P-type Ca²⁺ current in crayfish peptidergic neurones

Entry of Ca²⁺ through voltage-dependent Ca²⁺ channels plays an important role in excitable cells and is related to their electrical activity. The opening of Ca²⁺ channels transiently increases the intracellular free Ca²⁺ concentration, which acts as messenger and may control ion channel gating, enzyme activation, gene expression, transmitter release, neurosecretion and other cell functions (for reviews, see Llinás et al., 1992; Miller, 1992; Tsien and Tsien, 1990).

Several classes of voltage-dependent Ca²⁺ channel have been identified in vertebrates on the basis of electrophysiological and pharmacological criteria and by molecular cloning (Sather et al., 1993; Snutch and Reiner, 1992). They are classified as low-voltage-activated (LVA) and high-voltage-activated (HVA) Ca²⁺ channels (Carbone and Lux, 1984; Fox et al., 1987; Swandulla et al., 1991; Tsien et al., 1988). LVA or T-type Ca²⁺ channels are sensitive to Ni²⁺, amiloride, ethosuximide and octanol and are insensitive to 1,4-dihydropyridines (DHPs), ω-conotoxin-GVIA and ω-agatoxin-IVA (Coulter et al., 1989; Herrington and Lingle, 1992; Mori, 1994; Tang et al., 1988). HVA Ca²⁺ channels have been subdivided into five types: (a) L-type channels sensitive to DHPs (Fox et al., 1987; Hess et al., 1984); (b) N-type channels insensitive to DHPs and irreversibly blocked by ω-conotoxin-GVIA (Aosaki and Kasai, 1989; Plummer et al., 1989); (c) P-type channels insensitive to DHPs and to ω-conotoxin GVIA but selectively blocked by nanomolar concentrations of ω-agatoxin-IVA (<30 nmol l⁻¹) (Llinás et al., 1989; Mintz et al., 1992) and by funnel-web spider toxin (Wang and Lemos, 1994); (d) Q-type channels blocked by ω-agatoxin-TK and by higher concentrations of ω-agatoxin IVA (>100 nmol l⁻¹) and which differ from P-type channels in their sensitivity to ω-conotoxin-MVIIC (<150 nmol l⁻¹) (Hillyard et al., 1992; Wang et al., 1997); and finally (e) R-type channels insensitive to organic compounds and blocked by Ni²⁺ (Zhang et al., 1993).

In contrast, few studies have been carried out on Ca²⁺ channels in invertebrate neurones. Both LVA and HVA Ca²⁺ current types have been identified in leech (Angstadt and Calabrese, 1991), snail (Haydon and Man-Son-Hing, 1988), Aplysia californica (Fossier et al., 1994), squid (Llinás et al., 1989) and crustacean neurones (Meyers et al., 1992; Onetti et al., 1990; Richmond et al., 1995, 1996). Recently, a P-type Ca²⁺ channel has been described in crayfish motoneurones (Hong and Lnenicka, 1997) that is less sensitive to ω-agatoxin-IVA (600 nmol l⁻¹) than that in vertebrate neurones (Randall and Tsien, 1995).

Spontaneous neuronal firing in the X-organ sinus gland system is related to hormonal release (Stuenkel, 1985). Action potentials are Ca²⁺-dependent (Iwasaki and Satow, 1971) and the intracellular free [Ca²⁺] modulates K⁺ channels (Martĺnez et al., 1991) and the negative slope conductance (Onetti et al., 1990). In the present study, we characterize the biophysical and pharmacological properties of the Ca²⁺ current in freshly dissociated XO neurones of the crayfish Procambarus clarkii and demonstrate that it corresponds to an HVA P-type Ca²⁺ current.

Procambarus clarkii (Girard) were collected from Rĺo Conchos, Chihuahua, México, and acclimated to laboratory conditions for 2 weeks under a 12 h:12 h L:D photoperiod. Eyestalks were excised and placed in chilled saline solution containing (in mmol l⁻¹): 205 NaCl, 5.4 KCl, 2.6 MgCl₂, 13.5 CaCl₂ and 10 Hepes, adjusted to pH 7.4 with NaOH. The exoskeleton, muscles and connective tissue were carefully removed under a dissecting microscope, and the neuronal somata were exposed. Isolated XOs were incubated with 200 μg ml⁻¹ collagenase-dispase (Boehringer Mannheim) for 1 h in modified Leibovitz L-15 culture medium containing (in mmol l⁻¹): 205 NaCl, 5.4 KCl, 13.5 CaCl₂, 2.5 MgCl₂, 10 Hepes, 5.5 glucose, 2 L-glutamine, and gentamycin (16 μg ml⁻¹, Schering Plough), streptomycin (5 μg ml⁻¹, Sigma) and penicillin (5 units ml⁻¹, Sigma). The enzyme was washed out, and the XO neurones were dissociated by gentle suction through fire-polished micropipettes (Garcĺa et al., 1990). Isolated neurones were plated individually onto a 200 μl recording chamber, precoated with Concanavalin A (Type III, Sigma). Cells were maintained at room temperature (22–24 °C) in modified Leibovitz L-15 medium.

Voltage-clamp experiments in the whole-cell configuration were performed using freshly dissociated XO neurones (2–6 h after plating). Recordings were made using an Axopatch-200A amplifier (Axon Instruments). Pipettes were constructed from borosilicate capillaries (Sutter Instruments) using a horizontal puller (P-87 Flaming Brown, Sutter Instruments) and fire-polished with a microforge (Narishige MF-90), to a final resistance of 2.5–4 MΩ. Pipettes were filled with a solution containing (in mmol l⁻¹): 215 CsCl, 2.86 CaCl₂, 2 Mg-ATP, 5.25 EGTA, 10 Hepes, adjusted to pH 7.4 with CsOH. Ca²⁺ currents were filtered at 5 kHz and acquired using commercially available hardware and software (Axon Instruments). Transient capacitative and leak currents were subtracted using the P/4 protocol (Almers et al., 1983), series resistance was generally compensated >70 %. X-organ neurones were continuously superfused with a solution containing (in mmol l⁻¹): 195 N-methyl-D-glucamine chloride (NMG-Cl), 20 tetraethylammonium chloride (TEA-Cl), 13.5 CaCl₂ and 10 Hepes, adjusted to pH 7.4 with NMG⁺. During permeability experiments, extracellular Ca²⁺ was substituted equimolarly by Ba²⁺, Sr²⁺, Mg²⁺, Mn²⁺, Cd²⁺, Ni²⁺ or Co²⁺; in blocking experiments, the divalent cations (Cd²⁺, Ni²⁺, Co²⁺ and Zn²⁺⁾ were added to the superfusing solution.

Stock solutions of DHPs (nifedipine, nitrendipine and Bay K-8644) were dissolved in 95 % ethanol and stored in the dark at 4 °C. The DHPs tested (2–20 μmol l⁻¹) were protected from light during experiments. Peptide toxins (Alomone) were dissolved in dimethyl sulphoxide, and the concentrations tested were from 50 nmol l⁻¹ to 5 μmol l⁻¹. Organic blockers were superfused with a low-[Ca²⁺] solution containing (in mol l⁻¹): 207 NMG-Cl, 20 TEA-Cl, 5 CaCl₂ and 10 Hepes (pH 7.4).

To characterize the properties of the Ca²⁺ current in freshly dissociated crayfish XO neurones, other membrane currents were minimized. The Na⁺ current was eliminated by replacing external Na⁺ with NMG⁺; K⁺ currents were blocked with TEA⁺ (20 mmol l⁻¹) in the external solution and by substituting intracellular K⁺ with Cs⁺. Under these ionic conditions, membrane-depolarizing voltage pulses activated an inward current identified as a Ca²⁺ current.

Fig. 1A shows a set of recordings of inward Ca²⁺ currents in response to 10 mV depolarizing voltage steps from a holding potential of —70 mV. The threshold for the activation of the Ca²⁺ current was approximately —40 mV (Fig. 1A,B). Both the onset and decay of the Ca²⁺ current became faster as depolarization increased. The Ca²⁺ current reached its peak in approximately 2 ms at 20 mV. The current decayed in a time-dependent manner that reflected mainly inactivation of Ca²⁺ channels. The maximal Ca²⁺ current was evoked at potentials between 20 and 30 mV. The current–voltage (I–V) relationship (Fig. 1B) shows that depolarizations greater than 20 mV resulted in a progressive decrease in the amplitude of the Ca²⁺ current; the reversal potential was more positive than 80 mV.

To establish the presence of LVA Ca²⁺ currents in XO neurones, recordings were made from several cells at three different holding potentials ranging from —90 to —40 mV. This manoeuvre is commonly used to remove the possible inactivation of such currents (Carbone and Lux, 1984; Kasai and Neher, 1992). The I–V relationships obtained from these experiments had the same threshold potential and current amplitudes, they also peaked at the same voltage, and the reversal potential was the same (data not shown). No differences were observed in the time course of deactivation when the holding potential was changed (Fig. 1C). The slow activation of the Ca²⁺ current at hyperpolarized potentials (—90 and —60 mV) may be due to the gating current and is explained by a sequential model in which the channel passes through several closed states before the open state (Armstrong, 1981). To explore the voltage-dependent inactivation of the Ca²⁺ current, it was measured by maintaining the neurones for 15 s at a defined holding potential (from —90 to 0 mV). A test pulse to 30 mV was then applied for 5 ms, and the membrane was again repolarized to the holding potential. The maximal amplitude was obtained at —50 mV (Fig. 1D). At holding potentials between —90 and —50 mV, the Ca²⁺ current amplitude showed a small reduction (10 % at —90 mV), probably due to a slow inactivation. The Ca²⁺ current amplitude was constant between —50 and —35 mV, whereas it was abruptly reduced between —30 and 0 mV (Fig. 1D). All these results indicate that the Ca²⁺ current in crayfish XO neurones is generated by HVA Ca²⁺ channels.

In the steady state, the fraction of Ca²⁺ conductance (m_∞^x) can be estimated by measuring the tail current amplitude on repolarization to —60 mV after a 5 ms activating pulse. This time allowed full activation of the Ca²⁺ current, with minimal contamination by other currents. The steady-state activation curve has a sigmoidal form (Fig. 2A) and was fitted by the Boltzmann expression:

The activation of the Ca²⁺ current followed a sigmoidal time course with a variable delay in response to depolarizing pulses (Fig. 2D). The contributions of gating currents, which dominate the earlier part of the activation, were minimized by subtracting the corresponding responses in the presence of 1 mmol l⁻¹ Cd²⁺.

In voltage-clamped crustacean neurones, as in other preparations, Ca²⁺ currents evoked by depolarizing pulses gradually decay from an initial peak as a result of Ca²⁺ channel inactivation (Hagiwara and Byerly, 1981; Richmond et al., 1995; Branchaw et al., 1997). The extent of inactivation was manifested by a decrease of the tail current amplitude on repolarization following command pulses of various durations (Fig. 4A), which bears a close relationship with the time course of decay of the Ca²⁺ current.

Moreover, the time course of the decay of the Ca²⁺ current depends on the cation that carries the current. Fig. 4B shows three representative recordings obtained from the same neurone, in which the inward current was carried by Ca²⁺, Sr²⁺ or Ba²⁺. When the current was carried by Ba²⁺, it did not decay during the command pulse, while the Sr²⁺ current decayed more slowly than the Ca²⁺ current, suggesting that Sr²⁺ partially inactivated Ca²⁺ channels. These findings strongly support the idea that the inactivation of the Ca²⁺ current in XO neurones is mediated by Ca²⁺.

To study the Ca²⁺-dependent inactivation in more detail, the relationship between Ca²⁺ entry and Ca²⁺ current inactivation was explored using a double-pulse protocol in which the potential of the prepulse varied from —60 to 110 mV and its effects on a fixed test pulse to 30 mV was evaluated (Fig. 4Ci). Representative traces obtained at prepulse values of 0, 30 and 60 mV are illustrated in the Fig. 4Cii (0 mV at the top, 60 mV at the bottom). The I–V relationship obtained with the Ca²⁺ current evoked by the prepulse is illustrated in Fig. 4Ciii, and the inactivation curve (Fig. 4Civ) was obtained by plotting the Ca²⁺ current amplitude evoked by the test pulse as a function of prepulse potential. At the prepulse potential at which Ca²⁺ entry was maximal (30 mV), the inactivation of the Ca²⁺ current was also maximal (35 %). These results indicate that the time-dependent inactivation of the Ca²⁺ current in XO neurones is dependent on the entry of Ca²⁺.

Recovery from inactivation is another mechanism that can be influenced by intracellular Ca²⁺ concentration (Gutnick et al., 1989). To study this process, we evoked the Ca²⁺ current with prepulses to 30 mV, and after a variable interval at the holding potential (—50 mV) a test pulse with the same characteristics was applied. Short prepulses (10 ms) had less effect on the Ca₂⁺ current amplitude evoked by the test pulse, reducing it by 25% when the interpulse interval was 5 ms. A longer (20 ms) prepulse reduced the current amplitude evoked by the second pulse at the same interpulse interval by 40% (Fig. 4Di,ii). The time courses of the recovery were fitted with single-exponential curves; the time constants were 17° 2.6 for a 10 ms prepulse and 31° 3.2 ms for a 20 ms prepulses (means ° S.E.M., N=4) (Fig. 4Diii). These results suggest that recovery from the inactivation of the Ca₂⁺ current in XO neurones is dependent on the intracellular free Ca₂⁺ concentration. Similar results have been reported in neurohypophysis terminals where two mechanisms of Ca₂⁺ channel inactivation are implicated, one depending on voltage and the other depending on the intracellular free Ca2⁺ concentration (Branchaw et al., 1997).

To explore the selectivity sequence of divalent cations through Ca²⁺ channels in XO neurones, extracellular Ca²⁺ (13.5 mmol l⁻¹) was equimolarly replaced by Ba²⁺, Sr²⁺, Mg²⁺, Mn²⁺, Co²⁺, Ni²⁺ or Cd²⁺. In a series of experiments, I–V relationships were obtained from a holding potential of —60 mV with 50 ms depolarizing pulses at 10 mV intervals. Ba²⁺, Sr²⁺ and Mg²⁺ were able to generate currents, but Mn²⁺, Co²⁺, Ni²⁺ and Cd²⁺ failed to generate currents (Fig. 5; results for Co²⁺, Ni²⁺ and Cd²⁺ not shown). The selectivity sequence for permeable divalent cations was 1.4 (Ba²⁺), 1.0 (Sr²⁺) and 0.2 (Mg²⁺). These values were obtained by normalizing the current carried by any one of the divalent cations with respect to the maximal Ca²⁺ current. The activation threshold of Ba²⁺ and Sr²⁺ currents was shifted in the hyperpolarizing direction by 20 mV, and the maximum current values were shifted in the hyperpolarizing direction by 10 V, in comparison with the activation threshold of the Ca²⁺ current.

Blocking of the Ca²⁺ current in XO neurones by divalent cations was explored by superfusing Co²⁺, Ni²⁺, Cd²⁺ or Zn²⁺, at low concentrations, in the presence of the normal Ca²⁺ concentration. Ca²⁺ currents were recorded first in the absence and then in the presence of the blocking ion, from a holding potential of —50 mV with 10 mV depolarizing steps for 50 ms. The Ca²⁺ current was more effectively blocked by Cd²⁺ and Zn²⁺ than by Co²⁺ and Ni²⁺ (Fig. 6A). The maximal blocking effect on the Ca²⁺ current of Zn²⁺ or Cd²⁺ (100 μmol l⁻¹) averaged 80 %, and the effects were partially reversible. Blockage by Ni²⁺ or Co²⁺ (2 mmol l⁻¹) was less effective (40 %). Furthermore, the block caused by Co²⁺ and Cd²⁺ was less potent at negative potentials, whereas the opposite was true for Ni²⁺ and Zn²⁺ (Fig. 6B), suggesting a voltage-dependent mechanism.

To explore the pharmacological properties of the Ca²⁺ current and to identify the possible multiple components of HVA Ca²⁺ channels involved in generating the Ca²⁺ current, we performed a series of experiments in which we applied several derivatives of DHPs, such as nitrendipine, nifedipine, nimodipine or Bay K-8644, as well as derivatives of phenylalkylamines, such as verapamil or D-600. None of these compounds, applied extracellularly at concentrations up to 20 μmol l⁻¹, affected the Ca²⁺ current.

The Ca²⁺ current was blocked by the P-type Ca²⁺-channel-selective peptide ω-agatoxin-IVA at a low extracellular Ca²⁺ concentration (5 mmol l⁻¹). Fig. 7A illustrates the effect of 200 nmol l⁻¹ ω-agatoxin-IVA on the amplitude of the Ca²⁺ current. In three neurones, the Ca²⁺ current amplitude was rapidly reduced by 85 % by the toxin to reach a maximal level of block at approximately 1 min; the block persisted after the peptide had been washed out. However, when two or three trains of depolarizing pulses (2 ms) to 60 mV at 5 Hz were applied for 5 s, after the maximal effect of the toxin, the blockage was partially removed, 80 % of the Ca²⁺ current amplitude being recovered (Fig. 7A), suggesting that the effects exerted by the toxin are voltage-dependent, as has been reported previously in vertebrate neurones (Mintz et al., 1992). The degree of block was dependent on the toxin concentration, 50 and 100 nmol l⁻¹ reduced the Ca²⁺ current amplitude by 10 and 35 %, respectively. In contrast, at the normal extracellular Ca²⁺ concentration, the peptide was ineffective in blocking the current at concentrations up to 500 nmol l⁻¹. Other peptide toxins specific for N-type (ω-conotoxin-GVIA, 2 μmol l⁻¹) and Q-type (ω-conotoxin-MVIIC, 5 μmol l⁻¹) Ca²⁺ channels had no effect on the Ca²⁺ current in XO neurones (Fig. 7B,C).

Finally, when the fraction of Ca²⁺ current resistant to ω-agatoxin-IVA (Fig. 8A) was scaled up to match the control recording, we observed that the kinetics of activation and deactivation were the same (Fig. 8C). The transient outward current at the beginning of the pulse (gating current) was not affected by the toxin. Furthermore, the residual Ca²⁺ current showed an I–V relationship with the same characteristics as the control Ca²⁺ current: the same threshold for activation and a maximum value at 20 mV (Fig. 8B). These results suggest that crayfish XO neurones express a single class of P-type Ca²⁺ channels.

When the Ca²⁺ current is activated, other physiological events such as action potentials occur, with the entry of Ca²⁺ participating in the firing pattern (Meyers et al., 1992). This phenomenon is probably important in neurosecretory cells, since during normal electrical activity Ca²⁺ channels are activated and allow the influx of Ca²⁺, a process crucial for cellular functions such as the inactivation of the Ca²⁺ channels themselves, the modulation of K⁺ channels (Martĺnez et al., 1991; Onetti et al., 1996), participation in the negative slope conductance (Onetti et al., 1990) and neuropeptide secretion (Renaud, 1988; Wang et al., 1997).

LVA Ca²⁺ currents seem to be absent in crayfish XO neurones because (a) the inactivation and deactivation kinetics of the Ca²⁺ current could be fitted to single-exponential functions, (b) the I–V relationship did not have an additional peak at hyperpolarized potentials values and (c) the Ca²⁺ tail current amplitude did not change when it was evoked from different holding potentials (see Fig. 1C). Similar results have been reported in XO neurones from the crab Cardisoma carnifex (Richmond et al., 1995).

Activation of the Ca²⁺ current has been described in a variety of preparations using the Hodgkin and Huxley model (m^x), in which the exponent x accounts for the speed of the activation (Adams and Gage, 1979; Llinás et al., 1981; Sala, 1991). In the present study, the onset of the Ca²⁺ current clearly follows the predicted kinetics at all the potentials tested. Additionally, the tail current data are consistent with the kinetics, implying a linear sequential model with two closed states and one open state. The Ca²⁺ current in XO neurones activates in a similar manner to that in in pituitary cells (Hagiwara and Ohmori, 1982) and in sympathetic and hippocampal neurones (Belluzzi and Sacchi, 1989; Kay and Wong, 1987; Sala, 1991). According to the model, the predicted mean open time of the Ca²⁺ channel at —20 mV would be 568 μs, a value faster than that in sympathetic neurones (Sala, 1991) and in chromaffin cells (Fenwick et al., 1982).

The small reduction of the Ca²⁺ current amplitude evoked from holding potentials between —90 and —35 mV does not appear to be due to inactivation, because tail current amplitudes and their time courses were identical for three different holding potentials (Fig. 1C), indicating that the number of channels that remained open at the time of repolarization was the same in the three cases. An explanation for this reduction may be the presence of an outward gating current superimposed on Ca²⁺ current activation, the presence of another type of Ca²⁺ channel that inactivates slightly in this voltage range or a very slow inactivation of the Ca²⁺ channels. To differentiate between these, it will be necessary to explore this process in more detail.

The time-dependent decay of the Ca²⁺ current during depolarizing command pulses could be due to (a) depletion of extracellular Ca²⁺, (b) contamination with outward currents, or (c) inactivation of Ca²⁺ channels. The first two possibilities were rejected because no effects were observed on the time course of the Ca²⁺ current when the cells were continuously superfused with a solution containing 13.5 mmol l⁻¹ Ca²⁺ or when Cl⁻ was replaced with CH₃SO₄⁻ or Cs⁺ was replaced with TMA⁺. The decay of the Ca²⁺ current in XO neurones is therefore probably due to the inactivation of Ca²⁺ channels. At least two mechanisms have been described for inactivating the Ca²⁺ current, one that depends on membrane potential and another that depends on [Ca²⁺]_i (Chad and Eckert, 1984). An increase in intracellular free Ca²⁺ concentration favours the inactivation of Ca²⁺ channels in a variety of preparations (Ashcroft and Stanfield, 1982; Gutnick et al., 1989; Plant et al., 1983). It appears that the mechanism of inactivation of the Ca²⁺ current in XO neurones is mediated mainly by Ca²⁺. This is supported by the observation that the current carried by Ba²⁺ did not decay, at least during the 50 ms command pulses, and that the amplitude of the Ca²⁺ current depended on previous influx of Ca²⁺, as in other crustacean peptidergic neurones (Meyers et al., 1992; Richmond et al., 1995). The U-shaped curve obtained from a double-pulse protocol (Fig. 4Civ) is a hallmark of Ca²⁺-dependent inactivation (Chad and Eckert, 1984) and is due to accumulation of Ca²⁺ in a submembrane compartment, probably binding to the open Ca²⁺ channels.

In a variety of cells, Ba²⁺ is more effective as a charge carrier than Ca²⁺, probably because of its lower affinity for the channel (Hagiwara and Byerly, 1981). On the basis of the peak current for the I–V relationships with different divalent cations, the selectivity sequence in crayfish XO neurones was: Ba²⁺>Sr²⁺≈Ca²Mg²⁺, in comparison with crab peptidergic neurones, in which the reported sequence was Ba²⁺>Sr²⁺>Ca²⁺ (Richmond et al., 1995). The shift of the I–V relationship of the Ba²⁺ current may be explained because Ba²⁺ is less effective than Ca²⁺ in shielding the membrane charge (Frankenhaeuser and Hodgkin, 1957). Other divalent cations that act as Ca²⁺ channel blockers (Mn²⁺, Co²⁺, Ni²⁺ and Cd²⁺) can also carry inward current through Ca²⁺ channels (for a review, see Hagiwara and Byerly, 1981). However, in crayfish XO neurones, we found that these cations did not generate currents, as they do in other crustacean neurones (Meyers et al., 1992; Richmond et al., 1995).

Blockage of the Ca²⁺ current in XO neurones by the divalent cations tested here depended on their concentration and on the membrane potential. We found that Cd²⁺ and Zn²⁺ were more effective in blocking the Ca²⁺ current than Co²⁺ and Ni²⁺, suggesting that they have a higher affinity for the Ca²⁺ channel than do Co²⁺ and Ni²⁺. In contrast, Ni²⁺ and Zn²⁺ blocked the Ca²⁺ current more effectively at hyperpolarized values (—10 to 50 mV), whereas block of the Ca²⁺ current by Co²⁺ and Cd²⁺ was more effective at depolarizing potentials (30–70 mV). The voltage-dependence of Ca²⁺ current blockage by divalent cations suggests that they interact within the ion channel. Thus, two additional criteria indicate that the inward Ca²⁺ current in XO neurones is generated by HVA channels: (a) the permeability to Ba²⁺ was higher than that to Ca²⁺ and (b) the blockage exerted by Cd²⁺ was greater than that exerted by Ni²⁺ (Tsien et al., 1988).

Dihydropyridines (DHPs), at concentrations of 10 μmol l⁻¹, block nearly all L-type Ca²⁺ current in ventricular myocytes and photoreceptors (Barnes and Hille, 1989; Balke et al., 1992). However, DHPs (nitrendipine, nifedipine, nimodipine or Bay K-8644) at concentrations of 20 μmol l⁻¹ and phenylalkylamines (verapamil or D-600) at concentrations of 20 μmol l⁻¹ had no effect on the Ca²⁺ current, indicating that the crayfish XO neurones do not express an L-type Ca²⁺ current.

One common feature of L-, N-, P-, Q- and R-type currents is their activation over the same range of membrane potentials, and it is difficult to distinguish them by their kinetics, permeability and blockage by inorganic cations (Usowics et al., 1992). Thus, the identification of HVA Ca²⁺ currents is based mainly on pharmacological tests. In the XO neurone somata, relatively high concentrations of ω-agatoxin-IVA (200 nmol l⁻¹) blocked 85 % of the total Ca²⁺ current. These results suggest that the Ca²⁺ channels expressed in XO neurones share the pharmacological properties of P-type Ca²⁺ channels, with less specificity for ω-agatoxin-IVA than the Ca²⁺ channels of mammalian neurones (Brown et al., 1994; Llinás et al., 1992; Mintz et al., 1992). For instance, in cerebellar Purkinje neurones, ω-agatoxin-IVA has a high selectivity for P-type Ca²⁺ channels with a dissociation constant K_d of 2–10 nmol l⁻¹ (Mintz et al., 1992). However, in the marine crab Cardisoma carnifex, high concentrations of ω-agatoxin-IVA (500 nmol l⁻¹) did not affect the Ca²⁺ current or peptide release in either freshly dissociated XO neurones (Richmond et al., 1995) or peptidergic terminals (Richmond et al., 1996), even at the low extracellular Ca²⁺ concentration at which the toxin appears to be more effective. In contrast, transmitter release at the neuromuscular junction (Araque et al., 1994) and the Ca²⁺ current in motoneurones (Hong and Lnenicka, 1997) from the freshwater crayfish are affected by concentrations of ω-agatoxin-IVA between 30 and 600 nmol l⁻¹. In the present study, the Ca²⁺ current in crayfish XO neurones was blocked by ω-agatoxin-IVA at a low extracellular Ca²⁺ concentration (5 mmol l⁻¹); at a normal extracellular Ca²⁺ concentration (13.5 mmol l⁻¹), ω-agatoxin-IVA was ineffective. Cardisoma carnifex XO neurones, in which peptide toxins have no effect on the Ca²⁺ current, may express R-type Ca²⁺ channels differing from the Ca²⁺ channels expressed in the crayfish XO neurones, which appear to be P-type Ca²⁺ channels.

Recently, R. Alvarado-Alvarez, E. Becerra and U. Garcĺa (in preparation) have suggested that the P-type Ca²⁺ current participates in secretory activity, and they have developed a sensitive bioassay to demonstrate this. Their assay is based on the pigmentary matrix retraction of erythrophores cultured together with identified XO neurones that produce red pigment concentrating hormone. Both neuronal Ca²⁺ currents, evoked by depolarizing command pulses, and neuronal firing, induced by depolarizing current injection, were able to induce aggregation on the target cells.

We discounted the existence of N- and Q-type Ca²⁺ channels in crayfish XO neurones because ω-conotoxin-GVIA, which acts selectively on N-type Ca²⁺ channels, did not affect the magnitude or the time course of the Ca²⁺ current in these neurones (Fig. 7B,C). Furthermore, the Q-type channel is resistant to low doses of ω-agatoxin-IVA (30 nmol l⁻¹), but is sensitive to micromolar concentrations of ω-conotoxin-MVIIC (Ellinor et al., 1993), whereas the opposite occurs in XO neurones.

These results do not exclude the existence of R-type Ca²⁺ channels in crayfish XO neurones, since 15 % of the total Ca²⁺ current was resistant to ω-agatoxin-IVA, even at 500 nmol l⁻¹. We suggest that the remaining current could be similar to that described in crab XO neurones (Richmond et al., 1995).

We thank Leopoldo Gonzalez Santos and Benita Mendiola for valuable technical assistance. U.G. was supported by CONACyT grant 1402-N9206.