Cyclic AMP-dependent regulation of the number of [3H]batrachotoxinin benzoate binding sites on rat cardiac myocytes.

We sought to assess the effect of an increase in cAMP on sodium channels on adult rat cardiac ventricular myocytes. Sodium channels were studied with the use of the radiolabeled sodium channel-specific toxin [3H] batrachotoxinin benzoate ([3H]BTXB). Forskolin, isoproterenol, prostaglandin E1, cholera toxin, and pertussis toxin each increased cAMP levels and decreased the number of [3H]BTXB binding sites without changing the affinity of [3H]BTXB for the sodium channel. The cAMP analog 8-bromo-cyclic AMP (8-Br-cAMP) reduced the number of [3H]BTXB binding sites from 19 fmol/10(5) cells to 11 fmol/10(5) cells. [3H]BTXB binding site down-regulation was reversible, cAMP dose-dependent, and time-dependent. To test the hypothesis that the cAMP effect was mediated by cAMP-dependent phosphorylation, we determined the effect of 8-Br-cAMP on [3H]BTXB binding after preincubation of myocytes with N-(2-(methylamino)ethyl)-5-isoquinolinesulfonamide dihydrochloride (H8), a protein kinase A inhibitor. H8 inhibited 70% of the decrease in the number of [3H]BTXB binding sites induced by 8-Br-cAMP. Thus increases in intracellular cAMP in cardiac myocytes reversibly induced a decrease in the number of [3H]BTXB binding sites via cAMP-dependent protein phosphorylation, possibly of the sodium channel.

There is accumulating evidence that the cardiac sodium channel may be regulated by a cyclic adenosine 3',5'-monophosphate (CAMP)-dependent mechanism. For example, Ono and colleagues (1) have shown that isoproterenol alters the gating mechanism of the sodium channel in a CAMP-dependent, G protein-dependent fashion. In earlier work, Catterall and colleagues (2) showed that the overall numbers of skeletal muscle sodium channels are responsive to the frequency at which they are electrically active and to interventions which alter intracellular levels of calcium and CAMP. This suggests that both the overall number of sodium channels and their gating mechanism(s) might be responsive to CAMP.
We have developed a biochemical approach to this problem * This study was supported in part by grants from the Medical Research Council of Canada, Alberta Heritage Foundation for Medical Research, and Alberta Heart and Stroke Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement'' in accordance with 18  with the use of a radioligand assay for the cardiac sodium channel. The radiolabeled sodium channel-specific toxin [3H] batrachotoxinin benzoate ([3H]BTXB)' (3)(4)(5) is used to measure sodium channel numbers on freshly isolated cardiac myocytes (6, 7). The [3H]BTXB binding site has many features characteristic of the cardiac sodium channel ( 5 ) and has been shown to be associated with the receptor for Class I antiarrhythmic drugs (8).
The purposes of this work were to determine whether cAMP regulated [3H]BTXB binding to myocytes and to determine whether G proteins and protein kinase A might be involved in this regulation. This was done by assessing the effect of a variety of hormones and toxins on the binding of [3H]BTXB to freshly isolated cardiac myocytes.

MATERIALS AND METHODS
Myocyte Preparation-Cardiac ventricular myocytes were isolated from adult male Sprague-Dawley rat ventricles (200-250 g) by collagenase dispersion using the method of Kryski et al. (6). This method routinely yielded about 60 mg (dry weight) of myocytes, which corresponds to 1.2 X lo7 cells (6, 7). The cells were 85-90% viable rodshaped cells which excluded trypan blue and were tolerant of 1 mM calcium.
The rationale for the incubation and filtration conditions have previously been described (5). The conditions provide a maximal reduction in background and scatter with a minimal reduction in specific binding. The total wash time is 45 s. Initial control experiments showed that under these conditions less than 10% of the specifically bound [3H]BTXB dissociated from the complexes. Under these reaction conditions about 60-75% of total radioactivity retained on the filters is bound specifically to the [3H]BTXB binding site. Intracellular Cyclic AMP Levels-Cardiac myocytes (4 X lo5 cells/ ml) were incubated with or without agents which stimulate cAMP production for 30 min. Each tube was then treated with 5% perchloric acid. The samples were centrifuged at 12,000 X g for 10 min, and the pellet was discarded. Thereafter, 75 p1 of 5 M NaH,CO, were added to 800 pl of the supernatant fluid. After 15 min on ice the samples were centrifuged for 10 min at 12,000 X g. Cyclic AMP levels were measured by a radioimmunological method using a cAMP assay kit.

RESULTS
Increases in CAMP Levels-Agents were studied which were thought likely to increase intracellular cAMP levels by a variety of modes of action. These agents included isoproterenol (a /3-adrenoceptor agonist) (lo), prostaglandin El (lo), pertussis toxin (a Gi blocker) ( l l ) , cholera toxin (a G, activated state blocker) (12-14), forskolin (an adenylylcyclase activator) (15) and IBMX (a CAMP-dependent phosphodiesterase inhibitor) (16). Fig. 1 illustrates a typical experiment in which isoproterenol, PGE1, and forskolin increased cAMP levels in a dose-dependent fashion. All of the agents which were tested increased intracellular cAMP levels in rat cardiac myocytes (Table I). Forskolin was the most potent, increasing the cAMP level at least 10-fold. Isoproterenol, PTX, CTX, and PGE, also increased cAMP levels 6-8-fold.  Effect of CAMP on the Number of PHIBTXB Binding Sites-The Kd and the B,,, of [3H]BTXB binding were assessed during treatments which increased intracellular cAMP levels. Fig. 2 illustrates typical experiments in which treatment with forskolin ( Fig. 2 A ) , IBMX (Fig. 2B), and CTX and PTX (Fig. 2C)  These results suggested that increases in intracellular cAMP might be responsible for the reduction in the BmaX of   Rat cardiac myocytes were preincubated in the presence or absence (0) of 200 p~ forskolin for 30 min at 37 "C. The myocytes treated with forskolin were then divided into two samples. The first sample continued to be incubated with forskolin (0) for 90 min, while the second sample of pretreated cells (A) was washed extensively, then incubated without forskolin for 90 min. At that time the [3H]BTXB binding assay was performed as described under "Materials and Methods." The original control cells (which had not been exposed to forskolin) were also incubated for 90 min in the absence of forskolin. The data are expressed in Scatchard rearrangements of [3H]BTXB binding curves and the lines plotted by linear regression analysis.
Reversibility of the cAMP Effect on $H]BTXB Binding-To determine whether the reduction in sodium channel number by cAMP was reversible we assessed whether [3H]BTXB binding returned to pretreatment levels when forskolin was removed. Cardiac myocytes were first incubated for 30 min in the presence or absence of forskolin. The cell suspension treated with forskolin then was divided in two. The first sample continued to be incubated with forskolin, while the second sample was rinsed three times with control buffer, then incubated in the absence of forskolin for 90 min.
[3H] BTXB binding then was assessed by Scatchard analysis. The original control cells continued to incubate in the absence of forskolin for 90 min. Fig. 5 shows that 200 p~ forskolin reduced B, , , from 25 to 12 fm01/105 cells and that this effect was completely reversed after forskolin had been removed. Thus the diminution of [3H]BTXB by increased intracellular ** cAMP is quickly reversed as cAMP levels return to normal.
Involvement of Protein Kinase A in CAMP-Regulation of C'HIBTXB Binding-To assess whether cAMP decreased the number of ['HIBTXB binding sites by activating protein kinase A we determined whether H8, a protein kinase A inhibitor (17), blocked this CAMP-mediated increase in binding site number. Cardiac myocytes were preincubated for 60 min with 50 PM H8, then incubated with various concentrations of 8-Br-CAMP. Fig. 6 shows the relationship between ["HIBTXB binding and extracellular 8-Br-CAMP binding in the presence or absence of H8. In the absence of H8, 8-Br-cAMP decreased ['HIBTXB binding in a concentration-dependent manner. However, this reduction in sodium channel number was significantly inhibited by preincubation with H8 ( p < 0.01). Thus the inhibition of protein kinase A markedly reduced the impact of 8-Br-CAMP on ["HIBTXB, suggesting that protein phosphorylation is involved in the CAMP-dependent modulation of ['HIBTXB binding to cardiac myocytes.

DISCUSSION
The principal findings of this study are that, in freshly isolated cardiac myocytes, 1) agents which increase intracellular cAMP level by a variety of mechanisms decreased the number of ["HIBTXB binding sites; 2) the cAMP analog 8-Br-CAMP decreased binding site number; 3) the sodium channel number was closely, linearly and inversely correlated with the intracellular cAMP concentration; 4) the effects of cAMP were time-dependent and reversible; and 5 ) inhibition of protein kinase A blocked this cAMP effect.
There is an accumulating body of data which attests to the importance of neuromodulation of sodium channel activity. Both Ono et al. (1) and Schubert et al. (18) have shown that isoproterenol inhibits voltage-dependent gating of the cardiac sodium channel. These adrenergic effects were mimicked by 8-Br-CAMP and forskolin. Thus cAMP appears to be an important link between adrenergic stimulation and changes in sodium channel activity, suggesting that sympathetic neuroeffectors interact with the sodium channel via CAMP.
Our data confirm and extend the work of Ono et al. (1) and Schubert et al. (18). Data from this study show that not only does isoproterenol decrease sodium channel number, but cAMP is a critical link in this interaction.
Moreover, this effect is not limited to the adrenergic receptor but can be activated by PGE,. In addition, we have demonstrated that both the adenylylcyclase inhibitory protein GI and the adenylylcyclase activating protein G. also alter sodium channel B,,, likely through a CAMP-mediated event. Indeed, the ability of pertussis toxin to decrease sodium channel B,,, by inhibiting a Gi protein is a novel observation. Finally, we have shown for the first time that inhibition of protein kinase A blocks the CAMP-mediated decrease in sodium channel B,,,. Thus, the sodium channel may be modulated not only by the adrenergic system but also by other receptors which interact with G proteins; cAMP and protein kinase A appear to be in the final common pathway for this modulation of the cardiac sodium channel.
Previous studies have shown that the sodium channel has four phosphorylation sites (19). As well, forskolin and 8-Br-cAMP cause phosphorylation of the sodium channel (20). Thus protein kinase-mediated phosphorylation of the sodium channel may mediate the change in B,,, observed in this study and possibly mediate the decrease in sodium current seen in voltage clamp studies as well. Although a change in overall sodium channel number appears to be produced by phosphorylation of the sodium channel, it is possible that phosphorylation of the channel may alter channel conformation and thereby allosterically decrease access to radioligand binding sites. Reductions in the B,,, of radioligand binding are uncommon but acknowledged effects of allosteric inhibitors (21)(22)(23)(24)(25). In particular, lidocaine inhibits ['HIBTXB binding by decreasing ['HIBTXB B,,,, yet is an allosteric inhibitor of the sodium channel (26). Thus at this time we cannot distinguish between these two hypotheses.
In conclusion, an increase in intracellular cAMP concentration induced a significant decrease in B,,, of ['HIBTXB binding sites on rat cardiac myocytes without changing the Kd of ['HIBTXB. The CAMP-dependent effect on ['HIBTXB binding was reversible, dose-dependent, and time-dependent. Finally H8, an inhibitor of CAMP-dependent protein kinase A, inhibited the effect of cAMP on sodium channel B,,,, suggesting that channel phosphorylation is involved in the decrease in the number of sodium channel binding sites.