INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE NATRIURETIC PEPTIDE family is composed of atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP). ANP and BNP are circulating peptides of primarily cardiac origin and activate natriuretic peptide receptor (NPR)-A. CNP, isolated by Sudoh et al. in 1990, is present mainly in the brain and vascular endothelium and appears to play a role in the local regulation of vascular tone (9). The diuretic and natriuretic effects of CNP are weak compared with those of ANP (10). It is actually doubtful that CNP has any endocrine function, but cardiac paracrine and autocrine actions have been proposed (10, 26). In support of these possibilities, it was found that the genes coding for the NPR-B, for which CNP is the natural ligand, are expressed in the rat heart, where NPR-B mRNA was most abundant in the atria (15). Activation of NPR-B receptors stimulates production of guanosine 3',5'-cyclic monophosphate (cGMP), which is thought to mediate the biological effects of CNP.

Boineau et al. (6) have put forth the concept of a widely distributed atrial pacemaker complex, with dominant pacemaker shifts occurring in response to sympathetic and parasympathetic nerve stimulation as well as to adrenergic and cholinergic agonists. Dominant pacemaker shifts could be related to differential nerve inputs to the atria as well as to differential receptor densities between the sinoatrial node (SAN) and other areas of the atrial pacemaker complex (3, 25). Recent evidence suggests that peptides, especially neuropeptides, play a role in heart rate regulation (4, 11, 14, 19, 22, 23). Calcitonin gene-related peptide (11), vasoactive intestinal peptide (23), and angiotensin II (ANG II) (14) produce positive chronotropic effects in various animal preparations, whereas bradykinin (22) and endothelin-1 (19) exert negative chronotropic actions. The influence of peptides on the dominant pacemaker location within the atrial pacemaker complex has not been investigated.

We have recently reported that CNP causes direct sustained positive chronotropic effects when injected into the SAN artery of canine preparations, whereas ANP does not influence heart rate (4). These effects of CNP were independent of changes in systemic arterial blood pressure and of -adrenergic and muscarinic receptors. The present study was conducted to further evaluate the effects of CNP on cardiac functions as well as to investigate the possible mechanisms mediating these effects in the dog. Our first goal was to map the dominant pacemaker location during maximal positive chronotropic responses to CNP injected into the SAN artery of in situ preparations compared with dominant pacemaker shifts occurring in response to the -adrenergic agonist isoproterenol and ANG II. Second, the presence of NPR-A and -B mRNAs was evaluated in tissue samples collected from the SAN area and from a trabecula located in a portion of the right atrium (lateral wall) not involved in pacemaker activity. Third, we studied the effects of CNP, injected into the SAN artery, on transmembrane action potentials recorded from the SAN area in multicellular right atrial preparations. Finally, the effects of CNP on atrial contractility were also evaluated in in vitro preparations using constant drive to assess inotropic responses independent of the chronotropic effects.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Atrial mapping. A median sternotomy was performed on eight adult mongrel dogs (20-30 kg), anesthetized with thiopental sodium (25 mg/kg iv) followed by -chloralose (15 mg · kg¹ · h¹ iv) and heparinized (250 IU/kg iv followed by 1,000 IU/h) for the purpose of direct atrial mapping. The SAN artery and both femoral arteries and veins were cannulated. A fluid-filled cannula (PE-50) was used to cannulate the SAN artery. These cannulations were followed by bilateral cervical vagotomy and later blockade of -receptors (propranolol, 1 mg/kg iv; Sigma, St. Louis, MO) to abolish the baroreflex. Activation sequences of both atria were obtained from 192 unipolar recording electrodes made of stainless steel (75 µm) and insulated with Teflon. These electrodes were fixed to five flexible templates (silicone plaque) designed to fit the epicardium of the right atrial free wall, the posteroinferior wall of the left atrium and coronary sinus, the posterior aspect of the left atrium, the left atrial free wall, and the interatrial band (20). The anatomic mapping grid is shown in Fig. 1. The 192 unipolar signals were simultaneously recorded along with an electrocardiogram. Isochronal activation maps were constructed from the activation times detected during a selected time window; the earliest excitation detected was used as the time reference (t = 0). The mapping system used a micro-VAX host computer (Digital Equipment) and custom-made software (Cardiomap, Institut de Génie Biomédical, École Polytechnique, Université de Montréal). Signals were amplified by programmable-gain analog amplifiers, sampled at 1,000 Hz, and converted to a 12-bit digital format.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Anatomic grid for atrial epicardial mapping in which position of 5 flexible templates used is shown along with important vascular structures. , Position of 192 unipolar electrodes. RA, right atrium; LA, left atrium; LA post, posterior aspect of left atrium (surrounded by 4 pulmonary veins); LA cs, posteroinferior wall of left atrium and coronary sinus; IAB, interatrial band; SVC, superior vena cava; IVC, inferior vena cava; SAN art, sinoatrial node artery.

Cardiac NPR-A and -B mRNAs. Total RNA (0.5 µg) was extracted from dog tissue samples located in the lung (reference), right atrium (lateral wall), and SAN (n = 2) by the acid-guanidinium thiocyanate-phenol-chloroform method (8), and the quality of the preparations was verified by gel electrophoresis. Preliminary histochemical studies (cholinesterase staining using acetylthiocholine as substrate) were performed to help identify the canine SAN region near the junction between the superior vena cava and the right atrium. Furthermore, the SAN tissue samples were taken from an area situated between the two branches of the SAN artery as it bifurcates while coursing along the atriocaval junction. The fibrous connective tissue from the epicardial surface and the superficial layer from the endocardial surface (including endocardial endothelium) were excised while atrial and nodal tissue samples were obtained. However, despite these precautions, a contribution from nonmyocytic cell types, such as endothelial cells and fibroblasts, cannot be excluded. RNA concentrations were measured by ultraviolet spectrophotometry. NPR-A and -B mRNA levels were measured by semiquantitative reverse-transcription (RT) polymerase chain reaction (PCR) assay using mutated NPR-A and -B fragments as internal standards. This method was first designed for the quantification of NPR-A and -B mRNA levels in rat organs (12). In the present experiments, the same primers and internal standards were used, since the NPR-A and -B receptors have not been cloned and sequenced in the dog. From the high degree of homology found between the sequences of human and rat NPRs, it was assumed that the sequences of rat and dog were sufficiently homologous to proceed by this way. For the quantification of NPR-B, the PCR conditions were slightly modified to obtain optimal amplification of NPR-B mRNA from dogs. The sequences of the internal standards differ from native NPR-A or -B cDNAs by only two bases, introducing a new EcoR I restriction site in the middle of each sequence. Tissue RNA and a known amount of mutated cRNA were mixed, reverse transcribed, and coamplified by PCR. PCR products were digested by EcoR I and electrophoresed on 1.5% agarose gel. The densities of the nondigested and digested amplicons were quantified by PhosphorImager (Molecular Dynamics). The levels of NPR-A and -B mRNAs in each organ were expressed as a ratio to the corresponding standard. Multiple determinations were made for each animal.

Electrophysiological study. Six adult mongrel dogs of either sex (20-30 kg) were anesthetized with thiopental sodium (25 mg/kg iv) and artificially ventilated with room air. The heart was exposed through a left thoracotomy, and the pericardium was opened. One minute after the injection of heparin (1,000 IU in 1 ml) into the left ventricle, the heart was excised and quickly immersed in cooled Tyrode solution. Right atrial preparations consisting of the anterior free wall of the atrium, the right atrial appendage, and the SAN and its artery were isolated and bathed in Tyrode solution. Preparations were perfused through the SAN artery (6 ml/min) and superfused (20 ml/min) with Tyrode solution kept at 37.5 ± 0.5°C and gassed with a 95% O₂-5% CO₂ mixture (pH 7.4 ± 0.1). The composition of the Tyrode solution used in all experiments was (in mM) 128 NaCl, 4.7 KCl, 1.2 MgSO₄, 20.1 NaHCO₃, 0.5 NaH₂PO₄, 2.3 CaCl₂, and 11.1 dextrose (all from Sigma). Atrial preparations were fixed with the endocardium facing upward.

Atrial contractility. Isolated right atrial preparations, obtained as described for the electrophysiological study, were used (n = 6). The superior part of the atrial preparation (near crista terminalis) was connected to an isometric force transducer (model FT03, Grass Instrument, Quincy, MA) by a silk thread. Two pairs of bipolar silver electrodes were brought into contact with the surface of the preparation. One pair was used to record the atrial electrogram, from which atrial rate was derived. The other pair was used to drive the preparation with suprathreshold electrical stimuli (1.5 times diastolic threshold intensity) of 2-ms duration generated by use of a stimulus isolation unit (model A385, World Precision Instrument, Sarasota, FL) synchronized with a pulse stimulator (model SD9, Grass Instrument). After determination of the spontaneous atrial rate, the preparation was paced at a rate 20% greater than basal rate to eliminate the influence of heart rate changes on contractility measurements. Isometric tension was recorded on a polygraph (model WI-641G, Nihon Kohden), and the atrial muscle was stretched to a resting tension of 2 g. In each preparation, after a 30-min period of stabilization, Tyrode solution was first injected into the SAN artery as control (0.5 ml in 1 min) followed by the injection of CNP (25 µg in 0.5 ml of Tyrode in 1 min) to observe its inotropic effect.

Statistics. Data are expressed as means ± SE. Comparison between groups in the cardiac mapping, electrophysiological, and contractility studies was made using Student's paired t-tests or one-way analysis of variance with the Bonferroni correction for multiple comparisons. The critical level of significance was set at P  0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Atrial mapping. The atrial activation sequences determined during sinus rhythm showed that, typically, the impulse originated in the area supplied by the SAN artery, as illustrated in Fig. 2, showing activation times at the recording sites carried by the right atrial electrode template. In all preparations, the impulse originating in the vicinity of the SAN activated the right atrium along the pathway of the crista terminalis as well as the left atrium along the pathway of the interatrial band; the site of latest activation was localized near the farthest edge of the left atrial free wall (70-ms isochrone, not shown, to provide greater resolution in region of pacemaker complex).

View larger version (49K): [in this window] [in a new window]   Fig. 2.   In situ right atrial mapping showing site of earliest extracellular electrical activation before and after injection (1 min) into SAN artery of drug and peptides dissolved in 1 ml of Tyrode. A: control (Tyrode); B: isoproterenol (10 µg); C: ANG II (5 µg); D: C-type natriuretic peptide (CNP, 40 µg). Time scale of activation is indicated in ms. Injection of isoproterenol and ANG II (producing an increase in heart rate of 31 and 10 beats/min, respectively) shifted site of earliest activation to superior region of SAN complex, whereas CNP (increase in heart rate of 10 beats/min) did not modify atrial activation patterns. PV, pulmonary veins.

Cardiac NPR-A and -B mRNAs. NPR-A and -B mRNA levels were determined in atrial and nodal tissue samples of two animals and compared with lung tissues known to express both types of receptors (Fig. 3). The levels of both receptors were expressed as a ratio to a known concentration of corresponding internal standard (5 × 10⁶ mutated NPR-A cRNA and 5 × 10⁷ mutated NPR-B cRNA) for an equal amount of total RNA (0.5 µg) extracted from each tissue sample. Because of potential differences in PCR amplification between dog tissue and rat standards, comparisons were made between the three tissues for the same receptor subtype but not between the two receptor subtypes. Both atrial and nodal tissues showed high expression of NPR-B receptors compared with lung (reference) tissue. For the NPR-A receptor, mRNA levels appeared higher in nodal than in atrial tissue, but these levels appeared lower than in the lung.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Natriuretic peptide receptor (NPR)-A (A) and -B (B) mRNA levels measured by semiquantitative reverse-transcription polymerase chain reaction (PCR). An aliquot of 0.5 µg of total RNA and 5 × 106 molecules of mutated NPR-A cRNA or 5 × 107 molecules of mutated NPR-B cRNA were mixed, reverse transcribed, and coamplified by PCR. PCR products were digested by EcoR I to cut standard and electrophoresed on 1.5% agarose gel. After quantification of 2 bands (endogenous RNA and mutated cRNA), levels of NPR-A and -B mRNAs in each organ were expressed as a ratio to corresponding standard.

Electrophysiological study. Transmembrane action potentials displaying the typical features of dominant cardiac pacemaker cells, with a smooth transition between phase 4 and the action potential upstroke, were recorded in all six isolated right atrial preparations. Figure 4A shows the configuration of such action potentials recorded before CNP injection into the SAN artery, and Fig. 4, B and C, depicts spontaneous action potentials recorded from the same nodal cell during and after the injection of CNP. The injection of CNP produced an increase in atrial rate (i.e., decrease in cycle length) and significantly modified some variables of the SAN action potentials as presented in Table 1. Specifically, CNP significantly increased the rate of diastolic depolarization and dV/dt_max and significantly decreased the cycle length, APD₇₅, and APD₉₀. The takeoff potential, action potential amplitude, and maximal diastolic potential remained unchanged. Whether the significant reduction in action potential duration might have been caused by the increase in rate alone and/or by a direct effect of CNP was not determined, since the action potentials were not recorded during constant drive.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Typical sinoatrial action potentials recorded, using a floating glass microelectrode, in same nodal cell before and after injection into SAN artery of CNP (25 µg in 2 min) dissolved in 1 ml of Tyrode. A: basal heart rate (86 beats/min); B: end of CNP injection (99 beats/min); C: maximal effect 30 s after end of CNP injection (103 beats/min).

Atrial contractility. The injection of CNP directly into the SAN artery of six isolated right atrial preparations, during pacing at a rate 20% greater than the spontaneous beating rate, induced a positive inotropic effect with a significant increase in contractile force from 0.9 ± 0.3 to 1.7 ± 0.6 g (Fig. 5). This inotropic effect was maximal 76 ± 13 s after the beginning of CNP injections and lasted for 326 ± 12 s before complete return to baseline values. The spontaneous rate during CNP injections never exceeded the pacing rate in any of the preparations.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Contractile force recorded, using an isometric transducer, from an isolated RA preparation paced at 120 beats/min before and after injection into SAN artery of CNP (25 µg in 1 min) dissolved in 0.5 ml of Tyrode.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Traditionally, autonomic cardiac control has been attributed solely to adrenergic and cholinergic neurotransmitters. However, there is accumulating evidence that neuropeptides within the cardiovascular system function as hormones, transmitters acting via their own receptors, neuromodulators influencing the release and/or the action of classic or other transmitters, and trophic agents (23, 29). These peptides can be localized as follows: 1) within autonomic efferents or sensory neurons innervating the heart (coexistence of peptides with classic neurotransmitters is now well established and indeed chemical coding of neurons for a particular peptide may reflect its participation in a well-defined physiological event); 2) in vascular cells [intimate contact between endothelial cells and cardiac myocytes, with no cardiomyocytes being >2-3 µm from a coronary vascular endothelial cell (26), certainly makes paracrine interactions between these two types of cells highly plausible]; and 3) in endocardial endothelial cells. It is now recognized that the endocardial endothelium regulates the performance of underlying cardiac muscle by the release of various factors (17). Among the factors released by endothelial cells, numerous endogenous substances, such as nitric oxide, calcitonin gene-related peptide, substance P, vasoactive intestinal peptide, ANG II, bradykinin, endothelin-1, prostaglandins, adenyl purines, neuropeptide Y, somatostatin, enkephalin, neurotensin, and neurokinins have been shown to modulate cardiac functions and, in particular, sinoatrial activity (4, 11, 14, 17, 19, 22, 23, 26, 29). Results reported herein as well as in a previous study (4) provide evidence that CNP, which is not currently considered as a cardiac peptide, might be involved in the modulation of cardiac functions through positive chronotropic and inotropic responses, contrasting with ANP.

The site of earliest extracellular electrical activation is known to shift in response to autonomic stimuli: with infusion of acetylcholine, it shifts mainly inferiorly, whereas with infusion of norepinephrine, it consistently shifts to the superior region of the SAN (25). The mechanisms underlying this phenomenon as well as the determinants of the location of dominant pacemaker activity have not been elucidated. However, Boineau et al. (5) reported that, in the case of -adrenergic stimulation, the distance of the shift is proportional to the intensity of the chronotropic response. Moreover, Beau et al. (3) have recently suggested that regional variations in receptor density are related to differential responsiveness to autonomic stimuli and may help to determine the spatial localization of pacemaker activity. In agreement with the results of others, we showed that isoproterenol infusion into the SAN artery shifted the dominant pacemaker from a midlevel position between the venae cavae toward the superior vena cava in all experiments. Such shifts were obtained in association with substantial chronotropic responses (33 ± 5 beats/min), which were the smallest that we could achieve in response to isoproterenol under our experimental conditions. In the same experiments, the pacemaker location remained constant, at the level of resolution afforded by our electrode array during maximal responses to CNP (11 ± 2 beats/min). The question then arises whether the absence of a shift might reflect an intrinsic property of CNP or might have been related to the relatively weak chronotropic responses to CNP. ANG II was studied at doses meant to produce responses of a similar magnitude to those to CNP. In most of the preparations, ANG II infusion produced a pacemaker displacement in the superior direction, an action associated with a positive chronotropic effect. Therefore, the fact that chronotropic responses to ANG II of a similar magnitude are associated with upward shifts of the dominant pacemaker, whereas those to CNP are not, is in agreement with the hypothesis that the actions of CNP are more localized to the dominant SAN. This hypothesis is further supported by the fact that transmembrane potentials showing the typical characteristics of pacemaker cells under basal conditions still presented such characteristics at the same location during the chronotropic response to CNP in the isolated atrial preparations. However, the earliest activation site did shift in response to CNP or ANG II injections in three of eight dogs. Two of the three dogs for which a shift occurred in response to CNP or did not occur in response to ANG II were actually the same dogs. Why these two dogs gave different results from the others is unclear but possibly they were "atypical." Therefore, the shift of the dominant pacemaker in response to substances producing positive chronotropic effect seems highly variable and may underline differences among the various substrates in their respective receptor density. Unlike what is found with isoproterenol and ANG II, when latent pacemakers appear more sensitive than the primary one, our results suggest that, in most dogs investigated in the present study, the SAN is functionally homogeneous with respect to its response to CNP or that a greater density of receptors for CNP is present in the lower SAN area, which is the dominant pacemaker in the baseline condition.

Specific cardiac binding sites for natriuretic peptides have been detected by ligand binding studies and autoradiography (7). The presence of NPR-B mRNA transcripts has been clearly shown in the monkey heart using in situ hybridization (30) as well as in rodent and human cardiac tissues using cDNA amplification with RT-PCR (18). Furthermore, Lin et al. (15), using Northern blot analysis for specific detection and quantification of all three NPR subtypes in the rat heart, have reported that NPR-B mRNA was most abundant in the atria. These authors, using a combination of cell isolation and RT-PCR, also demonstrated that NPR-B mRNA was retrieved in cardiomyocytes as well as in fibroblasts. Using RT-PCR, we have shown that NPR-A and -B mRNAs are both present in atrial and nodal tissues. If these mRNA transcripts are translated into functional receptors in the dog heart, our results suggest that natriuretic peptides exert direct effects on cardiac functions via their own receptors.

The mechanisms underlying the actions of CNP on the atrium and the SAN are still not known. In the SAN, at least three currents have been implicated in pacemaker activity. First, a decrease in the delayed rectifier potassium current seems to be predominantly responsible for the first half of the spontaneous depolarization phase of sinoatrial pacemaker cells (13). The second half is mainly caused by calcium currents, which become activated at about 60 to 50 mV. The transient T-type calcium channel is responsible for the early phase of the second half of slow diastolic depolarization, whereas long-lasting L-type calcium channel is involved in the latter phase. Third, an inward current appears to play a role especially at higher diastolic membrane potentials. The positive chronotropic responses to CNP reported herein were associated with an increase in the slope of the spontaneous diastolic depolarization, and the dV/dt_max of action potentials recorded from the SAN region using conventional floating microelectrodes was increased. Augmentation of contractile activity was also elicited in response to CNP in the right atrial preparations. Taken together, these findings suggest that CNP enhances calcium currents activity in the heart atrium, but the use of specific calcium channel blockers is needed to confirm this hypothesis.

Classically, activation of NPR-B receptors stimulates the production of cGMP. Paradoxically, an increase in cGMP is usually associated with negative inotropic or chronotropic effects (27) as well as with the inhibition of calcium channels via the activation of protein kinase G (24). On the other hand, cGMP analogs have been reported by some investigators to produce a positive inotropic effect (16). To explain the controversial effects that were noted, Fabiato (personal communication quoted in Ref. 16) suggested that cGMP has a biphasic effect: inhibition at supraphysiological levels and enhancement of contractility at more physiological levels (<1 µM). This hypothesis was recently confirmed by a study from Mohan et al. (17), demonstrating that, in papillary muscle of cat, intracellular cGMP causes a concentration-dependent biphasic contractile response. If this is the case and if CNP positive chronotropic and inotropic effects are via stimulation of NPR-B receptors and cGMP production, we can then assume that the doses of CNP used in this study were within physiological range.

Finally, CNP has been shown to act through mechanisms other than the production of cGMP. It is therefore possible that some of its biological actions are not mediated by a direct coupling to guanylate cyclase receptors. Among other alternatives, it has been reported that CNP 1) causes relaxation of isolated porcine coronary arterial rings through activation of potassium channels and membrane hyperpolarization (28), 2) causes relaxation of canine femoral veins by activation of large-conductance calcium-activated potassium channels (2), and 3) dilates the afferent renal arterioles via the prostaglandin-nitric oxide pathway (1). Thus, although stimulation of NPR-B receptors located in the SAN and right atrium might explain the effects of CNP reported in this study, since it also acts via other mechanisms, further studies, including voltage-clamp techniques, are obviously necessary to definitely elucidate this matter.

In conclusion, we have shown that CNP positively modulates pacemaker activity when injected into the canine SAN artery and increases the contractile force of the right atrium. Furthermore, if variations of the dominant pacemaker location in response to substances producing positive chronotropic effects are related to differences in receptor density, the results presented herein indicate that the primary SAN pacemaker is more sensitive to CNP, whereas subsidiary pacemakers are more sensitive to isoproterenol and ANG II. We suggest that these positive chronotropic and inotropic effects of CNP are the results of an increase in cardiac calcium channel activity and that NPR-B receptors present in right atrial and nodal tissues mediate these actions. Other studies are needed to clarify the signaling pathway and the currents involved in the responses to CNP as well as their physiological significance. In addition to the classical neurotransmitters, peptidergic transmitters are clearly present in the cardiovascular system and have been shown to exert direct and indirect actions on cardiac functions that may have both diagnostic and therapeutic implications in humans. These new concepts of cotransmission, neuromodulation, and "cross-talk" show that there is remarkable degree of interaction between the autonomic nervous system and neuropeptides at a specific target organ.