in medium containing 0.2 μM gefitinib for 7 days. After exposure to gefitinib, they were washed and cultured in drug-free medium for 14 days. When variable cells had increased, they were seeded in medium containing 0.3–0.5 μM of gefitinib on 96-well cultured plates for subcloning. After 21–28 days, the colonies were harvested and a single clone was obtained. The subcloned cells exhibited an 182-fold increase in resistance to the growth-inhibitory effect of gefitinib as determined by MTT assay, and the resistant phenotype has been stable for at least 6 months under drug-free conditions.
In vitro growth-inhibition assay
The growth-inhibitory effects of cisplatin, carboplatin, adriamycin, irinotecan, gemcitabine, vindesine, paclitaxel, genistein, K252a, staurosporin, AG-825, AG-1478, Tyrophostin 51, RG-14620, Lavendustin A and gefitinib in PC-9 and PC-9/ZD cells were examined by using a 3-(4,5-dimethylthiazol-2-yl)- 2,5-diphenyltetrazolium bromide (MTT) assay.22 A 180 μl volume of an exponentially growing cell suspension (6 × 103 cells/ml) was seeded into a 96-well microtiter plate, and 20 μl of various concentrations of each drug was added. After incubation for 72 hr at 37°C, 20 μl of MTT solution (5 mg/ml in PBS) was added to each well, and the plates were incubated for an additional 4 hr at 37°C. After centrifuging the plates at 200g for 5 min, the medium was aspirated from each well and 180 μl of DMSO was added to each well to dissolve the formazan. Optical density was measured at 562 and 630 nm with a Delta Soft ELISA analysis program interfaced with a Bio-Tek Microplate Reader (EL-340, Bio-Metallics, Princeton, NJ). Each experiment was carried out in 6 replicate wells for each drug concentration and carried out independently 3 or 4 times. The IC50-value was defined as the concentration needed for a 50% reduction in the absorbance calculated based on the survival curves. Percent survival was calculated as: (mean absorbance of 6 replicate wells containing drugs − mean absorbance of six replicate background wells)/(mean absorbance of 6 replicate drug-free wells − mean absorbance of 6 replicate background wells) × 100.
In vivo growth-inhibition assays
Experiments were carried out in accordance with the United Kingdom Coordinating Committee on Cancer Research Guidelines for the welfare of animals in experimental neoplasia (2nd ed.). Female BALB/c nude mice, 6-weeks-old, were purchased from Japan Charles River Co. Ltd (Atsugi, Japan). All mice were maintained in our laboratory under specific-pathogen-free conditions. In vivo experiments were scheduled to evaluate the effect of oral administration of gefitinib on pre-existing tumors. Ten days before administration, 5 × 106 PC-9 or PC-9/ZD cells were injected subcutaneously (s.c.) into the back of the mice, and gefitinib (12.5, 25 or 50 mg/kg, p.o.) was administered to the mice on Days 1–21. Tumor diameter was measured with calipers on Days 1, 4, 8, 11, 14, 19 and 22 to evaluate the effect of treatment, and tumor volume was determined by using the following equation: tumor volume = ab2/2 (mm3) (where a is the longest diameter of the tumor and b is the shortest diameter). Day “x” denotes the day on which the effect of the drugs was estimated, and Day “1” denotes the first day of treatment. All mice were sacrificed on Day 22, after measuring their tumors. We considered absence of a tumor mass on Day 22 to indicate a cure. Differences in tumor sizes between the treatment groups and control group at Day 22 were analyzed by the unpaired t-test. A p-value of <0.05 was considered statistically significant.
cDNA expression array
The gene expression profile of PC-9/ZD was assessed with an Atlas Nylon cDNA Expression Array (BD Bioscience Clontech, Palo Alto, CA). Total RNA was extracted by a single-step guanidinium thiocyanate procedure (ISOGEN, Nippon Gene, Tokyo, Japan). An Atlas Pure Total RNA Labeling System was used to isolate RNA and label probes. The materials provided with the kit were used, and the manufacturer's instructions were followed for all steps. Briefly, streptavidin-coated magnetic beads and biotinylated oligo(dT) were used to isolate poly A RNA from 50 μg of total RNA and the RNA obtained was converted into 32P-labeled first-strand cDNA with MMLV reverse transcriptase. The 32P-labeled cDNA fraction was purified on NucleoSpin columns and was added to the membrane on which fragments of 777 genes were spotted. Hybridization was allowed to proceed overnight at 68°C. After washing, the radiolabeled spots were visualized and quantified by BAS-2000II and Array Gauge 1.1 (Fuji Film Co., Ltd., Tokyo, Japan). The data were adjusted for the total density level of each membrane.
Quantitative real-time RT-PCR analysis
Total RNAs extracted from PC-9 cells and PC-9/ZD cells (1 × 106 cells each) were incubated with DNase I (Invitrogen, Carlsbad, CA) for 30 min. First-strand cDNA synthesis was carried out on 1 μg of RNA in 10 μl of a reaction mixture with 50 pmol of Random hexamers and 50 U of M-MLV RTase. Oligonucleotide primers for human EGFR were obtained from Takara (HA003051, Takara Bio Co., Tokyo, Japan). For PCR calibration, we generated a calibrator dilution series for EGFR cDNA in pUSEamp vector (Upstate, Charlottesville, VA) ranging from 108–102 copies/1 μl. A total of 2 μl of reverse transcriptase products was used for PCR amplification using Smart Cycler system (Takara) according to manufacturer's instructions. Absolute copy numbers were calculated back to the initial cell numbers, which were set into the RNA extraction. As a result we obtained copies/cell:ratio representing the average EGFR RNA amount per cell.
Immunoprecipitation and immunoblotting
The cultured cells were washed twice with ice-cold PBS, and lysed in EBC buffer (50 mM Tris-HCI, pH 8.0, 120 mM NaCl, 0.5% Nonidet P-40, 100 mM NaF, 200 mM Na orthovanadate, and 10 mg/ml each of leupeptin, aprotinin, pepstatin A and phenylmethylsulphonyl fluoride). The lysate was cleared by centrifugation at 15,000 r.p.m. for 10 min, and the protein concentration of the supernatant was measured by BCA protein assay (Pierce, Rockford, IL). The membrane was probed with antibody against EGFR (1005; Santa Cruz, Santa Cruz, CA), HER2/neu (c-18; Santa Cruz), HER3 (c-17; Santa Cruz), HER4 (c-18; Santa Cruz), PI3K (4; BD), Grb2 (81; BD), SOS1/2 (D-21; Santa Cruz), Shc (30; BD, San Jose, CA), PTEN (9552; Cell Signaling, Beverly, MA), AKT (9272; Cell Signaling), phospho-EGFR specific for Tyr 845, Tyr 992, Tyr 1045, and Tyr 1068 (2231, 2235, 2237, 2234; Cell Signaling), phospho-AKT (Ser473) (9271; Cell Signaling), phospho-Erk (9106; Cell Signaling), and phospho-Tyr (PY-20; BD) as the first antibody, and then with by horseradish-peroxidase-conjugated secondary antibody. The bands were visualized by enhanced chemiluminescence (ECL Western Blotting Detection Kit, Amersham, Piscataway, NJ). For Immunoprecipitation, 5 × 106 cells were washed, lysed in EBC buffer, and centrifuged, and the supernatants obtained (1,500 μg) were incubated at 4°C with the anti-EGFR (1005), -HER2 (c-18), and -HER3 (c-17) Ab overnight. The immunocomplexes were absorbed onto protein A/G-Sepharose beads, washed 5 times with lysate buffer, denatured, and subjected to electrophoresis on a 7.5% polyacrylamide gel.
Analysis of the genes of the HER families by direct sequencing
Total RNAs were extracted from PC-9 and PC-9/ZD cells with ISOGEN (Nippon Gene) according to manufacturer's instructions. First-strand cDNA was synthesized from 2 μg of total RNA by using 400 U of SuperScript II (Invitrogen, Carlsbad, CA). After reverse transcription with oligo (dT) primer (Invitrogen) or random primer (Invitrogen), the first-strand cDNA was amplified by PCR by using specific primers for EGFR, HER2 and HER3. The reaction mixture (50 μl) contained 1.25 U AmpliTaq DNA polymerase (Applied Biosystem, Foster City, CA), and amplification was carried out by 30 cycles of denaturation (95°C, 30 sec), annealing (55–59°C, 30 sec), and extension (72°C, 30 sec) with a GeneAmp PCR System 9600 (Applied Biosystem). After amplification, 5 μl of the RT-PCR products was subjected to electrophoretic analysis on a 2% agarose gel with ethidium bromide. DNA sequencing of the PCR products was carried out by the dideoxy chain termination method using the ABI PRISM 310 Genetic Analyzer (Applied Biosystem).
Chemical cross-linking in intact cells was carried out as described previously.23 In brief, after 6 hr exposure to 0.2 μM gefitinib, cells were washed with PBS and incubated for 25 min at 4°C in PBS containing 1.5 mM of the nonpermeable cross-linker bis (sulfosuccinimidyl) substrate (Pierce, Rockford, IL). The reaction was terminated by adding 250 mM glycine for 5 min while rocking. Cells were washed in EBC buffer and 20 μg of protein was resolved by 5–10% gradient SDS-PAGE, and then immunoblot analyzed for EGFR, HER2, HER3 and P-Tyr.
Sensitivity of PC-9/ZD cells to cytotoxic agents and tyrosine kinase inhibitors
No significant difference between PC-9 and PC-9/ZD cells was observed in in vitro cell growth (doubling time of 20.3 hr and 21.4 hr, respectively) and microscopic morphology. Figure 1 shows the growth-inhibitory effect of gefitinib on the parent PC-9 cell line and its resistant subline, PC-9/ZD. The IC50-value of gefitinib in PC-9 cells was 0.039 μM, as compared to 7.1 μM in PC-9/ZD cells (182-fold resistance). PC-9/ZD cells exhibited no cross-resistance to other conventional anticancer agents, including cisplatin, carboplatin, adriamycin, vindesine, paclitaxel and irinotecan. We also examined the growth-inhibitory effect of the EGFR tyrosine kinase inhibitors AG-1478, RG-14620 and Lavendustin A and other tyrosine kinase inhibitors in PC-9 and PC-9/ZD cells. PC-9/ZD cells show cross-resistance to AG1478, but not to all of the tyrosine kinase inhibitors (Tables I, II). It is likely that PC-9/ZD would also be resistant to EGFR-targeted quinazoline derivatives including gefitinib and erlotinib.20
|Drug||IC50 values (μM)1||RR2 1.6|
|Cisplatin||1.9 ± 0.7||3.1 ± 1.5||2.0|
|Carboplatin||25 ± 21||49 ± 23||1.3|
|Adriamycin||0.16 ± 0.13||0.20 ± 0.15||2.2|
|Irinotecan||15 ± 10||32 ± 11||1.5|
|Etoposide||4.5 ± 1.5||6.6 ± 1.3||1.5|
|Gemcitabine||18 ± 1.5||27 ± 1.5||0.7|
|Vindesine||0.0046 ± 0.0004||0.0032 ± 0.0009||1.2|
|Paclitaxel||0.0041 ± 0.0011||0.0048 ± 0.0004||1.6|
|Inhibitor||Target||IC50 values (μM)||RR2|
|AG-1478||EGFR||0.052 ± 0.02||6.0 ± 0.8||117|
|RG-14620||EGFR||13 ± 1.0||13 ± 2.5||1.0|
|Lavendustin A||EGFR||20 ± 4.6||27 ± 2.6||1.3|
|Genistein||TK||18 ± 1.5||27 ± 1.5||1.5|
|K252a||PKC||0.47 ± 0.17||0.63 ± 0.04||1.3|
|Staurosporin||PKC||0.0036 ± 0.0019||0.004 ± 0.0014||1.1|
PC-9/ZD cells show significant resistance to gefitinib in an in vivo model
To ascertain whether the resistance of PC-9/ZD occurs in vivo, we investigated the growth-inhibitory effect of gefitinib on PC-9 cells and PC-9/ZD cells in a xenotransplanted model. There was no significant difference in the size of the of PC-9 and PC-9/ZD cell tumor masses in nude mice before the start of gefitinib injection. Figure 2 shows the growth-inhibition curve of PC-9 (Fig. 2a) and PC-9/ZD (Fig. 2b) cells in vivo during the observation period. The PC-9 tumor masses decreased markedly in volume at all doses of gefitinib. In the 50 mg/kg/day p.o. group, the PC-9 masses were eradicated in all mice and did not regrow within the observation period. Growth of the PC-9/ZD masses, on the other hand, was inhibited by gefitinib administration in a dose-dependent manner, but significant tumor reduction was observed only in the 25 and 50 mg/kg/day groups, and the PC-9/ZD masses were not eradicated even in 50 mg/kg/day group. These results clearly demonstrate the significant in vivo resistance of PC-9/ZD cells to gefitinib.
Expression of HER family members and related molecules in PC-9 and PC-9/ZD cells
We examined the gene expression and protein levels of HER family members and related molecules by cDNA expression array (followed by confirmation using RT-PCR, data not shown) and immunoblotting. The ratios of the protein expression levels of PC-9 cells to PC-9/ZD cells almost paralleled the expression levels of their genes (Fig. 3a). The basal level of EGFR was comparable or slightly higher in PC-9/ZD cells (Fig. 3a,b), whereas the HER3 and AKT levels were lower in resistant cells.
We carried out quantitative RT-PCR to measure the copy numbers of EGFR. Estimated transcript levels of EGFR were 786.3 and 712.1 copies/cell for PC-9 cells and PC-9/ZD cells, respectively (Fig. 3d). Relative ratio of EGFR expression levels in PC-9 cells and PC-9/ZD cells is 1.104. Microarray analysis using CodeLink Bioarray (Amersham Bio, Piscataway, NJ) confirmed equivalent gene expression of EGFR with ratio of 1.002 between PC-9 and PC-9/ZD cells (data not shown).
Expression of PI3K, Grb2, SOS, and Shc, the adaptor proteins of EGFR, and PTEN was almost the same in PC-9 and PC-9/ZD cells, and no change in the protein levels was observed after exposure to gefitinib (data not shown). The relative densitometric units of each protein are shown in Figure 3c. These results suggest that the difference in protein levels of EGFR, HER2, and related proteins can not explain the high resistance of PC-9/ZD cells to gefitinib.
Sequence of HER family member in PC-9/ZD cells
Several reports suggest that the resistance to receptor tyrosine kinase inhibitor STI-571 is partially due to mutations in the ATP-binding site of the Bcr-Abl, the target of the drug.24, 25, 26, 27 We analyzed the sequences of the cDNAs of EGFR, HER2, and HER3, but found no differences in their sequences between PC-9 and PC-9/ZD cells. We did detect a deleted position of EGFR in both cell lines that results in deletion of 5 amino acids (Glu722, Leu723, Arg724, Glu725, and Ala726) (Fig. 4). Our findings indicate that the deletion does not directly contribute to the cellular resistance.
Inhibitory effect of gefitinib on autophosphorylation of EGFR in PC-9/ZD cells
Phosphorylation of EGFR is necessary for EGFR-mediated intracellular signaling. Although the EGFR phosphorylation levels of tumors were thought to be correlated with sensitivity to gefitinib, the basal level of phosphorylated EGFR in PC-9 and PC-9/ZD cells is almost the same. Gefitinib inhibited EGFR autophosphorylation in a dose-dependent manner and completely inhibited its phosphorylation at 0.2–2 μM in PC-9 cells (Fig. 5a), but its inhibitory effect on autophosphorylation of EGFR in PC-9/ZD cells was less than in PC-9 cells (Fig. 5a). Because each phosphorylation site of EGFR has a different role in the activation of downstream signaling molecules, we examined the inhibitory effect of gefitinib on site-specific phosphorylation of EGFR. Phosphorylation of several different EGFR tyrosine residues (Tyr845, Tyr992 and Tyr1068) was dose-dependently inhibited by gefitinib in PC-9 cells, whereas no clear inhibitory effects of gefitinib on phosphorylation at Tyr 845 and Tyr1068 residues in PC-9/ZD cells was observed (Fig. 5b,c,e). The most marked difference of in inhibition between the cells was observed at Tyr1068 (Fig. 5e). Tyr1045 showed resistance to inhibition of autophosphorylation by gefitinib in both PC-9 and PC-9/ZD cells (Fig. 5d).
Complex formation of EGFR and its adaptor proteins
Tyr1068 of EGFR is the tyrosine that is most resistant to inhibition of autophosphorylation by gefitinib in PC-9/ZD cells. Because the Tyr 1068 is a direct binding site for the GRB2/SH2 domain, and its phosphorylation is related to the complex formation of EGFR-adaptor proteins and their signaling, we examined complex formation between EGFR and the adaptor proteins GRB2, SOS, Shc, and PI3K by immunoprecipitation. The level of expression of these proteins in PC-9 and PC-9/ZD cells were similar (Fig. 3a). A smaller amount of EGFR-GRB2 complex was observed in PC-9/ZD cells and no EGFR-SOS complex was detected at all (Fig. 6). The amount of HER2- or HER3-GRB2 complex in PC-9 and PC-9/ZD cells was similar, and no decreases in complex formation were observed after exposure to gefitinib. A decreased amount of HER2-SOS complex and inability to detect HER3-SOS complex were also observed in PC-9/ZD cells. HER2-PI3K complex increased in PC-9/ZD. There are no significant differences in complex formation between SHC and EGFR, HER2, or HER3 between PC-9 and PC-9/ZD cells. These results suggest that GRB2-SOS-mediated signaling may be inactivated in PC-9/ZD cells.
Heterodimerization of HER family member in PC-9/ZD cells
Dimerization of members of the HER family is essential for activation of their catalytic activity and their signaling. We examined the effect of gefitinib on the dimerization of HER family members by immunoblotting, immunoprecipitation and chemical cross-linking analysis (Figs. 3a, 5a, 7a). The expression levels of EGFR and HER2 were similar and the HER3 level was lower in PC-9/ZD cells by immunoblotting (Fig. 3a). A chemical cross-linking assay showed that in the absence of gefitinib the amount of high molecular weight complexes (∼400 kDa) that are recognized by anti-EGFR antibody (EGFR dimers), including formations of homodimers and heterodimers (EGFR-EGFR, EGFR-HER2 or EGFR-HER3), was almost the same in PC-9 and PC-9/ZD cells, whereas HER2 dimerization detected by anti-HER2 antibody was remarkably lower in PC-9/ZD cells (Fig. 7a). Increased EGFR/HER2 (and EGFR/HER3) heterodimer formation was detected in PC-9/ZD cells by immunoprecipitation analysis (Fig. 5a). The proportion of EGFR heterodimer to homodimer is increased significantly in PC-9/ZD (Fig. 7b). When exposed to gefitinib at a concentration of 0.2 μM for 6 hr the amount of dimer-formation increased similarly in PC-9 and PC-9/ZD cells (Fig. 7a), whereas marked induction of hetero-dimerization of EGFR-HER2 was observed only in PC-9 cells (Fig. 5a). These results suggest that a difference in hetero- or homo-dimerization is a possible determinant factor of gefitinib sensitivity.
AKT and MAPK pathways in PC-9/ZD cells
Because phosphorylation at Tyr 1068 of EGFR plays an important role for transduction of the signal to downstream of MAPK and AKT pathway,28, 29 we examined the difference between PC-9 and PC-9/ZD cells in downstream signaling. The basal level of phosphorylated AKT is higher in PC-9 cells than in PC-9/ZD cells, and although gefitinib inhibited AKT phosphorylation in a dose-dependent manner (Fig. 8a), the inhibitory effect of gefitinib on phosphorylation of AKT in PC-9/ZD cells was significantly less than in PC-9 cells (Fig. 8a). This difference in the inhibitory effect of gefitinib on AKT phosphorylation between PC-9 and PC-9/ZD cells is very similar to the difference in effect on EGFR autophosphorylation. No inhibition of phosphorylation of MAPK by gefitinib was observed in either cell line (Fig. 8b). These results suggest that downregulation of activated AKT is closely correlated with the cellular sensitivity to gefitinib, but that inhibition of the MAPK pathway does not contribute to drug sensitivity.
Interest in resistance to target-based therapy (TBT) has been growing ever since clinical efficacy was first demonstrated.11, 12, 13 Although CML patients respond to STI-571 well at first, most patients eventually relapse in the late stage of the disease.25, 26, 27 It has been reported that some patients in whom treatment with gefitinib is effective at first, ultimately become refractory.30 Resistance is likely to remain a hurdle that limits the long-term effectiveness of TBT. PC-9 had a deletion mutation within the kinase domain of EGFR and is highly sensitive. These characters are similar to those of NSCLC with clinical responsiveness to gefitinib. Analyzing the mechanism of resistance of PC-9/ZD subline might be clinically meaningful.
The mechanism of drug resistance is thought to be multifactorial. Because the growth-inhibitory assay in our present study showed no cross resistance to a variety of cytotoxic agents, the mechanism of the resistance differs from the mechanism of multidrug resistance patterns. Although expression of BCRP, one of the multidrug-resistance-related proteins has been reported to contribute to the resistance to gefitinib,31 expression of BCRP mRNA is observed only in PC-9 cells (data not shown). Although mutations in the ATP-binding pocket of BCR-ABL gene have been identified recently in cells from CML patients who were refractory to STI-571 treatment or relapse,25, 26, 27 there have been no reports of any such mutations for gefitinib resistance. PC-9/ZD also became refractory to gefitinib without secondary mutation in EGFR cDNA. These suggest the possibility of refractory tumor after treatment of gefitinib including this kind of phenotype.
There is no significant difference in expression level of EGFR between PC-9 and PC-9/ZD. Does the antitumor effect of gefitinib require EGFR expression? Naruse et al.32 suggested that the high sensitivity of K562/TPA to gefitinib is due to acquired EGFR expression. In their study autophosphorylation of EGFR in K562/TPA cells was inhibited by 0.01 μM gefitinib, and the IC50-value of gefitinib in parental K562 cells, which do not express EGFR, was approximately 400-fold higher than that in the K562/TPA subline. Furthermore, most patients who responded to gefitinib therapy have EGFR mutation in lung tumor.18, 19 These findings suggest strongly that gefitinib exerts its antitumor effect through an action on EGFR. Our present study showed similar EGFR expression and autophosphorylation levels in PC-9 and PC-9/ZD cells. The inhibitory effect of gefitinib on phosphorylation of EGFR is different. PC-9/ZD did not show cross-resistance to the specific EGFR TK inhibitors RG-14620 and Lavendustin A in an MTT assay, nor did inhibit the phosphorylation of EGFR at the cellular level (data not shown). Paez et al.18 reported that phosphorylation of EGFR in gefitinib-resistant cell lines was inhibited only when gefitinib was present at high concentration. These findings suggest that the difference in the inhibitory-effect on EGFR phosphorylation may determine the efficacy of the drug.
The inhibitory effect of gefitinib on EGFR phosphorylation is not significant in PC-9/ZD cells despite the absence of differences in the sequences of EGFR, HER2, and HER3. There are several possible explanations for the difference in inhibitory effect. First, the avidity of gefitinib for the ATP-binding site of EGFR may be decreased in PC-9/ZD cells due to a protein-protein interaction, i.e., EGFR and a certain protein prevent gefitinib from binding to EGFR. Second, a change in the activity of specific protein-tyrosine kinase or phosphatase of EGFR in PC-9/ZD cells, especially after exposure to gefitinib, may result in resistance to inhibition of EGFR phosphorylation. The phosphorylation level is maintained in exquisite balance by the reciprocal activities of kinase and phosphatase,33, 34 and Wu reported that phosphatase plays a role in STI571-resistance.35 Third, increased heterodimer formation by EGFR with other members of the HER family results in the limited inhibition. Heterodimer formation is increased in PC-9/ZD cells under basal conditions, and no increase in formation was observed after exposure to gefitinib, although marked heterodimer induction was observed in PC-9 cells. Calculations in in vitro studies have shown that the IC50-value for inhibition of the tyrosine kinase activity of EGFR is 0.023–0.079 μM, whereas the IC50-value for inhibition of HER2 is 100-fold higher.36 We estimate that the inhibitory effect of gefitinib depends on the ratio of homodimer formation to heterodimer formation, and the heterodimer may be one of the routes of escape from the action of gefitinib.
Signal transduction by the HER family member is mediated by 2 major pathways, the MAPK signaling pathway and the AKT signaling pathway, which regulate cell proliferation and survival. Because phosphorylated AKT was inhibited completely by gefitinib in PC-9 cells, but inhibition of phosphorylated MAPK was not significant, inhibition of the AKT pathway may be more important to cell sensitivity than inhibition of MAPK. Moasser et al.37 reported consistent results, showing that downregulation of AKT activity is predominantly seen in tumors that are sensitive to gefitinib. The phosphorylation of AKT and MAPK was not inhibited significantly by gefitinib in PC-9/ZD cells. This finding might be attributable to inactivation of Tyr 1068-GRB2-SOS-mediated signaling.
Based on the results of this comparative study, EGFR-GRB2-SOS complex formation, phosphorylation of Tyr1068, the ratio of the amount of homodimer formation to heterodimer formation, and the AKT signaling pathway are possible predictive biomarkers for gefitinib sensitivity. As a different approach, we are now looking for the genes associated with gefitinib resistance in PC-9/ZD cells compared to PC-9 cells by subtractive cloning.
‘Iressa’ is a trademark of the AstraZeneca group of companies.
- 1Addressing the optimal duration of therapy in advanced, metastatic non-small-cell lung cancer. In: PerryMC,eds. American Society of Clinical Oncology Educational Book. Alexandria: Lisa Greaves, 2003; 144–52..
- 4Epidermal growth factor-related peptides and their receptors in human malignancies. Crit Rev Oncol Hematol 1995; 19: 182–232., , , .
- 32Antitumor activity of the selective epidermal growth factor receptor- tyrosine kinase inhibitor (EGFR-TKI) Iressatm (ZD1839) in a EGFR-expressing multidrug resistant cell line in vitro and in vivo. Int J Cancer 2002; 98: 310–5., , , , , , , .
- 35Tyrosine kinases and phosphatase play a role in STI571-mediate apoptosis of chronic myelogenous leukemia cells. Proc Am Assoc Cancer Res 2003; 44., , .