INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A CARDIAC-RELATED RHYTHM COMMONLY appears in sympathetic nerve discharge (SND) of baroreceptor-innervated animals (12, 17, 37, 39-41). This rhythm is thought to result from baroreceptor-induced entrainment of centrally generated irregular low-frequency (6 Hz) oscillations in a one-to-one relationship to the arterial pulse (AP; 16, 28, 39). A recent study by us (28) showed that the medullary lateral tegmental field (LTF) contains a critical synaptic relay in the baroreceptor reflex pathway that controls SND of urethan-anesthetized cats. Specifically, blockade of N-methyl-D-aspartate (NMDA) receptors in the LTF virtually eliminated the cardiac-related rhythm in SND and prevented the inhibition of SND produced by abrupt obstruction of the abdominal aorta. As determined by using power-density spectral analysis, these effects occurred without a significant change in the total power (cardiac-related plus background) in the 0- to 6-Hz band of SND. Thus blockade of NMDA receptors in the LTF apparently converted the cardiac-related rhythm to irregular low-frequency oscillations. Although blockade of non-NMDA excitatory amino acid (EAA) receptors in the LTF also decreased power in the cardiac-related band of SND, the baroreceptor reflex remained functional, as evidenced by the persistence of the inhibition of SND during aortic obstruction and high coherence between SND and the AP. There was another striking difference between the effects produced by blocking the two types of EAA receptors in the LTF. In contrast to the results obtained by NMDA-receptor blockade, blockade of non-NMDA receptors significantly reduced total power in the 0- to 6-Hz band of SND. This finding implies that the central drive to sympathetic nerves depends, in part, on activation of non-NMDA receptors in the LTF. The first major goal of the current study was to investigate this possibility more fully. We reasoned that if activation of non-NMDA receptors in the LTF is involved in setting the level of excitatory drive to sympathetic nerves, bilateral microinjection of 1,2,3,4-tetrahydro-6-nitro-2,3-dioxobenzo-[f]- quinoxaline-7-sulfonamide (NBQX), a non-NMDA-receptor antagonist, into the LTF of baroreceptor-denervated cats should decrease SND and mean arterial pressure (MAP). Moreover, if NMDA receptors in the LTF are predominantly involved in mediating baroreceptor-induced entrainment of sympathetic nerve slow waves to the AP, bilateral microinjection of D()-2-amino-5-phosphonopentanoic acid (D-AP5), an NMDA-receptor antagonist, into the LTF should have little, if any, effect on SND or MAP after baroreceptor denervation.

The second goal of the current study was to estimate the degree to which EAA-mediated synaptic drive accounts for the naturally occurring discharges of rostral ventrolateral medullary (RVLM) sympathoexcitatory neurons (5, 10, 27, 37, 42). There is compelling evidence that RVLM-spinal neurons are the major direct source of excitatory drive to preganglionic sympathetic neurons [see Reviews by Dampney (14), Guyenet (20), and Ross et al. (30)]. However, the extent to which the basal discharges of RVLM neurons depends on intrinsic pacemaker activity versus synaptic inputs from other brain stem regions is controversial. Sun and colleagues (37, 38) have recorded pacemaker activity in sympathoexcitatory RVLM-spinal neurons of the rat in vivo after microinjection of kynurenate, a broad-spectrum EAA-receptor antagonist, and in untreated slices. On the other hand, Lipski et al. (27) reported that the spontaneous activity of RVLM-spinal neurons in vivo can be attributed primarily, if not solely, to their synaptic inputs. Barman and Gebber (5) proposed that one source of synaptic input to the RVLM is the LTF. They showed that the axons of LTF sympathoexcitatory neurons project to, and likely terminate in, the vicinity of RVLM-spinal neurons. LTF sympathoexcitatory neurons are defined as those for which naturally occurring discharges are correlated to the cardiac-related rhythm in SND and for which firing rates are decreased in parallel to SND during baroreceptor reflex activation (5, 17, 40, 41). Importantly, such LTF neurons fire earlier than RVLM sympathoexcitatory neurons during the cardiac-related slow wave in inferior cardiac SND (4, 5, 17). The results of the current study are consistent with the view that a significant portion of the basal discharges of RVLM-spinal sympathoexcitatory neurons depends on EAA synaptic inputs, some of which may be from LTF neurons.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General procedures. The protocols used in these studies on 25 adult cats (1.8-4.0 kg) were approved by the All-University Committee on Animal Use and Care of Michigan State University. Cats were initially anesthetized with 2.5% isoflurane mixed with 100% O₂. The right femoral artery and left and right femoral veins were cannulated to measure arterial pressure and to administer drugs, respectively. Urethan (1.1-1.8 g/kg iv, initial dose) was then administered, and isoflurane inhalation was stopped. Supplemental doses (0.2 g/kg iv) of urethan were given every 4-6 h. The initial dose of urethan has been shown to maintain a surgical level of anesthesia for 8-10 h in cats (16). The frontal-parietal electroencephalogram (EEG) showed a mixture of 7- to 13-Hz spindles and delta slow waves, indicative of unconsciousness and blockade of information transfer through the thalamus (34, 35). Noxious stimuli (e.g., pinch, cauterizing muscle) did not change the EEG pattern. As reported by Barman et al. (7), coherence analysis showed that there was no correlation between SND and either the EEG spindles or delta slow waves in urethan-anesthetized cats. Cats were paralyzed (gallamine triethiodide, 4 mg/kg iv, initial dose), pneumothoracotomized, and artificially respired with room air. End-tidal CO₂ was held near 4% (Traverse Medical Monitors Capnometer, model 2200), and rectal temperature was kept near 38°C with a heat lamp.

Baroreceptor denervation. The carotid sinus, aortic depressor, and vagus nerves were sectioned bilaterally. Two observations verified the completeness of baroreceptor denervation in these experiments. First, spectral analysis failed to show a cardiac-related rhythm in SND (as illustrated in Fig. 3). Specifically, there was no sharp peak in the autospectrum of SND at the frequency of the heart beat (top traces), and coherence values relating SND to the AP (bottom traces) were <0.1 at this frequency. Second, SND was not reflexly inhibited during the pressor response produced by an injection of norepinephrine bitartrate (1-2 µg/kg iv; not shown).

Neural recordings. The methods used to make monophasic recordings of left inferior cardiac postganglionic SND and the EEG can be found in earlier reports (5-7, 17). The preamplifier band pass was 1-1,000 Hz. The synchronized discharges of sympathetic nerve fibers appear as slow waves (i.e., envelopes of spikes) when this band pass is used (17). The slow waves occurred primarily at frequencies between 2 and 6 Hz (as shown by the oscillographic records in Fig. 2C), and there was no sign of periodic variation in their amplitude on a time scale of ~2-4 s as would be expected if SND was respiratory related (46). Although SND often contains a 10-Hz rhythm after baroreceptor denervation in urethan-anesthetized cats (7, 19, 29), for this study, we selected only cats in which this rhythm was absent. This was done to avoid the potential complication of producing changes in the 0- to 6-Hz band of SND indirectly as a result of changes in the 10-Hz rhythm (29).

Microinjections. The procedure and doses of EAA-receptor antagonists used for microinjections into the medulla were the same as those used by us in an earlier study (28). The competitive non-NMDA-receptor antagonist NBQX or the competitive NMDA-receptor antagonist D-AP5 was microinjected bilaterally into the LTF or RVLM (sites defined below) through a glass micropipette (~40-µm tip diameter) superglued to the needle of a 5-µl Hamilton syringe. The syringe and micropipette were filled with a 1 mM solution of NBQX or a 3 mM solution of D-AP5. Drugs were diluted in 0.9% saline, and the solution was adjusted to a pH of six to eight (litmus paper test) that assured solubility of the antagonists. The syringe and micropipette were mounted on a microinjection unit (David Kopf Instruments, model 5000). A 100-nl injection of NBQX (100 pmol) or D-AP5 (300 pmol) was made slowly (~20 s) at each medullary site by turning the calibrated micrometer on the microinjection unit.

No attempt was made in the current study to test directly the selectivity of these drugs on different classes of EAA receptors. However, the concentrations and doses of these drugs used here were similar to, or much lower than, those used in other studies in which their ability to antagonize selectively NMDA or non-NMDA receptors was demonstrated (11, 15, 32). We did not construct dose-response curves for D-AP5 and NBQX; the long duration of action of the drugs would have made it difficult to do this in an individual experiment. NBQX and D-AP5 were purchased from Research Biochemicals International (Natwick, MA).

The dorsal surface of the brain stem was exposed by removing portions of the occipital bone and cerebellum. The midline, obex, and dorsal medullary surface were used as landmarks for placement of the micropipette. Microinjections of EAA-receptor antagonists were made at four sites on each side of the medulla in either the LTF or RVLM (total of 400 nl/side for both regions). Figure 1 shows the target sites around which these injections were made. Multiple injections were made because neurons with activity correlated to SND are distributed over several millimeters within both the LTF and RVLM (4-6, 17, 40, 41). The micropipette was positioned into the LTF in tracks located 2 and 3 mm rostral to the obex and 2.8 mm lateral to the midline (Fig. 1, bottom 2 cross sections). Microinjections were made bilaterally at depths of 3 and 4.5 mm from the dorsal surface. The micropipette was positioned into the RVLM in tracks located 4.5 and 5.5 mm rostral to the obex and 3.5 mm lateral to the midline (Fig. 1, top 2 cross sections). Microinjections were made bilaterally at depths of 4 and 5.5 mm from the dorsal surface.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Sites of microinjections of 1,2,3,4-tetrahydro-6-nitro-2,3-dioxobenzo-[f]quinoxaline-7-sulfonamide (NBQX) and D()-2-amino-5-phosphonopentanoic acid (D-AP5).  mark target sites around which these injections were made in medullary lateral tegmental field (LTF) and rostral ventrolateral medulla (RVLM).  mark target sites medial to LTF in paramedian reticular nucleus (PR) or dorsal to LTF in vicinity of nucleus of solitary tract (NTS). +2-5.5 Refer to approximate distance (in mm) from obex according to stereotaxic atlas of Berman (9). IO, inferior olive; 5Sp, spinal trigeminal nucleus. Calibration, 2 mm.

The protocol used in these studies was as follows. An 80-s control data sample was collected after the micropipette was positioned at the first site to be injected. A complete set of injections was then made on the left and right sides of the medulla. Test 80-s data blocks were collected within 1-2 min after the last injection, 10-15 min later, and then at 15-30 min intervals for up to 1.5 h to allow for partial or full recovery. The data block collected 10-15 min after the microinjection was used to quantify the effects of D-AP5 or NBQX on SND and MAP. By this time, the maximum changes in SND had occurred, and SND had reached a steady-state level. Changes in MAP were also quantified at this time. The time courses of action of NBQX and D-AP5 were similar to those reported by others when cardiovascular or respiratory changes were monitored after the microinjection of these drugs into the brain (14, 21, 28, 32, 33).

Two types of control experiments were performed. In the first set, vehicle (saline adjusted to a pH of 6-8) was injected bilaterally into the LTF or RVLM; sites of vehicle injection were the same as those for NBQX and D-AP5 microinjections. In the second set of control experiments, NBQX was microinjected bilaterally into sites medial to the LTF injection sites (1 mm lateral to the midline) or dorsal to the LTF injection sites (1.3-1.5 mm below the dorsal surface; Fig. 1, bottom). The injection sites medial to the LTF were in the paramedian reticular nucleus, and those dorsal to the LTF were in the vicinity of the nucleus of the solitary tract (NTS).

Some cats were used for more than one set of microinjections, but in no case were two sets of injections made into the same medullary region. For example, saline injections into the LTF were made in cats in which an EAA-receptor antagonist was subsequently microinjected into the RVLM (and vice versa). If microinjection of an EAA-receptor antagonist produced a change in SND, we waited for recovery from the first set of injections before injecting the same drug into the second region. If the first set of injections of a drug caused no changes in SND, we waited a minimum of 30 min before microinjecting the same drug into the second region.

Data analysis. Before all analyses on a Dell Optiplex 500 MHz Pentium III computer, SND was low-pass filtered at 100 Hz. Data were acquired (5-ms sampling interval) with software and an analog-to-digital converter board from RC Electronics (Santa Barbara, CA). Fast Fourier transform was performed on 32 5-s windows of data with 50% overlap (80-s data block) to construct autospectra of left inferior cardiac SND and the AP and coherence functions relating SND to the AP. A modified version (19) of the software of Cohen et al. (13) and Kocsis et al. (23) was used for these analyses. Spectral analyses were done over a frequency band of 0-100 Hz with a resolution of 0.2 Hz/bin. The spectra are displayed on a scale of 0-20 Hz, which contains >90% of the total power in SND (23). The autospectrum of a signal shows how much power (voltage squared) is present at each frequency. The coherence function (normalized cross spectrum) is a measure of the strength of linear correlation of two signals at each frequency. The squared coherence value (referred to as coherence value) is one in the case of a perfect linear relationship and zero if two signals are unrelated. A coherence value 0.1 reflects a statistically significant relationship when 32 windows are averaged (8).

ASCII files of the spectra were saved for transfer to spreadsheet, graphics, and statistical programs (Statmost for Windows 3.0; DataMost). The autospectra of SND before and after microinjection of D-AP5 or NBQX were displayed on the same power scale. We quantified power in SND at frequencies 6 Hz, because this is the component of SND that can become entrained to the AP when baroreceptor afferents are intact (17, 28, 39). The 0- to 6-Hz power was calculated by arithmetically summing the values for the bins in this frequency range from the ASCII files. We also measured total power in SND, as reflected by arithmetically summing the values for the bins in the 0- to 20-Hz frequency range.

Statistical analysis. When 0- to 6-Hz power, total power in SND, and MAP were evaluated at three time points (control, 10-15 min after drug injection, and recovery), we used repeated-measures analysis of variance followed by the Student's paired t-test (31). When data for only two time points were compared (control and 10-15 min after drug injection), we used the Student's paired t-test. Raw values of power were used for statistical analyses, but changes in SND are expressed as percent of control in the text and in Table 1. Values are means ± SE. P  0.05 indicated statistical significance.

Histology. The brain stem was removed and fixed in 10% buffered Formalin. Frontal sections of 30-µm thickness were cut and stained with cresyl violet. Sites of microinjection were identified with reference to the bottom of the tracks made with the micropipette and the stereotaxic planes of Berman (9). Figure 2B shows an example of the tracks made with the micropipette in the LTF in one of these experiments.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Effects of microinjection of NBQX into medullary LTF of a baroreceptor-denervated cat. A: oscillographic records of arterial pressure (AP; in mmHg) and inferior cardiac sympathetic nerve discharge (SND) during microinjection of NBQX. Injections into left side of medulla were begun at arrow. Set of injections into right side of medulla was made before start of this trace. Vertical calibration is 100 µV; horizontal calibration is 20 s. B: histological cross section of medulla, ~3 mm rostral to obex, showing tracks made with micropipette. C: oscillographic traces of SND on an expanded time scale before (top) and 10 min after (bottom) bilateral microinjection of NBQX into LTF. Vertical calibration is 100 µV; horizontal calibration is 1 s.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of NBQX microinjection into LTF. The competitive non-NMDA EAA-receptor antagonist NBQX (100 pmol/100 nl at each injection site) was microinjected bilaterally into the LTF of seven baroreceptor-denervated cats with an MAP of 103 ± 8 mmHg. The injection sites are shown in the bottom two cross sections in Fig. 1. SND and MAP typically began to decrease during or within 1-2 min of completing the set of eight injections of NBQX into the LTF. In the example shown in Fig. 2A, microinjections into the LTF on the right side of the medulla (contralateral to the nerve recording) had already been completed; shortly after starting to inject NBQX into the LTF on the left side, SND and MAP began to decrease.

Figure 2C shows 10-s traces of SND before and 10 min after bilateral microinjection of NBQX into the LTF. Note that the irregular oscillations in SND were markedly reduced in amplitude after the microinjection. We used spectral analysis to quantify the changes in SND. Figure 3 shows the results from a representative experiment. The autospectrum of SND (Fig. 3A) before microinjection of NBQX showed a dispersed band of power, primarily at frequencies 6 Hz. As shown by coherence analysis (Fig. 3A), SND was not correlated to the AP (coherence value at the frequency of the heart beat was <0.1). This is evidence for complete baroreceptor denervation. The data in Fig. 3B were obtained 10 min after NBQX was microinjected bilaterally into the LTF. At this time, 0- to 6-Hz power and total power in SND were decreased to 43 and 51% of control, respectively. The reduction in SND was accompanied by a fall in MAP from 95 to 83 mmHg. As shown in Fig. 3C, 1 h after the microinjection, both 0- to 6-Hz power and total power in SND were 109% of control; and MAP returned to 92 mmHg.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Effects of bilateral microinjection of NBQX into medullary LTF on SND of a baroreceptor-denervated cat. Traces (top to bottom) are autospectra (AS) of SND and AP and corresponding coherence function before (A), 10 min after (B), and 1 h after (C) microinjection. Spectra are based on 32 5-s windows with 50% overlap; frequency resolution is 0.2 Hz/bin. AS of SND before and after microinjection are displayed on same power scale here and in subsequent figures. MAP, mean arterial pressure.

As summarized in Table 1, 0- to 6-Hz power and total power in SND were significantly decreased 10-15 min after bilateral microinjection of NBQX into the LTF of seven baroreceptor-denervated cats. The reduction in SND was accompanied by a significant fall in MAP. In three of these cats, we raised MAP back to control level by a slow intravenous infusion of a mixture of norepinephrine bitartrate and dextran. This procedure did not reverse the decrease in SND. In five of the seven experiments, we monitored SND and MAP for as long as 60 min after injection of NBQX. At this time, 0- to 6-Hz and total power had recovered to 99 ± 14 and 101 ± 15% of control, respectively; MAP had returned to control level (100 ± 7 mmHg).

Effects of D-AP5 microinjection into LTF. The competitive NMDA-receptor antagonist D-AP5 (300 pmol/100 nl at each injection site) was microinjected bilaterally into the LTF of eight baroreceptor-denervated cats with an MAP of 91 ± 5 mmHg. In six of these cats, 0- to 6-Hz power and total power in SND were not changed by bilateral microinjection of D-AP5 into the LTF. In the example shown in Fig. 4, 0- to 6-Hz power and total power in SND were 101 and 98% of control, respectively, 15 min after bilateral microinjection of NBQX (compare Fig. 4, top and bottom); MAP was also essentially unchanged (95 vs. 100 mmHg). Comparable data were obtained from five other cats. In the other two cats, 0- to 6-Hz power in SND was increased to 124 and 153% of control 10-15 min after microinjection of D-AP5. Total power was also elevated (123 and 125% of control), and MAP was increased by 10 and 17 mmHg in these two cats. When the data from the eight experiments were pooled (Table 1), there were no significant changes in SND or MAP.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effects of bilateral microinjection of D-AP5 into LTF on SND of a baroreceptor-denervated cat. Top: AS of SND before microinjection. Bottom: AS 15 min after microinjection.

Effects of NBQX microinjection into RVLM. NBQX (100 pmol/100 nl at each injection site) was microinjected bilaterally into the RVLM of four baroreceptor-denervated cats with an MAP of 96 ± 4 mmHg. The injection sites are shown in the top two cross sections in Fig. 1. In the representative experiment shown in Fig. 5A, 0- to 6-Hz power and total power in SND were decreased to 46 and 47% of control, respectively, 11 min after bilateral microinjection of NBQX (compare Fig. 5A, 1 and 2). The reduction in SND was accompanied by a fall in MAP from 107 to 85 mmHg. The 0- to 6-Hz power and total power in SND partially recovered to 81 and 73% of control, respectively, 1.5 h after completing the injections (Fig. 5A3); MAP was 97 mmHg at that time.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Effects of bilateral microinjection of NBQX (A) and D-AP5 (B) into RVLM on SND of baroreceptor-denervated cats. A1 and B1: AS of SND before microinjection. A2 and A3: AS of SND 10 and 80 min after microinjection of NBQX, respectively. B2 and B3: AS of SND 10 and 45 min after microinjection of D-AP5, respectively.

As shown in Table 1, both 0- to 6-Hz power and total power in SND were significantly reduced 10-15 min after bilateral microinjection of NBQX into the RVLM of four baroreceptor-denervated cats. The reduction in SND was accompanied by a significant decrease in MAP. Raising MAP to control level by intravenous infusion of norepinephrine and dextran in two of these cats did not reverse the changes in SND. The effects of microinjection of NBQX into the RVLM were similar to those produced by microinjection of this drug into the LTF.

Effects of D-AP5 microinjection into RVLM. D-AP5 (300 pmol/100 nl at each injection site) was microinjected bilaterally into the RVLM of six baroreceptor-denervated cats with an MAP of 102 ± 6 mmHg. In the representative experiment shown in Fig. 5B, both 0- to 6-Hz power and total power in SND were reduced to 55% of control 10 min after bilateral microinjection of D-AP5 into the RVLM (compare Fig. 5B, 1 and 2). MAP was reduced from 82 to 75 mmHg at this time. The 0- to 6-Hz power and total power in SND partially recovered to 81 and 78% of control, respectively, 45 min later; and MAP was 85 mmHg at that time (Fig. 5B3).

As summarized in Table 1, bilateral microinjection of D-AP5 into the RVLM of six baroreceptor-denervated cats significantly reduced 0- to 6-Hz power and total power in SND. There was a tendency for MAP to decrease, but this change was not statistically significant.

Control injections. Saline was microinjected bilaterally into the LTF of three cats with an MAP of 105 ± 10 mmHg and into the RVLM of three other cats with an MAP of 97 ± 11 mmHg. Neither SND nor MAP were affected by these injections. The data are summarized in Table 1.

NBQX was microinjected into sites medial (n = 3) or dorsal (n = 3) to the LTF in baroreceptor-denervated cats with an MAP of 105 ± 1 and 107 ± 2 mmHg, respectively. These injection sites are indicated in Fig. 1; they were in the paramedian reticular nucleus and the vicinity of the NTS. The results from these two sets of experiments are summarized in Table 1. NBQX failed to significantly affect SND and MAP when microinjected into these regions, although there was a tendency for SND to increase after injection into the paramedian reticular nucleus.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current study is the first to show that medullary LTF neurons play a key role in setting the basal level of activity in sympathetic nerves with cardiovascular targets in baroreceptor-denervated cats. Specifically, microinjection of NBQX bilaterally into the LTF reversibly decreased 0- to 6-Hz power as well as total power in SND to ~50% of control. Importantly, the reduction in SND was accompanied by a significant fall in MAP. In contrast, microinjection of D-AP5 into the LTF did not affect SND or MAP. Thus non-NMDA, but not NMDA receptor-mediated activation of LTF neurons, appears to be imprtant in determining the resting level of SND in baroreceptor-denervated cats. The source of this EAA-mediated excitatory drive to LTF neurons remains to be determined.

The decreases in SND and MAP produced by microinjection of NBQX into the LTF may be due, at least in part, to removal of synaptic drive (i.e., disfacilitation) from this region to RVLM-spinal sympathoexcitatory neurons. This possibility stems from earlier work from this laboratory on LTF neurons with sympathetic nerve-related activity (5, 6, 17). The naturally occurring discharges of ~20% of LTF neurons are correlated to the cardiac-related rhythm in SND of baroreceptor-innervated cats or the irregular low-frequency oscillations in SND that replace the cardiac-related rhythm after baroreceptor denervation. Some of these neurons have been classified as sympathoexcitatory because their firing rates are decreased in parallel to SND during baroreceptor reflex activation (5, 17). Importantly, these LTF neurons fire significantly earlier than RVLM neurons during the inferior cardiac sympathetic nerve slow wave (4, 5, 17), and their axons project to and likely terminate in the vicinity of RVLM-spinal neurons (5). When considered in conjunction with these previous findings, the marked decrease in SND produced by microinjection of NBQX into the LTF supports the view that excitatory drive from the LTF is an important source of the discharges of RVLM-spinal sympathoexcitatory neurons.

The magnitude of the decrease in 0- to 6-Hz power in SND that occurred after the microinjection of NBQX into the LTF of baroreceptor-denervated cats was similar to that of the decrease in total power (cardiac-related plus background) in the 0- to 6-Hz band of SND of baroreceptor-innervated cats (28). In the earlier study, we could not assess the effects of NBQX microinjection into the LTF on MAP, because we purposely kept MAP constant throughout the experiment. This was done to avoid changes in cardiac-related power in SND that might have resulted from a change in the level of baroreceptor afferent nerve activity. Thus the current study is the first to show that the decrease in SND produced by blockade of non-NMDA receptors in the LTF is sufficient to cause a significant fall in MAP. One might argue that the decrease in SND resulted secondarily from the fall in MAP. This is unlikely because the change in SND was not reversed by returning MAP to control level by an intravenous infusion of a mixture of norepinephrine and dextran.

The possibility should be considered that the decreases in SND and MAP after microinjection of NBQX into the LTF resulted from spread of the injectate to the RVLM. Indeed, microinjection of the same dose of NBQX into the RVLM produced similar changes in SND and MAP. Although we did not directly assess the degree of spread of injectate from the LTF, several observations imply that the drug did not spread to the RVLM (a distance of 3 mm from the LTF) or to regions adjacent to the LTF that are involved in cardiovascular regulation. First, microinjection of NBQX at sites 1.8 mm medial to the LTF in the paramedian reticular nucleus or 1.5 mm dorsal to the LTF in the vicinity of the NTS did not decrease SND. In fact, there was a tendency for SND to increase after blockade of non-NMDA receptors in the paramedian reticular nucleus. Second, in an earlier study (28), we showed that microinjection of the same dose of NBQX into sites 1.2 mm lateral to the LTF in the caudal ventrolateral medulla of baroreceptor-innervated cats significantly increased power in the cardiac-related band of SND as well as total power in SND. These components of SND were significantly decreased when NBQX was microinjected into the LTF. Third, because the concentration of a drug would decrease at sites remote from the center of the injection (26), we should have needed to inject a higher dose of NBQX into the LTF than into the RVLM to produce equivalent reductions in SND. Fourth, if NBQX spread from the LTF to the RVLM, the same should be true for D-AP5. However, microinjection of D-AP5 into the LTF did not mimic the effects of microinjection of this drug into the RVLM. Specifically, D-AP5 significantly decreased 0- to 6-Hz power and total power in SND when injected into the RVLM, but SND either did not change or increased when this drug was injected into the LTF. Taken together, these observations support the view that NBQX acted within the LTF to induce changes in SND and MAP.

There is evidence that medullary regions other than the RVLM also contribute to the maintenance of vasomotor tone in rats. Korkola and Weaver (24) reported that bilateral microinjection of glycine into the dorsal medullary reticular formation produced transient decreases in renal SND, MAP, and heart rate of anesthetized rats. Recently, Campos and McAllen (10) reported that microinjection of glycine into the caudal pressor area of the rat significantly reduced MAP and the firing rate of RVLM sympathetic premotor neurons. Whether either of these regions are the counterpart of the LTF in the cat remains to be determined.

Microinjection of D-AP5 into the LTF of baroreceptor-innervated cats reversibly eliminated the cardiac-related rhythm in SND without significantly affecting total power at frequencies 6 Hz (28). This implied that blockade of NMDA receptors in the LTF simply converted cardiac-related bursts to irregular low-frequency oscillations by interfering with transmission in the baroreceptor reflex arc. If so, one would predict that microinjection of D-AP5 into the LTF of baroreceptor-denervated cats would not affect SND or MAP. This was indeed the case in six of eight experiments. In the other two cats, SND was increased to 124 and 153% of control and MAP was increased 10 mmHg after bilateral microinjection of D-AP5 into the LTF. These changes might reflect blockade of NMDA receptor-mediated activation of LTF sympathoinhibitory neurons by nonbaroreceptor inputs. In baroreceptor-innervated cats, the naturally occurring discharges of LTF sympathoinhibitory neurons are correlated to the cardiac-related rhythm in SND, their firing rates increase during the inhibition of SND produced by baroreceptor reflex activation, and their axons project to the vicinity of caudal medullary raphespinal sympathoinhibitory neurons (6, 17). Importantly, Gebber and Barman (17) showed that the discharges of these neurons remain correlated to the irregular low-frequency oscillations in SND after baroreceptor unloading. Thus such sympathoinhibitory neurons are synaptically driven by inputs of nonbaroreceptor as well as baroreceptor origin.

Microinjection of either NBQX or D-AP5 into the RVLM significantly decreased 0- to 6-Hz power and total power in SND. Thus EAA neurotransmission in the RVLM is required for maintenance of SND. It follows that synaptic inputs must be important in generating the basal activity of RVLM-spinal sympathoexcitatory neurons. It is not known whether intrinsic pacemaker properties also contribute to their basal activity as has been proposed for RVLM-spinal sympathoexcitatory neurons of the rat (37, 38). The decreases in SND and MAP produced by blockade of EAA neurotransmission in the RVLM was not surprising, because microinjection of the nonspecific glutamate-receptor antagonist kynurenate has been shown to reduce SND and MAP in cats (1, 18). However, this is the first study to show that both NMDA and non-NMDA receptor-mediated neurotransmission in the RVLM are involved in setting the level of basal SND. The role of EAA-mediated synaptic transmission in the RVLM of the rat is less clear. Blockade of EAA receptors in the RVLM of the rat has been reported either not to affect (2, 18, 22, 36) or to reduce (25, 43) SND and MAP.

Microinjections of NBQX into the LTF or RVLM reduced MAP from an average of 103 to 87 mmHg or from 96 to 81 mmHg, respectively. These changes seem modest in the presence of a ~50% reduction in SND. However, after complete cervical spinal cord transection in cats, MAP is reduced to ~70 mmHg at a time when SND is decreased to ~10% of control (44, 45). When viewed in this light, the ~15-mmHg fall in MAP produced by NBQX microinjection into the LTF or RVLM is not surprising.

The decreases in SND and MAP produced by NBQX microinjection into the LTF were comparable to those produced by microinjection of this EAA-receptor antagonist into the RVLM. This does not necessarily mean that these regions are of equal importance in setting the basal level of SND. First, because we did not determine dose-response relationships for EAA-receptor antagonists, we do not know whether the dose of NBQX used produced different degrees of blockade of non-NMDA receptors in the LTF and RVLM. Second, it is possible that combined blockade of non-NMDA and NMDA receptors in the RVLM would have lowered SND to a greater extent than blockade of EAA neurotransmission in the LTF. This is because microinjection of either NBQX or D-AP5 into the RVLM significantly reduced SND. Third, we do not know the extent to which neurotransmitters other than EAAs or intrinsic pacemaker properties determine the level of activity of LTF and RVLM neurons and, therefore, the level of SND. Indeed, blockade of EAA receptors in either the LTF or the RVLM did not reduce SND to the level (<10% of control) seen after complete cervical spinal cord transection (44, 45), implying that mechanisms other than EAA-mediated activation of neurons in these regions contribute to basal SND. Fourth, we might have actually underestimated the effects on SND and MAP produced by blockade of non-NMDA receptors on LTF sympathoexcitatory neurons. In this regard, it is possible that the decrease in SND due to inactivation of these neurons with NBQX was partially masked by a simultaneous blockade of EAA receptors on LTF sympathoinhibitory neurons that would have increased SND (28). Thus the data available do not allow us to compare the relative degrees to which these two regions determine the resting level of SND and MAP.

Perspectives