Modulation of the Cardiac Sodium Current by Inhalational Anesthetics in the Absence and Presence of β-Stimulation 

After approval by the institutional animal care and use committee, single ventricular cells were enzymatically isolated from adult guinea pigs weighing 200–300 g. The procedure of the cell isolation is the same as that described previously. [16]Briefly, single cardiac myocytes were obtained by retrograde perfusion of guinea pig hearts with an enzyme (collagenase; Gibco, Life Technologies, Grand Island, NY). Isolated cells were then transferred to a plexiglass chamber mounted on the stage of an inverted microscope (Olympus IMT-2, Tokyo, Japan). Only the rod-shaped cells with clear borders and striations were selected for these experiments and were used within 12 h of isolation.

High-resistance seals and voltage clamp were attained in modified Tyrode solution containing 132 mM NaCl, 4.8 mM KCl, 1.2 mM MgCl sub 2, 10 mM HEPES, 5 mM dextrose, 1 mM CaCl², pH = 7.4 with NaOH. After establishment of a whole-cell voltage clamp, the external bath solution was changed to one that isolates the Na sup + channel current. This solution contained 115 mM CsCl, 10 mM NaCl, 10 mM HEPES, 1 mM MgCl sub 2, 1.8 mM CaCl², 5.5 mM glucose, and 3 mM CoCl, pH = 7.2 with CsOH. Cobalt and cesium were used as blockers for the L-type calcium channel and potassium channel currents, respectively. The standard pipette solution contained 11 mM EGTA, 1 mM CaCl², 10 mM HEPES, 2 mM Mg-ATP, 90 mM CsF, 60 mM CsCl, 10 mM NaF, pH = 7.3 with CsOH. In experiments in which G-protein activation was blocked, GDP beta S (guanosine 5'-O-[2-thiodiphosphate], trilithium salt; 20 mM) was included in the pipette solution. In experiments in which protein kinase A activity was blocked, protein kinase inhibitor-(6–22)-amide (PKI, 10 micro Meter) was added to the standard pipette solution. The PKI used in our experiments was heat stable with K of 1.7 nM. [17]To compare halothane and isoflurane at equianesthetic concentrations, agents were prepared in the following final bath concentrations: 1.2 mM halothane and 1 mM isoflurane. These concentrations at 22 [degree sign] Celsius in Tyrode solution are equivalent to the following percentages in the gas phase (vol%)[18,19]: 1.96% for halothane and 2.60% for isoflurane. Anesthetic potencies are expressed as aqueous concentrations, which are relatively insensitive to temperature. [20]For the patch-clamp experiments, a system of glass syringes connected to a roller pump and a specially designed superfusion bath were used to allow a fast change of extracellular solution (< 1 s). Glass syringes were used to ensure a constant inhalational anesthetic concentration superfused at 1.5 ml/min. Anesthetics were mixed by adding known aliquots of concentrated anesthetics to graduated syringes with extracellular solution. One milliliter of the superfusate was collected in a metal-capped 2-ml glass vial to prevent anesthetic losses. The superfusate concentrations of the inhalational anesthetics were determined by gas chromatography (head-space analysis using flame ionization detection Perkin-Elmer, Norwalk, CT; Sigma 3B gas chromatograph) as reported previously. [21]Stock solutions of isoproterenol-hydrochloride (1 mM) and forskolin (water-soluble, 7beta-Deacetyl-7beta-[Y-N-methylpiperazino]-butyryl, diHCl; 5 micro Meter) were prepared every day and diluted in the external bath solution. GDP beta S and forskolin were obtained from Calbiochem (La Jolla, CA), and isoproterenol-hydrochloride and PKI inhibitor peptide were purchased from Sigma Chemical Company (St. Louis, MO).

Currents were recorded in the whole-cell configuration of the patch-clamp technique, as described by Hamill et al. [22]Due to the high density of Na sup + current in these ventricular cells, the following modifications were made to prevent loss of voltage control:(1) small cells (60–80 pF) were selected to decrease membrane capacitance;(2) external Na sup + concentration was reduced to 10 mM to decrease the magnitude of Na sup + current (< 2 nA);(3) pipette resistances ranged from 1.0 to 1.5 M Omega;(4) experiments were performed at room temperature (22 [degree sign] Celsius); and (5) series resistance compensation was adjusted (to approximately 80%) to give the fastest possible capacity transient without feedback instability in the voltage clamp circuit. Under these conditions, the voltage error was < 3 mV. Pipettes were pulled from borosilicate glass using a two-stage puller (Sachs-Flaming, PC-84, Sutter Instruments, Novato, CA) and heat polished (Narishige microforge, MF-83, Tokyo, Japan). Current was recorded by a List EPC-7 patch-clamp amplifier (Adams & List Associates, Great Neck, NY), and the output was lowpass filtered at 3 kHz to reduce high-frequency noise. All data were digitized and stored for later analysis with the pCLAMP software package (Axon Instruments, Foster City, CA).

After rupturing the membrane and achieving whole-cell voltage-clamp conditions, we allowed the cell to equilibrate for at least 15 min before acquiring data. When the internal perfusion of drugs was required, recordings were made after at least 25 min had elapsed to allow for diffusional exchange between the pipette and the cell. To determine peak I^Na, cells were held at a holding potential (V^H) of -80 mV, and whole cell currents were elicited by 30 ms test depolarizations from -80 mV to +30 mV in 10-mV increments. To monitor changes in the peak I^Na, 50-ms pulses were applied every 15 s from a holding potential of -80 mV to the test potential of -20 mV or -30 mV, the potential at which the peak I^Naoccurred. Recordings were made from cells when the current amplitude during control was stable for at least 5 min. To test for effects of drugs (isoproterenol or forskolin) in combination with anesthetics, drugs were applied after maximal depressant effects of anesthetics had occurred. To compare the additional reduction of I^Naby these agents, data were analyzed from the steady-state during anesthetic exposure. Thus the current obtained after maximal effect of anesthetic served as the “new” control. To monitor steady-state inactivation of I^Na, cells were subjected to 500-ms preconditioning pulse potentials between -130 (the holding potential) and +20 mV and then stepped to the test pulse potential of -20 mV (24 ms), where peak Na sup + current occurred. Peak currents evoked by the test potential were normalized to the maximum current (I sub max) obtained during the test potentials. Steady-state inactivation was fitted to a Boltzmann distribution [23]described by:Equation 1where V is the preconditioning potential, V¹/2 is the potential at which half maximum inactivation occurs, and k is the slope factor. Observed shifts in V¹/2 for the steady-state inactivation were corrected for spontaneous background shift over time as described previously. [16,24] 

Data are expressed as means +/- SEM. Statistical differences within one experimental group were determined from raw data using one-way repeated measures analysis of variance. Differences between treatment means were evaluated using the Bonferroni test. However, in some cases the tests for normality and equality of variance within groups were not satisfied. For those cases, a one-way repeated measure analysis of variance on ranks (post hoc Student-Newmann-Keuls test) was used. When different groups of anesthetics were compared, data were expressed as a percentage change and a two-way repeated measures analysis of variance was performed. Differences between group means were evaluated using the Bonferroni test. Statistical analysis was computed using commercially available software (SigmaStat, Jandel Scientific, Corte Madera, CA; SuperANOVA; Abacus Concepts, Berkley, CA). A test result was considered significant when P < 0.05.

(Figure 1) illustrates the effects of halothane or isoflurane and isoproterenol on the peak inward Na sup + current monitored over time. As shown by recordings from representative cells, I sub Na was initially depressed by halothane (1.2 mM;Figure 1(A)) or isoflurane (1 mM;Figure 1(B)), respectively. As demonstrated, efficacy of I^Nablock by isoflurane was less than that of halothane at approximately equianesthetic concentrations. For both halothane and isoflurane, the maximal effects were observed within 3 min after drug application. To determine the effect of anesthetic plus beta-adrenergic stimulation on peak I^Na, isoproterenol (1 micro Meter) was added to the superfusing solution in addition to anesthetic after the maximal anesthetic effects were achieved. To compare the additional reduction of I^Naby isoproterenol, data were analyzed from the steady state obtained during anesthetic exposure. Thus the current obtained after the maximal effect of anesthetic served as the new “control” as shown by the dotted lines in Figure 1. In the continued presence of anesthetics, isoproterenol further decreased peak I^Na. The additional reduction induced by isoproterenol was much more pronounced in combination with halothane than with isoflurane. A noticeable feature of the effects of both anesthetics on I^Nawas that the effects were not completely reversible. This was not due to a “run down” of I^Na. Under control conditions, in the absence of drugs, I^Nawas stable for at least 30 min (n = 6 cells, data not shown). This partial reversibility may be due to the anesthetics stabilizing the inactivated state of the channel. [25] 

To test whether G-protein pathways are involved in the anesthetic effects during the beta-adrenergic stimulation, we activated AC directly by forskolin, thus bypassing the G-protein pathway. The experimental protocol used was similar to that shown in Figure 1. After the maximal effect of anesthetic on I^Nareached a steady state, forskolin (5 micro Meter), in the continued presence of halothane or isoflurane, further decreased I^Na(Figure 2).

(Figure 3) summarizes the effects of isoproterenol or forskolin either alone or combined with anesthetics on I^Na. Isoproterenol (1 micro Meter) alone reduced peak I^Nain the absence of anesthetics by 14.6 +/- 1.7%(Figure 3(A)). In combination with halothane, the depressant effect of isoproterenol was approximately three times greater, reducing I^Naby 34.7 +/- 4.1%, indicating some form of synergistic interaction between halothane and isoproterenol. This reduction was measured after the anesthetic effect reached steady-state. In contrast, the Na sup + current was depressed by 12.6 +/- 4.9% by isoproterenol in combination with isoflurane. The effects of forskolin in the absence or presence of either halothane or isoflurane were rather unchanged (no significant difference;Figure 3(B)). Forskolin alone decreased I^Naby 13.8 +/- 2.4%. In contrast to the dramatic enhanced effect of isoproterenol in the presence of halothane, the effect of forskolin on I^Nawas not altered in the presence of halothane. These findings suggest that a G-protein mechanism not mediated through AC may be involved in the synergistic interaction between halothane and isoproterenol. For isoflurane, no differences in the reduction of peak I^Nawere observed with either isoproterenol or forskolin. Thus the effects of isoproterenol and forskolin, in combination with isoflurane, are not enhanced.

Further tests for the involvement of G proteins in the modulation of I^Naby anesthetics alone and during beta-adrenergic stimulation were done with GDP beta S in the intracellular (pipette) solution. GDP beta S, a nonhydrolyzable GDP analog, competitively inhibits G-protein activation by GTP and GTP analogs. [26] Figure 4(A) shows in representative cells the time course of peak I^Nain the presence of GDP beta S in the pipette during control exposure to anesthetics and isoproterenol in combination with halothane or isoflurane. In control experiments, GDP beta S at 20 mM prevented I^Nainhibition by extracellular application of isoproterenol (n = 6, data not shown). GDP beta S alone at this concentration had no effect on I sub Na over 30 min (n = 4, data not shown).

beta-adrenergic stimulation increases cytosolic cAMP. Thus phosphorylation of the channel protein by protein kinase A, as a result of increased cAMP, may be involved in anesthetic action. Therefore we tested for the involvement of the cAMP-dependent pathway in the modulation of cardiac Na sup + current by anesthetics in the absence and presence of isoproterenol. Experiments were carried out with PKI-(6–22)-amide (10 micro Meter), an inhibitor of protein kinase A, in the pipette solution. The inhibitor peptide, PKI, did not affect I^Naunder control conditions for at least 30 min (n = 4; data not shown). Figure 4(B) shows examples of the effects of halothane or isoflurane and isoproterenol on peak I^Nain the presence of PKI.

(Figure 5) summarizes the effects of halothane and isoflurane alone and in the presence of GDP beta S or PKI. Halothane alone (1.2 mM) decreased I^Naby 42.1 +/- 3.4%. With GDP beta S in the pipette, the depressant effect of halothane was significantly diminished. When G-protein activity was blocked, halothane inhibited I sub Na by 24.2 +/- 3.3% compared with control. Similarly, in the presence of PKI, halothane decreased I^Naby 23.9 +/- 2.4%. With isoflurane, the anesthetic alone decreased the peak I^Naby 24.1 +/- 3.9%. When combined with GDP beta S, the blocking effect of isoflurane was significantly diminished, decreasing I^Naby 15.0 +/- 2.2%. However, unlike the effect of halothane, the effect of isoflurane on I sub Na was similar in the absence and presence of PKI-(6–22)-amide. There was no significant difference in the sensitivity of I^Nato isoflurane with or without inhibition of protein kinase A activity.

(Figure 6) summarizes the effects of isoproterenol on I^Nain combination with either halothane or isoflurane. Steady-state obtained during exposure to anesthetics alone served as a new “control”(shown by dotted lines in Figure 1and Figure 4). In the presence of halothane, isoproterenol decreased peak I^Naby 34.7 +/- 4.1 %. With GDP beta S or PKI-(6–22)-amide in the pipette solution, the effect of isoproterenol in combination with halothane was significantly but not completely attenuated, decreasing I^Naby 5.5 +/- 3.3% and 10.1 +/- 3.6%, respectively. Extracellular application of isoproterenol alone in the presence of PKI did not decrease I^Nasignificantly (1.9 +/- 0.8%; n = 9; data not shown). Thus, even with inhibition of protein kinase A activity, the enhanced effect of isoproterenol plus halothane is still evident. Adding isoproterenol to isoflurane did not affect I sub Na in the presence of GDP beta S (0.9 +/- 3.7%). However, in the presence of PKI, the depression of I^Naby isoproterenol in combination with isoflurane is still evident, decreasing current amplitude by 9.4 +/- 3.1%.

We further investigated the effects of anesthetics and isoproterenol on the steady-state inactivation parameters of the Na channel. Spontaneous hyperpolarizing shifts in steady-state inactivation in the absence of drugs have been reported. [24,27]To distinguish drug-induced and spontaneous shifts, steady-state inactivation curves were evaluated over time under control conditions. We previously reported a rate of shift in steady-state inactivation of I^Naof -0.27 +/- 0.01 mV/min under control (drug-free) conditions. [16]In the presence of intracellular PKI, the rate of shift remained unchanged (n = 6; data not shown). However, in the presence of GDP beta S, no spontaneous shift in steady-state inactivation was observed in six cells. In both cases, steady-state inactivation was monitored after allowing for the diffusional exchange of PKI or GDP beta S into the cell (approximately 25 min). Thus spontaneous shifts were corrected for except in the presence of GDP beta S. Table 1and Table 2summarize the shifts in V¹/2 of the steady-state inactivation curve. The shift in the steady-state inactivation induced by halothane was significantly diminished in the presence of either GDP beta S or PKI. In contrast, the shift in inactivation by isoflurane was significantly less when GDP beta S was present, but not with PKI. Shifts in the steady-state inactivation for either isoproterenol or forskolin, obtained from control (without anesthetic) or measured after maximal effects of halothane or isoflurane have occurred, were not significantly different. Thus the dramatically enhanced effect of decreasing I^Naby isoproterenol in combination with halothane cannot be explained by an enhanced shift in steady-state inactivation. No statistically significant changes in the slope factor k were observed.

Modulation of the cardiac Na sup + channel intimately involves phosphorylation of the channel protein. beta-adrenergic stimulation produced by isoproterenol affects the cardiac Na channel by both membrane-delimited (direct) and cytoplasmic (indirect) pathways via the stimulatory G protein (G^s). [9,13]Although the direct pathway bypasses the second-messenger cascade, the indirect pathway involves the stimulation of cAMP mediated by protein kinase A (PKA), whose catalytic component can phosphorylate and modulate the functional activity of the Na channel. [11,28]Phosphorylation of the cardiac Na channel is also mediated by protein kinase C. [29]Although recent studies have shown that volatile anesthetic action on the cardiac Na channel may, in part, be a result of direct interaction between the anesthetic and the channel protein, [20,30]there is considerable evidence of volatile anesthetic action on signal transduction pathways. Consequently, the anesthetic effects on the cardiac Na channel also may be secondary to changes in the second messenger system.

In the present study, the depressant effects of halothane on I^Naare significantly diminished, but not completely abolished, in the presence of GDP beta S, suggesting the involvement of G proteins. Similar results in the presence of PKI, a PKA inhibitor peptide, suggest that modulation of I^Naby halothane may be via a G protein, cAMP-dependent pathway. This is further supported by findings that halothane stimulates AC activity by inhibiting the function of the inhibitory G proteins (Gⁱ). [14]Yet the exact nature of the involvement of G proteins in the modulation of I^Naby halothane is not clear. Halo-thane has been shown to produce a dose-dependent inhibition of the stimulatory action of catecholamines on AC but to have no effect on basal, nonstimulated AC activity. [31]In contrast, Anthony et al. [15]have found that halothane stimulates the basal AC activity but does not affect stimulation of AC activity produced by either forskolin or isoproterenol. [32]These discrepancies have been attributed to either a loss of some component(s) during membrane isolation or to species differences. [32]In the present study, the anesthetic agent was always applied first followed by forskolin or isoproterenol. Thus the effect of halothane was present before AC stimulation. Therefore, our findings correspond with the results of Anthony et al. [15]Because the presence of GDP beta S or PKI inhibitor fragment did not completely abolish the effect of halothane on I^Na, a pathway or mechanism beyond G proteins and PKA may also be involved.

In the presence of halothane, isoproterenol produces a disproportionately large decrease in peak inward Na sup + current. Based on the 14.6% effect of isoproterenol alone, we would predict another 8.5%(0.579 x 0.146) additional blockade from the 42.1% blockade observed in the presence of halothane alone. Instead, a 20.1% additional block was observed. An additive effect would have predicted that isoproterenol exerts the same proportional effect in the presence of anesthetics as it does in its absence. This suggests the presence of some sort of synergistic interaction between halothane and beta-adrenergic stimulation. Cardiac beta-adrenoceptors are coupled through a stimulatory guanine nucleotide binding protein (G^s) to AC. [26]When bypassing the G-protein cascade by using forskolin, which directly stimulates AC, the positive interaction between halothane and isoproterenol is abolished. This finding suggests that a G-protein-dependent but cAMP-independent pathway may be involved in the interaction between halothane and isoproterenol. The direct membrane-delimited pathway, which modulates I^Navia G^s, might be responsible for the augmented effect of isoproterenol in the presence of halothane. This hypothesis is supported by the result that GDP beta S, but not PKI, fully prevented the isoproterenol effect on I^Nain the presence of halothane. Thus, even when protein kinase A activity is blocked, the membrane-delimited pathway may be available for I^Namodulation. The mechanism by which a G-protein pathway results in the enhanced effect of isoproterenol in the presence of halothane remains unclear. A potential explanation could be either an externalization of beta-receptors, resulting in an increased number of beta-receptors on the cardiac cell surface or a facilitated beta-receptor-G^sprotein coupling. However, radioligand studies showing no effects of halothane on binding to cardiac beta-adrenoceptors and findings that halothane uncouples rather than facilitates coupling of beta-adrenoceptors exclude these possibilities. [33] 

Differences in the action of isoflurane and halothane on the cardiac I^Naprovide evidence of distinct pathways in the modulation of cardiac Na sup + current by these two anesthetics. In contrast to halothane's effect, the suppressive effect of isoflurane on I^Nain the presence of PKI in the pipette is unchanged compared with the effect of isoflurane alone. Therefore, isoflurane may not act on I^Navia a cAMP-dependent pathway. This view is supported by another study showing that isoflurane, in contrast to halothane, has no effect on basal AC activity. [15]However, the depressant effect of isoflurane on I^Nais significantly diminished in the presence of GDP beta S. Thus the results indicate that isoflurane, like halothane, acts through a G protein-dependent pathway; however, isoflurane action on I^Namay involve a cAMP-independent pathway. A possible mechanism of action is that isoflurane acts on cardiac Na channel through the direct, membrane-delimited pathway via the G^sprotein. Another possible pathway may be protein kinase C. [34] 

Another difference between halothane and isoflurane is that the isoproterenol effect is not enhanced in combination with isoflurane. However, surprisingly, when protein kinase A was blocked, the effect of isoproterenol in combination with isoflurane had a small but significant effect on I^Na. This is an unexpected result because our findings suggest a cAMP-independent pathway for the effect of isoflurane on I^Na. Yet blocking PKA activity reveals some sort of synergistic interaction between isoflurane and isoproterenol. Although this result is puzzling, it suggests that the mechanism involved in the interaction between beta-adrenergic stimulation and isoflurane may not be identical to the one involved in the absence of beta-simulation. This notion is also supported by the action of halothane, in which the positive interaction with isoproterenol may be via a cAMP-independent pathway. In contrast, in the absence of beta-adrenergic stimulation, the effect of halothane on I^Nainvolved a PKA pathway.

The effects of anesthetics and beta-adrenergic stimulation on I^Nacannot be explained by shifts in steady-state inactivation. Similar to the results of Ono et al., [28]the hyperpolarizing background shift in our study was unaffected by PKA inhibition. However, the hyperpolarizing background shift was not evident in the presence of GDP beta S. These results provide evidence that G proteins, independent of PKA activity, may be involved in the spontaneous hyperpolarizing shift of steady-state inactivation V¹/2. [24]Shifts in the presence of halothane and isoflurane, either in the absence or presence of GDP beta S or PKI inhibitor fragment, can account only in part for the effects on peak I^Na. Furthermore, the shifts in steady-state inactivation were found to be similar between either isoproterenol or forskolin alone and isoproterenol or forskolin in the presence of anesthetics. Thus the positive interaction on suppression of peak I^Nabetween halothane and beta-adrenergic stimulation cannot be explained by the shifts in steady-state inactivation. Interestingly, in the presence of GDP beta S or PKI, we observed small but significant differences between shifts in V¹/2 in the presence of isoproterenol alone and isoproterenol in combination with anesthetics. The significance of these findings is not known.

In the myocardium, volatile anesthetics differ substantially in their ability to potentiate the catecholamines-induced dysrhythmias in patients and laboratory animals, and halothane is more “sensitizing” then either isoflurane or enflurane. [8,35]The cellular mechanism(s) of this catecholamine-sensitizing effect are not completely understood. As previously reported, epinephrine or norepinephrine and halothane have synergistic negative dromotropic effects, slowing conduction of cardiac impulses to a greater extent than does isoflurane. [36,37]Our studies show some type of synergistic interaction between halothane and isoproterenol in decreasing I^Na, which may contribute to the prodysrhythmogenic effects of halothane. [8,37] 

In summary, the present study provides evidence that intracellular second messenger systems are involved in the regulation of the fast cardiac Na sup + current by volatile anesthetics in ventricular myocytes obtained from adult guinea pig hearts. Furthermore, our results suggest that different mechanisms are responsible for the modulation of I^Naby halothane and isoflurane. The effect of halothane involves the signal-transducing G protein that acts through a cytoplasmic, cAMP-dependent pathway and for isoflurane a G protein-dependent but cAMP-independent pathway. Our studies also show that the inhibition of cardiac I^Naduring beta-adrenergic stimulation by isoproterenol is more pronounced in the presence of halothane than isoflurane, indicating some form of synergistic interaction between halothane and isoproterenol.

*Weigt HU, Kwok W-M, Kampine JP, Bosnjak ZJ: Gating shifts induced by inhalational anesthetics on the cardiac sodium channel [Abstract]. Anesthesiology 1995; 83:A529.

**Murray KT, Snyders DS, Bennet PB: Isoproterenol increases the cardiac sodium current in the presence of buffered Ca sup ++. Circulation 1990; 82(Suppl III):III-522A.