G protein-mediated FMRFamidergic modulation of calcium influx in dissociated heart muscle cells from squid, Loligo forbesii

Abstract

1. The actions of the neuropeptide FMRFamide (Phe-Met-Arg-Phe-NH₂) on the L-type (I_Ca,L) and T-type (I_Ca,T) calcium currents were investigated in muscle cells dissociated from the heart of squid, Loligo forbseii.

2. The heart muscle cells could be divided into type I and type II cells, on the basis of morphological differences in the dissociated myocytes. FMRFamide induced a substantial block of the L-type calcium current seen in type I cells; this inhibition was rapid, reversible and dose dependent (IC₅₀ = 0.1 μm). FMRFamide induced an increase in the amplitude of the L-type calcium current in the type II heart muscle cells, but had no effect on the T-type calcium current in either type of dissociated heart muscle cell, even at concentrations much higher than those found to affect the L-type calcium current.

3. Internal dialysis of isolated type I heart muscle cells with guanosine 5′-O-(3-thiotriphosphate (GTPγS, 100 μm), a non-hydrolysable GTP analogue, mimicked the FMRFamide inhibition of the Ca²⁺ current and occluded any further FMRFamide-induced inhibition. Internal dialysis of these cells with guanosine 5′-O-(2-thiodiphosphate) (GDPβS, 100 μm) reduced the FMRFamide-induced inhibition of the peak Ca²⁺ current. The inhibitory effects of FMRFamide were abolished by pre-incubation of the cells with pertussis toxin (200 ng ml⁻¹).

4. The activation kinetics of I_Ca,L were not affected by FMRFamide application, nor by internal perfusion with GTPγS, and the FMRFamide-induced reduction in I_Ca,L was not relieved by large depolarising prepulses. These data indicate that FMRFamide can modulate I_Ca,L, but not I_Ca,T, in squid heart muscle cells, and that the underlying G protein pathway is dissimilar to that commonly associated with transmitter modulation of channel activity.

5. The FMRFamide-modulated increase in I_Ca,L seen in the type II heart muscle cells was not mediated by a PTX-sensitive G protein pathway.

FMRFamide-related neuropeptides have a broad distribution in both vertebrate and invertebrate nervous systems (Boer et al. 1980; Greenberg & Price, 1983; Dockray et al. 1983; O'Donohue et al. 1984; Sorenson et al. 1984; Price et al. 1985) and have been found to be involved in a wide variety of physiological roles. For example, in mammals FMRFamide can have quite specific actions, such as the depression of spinal cord reflexes (Huang et al. 1998), as well as more complex effects, such its anxiolytic action on rat social behaviour (Muthal & Chopde, 1994). The mechanisms by which these peptides act have not yet been fully elucidated, but from studies on neurones, it is clear that they can act directly on membrane conductances (Cottrell et al. 1984; Colombaioni et al. 1985; Brezina et al. 1985) or through second messenger systems (Raffa & Stone, 1996).

It is also well recognised that FRMFamide can act on muscle, particularly heart muscle, again either directly on ionic conductances (Brezden et al. 1991) or through second messenger systems (Reich et al. 1997). However, there is very little information on the mechanisms involved. In this paper, using an invertebrate model, we describe the actions of FMRFamide on ionic conductances in heart muscle cells and demonstrate that this peptide has a selective effect on the L-type calcium current, and that this action is mediated predominantly via a G protein pathway.

The excitatory effects of FMRFamide-related peptides on cardiac muscle, first demonstrated by Price & Greenberg (1977), have been confirmed for many other muscle systems (e.g. Cottrell et al. 1983; Evans & Mayers, 1983), but cardioinhibitory effects have also been reported (Painter & Greenberg, 1982). In cephalopods, FMRFamide produces a positive inotropic effect on perfused hearts (Kling & Jakobs, 1987) and recent immunocytochemical studies have demonstrated FMRFamide-like immunoreactivity in nerves innervating squid hearts (Chrachri et al. 1997), thus indicating a putative neurotransmitter role for this peptide.

Dissociated squid heart muscle cells display six specific voltage-sensitive, membrane conductances that influence their electrical behaviour; these include, a high voltage-activated (HVA) calcium current, I_Ca,L, a low voltage-activated (LVA) calcium current, I_Ca,T, a transient sodium current, I_Na, a voltage-gated potassium current, I_K(V), a transient potassium current, I_A, and finally a potassium current which depends on calcium influx, I_K(Ca) (Ödblom & Williamson 1995; Ödblom et al. 2000). All are similar to currents found in a variety of other cell types (e.g. see Connor & Stevens, 1971a,b; Meech & Standen, 1975; Thompson, 1977; Nowycky et al. 1985). Using whole-cell recording techniques (Hamill et al. 1981), we here examine the effects of FMRFamide on the L-type and T-type calcium currents observed in freshly dissociated heart muscle cells from squid, Loligo forbesii. This preparation allows good isolation of these two Ca²⁺ currents on the basis of their pharmacological properties and voltage dependence (Ödblom et al. 2000). We also identify two separate types of heart muscle fibres and investigate whether G proteins represent the link between FMRFamide receptors and the responses seen in the voltage-dependent Ca²⁺ channels. We present evidence that the inhibition of I_Ca,L by FMRFamide is abolished by pertussis toxin, a protein isolated from Bordetella pertussis, which selectively ADP ribosylates the GTP-binding proteins G_i and G_o (Sternweis & Robishaw, 1984) and suppresses the ability of G_i to couple inhibitory receptors to adenylate cyclase (Kurose et al. 1983).

Some of these data have been reported in abstract form (Chrachri et al. 1997).

METHODS

Cell dissociation

For the experiments, squid, Loligo forbseii, were killed by decapitation after brief anaesthesia in 3 % ethanol. Heart muscle cells were isolated enzymatically with trypsin and collagenase P from the ventricular heart muscle. Tissue was incubated for 30 min in 2 mg ml⁻¹ trypsin (Boehringer Mannheim Biochemicals) in calcium-free artificial sea water (ASW), and then for another 20–30 min in 8 mg ml⁻¹ collagenase P (Boehringer Mannheim Biochemicals) in calcium-free ASW. After careful washing with fresh calcium-free ASW, heart muscle cells were dissociated by gentle trituration. The isolated cells were then transferred to the recording chamber and viewed with a Nikon inverted microscope.

Voltage clamp recording

Dissociated heart muscle cells were studied at room temperature (18-20°C) with the whole-cell patch clamp technique (Hamill et al. 1981), using an Axopatch 200A amplifier (Axon Instruments). Pipettes were pulled from soda glass capillaries (Intracel, 1.5 mm o.d., 0.86 mm i.d.). Tip resistances were typically between 2 and 6 MΩ when filled with potassium or caesium aspartate. Currents were filtered at 5 kHz by a low-pass filter. Pulse generation and data acquisition were controlled by a PC using the Patch software suite (Cambridge Electronic Design, UK).

Series resistance was electronically compensated (usually to > 90 %). Thus, voltage errors of only a few millivolts occurred at peak current levels. Liquid junction potentials were measured following the method of Neher (1992) and found to be less than 4 mV with potassium as the pipette internal solution and less than 5 mV with caesium as the internal solution.

Data analysis

Percentage inhibition of currents was quantified with the expression 100 × (1 –I_drug/I_control). We used the software Sigmaplot to calculate the mean, standard deviation (s.d.) and standard error (s.e.m.). Averaged data were presented as means ±s.e.m. The Igor software package (Wavemetrics, Inc.) was used to fit exponential curves to the data to obtain the time constants for activation.

Solutions and drugs

The bath solution of artificial seawater (ASW), contained (mM): 430 NaCl; 10 KCl; 10 CaCl₂; 50 MgCl₂; 10 Hepes, 0.002 tetrodotoxin; pH was adjusted to 7.6 with sodium hydroxide; osmolality was 997 mosmol (kg H₂O)⁻¹. Other solutions were made by equimolar substitution of this basic formula. For the calcium-free ASW magnesium was substituted for calcium. Recording pipettes were filled with a solution containing (mM): 500 D-aspartate; 10 NaCl; 4 MgCl₂; 3 EGTA; 20 Hepes, titrated with caesium hydroxide to a pH of 7.4 (giving a final caesium concentration of about 500 mM); osmolality was 870 mosmol (kg H₂O)⁻¹ to block the outward currents.

FMRFamide (Sigma Chemicals) was bath applied, leaving at least 10 min between successive applications. Usually the first application of FMRFamide was approximately 5–10 min after whole-cell membrane seal break-through. For experiments designed to examine the involvement of G proteins, 100 μM guanosine 5′-O-(3-thiotriphosphate (GTPγS) or 100 μM guanosine 5′-O-(2-thiodiphosphate) (GDPβS) was added to the pipette solution, and in another series, isolated cells were pre-incubated in 200 ng ml⁻¹ pertussis toxin (Bordetella pertussis; PTX) for 12–18 h immediately before the experiments.

RESULTS

Two distinct components of I_Ca: transient and sustained currents

Whole-cell patch clamp recordings of membrane current in dissociated squid heart muscle cells were performed with external and internal solutions that effectively isolated Ca²⁺ channel currents. The inward Na⁺ current was eliminated by including 2 μM tetrodotoxin (TTX) in the external bathing solution and outward K⁺ channel currents were minimised by replacing K⁺ in the internal pipette solution with Cs⁺. Under these conditions, and using 10 mM Ca²⁺ in the external solution as the charge carrier, prominent Ca²⁺ channel currents (I_Ca) with two components, one transient and the other sustained were observed (Fig. 1). Both of these components could be abolished by adding 4 mM Co²⁺ to the bath (not shown), strongly supporting the view that these currents were carried through voltage-dependent calcium channels. In all of the cells tested, I_Ca was stable without any run-down for a minimum of 30 min after breaking into the cell. The calcium current components could be separated on the basis of their voltage dependence and pharmacological properties.

Figure 1. Voltage dependence of T-type Ca²⁺ currents.

A, representative example of membrane currents during a test pulse to −30 mV from a holding potential of −80 mV (a) or −60 mV (b); the difference current (a – b) is the T-type Ca²⁺ current, I_Ca,T. B, peak current-voltage (I–V) relations of the same dissociated heart muscle cell plotted for V_h =−80 mV (•) and V_h =−60 mV (○); ▴, difference current.

Figure 1A (trace b) shows a long-lasting inward current evoked by a depolarisation to −30 mV from a holding potential (V_h) of −60 mV. The current-voltage (I–V) curve shown in Fig. 1B (○) demonstrates that this inward current first appeared at about −30 mV, grew to maximal amplitude at about +10 mV, and became smaller with progressively stronger depolarising pulses. We refer to this component as an L-type Ca²⁺ current (I_Ca,L) because of its long-lasting time course and its blockade by the dihydropyridine antagonist nifedipine (Fig. 2D).

Figure 2. Selective effects of substitution of equimolar Ba²⁺ for Ca²⁺ and of the dihydropyridine antagonist nifedipine on both I_Ca,T and I_Ca,L.

A and B, equimolar Ba²⁺ substitution for Ca²⁺ greatly enhanced the sustained I_Ca component (B), whereas the transient component was reduced (A). C and D, with Ca²⁺ as a charge carrier, nifedipine (Nif.; 2 μM) inhibited 58.8 % of the transient current evoked by a voltage step to −30 mV from a V_h of −80 mV (C), and 81.8 % of the sustained current at 0 mV (D).

A second, transient Ca²⁺ current component could be seen only with test pulses from a more negative holding potential (V_h =−80 mV). With 10 mM Ca²⁺ as the charge carrier, the current-voltage (I–V) curve shown in Fig. 1B (•) demonstrates that this component of the current began to activate at around −50 mV and its amplitude increased progressively with stronger depolarisations, reaching a plateau between −30 and −20 mV. We refer to this component as T-type Ca²⁺ current (I_Ca,T) because of its transient time course. The voltage dependence and amplitude of the squid heart muscle I_Ca,T can be demonstrated by comparing the currents obtained at a hyperpolarised holding potential with those obtained at a relatively more depolarised holding potential. The rapid, transient component of I_Ca was prominent at a holding potential of −80 mV, as was the sustained component. However, at a holding potential of −60 mV, the transient component was no longer present, whereas the sustained current was very little changed. A representative example is shown in Fig. 1.

Ca²⁺ channel types can sometimes be differentiated on the basis of variations in their permeability to different charge carriers (Fox et al. 1987). When the external calcium was replaced with equimolar barium, it was found that the amplitude of I_Ca,L more than doubled for voltage steps to 0 mV from a membrane holding potential of −80 mV (Fig. 2B, n = 10). This enhancement of I_Ca,L when calcium is replaced by barium as the charge carrier has been described in many other preparations (e.g. Carbone & Lux, 1987; Fox et al. 1987; Liu & Lasater, 1994).

On the other hand, under the same conditions, I_Ca,T was reduced by substitution of Ba²⁺ for Ca²⁺ (Fig. 2A, n = 6). This differential influence of Ba²⁺ and Ca²⁺ on the amplitude of the transient and sustained components adds further credence to the view that the two currents are carried through separate Ca²⁺ channel populations.

The dihydropyridines have been widely used to help characterise Ca²⁺ channel types, particularly the long-lasting sustained component seen in various preparations (Fox et al. 1987; Plummer et al. 1989; Liu & Lasater, 1994). The dihydropyridine antagonist nifedipine has been shown to block preferentially the L-type Ca²⁺ current in many preparations (Fox et al. 1987; Chrachri & Williamson, 1997). We found in the present experiments that nifedipine at a concentration of 2–5 μM blocked I_Ca,L (n = 6), but surprisingly, in five out of five cells expressing both components of I_Ca, nifedipine also inhibited the transient current. However, the reduction of I_Ca,L was usually greater than that seen for the transient I_Ca. Figure 2C and D illustrates an example of the effect of nifedipine on a cell expressing both components of I_Ca. Nifedipine (2 μM) blocked 58.8 % of the transient current at a membrane potential of −30 mV (Fig. 2C) and 81.8 % of the sustained component at 0 mV (Fig. 2D).

Nickel has been reported to suppress preferentially the transient component of I_Ca (Hagiwara et al. 1988). In the present experiments (n = 4), 40 μM Ni²⁺ was sufficient to block the transient component without having any effect on the long-lasting sustained component (data not shown). These results support further the finding that squid heart muscle cells express two I_Ca components.

Muscle cell types and FMRFamide-mediated attenuation of the L-type calcium currents (I_Ca,L)

The cell dissociation protocol used in these experiments produced two morphologically separate types of heart muscle cells. The majority of the dissociated muscle cells, here called type I, were between 100 and 350 μm in length, about 10 μm in width, and had smooth surface membranes (e.g. Fig. 3A). Type II cells made up less than 30 % of the recovered cells and were of about the same dimensions as the type I cells, but had heavily invaginated surface membranes (e.g. Fig. 7A and B). Both cell types showed occasional spontaneous rhythmic contractions and both displayed the two components of the calcium current described above. However, the different cell types responded very differently to the application of FMRFamide as described below.

Figure 3. FMRFamide-mediated inhibition of I_Ca,L in a type I muscle cell.

A, morphology of a type I dissociated heart muscle cell. Scale bar, 50 μm. B, current traces recorded in response to membrane depolarisation to a voltage step of −10 mV from a holding potential of −40 mV before (control) and after bath application of 1 μM FMRFamide (FMRFa). The effect of this neuropeptide on the calcium current was reversible (wash). C, time course of the onset, and recovery from, the effect of FMRFamide on I_Ca,L. D, concentration-response relationship for inhibition of peak Ca²⁺ currents by FMRFamide. Data points represent means ±s.e.m. Number of heart muscle cells tested is indicated in parentheses next to the data points. The curve is the best fit to the data points obtained using a non-linear regression equation for the inhibition of I_Ca = 86.9(1 + IC₅₀/[FMRFa])⁻¹, where [FMRFa] is the concentration of FMRFamide. The concentration of FMRFamide required to produce half-maximal inhibition (IC₅₀) was 0.1 μM.

Figure 7. FMRFamide-mediated increase in I_Ca,L in type II muscle cells.

A, morphology of a type II dissociated heart muscle cell. B, same cell at a higher magnification. Scale bars in A and B, 50 μm. Left traces in C, current traces in response to a voltage step to −40 mV from a holding potential of −80 mV showing that FMRFamide had no apparent effect on the transient component of I_Ca. Right traces in C, current traces recorded in response to membrane depolarisation to a voltage step of 0 mV from a holding potential of −80 mV before (control) and after bath application of 1 μM FMRFamide demonstrating that FMRFamide increases the amplitude of I_Ca,L in type II heart cells. D, I–V relationship of calcium currents under control conditions (○), and after bath application of 1 μM FMRFamide (•).

In 83 % of our experiments, using type I dissociated heart muscle cells, bath application of 1 μM FMRFamide reversibly decreased the amplitude of the I_Ca,L by ≥ 60 % (n = 53). An example is illustrated in Fig. 3B, where the inhibition of I_Ca,L was about 74 %. The magnitude of this inhibition did not change on repeated application of FMRFamide to the same dissociated heart muscle cell (data not shown). Figure 3C illustrates the time course of the onset of the inhibition of I_Ca,L by FMRFamide and the subsequent recovery by washing, and demonstrates that the inhibition was relatively rapid and fully reversible. The I–V curves show that FMRFamide reduced the amplitude of I_Ca,L over most of the voltage ranges (Fig. 9B), but this reduction was less pronounced for voltage steps more positive than +10 mV.

Figure 9. FMRFamide has no effect on I_Ca,T in type I heart muscle cells.

A, current traces recorded in response to membrane depolarisation induced by a voltage step to −40 mV from a holding potential of −80 mV before and after bath application of FMRFamide, demonstrating that it has no effect on the T-type calcium current. B, I–V relationship of both I_Ca,L and I_Ca,T before (○) and after bath application of 2 μM FMRFamide (•).

FMRFamide concentrations in the range from 10⁻⁹ to 10⁻⁴ M were used to study the effect of FMRFamide on the amplitude of I_Ca,L in dissociated heart muscle cells. The decrease in the amplitude of this current by FMRFamide was dose dependent as illustrated in Fig. 3D. Bath application of 10⁻⁹ M FMRFamide had no effect on I_Ca,L. When the concentration of FMRFamide was increased to 10⁻⁸ M, the amplitude of I_Ca,L was reduced by 6 ± 4 % (n = 3). On further increasing the concentration to 10⁻⁷ M, the amplitude of I_Ca,L was reduced by 43 ± 15 % (n = 5), and at a concentration of 10⁻⁶ M, the amplitude of I_Ca,L was reduced by 82 ± 20 % (n = 8). There was very little further increase in the extent of calcium current suppression with higher concentrations of FMRFamide. The threshold concentration for the inhibitory effect on the amplitude of I_Ca,L was around 10⁻⁸ M and the IC₅₀ was about 0.1 μM (Fig. 3D). In all subsequent experiments, unless otherwise stated, a concentration of 1 μM FMRFamide was used.

No relief of FMRFamide-induced inhibition by strong conditioning prepulses

In many cell types, strong depolarisations have been shown to enhance or facilitate HVA calcium currents (Dolphin, 1996), as well as to reverse neurotransmitter modulations (Howe & Surmeier, 1995). These are thought to be related phenomena resulting from the disruption of an interaction between a G protein and the channel (Bean, 1989).

The voltage dependence of FMRFamide-induced inhibition was examined by comparing the current elicited by a 57 ms duration test pulse (TP₁) to 0 mV with the current elicited by the same test pulse (TP₂) when it was preceded by a 76 ms duration positive prepulse to +100 mV (protocol in Fig. 4A). In all cells tested, there was no significant change in the FMRFamide-depressed current, even with strongly depolarising conditioning prepulses. Without a prepulse, FMRFamide inhibited 61.38 ± 3.84 % (n = 12) of the peak Ca²⁺ currents whereas after a conditioning prepulse the inhibition was 52.23 ± 3.59 % (n = 12, Fig. 4B). Using Welch's test, the data obtained before and after the depolarising pulse were found to be not significantly different (P = 0.1). In the absence of FMRFamide, the conditioning prepulse produced a small, but significant 10 ± 2.15 % (P = 0.01, n = 12) increase in the amplitude of the peak Ca²⁺ current (Fig. 4A).

Figure 4. Effect of a conditioning prepulse on FMRFamide-induced inhibition of I_Ca,L in a type I muscle cell.

Upper traces in A show the experimental voltage protocol (mV). Lower left traces in A show that with no prepulse FMRFamide inhibited Ca²⁺ current by 68 % in this cell; the lower right traces show that after a conditioning prepulse, the inhibition was not significantly attenuated and was about 63 %. TP₁ and TP₂, first and second test pulses. B, summary of inhibition of peak Ca²⁺ currents by FMRFamide without and with a conditioning prepulse (PP). In B, data represent means ±s.e.m. Number of cells tested is indicated in parentheses.

G proteins and FMRFamide-induced inhibition of I_Ca,L in type I muscle cells

To investigate the involvement of G proteins in the FMRFamide-induced inhibition of I_Ca,L in dissociated heart muscle cells of squid, a group of type I dissociated cells was dialysed with GTPγS (100 μM). GTPγS binds tightly to the α subunits of G proteins and confers a receptor-independent, irreversible G protein activation of I_Ca,L. During an 11 min dialysis with GTPγS in the absence of FMRFamide, peak Ca²⁺ currents showed a time-dependent decrease in amplitude, but little change in the activation kinetics (see below; Fig. 5A). This phenomenon was unlikely to be due to ‘run-down’ of the Ca²⁺ current for no run-down was observed under normal whole-cell clamp conditions or when the internal pipette solution contained chemicals such as ATP and GTP (not shown). After maximal G protein activation of I_Ca,L by GTPγS, application of FMRFamide inhibited the remaining Ca²⁺ current by only 8.26 ± 1.32 % (n = 5), which is significantly smaller (P = 0.001) than the inhibition of 79.54 ± 7.04 % when dissociated heart muscle cells of the same preparation were dialysed with control pipette solution (Fig. 5B).

Figure 5. Effect of GTPγS and GDPβS on FMRFamide-induced inhibition of I_Ca,L in type I muscle cells.

A, superimposed Ca²⁺ current traces at the start (0 min) and after 11 min of internal dialysis with GTPγS (100 μM). FMRFamide was applied after a steady-state inhibition of the peak Ca²⁺ current by GTPγS was attained. B, comparison of percentage inhibition of peak Ca²⁺ current in dissociated heart muscle cells dialysed with control internal patch pipette solution (□) and 100 μM GTPγS (▪). C, superimposed Ca²⁺ current traces evoked in the absence and presence of FMRFamide after 10 min of dialysis with GDPβS (100 μM). D, summary of percentage inhibition of peak Ca²⁺ current in dissociated heart muscle cells dialysed with control internal patch pipette solution (□) and 100 μM GDPβS (▪). For B and D, data represent means ±s.e.m. Number of cells tested is indicated in parentheses.

The non-hydrolysable GDP analogue GDPβS has been found to abolish the G protein-mediated effects of transmitters by acting as a competitive inhibitor of GTP and binding to the α subunits of G proteins (Holz et al. 1986). When GDPβS (100 μM) was added to the patch pipette solution, the peak I_Ca,L amplitude was found to be similar to that recorded in dissociated heart muscle cells dialysed with control pipette solution. Subsequent bath application of FMRFamide had no effect on the peak I_Ca,L amplitude or activation kinetics (Fig. 5C). When muscle cells were dialysed with GDPβS for 10 min, FMRFamide-induced inhibition of the peak Ca²⁺ currents was significantly attenuated to only 13.87 ± 3.68 % of pre-dialysis levels compared with 75.73 ± 6.11 % attenuation (n = 6, P < 0.0001) in heart muscle cells of the same preparation dialysed with control pipette solution (Fig. 5D).

To identify the nature of the G protein coupling FMRFamidergic receptors to Ca²⁺ channels, isolated heart muscle cells were incubated overnight in a medium containing 200 ng ml⁻¹ PTX. This toxin has been shown to uncouple G_i/o proteins from receptors by ADP ribosylation (Sternweis & Robishaw, 1984). Figure 6A shows that PTX treatment significantly (P < 0.0001) attenuated the inhibition of peak Ca²⁺ currents by FMRFamide. In control heart muscle cells incubated in PTX-free medium overnight, the mean inhibition of peak Ca²⁺ currents by FMRFamide was 76.13 ± 5.21 % of pre-incubation levels (n = 6), whereas in cells treated with PTX overnight, the inhibition of I_Ca,L was attenuated to only 14.17 ± 3.11 % (n = 6) (Fig. 6B).

Figure 6. Involvement of pertussis toxin (PTX)-sensitive G protein in FMRFamide-induced suppression of I_Ca,L in type I muscle cells.

A, current traces recorded from PTX pre-treated heart muscle cell in response to membrane depolarisation induced by a voltage step to +10 mV from a holding potential of −60 mV before (control) and after bath application of 1 μM FMRFamide. This figure shows that PTX pre-treated cells no longer responded to FMRFamide application. B, summary of percentage inhibition of peak Ca²⁺ current in dissociated heart muscle cells incubated overnight in a PTX-free medium (□) and in medium containing 200 ng ml⁻¹ PTX (▪). Data represent means ±s.e.m. Number of cells tested is indicated in parentheses.

Thus, the results obtained with GTPγS, GDPβS and PTX strongly suggest that a PTX-sensitive G protein mediates the FMRFamidergic inhibition of the L-type Ca²⁺ currents in dissociated, type I heart muscle cells of squid.

No FMRFamide-induced change in I_Ca,L activation kinetics

G protein-mediated inhibition of voltage-sensitive calcium currents is often characterised by a slowing of the activation kinetics (Dolphin, 1996). However, in the case of dissociated squid heart cells, there was no significant change in the speed of activation of I_Ca,L when FMRFamide was present (τ = 2.69 ms) compared to control values (τ = 2.81 ms) (Wilcoxon test, P = 0.84, n = 8). Similarly, there was no slowing of the activation kinetics when GTPγS was used to activate the G protein directly by including it in the pipette solution (e.g. Fig. 5A: in the experiments illustrated, τ was 2.1 ms for activation immediately after whole-cell break-through and 1.73 ms for activation 10 min later; P = 0.74, n = 8).

FMRFamide-mediated enhancement of I_Ca,L in type II cells

The second class of heart muscle cells seen after tissue dissociation were of similar sizes to the type I cells but had highly invaginated cell membranes (Fig. 7A and B). Whole-cell recordings from these type II cells, which represent the remaining 17 % of our experiments (n = 11), demonstrated that bath application of 1 μM FMRFamide this time increased the amplitude of I_Ca,L by about 57 % (Fig. 7C, I_Ca,L). Current-voltage (I–V) curves obtained from a holding potential of −80 mV before (○) and after (•) FMRFamide exposure in the same dissociated heart muscle cell showed that FMRFamide increased the amplitude of I_Ca,L over most of the voltage range without changing the potential at which the amplitude was maximum. However, the increase induced by FMRFamide was less pronounced for steps to depolarised potentials more positive than +20 mV (Fig. 7D).

In order to investigate whether this enhancement of I_Ca,L was mediated by a PTX-sensitive G protein, type II cells were also incubated overnight in medium containing 200 ng ml⁻¹ PTX and their subsequent responses to FMRFamide recorded. Figure 8A illustrates an example of the response to the application of FMRFamide (2 μM), demonstrating a clear increase (61.53 ± 2.98 %; n = 3) in the magnitude of I_Ca,L in PTX-treated heart muscle cells (Fig. 8B). In control heart muscle cells, incubated in PTX-free medium overnight from the same preparation, a similar mean increase in peak Ca²⁺ current induced by FMRFamide application of 62.37 ± 2.09 % (n = 3) was recorded. Therefore, PTX treatment had no significant effect (P = 0.8) on the increase in I_Ca,L in dissociated heart muscle type II cells, implying that, unlike the inhibition of I_Ca,L in type I heart cells, this enhancement of I_Ca,L is not mediated via a PTX-sensitive G protein.

Figure 8. The effect of pertussis toxin on FMRFamide-induced enhancement of I_Ca,L in type II muscle cells.

A, current traces recorded in response to membrane depolarisation induced by a voltage step to 10 mV from a holding potential of −60 mV before (control) and after bath application of 1 μM FMRFamide. This figure shows that FMRFamide increased I_Ca,L, even in PTX pre-treated type II heart muscle cells, indicating that the increase of the calcium current does not involve a PTX-sensitive G protein. B, summary of percentage increase of peak Ca²⁺ current in type II cells incubated overnight in a PTX-free medium (□) and in medium containing 200 ng ml⁻¹ PTX (▪). Data represent means ±s.e.m. Number of cells tested is indicated in parentheses.

Effect of FMRFamide on I_Ca,T

FMRFamide had no significant effect on the T-type calcium current even when the concentration applied was higher than that found to affect the L-type calcium current, and this was regardless of whether FMRFamide inhibited (type I cells) or enhanced (type II cells) I_Ca,L. Figure 9 illustrates an experiment from a type I heart cell where bath application of 1 μM FMRFamide was shown to have no significant effect on the amplitude of the evoked I_Ca,T (Fig. 9A, I_Ca,T). By contrast, in the same dissociated heart muscle cell, 1 μM FMRFamide reduced the amplitude of I_Ca,L (traces not shown). The I–V curve provides further evidence that FMRFamide selectively affects the amplitude of I_Ca,L (Fig. 9B). This was also true when recordings were made from type II heart muscle cells where application of 1 μM FMRFamide had no significant effect on I_Ca,T (Fig. 7C, I_Ca,T) but increased the amplitude of peak I_Ca,L (Fig. 7D, I_Ca,L).

DISCUSSION

These experiments have demonstrated that FMRFamide modulates I_Ca,L, but not I_Ca,T in cardiac myocytes from squid, and also that the peptide has opposing actions on two morphologically different types of myocytes. The suppression of I_Ca,L in type I cells operates through a G protein, but the mechanism is dissimilar to those commonly described for peptide modulation in neuronal tissue. The enhancement of I_Ca,L in type II cells does not appear to operate through a G protein pathway.

Identification of the calcium channel subtypes

We identified two voltage-sensitive calcium currents in the myocytes from squid heart cells. One is termed I_Ca,L since it shows features similar to those seen in the L-type (Nowycky et al. 1985) or high voltage-activated (HVA) current (Carbone & Lux, 1987) of other preparations in that it is activated at relatively depolarised voltage thresholds (-20 to −10 mV from a holding potential of −60 mV), has slow inactivation kinetics, has a higher permeability to Ba²⁺ than to Ca²⁺, and is inhibited by the dihydropyridine antagonist nifedipine.

The second voltage-sensitive calcium current we identified as I_Ca,T, as its characteristics are similar to the transient, low-threshold T-type (Nowycky et al. 1985) or low voltage activated (LVA) current (Carbone & Lux, 1987) found in many other preparations (e.g. Fox et al. 1987; Liu & Lasater, 1994). In comparison with I_Ca,L, this current was activated at more negative potentials, inactivated rapidly, was reduced when barium was substituted as the charge carrier, and was blocked by nickel. It is notable that I_Ca,T in squid myocytes was blocked by micromolar concentrations of the dihydropyridine antagonist nifedipine, which has been considered to be specific for the L-type Ca²⁺ channels (Scott et al. 1991). However, similar results have been reported by Bean (1985) in atrial cells and Liu & Lasater (1994) in retinal ganglion cells.

Selective FMRFamide modulation of I_Ca

FMRFamide selectively decreased the amplitude of I_Ca,L in type I squid myocytes, but increased its amplitude in type II cells. It is not yet known if these two muscle cell types show other differences but Kling & Schipp (1987) described two muscle types in the octopus heart: central muscle bundles which showed high levels of succinate dehydrogenase activity, and peripheral muscle bundles with much lower activity levels. It may be that these two types correlate with the observed responses to FMRFamide in which case it is likely that the more numerous type I cells will equate with the more numerous centrally located, high enzyme activity cells.

In both cell types, we found that FMRFamide had no effect on I_Ca,T. Selective modulability by neurotransmitters has been reported for L- and T-type channels (Marchetti et al. 1986), and for the L- and N-type channels following the activation of a type B γ-aminobutyric acid receptor in dorsal root ganglion neurones (Scott & Dolphin, 1986) or activation of a muscarinic receptor in rat sympathetic neurones (Wanke et al. 1987). Recently, Dreijer et al. (1995) reported selective modulability by FMRFamide of a high-voltage-activated calcium current in isolated neuroendocrine caudodorsal cells in the mollusc Lymnea stagnalis.

Mechanisms of FMRFamide-induced modulation of I_Ca,L

In many instances, where neurotransmitter receptors are known to be coupled to voltage-activated Ca²⁺ channels, this is achieved via membrane delimited PTX-sensitive G protein pathways (for reviews see Dolphin, 1995; Hille, 1994). This also appears to be the case for the FMRFamide-induced suppression of the Ca²⁺ currents in squid heart cells for the following reasons. First, in the above experiments, the internal dialysis of the non-hydrolysable GTP analogue GTPγS could mimic the inhibition of the Ca²⁺ current. Furthermore, if FMRFamide was then applied, its inhibitory action was significantly reduced. Second, the internal dialysis of the non-hydrolysable GDP analogue GDPβS significantly attenuated the FMRFamide-induced response. Third, overnight incubation of isolated heart muscle cells in a PTX-containing medium also significantly attenuated the FMRFamide response.

However, there were also differences between the responses observed here for FMRFamide and those commonly seen in other G protein-activated systems. Firstly, a strong depolarising prepulse did not result in the relief of the FMRFamide-induced inhibition in dissociated squid heart cells. Secondly, no slowing of I_CaL activation was observed following FMRFamide application or intracellular perfusion with GTPγS. These points indicate that there is a second, less common, mode of inhibition of Ca²⁺ currents, which is mediated by a GTP-binding protein, without affecting the activation kinetics. Our data therefore, resemble those of Toselli & Taglietti (1995), who found that muscarine produced a decrease in Ca²⁺ current amplitude via a pertussis toxin-sensitive G protein without any change in the current activation kinetics or relief of the Ca²⁺ current inhibition after a strong depolarising pulse. Similarly, Brezina et al. (1985) demonstrated that the FMRFamide-induced inhibition of the Ca²⁺ current in Aplysia neurones is mediated by activation of a GTP-binding protein without any apparent slowing of the activation kinetics.

PTX-sensitive G proteins are known to be involved in the inhibitory actions of transmitters on calcium currents in many types of neurones. For example, in molluscs, dopamine inhibits voltage-dependent calcium currents in identified neurones D1, D2, F76, and F77 of Helix aspera through a G_o protein (Harris-Warrick et al. 1988). In vertebrates, the N-type and P/Q-type calcium currents have been found to be subject to direct PTX-sensitive G protein inhibition (for reviews, see Hille, 1994; Dolphin, 1995). In embryonic rat hippocampal neurones, Toselli & Lux (1989) have demonstrated that ACh inhibits the voltage-dependent calcium channels through a GTP-binding protein, and in pituitary cell lines, Lewis et al. (1986) demonstrated that somatostatin reduces the voltage-dependent Ca²⁺ current via a PTX-sensitive G protein.

There are few reports of the action of FMRFamide on muscle ion channels but there is evidence for the direct action of FMRFamide on neuronal ion channels. Brezina et al. (1989) demonstrated that FMRFamide-induced inhibition of Ca²⁺ currents in Aplysia abdominal ganglion neurones through a G protein pathway, and Dreijer et al. (1995) found that FMRFamide evoked an inhibition of the slowly inactivating L-type current in Lymnea caudodorsal neurones. Similarly, Man-Son-Hing & Haydon (1989) reported the inhibition of HVA currents in growth cones of isolated B5 neurones of Helisoma by FMRFamide, and this could be blocked by GDPβS and mimicked by intracellular dialysis of GTPγS.

Because of their relatively infrequent occurrence, we have yet to study in detail the mechanism underlying the FMRFamide-induced enhancement of I_Ca,L which we observed in type II squid heart muscle cells. However, our data indicate that this increase in current amplitude was not mediated by a PTX-sensitive G protein. A similar observation was made by Yakel (1991) who found that FMRFamide enhanced the calcium currents in around 20 % the Helix central neurones examined, but that this was not mediated by a PTX-sensitive G protein.

Functional significance

Recent immunocytochemical studies have identified FMRFamide-like immunoreactivity in cells innervating the squid heart (Chrachri et al. 1997), thus supporting the view that the effects of exogenous FMRFamide reported here probably mirror the endogenous modulation of heart muscle cells by FMRFamidergic nerve fibres. Exogenous FMRFamide applied to isolated heart muscle cells either reduces or enhances I_Ca,L without apparently changing its voltage dependence. The fact that we observed two different effects in response to FMRFamide in isolated heart muscle cells may be attributed to the morphological (compare Figs 3A and 7B) and biochemical (Kling & Jakobs, 1987) diversity of the cell types present.

It is not surprising that FMRFamide affected only I_Ca,L and not I_Ca,T in dissociated heart muscle cells, because, when applied to an isolated whole heart, FMRFamide induces a long-lasting change in the amplitude of the heart beat, without any significant modulation of frequency (Kling & Jakobs, 1987). Since I_Ca,T has been proposed to control rhythmic membrane oscillations in various central neurones (Llinas & Yarom, 1981), the lack of effect of FMRFamide on this type of Ca²⁺ current might have been expected from the finding that FMRFamide had no significant effect on the frequency of heart beat when applied to a whole heart (Kling & Jakobs, 1987).

In conclusion, these experiments have shown that FMRFamide can modulate the L-type Ca²⁺ current, but not the T-type Ca²⁺ current, in dissociated squid heart muscle cells, and that the inhibitory effects of FMRFamide are mediated by a pertussis toxin-sensitive G protein pathway. However, the underlying mechanism is dissimilar to that commonly associated with transmitter modulation of G protein activity. The results extend our understanding of the action of FMRFamide and demonstrate the parallels between its mode of operation on neurones and muscles.