Background
Lung cancer (LC) is a major cause of cancer deaths in the Western world. Based on the histo-pathological features, LC is divided into small cell lung carcinoma (SCLC), and non-small cell lung carcinoma (NSCLC), which account for 25 and 75% of bronchogenic carcinomas, respectively. In contrast to NSCLC, SCLC is characterized by relatively high sensitivity to treatment with anticancer drugs and radiation. However, despite the initial responsiveness, relapses occur in most cases, accompanied by the fast development of severe resistance to treatments during the course of disease. SCLC represents a highly malignant and particularly aggressive form of cancer, with early and widespread metastases, and poor prognosis. Mechanisms responsible for the intrinsic and acquired resistance to treatment involve the defects/dysregulations of the apoptotic programme [
1,
2]. The avoidance of apoptosis is considered as one of the hallmarks of cancer cells, and represents a significant clinical problem. Therefore, elucidation of the mechanisms and molecules responsible for the resistance is essential for proper targeting of anticancer therapy.
The tumour necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL), a member of TNF family, is particularly interesting because of its unique ability to induce cancer cell death while sparing the most of normal cells. This implies its potential promise as an anti-cancer agent [
3]. TRAIL can interact with different receptors. Only two of them, namely, death receptors (DR) contain apoptosis-related death domain (DD): DR4 (TRAIL-R1) and DR5 (TRAIL-R2). Decoy receptors DcR1 (TRAIL-R3) and DcR2 (TRAIL-R4) either lack or posses truncated DD and are, therefore, not able to transmit apoptotic signal. Osteoprotegerin (OPG, TRAIL-R5) is a soluble receptor with the lowest afinity to TRAIL [
4]. TRAIL binding to DR4 and DR5 results in triggering of the extrinsic pathway, initiated by formation of the death-inducing signalling complex (DISC) consisting of Fas-associated DD protein (FADD) and pro-caspase-8. Activation of caspase-8 at the DISC level plays a crucial role in the DR-mediated pathway, and can be efficiently regulated by its competitive inhibitor cFLIP (FLICE-like inhibitory protein). Caspase-8 activation is followed by cleavage of effector caspases and apoptosis execution (characteristic for type I cells). In some cases, caspase-8 can also cleave Bid, which is responsible for translocation of apoptotic signal to mitochondria. Subsequent amplification of the death signal at the level of these organelles is essential in so-called type II cells [
5,
6].
TRAIL triggers apoptosis in a broad spectrum of cancer cell lines
in vitro and
in vivo[
7,
8]. However, failure to undergo apoptosis in response to TRAIL has been demonstrated in majority of SCLC cells [
9,
10]. Significant perturbances of apoptosis programme such as downregulation/absence of some proapoptotic proteins and/or overexpression of anti-apoptotic proteins have been shown to be a characteristic feature of SCLC cells [
11]. The higher rates of loss of expression of caspase-8, caspase-10, DR4, DR5, Fas, and FasL have been found in SCLC compared to NSCLC cells [
9,
12]. A relationship between the inactivation of some DISC components and Myc oncogene amplification, which is a common event in SCLC, has also been reported [
9].
Majority of chemotherapeutic agents are typical activators of mitochondria-mediated (intrinsic) apoptotic pathway, where release of cytochrome
c from mitochondrial intermembrane space is followed by formation of apoptosome complex (cytochrome
c, Apaf-1, dATP, pro-caspase-9), activation of initiator caspase-9 and downstream effector caspases. The described events can be effectively modulated by pro-apoptotic (
e.g. Bid, Bax, Bak) and/or anti-apoptotic (
e.g. Bcl-2, Mcl-1, Bcl-X
L) members of Bcl-2 family [
13]. Caspase-2 has been shown as an important link between DNA damage and the engagement of the mitochondrial pathway [
14].
The clinically relevant concentrations of chemotherapeutic drugs might restore the apoptotic response to TRAIL in various cancer cells through different mechanisms and, therefore, sensitize these cells to TRAIL treatment. Among them, upregulation of the DRs, facilitation of DISC formation, downregulation of anti-apoptotic proteins, enhancement of activation of mitochondrial pathway and caspase cascade are particularly interesting [
15,
16]. However, respective data concerning SCLC treatments are still missing. The usual therapeutic regimes used for this type of cancer include
e.g. doxorubicin, etoposide or cisplatin. As TRAIL monotherapy has been shown to be ineffective in SCLC, in present study we explored the potential use of combination of TRAIL and doxorubicin or etoposide in order to provide a tool for triggering apoptosis in resistant cancer cells.
Materials and methods
Cell culture and treatments
Human SCLC cell lines - H69 (ECACC), H82 (ATCC), U1285, U1690, and U1906 [
17] were grown in RPMI 1640 medium supplemented with 10% of heat-inactivated fetal bovine serum (FBS), glutamine (2 mM), penicillin (100 U/ml) and streptomycin (100 μg/ml) (all from Gibco) in 37°C, 5% CO
2 and 95% humidity. Twenty-four hours after seeding, cells were treated with human recombinant Killer TRAIL (provided by Dr. L. Andera), doxorubicin (Sigma), etoposide (Vepesid, Bristol-Myers), PKC412 (CGP41251, Novartis), cycloheximide (Sigma), pan-caspase inhibitor z-VAD-fmk (Enzyme Systems Products) or vehicle (DMSO) for the time points and concentrations indicated in figure legends.
Surface TRAIL receptor expression
The level of surface TRAIL receptors was detected in cells after incubation with specific antibodies using flow cytometry [
18]. Briefly, after washing with PBS + 5% FBS, cells were incubated (20 minutes, 4°C) with primary antibody (anti-TRAIL-R1 to -4 flow cytometry set, Axorra; HS101, HS201, HS301, HS402 antibodies, respectively). Cells were then washed twice, and incubated (20 min, 4°C) with secondary antibody (FITC-conjugated goat anti-mouse IgG1, ALX-211-200, Axorra or AlexaFluor-488-conjugated donkey anti-mouse-IgG, A21202, Molecular Probes). After washing twice, the cells were stained (15 min, 4°C) with 7-AAD (1 μg/ml, Molecular Probes) and analyzed using flow cytometry (FACScan, Becton Dickinson). The 7-AAD negative cells were subjected to receptor analysis (Cell Quest software). Results are expressed as histograms (green fluorescence indicating the amount of the receptor present at the cell surface versus cell counts), and related to appropriate control lacking the specific primary antibody.
Analysis of Bax activation
Cells were washed in PBS, fixed in 1% paraformaldehyde (10 minutes, 4°C), washed again, and incubated (30 min, 4°C) with mouse anti-Bax antibody (BD556467, Becton Dickinson) diluted in PBS containing 1% BSA and 0.1% saponin. After additional washing, cells were incubated (30 min, 4°C) with AlexaFluor-488-conjugated donkey anti-mouse IgG secondary antibody (A21202, Molecular Probes) diluted in PBS + 1% BSA + 0.1% saponin, washed twice, and resuspended in PBS. For each sample, controls lacking the specific primary antibody were prepared. Analysis was performed using flow cytometer (FACScan, Becton Dickinson) and CellQuest software. Ten thousand cells per sample were analysed and the results were expressed as percentage of cells with active Bax.
Analysis of mitochondrial membrane potential (MMP)
After washing in PBS, cells were incubated (20 minutes) in HBSS with 25 nM of tetramethylrhodamine ethyl ester perchlorate (TMRE, Molecular Probes), washed twice with HBSS, and analysed (2 × 104 cells per sample) by flow cytometry (FACScan, Becton Dickinson). Forward and side scatters were used to gate the viable cell population. The data were evaluated using Cell Quest software as a percentage of the cells with decreased MMP.
Caspase activity assay
Cells were lysed in appropriate lysis buffer, incubated with caspase-3, caspase-8 or caspase-2 substrates and analysed as described previously [
19]. The values of caspase activity were related to the total protein amount. Caspase activity was expressed as a fold-increase to appropriate control.
Immunoblotting
Cells were lysed in Complete lysis buffer (Roche) with protease inhibitor cocktail (PIC, Complete-M, Roche). The protein concentration was determined in samples (BCA protein assay, Pierce), which were mixed with Laemmli buffer and subjected to SDS-PAGE and western blotting [
19]. For immunodetection, following antibodies were used: anti-cleaved PARP (CS9546), anti-cleaved lamin A (CS2036), anti-cleaved caspase-3 (CS9661), anti-Bid (CS2002) (Cell Signaling Technology), anti-DR4 (D3813), anti-DR5 (D3938), anti-Mcl-1 (M8434) (Sigma), anti-caspase-10 (M059-3, MBL), anti-FLIP (NF6, 804-428, Axorra), anti-caspase-2 (BD611022), anti-FADD (BD556402), anti-cytochrome
c (BD556433), and anti-Bax (BD556467) (Becton Dickinson), anti-survivin (ab469, Abcam), anti-caspase-8 (provided by Prof. P. Krammer), anti-Bcl-2 (sc492), anti-Bcl-X
L (sc-634) (Santa Cruz Biotechnology). The recognized proteins were detected using horseradish peroxidase-labelled secondary antibodies: anti-mouse IgG (31430, Pierce), anti-rabbit IgG (31460, Pierce), and enhanced chemiluminescence kit (Western blot detection reagent, GE Healthcare UK Limited). An equal loading was verified using anti-β-actin (A2066, Sigma), anti-G3PDH (2275, Trevingen) or anti-TOM40 (sc-11414, Santa Cruz Biotechnology, for mitochondrial fraction) antibodies. For detection of cytosolic and mitochondrial cytochrome
c and Bax, the cells were washed twice with PBS, incubated (5 min, 4°C) in buffer (250 mM sucrose, 70 mM KCl, 100 μg/ml digitonin in PBS), centrifuged (5 min, 7000 g), and the supernatant was collected (cytosolic fraction). Mitochondrial fraction was prepared by lysis of the pellet in Complete Lysis-M buffer with PIC (Roche).
cFLIP and caspase-8 overexpression
Twenty four hours after seeding, cells were transfected with DNA:Lipofectamine 2000 (Invitrogen) mixture at ratio 2 μg:1 μl according to manufacturer's instructions. Following next 24 h, medium was exchanged and the cells were treated (8 or 24 h) with TRAIL (100 ng/ml), doxorubicin (1 μM) or their combination. The following plasmids were used: pcDNA3 (Invitrogen), pcDNA3-Flag-cFLIPS (provided by Prof. P. Krammer and Dr I. Lavrik), pcDNA-MACH alpha1 (provided by Prof. D. Wallach), pMSCV-IRES-GFP, and cFLIPL-pMSCV-IRES-GFP (provided by Dr A. Grandien).
cFLIP and caspase-8 siRNA experiments
Three different specific siRNAs were used to downregulate (a) both long and short cFLIP, (b) the long form, or (c) the short form only. The non-targeting siRNA was used as control. All the siRNAs were synthesized by Qiagen according to Galligan et al. [
20]. The siRNA was diluted in 100 μl of Opti-MEM I medium (Gibco), and 1 μl of Lipofectamine 2000 reagent (Invitrogen) was added to other 100 μl of OptiMEM I medium. Diluted siRNA and Lipofectamine 2000 were then mixed and incubated for 20 min. Transfection complexes were added (200 μl per well) to cells 24 h after seeding in 12 well plates. The final concentration of siRNA was 100 nM. After 24 h, the medium was exchanged, and cells were treated (24 h) with TRAIL (100 ng/ml). For caspase-8 siRNA transfection, the same experimental protocol was used, and the cells were treated with TRAIL and/or doxorubicin 48 h after transfection. Caspase-8 siRNA (L-003466-00) and control siRNA (D-001810-10) were obtained from Dharmacon.
MTS assay
Cells were seeded in 96-well plates at a density of 104 cells per well and 24 h later treated with various concentrations of the drugs. After next 24 h, CellTiter 96 AQueous MTS Reagent Solution (Promega) was added to each well according to the manufacturer's instructions, and in 1 h the cell viability was determined by measuring the absorbance at 490 nm using Labsystems Multiscan MCC/340 plate reader.
Analysis of nuclear morphology
Cells were washed in PBS, fixed in ethanol (70%), stained (30 min) with 4,6-diamidino-2-phenyl-indole (DAPI, Fluka, 1 μg DAPI/ml ethanol) and mounted in Mowiol. The percentage of apoptotic cells (with chromatin condensation and fragmentation) was calculated from a total number of 200 cells using LSM 510 META spectral laser scanner microscope (Zeiss).
Analysis of DNA content
For DNA analysis, cells were fixed in 70% cold ethanol overnight, washed, incubated 1 h at 37°C with RNAse A (100 μg/ml), and exposed to PI (50 μg/ml) in phosphate-buffered saline (pH 7.4) for 30 min. The cells were analyzed by flow cytometry (FACScan, Becton Dickinson) and CellQuest software. The cell death was monitored by evaluation of percentage of cells with subdiploid amount of DNA.
Statistical evaluation
The results of three independent experiments were expressed as the means ± S.E.M. Statistical significance (p < 0.05) was determined by one-way ANOVA followed by Tukey test or by non-parametric Mann-Whitney test.
Discussion
SCLC is a tumour entity where TRAIL monotherapy is not efficient. The loss of some DISC components, associated with inactivation of DR pathway make sensitization of the SCLC cells to TRAIL very difficult [
9,
10,
12,
21]. Caspase-8 is frequently silenced in SCLC and other tumours of neuroendocrine origin, usually by aberrant promoter methylation [
22,
23]. According to our results, 3 out of 5 studied SCLC cell lines were deficient for caspase-8, all lacked caspase-10, and only 1 cell line expressed surface DR4. Previously, restoration of caspase-8 expression by demethylation or gene transfer was shown to sensitize neuroblastoma cells to DR-mediated apoptosis [
23]. It has also been suggested that the resistance of SCLC cells lacking caspase-8 to TRAIL might be partially eliminated by combination of demethylation agents and treatment with IFNγ [
10]. However additional studies should be performed in order to elucidate how general this phenomenon is.
Notably, we and others found that caspase-8 expressing SCLC cells are also resistant to TRAIL-induced apoptosis. Therefore, we aimed to investigate whether the cytotoxic effects of TRAIL in SCLC could be restored by doxorubicin or etoposide, the conventionally used drugs for treatment of SCLC, and what are the mechanisms responsible for apoptosis resistance of caspase-8 expressing SCLC cells. Our results demonstrated that combination of these drugs with TRAIL might significantly improve efficiency to kill this type of SCLC cells as compared with drugs used alone.
The presence of TRAIL death receptors DR4 and/or DR5 at the cell surface is not always sufficient to trigger apoptosis and that might be due to an existence of the defects of intracellular signaling pathways or high expression of inhibitory proteins. Upregulation of DRs following treatment with chemotherapeutic agents has been demonstrated in some cancer cells, while no effects were apparent in others [
24,
25]. This upregulation has been shown to be responsible for the synergistic action of TRAIL and chemotherapy in several cell lines [
26‐
28]. In our experiments, untreated U1690 cells expressed a relatively low level of DR5 and no DR4 at the surface. Significant increase of surface DR5, but not DR4 was observed after treatment with doxorubicin, and no significant changes of DcRs level (not shown) as has been demonstrated by others [
29]. Our results, therefore, suggest that DR5 plays an exclusive role in mediation of signals triggered by TRAIL in these cells. Furthermore, relocalisation and clustering of DR5 within ceramide-enriched membrane platforms has been reported to affect TRAIL-induced apoptosis in cells treated with doxorubicin [
30]. Similar changes at the level of plasma membrane might be also involved in enhancement of TRAIL apoptotic signaling in SCLC cells.
Caspase-8 activation at the DISC proceeds through two subsequent cleavage steps that can be effectively regulated by cFLIP proteins. While cFLIP
L allows the first cleavage of pro-caspase-8, cFLIP
S can completely inhibit both of them [
31]. cFLIP is known to play an important role in regulation of TRAIL- and chemotherapy-induced apoptosis [
20,
32,
33]. Doxorubicin-induced downregulation of cFLIP
S contributed to sensitization of prostate cancer cells to apoptotic effects of TRAIL [
34,
35]. Significant decrease of cFLIP
L/S level and appearance of specific p43-cFLIP
L fragment following doxorubicin or TRAIL and doxorubicin treatment indicated their involvement in modulation of apoptotic response of U1690 cells to TRAIL. The role of p43-cFLIP
L in regulation of caspase-8 activation is still controversial, as both anti- and pro-apoptotic functions of this fragment have been demonstrated [
36,
37]. CHX-mediated downregulation of cFLIP level previously published also in some other cancer cell types [
38,
39], resulted in sensitization of U1690 cells to TRAIL-induced apoptosis, showing that resistance of these cells may be at least partially induced by newly synthesized proteins acting upstream of caspase-8. The siRNA-mediated downregulation of cFLIP
L/S sensitized U1690 cells to TRAIL-induced apoptosis, although to a significantly lesser extent than combination of TRAIL with doxorubicin or etoposide. Moreover, overexpression of cFLIP
L or cFLIP
S only partially protected the cells from apoptosis induced by this combinatory treatment. Recently, it has also been shown that silencing of cFLIP induced caspase-8 activation via increased co-localisation of DR5 and ceramide, and significantly enhanced TRAIL-induced apoptosis [
40]. Therefore, our results suggest that elimination of cFLIP itself as a single factor may not be sufficient to fully restore the sensitivity to TRAIL-induced apoptosis, and modulation(s) at the level of other molecule(s) such as DR5 need to be involved.
Various chemotherapeutic drugs are known to trigger intrinsic apoptotic pathway [
41] and/or increase a susceptibility of mitochondria to apoptotic signals translocated from DRs [
25,
42,
43]. In our experiments we detected an increase of caspase-2 activity that was previously shown as a critical component of DNA damage-induced apoptotic cascade, being activated upstream of mitochondria [
44]. Role of caspase-2 as initiator caspase was also observed in apoptosis triggered by other chemotherapeutic drugs [
14,
45]. In addition to acting as an initiator caspase primarily activated in DR-mediated apoptosis at the DISC level, caspase-8 has also been reported to be activated by other caspases downstream of mitochondria during drug-induced apoptosis [
46‐
48]. This led us to investigate the role of mitochondria and the ordering of caspase activation. Since z-VAD-fmk efficiently blocked TRAIL and doxorubicin-induced cleavage of caspase-2, -3, PARP, lamin A, apoptotic nuclear morphology changes, but not generation of caspase-8 p41/43 kDa fragment, we suggested that the first caspase-8 cleavage step might occur prior to and independently on the activation of other caspases. Moreover, our results demonstrated that siRNA-mediated downregulation of caspase-8 resulted in marked decrease of TRAIL and doxorubicin-induced processing of caspase-2 and -3, and overall apoptosis, while re-expression of caspase-8 fully reverted the apoptotic phenotype. Furthermore, a rapid processing of caspase-8 was a relatively early event in the course of TRAIL and doxorubicin-induced apoptosis compared to attenuated kinetics of mitochondrial events, such as cytochrome
c release. Therefore, we assume that chemotherapy-mediated modulation of events leading to caspase-8 activation following TRAIL treatment occurs upstream of mitochondria and effector caspases.
As TRAIL by itself did not induce caspase-8 processing, it is likely that the resistance to TRAIL-induced apoptosis is regulated by protein(s) acting upstream of this molecule. Here we demonstrated that doxorubicin was efficient in sensitizing the cells to apoptotic effects of TRAIL that was associated with significant increase of surface and total DR5 level, activation of caspase cascade through processing of caspase-8, specific cleavage of cFLIPL, and decrease of cFLIPS. Cellular apoptosis was accompanied by cleavage of Bid, Bax activation, decrease of MMP, cytochrome c release, decrease of survivin level, and effector caspase activation. Based on the caspase-8 siRNA approach and experiments with z-VAD-fmk we concluded that combination of TRAIL and doxorubicin facilitates caspase-8 processing primarily at the DISC level rather than being secondary result of activation of mitochondrial pathway and/or effector caspases. We suggest the possibility of sensitization of SCLC cells to TRAIL by modulation of the apical events in TRAIL apoptotic signaling at the level of surface DR5 and intracellular inhibitory molecules such as cFLIP.
In summary, although TRAIL monotherapy is completely inefficient in SCLC cells due to defects in initial steps of the DR-mediated pathway, combined treatment with TRAIL and conventional chemotherapeutic drugs, such as doxorubicin and etoposide, might be a promising therapeutic strategy for SCLC expressing caspase-8. Our results showed that doxorubicin and etoposide significantly sensitized these cells to TRAIL-induced apoptosis by modulation of events that facilitate activation of caspase-8. Our study highlights the potential applicability of this combination in chemotherapy of SCLC.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AV and VK equally contributed to this work. They were involved in planning and performing experiments, carried out analysis of data and participated in writing of the MS. EJ and OS carried out experiments. BZ participated in the design of the experiments and writing the manuscript. All authors read and approved the final manuscript.