Sympathectomy and Cardiovascular Spectral Components in Conscious Normotensive Rats
Abstract We examined the extent to which sympathetic influences are reflected by spectral powers of blood pressure and pulse interval in specific frequency bands in spontaneously behaving Wistar-Kyoto rats subjected to continuous intra-arterial blood pressure recording. The rats were pretreated with 6-hydroxydopamine (150 mg/kg twice in 1 week, n=19) to produce chemical sympathectomy or received vehicle (n=15). In the sympathectomized group, additional monitoring sessions were performed with rats under α-adrenergic receptor blockade with phenoxybenzamine (n=8), β-receptor blockade with propranolol (n=7), or cholinergic receptor blockade with atropine (n=8). Blood pressure signals were analyzed by a computer to calculate spectral powers (fast Fourier transform) in the low-frequency (0.025 to 0.1 Hz), mid-frequency (0.1 to 0.6 Hz), and high-frequency (0.8 to 3.0 Hz) bands. In sympathectomized rats, low-frequency power of blood pressure was 70% greater than in intact rats, whereas mid-frequency power was 60% smaller (P<.05 for both) and high-frequency power was unchanged. High-frequency power of pulse interval was also unchanged in sympathectomized rats, whereas low- and mid-frequency powers were reduced by approximately 50% (P<.05). No further alterations in spectral powers were observed by adding α- or β-adrenergic blockade to sympathectomy, whereas adding cholinergic blockade caused a striking reduction in all pulse interval powers. Thus, mid-frequency blood pressure power depends on sympathetic but also to a substantial extent on nonsympathetic influences. Sympathetic influences do not contribute to low-frequency blood pressure power, having instead a restraining effect. The low- and mid-frequency pulse interval powers depend on both sympathetic and vagal influences. Thus, no blood pressure or pulse interval power in the mid- and low-frequency ranges can be regarded as a specific marker of sympathetic activity.
Spectral analysis has shown in animals and humans that blood pressure and heart rate fluctuate over a wide range of frequencies. Most studies, however, have focused on discrete frequency bands, defined as high-frequency (HF), mid-frequency (MF), and low-frequency (LF), that have been thought to carry important information on cardiovascular control mechanisms.1 2 3 4 5 6 In particular, the LF and MF fluctuations have been deemed to reflect sympathetic influences because these fluctuations increase and decrease under conditions in which sympathetic cardiovascular drive rises and falls, respectively.6 7 8 9 However, the ability of LF and MF blood pressure and heart rate powers to represent precise markers of sympathetic cardiovascular influences is not consistent with several other observations. First, LF and MF heart rate powers were shown to depend also on vagal cardiac influences, and LF and MF blood pressure powers were not consistently attenuated or abolished by sympatholytic drugs.3 10 11 12 Second, in humans, MF heart rate power was not found to increase during a sympathoexcitatory maneuver such as mental arithmetic,13 and under resting conditions, a limited correlation was reported to exist between heart rate and blood pressure fluctuations in the LF and MF bands and directly recorded sympathetic nerve activity in both humans and rats.14 15 Finally, preliminary experiments performed in our laboratory in rats subjected to sympathectomy by 6-hydroxydopamine suggested that this intervention reduces but does not abolish the MF power of systolic blood pressure (SBP), the concomitant LF SBP power paradoxically being increased.16
These experiments, however, were difficult to interpret because 6-hydroxydopamine spares the adrenal medulla and may not destroy all peripheral sympathetic nerves. In addition, data analysis was limited to SBP. In the present study, we attempted to overcome these limitations and thus provide conclusive evidence on whether the use of MF and/or LF blood pressure powers as specific indexes of sympathetic activity is justified. We pursued this by performing spectral analysis of SBP and diastolic blood pressure (DBP) in a large number of intact and sympathectomized rats; the latter were examined before and after superimposition of α- or β-adrenergic blockade to completely remove cardiovascular adrenergic influences. A further goal of the study was to perform spectral analysis of pulse interval in rats studied before and after sympathectomy and additional cholinergic blockade and determine to what degree the MF and LF heart rate powers are specific sympathetic markers and whether vagal influences may also contribute.
Thirty-four normotensive Wistar-Kyoto male rats (Charles River Calco) were housed in groups of two to four and received standard chow and water ad libitum. At the time of the study, rats were 10 to 12 weeks old. Intact rats weighed 218±9 g, and sympathectomized rats weighed 221±8 g (mean±SEM).
Destruction of efferent sympathetic nerve endings was accomplished in 19 rats by administration of 6-hydroxydopamine (Sigma Chemical Co) injected twice at 150 mg/kg body wt over a period of 5 to 7 days.17 18 The first dose was injected intraperitoneally, and the second dose was injected intravenously after chronic instrumentation of the rats (see below). Fifteen intact rats injected with vehicle alone were used as controls. Before the blood pressure monitoring session (see below) was started, the effectiveness of sympathectomy was tested by comparing the pressor and tachycardic responses to tyramine (ie, a drug that releases norepinephrine from sympathetic endings) at 150 μg/kg IV in treated versus untreated rats. The example of Fig 1⇓ and the average responses summarized in Table 1⇓ indicate that 6-hydroxydopamine administration produced a subtotal (85% to 90%) sympathectomy.
After anesthesia with ketamine HCl at 80 mg/kg IP, each rat was instrumented with polyethylene catheters inserted into a femoral artery and femoral vein. The catheters were tunneled subcutaneously and exteriorized at the interscapular region, their patency being maintained by periodic flushing with minute amounts of heparinized saline solution. After instrumentation, rats were individually housed in a wide cage for 24 hours to allow them to recover from surgery and become acclimated to the environment.
In both sympathectomized and control rats, the arterial catheter was connected to a P23Dc pressure transducer (Gould-Statham); this catheter-transducer system had a flat frequency response up to 30 Hz. This avoided distortions of the harmonic components of the pressure wave smaller than 6; ie, it allowed accurate recordings of components accounting for virtually all signal variance. The blood pressure signal was displayed on a chart recorder (7D polygraph, Grass Instrument Co), and heart rate was also displayed by tachographic beat-to-beat conversion of the pulsatile arterial blood pressure signal. Initially, tyramine was injected via the venous catheter (see above). The rat was then allowed to recover from the effects of tyramine for no less than 30 minutes. Finally, arterial blood pressure was monitored for 90 minutes, the signal being stored on a high-fidelity FM magnetic tape recorder (Racal Store 4), while the rats could freely move, explore, eat, and sleep within their cages. During the recording period, acoustic disturbances were avoided, and the cage was kept at constant temperature and moderate degree of illumination. The rats were kept under constant visual surveillance; no appreciable behavioral differences were perceived in the rat groups subjected to the various interventions. In 8 sympathectomized rats, an additional 90-minute recording period was carried out after administration of phenoxybenzamine at 1 mg/kg. This was done to remove possible influences on α-adrenergic receptors exerted by (1) residual norepinephrine release from the sympathetic endings that may not have been destroyed by 6-hydroxydopamine and (2) catecholamines released by the adrenal medulla, which is spared by this model of sympathectomy.19 Similarly, since residual catecholamine release may also affect variability by stimulating vascular and cardiac β-adrenoceptors, 7 sympathectomized rats were subjected to a further 90-minute blood pressure recording after administration of 1 mg/kg propranolol. Finally, in 8 sympathectomized rats, an additional 90-minute recording period was carried out after 0.8 mg/kg IV atropine. This was done to establish the contribution of vagal modulation of the cardiovascular system to pulse interval and blood pressure powers after sympathectomy. Four of the sympathectomized rats were used for two drug interventions, the atropine session being carried out immediately after baseline recording and the phenoxybenzamine session being carried out 24 hours later.
The effectiveness of α-adrenergic blockade was tested by the disappearance of the pressor effect of phenylephrine: at 2 μg/kg IV, this drug induced an increase in mean arterial pressure always greater than 20 mm Hg before and never greater than 2 mm Hg after phenoxybenzamine. The effectiveness of β-adrenergic blockade was tested by the disappearance of the tachycardic effect of isoproterenol: at 4 μg/kg IV, the drug induced an increase of heart rate always greater than 40 beats per minute before and never greater than 5 beats per minute after propranolol. The cholinergic blockade was regarded as effective if the reflex bradycardic response to a phenylephrine-induced (2 μg/kg IV bolus) increase in blood pressure was reduced by at least 90%.
The experimental procedures were performed in accordance with Italian government directions concerning the protection of animals used for scientific purposes.
A detailed description of the spectral analysis technique we used has been reported elsewhere.20 In brief, the 90-minute tape-recorded arterial pressure signal was fed into a computer (Olivetti XP5), sampled at 250 Hz, digitized at 12 bits, and edited for artifacts as well as for nonstationarities caused by the occurrence of arrhythmias. Systolic and diastolic values were identified for each cardiac cycle and computed on a beat-to-beat basis; pulse interval was computed by measuring the interval between two consecutive systolic peaks. To improve the time resolution of the pulse interval estimate, the systolic portion of each sampled pulse wave was interpolated by a parabolic curve, the apex of which was taken as the true time of occurrence of the systolic peak. We have previously shown that in normotensive and hypertensive humans under concomitant 24-hour blood pressure and electrocardiographic recordings, the use of pulse interval (the time interval between two consecutive systolic peaks) and of RR interval (electrocardiographic) provided similar spectral powers.6 Pulse interval and corresponding RR interval were similar also in our conscious rats. This is exemplified in Fig 2⇓, which shows a 30-minute recording of heart interval by the two methods. The average heart interval values of the 30-minute period were both 167.3 milliseconds when estimated by pulse interval and RR interval, the corresponding standard deviations being 4.45 and 4.36 milliseconds, respectively. The absolute difference (regardless of the sign) was 0.92 millisecond.
For each 90-minute recording period, average SBP, DBP, and pulse interval along with the respective variances were computed. In addition, each SBP, DBP, and pulse interval beat-to-beat series was divided into consecutive 100-second periods, and for each of the three variables, the spectra were calculated by the fast Fourier transform technique. This approach provided spectra including the power related to both rhythmic and nonrhythmic variations in line with what is usually done in spectral studies on rats.3 8 10 11 12 Since SBP, DBP, and pulse interval are defined on a beat-to-beat basis, they were irregularly sampled in time (because of the variability in heart rate) but regularly spaced as a function of the beat number. Thus, the obtained spectra, which should be estimated from evenly sampled data, had frequencies expressed as cycles per beat. To transform the results into hertz (cycles per second), the original frequencies were divided by the average pulse interval computed over the respective 100-second period. This approach could be used without risk of distortion in the spectra, instead of procedures based on the interpolation and resampling of the data, because the pulse interval variations in a single period were largely lower than the average pulse interval value.21 From each spectrum thus obtained, power spectral densities were integrated in the HF band (3.0 to 0.8 Hz), MF band (0.6 to 0.1 Hz), and LF band (0.1 to 0.025 Hz) based on (1) preferential clustering of power in these frequency regions, in agreement with observations made in earlier spectral studies in rats,3 8 10 11 12 22 and (2) correspondence of these bands to the HF, MF, and LF bands identified within a lower frequency range (between 1 and 0.07 Hz) in spectral studies on humans and large laboratory animals.8
Individual data were averaged separately for intact and sympathectomized rats. Statistical comparisons of the differences between the two groups were performed by the t test for unpaired observations. The t test for paired observations was used to statistically evaluate the effects of α-adrenergic blockade and cholinergic blockade. The level of statistical significance was set at a value of P<.05.
Sympathectomy and Spectral Powers
Table 2⇓ shows that sympathectomized rats had lower SBP and DBP than control rats, whereas pulse interval was not significantly different in the two groups. SBP and DBP variances were significantly larger in sympathectomized rats, whereas pulse interval variance was not different in sympathectomized and control rats.
Fig 3⇓ shows a typical spectral profile of SBP in one control and one sympathectomized rat, with the average spectral data reported in Fig 4⇓. For both SBP and DBP, the power of the HF band was similar in control and sympathectomized rats. In sympathectomized rats, however, the MF power for both SBP and DBP was significantly lower than in control rats, and the LF power was significantly higher. The decrease in the MF power and the increase in the LF power displayed by the sympathectomized rats amounted to approximately −60% and +70%, respectively.
The HF power of pulse interval was similar in control and sympathectomized rats, and the MF and LF powers were significantly lower in the latter compared with the former group.
α-Adrenergic Receptor Blockade and Spectral Powers
In sympathectomized rats, additional α-adrenergic blockade was accompanied by no significant change in SBP and DBP and by a slight but significant reduction in pulse interval (Table 3⇓). On the other hand, the variance of all three variables was unchanged. This was the case also for the LF, MF, and HF powers, which for SBP, DBP, and pulse interval were not significantly different before and after α-adrenergic blockade (Fig 5⇓).
β-Adrenergic Receptor Blockade and Spectral Powers
In sympathectomized rats, additional β-adrenergic blockade was accompanied by a slight but significant increase in SBP and by the expected increase in pulse interval (Table 3⇑). On the other hand, the variance of all three variables was unchanged. This was the case also for LF, MF, and HF powers of SBP and DBP, which were not significantly different before and after β-adrenergic blockade. LF and MF powers of pulse interval showed a trend toward a further reduction after propranolol although the difference was not statistically significant in either case with the prepropranolol data (Fig 6⇓).
Cholinergic Receptor Blockade and Spectral Powers
In sympathectomized rats, additional cholinergic blockade was accompanied by no significant changes in SBP and DBP and by the expected reduction in pulse interval (Table 3⇑). Both SBP and DBP variances were increased, whereas pulse interval variance was strikingly reduced. SBP and DBP powers were unaffected in the HF and MF bands, and they were significantly increased in the LF band; pulse interval power was drastically reduced in all three frequency bands (Fig 7⇓).
Our study shows that in spontaneously behaving unanesthetized Wistar-Kyoto rats, sympathectomy by 6-hydroxydopamine is associated with an increase in LF and a decrease in MF powers of SBP and DBP. Both the increased LF and decreased MF blood pressure powers were unaffected by additional α- or β-adrenergic blockade, ie, by removal of possible residual vascular adrenergic influences. It can thus be concluded that the LF blood pressure power is not produced by sympathetic influences, which instead exert a net counteracting effect on its amplitude. In contrast, the MF blood pressure power does depend on sympathetic mechanisms, as suggested by previous studies.11 22 It should be emphasized, however, that about a third of the MF blood pressure power survived sympathectomy without or with additional α- or β-adrenergic blockade, which means that a substantial proportion of this power is generated by nonsympathetic factors. Thus in spontaneously behaving rats, the specificity of MF blood pressure power as a marker of sympathetic activity is limited.
Another new finding of our study is that sympathectomy caused a marked reduction in the MF but also in the LF powers of pulse interval. Thus, sympathetic influences are more extensively involved in the production of heart rate powers than they are in the production of blood pressure powers. Also for heart rate, however, the specificity of these powers as sympathetic markers is limited because a noticeable proportion of the LF (50%) and MF (30%) powers seen in control rats was still present in sympathectomized rats. Our study did not allow us to establish the extent to which the observed sympathetic influences on spectral powers originate from central neural rather than from baroreflex mechanisms. However, the spectral effects of sympathectomy seen in the present study were similar to those caused by sinoaortic denervation in a previous study on conscious cats,23 suggesting that reflex influences may be significantly involved. We also have no information on the nature of the nonsympathetic influences accounting for the residual MF and enhanced LF blood pressure powers seen after sympathectomy and α- or β-adrenergic blockade. It is clear from our atropine data, however, that these influences are not parasympathetic. We can speculate that fluctuations in (1) circulating vasoactive substances such as angiotensin II or vasopressin, (2) receptor responsiveness, and/or (3) myogenic activity in the vessel wall cause changes in systemic vascular resistance and blood pressure within the MF and LF bands; indeed, several previous studies have implicated the renin-angiotensin system in the genesis of LF blood pressure fluctuations.3 5 12 24
Several other aspects of our study deserve to be discussed. First, one can speculate that part of the increase in LF power and the reduction in MF blood pressure power produced by sympathectomy were related to each other, ie, that this pattern reflected a redistribution of power from the MF to lower-frequency regions and thus an entraining effect of sympathetic influences toward the MF band. This entraining effect may not be limited to the LF and MF powers but may extend over a lower frequency range. Namely, sympathectomy may enhance spectral powers to a progressively greater extent at frequencies progressively lower than LF, sympathetic influences thereby reducing the degree to which blood pressure power normally increases as frequency diminishes (the so-called 1/f slope).25 Second, the unchanged HF power and reduced LF and MF powers of pulse interval seen after sympathectomy were markedly blunted by additional cholinergic blockade by atropine, in line with previous reports that the spectral expression of vagal cardiac control extends well below the HF range.3 4 22 It is thus clear that pulse interval fluctuations in the MF and LF regions of the spectrum are generated by concurrent vagal and sympathetic influences and that none of these fluctuations represent specific sympathetic or vagal markers.
The third and final point concerns the methodology of power spectral analysis, ie, the fact that cardiovascular spectral data are often analyzed and presented as HF and LF bands only, the LF band including the MF range as well. However, previous data in dogs, cats, and rats indicate that interfering with autonomic cardiovascular influences may affect LF and MF blood pressure powers in a quantitatively and qualitatively different fashion.3 10 11 22 23 This is fully confirmed in the present study in which LF and MF blood pressure powers were affected in an opposite direction by chemical sympathectomy. Thus, lumping MF and LF blood pressure powers together may conceal information on their determinants that can instead be obtained by analyzing each of the two frequency components separately.
In conclusion, our study provides evidence that in rats MF blood pressure power originates from sympathetic but to a sizable extent also from nonautonomic influences. Therefore, its specificity as a sympathetic marker is limited. In contrast, LF blood pressure power is not produced by sympathetic influences, which are likely to oppose LF blood pressure oscillations. Both MF and LF powers of pulse interval are markers of cardiac sympathetic control, but their specificity is also limited because of concurrent vagal influences impinging on the same frequency range.
Revision accepted January 30, 1995.
Kitney RI, Rompelman O. In: The Study of Heart Rate Variability. Oxford, UK: Claredon; 1980:59-107.
Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat to beat cardiovascular control. Science. 1981;213:220-223.
Pomeranz B, Macaulay RJB, Candill MA, Kutz I, Adam D, Gordon D, Kilborn KM, Barger AC, Shannon DC, Cohen RJ, Benson H. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol. 1985;248:H151-H153.
Koepchen HP. History of studies and concepts of blood pressure waves. In: Miyakawa K, Koepchen HP, Polosa C, eds. Mechanisms of Blood Pressure Waves. Tokyo, Japan: Japan Sci Soc Press; Berlin, Germany: Springer Verlag; 1984:3-23.
Parati G, Castiglioni P, Di Rienzo M, Omboni S, Pedotti A, Mancia G. Sequential spectral analysis of 24-hour blood pressure and pulse interval in humans. Hypertension. 1990;16:414-421.
Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfatto G, Dell’Orto S, Piccaluga E, Turiel M, Baselli G, Cerutti S, Malliani A. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res. 1986;59:178-193.
Akselrod S, Eliash S, Oz O, Cohen S. Hemodynamic regulation in SHR: investigation by spectral analysis. Am J Physiol. 1987;253:H176-H183.
Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation. 1991;84:482-492.
Akselrod S, Gordon D, Madwed JB, Snidam NC, Shannon DC, Cohen RJ. Hemodynamic regulation: investigation by spectral analysis. Am J Physiol. 1985;249:H867-H875.
Cerutti C, Gustin MP, Paultre CZ, Lo M, Julien C, Vincent M, Sassard J. Autonomic nervous system and cardiovascular variability in rats: a spectral analysis approach. Am J Physiol. 1991;261:H1292-H1299.
Eliash S, Oz O, Cohen S, Akselrod S. The role of the renin-angiotensin and alpha control in the regulation of blood pressure in a normotensive versus in a hypertensive system. Proc Computers in Cardiology, Chicago, 1990. Los Alamitos, Calif: IEEE Comput Society Press. 1991:155-158.
Langewitz W, Ruddel H. Spectral analysis of heart rate variability under mental stress. J Hypertens. 1989;7:S32-S33.
Saul JP, Rea RF, Eckberg DL, Berger RD, Cohen RJ. Heart rate and muscle sympathetic nerve variability during reflex changes of autonomic activity. Am J Physiol. 1990;258:H713-H721.
Persson PB, Stauss H, Chung O, Wittmann U, Unger T. Spectrum analysis of sympathetic nerve activity and blood pressure in conscious rats. Am J Physiol. 1992;263:H1348-H1355.
Daffonchio A, Franzelli C, Di Rienzo M, Castiglioni P, Ramirez AJ, Parati G, Mancia G, Ferrari AU. Effect of sympathectomy on blood pressure variability in conscious rat. J Hypertens. 1991;9(suppl 6):S70-S71.
Provoost AP. Sympathectomy and the development of hypertension in rats. In: De Jøng W, ed. Handbook of Hypertension, Volume 4: Experimental and Genetic Models of Hypertension. Amsterdam, Netherlands: Elsevier Science Publishers; 1984:495-517.
Ferrari AU, Daffonchio A, Franzelli C, Mancia G. Potentiation of the baroreceptor–heart rate reflex by sympathectomy in conscious rats. Hypertension. 1991;18:230-235.
Kostrezewa RM, Jacobowitz DM. Pharmacological actions of 6-hydroxydopamine. Pharmacol Rev. 1974;26:199-275.
Di Rienzo M, Parati G, Castiglioni P, Omboni S, Ferrari AU, Ramirez AJ, Pedotti A, Mancia G. Role of sinoaortic afferents in modulating BP and pulse-interval spectral characteristics in unanesthetized cats. Am J Physiol. 1991;261:H1811-H1818.
Akselrod S, Oz O, Eliash S. Neural and humoral factors in regulation of blood pressure in the investigation of essential hypertension. In: Di Rienzo M, Mancia G, Parati G, Pedotti A, Zanchetti A, eds. Blood Pressure and Heart Rate Variability. Amsterdam, Netherlands: IOS Press; 1993:192-206.
Marsh DJ, Osborn JL, Cowley AW. 1/f fluctuations in arterial pressure and regulation of renal blood flow in dogs. Am J Physiol. 1990;258:F1394-F1400.