Skip to main content
PMC home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2004 Jun 11;558(Pt 3):975–983. doi: 10.1113/jphysiol.2004.064527

Acute shifts in baroreflex control of renal sympathetic nerve activity induced by REM sleep and grooming in rats

Satsuki Nagura 1, Tamaki Sakagami 1, Ai Kakiichi 1, Misa Yoshimoto 1, Kenju Miki 1
PMCID: PMC1665029  PMID: 15194739

Abstract

The present study aimed to determine the impact of REM sleep and grooming on the baroreflex stimulus–response curve for renal sympathetic nerve activity (RSNA). At least 3 days before study, Wistar female rats (n = 12) were chronically implanted with catheters to measure systemic arterial pressure (Pa) and to intravenously infuse vasoactive drugs. In addition, electrodes were placed for measurements of RSNA, electroencephalogram, trapezius electromyogram and electrocardiogram. The baroreflex curve for RSNA was determined by changing Pa using rapid intravenous infusions of phenylephrine and nitroprusside and then fitted to an inverse sigmoid function curve. REM sleep induced a vertical suppression of the Pa–RSNA baroreflex curve, which was characterized by significant decreases in the maximum response (by 72.0%, P < 0.05) and the maximum gain (by 4.02% mmHg−1, P < 0.05) compared with NREM sleep level. Grooming shifted the Pa–RSNA baroreflex curve upward and to the right, which was associated with increases in the maximum response (by 45.2%, P < 0.05), the minimum response (by 20.7%, P < 0.05) and the pressure at the centring point (by 11.1 mmHg, P < 0.05). These data suggest that the Pa–RSNA baroreflex curve was shifted acutely and differently in a state-dependent manner during natural sleep and wake cycle in rats.

The arterial baroreflex is the primary closed-loop feedback mechanism participating in short-term control of systemic arterial pressure (Pa). Baroreflex modulation of sympathetic nerve activity is believed to be critical in altering and stabilizing Pa. Changes of behavioural state affect both steady state and stability of Pa (Mancia & Zanchetti, 1980; Rowell et al. 1996). During non-rapid eye movement (NREM) sleep, Pa and heart rate (HR) are well controlled and quite stable, showing a constant level throughout NREM sleep (Sei & Morita, 1999). In contrast, the transition from NREM to REM sleep causes a significant increase in Pa in human (Somers et al. 1993) and rat (Mion & Krieger, 1988; Sei & Morita, 1999; Miki et al. 2003a), with a more labile Pa which is characterized by abrupt fluctuations, suggesting that the baroreflex regulation of sympathetic nerve activity becomes unstable during REM sleep (Junqeira & Krieger, 1976; Parmegianni & Morrison, 1990). However, little is known of how REM sleep modulates baroreflex control of sympathetic nerve activity. Similarly, exercise causes an increase in both the steady-state level and variability of Pa (Rowell et al. 1996). In contrast to REM sleep, exercise increases Pa associated with tachycardia (Combs et al. 1986; Krieger et al. 1998) and induces a sustained increase in sympathetic nerve activity (Matsukawa et al. 1991; O'Hagan et al. 1993; Miki et al. 2002, 2003b). Since parallel increases in Pa, HR and sympathetic nerve activity during exercise could not be explained by a single baroreflex curve (Combs et al. 1986), an acute shift in baroreflex control of sympathetic nerve activity has been hypothesized (Rowell et al. 1996). This hypothesis has been confirmed during exercise in men (Fadel et al. 2001; Kamiya et al. 2001) and rats (Miki et al. 2003b). These observations obtained during exercise impinge on the question of whether acute shifts in the baroreflex control of sympathetic nerve activity occur during REM sleep, raising the hypothesis that each behavioural state may be associated with a characteristic baroreflex control of sympathetic nerve activity. Nonetheless, little is known about state-related modulation of baroreflex control of sympathetic nerve activity during the sleep–wake cycle.

The aim of the present study was to generate the entire baroreflex stimulus–response curve for sympathetic nerve activity during the natural sleep–wake cycle in rats. To achieve this aim, rats were chronically instrumented for determination of vigilance state, cardiovascular function and renal sympathetic nerve activity (RSNA). Artificial changes in Pa were made by intravenous administration of vasoactive drugs during NREM, REM sleep and grooming state. The entire baroreflex curve for RSNA and HR was quantified by fitting the data to a logistic model. In this way, we studied the hypothesis of whether baroreflex control of RSNA could be altered in a state-dependent manner.

Methods

Animals

Experiments were performed on 12 adult female Wistar rats weighing 267.9 ± 2.8 g (mean ± s.e.m.). Animals were housed individually and kept in a temperature (24°C)- and humidity (60%)-controlled chamber with a 12 h: 12 h light–dark cycle (light on at 07.00 h) (Espec, Osaka, Japan). Food and water were available ad libitum. All procedures were in accordance with the Guiding Principles in the Care and Use of Animals in the Fields of Physiological Sciences published by the Physiological Society of Japan (Physiological Society of Japan, 2002) with the prior approval of the Animal Care Committee of Nara Women's University.

Instrumentation of animals

The animals were operated on in two stages. All procedures were performed aseptically in an operating theatre. The rats were anaesthetized with pentobarbital sodium (45 mg kg−1 i.p.). During the first surgical procedure, the electroencephalogram (EEG), electrocardiogram (ECG), and electromyogram (EMG) electrodes were implanted. EEG electrodes were implanted over the frontal cortex (anteroposterior +2 mm, mediolateral −2 mm from bregma), the parietal cortex (anteroposterior −3 mm, mediolateral −2 mm from bregma) and over the cerebellum (1.5 mm posterior to lambda). Three stainless steel miniature screws (1.0 mm diameter), which served as electrodes, were screwed into the skull and secured with dental cement. The bipolar EMG electrodes were implanted bilaterally in both trapezius muscles. The bipolar ECG electrode was implanted under the skin at the manubrium of the sternum and xiphoid process. The electrodes were exteriorized between the ears and passed through the centre of a circled-cut Ducron sheet, fixed into place and sutured to the skin.

At least 5 days after the first surgery, the catheters and electrodes were implanted as described in our previous reports (Miki et al. 2002, 2003b). Briefly, the arterial catheter was placed into the abdominal aorta via the tail artery. Two small venous catheters were placed into the superior vena cava via the common jugular vein. A pair of 60 cm polyethylene tubes (PE 10, Intramedic, Sparks, MD, USA) were tied and advanced into the superior vena cava so that the tips lay just above the right atrium. One of the venous catheters was used for the infusion of phenylephrine hydrochloride and the other for sodium nitroprusside infusion. RSNA was recorded from an implanted bipolar electrode. The left kidney was exposed retroperitoneally through a left flank incision, a branch of the renal nerve running on or beside the renal artery was carefully isolated, bipolar stainless steel wire electrodes were hooked onto the nerve and both were embedded in a two-component silicone rubber (see Miki et al. 2002 for full details). The electrodes and catheters were also exteriorized between the ears and protected by a plastic tube.

On completion of surgery, antibiotics were given intraperitoneally (Fradiomycin, Mochida-Seiyaku, Tokyo, Japan). A blanket and jelly were provided after the surgery. Animals were examined at least twice a day. The examination score included posture, activity, breathing, coat and eyes, body weight, food intake, urine volume, faecal volume. For the control of postoperative pain, a non-steroidal anti-inflammatory drug (diclofenac sodium, Voltaren; 0.5–3 mg kg−1, Novartis Japan, Tokyo, Japan) mixed with jelly was given orally when necessary. Arterial and venous catheters were filled with heparin sodium solution (1000 i.u. ml−1) and were flushed every day. The animals were housed individually in transparent plastic cages and allowed standard laboratory rat chow and water ad libitum thereafter. We terminated experiments if the body weight of the animal decreased by more than 20% of the pre-surgical body weight. In the present study, no such situation arose. At the end of the study protocol, rats were humanely killed using an intravenous overdose of pentobarbital sodium (> 200 mg kg−1).

Measurements

EEG, ECG, EMG and RSNA signals were amplified by a differential amplifier (MK-2, Biotech, Kyoto, gain: × 10 000 and bandwidth: 0.16–50 Hz for EEG; gain: × 1000 and bandwidth: 0.16–150 Hz for ECG; gain: × 100 and bandwidth: 100–2000 Hz for EMG; gain: × 10 000 and bandwidth: 150–2000 Hz for RSNA). Pa was measured by connecting the arterial catheters to pressure transducers (DX-100, Nihon Kohden, Tokyo). Heart rate (HR) was determined with a cardiotachometer (AT-601G, Nihon Kohden, Tokyo) triggered by the ECG. The amplified RSNA was integrated using a voltage integrator with a time constant of 100 ms (AD-600G, Nihon Kohden, Tokyo, Japan). The signals for ECG, EEG, EMG, Pa, HR, RSNA, and integrated RSNA were displayed continuously on an oscilloscope and recorded simultaneously on a thermal head paper recorder (ORP1200, Yokogawa, Tokyo, Japan) and a magnetic tape recorder (RX-8016, TEAC, Tokyo, Japan). The EEG, EMG, Pa, HR and integrated RSNA signals were sampled for analog-to-digital conversion at 1 ms intervals. The digitized EEG signal was Fourier analysed in 1 s epochs using a computerized data acquisition program (Visual Designer 4.0, Intelligent Instrumentation, Tucson, AZ, USA). The power spectrum was averaged simultaneously in two frequency bands: delta (0.5–4 Hz) and theta (6–9 Hz). The digitized EMG signal was simultaneously converted to the mean square root value. With the aid of the data acquisition program (Visual Designer 4.0), data for the power spectrums of EEG, Pa, the root mean square value of EMG, HR and integrated RSNA were averaged simultaneously and continuously every 1 s, displayed on the computer monitor, and stored on a hard disk.

Experimental protocols

After surgery, rats were allowed a minimum of 3 days to recover. Experiments were performed with the animals in their home cage, and they were given free access to food and water. Each experiment was performed over 2 or 3 h following an hour's rest after all electrodes, probes and catheters had been connected to a measuring instrument. The behaviour of the animal was monitored by the investigator through a small acrylic window of the chamber.

Behavioural patterns were classified as REM sleep, NREM sleep, and grooming by the investigators based on the EEG, EMG data displayed on the computer monitor and visual observation throughout the experimental period. During REM sleep, the EEG was characterized by the predominant theta activity and there was a dramatic suppression of the EMG. During NREM sleep, the EEG amplitude was larger and dominated by both delta and theta-frequency components and the EMG was low. Grooming behaviour was identified by visual observation.

To quantify baroreflex control of HR and RSNA, artificial perturbations to Pa were made by intravenous bolus infusions of phenylephrine hydrochloride (10 μg in 200 μl saline solution over 40 s) to increase Pa up to ∼170 mmHg, and sodium nitroprusside (10 μg in 200 μl saline solution over 40 s) to decrease Pa to ∼60 mmHg, during each state. Figure 1 depicts typical recordings of ECG, EEG, EMG, Pa, HR, RSNA and integrated RSNA obtained during REM sleep with pharmacological manipulations of Pa. Since the REM sleep periods lasted ∼30–90 s in most instances, the pharmacological manipulations were undertaken in separate REM episodes. The data for the RSNA and HR responses to the infusions of phenylephrine and nitroprusside were pooled for each experiment and then fitted to the sigmoidal logistic equation.

Figure 1. Typical recording from an individual rat of electrocardiogram (ECG), systemic arterial pressure (Pa), electroencephalogram (EEG), heart rate (HR), renal sympathetic nerve activity (RSNA) and integrated RSNA during pharmacological manipulation of Pa during transition between non-rapid eye movement (NREM) sleep and REM sleep.

Figure 1

Open in a new tab

Bolus intravenous infusions of phenylephrine (10 μg) and nitroprusside (10 μg) were given to generate the stimulus–response curve for RSNA and HR. Data are presented at two different recording speeds.

Data analysis

A logistic sigmoid function described by Kent et al. (1972) was used to analyse baroreflex curves:

graphic file with name tjp0558-0975-m1.jpg (1)

where Y is RSNA or HR, X is Pa, A1 is the response range for Y (maximum response minus minimum response), A2 is the gain coefficient, A3 is the pressure at the midrange of the curve (centring point), and A4 is the minimum response of Y. In each animal, Pa and RSNA or HR data were fitted to the logistic function to generate parameters A1, A2, A3 and A4 using graphics software (DeltaGraph, SPSS Chicago, IL, USA). We calculated the maximum response for HR and RSNA, saturation pressure for Pa (Pa,sat), threshold pressure for Pa (Pa,thr), operating range for Pa, and maximal gain according to the following equations (Potts et al. 1993; Miki et al. 2003b):

graphic file with name tjp0558-0975-m2.jpg
graphic file with name tjp0558-0975-m3.jpg
graphic file with name tjp0558-0975-m4.jpg

The maximum response is the upper plateau of the curve. Pa,thr and Pa,sat are the Pa at which HR or RSNA was within 5% of its maximum or minimum response, respectively. The operating range implies the range of Pa over which HR and RSNA responded.

To avoid the effects of uneven density of Y (HR or RSNA) axis data along the X (Pa) axis, all data were averaged over each 5.0 mmHg bin of Pa. Mean values of RSNA, HR and Pa within every 5.0 mmHg bin of Pa were used for curve fitting. Baroreflex response curves were constructed, and their parameters were calculated for each trial of the pharmacological manipulation of arterial pressure in each animal and then averaged across the animals. The averaged A1, A2, A3 and A4 were then used to generate average baroreflex curves (Figs 3 and 4) and first derivatives of the curves (Fig. 5).

Figure 3. Shifts in the baroreflex curves for renal sympathetic nerve activity (RSNA) obtained during rapid eye movement (REM) sleep, non-REM (NREM) sleep and grooming periods.

Figure 3

Open in a new tab

Curves reflect data averaged from 12 animals and symbols and error bars indicate mean ± s.e.m., respectively, estimated over each 5.0 mmHg bin of systemic arterial pressure (Pa).

Figure 4. Shifts in the baroreflex curves for heart rate (HR) obtained during rapid eye movement (REM) sleep, non-REM (NREM) sleep and grooming periods.

Figure 4

Open in a new tab

Curves reflect data averaged from 12 animals and symbols and error bars indicate mean ± s.e.m., respectively, estimated over each 5.0 mmHg bin of systemic arterial pressure (Pa).

Figure 5. First derivative of the baroreflex curves shown in Figs 3 and 4.

Figure 5

Open in a new tab

A, first derivative of systemic arterial pressure (Pa)–RSNA (renal sympathetic nerve activity) curve. B, first derivative of Pa–heart rate (HR) curve. Symbols indicate steady-state levels of Pa in Fig. 1, which were obtained before the pharmacological manipulations of Pa.

Statistical analysis

Statistical analysis was performed using analysis of variance (ANOVA) for repeated measures. When the F values were significant (P < 0.05), individual comparisons were made using the Fisher's least significant difference test (Sachs, 1982). Values are reported as means ± s.e.m.P < 0.05 was taken to indicate a significant difference.

Results

Figure 2 presents changes in mean values of Pa, HR and RSNA during REM sleep, NREM sleep and grooming. The mean values of Pa during REM sleep, NREM sleep and grooming periods were 112.6 ± 1.3, 107.9 ± 1.4, and 114.1 ± 2.1 mmHg, respectively. Pa increased significantly (P < 0.05) during both REM sleep and grooming compared with that during NREM sleep and there was no significant difference in Pa between REM sleep and grooming states. HR and RSNA increased significantly (P < 0.05) and progressively with an increase in physical activity. Mean values of HR during REM sleep, NREM sleep and grooming were 392.7 ± 3.1, 414.5 ± 3.8 and 464.5 ± 6.1 beats min−1, respectively. The mean values of RSNA during REM sleep, NREM sleep and grooming were 54.9 ± 6.3, 99.9 ± 7.4, and 146.8 ± 14.9%, respectively. Mean values of HR and RSNA were significantly different (P < 0.05) in all states.

Figure 2. Average level of systemic arterial pressure (Pa), heart rate (HR) and renal sympathetic nerve activity (RSNA) during rapid eye movement (REM) sleep, non-REM (NREM) sleep and grooming.

Figure 2

Open in a new tab

Each column and error bar represents mean +s.e.m. (n = 12). *P < 0.05.

REM sleep shifted the baroreflex curve for RSNA downwards (Fig. 3, Tables 1 and 2). This was characterized by significant decreases in the response range (A1, by 68.0%, P < 0.05), the maximum response (A1 +A4, upper plateau level, by 72.0%, P < 0.05), and the maximal gain (by 4.02% mmHg−1, i.e. 30.2%, P < 0.05; Fig. 5A) relative to the NREM level. The baroreflex curve for HR obtained during REM sleep was almost identical with that obtained during NREM sleep (Fig. 4).

Table 1.

Logistic model parameters describing baroreflex curve for RSNA and HR

Pa–RSNA reflex A1 (%) A2 (mmHg−1) A3 (mmHg) A4 (%)
REM sleep 136.6 ± 8.6* 0.253 ± 0.042 106.7 ± 2.2 7.6 ± 2.0
NREM sleep 204.7 ± 5.8 0.237 ± 0.026 105.4 ± 2.0 11.5 ± 2.1
Grooming 229.1 ± 17.7† 0.240 ± 0.020 116.5 ± 2.8*† 32.2 ± 8.7*†
Pa–HR reflex A1 (beats min−1) A2 (mmHg−1) A3 (mmHg) A4 (beats min−1)
REM sleep 176.6 ± 7.0 0.118 ± 0.009 124.3 ± 1.8 263.3 ± 6.0
NREM sleep 183.0 ± 6.4 0.116 ± 0.007 123.6 ± 1.5 268.9 ± 5.7
Grooming 159.7 ± 9.8* 0.167 ± 0.021*† 129.3 ± 2.3* 330.2 ± 8.4*†
Open in a new tab

Values are means ± s.e.m. (n = 12). Pa, systemic arterial pressure; RSNA, renal sympathetic nerve activity; HR, heart rate; A1, response range; A2, gain coeffient; A3, pressure at the midrange of the curve (centring point); A4, minimum response (lower plateau).

*

P < 0.05, vs. NREM sleep;

†

P < 0.05, grooming vs. REM sleep

Table 2.

Derived variables describing the baroreflex control of RSNA and HR

Pa–RSNA reflex Maximum response(%) Pa,thr(mmHg) Pa,sat(mmHg) Operating range(mmHg) Maximal gain(% mmHg−1)
REM sleep 144.2 ± 8.4* 95.0 ± 3.2 118.4 ± 2.8 23.4 ± 4.0 −8.25 ± 1.33*
NREM sleep 216.1 ± 6.2 94.6 ± 2.6 116.2 ± 2.5 21.6 ± 3.1 −12.27 ± 1.49
Grooming 261.3 ± 23.5*† 107.7 ± 3.2*† 125.4 ± 2.7* 17.8 ± 1.7 −13.69 ± 1.31†
Pa–HR reflex Maximum response(beats min−1) Pa,thr(mmHg) Pa,sat(mmHg) Operating range(mmHg) Maximal gain(beats min−1 mmHg−1)
REM sleep 440.0 ± 5.1 104.8 ± 2.5 143.9 ± 1.6 39.1 ± 2.2 −5.11 ± 0.37
NREM sleep 451.9 ± 4.3 103.8 ± 1.7 143.5 ± 2.0 39.7 ± 2.3 −5.20 ± 0.33
Grooming 489.9 ± 8.4*† 114.3 ± 2.5*† 144.3 ± 3.0 30.0 ± 3.1*† −6.69 ± 0.91*†
Open in a new tab

Values are means ± s.e.m. (n = 12). Pa, systemic arterial pressure; RSNA, renal sympathetic nerve activity; HR, heart rate; Pa,thr, threshold pressure; Pa,sat, saturation pressure.

*

P < 0.05, vs. NREM sleep;

†

P < 0.05, grooming vs. REM sleep.

Grooming shifted the baroreflex curve for RSNA to the right and upward relative to NREM sleep (Fig. 3, Tables 1 and 2); the upward shift was characterized by a significant increase in the maximum response (by 45.2%, P < 0.05), and the rightward shift was characterized by significant (all, P < 0.05) increases in the centring point of the reflex (by 11.1 mmHg), Pa,thr (by 13.0 mmHg) and Pa,sat (by 9.2 mmHg) relative to the NREM level.

The shift in the baroreflex curve for RSNA (Fig. 3, Tables 1 and 2) observed during grooming was greater than that during REM sleep, causing significant increases in the minimum response (by 24.7%, P < 0.05), the response range (by 92.5%, P < 0.05), the maximum response (by 117.1%, P < 0.05), Pa,thr (by 12.7 mmHg, P < 0.05), the centring point (A3, by 9.9 mmHg, P < 0.05) and the maximal gain (by 5.43 mmHg−1, i.e. 59.5%, P < 0.05; Fig. 5A).

Grooming shifted the baroreflex curve for HR upward relative to NREM sleep (Fig. 4, Tables 1 and 2); the parameters relating to the shift in the Y-axis, the maximum and minimum responses, increased by 38.0 beats min−1 (P < 0.05) and by 61.2 beats min−1 (P < 0.05), respectively, while the response range decreased by 23.2 beats min−1 (P < 0.05); as to the parameters relating to the shift in the X-axis, Pa,thr increased by 10.5 mmHg (P < 0.05) while Pa,sat did not change, resulting in a reduction of the operating range by 9.7 mmHg (P < 0.05); the gain coefficient and the maximal gain, which depend on the response range and operating range, increased significantly, by 0.051 mmHg−1 (i.e. 41.1%, P < 0.05) and 1.49 beats min−1 mmHg−1 (i.e. 31.6%, P < 0.05, Fig. 5B), respectively.

When the baroreflex curve for HR obtained during grooming was compared with the curves obtained during REM sleep (Fig. 4, Tables 1 and 2) and NREM sleep, the magnitude of the shift was similar. Thus there were significant increases in the minimum response (A4, by 66.8 beats min−1, P < 0.05) and the maximum response (by 49.9 beats min−1, P < 0.05), with no changes in the response range; Pa,thr increased by 9.5 mmHg (P < 0.05) but there were no changes in Pa,sat, resulting in the operating range being decreased by 9.0 mmHg (P < 0.05), which in turn caused an increase in the maximal gain by 1.58 beats min−1 mmHg−1 (i.e. 41.9%, P < 0.05, Fig. 5B).

Discussion

In the present study, we succeeded in generating the full range of the baroreflex curve for RSNA and HR during REM sleep, NREM sleep and grooming in rats. We demonstrated that during REM sleep the baroreflex curve for RSNA was shifted downward relative to NREM sleep while it was shifted right and upward during grooming relative to the NREM sleep. These acute shifts in the Pa–RSNA baroreflex curve would explain the state-dependent changes in the average level of RSNA during REM sleep, NREM sleep and grooming. Interestingly, the shifts in the Pa–RSNA baroreflex curve induced by changes in behavioural state were not simply correlated to the state-related changes in Pa.

Shift in baroreflex curve for RSNA

REM sleep results in a profound state-dependent modulation of Pa and HR, suggesting that sympathetic nerve activity plays a crucial role in modulating Pa and HR during this state (Baust et al. 1968; Futuro-Neto & Coote, 1982; Parmegianni & Morrison, 1990). Indeed, in the present study, the average level of RSNA decreased significantly by 42.0% during REM sleep relative to NREM sleep, which is consistent with previous report in cats (Baust et al. 1968; Futuro-Neto & Coote, 1982) and rats (Miki et al. 2003a). There is a lack of data on the full range of baroreflex curve parameters for RSNA during REM sleep. We demonstrated in the present study that REM sleep shifted the Pa–RSNA baroreflex curve downward, which could explain the decrease in the average level of RSNA during REM sleep. Because the entire baroreflex curve for Pa–RSNA during REM sleep was simply shifted downward without altering the minimum response or any parameters related to a horizontal shift (Figs 3 and 5A), the outcome was that RSNA during REM sleep was lower relative to NREM sleep when the systemic arterial pressure was below 110 mmHg.

By contrast, grooming shifted the Pa–RSNA baroreflex curve to the right and upward compared to the curve obtained during NREM sleep (Fig. 3), which is essentially consistent with previous reports during moderate exercise in humans (Fadel et al. 2001; Kamiya et al. 2001) and during treadmill exercise in the rat (Miki et al. 2003b). It is well established that sympathetic nerve activity increases during exercise in the rat (Miki et al. 2002, 2003b), rabbit (O'Hagan et al. 1993), cat (Matsukawa et al. 1991) and man (Fadel et al. 2001; Kamiya et al. 2001). In agreement with previous studies, we observed an increase in RSNA by 49.9% during grooming, which could be well explained by the right and upward shift of the Pa–RSNA baroreflex curve induced by grooming (Fig. 3). This shift in the Pa–RSNA baroreflex curve during grooming would cause the RSNA to increase at all levels of Pa compared with that during NREM sleep. Although, the increase in Pa occurring during grooming would act to suppress RSNA via the baroreflex, it was not sufficient to overcome the rise in RSNA caused by the acute shift of the Pa–RSNA baroreflex curve. This could explain the simultaneous increases in Pa and RSNA during grooming. Together, the above results demonstrated that baroreflex control of RSNA is most likely altered acutely in a state-dependent manner (Figs 3 and 5A), influencing the average level of RSNA during REM sleep, NREM sleep and grooming.

Shift of baroreflex curve for HR

The sensitivity of the baroreflex control of HR during REM sleep has been estimated by a linear regression analysis between Pa and HR, obtained during either spontaneous Pa and HR fluctuations or infusion of vasoactive drugs in rats (Zoccoli et al. 2001), cats (Knuepfer et al. 1986), and men (Nakazato et al. 1998; Legramante et al. 2003). The sensitivity of the baroreflex control of HR during REM sleep has been reported not to change (Nakazato et al. 1998; Zoccoli et al. 2001) or to decrease (Knuepfer et al. 1986). The present study successfully described the full range of the baroreflex curve for HR and demonstrated that the Pa–HR baroreflex curve generated in REM sleep was almost identical to that measured during NREM sleep (Figs 4 and 5B). These data indicate that the sensitivity of the baroreflex control of HR remains unchanged during REM sleep, and are consistent with previous reports in the rat (Zoccoli et al. 2001), where baroreflex-mediated fluctuations in HR (BRSP) did not change during REM sleep, and in the human (Nakazato et al. 1998), where the relationship between heart beat interval and systolic blood pressure remained unchanged during REM sleep. It is of interest that the operating pressure (112.6 mmHg) during REM sleep observed in the present study, which denotes an average level before pharmacological manipulation of Pa, was depressed to a point below the centring pressure (124.3 mmHg), where the gain is maximal (Fig. 5B). This may be one of the reasons why the slope of the regression line between Pa and HR obtained during a dynamic increase in Pa was different from that obtained during the descending phase of Pa, when Pa and HR fluctuate spontaneously (Legramante et al. 2003). If the baroreflex sensitivity for HR is estimated using a linear regression method, when Pa changes within only a part of the operating range, the estimated slope (baroreflex sensitivity) would be different, even though the full range of the Pa–HR baroreflex curve remains unchanged. This may explain, in part, the inconsistent reports on the sensitivity of baroreflex control for HR during REM sleep in previous reports (Knuepfer et al. 1986; Nakazato et al. 1998; Zoccoli et al. 2001; Legramante et al. 2003).

State-dependent modulation of baroreflex control of RSNA and cardiovascular function

It was surprising that Pa increased significantly despite the reductions of RSNA and HR and the vertical suppression in the Pa–RSNA baroreflex curve which occurred simultaneously during REM sleep relative to NREM sleep. The mechanisms underlying the increase in Pa during REM sleep are not evident from the present study. However, one possible explanation could involve regional differences in sympathetic nerve activities that may occur during REM sleep. Indeed, this was first described by Futuro-Neto & Coote (1982), and was supported by our recent publication (Miki et al. 2003a). By contrast, the cause-and-effect relationship between RSNA, HR and the Pa–RSNA baroreflex curve during grooming seems to be relatively simple compared with that during REM sleep. It is likely that the central neural processing leading to grooming behaviour and/or afferent input originating from muscle mechano- and chemoreceptors could exert excitatory influences on the baroreflex control of RSNA, resulting in the right–upward shift of the Pa–RSNA baroreflex curve. This in turn would increase RSNA, possibly including cardiac and visceral sympathetic nerve activities and then cause an increase in cardiac performance and vasoconstriction of visceral and non-contracting muscle, leading finally to a rise in Pa. Although we have not measured baroreflex control of RSNA during other behavioural states, for example drinking, eating and sexual behaviour, it might be safe to extend the present results and conclude that the baroreflex control of RSNA is likely to be altered in state-dependent manner (Figs 3 and 5A).

Acknowledgments

The authors thank Drs Edward J. Johns (Department of Physiology, University College Cork, Ireland) and Craig G. Crandall (Department of Internal Medicine, University of North Texas Health Science Center, Dallas, USA) for their critical reading of the manuscript. This study was supported in part by grants-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology, Japan, by 'Ground-Based Research Announcement for Space Utilization' promoted by the Japan Space Forum, and by the Kitsuen Kagaku foundation of Japan.

References

  1. Baust W, Weidinger H, Kirchner F. Sympathetic activity during natural sleep and arousal. Arch Ital Biol. 1968;106:379–390. [PubMed] [Google Scholar]
  2. Combs CA, Smith OA, Astley CA, Feigl EO. Differential effect of behavior on cardiac and vasomotor baroreflex responses. Am J Physiol. 1986;251:R126–R136. doi: 10.1152/ajpregu.1986.251.1.R126. [DOI] [PubMed] [Google Scholar]
  3. Fadel PJ, Stromstad M, Hansen J, Sander M, Horn K, Ogoh S, Smith ML, Secher NH, Raven PB. Arterial baroreflex control of sympathetic nerve activity during acute hypotension: effect of fitness. Am J Physiol Heart Circ Physiol. 2001;280:H2524–H2532. doi: 10.1152/ajpheart.2001.280.6.H2524. [DOI] [PubMed] [Google Scholar]
  4. Futuro-Neto HA, Coote JH. Changes in sympathetic activity to heart and blood vessels during desynchronized sleep. Brain Res. 1982;252:259–268. doi: 10.1016/0006-8993(82)90393-6. [DOI] [PubMed] [Google Scholar]
  5. Junqeira FL, Jr, Krieger EM. Blood pressure and sleep in the rat in normotension and in neurogenic hypertension. J Physiol. 1976;259:725–735. doi: 10.1113/jphysiol.1976.sp011491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Kamiya A, Michikami D, Fu Q, Niimi Y, Iwase S, Mano T, Suzumura A. Static handgrip exercise modifies arterial baroreflex control of vascular sympathetic outflow in humans. Am J Physiol Regul Integr Comp Physiol. 2001;281:R1134–R1139. doi: 10.1152/ajpregu.2001.281.4.R1134. [DOI] [PubMed] [Google Scholar]
  7. Kent BB, Drane JW, Blumenstein B, Manning JW. A mathematical model to assess changes in the baroreceptor reflex. Cardiology. 1972;57:295–310. doi: 10.1159/000169528. [DOI] [PubMed] [Google Scholar]
  8. Knuepfer MM, Stumpf H, Stock G. Baroreceptor sensitivity during desynchronized sleep. Exp Neurol. 1986;92:323–334. doi: 10.1016/0014-4886(86)90084-1. [DOI] [PubMed] [Google Scholar]
  9. Krieger EM, Brum PC, Negrao CE. Role of arterial baroreceptor function on cardiovascular adjustments to acute and chronic dynamic exercise. Biol Res. 1998;31:273–279. [PubMed] [Google Scholar]
  10. Legramante JM, Marciani MG, Placidi F, Aquilani S, Romigi A, Tombini M, Massaro M, Galante A, Iellamo F. Sleep-related changes in baroreflex sensitivity and cardiovascular autonomic modulation. J Hypertens. 2003;21:1555–1561. doi: 10.1097/00004872-200308000-00021. [DOI] [PubMed] [Google Scholar]
  11. Mancia G, Zanchetti A. Cardiovascular regulation during sleep. In: Orem J, Barnes CD, editors. Physiology in Sleep. New York: Academic Press; 1980. pp. 1–55. [Google Scholar]
  12. Matsukawa K, Mitchell JH, Wall PT, Wilson LB. The effect of static exercise on renal sympathetic nerve activity in conscious cats. J Physiol. 1991;434:453–467. doi: 10.1113/jphysiol.1991.sp018480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Miki K, Kato M, Kajii S. Relationship between renal sympathetic nerve activity and arterial pressure during REM sleep in rats. Am J Physiol Regul Integr Comp Physiol. 2003a;284:R467–R473. doi: 10.1152/ajpregu.00045.2002. [DOI] [PubMed] [Google Scholar]
  14. Miki K, Kosho A, Hayashida Y. Method for continuous measurements of renal sympathetic nerve activity and cardiovascular function during exercise in rats. Exp Physiol. 2002;87:33–39. doi: 10.1113/eph8702281. [DOI] [PubMed] [Google Scholar]
  15. Miki K, Yoshimoto M, Tanimizu M. Acute shifts of baroreflex control of renal sympathetic nerve activity induced by treadmill exercise in rats. J Physiol. 2003b;548:313–322. doi: 10.1113/jphysiol.2002.033050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Mion D, Jr, Krieger EM. Blood pressure regulation after deprivation of rapid-eye-movement sleep in rats. J Hyperten Suppl. 1988;6:S74–S76. doi: 10.1097/00004872-198812040-00019. [DOI] [PubMed] [Google Scholar]
  17. Nakazato T, Shikama T, Toma S, Nakajima Y, Masuda Y. Nocturnal variation in human sympathetic baroreflex sensitivity. J Auton Nerv Syst. 1998;70:32–37. doi: 10.1016/s0165-1838(98)00024-1. [DOI] [PubMed] [Google Scholar]
  18. O'Hagan KP, Bell LB, Mittelstadt SW, Clifford PS. Effect of dynamic exercise on renal sympathetic nerve activity in conscious rabbits. J Appl Physiol. 1993;74:2099–2104. doi: 10.1152/jappl.1993.74.5.2099. [DOI] [PubMed] [Google Scholar]
  19. Parmegianni PL, Morrison AR. Alterations in autonomic functions during sleep. In: Loewy AD, Spyer KM, editors. Central Regulation of Autonomic Functions. New York: Oxford University Press; 1990. pp. 367–386. [Google Scholar]
  20. Physiological Society of Japan. Guiding principles for the care and use of animals in the field of physiological sciences. J Physiol Soc Japan. 2002;64:140–146. [PubMed] [Google Scholar]
  21. Potts JT, Shi XR, Raven PB. Carotid baroreflex responsiveness during dynamic exercise in humans. Am J Physiol Heart Circ Physiol. 1993;265:H1928–H1938. doi: 10.1152/ajpheart.1993.265.6.H1928. [DOI] [PubMed] [Google Scholar]
  22. Rowell LB, O'Leary DS, Kellogg DL. Integration of cardiovascular control systems in dynamic exercise. In: Rowell LB, Shepherd JT, editors. Handbook of Physiology, Exercise: Regulation and Integration of Multiple Systems. Oxford UK: Oxford University Press; 1996. pp. 770–838. [Google Scholar]
  23. Sachs L. Applied Statistics. New York: Springer-Verlag; 1982. [Google Scholar]
  24. Sei H, Morita Y. Why does arterial blood pressure rise actively during REM sleep. J Med Invest. 1999;46:11–17. [PubMed] [Google Scholar]
  25. Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic-nerve activity during sleep in normal subjects. N Eng J Med. 1993;328:303–307. doi: 10.1056/NEJM199302043280502. [DOI] [PubMed] [Google Scholar]
  26. Zoccoli G, Andreoli E, Bojic T, Cianci T, Franzini C, Predieri S, Lenzi P. Central and baroreflex control of heart rate during the wake-sleep cycle in rat. Sleep. 2001;24:753–758. [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES