|
|
||||||||
CALL FOR PAPERS
Physiological and Molecular Mechanisms Implicated in the Neural Control of Circulation
Université de Lyon, Lyon, F-69008, France; Université Lyon 1, Lyon, F-69008, France, Centre National de la Recherche Scientifique FRE 3075
Submitted 29 February 2008 ; accepted in final form 30 April 2008
| ABSTRACT |
|---|
|
|
|---|
arterial pressure; baroreceptor reflex; sympathetic nervous system; transfer function
Recently, the concept has emerged that the characteristics of the baroreflex control of SNA might continuously change as part of normal autonomic responses accompanying natural behaviors (23), physical exercise (20), and exposure to emotional stress (13). All these data were obtained using the pharmacological method. Although this method is a valuable tool to describe the characteristics of the baroreflex function curves, it suffers from some important limitations. First, a proper use of the method requires stationary conditions, i.e., it should be applied only when changes in cardiovascular variables have reached a steady state. Secondly, its temporal resolution is low (several minutes or tens of minutes). For these reasons, the pharmacological method is not suitable for capturing dynamic and short-lasting changes in the functional characteristics of the sympathetic baroreceptor reflex. To circumvent these limitations, we have recently described a new method to assess the renal sympathetic baroreflex sensitivity (BRS) that does not necessitate the administration of vasoactive drugs. It is based on the computation of the gain of the transfer function relating RSNA and AP at heart rate (HR) frequency (24). The method has been validated in partially (aortic) baroreceptor denervated rats (14). One major advantage of the method is that its temporal resolution is about 1 min. Therefore, it provides a continuous estimate of sympathetic BRS and has already been used to show that the renal BRS exhibits minute-to-minute spontaneous variations (14).
The objective of the present study was to compare the baroreflex control of RSNA and LSNA in the same conscious rats. The lumbar nerve was chosen because it innervates mainly the skeletal muscle circulation of the hindlimb (1, 32). The majority of studies performed in humans are based on recordings of muscle SNA from the peroneal nerve in the leg. Comparing the baroreflex control of LSNA and RSNA is therefore clinically relevant. LSNA and RSNA were simultaneously recorded in conscious freely moving rats. Pharmacological estimates of the sympathetic BRS were obtained by means of the sequential administration of sodium nitroprusside and phenylephrine (7, 13). Spontaneous renal and lumbar BRS were assessed with the transfer function method (14).
| METHODS |
|---|
|
|
|---|
One day before the study, rats (n = 6) were anesthetized with isoflurane (2% in oxygen) and received an injection of penicillin G (50,000 IU sc). Femoral arterial and venous polyethylene catheters were then inserted. Rats were reanesthetized 4 to 6 h later, with pentobarbital sodium (60 mg/kg iv, supplemented with 10 mg/kg iv as needed) and received an injection of ketoprofen (5 mg/kg ip). The lumbar chains were accessed through a midline abdominal incision after gentle retraction of the abdominal aorta and vena cava. One of the chains was dissected and placed on a bipolar electrode at the L3-L4 level. The nerve-electrode preparation was embedded in a silicone gel and fixed to the surrounding tissue with
-cyanoacrylate glue, as described by Miki et al. (19). The electrode cable was sutured to the psoas muscle and guided subcutaneously to exit at the back of the neck at the same site as the catheters. The abdominal wall was then closed in layers and the rat was turned onto its right side. Through a flank incision, a branch of the left renal nerve was dissected, and a recording electrode was implanted as previously described in detail (7, 13, 14). The electrode cable was also exteriorized at the back of the neck. Both catheter ends and electrode plugs were protected in a small cap sewn to the skin. Each electrode had a separate ground wire that was left under the skin. Rats were allowed to recover from anesthesia and surgery during 16 to 18 h in the recording room before starting the study.
Data collection and experimental protocol. AP was measured using a pressure transducer (model TNF-R; Becton Dickinson, Sandy, UT) coupled to an amplifier (model 13-4615-52; Gould, Cleveland, OH). Both SNAs were band-pass filtered (300–3,000 Hz) and amplified with a gain of 100,000 (model P-511J; Grass, Quincy, MA). AP and SNA signals were sampled at 500 and 10,000 Hz, respectively, by a computer equipped with a 12-bit A/D converter (model AT-MIO-16; National Instruments, Austin, TX) and LabView 5.0 software (National Instruments).
After an acclimatization period, AP, LSNA, and RSNA were continuously recorded for
4 h. Thereafter, the baroreflex control of LSNA and RSNA was assessed in awake, resting animals by means of sequential intravenous injections of sodium nitroprusside (100 µg/kg) and phenylephrine (50 µg/kg/min for 1 min), as previously described (13). Finally, the mean level of LSNA and RSNA was recorded after administration of the ganglionic blocker chlorisondamine (2.5 mg/kg iv). As the lumbar and renal nerves were not cut distal to the electrode, both remaining SNAs incorporated afferent nerve activity, if present. On completion of the experiment, rats were euthanized with an intravenous overdose of pentobarbital sodium and the residual electrical activities were recorded for an additional 20-min period, which provided estimates of the background noise.
Off-line data analysis. SNA signals were full-wave rectified, low-pass filtered (cut-off frequency: 150 Hz), and resampled at 5,000 Hz with LabView 6i software. For each SNA, the corresponding background noise was subtracted from the time series. The spontaneous sympathetic BRS were calculated as previously described (14). AP, LSNA, and RSNA were resampled at 50 Hz and time series were segmented into 512-point (10.24 s) consecutive data sets. Coherence and transfer functions were calculated between AP (input signal) and SNA (output signal) over 11 segments overlapping by one-half (total duration of 61.4 s) using a fast Fourier transform algorithm. Autospectra of AP served to locate the frequency at which maximum power spectral density occurred in the frequency band encompassing HR (usually 5–8 Hz). For each SNA, coherence, gain, and phase were noted at this particular frequency. Only gain and phase values associated with a significant coherence (0.348 at P < 0.05) (2) were retained for further calculations. In each rat, the mean of the coherence and transfer gain and phase were calculated over a 219-min baseline period, which yielded a maximum of 213 values (in case all coherence values were significant). The mean LSNA and RSNA values were calculated over the same 219-min period and were used as reference values (set to 100%) for normalizing SNA data in spontaneous gain computations.
Pharmacological estimates of the sympathetic BRS were obtained by fitting a sigmoid function to AP-LSNA and AP-RSNA data pairs collected from the beginning of the nitroprusside-induced fall in AP up to the maximum phenylephrine-induced rise in AP (7). This was possible because each SNA followed similar trajectories during the falling and rising phases of AP, i.e., there was no apparent hysteresis in the sympathetic responses. AP, LSNA, and RSNA time series were resampled at 1 Hz by averaging over consecutive 1-s periods. For each SNA, data were normalized by their respective mean value calculated over the 2-min period preceding baroreflex testing (set to 100%). A four-parameter logistic function was then fitted to AP-SNA data pairs (SigmaPlot 2000; SPSS, Chicago, IL) by an interactive least-squares procedure: SNA = P1/{1 + exp[P2(AP-P3)]} + P4, where P1 is the full range of SNA variation, P2 is a slope coefficient, P3 is AP at one-half the SNA range, and P4 is the lower plateau of the curve. From these parameters, the upper plateau (P1 + P4), the threshold AP (P3 – 2/P2), the saturation AP (P3 + 2/P2), and the operating range (saturation pressure – threshold pressure) were derived (6, 13). Threshold and saturation AP values lay 11.9% (of the SNA range) below and above the upper and lower plateaus, respectively (18). The operating range indicates the range of AP over which SNA is responsive. The first derivative of the sigmoid function was computed to determine the baroreflex gain across the full range of AP variation, including the maximum gain (measured at P3) and the operational gain (measured at the reference AP, i.e., the 2-min average value before baroreflex testing). Only the absolute values of gain are presented.
Statistics. All data are presented as means ± SE. Statistical comparisons were performed using the Wilcoxon signed-rank test.
| RESULTS |
|---|
|
|
|---|
Pharmacological assessment of baroreflex control of regional SNAs. Fig. 1 shows an example of the reflex responses of LSNA and RSNA to drug-induced changes in AP. The sigmoid equation could be fitted satisfactorily to each set of AP-SNA data pairs, as indicated by highly significant, similar coefficients of determination (Fig. 1D and Table 1). Overall, LSNA appeared less reactive to AP changes than RSNA (Fig. 1). This was reflected in a lesser maximum reflex sympathoexcitation and a lesser maximum reflex sympathoinhibition (Fig. 2A), which resulted in a range-dependent decrease in the maximum gain for LSNA compared with RSNA (Fig. 2B and Table 1). The operational gain was also lower for LSNA than for RSNA. On the other hand, the operating range of baroreflex function curves did not differ between LSNA and RSNA (Table 1).
|
|
|
20 ms.
|
|
| DISCUSSION |
|---|
|
|
|---|
Methodological issues. Surgery for electrode implantation was performed in two steps. We did not succeed in chronically implanting both electrodes through a single approach, either abdominal or retroperitoneal. In addition, the recovery time from surgery was relatively short when considering the extensive surgery rats underwent. In our hands, the recording of both SNAs at 24 h postsurgery was possible in about one-half of the rats having acutely recordable signals. This percentage was drastically reduced at 48 h postsurgery. Notwithstanding these limitations, it should be noted that rats of the present study showed reasonably low HR values (<400 beats/min), which is a good indicator of an acceptable level of surgical stress, possibly owing to the administration of a long-acting nonsteroidal anti-inflammatory drug before surgery. It is important to note that this class of drugs has been reported not to impair the baroreceptor sensitivity in rats (28) or the sympathetic BRS in humans (21).
The normalization procedure is an important issue that should be considered when comparing results from different studies. For each SNA, we assigned the value 100% to a mean level recorded over 219 min under baseline conditions for spontaneous BRS computation, and during 2 min of rest for pharmacological baroreflex testing. Under these conditions, any spontaneous or evoked change reflected a variation around the mean. In a recent study performed on conscious rabbits (27), baroreflex function curves for lumbar and RSNA were compared, and the conclusion drawn was that BRS did not differ between the two SNAs. In this study, the background noise level was taken as the minimum SNA level observed during phenylephrine infusion, and data were normalized by the maximum value reached during sodium nitroprusside infusion. Consequently, the SNA range was constrained to 100% for both SNAs so that the gain only depended on the slope coefficient P2. In the present study, it was observed first that phenylephrine was far from eliminating discharges in LSNA, and secondly that sodium nitroprusside induced much more robust increases in RSNA than in LSNA (Fig. 1). With the normalization procedure used, the range of SNA variation was smaller for LSNA than for RSNA, which accorded with the visual inspection of raw data. As a consequence, and although the slope coefficient P2 did not differ between the two SNAs, the LSNA gain was markedly lower than the RSNA gain, which is in agreement with previous studies in anesthetized rats (29, 30).
Physiological issues. SNA was recorded from the lumbar chain at the L3-L4 level assuming that sympathetic nerves at this level mostly innervate blood vessels in skeletal muscles of the hindlimb. This assumption is supported by anatomical (1, 32) and functional (15, 19) studies. In particular, Miki et al. (19) found an inverse relationship between LSNA and iliac vascular conductance during the various stages of sleep in rats. An inverse relationship with triceps surae muscle vascular conductance was also found during the early stage of fictive locomotion induced by electrical stimulation of the mesencephalic locomotor region in decerebrate, paralyzed rats (15). However, it should not be ignored that a small proportion of cutaneous nerve fibers is present in the lumbar chain at this level (11, 31). Regarding the post- vs. preganglionic nature of LSNA, it is of note that the residual LSNA recorded after ganglionic blockade amounted to almost 10% of baseline activity, which would point to a significant contribution from preganglionic axons to overall LSNA, in accordance with a previous study (29). However, this contribution is probably overestimated because of the reflex increase in preganglionic SNA that was elicited by the large AP fall caused by ganglionic blockade. In the present study, it was indeed observed that, after chlorisondamine administration, residual LSNA could be almost halved by raising AP above baseline levels with a phenylephrine infusion (data not shown).
Baroreflex control of regional SNAs. LSNA showed a strong rhythmicity at HR frequency that was strongly correlated to the cardiac beat as shown by the high coherence values observed between AP and LSNA. These values were comparable to those observed between AP and RSNA, and thus allowed calculating a reliable spontaneous BRS index for both SNAs. This method allowed demonstrating that both lumbar and renal spontaneous BRSs fluctuate widely over time. These variations are probably dependent, at least in part, on the behavioral state of the rats (13, 19, 20, 23). Most of the time, spontaneous fluctuations of the renal and lumbar indices of sympathetic BRS were positively correlated. However, opposite variations were sometimes observed. This was particularly apparent in one rat that did not show a significant correlation between the two BRS indices. This was mainly due to outlying points that corresponded to situations when increases in LSNA gain were associated with decreases in RSNA gain. During these periods, it was observed that AP slightly increased, while HR decreased, LSNA slightly increased, and RSNA decreased. This particular pattern of cardiovascular changes has been shown to occur during rapid eye movement sleep in rats (19, 23).
Both methods for assessing sympathetic baroreflex control revealed a significantly lower BRS for LSNA than for RSNA, which is in agreement with earlier observations in anesthetized rats (29, 30). It is known that most, if not all, renal sympathetic nerve fibers are responsive to the stimulation of arterial baroreceptors (9, 25). On the other hand, the lumbar sympathetic chain contains a proportion of cutaneous nerve fibers that are poorly barosensitive (11). It is therefore possible that cutaneous SNA would lower overall BRS measured at the level of the lumbar chain.
Machado et al. (17) reported that electrical stimulation of the aortic depressor nerve produces a large reduction in hindlimb vascular resistance with little or no change in renal vascular resistance, thus suggesting that aortic baroreceptors play a predominant role in the regulation of hindlimb vascular resistance in rats. This is in apparent disagreement with the results of the present study, which point to a more powerful baroreflex control of RSNA than of LSNA. Although the reasons for this discrepancy are unclear, it might be suggested that sympathetic withdrawal is not the sole mechanism of baroreflex vasodilatations in the hindlimb circulation and that active phenomena are also involved. It has indeed been shown that the hindlimb circulation of the rat receives a nitroxidergic innervation (8) that might be, at least partly, under baroreflex control (26). During stimulation of the aortic depressor nerve, both sympathetic withdrawal and active neurogenic vasodilatation could act in concert, evoking stronger vasodilatory effects in the hindlimbs than in the kidney.
Relationships between SNA and vascular conductance are not necessarily identical between different vascular beds, i.e., a given change in SNA will not necessarily result in identical changes in vascular conductance. Pharmacological studies indicate that the hindquarter circulation of the conscious rat is less sensitive to the vasoconstrictor effects of exogenously administered norepinephrine than the renal circulation (4). It remains, however, that in conscious rats, both iliac and renal vascular conductance are inversely related to LSNA and RSNA, respectively (19, 33).
Perspectives and Significance
Owing to the simultaneous recording of LSNA and RSNA in conscious rats, and to the use of pharmacological and spontaneous approaches, we have shown that the sympathetic BRS is lower when assessed using LSNA than when using RSNA. In addition, the study shows that renal and lumbar BRS fluctuate widely over time and that these spontaneous fluctuations are mostly coordinated. The mechanism and physiological significance of these fluctuations, however, remains to be established. In particular, further studies are needed to clarify the effect of behavior on the baroreflex control of regional SNAs and its modulation.
| GRANTS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. Montano, R. Furlan, S. Guzzetti, R. M McAllen, and C. Julien Analysis of sympathetic neural discharge in rats and humans Phil Trans R Soc A, April 13, 2009; 367(1892): 1265 - 1282. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |