| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Dipartimento di Medicina Clinica, Prevenzione e Biotecnologie Sanitarie, and 3Divisione di Riabilitazione Cardiologica, Ospedale San Gerardo, Monza; and 2Centro Interuniversitario di Fisiologia Clinica e Ipertensione, Università di Milano-Bicocca, and 4Centro di Bioingegneria, Fondazione Don Gnocchi, Milan, Italy
Submitted 21 July 2005 ; accepted in final form 15 August 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-, but not 1- or 
-, adrenergic receptor blockade and by Ba2+-sensitive K+ channel or L-type Ca2+ channel, but not by other ion channel, blockers. We conclude that, in the conscious, unrestrained ganglion-blocked rat, catecholamine infusion generates VLF fluctuations in BP through stimulation of 2-receptors and activation of Ba2+-sensitive K+ channels. These fluctuations may have (patho)physiological relevance under conditions of disrupted circulatory homeostasis.
catecholamines; spectral analysis; ganglionic blockade; adrenergic receptors
With regard to the possible adrenergic contribution to these phenomena, our attention was drawn by the markedly enhanced amplitude of VLF oscillations in blood pressure in rats subjected to chronic chemosympathectomy by 6-hydroxydopamine, a model extensively used in our laboratory (9, 10, 14). We considered that in these animals there is an 80–90% suppression of neural cardiovascular sympathetic influences, but there is at the same time a clear-cut increase in circulating epinephrine (9), which is compatible with the notion that adrenergic influences may contribute significantly to the origin of VLF oscillations in blood pressure and that these oscillations are preferentially promoted by circulating, rather than neurally released, catecholamines.
We therefore designed the present study to test the hypotheses that 1) VLF oscillations in systemic arterial pressure disappear in rats in which all adrenergic influences are suppressed and 2) intravenous infusion of catecholamines can restore or even magnify VLF oscillations in blood pressure under these conditions.
Adrenergic influences were suppressed by acute ganglionic blockade, rather than chronic sympathectomy. To avoid possible confounding effects of anesthesia on modulation of vascular tone, all experiments were performed in conscious, unanesthetized, free-moving animals.
Because the data showed circulating catecholamine-dependent VLF fluctuations in blood pressure, additional goals of the study were to determine which adrenergic receptor(s) and which membrane ion channel(s) are involved.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
24 h were allowed for the animal to recover and acclimate to the experimental environment, which consisted of a wide cage in which the rat could walk, explore, eat, and drink ad libitum. Experimental protocol. All experimental sessions were carried out during daytime in the unanesthetized condition. The arterial catheter was connected to a pressure transducer (model P23 Dc, Gould-Statham, Oxnard, CA). The blood pressure signal was continuously displayed on a chart recorder (7D polygraph, Grass Instruments, Quincy, MA). Heart rate was visualized from the pulsatile pressor signal via tachographic beat-to-beat conversion. The pulsatile blood pressure signal was simultaneously tape recorded for off line computer acquisition and analysis.
Baseline and withdrawal of adrenergic influences.
To assess the effect of withdrawal of adrenergic influences, blood pressure was continuously recorded for two 15-min periods in 12 animals: 1) in the absence of pharmacological intervention to obtain hemodynamic data in the intact condition and 2) after induction of ganglionic blockade by hexamethonium (30 mg/kg bolus followed by continuous infusion at 1.5 mg·kg–1·min–1). The infusion was continued throughout the experimental protocols involving administration of additional drugs (see below), i.e., for 
30 min.
Blood pressure restoration with "adrenergic" and "nonadrenergic" vasoconstrictor agents.
In a first series of experiments performed on another group of 23 ganglion-blocked rats, blood pressure was restored by infusion of adrenergic and nonadrenergic agents in addition to hexamethonium as follows: epinephrine (10 µg·kg–1·min–1, 8 rats), norepinephrine (1 µg·kg–1·min–1, 4 rats), and arginine vasopressin, a nonadrenergic vasoconstrictor (0.5 ng·kg–1·min–1, 6 rats). Clonidine (40 µg·kg–1·min–1, 5 rats) was infused to verify the ability of selective peripheral 2-adrenergic stimulation to originate slow blood pressure oscillations.
Infusion of the above-mentioned drugs with hexamethonium was maintained for 12–15 min; although the presence or absence of blood pressure oscillations invariably became evident within 1 or 2 min of drug infusion, the observation period was prolonged to a duration feasible for computer analysis and spectral power extraction over stationary segments of the blood pressure recording.
Blockade of adrenergic receptors, angiotensin receptors, and nitric oxide synthesis.
In a second series of experiments performed in another group of 29 rats, adrenergic receptor antagonists were injected with hexamethonium plus epinephrine (starting a few minutes after appearance of blood pressure oscillation) to identify the adrenergic receptor subtypes involved in the genesis of blood pressure oscillations. In addition, the possible involvement of the renin-angiotensin system and nitric oxide in catecholamine-dependent oscillations was examined by intravenous injection of a bolus of the appropriate antagonist. The scheme of this second series of experiments was as follows: infusion of 1) propranolol (1 mg/kg), a -adrenergic receptor blocker (10 rats); 2) prazosin (0.1 mg/kg), an 1-adrenergic receptor blocker (5 rats); 3) yohimbine (10 mg/kg), an 2-adrenergic receptor blocker (5 rats); 4) losartan (30 mg/kg, 4 rats); and 5) nitro-L-arginine methyl ester (L-NAME, 100 mg/kg, 5 rats). The observation period (drug of interest in addition to hexamethonium + epinephrine) was maintained for 12–15 min.
Ion channel blockade.
In a third series of experiments aimed at identifying the ion channels involved in the genesis of VLF oscillations, the following specific blockers were added in addition to hexamethonium plus epinephrine (17 rats): 1) nifedipine (1 mg/kg iv bolus), a voltage-dependent Ca2+ entry blocker (3 rats); 2) glibenclamide (3.5 mg/kg iv bolus), an ATP- dependent K+ channel blocker (4 rats); 3) iberiotoxin (50 g/kg iv bolus followed by 1 g·kg–1·min–1 infusion), a Ca2+-activated K+ channel blocker (3 rats); 4) 4-aminopyridine (3 mg·kg–1·min–1), a voltage-dependent K+ channel blocker (3 rats); and 5) BaSO4 (10 mg/kg iv bolus), a Ba2+-sensitive K+ channel blocker (4 rats). The duration of the observation period was 12–15 min, except in a few animals (especially those receiving iberiotoxin or BaSO4) in which continuous drug administration was associated with phenomena such as agitation and muscle twitches; it was, at any rate, possible to obtain in each animal 10 min of blood pressure recordings in the absence of any apparent behavioral disturbances.
Effectiveness of channel blockade was demonstrated by the disappearance of the VLF oscillations in blood pressure (nifedipine and BaSO4) or by the induction of significant systemic vasoconstriction, despite lack of abolition of VLF oscillations (glibenclamide, iberiotoxin, and 4-aminopyridine).
At the end of the blood pressure recording period, blood was drawn from the arterial catheter to assay circulating catecholamines (HPLC).
Signal analysis. The tape-recorded arterial pressure signal was sampled at 250 Hz and digitized at 12 bits. Systolic and diastolic blood pressure (SBP and DBP) values were identified beat by beat. Pulse interval (PI), the reciprocal of heart rate, was computed as the interval between two consecutive systolic peaks. To improve the time resolution of the PI estimate, the highest blood pressure sample during systole and its two neighboring samples were interpolated by a parabola, the apex of which was taken as the true systolic peak. For each experimental condition, we visually selected a steady-state subperiod from the blood pressure recording for spectral analysis. The selected subperiods were free from artifacts or nonstationarities after bolus injections or the start of drug infusions. Mean values of SBP, DBP, and PI were computed in each period.
Beat-by-beat series of blood pressure and PI data are sampled unevenly: sampling frequency was bound to the instantaneous heart rate. The irregular sampling frequency may cause distortions in the power spectra, which can be avoided by interpolation and resampling at constant frequency (38). To apply spectral analysis on evenly sampled time series, we followed a procedure previously used in humans (6) and animals (11) that consists of 1) linearly interpolating the SBP, DBP, and PI beat-by-beat values; 2) low-pass filtering the interpolated series with a cutoff frequency at 3 Hz, which is sufficiently high to preserve all the fluctuations of spontaneous beat-by-beat variabilities; and 3) resampling the series at the constant frequency of 10 Hz, i.e., well above the Nyquist rate. The effects of hexamethonium on spectral powers were assessed by computing a fast Fourier-transformed (FFT) spectrum over the identified steady-state subperiods of the intact and hexamethonium recordings and by integrating each spectrum over the VLF (0.01–0.15 Hz, oscillations between 7 and 100 s), LF (0.15–0.8 Hz, oscillations between 7 and 1.25 s), and HF (0.8–2.5 Hz, oscillations between 1.25 and 0.4 s) bands (38).
The integral of the FFT power spectrum over a given frequency band quantifies the spectral power, independently of the structure of the power spectrum over the band, i.e., whether it is white or 1/f noise, rather than a periodic oscillation. Therefore, to quantify the power contribution of an oscillation possibly present in the VLF band, we also applied a parametric spectral estimation (3). For each resampled time series, an autoregressive power spectrum was calculated over the previously selected steady-state subperiods by means of the Burg algorithm (23). This method models each spectral peak by a couple of complex conjugate poles, the total number of poles depending on the order of the autoregressive model. If the model order is correctly selected, it is possible to assess whether at least one spectral peak is present in a given frequency band and to calculate the power of this peak. We selected the autoregressive model order on the basis of the Akaike information criterion. The power of each spectral peak was computed as the residue associated with each couple of poles of the autoregressive model (18, 23). The power of the VLF oscillation was estimated as the power of the peak, with the highest power among those falling within the VLF band. If no poles fell in this band, it was assumed that no VLF oscillations were present, and a power equal to zero was estimated.
Statistical analysis. Individual SBP, DBP, and PI values and plasma catecholamine concentrations were averaged in each experimental group of animals. Differences between groups were assessed by paired or unpaired Student’s t-test, as appropriate.
Because spectral powers do not follow a normal distribution, groups were statistically compared by nonparametric methods. Medians and first and third quartiles were estimated in each group of animals and compared by the Mann-Whitney U (independent samples) test or Wilcoxon’s matched pairs (dependent samples) test (37) with STATISTICA 6.0 software (StatSoft, Tulsa, OK). The level of statistical significance was set at P = 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
Blockade of adrenergic receptors, angiotensin receptors, and nitric oxide synthesis. The effects of antagonists of adrenergic and angiotensin receptors and nitric oxide synthesis on average blood pressure and PI values are shown in Table 1. The effects of these agents on the VLF oscillation of blood pressure are shown in Figs. 3 and 4.
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-receptors; and 5) the membrane ion channels involved are voltage-operated Ca2+ channels and Ba2+-sensitive K+ channels.
Additional characteristics of the oscillations described in the present experiments are worthy of comment. 1) Occurrence of the oscillations is by no means dependent (within wide limits) on the prevailing mean values of blood pressure, because even when blood pressure was markedly elevated (e.g., by infusion of L-NAME or propranolol) or reduced (e.g., by infusion of losartan or prazosin), the catecholamine-induced blood pressure oscillations were maintained. 2) VLF oscillations of blood pressure occurred in the absence of parallel oscillations of PI, so that their hemodynamic genesis is largely ascribed to oscillations of total peripheral resistance, rather than cardiac output. 3) No apparent contribution of humoral influences, other than adrenergic, could be identified, inasmuch as no effects on VLF oscillations of blood pressure were produced by administration of losartan, vasopressin, or L-NAME, i.e., specific antagonists or agonists of major vasoactive systems such as the renin-angiotensin, vasopressin, and nitric oxide systems, respectively. 4) The catecholamine-dependent blood pressure oscillations represent a quantitatively striking phenomenon, which indicates that they depend on rhythmic and coordinated oscillations of all (or at least the majority) of circulatory territories. 5) The 2-adrenergic receptors responsible for the appearance of VLF oscillations of blood pressure are most likely located postsynaptically, because the oscillations were observed under the ganglion-blocked condition and, thus, were independent of presynaptic modulation of neurotransmitter release.
Before commenting on the mechanistic and possible (patho)physiological implications of our findings, we should mention a possible limitation of our study, i.e., that the small number of animals included in some experimental groups might have generated type II statistical errors, especially for the experiments with K+ channel-blocking agents that failed to interfere with blood pressure oscillations. However, two considerations suggest that such errors are unlikely: 1) In the above experiment, as well as in all other experiments of the present study, the ability or inability of a given substance to generate VLF oscillations of blood pressure or to abolish ongoing VLF oscillations of blood pressure was visually and reproducibly apparent within a couple of minutes of infusion, indicating that we were dealing with qualitative (i.e., the presence or absence of blood pressure oscillations), rather than quantitative (i.e., greater or smaller amplitude), differences. 2) In our experimental setting, whenever an agent proved to be effective in generating or abolishing the oscillations, statistical analysis was associated with highly significant P values, even though the number of animals was low [e.g., norepinephrine (n = 4) and yohimbine (n = 4)].
Our attempts to clarify the mechanisms underlying catecholamine-induced LF oscillations of blood pressure revealed the involvement of 2-adrenergic receptors, downstream from which different membrane ion channels, i.e., the Ca2+ voltage-dependent and the Ba2+-sensitive K+ channels, are also involved (25, 28). We have, on the other hand, no experimental evidence about the location of the above-mentioned channels (endothelium vs. vascular smooth muscle cells) (26, 34) or the mechanism(s) responsible for the genesis of the vascular oscillatory phenomenon; one can speculate that we are dealing with a systemic extension of the vasomotion phenomena that have been described in the microcirculation (4, 8, 33, 35) and/or a cyclic variation in the responsiveness of 2-receptors, which are notoriously sensitive to local metabolic conditions (39). It is, however, unlikely that deranged local conditions (e.g., tissue hypoperfusion and hypoxia) potentially able to trigger the above phenomena (5, 13, 36) may have occurred in our animals. A more likely (although no less speculative) explanation would be that catecholamine-induced activation of 2-receptors may have triggered opening of voltage-dependent Ca2+ channels and vasoconstriction, which would in turn mechanically affect vascular smooth muscle cells with opening of inwardly rectifying K+ channels, cell repolarization, and vasodilation (12, 17, 29, 41).
Whatever the fine molecular basis of catecholamine-induced VLF oscillations of blood pressure, it is important to consider whether they represent a laboratory curiosity or may, instead, have clinical implications, i.e., whether they may affect any naturally occurring (patho)physiological conditions. This issue specifically arises in relation to the two major experimental interventions of our protocol: 1) ganglionic blockade and 2) catecholamine infusion. As to the former, it must be admitted that total abolition of neural circulatory control, such as that determined by hexamethonium administration, rarely if ever occurs in the clinical setting, but it is also true that at least partial neurocirculatory "denervation" is a feature of various clinical conditions such as diabetic neuropathy, dysautonomias, neurodegenerative diseases, and baroreceptor reflex dysfunction. As to the latter, it is to be emphasized that the circulating levels of epinephrine or norepinephrine associated with the doses administered in our experiments may not uncommonly be attained pathophysiologically (e.g., circulatory shock, severe heart failure, and phaeochromocytoma) or iatrogenically, especially in the emergency and operating rooms. Also VLF oscillations of blood pressure, although with much less amplitude than elicited under our experimental conditions, occur physiologically, as documented by the sizable spectral power observed within this frequency range in intact compared with ganglion-blocked rats (Fig. 1 and Table 2). Thus vascular adrenergic influences have the intrinsic potential to elicit VLF oscillations of blood pressure. In our animals subjected to ganglionic blockade and catecholamine infusion, the VLF oscillations of blood pressure were strikingly wider than in the absence of any pharmacological intervention, perhaps because of the absence of baroreflex buffering and/or a genuinely different vasomotor effect of a preferential activation of 2- vs. 1-adrenergic receptors, the former predominantly occurring under our experimental conditions (with vascular tone being largely due to blood-borne catecholamines) and the latter under normal conditions (with vascular tone being largely due to neurally released catecholamines).
Whatever the magnitude of this phenomenon under (patho)physiological conditions, it may be interesting to consider that oscillations of vascular diameter may have a favorable hemodynamic implication, insofar as, for the same mean diameter, an "oscillating" vascular bed will have greater conductance than a "fixed" one (24); thus, under critical circulatory conditions, it may be that vascular diameter oscillations may favor tissue perfusion. Along this line of reasoning, there is also evidence that in medium or large arteries, the application of pulsatile, rather than constant, stretch is accompanied by enhanced release of endothelium-derived relaxing factor and greater vessel compliance (27, 32, 35, 40).
In summary, we documented in the unanesthetized free-moving rat that 1) VLF oscillations of blood pressure are abolished by removal of autonomic influences by acute ganglionic blockade and 2) under such areflexic conditions, exogenous catecholamines have the ability to generate wide VLF oscillations of blood pressure, predominantly arising from 2-receptor activation coupled with Ba2+-sensitive K+ channels and voltage-dependent Ca2+ channels.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
1a- and 2-adrenoceptors in vascular myocytes: involvement of different transduction pathways. Cell Signal 7: 471–479, 1995.[CrossRef][ISI][Medline]1- and 2-adrenoceptors and endothelium-derived relaxing factor. Circ Res 71: 188–200, 1992.2- but not 1-adrenoceptor constriction by acidosis and hypoxia in vitro. Am J Physiol Heart Circ Physiol 268: H2068–H2076, 1995.This article has been cited by other articles:
|
|
 |
|
  Y. Bai, K. L. Siu, S. Ashraf, L. Faes, G. Nollo, and K. H. Chon Nonlinear coupling is absent in acute myocardial patients but not healthy subjects Am J Physiol Heart Circ Physiol, August 1, 2008; 295(2): H578 - H586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Langager, B. E. Hammerberg, D. L. Rotella, and H. M. Stauss Very low-frequency blood pressure variability depends on voltage-gated L-type Ca2+ channels in conscious rats Am J Physiol Heart Circ Physiol, March 1, 2007; 292(3): H1321 - H1327. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
