The NSCLC cell lines H460, A549, and SW1573 were obtained from the American Type Culture Collection (ATCC) (Manassas, VA). The cell lines were tested annually for authenticity by short tandem repeat profiling DNA fingerprinting (Baseclear, Leiden, The Netherlands) and for mycoplasma by PCR. Cells were only used when mycoplasma negative. H460 and A549 cells were cultured in RPMI 1640 medium (Gibco, Paisley, Scotland), and SW1573 cells in DMEM medium (Gibco). Culture media were supplemented with 10% fetal calf serum (FCS) (Sanbio, Uden, The Netherlands), 100 U/mL penicillin, and 100 µg/mL streptomycin (Invitrogen, Breda, The Netherlands). Cells were grown at 37°C in a humidified atmosphere containing 5% CO₂. Cell lines were harvested by trypsinization and passaged twice weekly.

RhTRAIL, TRAIL-R1 specific variant (S159R) [23], and TRAIL-R2 specific variant (D269H/E195R) [24] were produced non-commercially as described earlier [25]. Bortezomib (Velcade™, Janssen-Cilag BV, Tilburg, The Netherlands) was provided as a pure substance and dissolved in 0.9% sodium chloride. Primary antibodies for cleaved caspase-8, -9, and -3, and Bid were purchased from Cell Signaling Technology (Danvers, MA, USA). Primary antibody for cFLIP was from Transduction Laboratories (Lexington, KY, USA), and mouse anti-XIAP clone 2F1 from MBL International (Woburn, USA). Mouse anti-β-actin was purchased from Sigma-Aldrich (Zwijndrecht, The Netherlands). The secondary antibodies used were goat-a-mouse-IRDye (1:10,000, 800CW; #926-32210 and 680; #926-32220), or goat-a-rabbit-IRDye (1:10,000, 800CW; #926-32211 and 680; #926-32221), all purchased from LI-COR Biosciences (Lincoln, Nebraska, USA).

Western blot analysis was performed as previously described [21, 22, 26]. In brief, the cells were disrupted in lysis buffer (Cell Signaling Technology Inc., Danvers/Boston, MA, USA), protein was separated on an SDS-PAGE, and electroblotted onto polyvinylidene difluoride (PVDF) membranes (Millipore, Amsterdam, The Netherlands). Next, membranes were blocked for one hour at room temperature in InfraRedDye blocking buffer (Rockland Inc., Pennsylvania, USA) and incubated overnight at 4°C with the primary antibodies diluted in 1:1 InfraRedDye blocking buffer with PBS-T (PBS with 0.05% Tween-20). The membrane was incubated with the secondary antibodies for one hour at room temperature in the dark. Protein expression was detected using the Odyssey Infrared Imager (LI-COR Biosciences, Lincoln, Nebraska, USA), 84 µm resolution, 0 mm offset, and high quality.

Apoptosis was measured using the FITC Annexin V Apoptosis Detection kit 1 (BD Biosciences, Franklin Lake, USA) according to the manufacturer’s protocol. Briefly, after exposure to rhTRAIL and/or bortezomib, cells were harvested and washed with cold PBS. Following resuspension in 1× binding buffer, 5 µL of FITC Annexin V and 10 µL of propidium iodide solution were added. Cells were incubated for 15 minutes at room temperature in the dark. After the addition of 400 µL 1× binding buffer, apoptosis was analyzed by flow cytometry using the FACS Calibur (Becton Dickinson) with an acquisition of 10,000 events. Cells were considered non-apoptotic (Annexin V-FITC negative/PI negative), in early apoptotic phase (Annexin V-FITC positive/PI negative), in late apoptotic phase (Annexin V-FITC positive/PI positive), or dead (Annexin V-FITC negative/PI positive).

Clonogenic survival was performed as described previously [26]. Concentrations of the drugs were based on drug sensitivity experiments (performed by using the MTT assay and clonogenic survival) as described earlier [19, 26]. Briefly, cells were seeded at a density of 250 cells/well and treated with 100 ng/mL TRAIL and/or 50 nM Bortezomib. After 6 hours of exposure, the medium was refreshed, and the cells were incubated for 1–2 weeks to form colonies. Upon the formation of colonies, cells were washed with PBS, fixed with 99% ethanol, and washed again. Colonies were stained with 10% Giemsa (Merck, Darmstadt, Germany), and colonies consisting of > 50 cells were counted. The surviving fraction was determined by dividing the number of colonies by the number of cells plated. The surviving fraction of untreated cells was set to 1.

Cell cycle analysis and cell death measurements were performed as described previously [27]. Briefly, 100,000 cells were seeded in a 6-well plate. After exposure to TRAIL, bortezomib, or their combination, cells were trypsinized, resuspended in medium collected from the matching sample, and centrifuged for 5 min at 1,200 rpm. Cells were washed with PBS and subsequently stained with propidium iodide-containing buffer [50 µg/mL propidium iodide, 0.1% (Tri-) Sodium Citrate, 0.1% Triton X-100, 0.1 mg/mL RNAse]. DNA content was analyzed by a FACS Calibur (Becton Dickinson) with an acquisition of 10,000 events. Cell death was quantified by calculating the area of the sub-G1 peak.

Sequences for small interfering RNA (siRNA) molecules were for Bid: 5’-GAA UAG AGG CAG AUU CUG AdTdT-3’ (sense) and 5’-UCA GAA UCU GCC UCU AUU CdTdT-3’ (anti-sense), for XIAP: 5’-GUG GUA GUC CUG UUU CAG CdTdT-3’ (sense) and 5’-GCU GAA ACA GGA CUA CCA CdTdT-3’ (anti-sense) and for cFLIP: 5’-GAG GUA AGC UGU CUG UCG GdTdT-3’(sense) and 5’-CCG ACA GAC AGC UUA CCU CdTdT-3’(anti-sense) as used before [28]. The siRNA control, without any known homology with the human genome, was purchased from Eurogentech (Seraing, Belgium). H460, A549, and SW1573 cells were transfected in 6-well plates with 10 µL of 20 µM siRNA duplexes using Oligofectamine reagent according to the manufacturer’s instructions (Invitrogen BV, Breda, The Netherlands). After 24 hours, cells were replated for apoptosis assay (96-well culture plate).

For apoptosis measurements, 3,000 cells were plated in 96-well culture plates. Cells were continuously incubated with bortezomib and/or rhTRAIL or TRAIL-receptor specific variants (S159R and D269H/E195R) at various concentrations, which were chosen based on data published earlier [19, 26]. Staining was performed 6 or 24 hours after the start of incubation. Acridine orange (5 mg/mL) was added to each well to distinguish apoptotic cells from viable cells. Staining pattern was determined by fluorescence microscopy, where apoptosis was defined as the appearance of apoptotic bodies and/or chromatin condensation. Results are expressed as the percentage of apoptotic cells in a culture by counting at least 300 cells per well.

All experiments were performed at least in triplicate. Data were analyzed by the student’s t-test for unpaired data; in cases appropriate, we used a paired t-test; p-values < 0.05 were considered statistically significant, and exact numbers are provided in the legends. In case no specification is given, there is no statistical difference. All data were normally distributed, and only comparisons between two groups were evaluated statistically. Statistical analysis was performed using GraphPad Prism (version 5.0 GraphPad Software, San Diego, CA, USA).

Since TRAIL rapidly induces apoptosis, we determined caspase activation at early time points (1, 2, 3, and 6 hours) in TRAIL-sensitive H460 and TRAIL-resistant A549 and SW1573 NSCLC cells upon exposure to bortezomib and/or rhTRAIL (Figure 1). Bortezomib as a single agent did not induce caspase cleavage within 6 hours of treatment in the 3 cell lines tested. In sensitive H460 cells, the combination of bortezomib with rhTRAIL did not affect the kinetics of rhTRAIL-induced caspase-8, -9, and -3 activation (Figure 1A). Caspase cleavage in these cells was already detectable after 1 hour, and increased at later time points upon rhTRAIL and combined treatment. A p37 cleavage product of caspase-9 was detected after 1 hour treatment, followed by a strong increase in both p37/35 cleavage products at later time points. Since caspase-9 p37 is generated by caspase-3-dependent cleavage, this suggests early initiation of the extrinsic apoptotic pathway and subsequent caspase-9 cleavage [29]. The p35 product, detected after 2 hours of treatment, has been linked with auto-cleavage of caspase-9 via the intrinsic pathway. Caspase-3 cleavage was initially marked by the appearance of p20 and p19 cleavage products at 1 hour post-treatment. By 6 hours, a clear increase in the p17 cleavage product, representing a more active form of caspase-3, was observed along with a concomitant decrease in full-length caspase-3 (Figure 1A).

In contrast to H460 cells, A549 cells displayed overall lower levels of caspase cleavage upon bortezomib and/or rhTRAIL treatment. Particularly, combined treatment induced weak caspase-8 cleavage that increased further at 6 hours post-treatment (Figure 1B). Only weak caspase-9 cleavage was detected following treatment with rhTRAIL alone or combined with bortezomib. Also, marginal caspase-3 cleavage was observed, although this did not yield the p17 product (Figure 1B). Cleavage of these caspases was detected up to 3 hours post-treatment and was hardly detectable after 6 hours.

Finally, in SW1573 cells (Figure 1C), exposure to rhTRAIL caused rapid cleavage of caspase-8, although at lower levels than seen in H460 cells. Caspase-8 cleavage was slightly increased upon combined rhTRAIL/bortezomib treatment. Furthermore, caspase-9 cleavage yielded mainly the p37 product that is generated by caspase-3-mediated caspase-9 cleavage. The p17 cleaved fragment of caspase-3 was detectable at 6 hours post-treatment, although a large proportion of full-length caspase-3 remained uncleaved (Figure 1C). In summary, rhTRAIL-induced caspase-8 cleavage is detectable at early time points in all cell lines, with high levels observed in sensitive cells and low levels in resistant cells. Combination treatment with bortezomib slightly increased caspase-8 cleavage. The cleavage patterns of caspase-3 and -9 were markedly different between the sensitive and resistant NSCLC cell lines, as well as between the resistant cells. Since in another study, the pan-caspase inhibiter zVAD completely prevented induction of cell death [30] and the caspase-8 inhibitor zIETD partially (manuscript submitted), we concluded that in these cell lines cell death was caspase dependent and we did not further investigate non-caspase induced cell death.

To determine whether caspase activation leads to apoptosis induction in H460, A549, and SW1573 cells, FACS analysis was performed using Annexin V and propidium iodide (PI). This approach enabled the identification of early apoptotic cells (Annexin V-positive/PI-negative), late apoptotic cells (Annexin V-positive/PI-positive), and necrotic or late-stage apoptotic cells (PI-positive only). Apoptosis was first examined after 6 hours of treatment. As expected, H460 cells were sensitive to 50 ng/mL rhTRAIL, showing early (13%) and late (3%) apoptosis and dead cells (7%) (Figure 2A). Bortezomib at 50 nM did not trigger cell death, whereas combined rhTRAIL/bortezomib exposure resulted in 18% early, 4% late apoptosis, and 9% dead cells. In A549 cells, 100 ng/mL rhTRAIL and/or bortezomib exposure resulted in similar levels of early and late apoptosis as seen in H460 cells, however, no cell death was induced within 6 hours of exposure (Figure 2A). SW1573 cells showed no induction of apoptosis and cell death after rhTRAIL or combined treatment (Figure 2A).

Next, we assessed apoptosis by maintaining the cells for 18 hours following the initial 6-hour exposure to bortezomib and/or rhTRAIL. The exposure of H460 cells to rhTRAIL alone resulted in 12%, 37% and 21% of early-, late-apoptotic and dead cells, respectively. The combination of bortezomib and rhTRAIL strongly increased late apoptotic (38%) and dead cells (49%). Both A549 and SW1573 cells displayed minimal induction of apoptosis when exposed to either rhTRAIL or bortezomib alone. However, combined rhTRAIL/bortezomib treatment showed around 15% late apoptotic cells and 35% dead cells (Figure 2B). The antitumor effect of rhTRAIL, bortezomib and combined treatment was also tested in clonogenic assays. In all three NSCLC cells, bortezomib had little effect (Figure 2C). Clonogenic growth was almost completely suppressed by rhTRAIL and combined treatment in H460 cells, whereas in A549 and SW1573 cells, only combined treatment significantly reduced clonogenic growth. Thus, 6 hours combined rhTRAIL/bortezomib treatment is sufficient to initiate apoptotic cell death at later time points and was able to strongly suppress clonogenic growth.

To investigate the mechanisms underlying bortezomib-dependent sensitization to TRAIL, we first explored the involvement of the extrinsic and intrinsic apoptosis pathways by inhibiting Bid. The cleavage and activation of Bid by caspase-8 connect the extrinsic- with the intrinsic apoptotic pathway, the so-called amplification loop. The three NSCLC cell lines were treated continuously with rhTRAIL, bortezomib or the combination for 24 hours. rhTRAIL resulted in over 50% cell death (measured as the sub-G1 fraction) in H460, while A549 and SW1573 showed less than 10% apoptosis (Figure 3A). Combined rhTRAIL/bortezomib treatment strongly increased cell death to approximately 30% in A549 and SW1573 cells and 70% H460 cells. Silencing of Bid strongly reduced rhTRAIL-induced cell death in H460 cells from around 55% to 5%, whereas no significant effects were seen in A549 and SW1573 cells. Upon combined treatment Bid silencing strongly inhibited cell death in H460 cells (Figure 3A). In A549 cells cell death activation by combined rhTRAIL/bortezomib exposure was effectively inhibited in Bid silenced cells. Interestingly, in SW1573 cells Bid silencing had no effect on rhTRAIL/bortezomib induced cell death. Together, these findings indicate that bortezomib sensitizes rhTRAIL induced apoptosis predominantly via the Bid-dependent amplification loop in H460 and A549 cells, whereas in SW1573 cells sensitization occurs by stimulating the extrinsic apoptotic pathway.

Since we observed distinct cleavage patterns for caspase-9 and caspase-3 in the 3 cell lines (Figure 1). We also examined the involvement of the caspase-9 and -3/-7 inhibitor XIAP in bortezomib-dependent sensitization. Silencing of XIAP increased rhTRAIL-induced cell death further in H460 cells from around 55% to 70% (Figure 3B). Interestingly, the somewhat increased level of cell death after combined treatment did not increase further after XIAP silencing, and showed actually some decrease in apoptosis. In XIAP-silenced A549 cells, no significant increase in rhTRAIL but a significant increase in rhTRAIL/bortezomib-induced cell death was seen compared to control siRNA transfected cells. Interestingly, in SW1573 cells XIAP silencing strongly enhanced rhTRAIL-induced cell death from around 5% to 45%. Also combined treatment increased cell death from around 30% to 60%. This suggests that in SW1573 cells, the intrinsic pathway is suppressed by XIAP, and the bortezomib-induced enhancement, primarily mediated by the extrinsic pathway, is further amplified by the derepression of the intrinsic apoptosis pathway.

Next, we examined the combined involvement of Bid and XIAP in bortezomib-dependent rhTRAIL sensitization, hence, under conditions in which the caspase-8/Bid amplification loop is inhibited and caspase-9 and downstream caspases are derepressed. Whereas Bid silencing suppressed rhTRAIL and rhTRAIL/bortezomib induced apoptosis in H460 cells, combined Bid/XIAP depletion restored cell death levels to around 50% and 70%, respectively (Figure 3A, C). Depression of the caspases appears sufficient to trigger effective rhTRAIL-induced apoptosis, even when the Bid-dependent pathway is disrupted. In A549 cells, where bortezomib sensitization was also dependent on the Bid-amplification loop, simultaneous silencing Bid and XIAP resulted in higher levels of apoptosis than seen upon Bid depletion alone. This resembles that of H460 cells, although apoptotic levels are obviously lower in these resistant cells. Silencing of Bid/XIAP in SW1573 cells resulted in similar apoptosis levels as seen in Bid-depleted cells and around 20% lower than in XIAP silenced cells (Figure 3A, B, C). This suggests that although rhTRAIL/bortezomib induced apoptosis in SW1573 is mainly dependent on the extrinsic apoptotic pathway, derepression of caspase-9 and downstream caspases will further increase apoptosis.

We next investigated the possible involvement of cFLIP in regulating rhTRAIL-induced apoptosis and sensitization to bortezomib. cFLIP is a potent inhibitor of extrinsic apoptosis by blocking caspase-8 activation. In H460 cells, silencing of cFLIP markedly increased rhTRAIL-induced apoptosis from around 40% to 70%. Also, upon combined rhTRAIL/bortezomib treatment, apoptosis levels increased from 60% to 85% in cFLIP siRNA-transfected cells (Figure 4A). In SW1573 cells, down-regulation of cFLIP strongly increased rhTRAIL-induced apoptosis from 10% to 50%, indicating that cFLIP is a major determinant of rhTRAIL-resistance in these cells (Figure 4B). Combined rhTRAIL/bortezomib exposure resulted in an additional increase in apoptosis levels from 50% to 75% in cFLIP-suppressed SW1573. Taken together, this indicates that in both NSCLC cells, caspase-8 activation is an important rate limiting step for rhTRAIL-induced apoptosis, which can be overcome by bortezomib. Bortezomib also enhanced apoptosis in cFLIP silenced cells, indicating sensitizing effects also downstream of caspase-8 activation.