- Email: [email protected]
Caulerpenyne, a toxin from the seaweed Caulerpa taxifolia, depresses afterhyperpolarization in invertebrate neurons
PII: S 0 3 0 6 - 4 5 2 2 ( 0 1 ) 0 0 3 6 5 - 7
Neuroscience Vol. 107, No. 3, pp. 519^526, 2001 ß 2001 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522 / 01 $20.00+0.00
www.neuroscience-ibro.com
CAULERPENYNE, A TOXIN FROM THE SEAWEED CAULERPA TAXIFOLIA, DEPRESSES AFTERHYPERPOLARIZATION IN INVERTEBRATE NEURONS R. MOZZACHIODI,1 R. SCURI, M. ROBERTO and M. BRUNELLI* Department of Physiology and Biochemistry `G. Moruzzi', University of Pisa, Via S. Zeno 31, 56127 Pisa, Italy
AbstractöThe massive invasion of the Mediterranean Sea by the tropical seaweed Caulerpa taxifolia (Vahl) C. Agardh has stimulated several investigations in order to test the environmental risk from an ecotoxicological point of view. The studies carried out on various experimental models have shown that caulerpenyne, the major metabolite synthesized by the seaweed, a¡ects several cellular and molecular targets. In addition, neurological disorders have been reported in patients who accidentally ate C. taxifolia, but no evidence about the potential e¡ects of the seaweed and of its metabolites on nerve cells were up to now available. Herein we describe that caulerpenyne modi¢es the electrical properties of touch mechanosensory cells of the leech Hirudo medicinalis. The physiological ¢ring of these cells causes an afterhyperpolarization that is mainly due to the activity of the Na /K -ATPase and to a lesser extent to a calcium-dependent potassium current. Caulerpenyne depressed this afterhyperpolarization; the e¡ect was dose-dependent and partially reversible. Experiments have been carried out in order to understand the mechanism through which caulerpenyne reduced the afterhyperpolarization. The action of the biotoxin has been tested in the presence of pharmacological blockers of calcium-dependent potassium channels such as cadmium and apamin. In these experimental conditions, caulerpenyne still reduced the residual afterhyperpolarization, suggesting a direct e¡ect of the toxin on the Na /K -ATPase. In order to test this hypothesis, we have performed experiments where the Na /K -ATPase was activated by the intracellular injection of sodium and where also its basal activity was modi¢ed as well. From the data collected we suggest that caulerpenyne inhibits both the basal and the sodium-induced activity of the Na /K -ATPase in leech touch neurons. ß 2001 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: Hirudo medicinalis, Na /K -ATPase, sesquiterpene, T sensory neurons.
et al., 1996). In this second case, the toxin reduces the cytosolic ATP-dependent Ca2 accumulation but does not provoke a release of sequestered Ca2 (Pesando et al., 1996). This e¡ect is similar to that due to thapsigargin, a speci¢c inhibitor of reticular Ca2 -ATPase (Pesando et al., 1996). It has been proved that CYN blocks the mitotic cycle of sea urchin embryos at metaphase and also inhibits the stimulation of mitogen-activated protein kinase (MAPK) (Pesando et al., 1998, 1999). In addition, antineoplastic activity of CYN has been described. The sesquiterpene has demonstrated growth-inhibitory e¡ects in eight cancer cell lines of human origin, but its potential targets still remain to be identi¢ed (Fischel et al., 1995). De Haro et al. (1993) reported that patients with food poisoning due to the ingestion of Sarpa salpa, a ¢sh consuming C. taxifolia, showed neurological disorders such as amnesia, vertigo, and hallucinations; these symptoms have been ascribed to CYN. Following this clinical observation, we have begun to analyze CYN e¡ects on neural activity. We used the ganglionic nervous system of the leech Hirudo medicinalis. This invertebrate has proved to be very useful for studying cellular physiology, synaptogenesis (Fernandezde-Miguel and Drapeau, 1995), pharmacology (Kleinhaus and Angstadt, 1995) and also molecular
The seaweed Caulerpa taxifolia (Vahl) C. Agardh has been accidentally introduced into the Mediterranean Sea and rapidly invaded speci¢c coast areas. It contains mono- and sesquiterpenes such as taxifolial A, B, C and D in larger amounts than Caulerpa species endemic in the tropics (Guerriero et al., 1992). The major metabolite produced by C. taxifolia is the sesquiterpene caulerpenyne (CYN). In order to screen the environmental damage caused by the anomalous spreading of C. taxifolia into the Mediterranean, the metabolites of CYN have been tested on di¡erent experimental models. The data collected provide evidence that CYN (1) exhibits antiproliferative activity in bacterial cultures of Planococcus (Giannotti et al., 1994) and (2) inhibits the ¢rst cleavage of sea urchin eggs without a¡ecting fertilization (Pesando 1 Present address: Department of Neurobiology and Anatomy, University of Texas ^ Houston, School of Medicine, 6431 Fannin, Houston, TX 77030, USA. *Corresponding author. Tel : +39-50-554030; fax: +39-50-552183. E-mail address: [email protected] (M. Brunelli). Abbreviations : 5HT, 5-hydroxytryptamine; AHP, afterhyperpolarization; CYN, caulerpenyne; DMSO, dimethylsulfoxide ; IK=Ca , Ca2 -dependent K current ; MAPK, mitogen-activated protein kinase ; Rm , input resistance ; T, touch-sensitive; Vm , transmembrane potential.
519
NSC 5206 12-11-01
520
R. Mozzachiodi et al.
mechanisms underlying complex behaviors (Brodfuehrer et al., 1995). The leech nervous system is made up of a chain of 21 segmental ganglia connected with a cephalic and a caudal ganglion. Each segmental ganglion contains about 400 cells including three classes of sensory neurons responding to mechanical stimuli of di¡erent intensity: touch (T), pressure and nociceptive cells. These neurons can be identi¢ed within the ganglion by the size and location of their somata, and by the shape and amplitude of their action potentials (Nicholls and Baylor, 1968). By testing CYN on di¡erent neurons of the ganglia, we have detected important e¡ects on the electrical properties of T sensory cells. Therefore, we have focused our investigations on these neurons. In T cells, the discharge of action potentials induced either by receptive ¢eld activation or by intracellular repetitive stimulation elicits an afterhyperpolarization (AHP) due both to Ca2 -dependent K currents (IK=Ca ) and mainly to the activity of the electrogenic Na /K -ATPase (Baylor and Nicholls, 1969a; Jansen and Nicholls, 1973). Therefore, changes of AHP amplitude can be directly related to modi¢cations of the Na /K -ATPase activity and T neurons can represent a cellular model for monitoring the activity of the Na /K -ATPase with an electrophysiological approach. Previously, we have demonstrated that also the endogenous neurotransmitter serotonin (5-hydroxytryptamine or 5HT) reversibly reduces AHP through the inhibition of the electrogenic pump (Belardetti et al., 1984; Catarsi and Brunelli, 1991). In this paper we report evidence that CYN depresses the AHP amplitude through a selective inhibition of the Na /K -ATPase. A preliminary account of some of these ¢ndings has been presented at a meeting of the Society for Neuroscience (Brunelli et al., 1998).
Fig. 1. E¡ects of CYN on the electrical properties of T neurons. Dose-dependent e¡ect of CYN on AHP amplitude, Vm and Rm . CYN was tested at two di¡erent concentrations : 10 WM (n = 5) and 50 WM (n = 13). In the graphs of the ¢rst ¢ve ¢gures, the abscissas always describe the timing of the experimental phases; the means of the parameter values are expressed in the ordinates ; all data have been normalized to the initial value recorded in saline (control value). Statistical di¡erence (*P 6 0.05) was determined by performing Wilcoxon matched-pairs signed-ranks test.
EXPERIMENTAL PROCEDURES
Animals and surgery Adult leeches (H. medicinalis) purchased from a local supplier were maintained at 16³C, under natural daylight rhythm. Leeches were anesthetized in 10% ethanol in tap water for 10 min and then the nerve cord was exposed by cutting the body wall along the midline and by opening the ventral sinus. The surgery was performed at room temperature. A short chain of ganglia was removed at mid-body level and kept at 16³C in saline solution (see below). For each experiment a single ganglion was pinned with the ventral side up (where the sensory neurons are located) to the bottom of a small recording chamber coated with silicone elastomer (Sylgard0 , Dow Corning, Midland, MI, USA). Solutions Unless otherwise stated, the following saline solution was used, for both surgery and electrophysiological recordings: 115 mM NaCl, 4 mM KCl, 1.8 mM CaCl2 , 10 mM glucose, bu¡ered at pH 7.4 with 10 mM tris(hydroxymethyl)aminomethane^maleate. In one group of experiments, K -free saline was obtained by replacing 4 mM KCl with 4 mM NaCl in order to block the basal activity of the sodium pump. 4 mM CsCl was added to K -free saline (Schlue and Deitmer, 1984; Catarsi and Brunelli, 1991; Catarsi et al., 1993) to re-activate the Na /K pump. CYN, from fresh algae collected at Imperia (Liguria, Italy),
was puri¢ed with the extraction procedure used by Pesando et al. (1996). During electrophysiological experiments ganglia were constantly perfused with the saline solution at a rate of 1.5 ml/ min by means of a peristaltic pump (Master£ex, Cole-Palmer Instrument, Chicago, IL, USA) and all the drugs (from Sigma, St. Louis, MO, USA) were freshly dissolved in appropriate saline just before their application. CYN was previously dissolved in dimethylsulfoxide (DMSO) 1:1000, sonicated and then added to the saline solution. In control experiments, DMSO at that concentration did not a¡ect the electrical properties of T cells (data not shown). Electrophysiological technique T neurons were identi¢ed by the size and location of their cell bodies as well as by their ¢ring pattern (Nicholls and Baylor, 1968). Intracellular recordings were performed in current clamp mode by using borosilicate microelectrodes ¢lled with 4 M potassium acetate (60^80 M6 impedance). In a series of investigations, neurons were impaled with 2 M sodium acetate microelectrodes in order to increase the activity of Na /K -ATPase. All the experiments were carried out at room temperature. AHP was induced by injecting 10-s trains of intracellular depolarizing impulses (200 ms, 2.5 Hz). The discharge frequency of T cells was kept constant by adjusting the amount of injected current (0.4^0.8 nA). Since AHP is very sensitive to the quality of the impalement, only cells with AHP of at least 10^15 mV were used
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
for analysis. All the experiments were stored on a video tape recorder connected to a pulse code modulator (PCM 501 ES, Sony) for further data analysis. Statistical evaluation The amplitude of AHP was measured from the baseline of the resting potential before the stimulation, to the peak of maximal hyperpolarization. The action potential amplitude was measured relative to the prespike membrane potential. Action potential width was measured from a ¢xed fraction of the spike at a point corresponding to 35% of its maximal amplitude. For the statistical analysis of the data, Student's t-test was used when the Gaussian distribution was satis¢ed and non-parametric tests (Wilcoxon matched-pairs signed-ranks test and Mann^Whitney U-test) when the normal distribution could not be satis¢ed. The Wilcoxon matched-pairs signed-ranks test is similar to t-test for paired samples and it is used in order to test the same group of data twice and to anticipate a relationship. The Mann^Whitney U-test is used to compare two di¡erent groups in a study where the raw data were converted into ranks. In all the graphs, signi¢cance (P 6 0.05) is indicated with an asterisk. Statistical analysis was performed with the SPSS software package (Advanced Models tm 9.0).
521
potential (Vm ), input resistance (Rm , evaluated by injecting 0.5 nA hyperpolarizing impulses of 200 ms), amplitude and duration of action potential and amplitude of AHP. Fifteen minutes after application, CYN produced a reduction of the AHP amplitude. This e¡ect was dosedependent, became maximal after 15 min wash with saline solution (P 6 0.01) and started to recover after 30 min wash (Figs. 1 and 2A). A statistically signi¢cant depolarization of Vm (P 6 0.01) and a substantial reduction of Rm (P 6 0.01) were detected in 50 WM CYN (Fig. 1). We did not observe consistent changes in the duration of the action potential, whereas a signi¢cant reduction of the spike amplitude occurred during 50 WM CYN application (P = 0.018, Fig. 3). Since Rm changes a¡ect the AHP amplitude (shunt e¡ect), we compared the AHP amplitude with Rm changes and we found that the reduction of the AHP amplitude was signi¢cantly larger (30%, P = 0.0015) than the decrease of Rm (Fig. 2B). This result suggested a direct action of CYN on the components generating the AHP. E¡ect of CYN on AHP amplitude after application of CdCl2
RESULTS
E¡ects of CYN on T neuron activity We have detected the e¡ects of di¡erent concentrations of CYN (10 and 50 WM) on the following electrophysiological parameters of T neurons: membrane
Experiments were performed using 0.1 mM CdCl2 in normal saline to block IK=Ca (Stewart et al., 1989; Catarsi and Brunelli, 1991), one of the two components generating the AHP. Five minutes application of CdCl2 produced a reduction in the AHP amplitude down to
Fig. 2. E¡ects of 50 WM CYN on AHP amplitude and Rm . (A) Traces showing AHPs recorded from a single neuron. The action potentials generating AHP are clipped and indistinguishable from each other at the time scale shown. (B) Comparison between the e¡ect of 50 WM CYN on AHP amplitude and on Rm (n = 13). Statistical analysis was performed between the two curves with Mann^Whitney U-test (*P 6 0.05).
NSC 5206 12-11-01
522
R. Mozzachiodi et al.
Fig. 3. E¡ect of 50 WM CYN on the spike amplitude. (A) Traces showing action potentials evoked in a T cell before (control), after 15 min 50 WM CYN perfusion and during the washout. (B) Graph showing the e¡ect of 50 WM CYN on the amplitude of the evoked action potentials (n = 9). The statistical analysis was performed by Wilcoxon matched-pairs signedranks test (*P 6 0.05).
60% of the control value recorded in saline. No further decrease of the AHP was detectable when CdCl2 application was prolonged as previously demonstrated by Catarsi and Brunelli (1991). After the initial 5 min perfusion with CdCl2 , 50 WM CYN was applied for 15 min while the perfusion with the blocking agent was continued. A further signi¢cant reduction of the AHP amplitude down to 25% of the initial value was observed (P = 0.0277). After 60 min wash, ¢rst with CdCl2 solution and then with saline, CYN e¡ects recovered to about 60% (Fig. 4A, B). As with CYN alone, a reduction of Rm was detected, but the AHP depression was signi¢cantly larger than the decrease of Rm (P = 0.0281, Fig. 4B). E¡ect of CYN on AHP amplitude after application of apamin Since cadmium inhibited IK=Ca through the block of Ca2 channels also a¡ecting Rm and producing a strong shunt of the AHP current, we used a more speci¢c IK=Ca inhibitor, the peptide apamin, which has been successfully used to block IK=Ca in several neurons (Hugues et al., 1982; Zhang and Krnjevic, 1987). During preliminary experiments we found that 15 min perfusion with 1 nM apamin produced a 25% reversible reduction of the AHP amplitude without a¡ecting Rm . More prolonged applications of apamin did not further decrease the AHP amplitude (Fig. 5B). When 50 WM CYN was added to saline containing apamin, an evident reduction of the residual AHP was observed (P = 0.0117, Fig. 5A, C).
The e¡ect of CYN was maximal after 10 min wash with apamin, in agreement with the kinetics shown by CYN alone, and recovered to about 70% during 60 min wash. The maximal AHP depression detected was signi¢cantly larger than the Rm reduction (33%, P 6 0.001, Fig. 5C). E¡ect of CYN on Na+-stimulated activity of the Na+/K +-ATPase Since CYN still a¡ected the residual AHP amplitude when IK=Ca was blocked with di¡erent pharmacological agents, in a further group of experiments we tested the hypothesis of a direct inhibition of the Na /K -ATPase by CYN. When a T neuron was impaled with 2 M sodium acetate microelectrode, the leakage of Na into the soma rapidly induced a clear-cut hyperpolarization due to the activation of the electrogenic pump (Baylor and Nicholls, 1969a; Catarsi and Brunelli, 1991; Catarsi et al., 1993): Vm hyperpolarized by about 12 mV from the resting potential (b in Fig. 6A) and then reached a steady-state level (a in Fig. 6A). When CYN was added to the perfusion bath, the hyperpolarization turned into a depolarization (b in Fig. 6A) suggesting an inhibition of the Na /K -ATPase. In some experiments we perfused the ganglia with 50 WM CYN for 15 min before penetrating T cells with 2 M sodium acetate microelectrodes and we did not measure any detectable change of Vm during a further 10 min application of 50 WM CYN. Only when CYN was washed out did that neuron hyperpolarize (Fig. 6B).
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
523
Fig. 4. E¡ects of 50 WM CYN on AHP amplitude and on Rm during perfusion with 0.1 mM CdCl2 . (A) Traces showing AHPs recorded from a single T neuron. (B) Graph showing means of AHP amplitudes and Rm (n = 8). Cadmium perfusion did not prevent the e¡ect of CYN on AHP amplitude ; the depression of AHP driven by CYN was still present and was similar to that produced by CYN alone. Asterisk indicates a signi¢cant di¡erence (P 6 0.05) between AHP reduction and Rm change (Mann^Whitney U-test).
E¡ect of CYN on the basal activity of the Na+/K +-ATPase We also tested the e¡ect of CYN on the basal activity of the Na /K -ATPase. Perfusion with K -free saline blocked the Na /K -ATPase and led to a biphasic change of Vm in T cells (b in Fig. 7A, B): an early hyperpolarization was followed by a delayed depolarization due to the inactivation of the electrogenic pump as previously described by Schlue and Deitmer (1984). When the sodium pump was blocked, the depolarization phase reached a steady-state (dashed line in Fig. 7A, B) (Catarsi and Brunelli, 1991). Ten minutes perfusion with 4 mM CsCl, a Na /K -ATPase activator (Skou, 1965), shifted the Vm polarization back to the initial value as a consequence of the reactivation of the pump (b in Fig. 7A, B). Once the membrane potential had settled to its new value, the Vm maintained this polarized value as long as the T cell was perfused with K -free saline containing 4 mM CsCl (b in Fig. 7A). When 50 WM CYN was added to K -free solution containing 4 mM CsCl, Vm showed a remarkable depolarization (a in Fig. 7A). In another group of experiments, CYN was applied together with CsCl right after Vm reached the steady-state depolarization during perfusion with K -free solution (a in Fig. 7B); the e¡ect of CsCl was blocked by the toxin and the expected repolarization turned into a depolarization. The time course of Vm dur-
ing perfusion with CsCl alone (b in Fig. 7B) was statistically di¡erent from that produced by 4 mM CsCl plus 50 WM CYN (a in Fig. 7B, P = 0.034, P = 0.023).
DISCUSSION
This study reports the ¢rst evidence that CYN, the major metabolite synthesized by the so-called killer seaweed C. taxifolia, a¡ects the activity of nerve cells. We have tested the e¡ects of CYN on the electrical properties of T sensory neurons of the leech H. medicinalis, an experimental model extensively used in neurobiology. We have observed that CYN produces a clear-cut depression of the AHP that characterizes the ¢ring pro¢le of these neurons. This e¡ect is dose-dependent and coupled with a reduction of Rm and with a slight depolarization of Vm . Although a decrease of Rm a¡ects AHP amplitude (assuming Ohm's law: V = RI), we have detected a 70% AHP depression vs. a 40% Rm reduction when the AHP decrease was compared with Rm changes during CYN perfusion. This indicates that a simple linear short circuit of the currents generating the AHP failed to fully explain the action of CYN. After 60 min wash, the AHP amplitude recovered towards the control value suggesting that CYN e¡ects were long lasting but at least partially reversible. Two series of experiments provided evidence for an
NSC 5206 12-11-01
524
R. Mozzachiodi et al.
Fig. 5. E¡ects of 50 WM CYN on AHP amplitude and on Rm during 1 nM apamin application. (A) Traces showing AHPs recorded from a single T neuron. (B) The graph shows the means of AHP amplitude and Rm during apamin application (n = 5). Fifteen minutes apamin application depressed AHP to about 75% of the initial response. Prolonged applications (up to 30 min) did not further reduce the AHP amplitude. No signi¢cant change in Rm was detected. Asterisks on the AHP curve represent statistical di¡erence (P 6 0.05) determined by Wilcoxon matched-pairs signed-ranks test. (C) The graph shows the means of AHP amplitude and Rm (n = 9) during apamin and CYN application. Asterisks indicate the signi¢cant di¡erence (P 6 0.05) between AHP reduction and Rm change (Mann^Whitney U-test).
actual e¡ect of CYN on the components generating AHP. In the ¢rst one, the IK=Ca was blocked with two distinct pharmacological agents: cadmium and apamin. CdCl2 blocks the voltage-gated calcium channels from the extracellular side of the plasma membrane thus preventing the activation of IK=Ca . Apamin is a bee venom peptide that selectively blocks Ca2 -dependent K channels involved in afterpotentials (Hugues et al., 1982; Zhang and Krnjevic, 1987). In both experimental conditions, CYN was still capable of reducing the residual AHP due to the Na /K -ATPase. These results strongly support the hypothesis that CYN exerted an inhibitory action on the Na /K -ATPase. In addition, if CYN a¡ected the IK=Ca the AHP depression should occur through the closure of K or Ca2 channels with an increase of Rm . In contrast, we always measured a decrease of Rm that was signi¢cantly smaller than the maximal AHP depression in all the experimental conditions. During perfusion with apamin alone we did not detect changes in Rm , whereas a 20% reduction of Rm was observed during application of CdCl2 alone. This phenomenon does not seem to be ascribable to the action of cadmium on some membrane conductance, but it might be due in part to a decrease of the junctional re-
sistance through which T sensory neurons are coupled in the segmental ganglia (Baylor and Nicholls, 1969b). A decrease of the cytosolic calcium concentration, which can occur during CdCl2 perfusion, might lead to a reduction of junctional resistance and, consequently, of Rm (Rose and Loewenstein, 1975). In the second series of experiments, we gained insights for a direct inhibition of the Na /K -ATPase by CYN. This hypothesis could be strengthened using speci¢c blockers of the pump (such as ouabain or strophantidin), but the application of these blockers does not make it possible to keep a constant discharge frequency leading to the AHP, probably because of the excess of sodium in the T cell that cannot be pumped out (Catarsi and Brunelli, 1991). Therefore, we used di¡erent approaches such as ion manipulation of sodium pump activity. We observed that during CYN treatment the Na /K -ATPase became insensitive to intracellular injection of Na ions. In fact, the injection of Na into T neurons during perfusion with CYN turned the normal ouabain-sensitive hyperpolarization (Baylor and Nicholls, 1969a) into a depolarization. This e¡ect may be explained by the inhibition of the Na -stimulated activity of the pump: Na ions leaked into the cell could not be pumped out. When
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
525
result suggested that CYN hindered cesium from activating the Na /K -ATPase. Overall, these data provide clear evidence that the Na /K -ATPase is the molecular target of CYN. During CYN application we did not observe any change in spike duration, whereas we detected a signi¢cant decrease of spike amplitude which might be ascribed
Fig. 6. E¡ects of CYN on Na -induced activity of the Na /K ATPase. (A) Na injection induced a rapid hyperpolarization (b). CYN 50 WM shifted the hyperpolarized Vm towards more depolarized values; asterisks represent signi¢cant di¡erences (P 6 0.05; Student's t-test for unpaired samples) between mean points obtained during CYN application (b, n = 11) and control value recorded in neurons impaled with sodium acetate microelectrodes and constantly perfused with saline (a, n = 9). (B) CYN prevented the Na -dependent hyperpolarization when T cells were pre-incubated with the toxin (b). Asterisks represent signi¢cant di¡erences (P 6 0.05; Student's t-test for unpaired samples) between mean points obtained during a further 10 min CYN application (b, n = 5) and control value recorded in neurons impaled with sodium acetate microelectrodes and constantly perfused with saline (a, n = 9 as in A). Arrowheads in A and B indicate the beginning of recording with the sodium acetate electrode.
T cells were pre-incubated with CYN, the Na injection failed to hyperpolarize Vm : the inhibitory e¡ect of CYN antagonized the activation of the Na /K -ATPase by sodium also in this condition. When the basal activity of the Na /K -ATPase was blocked with K -free perfusion, typical changes of Vm were detected (Schlue and Deitmer, 1984). In these experimental conditions, cesium application induced the repolarization of Vm following the reactivation of the electrogenic pump (Skou, 1965), whereas CYN converted this expected repolarization into a depolarization; this
Fig. 7. E¡ects of CYN on the basal activity of the Na /K -ATPase. The symbols represent the means of the shift of Vm from the resting potential. (A) In T neurons perfused with K -free saline (b, n = 10), Vm showed a hyperpolarization followed by depolarization. Dashed lines represent the steady-state depolarization when the cells were constantly perfused with K -free saline. Perfusion with saline in which 4 mM KCl was replaced with 4 mM CsCl induced a repolarization (b). After repolarization was settled ¢ve cells were treated for 10 min with K -free saline plus 4 mM CsCl plus 50 WM CYN (a). Depolarization occurred. Asterisks indicate a signi¢cant di¡erence (P 6 0.05) between Vm recorded in T cells perfused with CYN and Vm recorded in T cells repolarized during CsCl perfusion (Student's t-test for unpaired samples). (B) Thirteen neurons were treated with K -free saline (b). After Vm reached the steady-state depolarization, a repolarization was evident in the ¢ve cells that were perfused with K -free saline plus 4 mM CsCl (b). When 50 WM CYN was added to the K -free saline plus 4 mM CsCl, the repolarization did not occur (a, n = 8). Asterisks indicate that a signi¢cant di¡erence occurred between the Vm recorded in the two groups of cells (Student's t-test for unpaired samples).
NSC 5206 12-11-01
526
R. Mozzachiodi et al.
to the Vm depolarization following the inhibition of the basal activity of the pump. Concerning the reduction of Rm and the depolarization of Vm detected when 50 WM CYN was applied, we can speculate that CYN, besides its inhibitory e¡ect on the Na /K -ATPase, might also a¡ect some other targets in the plasma membrane such as ion channels. For example, the activation of a not yet identi¢ed cationic conductance could account for both the depolarization and the decrease of Rm . Furthermore, we cannot rule out that CYN modulates the electrical junctions which couple each other T cells placed in the same ganglion and in adjacent ones (Baylor and Nicholls, 1969b). The CYN action on the Na /K -ATPase occurred with a certain latency suggesting the activation of a putative receptor and the involvement of a biochemical cascade. We have previously showed that in T neurons 5HT depressed AHP amplitude through the inhibition of the electrogenic pump (Catarsi and Brunelli, 1991). 5HT stimulates cAMP production in T neurons (Catarsi et al., 1993). The magnitude and the time course of the 5HT e¡ect are very similar to that produced by CYN,
but CYN does not act through serotonergic receptors and does not activate the cAMP pathways (Brunelli et al., 1998). Therefore it will be very enticing to clarify the molecular mechanism underlying CYN e¡ects. At present, some hypotheses can be made. CYN is a sesquiterpene capable of permeating the plasma membrane. For example, in sea urchin eggs CYN acts in the cytosol and inhibits the ¢rst cleavage by blocking the mitotic cycle at metaphase (Pesando et al., 1996). In this case CYN a¡ects the intracellular ATP-dependent Ca2 accumulation (Pesando et al., 1996) in a thapsigargin-like manner, and also inhibits the MAPK (Pesando et al., 1998, 1999). We may hypothesize that in T neurons CYN penetrates the membrane and a¡ects the Na / K -ATPase from the intracellular side or triggers molecular events leading to the changes of the Na /K -ATPase activity. In future researches, it will be interesting to investigate whether the inhibitory e¡ect of CYN on the Na /K ATPase detected in T cells of the leech can be observed in other tissues where the electrogenic pump plays pivotal roles.
REFERENCES
Baylor, D.A., Nicholls, J.G., 1969a. After e¡ects of nerve impulse on signaling in the central nervous system of the leech. J. Physiol. 203, 571^589. Baylor, D.A., Nicholls, J.G., 1969b. Chemical and electrical synaptic connections between cutaneous mechanoreceptor neurones in the central nervous system of the leech. J. Physiol. 203, 591^609. Belardetti, F., Brunelli, M., Demontis, G.C., Sonetti, D., 1984. Serotonin and Retzius cell depress the hyperpolarization following impulses of the leech touch cell. Brain Res. 300, 91^102. Brodfuehrer, P.D., Debski, A.E., O'Gara, B.A., Friesen, W.O., 1995. Neural control of leech swimming. J. Neurobiol. 27, 403^418. Brunelli, M., Garcia-Gil, M., Mozzachiodi, R., Roberto, M., Scuri, R., Traina, G., Zaccard, M.L., 1998. Caulerpenyne, a seaweed biotoxin, provokes the inhibition of the Na /K electrogenic pump in neurons. Soc. Neurosci. Abstr. 24, 231. Catarsi, S., Brunelli, M., 1991. Serotonin depresses the after-hyperpolarization through the inhibition of the Na /K electrogenic pump in the sensory neurones of the leech. J. Exp. Biol. 155, 261^273. Catarsi, S., Scuri, R., Brunelli, M., 1993. Cyclic AMP mediates inhibition of the Na /K electrogenic pump by serotonin in tactile sensory neurons of the leech. J. Physiol. 462, 229^242. De Haro, L., Tre¡ot, M.J., Jouglard, J., Perringue, C., 1993. Trois cas d'intoxication de type ciguateresque apre©s ingestion de sparidae de Mediterrane¨e. Ictyol. Physiol. Acta 16, 133^146. Fernandez-de-Miguel, F., Drapeau, P., 1995. Synapse formation and function: insights from identi¢ed leech neurons in culture. J. Neurobiol. 27, 367^379. Fischel, J.L., Leme¨e, R., Formento, P., Caldani, C., Moll, J.L., Pesando, D., Meinesz, A., Grelier, P., Pietra, F., Guerriero, A., Milano, G., 1995. Cell growth inhibitory e¡ects of caulerpenyne, a sesquiterpenoid from the marine alga Caulerpa taxifolia. Anticancer Res. 15, 2155^2160. Giannotti, A., Ghelardi, E., Senesi, S., 1994. Characterization of seawater bacterial communities within environments colonized by the tropical green seaweed Caulerpa taxifolia. In: Boudouresque, C.F., Meinesz, A., Gravez, V. (Eds.), First International Workshop on Caulerpa taxifolia. GIS Posidonie, Marseille, pp. 197^201. Guerriero, A., Meinesz, A., D'Ambrosia, M., Pietra, F., 1992. Isolation of toxic sesqui- and monoterpenes from tropical green seaweed Caulerpa taxifolia which has invaded the region of Cap Martin and Monaco. Helv. Chim. Acta 75, 689^695. Hugues, M., Romey, G., Duval, D., Vincent, J.P., Landuzki, M., 1982. Apamin as a selective blocker of the calcium-dependent potassium channel in neuroblastoma cells: Voltage-clamp and biochemical characterization of the toxin receptor. Proc. Natl. Acad. Sci. USA 79, 1308^1312. Jansen, J.K.S., Nicholls, J.G., 1973. Conductance changes, an electrogenic pump and the hyperpolarization of leech neurones following impulses. J. Physiol. 229, 635^655. Kleinhaus, A.L., Angstadt, J.D., 1995. Diversity and modulation of ionic conductances in leech neurons. J. Neurobiol. 27, 419^433. Nicholls, J.G., Baylor, D.A., 1968. Speci¢c modulation and receptive ¢elds of sensory neurones in CNS of the leech. J. Neurophysiol. 31, 740^756. Pesando, D., Huitorel, P., Dolcini, V., Amade, P., Girard, J.P., 1998. Caulerpenyne interferes with microtubule-dependent events during the ¢rst mitotic cycle of sea urchin eggs. Eur. J. Cell Biol. 77, 19^26. Pesando, D., Leme¨e, R., Ferrua, C., Amade, P., Girard, J.P., 1996. E¡ects of caulerpenyne, the major toxin from Caulerpa taxifolia on mechanisms related to sea urchin egg cleavage. Aquat. Toxicol. 35, 139^155. Pesando, D., Pesci-Bardon, C., Huitorel, P., Girard, J.P., 1999. Caulerpenyne blocks MBP kinase activation controlling mitosis in sea urchin eggs. Eur. J. Cell Biol. 78, 903^910. Rose, B., Loewenstein, W.R., 1975. Permeability of cell junction depends on local cytoplasmic calcium activity. Nature 254, 250^252. Schlue, W.R., Deitmer, J.W., 1984. Potassium distribution and membrane potential of sensory neurons in the leech nervous system. J. Neurophysiol. 51, 689^704. Skou, J.C., 1965. Enzymatic basis for active transport of Na and K across cell membrane. Physiol. Rev. 45, 596^617. Stewart, R.R., Nicholls, J.G., Adams, W.B., 1989. Na, K and Ca currents in identi¢ed leech neurones in culture. J. Exp. Biol. 141, 1^20. Zhang, L., Krnjevic, K., 1987. Apamin depresses selectively the afterhyperpolarization of cat spinal motoneurons. Neurosci. Lett. 74, 58^62. (Accepted 29 May 2001)
NSC 5206 12-11-01
Neuroscience Vol. 107, No. 3, pp. 519^526, 2001 ß 2001 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522 / 01 $20.00+0.00
www.neuroscience-ibro.com
CAULERPENYNE, A TOXIN FROM THE SEAWEED CAULERPA TAXIFOLIA, DEPRESSES AFTERHYPERPOLARIZATION IN INVERTEBRATE NEURONS R. MOZZACHIODI,1 R. SCURI, M. ROBERTO and M. BRUNELLI* Department of Physiology and Biochemistry `G. Moruzzi', University of Pisa, Via S. Zeno 31, 56127 Pisa, Italy
AbstractöThe massive invasion of the Mediterranean Sea by the tropical seaweed Caulerpa taxifolia (Vahl) C. Agardh has stimulated several investigations in order to test the environmental risk from an ecotoxicological point of view. The studies carried out on various experimental models have shown that caulerpenyne, the major metabolite synthesized by the seaweed, a¡ects several cellular and molecular targets. In addition, neurological disorders have been reported in patients who accidentally ate C. taxifolia, but no evidence about the potential e¡ects of the seaweed and of its metabolites on nerve cells were up to now available. Herein we describe that caulerpenyne modi¢es the electrical properties of touch mechanosensory cells of the leech Hirudo medicinalis. The physiological ¢ring of these cells causes an afterhyperpolarization that is mainly due to the activity of the Na /K -ATPase and to a lesser extent to a calcium-dependent potassium current. Caulerpenyne depressed this afterhyperpolarization; the e¡ect was dose-dependent and partially reversible. Experiments have been carried out in order to understand the mechanism through which caulerpenyne reduced the afterhyperpolarization. The action of the biotoxin has been tested in the presence of pharmacological blockers of calcium-dependent potassium channels such as cadmium and apamin. In these experimental conditions, caulerpenyne still reduced the residual afterhyperpolarization, suggesting a direct e¡ect of the toxin on the Na /K -ATPase. In order to test this hypothesis, we have performed experiments where the Na /K -ATPase was activated by the intracellular injection of sodium and where also its basal activity was modi¢ed as well. From the data collected we suggest that caulerpenyne inhibits both the basal and the sodium-induced activity of the Na /K -ATPase in leech touch neurons. ß 2001 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: Hirudo medicinalis, Na /K -ATPase, sesquiterpene, T sensory neurons.
et al., 1996). In this second case, the toxin reduces the cytosolic ATP-dependent Ca2 accumulation but does not provoke a release of sequestered Ca2 (Pesando et al., 1996). This e¡ect is similar to that due to thapsigargin, a speci¢c inhibitor of reticular Ca2 -ATPase (Pesando et al., 1996). It has been proved that CYN blocks the mitotic cycle of sea urchin embryos at metaphase and also inhibits the stimulation of mitogen-activated protein kinase (MAPK) (Pesando et al., 1998, 1999). In addition, antineoplastic activity of CYN has been described. The sesquiterpene has demonstrated growth-inhibitory e¡ects in eight cancer cell lines of human origin, but its potential targets still remain to be identi¢ed (Fischel et al., 1995). De Haro et al. (1993) reported that patients with food poisoning due to the ingestion of Sarpa salpa, a ¢sh consuming C. taxifolia, showed neurological disorders such as amnesia, vertigo, and hallucinations; these symptoms have been ascribed to CYN. Following this clinical observation, we have begun to analyze CYN e¡ects on neural activity. We used the ganglionic nervous system of the leech Hirudo medicinalis. This invertebrate has proved to be very useful for studying cellular physiology, synaptogenesis (Fernandezde-Miguel and Drapeau, 1995), pharmacology (Kleinhaus and Angstadt, 1995) and also molecular
The seaweed Caulerpa taxifolia (Vahl) C. Agardh has been accidentally introduced into the Mediterranean Sea and rapidly invaded speci¢c coast areas. It contains mono- and sesquiterpenes such as taxifolial A, B, C and D in larger amounts than Caulerpa species endemic in the tropics (Guerriero et al., 1992). The major metabolite produced by C. taxifolia is the sesquiterpene caulerpenyne (CYN). In order to screen the environmental damage caused by the anomalous spreading of C. taxifolia into the Mediterranean, the metabolites of CYN have been tested on di¡erent experimental models. The data collected provide evidence that CYN (1) exhibits antiproliferative activity in bacterial cultures of Planococcus (Giannotti et al., 1994) and (2) inhibits the ¢rst cleavage of sea urchin eggs without a¡ecting fertilization (Pesando 1 Present address: Department of Neurobiology and Anatomy, University of Texas ^ Houston, School of Medicine, 6431 Fannin, Houston, TX 77030, USA. *Corresponding author. Tel : +39-50-554030; fax: +39-50-552183. E-mail address: [email protected] (M. Brunelli). Abbreviations : 5HT, 5-hydroxytryptamine; AHP, afterhyperpolarization; CYN, caulerpenyne; DMSO, dimethylsulfoxide ; IK=Ca , Ca2 -dependent K current ; MAPK, mitogen-activated protein kinase ; Rm , input resistance ; T, touch-sensitive; Vm , transmembrane potential.
519
NSC 5206 12-11-01
520
R. Mozzachiodi et al.
mechanisms underlying complex behaviors (Brodfuehrer et al., 1995). The leech nervous system is made up of a chain of 21 segmental ganglia connected with a cephalic and a caudal ganglion. Each segmental ganglion contains about 400 cells including three classes of sensory neurons responding to mechanical stimuli of di¡erent intensity: touch (T), pressure and nociceptive cells. These neurons can be identi¢ed within the ganglion by the size and location of their somata, and by the shape and amplitude of their action potentials (Nicholls and Baylor, 1968). By testing CYN on di¡erent neurons of the ganglia, we have detected important e¡ects on the electrical properties of T sensory cells. Therefore, we have focused our investigations on these neurons. In T cells, the discharge of action potentials induced either by receptive ¢eld activation or by intracellular repetitive stimulation elicits an afterhyperpolarization (AHP) due both to Ca2 -dependent K currents (IK=Ca ) and mainly to the activity of the electrogenic Na /K -ATPase (Baylor and Nicholls, 1969a; Jansen and Nicholls, 1973). Therefore, changes of AHP amplitude can be directly related to modi¢cations of the Na /K -ATPase activity and T neurons can represent a cellular model for monitoring the activity of the Na /K -ATPase with an electrophysiological approach. Previously, we have demonstrated that also the endogenous neurotransmitter serotonin (5-hydroxytryptamine or 5HT) reversibly reduces AHP through the inhibition of the electrogenic pump (Belardetti et al., 1984; Catarsi and Brunelli, 1991). In this paper we report evidence that CYN depresses the AHP amplitude through a selective inhibition of the Na /K -ATPase. A preliminary account of some of these ¢ndings has been presented at a meeting of the Society for Neuroscience (Brunelli et al., 1998).
Fig. 1. E¡ects of CYN on the electrical properties of T neurons. Dose-dependent e¡ect of CYN on AHP amplitude, Vm and Rm . CYN was tested at two di¡erent concentrations : 10 WM (n = 5) and 50 WM (n = 13). In the graphs of the ¢rst ¢ve ¢gures, the abscissas always describe the timing of the experimental phases; the means of the parameter values are expressed in the ordinates ; all data have been normalized to the initial value recorded in saline (control value). Statistical di¡erence (*P 6 0.05) was determined by performing Wilcoxon matched-pairs signed-ranks test.
EXPERIMENTAL PROCEDURES
Animals and surgery Adult leeches (H. medicinalis) purchased from a local supplier were maintained at 16³C, under natural daylight rhythm. Leeches were anesthetized in 10% ethanol in tap water for 10 min and then the nerve cord was exposed by cutting the body wall along the midline and by opening the ventral sinus. The surgery was performed at room temperature. A short chain of ganglia was removed at mid-body level and kept at 16³C in saline solution (see below). For each experiment a single ganglion was pinned with the ventral side up (where the sensory neurons are located) to the bottom of a small recording chamber coated with silicone elastomer (Sylgard0 , Dow Corning, Midland, MI, USA). Solutions Unless otherwise stated, the following saline solution was used, for both surgery and electrophysiological recordings: 115 mM NaCl, 4 mM KCl, 1.8 mM CaCl2 , 10 mM glucose, bu¡ered at pH 7.4 with 10 mM tris(hydroxymethyl)aminomethane^maleate. In one group of experiments, K -free saline was obtained by replacing 4 mM KCl with 4 mM NaCl in order to block the basal activity of the sodium pump. 4 mM CsCl was added to K -free saline (Schlue and Deitmer, 1984; Catarsi and Brunelli, 1991; Catarsi et al., 1993) to re-activate the Na /K pump. CYN, from fresh algae collected at Imperia (Liguria, Italy),
was puri¢ed with the extraction procedure used by Pesando et al. (1996). During electrophysiological experiments ganglia were constantly perfused with the saline solution at a rate of 1.5 ml/ min by means of a peristaltic pump (Master£ex, Cole-Palmer Instrument, Chicago, IL, USA) and all the drugs (from Sigma, St. Louis, MO, USA) were freshly dissolved in appropriate saline just before their application. CYN was previously dissolved in dimethylsulfoxide (DMSO) 1:1000, sonicated and then added to the saline solution. In control experiments, DMSO at that concentration did not a¡ect the electrical properties of T cells (data not shown). Electrophysiological technique T neurons were identi¢ed by the size and location of their cell bodies as well as by their ¢ring pattern (Nicholls and Baylor, 1968). Intracellular recordings were performed in current clamp mode by using borosilicate microelectrodes ¢lled with 4 M potassium acetate (60^80 M6 impedance). In a series of investigations, neurons were impaled with 2 M sodium acetate microelectrodes in order to increase the activity of Na /K -ATPase. All the experiments were carried out at room temperature. AHP was induced by injecting 10-s trains of intracellular depolarizing impulses (200 ms, 2.5 Hz). The discharge frequency of T cells was kept constant by adjusting the amount of injected current (0.4^0.8 nA). Since AHP is very sensitive to the quality of the impalement, only cells with AHP of at least 10^15 mV were used
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
for analysis. All the experiments were stored on a video tape recorder connected to a pulse code modulator (PCM 501 ES, Sony) for further data analysis. Statistical evaluation The amplitude of AHP was measured from the baseline of the resting potential before the stimulation, to the peak of maximal hyperpolarization. The action potential amplitude was measured relative to the prespike membrane potential. Action potential width was measured from a ¢xed fraction of the spike at a point corresponding to 35% of its maximal amplitude. For the statistical analysis of the data, Student's t-test was used when the Gaussian distribution was satis¢ed and non-parametric tests (Wilcoxon matched-pairs signed-ranks test and Mann^Whitney U-test) when the normal distribution could not be satis¢ed. The Wilcoxon matched-pairs signed-ranks test is similar to t-test for paired samples and it is used in order to test the same group of data twice and to anticipate a relationship. The Mann^Whitney U-test is used to compare two di¡erent groups in a study where the raw data were converted into ranks. In all the graphs, signi¢cance (P 6 0.05) is indicated with an asterisk. Statistical analysis was performed with the SPSS software package (Advanced Models tm 9.0).
521
potential (Vm ), input resistance (Rm , evaluated by injecting 0.5 nA hyperpolarizing impulses of 200 ms), amplitude and duration of action potential and amplitude of AHP. Fifteen minutes after application, CYN produced a reduction of the AHP amplitude. This e¡ect was dosedependent, became maximal after 15 min wash with saline solution (P 6 0.01) and started to recover after 30 min wash (Figs. 1 and 2A). A statistically signi¢cant depolarization of Vm (P 6 0.01) and a substantial reduction of Rm (P 6 0.01) were detected in 50 WM CYN (Fig. 1). We did not observe consistent changes in the duration of the action potential, whereas a signi¢cant reduction of the spike amplitude occurred during 50 WM CYN application (P = 0.018, Fig. 3). Since Rm changes a¡ect the AHP amplitude (shunt e¡ect), we compared the AHP amplitude with Rm changes and we found that the reduction of the AHP amplitude was signi¢cantly larger (30%, P = 0.0015) than the decrease of Rm (Fig. 2B). This result suggested a direct action of CYN on the components generating the AHP. E¡ect of CYN on AHP amplitude after application of CdCl2
RESULTS
E¡ects of CYN on T neuron activity We have detected the e¡ects of di¡erent concentrations of CYN (10 and 50 WM) on the following electrophysiological parameters of T neurons: membrane
Experiments were performed using 0.1 mM CdCl2 in normal saline to block IK=Ca (Stewart et al., 1989; Catarsi and Brunelli, 1991), one of the two components generating the AHP. Five minutes application of CdCl2 produced a reduction in the AHP amplitude down to
Fig. 2. E¡ects of 50 WM CYN on AHP amplitude and Rm . (A) Traces showing AHPs recorded from a single neuron. The action potentials generating AHP are clipped and indistinguishable from each other at the time scale shown. (B) Comparison between the e¡ect of 50 WM CYN on AHP amplitude and on Rm (n = 13). Statistical analysis was performed between the two curves with Mann^Whitney U-test (*P 6 0.05).
NSC 5206 12-11-01
522
R. Mozzachiodi et al.
Fig. 3. E¡ect of 50 WM CYN on the spike amplitude. (A) Traces showing action potentials evoked in a T cell before (control), after 15 min 50 WM CYN perfusion and during the washout. (B) Graph showing the e¡ect of 50 WM CYN on the amplitude of the evoked action potentials (n = 9). The statistical analysis was performed by Wilcoxon matched-pairs signedranks test (*P 6 0.05).
60% of the control value recorded in saline. No further decrease of the AHP was detectable when CdCl2 application was prolonged as previously demonstrated by Catarsi and Brunelli (1991). After the initial 5 min perfusion with CdCl2 , 50 WM CYN was applied for 15 min while the perfusion with the blocking agent was continued. A further signi¢cant reduction of the AHP amplitude down to 25% of the initial value was observed (P = 0.0277). After 60 min wash, ¢rst with CdCl2 solution and then with saline, CYN e¡ects recovered to about 60% (Fig. 4A, B). As with CYN alone, a reduction of Rm was detected, but the AHP depression was signi¢cantly larger than the decrease of Rm (P = 0.0281, Fig. 4B). E¡ect of CYN on AHP amplitude after application of apamin Since cadmium inhibited IK=Ca through the block of Ca2 channels also a¡ecting Rm and producing a strong shunt of the AHP current, we used a more speci¢c IK=Ca inhibitor, the peptide apamin, which has been successfully used to block IK=Ca in several neurons (Hugues et al., 1982; Zhang and Krnjevic, 1987). During preliminary experiments we found that 15 min perfusion with 1 nM apamin produced a 25% reversible reduction of the AHP amplitude without a¡ecting Rm . More prolonged applications of apamin did not further decrease the AHP amplitude (Fig. 5B). When 50 WM CYN was added to saline containing apamin, an evident reduction of the residual AHP was observed (P = 0.0117, Fig. 5A, C).
The e¡ect of CYN was maximal after 10 min wash with apamin, in agreement with the kinetics shown by CYN alone, and recovered to about 70% during 60 min wash. The maximal AHP depression detected was signi¢cantly larger than the Rm reduction (33%, P 6 0.001, Fig. 5C). E¡ect of CYN on Na+-stimulated activity of the Na+/K +-ATPase Since CYN still a¡ected the residual AHP amplitude when IK=Ca was blocked with di¡erent pharmacological agents, in a further group of experiments we tested the hypothesis of a direct inhibition of the Na /K -ATPase by CYN. When a T neuron was impaled with 2 M sodium acetate microelectrode, the leakage of Na into the soma rapidly induced a clear-cut hyperpolarization due to the activation of the electrogenic pump (Baylor and Nicholls, 1969a; Catarsi and Brunelli, 1991; Catarsi et al., 1993): Vm hyperpolarized by about 12 mV from the resting potential (b in Fig. 6A) and then reached a steady-state level (a in Fig. 6A). When CYN was added to the perfusion bath, the hyperpolarization turned into a depolarization (b in Fig. 6A) suggesting an inhibition of the Na /K -ATPase. In some experiments we perfused the ganglia with 50 WM CYN for 15 min before penetrating T cells with 2 M sodium acetate microelectrodes and we did not measure any detectable change of Vm during a further 10 min application of 50 WM CYN. Only when CYN was washed out did that neuron hyperpolarize (Fig. 6B).
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
523
Fig. 4. E¡ects of 50 WM CYN on AHP amplitude and on Rm during perfusion with 0.1 mM CdCl2 . (A) Traces showing AHPs recorded from a single T neuron. (B) Graph showing means of AHP amplitudes and Rm (n = 8). Cadmium perfusion did not prevent the e¡ect of CYN on AHP amplitude ; the depression of AHP driven by CYN was still present and was similar to that produced by CYN alone. Asterisk indicates a signi¢cant di¡erence (P 6 0.05) between AHP reduction and Rm change (Mann^Whitney U-test).
E¡ect of CYN on the basal activity of the Na+/K +-ATPase We also tested the e¡ect of CYN on the basal activity of the Na /K -ATPase. Perfusion with K -free saline blocked the Na /K -ATPase and led to a biphasic change of Vm in T cells (b in Fig. 7A, B): an early hyperpolarization was followed by a delayed depolarization due to the inactivation of the electrogenic pump as previously described by Schlue and Deitmer (1984). When the sodium pump was blocked, the depolarization phase reached a steady-state (dashed line in Fig. 7A, B) (Catarsi and Brunelli, 1991). Ten minutes perfusion with 4 mM CsCl, a Na /K -ATPase activator (Skou, 1965), shifted the Vm polarization back to the initial value as a consequence of the reactivation of the pump (b in Fig. 7A, B). Once the membrane potential had settled to its new value, the Vm maintained this polarized value as long as the T cell was perfused with K -free saline containing 4 mM CsCl (b in Fig. 7A). When 50 WM CYN was added to K -free solution containing 4 mM CsCl, Vm showed a remarkable depolarization (a in Fig. 7A). In another group of experiments, CYN was applied together with CsCl right after Vm reached the steady-state depolarization during perfusion with K -free solution (a in Fig. 7B); the e¡ect of CsCl was blocked by the toxin and the expected repolarization turned into a depolarization. The time course of Vm dur-
ing perfusion with CsCl alone (b in Fig. 7B) was statistically di¡erent from that produced by 4 mM CsCl plus 50 WM CYN (a in Fig. 7B, P = 0.034, P = 0.023).
DISCUSSION
This study reports the ¢rst evidence that CYN, the major metabolite synthesized by the so-called killer seaweed C. taxifolia, a¡ects the activity of nerve cells. We have tested the e¡ects of CYN on the electrical properties of T sensory neurons of the leech H. medicinalis, an experimental model extensively used in neurobiology. We have observed that CYN produces a clear-cut depression of the AHP that characterizes the ¢ring pro¢le of these neurons. This e¡ect is dose-dependent and coupled with a reduction of Rm and with a slight depolarization of Vm . Although a decrease of Rm a¡ects AHP amplitude (assuming Ohm's law: V = RI), we have detected a 70% AHP depression vs. a 40% Rm reduction when the AHP decrease was compared with Rm changes during CYN perfusion. This indicates that a simple linear short circuit of the currents generating the AHP failed to fully explain the action of CYN. After 60 min wash, the AHP amplitude recovered towards the control value suggesting that CYN e¡ects were long lasting but at least partially reversible. Two series of experiments provided evidence for an
NSC 5206 12-11-01
524
R. Mozzachiodi et al.
Fig. 5. E¡ects of 50 WM CYN on AHP amplitude and on Rm during 1 nM apamin application. (A) Traces showing AHPs recorded from a single T neuron. (B) The graph shows the means of AHP amplitude and Rm during apamin application (n = 5). Fifteen minutes apamin application depressed AHP to about 75% of the initial response. Prolonged applications (up to 30 min) did not further reduce the AHP amplitude. No signi¢cant change in Rm was detected. Asterisks on the AHP curve represent statistical di¡erence (P 6 0.05) determined by Wilcoxon matched-pairs signed-ranks test. (C) The graph shows the means of AHP amplitude and Rm (n = 9) during apamin and CYN application. Asterisks indicate the signi¢cant di¡erence (P 6 0.05) between AHP reduction and Rm change (Mann^Whitney U-test).
actual e¡ect of CYN on the components generating AHP. In the ¢rst one, the IK=Ca was blocked with two distinct pharmacological agents: cadmium and apamin. CdCl2 blocks the voltage-gated calcium channels from the extracellular side of the plasma membrane thus preventing the activation of IK=Ca . Apamin is a bee venom peptide that selectively blocks Ca2 -dependent K channels involved in afterpotentials (Hugues et al., 1982; Zhang and Krnjevic, 1987). In both experimental conditions, CYN was still capable of reducing the residual AHP due to the Na /K -ATPase. These results strongly support the hypothesis that CYN exerted an inhibitory action on the Na /K -ATPase. In addition, if CYN a¡ected the IK=Ca the AHP depression should occur through the closure of K or Ca2 channels with an increase of Rm . In contrast, we always measured a decrease of Rm that was signi¢cantly smaller than the maximal AHP depression in all the experimental conditions. During perfusion with apamin alone we did not detect changes in Rm , whereas a 20% reduction of Rm was observed during application of CdCl2 alone. This phenomenon does not seem to be ascribable to the action of cadmium on some membrane conductance, but it might be due in part to a decrease of the junctional re-
sistance through which T sensory neurons are coupled in the segmental ganglia (Baylor and Nicholls, 1969b). A decrease of the cytosolic calcium concentration, which can occur during CdCl2 perfusion, might lead to a reduction of junctional resistance and, consequently, of Rm (Rose and Loewenstein, 1975). In the second series of experiments, we gained insights for a direct inhibition of the Na /K -ATPase by CYN. This hypothesis could be strengthened using speci¢c blockers of the pump (such as ouabain or strophantidin), but the application of these blockers does not make it possible to keep a constant discharge frequency leading to the AHP, probably because of the excess of sodium in the T cell that cannot be pumped out (Catarsi and Brunelli, 1991). Therefore, we used di¡erent approaches such as ion manipulation of sodium pump activity. We observed that during CYN treatment the Na /K -ATPase became insensitive to intracellular injection of Na ions. In fact, the injection of Na into T neurons during perfusion with CYN turned the normal ouabain-sensitive hyperpolarization (Baylor and Nicholls, 1969a) into a depolarization. This e¡ect may be explained by the inhibition of the Na -stimulated activity of the pump: Na ions leaked into the cell could not be pumped out. When
NSC 5206 12-11-01
Caulerpenyne depresses the AHP in leech neurons
525
result suggested that CYN hindered cesium from activating the Na /K -ATPase. Overall, these data provide clear evidence that the Na /K -ATPase is the molecular target of CYN. During CYN application we did not observe any change in spike duration, whereas we detected a signi¢cant decrease of spike amplitude which might be ascribed
Fig. 6. E¡ects of CYN on Na -induced activity of the Na /K ATPase. (A) Na injection induced a rapid hyperpolarization (b). CYN 50 WM shifted the hyperpolarized Vm towards more depolarized values; asterisks represent signi¢cant di¡erences (P 6 0.05; Student's t-test for unpaired samples) between mean points obtained during CYN application (b, n = 11) and control value recorded in neurons impaled with sodium acetate microelectrodes and constantly perfused with saline (a, n = 9). (B) CYN prevented the Na -dependent hyperpolarization when T cells were pre-incubated with the toxin (b). Asterisks represent signi¢cant di¡erences (P 6 0.05; Student's t-test for unpaired samples) between mean points obtained during a further 10 min CYN application (b, n = 5) and control value recorded in neurons impaled with sodium acetate microelectrodes and constantly perfused with saline (a, n = 9 as in A). Arrowheads in A and B indicate the beginning of recording with the sodium acetate electrode.
T cells were pre-incubated with CYN, the Na injection failed to hyperpolarize Vm : the inhibitory e¡ect of CYN antagonized the activation of the Na /K -ATPase by sodium also in this condition. When the basal activity of the Na /K -ATPase was blocked with K -free perfusion, typical changes of Vm were detected (Schlue and Deitmer, 1984). In these experimental conditions, cesium application induced the repolarization of Vm following the reactivation of the electrogenic pump (Skou, 1965), whereas CYN converted this expected repolarization into a depolarization; this
Fig. 7. E¡ects of CYN on the basal activity of the Na /K -ATPase. The symbols represent the means of the shift of Vm from the resting potential. (A) In T neurons perfused with K -free saline (b, n = 10), Vm showed a hyperpolarization followed by depolarization. Dashed lines represent the steady-state depolarization when the cells were constantly perfused with K -free saline. Perfusion with saline in which 4 mM KCl was replaced with 4 mM CsCl induced a repolarization (b). After repolarization was settled ¢ve cells were treated for 10 min with K -free saline plus 4 mM CsCl plus 50 WM CYN (a). Depolarization occurred. Asterisks indicate a signi¢cant di¡erence (P 6 0.05) between Vm recorded in T cells perfused with CYN and Vm recorded in T cells repolarized during CsCl perfusion (Student's t-test for unpaired samples). (B) Thirteen neurons were treated with K -free saline (b). After Vm reached the steady-state depolarization, a repolarization was evident in the ¢ve cells that were perfused with K -free saline plus 4 mM CsCl (b). When 50 WM CYN was added to the K -free saline plus 4 mM CsCl, the repolarization did not occur (a, n = 8). Asterisks indicate that a signi¢cant di¡erence occurred between the Vm recorded in the two groups of cells (Student's t-test for unpaired samples).
NSC 5206 12-11-01
526
R. Mozzachiodi et al.
to the Vm depolarization following the inhibition of the basal activity of the pump. Concerning the reduction of Rm and the depolarization of Vm detected when 50 WM CYN was applied, we can speculate that CYN, besides its inhibitory e¡ect on the Na /K -ATPase, might also a¡ect some other targets in the plasma membrane such as ion channels. For example, the activation of a not yet identi¢ed cationic conductance could account for both the depolarization and the decrease of Rm . Furthermore, we cannot rule out that CYN modulates the electrical junctions which couple each other T cells placed in the same ganglion and in adjacent ones (Baylor and Nicholls, 1969b). The CYN action on the Na /K -ATPase occurred with a certain latency suggesting the activation of a putative receptor and the involvement of a biochemical cascade. We have previously showed that in T neurons 5HT depressed AHP amplitude through the inhibition of the electrogenic pump (Catarsi and Brunelli, 1991). 5HT stimulates cAMP production in T neurons (Catarsi et al., 1993). The magnitude and the time course of the 5HT e¡ect are very similar to that produced by CYN,
but CYN does not act through serotonergic receptors and does not activate the cAMP pathways (Brunelli et al., 1998). Therefore it will be very enticing to clarify the molecular mechanism underlying CYN e¡ects. At present, some hypotheses can be made. CYN is a sesquiterpene capable of permeating the plasma membrane. For example, in sea urchin eggs CYN acts in the cytosol and inhibits the ¢rst cleavage by blocking the mitotic cycle at metaphase (Pesando et al., 1996). In this case CYN a¡ects the intracellular ATP-dependent Ca2 accumulation (Pesando et al., 1996) in a thapsigargin-like manner, and also inhibits the MAPK (Pesando et al., 1998, 1999). We may hypothesize that in T neurons CYN penetrates the membrane and a¡ects the Na / K -ATPase from the intracellular side or triggers molecular events leading to the changes of the Na /K -ATPase activity. In future researches, it will be interesting to investigate whether the inhibitory e¡ect of CYN on the Na /K ATPase detected in T cells of the leech can be observed in other tissues where the electrogenic pump plays pivotal roles.
REFERENCES
Baylor, D.A., Nicholls, J.G., 1969a. After e¡ects of nerve impulse on signaling in the central nervous system of the leech. J. Physiol. 203, 571^589. Baylor, D.A., Nicholls, J.G., 1969b. Chemical and electrical synaptic connections between cutaneous mechanoreceptor neurones in the central nervous system of the leech. J. Physiol. 203, 591^609. Belardetti, F., Brunelli, M., Demontis, G.C., Sonetti, D., 1984. Serotonin and Retzius cell depress the hyperpolarization following impulses of the leech touch cell. Brain Res. 300, 91^102. Brodfuehrer, P.D., Debski, A.E., O'Gara, B.A., Friesen, W.O., 1995. Neural control of leech swimming. J. Neurobiol. 27, 403^418. Brunelli, M., Garcia-Gil, M., Mozzachiodi, R., Roberto, M., Scuri, R., Traina, G., Zaccard, M.L., 1998. Caulerpenyne, a seaweed biotoxin, provokes the inhibition of the Na /K electrogenic pump in neurons. Soc. Neurosci. Abstr. 24, 231. Catarsi, S., Brunelli, M., 1991. Serotonin depresses the after-hyperpolarization through the inhibition of the Na /K electrogenic pump in the sensory neurones of the leech. J. Exp. Biol. 155, 261^273. Catarsi, S., Scuri, R., Brunelli, M., 1993. Cyclic AMP mediates inhibition of the Na /K electrogenic pump by serotonin in tactile sensory neurons of the leech. J. Physiol. 462, 229^242. De Haro, L., Tre¡ot, M.J., Jouglard, J., Perringue, C., 1993. Trois cas d'intoxication de type ciguateresque apre©s ingestion de sparidae de Mediterrane¨e. Ictyol. Physiol. Acta 16, 133^146. Fernandez-de-Miguel, F., Drapeau, P., 1995. Synapse formation and function: insights from identi¢ed leech neurons in culture. J. Neurobiol. 27, 367^379. Fischel, J.L., Leme¨e, R., Formento, P., Caldani, C., Moll, J.L., Pesando, D., Meinesz, A., Grelier, P., Pietra, F., Guerriero, A., Milano, G., 1995. Cell growth inhibitory e¡ects of caulerpenyne, a sesquiterpenoid from the marine alga Caulerpa taxifolia. Anticancer Res. 15, 2155^2160. Giannotti, A., Ghelardi, E., Senesi, S., 1994. Characterization of seawater bacterial communities within environments colonized by the tropical green seaweed Caulerpa taxifolia. In: Boudouresque, C.F., Meinesz, A., Gravez, V. (Eds.), First International Workshop on Caulerpa taxifolia. GIS Posidonie, Marseille, pp. 197^201. Guerriero, A., Meinesz, A., D'Ambrosia, M., Pietra, F., 1992. Isolation of toxic sesqui- and monoterpenes from tropical green seaweed Caulerpa taxifolia which has invaded the region of Cap Martin and Monaco. Helv. Chim. Acta 75, 689^695. Hugues, M., Romey, G., Duval, D., Vincent, J.P., Landuzki, M., 1982. Apamin as a selective blocker of the calcium-dependent potassium channel in neuroblastoma cells: Voltage-clamp and biochemical characterization of the toxin receptor. Proc. Natl. Acad. Sci. USA 79, 1308^1312. Jansen, J.K.S., Nicholls, J.G., 1973. Conductance changes, an electrogenic pump and the hyperpolarization of leech neurones following impulses. J. Physiol. 229, 635^655. Kleinhaus, A.L., Angstadt, J.D., 1995. Diversity and modulation of ionic conductances in leech neurons. J. Neurobiol. 27, 419^433. Nicholls, J.G., Baylor, D.A., 1968. Speci¢c modulation and receptive ¢elds of sensory neurones in CNS of the leech. J. Neurophysiol. 31, 740^756. Pesando, D., Huitorel, P., Dolcini, V., Amade, P., Girard, J.P., 1998. Caulerpenyne interferes with microtubule-dependent events during the ¢rst mitotic cycle of sea urchin eggs. Eur. J. Cell Biol. 77, 19^26. Pesando, D., Leme¨e, R., Ferrua, C., Amade, P., Girard, J.P., 1996. E¡ects of caulerpenyne, the major toxin from Caulerpa taxifolia on mechanisms related to sea urchin egg cleavage. Aquat. Toxicol. 35, 139^155. Pesando, D., Pesci-Bardon, C., Huitorel, P., Girard, J.P., 1999. Caulerpenyne blocks MBP kinase activation controlling mitosis in sea urchin eggs. Eur. J. Cell Biol. 78, 903^910. Rose, B., Loewenstein, W.R., 1975. Permeability of cell junction depends on local cytoplasmic calcium activity. Nature 254, 250^252. Schlue, W.R., Deitmer, J.W., 1984. Potassium distribution and membrane potential of sensory neurons in the leech nervous system. J. Neurophysiol. 51, 689^704. Skou, J.C., 1965. Enzymatic basis for active transport of Na and K across cell membrane. Physiol. Rev. 45, 596^617. Stewart, R.R., Nicholls, J.G., Adams, W.B., 1989. Na, K and Ca currents in identi¢ed leech neurones in culture. J. Exp. Biol. 141, 1^20. Zhang, L., Krnjevic, K., 1987. Apamin depresses selectively the afterhyperpolarization of cat spinal motoneurons. Neurosci. Lett. 74, 58^62. (Accepted 29 May 2001)
NSC 5206 12-11-01