INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE VON BEZOLD-JARISCH REFLEX is an inhibitory cardiovascular reflex characterized by increased parasympathetic drive to the heart and reduced sympathetic drive to the heart and blood vessels, resulting in bradycardia and hypotension (11, 18). The reflex is evoked by stimulation of cardiopulmonary receptors that are the sensory endings of unmyelinated (C fiber) vagal afferents located predominantly on the ventricular epicardium (11). Most of these cardiac receptors respond to a range of chemical substances including veratrum alkaloids, bradykinin, prostaglandins, capsaicin, serotonin [5-hydroxytryptamine (5-HT)], and phenylbiguanide (PBG) (11). Experimentally, the latter two agents are used most commonly to stimulate the reflex in vivo.

The heart produces its own hormones in myocytes, atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP), which are released into the circulation in response to physiological stimuli that increase atrial or ventricular stretch, respectively (7), and in association with pathophysiological conditions such as congestive heart failure, hypertension, and myocardial infarction (23). C-type natriuretic peptide (CNP), the other member of the natriuretic peptide family, is found in endothelial cells and the brain and appears to be regulated differently from ANP and BNP. It is considered to be a paracrine regulator of the endothelium and a neuropeptide (21). Plasma CNP levels are increased in septic shock and severe hypovolemic conditions (19), but little changed under the physiological or pathophysiological conditions that are known to be associated with elevated levels of ANP and/or BNP.

In early studies using partially purified cardiac tissue extracts, Ackermann and co-workers (1) demonstrated that atrial, but not ventricular, extracts injected into anesthetized rats had a 5-HT-like action causing bradycardia and hypotension, suggesting that an atrial factor may sensitize cardiac chemically sensitive afferents such as those involved in the von Bezold-Jarisch effect. Subsequently, Thoren et al. (31) and Schultz et al. (24) provided evidence from anesthetized animal studies that, through actions on cardiac vagal afferents, ANP inhibited reflex activation of sympathetic nerve activity normally expected with hypotension. In more recent studies, we reported that in conscious rats, ANP enhanced the sensitivity of von Bezold-Jarisch reflex bradycardia as well as cardiopulmonary baroreflex bradycardia but not arterial baroreflex sensitivity (28, 30, 35). These findings supported our unifying hypothesis (28-30) that ANP has a cardioprotective function via a selective action on cardiac vagal afferents, both chemosensory (1) and mechanosensory (24, 31). Recently, we reported (29) that BNP and CNP administered to conscious rats also enhanced cardiac vagal baroreflex sensitivity in response to rapid injections of phenylephrine, a stimulus that preferentially activates cardiac rather than arterial baroreceptors (10), providing the first evidence that all three natriuretic peptides have a similar action on cardiac vagal baroreflexes. From the original study of Ackermann et al. (1), it seemed that BNP may not act like ANP to enhance the von Bezold-Jarisch reflex, because ventricular extracts, which should have contained a high concentration of BNP, did not cause hypotension and bradycardia. Hence, the primary aim of the present study was to determine whether equimolar infusions of all three natriuretic peptides, ANP, BNP, or CNP, could modify the von Bezold-Jarisch reflex. Because all previous evidence of an interaction between ANP and the von Bezold-Jarisch reflex was provided from experiments in rats, much of it in anesthetized animals, in the present study, we established a model to activate this reflex in a conscious animal with resting heart rate (HR) closer to that of humans.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Adult Merino ewes (Howard Florey Institute, Melbourne, Australia) were housed in individual cages in an air-conditioned room (constant 21°C) with natural light and dark cycles throughout the study. The sheep were fed daily with 800 g of lucerne-oaten chaff, supplemented once a week with fresh lucerne and every 4 wk with vitamin and mineral supplements (Min-A-Vit; Welcome). Water was provided ad libitum.

Surgical preparation. At least 3 wk before experimentation, all sheep (weight range 38-48 kg) underwent surgery for oopherectomy and bilateral exteriorization of carotid arteries for subsequent ease of arterial cannulation. At least 1 day before experimentation, under local anesthesia (0.5 ml of 2% Xylocaine, Astra, Australia), two polyethylene cannulas (ID 0.58 mm, OD 0.98 mm) were introduced into the jugular vein and advanced toward the heart. One catheter was used to administer drugs close to the right heart to activate the von Bezold-Jarisch reflex. The other catheter was used for natriuretic peptide infusion. A tygon cannula (ID 1.5 mm, OD 2.08 mm) was also inserted into one of the exteriorized carotid arteries to record mean arterial blood pressure (MAP) and heart period (HP). All cannulas were secured and, when not in experimental use, were constantly infused with heparinized saline (25 IU/ml, at 3 ml/h) to maintain patency of the lines for up to 2 wk.

Hemodynamic measurements. On each experimental day, the arterial catheter was connected to a Cobe disposable pressure transducer (Lakewood, CO) to measure mean and phasic systemic arterial blood pressure (BP). HP was measured using a tachometer (Baker Medical Research Institute, Melbourne, Australia) triggered by the arterial pulse pressure signal or from the R wave of the electrocardiogram (ECG; ECG 100, Biopac Systems; Goleta, CA). ECG was measured through recording leads across the heart connected to previously implanted subcutaneous stainless steel wires. All hemodynamic variables were recorded continuously on an eight-channel Graphtec chart recorder (Linearcorder No. WR3310), and the signals were digitized and recorded at a sampling rate of 50 Hz using the AcqKnowledge data-acquisition system (Biopac Systems) connected to a Pentium computer.

Von Bezold-Jarisch reflex technique. Cardiopulmonary chemoreflex activation (von Bezold-Jarisch reflex) was elicited by bolus injections of the 5-HT₃ agonist PBG (Sigma Chemical) into the jugular vein or coronary artery in the range of 10-90 µg/kg. Doses of PBG were injected in ascending order at 10-min intervals. A minimum of five PBG doses was administered to cover the range of responsiveness from subthreshold to maximum in each animal. From these, three doses were chosen as representative "low", "medium," and "high" doses, which we have previously described for the von Bezold-Jarisch reflex in rats (28, 30). Because it is not uncommon for responsiveness to the agonist to vary between animals (17), analysis of the von Bezold-Jarisch reflex was individualized by selecting doses of PBG that elicited threshold, intermediate, and submaximal HR responses in each animal. Although doses of PBG varied according to the sensitivity of each sheep, the same three doses of PBG were always given in the absence and presence of the natriuretic peptides or saline in any given sheep. Maximum changes in HP were determined within 12-14 s after the start of each dose of PBG. Changes in MAP to PBG were recorded at the same time as the maximum bradycardic responses.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Comparison of intracoronary (1 mg in 1 ml saline; A) vs. intravenous (6 mg in 3 ml saline; B) administration of phenylbiguanide (PBG) on arterial blood pressure (BP) and heart period (HP) in 1 sheep on different days.

Experimental protocol. In each sheep included in the final study (n = 5), the effect of equimolar doses (10 pmol · kg¹ · min¹; infused at 12 ml/h) of -human-ANP (1---28) (Auspep; Melbourne, Australia), porcine-BNP-32 (Bachem; Bubendorf, Switzerland), human, porcine, rat-CNP (32---53) (Bachem), or vehicle (isotonic saline, 12 ml/h; Baxter, Australia) on von Bezold-Jarisch reflex responsiveness was determined. On the basis of studies from Charles and colleagues (4), infusions of natriuretic peptides at 10 pmol · kg¹ · min¹ into normal sheep should result in increased circulating levels of plasma ANP, BNP, and CNP in the order of ~100 pmol/l above baseline values, which is about 20 times resting levels. Experiments were performed on separate days, in randomized order, at least 3 days apart. In each experiment, baseline hemodynamic measurements were recorded for 20 min, and this was followed by five to seven doses of PBG. The time to complete all doses of PBG, with recovery times in between doses, was 45-60 min. The same doses of PBG were then repeated in the presence of an infusion of one of the natriuretic peptides or vehicle. The infusions lasted 65-80 min, incorporating a 20-min "run-in" period to allow infused natriuretic peptides to reach steady-state levels in the circulation before HR-reflex testing.

Statistical analysis. All HP measurements were converted to HR for statistical analysis. Significant effects of natriuretic peptides or saline infusions on resting BP and HR were determined by paired t-tests. For all other comparisons, data were analyzed by a two-way, repeated-measure ANOVA. To determine effect of treatment with natriuretic peptide or saline on von Bezold-Jarisch reflex, ANOVA of data from each experimental day compared all three doses of PBG before and after each treatment. Where appropriate, Bonferroni adjustment of probability (P) values was made for multiple comparisons. P values of <0.05 were regarded as statistically significant. Values quoted are means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of natriuretic peptide infusions on resting hemodynamics. Average resting MAP and HP in the sheep were 75 ± 1 mmHg and 927 ± 31 ms (HR = 68 ± 2 beats/min), respectively. There was a small, significant (P = 0.024) fall in resting BP of 9.5 ± 2.0 mmHg during BNP infusion. However, there was no significant change in resting BP in the presence of ANP (6.4 ± 1.8 mmHg, P = 0.063), CNP (3.6 ± 1.1 mmHg, P = 0.125), or saline (+3.0 ± 1.6 mmHg, P = 0.136). Associated with these changes in resting BP, there was a small but significant increase in resting HR with ANP infusion (+10 ± 3 beats/min, P = 0.028) but no change with BNP (+13 ± 10 beats/min, P = 0.246), CNP (+8 ± 3 beats/min, P = 0.062), or saline infusion (0.2 ± 1.0 beats/min, P = 0.869).

Von Bezold-Jarisch reflex in sheep. Unlike our previous studies in rats (28, 30), 5-HT was not a suitable agent to evoke the von Bezold-Jarisch reflex in sheep due to a lack of reproducibility of bradycardic responses both within and between animals. We also tried nicotine directly into the coronary circulation, and it did not cause bradycardia. In contrast, bolus injections of PBG produced rapid, reproducible, dose-related bradycardia in all the sheep included in the study (for example, see Fig. 2). For the most part, the sheep were undisturbed by PBG injections. However, at the highest doses of PBG, some animals moved during the injection or immediately after [possibly due to the stimulation of pulmonary vagal nociceptor fibers (11)], and some animals experienced syncopal symptoms when the HR fall was prolonged. In most animals, bradycardia began within 6-7 s after the start of each PBG injection, reached a maximum 1-2 s later (~8-9 beats after injection), and was sustained for a further 3-4 s (occasionally longer at the highest doses, particularly in the presence of an infusion of natriuretic peptide, as described later; Fig. 2).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Sensitizing effect of B-type natriuretic peptide (BNP; 10 pmol · kg1 · min1) on the von Bezold-Jarisch reflex in a conscious sheep. Dose-related HP and arterial BP responses measured with the same 3 doses of PBG into the jugular vein in the absence (A) and presence (B) of BNP.

Effect of natriuretic peptide infusions on the von Bezold-Jarisch reflex. Equimolar ANP, BNP, and CNP infusions enhanced the magnitude of von Bezold-Jarisch reflex bradycardic responses in sheep (Figs. 2 and 3). Figure 2 illustrates, in one sheep, the greater reflex rise in HP at each of the three doses of PBG during BNP infusion (B compared with A). Figure 3 illustrates, in another sheep, enhanced reflex bradycardia to a single ("middle") dose of PBG during infusion of each of the natriuretic peptides compared with saline control. The grouped data for the effect of three separate doses of PBG on HP are illustrated in Fig. 4A. When responses to all three doses of PBG were contrasted before and after treatment, average falls in HR to PBG were significantly enhanced in the presence of ANP by 94 ± 18% (16 ± 4 beats/min, P = 0.0003), BNP by 142 ± 55% (20 ± 4 beats/min, P = 0.0002), and CNP by 61 ± 16% (12 ± 3 beats/min, P = 0.0013). Saline infusion did not significantly change bradycardic responses to PBG (25 ± 12% or 4 ± 2 beats/min, P = 0.23; Fig. 4).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   A pronounced example in 1 sheep of the natriuretic peptides enhancing the von Bezold-Jarisch reflex. Responses shown are "middle dose" of PBG (750 µg) in the presence of saline infusion (12 ml/h) or equimolar (10 pmol · kg1 · min1) doses of atrial natriuretic peptide (ANP), BNP, and C-type natriuretic peptide (CNP).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   A: illustrates averaged dose responses to PBG, measured as maximum changes in HP before (Control, open circles) and during equimolar natriuretic peptide (10 pmol · kg1 · min1) or saline (12 ml/h) infusion (filled circles). B and C: changes in HR and BP, respectively, averaged over all 3 doses of PBG before (Control, open circles) and during (filled circles) natriuretic peptide or saline infusion. *Significant (P < 0.05) effect of treatment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The major finding of the present study was that the three structurally related natriuretic peptides, ANP, BNP, and CNP, markedly enhanced vagal cardiopulmonary chemoreflex control of HR in awake, normal sheep. We established methodology to activate the von Bezold-Jarisch cardiopulmonary reflex in these animals using PBG, a serotonin agonist selective for 5-HT₃ receptors. Sensitivities of the bradycardic responses to PBG were increased by ~60% with CNP, ~95% with ANP, and ~140% with BNP in equimolar doses. We recently reported that each of these peptides enhanced reflex bradycardia from cardiac mechanoreceptor activation (cardiac baroreflex) in rats stimulated by rapid injection of phenylephrine (29) but not on arterial baroreflex pathways (28, 30, 35, and C. J. Thomas and R. L. Woods, unpublished observations). Together, our findings provide support for a general hypothesis that ANP, BNP, and CNP act on cardiac vagal sensory afferent pathways, whether activated chemically (28, 30, and present study) or by pressure (29).

This study provides the first evidence that any of the natriuretic peptides modulates von Bezold-Jarisch-like reflex sensitivity in a nonrodent species with resting HR close to that of humans. Despite the early study from Ackermann and co-workers (1) showing that atrial but not ventricular extracts activated von Bezold-Jarisch-like hemodynamic responses, it is clear that BNP, primarily a ventricular product, is just as effective as ANP at sensitizsing bradycardic reflexes activated by PBG. The similar enhancement of von Bezold-Jarisch reflex bradycardic action by the three natriuretic peptides contrasts with the different biological actions of these hormones on other systems. For example, infusions of ANP and BNP are natriuretic and diuretic and increase mesenteric vascular resistance and hematocrit, whereas infused CNP has much weaker effects on renal function or on arterial vascular tone (33, 36).

The mechanism of action and target site for sensitization of these cardiopulmonary vagal reflexes by the natriuretic peptides have not been elucidated. In rat cardiac cells, mRNA for natriuretic peptide receptors (NP_A, NP_B, and NP_C) has been detected by RT PCR, and cGMP accumulation has been measured in response to treatment with different peptides (15). From these data, it appears that NP_A and NP_C receptors predominate in cardiac ventricular myocytes, whereas all three subtypes are present in cardiac fibroblasts (15). There is no information regarding specific expression of NP_A, NP_B, and NP_C receptors within cardiac sensory afferents, and there is no published information regarding natriuretic peptide binding or receptor mRNA expression in sheep heart. The NP_C receptor may be a possible target, because all three natriuretic peptides recognize it and with not too widely differing affinities (26). Given that the selectivity profile of the NP_A receptor is in the order of ANP  BNP > CNP (26) and for the NP_B receptor is in the order of CNP > ANP  BNP (26), it is less likely that enhanced reflex sensitization is through either the NP_A or the NP_B receptor. If the myocytes or other nonneural tissues are the target site for natriuretic peptide action, a secondary agent, such as a prostaglandin, nitric oxide, bradykinin, or serotonin, may then be released to sensitize the vagal afferents. Indeed, Deliva and Ackermann (8) have suggested that ANP acts via the release of a 5-HT₃ agonist, because bolus administration of ANP directly into the pericardium of anesthetized rats did not cause the hypotension and sympathoinhibition evoked by an intravenous bolus of the same dose of ANP. An alternative interpretation is that the lack of effect with intrapericardial administration of ANP (8) implicates a central site of action for the peptide. Whether BNP or CNP is similarly ineffective when applied directly to the epicardial surface of the heart is unknown. Nevertheless, for any of the natriuretic peptides, we cannot exclude the possibility of a site of action in the central nervous system such as one of the circumventricular organs that can be accessed by blood-borne peptides.

In contrast to previous studies using bolus administration of ANP (8) or atrial extract (1) into anesthetized rats, we found no evidence that ANP itself activates von Bezold-Jarisch-like responses. The dose of ANP that we infused into conscious sheep should elevate circulating levels by ~100 pmol/l (4), which is within the pathophysiological range achieved by endogenous ANP release (22). It was not accompanied by any significant hypotension or bradycardia. Infused BNP and CNP at the same dose also caused minimal changes in resting hemodynamics, similar to our previous findings in conscious rats (29) or dogs (36). Thus we conclude that ANP, BNP, and CNP in this dose range do not directly activate von Bezold-Jarisch reflexes, but they potentiate the effects of cardiopulmonary reflex activity evoked by other means [PBG in the present study, 5-HT in rats (28, 30)], although the site of this facilitation is yet to be determined.

An unexpected finding from our experiments was that only half the sheep tested displayed a "classical" von Bezold-Jarisch reflex to PBG; that is, dose-related bradycardia. From the 10 sheep tested, 5 were excluded because PBG caused not bradycardia, but tachycardia. This tachycardia appeared to be associated with activation of nociceptors, as evidenced by these (and only these) animals' aversive responses to PBG administration. We do not believe that those sheep lack the PBG-sensitive vagal afferent fibers that drive the von Bezold-Jarisch reflex, because anesthetized sheep invariably show a bradycardia in response to PBG (n = 10, unpublished observations). It appears, rather, that the excitatory reflex is silenced by anesthesia. Thus it seems that sheep may have a variable population of PBG-sensitive excitatory afferents that, in some conscious animals, can override PBG-induced vagal afferent activation. This is not a species-dependent phenomenon, because in the conscious dog, vagal cooling exposed veratridine-sensitive excitatory cardiac afferent activity (3). Clearly, any further interpretation of our observations in the sheep will require a dedicated study of the excitatory/nociceptive afferents.

A limitation to the interpretation of our study is that for most of the experiments, PBG was administered via the jugular vein, allowing potential activation of both pulmonary and ventricular afferents. We did not observe, however, any obvious change in the pattern of respiration, which is the hallmark of the reflex response to pulmonary afferent stimulation (2). Moreover, the bradycardic response to intracoronary administration of PBG was virtually identical to that after jugular administration, except that it occurred at a shorter latency. The 2- to 3-s time difference is consistent with the transit time of the pulmonary circulation. It was clear many years ago that the major hemodynamic responses to intravenous agents activating the von Bezold-Jarisch reflex were due to left ventricular sensory afferents (6). Any changes in rate or depth of respiration, on the other hand, were due to activation of pulmonary afferents that run in the vagi (2). Nevertheless, we cannot exclude the possibility that all three natriuretic peptides act on afferent pathways deriving from pulmonary vascular as well as left ventricular coronary chemosensitive receptors to enhance bradycardic reflexes.

In summary, in the present experiments, we described the cardiopulmonary chemoreflex (von Bezold-Jarisch reflex) in normal adult sheep for the first time. We also showed that equimolar intravenous infusions of ANP, BNP, and CNP at a dose that caused minimal changes in resting BP or HR markedly enhanced cardiac slowing in response to cardiopulmonary chemogenic receptor activation. The shared ability of these hormones to facilitate the activity of parasympathetic reflexes may indicate a common cardioprotective action of the natriuretic peptide family. Further experimentation is required to elucidate the site, mechanism of action, and receptor(s) responsible for this potentially important cardiovascular function.

Perspectives