Kahalalide F, a new marine-derived compound, induces oncosis in human prostate and breast cancer cells Free

Kahalalide F (KF) (C₇₅H₁₂₄N₁₄O₁₆, M_r = 1476; Fig. 1A) is one of the families of natural depsipeptides isolated from the Hawaiian herbivorous marine mollusk Elysia rufescens (1, 2). Like other kahalalides, it is probably a secondary metabolite synthesized by the mollusk from peptides produced by a diet of the green algae Bryopsis pennata. KF has potent cytotoxic activity in vitro against cell lines from solid tumors including prostate, breast and colon carcinomas, neuroblastoma, chondrosarcoma, and osteosarcoma (3–6). In animal models in vivo, KF has also shown activity against human prostate cancer xenografts (4). Cytotoxicity against human tumor specimens has been seen with breast, colon, non-small cell lung, and ovarian carcinomas (7). Two phase I clinical trials in adult patients bearing advanced prostate cancer and pretreated solid tumors have identified the recommended dose for two different schedules (once a week and five consecutive days every 3 weeks) (8). Dose-limiting toxicity was elevation of aminotransferases in both trials. Recently, KF has entered into a phase II clinical trial.

The mechanism of KF action is mostly unknown. KF is a National Cancer Institute COMPARE-negative compound, which indicates that its cytotoxicity might be related to a unique mode of action. Human HeLa cervical cancer cells and monkey COS-1 fibroblasts treated with KF became dramatically swollen and developed large vacuoles that appeared to be consequence of changes in lysosomal membranes (9). However, the nature and characteristics of the cytotoxic effect of KF and mechanisms of cell death induction remain to be elucidated.

Clearly, the clinical use of KF will greatly benefit from the study of specificity and molecular determinants of its activity. Here we examine the action of KF in human prostate and breast cancer cells. KF rapidly induces a strong cytotoxic effect with IC₅₀ in the submicromolar range. We have examined the involvement of BCL-2, MDR1, and HER2/NEU genes in the mechanism of cytotoxicity of this compound. In addition, by flow cytometry and confocal and electron microscopy, we have characterized the cell death process induced by KF treatment. Together, our data characterize KF as a potent antitumor drug that induces predominantly necrotic cell death.

Human prostate cancer PC3, DU145, and LNCaP cells were grown in RPMI 1640 supplemented with 10% FCS, breast cancer SKBR-3, BT474, MCF7 and MDA-MB-231 cells, colon cancer LoVo cells, and IMR90 human diploid fibroblasts (supplied by Dr. M. Serrano, Centro Nacional de Biotecnologı́a, Madrid, Spain) were grown in DMEM supplemented with 10% FCS and 1 mm glutamine (all from Life Technologies, Inc.-Invitrogen, Paisley, Scotland). MCF10A normal human mammary epithelial cells (supplied by Dr. D. Salomon, National Cancer Institute, Bethesda, MD) were cultured in a mixture (1:1) of DMEM:F12 media supplemented with 5% horse serum, 20 ng/ml human recombinant EGF (Upstate Biotechnology, Lake Placid, NY) and 10 μg/ml insulin, 5.5 μg/ml transferrin, 6.7 ng/ml sodium selenite (all from Life Technologies-Invitrogen), and 0.5 μg/ml hydrocortisone. Endothelial HMEC-1 (dermal microvascular) and HUVEC (umbilical cord) cells (supplied by Dr. B. Jiménez, Instituto de Investigaciones Biomédicas “Alberto Sols”, Madrid, Spain) were grown respectively in DMEM supplemented with 10% FCS and M199 medium supplemented with 20% FCS, 10 ng/ml human recombinant EGF (Upstate Biotechnology), 20 mm HEPES, 10 μg/ml hydrocortisone, and 100 μg/ml heparin. MCF7/HER2 cells expressing an exogenous HER2/NEU gene (10) were supplied by Dr. A. Pandiella (Centro de Investigación del Cáncer, Salamanca, Spain); MDA-MB-231/BCL-2 cells expressing an exogenous BCL-2 gene (11) were supplied by Dr. A. López (Instituto López-Neyra de Biomedicina y Parasitologı́a, Granada, Spain); and Lovo/DOX cells overexpressing the MDR1 gene were donated by Dr. Lola Garcı́a-Grávalos (Pharma Mar S.A., Madrid, Spain), and were cultured as their parental cells.

KF is manufactured by Pharma Mar. Stock solutions were freshly prepared in dimethyl sulfoxide:ethanol (1:1) and diluted in the cell culture to final concentrations as indicated. Actinomycin D, cycloheximide, chloroquine, and valinomycin were from Sigma Chemical Co. (St. Louis, MI), and the Z-VAD-fmk general caspase inhibitor was from Calbiochem-Merck Biosciences (Darmstadt, Germany).

Primary antibodies used were: mouse monoclonal antibodies antihuman BCL-2 (DAKO, Copenhagen, Denmark), anti-MDR1 (Neo Markers, Fremont, CA), anti-HER2/NEU (Oncogene Research Products, Boston, MA) and anti-pan-Histone (clone H11-4, Roche Diagnostics, Indianapolis, IN); and rabbit polyclonal antibodies anti-calreticulin (Calbiochem-Merck) and anti-lamin B (Dr. S. Georgatos, University of Crete, Greece). As secondary antibodies, we used goat anti-mouse or anti-rabbit antibodies conjugated with FITC or Texas Red (Jackson ImmunoResearch Laboratories, West Grove, PN).

We used a FACScan flow cytometer (Becton Dickinson, San Jose, CA) equipped with a 488-nm argon ion laser. For cell-cycle analysis, asynchronous cells (500,000/ml) were cultured in the presence of different concentrations of KF. On incubation, the cells were harvested and washed in PBS, fixed in 70% cold ethanol for 30 min at −20°C, washed again in PBS, and incubated for 1 h in PBS containing 100 μg/ml RNase (Sigma) and 50 μg/ml propidium iodide (PI) (Sigma). Ten thousand events per sample were acquired for data analysis using CELLQUEST software by selective gating to exclude doublet cells. The results of the DNA content were modeled using ModFit (VERITY Software House, Inc., Topsham, ME). Lysosomal integrity was analyzed by the LysoTracker Green-uptake assay. The acidotropic dye LysoTracker Green (Molecular Probes, Eugene, OR), which accumulates in intact lysosomes, was diluted in culture media. Cells were cultured in the presence or absence of KF for the desired period of time and then medium containing 75 nm Lysotracker Green was added and cells were incubated for a further 30 min at 37°C. Cells were harvested, washed in PBS, and resuspended in 1 ml PBS. Green fluorescence of 10,000 events per sample was acquired for the data analysis using the FL1 channel. In some instances, cells were labeled for a further 15 min in the presence of 15 μg/ml PI to allow discrimination between live (PI non-permeant, no red fluorescence) and dead cells (PI permeant, high red fluorescence). For Acridine Orange (AO) uptake assays, cells were incubated as above but exposed to 5 μg/ml AO (Sigma) for 15 min at 37°C. AO is a lysomotropic base and metachromatic fluorochrome exhibiting red fluorescence at high concentrations, when retained in its charged form in intact lysosomes. Red lysosomal fluorescence of 10,000 events per sample was determined by flow cytometry using FL3 channel. To measure mitochondrial membrane potential (ΔΨm), mitochondria were selectively probed with potential-sensitive JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethyl-benzimidazolylcarbocyanine iodide) (Molecular Probes). JC-1 was stocked as 1 mg/ml in DMSO and freshly diluted with complete culture medium. After treatments, cells were harvested and incubated with medium containing JC-1 (10 μg/ml) for 10 min. Finally, cells were washed and resuspended in 1 ml PBS for flow cytometry analysis. JC-1 exhibits ΔΨm-dependent accumulation in mitochondria, indicated by a shift in its fluorescence emission from green to red due to the formation of J-aggregates. Mitochondrial depolarization is so indicated by a decrease in red/green fluorescence ratio.

Thirty thousand cells were seeded in 24-well dishes. After overnight incubation, cells were treated or not with KF in normal growth medium and pulsed with 1 μCi/ml [³H]thymidine (Amersham-Pharmacia Biotech, Piscataway, NJ) for 4 h at the indicated times posttreatment. At the end of the labeling period, the medium was removed and the cells were rinsed twice in PBS and fixed with chilled 10% trichloroacetic acid for 10 min. Trichloroacetic acid was then removed and the monolayers were washed in ethanol and air-dried at room temperature for 20 min. Thereafter, precipitated macromolecules were dissolved in 500 μl of 0.5 N NaOH-0.1% SDS and 450 μl of each sample were diluted in 5 ml of scintillation solution OptiPhase HighSafe (Wallac Scintillation Products). Radioactivity was measured on a 1209 RackBeta counter (LKB Wallac, Turku, Finland).

To estimate cell mass, medium was removed and 24-well dishes were washed in PBS, fixed with 1% glutaraldehyde for 15 min, washed again in PBS twice, and stained with 200 μl 0.1% aqueous crystal violet for 20 min. Dishes were rinsed 4 times in tap water and allowed to dry. One hundred microliters 10% acetic acid were added and the content of each well was mixed before reading A₅₉₅.

After treatments, cells were harvested and washed in PBS. Cellular DNA was then extracted after overnight incubation at 37°C in a lysis buffer containing 25 mm EDTA, 100 mm NaCl, 0.5 mg/ml proteinase K, 0.5% SDS, and 10 mm Tris-HCl (pH 8.0). DNA was precipitated by 2 vol of 100% ethanol and 1/10 vol of 3 m sodium acetate. After centrifugation, samples were resuspended and incubated for 1 h at 37°C in TE buffer (10 mm Tris-HCl, 0.2 mm EDTA, pH 7.5) with 50 μg/ml RNase. Supernatants were extracted with a 1:1 mixture of phenol:chloroform, precipitated as above and resuspended in TE buffer. Electrophoresis was carried out in 2% agarose gels in TAE buffer containing 0.5 μg/ml ethidium bromide. Gels were examined under UV light and photographed.

For indirect immunofluorescence, culture cells were washed twice in PBS, fixed with 3.7% formaldehyde (freshly prepared from paraformaldehyde) in PBS for 10 min at room temperature, and subsequently permeabilized with 0.5% Triton X-100 in PBS for 20 min at room temperature. Before immunostaining, fixed culture cell samples were sequentially incubated with 0.1 m glycine in PBS for 30 min, 1% BSA in PBS for 15 min, and 0.01% Tween 20 in PBS for 5 min. For immunolabeling, cells were rinsed in PBS containing 0.05% Tween 20 (PBS-Tw), incubated for 2 h at room temperature with primary antibodies diluted in PBS, washed in PBS-Tw, and incubated for 45 min with the appropriate secondary antibodies conjugated to FITC, or Texas Red (Jackson ImmunoResearch Laboratories). Some samples were counterstained with PI for the cytochemical demonstration of nucleic acids. Finally, the coverslips were mounted in VectaShield (Vector Laboratories, Peterborough, United Kingdom) and sealed with nail polish. Confocal microscopy was performed with a Bio-Rad MRC-1024, using excitation wavelengths of 488 nm (for FITC) and 543 nm (for Texas Red). Each channel was recorded independently and pseudocolor images were generated and superimposed. Images were processed by the Adobe Photoshop 7.0 software (Adobe Systems, Inc., San Jose, CA).

PC3 and SKBR-3 cells were fixed with 3% glutaraldehyde in 0.1 m cacodylate buffer for 15 min at room temperature. The cells were scraped, transferred to an Eppendorf tube, and centrifuged for 10 min. The pellets were washed with 0.1 m cacodylate buffer, postfixed in 2% osmium tetroxide, dehydrated in acetone, and embedded in araldite (Durcupan; Fluka, Switzerland). Ultrathin sections stained with uranyl acetate and lead citrate were examined with a Philips EM-208 electron microscope operated at 60 kV.

First, we studied the effect of KF on the proliferation of a panel of human prostate (PC3, DU145, LNCaP) and breast (SKBR-3, MCF7, BT474) cancer cell lines by estimating the level of DNA synthesis 24 h after treatment with different doses. Under the phase-contrast microscope, cells showed a dramatic morphological change soon after KF exposure, with intense swelling followed by cytoplasmic disintegration but no visible nuclear breakdown (Fig. 1B). In all cases, a dose-dependent inhibition of DNA synthesis was found, with IC₅₀ around 0.2 μm (Fig. 1C) except for the androgen-independent PC3 line, which was more sensitive (IC₅₀ = 0.07 μm). Wash-out experiments showed that short-pulse exposure to KF was sufficient to induce cytotoxicity, because 45 min treatment reduced cell survival as much as 24 h treatment, and 15 min treatment had a half-maximal effect (Fig. 1D).

Next, we examined whether KF cytotoxicity depends on the level of expression of three genes that are relevant for the action of many antitumor drugs: (a) BCL-2, an apoptosis inhibitor oncogene (12); (b) MDR1, encoding the P-glycoprotein responsible for extrusion of and resistance to multiple drugs, the overexpression of which is associated with poor clinical outcome (13, 14); and (c) HER2/NEU oncogene coding for a plasma membrane tyrosine kinase, the overexpression of which is linked to high malignancy and poor patient prognosis in breast cancer and other neoplasias (Ref. 15 and references therein). Only slight differences in KF sensitivity were found between MDA-MB-231 breast cancer cells overexpressing BCL-2 protein (MDA-MB-231/BCL-2; IC₅₀ = 0.74 μm) and the empty vector-transfected cells (IC₅₀ = 0.39 μm) (Fig. 2A). Expression of an exogenous MDR1 gene did not significantly alter the activity of KF against LoVo colon cancer cells (Fig. 2B). Likewise, KF action was also independent of the level of HER2/NEU expression, because no differences in drug sensitivity were found between parental and HER2/NEU-overexpressing MCF7 breast cancer cells (Fig. 2C). This result was confirmed by the finding that cell lines expressing either high (SKBR-3, BT-474), intermediate (LNCaP), or low (DU145, MCF7, MDA-MB-231) levels of HER2/NEU protein in Western blots (not shown) showed no significant differences in KF sensitivity (Fig. 1B). Moreover, PC3 cells showing highest sensitivity to KF express only moderate levels of HER2/NEU protein (not shown).

To assess whether gene expression is required for KF action, we treated PC3 cells in the presence of inhibitors of transcription (actinomycin D) or translation (cycloheximide). At periods (6 h) causing no toxicity per se, neither of these two inhibitors affected the cytotoxic activity of KF (Fig. 3A). Likewise, pretreatment (24 h) with the general caspase inhibitor Z-VAD-fmk did not inhibit KF action, suggesting that caspases are not involved in its mechanism of action (Fig. 3B). In these same cells, the analysis by flow cytometry revealed a time-dependent increase in highly autofluorescent cell debris (as evidenced by forward and side-scatter properties) following KF exposure. In contrast, no hypodiploid peak (a hallmark of apoptosis) or arrest in any phase of the cell cycle was found. Furthermore, DNA analysis by gel electrophoresis showed that KF treatment did not induce the ladder-type of DNA degradation that is characteristic of apoptosis (Fig. 3D).

Mitochondrial depolarization is a common event on treatment of cells with antitumor drugs (16). Using the cationic dye JC-1, which exhibits potential-dependent accumulation in mitochondria, we found that KF induced a rapid and transient loss of membrane potential (ΔΨm) which is detected as soon as 2 h on KF addition by an increase in green and a decrease in red fluorescence emission (Fig. 4). At 24 h posttreatment, abundant non-specific highly auto-fluorescent cell debris was found, concomitantly with a decrease in the cell population undergoing mitochondrial depolarization, suggesting the induction by KF of ΔΨm loss in a proportion of cells that are prone to die.

Putative effects of KF on lysosomes were studied using the fluorescent acidotropic probes LysoTracker and AO, which are retained in intact organelles. KF exposure led to a dose-dependent appearance of “pale cells” (deficient in intact lysosomes or with damaged lysosomes) that showed very low LysoTracker Green or AO fluorescence, respectively, indicating alteration of the lysosomal membrane permeability (Fig. 5A). Double labeling with LysoTracker Green and PI showed that the population of cells with reduced LysoTracker Green fluorescence incorporated a high-level PI, which corresponds to dying cells with heavily altered plasma membrane permeability (Fig. 5B). Emphasizing its effect on lysosomes, KF also enhanced the effect of the lysosomotropic agent chloroquine on the organelle integrity (Fig. 5C).

First, we examined the alterations caused by KF in PC3 cells by immunostaining with specific antibodies followed by confocal laser microscopy analysis. KF drastically affected the ER, causing microvesiculation and disruption of the normal organization of the ER network observed in untreated cells using an anti-calreticulin antibody (Fig. 6, right panels). In contrast to what is usually observed during apoptosis, the staining with an anti-lamin B antibody showed that the integrity of nuclear envelope is preserved in KF-treated cells (Fig. 6, left panels). Likewise, the use of an anti-pan-Histone antibody revealed that KF induces chromatin condensation in small irregular aggregates distributed throughout the nucleus, but not the formation of peripheral, sharply delineated masses of condensed chromatin or apoptotic bodies, which are characteristic structural features of apoptosis (Fig. 6, middle panels).

To gain further insight into alterations caused by KF, we performed electron microscopy studies. In prostate PC3 and also in breast SKBR-3 cells, KF exposure leads to a rapid cytoplasmic swelling, with dramatic disruption of the normal architecture (Fig. 7). This includes extensive vesiculation of cytoplasmic organelles, dilation of the endoplasmic reticulum (ER) elements, cytoskeletal degradation, and rupture of the plasma membrane (Fig. 7, C, D, G, and H). In addition, mitochondria of PC3 cells were severely damaged, showing increased matrix density, some swelling of cristae, and wrapping by a single rough ER cistern, suggesting autophagy (Fig. 7B). In the case of the SKBR-3 cells, some mitochondria had increased electron density and showed partial association with short cisterns of rough ER (Fig. 7F), but typical images of autophagy were not found. Cell nuclei were less affected in both PC3 and SKBR-3 cell lines. Main nuclear changes include irregular clumping of chromatin and the appearance of cleared nuclear domains free of chromatin (Fig. 7C). The nuclear envelope was preserved until the final stages of necrosis when cellular disintegration takes place (Fig. 7, C and F–H). Taken together, these results indicated that KF treatment causes a primary and preferential disruption of the cytoplasm architecture, while the cell nucleus is less affected at the early stages of drug exposure.

To test whether KF has a preferential action on tumor cells versus normal cells that could provide a therapeutic window in vivo, we studied its inhibitory effect on DNA synthesis in four normal, non-tumor cell lines: IMR90, MCF10A, HMEC-1, and HUVEC (Fig. 8). As compared with the panel of tumor cell lines previously analyzed (Fig. 1), KF had a diminished inhibitory effect on normal cells. The comparison of the respective IC₅₀ showed that normal cells were 5–40 times less sensitive to KF (Figs. 1 and 8).

In this work we have examined the action of KF, a novel antitumor agent of marine origin under clinical investigation, in human prostate and breast cancer cells. KF showed potent cytotoxic activity against both types of tumor cells. This action is triggered very rapidly and does not require continuous presence of the drug. Soon after exposure to KF, cells initiate a death process including marked swelling and a series of profound morphological alterations that affect many cytoplasm organelles and the plasma membrane. These features are typical of the process named oncosis, a term describing the progression of cellular events leading to necrotic cell death (16–18). The cellular architecture is greatly affected as early as 1–3 h after KF treatment, and the integrity of crucial organelles such as mitochondria, ER, or lysosomes is severely compromised. In contrast, the nuclear structure is preserved, and no drastic alteration of chromatin or DNA degradation is detected. Under the electron microscope, a proportion of damaged mitochondria in PC3 cells appeared surrounded by rough endoplasmic cisterns, suggesting that KF may to some extent induce mitochondrial autophagy. The rapid cytoplasmic swelling and the severe dysfunction and subsequent disintegration of endomembranes and plasma membrane lead to necrotic cell death. The alterations found suggest changes in the osmotic balance of the cell, possibly as a result of damage to cellular membranes, which we are currently investigating.

Our data indicate that KF has a novel mechanism of action among antitumor drugs, which is in line with the negative result of the NCI COMPARE program. Most anticancer drugs have been considered to induce apoptosis as a result of nuclear DNA damage, activation of membrane death receptors, or cytoskeletal alteration (Refs. 19–22 and references therein). Apoptosis is defined as an active, fixed-pathway process of cell death characterized by cell shrinking, cytoplasmic condensation, ladder DNA degradation, and nuclear fragmentation resulting in the formation of apoptotic bodies. In contrast, oncosis was considered a passive death process, a consequence of the unspecific toxicity of life-incompatible physical conditions or chemical poisons (16). Today, however, apoptosis and oncosis are no longer believed to be mutually exclusive processes (18, 23) and, moreover, apoptosis is probably not the only, or even the main mechanism by which tumor cells are killed by anticancer drugs or radiotherapy (24). It is accepted that apoptosis and oncosis share certain mechanisms and alterations such as loss of mitochondria permeability and membrane potential, and that cell fate is mostly due to the intensity and duration of the death signal and the cell genetic and metabolic status (16). Likewise, caspase-independent apoptosis has been described, which challenges the hypothesis that caspases activity is a definitive hallmark of the apoptotic process (25, 26). Our data indicate that KF action does not involve caspase activity but a rapid and severe destruction of cell organelles such as mitochondria and also ER and lysosomes. This is in line with the recent idea that the ER, Golgi apparatus, and lysosomes also play important roles in the integration and sensing of cellular damage in the initiation of cell death (27). We observed that KF induces lysosome damage, which agrees with a previous study in HeLa and COS-7 cells (9), and may contribute to necrotic cell death.

In addition to the morphological changes (cell swelling instead of shrinking), several biochemical results support the induction by KF of oncosis as opposed to apoptosis. First, the fact that no gene expression is required for KF to induce its effects. Second, that this family of proteases is not involved in KF action, as shown by the use of a general caspases inhibitor. Third, the absence of effects of KF on the cell cycle, because it does not induce the sub-G₀ hypodiploid peak that is characteristic of apoptosis or any phase-specific arrest. Fourth, the lack of DNA ladder degradation. And fifth, overexpression of the antiapoptotic BCL-2 gene does not significantly affect KF action.

The finding that high levels of expression of BCL-2, MDR1, or HER2/NEU genes did not significantly reduce the cytotoxic activity of KF is relevant, given that overexpression of these genes is common in human carcinogenesis and involved in the pattern of primary and acquired resistance to anticancer treatment. These results also support a non-apoptotic mode of action (as discussed above for BCL-2), and imply differences in the mechanism of action, and probably of resistance, with most available antitumor drugs. The lack of effect of MDR1 overexpression raises the possibility of combined treatments of KF with other drugs and of the use of KF as a second-line treatment in patients developing resistance to conventional drugs mediated by P-glycoprotein. Additionally, KF may be useful against tumors overexpressing HER2/NEU that are usually linked to high malignancy and bad prognosis.

The comparable activity of KF on the two series of cancer cell lines studied indicates that its action is mostly unrelated to p53 status (wild-type in MCF7 and LNCaP; mutated in PC3, DU145, MDA-MB-231, BT474, SKBR-3), and expression of androgen receptor (LNCaP are positive; PC3 and DU145 are negative) and estrogen receptor (MCF7 and BT474 are positive; MDA-MB-231 and SKBR-3 are negative).

Oncosis induction is sometimes thought to be a less specific mechanism than apoptosis, which may indicate potential side effects based on the features of necrotic cell death. Against this idea, the comparative study done in normal and tumor human cells shows that KF has a preferential cytotoxic activity on neoplastic versus normal cells. This result agrees with the promising toxicity data obtained in preclinical (28) and phase I clinical trials (8, 29).

Our results show that KF is a very potent cytotoxic drug that displays an unusual and interesting mode of action. We have characterized several key features of KF action in human prostate and breast cancer cells at the cellular and molecular level, which constitute the basis for additional in-depth studies and analysis of KF effects on other types of cancer cells. More studies are needed to fully characterize the antitumor action of KF including the identification of its primary target(s), the precise molecular mechanism of cytoplasmic organelle damage, or the basis for its preferential effect on tumor cells.

We thank those who appear in “Materials and Methods” for providing us with cells or antibodies, Robin Rycroft for his valuable assistance in the preparation of the English manuscript, and Teresa Martı́nez for her expert technical assistance.