INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

CARDIOVASCULAR REGULATION is compromised to a large extent by upper thoracic and cervical level spinal cord injuries. As a result, quadriplegics are subject to hypotension, characterized by low levels of tonic sympathetic outflow (16, 19), as well as life-threatening bouts of hypertension. These acute hypertensive periods, known as autonomic dysreflexia, can be triggered by a number of stimuli of somatic and visceral origin, but urinary bladder and rectal distension are the most common irritants (1, 17). In addition to hypertension, symptoms include excessive sweating, bradycardia, flushing of the face, cutaneous vasodilation above the lesion and vasoconstriction below the lesion, nasal obstruction, severe throbbing headache, and piloerection. Generally, the onset of symptoms occurs following the spinal shock phase (11) and may affect as many as 85% of quadriplegics and upper level paraplegics (17). Among this population are patients who may experience periods of hypertension without any of the characteristic symptoms, or "silent" autonomic dysreflexia (9).

Autonomic dysreflexia is characterized by a massive, generalized sympathetic response to stimuli, which are largely innocuous in neurologically intact persons. There is a high incidence of autonomic dysreflexia among patients with lesions above T₆, suggesting that splanchnic sympathetic outflow is involved (2). The substrates that mediate autonomic dysreflexia are presently unknown, but it is generally believed that afferent fibers transmit proprioceptive and nociceptive information from visceral and somatic regions to the spinal cord, stimulating reflex sympathetic vasoconstrictor activity to smooth muscle (7). Activation of the spinal reflex arc results in an exaggerated sympathetic response, but the specific mechanisms, whether central or peripheral, are unclear. Direct recordings from sympathetic efferent fibers to cutaneous and somatic regions in quadriplegics have shown a virtual absence of nerve activity at rest and only moderate increases in response to bladder distension (16, 19). In addition, quadriplegics have significantly lower catecholamine levels and blood pressure than intact subjects at rest (10, 12). However, during hypertensive periods, plasma norepinephrine levels are increased significantly but epinephrine levels are unchanged (10). It is important to note that whereas norepinephrine levels rise significantly during episodic hypertension, these levels never exceed norepinephrine levels in intact subjects at rest. This suggests that, in quadriplegics, alterations in peripheral catecholamine sensitivity may occur, which facilitate a massive sympathetic response. The purpose of this study was to evaluate in a chronic rat model potential peripheral changes in -adrenergic receptor sensitivity as a contributor to the exaggerated sympathetic response characteristic of autonomic dysreflexia.

During the process of developing a chronic animal model, we unexpectedly observed that lesioned Sprague-Dawley rats did not become as hypertensive in response to colorectal distension (CRD) as we expected from previous reports in the literature (3, 14, 15). When lesioned Wistar rats were subjected to the same protocol, they did become hypertensive. From these findings, all experiments were conducted in both Sprague-Dawley and Wistar rats.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Experiments were performed on 14 male Sprague-Dawley rats (Harlan Sprague Dawley, Frederick, MD) and 12 male Wistar rats (Charles Rivers, Wilmington, MA) ranging in weight from 340 to 500 g (mean weight of 417.4 g). Six neurologically intact rats from each strain served as controls, and the remaining rats were given upper thoracic spinal cord lesions. The following groups of rats were subjected to all protocols: intact Sprague-Dawley, lesioned Sprague-Dawley, intact Wistar, and lesioned Wistar. All procedures were approved by the University of Oklahoma Health Sciences Center Institutional Animal Care and Use Committee.

Spinal Cord Transection

Care of Spinal Cord-Injured Rats

Chronic Catheterization

CRD

Experimental Protocol

Intravenous PE Infusion

Data Analysis

For comparisons of responses within each strain of rats, data were analyzed with a three-factor analysis of variance (ANOVA). The three factors were 1) type (lesioned vs. intact), 2) intervention (CRD at all pressures or PE infusion at all doses), and 3) prestimulus MAP or HR vs. MAP or HR elicited in response to stimulation (abbreviated as C or E for the prestimulus control or the experimental values). Factors 2 and 3 were treated as repeated measures. Separate analyses were performed for the Wistar rats and the Sprague-Dawley rats. A two-factor ANOVA, with factors type and intervention, was used to compare changes in resting and response values within the Wistar strain. A four-factor ANOVA, consisting of the same factors for the three-factor ANOVA plus the additional factor of strain, was used to compare responses between strains. Post hoc comparisons were completed using the Newman-Keuls test. All statistical analyses were computed with the CSS Statistica program. Data are presented as the group means ± SE. Differences were considered significant if P < 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

The primary focus of this study was to investigate the potential development of ₁-adrenergic receptor supersensitivity as a mechanism for autonomic dysreflexia in an appropriate animal model. We defined autonomic dysreflexia as an increase in MAP elicited by CRD in spinalized rats that was significantly greater than the resting MAP in intact rats. This definition was based on two factors: 1) the clinical observation that during periods of autonomic dysreflexia, MAP in a quadriplegic person not only increases significantly from resting levels but exceeds those levels considered hypertensive for neurologically intact persons (10); and 2) the use of an animal model previously developed for autonomic dysreflexia in Wistar rats (3). In the process of establishing our chronic model, however, we were surprised to observe that spinalized Sprague-Dawley rats did not meet our definition of autonomic dysreflexia, but Wistar rats did develop dysreflexia in response to CRD. Therefore, Wistar rats provide a better model for autonomic dysreflexia when CRD is used as the stimulus for evoking hypertension. From this finding, the data applicable to the question of autonomic dysreflexia and ₁-receptor supersensitivity are presented only in Wistar rats. These data comprise the first part of RESULTS and include Figs. 1-3. The concluding portion of RESULTS contains data pertaining to Sprague-Dawley rats (Figs. 4 and 5) and the observed differences in responses between Wistar and Sprague-Dawley rats. A summary of the statistical results of the three-way analyses of variance is presented in Table 1.

Effects of CRD on MAP and HR in Wistar Rats

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Response to colorectal distension (CRD) in Wistar rats. Top: there were no significant differences in resting mean arterial pressure (MAP) values within each group. Resting MAP was significantly lower in lesioned rats relative to intact rats. In response to CRD, MAP was significantly increased above resting levels at distension pressures of 20, 40, and 80 mmHg in lesioned rats. Peak MAP was significantly increased above resting levels at distension pressures of 40 and 80 mmHg in intact rats. Bottom: there were no significant differences in resting heart rate (HR) values within each group. Resting HR was significantly increased in lesioned rats relative to intact rats. CRD did not significantly affect HR in lesioned rats at any pressure. HR in intact rats was significantly increased at 80 mmHg.

Elevations in MAP usually reached a plateau within the first 10-15 s of the stimulus and remained at this level for the duration of the 50-s distension period. MAP began to drop immediately after the release of the stimulus and within 2-3 min had returned to prestimulus levels. During the distension procedure, most rats did not visibly react to the stimulus; however, some rats (intact and lesioned) exhibited movements of their hindquarters for a few seconds upon release of distension. Because these movements occurred on release of distension, they did not contribute to the MAP or HR measurements, which were taken before distension (resting value) and during distension (peak response value).

Heart rate. Figure 1 also summarizes the HR values at rest and in response to CRD in intact and lesioned Wistar rats. The results of the ANOVA (Table 1) indicate that the HR response to CRD depended on whether the rats were intact or lesioned. Post hoc analyses revealed the following specific significant differences. Thirty days after transection, resting levels of HR were significantly increased in lesioned rats relative to resting values for intact rats. In response to CRD, HR was not significantly different from resting values at any distension pressure in lesioned rats. In neurologically intact rats, HR was increased significantly above resting values at 80 mmHg but was not changed in response to any other distension pressure.

Figure 2 profiles raw data taken from a single animal from the intact Wistar group (Fig. 2A) and the lesioned Wistar group (Fig. 2B), which are representative of the group MAP responses to CRD at 80 mmHg. Note that the MAP response is greater in the lesioned rat compared with the intact rat. Although in this example HR decreased in the lesioned animal, overall the HR change was not significant. The intact animal experienced an increase in HR.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   MAP and HR responses to CRD at 80 mmHg in Wistar and Sprague-Dawley rats. Time between arrows in each panel indicates 50-s period of CRD. Each panel displays response from a single animal, which is representative of the group peak MAP response to distension pressure at 80 mmHg. A: intact Wistar; B: lesioned Wistar; C: intact Sprague-Dawley; D: lesioned Sprague-Dawley.

Effects of PE on MAP and HR in Wistar Rats

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Responses to phenylephrine (PE) infusion in Wistar rats. Top: there were no significant differences in resting MAP values within each group. In response to PE infusion, peak MAP was significantly increased above resting values at all doses except 0.25 µg/kg iv in both intact and lesioned rats. Bottom: there were no significant differences in resting HR values within each group. In response to PE, at a dose of 4.0 µg/kg, HR was significantly decreased from resting level.

The lack of a significant interaction effect between either type and PE dose or type and C or E suggests that lesioned rats did not exhibit a supersensitive response relative to the intact rats. Because the interaction effect between type and C or E did approach statistical significance (P = 0.085), we performed a two-factor (type and PE dose) ANOVA on the differences between the resting MAP and the MAP in response to PE. This had the effect of emphasizing the change in MAP achieved with PE infusion. The main effect of type was not significant (P = 0.085), confirming that, across the entire doseresponse curve, there was no enhanced response to PE.

On the other hand, the combination of a significant main effect of PE dose (P = 0.001) and a significant interaction effect (P = 0.002) indicated that the dose-response curves in the intact and lesioned rats were not parallel. Post hoc analyses indicated that at the two highest doses of PE, the lesioned rats exhibited larger responses than intact rats. The magnitude of the enhanced responses (mean of 12 and 13 mmHg at doses of 2.0 and 4.0 µg/kg, respectively) were, however, approximately one-half of the reduction in MAP caused by the transection.

Heart rate. Figure 3 also characterizes HR at rest and in response to intravenous infusions of PE in intact and lesioned Wistar rats. The results of the ANOVA (Table 1) indicate that PE produced decreases in HR proportional to the dose. Post hoc analyses revealed that significant decreases in HR were observed in both intact and lesioned rats in response to PE at 4.0 µg/kg, but no significant changes were seen at any other dose.

Effects of CRD on MAP and HR in Sprague-Dawley Rats

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Response to CRD in Sprague-Dawley rats. Top: there were no significant differences in resting MAP values within each group. In response to CRD, peak MAP was significantly increased in intact rats at distension pressures of 20, 40, and 80 mmHg. Lesioned rats experienced significant increases in peak MAP at 40 and 80 mmHg. Bottom: there were no significant differences in resting HR values within each group. In response to CRD, peak HR values were not significantly different from resting values at any distension pressure in either intact or lesioned groups.

Heart rate. In addition, Fig. 4 illustrates HR values at rest and in response to CRD in intact and lesioned Sprague-Dawley rats. ANOVA (Table 1) revealed that the HR response to CRD was dependent on whether the rat was lesioned or intact. Post hoc analyses revealed that resting HR was significantly increased in lesioned rats relative to intact rats when assessed 30 days after a complete spinal transection. However, peak HR responses were not significantly different from resting values at any distension pressure in either intact or lesioned rats.

Figure 2 profiles raw data taken from a single animal from the intact Sprague-Dawley group (Fig. 2C) and the lesioned Sprague-Dawley group (Fig. 2D), which are representative of the group MAP responses to CRD at 80 mmHg. Note that the MAP response was greater in the intact Sprague-Dawley rat compared with the MAP response of the lesioned rat. HR initially increased in the lesioned animal but then decreased during the period of peak MAP response. Conversely, the intact animal experienced an increase in HR concomitant with increased MAP; however, the HR changes for the groups were not significantly different.

Effects of PE on MAP and HR in Sprague-Dawley Rats

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Response to PE infusion in Sprague-Dawley rats. Top: there were no significant differences in resting MAP values within each group. In response to PE infusion, peak MAP was significantly increased above resting levels at all doses in intact rats. Peak MAP was increased in lesioned rats in response to all doses except 0.25 µg/kg iv. Bottom: there were no significant differences in resting HR values within each group. In response to PE, peak HR values were not significantly different at any dose in intact rats. In lesioned rats, peak HR was significantly decreased at 4.0 µg/kg iv.

Heart rate. Figure 5 also characterizes group means for HR at rest and in response to PE in intact and lesioned Sprague-Dawley rats. ANOVA (Table 1) indicated that HR decreased proportional to the dose of PE, and the HR response was dependent on whether the animal was intact or lesioned. Post hoc analysis revealed no significant changes in HR from resting values at any dose of PE in intact rats. In lesioned rats, HR was significantly decreased from resting values only at 4.0 µg/kg.

Differences in Responses Between Wistar and Sprague-Dawley Rats

There were no differences in the HR responses to CRD between strains, and there were no differences in either the MAP or HR responses to PE.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We evaluated pressor responses to PE infusion as an index of peripheral -adrenergic-receptor sensitivity. The major goal of this study was to determine whether a supersensitivity of postsynaptic -adrenergic receptors could account for the enhanced blood pressure responses to CRD following spinal transection. Supersensitivity of -adrenergic receptors was proposed as a possible mechanism to explain the hypertensive component of autonomic dysreflexia. The concept is that removal of the descending excitatory drive to sympathetic preganglionic neurons would lead to an effective sympathetic denervation, with the consequent development of postsynaptic supersensitivity to -adrenergic receptors. Two clinical studies evaluated this hypothesis, and both of them concluded that -adrenergic receptors exhibited enhanced responses to adrenergic agonists. In the first study (12), the arterial pressure responses to infusions of norepinephrine were used as an index of sensitivity of peripheral -adrenergic receptors. Patients with complete cervical spinal cord transections at the levels of C₃-C₈ did exhibit significantly larger increases in arterial pressure in response to infusions of norepinephrine over a wide range of doses (0.12-9.2 µg · kg¹ · min¹) than did intact controls. However, the use of norepinephrine to evaluate supersensitivity is problematic for several reasons. First, norepinephrine is not specific for postsynaptic -adrenergic receptors. Second, because norepinephrine is subject to reuptake under normal conditions, it is troublesome to use this agent to examine supersensitivity in spinal cord-injured populations in which uptake may be altered. Given this, it is unclear whether the enhanced blood pressure response is truly a result of increased peripheral adrenergic sensitivity or rather reductions in the uptake of norepinephrine at presynaptic nerve terminals. Dysfunctional reuptake would lead to greater concentrations of norepinephrine at postsynaptic receptors and increases in MAP. Mathias and Frankel (12) attempted to address this criticism by measuring plasma levels of norepinephrine in quadriplegics and intact subjects before and after infusion of norepinephrine at 0.1 µg · kg¹ · min¹. Before infusion, plasma norepinephrine levels were three times lower in quadriplegics relative to control subjects, but after a 5-min infusion, there were no significant differences between the two groups. It is of interest to note that the dose used for this experiment (0.1 µg · kg¹ · min¹) was lower than any of the doses used to evaluate MAP responses between quadriplegics and intact subjects (0.12-9.2 µg · kg¹ · min¹). Because this low dose was used, instead of a dose that had been demonstrated to produce a hypertensive response, one may speculate that differences in plasma norepinephrine levels would have been present if a higher dose had been used.

The second clinical study attempted to address the question of receptor specificity by comparing the responses of quadriplegics and intact control subjects to infusions of PE (₁-agonist), -methylnorepinephrine (₂-agonist), angiotensin II, and isoprenaline (6). To determine changes in receptor sensitivity, this group of researchers noted the dose that was required to achieve a rise in blood pressure of 20 mmHg or HR of 20 beats/min. Using these criteria, quadriplegics had an increased sensitivity to PE, -methylnorepinephrine, and angiotensin II but not to isoprenaline. Therefore, this study concluded that postsynaptic supersensitivity could account for the enhanced blood pressure responses during bouts of autonomic dysreflexia. PE, a selective ₁-agonist, is a better agent than norepinephrine for evaluating changes in postsynaptic receptor sensitivity. Additionally, because PE is not subject to neuronal uptake, the questions regarding potential dysfunctions in uptake are not an issue. Unfortunately, the conclusion that supersensitivity occurred is based on an arbitrary magnitude of response (20 mmHg). This conclusion would have been more convincing had an increased response been shown for an entire dose-response curve rather than a single point. Additionally, the levels of the lesions in these patients were not presented, which made it difficult to compare the results with those from other laboratories. Thus, considering these two clinical studies, the issue of whether there is a postsynaptic supersensitivity to -adrenergic receptors following a high spinal injury remains unresolved.

Before our investigation, the only animal study that had addressed the possibility of postsynaptic supersensitivity evaluated pressor responses to PE in Sprague-Dawley rats (14). A dose-response curve was constructed to analyze pressor responses to bolus injections of PE at doses ranging from 0.2 to 1.0 µg/kg, 7 days after cervical spinal transection. At 7 days postlesion, there were no significant differences between spinalized and intact rats. However, because autonomic dysreflexia cannot be evoked at 7 days after transection (3), we wanted to test postsynaptic receptor sensitivity in an animal model that more accurately reflected the clinical condition. We hypothesized that 30 days after transection, rats that became severely hypertensive in response to CRD would likewise be more responsive to PE infusions. Unexpectedly, we observed no consistent differences between lesioned and intact rats in response to PE. At low doses of PE the responses of intact and lesioned rats were similar, but at the two highest doses the lesioned rats exhibited larger responses than the intact rats. The magnitude of the enhanced response was, however, approximately one-half of the baseline change in MAP due to the lesion. Thus the shifted baseline might account for the enhanced responses to PE. One might suppose that if -adrenergic supersensitivity was the major mechanism contributing to the increased responses to CRD after spinal lesion, then the entire PE dose-response curve in lesioned animals would show supersensitive responses. On balance, then, our study fails to provide strong support for the concept that the hypertensive periods evoked by CRD in chronic, spinalized rats are a result of ₁-adrenergic-receptor changes in the periphery.

Three potential limitations to our study should be kept in mind. First, the numbers of animals in each group were not large, and some of the P values were close to significant (Table 1). If there were more animals per group, then these nonsignificant values might become significant. Although this is true, if one presumes that the mean values in response to each intervention were similar after the addition of animals to each group, then the major conclusions of this study would not be altered. Second, we did not determine the maximal effect for PE; therefore, complete dose-response curves were not obtained. Our conclusions regarding supersensitivity might have differed had complete curves been compared. On the other hand, the magnitude of the mean increase in MAP with the highest dose of PE was comparable to the change in MAP achieved with CRD. Therefore, we believe that higher doses of PE would not be "physiologically" relevant, and hence would not alter our conclusions. Third, the differences in baseline MAP between the intact and lesioned groups could complicate interpretation of data. As alluded to in the previous paragraph, we surmise that a lower baseline would facilitate the likelihood of finding evidence of supersensitivity, but this was not observed at all doses. Therefore, although there are potential limitations in this study, we do not believe that they materially affected the conclusions.

If the changes in neural function that lead to autonomic dysreflexia are not primarily localized in the periphery, then functional elements in the central nervous system must undergo a reorganization. Indeed, recent studies indicate that significant neuronal remodeling does occur after spinal cord injury in a chronic rat model (4, 5). Further investigation of the specific circuits involved in mediating autonomic dysreflexia may reveal that trauma to the spinal cord stimulates plastic changes in spinal reflex arcs that ultimately result in exaggerated sympathetic responses.

The chronic animal model used in our investigation was based largely on models used in two earlier studies (3, 15). In the process of developing our model, we observed that Sprague-Dawley rats did not exhibit autonomic dysreflexia (e.g., peak MAP response significantly greater than resting MAP) in response to CRD 30 days after spinal transection. In contrast, lesioned Wistar rats did become dysreflexic, with pressor responses near 50 mmHg at the most intense stimulus. To our knowledge there have been no studies examining autonomic dysreflexia in chronically lesioned Sprague-Dawley rats, but our results are consistent with the findings of an earlier study using Wistar rats (3). The difference in pressor response between strains suggests that a genetic component is present, which renders the Wistar rats more susceptible to periods of episodic hypertension following a chronic spinal cord injury. Thus Wistar rats provide a better model for autonomic dysreflexia than Sprague-Dawley rats.

Clinically, the onset of autonomic dysreflexia differs from patient to patient (8), but in general, autonomic dysreflexia does not occur until after the spinal shock phase has ended (11). Because of this, it was essential that our model include an appropriate period of recovery to ensure autonomic dysreflexia could be evoked. One study examined the time course of the development of autonomic dysreflexia in Wistar rats (3). Twenty-four hours after midthoracic spinal cord transection, MAP is increased ~40 mmHg in response to CRD. At days 5 and 7 in this same group of rats, the pressor responses were diminished significantly. However, at day 30 the pressor responses returned and even exceeded the magnitude of hypertension experienced at day 1 in response to CRD. The authors concluded that the depression of responses from days 5 to 7 suggests that different mechanisms may be responsible for the hypertension experienced at day 1 and day 30. The loss of descending inhibitory command to sympathetic outflow most likely results in the hypertensive response elicited during the first 24 h, whereas the depression of responses at days 5 through 7, followed by a vigorous return at day 30, may occur as a result of remodeling and/or reorganization of the spinal reflex arc (3).

In studies with Sprague-Dawley rats, MAP increased ~25 mmHg in response to bladder distension, one day after spinal transection of C₇-T₁ (15). At days 3, 5, and 7, the pressor response to bladder distension in this same group of rats was not significantly different from that seen at day 1. In another study, pressor responses to CRD were evaluated in Sprague-Dawley rats 1-3 days after spinal transection (13). Rats with transection at C₁ or T₆ demonstrated significantly depressed pressor responses relative to intact rats with a rise in MAP of only 10 mmHg. The different results of these two studies with Sprague-Dawley rats might be explained by the use of different stimuli to elicit changes in blood pressure. To our knowledge, there have been no recorded studies of this nature using spinalized Sprague-Dawley rats with recovery periods of more than 7 days. From the findings of these studies with Wistar and Sprague-Dawley rats, we decided a 30-day recovery period would provide an appropriate time frame to ensure that autonomic dysreflexia could be elicited.

The apparent strain difference in the response to CRD after spinal transection is, at present, unexplained. This difference does suggest that there is a genetic component to the pressor response to CRD. In humans, the incidence of autonomic dysreflexia after high spinal lesion is not 100%. An implication of our results is that a genetic component may partially account for the observation that some humans develop autonomic dysreflexia but others do not. The observations that lesioned Wistar rats exhibited greater pressor responses to CRD than intact Wistar rats and lesioned Sprague-Dawley rats were less responsive than intact rats lead to the hypothesis that the descending pathways that modulate reflex changes in blood pressure have a restraining influence in Wistar rats but a facilitatory influence in Sprague-Dawley rats.

With regard to HR, earlier studies using cervical spinalized rats have shown HR to be significantly decreased 24 h after transection (14, 15). Conversely, chronic spinalized rats were observed to have increased HR, 30 days after midthoracic transection (3). Our results are consistent with the findings of the latter study and indicate that resting HR may be increased to maintain cardiac output as a compensation for decreased MAP in lesioned animals. In response to CRD, HR was increased in intact rats concomitant with increased MAP; however, spinal transection eliminated increases in HR, suggesting that a supraspinal loop is involved in mediating the HR response to CRD. As expected, HR was significantly decreased in response to PE at a dose of 4 µg/kg in both intact and lesioned animals.

In summary, our study suggests that ₁-adrenergic supersensitivity does not contribute significantly to the enhanced blood pressure increases in response to CRD in rats 30 days after an upper thoracic spinal transection. In addition, the lack of similar blood pressure responses to CRD in lesioned rats in the Wistar and Sprague-Dawley strains suggests that there is a genetic component to the spinal reflex elicited by CRD.