Boe Sorensen, Department of Clinical Biochemistry, Aarhus University Hospital, Norrebrogade 44, DK-8000 Aarhus C, Denmark. e-mail: Boess@as.aaa.dk
To investigate the effect of the epidermal growth factor receptor (EGFR) on the induction of apoptosis by the chemotherapeutic agent etoposide (VP16), and to examine the effect of combining VP16 with gefitinib to see if the cell-survival mechanism can be prevented.
MATERIALS AND METHODS
The bladder cancer cell lines RT4 and T24, representing low- and high-malignancy grades respectively, were treated with VP16 (10 or 50 µm) and the level of apoptosis determined using a commercial kit. EGFR receptor activity was determined by western blotting using antibodies against phosphorylated EGFR. The EGFR was either activated by heparin-binding (HB)-EGF (1 nm) or inhibited with the specific EGFR inhibitor gefitinib (1 or 5 µm). The pan-caspase inhibitor Z-VAD (30 µm) was used to test the involvement of caspase activity.
Treatment of T24 bladder cancer cells with VP16 (50 µm) for 48 h induced phosphorylation of the EGFR and activation of the EGFR prevented the apoptosis induced by VP16. Thus, treatment of T24 cells with 50 µm VP16 for 48 h resulted in 19% apoptosis. However, activation of the EGFR with HB-EGF (1 nm) with VP16 (50 µm) significantly reduced the level of apoptosis by 25% (P < 0.05) showing that activating the EGFR has a cell-survival function. Inhibiting the EGFR with gefitinib (5 µm) blocked the VP16-induced activation of the EGFR. Combined treatment with gefitinib and VP16 resulted in 45% apoptotic cells, i.e. more than double the percentage of apoptotic cells with VP16 alone. This was found in both T24 and RT4 cells. Gefitinib used alone (1 and 5 µm) generated no apoptosis in the cells. Treatment of T24 cells with Z-VAD showed that apoptosis induced by both VP16 alone and VP16 with gefitinib was caspase-mediated.
These results suggest that activation of the EGFR induced a cell-survival function when bladder cancer cells were treated with the DNA-damaging drug VP16, and that combined treatment with VP16 and the EGFR inhibitor gefitinib might improve the efficacy of treatment.
The incidence of human bladder cancer is increasing and tumours are often significantly resistant to chemotherapy, with relapse to either intravesical or systemic treatment . This has lead to growing interest in developing new treatment strategies to improve the effect of chemotherapy. One important group of drugs used in bladder cancer treatment is the topoisomerase inhibitors. These classic genotoxic drugs inhibit topoisomerase I or II, leading to DNA-strand breaks, which ultimately result in apoptosis. Representatives of these drugs are doxorubicin and etoposide (VP16). Several studies describe mechanisms for resistance to these drugs [2,3], and our recent observation that VP16 induces the expression of the epidermal growth factor receptor (EGFR)-activating ligands , makes it interesting to examine if activation of the EGFR might represent a mechanism for developing resistance towards VP16, and if the novel EGFR inhibitor gefitinib can prevent this resistance.
The EGF system is involved in the proliferation of tumour cells in numerous epithelial cancers, including bladder cancer [1,5,6]. The EGF system consists of a family of four receptors, i.e. the EGFR, also called human EGF receptor (including HER-1, -2, -3 and -4). In addition there are 10 ligands, i.e. EGF, amphiregulin, TGF-α, betacellulin, heparin-binding EGF-like growth factor (HB-EGF), epiregulin and four neuregulins, which activate the receptors. The EGFR, which has been most extensively studied, is a 170-kDa transmembrane glycoprotein that contains an intrinsic tyrosine kinase domain. The receptors are only active in a dimeric form consisting of either homo- or heterodimers . Activation results in conformational changes within the receptor complex leading to phosphorylation of specific tyrosine residues in the tyrosine kinase domain. This in turn transmits the signal to important downstream molecules like signal transducers and activators of transcription (STAT), the mitogen-activated protein kinase (MAPK), protein kinase B (Akt) and more. These pathways have all been linked to cell proliferation and cell survival in cell-culture studies, influencing important factors that decide whether a cell should proliferate or enter the apoptotic pathway [8–11]. Interestingly, in patients with bladder cancer it was reported that there is a significant correlation between a poor prognosis and both the EGFR and HER-2, and several ligands from the EGF system activating these two receptors [12,13], while the presence of the receptors HER-3 and HER-4 and their activating ligands are associated with a good prognosis . The identification of any correlation between a poor prognosis and elevated expression of the EGFR in several cancers has lead to the development of several drugs targeting the EGFR and thereby inhibiting its activity . Gefitinib is one of these compounds and belongs to the group of small tyrosine kinase inhibitors that specifically target the EGFR. This drug has been particularly efficient in blocking the EGFR by preventing phosphorylation in the ATP-binding pocket of the receptor. In patients, gefitinib has been investigated for treating non-small lung cancer as a second- or third-line therapy .
We previously showed that treating the bladder cancer cell lines HCV29 and T24 with the genotoxic drug VP16 leads to a significant increase in both mRNA and protein levels of several ligands, especially HB-EGF and amphiregulin from the EGF system, while the level of mRNA coding for the receptors remain unchanged . In the present study we tested if the activity of the EGFR was involved in a cell-survival mechanism against VP16-induced apoptosis in bladder cancer cells. Furthermore, we examined the effect of combining VP16 with gefitinib to see if the cell survival mechanism could be prevented.
MATERIALS AND METHODS
Gefitinib (ZD1839, Iressa) was purchased from proteinkinase.de (Kassel, Germany); VP16 was obtained from Bristol Meyers (New York, USA); the pan-caspase inhibitor Z-VAD.fmk was obtained from R&D Systems (Abingdon, UK); and antibodies used were p-EGFR (Tyr1173) sc-12351 and antiβ-actin (AC-74).
Two bladder-cancer cell lines (RT4 and T24) were obtained from the American Type Culture collection (ATCC, Rockville, MD, USA); they represent low and high malignancy, respectively, and both express considerable amounts of the EGFR [17,18]. The cell lines were grown in complete Dulbecco’s modified Eagle medium (Gibco, Invitrogen, Taastrup, Denmark) supplemented with 10% fetal bovine serum and streptomycin (100 mg/mL), and penicillin (100 U/mL) in a humidified atmosphere of 5% CO2 at 37 °C.
Gefitinib in stock concentrations of 2 and 10 mm dissolved in dimethylsulphonic acid (DMSO) were used to make 1 and 5 µm concentrations in the media; the total concentration of DMSO was therefore kept at 0.05%. VP16 was diluted in media to 10 and 50 µm. Combined treatments with 10 µm VP16 + 5 µm gefitinib, 50 µm VP16 + 1 µm gefitinib, and 50 µm VP16 + 5 µm gefitinib were also used. Finally VP16-treated cells were incubated in the presence of 1 nm HB-EGF and compared to 50 µm VP16 + 5 µm gefitinib + 1 nm HB-EGF. All samples were treated with the same concentration of vehicle.
For Western blot analysis, cells were cultured and treated as described earlier, but cells were starved 24 h before treatment in serum-free medium. After treatment the cells were washed, scraped and handled as described previously . Cells were collected at 48 h and the protein concentration determined using the BCA protein kit (Pierce, Rockford, USA) following the manufacturer’s manual. Protein samples were boiled in SDS sample buffer (60 mm Tris, pH 6.8, 2% SDS, 10% glycerol, 5%β-mercaptoethanol, 0.1% bromophenol blue) for 2 min, 7.5% denaturing polyacrylamide gels were used and each lane was loaded with 35 µg protein; PAGE was run overnight. The separated proteins were transferred to an NMDF membrane (Millipore, Copenhagen, Denmark) and blocked for 2 h in 5% skimmed milk (Biorad, Herlev, Denmark). After blocking, the membranes were cut into suitable pieces and incubated overnight at 4 °C with the appropriate antibodies. The membranes were then washed with Tris-buffered saline-Tween (TBST, 25 mm Tris, pH 7.5, 150 mm NaCl, 0.05% w/v, Tween-20) for 20 min, repeated three times. Horseradish peroxidase-conjugated anti-rabbit or -mouse secondary antibodies (Dako, Glostrup, Denmark) were then added and incubated for 1 h at room temperature, followed by another series of washings with TBST. Finally membranes were submerged in ECL (Amersham, Buckinghamshire, UK) and immunoreactive bands were detected on X-ray film.
To detect apoptosis, the APOPercentageTM system was obtained from Biocolor (Dublin, Ireland); this assay is dye-based and takes advantage of the phosphatidylserine switch that takes places during apoptosis. When this occurs the dye is able to penetrate the cell membranes and is trapped inside the cell. Living and necrotic cells will not absorb the dye. The cells were seeded in flat-bottomed 96-well plates and 20 000 RT4 cells/well were seeded on day 1 in 100 µL gelatine 0.4%, to prevent cells from detaching during apoptosis, and 200 µL of medium. The day before treatment 20 000 T24 cells were seeded/well; on day 3 the cells were treated as described previously. RT4 cells were treated for 24 h and T24 cells for 48 h. According to the manufacturer’s manual, with minor modifications, the dye was added 2 h before analysis. RT4 cells were treated with the included dye-release agent and analysed using a calorimetric plate reader. RT4 samples were pooled to a total of 300 µL and read at 540 nm/690 nm. RT4 cells were analysed calorimetrically as they are slow-growing in small islands, and not in a single layer. However, T24 cells are fast-growing in a single layer, making them more suitable for manual counting by microscopy. T24 cells were photographed using a Eclipse TE300 mounted with a Coolpix 4500 Camera (Nikon, Japan); four pictures were taken from across each well at × 20, and the cells were later counted on a computer. Both experiments were done at least three times in quadruplets.
Analysis of caspase-dependent apoptosis was investigated in T24 cells using the pan-caspase inhibitor Z-VAD.fmk; 30 µm of Z-VAD.fmk was added to 50 µm VP16-treated cells and compared with VP16 treatment alone. Also, the combined treatment 50 µm VP16 + 5 µm gefitinib was compared with and without 30 µm Z-VAD.fmk. DMSO was kept at <0.25% in all experiments and showed no toxicity in control cells. The APOPercentage assay was used as described and the cells counted manually.
The plots are shown as the mean (sem) to reflect the precision of the in vitro results; results were compared using an unpaired t-test, with all P values two-tailed.
Treatment of T24 cells for 48 h with 50 µm VP16 resulted in pronounced activation of the EGFR (Fig. 1, upper panel), but this increased phosphorylation of the EGFR was completely inhibited when the cells were treated with 50 µm VP16 + 5 µm gefitinib. β-actin was used as a loading control (Fig. 1 lower panel), showing that equal amounts of protein were loaded.
T24 bladder cancer cells were treated with either 50 µm VP16 alone or combined 50 µm VP16 + 1 nm HB-EGF for 48 h. This showed that 50 µm VP16 induced apoptosis in the cells, but combined treatment with 50 µm VP16 + 1 nm HB-EGF reduced the fraction of cells entering apoptosis, giving a 25% reduction in the level of apoptosis compared with that induced by VP16 alone (P = 0.024; Fig. 2). This shows that stimulating bladder cells with exogenous ligand has a cell survival effect in the presence of VP16.
T24 cells were treated for 48 h with VP16 and gefitinib and the number of apoptotic cells counted manually, based on four photographs taken from across each well at the same resolution. Figure 3A is an example of the results, showing that there was no dye uptake in cells treated with control and gefitinib (5 µm) after 48 h. In cells treated with VP16 (50 µm) apoptosis was induced, and was significantly enhanced when gefitinib was combined with VP16. Figure 3B shows the results obtained with the T24 cells after treatment for 48 h, as the percentage of apoptotic cells relative to the total cell count; 10 µm VP16 generated 4.7% apoptosis, but adding 5 µm gefitinib significantly increased apoptosis to 9.8% (P = 0.005), and 50 µm VP16 resulted in 19% apoptosis, but adding gefitinib at 1 and 5 µm resulted in an increase to 36% (P = 0.001) and 45% (P < 0.0001), respectively. Strikingly, gefitinib at the highest concentration more than doubled the effect of VP16. In RT4 cells the number of apoptotic cells could not be directly counted due to the growth pattern of these cells, and apoptosis was instead monitored spectrophotometrically. The results were similar in the RT4 bladder cell line as in the T24 cells, with no apoptosis in control- and gefitinib- (1 µm or 5 µm) treated cells, whereas treatment with 50 µm VP16 resulted in a moderate level of apoptosis, which was enhanced by adding either 1 or 5 µm gefitinib (Fig. 4).
Using Z-VAD.fmk, the results clearly show that in T24 cells the apoptosis induced by VP16 is caspase-dependent, as there was virtually no apoptosis when adding 30 µm Z-VAD.fmk combined with 50 µm VP16 for 48 h (data not shown), or when the incubation time was prolonged to 72 h to account for a potentially delayed non-caspase-dependent apoptosis (Fig. 5). Interestingly, the additional apoptosis in the combined treatment with 50 µm VP16 and 5 µm gefitinib was also abolished by 30 µm Z-VAD.fmk, showing that the additional apoptosis induced is also caspase-dependent (Fig. 5).
Genotoxic drugs like VP16 are used in the treatment of bladder cancers. We show that VP16 treatment of bladder cancer cells results in increased activity of the EGFR. The importance of activating EGFR in modulating the effect of VP16 in T24 bladder cancer cells was further substantiated because the activation of EGFR on adding one of its ligands, HB-EGF, induced a significant cell-survival effect in VP16-treated cells. Interestingly, we showed that treating the cells with the EGFR-specific inhibitor gefitinib renders them more sensitive to VP16, suggesting that in bladder cancer cells the activity of the EGFR influences the effect of genotoxic drugs like VP16.
Drug resistance is a major obstacle in the treatment of locally or advanced bladder cancer. It is evident that part of the drug resistance in human cancer is due to up-regulation of certain proteins like P-glycoprotein and multidrug resistance protein-1, which can transport various compounds, including chemotherapeutic drugs, out of the cells . The present results are supported by a study showing an increase in apoptosis when other cytostatic drugs are used in combination with gefitinib in an ovarian and a colon cancer cell line . The molecular mechanism of this action of gefitinib is not known, but an inhibition of repair of the DNA strand breaks generated by the anticancer agents was suggested .
The activation of the EGFR by VP16 in the present study is corroborated by another study where the EGFR was also activated by cisplatin, another type of DNA-damaging drug . This effect of cisplatin was shown in glioma and breast cancer cell lines known to have a greater expression of the EGF system, and the cisplatin-induced signalling was dependent on the c-SRC kinase. Furthermore, cisplatin was able to activate the EGFR in a ligand-independent manner. Irrespective of whether the bladder cancer cells investigated here activate the EGFR by a ligand-dependent or -independent mechanism, or a combination of the two, we showed that the EGFR inhibitor gefitinib has the potential to inhibit the drug-induced activation of the receptor.
Recent discoveries show that apoptosis can be divided into two distinct forms, caspase-dependent and -independent [24,25]. It was argued that the phylogenetically older non-caspase-dependent system might not be as efficient as caspase-dependent apoptosis in killing cells . When the pan-caspase inhibitor Z-VAD.fmk was added to the present cell lines there was virtually no dye uptake either on VP16 treatment or when VP16 was combined with gefitinib. This shows that VP16 induces apoptosis through the caspase-dependent pathway and that the additional apoptosis when VP16 and gefitinib are combined is also caspase-dependent.
Novel findings show that uncoupling of the EGFR and downstream signalling molecules can cause gefitinib resistance in bladder cancer cells . A large proportion of bladder tumours harbour mutations in the RAS oncogene  which therefore might render these bladder tumours resistant to EGFR inhibition. The bladder cancer cell line T24 used in the present study contains a H-RAS mutation (data not shown) and others also reported this [29,30]. Despite this we showed that gefitinib has the potential to inhibit the cell-survival mechanism in T24 cells, suggesting that the RAS/MAPK pathway is not essential for the cell-survival effect of the EGFR, but that this effect can also be transferred through other signalling pathways.
In conclusion we suggest that the efficiency of genotoxic treatment could be improved by including EGFR inhibitors like gefitinib, which might prevent the cell-survival effect of the EGFR in bladder tumours.