Introduction

Top Abstract Introduction Methods Results Discussion References

Vagal stimulation usually evokes negative chronotropic, dromotropic and inotropic effects on the heartbeat (reviewed in Pappano, 1995; Pappano and Mubagwa, 1992) via pre- and postjunctional mAChR. Prejunctionally, the vagal transmitter ACh suppresses norepinephrine release from sympathetic adrenergic neurons. Postjunctionally, ACh-mediated signals are transduced by the G_i into inhibition of I_Ca via reduced synthesis and action of cAMP, activation of I_K(ACh) and reduced I_f. PTX chemically inactivates G_i and suppresses vagal and muscarinic inhibition (reviewed in Pappano and Mubagwa, 1992).

Vagus nerve activation or muscarinic agonist also stimulates the heartbeat (reviewed in Pappano, 1991, 1995). Examples include the reflex-induced phase of sinus tachycardia in humans (Prystowsky and Zipes, 1985) and the positive inotropic effects in rabbit atria (Endoh and Blinks, 1984), canine (Gilmour and Zipes, 1985) and sheep (Iacono and Vassalle, 1989) Purkinje fibers and guinea pig ventricle (Korth and Kühlkamp, 1985; Yang et al., 1996). Muscarinic agonists increase contraction force in embryonic chick ventricle (Protas, 1990) and in atria from embryonic (Protas, 1992) and hatched chicks (Webb and Pappano, 1995). These results are interesting for two reasons. First, they are initiated by ACh at mAChR and are catecholamine-independent. Second, muscarinic stimulation of the heart can occur in the absence of PTX. Indeed, PTX treatment may only serve to reveal the stimulant effects more clearly because G_i-transduced inhibition is lacking (reviewed in Pappano, 1991). However, there are reports of stimulant actions of CCh in guinea pig (Gallo, et al., 1993) and rat (Sharma et al., 1996) ventricular myocytes that apparently require PTX pretreatment.

The proposed mechanism(s) for muscarinic stimulation of the heart are varied. Intracellular Na⁺ increased in guinea pig ventricle (Korth and Kühlkamp, 1985; Yang et al., 1996) and in sheep cardiac Purkinje fibers (Iacono and Vassalle, 1989) in the presence of either ACh or CCh. Increased Na entry via TTX-resistant channels was suggested for ventricular muscle (Korth and Kühlkamp, 1985) and verified in myocytes (Matsumoto and Pappano, 1989) of guinea pig; inhibition of the Na/K pump was proposed for the muscarinic agonist stimulant effect in sheep Purkinje fibers (Iacono and Vassalle, 1989). Altered Na-Ca exchange secondary to increased Na⁺ entry has experimental support that could link it to increased contractions (Korth et al., 1988). Alternatively, an increased I_Ca by CCh was advanced as an explanation of the stimulation of guinea pig ventricular myocytes treated with PTX (Gallo et al., 1993). This hypothesis is assumed to underlie the increased intracellular Ca⁺⁺ transient by CCh in rat ventricular myocytes treated with PTX (Sharma et al., 1996). Increased release of Ca⁺⁺ secondary to augmentation of InsP₃ synthesis by muscarinic agonist also is a possible explanation (Pappano, 1991). Activation of pirenzepine-sensitive (M₁) receptors reportedly initiates muscarinic stimulation of I_Ca (Gallo et al., 1993) and intracellular Ca⁺⁺ transients (Sharma et al., 1996). In contrast, the CCh-induced background Na current is initiated at M₂ mAChR (Matsumoto and Pappano, 1991) as is the positive inotropic effect in human ventricular muscle (Du et al., 1995).

The divergent results and mechanisms for muscarinic stimulation could arise from differences in preparations. Our experiments describe a guinea pig ventricular myocyte model in which muscarinic agonist-induced stimulation of cell contractions is reproducibly obtained and which replicates that found in guinea pig papillary muscle. We evaluated the hypothesis that CCh increases contractions by augmenting I_Ca (Gallo et al., 1993) and also tested a hypothesis concerning the nature of the mAChR that initiates the increased contractions and ascertains the effect of CCh on intracellular Ca⁺⁺ transients. We chose ventricular myocytes because they illustrate the importance of mAChR signaling in a tissue that is usually considered not to have direct actions either to choline esters or to vagus nerve stimulation. A preliminary account of a portion of this work has been presented as an abstract (Protas et al., 1997).

	Methods

Top Abstract Introduction Methods Results Discussion References

Cell isolation. Single ventricular myocytes were isolated enzymatically from the hearts of male and female guinea pigs (250-350 g) anesthetized with sodium pentobarbital (30 mg/kg i.p.) and anticoagulated with heparin (400 IU i.p.). The heart was retrogradely perfused with Tyrode's solution for 5 min at a rate of 8 to 10 ml/min through an aortic cannula in a Langendorff apparatus. The composition of Tyrode's solution was (in mM): NaCl, 135; KCl, 5.4; CaCl₂, 1.8; MgCl_2, 1.0; NaH₂PO₄, 0.33; HEPES, 10; glucose, 27.5; pH was adjusted to 7.4 with NaOH. Next, Ca-free Tyrode's solution was perfused for 45 to 60 s until the heart stopped beating. Subsequently, the heart was perfused with Tyrode's solution having 30-40 µM Ca and containing 36 mg collagenase (Boehringer Mannheim, Indianapolis, IN, type A) and 3.6 mg protease (Sigma, St. Louis, MO, type IV) in 50 ml. This solution was recirculated for 10 min, after which the enzymes were washed out by perfusion with 50 ml of Recovery solution. Recovery solution contained (in mM): K aspartate, 130; K₂ATP, 5; HEPES, 5; glucose, 20; pH was adjusted to 7.4 with KOH. Afterward, the ventricles were cut off, the cells dispersed in Recovery solution and kept at 4°C for at least an hour. An aliquot of cell suspension was placed in a recording chamber (500 µl volume) mounted on the stage of an inverted microscope. After 10 min, it was superfused with Tyrode's solution (2 ml/min); the glucose concentration was reduced to 10 mM for experiments. Temperature was 36 ± 0.5°C unless otherwise indicated.

Electrophysiology. In some experiments, external stimuli were applied to the myocyte from a programmable stimulator (WPI Electronics, Sarasota, FL) equipped with an isolation unit. The stimulating electrode was a glass capillary microelectrode filled with Tyrode's solution; the tip was ~50 µm. The electrode was placed within 20 to 100 µm of the myocyte; the anode was a platinum wire immersed in the bath.

Cell contraction. Contractions of single myocytes were elicited by either external or internal stimuli. With the cell shortening along its long axis, displacement of one or both ends of the cell edge is an indicator of the extent of cell contraction. A video-edge detector system (Crescent Electronics; Sandy, UT) was used to track the motion of cell edge. A microscope-magnified (400×) cell image was observed continuously on a high-resolution black-and-white TV monitor via a sequential scanning video camera attached to a sideport of the microscope. The camera position was rotated so that the video monitor raster lines were parallel with the long axis of the cell. The video dimension analyzer monitored a selected raster line for light intensity differences between the end of the myocyte and the surrounding field. The temporal resolution of this detector was 16.7 ms, and motion of as little as 0.25 µm could be detected. The signal from the detector was sent to a strip chart recorder and to a VCR for storage and off-line analysis. The extent of shortening of each cell was measured and was expressed either as micrometer of displacement or as percent cell length.

Intracellular calcium transients. A suspension of myocytes was incubated in 2 ml Tyrode's solution (36°C) containing 2 to 3 µM fura-2 AM ester for 5 to 10 min. The incubation solution was decanted and the cells washed in Tyrode's solution for up to 30 min to remove extracellular dye and to allow time for its intracellular de-esterification. The epifluorescence system (PTI Deltascan) had two monochromators that passed excitation wavelengths of 340 and 380 nm reflected onto the myocyte through a dichroic mirror (410 nm). Emitted fluorescence passed through the mirror to a 510 nm bandpass filter and then to the photomultiplier. Cytoplasmic Ca⁺⁺ was reported as the 340/380 ratio rather than the concentration of free Ca⁺⁺ because the calibration may be complicated by dye compartmentation and incomplete de-esterification. Cell contractions and the 340/380 were recorded alternately under steady-state conditions.

Statistical analysis. Data are expressed as mean ± S.E.M. Significance between means was calculated by the Student's t test (paired or independent as appropriate).

	Results

Top Abstract Introduction Methods Results Discussion References

Effects of extracellular calcium [Ca⁺⁺]_o and stimulation frequency. We used combinations of two frequencies of extracellular stimulation (1 Hz and 0.2 Hz) and two [Ca⁺⁺]_o (1.8 mM and 5.4 mM) to find the conditions needed to detect a stimulant effect of CCh in single ventricular myocytes. The CCh concentration used in these experiments, 30 µM, was very close to the EC₅₀ found for the positive inotropic effect of CCh in multicellular heart preparations.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   An example of the increased contraction amplitude produced by CCh (100 µM) in a myocyte superfused with 5.4 mM Ca++-Tyrode's solution at 36°C and stimulated at 0.2 Hz. (A) Recording of contractions at slow chart speed. The time of exposure to CCh is shown by the thin line above the trace; time calibration is 2 min. (B) Single contractions recorded at fast chart speed taken at the times indicated by a, b and c in panel A. Contraction amplitude doubled in CCh whose effect is completely reversed by washout. Calibration for shortening in B applies to both panels.

Stimulus frequency-cell shortening relation and its importance for CCh effect. At the beginning of all experiments with extracellular stimulation, we checked the influence of stimulation frequency on the amplitude of myocyte shortening. Stimuli at 0.2 Hz and 1 Hz were applied to the cell and the ratio L_0.2Hz/L_{1 Hz} was calculated where L_0.2Hz and L_{1 Hz} are the cell shortenings at 0.2 Hz and 1 Hz, respectively. In more than half the experiments, the change from higher to lower stimulation frequency was accompanied by a decrease in contraction amplitude which is typical for guinea pig ventricle. The magnitude of the decrease, however, was variable. In figure 2, the effect of 100 µM CCh is presented as function of L_0.2Hz/L_{1 Hz} ratio. The results show that there is a correlation between the ratio and the response to CCh; the greater the ratio (the smaller the decrease of amplitude with lowering of stimulation rate), the smaller the response to CCh. In 4 of 12 myocytes, the contraction amplitude was the same at 0.2 and 1 Hz stimulus frequencies; in those myocytes application of CCh was without effect (fig. 2). We considered the myocytes with a flat frequency-shortening dependence to have maximally filled SR Ca⁺⁺ stores and rejected experiments in which L_0.2Hz/ L_{1 Hz} index was >0.8. All results obtained with extracellular stimulation, including those described above (fig. 1), were selected by this criterion.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Dependence of CCh-induced increase in cell contractions on the stimulus frequency-shortening relation. (A) Decrease in amplitude of cell contraction as stimulus frequency is reduced from 1 Hz to 0.2 Hz. (B) Ordinate, response to 100 µM CCh as % of control shortening at 0.2 Hz; abscissa, ratio of shortening at 0.2 Hz to that at 1.0 Hz obtained before addition of CCh. The linear regression is fit by y = 150 (±31)x + 140 (±22) with r = 0.84.

Temperature. We carried out experiments at 23°C with other conditions at which body temperature provided positive results, namely, a stimulus frequency of 0.2 Hz and 5.4 mM [Ca⁺⁺]_o. At 23°C, control shortening (10.9% cell length) was nearly twice as great as at 36°C. We used 100 µM CCh; however, even at this concentration, the effect of CCh was much smaller than at body temperature. In two experiments, CCh did not change the cell contraction; in the remaining four, the increase in shortening did not exceed 0.5% cell length (0.55 µm). The average shortening in CCh was 11 ± 1.6% of cell length and did not significantly differ from control shortening (P = .34).

Concentration dependence and sensitivity to muscarinic antagonists. External stimulated cells were exposed to CCh which was added by a discrete or cumulative method. In discrete addition, CCh was used only once or twice (for example, 3 µM and then 300 µM, 20-30 min after the first portion was washed out) on one cell. In the experiments with discrete addition, the maximal response occurred at 100 µM CCh and amounted to a 68 ± 18.5% increase above control shortening (+2.3 ± 0.36% of cell length; n = 7); the EC₅₀ of CCh was 17.4 µM.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Influence of the muscarinic receptor antagonists pirenzepine (1 µM, ) and AF-DX 116 (30 nM, ) on the stimulant effect of CCh. The control concentration-dependence relation () is taken from the data for cumulative addition and the EC50 is 18.1 µM. From the shift in the EC50 for CCh at each antagonist concentration, apparent dissociation constants (Kd) taken from the Schild equation (see text) are 1.6 × 107 M (pirenzepine) and 9.1 × 109 M (AF-DX 116).

7

9

Action potentials. The effect of CCh on excitation-contraction coupling was examined by simultaneous recording of action potentials and contractions under current clamp conditions. The bath contained 5.4 mM [Ca⁺⁺]_o in the modified Tyrode's solution (36°C); action potentials were initiated by currents delivered through the patch electrode at a frequency of 0.2 Hz. The electrode filling solution included 50 µg/ml nystatin; electrical access was achieved by perforation of the patch rather than by rupture. An example is shown in figure 4A. The resting potential was 67 mV before, in the presence of and after washout of 100 µM CCh. There was a small decrease in APD₅₀ which was 170 ms in the absence and 149 ms in the presence of CCh. At the same time, myocyte contractions increased by 26% from 8.8% of cell length in the absence of CCh to 11.1% of cell length in its presence (fig. 4A). After washout of CCh, myocyte contractions diminished to 8.0% of cell length; APD₅₀ increased to 165 ms during this time.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   The effect of CCh (100 µM) on the transmembrane action potential and contraction in ventricular myocytes. (A) Records from one cell showing the action potentials and contractions in the control period (a), at 5 min in 100 µM CCh (b) and after washout of the drug (c). The membrane potential was recorded with the nystatin perforated patch technique; time calibration applies to the membrane potential and contraction traces. (B) Summary of experiments in this series. Ordinates: APD50 in milliseconds (left-hand panel) and myocyte shortening as % cell length (right-hand panel). The control (Ctr) and CCh data are taken from nine cells each whereas the washout (WO) data are from six cells. See text for details.

Measurements of I_Ca. Before establishing patch clamp conditions, the frequency-shortening dependence was tested in every cell (index L_{0.2 Hz}/L_{1 Hz}; see the discussion on frequency-shortening relation) by external stimulation. Only those cells in which the index was <0.8 were selected for patch clamp recording of I_Ca together with shortening. Five cells were obtained with a mean index L_{0.2 Hz}/L_{1 Hz} of 0.56 ± 0.08. Experiments were carried out at 36°C in modified Tyrode's solution containing 5.4 mM Ca⁺⁺. The holding potential was 40 mV and I_Ca was induced by depolarizing steps to 0 mV for 300 ms every 5 s. I_Ca values in control, at the end of CCh exposure, and 5 min after CCh was washed out were 3248 ± 358 pA, 3064 ± 299 pA and 2550 ± 235 pA, respectively (not shown). The value for CCh differed significantly from the value obtained after washout (P = .04), but not from the control value (P = .37). The shortening in CCh (1.7 ± 0.5% cell length) was significantly smaller than in control (3.1 ± 0.9%) and greater than that after washout (0.9 ± 0.28%; data not shown). Although these differences were significant (P = .028 and 0.025, respectively), they reflected the common tendency of shortening to run down in the course of the experiment.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   The effect of CCh (100 µM) on myocyte contractions evoked by ICa. The voltage clamp protocol held the membrane at 80 mV between tests followed by a ramp voltage (1 s) to 30 mV to inactivate INa and T-type Ca++ current. Membrane voltage was jumped to 10 mV for 100 ms to elicit ICa and then returned to 30 mV. This was repeated at a frequency of 0.2 Hz. The Tyrode's solution contained 5.4 mM Ca++; bath temperature was 36°C. (A) ICa and contractions in control (a), at 5 min in CCh (b) and after 10 min washout (c). (B) mean values for ICa (left-hand panel) and contractions (right-hand panel) in control (Ctr), CCh and washout (WO) for this series of experiments. * indicates significant difference (P < .05) from control.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Ryanodine blocks phasic ICa-dependent myocyte contraction. The voltage clamp protocol to elicit ICa and its associated contraction is the same as in figure 5. CTR panel shows ICa and its contraction in Tyrode's solution. Middle panel (10 µM Ryn) shows the effect of ryanodine at 3 min; ICa is present and contraction is blocked. The right-hand panel (+100 µM CCh) is taken at 5 min in ryanodine the indicated concentration of CCh. ICa is present but contraction remains blocked. The zero current level marked on the left-hand panel and the current calibration in the upper right-hand panel apply to all current traces taken from this single cell. Time calibration applies to current and displacement traces; the displacement calibration applies to all lower traces.

Effect of CCh on intracellular Ca⁺⁺ transients and contractions. The lack of I_Ca stimulation by CCh in otherwise untreated myocytes pointed toward altered SR Ca⁺⁺ handling and/or myofilament Ca sensitivity as mechanisms for the positive inotropic action. To evaluate the first possibility, myocytes were incubated in Fura-2 AM ester (see "Methods"). Myocyte contractions were evoked by extracellular stimuli in the same manner as described previously (see fig.1) and an increase of cell shortening recorded on increasing stimulus frequency from 0.2 to 1.0 Hz (data not shown). The intracellular Ca⁺⁺ transient was recorded as the 340/380 ratio whereas the myocyte was stimulated continuously at 0.2 Hz. The results of a typical experiment are shown in figure 7. The amplitudes of the 340/380 ratio (intracellular Ca⁺⁺ transient) and cell contraction were 0.80 and 3 µm in Tyrode's solution with 1.8 mM [Ca⁺⁺]_o. Raising [Ca⁺⁺]_o to 5.4 mM was accompanied by increases of the 340/380 ratio and cell shortening to 0.86 and 4 µm, respectively. Addition of 100 µM CCh in the presence of 5.4 mM [Ca⁺⁺]_o increased the 340/380 ratio by 13% to 0.97 and cell contraction by 42% to 5.7 µm, respectively. Washout of CCh restored both variables to values seen initially in 5.4 mM [Ca⁺⁺]_o(fig. 7). The results of nine experiments showed that raising [Ca⁺⁺]_o from 1.8 to 5.4 mM increased the 340/380 ratio by 22 ± 4% (P = .002) and cell contraction by 130 ± 34% (P = .001). In 5.4 mM [Ca⁺⁺]_o, CCh further increased the 340/380 ratio by 13 ± 2.5% (P = .003) and myocyte shortening by 33 ± 4.8% (P = .001). Carbachol increased the ratio, signaling an increased intracellular Ca⁺⁺ transient, and cell shortening in each of the nine myocytes.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Increase of intracellular Ca++ transient and cell contraction by 100 µM CCh. The upper row shows individual contractions evoked by extracellular stimuli (0.2 Hz) in Tyrode's solution containing 1.8 mM Ca++ (CTR), at 8 min after raising [Ca++]o to 5.4 mM, at 3 min after adding 100 µM CCh on top of the 5.4 mM Ca-Tyrode's solution and 6 min after washout (W.O.) of CCh when the cell was superfused again with 5.4 mM Ca-Tyrode's solution. The time and displacement calibrations apply to all contraction records. The lower row shows Ca++ transients recorded from the same myocyte with a 340/380 ratio of Fura-2 fluorescence as the index of intracellular Ca++. Pairs of intracellular Ca++ transients are shown in control (CTR), at 7 min in 5.4 mM Ca-Tyrode's solution, at 2.5 min in 5.4 mM Ca-Tyrode's with 100 µM CCh and at 5.5 min after washout (W.O.). The cell contractions and intracellular Ca transients were recorded alternately.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Carbachol increased the extent of myocyte shortening significantly when the experimental conditions replicated those in the original description of the phenomenon in multicellular ventricular preparations of reserpine-treated guinea pigs (Korth and Kühlkamp, 1985). The half-maximal CCh concentration (EC₅₀) of 18 µM in our experiments is close to the value of 32 µM in guinea pig ventricle (Korth and Kühlkamp, 1987). One additional qualification is the occurrence of a positive staircase when increasing stimulus frequency. With this proviso, acutely dissociated single myocytes from guinea pig ventricle are a reliable system to evaluate the mechanism(s) for the stimulant effect of choline ester muscarinic agonists on ventricular myocyte contractions.

Pharmacological experiments (fig. 3) indicate that the mAChR initiating the contraction effect of CCh is indistinguishable with the presently available data from the M₂ subtype because AF-DX 116 (M₂-selective) is a more potent antagonist than pirenzepine (M₁-selective). The apparent K_d for pirenzepine (1.6 × 10⁷ M) is ~18 times greater than that of AF-DX 116 (9.1 × 10⁹ M). The CCh-induced background Na⁺ current, thought to underlie muscarinic stimulation of ventricular contractions, is also characterized as M₂ because AF-DX 116 was ~6 times more potent than pirenzepine as an antagonist (Matsumoto and Pappano, 1991). In human ventricle, ACh evoked a positive inotropic effect attributed to M₂ mAChR (Du et al., 1995). Expression of M₂ mRNA in Xenopus oocytes yielded mAChR with greater affinity for AF-DX 116 than for pirenzepine (Fukuda et al., 1987). The possibility that stimulation of ventricular myocyte contractions by CCh might also involve M₁ mAChR must also be considered. The detection of M₁ mRNA in guinea pig (Gallo et al., 1993) and rat ventricular myocytes (Sharma et al., 1996) has been linked to stimulant effects of CCh (see below). Although we favor the view that the M₂ mAChR more likely initiates the stimulant effect of CCh, we cannot exclude the possibility of some M₁ mAChR participation. Experiments done with selective agonists would help solve this matter.

Carbachol did not change the resting potential when it increased cell shortening; the APD sometimes decreased significantly (see also Arreola et al., 1994; Yang et al., 1996). The latter effect was observed when the perforated patch technique was used to minimize diffusional loss of intracellular constituents. Muscarinic agonists activate K_(ACh) channels in ventricular myocytes (McMorn et al., 1993); however, some membrane hyperpolarization >1 mV (the limit of detection) would be expected because the average membrane potential was 16 mV less negative than the K⁺ equilibrium potential (88 mV). An increase of outward I_Na-Ca at plateau potentials (Protas et al., 1997) could explain the reduced APD by CCh detected by us in single myocytes and by others in guinea pig papillary muscle (Arreola et al., 1994; Yang et al., 1996). Alternatively, CCh might induce a Ca⁺⁺-activated Cl current (I_Cl(Ca)) and reduce APD. When oocytes were transfected with mRNA for M₂, ACh induced a monovalent cation current (Na⁺,K⁺), whereas in cells transfected with M₁ mRNA, ACh induced I_Cl(Ca) (Fukuda, et al., 1987). We favor the participation of I_Na-Ca because CCh acts through M₂ mAChR to increase cell contractions. If CCh also activated I_Cl(Ca) when it augmented contractions, this agonist would be interacting simultaneously with both M₁ and M₂ mAChR.

The L-type Ca⁺⁺ current is an important trigger for the release of Ca⁺⁺ from the SR (McDonald et al., 1994). In CCh, I_Ca was not significantly greater, yet myocyte shortening increased (fig. 5). This result confirms the lack of effect of CCh or ACh on the Ca-dependent action potential in K⁺-depolarized guinea pig ventricle (Korth and Kühlkamp, 1985) and in rabbit ventricle (Inui et al., 1982) during the positive inotropic effect of these choline esters. It is unlikely that the signal transduction pathway underlying the increased cell shortening by CCh in guinea pig ventricle involved an increase of I_Ca. The mechanism for the stimulant effect of CCh on guinea pig ventricular myocyte contractions must reside in other steps of E-C coupling.

Intracellular Ca⁺⁺ transients increased in the presence of CCh. This effect paralleled the stimulation of cell contractions by CCh and was reversible (fig. 7). This novel action for choline ester muscarinic agonists agrees with their positive inotropic effect (Korth and Kühlkamp, 1985; Yang et al., 1996). Ryanodine (10 µM) suppressed the fast component of the isotonic contraction (fig. 6) and CCh did not restore this contraction phase in the presence of ryanodine. Our results are consistent with the hypothesis that the increased contractions by CCh depend on SR Ca release. In isolated guinea pig ventricular myocytes, ryanodine (up to 10 µM) selectively suppressed the fast component of isotonic contractions, an indication that the contraction depended on Ca release from the SR (Shepherd et al., 1990). In unstimulated rat ventricular myocytes, CCh increased intracellular Ca⁺⁺ by a mechanism involving Na-Ca exchange but independent of SR Ca release, because it was not opposed by 1 µM ryanodine which blocked caffeine-induced Ca release (Korth et al., 1988). In electrically paced myocytes, we find that the size of the intracellular Ca⁺⁺ transient increases with CCh at constant stimulation rate. Previous attempts to record intracellular Ca⁺⁺ transients provided hints that these could be augmented by ACh. In rabbit atrial muscle, vagal stimulation produced an atropine-sensitive rebound positive inotropic effect accompanied by an increase of the aequorin light signal (Endoh and Blinks, 1984). In rabbit SA node, intracellular Ca⁺⁺ transients detected with Indo-1 were modestly increased when ACh reduced spontaneous beat frequency (Lee, et al., 1987).

Carbachol, in the presence of PTX, reportedly acts through M₁ mAChR to increase the amplitude of intracellular Ca⁺⁺ transients in paced rat ventricular myocytes (Sharma et al., 1996). This effect, attributed to calcium-induced calcium release in rat ventricular myocytes, was assumed to arise from a larger I_Ca because CCh increased I_Ca in guinea pig ventricular myocytes treated with PTX to suppress muscarinic inhibition of this current (Gallo et al., 1993). The stimulant effect of CCh on I_Ca was prevented by pirenzepine (Gallo et al., 1993) and attributed to activation of M₁ receptors whose mRNA was detected in myocytes (Gallo et al., 1993; Sharma et al., 1996). Although our results in guinea pig ventricular myocytes also report an increased Ca⁺⁺ release transient, the effect is initiated at M₂ mAChR in the absence of PTX and occurs without an increased I_Ca. On the one hand, the divergent results may be caused by the different conditions used. On the other hand, our results for stimulation of contractions replicate those found in multicellular preparations of guinea pig ventricle where the importance of intracellular Na⁺ for regulation of contractions has been emphasized (Korth and Kuhlkamp, 1985; Matsumoto and Pappano, 1989).

The similarity of the frequency-shortening relation in single myocytes (fig. 2) to the frequency-force relation in multicellular ventricular preparations from guinea pigs (Gibbons and Zygmunt, 1992) is an important qualification for the reliability of the isolated myocyte preparation. It may also provide a clue for the mechanism of CCh action. For example, increasing the frequency of voltage clamp pulses (80 to 0 mV for 100 ms) from 0.5 to 3 Hz augmented the extent of shortening of guinea pig ventricular myocytes (Harrison and Boyett, 1995). The increased contractions during the positive staircase are initiated by elevation of intracellular Na⁺ which, through the Na-Ca exchange mechanism, increases the Ca content of the SR. This allows increased contractions in the presence of an unchanged or reduced I_Ca. Carbachol elevates intracellular Na⁺ (Korth and Kühlkamp, 1985; Yang et al., 1996) by increasing Na⁺ entry via TTX-resistant channels (Matsumoto and Pappano, 1989, 1991). Therefore, if CCh acted like increased frequency to raise intracellular Na⁺, the effects of CCh and stimulus frequency on intracellular Na⁺ would be additive. A similar view applies to sheep Purkinje fibers, although the mechanism of muscarinic action underlying the increased intracellular Na⁺ is inhibition of the Na/K pump (Iacono and Vassalle, 1989).

At room temperature Na-Ca exchange contributes to SR Ca loading but does not trigger SR Ca release by action potentials (Sipido et al., 1995). That Na-Ca exchange might trigger SR Ca release was first proposed by Leblanc and Hume (1990) and was supported by the observation that contractions could be triggered by Na-Ca exchange at 35°C (Levi et al., 1994; Vornanen et al., 1994; Wasserstrom and Vites, 1996) but not at 22°C (Vornanen et al., 1994). Our experiments were not designed to evaluate the loading versus triggering functions of Na-Ca exchange. In preliminary studies, we find that CCh increased the magnitude of caffeine-induced I_Na-Ca, an indication that SR Ca⁺⁺ stores are increased by CCh (Protas et al., 1997). Another bit of evidence for the Na-Ca exchanger as a link between increased intracellular Na⁺ and increased contraction during the positive staircase is voltage dependence (see above). The positive staircase was absent when voltage jumps to 0 mV began from a holding potential of 40 mV because the cells did not gain intracellular Na⁺ under this condition (Harrison and Boyett, 1995). We find that CCh has no effect on contractions when the membrane holding potential is kept continuously at 40 mV between voltage jumps to +10 mV.

An increase of myofilament sensitivity, if it occurred, would provide an additional mechanism for muscarinic agonist stimulation of cell contractions. Our experiments did not address this issue. The proportionately greater increase of contractions than of Ca_i⁺⁺ transients by CCh might indicate that myofilament Ca⁺⁺ sensitivity increased. However, complexities (compartmentation, incomplete de-esterification, nonlinear relation between fluorescence ratio and Ca_i⁺⁺) arising from the use of Fura 2 ester preclude any conclusions regarding this possibility. Carbachol reportedly increased myofilament Ca⁺⁺ sensitivity in guinea pig papillary muscle (Yang et al., 1996). The mechanism for this effect is not known but the increase of [Na⁺]_i is PKC-independent.

Divergent results have been reported for regulation of myofilament Ca⁺⁺ sensitivity by cGMP and PKC, each of which has been identified as an intracellular messenger for muscarinic agents (reviewed in Pappano, 1995). Myofilament Ca⁺⁺ sensitivity diminished in the presence of 8-bromo-cGMP whereas the intracellular Ca⁺⁺ transient was unchanged (Shah et al., 1994). In acutely dissociated guinea pig ventricular myocytes, CCh (10 µM) increased cGMP by 100% and cell shortening by 18%. Methylene blue suppressed the cGMP effect but not the increased contractions caused by CCh (Stein et al., 1993). This pattern of results also militates against a role of cGMP in the positive inotropic action of CCh under the conditions of our experiments.

Activation of PKC by phorbol ester elicited an increased myofilament Ca⁺⁺ sensitivity in permeabilized rat ventricular myocytes (reviewed in Terzic et al., 1993). However, phorbol esters have negative inotropic effects in many cardiac tissues and also reduce intracellular Ca⁺⁺ transients (Terzic et al., 1993). In chick atrial muscle (Webb and Pappano, 1995) and in guinea pig papillary muscle (Arreola et al., 1994), phorbol ester had a small negative inotropic effect and suppressed the positive inotropic action of CCh. In rat ventricular myocytes, phorbol ester stimulated Na-H exchange by activation of PKC (Korth et al., 1988). However, CCh did not mimic this action of phorbol ester, which indicates that PKC activation was not a component of the reaction mechanism whereby CCh stimulated ventricular contractions. Experiments with staurosporine confirmed that CCh did not regulate Na-H exchange (Yang et al., 1996). Carbachol also increases InsP₃ in cardiac muscle (reviewed in McDonald et al., 1994; Pappano, 1991). Although this intracellular messenger can release Ca⁺⁺ from the SR, it has no effect on myofilament Ca sensitivity (Nosek et al., 1986; Vites and Pappano, 1990).

Our results provide evidence for the suitability of a single myocyte model that replicates the features of cells in guinea pig papillary muscle. The increased contractions by CCh are apparently initiated at M₂ mAChR, the same mAChR subtype that mediates muscarinic inhibition. The increase of Ca⁺⁺ release transients and of contractions by CCh without a concomitant increase of I_Ca indicates that the effect of CCh on contractions resides in other steps connected to E-C coupling, perhaps the Na-Ca exchanger and myofilament Ca⁺⁺ sensitivity.