Cancer Cell Biology
α6 integrin subunit mediates laminin enhancement of cisplatin-induced apoptosis in testicular tumor germ cells
Article first published online: 4 MAY 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 117, Issue 1, pages 68–81, 20 October 2005
How to Cite
Andjilani, M., Droz, J.-P., Benahmed, M. and Tabone, E. (2005), α6 integrin subunit mediates laminin enhancement of cisplatin-induced apoptosis in testicular tumor germ cells. Int. J. Cancer, 117: 68–81. doi: 10.1002/ijc.21144
- Issue published online: 2 AUG 2005
- Article first published online: 4 MAY 2005
- Manuscript Accepted: 22 FEB 2005
- Manuscript Received: 7 SEP 2004
- Institut National de la Santé et de la Recherche Médicale
- Ligue Nationale Contre le Cancer, Comité Départemental de l'Ain, France
- testicular tumors;
Our study demonstrates that laminin potentiates cisplatin-induced apoptosis in NCCIT, a testicular tumor germ cell line. When cultured on laminin, NCCIT cells displayed a significantly higher susceptibility to cisplatin-induced apoptosis than on plastic or on other ECM components including fibronectin, Type IV collagen and vitronectin. This high cisplatin sensitivity observed on NCCIT cell cultured on laminin was mediated by the α6-integrin signaling. The knockdown of the α6-integrin subunit by small interfering RNAs suppressed the higher cisplatin-sensitivity supporting the existence of a crosstalk between laminin-α6-integrin signaling and cisplatin-induced apoptosis. Our findings indicate that in cisplatin-treated NCCIT cells, the laminin-α6-integrin signaling induces the activation of executioner procaspase-3 and -6 as well as apoptosis-inducing factor (AIF) transcription and expression. The ability of integrin-mediated specific stroma–tumor cell interactions to modulate the chemosensitive phenotype of a tumor cell might provide new insights to overcome cisplatin resistance of tumor cells. © 2005 Wiley-Liss, Inc.
Resistance to anti-tumor drugs represents a major cause of cancer death. Platinum-based combination therapy is a standard chemotherapy in many cancers. Cisplatin, a major cytotoxic drug is used for treatment of a number of malignancies including ovarian, lung, head and neck, bladder and testicular cancer. Unfortunately, some tumors are intrinsically resistant or acquire resistance during the treatment.1 Testicular germ cell tumors (TGCT) are curable even in an advanced stage, by cisplatin-based chemotherapy, with an overall cure rate of approximately 80%2, 3 but in a few cases, however, a resistance occurs. The mechanism of this high sensitivity to cytotoxic drugs is not yet elucidated. It is now well established that cisplatin exerts its cytotoxic action through induction of apoptosis, at least in TGCT cells.4, 5, 6 Cisplatin forms DNA intrastrand and interstrand cross-links that trigger cell cycle arrest and apoptosis in cancer cells.7 The molecular mechanism underlying cisplatin resistance is multifactorial but increased DNA repair, tolerance to DNA adducts and inhibition of the apoptotic pathways are among the primary causes of tumor resistance.8, 9
Apoptosis is a tightly controlled multi-step mechanism of programmed cell death. Two well-characterized apoptotic pathways have been identified. The first one is mediated by death receptors such as Fas or tumor necrosis factor (TNF) receptor and a second one requires the release of apoptogenic factors from mitochondria. In both apoptotic pathways, the activation of initiator caspases (caspase-8 and -9) leads to the proteolytic activation of downstream effector caspases (caspase-3, -6 and -7) that cleave specific substrates such as nuclear lamins, cytoskeletal proteins and DNA ensuring irreversible cell death.10 Recently the apoptosis-inducing factor (AIF), a mitochondrial intermembrane flavoprotein, has been found to induce apoptosis in a caspase-independent manner.11 AIF is early translocated directly from mitochondria to nucleus causing large-scale DNA fragmentation and a peripheral chromatin condensation that resembles the first stage of nuclear apoptosis. A large range of pro- and anti-apoptotic factors regulates these apoptotic independent pathways. The involvement of each factor depends on the cell type and on the apoptotic stimuli. The factors that prevent tumor cells from undergoing drug-induced apoptosis and thus help tumor to acquire drug resistance are at present largely unknown. The concept that the cell microenvironment and particularly the extracellular matrix (ECM) could play a critical role in the regulation of apoptosis in many cell types is now well accepted.12
The interaction of the cells with the ECM is mediated by cell surface receptors belonging to the integrin super family. Integrin receptors are heterodimers composed of α and β subunits that recognize and bind to components of the extracellular matrix. In addition to providing a link between the ECM and actin cytoskeleton, integrin receptors serve as signaling receptors that transduce informations from the ECM that affect cell behavior and gene expression.13, 14 Contradictory results were reported concerning the regulation of apoptosis by the ECM-integrin signaling pathway. In most cases, it has been shown that the tumor microenvironment, and in particular the ECM, provides protection against chemotherapeutic (including cisplatin) drug-induced apoptosis in many cell types.15, 16, 17, 18, 19, 20 This phenomenon is referred to as cell adhesion-mediated drug resistance or CAM-DR. In addition, recent studies pointed out that tumor cells can actively remodel their environment to create favorable ECM-integrin signals leading to decreased apoptosis.21 In other cases, however, ECM signals not only fail to protect cells from apoptosis but promote apoptosis.22, 23, 24 The molecular events that regulate ECM-mediated cell survival or cell death pathways are not well determined. The aim of our study is to provide evidence that an extracellular matrix component linked to its integrin receptor, is able to modulate the cisplatin-induced apoptosis phenotype in NCCIT, a human TGCT cell line, which is known to be moderately sensitive to cisplatin.25
We examined the apoptosis executors' expression profile when the cisplatin-treated NCCIT cells were cultured on various ECM proteins without any gene modification. Our results gave strong evidences that when NCCIT cells adhered on laminin via α6-integrin, they highly improved their sensitivity to cisplatin-induced apoptosis. This phenomenon was associated with a sequential activation of mitochondrial apoptotic effector AIF, caspase-3 and -6, 3 main components in the cell death machinery.
Material and methods
NCCIT is a human testicular embryonic carcinoma cell line obtained from the American Type Culture Collection. These TGCT (testicular germ cell tumor) adherent cells were grown in RPMI 1640 medium (Invitrogen Corp., Groningen, NL) supplemented with 15% FBS (SVF, Life Technology, Germany). They were maintained at 37°C in 5% CO2 atmosphere.
Coating and cytotoxicity assay
Plates coated with ECM protein or adhesion molecules were done by incubating 12-well plastic tissue culture plates with 8 μg/cm2 laminin-1 (formerly EHS laminin) (Sigma-Aldrich, St. Quentin Fallavier, France); 5 μg/cm2 fibronectin (Sigma-Aldrich); 7 μg/cm2 Type IV collagen (Sigma-Aldrich); 0.2 μg/cm2 vitronectin (Life Technology); 10 mg/cm2 heat-denatured BSA (Sigma-Aldrich) or 5 μg/cm2 poly-L-Lysine (Sigma Diagnostics Inc., St. Louis, MO) for 2 hr at 37°C. Plates were then rinsed with PBS 1× and blocked with heat-denatured BSA (2 mg/ml) for 1 hr. Plates were then twice rinsed with PBS 1× and cells (2 × 105 per well) were seeded onto these plates and incubated at 37°C in 5% CO2.
Cells from exponential culture phase were treated with 10 μM cisplatin [cis-platinum (II) diamine dichloride, DDP] (Merck, Fontenay-sous-bois, France) for 4 hr, washed with fresh culture medium and incubated in cisplatin-free medium. After 48 hr incubation, attached and floating cells were collected, pooled and suspended in PBS 1× and then cytocentrifugated (300g, 5 min) on adhesive slides. They were immediately fixed and apoptosis detection assay was carried out.
Kinetic studies in cisplatin-treated NCCIT cells
NCCIT cells grown on laminin, BSA or fibronectin until exponential culture phase (Control) were treated with cisplatin as indicated above. At 0, 12, 24 and 48 hr after the end of the treatment, cells were collected and total RNA and proteins were extracted for RT-PCR and Western blot analysis. Cells were also cytocentrifugated on adhesive slides for immunochemistry or in situ TUNEL analysis.
Cells proliferation was assessed using the tetrazolium salt XTT (sodium 3′-[1-[(phenylamino)-carbonyl]-3,4-tetrazolium]-bis(4-methoxy-6-nitro)benzenesulfonic acid hydrate). General procedure was carried out as described previously.26 Briefly, 2 × 104 cells were cultured in triplicate in 96-well flat-bottomed plates, which have been coated with laminin, BSA or fibronectin as described above. Treated and untreated cells were incubated in the mixture of XTT (0.2 mg/ml) and 25 μM PMS (Sigma, St. Quentin-Fallavier, France) at 37°C, 5% CO2 for 7–8 hr. The plates were then analyzed on a microplate reader (Dynex MRX, De Chantilli, VA) at 450 nm wavelengths.
In situ TUNEL
Cells were fixed with paraformaldehyde solution (4% in PBS, pH = 7.4) for 30 min at room temperature. Slides were rinsed with PBS 1× and incubated in permeabilization solution (0.1% Triton X-100, 0.1% sodium citrate) at 4°C for 2 min. After 3 rinses in PBS 1×, the cells were treated with a mixture of terminal deoxynucleotidyl transferase (TdT) and biotin-16-ddUTP prepared in TdT buffer, for 60 min at 37°C in a humidified chamber. After washing, a streptavidin-biotin-peroxidase complex was applied for 15 min at R.T. and peroxidase was revealed with 3,3′-diaminobenzidine (DAB) as chromogen and H2O2 as substrate. A brief nuclear counterstaining was carried out with Harris hematoxylin. Positive apoptotic cells exhibited a strong brown nuclear staining as observed by light microscopy.
Caspase inhibitors assay
Synthetic peptide inhibitors DEVD-fmk (caspase-3 inhibitor), VEID-fmk (caspase-6 inhibitor) and VAD-fmk, a general pan caspase inhibitor were used at the final concentration of 80 μM. All inhibitors were purchased from Chemicon (Chemicon, Temecula, CA). NCCIT cells cultured on laminin until exponential culture phase were treated with cisplatin (10 μM) for 4 hr in the presence of caspase inhibitor. Cells were washed as indicated above and incubated in cisplatin-free medium containing caspase inhibitor for 48 hr. Apoptotic cells were determined by in situ TUNEL analysis as indicated above.
Integrin expression was determined by immunohistochemistry using mouse monoclonal antibodies. Anti-α1 and anti-β1 (Chemicon, Temecula, CA) were 1:500 diluted whereas anti-α2, anti-α3, anti-α4, anti-α5 and anti-α6 (Dako S.A., Trappes, France) were 1:50 diluted. Cells in exponential culture phase were harvested and cytocentrifugated. Fixation and permeabilization were carried out by addition of methanol/PBS (6/1) solution precooled at −20°C for 24 hr. Slides were washed in PBS 1×, endogenous peroxidases were quenched in 3% H2O2 for 15 min. The primary antibody was incubated for 90 min at room temperature. Immunostaining was carried out using Dako LSAB2 system as recommended by the manufacturer. For activated caspase-3, -6 and -7 and laminin immunochemistry analysis, cells were fixed with paraformaldehyde solution (4% in PBS, pH = 7.4) for 30 min at room temperature. Slides were rinsed with PBS 1× and incubated in permeabilization solution (0.1% Triton X-100, 0.1% sodium citrate) at 4°C for 2 min. Slides were washed in PBS 1× and endogenous peroxidases were quenched in 3% H2O2 for 15 min. They were then incubated for 90 min at room temperature in laminin monoclonal antibody (Sigma) diluted at 1:50 and in caspase-3, -6 and -9 primary antibodies (Cell Signaling, Beverly, MA) respectively diluted at 1:600, 1:200 and 1:200 in antibody diluent (Dako, Trappes, France). Immunostaining was carried out using iVIEW DAB Detection Kit (Ventana Med. Sys., Tucson, AZ) according to the manufacturer instructions. Cells were rapidly counterstained with Harris hematoxylin and mounted.
Paraffin sections of Bouin-fixed tumor were cut onto silanized slides. The samples were deparaffinized, rehydrated, incubated in antigen retrieval solution (pH = 6 citrate buffer) and heat at 95°C for 50 min. They remained in the hot solution at room temperature for 20 min. The primary antibodies raised against the α6-integrin subunit (Santa Cruz Biotechnology, Santa Cruz, CA) and laminin (Sigma) were diluted at 1:50 in Dako antibody diluent. Immunostaining was carried out using iVIEW DAB Detection Kit (Ventana Med. Sys., Tucson, AZ) according to the manufacturer instructions.
Total RNA extraction
Cells were harvested, washed in cold PBS and lysed immediately in Trizol reagent, a monophasic solution of phenol and guanidine isothiocyanate (Invitrogen, Groningen, NL). This reagent is an improvement over the single-step RNA isolation method developed by Chomczynski and Sacchi.27 The quality of extracted total RNA was checked by spectrophotometry (ratio = A260/A280) and 1% agarose gel electrophoresis. The amount of RNA was estimated by spectrophotometer at 260 nm.
Semi-quantitative co-amplified RT-PCR
cDNAs were generated from reverse transcription (RT) of 2 μg of total RNA using random hexanucleotides as primers (50 μM) in presence of dNTP 250 μM; (Invitrogen), dithiothreitol (DTT, 10 μM) and Moloney murine leukemia virus (MMLV) reverse transcriptase (10 U/μl) for 1 hr at 37°C.
cDNAs were co-amplified (AIF with HPRT, caspase-3, -6 and -7 with β2 microglobulin) by PCR with Taq polymerase (0.05 U/μl), deoxynucleotide (50 μM), 0.75 μCi (α33P) deoxy-ATP, specific primers (1 μM) and HPRT or β2 microglobulin primer (0.4 or 0.2 μM). The mixture was heated at 95°C for 5 min and then × cycles of 95°C for 30 sec, Tm (melting temperature) for 30 sec, 72°C for 30 sec, then 72°C for 5 min (Table I). PCR products were analyzed on an 8% polyacrylamide gel. Dried gels were exposed to Cyclone™ storage phosphor screen for 30 min. Intensities of autoradiographic bands were estimated by densitometric scanning using OptiQuant software (Packard Instrument Company, Meriden, USA). The data were expressed as AIF or caspases: HPRT or β2 microglobulin mRNA ratios. PCR amplified product were checked by direct sequencing (ABI Prism; 310 Genetic Analyzer; Applied Biosystem, Courtaboeuf, France).
|AIF||R: 5′-gATgCTgCATgCTTCTACgA-3′||54°C||203 bp||24|
|Caspase-3||R: 5′-ggCAggCCTgAATAATgAAA-3′||60°C||200 bp||30|
|Caspase-6||R: 5′-gTCAggCTggTCTCgAACTC-3′||62°C||245 bp||32|
|Caspase-7||R: 5′-TggAAgTgTgggTCATCagA-3′||62°C||201 bp||29|
|β2 μglobulin||R: 5′-CCATgATgCTgCTTACATgTCTCTgATCC-3′||60 or 62°C||374 bp||29–32|
|HPRT||R: 5′-CCTgCTggATTACATTAAAgCACTg-3′||54°C||354 bp||24|
Western blot analysis
NCCIT cells were harvested using Trypsin-EDTA (Invitrogen, Groningen, NL) and pelleted by centrifugation at 300g for 5 min. Whole cell protein extracts were prepared by addition of cold lysis buffer on pelleted cells. Lysis buffer consist of 50 mM Tris (pH = 7.4), 250 mM NaCl, 5 mM EDTA and 50 mM NaF supplemented immediately before use with a cocktail of protease inhibitor (Sigma, St. Quentin-Fallavier, France). The protein concentration of cells lysates was determined using a colorimetric Bradford method.
Proteins (30–60 μg) were resolved by 12% SDS-PAGE and electro-blotted onto nitrocellulose membranes in transfer buffer (25 mM Tris, 185 mM glycine containing 20% methanol). The transfer was carried out at a constant voltage of 100 V for 1 hr. The membrane was blocked by soaking in PBS, 0.05% Tween 20 and 5% nonfat dried milk for 1 hr at room temperature. The membranes were then rinsed 3 times with PBS/0.05% Tween 20 and incubated overnight at 4°C, with anti pro-AIF (Santa Cruz Biotechnologies, Santa Cruz, CA), AIF (Sigma, St. Quentin-Fallavier, France), caspase-2 (Cell Signaling Technology), caspase-3, caspase-6 (Santa Cruz Biotechnology) and cleaved-caspase-7 (Biovision, Mountain View, CA), caspase-8 (Cell Signaling Technology) and cleaved-caspase-9 (Cell Signaling Technology) primary antibodies. Antibodies were diluted 1:400, 1:200, 1:100, 1:500, 1:200, 1:100, 1:100 and 1:100 respectively in PBS containing 1% blocking reagent (nonfat dried milk). Membranes were rinsed 3 times with PBS/0.05% Tween 20 and then incubated with horseradish peroxidase-labeled secondary antibodies (1:3,000).
Bound antibodies were detected by chemiluminescence using a Covalab kit and Hyperfilm ECL (Amersham Biosciences, Orsay, France). The blots were reprobed with an anti-actin antibody 1:4,000 diluted (Sigma, St. Quentin-Fallavier, France). The data were expressed as protein level:β-actin level ratio.
NCCIT cells were seeded at 3 × 105 cells per well into 6-well plates. Confluent cells (60–70%) were then transfected. SiRNA integrin-α6 transfection was conducted according to the manufacturer's specifications. Briefly, 3 μl of Lipofectamine (Invitrogen, USA) per well were added to serum-free RPMI for a final completing volume of 200 μl and vortexed. They were then incubated at room temperature for 15 min. A total of 10 μl of 20 μM solution of the Silencer™ Validated siRNA integrin-α6 (Ambion, Austin, TX) were added to the diluted Lipofectamine reagent, gently mixed, and then incubated at room temperature for 15 additional min. The transfection agent/siRNA complex mixture was added into the wells containing 800 μl serum-free RPMI and cells were incubated at 37°C, 5% CO2 for 4 hr. One milliliter RPMI containing 30% FBS was added to serum-free cell culture medium. Scrambled siRNA (Silencer™ Negative Control siRNA; Ambion), a non-homologous to any known gene sequence, was used as a negative control. An additional siRNA transfection was carried out 72 hr after the first one. Cells were harvested at different time points to measure the α6-integrin protein level by Western blot using an antibody raised against the α6-integrin subunit (Santa Cruz Biotechnology).
To test whether α6-integrin is involved in NCCIT cells sensitivity to cisplatin-induced apoptosis, NCCIT cells were first transfected with α6-integrin siRNA and after 48 hr of culture, cells were harvested and seeded on either laminin- BSA- or fibronectin-coated plates. Twenty-four hours later, cells were transfected once again and 24 hr after this second transfection, cells were treated with 10 μM cisplatin for 4 hr. In situ TUNEL was carried out to quantify apoptotic cells 48 hr after cisplatin treatment.
Data represent the means of at least 4 independent experiments. Statistical significance was determined by one-way ANOVA with Dunnett's post-test using Prism version 3.03 (GraphPad Software, Inc., San Diego, CA). Differences were accepted as significant at p < 0.05.
Effect of ECM proteins in cisplatin-induced apoptosis
Forty-eight hours after the end of the 10 μM cisplatin treatment, apoptotic cells were quantified using the in situ TUNEL method (Fig. 1). Positive NCCIT cells cultured on plastic (PLA), poly-L-lysine (LYS), BSA, vitronectin (VTN), or fibronectin (FBN) coated plates displayed low percentage (7% and 9%) of apoptotic cells. When NCCIT were cultured on laminin (LAM) coated plates, however, the percentage of apoptotic cells was significantly (p < 0.001) higher (20%) compared to treated cells cultured on the other substrata. Untreated control cells cultured on any substratum as well as treated NCCIT cells cultured on Type IV collagen (COL) did not undergo apoptosis, showing <1% of apoptotic cells.
Cell viability and proliferation
To evaluate the role of ECM proteins on NCCIT cells growth and sensitivity to cisplatin, we used XTT assay to determine the amount of viable cells. As shown in Figure 2a, there were no significant differences between untreated NCCIT cells grown on laminin as compared to untreated cells grown on BSA or fibronectin. The cisplatin treatment reduced the viability of NCCIT cells, the number of viable NCCIT cells decreased in a time- and substrata- dependent manner. As shown in Figure 2b, the viability of treated NCCIT cells cultured on laminin significantly (p < 0.01) decreased 24 and 48 hr after the end of the cisplatin treatment as compared to treated cell cultured on the other substrata. The in situ TUNEL results (Fig. 2c) were concordant with the data obtained with the XTT viability test. At 24 and 48 hr after the end of cisplatin treatment, the number of apoptotic NCCIT cells was significantly (p < 0.01) higher when the cells were grown on laminin compared to cells seeded on BSA and FBN.
Caspase-3 mRNA and protein expression levels
In NCCIT cells cultured on laminin, BSA or fibronectin and treated with cisplatin, semi quantitative co-amplification RT-PCR was carried out at 0, 12, 24 and 48 hr after the end of the cisplatin treatment. It showed no significant difference in caspase-3 mRNA expression in all tested conditions (Fig. 3a,b).
Western blotting analysis showed that the pro-form of the enzyme (procaspase-3) was highly expressed in untreated (C) as well as in treated cells. The level of procaspase-3 (Fig. 3c,e) remained elevated at least 48 hr after the end of cisplatin treatment, in NCCIT cells plated on laminin, BSA or fibronectin. In contrast, at 48 hr after the end of the treatment, the treated NCCIT cells grown on laminin, showed a highly increase (p < 0.01) of the activated p17 caspase-3 as compared to cells cultured on BSA or fibronectin coatings. The p20-caspase-3 isoform was weakly expressed in all the conditions (Fig. 3d,e).
Caspase-6 mRNA and protein expression levels
The data in Figure 4a,b show that caspase-6 mRNA was expressed in equal amount either in untreated control cells (C) or in cisplatin treated NCCIT cells. Furthermore, the various substrata (laminin, BSA or fibronectin) did not modify caspase-6 mRNA expression.
Procaspase-6 protein level evaluated by western blotting (Fig. 4c,e) was significantly increased (p < 0.05) 48 hr after the end of the cisplatin treatment in all substrata. In contrast, 48 hr after the end of the cisplatin treatment, active cleaved caspase-6 level was highly increased (p < 0.01) in NCCIT cells cultured on laminin, as compared to NCCIT cells cultured on BSA or fibronectin (Fig. 4d,e).
Caspase-7 mRNA and protein expression levels
As indicated in Figure 5a,b the caspase-7 mRNA expression was not modified either by the cisplatin treatment or by the substratum. Whereas in untreated NCCIT cells, active cleaved caspase-7 protein was weakly expressed, 48 hr post-cisplatin treatment, its level was 2-fold increased (Fig. 5c,d), but this phenomenon was substratum independent.
Caspase-2, -8 and -9 protein expression levels
In cisplatin-treated NCCIT cells cultured on laminin, BSA or fibronectin, Western blotting analysis was carried out at 0, 12, 24 and 48 hr after the end of the cisplatin treatment.
Pro-caspase-2 protein was highly expressed in untreated control cells as well as cisplatin treated cells. In contrast, active cleaved caspase-2 was weakly expressed in all tested conditions (Fig. 6a–c).
Pro-caspase-8 and active cleaved caspase-8 expression increased after cisplatin treatment on all substrata. High level of pro-caspase-8 was observed at 24 and 48 hr after the end of cisplatin treatment. In all conditions, active cleaved caspase-8 still weakly expressed compared to pro-caspase-8 levels (Fig. 6d–f).
Active cleaved caspase-9 expression showed an 5-fold level increased early after cisplatin treatment and remained highly expressed at least until 48 hr after the end of the treatment but this phenomenon was substratum independent (Fig. 6g,h).
Caspase-3, -6 and -7.
The immunohistochemical analysis of activated caspases was carried out at 48 hr after the end of cisplatin treatment. Activated caspase-3 immunodetection showed stronger positive staining in NCCIT cells cultured on laminin (Fig. 7b) compared to cells adhered on BSA (Fig. 7a). Similar results, although to lesser extend, were obtained on activated caspase-6 immunostaining of NCCIT cells seeded on laminin (Fig. 7d) compared to BSA (Fig. 7c). It is noticeable that this phenomenon concerned only a subpopulation of NCCIT cells. Active cleaved caspase-7 immunostaining showed no difference in NCCIT cell cultured on laminin (Fig. 7f) compared to cell cultured on BSA (Fig. 7e).
Laminin and α6 integrin.
The immunohistochemical analysis of laminin (Fig. 8a,c) and of α6 integrin (Fig. 8b,d) expression in human seminoma tumor (Fig. 8a,b) and human embryonal carcinoma tumor (Fig. 8c,d) showed that all tumor cells were positively stained for laminin and α6 integrin. Seminoma cells were highly positive for α6 antibody, however, whereas embryonal carcinoma cells were highly positive for laminin.
Effect of caspase inhibitors on cisplatin-induced apoptosisin NCCIT cells cultured on laminin
We evaluated the cisplatin-induced apoptosis in NCCIT cell cultured on laminin in presence of DEVD-fmk, a caspase-3 inhibitor; VEIED-fmk, a caspase-6 inhibitor and VAD-fmk, the general pan caspases inhibitor. In situ TUNEL analysis (Fig. 9) showed an at least 3-fold decrease in apoptotic cells number in presence of these inhibitors compared to cisplatin treated cell in absence of caspase inhibitor (Control). These caspase inhibitors were clearly able to rescue NCCIT cells cultured on laminin from cisplatin-induced apoptosis.
Expression of AIF mRNA and protein
We assessed, using semi-quantitative co-amplification RT-PCR, a time-sequence profile of AIF gene expression in NCCIT cultured on laminin, BSA or fibronectin coated wells and treated with cisplatin (Fig. 10a,b). We showed that untreated control cells (C) grown on laminin, fibronectin or BSA, displayed no significant difference in AIF gene expression. In NCCIT cells cultured on laminin, however, the level of AIF mRNA was significantly increased immediately after cisplatin treatment as compared to NCCIT cells cultured on BSA or fibronectin and remained higher up to 24 hr after the end of cisplatin treatment.
Analysis of whole cell extracts by Western blotting (Fig. 10c,d) indicated that the level of immature AIF protein (67 kDa) was low in control untreated NCCIT cells culture on either laminin, BSA or fibronectin coated plates. When NCCIT cells were treated with cisplatin, we observed on laminin coating only, a significant (p < 0.05) increase in immature AIF expression. The immature AIF protein level remained overexpressed at least 48 hr after the end of the treatment.
Untreated NCCIT cells displayed low level of mature 57-kDa AIF protein (Fig. 10e,f). Immediately after cisplatin treatment (Hour 0), the level of mature AIF protein was increased significantly in NCCIT cells cultured on laminin, as well as on BSA or fibronectin. Interestingly, 12 hr after the end of the cisplatin treatment, we observed NCCIT cells cultured on laminin exhibited an increased expression of mature AIF by at least 1.5-fold whereas in NCCIT cells cultured on BSA or fibronectin coated plates, mature AIF expression decreased rapidly after treatment. We observed a low level of mature AIF in the 3 tested conditions 24 and 48 hr after treatment (Fig. 10e).
Integrin expression in NCCIT cell line
The immunohistochemical analysis of integrin expression (Table II) showed a strong expression level of α6, β1 and β4 integrin subunits whereas α1, α3 and α5 subunits were poorly expressed. The subunits α2 and α4 were not detected in the NCCIT cell line.
Knockdown of α6 integrin by siRNA in NCCIT cells and apoptosis
The transfection of α6 siRNA in NCCIT cells resulted in a significant reduction (70%) in α6A and α6B integrin isoform protein levels as compared to control cells (non-transfected cells) (C). Irrelevant scrambled siRNA (Scr) transfection had no effect on α6-integrin synthesis (Fig. 11a). Furthermore, this knockdown of the α6 integrin subunit in NCCIT cells cultured on laminin reduced significantly (2-fold) the number of apoptotic cells as compared to nontransfected cells or cells transfected with scrambled negative control siRNA (Fig. 11b). The knockdown of the α6-integrin subunit had no significant effect in NCCIT cells cultured on BSA or fibronectin with regard to apoptosis. It is noteworthy that the apoptotic cell number of NCCIT cells with the knockdown of the α6-integrin subunit cultured on laminin was reduced at a level similar to that observed on NCCIT cells cultured on BSA or fibronectin.
We provided evidence that a single tumor cell line can change its sensitivity to an anticancer drug depending to the stromal protein it adhered. We reported that among the extracellular components, which represent immediate cell microenvironment, only laminin promoted the cell death process triggered by cisplatin in a tumor germ cell line, NCCIT. We showed that NCCIT cells, which are basically moderately cisplatin-sensitive in standard culture conditions,25 exhibited a significantly higher sensitivity to cisplatin when cultured on laminin. This effect was laminin-specific because it was not observed on plastic, BSA, or on other ECM components such as fibronectin, vitronectin or Type IV collagen. NCCIT cells were at least 2-fold more sensitive to cisplatin-induced apoptosis when they were adherent to laminin as compared to other substrata. Untreated NCCIT cells adherent on laminin did not undergo apoptosis suggesting that the induction of apoptosis is not a direct effect of laminin but, rather a promotion or a facilitation of the cisplatin-induced apoptosis in these TGCT cells. Interestingly, Type IV collagen completely prevented cisplatin-induced apoptosis. This observation emphasizes the importance of ECM-tumor cells interactions in the control of cisplatin cell sensitivity. The protective effect of Type IV collagen against apoptosis induction is in agreement with many recent papers that have already shown that the extracellular matrix can protect against drug-induced apoptosis.15, 16, 17, 18, 19, 20 Our study was focused on a relative unexplored area, the microenvironment mediated increase in the chemotherapeutic drug response as observed in NCCIT cells cultured on laminin.
It is known that cell growth and proliferation are closely regulated by ECM proteins.28 An increase in the proliferating rate could indirectly enhance the cisplatin-induced DNA damage and subsequently the number of apoptotic cells. Such increase in cisplatin sensitivity observed on laminin was not due to a growth stimulation of NCCIT cells, however, because the growing rate of the cells was not different as compared to other substrata. In standard conditions, without any drug treatment, no significant difference was observed in the number of viable NCCIT cells regardless of the nature of substratum on which they were grown.
We further investigated the signaling mechanisms whereby attachment to laminin can enhance NCCIT cells sensitivity to cisplatin-induced apoptosis. It has been shown previously29 that cisplatin-induced apoptotic pathway in TGCT cell lines is p53-independent and is not correlated with the Bcl-2/Bax rheostat. We specifically focused on both effector caspases and AIF that is believed to mediate caspase-independent cell death pathway.30 We showed that increase of cisplatin-induced apoptosis in the NCCIT cell cultured on laminin could be related to the upregulation of the activation of caspase-3 and caspase-6. Caspases are cysteine proteases expressed as inactive precursors (procaspases) that are activated upon proteolytic cleavage. Once cleaved by initiator caspases (caspase-2, -8 and -9), the executioner caspase-3,-6 and -7 cause cell death by cleavage of molecules crucial for the cell integrity.10, 31, 32 Active cleaved caspase-3 and -6 levels were specifically increased in NCCIT cells cultured on laminin whereas active cleaved caspase-7 was upregulated in a substratum independent manner. Our results suggest that laminin specifically stimulates the proteolytic activation of executioner caspase-3 and -6 after cisplatin treatment. Activated caspase-7 may be involved in cisplatin-induced apoptosis but in a substratum independent manner. Caspase-7 might be responsible for the observed basal level of apoptosis (<9%) 48 hr after cisplatin treatment on BSA and on fibronectin. Laminin additional activation of caspase-3 and caspase-6 allowed the NCCIT cells to amplify the cell death process. Procaspase-3 and -6 mRNA levels remained unchanged both in untreated and treated NCCIT cells on all tested substrata. Whereas procaspase-3 protein level was not modified in all the tested experimental conditions, procaspase-6 protein level was increased 48 hr after cisplatin treatment but in a substratum-independent manner. Because laminin enhancement of cisplatin-induced apoptosis was related to effector caspase activation, we investigated activator caspase expression in NCCIT cells cultured on laminin. Previous studies have shown that, in response to DNA damage, activation of caspase-2 is required before mitochondrial permeabilization. Caspase-2 is the earliest caspase activated in different cell types treated with anticancer drugs.33, 34, 35 In NCCIT cells, the level of activated caspase-2 was low in all tested conditions whereas pro-caspase-2 was highly expressed even in untreated cells. Because laminin did not change either its expression or activation, we suggest that caspase-2 does not play a role in the regulation of cisplatin-induced apoptosis in NCCIT cells.
Caspse-8 is a critical caspase in death receptor-mediated apoptosis triggered by Fas and TNF. It may play a role in determining sensitivity or resistance of tumor cells to anticancer drugs (etoposide and cisplatin) induced apoptosis.36, 37 Our results indicate that caspase-8 was activated at least at 48 hr after the end of cisplatin treatment but in substratum independent manner. The laminin did not change the expression of caspase-8 in NCCIT cells.
After DNA damage, mitochondrial release of cytochrome c and the subsequent activation of procaspase-9 are essential for activation of downstream apoptotic effectors. Caspase-9 is an important intracellular amplifier of caspase signaling downstream of mitochondria. As caspase-8, caspase-9, which is activated through an apoptosome-induced conformational change, processes the executioner caspase-3 and -7 to initiate the execution of apoptosis. A recent study by Mueller et al.6 demonstrated that the active caspase-9 was the trigger that starts the apoptotic program after cisplatin treatment. The failure to activate caspase-9 confers resistance to cisplatin-induced apoptosis in TGCT cells. We showed that cisplatin treatment of NCCIT cells highly increased active caspase-9 levels but in substratum independent manner and its level remained unchanged at least 48 hr after the end of cisplatin treatment. Laminin, as well as other substrata, did not change initiator caspase-9, -8 and -2 activation patterns.
Thus, with the cisplatin treatment, laminin does not seem to affect specifically the transcription and the translation of the effector caspase-3 and -6 but rather to stimulate their proteolytic activation leading to an increase of the active forms of these caspases in NCCIT cells, 48 hr after the end of the treatment. Indeed the active cleaved caspase-3 and, to a lesser extent, the active cleaved caspase-6 were markedly immunostained 48 hr after the end of the treatment in NCCIT cells adhered on laminin compared to cells cultured on BSA or fibronectin substrata (data not shown). It is noticeable that only a small number of cells were strongly immunostained for active caspase-3 and -6 corresponding to the 20% apoptotic cells, which were observed 48 hr after the cisplatin treatment. The findings that caspase inhibitors were able to suppress cell death enhanced by laminin confirm that caspases contribute to cisplatin hypersensitivity in NCCIT cell cultured on laminin. Because a single caspase inhibitor was as effective as the broad-spectrum caspase inhibitor to protect cell from apoptosis suggest that the cisplatin-induced apoptosis requires additional activities of caspase-3 and caspase-6 and probably of other apoptotic effectors. Surprisingly, the activation of executioner caspases -3 and -6 in NCCIT cultured on laminin was not related to differential activation of the upstream initiator caspase-2, -8 and -9. The nature of this laminin-related integrin specific signal and the mechanism by which it is coupled with caspase-3 and -6 proteolytic activation remain to be investigated.
Because several authors38, 39, 40 suggested the ability of AIF and of caspases to cooperate during the apoptotic process, we investigated the AIF expression. Interestingly, we found that AIF mRNA was early upregulated by cisplatin treatment only in NCCIT cells grown on laminin. This amount remained significantly elevated 24 hr after the end of cisplatin treatment. The increase in the apoptotic cell number at 24 hr after the end of cisplatin treatment with a very early accumulation of mature AIF protein in NCCIT cells after cisplatin treatment support the hypothesis41 that the decrease of AIF mRNA could be due to the chromatin condensation and the start of DNA degradation. This phenomenon, which occurred in the earliest stages of apoptotic nuclei degradation, could induce a decrease in the transcription process.
Furthermore, the adhesion of NCCIT cells to laminin allowed cisplatin to activate AIF mRNA translation. We found that 48 hr after the cisplatin treatment, the immature 67 kDa AIF protein was overexpressed in NCCIT cells grown on laminin as compared to cells cultured on fibronectin or BSA. This could be a result from the translation of highly expressed AIF mRNA templates preformed during the first hours after cisplatin treatment in the laminin growing NCCIT cells.
Immediately after cisplatin treatment, we showed a high-amount of mature AIF protein in NCCIT cells grown on laminin, fibronectin or BSA. The mature AIF protein decreased rapidly in NCCIT cells cultured on fibronectin or BSA, whereas in NCCIT cells seeded on laminin, it remained elevated for >12 hr. The delayed mature AIF degradation observed on laminin could be due to the overexpression of the immature AIF, modifying the ratio of AIF synthesis and maturation vs. degradation, in favor of the AIF synthesis in early time after treatment. Because AIF apoptotic effect occurs very early (from 0–6 hr) after cell death induction,41, 42 the prolonged accumulation of this molecule (12 hr) in NCCIT cell cultured on laminin, could allow it to translocate into the nuclei, bind to the DNA and cause DNA damage. Indeed, our study of AIF release from mitochondria and the temporal relationship between the AIF nuclear translocation and caspase activation may provide informations concerning the role of AIF in our model.
To acquire more insights in the mechanism whereby laminin enhanced cisplatin-induced apoptosis, we investigated which receptor mediates the observed effect. Integrins are the main trans-membrane receptors that mediate cell attachment on extracellular matrix proteins and transduce signals that regulate cell growth, survival, differentiation, proliferation or apoptosis.28, 43, 44 Several different integrins can bind to the same ECM component and subsequently induce different signals.13, 45 There have been contradictory findings in the role of ECM proteins and integrin signaling. For example, basement membrane has been shown to inhibit apoptosis in mammary epithelial cells in vitro and in vivo whereas isolated collagen or fibronectin failed to protect these cells from apoptosis.46 The integrin αvβ3 protects from apoptosis melanoma cells cultured on collagen gel47 but the inhibition of this integrin had no significant effects on apoptosis in Chinese hamster ovary cells (CHO cells).48 The integrin α5β1 is able to protect CHO cells adhered to fibronectin whereas αvβ1, another fibronectin, receptor was ineffective in this regard.49 Recent studies reported that in breast cancer cells,16 small-cell lung cancer cells,17 hepatocellular carcinoma cells or ovarian cancer cells,19 the cell-ECM interactions via integrins inhibit apoptosis induced by chemotherapeutic drugs, including cisplatin. In human hematopoietic cell lines22, 50 or in peripheral blood monocytes,24 fibronectin was able to induce apoptosis via interaction with α5β1 integrin. Furthermore, laminin linkage to α3β1 integrin induced apoptosis in human Jurkat T cells.23
We showed that NCCIT cells highly expressed the integrins α6β1 and α6β4. The knockdown of the α6-integrin subunit suppressed the ability of laminin to enhance cisplatin-induced apoptosis. The high cisplatin sensitivity observed in NCCIT cells cultured on laminin is mediated by the α6β1 or α6β4 linked to laminin. Our results suggest that these laminin receptor signaling may cooperate with apoptotic pathways to increase NCCIT cell sensitivity to cisplatin. It was shown recently that the cisplatin-induced apoptosis in NCCIT cells depends on integrin-mediated activation of MEK and ERK, 2 members of mitogen-activated protein kinase (MAPK) family.51, 52 Further studies should determine the specific laminin–α6-integrin downstream signaling molecules that improve the apoptotic response after the cisplatin treatment. Interestingly, the in vivo and in vitro expression of laminin and α6 integrin we found in a panel of human germ cell tumors is in agreement with our hypothesis that these molecules could play a crucial role in the well known susceptibility of these tumors to cisplatin-induced apoptosis.
Our results showed that when adherent on laminin via the α6 integrin subunit, the sensitivity of NCCIT cells to cisplatin is significantly increased and involves a sequential activation of AIF and caspase-3 and -6. In NCCIT cells, mature AIF protein upregulation occurred early after cisplatin treatment and before caspase-3 and caspase-6 activation. NCCIT cells adherent on laminin, which displayed a prolonged (>12 hr) accumulation of mature AIF, were able to activate caspase-3 and caspase-6. It has been shown that mature AIF can be released from mitochondria before cytochrome c and that neutralization of AIF can prevent mitochondrial release of cytochrome c, suggesting that AIF could be required for cytochrome c-dependent caspase activation cascade.11, 53, 54 We can postulate that, in NCCIT cells, AIF is required for the activation of the executioner caspase-3 and -6, by the cytochrome c-dependent pathway. We can not, however, exclude the possibility that AIF acts also as a caspase-independent death effector,11, 53, 55, 56 and on its own, triggers apoptosis by causing chromatin condensation and large-scale (50 kbp) DNA fragmentation.11, 57 The absence of differential expression of initiator caspase argues in favor of this hypothesis.
Our study shows clearly for the first time that the complex laminin/α6 integrin signaling can potentiate the cisplatin-inducible apoptosis in a human tumor germ cell line. This effect is reversible and not linked to any gene alteration. Laminin–α6-integrin signaling promotes high expression of AIF early after cisplatin treatment and later activation of caspase-3 and -6. Our results suggest that laminin–α6-integrin signaling enhances the cell response to drug-induced apoptosis by stimulating the activation of at least 2 apoptotic effectors. Systematic investigations of extracellular factors upon chemotherapy efficacy in GCT9 as well as of the molecular events mediating ECM-tumor interactions might give insights to cisplatin efficiency and provide new tools for cancer treatment. These results are important because they show that the chemosensitive phenotype of these cells is partially under the control of cell–stroma interactions. These data open a new field of study concerning the role of cell environment, integrin signaling, and the epigenetic events associated with chemotherapy resistance.58, 59 This concept may be translated in the clinics and particularly in clinical trials by using specific drugs, which may act on cell-ECM interactions.
M. Andjilani is a recipient of a grant from the Collectivité Départementale de Mayotte. The authors would like to thank M. Groleas and M. Hesch for their excellent technical assistance.
- 59Inhibition of integrin alpha5beta1 function with a small peptide (ATN-161) plus continuous 5-FU infusion reduces colorectal liver metastases and improves survival in mice. Int J Cancer 2003; 104: 496–503., , , , , , , , , .