We have explored the possibility that renal sympathetic nerve activity (RSNA) and vasomotor sympathetic nerve activity are differentially regulated. We measured sympathetic nerve activity (SNA) to the kidney and the hind limb vasculature in seven conscious rabbits 6–8 days after the implantation of recording electrodes. Acute infusion of NG-nitro-l-arginine methyl ester (l-NAME) (6 mg·kg−1·min−1 for 5 min) led to an increase in blood pressure (from 66 ± 1 to 82 ± 3 mmHg) and a decrease in heart rate (from 214 ± 15 to 160 ± 13 bpm). l-NAME administration caused a significantly greater decrease in RSNA than lumbar sympathetic nerve activity (LSNA) (to 68 ± 14% vs. 84 ± 4% of control values, respectively). Volume expansion (1.5 ml·kg−1·min−1) resulted in a significant decrease in RSNA to 66 ± 7% of control levels but no change in LSNA (127 ± 20%). There was no difference in the gain of the baroreflex curves between the LSNA and RSNA [maximum gain of −7.6 ± 0.4 normalized units (nu)/mmHg for LSNA vs. −7.9 ± 0.75 nu/mmHg for RSNA]. A hypoxic stimulus (10% O2 and 3% CO2) led to identical increases in both RSNA and LSNA (195 ± 40% and 158 ± 21% of control values, respectively). Our results indicate tailored differential control of RSNA and LSNA in response to acute stimuli.
- sympathetic activity
- blood pressure
early concepts of autonomic control suggested a global uniform activation of sympathetic nerve activity (SNA) to all organs in response to a “fight or flight” stimulus (3). The general convention was to refer to SNA as if the same signal travels to all organs, but this has been replaced with a more organized model that emphasizes differential control of sympathetic outflow to functionally specific targets (4, 22). Sympathetic nerve activity appears to be elevated in hypertension (9, 10), and half of the increase can be accounted for by increased norepinephrine spillover solely from the heart and the kidney (7, 8). The selective increase in cardiac and renal sympathetic nerve activity (RSNA) indicates the importance of studying the differential control of nerve activity to better understand the control of blood pressure. The majority of evidence supporting differential control of SNA has come from anesthetized experiments in which sympathetic outflow to different organs in response to stimuli, such as volume expansion and baroreceptor stimulation, were measured.
Studies in anesthetized rabbits have shown that volume expansion leads to an inhibition of RSNA (32). This reduction in RSNA with volume expansion appears to be differentially regulated compared with SNA to other nonrenal organs. Although volume expansion is also accompanied by a decrease in lumbar SNA (LSNA), this decrease appears to be differentially regulated compared with the RSNA response (33), although it must be mentioned that the two nerve activities were recorded in two separate sets of animals. It is pertinent to note that both of these studies were conducted in anesthetized rabbits. Recent studies have indicated that anesthesia can influence mean levels of SNA (31), dampen reflex responses (23), and alter the cardiovascular responses to various stimuli. If the resting levels of nerve activity in the above studies were influenced by anesthesia, it is possible that the differential responses observed may be a factor of the anesthesia as opposed to a real differential response in SNA. It is important that the differential responses in nerve activity to volume expansion are reassessed in a conscious animal setting, where the animals have recovered from the surgery and effects of stress are minimal.
In addition to volume expansion, researchers have also explored the differential control of SNA to baroreflex and chemoreflex stimulation. As indicated earlier, the level of anesthesia and stress state of the animal can influence the mean levels of nerve activity, and hence, it is important that experiments involving SNA recordings to two organs are conducted in the same animal. An inability to directly record SNA to different organs for a long period of time has meant that the majority of studies have been conducted in anesthetized animals. These studies examining differential control of nerve activity have been conducted in anesthetized rat preparations and support differential control of nerve activities in response to baroreceptor stimulation (5, 29, 35). The mean level of nerve activity recorded varies depending on the nature of the contact between the nerves and the electrode. This has necessitated nerve activity levels being regularly expressed as a percentage of the control resting levels. Given that anesthesia can influence the mean resting levels of nerve activity and dampen reflex responses (23), it is likely that the differential responses observed in response to baroreceptor stimulation are confounded by the anesthetized state of the animals. It is important that a conscious animal preparation be used to examine the differential control of SNA, and we have explored the possibility that RSNA and vasomotor SNA are differentially regulated.
To better understand differential control of nerve activity, it is also important that different stimuli be used, as some nerves can have a differential response to one set of stimuli, but the same nerves could have similar responses to other stimuli. Although volume expansion is associated with a decrease in RSNA, chemoreceptor stimulation is associated with an increase in RSNA in the anesthetized rabbit preparation (13). We are not aware of any studies that have examined the response in LSNA to hypoxia to indicate whether the response in RSNA is differentially modulated. Another aim of our experiments was to examine the differential control of SNA to chemoreceptor stimulation in a conscious animal preparation.
Recent evidence has indicated an important interaction between SNA and nitric oxide (NO) in modulating the mean levels of arterial pressure (30). It is unclear whether NO differentially modulates SNA to different organs. The only previous study exploring this question was conducted in an anesthetized rat preparation. Hirai et al. (12) observed similar reductions in both RSNA and LSNA with inhibition of NO in baroreceptor-intact animals. These responses in nerve activity have not been confirmed in a conscious animal preparation. As we have suggested previously, the presence of anesthesia can affect mean levels of SNA, and it is important to verify the differential responses in a conscious animal setting. To explore the possibility that RSNA is selectively regulated, we have measured RSNA and LSNA responses to different acute stimuli in a conscious animal setting. We hypothesized that different acute stimuli would be accompanied by tailored differential responses in RSNA and vasomotor SNA.
Experiments were conducted with New Zealand White rabbits that had initial weights of 2.2–3.5 kg and were reviewed and approved by the University of Auckland Animal Ethics Committee. The rabbits were housed individually in cages (height 45 cm, width 72 cm, and depth 72 cm). The rabbits were fed daily (100 g standard rabbit pellets, supplemented with hay, carrots, and apples) at 0900, and water was available ad libitum. The room was kept at a constant temperature (18°C) and light-dark cycle (lights on from 0600 to 1800).
Anesthesia was induced using intravenous administration of propofol (Diprivan, 10 mg/kg) followed by intubation and then maintenance with halothane. Nerve recording electrodes were implanted to record RSNA and LSNA. Implantation of electrodes to both regions was conducted in the same surgery. Procedures to implant the renal nerve electrodes have been described in detail previously (18). Briefly, a retroperitoneal incision was used to expose the left kidney, and the renal nerve was identified and placed within a pair of coiled electrodes. The electrode wires and nerve were then coated in a silicone elastomer (Kwik-sil, World Precision Instruments, Sarasota, FL). The lumbar nerves were also approached via a left flank incision and a retroperitoneal approach. The lumbar nerves in the vicinity of the 3rd to 5th lumbar vertebrae were identified and placed within a pair of coiled electrodes. As with the renal nerve, the electrode and nerves were coated in a silicone elastomer. The free ends of the electrode wires were tunneled to the back of the neck and buried subcutaneously. All incisions were closed, and the animal was then allowed to recover. After surgery the rabbits were treated prophylactically with an antibiotic (enrofloxacin, Baytril, Bayer, New Zealand; 5 mg/kg sc daily for 5 days) and analgesic (ketoprofen, Ketofen, Rhone Merieux, Essex, UK; 2 mg/kg sc daily for 3 days). As soon as the rabbits regained consciousness, they were returned to their home cages. A heating pad was placed in the cage for 24 h after the surgery.
At least 6 days were allowed for the animals to recover from the surgery and resume normal eating and drinking habits. When normal eating and drinking patterns were established and the weights of the animals had returned to that previous to surgery, the acute experiments were performed. On the day of the experiment the rabbit was placed in a small box, and catheters were inserted under local anesthesia into a central ear artery and connected to a pressure transducer for continuous arterial pressure measurements and into the marginal ear veins for administration of drugs. The renal and lumbar nerve electrode wires were exteriorized and connected to the recording equipment. RSNA and LSNA was amplified, filtered between 50 and 5,000 Hz, full-wave rectified, and integrated using a low-pass filter with a time constant of 20 ms. Once preparation was complete, rabbits were left for at least 30 min before starting the experiment. The animals were subjected to acute interventions over two separate days at least 48 h apart. On the first day, baroreflex responses and an acute infusion of NG-nitro-l-arginine methyl ester (l-NAME) were carried out. On the second day, the animals were subjected to hypoxia and volume expansion. Appropriate time control experiments were conducted to ensure that the SNA data collected were stable.
Experimental Protocol for Day 1
After collecting a 5-min period of baseline recording, baroreflex responses were determined in response to infusions of phenylephrine and sodium nitroprusside. Sodium nitroprusside (1 mg/ml) was slowly infused to reduce arterial pressure down to ∼45 mmHg at a rate of 0.5–1 mmHg/s; all variables were then allowed to return to baseline before phenylephrine (1 mg/ml) was infused to raise arterial pressure at a rate of 0.5–1 mmHg/s to between 120 and 140 mmHg (when SNA was silent). These sequences were repeated at least three times.
Acute blockade of NO.
Once recorded variables had returned to resting levels, a 5-min period of control data was collected before acute administration of l-NAME was carried out. l-NAME was infused at a dose of 6 mg·kg−1·min−1 (in 3 ml saline) for 5 min through a marginal ear vein. Once the l-NAME was administered, a further 15 min of data was collected. The rabbit was then returned to its home cage.
Experimental Protocol for Day 2
On a separate day, at least 48 h after the above protocols, the animals were prepared as described above and exposed to hypoxia and volume expansion.
The rabbit was placed within a box in a sealable Perspex chamber (volume: 30 liters). The animals were allowed to rest inside the Perspex chamber, which was initially open to room air for 10 min, while control baseline data were collected. Once baseline data were collected, the chamber was closed to room air and then perfused through a small hole with a gas mixture of 10% O2 and 3% CO2 (makes a normocapnic hypoxic mixture) delivered at 10 liters/min for 15 min. This was followed by a further 5 min of recovery data when the chamber was exposed to room air. Arterial blood samples (0.3 ml) were taken before and at the end of the hypoxic period.
Once the response to hypoxia was recorded and variables had returned to baseline resting levels, a further 15 min was allowed before commencing the volume expansion protocol. A polygeline/electrolyte solution [Gelofusine (composition; succinylated gelatin, sodium chloride, and sodium hydroxide: osmolality 283 mosmol/kgH2O), Health Support, Auckland, New Zealand] was used to increase plasma volume. This was administered at room temperature at a rate of 1.5 ml·min−1·kg−1 for 15 min (total volume infused was 55 ± 3 ml). Data continued to be collected for 15 min after the infusion was stopped.
All data were sampled at 500 Hz using an analog-to-digital data acquisition card (AT-MIO64E-3 National Instruments, Austin, TX). All subsequent data analysis was performed using a data acquisition program (Universal acquisition and analysis, ver. 11; Telemetry Research, Limited, Auckland, New Zealand). The 2-s averages of mean arterial pressure (MAP), LSNA, and RSNA during the baroreflex curves were collected, and a general nonlinear regression program was used (26) to fit the collected MAP-LSNA and MAP-RSNA data to a sigmoidal logistic function to produce baroreflex curves. The program uses a five-parameter nonlinear regression equation to produce the resultant baroreflex curves. For the baroreflex curves, the sympathetic nerve values were normalized with the upper plateau with sodium nitroprusside being 100%. The noise levels were taken to be the integrated value when blood pressure was high with phenylephrine and no bursts were evident in the raw SNA signal.
Apart from the baroreflex curves, all RSNA and LSNA values were normalized as a percent of the resting nerve activity recorded during the control period of each protocol. All data were analyzed using ANOVA. The sum of squares was completely partitioned to account for all the variability in the data. Data are shown as the means ± SE. P values < 0.05 were considered significant.
Original neurograms from RSNA and LSNA obtained under control conditions revealed a high degree of similarity between the bursting properties of each nerve activity (Fig. 1). Direct voltages cannot be compared between nerves because of the variable nature of the contact between the nerve and recording electrode and the lumbar nerve being, in general, slightly smaller than the renal. However, it was apparent that the individual frequency of bursts was very similar within each animal under control conditions.
Responses to Blockade of Endogenous NO
Both RSNA and LSNA displayed a significant decrease following l-NAME administration. However, the decrease in RSNA was significantly greater than the decrease in LSNA; reaching 68 ± 14% of control values 10 min after the infusion ended vs. 84 ± 4% for LSNA (P < 0.05). l-NAME administration led to a gradual increase in MAP from 66 ± 1 mmHg during control levels to 82 ± 3 mmHg 10 min after the end of the infusion (P < 0.05; Fig. 2). This was accompanied by a decrease in heart rate from 214 ± 15 to 160 ± 13 bpm 10 min after the end of the infusion.
Steady-State Changes During Volume Expansion
Volume expansion resulted in a significant decrease in RSNA (P < 0.05) but no change in LSNA (Fig. 3). In the 5-min period after the infusion, RSNA was 78 ± 10% of control levels, whereas LSNA was 110 ± 22% of control levels. RSNA decreased further to 66 ± 7% (maximum decrease of 43% and minimum decrease of 95% of control values), whereas LSNA was 127 ± 20% of control levels (there was a decrease in one animal to 87% of control) in the 10- to 15-min period after infusion was stopped (P < 0.05 vs. each other). Arterial pressure was unchanged (66 ± 1 mmHg), but heart rate was significantly increased by 18 ± 5 bpm (P < 0.05).
Baroreflex determination using rapid infusions of sodium nitroprusside and phenylepineprhine revealed that responses in the average levels of RSNA and LSNA were the same. When the levels of SNA were reflected as a percentage of the maximum SNA obtained during sodium nitroprusside infusion, it was apparent that the mean baroreflex curve was not significantly different between the nerves (Fig. 4 and Table 1) with a maximum gain of −7.6 ± 0.4 nu/mmHg for LSNA vs. −7.9 ± 0.75 nu/mmHg for RSNA.
Hypoxia (10% O2 and 3% CO2) caused PaO2 to fall from 100 ± 2 to 50 ± 5 mmHg. Significant increases in both LSNA and RSNA was evident in all of the animals (Fig. 5). However, the RSNA and LSNA responses were not different from each other. RSNA averaged 195 ± 40% (maximum of 290% and minimum of 110%) and LSNA averaged 158 ± 21% of control values (maximum of 210% and minimum of 120%) during the last 5 min of the 15-min hypoxia period. Both nerve activities returned back to normal when hypoxia was ceased. Hypoxia had no significant effect on either MAP (66 ± 2 mmHg during control) or heart rate (204 ± 17 mmHg during control).
We have investigated the reflex control of both RSNA and LSNA and have shown that RSNA and LSNA are differentially regulated in response to certain stimuli (volume expansion and blockade of endogenous NO). However, there is no difference in the response of RSNA and LSNA to baroreceptor and chemoreceptor stimulation, suggesting differential control of nerve activity does not occur in response to all stimuli. Our simultaneous recordings of RSNA and LSNA in conscious rabbits extend observations from previous studies conducted in anesthetized animals.
Differential Responses to Increased Blood Volume
Our results showing that RSNA is particularly sensitive to volume expansion is consistent with the conclusions drawn from previous studies conducted in anesthetized conditions. Studies in anesthetised rabbits have shown that volume expansion is accompanied with a decrease in RSNA (32). Studies in conscious rabbits have confirmed the decrease in RSNA with volume expansion (2). The reduction in RSNA seen in our study is similar to the one seen by Badoer et al. (2) in their experiment with conscious rabbits (50%). Whether volume expansion also results in a decrease in LSNA in conscious animals was not determined previous to this study. We have simultaneously shown that whereas RSNA decreased, LSNA was unchanged with volume expansion. This is in contrast to a previous study in anesthetized rabbits in which the authors observed a decrease in LSNA with volume expansion (33). The rate of volume expansion and the final volume administered in this study was comparable to our study. Volume expansion in both the previous anesthetized preparation studies (32, 33) was accompanied by an increase in MAP in contrast to our study that showed no change in arterial pressure. The increase in MAP would lead to a baroreflex-mediated decrease in nerve activity in these studies, which might help explain the decrease in LSNA with volume expansion. Indeed, sinoaortic denervation in these groups of animals greatly attenuated the LSNA response but did not affect the RSNA response, suggesting that the decrease in LSNA was primarily baroreflex mediated (32, 33). This might help explain the differences in LSNA responses between our study and the previous study conducted in anesthetized conditions. It is also unclear whether the anesthetized condition of the rabbit played a role in the different LSNA responses observed. The reduction in RSNA with volume expansion appears to be mediated by cardiopulmonary afferents as a section of the cardiopulmonary afferents in anesthetized rabbits abolished the response (32). Administration of intrapericardial procaine to block cardiac afferents in conscious rabbits also blocks the sympathoinhibition in RSNA to volume expansion (2). Our results suggest control of LSNA is not governed to any large extent by cardiopulmonary afferent activity, which supports findings from previous studies conducted in the anesthetized rabbit (33). Our results indicate volume expansion is accompanied by a tailored differential decrease in RSNA compared with no change in LSNA.
This differential inhibition of RSNA compared with LSNA does not appear to be confined to the rabbit. Weaver (36) showed in an anesthetized cat preparation that intravascular volume expansion inhibited RSNA but has no effect on LSNA. The inhibition in RSNA was confirmed in a conscious cat preparation by Schad and Seller (28). Karim et al. (17) showed in an anesthetized dog preparation that left atrial distension is accompanied with a decrease in RSNA but no change in LSNA. This differential SNA response in dogs has not been confirmed in a conscious preparation yet. It appears that in the rabbit, the cat and the dog, volume expansion is accompanied by a decrease in RSNA but no change in LSNA.
Differential Response to Blockade of Endogenous NO
Administration of l-NAME was accompanied by a significant decrease in both RSNA and LSNA (68% and 84% of control, respectively). The only previous study that examined the differential response in RSNA and LSNA to NO inhibition was conducted in an anesthetized rat preparation. Hirai et al. (12) examined the differential responses of lumbar and RSNA to NO inhibition and observed a reduction in both renal (−45% of control) and LSNA (−35% of control) with l-NAME administration. In contrast to our findings, they did not observe any significant differences between the responses in RSNA vs. LSNA. It is unclear whether the differences between the studies are due to the different species involved or the anesthetized condition of the rat preparation. The l-NAME administration in the rat study was also less (20 mg/kg) compared with our study (30 mg/kg), which may be another reason for the differences observed. We acknowledge that comparison between studies where l-NAME administration is used is further made difficult because of the complex central, as well as peripheral effects of NO on SNA (11, 34, 38).
One of the novel aspects of our study is that we have determined the baroreflex and NO effects on RSNA and LSNA in the same animal. We believe much of the decrease in SNA with l-NAME administration can be attributed to the arterial baroreflex. l-NAME caused blood pressure to increase by 16 mmHg; using data collected from baroreflex determination (Fig. 4) indicates that such an increase in blood pressure would be expected to decrease nerve activity by 70% of control nerve activity. Given the actual change in RSNA was 68% of control, we suggest that the decrease in RSNA is most likely baroreflex mediated. In support of a baroreflex-mediated decrease in RSNA, Liu et al. (19) showed that the decrease in RSNA in response to NO blockade could be reversed by abolishing the change in blood pressure in the conscious rabbit. The decline in LSNA (84% of control) was less than expected if the arterial baroreflex was the only contributing factor. As RSNA and LSNA had identical baroreflex responses (Fig. 4), we suggest that NO differentially modulates the LSNA and RSNA response. Our results support the possibility that NO differentially modulates SNA to the kidney and the hind limb vasculature. It is difficult to extend our results indicating differential modulation of SNA by NO to chronic conditions because of a lack of a chronic stimulus in our studies. Clearly, further studies are needed to explore the possibility of differential control of RSNA with long-term changes in blood volume and NO levels.
Baroreflex and Chemoreflex Control of SNA
The use of a conscious animal setting in our study in which nerve activity to both the kidney and the hind limb region is measured at least 6 days after surgery is a major improvement from previous studies, as SNA responses can be affected by anesthesia (6, 31). Indeed, differences in the gain of the RSNA response to baroreceptor stimulation has been observed between conscious and anesthetized rabbits (15). Our findings show that baroreflex control of RSNA and LSNA are identical (Fig. 4 and Table 1). It is problematic to compare nerve activities between animals, and this necessitates recordings of SNA from different organs in the same animal. Very few studies have examined the differential control of SNA by recording more than two nerve activities in the same animal, and these studies have primarily been conducted in anesthetized rats. In contrast to our study, Scislo et al. (29) observed a steeper gain for the RSNA baroreflex curve compared with the LSNA baroreflex curve in the anesthetized rat. The conscious state of the animal in our study is an improvement on this previous study as the anesthetized condition of the rat might lead to a depressed baroreflex response and may account for the differences observed. Apart from the steeper gain for the RSNA baroreflex, the authors conclude that there is no difference between the functions of RSNA and LSNA in the range of MAP close to the operating point, which is in agreement with our findings. We cannot discount the possibility that in the rat, the gain of the RSNA baroreflex curve is steeper compared with the LSNA baroreflex curve. To date, differential control of LSNA vs. RSNA has not been examined in the conscious rat, and it is unclear whether the species difference or the difference in the anesthetized state of the animal accounts for the different gains observed. The only other study in anesthetized rats compared RNSA vs. adrenal SNA in the anesthetized rat and found similar gain values (35). The different nerves being compared makes it difficult to make meaningful comparisons between the studies.
One previous study has examined differential control of SNA in a conscious animal preparation (24, 25). These authors examined the baroreflex control of SNA to the heart and the kidney in a conscious cat preparation and reported a significant difference in the gain of cardiac and RSNA baroreflex responses. It is difficult to compare the results of this study with our study as different nerves were being compared, and the species being studied is also different. In addition, the experimental approach used to calculate gain in these studies was different from our study. The baroreflex gain in these conscious animals was calculated with injection of norepinephrine, and the resultant changes in blood pressure and nerve activity were recorded. As such, the gain obtained would represent the portion of the baroreflex curve above the resting point. Our study included both an increase and a decrease in blood pressure, and we believe this gives a truer representation of baroreflex gain.
Exposure of the animals to hypoxia led to a similar increase in both RSNA and LSNA, suggesting no difference between the two nerves in the response to chemoreceptor activation. Our results confirm previous studies in anesthetized (13) and conscious (20, 21) rabbits, which have indicated an increase in RSNA with chemoreceptor stimulation. We are not aware of any studies that have examined responses in LSNA to chemoreceptor stimulation previous to this study. Interestingly, arterial hypoxia leads to inhibition of cardiac SNA in the anesthetized rabbit (13, 14) and the rat, as evidenced by a decrease in cardiac norepinephrine turnover (16). Thus, although our results showed no differential response in the RSNA and LSNA response, it does appear that moderate chemoreceptor stimulation leads to differential inhibition of cardiac SNA.
We have observed differential regulation of RSNA and LSNA in response to volume expansion and l-NAME administration. Recordings of LSNA include a significant number of preganglionic neurons, whereas renal nerves consist of almost entirely postganglionic neurons. It is possible that the differential regulation seen is attributable to differences in modulation of ganglionic transmission and therefore recorded RSNA and LSNA. Volume expansion will lead to suppression of ANG II, and ANG II can modulate ganglionic transmission (27). NO levels have been shown to modulate ganglionic transmission as well, although some controversy still exists (37). It is possible that changes in ANG II and NO levels may have influenced ganglionic transmission, and we cannot discount this possibility as the reason for the differential responses seen.
Differential regulation of sympathetic nerve activity could play an important role in the long-term control of blood pressure. There is now strong evidence to indicate an increase in sympathetic nerve activity, which may play a critical role in the development of hypertension (1, 9, 10). Interestingly, half of the increase in total noradrenaline spillover seen in essential hypertension can be accounted for by increased noradrenaline spillover from the heart and the kidney (7, 8). This highlights the differential control of nerve activity in which increases in SNA to selected organs may be a critical factor in the development of hypertension. The finding that sympathetic nerve activity to different beds is differentially regulated in the short term raises the possibility of whether RSNA could be differentially controlled in the long term. It is equally possible that LSNA might be differentially increased in the long term, leading to increased peripheral resistance, which may maintain the high blood pressure. Clearly, further studies are needed to explore the importance of differential control of RSNA in the long-term control of blood pressure.
Research in the authors' laboratory was funded by the Auckland Medical Research Foundation, the Maurice and Phyllis Paykel Trust, and the Health Research Council of New Zealand.
The authors acknowledge the assistance of Fiona McBryde and Kris Hillock.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society