Báo cáo khoa học: The cytoskeleton, a structure that is susceptible to the toxic mechanism activated by palytoxins in human excitable cells

The cytoskeleton, a structure that is susceptible to the toxic mechanism activated by palytoxins in human excitable cells M. Carmen Louzao1, Isabel R. Ares1, Mercedes R. Vieytes2, Iago Valverde3, Juan M. Vieites3, Takeshi Yasumoto4 and Luis M. Botana1 1 Departamento de Farmacologia, Facultad de Veterinaria, Universidad de Santiago de Compostela, Lugo, Spain 2 Departamento de Fisiologia Animal, Facultad de Veterinaria, Universidad de Santiago de Compostela, Lugo, Spain 3 ANFACO-CECOPESCA, Campus Universitario de Vigo, Vigo, Pontevedra, Spain 4 Japan Food Research Laboratories, Tama Laboratory, Tokyo, Japan Keywords actin cytoskeleton; laser-scanning cytometry; membrane potential; ostreocin-D; palytoxin Correspondence L. M. Botana, Departamento de Farmacologia, Facultad de Veterinaria, 27002 Lugo, Spain Fax⁄Tel: +34 982 252 242 E-mail: ffbotana@lugo.usc.es (Received 31 October 2006, revised 7 February 2007, accepted 14 February 2007) doi:10.1111/j.1742-4658.2007.05743.x Palytoxin is a marine toxin responsible for a fatal type of poisoning in humans named clupeotoxism, with symptoms such as neurologic distur-bances. It is believed that it binds to the Na+ ⁄K+-ATPase from the extra-cellular side and modifies cytosolic ions; nevertheless, its effects on internal cell structures, such as the cytoskeleton, which might be affected by these initial events, have not been fully elucidated. Likewise, ostreocin-D, an analog of palytoxin, has been only recently found, and its action on excit-able cells is therefore unknown. Therefore, our aim was to investigate the modifications of ion fluxes associated with palytoxin and ostreocin-D activ-ities, and their effects on an essential cytoskeletal component, the actin system. We used human neuroblastoma cells and fluorescent dyes to detect changes in membrane potential, intracellular Ca2+ concentration, cell detachment, and actin filaments. Fluorescence values were obtained with spectrofluorymetry, laser-scanning cytometry, and confocal micro-scopy; the last of these was also used for recording images. Palytoxin and ostreocin-D modified membrane permeability as a first step, triggering depolarization and increasing Ca2+ influx. The substantial loss of filamen-tous actin, and the morphologic alterations elicited by both toxins, are possibly secondary to their action on ion channels. The decrease in poly-merized actin seemed to be Ca2+-independent; however, this ion could be related to actin cytoskeletal organization. Palytoxin and ostreocin-D alter the ion fluxes, targeting pathways that involve the cytoskeletal dynamics of human excitable cells. Palytoxin is an extremely toxic product that was first isolated from the zoanthid Palythoa toxica [1]. This in dogs [4]. Ostreocins are not well-known congeners of the palytoxin found in Ostreopsis siamensis [5,6]. toxin is a large, water-soluble polyalcohol with a wide Likewise, almost nothing is known about a still distribution in algae, crabs, herbivorous fish, and the surgeonfish. Palytoxin enters the food chain and causes a fatal type of poisoning named clupeotoxism [2,3]. This compound is highly toxic for mammals, with an LD50 of 33 ngÆkg)1 when administered intravenously unnamed toxin produced by Ostreopsis ovata that may be related to ostreocins and palytoxins [5]. The putative palytoxin receptor is the Na+ ⁄K+-ATPase. The membrane enzyme Na+ ⁄K+-ATPase uses ATP hydrolysis to pump 2 K+ and 3 Na+ in Abbreviations bis-oxonol, bis-(1,3-diethylthiobarbituric acid) trimethine oxonol; F-actin, filamentous actin; LSC, laser-scanning cytometry. FEBS Journal 274 (2007) 1991–2004 ª 2007 The Authors Journal compilation ª 2007 FEBS 1991 Palytoxin-induced changes to the cytoskeleton M. C. Louzao et al. opposite directions, thereby creating an ionic asymmetry that is responsible for the electrochemical potential difference of Na+ and K+ across the cell membrane [7]. The marine toxin appears to bind to the Na+ ⁄K+-ATPase from the extracellular side and elicit a nonselective cation leak pathway, possibly by dis-rupting the strict coupling between the pump’s inner and outer gates, allowing them to both be open [8,9]. Several studies have shown that the palytoxin-induced channel permits the passage of cations such as Na+ and K+ but not Ca2+ [10–12]. Ouabain, a well-known inhibitor of the sodium pump, hampers the fluxes evoked by palytoxin [13] and seems to induce detach-ment of cells [14]. Likewise, the phenomena of cell– matrix and cell–cell adhesion have been related to 2 1.8 1.6 1.4 1.2 1 Ostreocin-D 75 nM Gramicidin Palytoxin 75 nM Toxin filamentous actin (F-actin) [15,16]. Until now, the effects of palytoxins on the actin cytoskeleton of nerve cells, which comprise one of its 0.8 0 200 400 600 800 1000 Time (s) main targets in mammals, have not been evaluated. The ability of excitable cells to generate all-or-none action potentials in response to depolarizing stimuli is well known. There are some reports describing the Fig. 1. Palytoxin-induced and ostreocin-D-induced changes in mem-brane potential in human neuroblastoma cells loaded with the membrane potential dye bis-oxonol. Results are expressed as relat- ive fluorescence ± SEM (n ¼ 4). membrane depolarization induced by palytoxin in neuronal cells [17–19]. However, surprisingly, this membrane potential and the increase in intracellular Na+ and further Ca2+ movements has never been linked with a possible effect of palytoxin on the cyto-architecture of excitable cells. On the basis of previous information, the specific aim of this work was to investigate and relate the effects of palytoxins (a) on ion fluxes (shown as changes in membrane potential or alterations in intracellular Ca2+) and (b) on actin cyto- therefore the probe only indicates changes in mem-brane potential [21]. Process conditions such as con-centration of dye were taken from our previous studies [20,22,23]. The dye was allowed to equilibrate with the cells for 15–20 min before any experiment was begun. Downward and upward deflections of the fluorescence tracings represent hyperpolarization and depolari- skeletal dynamics, morphology and cell attachment, by zation, respectively. Figure 1 shows that 75 nm using human neuroblastoma cells. Herein, we present new findings on the mode in which both compounds induce toxicity in a neuronal model. Knowledge of the key points involved in the cellular damage is funda-mental for establishing functional parameters with practical application in the design and development of in vitro detection techniques for seafood control pro-grams. This information is very useful, as it forms the basis for developing biotoxin detection methods that could provide an alternative to animal testing. Results In this study, we determined the electrophysiologic consequences of palytoxin action in excitable cells. We have previously developed an assay to detect changes in membrane potential by using the sensitive dye bis-(1,3-diethylthiobarbituric acid) trimethine oxonol (bis-oxonol) [20]. The negative charge of bis-oxonol prevents its accumulation in the mitochondria, and palytoxin or ostreocin-D induced depolarization in neuroblastoma cells. The next step was to investigate whether, together with this depolarization, there was an increase in cyto-solic Ca2+. It is known that the use of dual indicators to study intracellular Ca2+ minimizes the contribution of artefactual changes to the fluorescent signal that are not related to variations in Ca2+ [24,25]. Taking into account this fact, we used a combination of Fluo-3 and Fura red to perform this kind of assay. The Fluo-3 and Fura red fluorescence intensities are increased and reduced, respectively, when the intracellular free Ca2+ rises [26]; in consequence, an elevation in the Fluo-3⁄ Fura red fluorescence ratio is produced. Figure 2A shows an increase of cytosolic Ca2+ in neuroblastoma cells loaded with Fluo-3 and Fura red after addition of palytoxins to the solution. This ion change might result from the influx of external Ca2+ or Ca2+ release from internal stores. To clarify this, the same experi-ments were carried out in nominally Ca2+-free 1992 FEBS Journal 274 (2007) 1991–2004 ª 2007 The Authors Journal compilation ª 2007 FEBS M. C. Louzao et al. Palytoxin-induced changes to the cytoskeleton A 3 Control Palytoxin 75 nM 80 Ostreocin-D 75 nM 70 2 60 50 40 1 30 20 10 100 nM palytoxin 1 nM palytoxin Control 0 0 100 200 300 400 500 600 700 Time (s) 0 0.1 1 10 100 Time (h) B 3 Control Palytoxin 75 nM Ostreocin-D 75 nM 2 Fig. 3. Time course of cell detachment. Percentage of cell detach-ment induced by 1 nM and 100 nM palytoxin in neuroblastoma cells. Results are average ± SEM (n ¼ 4). 140 1 120 100 0 80 0 100 200 300 400 500 600 700 Time (s) Fig. 2. Calcium response of neuroblastoma cells to 75 nM palytoxin or 75 nM ostreocin-D in Ca2+-containing solution (A) and in nomin-ally Ca2+-free medium (B). Results are presented as Fluo-3⁄Fura red fluorescence ratios. The black arrow indicates addition of tox-ins. Fluorescent signals were collected during scans that were per-formed every 5 s for up to 10 min. Results are average ± SEM of at least 50 cells per assay (n ¼ 3). 60 40 20 0 –20 0.001 0.01 0.1 1 10 100 medium, and we found that neither palytoxin nor ost-reocin-D modified the cation concentration within neuroblastoma cells (Fig. 2B). These results suggest that toxins induce an extracellular Ca2+ influx. Modifications of intracellular Ca2+ levels could be associated with cytotoxic effects. In order to examine whether palytoxin affected the attachment of cells to the extracellular matrix, we determined, by additional fluorimetric microplate assays, the percentage of paly-toxin-induced cell detachment (Fig. 3). The time course [Palytoxin] nM Fig. 4. Effect of palytoxin on cytoskeleton architecture of neuro-blastoma cells. Dose-dependent disruption of the actin filament system after 24 h of incubation with palytoxin. Results are aver-age ± SEM (n ¼ 4). described in Experimental procedures. It is important to point out that a decrease in the fluorescent signal of this dye indicates F-actin disassembly. Treatment of cells with palytoxin for 24 h led to dose-dependent dis- of cell detachment experiments revealed that less than ruption of polymerized actin (Fig. 4), which was 10 h of treatment with an effective concentration of 100 nm palytoxin was sufficient to detach more than 60% of cells from the substratum. Because the cyto-skeleton is involved in cell attachment to the extra-cellular matrix, we also analyzed the effect of palytoxin on this structure. Neuroblastoma cells were incubated with the toxin, and their actin cytoskeletons were then stained with Oregon green 514 phalloidin, as already observed at 0.05 nm. Complete F-actin break-down was found with palytoxin concentrations ranging from 0.1 nm to 100 nm. On the basis of previous results, we investigated whe-ther 75 nm palytoxin and 75 nm ostreocin-D (concen-trations that clearly depolarize, increase cytosolic Ca2+, and cause complete F-actin disruption in 24 h) modified the actin filament system in a short period of time. The FEBS Journal 274 (2007) 1991–2004 ª 2007 The Authors Journal compilation ª 2007 FEBS 1993 Palytoxin-induced changes to the cytoskeleton M. C. Louzao et al. 10 min Control Palytoxin 63 50 Black: Palytoxin Grey: Control 37 25 20 0 Green Max Pixel 9000 Green Max Pixel 9000 Green Max Pixel 9000 10 min Control Ostreocin-D 76 60 Black: Ostreocin-D Grey: Control 45 30 15 0 Green Max Pixel 9000 Green Max Pixel 9000 Green Max Pixel 9000 Fig. 5. Representative diagrams obtained with LSC, showing the fluorescence distribution of control cells (left column) and cells incubated with palytoxin (upper central) or ostreocin-D (lower central) for 10 min in a Ca2+-containing medium. Spots in the scattergram plot corres-pond to microfilaments of neuroblastoma cells stained with fluorescent phalloidin. Comparative profiles of fluorescence distribution between controls and treated cells derived from the selected regions of scattergrams are shown in histograms (right column). fluorescence associated with the microfilaments of neuro-blastoma cells was measured in a large number of cells using laser-scanning cytometry (LSC). After 10 min of incubation with palytoxin or ostreocin-D (Fig. 5), the signal from cells was decreased with respect to controls. 10 min, although the disappearance of F-actin was intensified (Fig. 8). On the other hand, 30 min after addition of ostreocin-D to the solution, the cells began to lose their normal array of actin fibers. The intense peripheral bands vanished, being replaced by diffuse This loss of fluorescence intensity indicated a F-actin staining that did not seem to cause modifica- 20 ± 6.2% reduction in the actin cytoskeletons of paly-toxin-treated cells and a 19 ± 2% reduction in ostreo-cin-D-treated cells. At the same time, images captured through confocal microscopy allowed us to evaluate the outcome of F-actin alterations on the morphology of the cells. Control cells were flat and extended, with mod-erately long cell processes. In addition, their actin cytoskeletons were intact and distributed throughout tions of cell shape (Fig. 8). Finally, we evaluated the progress of the cytoskeletal effects caused by the toxins after hours of treatment. LSC examination (Fig. 9) showed that palytoxin and ostreocin-D induced, respectively, decreases in poly-merized actin of 51 ± 6.4% and 45 ± 4% in 4 h. Visualization of cells showed a substantial reduction of overall fluorescence, with a few, more intense, patches the cytoplasm. Neuroblastoma cells incubated for of F-actin in the cytoplasm; F-actin staining was rather 10 min with the toxins did not show relevant changes in shape. However, the F-actin bands surrounding the cell margin became less dense, and a small degree of loss of cytoplasmic fibers was also seen (Fig. 6). F-actin levels continued to fall in cells exposed for 30 min to the action of the toxins (Fig. 7). In this case, LSC revealed that palytoxin reduced the amount of F-actin by 27 ± 1%, whereas ostreocin-D reduced it by 35 ± 6%. Confocal images showed that neuroblastoma cells treated for 30 min with palytoxin were morphologically similar to those incubated for widespread, especially in cells treated with palytoxin. In addition, with this duration of incubation, it was found that both toxic compounds induced a retraction of the cellular body, as shown in transmission images (Fig. 10). This retraction is closely related to the cell detachment observed in Fig. 3. On the basis of the data obtained above, it was of interest to know whether the alterations in micro-filaments caused by palytoxins were related to their effects on intracellular Ca2 . Therefore, we decided to carry out actin cytoskeleton experiments in nominally 1994 FEBS Journal 274 (2007) 1991–2004 ª 2007 The Authors Journal compilation ª 2007 FEBS M. C. Louzao et al. Palytoxin-induced changes to the cytoskeleton Fig. 6. Representative images obtained by confocal microscopy of neuroblastoma cells whose F-actin was stained with Oregon green 514 phalloidin. Fluorescence images show the distribution pattern of microfilaments in control cells (upper left) and cells exposed to palytoxin (upper central) or ostreocin-D (upper right) for 10 min in a Ca2+-containing medium. Transmission images of the same cells are displayed (lower). Scale bar ¼ 50 lm. Ca2+-free medium, incubating neuroblastoma cells logic symptoms in humans. Palytoxin, isolated from with the toxins for 4 h (the time at which the larger decrease in polymerized actin levels was observed). Controls analyzed under these conditions showed iden-tical characteristics as when they were incubated in a Ca2+-containing medium; however, the response of treated cells was unexpected. In Fig. 11, scattergrams and histograms show that after 4 h of incubation of cells in nominally Ca2+-free medium, the F-actin con-tent was reduced by 47 ± 4.4% with palytoxin and by 40 ± 2.1% with ostreocin-D. No substantial differ-ences were observed in the palytoxin effect as com-pared to incubation with Ca2+-containing medium. Palythoa spp., is one of the most potent marine toxins known. We report here the effects of palytoxin on membrane potential and Ca2+ homeostasis, and how these events affect cytoskeletal dynamics in an excit-able cell model. Moreover, we studied ostreocin-D, a new analog of palytoxin, whose biological activity on a neuronal model is analyzed here for the first time. Na+ ⁄K+-ATPase is the suggested palytoxin receptor [13]. Na+ ⁄K+-ATPase is responsible for generating and maintaining transmembrane ionic gradients that are of vital importance for cellular function and activi-ties such as volume regulation, pH maintenance, gen- Confocal imaging of samples showed weak cell retrac- eration of action potentials, and secondary active tion but a substantial loss of microfilaments; otherwise, the marginal F-actin was more resistant, especially to ostreocin-D (Fig. 12). Fluorescence values presented in this section are graphed in Fig. 13. Discussion Many phycotoxins responsible for important seafood poisoning syndromes have been associated with neuro- transport. In this case, palytoxins may interfere with the function of Na+ ⁄K+-ATPase. For instance, paly-toxin, by binding to the pump, may transform it into a channel and stimulate ion fluxes, as has been previously suggested [8,27]. We observed that palytoxin and ostreo-cin-D induced depolarization in human neuroblastoma cells. It is interesting to point out that we also tested the effect of a purified extract of Ostreopsis ovata (kindly provided by T. Yasumoto). This toxic extract, which has been closely associated with palytoxins, also FEBS Journal 274 (2007) 1991–2004 ª 2007 The Authors Journal compilation ª 2007 FEBS 1995 ... - tailieumienphi.vn