Cytochalasin D reduces Ca²⁺ currents via cofilin-activated depolymerization of F-actin in guinea-pig cardiomyocytes

Abstract

1. L-type Ca²⁺ channel currents (I_Ca) were measured in guinea-pig ventricular myocytes (22 °C, 300 ms steps from -45 to +10 mV). Pulsing at 0.5 Hz reduced I_Ca within 5 min to 92 ± 3 % (mean ±s.e.m., n = 14) and within 10 min to 83 ± 4 % (‘run-down’ with reference to I_Ca after a 5 min equilibration period).

2. Bath-applied cytochalasin D (cytD, 10 μm) reduced I_Ca to 75 ± 4 % within 5 min and to 61 ± 4 % within 10 min (‘cytD reduction of I_Ca‘) by reduction of maximal Ca²⁺ conductance (suggested by fits of time course and of current-potential (I–V) curves).

3. Preincubation with phalloidin (bath applied, 100 μm, 5 h) prevented the cytD reduction of I_Ca. Since phalloidin specifically blocks F-actin depolymerization, cytD reduction of I_Ca is linked to depolymerization of F-actin.

4. CytD did not attenuate the β-adrenergic stimulation of I_Ca (30 nm isoproterenol), suggesting that A kinase anchoring proteins are unlikely to mediate the cytD reduction of I_Ca. The cytD reduction of I_Ca was abolished by extra-/intracellular acidosis (pH_o 6.9), by cell dialysis of 5 mm BAPTA, or by serine/threonine protein phosphatase inhibitors.

5. Actin-depolymerizing factor (ADF)/cofilin are proteins that bind to actin, mediate a pH-sensitive depolymerization of F-actin, and are activated by dephosphorylation. Western blots from hearts perfused with solutions containing zero or 10 μm cytD indicated that cytD reduces the ratio of phosphorylated to total ADF/cofilin content by 50 %.

6. The data support the concept that cytD mediates dephosphorylation and activation of ADF/cofilin, leading to depolymerization of F-actin with a subsequent reduction of I_Ca.

Modulation of ion channels and transporters by the actin and tubulin components of the cytoskeleton is a common phenomenon. In cardiac preparations, actin-dependent modulation has been reported for Na⁺-Ca²⁺ exchangers (Li et al. 1993), Na⁺-K⁺-ATPases (Shibayama et al. 1993), ryanodine receptors (Bourguignon et al. 1995) and voltage-gated Na⁺ channels (Srinivasan et al. 1988), and the influence of tubulin has been shown for Na⁺ and ATP-sensitive K⁺ (K_ATP) channels (Undrovinas et al. 1995; Brady et al. 1996). With regard to L-type Ca²⁺ channel current (I_Ca), evidence for cytoskeleton-dependent regulation was provided by increased Ca²⁺ channel activity induced by cell swelling in hyposmotic solutions (Matsuda et al. 1996; Xu et al. 1997) and by altered kinetics of single channel and whole-cell Ca²⁺ channel currents in the presence of compounds known to interfere with actin or tubulin (Johnson & Byerly, 1993; Galli & DeFelice, 1994; Nakamura et al. 2000). Despite this knowledge, the modulation of I_Ca by F-actin is still controversial for the cardiac ventricular myocyte, as two recent reports suggest that cytochalasin D (cytD) has no effect on I_Ca (Undrovinas & Maltsev, 1998; Pascarel et al. 1999).

CytD is a compound known to break down F-actin fibres by preventing the spontaneous polymerization of G-actin at the barbed ends of F-actin. Here, we show that cytD does reduce peak I_Ca, but only if cytosolic Ca²⁺ ions are not chelated by BAPTA or EGTA. We further show that the cytD reduction of I_Ca can be blocked by acidosis or by serine/threonine protein phosphatase inhibitors. The dependence of the reduction of I_Ca by cytD on Ca²⁺, pH and dephosphorylation points to the possible involvement of actin-depolymerizing factor (ADF)/ cofilin (Hayden et al. 1993). ADF/cofilin belong to a family of actin-binding proteins that are responsible for the rapid turnover of actin seen in vivo (Rosenblatt et al. 1997). ADF/cofilin proteins are activated by the Ca²⁺-dependent protein phosphatase PP2B and by the Ca²⁺-independent protein phosphatase PP1 by dephosphorylation on Ser³ (Agnew et al. 1995; Meberg et al. 1998). Binding of activated ADF/cofilin to actin is competitive with tropomyosin and phalloidin (Bernstein & Bamburg, 1982; Hayden et al. 1993). Binding of activated ADF/cofilin destabilizes the interactions between monomers of G-actin, resulting in the depolymerization of F-actin (McGough et al. 1997). Here, we demonstrate that cytD induces dephosphorylation of ADF/cofilin within the time course expected from the reduction of I_Ca. We suggest that depolymerization of F-actin by ADF/cofilin may be involved in the regulation of the cardiac L-type Ca²⁺ channel.

METHODS

Cell isolation and solutions

Adult guinea-pigs (≈300 g) were killed by cervical dislocation in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication no. 85-23, revised 1985) and with the local ethics committee. Ventricular myocytes were isolated by a standard collagenase dissociation technique. During the experiment, the cells were continuously superfused with an extracellular solution composed of (mm): 150 NaCl, 5.4 CsCl, 1.8 CaCl₂, 1.2 MgCl₂, 20 glucose and 5 Hepes (pH 7.4). Whole-cell patch-clamp recordings were performed with pipettes (1.5 MΩ resistance) filled with (mm): 140 CsCl, 5 NaCl, 0.5 MgCl₂, 0.005 EGTA and 10 Hepes, pH 7.4.

Measurement of whole-cell currents

Starting from a holding potential of -45 mV, 300 ms pulses to 0 mV were applied at 0.5 Hz. Currents were recorded with an RK300 amplifier (Biologic, Echirolle, France) connected to a personal computer via a CED-1401 interface (CED, Cambridge, UK). The peak current through L-type Ca²⁺ channels (I_Ca) was estimated as difference negative peak current minus late current at the end of the 300 ms pulse (see Isenberg & Klöckner, 1982). During the period of the experiment (≈15 min), access and seal resistance remained constant.

Toxins and drugs

CytD (10 mm) was dissolved in DMSO and diluted 1:1000 or 1:2000 to a final concentration of 10 or 5 μm. Phalloidin (100 μm), isoproterenol (isoprenaline, 30 nm) and protein phosphatase inhibitor cocktail I (1:100) were directly dissolved in physiological saline solution (PSS). All toxins and drugs were obtained from Sigma (Deisenhofen, Germany).

Protein isolation and Western blots

After perfusion of the spontaneously beating isolated heart for 10 min with either Tyrode solution or Tyrode solution plus 10 μm cytD, the ventricles were cut off and shock-frozen in liquid nitrogen. Frozen samples were ground and homogenized in lysis buffer composed of: 20 mm Tris, 250 mm sucrose, 3 mm EGTA, 20 mm EDTA, 0.5 % Triton X-100 (pH 8.0), protein phosphatase inhibitor cocktail I (1:100) and protease inhibitor cocktail (1:20; both from Sigma). After 30 min lysates were cleared by centrifugation and protein concentration was determined using the BCA assay (Pierce, Rockford, USA).

Proteins were separated via 12 % SDS-PAGE and subsequently transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad, Hercules, CA, USA). Immunological detection of total and phospho-ADF/cofilin was performed with primary antibodies that were characterized and kindly provided by Dr Bamburg (Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO 80523-1870, USA; Bamburg & Bray, 1987; Meberg et al. 1998). Blots were visualized with the ECL Western blotting analysis system (Amersham Pharmacia Biotech, Freiburg, Germany) as described by the manufacturer.

Statistics

Values are given as means ±s.e.m. Significance was determined by Student's t test or ANOVA Bonferroni test, and was assumed at P < 0.05.

RESULTS

Spontaneous run-down of I_Ca

The L-type Ca²⁺ current (I_Ca) of guinea-pig ventricular myocytes was measured 5 min after whole-cell access, a time sufficient to equilibrate the cytosol with the Cs⁺-containing electrode solution. In control conditions (i.e. in the absence of pharmacological interventions) peak I_Ca was not stable but slowly ran down (Table 1, Control), in accordance with the literature (Kameyama et al. 1997). Since the run-down of I_Ca superimposes on the possible effects of cytD, ‘control I_Ca’ was measured at identical times in the absence of cytD. Within 5 min peak I_Ca had fallen to 92 ± 3 % (n = 14) and within 10 min to 83 ± 4 % of control peak I_Ca.

Table 1.

Peak I_Ca under the influence of toxins and drugs

	Concentration	n	2.5 min	5 min	10 min
Control (no intervention)	—	14	96.4 ± 1.5	92.0 ± 3.2	82.8 ± 4.4
CytD	5 μM	6	89.3 ± 2.7	82.3 ± 4.8	73.7 ± 9.3
CytD	10 μM	15	85.4 ± 2.2*	75.4 ± 4.4*	60.6 ± 4.3*
Phalloidin	100 μM	6	96.0 ± 2.3	89.8 ± 1.0	81.7 ± 1.6
Phalloidin + cytD10	100 μM	6	93.8 ± 1.3†	86.3 ± 1.5†	78.8 ± 1.9†
BAPTA	5 mM	3	97.3 ± 2.7	91.7 ± 4.9	86.3 ± 5.7
BAPTA + cytD10	5 mM	7	90.9 ± 2.8	87.3 ± 4.7	85.7 ± 5.2†
PPase inhibitors	—	7	100.0 ± 2.5	97.6 ± 2.0	91.4 ± 3.4
PPase inhibitors + cytD10	—	7	98.0 ± 1.9†	93.1 ± 2.4†	84.6 ± 5.3†
pHo 6.9	—	6	95.3 ± 1.7	89.5 ± 3.8	79.3 ± 3.7
pHo 6.9 + cytD10	—	7	92.6 ± 2.1	86.3 ± 4.1	80.9 ± 5.6†

Values indicate percentage of I_Ca at the listed time with reference to I_Ca at t = 0 min (corresponds to 5 min equilibration). Means ±s.e.m., n = number of myocytes. CytD10, 10 μM cytD; PPase, protein phosphatase.

^*P < 0.05 vs. Control.

^†P < 0.05 vs. 10 μM cytD.

CytD reduction of I_Ca

CytD reduced the peak I_Ca; the extent of the reduction increased with concentration and exposure time (Table 1). CytD (10 μm) reduced peak I_Ca to 75 ± 4 % within 5 min and to 61 ± 4 % within 10 min (n = 15). Both values were significantly smaller than the I_Ca values in the absence of the drug (control run-down, P = 0.01).

CytD reduction of maximal conductance (G_max)

The cytD reduction of I_Ca did not significantly change the time constants of inactivation. This conclusion is derived from fits with two exponentials (Isenberg & Klöckner, 1982):

(1)

where A_s and A_f are the amplitudes and τ_s and τ_f the time constants of the slow and the fast component, respectively (Fig. 1). CytD reduced A_s and A_f without significant effects on the τ_f and τ_s values (Table 2). This result let us attribute the cytD reduction of I_Ca to the reduction of G_max (amplitudes A_s and A_f) rather than to a change in the inactivation kinetics. The idea that cytD reduces peak I_Ca via G_max was further supported by the effects of cytD on the voltage dependence of peak I_Ca. Fits according to:

Figure 1. Reduction of L-type I_Ca by 10 μm cytochalasin D (cytD).

Current traces due to 300 ms pulses from -45 to 10 mV, 5 min after rupture of the patch (A) and 10 min after addition of 10 μm cytD (B). Inactivation time course was fitted by sum of two exponentials (thin line) I_Ca(t) =A_sexp(-t/τ_s) +A_fexp(-t/τ_f). CytD reduced the amplitudes A_s (900 pA control, 700 pA cytD) and A_f (1800 pA control, 1100 pA cytD) without modifying the time constants τ_s (110 ms control, 115 ms cytD) and τ_f (15 ms control, 16 ms cytD). C, 2.5, 5 and 10 min after rupture of the patch, the I–V relationship of I_Ca was recorded and peak I_Ca was determined. Under control conditions, peak I_Ca fell to 82.8 % within 10 min (run-down, ○). Application of cytD dose-dependently accelerated the decline of peak I_Ca to 73.7 % (5 μm, •) and to 60.6 % (10 μm, ▾). *P < 0.05.

Table 2.

I_Ca, parameters of inactivation time course

	Control		CytD
Parameter	0 min	10 min	0 min	10 min
As (pA)	873.6 ± 189.2	783.0 ± 110.4	945.0 ± 103.7	734.2 ± 91.3
Af (pA)	1749.6 ± 369.7	1123.2 ± 276.8	1879.2 ± 261.7	1144.8 ± 115.8*
τs (ms)	111.8 ± 8.3	110.2 ± 8.8	105.2 ± 9.0	104.8 ± 7.9
τf (ms)	14.2 ± 1.9	16.6 ± 2.7	13.4 ± 2.0	16.6 ± 1.2

I_Ca analysed during 300 ms pulses at 10 mV. I_Ca data were fitted according to I_Ca(t) =A_sexp(–t/_s) +A_fexp(–t/τ_f), where A_s and A_f are amplitudes and τ_s and τ_f time constants of the slow and the fast component, respectively (see Fig. 1). Results indicate means ±s.e.m. of 5 control cells and 5 cytD-treated cells.

^*P < 0.05 vs. cytD at 0 min.

(2)

indicate that cytD reduced G_max whilst the reversal potential (V_rev), the slope (sl) and the potential of half-maximal activation (V_½) remained constant (Fig. 2, Table 3). The cytD reduction of G_max was significantly larger than the spontaneous decay of G_max during run-down (Table 3, Control).

Figure 2. Reduction of maximal Ca²⁺ conductance (G_max) by 10 μm cytD (10 min).

Pulses (300 ms) were applied from -45 mV to the potentials indicated on the abscissa. ○, peak I_Ca 5 min after rupture of the patch; •, peak I_Ca measured 10 min later. Voltage dependence of peak I_Ca was fitted according to:

A, during I_Ca run-down G_max fell from 0.91 to 0.76 nS. B, in the presence of 10 μm cytD G_max fell from 0.98 to 0.66 nS. The voltage dependence of the I–V curve, however, did not change, i.e. the reversal potential (V_rev= 52 mV), the potential of half-maximal activation (V_1/2= 1 mV) and the slope factor (sl = -5 mV) remained constant.

Table 3.

Peak I_Ca, voltage dependence

	Control			CytD
Parameter	0 min	5 min	10 min	0 min	5 min	10 min
Gmax(nS)	0.91 ± 0.07	0.83 ± 0.06	0.76 ± 0.06	0.98 ± 0.08	0.74 ± 0.07	0.66 ± 0.07*
Slope(mV)	−4.9 ± 0.2	−5.0 ± 0.2	−4.9 ± 0.4	−5.3 ± 0.2	−5.3 ± 0.2	−5.6 ± 0.2
V1/2(mV)	1.2 ± 1.6	0.3 ± 1.7	−0.4 ± 1.7	−0.3 ± 1.0	−0.7 ± 0.8	−0.8 ± 1.1
VCa(mV)	57.0 ± 1.1	55.0 ± 1.3	55.0 ± 1.4	56.0 ± 1.1	56.0 ± 1.0	56.0 ± 1.0

I_Ca data were fitted according to I_Ca=G_max(V – V_rev){1 – exp[(V – V_1/2)/sl]}^–1, yielding maximal Ca²⁺ conductance (G_max), reversal potential (V_rev), slope of the curve (sl) and the potential of half-maximal activation (V_1/2). Data from 14 control cells and 15 cytD-treated cells (means ±s.e.m.).

^*P < 0.05 vs. cytD at 0 min.

Possible pathways involved in cytD reduction of I_Ca

Phalloidin

Phalloidin is known to block the spontaneous depolymerization of F-actin. When the cells were preincubated with 100 μm phalloidin for 4-6 h, the spontaneous run-down of I_Ca was not modified, and it was indistinguishable from the decay of I_Ca following the addition of 10 μm cytD (Table 1). That is, preincubation with phalloidin blocked the cytD effect. Owing to the specificity of phalloidin in binding to F-actin, the cytD effect on I_Ca can be linked to the depolymerization of F-actin by cytD, and other non-specific cytD effects are unlikely to be important.

A kinase anchoring proteins (AKAPs)

AKAPs are thought to anchor protein kinase A (PKA) via the cytoskeleton near to Ca²⁺ channels (Zhong et al. 1999). Thus cytD depolymerization of F-actin could have reduced I_Ca by a dislocation of the PKA-AKAP complexes, with the consequent reduction of Ca²⁺ channel phosphorylation. Application of 30 nm isoproterenol for 1 min increased peak I_Ca (cascade: β-receptors, G_s-protein, adenylyl cyclase, cAMP and PKA-AKAPs, Ca²⁺ channel phosphorylation; see Hove-Madsen et al. 1996). Isoproterenol augmented peak I_Ca to 144.4 ± 27.9 % (n = 3) in the absence (control) and to 161.7 ± 40.2 % (n = 3) in the presence of cytD (myocytes preincubated with 10 μm cytD for 10 min). The stimulated PKA activity could have been limited by the availability of ATP, which was reduced due to cell dialysis with ATP-free electrode solution. When we repeated these experiments with an electrode solution containing 4 mm Na₂ATP (plus 4.5 mm MgCl₂), isoproterenol augmented I_Ca to 272 ± 23 % in control cells and to 290 ± 54 % in myocytes pretreated with cytD (no significant difference). Since cytD did not reduce the isoproterenol stimulation of I_Ca, a dislocation of PKA-AKAP complexes is unlikely to be the explanation for the cytD reduction of I_Ca.

Reduction of [Ca²⁺]_c by BAPTA

The present experiments employed electrode solutions containing only minimal EGTA (5 μm; pCa ≈7). Addition of 5 mm BAPTA and dialysis of this solution from the electrode into the cytosol increased peak I_Ca and slowed down the spontaneous run-down (Table 1). More importantly, 5 mm intracellular BAPTA blocked the reduction of I_Ca by cytD, i.e. peak I_Ca in the absence and presence of 10 μm cytD did not differ (Table 1). This result let us postulate that a Ca²⁺-dependent process is involved, with cytD causing depolymerization of F-actin and reduction of I_Ca.

Extra-/intracellular acidosis

Extracellular acidification has been reported to reduce intracellular pH and thereby the L-type I_Ca of both cardiac and vascular myocytes (Irisawa & Sato, 1986; Klöckner & Isenberg, 1994; Kiss & Korn, 1999). Superfusion with Tyrode solution of pH_o 6.9 instead of pH_o 7.4 reduced the control I_Ca at 0, 5 and 10 min. pH_o 6.9 blocked the cytD reduction of peak I_Ca (Table 1). The result indicates that the cytD reduction of I_Ca may involve a pH-dependent mechanism.

Protein phosphatases

Application of a cocktail of serine/threonine protein phosphatase inhibitors via the patch pipette reduced the run-down of I_Ca in the absence of cytD. In comparison with these experiments, there was a tendency towards a cytD-mediated reduction of I_Ca even in the presence of protein phosphatase inhibitors that did not reach statistical significance. Compared to the control run-down in the absence of phosphatase inhibitors, cytD no longer reduced I_Ca (Table 1). We conclude that cytD may have activated serine/threonine protein phosphatases, and that these protein phosphatases may be involved in the cytD-mediated reduction of I_Ca. These phosphatases are also likely to contribute to the spontaneous run-down.

Involvement of ADF/cofilin

The sensitivity of the cytD reduction of I_Ca to Ca²⁺ chelation, acidification and protein phosphatase inhibition suggests a possible involvement of the ADF/cofilin system. To test this possibility, we measured the putative cytD-mediated activation of ADF/cofilin as protein dephosphorylation within the time frame of the I_Ca measurements (10 min). A 10 min perfusion of 10 μm cytD through spontaneously beating Langendorff hearts reduced the ratio of phosphorylated ADF/cofilin to total ADF/cofilin by 50 % (Fig. 3). This result suggests that activation of ADF/cofilin by dephosphorylation may mediate the cytD-induced reduction of I_Ca.

Figure 3. Activation of ADF/cofilin by 10 μm cytD.

Langendorff hearts were perfused with or without 10 μm cytD for 10 min. Isolated proteins were separated via 12 % SDS-PAGE. Immunological detection of phosphorylated and total ADF/cofilin was performed on the same blot (A). Results of densitometric evaluation of the blots are shown in the bar diagram (B). The significant reduction in the phosphorylated form of ADF/cofilin (normalized to total ADF/cofilin) by 50 % is consistent with an activation of ADF/cofilin in response to cytD. *P < 0.05.

DISCUSSION

The present study of the possible F-actin-dependent regulation of L-type Ca²⁺ channels indicates that depolymerization of F-actin by cytD is linked to the reduced peak of I_Ca. We tentatively attribute these cytD effects to the activation of ADF/cofilin by dephosphorylation.

Our results indicate that the reduction of I_Ca, caused by 5 or 10 μm cytD, follows a faster time course than the spontaneous I_Ca run-down, and because of this difference the cytD effect could be separated from spontaneous run-down. In addition, the cytD-mediated reduction was significantly reduced by phalloidin, BAPTA, protein phosphatase inhibitors and acidic pH_o (Table 1), whilst the spontaneous run-down was not. The block of the cytD effect on I_Ca by phalloidin links it specifically to the depolymerization of the actin-based cytoskeleton because of the specific binding of phalloidin to F-actin, which confers stability (Dancker et al. 1975).

How could depolymerization of F-actin have caused the reduction of I_Ca? One possibility would be the dislocation of PKA-AKAP complexes and subsequent reduction of Ca²⁺ channel phosphorylation (Zhong et al. 1999). We consider this pathway unlikely because the PKA-mediated β-adrenergic stimulation of I_Ca by 30 nm isoproterenol (1 min) was not attenuated by cytD. The block of the cytD effect with 5 mm BAPTA led us to postulate the involvement of a Ca²⁺-dependent process. In addition, it explained the absence of a reduction of I_Ca by cytD in previous reports where the cells were dialysed with 5 or 10 mm EGTA (Undrovinas & Maltsev, 1998; Pascarel et al. 1999). Since the reduction of I_Ca by cytD depended on Ca²⁺ ions, one could speculate that cytD induces actin depolymerization via gelsolin, which is known to cap and sever F-actin in a Ca²⁺-dependent way (Yin & Stossel, 1979). However, recent experiments on I_Ca in ventricular myocytes from gelsolin -/- (knockout) mice indicated that gelsolin is not necessary for a reduction of I_Ca by cytD (Lader et al. 1999). Also, the attenuation of the cytD-induced decline of I_Ca by intracellular application of serine/threonine protein phosphatase inhibitors (present study) conflicts with the understanding that gelsolin is tyrosine phosphorylated (De Corte et al. 1999). A pH_i-sensitive step in the cytD signalling is suggested by the effective block of the cytD reduction of I_Ca by pH_o 6.9. In accordance with the literature we assume that pH_o 6.9 reduced the intracellular pH_i after a short delay (Austin & Wray, 1993), and speculatively we attribute the effect of pH_o 6.9 to intracellular acidosis (Klöckner & Isenberg, 1994; Kiss & Korn, 1999).

The observed block of the cytD effect by the different interventions, chelation of [Ca²⁺]_c, intracellular acidification and inhibition of serine/threonine protein phosphatases, is compatible with the idea that the cytD effect depends on a step mediated by ADF/cofilin (Fig. 4). ADF/cofilin belong to a family of actin-binding proteins that catalyse the rapid in vivo turnover of actin (Rosenblatt et al. 1997). The ADF/cofilin proteins are activated upon dephosphorylation on Ser³ (Agnew et al. 1995). This dephosphorylation is mediated by the Ca²⁺-dependent protein phosphatase PP2B or, in response to elevated cAMP levels, by the Ca²⁺-independent protein phosphatase PP1 (Meberg et al. 1998), properties that can explain the sensitivity of the cytD effect to BAPTA and serine/ threonine protein phosphatase inhibitors. The phosphatase inhibitor cocktail used in these experiments contained specific inhibitors for PP1 and PP2A, but not PP2B. Therefore, the observed tendency towards a cytD-mediated reduction of I_Ca in the presence of protein phosphatase inhibitors when compared to the effect of protein phosphatase inhibitors alone could be due to the lack of PP2B inhibition. Furthermore, near the membrane loading of activated ADF/cofilin with phosphatidylinositol bisphosphate (PIP₂) activity can occur, which attenuates F-actin binding of ADF/cofilin (Yonezawa et al. 1991). Therefore, hydrolysis of ADF/cofilin-bound PIP₂ by phospholipase C (PLC) is required (Goldschmidt-Clermont et al. 1991). Since PLC activity was reported to be Ca²⁺ dependent (Gusovsky et al. 1993; Ryan et al. 2000), the attenuation of the cytD effect by Ca²⁺ chelation could also be attributed to an inhibition of PLC activity. Binding of activated ADF/cofilin is competitive with binding of tropomyosin and phalloidin (Bernstein & Bamburg, 1982; Hayden et al. 1993) and results in a destabilization of actin monomer interactions, leading to the depolymerization of F-actin (McGough et al. 1997). This depolymerizing activity of ADF/cofilin is reduced by acidification (Hayden et al. 1993).

Figure 4. Putative modulation of L-type Ca²⁺ channels by ADF/cofilin.

Activation of ADF/cofilin by dephosphorylation is shown in response to cytD, which might occur via Ca²⁺-dependent (via PP2B) or -independent (via PP1) mechanisms. Binding of PIP₂ to activated ADF/cofilin is known to interfere with F-actin binding of ADF/cofilin. Hydrolysis of ADF/cofilin-bound PIP₂ requires PLC activity that is Ca²⁺-dependent. In concert with the capping activity of cytD, a net depolymerization of the subsarcolemmal F-actin by activated ADF/cofilin is suggested. Mechanical strain from the cytoskeleton, putatively applied via linker proteins (X and Y), has been suggested to modulate the Ca²⁺ channels of vascular and cardiac myocytes, hence a depolymerization of subsarcolemmal F-actin decreases the mechanical strain and thereby reduces the activity of the Ca²⁺ channel. Phalloidin, Ca²⁺ chelation, application of serine/threonine protein phosphatase inhibitors and acidification interfere with this proposed mechanism.

The postulated involvement of ADF/cofilin in the cytD-mediated decline of I_Ca requires the demonstration that application of cytD results in the dephosphorylation of ADF/cofilin along a similar time course. Indeed, a significant reduction of phosphorylated ADF/cofilin (normalized to total ADF/cofilin) was found in hearts perfused with cytD for 10 min, indicating an activation of ADF/cofilin by cytD. Thus, cytD will cause a net depolymerization of the subsarcolemmal actin cytoskeleton because the F-actin capping activity of cytD prevents F-actin polymerization in concert with the promotion of F-actin depolymerization by activated ADF/cofilin.

The mechanism of the cytD-induced dephosphorylation of ADF/cofilin remains to be elucidated. Dephosphorylation of ADF/cofilin via PP1 has been reported for an increase in [cAMP]_c and via PP2B for an increase in [Ca²⁺]_c (Meberg et al. 1998). Since the cytochalasins can elevate both messengers (Watson, 1990; Howarth et al. 1998), further studies are needed to analyse which of the phosphatases mediates the cytD-induced activation of ADF/cofilin.

Implications for a physiological role of F-actin modulation of I_Ca

In vascular and cardiac myocytes, cell shrinkage or swelling in hyper- or hyposmotic media has been reported to increase or decrease I_Ca, respectively (Langton, 1993; Matsuda et al. 1996; Xu et al. 1996). Possibly, augmentation of I_Ca by cell swelling is mediated by an increased mechanical strain from the cytoskeleton to the Ca²⁺ channel protein. If so, we would expect that volume modulation of the Ca²⁺ channel protein is attenuated when ADF/cofilin is activated by increments in Ca²⁺ (PP2B) or cAMP (PP1). Also, cardiac ischaemia may activate ADF/cofilin and reduce Ca²⁺ channel activity, in analogy to renal proximal tubule cells where ischaemia results in activated ADF/cofilin with subsequent microvillar actin alterations (Schwartz et al. 1999).