Malignant melanoma is the major cause of death from skin cancer, and the incidence has increased significantly the past few decades. If diagnosed early, it can be cured by surgical resection. However, metastatic malignant melanomas are generally drug resistant and have a very poor prognosis.1 The response rate for dacarbazine (DTIC), a chemotherapeutic drug commonly used to treat metastatic malignant melanoma, is less than 20% and the complete response rate is below 5%.2 Hence, there is an obvious need for new therapies, preferably approaches that kill drug resistant cancer cells and not normal cells.
Immunotoxins (ITs) are compounds that are being developed for cancer therapy. They consist of a toxin, derived from either plants or bacteria, which is linked to an antibody. The antibody recognizes cancer specific antigens. Once internalized, the IT induces cell death through 2 different mechanisms; inhibition of protein synthesis and induction of apoptosis.3, 4 We wanted to study the activity of the 9.2.27PE IT and to understand its mechanisms of action in melanoma cells.
A complex interplay between pro- and antiapoptotic molecules ensures the homeostatic balance within the cell. A shift in this balance can lead to apoptosis. Apoptosis may occur via the extrinsic pathway (death-receptor pathway) or the intrinsic pathway (mitochondria-dependent). Both pathways involve activation of caspases, a group of proteolytic enzymes. Caspases are grouped into initiator caspases (no. 1, 2, 4, 8, 9 and 10) that activates the effector caspases (no. 3, 6 and 7). Activation of the effector caspases inactivates PARP, a protein involved in DNA repair mechanisms.5, 6 Caspase-3 and PARP are commonly used as biochemical markers of apoptosis.
We have utilized pseudomonas exotoxin A (PE) chemically conjugated to the antibody 9.2.27, which recognizes the surface molecule HMW-MAA, an antigen expressed in most malignant melanomas and melanoma cell lines. In this study, we investigated the effect of the 9.2.27PE IT on 2 different malignant melanoma cell lines, FEMX and SKMEL-28. We show that 9.2.27PE-induced cell death in FEMX and SKMEL-28 primarily occurs through rapid inhibition of protein synthesis that is succeeded by some cell death associated apoptotic features, e.g., PARP inactivation, although very little caspase-3 activation was detected in the cell lines. DNA fragmentation, another typical but not essential, feature of apoptosis was absent. We chose to study the FEMX cells in more detail as these cells interestingly showed a nontransient hyperpolarization of the mitochondrial membrane after 9.2.27PE treatment.
The high affinity mAb 9.2.27 (obtained from Dr. R. Reisfeld, Scripps Research Institute, La Jolla, CA) recognizes an epitope on the high-molecular weight melanoma-associated antigen (HMW-MAA).7 The antibody was conjugated to PE (obtained from Dr. D. Galloway, University of Ohio, Columbus, OH) by a thioether bond formed with the reagent sulfo-SMCC (Pierce, Rockford, IL) as described previously.8 9.2.27PE was diluted in PBS 0.1% HSA and used in the experiments.
The IT 425.3PE has previously been described.3 It targets the epidermal growth factor receptor (EGFR) and has been shown to induce apoptotic cell death in carcinoma.3 425.3PE was used as a positive control of DNA fragmentation in MA-11 breast cancer cells. All control cells were given PBS 0.1% HSA as vehicle.
The general-caspase inhibitor Z-VAD-FMK, the cathepsin B/L inhibitor Z-FA-FMK, caspase-3 inhibitor II and caspase-8 inhibitor II (Calbiochem, La Jolla, CA) were dissolved in DMSO and lactacystin was dissolved in sterile water (both from Sigma-Aldrich, St. Louis, MO). Anti-caspase-3 was purchased from R&D systems (Minneapolis, MN). Anti-PARP and anti-α-tubulin were purchased from Calbiochem. Anti-lamin B was obtained from Oncogene (La Jolla, CA), anti-livin from Imgenex (San Diego, CA) and anti-Bcl-2 and anti-Mcl-1 from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-Bcl-xl was purchased from Cell Signaling Technology (Danvers, MA), anti-caspase-7 from BD Bioscience (San Jose, CA), anti-survivin and anti-XIAP from R&D systems and anti-β-actin and Staurosporine (STS) from Sigma-Aldrich. LY294002 was from Cell Signaling Technology and 3-methyladenine (3-MA) was from Sigma-Aldrich.
Establishment and characterization of the FEMX malignant melanoma cell line9 and the MA-11 breast cancer cell line has previously been described.10, 11 The SKMEL-28 and the T47D cell lines was obtained from the American Type Culture Collection (Rockville, MD). The 4 cell lines as well as the malignant melanoma cell lines LOX12 and Melmet#1 (in-house cell line) were kept in RPMI-1640 medium supplemented with 8% heat inactivated FCS, Hepes and Glutamax (Gibco, Paisley, UK) at 37°C. The human glioblastoma cell line U87MG was grown in DMEM (Cambrex, Verviers, Belgium) supplemented with 8% heat inactivated FCS and Glutamax at 37°C (100% humidity, 5% CO2, 95% air). All cell lines were routinely tested and found to be free from contamination with Mycoplasma species.
Protein synthesis inhibition assay
The ability of the 9.2.27PE to inhibit protein synthesis in FEMX and SKMEL-28 cells was evaluated by a [3H]-leucine incorporation assay. Depending on the cell type, 2–5 × 104 cells were seeded in 48-well plates (Nunc, Roskilde, Denmark) and kept overnight at 37°C. The medium was refreshed before adding different concentrations of 9.2.27PE for 5 and 22 hr. Protein synthesis inhibition was analyzed as previously described.3 Assays were performed in triplicate and repeated at least twice. Counts per minute (cpm) for treated cells were compared to cpm for untreated cells and reported as percentage leucine incorporation. The control value was set to 100%. To determine if caspases or cathepsins were involved in the decreased protein synthesis observed in the 9.2.27PE-treated FEMX cells, the cells were incubated with the general caspase inhibitor Z-VAD-FMK (50 μM) and/or the cathepsin Z-FA-FMK (50 μM) for 45–60 min prior to 9.2.27PE treatment (22 hr experiment only).
Cell viability of the FEMX, SKMEL-28, LOX, Melmet#1, U87MG and T47D cells after treatment of 9.2.27PE or STS was measured using CellTiter 96®AQueous One Solution Cell Proliferation Assay (MTS assay; Promega, Madison, WI). Cells were seeded in 96-well plates (Falcon, Becton Dickinson, Franklin Lakes, NJ) at 5,000–10,000 cells per well depending on the cell type and the exposure time to 9.2.27PE or STS and kept overnight at 37°C. The medium was refreshed before adding different concentrations of 9.2.27PE or STS (1 μM). FEMX, SKMEL-28, LOX, Melmet#1 and U87MG cells were exposed to 9.2.27PE for 24, 48 and 72 hr (FEMX cells only). Only FEMX cells were exposed to STS for 24 hr. CellTiter reagent was added to the culture and the assay plates were incubated for 1–3 hr at 37°C before absorbance was measured at 490 nm (Wallac Victor2™ 1420 Multilabel Counter). Values for treated cells were compared to those for untreated cells and reported as percentage cell viability. The values were corrected for background absorbance. The assays were performed in triplicate and repeated at least twice.
To determine if caspases, cathepsins or proteasomes were involved in the decreased cell viability in 9.2.27PE-treated FEMX cells, the cells were incubated with the general caspase inhibitor Z-VAD-FMK (50 μM) and/or the cathepsin Z-FA-FMK (50 μM) and/or lactacystin (10 μM) for 45–60 min prior to 9.2.27PE treatment. To rule out the involvement of autophagy, FEMX cells were incubated with the phosphoinositide 3-kinase (PI3 kinase) inhibitors 3-MA (0.5 mM) and LY294002 (10 μM) for 60 min prior to 24 hr treatment with 9.2.27PE. These inhibitors block PI3 kinase and subsequently autophagy.
The activity of caspase-3/7 and caspase-8 in FEMX cells was measured using the Caspase-Glo 3/7 and Capsase-Glo 8 assays from Promega. Ten thousand cells were seeded in 96-well optical bottom plates with white upper structure (Nunc) and kept overnight at 37°C. To determine if caspase-3 and caspase-8 were involved, the cells were incubated with caspase-3 inhibitor II and caspase-8 inhibitor II (50 μM) for 45 min prior to 9.2.27PE treatment (100 ng/ml) or STS (1 μM) treatment. The activity was measured after 24 hr in accordance with the manufacturer's instructions. Assays were performed in triplicate and repeated at least three times.
Mitochondrial membrane potential
The mitochondrial membrane potential, Δψm, was measured using flow cytometry. FEMX and SKMEL-28 cells (0.5–8 × 105 depending on the cell type) were seeded in 25 cm2 flasks and kept overnight at 37°C. The medium was refreshed before treatment with 100 ng/ml 9.2.27PE or 1 μM STS for 5 and 24 hr. Subsequently, both adherent and floating cells were harvested and pelleted by centrifugation. After suspending the cells in 1 ml of medium, 10 μl of 200 μM JC-1 (MitoProbe™ JC-1, Molecular Probes, Eugene, OR) were added and the cells were incubated for 30 min at 37°C. To confirm that the JC-1 response was sensitive to changes in the mitochondrial membrane potential, CCCP was included as a depolarization control. The cells were subsequently washed by adding PBS, pelleted and resuspended in 500 μl of PBS prior to FACS analysis. The LSR II flow cytometer (Becton Dickinson) was programmed to measure forward and side scatter, green and red fluorescence. A change in the mitochondrial membrane potential is indicated by a decrease in the red/green fluorescence intensity ratio (depolarization) or an increase in the same ratio (hyperpolarization). A total of 1 × 104 cells per sample were measured.
Western blot analysis
After treatment of the FEMX and SKMEL-28 cells with 100 ng/ml 9.2.27PE for the indicated time periods, both floating and adherent cells were collected. The amino terminal cell-binding domain of PE is intact in our IT, making it able to bind to the α-macroglobulin receptor on cells. We therefore included as controls free PE (100 ng/ml) and 9.2.27 antibody (100 ng/ml), and treated the FEMX cells for up to 24 hr. Total cellular protein was prepared as previously described.3 Lysates were snap-frozen in liquid N2 and kept at −80°C until used. Protein concentrations were measured using the BCA protein assay (Pierce, Rockford, IL).
Equal amounts of protein (15 μg) were separated by 10% NuPAGE Bis-Tris gel (Invitrogen, Carlsbad, CA), and subsequently transferred by electrophoresis to Immobilon membrane (Millipore, Bedford, MA). The membranes were blocked with 5% nonfat dry milk for at least 1 hr at room temperature followed by incubation with various primary antibodies overnight at 4°C. The membranes were washed before incubation with appropriate HRP-coupled secondary antibodies. Following several washes, the peroxidase activity was visualized with ECL (Pierce) on a Kodak Image Station 2000R.
DNA fragmentation analysis
Two methods were used to analyze DNA fragmentation in FEMX, SKMEL-28 and MA-11 cells. For the TUNEL assay, the 3 cell lines were treated with ITs (FEMX and SKMEL-28 with 100 ng/ml 9.2.27PE and MA-11 with 10 ng/ml 425.3PE). In addition, FEMX and MA-11 cells were treated with the apoptotic inducer STS (1 μM), for up to 48 hr. Both floating and adherent cells were collected and washed twice with PBS before the pellets were fixed in ice-cold MeOH and kept at −20°C. The terminal transferase assay kit (Boehringer Manheim, Manheim, Germany) was used as previously described13 to detect free 3′-OH ends of cleaved DNA. The cells were filtered through Falcon tubes with Cell-Strainer Cap (Becton Dickinson) before analyzed on the FACSVantage SE or LSR II flow cytometer (Becton Dickinson).
To detect DNA fragmentation in FEMX cells after 9.2.27PE treatment (up to 36 hr), both floating and adherent cells were collected and DNA was extracted and subjected agarose gel electrophoresis as previously described.3
Morphological studies of apoptosis of FEMX cells
Apoptotic cells were identified by (i) fluorescence microscopy based on nuclear morphology after staining the cells with Hoechst 33258 and (ii) electron microscopy (EM). For identification of condensed chromatin or fragmented DNA, 9.2.27PE (100 ng/ml) or STS (1 μM) treated FEMX cells were stained with Hoechst 33258 for nuclear staining. MA-11 breast cancer cells treated with either 425.3PE (10 ng/ml) or STS (1 μM) for 24 hr were used as positive control for DNA fragmentation. FEMX cells were seeded in 6-well plates at 5 × 105 cells per well and kept overnight at 37°C. The medium was refreshed before adding 9.2.27PE at a final concentration of 100 ng/ml. After 24 and 48 hr treatment periods, the cells were harvested, washed in PBS and subsequently fixed in ice-cold MeOH. For quantification, the cells were stained with Hoechst 33258, 200 cells of each sample were counted in a blinded manner and the percentage of cells with chromatin condensation was calculated. The experiment was repeated 3 times. For electron microscopy, the FEMX cells were treated with 100ng/ml 9.2.27PE and prepared as described previously.14 The experiment was repeated twice.
9.2.27PE-induced inhibition of protein synthesis and cell viability in FEMX and SKMEL-28 cells
The effect of increasing concentrations of 9.2.27PE on protein synthesis in FEMX cells is shown in Figure 1a. The 9.2.27PE induced a time- and dose-dependent inhibition of protein synthesis with an IC50 value of ∼7 ng/ml after 22 hr. Close to 100% inhibition of protein synthesis was observed after 22 hr treatment with 9.2.27PE (100 ng/ml). In parallel, this dose (100 ng/ml) decreased cell viability by ∼40% and close to ∼90% after 24 and 48 hr, respectively (Fig. 1c). Similar results were obtained after treating the malignant melanoma cell line SKMEL-28 with the same IT (Figs. 1b and 1d). The 9.2.27PE also induced decreased cell viability in the HMW-MAA expressing melanoma cell lines LOX and Melmet#1 and in the glioblastoma cell line U87MG. The decrease in cell viability seen in these cell lines was similar to the decrease seen in the FEMX and the SKMEL-28 cells (data not shown). As the FEMX cells do not express EGFR (data not shown), we used the 425.3PE IT that binds to the EGFR as a negative control. As expected, 425.3PE (100 ng/ml for 24 hr) had no effect on cell viability in the FEMX cells (data not shown). In addition, only minimal decrease in protein synthesis was observed (∼15 %) was observed after treating the HMW-MAA negative T47D breast cancer cells were treated with 100 ng/ml 9.2.27PE for 24 hr. PE and 9.2.27 alone (100 ng/ml) had no effect on the cell viability in FEMX cells (data not shown). 9.2.27PE concentrations ranging from 1 ng/ml to 100 ng/ml were selected for further experiments treatment of the FEMX and the SKMEL-28 cells.
9.2.27PE-induced activation of caspase-3 and PARP inactivation in FEMX and SKMEL-28 cells
We have previously shown that holo-PE-containing ITs induce cell death through both inhibition of protein synthesis as well as apoptosis.3 To determine whether apoptosis was involved in the decreased cell viability observed in FEMX cells, Western blot analysis was performed. The 9.2.27PE induced time-dependent decrease of procaspase-3 (Fig. 2a, top panel). Active caspase-3 fragment was detected after 24 hr, peaking at 32 hr whereafter it decreased. Hence, active caspase-3 did not accumulate over time, suggesting instability of this fragment. The levels of active caspase-3 were very low and long exposure times were needed to detect this protein. A time-dependent decrease of procaspase-7, also starting at 24 hr, was observed, with very low levels detectable after 48 hr. Caspase-7 is another executioner caspase with the ability to cleave PARP.15, 16 Cleavage of PARP, which is typically seen in apoptotic cells,5, 17 started at 24 hr, and after 48 hr PARP was totally inactivated (Fig. 2a, top panel). Similar results were obtained with 9.2.27PE-treatment of SKMEL-28 cells (Fig. 2c). 9.2.27PE induced time-dependent decrease of procaspase-3, the active caspase-3 fragment was detected after 24 hr, peaking at 32 hr whereafter it decreased. Again, long exposure times were needed to detect the p17 active caspase-3 (lower band). Inactivation of PARP was seen after 24 hr. The levels of α-tubulin in the SKMEL-28 9.2.27PE-treated cells were stable and used as loading control for this cell line, in contrast to the decreased levels seen in 9.2.27PE-treated FEMX cells. Taken together, both cell lines showed biochemical signs of apoptosis with late appearance of active caspase-3 (however, very low levels) and inactivation of PARP.
To determine whether caspases or cathepsins were effectors in the reduced cell viability, FEMX cells were incubated with the general caspase inhibitor Z-VAD-FMK and/or the cathepsin inhibitor Z-FA-FMK. At 24 hr the inhibitors did not have a significant effect on the inhibition of protein synthesis (data not shown) or the 9.2.27PE-induced cell death suggesting no/minor involvement of these proteases at this time point. We therefore chose to investigate their effect at the 48 hr in the FEMX cells. Both inhibitors were added 45–60 min prior to addition of 9.2.27PE (10 ng/ml). As shown in Figure 2b, Z-VAD-FMK and Z-FA-FMK both inhibited the 9.2.27PE-induced cell death after 48 hr, although the inhibitory effect was small. The combination of the two inhibitors reduced the 9.2.27PE-induced cell kill in FEMX cells by ∼50%.
We further wanted to investigate the effect of the general caspase inhibitor Z-VAD-FMK and the cathepsin B/L inhibitor Z-FA-FMK at the 48 hr in the FEMX cells using Western blot. The Z-VAD-FMK inhibitor did not prevent processing of procaspase-3 after 48 hr, but the active fragment of caspase-3 was blocked, indicated by a shift upwards of the active caspase-3 band from p17 to p20. The cathepsin B/L inhibitor had no effect on active caspase-3 levels. The combination of the two inhibitors had a slight effect on the levels of procaspase-3, procaspase-7 and active caspase-3, as well as partially inhibiting PARP cleavage (Fig. 2a, top panel). Lamin B, a structural component of the nuclear membrane and a target protein for active caspase-3 and -6,18 was only rarely detectable after 48 hr of 9.2.27PE treatment (Fig. 2a, top panel). Z-VAD-FMK prevented downregulation of Lamin B, more so than Z-FA-FMK, indicating that primarily caspases are responsible for decreased levels of this nuclear matrix protein, as also shown in previous studies.3, 19, 20 As controls, FEMX cells were treated with both free PE and 9.2.27 antibody. Activation of caspase-3 or inactivation of PARP with either 9.2.27 or PE alone was not observed (not shown). Taken together, apoptotic markers like activation of caspases and inactivation of PARP were induced by 9.2.27PE treatment of the FEMX cells, but no strong activation of caspases was observed. Caspases and cathepsins are only partly responsible for the observed decrease in cell viability after 9.2.27PE treatment.
To confirm the low caspase-3 levels detected on Western blot as well as the limited effect seen with the general caspase inhibitor Z-VAD-FMK in FEMX cells, caspase-3 activity assay was utilized. As seen in Figure 3, some caspase-3 activity was observed after 9.2.27PE treatment, but compared to the effect seen in other cell lines,3 the active caspase-3 level was low, thus confirming the Western blot results. To investigate whether the FEMX cells are nonsensitive to caspase-dependent cell death in general or if 9.2.27PE simply is unable to elicit caspase-dependent properties because of its intrinsic properties, the FEMX cells were treated with the apoptotic inducer STS. STS-treatment (1 μM for 24 hr) caused caspase-3/7 activity, but no caspase-8 activity (Fig. 3). In parallel, this dose STS caused 84% ± 6.6% reduction in cell viability after 24 hr.
9.2.27PEs effect on a panel of antiapoptotic molecules, α-tubulin and β-actin in FEMX cells
We further wanted to characterize proteins involved in the 9.2.27PE-induced cell death in FEMX cells. Expression of antiapoptotic proteins of the Bcl-2 family, Bcl-2, Bcl-xl and Mcl-1, and inhibitor of apoptosis (IAP) family members, XIAP, survivin and livin,21–23 was analyzed by Western blot.
XIAP is thought to inhibit apoptosis by binding to caspase-3 and caspase-7.24, 25 9.2.27PE-induced decreased levels of XIAP starting after 24 hr, with very low levels detected at 48 hr (Fig. 2a, central panel). XIAPs inhibitory effect on caspase-3 and -7 would therefore decline in a time-dependent fashion in 9.2.27PE-treated cells. The protein level of survivin decreased rapidly till 24 hr, whereafter the protein level stabilized. We also observed a time-dependent downregulation of livin starting 8 hr after 9.2.27PE treatment likely due to 9.2.27PE-induced inhibition of de novo livin protein synthesis, as livin is an unstable protein with a reported half-life of less than 4 hr.26 The Z-VAD-FMK and the Z-FA-FMK did not inhibit the downregulation of these protein levels significantly.
A slight downregulation of the Bcl-xl levels was observed at 24 hr. Similar to survivin, the Bcl-xl levels were stabilized after 24 hr. The decrease in Bcl-xl levels was inhibited by Z-VAD-FMK that indicates that caspases either directly or indirectly have an effect on Bcl-xl. In contrast, the levels of Bcl-2 increased slightly during 9.2.27PE treatment. Bcl-2 was cleaved after 32 and 48 hr of IT treatment, indicated by a 4 kDa smaller Bcl-2 fragment. Bcl-2 has a cleavage site for caspase-3 between amino acid 34 and 35 whereby the BH4 domain is cleaved off.27, 28 Studies have shown that the cleaved BH4 domain facilitates activation of additional proteases leading to cell death similar to Bax, a pro- apoptotic molecule.28 The general caspase inhibitor Z-VAD-FMK prevented cleavage of the BH4 domain, suggesting that caspases was responsible for cleavage of Bcl-2, as previously shown by others.28
The cytoskeleton consists of α-tubulin and β-actin among other proteins. As seen in Figure 2a, lower panel, the protein levels of both α-tubulin and β-actin were significantly decreased after 48 hr of 9.2.27PE treatment in FEMX cells. The decrease in levels of both the α-tubulin and the β-actin were inhibited by the combination of Z-VAD-FMK and Z-FA-FMK, indicating that caspases and cathepsins were involved. α-tubulin and β-actin are commonly used as loading control for Western blot. Because both α-tubulin and β-actin protein levels decreased during 9.2.27PE treatment, Bcl-2 protein levels were chosen as loading control in FEMX cells, as this was the most stable protein during 9.2.27PE treatment in this experiment.
At the 48 hr time point, all of the antiapoptotic protein levels investigated (except Bcl-2) as well as α-tubulin and β-actin levels were decreased. This is probably a consequence of inhibition of protein synthesis and the inevitable cell collapse resulting in cell death.
9.2.27PE did not induce depolarization of the mitochondrial membrane potential in FEMX and SKMEL-28 cells
Mcl-1 is mainly localized to the mitochondrial membrane29 and loss of Mcl-1 has been shown to be a key factor for the induction of apoptosis.3, 30 We have previously shown that another IT, 425.3PE, induced rapid downregulation of Mcl-1 protein expression in breast cancer MA-11 cells, and that this downregulation contributed to caspase-3 activation and PARP inactivation.3 As seen in Figure 4a, the protein level of Mcl-1 was noticeably decreased after 8 hr and not detectable after 24 hr of 9.2.27PE treatment, likely due to inhibition of de novo Mcl-1 protein synthesis and the rapid turnover rate of Mcl-1.31 Lactacystin inhibited the rapid turnover of Mcl-1 protein (8 hr time point) suggesting that Mcl-1 degradation takes place in the proteasomes as previously reported.3, 32 In the FEMX cell line, decreased Mcl-1 protein level was observed simultaneously with hyperpolarization of the mitochondrial membrane potential (Fig. 4b). We therefore speculate whether Mcl-1 might not be a key factor responsible for the induction of 9.2.27PE-induced cell death in FEMX cells. On the other hand, the rapid decrease of this antiapoptotic molecule might shift the homeostatic balance in favor of the proapoptotic molecules and induction of apoptosis.
Mitochondrial depolarization is often linked to apoptosis as it allows the release of proapoptotic molecules from the intermembrane space, which ultimately can lead to caspase-3 activation.33 To detect whether the mitochondrial membrane potential (Δψm) was affected by 9.2.27PE treatment of FEMX and SKMEL-28 cells, we utilized the fluorescent cation dye JC-1. JC-1 produces red-orange fluorescence in intact mitochondria, whereas mitochondrial depolarization causes JC-1 to accumulate in the cytoplasm where green fluorescence is produced. Thus, the higher the Δψm the more JC-1 accumulates in the mitochondrial matrix. Figure 4b illustrates the changes in the Δψm following 9.2.27PE treatment in FEMX cells. The quadrants were arbitrarily defined to highlight the changes in the red-green fluorescence, and the numbers in each quadrant represents the percentage of cells in that quadrant. The control represents normal Δψm in FEMX cells. 9.2.27PE induced a time-dependent increase in membrane potential indicated by an increase in percentage of red fluorescence in the upper two quadrants from 22% in the control cells to 55% after 24 hr of 9.2.27PE treatment. Less fluctuations were seen in green fluorescence (the 2 right quadrants) indicating that the JC-1 dye accumulated in the mitochondria. To exclude involvement of the death receptor/extrinsic apoptotic pathway through caspase-8 and BID/tBID effect on the mitochondrial membrane, caspase-8 activity was measured (Fig. 3). Practically no caspase-8 activity was detected, ruling out the involvement of the death receptor/extrinsic apoptotic pathway in 9.2.27PE-treated FEMX cells. Treating the SKMEL-28 cells with 9.2.27PE did not have a depolarization effect on the mitochondrial membrane. The increase in red fluorescence was close to zero after 5 and 24 hr, respectively. STS-treated FEMX cells showed some decrease in red fluorescence after 5 hr, indicating a slight depolarization of the mitochondrial membrane. Interestingly, after 24 hr there was a distinct increase in the red fluorescence intensity indicating hyperpolarization of the mitochondria.(date not shown). Hence, neither FEMX nor SKMEL-28 showed significant depolarization of the mitochondrial membrane during treatment, indicating that proapoptotic molecules like cytochrome c remain within the mitochondria and do not activate caspases.
9.2.27PE-induced chromatin condensation in FEMX cells
To detect whether 9.2.27PE-induced chromatin condensation in FEMX cells, the cells were fixed in MeOH and subsequently stained with Hoechst 33258 (Fig. 5). STS was used as a positive control in this experiment as it has been shown to induce apoptosis in other cell lines.34, 35 Untreated cells showed an even staining of the nuclear chromatin in FEMX cells. Both 9.2.27PE and STS treatment caused chromatin condensation. No apoptotic bodies were observed reflecting that no DNA fragmentation occurred. This is in contrast to the effect of 425.3PE IT on MA-11 breast cancer cells, where DNA fragmentation was induced.3 We therefore used STS and 425.3PE-treated MA-11 cell as positive control for DNA fragmentation. STS and 425.3PE treatment caused fragments of condensated chromatin compatible with DNA fragmentation in MA-11 cells (data not shown). To quantify the percentage of FEMX cells affected by 9.2.27PE treatment, the number of cells showing an even staining of Hoechst, as well as cells with chromatin condensation, was counted. As seen in Figure 5a, only 2% of the control cells showed chromatin condensation, compared to 21% and 82% after 9.2.27PE-treatment for 24 and 48 hr, respectively. STS (1 μM) induced chromatin condensation in 94% of the cells after 24 hr. No apoptotic bodies were observed. EM analysis of untreated and 9.2.27PE-treated FEMX cells was performed. Chromatin condensation was observed after 24 and 32 hr of 9.2.27PE treatment indicated by arrows (Fig. 5b). No apoptotic bodies were observed.
9.2.27PE-induced no DNA fragmentation in FEMX
DNA fragmentation was analyzed by TUNEL assay, where increased FITC staining indicates increased DNA fragmentation. As seen in Figure 6a, the FEMX cells did not show increased FITC intensity after 24 and 48 hr 9.2.27PE treatment. When treating FEMX cells with STS for 24 hr similar results were obtained, but after 48 hr treatment with STS a slight increase in FITC intensity was observed, consistent with increased DNA fragmentation. However, treatment with 1 μM STS for 48 hr is long compared to other studies where equal concentration of STS induced apoptosis and DNA fragmentation already after 2 and 8 hr.36, 37 The DNA fragmentation observed might be a result of disintegration of the cell and DNA, and not a direct effect of apoptosis.
SKMEL-28 cells showed low levels of DNA fragmentation after 24 hr and 48 hr treatment with 9.2.27PE (100 ng/ml), 8% and 12%, respectively (Fig. 6b). As previously shown for MA-11 breast cancer cells, treatment with 425.3PE results in DNA fragmentation.3 425.3PE and STS treatment of MA-11 cells used as a positive control resulted in a distinct shift in FITC intensity, indicating massive DNA fragmentation (Figs. 6a and 6b). DNA fragmentation was also analyzed by agarose gel electrophoresis, but no DNA fragmentation of 9.2.27PE-treated FEMX cells (up to 36 hr) were detected (data not shown), confirming the TUNEL assay results.
The response rate for malignant melanoma to the standard chemotherapeutic drug dacarbazine is very low,1 and hence, there is an obvious need for new therapies. These approaches should be highly effective and cancer cell specific. We have previously shown that 425.3PE and MOC31PE that binds to different antigens, kill breast cancer cells through inhibition of protein synthesis and induction of apoptosis.3 The 9.2.27PE IT that targets the HMW-MAA expressed on e.g., malignant melanoma cells is interesting to study, also for potential treatment of malignant melanoma in the future. Other groups have shown that a majority of the melanoma cell lines tested bound 9.2.27, whereas colon and ovarian carcinomas, LCL (Epstein Barr Virus immortalized B cell lines) and fibroblasts did not.38, 39 In addition, the HMW-MAA antigen was detected only in fetal skin and in one nipple of 54 normal tissues from adults, but among malignant lesions the HMW-MAA was expressed in melanomas, astrocytomas and skin carcinomas.40 In addition, 9.2.27PE induces decreased cell viability invitro in the glioblastoma cell line U87MG (data not shown) and in vivo.41 These data show that the HMW-MAA is expressed in several different malignant cells and can be used as a target for 9.2.27PE IT.
In this study, we investigated the underlying mechanisms for the 9.2.27PE IT-induced toxicity in the DTIC sensitive FEMX and the DTIC resistant SKMEL-28 malignant melanoma cells,42 and found that this IT efficiently induced cell death primarily through inhibition of protein synthesis, followed by some morphological and biochemical features of apoptosis.
The general caspase inhibitor Z-VAD-FMK and the cathepsin B/L inhibitor Z-FA-FMK reduced the 9.2.27PE-induced cell kill in FEMX cells by ∼50% (Fig. 2b). In addition, their effect on several proteins involved in apoptosis was moderate. The inhibitors did not have any effect on the 9.2.27PE-induced inhibition of protein synthesis. This indicates that 9.2.27PE-induced FEMX cell death initially occurs through inhibition of protein synthesis. Apoptotic features were observed late in the cell death process, in contrast to what was previously seen for another PE-based IT in carcinoma cells.3 For 9.2.27PE, autophagy and necrosis may be excluded as cell death mechanisms, as EM showed chromatin condensation but no cell swelling.43 In addition, 3-MA and LY294002, substances that block phosphoinositide 3-kinase (PI3 kinase) and subsequently autophagy, did not inhibit the cell kill induced by 9.2.27PE (not shown).
Two major caspase-dependent apoptotic pathways have been described in eukaryotic cells: extrinsic and intrinsic. The intrinsic/mitochondrial pathway is induced by cellular stress (e.g., DNA damage, ER stress).44 The stress signals converge on the mitochondria to induce mitochondrial membrane permeabilization, allowing release of cytochrome c which subsequently triggers caspase-9 and caspase-3 activation.33, 45 Cell death is therefore most commonly associated with the collapse of the mitochondrial membrane potential (Δψm).46 Notably, increased Δψm has been linked to survival signals, and the hyperpolarization of the mitochondrial membrane usually precedes caspase activation and total collapse of Δψm.47, 48 Even though we observed increased Δψm in FEMX cells, which suggests that cytochrome c is retained within the mitochondria, this increase was not transient as has been seen in other model systems.47, 48 9.2.27PE treatment of SKMEL-28 did not result in depolarization of the mitochondrial membrane potential. This could imply that these malignant melanoma cells upon mitochondrial stress exert a survival function through hyperpolarization/no depolarization of the mitochondrial membrane preventing the release of proapoptotic molecules like cytochrome c. The 9.2.27PE-induced apoptosis is therefore most likely neither activated via the intrinsic pathway (no depolarization) nor through the extrinsic pathway (no caspase-8 activity). These results indicate that the 9.2.27PE-induced cell death primarily occurs through inhibition of protein synthesis. STS has been shown to induce release of cytochrome c in malignant cells.37, 49 In FEMX cells, STS initially induced a slight depolarization followed by hyperpolarization of the mitochondria. Depolarization leads to release of proapoptotic molecules and subsequently activation of caspases. The caspase-3/7 activity observed after STS treatment is possibly an effect of the early depolarization of the mitochondrial membrane in the FEMX cells. The results indicate that the FEMX cells are able to undergo caspase dependent cell death, and that STS induce different mechanisms leading to cell death than the 9.2.27PE IT. Clearly, the FEMX cells have strong intrinsic apoptotic resistance mechanisms, as the cells need to be treated for 24 hr with STS for maximum decrease in cell viability, whereas others have shown that STS induces apoptosis after 2–8 hr in other cell lines.36, 37 More interestingly, at this time point (24 hr) the cells show hyperpolarization of the mitochondria.
The level of Bcl-2 was not significantly altered after 9.2.27PE treatment of FEMX cells. Bcl-2 and its family members regulate the mitochondrial membrane integrity,50 and subsequently the release of e.g., cytochrome c. Bcl-2 might therefore play an important role in maintaining the Δψm in the 9.2.27PE-treated FEMX cells, possibly counteracting the decreased protein levels of other antiapoptotic members like Bcl-xl and Mcl-1. Bcl-2 levels did not decrease even after 48 hr of treatment. At this time point, the levels of the other proteins studied were decreased or inactivated (PARP), and close to 100% of the cells showed no viability. The importance of Bcl-2 in 9.2.27PE-induced cell death will be further studied.
The IAP family members XIAP, livin and survivin have all been shown to be overexpressed in malignant melanoma.51–53 These molecules interact with caspase-3 and through this interaction caspase-activation is inhibited with subsequent inhibition of apoptosis. Others have shown that decreased protein levels of IAPs results in induction of cell death,54–56 and that high expression levels of these proteins are associated with advanced disease states and negative prognostic factors in many types of cancer.57, 58 Despite decreased protein levels of XIAP, livin and survivin after 9.2.27PE treatment, no strong caspase-3 activity was observed in FEMX cells.
Chromatin condensation and DNA fragmentation are respectively early and relatively late apoptotic effects on the nucleus compartment. 9.2.27PE-induced cell death in the FEMX cells revealed chromatin condensation, but no DNA fragmentation or formation of apoptotic bodies. Activated caspases (e.g., active caspase-3 and -7) indirectly activate CAD (caspase-activated deoxyribonuclease) that is involved in DNA fragmentation.20, 59, 60 The active caspase-3 protein levels were low in 9.2.27PE-treated FEMX cells, which might explain why no DNA fragmentation was detected. However, others have shown that cells may be able to undergo apoptosis, even though no DNA fragmentation takes place.43, 60
In conclusion, we show that 9.2.27PE-induced malignant melanoma cell death primarily occurs through inhibition of protein synthesis that is succeeded by cell death associated apoptotic features. Although only low levels of caspase-3 activation was detected, other observed aspects of 9.2.27PE-induced FEMX cell death were rounding up of the cells (data not shown), chromatin condensation and inactivation of PARP, all compatible with apoptosis. Interestingly, the lack of strong activation of caspases is likely due to the hyperpolarization (FEMX cells) or no depolarization (SKMEL-28 cells) of the mitochondrial membrane after IT treatment. This suggests that cytochrome c, a molecule that triggers activation of caspase-3, was retained within the mitochondria. Studies are underway to examine whether this effect is linked to the high intrinsic resistance against apoptosis in malignant melanomas. Nevertheless, 9.2.27PE IT successfully killed malignant melanoma cells, which can be ascribed to inhibition of protein synthesis followed by some morphological and biochemical features of apoptosis.
We thank Dr. Q. Peng, Ms. E. Hellesylt and Dr. A. Brech for execution and discussion of EM photos. We also thank Ms. I.J. Guldvik and Dr. B. Engesæther for help with experiments.