Int. J. Cancer: 119, 493–500 (2006) ' 2006 Wiley-Liss, Inc.
FAST TRACK The antimicrobial peptide, Lactoferricin B, is cytotoxic to neuroblastoma cells in vitro and inhibits xenograft growth in vivo Liv Tone Eliassen1, Gerd Berge1, Arild Leknessund2, Mari Wikman1, Inger Lindin1, Cecilie Løkke2, Frida Ponthan3, John Inge Johnsen3, Baldur Sveinbjørnsson4, Per Kogner3, Trond Flægstad2 and Øystein Rekdal1* 1 Department of Biochemistry, Faculty of Medicine, University of Tromsø, Tromsø, Norway 2 Department of Pediatrics, Faculty of Medicine, University of Tromsø, Tromsø, Norway 3 Childhood Cancer Research Unit, Department of Woman and Child Health, Karolinska Institutet, Stockholm, Sweden 4 Department of Experimental Pathology, Faculty of Medicine, University of Tromsø, Tromsø, Norway Antimicrobial peptides have been shown to exert cytotoxic activity towards cancer cells through their ability to interact with negatively charged cell membranes. In this study the cytotoxic effect of the antimicrobial peptide, LfcinB was tested in a panel of human neuroblastoma cell lines. LfcinB displayed a selective cytotoxic activity against both MYCN-amplified and non-MYCN-amplified cell lines. Non-transformed fibroblasts were not substantially affected by LfcinB. Treatment of neuroblastoma cells with LfcinB induced rapid destabilization of the cytoplasmic membrane and formation of membrane blebs. Depolarization of the mitochondria membranes and irreversible changes in the mitochondria morphology was also evident. Immuno- and fluorescence-labeled LfcinB revealed that the peptide co-localized with mitochondria. Furthermore, treatment of neuroblastoma cells with LfcinB induced cleavage of caspase-6, -7 and -9 followed by cell death. However, neither addition of the pan-caspase inhibitor, zVADfmk, or specific caspase inhibitors could reverse the cytotoxic effect induced by LfcinB. Treatment of established SH-SY-5Y neuroblastoma xenografts with repeated injections of LfcinB resulted in significant tumor growth inhibition. These results revealed a selective destabilizing effect of LfcinB on two important targets in the neuroblastoma cells, the cytoplasmic- and the mitochondria membrane. ' 2006 Wiley-Liss, Inc. Key words: bovine lactoferricin; neuroblastoma; mitochondria; cell death
Neuroblastoma, an embryonic tumor of the sympathetic nervous system, is the most common neoplasia during infancy. Neuroblastoma patients over 1 year of age with metastatic disease and those with MYCN-amplified tumors continue to have poor prognosis and often develop resistance to conventional therapy.1 Alternative treatment options for these patients are therefore urgently needed. Similar to prokaryotic cells, many cancer cells have a higher content of anionic phospholipids on their outer leaflet relative to normal eukaryotic cells. Several reports have demonstrated that antimicrobial peptides induce cell death of transformed cells, but are much less cytotoxic to non-transformed cells.2–10 Anticancer activity against solid tumors has been demonstrated after local treatment with antimicrobial peptides.4,5 Additionally, a cytotoxic peptide modified to resist enzymatic degradation was recently shown to inhibit lung metastasis in mice without detectable side effects.11 These results suggest that such peptides may have future value as anticancer drugs either alone or in combination with conventional therapies. The mechanisms of action of antimicrobial peptides against prokaryotic cells have been widely investigated,12–15 while much less is known regarding the interactions with and effects on eukaryotic cells. Pro-apoptotic effects induced by antimicrobial peptides that are conjugated to a functional domain, which allow receptor-mediated or receptor-independent internalization in eukaryotic cells, have been reported.16,17 After internalization these antimicrobial peptide conjugates caused mitochondrial membrane disruption resulting in cytochrome c release and induction of apoptosis.16,17 Other types of antimicrobial peptides translocate sponPublication of the International Union Against Cancer
taneously into the cell cytoplasm where they depolarize the inner mitochondrial membranes. Whether the mitochondria in intact cells are disrupted as a result of a direct effect of the peptides is unknown.18,19 The antimicrobial peptide, bovine lactoferricin (LfcinB), is a 25 amino acid long highly basic peptide with a disulfide bridge between two cysteines, thus giving it a cyclic twisted antiparallel b-sheet solution structure.20 LfcinB has been shown to inhibit liver and lung metastasis of both murine melanomas and lymphomas21 and to induce apoptosis in human leukemic and carcinoma cell lines.22,23 Recently, our group demonstrated that LfcinB inhibited the growth of solid murine Meth A sarcomas in vitro. Meth A cells were killed by LfcinB as a result of acute swelling, bursting and cell disruption, mechanisms that are frequently associated with necrosis.5 Here, we report that LfcinB displays a toxic effect against both MYCN-amplified and non-MYCN-amplified NB cell lines in vitro. LfcinB peptides were internalized into the cytoplasm, followed by a depolarization of the mitochondria membrane and activation of caspase-6, -7 and -9. LfcinB colocalized with small membrane blebs, indicating a direct destabilizing effect on the cytoplasmic membrane of the NB cells. However, since inhibitors of caspases failed to inhibit the cytotoxic activity of LfcinB, we conclude that even though both apoptotic and necrotic processes are observed by LfcinB treatment, a direct destabilization of the cytoplasmic membrane and a collapse of the outer and inner mitochondrial membranes are the main events responsible for the cytotoxic effect by LfcinB. Additionally, the growth of NB xenografts in nude rats was significantly reduced after LfcinB treatment. Material and methods Reagents The LfcinB was prepared by pepsin digestion of bovine lactoferrin, by the Centre for Food Technology (Queensland, Australia). Mouse polyclonal anti-LfcinB17–41 antibody was purchased from Pharos SA (Eurogentech, Belgium). 3-(4,5-Dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide (MTT), carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP), culture media and cold-water fish gelatine were obtained from Sigma-Aldrich AS (Oslo, Norway). Fetal bovine serum (FBS) was purchased from Abbreviations: ATCC, American type culture collection; DWm, mitochondrial membrane potential; FCCP, carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone; FSG, fish skin gelatine; IFN-g, interferon-g; JC1, 5,50 ,6,60 -tetrachloro-1,10 ,3,30 -tetraethylbenzimidazolecarbocyanine iodide; LfcinB, bovine lactoferricin; MDR, multiple drug resistance; MTT, 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NB, neuroblastoma; PI, propidium iodide; RP-HPLC, reversed phase high performance liquid chromatography; TRAIL, TNF-related apoptosis-inducing factor. *Correspondence to: Department of Biochemistry, Faculty of Medicine, University of Tromsø, Tromsø, Norway. Fax: 147-776-45350. E-mail:
[email protected] Received 29 September 2005; Accepted 19 January 2006 DOI 10.1002/ijc.21886 Published online 29 March 2006 in Wiley InterScience (www.interscience. wiley.com).
494
ELIASSEN ET AL.
Biochrom, KG (Berlin, Germany). Monoclonal Anti-rabbit caspase-3, caspase-6, caspase-7, caspase-8 and caspase-9 antibodies were purchased from Cell Signaling Inc. (Abingdon, UK). A mouse-isotype control was obtained from Zymed (San Francisco, CA). 5,50 ,6,60 -Tetrachloro-1,10 ,3,30 -tetraethylbenzimidazolecarbocyanine iodide (JC-1) probe and MitoTracker RedTM CM-H2XRos (M7513) were from Molecular Probes (Leiden, The Netherlands). Propidium iodide (PI) was obtained from Oncogene (San Diego), and the caspase-3 and -7 inhibitor zDEVD-fmk (FMK004), caspase-6 inhibitor zVEID-fmk (FMK006), caspase-9 inhibitor zLEHDfmk (FMK008) and the general caspase inhibitor zVAD-fmk (FMK001) were from R&D Systems Inc. (Abingdon, UK). Cell cultures The neuroblastoma cell lines Kelly, SK-N-DZ and IMR-32 were cultured in RPMI-1640, and SHEP-1 was grown in HAM’s F12:DMEM (1:1) with 1% nonessential amino acids and SH-SY5Y was grown in HAM’s F12 containing 1% nonessential amino acids. The human embryonic fibroblast cell line MRC-5 was maintained in MEM medium. All media used were without antibiotics, but supplemented with 10% FBS and 1% L-glutamine. The cell cultures were grown in a humidified 5% CO2 atmosphere at 37°C. All cell lines were regularly tested for the presence of Mycoplasma (ATCC Mycoplasma Detection Kit). Peptide synthesis, purification, analysis, cyclization and fluorescence-labeling Linear LfcinB was synthesized by solid-phase methods using standard Fmoc chemistry with a Pioneer Peptide synthesizer (Applied Biosystem, Foster City, CA), purified and analyzed as previously described.24 Cyclization of linear LfcinB labeled with 5-carboxyfluorescein (Novabiochem, Merck, KgaA, Darmstat, Germany) in the N-terminus was catalyzed by using the Ekathiox resin as described by the manufacturer (Sigma-Aldrich). In vitro cytotoxicity NB cells and fibroblasts were incubated with LfcinB or caspase inhibitors in concentrations as indicated and tested for cell viability using the MTT assay.25 The final results were recorded by averaging at least 6 parallel experiments. Flow cytometric detection of changes in the mitochondrial membrane potential Changes in the mitochondrial membrane potential (DWm) was evaluated using the lipophilic and cationic JC-1 dye on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA). Briefly, Kelly cells were seeded in complete medium in 12-well plates at a concentration of 1 3 105 cells/well and grown until a concentration of 3 3 105 cells/well. The cells were stained with 1.5 lg/ml JC-1 dissolved in complete medium and incubated for 20 min in darkness at room temperature, washed twice in PBS and incubated with LfcinB or linear LfcinB (40 lg/ml) dissolved in PBS. Parallel experiments, in which cells were stained with PI (10 lg/ml), incubated with LfcinB (40 g/ml) and analyzed in the flow cytometer showed that a time scale of 20 min was optimal for investigating cell viability. FCCP (50 ng/ml) was used as positive control. A gate was set on the forward scatter-FL-3 dot-plot for each sample to exclude all cells stained with PI. Samples of 10,000 cells were examined on a FL-1 (530 nm) versus FL-2 (590) dot plot. The changes in DWm were monitored with JC-1 for 20 min of LfcinB exposure in the flow cytometer, using the Cell Quest Pro software. FACS analysis was used to quantify apoptosis and counting of DAP1-stained nuclei was performed as described earlier.26 Confocal microscopy Laser scanning confocal microscopy was used to visualize changes in the DWm in Kelly cells exposed to LfcinB. Kelly cells grown in 8-well chamber slides (Nunc) were incubated with JC-1 (1.5 lg/ml, 300 ll/well) for 10 min in darkness and washed twice
with PBS. After addition of LfcinB (40 lg/ml), the fluorescence of JC-1 dye was immediately monitored. FCCP (50 ng/ml) was used as positive control and the inactive linear LfcinB peptide (40 lg/ ml) was used as negative control. Confocal microscopy studies of changes in the mitochondrial membrane potential images were collected after 5, 10, 15 and 20 min using a Zeiss Axiovert 200 microscope equipped with a LSM 510 confocal module and processed using the Adobe Photoshop software. Transmission electron microscopy (TEM) Kelly cells incubated with LfcinB (40 lg/ml) for different time points (5 and 15 min, 1, 3, 6, 12 and 24 h) were harvested and aspirated and fixed in Karnovsky’s fixative overnight, followed by post-fixation, dehydration and embedding in Epon-araldite according to standard procedures.27 Ultrathin sections were prepared and examined on a JEOL JEM-1010 transmission electron microscope (Tokyo, Japan). Immunolabeling Kelly cells were harvested, aspirated and incubated with LfcinB peptides diluted in serum-free medium (100 ll, 40 lg/ml) for different time periods (5 and 15 min, 1 and 3 h). Cell samples were fixed and prepared for cryosection and labeled as described earlier.28 To check any nonspecific binding during the labeling procedure, sections of cell samples without LfcinB were immunolabeled, and the unspecific binding of protein A-gold was checked by excluding the primary Ab and by use of mouse IgG isotype control. Immunoblotting and FACS analysis Protein was extracted from the LfcinB-treated (40 lg/ml) cells and the protein content of each lysate was measured, separated and probed with specific Ab against the different caspases as previously described.29 Caspase inhibitors (200 lM) were added 3 h before LfcinB (40 lg/ml) treatment. Quantification of apoptosis was performed by counting DAPI-stained nuclei, using a fluorescence microscope. DNA content was assessed by FACS analysis as previously described.26 Cellular imaging of fluorescence-labeled LfcinB Cells were incubated with red MitoTracker (250 nM) diluted in complete medium for 15 min at 37°C, washed twice with PBS and incubated with 40 lg/ml fluorescence-labeled LfcinB at different time periods (2, 4, 6 and 8 min). The cells were washed twice with PBS and fixed with 4% formaldehyde for 10 min on ice. After a second wash with ice-cold PBS, the cells were investigated by confocal microscopy. Xenografting Xenografting of SH-SY-5Y cells in nude rats (HsdHan, RNUrnu; Harlan Netherlands) was performed as described earlier.26,30 The rats were randomized and LfcinB treatment was started when a tumor in an animal had reached a volume of 0.3 ml (chosen as day 1). Seven tumors were randomized to each group. Group A was chosen as control group and given 0.1 ml physiological saline into the tumor, group B was given 1 mg LfcinB dissolved in physiologic saline in the tumor and group C achieved 2 mg LfcinB. The animals were given a single injection on days 1, 2 and 3. Statistical analysis was performed using the Student’s t-test for 2 independent samples. All p-values refer to a two-sided probability unless stated otherwise. Results Cytotoxic effects of LfcinB on neuroblastoma cells LfcinB was tested against MYCN-amplified NB cell lines (Kelly, IMR-32 and SK-N-DZ) and non-MYCN-amplified cell lines (SHEP-1 and SH-SY-5Y) (Table I). The effect was most prominent against the MYCN-amplified Kelly cell line (IC50 5 15.5 lM), with a 2- and 3-folds lower activity against IMR-32 and SK-N-DZ, respectively, while SH-SY-5Y and SHEP-1 were less
LACTOFERRICIN B IS CYTOTOXIC TO NEUROBLASTOMA CELLS IN VITRO AND INHIBITS XENOGRAFT GROWTH IN VIVO
495
TABLE I – THE CYTOTOXIC ACTIVITY OF LFCINB AGAINST NEUROBLASTOMA CELL LINES AND NORMAL FIBROBLASTS Cell lines
LfcinB IC501 (lM)
Kelly IMR-32 SK-N-DZ SHEP-12 SH-SY-5Y2 Fibroblasts (MRC-5)
15.5 29 37 45 60 >1603
1 The peptide concentration killing 50% of the cancer cells.–2NonMYCN-amplified neuroblastoma cell lines.–3The maximum concentration of the peptides tested was 500 lg/ml (160 lM).
sensitive. The specificity for the NB cell lines was more than 11fold higher than that for the human fibroblasts (MRC-5) used as control cells. The linear LfcinB did not display any activity against the cell lines when tested up to a concentration of 500 lg/ ml (160 lM) (data not shown). LfcinB affects the mitochondrial membrane potential To investigate if the cytotoxic effects of LfcinB were related to changes in the mitrochondrial transmembrane potential, the intensity and shift of fluorescence emission of the JC-1 dye was monitored by flow cytometry and confocal microscopy. Kelly cells incubated with JC-1 were treated with 40 lg/ml LfcinB and immediately monitored in the flow cytometer for a period of 20 min. In parallel experiments, in which Kelly cells were stained with PI, most of the cells were viable in this time period. The results illustrated in Figure 1 (upper panel) show a marked decline in red fluorescence (FL-2) and an increase in green fluorescence (FL-1) after 20 min of LfcinB exposure, relative to untreated cells used as negative control (Figs. 1a and 1b), indicating depolarization of the mitochondrial membrane. FCCP was used as positive control and acted similar to LfcinB, inducing a clear shift in the color of the JC-1 dye from red to green (Fig. 1c). The inactive linear LfcinB analogue was used as negative control and showed a general decline in both red and green fluorescence (data not shown). The confocal microscopy studies showed a punctuated red fluorescence pattern in JC-1-labeled Kelly cells, which is consistent with a mitochondrial distribution. After 20 min treatment with 40 lg/ml LfcinB, a decrease in the intensity of the red fluorescent spots and an increase in green fluorescence intensity in the cytoplasm were evident (Figs. 1d and 1e). FCCP induced a rapid decrease in the DWm, which was shown as a clear increase in green fluorescence (Figs. 1f and 1g). In non-treated Kelly cells or Kelly cells exposed to linear LfcinB, both incubated with JC-1, an expected decrease in both the red and green fluorescence was evident (data not shown). Taken together, the results from the flow cytometry and confocal studies indicate that LfcinB induced a rapid decrease in the mitochondrial DWm in the Kelly cells, suggesting that the mitochondria are one of the targets of LfcinB. LfcinB induces alteration in mitochondrial morphology Kelly cells treated with cyclic LfcinB (40 lg/ml) for different time periods (5 and 15 min, 1, 3, 6, 12 and 24 h) were prepared and studied by TEM (Fig. 2). Moderate swelling of the mitochondria was observed after 1 h (data not shown), whereas severe parallelization of the mitochondria inner membrane and vacuolization and swelling of the mitochondria were evident after 3 h compared to untreated cells (Figs. 2a and 2b). A fraction of cells displaying typical necrotic morphological features such as swelling and blebbing appeared after 6 h of peptide exposure (data not shown). After 12 h of LfcinB exposure, the micrographs showed that a relatively large fraction of cells had undergone necrosis compared to untreated control cells (Figs. 2c and 2d). Both FACS analyses and Dapi-staining did not reveal any significant apoptosis in LfcinBtreated Kelly cells (data not shown).
FIGURE 1 – LfcinB-induced change of the mitochondrial membrane potential in neuroblastoma cells in situ. The upper panel shows a three-dimensional representation of changes in fluorescence intensity of Kelly cells stained with JC-1. (a) Untreated cells, (b) 40 lg/ml LfcinB and (c) 50 ng/ml FCCP. The change in DWm is indicated by a decrease in red fluorescence intensity (FL-2) and an increase in green fluorescence (FL-1) as JC-1 aggregates emitting red fluorescence leakage into the cytoplasm forming monomers, which emit green fluorescence. The distribution of the cells in the green and red fluorescence channels is represented by the z-axis. The effects of LfcinB on Kelly cells probed with JC-1 dye imaged in the confocal microscope is represented in the lower panel, showing cells in which the mitochondria are seen as red dots representing JC-1 aggregates as seen in detail in upper left corner (d). The cytoplasm is pale green and represents JC-1 monomers. In cells being exposed to the peptide for 20 min (e), a decline in the red fluorescence was observed as a result of progressive unloading of JC-1 from the depolarized mitochondria, thus resulting in a brighter green cell cytoplasm. (g) The effect of the positive control FCCP on JC-1 treated Kelly cells. (f) Cells before addition of FCCP.
LfcinB triggers caspase activation in neuroblastoma cells Western-blot analysis of protein extracts indicated that LfcinBtreated Kelly cells expressed activated caspase-6, -7 and-9. A small amount of cleaved caspase-9 was detected after 30 min and higher amounts were evident with prolonged incubation of LfcinB (Fig. 3). The activation of caspase-6 and -7 were first detected after 4 h (Fig. 3). No activation of caspase-3 was detected, and procaspase-8 was not detected in Kelly cells (data not shown). Incubation of NB cells with the pan-caspase inhibitor zVAD-fmk and the specific inhibitors for caspase-3, -6, -7 and -9 did not influence the cell death induced by LfcinB (data not shown). Internalization of LfcinB and colocalization with the mitochondria The localization of LfcinB in Kelly cells was studied using both immunolabeled and fluorescence-labeled peptides. Cells treated
496
ELIASSEN ET AL.
FIGURE 2 – TEM micrographs showing differences in neuroblastoma cell and mitochondria morphology upon LfcinB-exposure. (a) Mitochondria in untreated cells versus the mitochondria of the cells treated with LfcinB (40 lg/ml) for 3 h (b), showing severe ultrastructural alterations such as disorganization and swelling of the mitochondria resulting in collapse of the organelle. The cells exposed to LfcinB (40 lg/ml) for 12 h (d) exhibited typically necrotic morphology such as cell swelling and bursting.
with LfcinB revealed gold-marked LfcinB peptides in membrane blebs on the cell surfaces after 15 min of treatment (Fig. 4a). Figure 4b shows LfcinB located in larger structures outside the cell, which could be mitochondria released from dead cells. Several control samples were analyzed to eliminate unspecific labeling of antibodies and protein A-gold. Samples not exposed to LfcinB were immunolabeled with Ab and protein A-gold, and samples exposed to LfcinB were labeled with protein A-gold only. Gold particles were not detected in any of these samples, indicating a specific staining of LfcinB (data not shown). Immunofluorescence studies using MitoTracker (red fluorescence) and green fluorescence-labeled LfcinB peptides showed that LfcinB colocalized with mitochondria close to the cell membrane after 4 min of exposure (Figs. 5a–5c). A more uniform distribution of the peptides in the cytoplasm was evident after 8 min of exposure as seen in Figures 5d–5f. In addition, the peptides colocalized with red-colored spots that appeared close to the cell membrane. A similar clustering of red spots was also observed in LfcinB-treated cells stained with JC-1. Fluorescence-labeled LfcinB was also detected in cells that did not show any staining for the mitochondria-specific MitoTracker. These cells could be apoptotic cells or apoptotic bodies. Apoptotic cells and apoptotic bodies expresses higher levels of negatively charged phosphatidylserine in their outer membranes31,32 and might attract a higher amount of peptides, compared to non-apoptotic cancer cells.
LfcinB treatment reduces the growth of SH-SY-5Y xenografts To investigate the effects of LcinB on neuroblastoma growth in vivo, we treated nude rats carrying SH-SY-5Y xenografts with once-daily injections of 1.0 or 2.0 mg LfcinB. Tumor growth was significantly inhibited after 2 days of LfcinB treatment, compared to untreated controls. The difference between 1 and 2 mg of LfcinB was significant only on day 11 (p < 0.04) (Fig. 6). Tumor weight was recorded at autopsy. Neuroblastoma xenograft tumors from rats treated with LfcinB displayed significant reduction in tumor weight compared with the control animals (p 5 0.016 for 1 mg LfcinB; p 5 0.002 for 2 mg LfcinB). There was no difference in tumor weight between the two treatment groups (p 5 0.17, data not shown).
Discussion LfcinB displayed a selective cytotoxicity to all NB cell lines investigated compared to non-transformed fibroblasts. This difference in sensitivity most likely reflects differences in cell membrane compositions through which LfcinB interacts. This hypothesis is based on the fact that negatively charged molecules such as phosphatidylserine, lipoproteins, O-glycocylated mucines and sialic acid are present at higher levels on the outer leaflet of several tumor cell membranes relative to non-trans-
LACTOFERRICIN B IS CYTOTOXIC TO NEUROBLASTOMA CELLS IN VITRO AND INHIBITS XENOGRAFT GROWTH IN VIVO
497
FIGURE 4 – Localization of immunolabeled LfcinB in neuroblastoma cells. Kelly cells treated with LfcinB (40 lg/ml) for different time-spans were prepared for TEM, marked with primary polyclonal Ab specific for LfcinB and protein A-gold. Figure (a) shows that the peptide associates with the membrane blebs after 5 min of treatment, visualized by gold particles that are shown as black dots. Figure (b) also shows localization of LfcinB to structures outside the cells after 15 min of treatment.
FIGURE 3 – Western-blot analysis of caspase activation in LfcinBexposed neuroblastoma cells. Kelly cells were incubated with 40 lg/ml LfcinB for 24 h and harvested, lysed and tested for caspase activity. The Western-blots show activation of procaspase-9 after 30 min, whereas activation of the procaspase-6 and -7 were detected after 6 h.
formed cells.8,33–37 Notably, the MYCN-amplified cell lines were more sensitive to LfcinB compared to the non-MYCNamplified cell lines (Table I), in agreement with earlier reports documenting similar results with other antimicrobial peptides tested against MDR cell lines and drug-sensitive tumor cell lines.38 LfcinB also significantly reduced the growth of SH-SY5Y NB xenografts in nude rats without showing any pathophysiological effects on the animals after treatment. The fact that LfcinB possessed a relatively high specificity for the NB cells and that several antimicrobial peptides do not seem to be affected by MDR phenotypes suggest a possible clinical potential in cancer therapy for antimicrobial peptides. To investigate whether LfcinB induced mitochondrial changes in NB cells, the mitrochondrial transmembrane potential in Kelly cells treated with LfcinB were examined. The flow cytometry studies revealed a leakage of the fluorescent dye JC-1 from the mitochondria, indicating a mitochondrial membrane destabilizing effect by LfcinB. This was further documented by confocal microscopy studies, showing punctuated red spots randomly spread within the cell cytoplasm before LfcinB exposure, and a marked increase in green fluorescence in the cell cytoplasm after LfcinB treatment. The observed effect on the mitochondria could result in release of apoptogenic factors such as cytochrome c.39 Released cytochrome c creates a high-molecular weight complex together with ATP or dATP, APAF-1 and pro-caspase-9, which then activates caspase-9 and downstream caspases.40 Hence, the effect of LfcinB on the caspase-machinery was examined by studying processing of caspase-3, -6, -7, -8 and -9 from the inactive zymogens to active proteases by western blot analysis of extracts from LfcinB-treated Kelly cells. Procaspase-3 was weakly upregulated but not activated, whereas caspase-8 was not detected (data not shown). Several MYCN-amplified NB cell lines fail to express caspase-8 mRNA because of silencing of the capase-8 gene through DNA methylation.41 Activation of caspase-9 indicated an LfcinB-induced activation of a mitochondria-dependent apoptosis. However, the specific caspase inhibitors and the pan-caspase inhibitor zVAD-fmk were not able to inhibit the cytotoxic activity against Kelly cells mediated by LfcinB, suggesting that caspase activation was not critical for the cytotoxic activity by LfcinB.
Since the activation of caspases was not responsible for the cytotoxic effect of LfcinB, in vitro morphological changes in Kelly cells treated with LfcinB were examined in more detail by transmission electron microscopy (TEM). The TEM studies revealed an abnormal mitochondrial morphology in Kelly cells after 3 h of LfcinB treatment, followed by a progression from a state of condensed morphology, ultrastructural parallelization of the mitochondrial inner membrane, vacuolization and swelling, to a final state of irreversible damage of the inner and outer mitochondria membrane (Figs. 2a and 2b). The effect observed on the mitochondria in Kelly cells indicates that an irreversible damage of the mitochondria membranes is at least partly responsible for the cytotoxic effect of LfcinB. A severe swelling and bursting of whole cells were evident after 12 h (Figs. 2c and 2d), a morphology characteristic for cells undergoing necrosis.39 Apoptotic membrane bodies, which are typical hallmarks of the apoptotic process, were not evident. Additionally, FACS analyses and DAPI-staining did not show significant apoptotic morphology of the LfcinB-treated Kelly cells. Taken together, these observations suggest that the neuroblastoma cells are killed by necrosis. Although an early effect of LfcinB on the mitochondria suggests a direct effect on these organelles, the possibility that the peptide killed the cells by targeting the cytoplasmic membrane, which then leads to secondary effects on the mitochondria, cannot be ruled out. The localization of LfcinB in the Kelly cells, using both immunolabeled and fluorescent-labeled LfcinB, by TEM and confocal microscopy, respectively, was therefore investigated to identify the target(s) for LfcinB. The TEM studies revealed that immunolabeled LfcinB peptides were localized close to the plasma membrane and in small membrane blebs after a few minutes of exposure. Formation of plasma membrane blebs has earlier been reported as a result of external disturbance such as anoxia, and has been explained to be a consequence of ATP depletion, cytoskeletal alterations and loss of volume control in the cells.42 The results revealed by the TEM studies indicate a direct effect on the cytoplasmic membrane by LfcinB. Fifteen minutes exposure to LfcinB leads to cytoplasmic sequestration of the peptide (Fig. 4). Moreover, the confocal microscopy studies documented that fluorescence-labeled LfcinB were colocalizing with red MitoTracker, indicating that LfcinB targets mitochondria (Fig. 5). Interestingly, relatively large red-colored spots that colocalized with fluorescence-labeled LfcinB were observed close to the cell membrane, suggesting that a clustering of the mitochondria had occurred after LfcinB treatment. A similar redistribution of the mitochondria was not observed in experiments using the positive control FCCP (Fig. 1). Clustering of the mitochondria has been
498
ELIASSEN ET AL.
FIGURE 5 – Localization of fluorescence-labeled LfcinB peptide in neuroblastoma cells using confocal microscopy. (a) shows mitochondria labeled with red MitoTracker, while (b) illustrates the localization of the fluorescence-labeled LfcinB (40 lg/ml) in Kelly cells after few min of incubation. Image (c) shows that the peptide colocalizes with the mitochondria. (d, e and f) Kelly cells incubated with the fluorescence-labeled peptide for 8 min, showing that the peptides were distributed more throughout the cell cytoplasm and colocalize in some hot spots close to the cell membrane.
FIGURE 6 – Treatment effects on neuroblastoma SH-SY-5Y xenografts. The SH-SY-5Y xenograft tumor volume in nude rats treated with LfcinB (1 mg and 2 mg) and control tumors treated with physiological saline on day 1, 2 and 3. Mean volumes for tumors at tumor take (day 1) and 10 days from start of are shown. Tumors from rats treated with LfcinB 1 mg (square), LfcinB 2 mg (triangle) and physiological saline (diamond) are shown. The SH-SY-5Y xenograft tumor weight at sacrifice 10 days from tumor take and start of LfcinB treatment. Tumors from rats treated with LfcinB 1 mg (treatment type 2, mean weight 3.13 g, range 0.84–4.74 g), LfcinB 2 mg (treatment type 3, mean 2.23 g, range 1.18–3.70 g) and physiological saline (treatment type 1, mean 5.42 g, range 3.22–7.64 g) (data not shown).
observed at an early apoptotic stage in cells exposed to a mitochondrial-targeted fatty acid analogue.43 The clustering and redistribution of the mitochondria after LfcinB exposure could be a result of changes in osmotic pressure in the cytosol because of leakage in the cell membrane. Taken together, our results suggest that antimicrobial peptides may have a dual mechanism of action against cancer cells, since it seems to directly affect both the mitochondria and the cytoplasmic membrane of Kelly cells. Differences in the level of cytoplasmic membrane destabilization among different cancer cell lines are most probably dependent on the number of target molecules for LfcinB expressed on the cell surface, e.g. PS and sialic acid,8,44 whereas the differences in the level of mitochondria membrane destabilization should be dependent on the number of LfcinB molecule targeting the mitochondria
membrane after internalization. The mitochondria are suggested as possible targets for LfcinB because of their rather negative surface and a highly negative transmembrane potential. It is the negatively charged lipid cardiolipin that causes the negative surface of the organelle. Cardiolipin is also abundantly concentrated in the inner membrane of the mitochondrion.45,46 Interestingly, recently it was demonstrated an altered molecular composition of the outer and inner mitochondrial membrane in cancer cells relative to healthy cells,47,48 thus strengthening that mitochondria are specific targets for LfcinB. DNA and RNA molecules are other negative charged components that possibly could function as targets for LfcinB. Both bovine and human lactoferrin protein are found to internalize eukaryotic cells and localize in the nucleus where they bind DNA and act as transcription factors.49–53 The proteins bind DNA through their highly basic N-terminal domains.54 It is this part of the proteins from which the lactoferricin peptides are derived. However, the peptides most probably do not interact with the lactoferrin receptors during any stage of cell internalization because of their small size and a molecular structure that is largely different when released from the intact protein. A more unspecific mode of action is suggested, in which electrostatic interactions in between positively charged residues in the peptide and negative lipid head groups are the main events. The mitochondria are involved in both apoptotic and necrotic death processes influenced by the ATP level and the activation of caspases.39 In mammals, caspases are frequently activated as a result of mitochondrial membrane permeabilization.40 However, if the mitochondrial damage has exceeded a certain threshold and ATP is depleted due to uncoupling of the respiration chain in the mitochondria, apoptosis is blocked and the cell dies of necrosis.39 The recent observation that inhibitors of caspases protect carcinoma and leukemia cells,23 but not Kelly cells in our experiments from LfcinB-induced killing, is probably due to a higher level of irreversible membrane damage in Kelly cells relative to the level of cytoplasmic and mitochondria membrane damage by LfcinB in the carcinoma and leukemia cells, thereby favoring a caspase-independent apoptotic cell death. The antimicrobial peptides magainin, cecropin and melittin have all been shown to affect mitochondrial coupling and respiration,55,56 and the antimicrobial peptide BMAP-28 causes depolarization of the inner mitochondrial membrane in human monocytic
LACTOFERRICIN B IS CYTOTOXIC TO NEUROBLASTOMA CELLS IN VITRO AND INHIBITS XENOGRAFT GROWTH IN VIVO
and erythroid cell lines by inducing formation of mitochondrial permeability transition pores and show a direct effect on isolated mitochondria.19 However, as far as we know, this is the first report documenting that an antimicrobial peptide is physically targeting the mitochondria in intact cells. In conclusion, this study demonstrates that LfcinB is cytotoxic to a number of NB cell lines in vitro and inhibits the growth of SH-SY5Y xenografts significantly in vivo. LfcinB possesses a cytotoxic effect on MYCN-amplified NB cells by directly targeting and destabilizing the cytoplasmic membrane and the mitochondria.
499
Acknowledgements The present study was supported by The National Programme for Research in Functional Genomics in Norway (FUGE) in The Research Council of Norway. Assistant professor Geir Bjørkøy is acknowledged for assistance and for valuable discussions during the confocal microscopy studies. Thanks to Eirin Listau Bertelsen and Bjørn Greve Moe for important technical assistance and discussions in doing the flow cytometric experiments. We also thank the Department of Electron Microscopy and especially Randi Olsen for skilful technical help and valuable discussions.
References 1. 2. 3.
4. 5. 6.
7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17.
18.
19.
20.
21.
Brodeur GM. Neuroblastoma: biological insights into a clinical enigma. Nat Rev Cancer 2003;3:203–16. Cruciani RA, Barker JL, Zasloff M, Chen HC, Colamonici O. Antibiotic magainins exert cytolytic activity against transformed cell lines through channel formation. Proc Natl Acad Sci USA 1991;88:3792–6. Chen HM, Wang W, Smith D, Chan SC. Effects of the anti-bacterial peptide cecropin B and its analogs, cecropins B-1 and B-2, on liposomes, bacteria, and cancer cells. Biochim Biophys Acta 1997;1336: 171–9. Baker MA, Maloy WL, Zasloff M, Jacob LS. Anticancer efficacy of Magainin 2 and analogue peptides. Cancer Res 1993;53:3052–7. Eliassen LT, Berge G, Sveinbjørnsson B, Svendsen JS, Vorland LH, Rekdal Ø. Evidence for a direct antitumor mechanism of action of bovine lactoferricin. Anticancer Res 2002;22:2703–10. Johnstone SA, Gelmon K, Mayer LD, Hancock RE, Bally MB. In vitro characterization of the anticancer activity of membrane-active cationic peptides. I. Peptide-mediated cytotoxicity and peptide-enhanced cytotoxic activity of doxorubicin against wild-type and p-glycoprotein over-expressing tumor cell lines. Anticancer Drug Des 2000; 15:151–4. Ohsaki Y, Gazdar AF, Chen HC, Johnson BE. Antitumor activity of magainin analogues against human lung cancer cell lines. Cancer Res 1992;52:3534–8. Utsugi T, Schroit AJ, Connor J, Bucana CD, Fidler IJ. Elevated expression of phosphatidylserine in the outer membrane leaflet of human tumor cells and recognition by activated human blood monocytes. Cancer Res 1991;51:3062–6. Yang N, Lejon T, Rekdal Ø. Antitumor activity and specificity as a function of substitutions in the lipophilic sector of helical lactoferrinderived peptides. J Pept Sci 2003;9:300–11. Yang N, Stensen W, Svendesen JS, Rekdal Ø. Enhanced antitumor activity and selectivity of lactoferrin-derived peptides. J Pept Res 2002; 60:187–97. Papo N, Shahar M, Eisenbach L, Shai Y. A novel lytic peptide composed of DL-amino acids selectively kills cancer cells in culture and in mice. J Biol Chem 2003;278:21018–23. Sitaram N, Nagaraj R. Interaction of antimicrobial peptides with biological and model membranes: structural and charge requirements for activity. Biochim Biophys Acta 1999;1462:29–54. Matsuzaki K, Sugishita K, Ishibe N, Ueha M, Nakata S, Miyajima K, Epand RM. Relationship of membrane curvature to the formation of pores by magainin 2. Biochemistry 1998;37:11856–63. Epand RM, Vogel HJ. Diversity of antimicrobial peptides and their mechanisms of action. Biochim Biophys Acta 1999;1462:11–28. Andreu D, Rivas L. Animal antimicrobial peptides: an overview. Biopolymers 1998;47:415–33. Mai JC, Mi Z, Kim SH, Ng B, Robbins PD. A proapoptotic peptide for the treatment of solid tumors. Cancer Res 2001;61:7709–12. Ellerby HM, Arap W, Ellerby LM, Kain R, Andrusiak R, Rio GD, Krajewski S, Lombardo CR, Rao R, Ruoslahti E, Bredesen DE, Pasqualini R. Anti-cancer activity of targeted pro-apoptotic peptides. Nat Med 1999;5:1032–8. Takeshima K, Chikushi A, Lee KK, Yonehara S, Matsuzaki K. Translocation of analogues of the antimicrobial peptides magainin and buforin across human cell membranes. J Biol Chem 2003;278:1310– 15. Risso A, Braidot E, Sordano MC, Vianello A, Macri F, Skerlavaj B, Zanetti M, Gennaro R, Bernardi P. BMAP-28, an antibiotic peptide of innate immunity, induces cell death through opening of the mitochondrial permeability transition pore. Mol Cell Biol 2002;22:1926–35. Vogel HJ, Schibli DJ, Jing W, Lohmeier-Vogel EM, Epand RF, Epand RM. Towards a structure-function analysis of bovine lactoferricin and related tryptophan- and arginine-containing peptides. Biochem Cell Biol 2002;80:49–63. Yoo YC, Watanabe S, Watanabe R, Hata K, Shimazaki K, Azuma I. Bovine lactoferrin and lactoferricin inhibit tumor metastasis in mice. Adv Exp Med Biol 1998;443:285–91.
22. Yoo YC, Watanabe R, Koike Y, Mitobe M, Shimazaki K, Watanabe S, Azuma I. Apoptosis in human leukemic cells induced by lactoferricin, a bovine milk protein-derived peptide: involvement of reactive oxygen species. Biochem Biophys Res Commun 1997;237:624–28. 23. Mader JS, Salsman J, Conrad DM, Hoskin DW. Bovine lactoferricin selectively induces apoptosis in human leukemia and carcinoma cell lines. Mol Cancer Ther 2005;4:612–24. 24. Strøm MB, Rekdal O, Svendsen JS. Antibacterial activity of 15-residue lactoferricin derivatives. J Pept Res 2000;56:265–74. 25. Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983;65:55–63. 26. Ponthan F, Johnsen JI, Klevenvall L, Castro J, Kogner P. The synthetic retinoid RO 13-6307 induces neuroblastoma differentiation in vitro and inhibits neuroblastoma tumour growth in vivo. Int J Cancer 2003;104:418–24. 27. Bozzola JJ, Russel LD. Specimen preparation for transmission electron microscopy. In: Bozzola JJ, Russel LD, eds. Electron microscopy. Sudbury, MA: Jones & Barlett, 1992 p. 15–38. 28. Haukland HH, Ulvatne H, Sandvik K, Vorland LH. The antimicrobial peptides lactoferricin B and magainin 2 cross over the bacterial cytoplasmic membrane and reside in the cytoplasm. FEBS Lett 2001;508: 389–93. 29. Johnsen JI, Pettersen I, Ponthan F, Sveinbjornsson B, Flægstad T, Kogner P. Synergistic induction of apoptosis in neuroblastoma cells using a combination of cytostatic drugs with interferon-g and TRAIL. Int J Oncol 2004;25:1849–57. 30. Nilsson S, Pahlman S, Arnberg H, Letocha H, Westlin JE. Characterization and uptake of radiolabelled meta-iodobenzylguanidine (MIBG) in a human neuroblastoma heterotransplant model in athymic rats. Acta Oncol 1993;32:887–91. 31. Blankenberg FG, Katsikis PD, Tait JF, Davis RE, Naumovski L, Ohtsuki K, Kopiwoda S, Abrams MJ, Darkes M, Robbins RC, Maecker HT, Strauss HW. In vivo detection and imaging of phosphatidylserine expression during programmed cell death. Proc Natl Acad Sci USA 1998;95:6349–54. 32. Bombeli T, Karsan A, Tait JF, Harlan JM. Apoptotic vascular endothelial cells become procoagulant. Blood 1997;89:2429–42. 33. Zwaal RF, Schroit AJ. Pathophysiologic implications of membrane phospholipid asymmetry in blood cells. Blood 1997;89:1121–32. 34. Williamson P, Schlegel RA. Back and forth: the regulation and function of transbilayer phospholipid movement in eukaryotic cells. Mol Membr Biol 1994;11:199–216. 35. Raval GN, Patel DD, Parekh LJ, Patel JB, Shah MH, Patel PS. Evaluation of serum sialic acid, sialyltransferase and sialoproteins in oral cavity cancer. Oral Dis 2003;9:119–28. 36. Miyagi T, Wada T, Yamaguchi K, Hata K. Sialidase and malignancy: a minireview. Glycoconj J 2004;20:189–98. 37. Cappelli G, Paladini S, D’Agata A. Tumor markers in the diagnosis of pancreatic cancer. Tumori 1999;85 (Suppl 1):S19–S21. 38. Jing N, Urry DW. Ion pair binding of Ca21 and Cl2 ions in micellarpackaged gramicidin A. Biochim Biophys Acta 1995;1238:12–21. 39. Kroemer G, Dallaporta B, Resche-Rigon M. The mitochondrial death/ life regulator in apoptosis and necrosis. Annu Rev Physiol 1998;60: 619–42. 40. Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome cdependent activation of caspase-3. Cell 1997;90:405–13. 41. Teitz T, Wei T, Valentine MB, Vanin EF, Grenet J, Valentine VA, Behm FG, Look AT, Lahti JM, Kidd VJ. Caspase 8 is deleted or silenced preferentially in childhood neuroblastomas with amplification of MYCN. Nat Med 2000;6:529–35. 42. Kim JS, He L, Lemasters JJ. Mitochondrial permeability transition: a common pathway to necrosis and apoptosis. Biochem Biophys Res Commun 2003;304:463–70. 43. Tronstad KJ, Gjertsen BT, Krakstad C, Berge K, Brustugun OT, Doskeland SO, Berge RK. Mitochondrial-targeted fatty acid analog
500
44. 45.
46.
47. 48.
49.
ELIASSEN ET AL.
induces apoptosis with selective loss of mitochondrial glutathione in promyelocytic leukemia cells. Chem Biol 2003;10:609–18. Risso A, Zanetti M, Gennaro R. Cytotoxicity and apoptosis mediated by two peptides of innate immunity. Cell Immunol 1998;189:107–15. Hovius R, Lambrechts H, Nicolay K, de Kruijff B. Improved methods to isolate and subfractionate rat liver mitochondria. Lipid composition of the inner and outer membrane. Biochim Biophys Acta 1990;1021: 217–26. de Kroon AI, Dolis D, Mayer A, Lill R, de Kruijff B. Phospholipid composition of highly purified mitochondrial outer membranes of rat liver and Neurospora crassa. Is cardiolipin present in the mitochondrial outer membrane? Biochim Biophys Acta 1997;1325:108–16. Villa AM, Doglia SM. Mitochondria in tumor cells studied by laser scanning confocal microscopy. J Biomed Opt 2004;9:385–94. Parlo RA, Coleman PS. Enhanced rate of citrate export from cholesterol-rich hepatoma mitochondria. The truncated Krebs cycle and other metabolic ramifications of mitochondrial membrane cholesterol. J Biol Chem 1984;259:9997–10003. Son KN, Park J, Chung CK, Chung DK, Yu DY, Lee KK, Kim J. Human lactoferrin activates transcription of IL-1b gene in mammalian cells. Biochem Biophys Res Commun 2002;290:236–41.
50. He J, Furmanski P. Sequence specificity and transcriptional activation in the binding of lactoferrin to DNA. Nature 1995;373:721–4. 51. Garre C, Bianchi-Scarra G, Sirito M, Musso M, Ravazzolo R. Lactoferrin binding sites and nuclear localization in K562(S) cells. J Cell Physiol 1992;153:477–82. 52. Baumrucker CR, Schanbacher F, Shang Y, Green MH. Lactoferrin interaction with retinoid signaling: cell growth and apoptosis in mammary cells. Domest Anim Endocrinol, in press. 53. Ashida K, Sasaki H, Suzuki YA, Lonnerdal B. Cellular internalization of lactoferrin in intestinal epithelial cells. Biometals 2004;17: 311–15. 54. Kanyshkova TG, Semenov DV, Buneva VN, Nevinsky GA. Human milk lactoferrin binds two DNA molecules with different affinities. FEBS Lett 1999;451:235–7. 55. Hugosson M, Andreu D, Boman HG, Glaser E. Antibacterial peptides and mitochondrial presequences affect mitochondrial coupling, respiration and protein import. Eur J Biochem 1994;223:1027–33. 56. Diaz-Achirica P, Prieto S, Ubach J, Andreu D, Rial E, Rivas L. Permeabilization of the mitochondrial inner membrane by short cecropin-A-melittin hybrid peptides. Eur J Biochem 1994;224:257–63.
