neural control of arterial pressure variability in the neuromuscularly blocked rat

ORIGINAL ARTICLE

Neural control of arterial pressure variabilityin the neuromuscularly blocked rat

Xiaorui Tang • Tian Hu

Received: 23 March 2011 / Accepted: 30 August 2011 / Published online: 23 September 2011

� Springer-Verlag 2011

Abstract The baroreflexes stabilize moment-to-moment

arterial pressure. Sinoaortic denervation (SAD) of the ba-

roreflexes results in a large increase in arterial pressure

variability (APV) across various species. Due to an

incomplete understanding of the nonlinear interactions

between central and peripheral systems, the major source

of APV remains controversial. While some studies sug-

gested that the variability is endogenous to the central

nervous system (CNS), others argued that peripheral

influences may be the main source. For decades, abnormal

cardiovascular variability has been associated with a

number of cardiovascular diseases including hypertension,

heart failure, and stroke. Delineating mechanisms of the

APV is critical for the improvement of current strategies

that use APV as a clinical tool for the diagnosis and

prognosis of cardiovascular diseases. In this study, with a

unique chronic neuromuscularly blocked (NMB) rat prep-

aration that largely constrains peripheral influences, we

determined the CNS contribution to the post-SAD APV.

First, we confirmed that SAD significantly increased APV

in the NMB rat, then demonstrated that post-SAD gangli-

onic blockade substantially reduced APV, and subsequent

intravenous infusions of phenylephrine and epinephrine (in

presence of ganglionic blockade) only slightly increased

APV. These data suggest that the CNS is an important

source, and skeletal activity, thermal challenges or other

forms of peripherally generated cardiovascular stress are

not required for the post-SAD APV. In addition, we

showed that bilateral aortic denervation produced a larger

increase in APV than bilateral carotid sinus denervation,

suggesting that the aortic baroreflex plays a more dominant

role in the control of APV than the carotid sinus.

Keywords Sinoaortic denervation � Arterial pressure

variability � Heart rate variability � Baroreflex �Neuromuscular block � Ganglionic blockade � Sympathetic

nervous system � Vascular tone

Introduction

For decades, abnormal cardiovascular variability has been

regularly described in, and used as a risk factor for, a

number of cardiovascular diseases including hypertension,

heart failure, myocardial infarction, and stroke (Casolo

et al. 1995; Frattola et al. 1993; Julius and Nesbitt 1996;

Malpas 2002; Pringle et al. 2003). A better understanding

of the factors involved in the generation of arterial pressure

variability (APV) will not only improve current strategies

for the use of APV as a clinical tool in the assessment of

cardiovascular function, but may also shed some light into

the processes involved in the development of cardiovas-

cular pathologies in general.

The arterial baroreflexes have a critical role in regulat-

ing and stabilizing the arterial pressure. Impairment or

interruption of the baroreflexes results in a significant

increase in APV (Dworkin et al. 2000; Martinka et al.

2005; Pringle et al. 2003; Schreihofer and Sved 1994;

Timmers et al. 2004a, b). However, due to complex and

often nonlinear interplays between the central and

Communicated by Susan A. Ward.

X. Tang (&)

Department of Neural and Behavioral Sciences, Pennsylvania

State University College of Medicine, Hershey, PA 17033, USA

e-mail: [email protected]

T. Hu

Department of Public Health, Pennsylvania State University

College of Medicine, Hershey, PA 17033, USA

123

Eur J Appl Physiol (2012) 112:2013–2024

DOI 10.1007/s00421-011-2160-4

peripheral systems, it is unclear whether the APV that

follows baroreflex disfunction arises from the central ner-

vous system (CNS) or from factors extrinsic to the sym-

pathetic nervous system. Some studies (Alper et al. 1987a,

b; Jacob et al. 1989; Trapani et al. 1986) suggested that

APV is produced primarily by influences from the CNS and

secondarily by peripheral non-neurogenic influences.

These conclusions stem from the observation in freely

moving sinoaortically denervated rats that blocking either

sympathetic ganglionic transmission or alpha-adrenergic

receptors substantially attenuated APV, but an additional

blockade of the renin-angiotensin and vasopressin systems

was required to further reduce APV.

In contrast, other studies (Barres et al. 1992; Jacob et al.

1988; Julien et al. 1993) suggested that APV is largely

determined by peripheral non-neurogenic mechanisms and

that the sympathetic nervous system only plays a permis-

sive role by maintaining vascular tone at a normal level.

These conclusions mainly resulted from observations in the

freely moving sinoaortically denervated rat that the low

APV produced by ganglionic blockade was largely restored

by infusion of pressor agents such as phenylephrine or

angiotensin II.

Over the past 30 years, our laboratory has developed and

utilized a unique chronic NMB rat preparation (Dworkin

and Dworkin 1990, 1995, 2004; Dworkin et al. 2000a, b;

Tang and Dworkin 2007a, b, 2009) that largely constrains

peripheral influences and significantly limits the complex

interactions between the central and peripheral nervous

systems. As such, this model provides a platform to clarify

the CNS’s contribution to a specific physiological phe-

nomenon because of minimal interference from peripheral

influences. The novelty of this current study resides in the

use of this unique NMB preparation to illustrate the role of

the CNS in generating post-SAD APV. Contrary to the idea

that APV is largely determined by peripheral nonneuro-

genic mechanisms (Barres et al. 1992; Jacob et al. 1988;

Julien et al. 1993), our results suggested that skeletal

activity, thermal challenges or other forms of peripherally

generated cardiovascular stress are not required for the post-

SAD APV. Two aims were addressed in this study. The first

aim was to evaluate the hypothesis that the CNS is a major

contributor to the observed large post-SAD APV. To this

end, we evaluated APV after (1) sino-aortic denervation

(SAD); (2) ganglionic blockade; and (3) ganglionic block-

ade followed by infusion of pressors in SAD and baroreflex-

intact rats. The second aim of this study was to examine the

relative contribution of the aortic and carotid sinus ba-

roreflexes in the control of APV. Therefore, we measured

APV following (1) bilateral aortic denervation followed by

SAD; (2) bilateral carotid sinus denervation followed by

SAD; and (3) unilateral aortic denervation. Finally, to better

understand detail characteristic features of the variability,

APV was analyzed both in the time domain as the standard

deviation (SD) and in the frequency domain with the Fast

Fourier Transform power spectral analysis of the hourly

systolic arterial pressure.

Methods

Sprague–Dawley female (255–390 g) rats were used in the

study. The experimental protocols were supervised and

certified to be in compliance with National Institute of

Health and American Physiology Society guidelines by the

Pennsylvania State University College of Medicine Insti-

tutional Animal Care and Use Committee (IACUC). On

average, each preparation lasted for *10 days, and mul-

tiple protocols were implemented in the same preparation.

Each rat was monitored and attended to around the clock.

The NMB rat was prepared as previously described in

(Dworkin et al. 2000; Tang and Dworkin 2007a). Briefly,

for maintaining the preparation, the following parenteral

solutions were infused continuously: (1) An intra-arterial

solution (0.6 ml/h) consisting of 50 ml H2O, 50 ml 0.5 N

lactated Ringer, 1.25 g oxacillin Na, 2.8 mg a-cobratoxin,

0.3 mg vitamin K (Synkavite), and 20 meq K? (as KCl);

(2) an intra-venous solution (0.6 ml/h) consisting of 50 ml

H2O, 50 ml 0.5 N lactated Ringer and 1.25 g oxacillin Na

for the initial 48 h, then glucose, 8.5% Travasol amino

acid, and intralipid (fat emulsion) were added subsequently

to provide necessary nutrients.

Surgery

The general surgical procedures are provided in detail in

Dworkin et al. (2000) and Tang and Dworkin (2007a).

Briefly, the rat was deeply anesthetized with [1.5% iso-

flurane and monitored to ensure that (1) the electroen-

cephalogram (EEG) was synchronized and dominated by

high-voltage low-frequency activity; (2) mean systolic

arterial pressure was\100 mmHg; (3) heart rate was\420

beats/min; and (4) arterial pressure, heart rate, and EEG did

not respond to surgical manipulation. General surgery

included implantation of a pair of subcutaneous precordial

electrocardiogram (EKG) silver wire electrodes, a transu-

rethral bladder cannula to allow drainage and measurement

of the rate of urine production, a femoral artery cannula for

arterial pressure measurement, and a femoral vein cannula

for recording venous pressure and administering parenteral

solutions. General surgery also included implantations of

EEG electrodes to assure maintenance of adequate levels of

anesthesia during the surgery and the regularity of sleep

and wakefulness cycles throughout the experiment. Core

temperature was servo-regulated at 37�C. Neuromuscular

block was induced with a 75 lg i.v. bolus of a-cobratoxin

2014 Eur J Appl Physiol (2012) 112:2013–2024

123

and maintained by continuous infusion at 250 lg/day. The

rat was ventilated with positive pressure at a rate of 72

breaths/min and at a volume of 200 cc/min. The air mix-

ture consisted of 48.5% O2, 47% N2, 3% CO2, and 1.5%

isoflurane. After the surgery, the rat was allowed to recover

for 2 days during which the isoflurane level was gradually

reduced to 0.5%. During the recovery, adequate level of

anesthesia was maintained by observing that the EEG,

arterial pressure, and heart rate signals were within the

physiological range. If any of these signals indicated signs

of distress during the reduction of anesthesia, the level of

isoflurane was elevated to its previous stage and main-

tained for a longer period of time before the next lowering

attempt. All experimental data used for the final analysis in

this study were acquired at 0.5% isoflurane level.

The aortic and carotid sinus denervation A modification

of Krieger’s sinoaortic denervation method (Krieger 1964)

was used. Briefly, the aortic denervation was achieved by

isolating and resecting the superior cervical ganglion,

superior laryngeal nerve, and aortic depressor nerve

(ADN). The carotid sinus denervation was achieved by

stripping off all connective tissue around the carotid sinus

region and painting the region with a solution of 10%

phenol in ethanol to remove any remaining baroreceptor

afferent fibers. Complete SAD was achieved by bilateral

aortic and carotid sinus denervation. This was confirmed by

the absence of bradycardia in response to a phenylephrine

challenge, which typically increased arterial pressure more

than 30 mmHg.

Experimental protocols

Protocols were arranged into two main categories corre-

sponding to the two major aims of the study: (1) evaluating

the CNS contribution to the post-SAD APV; and (2) evalu-

ating the relative contributions of the aortic and carotid sinus

baroreflexes in stabilizing cardiovascular variability.

Protocols for evaluating the CNS contribution

to the post-SAD cardiovascular variabilities

In the baroreflex denervated rats: Baseline ? SAD ?chlorisondamine administration ? phenylephrine and epi-

nephrine infusions Following recovery from the general

surgery, 6 h of baseline data were acquired before SAD.

Twenty-four hours following the SAD, a bolus of the gan-

glionic blocker, chlorisondamine (2.5 mg/kg, i.v.), was given.

Six hours later, phenylephrine and epinephrine were infused

continuously (i.v.) for 6 h each in a random order at a dose that

restored arterial pressure to or above baseline levels, as cus-

tomized for each NMB rat (phenylephrine, 5–73 lg kg-1

min-1; epinephrine, 2.6–36 lg kg-1 min-1). To achieve the

appropriate dose, the infusion rates of phenylephrine and

epinephrine were adjusted from 5 to 33 ll/min and from 3 to

13 ll/min, respectively. The purpose of using two different

pressors, i.e., phenylephrine (an alpha adrenergic agonist) and

epinephrine (an alpha and beta adrenergic agonist), was to

partially control for idiosyncratic effects of the particular

agent on the APV. Saline infusion at a rate of 10 ll/min was

used as a control.

In baroreflex-intact rats: Baseline ? chlorisondamine

administration ? phenylephrine and epinephrine infu-

sions Following recovery from the general surgery, a bolus

of chlorisondamine (2.5 mg/kg, i.v.) was given. Six hours

later, the pressors phenylephrine and epinephrine were

continuously infused (i.v.) for 6 h each in a random order

as described above. Saline infusion at a rate of 10 ll/min

was used as a control.

Protocols for evaluating the relative contributions

of the aortic and carotid sinus baroreflexes

in stabilizing the APV

Bilateral aortic or carotid sinus denervation: Base-

line ? bilateral aortic (or carotid sinus) denerva-

tion ? bilateral carotid (or aortic) denervation, i.e., SAD

Following recovery from the general surgery, bilateral

aortic (or carotid sinus) denervation was performed. This

initial denervation was chosen randomly on either pair of

the baroreceptors (i.e., half of the rats began with the aortic

then the carotid sinus denervation, and another half with

the carotid sinus then the aortic denervation). Twenty-four

hours later, further denervation was conducted on the

remaining pair of the baroreceptors.

Unilateral aortic denervation: Baseline ? left or right

aortic denervation: Following recovery from the general

surgery, aortic denervation was performed on the left aortic

baroreceptors.

Data acquisition and analysis

All experimental signals were continuously acquired 24 h a

day and 7 days a week throughout the entire experiment

using Spike2� and a Power 1401 data acquisition system

(Cambridge Electronic Design, Cambridge, UK). Five

hours of data were analyzed for each experimental condi-

tion in protocols evaluating the CNS contribution to the

post-SAD cardiovascular variability. Ten hours of data

were analyzed for each condition in protocols evaluating

the role of the aortic and carotid sinus baroreflexes in

stabilizing the APV. To eliminate potential transitional

effects of switching from one experimental condition to

another, signals were often acquired more than 5 or 10 h in

an experimental condition, and those from the last 5 or

10 h were analyzed for that condition. Data used for final

Eur J Appl Physiol (2012) 112:2013–2024 2015

123

analysis were always acquired at a 0.5% isoflurane level.

Arterial pressure variability (APV) and heart rate vari-

ability (HRV) were calculated as follows: First, beat-to-

beat systolic arterial pressure and heart rate were generated

from the arterial pressure and EKG waveforms, respec-

tively, using a peak detection algorithm in the Spike2�

software. APV and HRV were calculated in the time

domain as the standard deviations of systolic arterial

pressure and heart rate in each hour; and the standard

deviations across 5 or 10 h were averaged and reported for

each experimental condition. In the frequency domain, Fast

Fourier Transform (0.012 Hz resolution, Hanning window,

no overlapping samples) power spectral analysis was

conducted on systolic arterial pressure for each hour and

the very low-frequency (VLF: 0.01–0.15 Hz) and low-

frequency (LF: 0.15–0.6 Hz) powers were calculated by

integration of the areas in the corresponding frequency

ranges. The VLF and LF powers across 5 or 10 h were

averaged and reported for each experimental condition.

Statistical analysis

A separate analysis of variance (ANOVA) with repeated

measures was used to determine the effect of each exper-

imental intervention on APV, HRV, and the VLF and LF

powers of the systolic arterial pressure spectra; Newman–

Keuls post-hoc comparisons were applied when appropri-

ate. Differences were considered significant at p \ 0.05.

A Student t test was used to evaluate the effect of unilateral

aortic denervation on the APV, and differences were con-

sidered significant at p \ 0.05. All results are presented as

the mean ± SE.

Results

Results are arranged according to the two major aims of

this study: (1) evaluating the CNS contribution to the post-

SAD cardiovascular variability; and (2) evaluating the

relative contributions of the aortic and carotid sinus ba-

roreflexes in stabilizing the APV.

The CNS has a significant contribution to the post-SAD

APV

SAD increased and chlorisondamine decreased APV

in the NMB rat (Fig. 1, N = 9)

Systolic AP following baseline, SAD, chlorisondamine,

phenylephrine, and epinephrine infusion was analyzed using

one-way analyses of variance (ANOVAs) with repeated

measures across stage (Fig. 1B1, N = 9). The main effect

of stage was significant (F(4, 38) = 50.18, p \ 0.05).

Newman–Keuls post hoc test showed that, following SAD,

systolic AP was significantly increased from the baseline

(from 133 ± 4.9 to 151 ± 4.17 mmHg; p \ 0.05 vs. base-

line); the subsequent application of chlorisondamine

decreased systolic AP from 151 ± 4.17 to 67 ± 1.98

mmHg (p \ 0.05 vs. every other experimental stages). In the

presence of chlorisondamine, phenylephrine, and epineph-

rine infusion both returned systolic AP to the baseline control

level, from 67 ± 1.98 to 138 ± 7.72 mmHg for phenyl-

ephrine (p \ 0.05 vs. the chlorisondamine stage); and from

67 ± 1.98 to 153 ± 5.26 mmHg for epinephrine (p \ 0.05

vs. the chlorisondamine stage).

APV was analyzed using one-way repeated measures of

ANOVAs across stage (Fig. 1B2, N = 9). The main effect

of stage was significant (F(4, 38) = 84.31, p \ 0.05).


APV was significantly increased from the baseline (from

8.3 ± 1.4 to 25 ± 1.98 mmHg; p \ 0.05 vs. baseline); the

subsequent application of chlorisondamine decreased APV

from 25 ± 1.98 to 2.44 ± 0.36 mmHg (p \ 0.05 vs. every

other experimental stages). In the presence of chlorisond-

amine, phenylephrine, and epinephrine infusion both only

slightly increased APV, from 2.44 ± 0.36 to 7.18 ± 0.47

mmHg for phenylephrine (p \ 0.05 vs. the chlorisond-

amine stage); and from 2.44 ± 0.36 to 7.46 ± 0.71 mmHg

for epinephrine (p \ 0.05 vs. the chlorisondamine stage).

These data suggest that the CNS has a significant contri-

bution to the post-SAD APV. In corollary, these data also

suggest that the baroreflex is frequently engaged to restrain

moment-to-moment APV.

VLF power was analyzed using one-way repeated mea-

sures of ANOVAs across stage (Fig. 1B3, N = 9). The main

effect of stage was significant (F(4, 38) = 41.97, p \ 0.05).


VLF was significantly increased from the baseline (from

13.23 ± 2.07 to 30.89 ± 2.38 mmHg/HHz; p \ 0.05 vs.

baseline); the subsequent application of chlorisondamine

decreased VLF power from 30.89 ± 2.38 to 3.77 ± 0.51

mmHg/HHz (p \ 0.05 vs. every other experimental stages).

In the presence of chlorisondamine, phenylephrine and epi-

nephrine infusions both slightly increased VLF power, from

3.77 ± 0.51 to 11 ± 0.99 mmHg/HHz for phenylephrine

(p \ 0.05 vs. the chlorisondamine stage); and 3.77 ± 0.51

to 9.87 ± 0.85 mmHg/HHz for epinephrine (p \ 0.05 vs.

the chlorisondamine stage). VLF was identified as the prin-

cipal frequency region of the baroreflex system (Dworkin

et al. 2000). The VLF changes in response to each experi-

mental intervention corresponded well with the APV chan-

ges, further suggesting that a substantial proportion of the

APV likely arises from the CNS, and the baroreflex is fre-

quently engaged to restrain moment-to-moment APV.

LF power was analyzed using one-way repeated mea-

sures of ANOVAs across stage (Fig. 1B4, N = 9). The


123

Fig. 1 Effects of sinoaortic denervation (SAD), chlorisondamine

administration, intravenous infusions of phenylephrine, and epineph-

rine on the arterial pressure (AP) and arterial pressure variability

(APV) in NMB rats. a Shows representative traces of AP (A1) and

systolic AP (A2) and a frequency histogram of the systolic AP (A3)

from each experimental procedure. b Summarizes the results of

systolic AP (B1), APV (B2), very low-frequency (VLF: 0.01–0.15 Hz,

B3) and low-frequency (LF: 0.15–0.6 Hz, B4) powers of the systolic

AP spectra at baseline, following sinoaortic denervation, chlorisondamine

administration, and intravenous infusions of phenylephrine and

epinephrine. While SAD significantly increased APV and VLF

power, chlorisondamine decreased the values of both variables to

below their corresponding baseline levels; subsequent phenylephrine

and epinephrine infusions only slightly increased APV. APV was

calculated as standard deviation of the hourly systolic AP. Data are

mean ± SEM. Error bars SEM. *p \ 0.05. ##Statistically significant

difference from every other experimental stages


123

main effect of stage was significant (F(4, 38) = 20.63,

p \ 0.05). Newman–Keuls post hoc test showed that, fol-

lowing SAD, LF power was decreased from the baseline

(from 12.17 ± 1.68 to 6.34 ± 0.43 mmHg/HHz; p \ 0.05

vs. baseline); the subsequent application of chlorisond-

amine further decreased LF power from 6.34 ± 0.43 to

1.48 ± 0.11 mmHg/HHz (p \ 0.05 vs. every other

experimental stages). In the presence of chlorisondamine,

the phenylephrine and epinephrine infusion both slightly

increased LF power, from 1.48 ± 0.11 to 4.98 ± 0.59

mmHg/HHz for phenylephrine (p \ 0.05 vs. the chloris-

ondamine stage), and 1.48 ± 0.11 to 4.76 ± 0.45 mmHg/

HHz for epinephrine (p \ 0.05 vs. the chlorisondamine

stage). These data suggest that LF is the resonant (i.e.,

oscillation) region of the baroreflexes (Dworkin et al.

2000a). When denervation eliminated the negative feed-

back, it also eliminated the oscillation in this closed-loop

negative feedback system, which resulted in a power

decrease in the LF band of the APV spectrum.

Chlorisondamine, phenylephrine, and epinephrine

infusions produced similar effects in baroreflex-intact NMB

rats as in SAD NMB rats (Fig. 2)

Systolic AP, APV, VLF, and LF powers following base-

line, chlorisondamine, phenylephrine, and epinephrine

infusion in the baroreflex-intact and SAD rat were analyzed

separately using repeated measures of two-way ANOVAs

varying group and stage. There was a significant two-way

interaction varying group and stage in systolic AP, APV,

VLF, and LF (systolic AP, F(3, 33) = 3.82, p \ 0.05,

Fig. 2a; APV, F(3, 33) = 32.93, p \ 0.05, Fig. 2b; VLF, F

(3, 33) = 11.69, p \ 0.05, Fig. 2c; and LF, F(3,

33) = 9.76, p \ 0.05, Fig. 2d). Newman–Keuls post hoc

test showed that, similarly, in both SAD and baroreflex-

intact NMB rats, chlorisondamine significantly decreased

systolic AP, APV, VLF, and LF powers compared with the

pre-chlorisondamine baseline condition (p \ 0.05 vs. every

other experimental stages); the subsequent phenylephrine

and epinephrine infusion (in presence of the chlorisond-

amine) both returned systolic AP to the baseline condition

and increased APV, VLF, and LF power compared with the

chlorisondamine condition (p \ 0.05 vs. the chlorisond-

amine stage). Between the SAD and the baroreflex-intact

rats, the SAD rats had a higher systolic AP (Fig. 2a), APV

(Fig. 2b), and VLF power (Fig. 2c), but a lower LF power

(Fig. 2d) than the baroreflex-intact rats during the pre-

chlorisondamine baseline stage (p \ 0.05, for all vari-

ables). Note that, the baroreflexes were intact for the ‘Intact

Rats,’ and completely denervated for the ‘SAD Rats’ in the

baseline stage in Fig. 2. Interestingly, epinephrine infusion

produced a larger increase in APV, VLF power, and LF

power in the baroreflex-intact rats than in the SAD rats; the

significance of this increase, if any, has yet to be

determined.

Independent of prior status of the baroreflex system (i.e.,

SAD or intact), a bolus application of chlorisondamine

substantially decreased APV and VLF in both SAD and

baroreflex-intact NMB rats, suggesting that, by interrupting

the baroreflex efferent pathway, chlorisondamine blocked

the CNS contribution to the APV. Moreover, in presence of

the chlorisondamine, subsequent phenylephrine and epi-

nephrine infusions both returned systolic AP to or above

the baseline levels in both SAD and intact rats (Fig. 2a,

p \ 0.05). This restoration of systolic AP significantly

increased APV (Fig. 2b), VLF (Fig. 2c), and LF powers

(Fig. 2d) in both SAD and intact NMB rats, suggesting that

the APV is not only modulated by the CNS, but also by the

peripheral system.

The aortic baroreflex has a more dominant role

in the control of APV than the carotid sinus baroreflex

The aortic denervation produced a larger increase in APV

than the carotid sinus denervation in the NMB rat (Fig. 3,

N = 6 in each group)

APV, VLF power, and LF power following baseline,

bilateral aortic, or carotid sinus denervation and complete

SAD were analyzed using separate two-way analyses of

variance (ANOVAs) with repeated measures varying group

and stage. There was a significant two-way interaction

varying group and stage in APV and VLF (APV, F(2,

20) = 5.09, p \ 0.05, Fig. 3a; and VLF, F(2, 20) = 3.63,

p \ 0.05, Fig. 3b). Newman–Keuls post hoc test showed

that, for the APV (Fig. 3a), it was increased from the

baseline 7.02 ± 0.66 to 17.27 ± 0.82 mmHg following

bilateral aortic denervation (p \ 0.05 vs. baseline) and

further increased to 24.02 ± 2.07 mmHg following the

subsequent complete SAD (p \ 0.05 vs. baseline and

bilateral aortic denervation); the APV was increased from

the baseline 7.31 ± 0.96 to 12.14 ± 1.83 mmHg following

bilateral carotid sinus denervation (p \ 0.05 vs. baseline),

and further increased to 27.29 ± 1.12 mmHg following the

subsequent complete SAD (p \ 0.05 vs. baseline and

bilateral carotid sinus denervation). For the VLF power

(Fig. 3b), it was increased from the baseline 12.01 ± 1.5 to

24.69 ± 1.73 mmHg/HHz (p \ 0.05 vs. baseline) follow-

ing bilateral aortic denervation, and further increased to

30.63 ± 1.53 mmHg/HHz following the subsequent com-

plete SAD (p \ 0.05 vs. baseline and bilateral aortic

denervation); the VLF power was increased from

11.21 ± 1.31 to 19.12 ± 3.49 mmHg/HHz following

bilateral carotid sinus denervation (p \ 0.05 vs. baseline),

and further increased to 34.38 ± 1.85 mmHg/HHz fol-

lowing the subsequent complete SAD (p \ 0.05 vs.


123

baseline and bilateral aortic denervation). These data sug-

gest that while both aortic and carotic sinus baroreflexes

are important in stabilizing APV, the aortic baroreflex has a

more dominant role than the carotid sinus baroreflex. These

data also suggest that the effect of complete SAD on APV,

VLF, and LF variables is independent of the sequence of

the denervation.

A two-way analysis of variance with repeated measures

revealed a significant difference on main effect of stage in

LF (F(2, 20) = 21.33, p \ 0.05). Newman–Keuls post hoc

test indicated that, when combined, baseline LF was higher

in baseline than in the partial denervation (i.e., AD or CSD)

stage (Fig. 3c).

Single-side aortic denervation did not elicit a significant

increase in APV in the NMB rat (Fig. 4, N = 6)

Following unilateral (left or right only) aortic denervation

in the NMB rat, APV increased from 7.2 ± 1.03 to

8.87 ± 1.47 mmHg and VLF power from 9.54 ± 0.95 to

12.81 ± 1.89 mmHg/HHz. LF power decreased from

11.06 ± 2.04 to 8.58 ± 1.48 mmHg/HHz (N = 6 for all;

Fig. 2 Chlorisondamine administration, intravenous infusions of

phenylephrine, and epinephrine produced similar effects on the

systolic arterial pressure (AP), arterial pressure variability (APV),

VLF and LF powers of the systolic AP spectra in both SAD and

baroreflex-intact NMB rats. a The systolic AP. In both SAD and

baroreflex-intact rats, chlorisondamine significantly decreased sys-

tolic AP and subsequent infusions of phenylephrine and epinephrine

both returned systolic AP to or above the corresponding baseline

levels. b The APV. In both SAD and baroreflex-intact rats, chlorisond-

amine significantly decreased the APV and subsequent infusions of

phenylephrine and epinephrine both increased APV to the baroreflex-

intact baseline variability level. c VLF power of the systolic AP

spectrum. In both SAD and baroreflex-intact rats, chlorisondamine

significantly decreased the VLF power and subsequent infusions of

phenylephrine and epinephrine both increased the VLF power to the

baroreflex-intact baseline level. d LF power of the systolic AP

spectrum. In both SAD and baroreflex-intact rats, chlorisondamine

significantly decreased LF power and subsequent infusions of phenyl-

ephrine and epinephrine both increased the power to the SAD level.

Moreover, the SAD rat had a higher systolic AP, larger APV, and VLF,

but a lower LF power than the baroreflex-intact rat during the baseline

stage. In the baseline stage, the baroreflexes were completely dener-

vated for the ‘SAD Rats’ and un-manipulated for the ‘Intact Rats’. APV

was defined as standard deviation of the hourly systolic AP. Data are

mean ± SEM. Error bars SEM. *p \ 0.05. ##Statistically significant

difference from every other experimental stage


123

Fig. 4). None of the above variables were significantly

altered by unilateral aortic denervation. Taken together

with the results shown in Fig. 3, these data suggest that

with the carotid sinuses intact, only a single aortic limb was

required to stabilize the pressure.

SAD had no effect on the HRV (Fig. 5, N = 9)

In the NMB rat, baseline values of heart rate and HRV

were 373 ± 13 and 12 ± 1 beats/min, respectively. Fol-

lowing SAD, neither heart rate nor HRV was significantly

altered (p [ 0.05; Fig. 5). The different outcomes of SAD

on the APV and HRV suggest that different mechanisms

are involved in producing and stabilizing heart rate and

arterial pressure variability.

Discussion

In the present study we first confirmed that SAD produced

a substantial increase in APV in the NMB rat. We then

demonstrated that ganglionic blockade reduced the mean

systolic pressure and APV in both SAD and baroreflex-

intact NMB rats, and following ganglionic blockade,

Fig. 3 While bilateral aortic and carotid sinus denervation signifi-

cantly increased APV and VLF, bilateral aortic denervation produced

a larger increase in APV than carotid sinus denervation. a, b and

c depicted the effects of different denervation (i.e., AD, CSD, or

SAD) on APV, VLF, and LF, respectively. Bilateral aortic and carotid

sinus denervation both significantly increased APV and VLF power

compared with the baroreflex-intact baseline (p \ 0.05 vs. baseline,

for the APV and the VLF power); the subsequent SAD further

increased APV and VLF. (p \ 0.05 vs. baseline and AD or CSD

denervation stages). Bilateral aortic denervation produced a larger

increase in APV and VLF than carotid sinus denervation (p \ 0.05,

aortic vs. carotid sinus denervation). Data are mean ± SEM. Errorbars SEM. *p \ 0.05. APV was calculated as standard deviation of

the hourly systolic arterial pressure. VLF and LF: the very low-

frequency (VLF: 0.01–0.15 Hz) and the low-frequency (LF:

0.15–0.6 Hz) power of the systolic arterial pressure spectra, respec-

tively. AD aortic denervation, CSD carotid sinus denervation, SADsinoaortic denervation

Fig. 4 Unilateral aortic denervation had no effect on the APV. From

top to bottom, APV was characterized as the standard deviation of the

hourly systolic arterial pressure. The VLF and LF were the very low-

frequency (VLF: 0.01–0.15 Hz) and the low-frequency (LF:

0.15–0.6 Hz) power of the hourly systolic arterial pressure spectra,

respectively. Data are mean ± SEM. Error bars SEM. *p \ 0.05


123

pressor infusions restored mean systolic pressure to base-

line levels, but only slightly increased APV in both SAD

and baroreflex-intact NMB rats. These results suggest that

the CNS is an important source for the large APV fol-

lowing the SAD. However, pressor infusions following

chlorisondamine application also increased APV, indicat-

ing that the peripheral input also plays a role. The

peripheral input, however, is not necessarily required for

the manifestation of the large post-SAD APV, because

APV was substantially increased following SAD in the

NMB rat, in which the peripheral variance is largely

constrained.

We have also shown that (1) while bilateral aortic and

carotid sinus denervation both significantly increased APV,

the bilateral aortic denervation produced a larger increase

in APV than the bilateral carotid sinus denervation; and (2)

unilateral (left or right) aortic denervation alone had no

significant effect on the APV. These data suggest that the

aortic baroreflex has a more dominant role in stabilizing

moment-to-moment APV compared with the carotid sinus.

In addition, these data also suggest that, with a pair of

intact carotid sinus baroreflexes, a single aortic limb was

sufficient to stabilize the pressure.

Last, we showed denervation had no effect on HRV.

Different effects of baroreflex denervation on APV and

HRV suggest that different mechanisms are involved in

producing and stabilizing these variabilities.

The CNS is a major source of post-SAD APV

The NMB rat provides an ideal setting to analyze the

sources of APV: First, it is an extremely stable preparation.

We (Dworkin et al. 2000a) have shown previously that

baseline drift for arterial pressure for a typical NMB rat is

?0.01 mmHg/h (see bottom panel in Fig. 1 of Dworkin

et al. 2000a). Thus, for a typical 10-day experiment per-

formed in the current study, only 2–3 mmHg increase is

expected in arterial pressure over the entire experimental

duration. Also, APV was demonstrated to be consistent

over a 10-day period (bottom panel of Fig. 1 in Dworkin

et al. 2000a). Second, the NMB rat is largely free of

peripheral interference. Specifically, (1) chronic neuro-

muscular block removes the influence of skeletal muscle

activity on the regional blood flow and thereby regional

APV; (2) controlled ventilation at a precise rate and vol-

ume circumvents the influence of respiratory variations on

APV; (3) servo-regulated core temperature removes the

influence of temperature fluctuations on the APV; and (4)

experiments are conducted in a controlled, sound-isolated

environment which eliminates the presence of other

external stimuli. Finally, data in this study were acquired at

0.5% isoflurane level. It is well recognized that, at high

doses (3.0 MAC & 4.5%, Eger et al. 2003), isoflurane

produces direct cardiac depression, affecting both systolic

and diastolic function. However, no differences were found

in systolic and diastolic functions of rats between the low-

dose (\1 MAC) and isoflurane-free control conditions

(Skeehan et al. 1995). We concluded from over 30 years of

experience with the NMB rat that 0.5% isoflurane is a

comfortable analgesia level that minimally alters auto-

nomic cardiovascular function. At 0.5% isoflurane level,

the cardiovascular parameters (heart rate, blood pressure,

and even the post-SAD APV, etc.,) in NMB rats are well

within the range of those in freely moving rats, suggesting

that the NMB rat highly emulates the condition of freely

moving rats in the cardiovascular aspect.

Given the above, the evidence of a substantial increase

in APV following SAD in the NMB rat (Fig. 1) strongly

suggests that the CNS is an important source of the post-

SAD APV. The large increase in post-SAD APV also infers

that the intact baroreflex is frequently engaged to restrain

moment-to-moment APV.

The conclusion that the CNS is an important source of

post-SAD APV is further strengthened by the observation

that restoration of systolic pressure with pressor infusions

only slightly increased the APV levels in both SAD and

baroreflex-intact NMB rats. Chlorisondamine reduced APV

significantly, but also induced a significant drop in mean

Fig. 5 Sinoaortic denervation (SAD) had no effect on both the heart rate (a) and the heart rate variability (HRV) (b). Data are mean ± SEM.

Error bars SEM. *p \ 0.05


123

systolic pressure, and thus may have limited the range in

which arterial pressure could fluctuate. The fact that res-

toration of systolic pressure only slightly increased the

APV and VLF levels in both SAD and baroreflex-intact

NMB rats strongly supported the hypothesis that the CNS

is an important source of the post-SAD APV.

Furthermore, the result that chlorisondamine produced

similar effects on APV, VLF, and LF in the SAD and

baroreflex-intact NMB rat provided additional evidence

suggesting that the CNS is an important source for the post-

SAD APV. If the major source of post-SAD APV were in

the peripheral system, chlorisondamine application should

have produced different effects in the baroreflex-intact and

SAD rat. Chlorisondamine should have substantially

increased the APV in baroreflex-intact rats (particularly

when the systolic pressure was restored with the phenyl-

ephrine and epinephrine infusion) and had no substantial

effect on the post-SAD APV in the SAD rat. By blocking

the neural transmission in the efferent pathway of the

baroreflex system, chlorisondamine disrupts the baroreflex

control of the APV, which should result in a large increase

in the APV in baroreflex-intact rats. Conversely, since the

baroreflex system is already interrupted in the SAD rat,

further disruption of the system with chlorisondamine

should not further influence the APV (given that the major

source is at the periphery). Thus, the findings that chlo-

risondamine had similar effects on APV, VLF, and LF in

the SAD and baroreflex-intact NMB rat further suggest that

the CNS is an important source for the post-SAD APV.

Certainly, following chlorisondamine, the slight

increase of APV produced by phenylephrine and epi-

nephrine infusions in both SAD and baroreflex-intact rats

also suggests that peripheral input plays a role in the post-

SAD APV. Obviously, this input is not necessarily required

for the manifestation of the large post-SAD APV.

Aortic baroreflex has a more dominant role in control

of APV than the carotid baroreflex

In agreement with findings of Van Vliet et al. (1999), we

found that the aortic baroreflex has a more dominant role in

controlling APV than the carotid sinus baroreflex, because

aortic denervation produced a larger increase in APV than

the carotid sinus denervation. In support, Sanders et al.

(1988) showed that stimulation of the aortic baroreflexes,

but not carotid sinus reflexes, produced a substantial and

sustained inhibition of muscle sympathetic nervous activity

in humans. Prior studies (Dampney et al. 1971; Donald and

Edis 1971; Pelletier et al. 1972; Pickering et al. 2008;

Sanders et al. 1988) have also shown that the cardiac

baroreflex is critically dependent upon the aortic afferents

with relatively little contribution from the carotid sinus. All

together, these results indicate that the aortic baroreflex has

a more dominant role than the carotid baroreflex in overall

blood pressure regulation.

In the present study we reported that HRV was not sig-

nificantly altered by either partial or complete SAD. This is

similar to what was observed in previous reports (Alper et al.

1987a, b; Mancia et al. 1999; Tang and Dworkin 2009;

Trapani et al. 1986), indicating that HRV has no relevant role

in APV. The different effects of denervation on the HRV and

APV suggest that the stabilization of heart rate and arterial

pressure variability is controlled by different mechanisms.

The similar HRV observed between the denervation and

baseline stages suggests that HRV is unlikely to be the source

of the post-SAD APV. In support of these conclusions,

Ferrari et al. (Ferrari et al. 1987) showed that administration

of atropine to freely moving rats with intact baroreceptors

significantly reduced HRV while increasing APV. This

indicates that the vagally mediated fluctuations in heart rate

are probably triggered by the arterial baroreflexes, instead of

being the source of APV, and reflect the ability of the baro-

reflex system to buffer arterial pressure changes through

opposing changes in cardiac output.

Power spectral analysis of the APV

In this study, we found that a partial denervation (the aortic

or carotid sinus denervation) increased and a complete

SAD further increased VLF power of the APV spectra.

These results are consistent with the findings of our prior

study (Dworkin et al. 2000b) and also findings of others

(Burgess et al. 1997; Cerutti et al. 1994; Jacob et al. 1995;

Julien et al. 2003; Ramanathan et al. 1994). In a previous

study (Dworkin et al. 2000b), using electrical stimulation

of the aortic depressor nerve in NMB rats, we identified a

very low-frequency (VLF) band (0.01–0.2 Hz) as the

principal frequency region of the baroreflexes. Thus, when

the baroreflex was intact, the APV was well constrained by

the negative feedback mechanism of the baroreflex in the

VLF region. However, when the constraint was removed

step-by-step via a partial and then a complete baroreflex

denervation, the VLF power increased progressively. In

addition, in this study, we found that, in contrast to the

VLF, LF power was decreased in the APV spectra after the

denervation, which is, again, consistent with our prior

findings. Previously, we (Dworkin et al. 2000b) measured

an overall baroreflex transportation lag of *1.05 s in the

NMB rat, which determined a low-frequency (LF) band of

0.2–0.6 Hz as the resonant (i.e., oscillation) region of the

baroreflexes. SAD interrupted the baroreflex system, thus

removed the oscillation in this closed loop negative feed-

back system, and resulted in a power decrease in the LF

band of the APV spectrum.

In conclusion, the present study suggests that (1) the

CNS has a significant contribution to the increased APV


123

observed following baroreflex denervation. Although

peripheral interference contributes, it is not necessarily

essential. (2) Compared with the carotid baroreflexes, the

aortic baroreflexes have a more dominant role in stabilizing

moment-to-moment APV. With the carotid sinuses intact,

only a single aortic limb was required to stabilize the

pressure.

Acknowledgments The authors would like to thank Dr. Sean

Stocker for his input for the manuscript, Dr. A. R. Travagli for his

significant editing of the manuscript, and Dr. Christopher Freet and

Mr. James Stoner for their help in fine tuning the manuscript.

References

Alper RH, Jacob HJ, Brody MJ (1987a) Central and peripheral

mechanisms of arterial pressure lability following baroreceptor

denervation. Can J Physiol Pharmacol 65:1615–1618

Alper RH, Jacob HJ, Brody MJ (1987b) Regulation of arterial

pressure lability in rats with chronic sinoaortic deafferentation.

Am J Physiol 253:H466–H474

Barres C, Lewis SJ, Jacob HJ, Brody MJ (1992) Arterial pressure

lability and renal sympathetic nerve activity are dissociated in

SAD rats. Am J Physiol 263:R639–R646

Burgess DE, Hundley JC, Li SG, Randall DC, Brown DR (1997)

First-order differential-delay equation for the baroreflex predicts

the 0.4-Hz blood pressure rhythm in rats. Am J Physiol

273:R1878–R1884

Casolo GC, Stroder P, Sulla A, Chelucci A, Freni A, Zerauschek M

(1995) Heart rate variability and functional severity of conges-

tive heart failure secondary to coronary artery disease. Eur Heart

J 16:360–367

Cerutti C, Barres C, Paultre C (1994) Baroreflex modulation of blood

pressure and heart rate variabilities in rats: assessment by

spectral analysis. Am J Physiol 266:H1993–H2000

Dampney RA, Taylor MG, McLachlan EM (1971) Reflex effects of

stimulation of carotid sinus and aortic baroreceptors on hindlimb

vascular resistance in dogs. Circ Res 29:119–127

Donald DE, Edis AJ (1971) Comparison of aortic and carotid

baroreflexes in the dog. J Physiol 215:521–538

Dworkin BR, Dworkin S (1990) Learning of physiological responses:

I. Habituation, sensitization, and classical conditioning. Behav

Neurosci 104:298–319

Dworkin BR, Dworkin S (1995) Learning of physiological responses:

II. Classical conditioning of the baroreflex. Behav Neurosci

109:1119–1136

Dworkin BR, Dworkin S (2004) Baroreflexes of the rat. III. Open-

loop gain and electroencephalographic arousal. Am J Physiol

Regul Integr Comp Physiol 286:R597–R605

Dworkin BR, Dworkin S, Tang X et al (2000a) Carotid and aortic

baroreflexes of the rat: I. Open-loop steady-state properties and

blood pressure variability. Am J Physiol Regul Integr Comp

Physiol 279:R1910–R1921

Dworkin BR, Tang X, Snyder AJ, Dworkin S et al (2000b) Carotid

and aortic baroreflexes of the rat: II. Open-loop frequency

response and the blood pressure spectrum. Am J Physiol Regul

Integr Comp Physiol 279:R1922–R1933

Eger EI 2nd, Xing Y, Laster M, Sonner J, Antognini JF, Carstens E

(2003) Halothane and isoflurane have additive minimum alveolar

concentration (MAC) effects in rats. Anesth Analg 96:1350–

1353 (table of contents)

Ferrari AU, Daffonchio A, Albergati F, Mancia G (1987) Inverse

relationship between heart rate and blood pressure variabilities in

rats. Hypertension 10:533–537

Frattola A, Parati G, Cuspidi C, Albini F, Mancia G (1993) Prognostic

value of 24-hour blood pressure variability. J Hypertens

11:1133–1137

Jacob HJ, Barres CP, Machado BH, Brody MJ (1988) Studies on

neural and humoral contributions to arterial pressure lability. Am

J Med Sci 295:341–345

Jacob HJ, Alper RH, Brody MJ (1989) Lability of arterial pressure

after baroreceptor denervation is not pressure dependent.

Hypertension 14:501–510

Jacob HJ, Ramanthan A, Pan SG, Brody MJ, Myers GA (1995)

Spectral analysis of arterial pressure lability in rats with

sinoaortic deafferentation. Am J Physiol 269:R1481–R1488

Julien C, Zhang ZQ, Barres C (1993) Role of vasoconstrictor tone in

arterial pressure lability after chronic sympathectomy and

sinoaortic denervation in rats. J Auton Nerv Syst 42:1–10

Julien C, Chapuis B, Cheng Y, Barres C (2003) Dynamic interactions

between arterial pressure and sympathetic nerve activity: role of

arterial baroreceptors. Am J Physiol Regul Integr Comp Physiol

285:R834–R841

Julius S, Nesbitt S (1996) Sympathetic overactivity in hypertension. A

moving target. Am J Hypertens 9:113S–120S

Krieger EM (1964) Neurogenic hypertension in the rat. Circ Res

15:511–521

Malpas SC (2002) Neural influences on cardiovascular variability:

possibilities and pitfalls. Am J Physiol Heart Circ Physiol

282:H6–H20

Mancia G, Parati G, Castiglioni P, di Rienzo M (1999) Effect of

sinoaortic denervation on frequency-domain estimates of baro-

reflex sensitivity in conscious cats. Am J Physiol 276:H1987–

H1993

Martinka P, Fielitz J, Patzak A, Regitz-Zagrosek V, Persson PB,

Stauss HM (2005) Mechanisms of blood pressure variability-

induced cardiac hypertrophy and dysfunction in mice with

impaired baroreflex. Am J Physiol Regul Integr Comp Physiol

288:R767–R776

Pelletier CL, Clement DL, Shepherd JT (1972) Comparison of

afferent activity of canine aortic and sinus nerves. Circ Res

31:557–568

Pickering AE, Simms AE, Paton JF (2008) Dominant role of aortic

baroreceptors in the cardiac baroreflex of the rat in situ. Auton

Neurosci 142:32–39

Pringle E, Phillips C, Thijs L, Davidson C, Staessen JA, de Leeuw

PW, Jaaskivi M, Nachev C, Parati G, O’Brien ET, Tuomilehto J,

Webster J, Bulpitt CJ, Fagard RH (2003) Systolic blood pressure

variability as a risk factor for stroke and cardiovascular mortality

in the elderly hypertensive population. J Hypertens 21:2251–

2257

Ramanathan A, Pan SG, Gauthier M, Jacob HJ, Brody MJ, Myers GA

(1994) Mechanisms regulating blood pressure in the conscious

rat: investigation by Wigner-Ville analysis. Comput Cardiol

317–319

Sanders JS, Ferguson DW, Mark AL (1988) Arterial baroreflex

control of sympathetic nerve activity during elevation of blood

pressure in normal man: dominance of aortic baroreflexes.

Circulation 77:279–288

Schreihofer AM, Sved AF (1994) Use of sinoaortic denervation to

study the role of baroreceptors in cardiovascular regulation. Am

J Physiol 266:R1705–R1710

Skeehan TM, Schuler HG, Riley JL (1995) Comparison of the

alteration of cardiac function by sevoflurane, isoflurane, and

halothane in the isolated working rat heart. J Cardiothorac Vasc

Anesth 9:706–712


123

Tang X, Dworkin BR (2007a) Baroreflexes of the rat. IV. ADN-

evoked responses at the NTS. Am J Physiol Regul Integr Comp

Physiol 293:R2243–R2253

Tang X, Dworkin BR (2007b) Baroreflexes of the rat. V. Tetanus-

induced potentiation of ADN A-fiber responses at the NTS. Am J

Physiol Regul Integr Comp Physiol 293:R2254–R2259

Tang X, Dworkin BR (2009) The dmNTS is not the source of

increased blood pressure variability in baroreflex denervated

rats. Auton Neurosci 148:21–27

Timmers HJ, Buskens FG, Wieling W, Karemaker JM, Lenders JW

(2004a) Long-term effects of unilateral carotid endarterectomy

on arterial baroreflex function. Clin Auton Res 14:72–79

Timmers HJ, Wieling W, Karemaker JM, Lenders JW (2004b)

Cardiovascular responses to stress after carotid baroreceptor

denervation in humans. Ann N Y Acad Sci 515–519:1018

Trapani AJ, Barron KW, Brody MJ (1986) Analysis of hemodynamic

variability after sinoaortic denervation in the conscious rat. Am J

Physiol 251:R1163–R1169

Van Vliet BN, Chafe LL, Montani JP et al (1999) Contribution of

baroreceptors and chemoreceptors to ventricular hypertrophy

produced by sino-aortic denervation in rats. J Physiol 516(Pt

3):885–895


123

neural control of arterial pressure variability in the neuromuscularly blocked rat

Documents