Background
Tumor necrosis factor related apoptosis inducing ligand is a member of the tumor necrosis factor (TNF) superfamily. Recombinant human TRAIL (rhTRAIL) is currently drawing attention in the field of cancer therapy because of its specific action in inducing apoptosis in tumor cells. Five TRAIL-receptors have been identified to date. The death receptors DR4 and DR5 transduce the apoptotic signal, while three decoy receptors - decoy receptor (DcR1), decoy receptor 2 (DcR2) and osteoprotegerin (OPG) - block the signal and thereby inhibit TRAIL-mediated apoptosis [
1,
2]. Administration of rhTRAIL in tumor-bearing animals has been shown to induce significant tumor regression without systemic toxicity [
3,
4]. Furthermore, rhTRAIL in combination with chemotherapy or radiotherapy greatly enhances anti-tumor efficacy, both
in vitro and
in vivo [
5‐
8].
The TRAIL apoptotic pathway can also be stimulated by death receptor (DR)-specific agonistic antibodies. These anti-DR4 and anti-DR5 monoclonal antibodies, either used alone or in combination with chemotherapy (or irradiation), induce apoptosis in tumor cells
in vitro and
in vivo [
9‐
12]. Thus, both rhTRAIL and agonistic antibodies exhibit interesting preclinical anti-tumor properties. A phase I clinical study on rhTRAIL has been initiated [
13]. Several phase I-II clinical studies on agonistic DR4 antibodies, as well as a phase I study on agonistic DR5 antibodies, have also been performed [
2,
14,
15]. However, because rhTRAIL and DR-agonistic antibodies differently stimulate the apoptotic signaling cascade, drug-specific effects in the treatment of cancer patients are expected [
16‐
18]. rhTRAIL, which can bind to DR4 and DR5 but also to the decoy receptors, triggers cross-linking of these receptors into homo- and/or heterotrimers [
19,
20]. In contrast, agonistic DR4 or DR5 antibodies have been suggested to trigger the formation of multimeric complexes consisting of only one specific receptor, which consequently enables them to bypass the decoy receptors [
21,
22].
Not all tumor cells are sensitive to rhTRAIL, since intrinsic or acquired resistance to this ligand can occur. Very little is known about the specific properties of different DR agonists when it comes to the downstream activation signaling pathways (e.g. NFκB) and resistance to rhTRAIL. However, rhTRAIL and agonistic anti-DR5 antibodies are known to exhibit different abilities to induce the conformational changes in DR5 which are required to facilitate FADD recruitment [
23].
The cytokine IFN-γ, and also proteasome inhibitors, are both known to modulate components of the apoptotic signaling pathway involved in TRAIL resistance [
24‐
26]. Combinations of these drugs with TRAIL and/or agonistic death receptor antibodies can enhance TRAIL-induced apoptosis and overcome TRAIL resistance in tumor cells [
27‐
32]. However, potential receptor specific effects on the development of resistance to rhTRAIL have not been investigated. This is of interest, as it has not yet been established which of the agents of interest - DR4 antibodies, DR5 antibodies or rhTRAIL - exhibit superior anti-tumor activity in the clinic. Moreover, it is still unknown which biomarkers should be used to select patients for therapies specifically targeting DR4 and DR5.
In the present study we used agonistic monoclonal antibodies to individually evaluate the roles of DR4 and DR5 in rhTRAIL sensitivity and TRAIL resistance. We compared apoptosis induced by rhTRAIL with apoptosis induced by agonistic DR4- and DR5- antibodies, taking as our model an rhTRAIL-sensitive human colon cancer cell line and its rhTRAIL-resistant sub-line [
33]. Furthermore, we analyzed whether the effects of IFN-γ and a proteasome inhibitor in modulating apoptosis differed depending on how apoptosis was induced; that is, we compared the modulatory effects of these compounds after using agonistic DR4 and DR5 antibodies, or rhTRAIL, to initiate apoptosis.
Methods
Reagents
We used RPMI 1640 medium obtained from Life Technologies (Breda, the Netherlands) and fetal calf serum (FCS) from Bodinco BV (Alkmaar, the Netherlands). 3-(4,5-dimethyl-thiazol-2-yl) 2,5-diphenyltetrazolium bromide (MTT)-solution and CHX were purchased from Sigma-Aldrich Chemie BV (Zwijndrecht, the Netherlands). rhTRAIL was produced non-commercially in cooperation with IQ-Corporation (Groningen, the Netherlands) following a protocol described earlier [
34]. To stimulate DR4, an agonistic anti-DR4 antibody (HGS-ETR1) was used. The agonistic anti-DR5 antibodies HGS-ETR2 or TR2J were used to stimulate DR5 with similar results. HGS-ETR1, HGS-ETR2 and TR2J were a kind gift from Human Genome Sciences (HGS, Rockville, MD, USA). The inhibiting anti-DR4 (HS 101) and anti-DR5 (HS 201) antibodies were purchased from Alexis (10 P's BVBA, Breda, the Netherlands). The proteasome inhibitor MG-132 was obtained from Calbiochem (Breda, the Netherlands), and IFN-γ was purchased from Roche Diagnostics (Mannheim, Germany). The TRAIL receptor agonistic antibodies used for flow cytometry were obtained from Immunex Corporation (Seattle, WA, USA).
Cell lines
The rhTRAIL-sensitive colon carcinoma SW948 cell line [
35] was cultured as described previously [
36]. The rhTRAIL-resistant SW948-TR cell lines was generated and cultured as described recently [
33].
SDS-polyacrylamide gel electrophoresis and Western blotting
Protein lysate preparation and Western blot analysis were performed as described previously [
36].
Co-Immunoprecipitation of TRAIL-, DR4- or DR5-DISC
DISC immunoprecipitation after TRAIL-receptor ligation was performed according to Bodmer
et al. [
37] with some modifications. Briefly, 50.10
6 cells per condition were grown, harvested and collected by centrifugation. The cell pellet was resuspended in 1 ml pre-warmed medium, and the tube was placed in a 37°C incubator. Recombinant human soluble flag-tagged TRAIL and the anti-Flag monoclonal antibody M2 were premixed for 15-30 min on ice. Cells were stimulated in a final volume of 1 ml with 500 ng/ml Flag-tagged TRAIL and 1.5 μg/ml M2. In unstimulated cells, the Flag-tagged TRAIL and M2 premix were added after lysis to immunoprecipitate non-stimulated TRAIL-receptors. Cell suspensions were incubated for 30 min at 37°C, and the reaction was stopped by the addition of 10 ml ice-cold phosphate buffered saline (6.4 mM Na
2HPO
4; 1.5 mM KH
2PO
4; 0.14 mM NaCl; 2.7 mM KCl; pH = 7.2). The cells were immediately washed with ice-cold PBS and lysed in 1 ml lysis buffer (30 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride) with complete protease inhibitors (Roche Diagnostics, Almere, the Netherlands) for 15 min on ice. After centrifugation (2,500 ×
g) at 4°C for 10 min, the lysates were pre-cleared with 20 μl Sepharose-6B (Pharmacia, Uppsala, Sweden) for 2 h at 4°C and immunoprecipitated with 30 μl protein-A sepharose beads for 4 h-overnight at 4°C. Beads were washed three times with 1.5 ml lysis buffer, resuspended in standard Western blot sample buffer, and boiled for 5 min. Immunoprecipitated proteins were separated with SDS-PAGE. Western blot analysis for FADD, caspase 8, c-FLIP, DR4, DR5 and TRAIL was performed as described in the SDS-polyacrylamide gel electrophoresis and Western blotting section. Goat HRP-conjugated secondary antibody specific for mouse IgG
1 and donkey-anti-goat-HRP were used for the detection of caspase 8, FADD, c-FLIP or DR4.
DR4 and DR5-DISC co-immunoprecipitation was performed as described above using the respective Human Genome Sciences antibodies (HGS-ETR1 and TR2J), with some modifications. Briefly, 50.106 cells per condition were grown in medium harvested and re-suspended in fresh medium. Cells were stimulated with 5 μg/ml antibody in a final volume of 2 ml. The antibodies were added after cell lysis for the control treatment. Cell suspensions were incubated for 15 min at 37°C, and the reaction was stopped by the addition of 10 ml ice-cold phosphate buffered saline (PBS). The cells were immediately washed with ice-cold PBS and lysed in 1 ml lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.2% Nonidet P-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride) with complete protease inhibitors (Roche Diagnostics, Almere, the Netherlands) for 30 min on ice. After centrifugation (12,000 × g) at 4°C for 10 min, the lysates were pre-cleared with 20 μl Sepharose-6B (Pharmacia, Uppsala, Sweden) for 2 h at 4°C and immunoprecipitated with 50 μl protein-G agarose beads (Roche Diagnostics, Mannheim, Germany) for 3 at 4°C. Beads were washed two times with 1 ml lysis buffer and one time with PBS before resuspension in standard Western blot sample buffer and boiling for 5 min. Immunoprecipitated proteins were separated with SDS-PAGE. Western blot analysis for FADD and c-FLIP was performed as described below. DR4 and DR5 were detected using rabbit anti-DR4 and rabbit anti-DR5 from ProSci Inc. (Poway, CA, USA). Caspase 8 was detected with rabbit anti-caspase 8 (Abcam plc, Cambridge, UK). Goat anti-rabbit HRP-conjugated secondary antibody and rabbit-anti-mouse-HRP were used for the detection of DR4, DR5, FADD, caspase 8 or c-FLIP.
Apoptosis assay
Apoptosis was assessed using acridine orange (AO) staining, using a method described earlier [
36].
Cytotoxicity assay
The microculture tetrazolium (MTT) assay was used to determine cytotoxicity. SW948 and SW948-TR cells were incubated in a total volume of 200 μl. After an incubation period of 96 h, 20 μl of 5 mg/ml MTT solution diluted in PBS was added for 3.75 h. Subsequently, plates were centrifuged and the supernatant aspirated. After dissolving the formazan crystals by adding dimethyl sulfoxide (Merck, Amsterdam, the Netherlands), plates were read immediately at 520 nm using a microtiter well spectrometer (Bio-Rad microplate reader, Bio-Rad laboratories BV, Veenendaal, the Netherlands). Controls consisted of media without cells. Cell survival was defined as the growth of treated cells compared with untreated cells. IC50 was the concentration of drug inhibiting survival by 50%. Mean cytotoxicity was determined in three independent experiments where each condition was performed in quadruplicate.
Flow cytometry
Analysis of DR4 and DR5 membrane expression was performed as described earlier [
36]. For competition experiments with rhTRAIL, SW948 cells (2.10
6/condition) were harvested, resuspended in 2.5 ml ice cold PBS and incubated on ice with various amounts of rhTRAIL for 1 h. Cells were then washed twice with PBS before fluorescent staining for DR4 and DR5 as described above.
Discussion
DISC formation is the first step toward apoptosis after engagement of DR4 and DR5 by their ligands. Not surprisingly, resistance to TRAIL-mediated apoptosis is often initiated at the level of the DISC [
42]. We observed that upon stimulation with TRAIL, all canonical DISC proteins were recruited to DR4 and DR5 in SW948 and SW948-TR cells. In relation to the other DISC proteins, smaller amounts of cleaved caspase 8 were found in the TRAIL-DISC of SW948-TR cells as compared with SW948 cells. These results were in agreement with the lower basal caspase 8 protein levels found in SW948-TR. Altogether, these findings suggest less active recruitment and processing of caspase 8 in our resistant cell line. We also describe a difference in apoptosis-inducing ability between DR4 and DR5. Despite equivalent initial DISC formation at the level of each receptor, DR4 stimulation achieves superior caspase 8 processing than DR5.
Antagonistic DR4 or DR5 antibodies were used to specifically block the function of their respective receptors during rhTRAIL-induced apoptotic signaling. Experiments with these blocking antibodies proved that DR4 was critical to rhTRAIL-induced apoptotic signaling in both cell lines; DR5 was not. It was previously reported that apoptosis induction in keratinocytes with leucine zipper TRAIL was also mainly mediated by DR4 [
43]. In contrast, Kelley
et al. generated receptor-selective mutants of TRAIL, with three to six ligand amino acid substitutions [
17], and found DR5 to be more important for apoptotic signaling than DR4 in cancer cells expressing both receptors (including colon cancer cells). Van der Sloot
et al. demonstrated that DR5-selective TRAIL variants did not induce apoptosis in cell lines mostly responsive to DR4 stimuli, while they greatly stimulated apoptosis in DR5-responsive cancer cell lines [
44]. Little is known about the mechanisms underlying these differences. Blocking of both DR4 and DR5 prevented apoptosis more efficiently than single blocking of DR4, which suggests that DR5 also contributes to rhTRAIL-induced apoptosis. While DR4 was more potent in transducing apoptosis, DR5 might, at least in our model, increase the overall apoptotic stimulus compared with DR4 stimulation alone.
We found that in SW948 and SW948-TR cells, sensitivity to agonistic DR4 and DR5 antibody mostly reflected the sensitivity of each receptor to rhTRAIL. Although the agonistic DR4 antibody was more potent than its anti-DR5 counterpart, agonistic DR5 antibody could induce apoptosis to some extent in SW948 cells, and also in SW948-TR following pre-sensitization with CHX. This strongly suggests that DR5 is functional in both cell lines, as also indicated by rhTRAIL binding and comparable DISC formation between DR4 and DR5. Georgakis
et al. found that in primary non-Hodgkin's lymphoma samples, an agonistic DR4 antibody was also more effective than an agonistic DR5 antibody [
16].
We investigated the ability of each receptor to initiate DISC formation and caspase 8 cleavage in SW948 by co-immunoprecipitating DR4- and DR5-DISC. Similar DISC formation and caspase 8 cleavage were triggered at the level of both receptors, although the rate of caspase 8 processing was much higher in whole cell lysates when cells were stimulated with agonistic DR4 antibody instead of agonistic DR5 antibody. This is consistent with our recent findings that caspase 8 cleavage at the level of DR5-DISC in SW948 is limited due to a lower turn-over of DISC components rather than to decreased DISC formation [
45]. Data on such DISC protein turn-over following TRAIL receptor stimulation are limited. However, studies by McDonald
et al. and Jin
et al. have suggested that caspase 8 ubiquitination influences processing of the available caspase 8 cellular pool [
46,
47]. Brief pre-incubation with CHX could also restore DR5-mediated caspase 8 cleavage and sensitivity; a short-lived protein might, therefore, be involved in this process.
IFN-γ is known to increase caspase 8 expression and, consequently, to sensitize cancer cells (including colon cancer cells) to rhTRAIL [
24,
41,
48‐
50]. Our results are in slight contrast to these findings, as upregulation of caspase 8 by IFN-γ did not further enhance apoptosis induction in the highly sensitive SW948 cells. Caspase 8 levels might not be a limiting factor in these cells, since procaspase 8 recruitment to the DISC is first determined by the amount of FADD molecules available there [
51]. IFN-γ did, however, increase apoptosis induction by rhTRAIL and agonistic DR4 antibody in SW948-TR cells, which express lower caspase 8 levels than their parental cells. In SW948-TR, IFN-γ induced a marked increase in both pro- and intermediate forms of caspase 8, which were cleaved to the active form upon DR4 stimulation. Downregulation of caspase 8 levels in SW948 - that is, to a level comparable to that which is normally observed in SW948-TR cells - was sufficient to induce TRAIL resistance, indicating the importance of the caspase-8/c-FLIP ratio in these cells [
33]. It is noteworthy that IFN-γ, while inducing an increase in caspase 8 levels, did not change c-FLIP expression in SW948-TR. In contrast, IFN-γ failed to enhance the appearance of active caspase 8 following DR5 stimulation. This was shown using agonistic DR5 antibody, but also rhTRAIL in combination with DR4 blocking antibodies. The slight reduction in DR5 surface expression may explain the lack of increase in DR5-mediated apoptosis.
Unlike IFN-γ, MG-132 markedly increased both DR4- and DR5-mediated sensitivity in SW948-TR cells. Another proteasome inhibitor, bortezomib, was previously reported to enhance the effect of agonistic DR4 and agonistic DR5 antibody in Hodgkin's disease cell lines [
18,
52]. MG-132 could increase TRAIL-induced apoptosis in both Bax-deficient and proficient colon cancer cells, which is suggestive of a sensitization mechanism independent of the mitochondrial pathway of apoptosis [
24]. DR4 and DR5 upregulation by proteasome inhibitors is thought to be a major factor contributing to TRAIL-induced apoptosis sensitization. However, this hypothesis has been subject to an extensive debate [
53]. We found that upon MG-132 treatment, DR4-mediated sensitivity increased in the absence of DR4 upregulation. The intermediate form of caspase 8 was upregulated by MG-132, an event which we found was sufficient to restore sensitivity to agonistic DR4 antibody or rhTRAIL following IFN-γ treatment. Importantly, MG-132 also increased caspase 8 cleavage and enhanced DR5-mediated apoptosis. MG-132 has previously been shown to modulate DR5 expression [
24,
54]. Our flow cytometry analysis only showed negligible DR5 upregulation by MG-132, which again points to a mechanism of sensitization independent of this effect. MG-132 upregulated caspase 8 in SW948-TR, but we demonstrated using IFN-γ that restoring caspase 8 levels was not sufficient to increase DR5-mediated caspase 8 cleavage and apoptosis. Taken together, these results support the existence of additional sensitizing mechanisms of MG-132 at the level of DR5-DISC, beyond DR5 and caspase 8 upregulation. Many of the proteins involved in the DR-mediated apoptotic pathways are regulated by ubiquitination [
55], and some of these proteins could play a crucial role in specific regulation of DR5 signaling.
In particular, understanding how MG-132 can overcome DR5-resistance may be of crucial importance, since Johnsen
et al. previously reported that IFN-γ failed to increase TRAIL-induced apoptosis in 3 out of 8 neuroblastoma cancer cell lines; this result occurred despite caspase 8 upregulation and in the presence of all the proteins known to be necessary for DISC formation [
41]. It would be interesting to verify whether DR4 signaling remains functional in the non-responding cells. Furthermore, in view of the large number of agents in clinical development targeting DR5, as opposed to DR4, a deeper knowledge on DR5-mediated signaling regulation is critical for a more rational design of targeted therapies.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
CvG and BP performed most of the experiments and drafted the manuscript. PL carried out the MTT assays. EdV and SdJ participated in the design and coordination of the study and helped to draft the manuscript. All authors read and approved the final manuscript.