INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

FUTURO-NETO AND COOTE (13) demonstrated that in the decerebrate cat, electrical stimulation within a small region of the medullary raphe caudal to the obex results in a differential pattern of spinal sympathetic outflow. They noted a decrease in both postganglionic inferior cardiac (CN) and renal (RN) sympathetic nerve activity and an increase in the activity of sympathetic vasoconstrictor fibers to the gastrocnemius muscle. In another study (12) they observed a qualitatively similar pattern of changes in sympathetic nerve discharge (SND) after intravenous administration of physostigmine. These changes were accompanied by rapid eye movements (REM), muscle atonia, and pontogeniculooccipital spikes, suggestive of REM sleep. The observed changes in SND were consistent with the regional cardiovascular changes reported during REM sleep in the cat; that is, decreased heart rate (8, 12), decreased skeletal muscle blood flow (25), and increased blood flow to the kidney and mesentery (21). Futuro-Neto and Coote therefore termed the reciprocal changes in SND produced by electrical stimulation of the caudal medullary raphe or intravenous physostigmine as "REM-like." In a subsequent study, Yusof and Coote (29) reported that electrical stimulation within the caudal medullary raphe of the rat elicited a similar differential pattern of spinal sympathetic outflow.

Futuro-Neto and Coote and colleagues (9, 12, 13) did not investigate whether the changes in SND induced by caudal medullary raphe stimulation were restricted to a particular frequency band. In this context, pre- and postganglionic SND contains a variable mixture of cardiac-related and 10-Hz rhythms in cats with intact baroreceptor nerves (4, 7). Whereas the cardiac-related rhythm is thought to reflect the ability of pulse-synchronous baroreceptor afferent nerve activity to entrain irregular low-frequency oscillations in a 1:1 relationship to the cardiac cycle (14, 28), the 10-Hz rhythm is generated in the brain stem by a system of coupled oscillators, each of which preferentially controls a different portion of the spinal sympathetic outflow (15, 19). After baroreceptor denervation in the cat, the cardiac-related rhythm is replaced by irregular oscillations at frequencies 6 Hz, and the 10-Hz rhythm usually becomes stronger (2, 4).

The issue of whether the changes in SND in response to caudal medullary raphe stimulation are band specific lies at the heart of the current study. The rationale for addressing this problem is based on two studies from our laboratory that described a defenselike pattern of spinal sympathetic outflow produced by electrical stimulation of the midbrain periaqueductal gray (PAG) in urethan-anesthetized, baroreceptor-denervated cats (16, 17). During PAG stimulation, CN and RN activities increased, with a decrease in the activity of the vertebral nerve (VN), which carries vasoconstrictor fibers to the forelimb. These changes in SND are consistent with the changes in regional blood flows that occur during the defense response (18), which are in the opposite direction to those seen in REM sleep (8). The reciprocal changes in VN vs. CN and RN activities in response to PAG stimulation were restricted to the 10-Hz band and were accompanied by significant lengthening of the CN-VN phase angle and shortening of VN-RN phase angle at the frequency of peak coherence in the 10-Hz band. These responses were observed during high-frequency (25 Hz) stimulation or when the 10-Hz rhythm was entrained in a 1:1 fashion by stimulation at frequencies equal to or slightly above that of the free-running rhythm. However, during 1:1 entrainment at frequencies just below that of the free-running rhythm, an increase in 10-Hz power occurred in VN as well as CN and RN discharges. There was no significant change in either the CN-VN or the VN-RN phase angle in cases in which PAG stimulation uniformly increased the 10-Hz discharges of the three nerves. Under the assumption that the state of coupling of multiple brain stem 10-Hz oscillators is reflected by the phase angles relating the discharges of the postganglionic nerves that they target, Gebber et al. (16) proposed that changes in the state of coupling of the oscillators lead to differential patterns of spinal sympathetic outflow.

In the present study, we sought to examine whether electrical stimulation of the caudal medullary raphe elicits reciprocal changes in VN vs. CN and RN discharges using the same mechanism as that proposed for the defenselike pattern elicited by PAG stimulation. If such a mechanism were responsible for the differential pattern of spinal sympathetic outflow elicited by caudal medullary raphe stimulation, then the reciprocal changes in VN vs. CN and RN activities should be restricted to the 10-Hz band and be accompanied by changes in phase angle in the 10-Hz band. In addition, the pattern should be elicited by some frequencies of raphe activation that entrain the 10-Hz rhythm in a 1:1 relation to the stimuli, but not others.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental subjects and anesthesia. The protocols used on 16 cats of either sex (2.35-5.05 kg) were approved by the All-University Committee on Animal Use and Care of Michigan State University. The cats were initially anesthetized with 2.5-3.5% isoflurane in oxygen. Urethan (1.2-1.4 g/kg iv) was administered via a catheter in the femoral vein, and isoflurane inhalation was terminated. This dose range of urethan has been reported to maintain a surgical level of anesthesia for a period of 8-10 h (11), which exceeded the duration of our experiments. Blood pressure was measured from a catheter inserted into either the femoral or brachial artery. The cats were paralyzed and artificially ventilated with room air; a bilateral pneumothoracotomy was performed. End-tidal CO₂ was maintained between 3.5 and 4.0% by adjusting the parameters of ventilation. Rectal temperature was maintained near 38°C by using a heat lamp. Baroreceptor denervation was performed by bilateral sectioning of the carotid sinus, aortic depressor, and vagus nerves (4). Sectioning of these nerves eliminated the cardiac-related rhythm in SND and the inhibition of SND produced by raising blood pressure either with a bolus injection of norepinephrine or by abdominal aortic obstruction induced by inflation of the balloon-tipped end of a Fogarty embolectomy catheter (model 4F). In the later case, blood pressure was measured from a brachial artery. SND contained a variable mixture of the 10-Hz rhythm and irregular low-frequency (6 Hz) oscillations after baroreceptor denervation.

Neural recordings and central stimulation. As previously described (4), bipolar platinum electrodes were used to record monophasically from the central ends of the cut CN and VN near their exits from the left stellate ganglion and from the left RN. Nerve recordings were made with a band-pass filter set at 1-1,000 Hz (Grass Instruments 7P3 preamplifier), so that envelopes of multiunit spikes appeared as slow waves (4, 7). Data were stored on magnetic tape.

Data analysis. Frequency- and time-domain methods used in this study have been described in detail previously (16). Briefly, spectral analysis was performed by using fast Fourier transform (FFT) after SND had been low-pass filtered at 100 Hz (original recordings). The sampling rate of 200 Hz gave a resolution of 0.2 Hz/bin. The power spectra (autospectra) were averages of 29 5-s data windows, with 75% overlap. The autospectrum of a signal shows how much power is present at each frequency. Although FFT was performed over a bandwidth of 0-100 Hz, the spectra are displayed on a scale of 0-20 Hz, which contained >90% of the total power in SND (4).

Statistical analysis. Data are presented as means ± SD and analyzed by using ANOVA for comparison between the three nerves and paired t-test for comparison of the same variable before and after stimulation. P values  0.05 were considered significant. Statistical analysis was performed by using Statview 5.0 (Abacus Concepts).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Medullary raphe sites yielding a differential pattern of spinal sympathetic outflow in baroreceptor-denervated cats. The differential pattern of spinal sympathetic outflow shown in Fig. 1C was elicited by electrical stimulation of a site in the medullary raphe, on the midline, 1 mm caudal to the obex, and 1 mm below the dorsal surface (Fig. 1A). In this case, the frequency of stimulation (10.2 Hz) was equal to that of the free-running 10-Hz rhythm, which was the predominant component of SND in this baroreceptor-denervated cat. The pattern produced by caudal raphe stimulation included an increase in VN 10-Hz activity and marked decreases in CN and RN 10-Hz activities, accompanied by a fall in mean arterial blood pressure (MAP) of 35 mmHg. This pattern was observed in each of the 16 cats in which the 10-Hz rhythm was predominant. In four animals, we mapped the extent of the caudal medulla from which this pattern could be elicited by stimulation at frequencies near that of the free-running 10-Hz rhythm (which ranged from 8.8 to 11.2 Hz in these 4 cats). The stimulating electrode was moved in 1-mm steps, from 2 mm rostral to 3 mm caudal to the obex, from 0-2 mm lateral to the midline, and from 1 to 4 mm below the dorsal surface. This region is projected onto the dorsal surface of the medulla in Fig. 1B. The sites from which we observed reciprocal changes in VN vs. CN and RN activities were located 1-2 mm below the dorsal surface at 0-2 mm caudal to the obex (on the midline) and at 1 mm caudal to the obex (1 mm lateral to the midline). These sites (Fig. 1B) were located in nucleus raphe obscurus. Reciprocal changes in VN vs. CN and RN discharges were produced in all four animals by stimulation at 1 mm caudal to the obex on the midline but only in two of four cats at the four other sites indicated. These sites are referred to as pattern-producing caudal raphe (PPCR), without reference to the cell type(s) and/or pathway(s) that were electrically activated. Caudal medullary sites from which we did not observe reciprocal changes in nerve activity are described later.

View larger version (48K): [in this window] [in a new window]   Fig. 1.   Sites of stimulation and associated patterns of spinal sympathetic outflow. A: histological section showing site of raphe stimulation leading to differential pattern of spinal sympathetic outflow. Electrode tract is identified by the arrow. B: schematic view of the dorsal surface of the brain stem from 5 mm rostral to 5 mm caudal to the obex. , Sites projected onto dorsal medullary surface from which differential patterns in spinal sympathetic outflow were elicited. , Sites from which uniform increases in 10-Hz activity of all 3 nerves were observed. C and D: original recordings (low-pass filtered at 100 Hz) of discharges of inferior cardiac (CN), vertebral (VN), and renal (RN) postganglionic nerves. In C, stimulation at a site 1 mm caudal to the obex, on the midline, at a depth of 1 mm was begun at the arrow. Stimulus frequency (10.2 Hz) was the same as that of the free-running rhythm. In D, a site 2 mm rostral to the obex, on the midline, at a depth of 2 mm was stimulated at a frequency (8.8 Hz) equal to that of the free-running rhythm. Data in C and D are from different baroreceptor-denervated cats.

Stimulation (7-12 Hz) of PPCR in baroreceptor-denervated cats. PPCR stimulation at frequencies between 7 and 12 Hz elicited a significant increase in VN discharge and a significant decrease in discharges of CN and RN. Frequency-response relationships from one of six experiments are illustrated in Figs. 2 and 3. The stimulus intensity was 200 µA at each stimulus frequency. The autospectra of the discharges of each of the three nerves before (trace 1) and during (trace 2) PPCR stimulation at two frequencies are shown in Fig. 2. In this experiment, the frequency of the predominant free-running rhythm in SND was 10.2 Hz. Power in the 10-Hz band was completely moved to the frequency of PPCR stimulation, which was 8.4 Hz in Fig. 2A and 11.4 Hz in Fig. 2B. This indicates that PPCR activation in this frequency range entrained the 10-Hz rhythm in a 1:1 relationship to the stimuli. Note that VN 10-Hz power was increased, and CN and RN 10-Hz powers were decreased at both frequencies of stimulation. Figure 3A shows the complete frequency-response curve for the 10-Hz band of activity of the three nerves in the same animal. CN and RN 10-Hz power decreased and VN 10-Hz power increased at all four frequencies of stimulation. Table 1 summarizes the changes in 10-Hz power for CN, VN, and RN of the six cats in which we stimulated the PPCR at frequencies 1 Hz below, equal to, and 1 Hz above the frequency of the free-running 10-Hz rhythm (at constant stimulus intensity). Whereas VN 10-Hz power was significantly increased, significant decreases were observed for CN and RN 10-Hz power at these stimulus frequencies.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Frequency-domain analysis of the changes in CN, VN, and RN discharges elicited by activation of the pattern-producing caudal raphe (PPCR) at 2 frequencies of stimulation in a baroreceptor-denervated cat. The frequency of the free-running rhythm was 10.2 Hz. In each panel, the control (trace 1) and test (during PPCR stimulation, trace 2) autospectra are superposed on the same scale. The frequency of stimulation was 8.4 Hz in A and 11.4 Hz in B, both at an intensity of stimulation of 400 µA. The spectra (averages of 29 5-s windows with 75% overlap) were calculated from original recordings. Frequency resolution was 0.2 Hz.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Frequency-response curves under the condition of 1:1 entrainment of the 10-Hz rhythm to PPCR stimuli. A: power (% of control) in the 10-Hz band; B: power in the low-frequency (6 Hz) range; C: total power (0-20 Hz) for CN, VN, and RN. Stimulus intensity was 400 µA at all frequencies of PPCR activation. The frequency of the free-running rhythm was 10.2 Hz. Data are from the same experiment as Fig. 2.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Frequency-domain analysis of the effects of PPCR activation on the discharges of CN, VN, and RN at 2 intensities of stimulation in a baroreceptor-denervated cat. The control (trace 1) and test (during PPCR stimulation, trace 2) autospectra are superposed on the same scale. The intensity of stimulation was 100 µA in A and 500 µA in B, both at a frequency of stimulation of 9 Hz. Data processing was as in Fig. 2.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Intensity-response curves under the condition of 1:1 entrainment of the 10-Hz rhythm to PPCR stimuli. A: power (% of control) in the 10-Hz band; B: power in the low-frequency range; C: total power for CN, VN, and RN. Stimulus frequency was 9 Hz; free-running 10-Hz rhythm was at a frequency of 8.8 Hz. Data are from the same experiment as Fig. 4.

View larger version (45K): [in this window] [in a new window]   Fig. 6.   Time-domain analysis of the effects of PPCR stimulation in a baroreceptor-denervated cat. Time series showing cycle-by-cycle measurements of amplitudes (peak to trough) of 10-Hz slow waves in discharges of CN, VN, and RN (top to bottom on left) and phase angles relating the 10-Hz slow waves in CN-VN, VN-RN, and CN-RN pairs (top to bottom on right). The slow waves were extracted from original recordings by digital band-pass filtering (see METHODS). Vertical lines mark the start of stimulation at a frequency (10 Hz) near that (10.4 Hz) of the free-running 10-Hz rhythm. Stimulus intensity was 500 µA. Amplitudes are normalized on a scale of 0 to 1 relative to the largest slow wave in each time series.

Activation of PPCR with single shocks in baroreceptor-denervated cats. Single shocks (200-600 µA) delivered to PPCR at a frequency of 0.5 Hz elicited an initial biphasic response in all three nerves, consisting of an increase in activity (upward negative potential) followed by a decrease in activity (downward positive wave). An example is shown in Fig. 7; the traces are averages of 101 CN, VN, and RN responses, with the stimulus applied at time 0. In this case, the onset latencies of the negative potentials were 51 ms for the CN and 71 ms for the VN and the RN. In six cats, the onset latency of the negative potential was significantly shorter in CN than in VN (60 ± 17 ms vs. 78 ± 15 ms, P = 0.01, ANOVA), but RN onset latency (74 ± 14 ms) was not significantly different from that of either of the other two nerves. The initial biphasic response was followed by damped oscillations in SND (illustrated for CN and VN in Fig. 7) at a frequency corresponding to that of the free-running 10-Hz rhythm. The amplitude of these oscillations exceeded the deflections in the traces before application of the stimulus (left of time 0) by a factor of more than two. These damped oscillations indicate that PPCR stimulation reset the 10-Hz rhythm in these two nerves. In this animal, the free-running 10-Hz rhythm was absent in RN activity, and presumably for this reason there were no damped 10-Hz oscillations in the RN after the biphasic response. In all nerves in which 10-Hz activity was present, there was evidence of resetting of the 10-Hz rhythm in response to single shocks applied to the PPCR.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Responses of CN, VN, and RN to single PPCR shocks applied every 2 s in a baroreceptor-denervated cat. Traces are averages of 101 trials. Stimulus was delivered at time 0. Averages were calculated from original recordings sampled at 200 Hz.

High-frequency (25 Hz) stimulation of PPCR in baroreceptor-denervated cats. Figure 8 shows the autospectra of SND before (trace 1) and during (trace 2) high-frequency PPCR stimulation. In contrast to the increase in VN 10-Hz power induced by stimulation of the same site at frequencies near that of the free-running rhythm, high-frequency activation decreased 10-Hz power in VN as well as in CN and RN. We investigated the effects of high-frequency PPCR stimulation in seven cats. We noted a significant decrease in power in the 10-Hz band for all three nerves (CN 15 ± 27% of control, P = 0.005 paired t-test; VN 35 ± 29% of control, P = 0.003; and RN 16 ± 14% of control, P = 0.0001). In CN and RN, the decrease in 10-Hz power was accompanied by a significant decrease in low-frequency power (CN 46 ± 31% of control, P = 0.01; RN 56 ± 36% of control, P = 0.05). In contrast, there was a small but significant increase in VN low-frequency power (VN 119 ± 18% of control, P = 0.05). Total power was significantly reduced by high-frequency stimulation in CN (33 ± 23% of control P = 0.001) and RN (37 ± 33% of control, P = 0.01) but was not significantly changed in VN (95 ± 18% of control). The changes in SND produced by high-frequency PPCR stimulation were accompanied by a fall in MAP from 122 ± 20 to 83 ± 26 mmHg (P = 0.005, paired t-test).

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Frequency-domain analysis of the effects of high-frequency (25 Hz) PPCR stimulation on the discharges of CN, VN, and RN in a baroreceptor-denervated cat. The control (trace 1) and test (during PPCR stimulation, trace 2) autospectra are superposed on the same scale. Data processing was as in Fig. 2.

High-frequency (25 Hz) stimulation of PPCR in baroreceptor-intact cats. In baroreceptor-intact cats, the predominant rapid rhythm in SND is cardiac related (14, 28). That is, bursts of SND are locked in a 1:1 relationship to the cardiac cycle (Fig. 9A). Under this condition, we examined the effects of high-frequency PPCR stimulation on CN, VN, and RN discharges in six cats. To isolate the direct effects of PPCR stimulation on SND from baroreceptor reflex-mediated effects caused by changes in blood pressure, we maintained blood pressure at control levels (MAP 145 ± 19 mmHg) during stimulation. This was done by partial abdominal aortic obstruction induced by inflation of the balloon-tipped end of a Fogarty embolectomy catheter. In these animals, blood pressure was measured from the brachial artery. High-frequency PPCR stimulation dramatically decreased cardiac-related activity in CN and RN. This is shown by the time series in Fig. 9A and the superposed autospectra of SND in Fig. 9B. Although VN cardiac-related activity was initially somewhat reduced (Fig. 9A), there was little change in VN cardiac-related activity over the 40-s period of stimulation (Fig. 9B). In the six animals studied, high-frequency PPCR stimulation at a constant blood pressure produced a significant decrease in cardiac-related power in CN (24 ± 11% of control, P = 0.0001, paired t-test) and RN (13 ± 14% of control, P = 0.0001). In contrast, VN cardiac-related activity was not significantly changed by high-frequency PPCR stimulation (95 ± 22% of control).

View larger version (34K): [in this window] [in a new window]   Fig. 9.   Original recordings and frequency-domain analysis of the effects of high-frequency (25 Hz) PPCR stimulation on the discharges of CN, VN, and RN in a baroreceptor-intact cat. A: original recordings of arterial blood pressure (BP) and CN, VN, and RN discharges. High-frequency PPCR stimulation commenced at the vertical line. Note the 1:1 relationship between BP waves and sympathetic nerve activity. B: frequency-domain analysis. The primary peak in each autospectrum is at the cardiac frequency. The control (trace 1) and test (during PPCR stimulation, trace 2) autospectra are superposed on the same scale. Data processing was as in Fig. 2.

Stimulation (7-12 Hz) of caudal medullary sites that did not produce a differential pattern of spinal sympathetic outflow. The changes in CN, VN, and RN discharges shown in Fig. 1D were elicited by electrical stimulation of a site in nucleus raphe obscurus (2 mm rostral to the obex, on the midline, at a depth of 2 mm below the dorsal surface) at a frequency equal to that of the free-running rhythm. The uniform increases in CN, VN, and RN activities were accompanied by an increase in MAP of 17 mmHg. A similar pattern of increased 10-Hz activity in all 3 nerves was induced from 400 sites in 4 animals by stimulation at frequencies between 7 and 12 Hz. These sites are projected onto the dorsal medullary surface in Fig. 1B. The sites on the midline rostral to the obex were located in nucleus raphe obscurus, and those caudal to the obex were in the region of the decussation of the medial lemniscus. The other sites rostral to the obex were in the paramedian reticular nucleus (1 mm lateral to the midline) and nucleus reticularis parvocellularis and nucleus reticularis ventralis (2 mm lateral). The sites 1-3 mm caudal to the obex and lateral to the midline were in the medial reticular formation. In no case did altering the stimulus current (range 50-1,000 µA) change the increases in CN and RN 10-Hz discharges to decreases. Figure 10 shows representative autospectra of the discharges of the three nerves before (trace 1) and during (trace 2) stimulation at 8.8 Hz at a site 2 mm rostral to the obex, on the midline, 2 mm below the dorsal surface. Stimulation at this site dramatically increased 10-Hz power in all three nerves. There was a tendency for low-frequency power to increase, but this change was relatively small.

View larger version (10K): [in this window] [in a new window]   Fig. 10.   Frequency-domain analysis of the effects of stimulation of a medullary site 2 mm rostral to the obex, on the midline, at a depth of 2 mm on the discharges of CN, VN, and RN in a baroreceptor-denervated cat. Stimulus frequency (8.8 Hz) was the same as that of the free-running rhythm. The control (trace 1) and test (during PPCR stimulation, trace 2) autospectra are superposed on the same scale. Data processing was as in Fig. 2.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gebber et al. (16, 17) noted an increase in CN and a decrease in VN 10-Hz discharges during PAG stimulation and suggested that the mechanism underlying this defenselike response, at least in the urethan-anesthetized baroreceptor-denervated cat, is a reorganization of the coupling between brain stem oscillators that generate the 10-Hz rhythm. Three major observations were presented in support of this model. First, the responses to PAG stimulation were constrained to the 10-Hz band of SND. Second, the change in phase angle relating the 10-Hz discharges of CN and VN was proportional to the degree to which PAG differentially affected the 10-Hz discharges of these two nerves. Third, the response to PAG stimulation was frequency dependent in that reciprocal changes in CN and VN were produced by stimulation at frequencies at or above, but not below, the frequency of the free-running 10-Hz rhythm. In the current study, we tested the hypothesis that alterations of the coupling of brain stem 10-Hz oscillators also account for the differential pattern of spinal sympathetic outflow elicited by PPCR stimulation. We have rejected this hypothesis on the basis of the following observations. First, the response to PPCR stimulation was not band specific, because the reductions in CN and RN discharges were due to decreases in both low-frequency power and 10-Hz power. Second, the reciprocal changes in the discharges of VN vs. CN and RN were not, in the majority of cases, associated with changes in phase angle between the 10-Hz discharges of either the CN-VN or the VN-RN nerve pairs. Changes in phase angles would have been expected if PPCR activation reorganized the coupling among brain stem 10-Hz oscillators with different targets. Third, the pattern produced by PPCR activation was not frequency dependent, because we found no qualitative differences in PPCR response within the stimulus range in which 1:1 entrainment of the free-running 10-Hz rhythm occurred. Thus the mechanism responsible for the differential pattern of spinal sympathetic outflow elicited by PPCR stimulation is distinct from that responsible for the defenselike pattern elicited by PAG activation.

Figure 11 provides an alternative model to explain the increase in VN activity and decreases in CN and RN activities elicited by PPCR stimulation. In this model, PPCR projects to the system of coupled oscillators responsible for the 10-Hz rhythm in SND as well as to follower neurons located in the brain stem and/or spinal cord. These follower neurons receive convergent inputs from the 10-Hz rhythm generator and from the generators of low-frequency SND. Previous work from our laboratory has demonstrated that the 10-Hz and low-frequency components of SND are generated by different pools of brain stem neurons (1, 5), but they are carried to sympathetic nerves by the same pools of bulbospinal neurons (2, 3). We propose that PPCR inputs to the 10-Hz oscillators lead to resonance (that is, enhanced forced oscillations that reach follower circuits controlling CN, VN, and RN) when they are in the frequency range that entrains the 10-Hz rhythm in a 1:1 relationship to the stimuli (7-12 Hz). We further propose that PPCR inputs to bulbospinal and/or spinal follower neurons differentially inhibit the circuits governing CN, VN, and RN discharges.

View larger version (23K): [in this window] [in a new window]   Fig. 11.   Proposed model explaining the differential pattern of spinal sympathetic outflow elicited by PPCR stimulation. Inputs from PPCR entrain the 10-Hz oscillators and produce resonance when the frequency of stimulation is between 7 and 12 Hz. At or below the level at which the low-frequency generator and the 10-Hz oscillators have converged onto follower neurons, PPCR inputs inhibit the outflow to CN and RN, but not to the VN. +, Excitation; , inhibition; R, input capable of producing resonance. See METHODS for details.

Two observations support an input from the PPCR to the 10-Hz oscillators. First, PPCR activation at frequencies between 7 and 12 Hz entrained the 10-Hz discharges of all three nerves in a 1:1 relationship to the stimuli. As a consequence, peak power in the 10-Hz band was moved from the frequency of the free-running rhythm to the frequency of stimulation. Second, single shocks delivered to PPCR reset the 10-Hz rhythm, as evidenced by damped oscillations in SND after the initial excitatory response. We propose that entrainment of the 10-Hz oscillators led to resonance, because stimulation at all intensities in the 7- to 12-Hz range increased power in the 10-Hz band of VN activity without affecting power in the low-frequency band and, in addition, low-intensity stimulation of the PPCR at frequencies of 7-12 Hz increased power in the 10-Hz band of CN and RN, despite reducing low-frequency power. The decrease in CN and RN 10-Hz activity observed as stimulus current was raised may reflect inhibition occurring at the level of follower circuits that overshadowed resonance occurring at the level of the oscillator.

We propose that the inhibition of CN and RN discharges produced by PPCR stimulation occurred at the level of follower circuits because the decreases in CN and RN discharges were not band specific. In the baroreceptor-denervated cat, high-frequency PPCR stimulation decreased both low-frequency and 10-Hz components of CN and RN activities, as did high-intensity stimulation in the 7- to 12-Hz frequency range. In the baroreceptor-intact cat, high-frequency PPCR stimulation markedly reduced the cardiac-related discharges of these two nerves. Because the inhibition of CN and RN activity was not band specific, it follows that the site of inhibition occurs at a level at which the two rhythms converged onto groups of follower neurons, either in the brain stem or in the spinal cord.

We further propose that the inhibitory effects of PPCR stimulation were differentially distributed to the follower circuits controlling VN, CN, and RN, because neither 7- to 12-Hz or high-frequency stimulation reduced VN low-frequency power in the baroreceptor-denervated cat and high-frequency stimulation did not decrease VN cardiac-related power in the baroreceptor-intact cat. The slight increase in VN low-frequency activity in response to high-frequency stimulation in the baroreceptor-denervated cat could indicate some excitatory input either to the low-frequency generators or the follower neurons that control this nerve. However, this possibility is not consistent with the absence of a change in VN cardiac-related activity during high-frequency stimulation in the baroreceptor-intact cat. The only significant decrease in VN activity that we observed was the loss of 10-Hz activity in response to high-frequency PPCR stimulation in the baroreceptor-denervated cat. This decrease in 10-Hz power may represent disruption of 10-Hz rhythm generation by high-frequency inputs that upset the intrinsic oscillator, as occurs when high-frequency stimulation of the reticular formation blocks electroencephalogram sleep spindles (see review by Steriade and Llinas, Ref. 27).

The site of inhibition of CN and RN discharges could be in either the brain stem or the spinal cord. The caudal medullary raphe region is one of the major sources of descending inputs to the intermediolateral cell column (IML) of the thoracolumbar spinal cord (10, 20, 23). Indeed, sympathoinhibitory neurons have been located in the caudal medullary raphe rostral to the obex of the cat (22-24), and the axons of these neurons project to the IML (23, 24). Raphe neurons caudal to the obex have received less attention. Futuro-Neto and Coote (13) published the only other study we are aware of in which the PPCR was stimulated electrically in the cat while SND was recorded. The discrete region identified as PPCR in the current study is consistent with the region described by Futuro-Neto and Coote (13). In subsequent work, Coote et al. (9) showed that the PPCR region contains serotonergic neurons and that the disruption of serotonergic function eliminated inhibition of RN activity produced by intravenous physostigmine. In addition, they found that small cuts in the dorsolateral funiculus of the cervical spinal cord also eliminated the inhibition of RN activity produced by physostigmine. This led them to suggest that the inhibition was mediated in the spinal cord (9). The response to PPCR stimulation was considered by Futuro-Neto and Coote (13) to be analogous to that produced by intravenous physostigmine (9, 12). Although we did not investigate the site of inhibition of CN and RN discharges produced by PPCR stimulation in the present study, our data would be consistent with a spinal locus of inhibition, because both low-frequency and 10-Hz components of SND were reduced.

The possibility should be considered that rather than activating a homogenous cell population, PPCR stimulation coincidentally excited two functionally unrelated groups of neurons (or fibers of passage through this region), one of which provides input to the 10-Hz oscillator and the other of which selectively inhibits CN and RN activities. Although we cannot definitively exclude such a possibility, it seems improbable that there would be two independent groups of neurons located in a highly discrete raphe region whose electrical activation induced a patterned response that mimics that produced by systemically administered physostigmine. It is also unlikely that current spread to sites outside the caudal raphe accounted for the inhibition of CN and RN discharges produced by high-intensity PPCR stimulation. This follows from the observation that stimuli (7- to 12-Hz range) applied to adjacent sites in the paramedian nucleus and medial reticular formation enhanced rather than reduced the 10-Hz discharges of all three nerves.

Coote and coworkers (8, 9, 12, 13, 29) have proposed that the differential pattern of spinal sympathetic outflow elicited by PPCR stimulation is related to the cardiovascular changes that occur during REM sleep. In our experiments, the changes observed in CN and RN discharges were consistent with the changes in heart rate and blood flow to the kidney reported in cats during REM sleep (8, 21). Increased VN activity would also be consistent with the observed decrease in skeletal muscle blood flow during REM sleep (25) that, at least in some species, is mediated by an increase in sympathetic activity (26). It should be noted, however, that VN activity was increased only at frequencies of PPCR stimulation that entrained the 10-Hz rhythm. Whether population activity in the PPCR occurs naturally in the 7- to 12-Hz range during REM sleep is unknown. Thus the association of the pattern of spinal sympathetic outflow produced by PPCR stimulation in the current study with the cardiovascular changes seen during REM sleep remains speculative and requires further investigation.

Perspectives