INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

STATIC CONTRACTION AND STRETCH of skeletal muscle can reflexly increase heart rate (HR) and mean arterial blood pressure (MAP), and this response is commonly referred to as the "exercise pressor reflex" or the "muscle pressor reflex" (13, 19, 27). The increases in cardiovascular function associated with this reflex are mediated by activation of group III and IV muscle afferent fibers (17). Static contraction activates both mechanically and metabolically sensitive afferent fibers, whereas muscle stretch preferentially activates mechanically sensitive neurons (14, 15, 20). Many of these muscle afferent neurons synapse in the dorsal horn of the spinal cord, activating neurons that send projections to medullary regions and ultimately causing cardiovascular function to increase (5, 13, 18, 27).

Although a spinal synapse is involved, whether or not this dorsal horn synapse is obligatory for the muscle pressor reflex has not been directly tested. In other words, do all afferent neurons mediating the pressor reflex synapse in the dorsal horn, or do some of these primary afferents course rostrally to medullary regions and/or sympathetic preganglionic neurons without synapsing in the dorsal horn? Particularly germane to this question is anatomic and electrophysiological data showing that primary afferent neurons typically bifurcate into ascending and descending branches on entering the spinal cord (3, 4, 21, 22). These branches often travel beyond the segment of entry without synapsing by coursing through several different spinal regions, including Lissauer's tract. Previous studies have demonstrated that multiple spinal segments are involved in producing the muscle pressor reflex, despite limiting the afferent input to a single spinal segment (9, 10, 23, 24, 28). This multisegmental integration could be due to afferent branching. However, it is also possible that intraspinal pathways are responsible for the multisegmental integration of the muscle pressor reflex. In other words, second-order neurons within the segment of entry receiving input from muscle afferent fibers may activate dorsal horn cells in adjacent segments, and these third-order neurons project to medullary and/or spinal sympathetic preganglionic neurons.

The purpose of this study was to test the following hypotheses: 1) the dorsal horn of the lumbar spinal cord is an obligatory synaptic site for the muscle pressor reflex; and 2) branches of primary afferent neurons are responsible, at least in part, for the multisegmental integration of the muscle pressor reflex. To test these hypotheses, a "spinal well" was created over the exposed dorsal surface of the spinal cord. This allowed us to administer the GABA_A agonist muscimol over a single spinal segment or multiple spinal segments (6, 16). Activation of GABA receptors inhibits dorsal horn cells and can cause presynaptic inhibition of primary afferent neurons (16). Thus muscimol was given as a means of blocking synaptic excitation within the dorsal horn but not fibers of passage. Furthermore, afferent input mediating the muscle pressor reflex was limited to the L₇ spinal segment.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surgical preparation. Adult cats (mean weight 3.9 ± 0.2 kg) were anesthetized by inhaling an isoflurane-oxygen mixture. A catheter was inserted into the cephalic vein, the isoflurane-oxygen mixture was removed, and anesthesia was maintained by an intravenous (iv) injection of -chloralose (80 mg/kg) and urethane (100 mg/kg). Polyethylene catheters were inserted into an external jugular vein and a carotid artery, and the trachea was exposed. An endotracheal tube was inserted into the airway via a tracheotomy. If resting HR and/or arterial blood pressure increased, and/or a corneal reflex appeared, additional -chloralose (5-10 mg/kg) was provided. The animals were mechanically ventilated. Arterial blood gases were periodically measured and maintained within normal limits by adjusting the ventilator, administering sodium bicarbonate intravenously, and/or providing supplemental oxygen. Body temperature was continuously monitored with the use of a rectal thermometer (YSI series 400) and maintained between 36.0 and 38.0°C by a heating pad and lamp.

Protocols. Static muscle contractions were evoked by electrically stimulating the L₇ and S₁ ventral roots at three times motor threshold (0.1 ms, 40 Hz). Resting tension on the muscle is set at 1 kg (optimal length in this preparation). To activate muscle mechanoreceptors, the muscle was stretched by manually displacing the force transducer by 1.5 cm, which likely represents the upper limit of the physiological range (7). The contractions and stretches were 1 min in duration and were performed in a counterbalanced manner.

Measured and calculated variables. Arterial blood pressure was measured by connecting the common carotid artery catheter to a pressure transducer (model P23ID, Statham), and muscle tension was measured by using the force transducer. The arterial blood pressure and tension variables were continuously monitored on a four-channel chart recorder (model 7400, Astromed) and a personal computer (Biopac acquisition system). MAP and HR were obtained from the arterial blood pressure signal. Baseline values were determined by averaging at least 30 s of data before the muscle contraction or stretch. The peak value for MAP, HR, and tension represents the peak level that the variable reached during the 1 min of muscle contraction or stretch. The peak change in a given variable represents the difference between the peak and baseline values.

Data analysis. Data are expressed as means ± SE. Dose-response changes in MAP and HR were analyzed by using a one-way ANOVA with repeated measures. Baseline and peak hemodynamic and tension data were analyzed by using a two-way ANOVA with repeated measures, with the two factors being time (baseline and peak) and drug (pre- and postmuscimol). A Tukey test was used for post hoc comparisons when applicable. The changes in the hemodynamic and tension data before and after muscimol were compared by using a paired t-test. For all analyses, P < 0.05 was used as the level of statistical significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Figure 1 illustrates the dose-response data using the two different dose ranges of muscimol. Figure 1A, demonstrates that the 0.01-1.0 µM dose range had no effect on the pressor reflex evoked by static contraction. On the other hand, Fig. 1B, shows that 0.1 mM markedly attenuated the pressor reflex, whereas the 1.0 mM dose abolished it. There were no differences in baseline hemodynamic data or developed tensions across the various doses. However, the peak MAP and HR levels were reduced across the doses in the higher dose range tested.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effect of increasing doses of muscimol on the peak changes () in mean arterial blood pressure (MAP) evoked by static contraction of skeletal muscle. A: results from the lower dose range tested (n = 2). B: dose response using the higher dose range (n = 3). Values are means ± SE. * P < 0.05 compared with premuscimol (control; dose = 0).

Figure 2 is an original record from one cat showing the reflex cardiovascular responses to static muscle contraction before and after placing 1.0 mM muscimol into the well. Before muscimol, static contraction increased MAP and HR. However, note that after muscimol these increases were abolished. In fact, MAP fell during the muscle contraction, which was a typical result. The individual and mean data for the contraction-evoked changes in MAP and HR are illustrated in Fig. 3. Note that, with one exception, muscimol abolished the pressor and tachycardic responses evoked by static contraction. In this one exception, the pressor response was minimal and very transient. In the other four animals, MAP stayed constant until ~20 s into the contraction. Thereafter, MAP tended to fall, similar to the original record depicted in Fig. 1. Similar to contraction, muscimol abolished the cardiovascular responses evoked by muscle stretch in all but one animal (Fig. 4). Again, the increase in this animal (same animal as for the contraction data) was minimal and very transient. In contrast to the contraction data, MAP and HR remained stable throughout the stretch. For both the contraction and stretch data, baseline MAP was significantly lower, and baseline HR was lower for the contraction data but not the stretch data (Table 1). There were no differences in tension before and after muscimol. When the contraction and stretch were repeated after the muscimol was replaced with PBS, the reflex cardiovascular responses showed little to no evidence of recovery, even when it was elicited >2 h later.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Original record from 1 cat showing the reflex cardiovascular responses evoked by static contraction of skeletal muscle before (A) and ~30 min after (B) administration of 1.0 mM muscimol into the spinal well. Note that muscimol abolishes the muscle pressor reflex. HR, heart rate; bpm, beats/min.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   MAP (top) and HR (bottom) evoked by static contraction of skeletal muscle before (pre) and after (post) 1.0 mM muscimol was administered into the spinal well. Open symbols, individual responses (n = 5); , mean data (±SE). * P < 0.05, pre vs. post.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   MAP (top) and HR (bottom) evoked by muscle stretch before (pre) and after (post) 1.0 mM muscimol was administered into the spinal well. Open symbols, individual responses (n = 5); , mean data (±SE). * P < 0.05, pre vs. post.

Administration of muscimol to the L₇ segment attenuated, but did not abolish, the cardiovascular responses evoked by static contraction (Fig. 5). Both the MAP and HR increases were reduced with no difference in developed tension (10.6 ± 0.8 kg before and 10.3 ± 0.8 kg after). Muscimol administration to L₇ only had a similar effect on the cardiovascular increases evoked by muscle stretch, although the reductions did not reach statistical significance (Fig. 5). There was no difference in recorded tension for the stretch procedure (8.8 ± 0.5 kg before and 8.7 ± 0.5 kg after). There were also no differences in baseline hemodynamic data for either contraction or stretch (Fig. 5).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Mean hemodynamic data [MAP (A and B) and HR (C and D)] elicited by static muscle contraction (A and C) and muscle stretch (B and D) before (solid bars) and after (open bars) muscimol (1 mM) when the drug was confined to the L7 segment of the spinal cord. Nos. underneath bars indicate baseline values for that variable. Values are means ± SE; n = 4. * P < 0.05, before vs. after muscimol.

Injecting muscimol iv had no effect on the cardiovascular increases evoked by static contraction or muscle stretch for the three cats in which it was tested. Before iv muscimol, contraction increased MAP by 52 ± 13 mmHg and HR by 24 ± 15 beats/min. Approximately 30 min after muscimol, contraction increased MAP by 53 ± 16 mmHg and HR by 29 ± 13 beats/min. The stretch-evoked increases in MAP and HR were 57 ± 17 mmHg and 23 ± 11 beats/min before and 59 ± 15 mmHg and 24 ± 8 beats/min after iv muscimol.

Postmortem examination of the Prussian reaction showed a strong blue stain on the surface of the spinal cord that was within the well. Even though the well typically leaked some fluid by the end of the day, this fluid was confined to the musculature surrounding the spinal canal, as blue was never seen on regions covered with the VPS. For the L₇ administration, the blue was clearly confined to this region of the spinal cord. Blue reaction product was rarely seen below the surface of the cord, regardless of the duration of FeCl₃ exposure, although there were two cats in which blue could be seen as deep as about laminae III.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

One purpose of this study was to determine whether an obligatory synapse exists in the dorsal horn of the spinal cord for the muscle pressor reflex. In this regard, blocking synaptic excitation by administering the GABA_A agonist muscimol over the L₄-S₃ spinal segments abolished the muscle pressor reflex. This was true whether this reflex was evoked by static contraction or muscle stretch. Thus these data suggest that the afferent fibers mediating the pressor reflex synapse in the dorsal horn of the spinal cord and that the ascending projections responsible for this reflex are at least second-order neurons. The second purpose was to determine whether afferent branching is responsible, at least in part, for the fact that multiple spinal segments are involved in producing this reflex. When muscimol administration was confined to the L₇ segment, the segment receiving all of the afferent input, the reflex cardiovascular increases to both contraction and stretch were attenuated but not abolished. These results provide evidence that afferent branching is at least partially responsible for the multisegmental integration of the muscle pressor reflex.

There is abundant evidence to indicate that a synapse in the dorsal horn is involved in mediating the muscle pressor reflex. Administration of antagonists to excitatory amino acid (EAA) receptors into this region attenuates the pressor reflex (1, 9, 10, 12, 23). Consistent with these studies are the data from Hand et al. (8) showing that static muscle contraction increases the release of the EAA glutamate and aspartate into the dorsal horn. In addition, it has been demonstrated that blockade of neurokinin-1 receptors in the dorsal horn reduces the pressor reflex (11, 28). Substance P is the endogenous ligand for neurokinin-1 receptors, and previous work has demonstrated that static contraction causes the release of neuropeptide in the dorsal horn (25, 26). Furthermore, several studies have shown that the muscle pressor reflex can be modulated at the level of the dorsal horn (13, 27). Thus it was not the intent of the present study to simply implicate that a spinal synapse is involved. Rather, this study examined whether or not the dorsal horn synapse is requisite for producing the muscle pressor reflex. The results show that indeed it is.

Muscimol was chosen because it is a potent GABA_A agonist (6, 16). It is well known that activation of GABA receptors opens a chloride channel, and GABA is considered to be an important inhibitory neurotransmitter in the central nervous system (16). Thus activation of GABA receptors inhibits synaptic transmission but not action potentials in the axon. If some of the afferent neurons mediating this reflex did not synapse in the dorsal horn but instead had direct projections to higher central nervous system sites, then the action potentials occurring in these fibers would have been unaltered by muscimol. However, the fact that muscimol abolished the reflex suggests that these fibers do not exist or, if they do, their input alone is insufficient to trigger changes in cardiovascular function. Regardless, these results indicate that synaptic activation of dorsal horn cells is requisite for producing the cardiovascular changes that occur in response to static contraction or muscle stretch.

In the present study, the afferent neurons mediating the muscle pressor reflex were confined to the L₇ segment. We have previously shown that, given this condition, receptor blockade in adjacent segments effectively blunts the muscle pressor reflex (9, 10, 23, 24, 28). Thus synaptic activation occurs within the dorsal horn at segments beyond the segment of afferent neuron entry. Because of this, the well in the present study was created to encompass the L₄-S₃ segments. Although not rigorously tested, it was observed (data not shown) that, when muscimol was placed in the well that covered the L₅-S₂ segments, the muscle pressor reflex was attenuated but not abolished. It was not the purpose of this study to identify the segments involved precisely. Instead, this study was designed to determine whether a dorsal horn synapse is required for expression of the pressor reflex. Nevertheless, the fact that the well had to encompass the L₄-S₃ segments indicates that a considerable range of segments is activated when muscle afferent fibers are stimulated. If one considers the intact, behaving animal, then this result indicates that afferent neurons entering a single segment activate dorsal horn cells over a considerable number of segments. Furthermore, since most areas of the body have overlapping dermatomes and myotomes, a given muscle has afferent input entering at multiple segments, and this input expands over an even greater number of segments. This underscores the complexity of the system and further shows the integrative action of the spinal cord in this reflex.

This multisegmental integration could be the result of afferent branching and/or intraspinal pathways. As indicated above, afferent fibers typically branch on entering the spinal cord, sending rostral and caudal projections (3, 4, 21, 22). On the other hand, second-order neurons in L₇ could activate third-order neurons involved in this reflex in segments beyond the entry one. If this latter suggestion were the sole source of the multisegmental integration, then blocking synaptic activity in L₇ would eliminate this pathway and thus eliminate the pressor reflex. However, the results show that the muscle pressor reflex was only moderately attenuated when muscimol was administered to L₇ alone. Although these data do not completely eliminate the possibility that this intraspinal pathway is important, they do provide compelling evidence that afferent branching is an important component of the multisegmental activation.

The interpretation of the results described in the preceding paragraph are predicated on the belief that muscimol activated a sufficient number of GABA receptors to block synaptic transmission in this region. This seems quite likely. When given over the L₄-S₃ segments, muscimol eliminated the muscle pressor reflex, providing evidence that, indeed, synaptic transmission was blocked. The same dose of muscimol (1.0 mM) was applied to the L₇ region and thus should have had the same effect in this portion of the spinal cord, i.e., blocked synaptic transmission. In addition, this dose most likely represents a tissue concentration of muscimol that is higher than the EC₅₀. The spinal well methodology used in this study was patterned after the work of Beck et al. (2). Using a rat model, they showed that, at a depth of 500 µm below the cord surface, the tissue concentration of neurokinin A was 25-70 times less than the well concentration. Thus it is expected that the tissue concentration of muscimol in the present study is considerably less than the well concentration. If one assumes a twofold decrease in concentration at the most superficial levels, then the concentration of muscimol (10 µM) is still considerably higher than the reported EC₅₀ (10 nM) for this drug (6). Thus in the present study, muscimol may have activated not only GABA_A receptors, but GABA_B receptors as well. In addition, this high tissue concentration of muscimol makes it very unlikely that the drug failed to block synaptic transmission when applied to L₇ alone. However, whether or not a loss of GABA_A specificity occurred is not germane to the present study. Muscimol was used to block synaptic activity in the dorsal horn, regardless of the subtype of receptor activated. Another potential concern using this spinal well methodology is absorption of drug into the bloodstream. However, it was shown that iv muscimol, at a dose that approximates blood uptake (2), had no impact on the pressor reflex. Thus it seems unlikely that circulating muscimol influenced the results.

The inability of the muscle pressor reflex to recover when muscimol was removed from the well may be related to the relatively high tissue concentration of the drug that likely occurred. As indicated, muscimol is a potent GABA_A agonist, and if the tissue concentration were high, it would take a considerable length of time before its actions would terminate (6). We have previously shown that the muscle pressor reflex is stable for a longer time than in the present experiments; therefore, the lack of recovery does not simply reflect an unstable preparation (24, 27, 28). Recovery of the muscle pressor reflex may have occurred if a less potent agent or a lower dose of muscimol had been used. However, it was imperative in this study that a sufficient amount of drug reach the tissue to eliminate synaptic transmission, and thus erring on the side of too much muscimol and thereby preventing recovery was unavoidable. Despite the plausibility that the muscimol tissue concentration was high, the fact that L₇ application only modestly attenuated the muscle pressor reflex demonstrates that axonal conduction remained intact.

Despite the potential disadvantages mentioned above, the methodology employed in this study might be very useful for studying the dorsal horn neurochemistry as it pertains to the muscle pressor reflex. Because the drug solution resides on the surface of the cord for a period of time, this should provide a more constant and even tissue concentration of drug than an intrathecal injection. In addition, this well can encompass several spinal segments simultaneously or be confined to a single spinal segment, as was done in the present study. This is particularly important because multiple neurochemicals, receptors, and segments are involved in producing this reflex (13, 27). Furthermore, because sympathetic preganglionic neurons are absent below approximately L₃, this method does not directly interfere with the activation of sympathetic fibers. Finally, we have preliminary data (not shown) that EAA agonist administration evokes robust and reproducible cardiovascular responses, thereby allowing investigations into mechanisms and/or modulations of EAA-evoked cardiovascular response from the dorsal horn.

In summary, blocking synaptic activity in the dorsal horn of the spinal cord using muscimol abolishes the muscle pressor reflex. This is direct evidence that synaptic transmission in the dorsal horn is a required component of the muscle pressor reflex. This study also showed that blocking synaptic transmission in the dorsal horn at the segment of afferent entry caused a modest attenuation of the pressor reflex. This indicates that the multisegmental processing of the muscle pressor reflex is mediated, at least in part, by branches of the afferent neurons entering the spinal cord.