Serotonin2 receptors mediate respiratory recovery after cervical spinal cord hemisection in adult rats
http://www.100md.com
《应用生理学杂志》
Department of Anatomy and Cell Biology, Wayne State University School of Medicine, Detroit, Michigan 48201
ABSTRACT
The aim of the present study was to specifically investigate the involvement of serotonin [5-hydroxytryptamine (5-HT2)] receptors in 5-HT-mediated respiratory recovery after cervical hemisection. Experiments were conducted on C2 spinal cord-hemisected, anesthetized (chloral hydrate, 400 mg/kg ip), vagotomized, pancuronium- paralyzed, and artificially ventilated female Sprague-Dawley rats in which CO2 levels were monitored and maintained. Twenty-four hours after spinal hemisection, the ipsilateral phrenic nerve displayed no respiratory-related activity indicative of a functionally complete hemisection. Intravenous administration of the 5-HT2A/2C-receptor agonist (±)-2,5-dimethoxy-4-iodoamphetamine hydrochloride (DOI) induced respiratory-related activity in the phrenic nerve ipsilateral to hemisection under conditions in which CO2 was maintained at constant levels and augmented the activity induced under conditions of hypercapnia. The effects of DOI were found to be dose dependent, and the recovery of activity could be maintained for up to 2 h after a single injection. DOI-induced recovery was attenuated by the 5-HT2-receptor antagonist ketanserin but not with the 5-HT2C-receptor antagonist RS-102221, suggesting that 5-HT2A and not necessarily 5-HT2C receptors may be involved in the induction of respiratory recovery after cervical spinal cord injury.
keywords:spinal cord injury; phrenic motoneurons; plasticity; ketanserin; respiration
INTRODUCTION
CERVICAL SPINAL CORD HEMISECTION at the C2 level disrupts brain stem bulbospinal impulses from the rostral ventral respiratory group to the phrenic motoneurons (PMNs) located in the ventral horns at midcervical levels (4, 10). This type of lesion interrupts normal respiratory drive to the PMNs, resulting in a quiescent ipsilateral phrenic nerve and paralysis of the ipsilateral hemidiaphragm. Anatomically, descending fibers from the rostral ventral respiratory group make ipsilateral, contralateral, and bilateral axonal connections with PMNs (8, 9), yet immediately after cervical hemisection, the spinal decussating axons cannot depolarize PMNs (9). Within hours after hemisection, however, plasticity converts these functionally ineffective synapses to functionally latent connections in that they do not restore reflex activity under normal conditions. The latent connections become functionally effective only after contralateral phrenicotomy or under conditions of asphyxia induced several hours after spinal hemisection. The asphyxia results in functional recovery of the hemidiaphragm paralyzed by spinal cord injury (9). The connections form the "crossed phrenic pathway" (CPP) and escape injury by descending in the spinal cord contralateral to the lesion before crossing the midline to innervate PMNs ipsilateral and caudal to the C2 hemisection site (8, 9, 26). The recovered respiratory-related activity in the initially quiescent phrenic nerve under the above conditions has been referred to as "crossed phrenic nerve activity" (CPNA) (29).
Whereas the complete mechanisms driving CPNA are not currently known, evidence from our laboratory suggests that it may be mediated, in part, by serotonin [5-hydroxytryptamine (5-HT)] neurotransmission (11, 39, 40). Serotonin-containing fibers have been anatomically identified to project to, and distribute terminals near, PMNs (3, 16, 17, 31, 39). Interestingly, after C2 hemisection, the 5-HT afferents demonstrate elements of plasticity evidenced by increases in axodendritic and axosomatic terminals within the ipsilateral phrenic nucleus (37). The plasticity may enhance CPNA and thus contribute to recovery of the paralyzed hemidiaphragm. Physiologically, reduction of 5-HT with the 5-HT synthesis inhibitor p-chlorophenylalanine before cervical hemisection attenuates the normal asphyxia-induced CPNA, suggesting that the recovery is dependent on sufficient levels of 5-HT (11). This hypothesis was recently supported in our laboratory with data demonstrating that administration of the 5-HT precursor 5-hydroxytryptophan (5-HTP) induced CPNA when CO2 levels were maintained and augmented the asphyxia-induced CPNA in rats after cervical hemisection (40, 41). The increases were prevented with the general 5-HT-receptor antagonist methysergide, suggesting that 5-HT has the relevant characteristics to induce CPNA after cervical spinal cord hemisection. Given the broad-spectrum affinity of methysergide for multiple 5-HT receptors, the present study was designed to specifically target 5-HT2 receptors as a potential mechanism mediating hemidiaphragm recovery after cervical spinal cord hemisection. To investigate this, phrenic nerve activity in hemisected rats was monitored after the administration of the 5-HT2A/2C-receptor agonist (±)-2,5-dimethoxy-4-iodoamphetamine hydrochloride (DOI), in the presence or absence of 5-HT2- (ketanserin) and 5-HT2C- (RS-102221) receptor antagonists.
MATERIALS AND METHODS
Surgical procedures. Adult female Sprague-Dawley rats (n = 27; 250-350 g; Harlan) were used in these studies in accordance with the National Institutes of Health Guide to the Care and Use of Laboratory Animals and institutional approval by Wayne State University Animal Investigation Committee. Rats were randomly assigned to groups and injected with atropine sulfate (0.1 mg/kg im) 10 min before being anesthetized with 4% chloral hydrate (400 mg/kg ip). Animals were then prepared for cervical hemisection. After a dorsal midline incision in the neck, the paravertebral muscles were retracted, and a laminectomy of the second cervical vertebra was performed. The cervical spinal cord was exposed after a small incision of the dura mater and arachnoid. With the use of microscissors, the spinal cord was hemisected just caudal to the left dorsal root at C2. The muscle layers were sutured using interrupted silk, wound clips were utilized to close the skin incision, and the antibiotic Betadine was applied to the incision site. One day after surgery, animals were again pretreated with atropine and anesthetized as described above. Catheters were inserted into the left femoral artery and vein for recording arterial blood pressure (pressure monitor BP-1, World Precision Instruments) and intravenous drug administration, respectively. After insertion of a tracheal cannula, animals were paralyzed with pancuronium bromide (0.5 mg/kg iv) and then artificially ventilated with room air by a small-animal ventilator (Harvard Apparatus rodent ventilator). CO2 levels were recorded using a CO2 monitor (Datex, Normocap). A homeothermic blanket control unit (Harvard) was used to maintain the body temperature at 37 ± 1°C. Animals were bilaterally vagotomized to avoid locking respiratory drive to the cycle of the ventilator.
Neural recordings. Both left (ipsilateral to hemisection) and right (contralateral to hemisection) phrenic nerves were exposed in the neck via a ventral approach. The nerves were isolated, desheathed, and transected. A suture was passed through the cut end of the central stump of the phrenic nerve. Neural recordings were made monophasically with the suture tied to the caudal pole and the central nerve stump contacting the rostral pole of platinum bipolar electrodes. This procedure was adopted to minimize movement of the nerve during recording, especially as predrug and postdrug quantitative comparisons of nerve activity were conducted. Neural activity was filtered (bandwidth: 0.1-3 kHz), amplified (520 Tektronix, gain 5-20 K), and displayed on-line using a Cambridge Electronic Design (CED) 1401 data-acquisition system. Signals were also fed into a videotape recorder for off-line data analysis with Spike 2 (CED) software.
Experimental protocols. The functional completeness of each hemisection was verified first. After a left spinal cord hemisection at C2, the right phrenic nerve shows inspiratory-related activity. The left phrenic nerve, however, typically shows no respiratory-related activity because of the disruption of the bulbospinal respiratory pathways after cervical hemisection. Hemisections were considered functionally complete if the ipsilateral phrenic nerve showed a complete absence of respiratory-related activity. Only those animals showing a functionally complete hemisection were selected for the experiments. In the experiments designed to assess drug effects under conditions in which CO2 levels were monitored and maintained, CO2 levels were maintained by adjusting the rate (60-80 breaths/min) or stroke volume (3-5 ml) of the ventilator.
Other experiments were designed to assess drug effects during asphyxia. Briefly, the ventilation rate was lowered to increase CO2 levels >35 Torr for 60 s, and then asphyxia (hypoxia and hypercapnia) was induced by turning off the ventilator. During asphyxia, increasing respiratory-related activity in the right phrenic nerve reached a maximum just before the activity terminated. The ventilator was turned back on a few seconds after burst activity terminated to resuscitate the animal. In addition, CPNA was observed in the left phrenic nerve. Experiments were carried out for 5-HT2-receptor effects only in those animals that had similar CPNA results from at least two respiratory stress tests, separated by a 15-min interval. During drug testing, asphyxia was induced 5 min after each drug administration, and the animal was then allowed to recover for at least 5 min before the next drug administration.
Drug treatments. All drug treatments were administered systemically (via the femoral vein) to rats that received left cervical (C2) spinal hemisections 24 h earlier. All compounds were obtained from Tocris Cookson and included the 5- HT2A/2C-receptor agonist DOI, the 5-HT2-receptor antagonist ketanserin, the 5-HT2C-receptor antagonist RS-102221, the 5-HT2C-receptor agonist MK-212 hydrochloride, and the general 5-HT-receptor antagonist methysergide. To test the effects of 5-HT2A/2C-receptor stimulation on CPNA during asphyxia, DOI (0.2 mg/kg iv) was infused 5 min after asphyxia-induced CPNA. CPNA was assessed 5 min after DOI infusion, followed by administration of methysergide (4.0 mg/kg). The effects of methysergide on the DOI-induced CPNA in asphyxic rats were then recorded 5 min later.
In the group of animals designated to assess dose-dependent effects (CO2 is maintained), DOI was given in initial doses of 0.05 mg/kg and increased in successive dose increments at 15-min intervals, resulting in cumulative doses of 0.1, 0.2, 0.5, 1.0, and 2.0 mg/kg. Phrenic nerve activity was recorded from both nerves at each dose. To assess the temporal maintenance of respiratory recovery after DOI administration, phrenic nerve activity was monitored 1, 3, 5, 15, and 25 min after a single dose of DOI (0.2 mg/kg). A final time point was recorded 2 h after DOI administration (data not shown). From the DOI dose-response data, the dose (0.2 mg/kg) was chosen based on its ability to induce phrenic motor recovery effectively.
To discriminate the effects of DOI from 5-HT2A and/or 5-HT2C receptors, the antagonists ketanserin (5-HT2) and RS-102221 (5-HT2C) were utilized under conditions in which CO2 levels were monitored and maintained. CPNA was assessed 5 min after a single injection of DOI (1.0 mg/kg). Ketanserin (2.0 mg/kg dose; Ref. 19) or RS-102221 (2.0 mg/kg dose; Ref. 2) was subsequently infused, and the effects on DOI-mediated CPNA were recorded 5 min later. To counteract the adverse effects of ketanserin on blood pressure, epinephrine (1:5,000) was administered. In a separate group of animals, ketanserin was also administered to rats before DOI injection.
Last, to assess the impact of 5-HT2C-receptor stimulation on intact respiratory function, MK-212 (0.1 mg/kg) was administered, and phrenic nerve activity was recorded 3, 5, 10, and 50 min later. The dose utilized (0.1 mg/kg) was chosen through preliminary dose-response experiments in the laboratory before the study, which showed that any dose administered >0.1 mg/kg shut down respiratory activity. It should be noted that no animal in the study received postsurgical analgesia. One might argue that pain-induced stress after surgery may affect spinal synaptic pathways and is responsible for some of the observed results of the study. This is not likely because predrug phrenic nerve recordings, taken under general anesthesia in this study, showed no apparent pain-induced respiratory activity (see RESULTS).
Statistical analysis. Filtered nerve activity was rectified and integrated (time constant, 100 ms) by a moving averager (MA-821, CWE) and then quantitatively analyzed using the Spike2 (CED) software. For evaluation of the effect of chemical injection on phrenic nerve activity, comparisons of percent changes in intensity of inspiratory-related activity before and after drug administration were made in each nerve. The intensity was estimated by determining the area under the integrated curve after subtracting background activity (noise plus spontaneous tonic activity). Background activity level was first determined by measuring activity during the expiratory phases. Background activity during the inspiratory phase was estimated by extrapolating a line between the adjacent background activity level in the expiratory phases. The areas of the last five consecutive inspiratory-related bursts (gasps) during asphyxia were then measured automatically by the computer software after the cursors was set between the onset of the first burst and the termination of the fifth burst. The last five bursts before the onset of apnea were usually the largest. From this value (total area between the cursors), background activity was subtracted. Measurements of nerve activity are expressed as a percent change from either predrug, intact right phrenic nerve activity (R-PNA) or as a percentage of the respective asphyxia-induced maximal nerve response.
Values in the text are expressed as means ± SE. For multiple-group comparisons involving dose response and 5-HT2A/2C-receptor antagonists, as well as analysis of respiratory frequency (bursts/min) and blood pressure changes, a repeated-measures ANOVA was conducted to assess the overall effects of drug treatments with F ratios reported (GB-STAT for MS Windows, version 5.4). Individual comparisons between groups were subsequently conducted using a Fisher least significant difference post hoc analysis (significance at P < 0.05). The present study utilized area under the curve (AUC) analysis as a measure of motor activity instead of conventional peak amplitudes of the phrenic nerve bursts, because AUC analysis is a more direct measure of the entire phrenic nerve motor output profile, whereas amplitude analysis alone only measures a single component of motor activity. In fact, previous work from our laboratory utilizing both measures found no significant differences in conclusions or outcome between amplitude alone or AUC analysis (11). Eldridge (6) showed that phrenic nerve measures using the "moving average" yielded values with the same form and meaning as the episodic true integration, concluding that the moving average can provide a good index of changes in respiratory and/or phrenic nerve output and/or function.
RESULTS
Effects of DOI on CPNA during asphyxia. Neurogram tracings in Fig. 1 are qualitative but are representative of similar results obtained from five animals. Specifically, Fig. 1 shows the influence of 5-HT2A/2C-receptor stimulation with DOI on CPNA induced by temporary asphyxia 24 h after C2 spinal cord hemisection. Figure 1A shows that there was no obvious respiratory-related activity in the left (ipsilateral to hemisection) phrenic nerve before drug administration, indicative of a functionally complete cervical spinal hemisection. After cessation of mechanical ventilator support to induce asphyxia, and before drug administration, CPNA in the left phrenic nerve was induced (Fig. 1B). Five minutes after systemic administration of DOI (0.2 mg/kg), the asphyxia-induced CPNA in the left phrenic nerve is apparently augmented to levels above that induced by temporary asphyxia alone (Fig. 1C). The DOI-enhanced CPNA under conditions of asphyxia was blocked 5 min after infusion of the general 5-HT-receptor antagonist methysergide (4.0 mg/kg; Fig. 1D). Whereas this paradigm utilizing repetitive bouts of asphyxia may be perceived to alter subsequent respiratory responses (e.g., tachyphylaxis), previously published data from our laboratory yielded reproducible results under a similar asphyxia-based paradigm (41).
Effects of DOI on CPNA without asphyxia. Figure 2 demonstrates the effects of systemic DOI administration on respiratory activity in a rat in which CO2 levels were monitored and maintained at a constant level. Figure 2 exemplifies the dose-dependent effects of systemic DOI on phrenic nerve activity in rats after C2 spinal cord hemisection. Predrug baseline demonstrates pronounced respiratory-related activity in the right phrenic nerve and an absence of respiratory-related activity in the left phrenic nerve, indicative of a functionally complete cervical spinal cord hemisection. DOI at cumulative doses of 0.05, 0.1, 0.2, 0.5, 1.0, and 2.0 mg/kg induced and gradually increased CPNA in the left phrenic nerve under conditions of maintained CO2 levels. At the higher doses (1.0-2.0 mg/kg), DOI did not significantly alter burst amplitude but increased burst rate activity (bursts/min) noticeably in the right, contralateral phrenic nerve compared with predrug controls (Figs. 2 and 3). Interestingly, DOI-induced respiratory recovery in the left nerve was sustained after a single injection (Fig. 3). Figure 3F demonstrates that, after a single administration of DOI (0.2 mg/kg), CPNA is maintained up to 25 min later. In fact, DOI-induced CPNA in the left phrenic nerve persisted for up to 2 h after a single administration (data not shown).
Effects of DOI and 5-HT2A/2C-receptor antagonists on respiratory recovery. Figures 4 and 5 demonstrate the results of DOI-induced CPNA under conditions of maintained CO2 in the presence or absence of antagonist compounds for both 5-HT2 (ketanserin) and 5-HT2C receptors (RS-102221). Figure 4 qualitatively demonstrates neurogram tracings for each group, whereas a quantitative summary of the results is shown in Fig. 5. Predrug analysis demonstrates a complete absence of respiratory-related activity in the left phrenic nerve, once again indicative of a functionally complete cervical spinal hemisection (Fig. 4A). Five minutes after a single intravenous injection of DOI (1.0 mg/kg), CPNA is detected in the left phrenic nerve (Fig. 4B). Subsequent administration of the 5-HT2C-receptor antagonist RS-102221 (2.0 mg/kg) appeared to decrease qualitatively respiratory burst rate induced by DOI but did not quantitatively alter the DOI-induced CPNA significantly (Fig. 4C and Table 1). Conversely, 5 min after a single dose of the 5-HT2-receptor antagonist ketanserin (2.0 mg/kg), the DOI-mediated CPNA was significantly attenuated but not completely abolished (Fig. 4D). These burst-amplitude data are consistent with the effects of DOI and antagonists on burst rate (bursts/min; Table 1). Specifically, DOI at all doses significantly increased respiratory rate (range: 63.3 ± 4.5 to 74.8 ± 6.2 bursts/min) compared with predrug controls (44.3 ± 4.8 bursts/min; Table 1). Interestingly, injection of RS-102221 did not significantly decrease the DOI-mediated increases in respiratory rate (64.1 ± 6.7 bursts/min); however, ketanserin reduced the DOI-evoked increases in respiratory rate to levels not significantly different from that of predrug controls (52.4 ± 4.4 bursts/min). It should also be noted that, whereas RS-102221 appeared to reduce blood pressure in Fig. 4C, Table 1 demonstrates no significant quantitative changes in blood pressure. Conversely, because ketanserin alone significantly lowers blood pressure (70.3 ± 4.6 mmHg; Table 1), epinephrine (1:5,000) was also given to rule out the possible effects of blood pressure on the attenuation of DOI-mediated CPNA. Figure 4E demonstrates that the addition of epinephrine, while normalizing blood pressure, did not significantly alter the ketanserin-mediated attenuation of the DOI-induced CPNA from that of ketanserin alone. For comparative purposes, Fig. 4F demonstrates the CPNA induced by asphyxia before any drug administration. It should also be noted that ketanserin, when administered in animals before DOI treatment, did not alter predrug neurogram profiles in either nerve yet attenuated the DOI-mediated respiratory recovery after cervical spinal cord hemisection (data not shown).
Figure 5 quantitatively summarizes the dose-dependent effects of DOI as well as the influences of 5- HT2A/2C-receptor antagonism on respiratory activity in both phrenic nerves. Figure 5A summarizes the data as a percentage of predrug R-PNA under conditions in which CO2 levels were maintained. Data are expressed as a percentage of the intact contralateral (right) nerve, because predrug respiratory activity in the ipsilateral (left) phrenic nerve is zero. Asterisks show significantly different levels compared with predrug R-PNA. Left phrenic nerve activity (L-PNA) in predrug controls showed a complete absence of functional activity (0.0% of R-PNA). However, DOI significantly increased L-PNA compared with predrug R-PNA at doses of 0.05 (8.6 ± 1.5%), 0.1 (11.0 ± 1.9%), 0.2 (23.4 ± 3.8%), 0.5 (27.3 ± 4.8%), 1.0 (33.6 ± 8.5%), and 2.0 mg/kg (30.3 ± 12.3% of predrug R-PNA), but all DOI-treated group levels, including antagonist treated and asphyxia induced, remained significantly less than R-PNA levels. RS-102221 did not significantly alter DOI-induced CPNA (25.2 ± 3.7%), whereas ketanserin significantly attenuated the DOI-evoked recovery at doses of 0.5-2.0 mg/kg (significantly different from DOI + ketanserin) but did not completely abolish the DOI-mediated CPNA (6.1 ± 1.7%). Interestingly, at higher doses (0.2-2.0 mg/kg), DOI-induced increases in CPNA were not significantly different from predrug asphyxia-induced levels in the left phrenic nerve. R-PNA was not significantly altered under any drug combination compared with predrug right phrenic nerve controls, with the exception of asphyxia-induced increases (198.4 ± 30.9%; Fig. 5A).
Figure 5B summarizes the drug-induced data derived from each nerve under conditions in which CO2 levels were maintained and presented as a percentage of the predrug asphyxic response from each respective nerve. Data were compared with the respective asphyxic control, representing maximal respiratory response in each nerve to assess what percentage of the maximal effect is elicited after 5-HT2A/2C-receptor stimulation. By setting the asphyxic response as maximum (100%) CPNA, activity in the left nerve predrug (1.0 ± 0.9%) or after DOI doses at 0.05 mg/kg (24.1 ± 2.8%) and at 0.1 mg/kg (31.9 ± 6.7%) was significantly less than asphyxia-induced activity in the respective left control nerve as indicated. However, L-PNA, after DOI at 0.2, 0.5, 1.0, and 2.0 mg/kg, was not significantly different from asphyxia-induced activity in the respective left control nerve. Subsequent application of RS-102221, although trending above DOI-induced levels, did not significantly alter DOI-induced L-PNA from maximal asphyxic control levels. However, after ketanserin, DOI-induced normalization (in 0.2-2.0 mg/kg DOI groups) was significantly different and/or attenuated (17.7 ± 8.1%; compared with DOI + ketanserin) but not completely abolished. R-PNA was significantly less in all experimental and predrug (range: 33.7 ± 9.7 to 61.2 ± 8.6%) groups compared with respective right phrenic asphyxia-response controls.
5-HT2C-receptor stimulation temporarily depresses respiratory activity. Figure 6 demonstrates the qualitative effects of 5-HT2C-receptor stimulation on respiratory activity at various time points after a single injection of the 5-HT2C-receptor agonist MK-212 (0.1 mg/kg). Burst amplitude in the intact, right phrenic nerve is apparently depressed at 3, 5, and 10 min after a single injection yet appeared to return to predrug levels within 50 min. However, burst rate 50 min after a single injection of MK-212 was qualitatively increased (Fig. 6). The effects of MK-212 on the quiescent left phrenic nerve after cervical hemisection are not shown because no noticeable effects could be detected.
DISCUSSION
Serotonin2A, but not serotonin2C, receptors are likely involved in respiratory recovery after cervical spinal cord hemisection. The present study demonstrates that the 5-HT2A/2C-receptor agonist DOI elicits CPNA in the ipsilateral phrenic nerve rendered quiescent after a C2 spinal cord hemisection. DOI-mediated increases in burst amplitude and respiratory rate (bursts/min) were attenuated with ketanserin and not with RS-102221. Although not absolutely conclusive, these results strongly implicate 5-HT2A- and not 5-HT2C-receptor linkage to respiratory recovery in this model. Elevated amplitude in phrenic nerve burst activity is thought to be related to an increase in tidal volume (6) and is a normal response to asphyxia (36). The present study demonstrated an increase in integrated phrenic nerve waveforms recorded from both contralateral and ipsilateral nerves during asphyxia. Qualitatively, DOI augmented burst amplitudes in the left nerve above those of asphyxia-induced levels alone, replicating previous data that suggested that increases in 5-HT neurotransmission may also augment asphyxia-induced CPNA (40). Functionally, the results suggest that PMNs ipsilateral to hemisection likely do not reach maximum output capacity during asphyxia, and thus there is an opportunity for 5-HT2 receptors to enhance the CPNA.
Stimulation of 5-HT2 receptors in the present study also activated CPNA under conditions in which CO2 levels were maintained. This suggests that 5-HT2 receptors do not require asphyxia-induced cues to activate the CPP. In fact, under these conditions, DOI induced CPNA <1 min after injection and maintained burst activity for up to 2 h. The DOI-mediated increases in CPNA under the above conditions matched levels induced by asphyxia, indicating that 5-HT2 receptors alone are capable of completely activating the CPP. DOI-induced CPNA is dose dependent and leads to a gradual increase in burst amplitude and frequency, both of which are attenuated by ketanserin and not RS-102221. Interestingly, at higher doses, DOI evoked prolonged inspiratory burst duration. Because the medullary raphe nuclei have postinspiratory and expiratory-related bulbospinal projections (7) to the PMNs (18), it is possible that 5-HT2 receptors associated with these bulbospinal pathways were activated. Alternatively, it is possible that 5-HT2 receptors produced a long-lasting depolarization of the PMN membrane potential, because 5-HT has been shown to induce this excitatory effect at the spinal level (23). This is very likely because our laboratory recently demonstrated that PMNs colocalize 5-HT2A- and not 5-HT2C-receptor mRNA expression (1), suggesting that PMNs are capable of 5-HT2A-receptor biosynthesis, which may be focally linked with the postsynaptic regulation of PMN activity. Therefore, it is possible that higher doses of DOI may lead to overstimulation of 5-HT2 receptors and PMN excitation, leading to prolonged burst activity. The notion that 5-HT2 receptors may be regulating CPNA at the level of the spinal cord is supported by studies showing that microapplication of 5-HT into the phrenic nucleus of the rabbit increased the peak integrated amplitude of phrenic nerve activity, an effect prevented by the 5-HT-receptor antagonist methysergide (33). Moreover, excitation of PMNs, after electrical stimulation of the raphe obscuris (13-15), is blocked with methysergide (14) and more specifically with the 5-HT2-receptor antagonists ketanserin (23) or SR-46349B (12). Together, these observations, combined with the present results, point to the likely possibility that 5-HT2A receptors may directly regulate the excitation of PMNs at the level of the spinal cord. Stimulation of these receptors leads to an enhancement of phrenic burst amplitude.
In the present study, DOI also mediated increases in respiratory rate (bursts/min), under conditions of both maintained CO2 levels and asphyxia. An increase in respiratory frequency during asphyxia in rats is a normal response to hypercapnia (35, 36). DOI-mediated increases in respiratory rate under conditions of maintained CO2 and asphyxia were attenuated with ketanserin but not RS-102221. This finding also points to the possibility that 5-HT2A receptors may also regulate CPNA at medullary respiratory centers. The likelihood is supported by previous work in which application of 5-HT (27) or 5-HTP (28) to an in vitro bath containing rat brain stem-spinal cord explant preparations from newborn rats increased respiratory-related discharge frequencies recorded from the cervical ventral roots. The increases were prevented either by reduction in 5-HT with the 5-HT synthesis inhibitor p-chlorophenylalanine, or by general 5-HT-receptor antagonism with methysergide (28). Moreover, 5-HT microinjection into the rostral ventrolateral medulla also increased the frequency of respiratory-related activity in the cervical ventral roots (5). Electrical stimulation of the nucleus raphé obscuris increased the frequency of respiratory activity recorded from the ipsilateral phrenic nerve, implicating the influence of 5-HT afferents on medullary respiratory centers that regulate respiratory rhythm (13, 25). The above observations, combined with the present results, suggest that 5-HT2 receptors, most likely 2A, may also mediate effects at the level of the medulla after C2 spinal cord hemisection.
It should be noted that ketanserin produces hypotension after intravenous administration. To rule out the possibility that hypotension caused the attenuation of DOI-induced CPNA, epinephrine was coadministered with ketanserin in the present study. The administration of epinephrine normalized blood pressure to levels seen before ketanserin but did not alter the depression of the DOI-induced CPNA compared with ketanserin alone. It is generally accepted that ketanserin is more selective for 5-HT2A receptors but is a less effective blocker of 5-HT2C receptors (21, 22, 32, 38). Thus 5-HT2A- rather than 5-HT2C-receptor activation is most likely involved in the induction of the CPNA after C2 spinal cord hemisection. It is noteworthy to mention that, although ketanserin is a poor 5-HT2C-receptor antagonist, it also shows only moderate binding affinity for histamine H1 and 1-adrenergic-receptor binding sites, binds very weakly to dopamine receptors, and is inactive in other known neurotransmitter-receptor binding assays (21). Therefore, given the binding characteristics of ketanserin, the results of the present study strongly implicate 5-HT2A-receptor-mediated mechanisms yet do not fully rule out the involvement of 5-HT2C receptors. However, with the growing availability of more specific 5-HT2A-receptor antagonists (MDL-100907), future studies may be able to discriminate more selectively 5-HT2A-receptor-mediated mechanisms in the spinal cord.
In the present study, RS-102221, a 5-HT2C-receptor antagonist, did not significantly attenuate either burst amplitude or rate after DOI-induced CPNA. This result, combined with a lack of colocalization of 5-HT2C-receptor mRNA expression with PMNs (1), suggests that 5-HT2C receptors may not be functionally linked to PMNs at the level of the spinal cord. Interestingly, specific stimulation of 5-HT2C receptors with MK-212 temporarily depressed phrenic nerve burst amplitudes in the intact right phrenic nerve. Similar observations have been reported in which 5-HT2C-receptor stimulation depressed respiratory rate in cats (20) and in vitro brain stem preparations (30). In addition, it has recently been shown that MK-212 inhibited neurons of the nucleus tractus solitarius, an effect blocked by RS-102221 (34). Together, these observations suggest that the influence of 5-HT2C receptors on respiratory activity may be isolated to brain stem respiratory centers with little or no direct regulation over PMNs.
Functional importance of serotonin2-receptor-mediated respiratory recovery. The present study provides strong evidence that 5-HT2, most likely 2A and not 2C, receptors may be functionally linked to the activation of respiratory pathways that become disrupted after cervical spinal cord injury. These findings are in direct agreement with previous work that implicated 5-HT in the activation of latent CPPs (24). Electrical stimulation of the lateral funiculus, contralateral to cervical hemisection, prompted short-latency responses in the ipsilateral phrenic nerve (24). Administration of 5-HTP induced tonic activity in the phrenic nerve, as well as two long latency responses. Similar findings were recently reported from our laboratory in which 5-HTP induced CPNA ipsilateral to cervical hemisection. This effect was prevented with methysergide (40, 41). Collectively, these data suggest that 5-HT may convert ineffective CPPs in the spinal cord to effective pathways capable of inducing CPNA ipsilateral to cervical hemisection. The present data, showing increases in burst rate and amplitude, implicate excitatory 5-HT2-receptor mechanisms, likely mediated at both medullary and spinal cord levels. Whereas the pharmacological outcomes of the present study strongly point to a 5-HT2A-receptor-mediated mechanism, additional studies will be required with more specific 5-HT2A-receptor antagonists to prove that the 5-HT2A receptor alone is responsible for these effects.
ACKNOWLEDGEMENTS
This work was supported by National Institute of Child Health and Human Development Grant HD-31550 (to H. G. Goshgarian).
FOOTNOTES
Address for reprint requests and other correspondence: H. G. Goshgarian, Dept. of Anatomy and Cell Biology, Wayne State Univ.-School of Medicine, 540 East Canfield Ave., Detroit, MI 48201 (E-mail: hgoshgar@med.wayne.edu).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 24 May 2001; accepted in final form 8 August 2001.
REFERENCES
1.Basura, GJ, Zhou SY, Walker PD, and Goshgarian HG. Distribution of serotonin 2A and 2C receptor mRNA expression in the cervical ventral horn and phrenic motoneurons following acute spinal cord hemisection. Exp Neurol 169: 255-263, 2001.
2.Bonhaus, DW, Weinhardt KK, Taylor M, Desouza A, McNeely PM, Szczepanski K, Fontana DJ, Trinh J, Rocha CL, Dawson MW, Flippin LA, and Eglen RM. RS-102221: a novel high affinity and selective 5-HT2C receptor antagonist. Neuropharmacology 36: 621-629, 1997.
3.Bowker, RM, Westlund KN, and Coulter JD. Origins of serotonergic projections to the spinal cord in rat: an immunocytochemical-retrograde transport study. Brain Res 226: 187-199, 1981.
4.DeVries, KL, and Goshgarian HG. Spinal cord localization and characterization of the neurons which give rise to the accessory phrenic nerve in the adult rat. Exp Neurol 104: 88-90, 1989.
5.Di Pasquale, E, Morin D, Monteau R, and Hilaire G. Serotonergic modulation of the respiratory rhythm generator at birth: an in vitro study in the rat. Neurosci Lett 143: 91-95, 1992.
6.Eldridge, FL. Quantification of electrical activity in the phrenic nerve in the study of ventilatory control. Chest 70: 154-157, 1976.
7.Gilbey, PM, Futuro-Neto HA, and Zhou SY. Respiratory-related discharge patterns of caudal raphe neurones projecting to the upper thoracic spinal cord in the rat. J Auton Nerv Syst 50: 263-273, 1995.
8.Goshgarian, HG, Ellenberger HH, and Feldman JL. Decussation of bulbospinal respiratory axons at the level of the phrenic nuclei in adult rats: a possible substrate for the crossed phrenic phenomenon. Exp Neurol 111: 135-139, 1991.
9.Goshgarian, HG, and Guth L. Demonstration of functionally ineffective synapses in the guinea pig spinal cord. Exp Neurol 57: 613-621, 1977.
10.Goshgarian, HG, and Rafols JA. The phrenic nucleus of the albino rat: a correlative HRP and Golgi study. J Comp Neurol 201: 441-456, 1981.
11.Hadley, SD, Walker PD, and Goshgarian HG. Effects of the serotonin synthesis inhibitor p-CPA on the expression of the crossed phrenic phenomenon 4 h following C2 spinal cord hemisection. Exp Neurol 160: 446-453, 1998.
12.Hilaire, G, Bou C, and Monteau R. Serotonergic modulation of central respiratory activity in the neonatal mouse: an in vitro study. Eur J Pharmacol 329: 115-120, 1997.
13.Holtman, JR, Anastasi NC, Norman WP, and Dretchen KL. Effect of electrical and chemical stimulation of the raphe obscurus on phrenic nerve activity in the cat. Brain Res 362: 214-220, 1986.
14.Holtman, JR, Dick TE, and Berger AJ. Involvement of serotonin in the excitation of phrenic motoneurons evoked by stimulation of the raphe obscurus. J Neurosci 6: 1185-1193, 1986.
15.Holtman, JR, Dick TE, and Berger AJ. Serotonin-mediated excitation of recurrent laryngeal and phrenic motoneurons evoked by stimulation of the raphe obscurus. Brain Res 417: 12-20, 1987.
16.Holtman, JR, Norman WP, and Gillis RA. Projections from the raphe nuclei to the phrenic motor nucleus in the cat. Neurosci Lett 44: 105-111, 1984.
17.Holtman, JR, Norman WP, Skirboll L, Dretchen KL, Cuello C, Visser TJ, H?kfelt T, and Gillis RA. Evidence for 5-hydroxytryptamine, substance P, and thyrotropin-releasing hormone in neurons innervating the phrenic motor nucleus. J Neurosci 4: 1064-1071, 1984.
18.Hosogai, M, Matsuo S, Sibahara T, and Kawai Y. Projection of respiratory neurons in rat medullary raphe nuclei to the phrenic nucleus. Respir Physiol 112: 37-50, 1998.
19.Kinkead, R, and Mitchell GS. Time-dependent hypoxic ventilatory responses in rats: effects of ketanserin and 5-carboxamidotryptamine. Am J Physiol Regulatory Integrative Comp Physiol 277: R658-R666, 1999.
20.Lalley, PM. The excitability and rhythm of medullary respiratory neurons in the cat are altered by the serotonin receptor agonist 5-methoxy-N,N,dimethyltryptamine. Brain Res 648: 87-98, 1994.
21.Leysen, JE, Awouters F, Kennis L, Laduron PM, Vandenberk J, and Janssen PAJ Receptor binding profile of R 41 468, a novel antagonist at 5-HT2-receptors. Life Sci 28: 1015-1022, 1981.
22.Leysen, JE, Niemegeer CJE, Van Neuten JM, and Laduron PM. [3H]ketanserin (R 41 468), as selective 3H-ligand for serotonin2 receptor binding sites: binding properties, brain distribution and functional role. Mol Pharmacol 21: 304-314, 1982.
23.Lindsay, AD, and Feldman JL. Modulation of respiratory activity of neonatal rat phrenic motoneurons by serotonin. J Physiol 461: 213-233, 1993.
24.Ling, L, Bach KB, and Mitchell GS. Serotonin reveals ineffective spinal pathways to contralateral phrenic motoneurons in spinally hemisected rats. Exp Brain Res 101: 35-43, 1994.
25.Millhorn, DE. Stimulation of raphe (obscurus) nucleus causes long-term potentiation of phrenic nerve activity in cat. J Physiol 381: 169-179, 1986.
26.Moreno, DE, Yu X, and Goshgarian HG. Identification of the axon pathways which mediate functional recovery of a paralyzed hemidiaphragm following spinal cord hemisection in the adult rat. Exp Neurol 116: 219-228, 1992.
27.Morin, D, Hennequin S, Monteau R, and Hilaire G. Serotonergic influences of central respiratory activity: an in vitro study in the newborn rat. Brain Res 535: 281-287, 1990.
28.Morin, D, Monteau R, and Hilaire G. 5-Hydroxytryptamine modulates central respiratory activity in the newborn rat: an in vitro study. Eur J Pharmacol 192: 89-95, 1991.
29.O'Hara, TE, and Goshgarian HG. Quantitative assessment of phrenic nerve functional recovery mediated by the crossed phrenic reflex at various time intervals after spinal cord injury. Exp Neurol 111: 244-250, 1990.
30.Onimaru, H, Shamoto A, and Homma I. Modulation of respiratory rhythm by 5-HT in the brainstem spinal cord preparation from newborn rat. Pflügers Arch 435: 485-494, 1998.
31.Pilowsky, PM, Castro D, Llewellyn-Smith I, Lipski J, and Voss MD. Serotonin immunoreactive boutons make synapses with feline phrenic motoneurons. J Neurosci 10: 1091-1098, 1990.
32.Sanders-Bush, E, and Breeding M. Putative selective 5-HT2 antagonists block serotonin 5-HT1C receptors in the choroid plexus. J Pharmacol Exp Ther 247: 169-173, 1988.
33.Schmid, K, Bohmer G, and Merkelbach S. Serotonergic control of phrenic motoneuronal activity at the level of the spinal cord of the rabbit. Neurosci Lett 116: 204-209, 1990.
34.Sevoz-Couche, C, Spyer KM, and Jordan D. Inhibition of rat nucleus tractus solitarius neurones by activation of 5-HT2C receptors. Neuroreport 11: 1785-1790, 2000.
35.St. John, WM, and Bianchi AL. Responses of bulbospinal and laryngeal respiratory neurons to hypercapnia and hypoxia. J Appl Physiol 59: 1201-1207, 1985.
36.St. John, WM, and Wang SC. Response of medullary respiratory neurons to hypercapnia and isocapnic hypoxia. J Appl Physiol 43: 812-821, 1977.
37.Tai, Q, Palazzolo KL, and Goshgarian HG. Synaptic plasticity of 5-hydroxytryptamine-immunoreactive terminals in the phrenic nucleus following spinal cord injury. A quantitative electron microscopic analysis. J Comp Neurol 386: 613-624, 1997.
38.Wolf, WA, and Schutz LJ. The serotonin 5-HT2C receptor is a prominent serotonin receptor in basal ganglia: evidence from functional studies on serotonin-mediated phosphoinositide hydrolysis. J Neurochem 69: 1449-1458, 1997.
39.Zhan, WZ, Ellenberger HH, and Feldman JL. Monoaminergic and GABAergic terminations in phrenic nucleus of rat identified by immunohistochemical labeling. Neuroscience 31: 105-113, 1989.
40.Zhou, SY, and Goshgarian HG. Effects of serotonin on crossed phrenic nerve activity in cervical spinal cord hemisected rats. Exp Neurol 160: 446-453, 1999.
41.Zhou, SY, and Goshgarian HG. 5-Hydroxytryptophan-induced respiratory recovery after cervical spinal cord hemisection in rats. J Appl Physiol 89: 1528-1536, 2000.(Shi-Yi Zhou, Gregory J. B)
ABSTRACT
The aim of the present study was to specifically investigate the involvement of serotonin [5-hydroxytryptamine (5-HT2)] receptors in 5-HT-mediated respiratory recovery after cervical hemisection. Experiments were conducted on C2 spinal cord-hemisected, anesthetized (chloral hydrate, 400 mg/kg ip), vagotomized, pancuronium- paralyzed, and artificially ventilated female Sprague-Dawley rats in which CO2 levels were monitored and maintained. Twenty-four hours after spinal hemisection, the ipsilateral phrenic nerve displayed no respiratory-related activity indicative of a functionally complete hemisection. Intravenous administration of the 5-HT2A/2C-receptor agonist (±)-2,5-dimethoxy-4-iodoamphetamine hydrochloride (DOI) induced respiratory-related activity in the phrenic nerve ipsilateral to hemisection under conditions in which CO2 was maintained at constant levels and augmented the activity induced under conditions of hypercapnia. The effects of DOI were found to be dose dependent, and the recovery of activity could be maintained for up to 2 h after a single injection. DOI-induced recovery was attenuated by the 5-HT2-receptor antagonist ketanserin but not with the 5-HT2C-receptor antagonist RS-102221, suggesting that 5-HT2A and not necessarily 5-HT2C receptors may be involved in the induction of respiratory recovery after cervical spinal cord injury.
keywords:spinal cord injury; phrenic motoneurons; plasticity; ketanserin; respiration
INTRODUCTION
CERVICAL SPINAL CORD HEMISECTION at the C2 level disrupts brain stem bulbospinal impulses from the rostral ventral respiratory group to the phrenic motoneurons (PMNs) located in the ventral horns at midcervical levels (4, 10). This type of lesion interrupts normal respiratory drive to the PMNs, resulting in a quiescent ipsilateral phrenic nerve and paralysis of the ipsilateral hemidiaphragm. Anatomically, descending fibers from the rostral ventral respiratory group make ipsilateral, contralateral, and bilateral axonal connections with PMNs (8, 9), yet immediately after cervical hemisection, the spinal decussating axons cannot depolarize PMNs (9). Within hours after hemisection, however, plasticity converts these functionally ineffective synapses to functionally latent connections in that they do not restore reflex activity under normal conditions. The latent connections become functionally effective only after contralateral phrenicotomy or under conditions of asphyxia induced several hours after spinal hemisection. The asphyxia results in functional recovery of the hemidiaphragm paralyzed by spinal cord injury (9). The connections form the "crossed phrenic pathway" (CPP) and escape injury by descending in the spinal cord contralateral to the lesion before crossing the midline to innervate PMNs ipsilateral and caudal to the C2 hemisection site (8, 9, 26). The recovered respiratory-related activity in the initially quiescent phrenic nerve under the above conditions has been referred to as "crossed phrenic nerve activity" (CPNA) (29).
Whereas the complete mechanisms driving CPNA are not currently known, evidence from our laboratory suggests that it may be mediated, in part, by serotonin [5-hydroxytryptamine (5-HT)] neurotransmission (11, 39, 40). Serotonin-containing fibers have been anatomically identified to project to, and distribute terminals near, PMNs (3, 16, 17, 31, 39). Interestingly, after C2 hemisection, the 5-HT afferents demonstrate elements of plasticity evidenced by increases in axodendritic and axosomatic terminals within the ipsilateral phrenic nucleus (37). The plasticity may enhance CPNA and thus contribute to recovery of the paralyzed hemidiaphragm. Physiologically, reduction of 5-HT with the 5-HT synthesis inhibitor p-chlorophenylalanine before cervical hemisection attenuates the normal asphyxia-induced CPNA, suggesting that the recovery is dependent on sufficient levels of 5-HT (11). This hypothesis was recently supported in our laboratory with data demonstrating that administration of the 5-HT precursor 5-hydroxytryptophan (5-HTP) induced CPNA when CO2 levels were maintained and augmented the asphyxia-induced CPNA in rats after cervical hemisection (40, 41). The increases were prevented with the general 5-HT-receptor antagonist methysergide, suggesting that 5-HT has the relevant characteristics to induce CPNA after cervical spinal cord hemisection. Given the broad-spectrum affinity of methysergide for multiple 5-HT receptors, the present study was designed to specifically target 5-HT2 receptors as a potential mechanism mediating hemidiaphragm recovery after cervical spinal cord hemisection. To investigate this, phrenic nerve activity in hemisected rats was monitored after the administration of the 5-HT2A/2C-receptor agonist (±)-2,5-dimethoxy-4-iodoamphetamine hydrochloride (DOI), in the presence or absence of 5-HT2- (ketanserin) and 5-HT2C- (RS-102221) receptor antagonists.
MATERIALS AND METHODS
Surgical procedures. Adult female Sprague-Dawley rats (n = 27; 250-350 g; Harlan) were used in these studies in accordance with the National Institutes of Health Guide to the Care and Use of Laboratory Animals and institutional approval by Wayne State University Animal Investigation Committee. Rats were randomly assigned to groups and injected with atropine sulfate (0.1 mg/kg im) 10 min before being anesthetized with 4% chloral hydrate (400 mg/kg ip). Animals were then prepared for cervical hemisection. After a dorsal midline incision in the neck, the paravertebral muscles were retracted, and a laminectomy of the second cervical vertebra was performed. The cervical spinal cord was exposed after a small incision of the dura mater and arachnoid. With the use of microscissors, the spinal cord was hemisected just caudal to the left dorsal root at C2. The muscle layers were sutured using interrupted silk, wound clips were utilized to close the skin incision, and the antibiotic Betadine was applied to the incision site. One day after surgery, animals were again pretreated with atropine and anesthetized as described above. Catheters were inserted into the left femoral artery and vein for recording arterial blood pressure (pressure monitor BP-1, World Precision Instruments) and intravenous drug administration, respectively. After insertion of a tracheal cannula, animals were paralyzed with pancuronium bromide (0.5 mg/kg iv) and then artificially ventilated with room air by a small-animal ventilator (Harvard Apparatus rodent ventilator). CO2 levels were recorded using a CO2 monitor (Datex, Normocap). A homeothermic blanket control unit (Harvard) was used to maintain the body temperature at 37 ± 1°C. Animals were bilaterally vagotomized to avoid locking respiratory drive to the cycle of the ventilator.
Neural recordings. Both left (ipsilateral to hemisection) and right (contralateral to hemisection) phrenic nerves were exposed in the neck via a ventral approach. The nerves were isolated, desheathed, and transected. A suture was passed through the cut end of the central stump of the phrenic nerve. Neural recordings were made monophasically with the suture tied to the caudal pole and the central nerve stump contacting the rostral pole of platinum bipolar electrodes. This procedure was adopted to minimize movement of the nerve during recording, especially as predrug and postdrug quantitative comparisons of nerve activity were conducted. Neural activity was filtered (bandwidth: 0.1-3 kHz), amplified (520 Tektronix, gain 5-20 K), and displayed on-line using a Cambridge Electronic Design (CED) 1401 data-acquisition system. Signals were also fed into a videotape recorder for off-line data analysis with Spike 2 (CED) software.
Experimental protocols. The functional completeness of each hemisection was verified first. After a left spinal cord hemisection at C2, the right phrenic nerve shows inspiratory-related activity. The left phrenic nerve, however, typically shows no respiratory-related activity because of the disruption of the bulbospinal respiratory pathways after cervical hemisection. Hemisections were considered functionally complete if the ipsilateral phrenic nerve showed a complete absence of respiratory-related activity. Only those animals showing a functionally complete hemisection were selected for the experiments. In the experiments designed to assess drug effects under conditions in which CO2 levels were monitored and maintained, CO2 levels were maintained by adjusting the rate (60-80 breaths/min) or stroke volume (3-5 ml) of the ventilator.
Other experiments were designed to assess drug effects during asphyxia. Briefly, the ventilation rate was lowered to increase CO2 levels >35 Torr for 60 s, and then asphyxia (hypoxia and hypercapnia) was induced by turning off the ventilator. During asphyxia, increasing respiratory-related activity in the right phrenic nerve reached a maximum just before the activity terminated. The ventilator was turned back on a few seconds after burst activity terminated to resuscitate the animal. In addition, CPNA was observed in the left phrenic nerve. Experiments were carried out for 5-HT2-receptor effects only in those animals that had similar CPNA results from at least two respiratory stress tests, separated by a 15-min interval. During drug testing, asphyxia was induced 5 min after each drug administration, and the animal was then allowed to recover for at least 5 min before the next drug administration.
Drug treatments. All drug treatments were administered systemically (via the femoral vein) to rats that received left cervical (C2) spinal hemisections 24 h earlier. All compounds were obtained from Tocris Cookson and included the 5- HT2A/2C-receptor agonist DOI, the 5-HT2-receptor antagonist ketanserin, the 5-HT2C-receptor antagonist RS-102221, the 5-HT2C-receptor agonist MK-212 hydrochloride, and the general 5-HT-receptor antagonist methysergide. To test the effects of 5-HT2A/2C-receptor stimulation on CPNA during asphyxia, DOI (0.2 mg/kg iv) was infused 5 min after asphyxia-induced CPNA. CPNA was assessed 5 min after DOI infusion, followed by administration of methysergide (4.0 mg/kg). The effects of methysergide on the DOI-induced CPNA in asphyxic rats were then recorded 5 min later.
In the group of animals designated to assess dose-dependent effects (CO2 is maintained), DOI was given in initial doses of 0.05 mg/kg and increased in successive dose increments at 15-min intervals, resulting in cumulative doses of 0.1, 0.2, 0.5, 1.0, and 2.0 mg/kg. Phrenic nerve activity was recorded from both nerves at each dose. To assess the temporal maintenance of respiratory recovery after DOI administration, phrenic nerve activity was monitored 1, 3, 5, 15, and 25 min after a single dose of DOI (0.2 mg/kg). A final time point was recorded 2 h after DOI administration (data not shown). From the DOI dose-response data, the dose (0.2 mg/kg) was chosen based on its ability to induce phrenic motor recovery effectively.
To discriminate the effects of DOI from 5-HT2A and/or 5-HT2C receptors, the antagonists ketanserin (5-HT2) and RS-102221 (5-HT2C) were utilized under conditions in which CO2 levels were monitored and maintained. CPNA was assessed 5 min after a single injection of DOI (1.0 mg/kg). Ketanserin (2.0 mg/kg dose; Ref. 19) or RS-102221 (2.0 mg/kg dose; Ref. 2) was subsequently infused, and the effects on DOI-mediated CPNA were recorded 5 min later. To counteract the adverse effects of ketanserin on blood pressure, epinephrine (1:5,000) was administered. In a separate group of animals, ketanserin was also administered to rats before DOI injection.
Last, to assess the impact of 5-HT2C-receptor stimulation on intact respiratory function, MK-212 (0.1 mg/kg) was administered, and phrenic nerve activity was recorded 3, 5, 10, and 50 min later. The dose utilized (0.1 mg/kg) was chosen through preliminary dose-response experiments in the laboratory before the study, which showed that any dose administered >0.1 mg/kg shut down respiratory activity. It should be noted that no animal in the study received postsurgical analgesia. One might argue that pain-induced stress after surgery may affect spinal synaptic pathways and is responsible for some of the observed results of the study. This is not likely because predrug phrenic nerve recordings, taken under general anesthesia in this study, showed no apparent pain-induced respiratory activity (see RESULTS).
Statistical analysis. Filtered nerve activity was rectified and integrated (time constant, 100 ms) by a moving averager (MA-821, CWE) and then quantitatively analyzed using the Spike2 (CED) software. For evaluation of the effect of chemical injection on phrenic nerve activity, comparisons of percent changes in intensity of inspiratory-related activity before and after drug administration were made in each nerve. The intensity was estimated by determining the area under the integrated curve after subtracting background activity (noise plus spontaneous tonic activity). Background activity level was first determined by measuring activity during the expiratory phases. Background activity during the inspiratory phase was estimated by extrapolating a line between the adjacent background activity level in the expiratory phases. The areas of the last five consecutive inspiratory-related bursts (gasps) during asphyxia were then measured automatically by the computer software after the cursors was set between the onset of the first burst and the termination of the fifth burst. The last five bursts before the onset of apnea were usually the largest. From this value (total area between the cursors), background activity was subtracted. Measurements of nerve activity are expressed as a percent change from either predrug, intact right phrenic nerve activity (R-PNA) or as a percentage of the respective asphyxia-induced maximal nerve response.
Values in the text are expressed as means ± SE. For multiple-group comparisons involving dose response and 5-HT2A/2C-receptor antagonists, as well as analysis of respiratory frequency (bursts/min) and blood pressure changes, a repeated-measures ANOVA was conducted to assess the overall effects of drug treatments with F ratios reported (GB-STAT for MS Windows, version 5.4). Individual comparisons between groups were subsequently conducted using a Fisher least significant difference post hoc analysis (significance at P < 0.05). The present study utilized area under the curve (AUC) analysis as a measure of motor activity instead of conventional peak amplitudes of the phrenic nerve bursts, because AUC analysis is a more direct measure of the entire phrenic nerve motor output profile, whereas amplitude analysis alone only measures a single component of motor activity. In fact, previous work from our laboratory utilizing both measures found no significant differences in conclusions or outcome between amplitude alone or AUC analysis (11). Eldridge (6) showed that phrenic nerve measures using the "moving average" yielded values with the same form and meaning as the episodic true integration, concluding that the moving average can provide a good index of changes in respiratory and/or phrenic nerve output and/or function.
RESULTS
Effects of DOI on CPNA during asphyxia. Neurogram tracings in Fig. 1 are qualitative but are representative of similar results obtained from five animals. Specifically, Fig. 1 shows the influence of 5-HT2A/2C-receptor stimulation with DOI on CPNA induced by temporary asphyxia 24 h after C2 spinal cord hemisection. Figure 1A shows that there was no obvious respiratory-related activity in the left (ipsilateral to hemisection) phrenic nerve before drug administration, indicative of a functionally complete cervical spinal hemisection. After cessation of mechanical ventilator support to induce asphyxia, and before drug administration, CPNA in the left phrenic nerve was induced (Fig. 1B). Five minutes after systemic administration of DOI (0.2 mg/kg), the asphyxia-induced CPNA in the left phrenic nerve is apparently augmented to levels above that induced by temporary asphyxia alone (Fig. 1C). The DOI-enhanced CPNA under conditions of asphyxia was blocked 5 min after infusion of the general 5-HT-receptor antagonist methysergide (4.0 mg/kg; Fig. 1D). Whereas this paradigm utilizing repetitive bouts of asphyxia may be perceived to alter subsequent respiratory responses (e.g., tachyphylaxis), previously published data from our laboratory yielded reproducible results under a similar asphyxia-based paradigm (41).
Effects of DOI on CPNA without asphyxia. Figure 2 demonstrates the effects of systemic DOI administration on respiratory activity in a rat in which CO2 levels were monitored and maintained at a constant level. Figure 2 exemplifies the dose-dependent effects of systemic DOI on phrenic nerve activity in rats after C2 spinal cord hemisection. Predrug baseline demonstrates pronounced respiratory-related activity in the right phrenic nerve and an absence of respiratory-related activity in the left phrenic nerve, indicative of a functionally complete cervical spinal cord hemisection. DOI at cumulative doses of 0.05, 0.1, 0.2, 0.5, 1.0, and 2.0 mg/kg induced and gradually increased CPNA in the left phrenic nerve under conditions of maintained CO2 levels. At the higher doses (1.0-2.0 mg/kg), DOI did not significantly alter burst amplitude but increased burst rate activity (bursts/min) noticeably in the right, contralateral phrenic nerve compared with predrug controls (Figs. 2 and 3). Interestingly, DOI-induced respiratory recovery in the left nerve was sustained after a single injection (Fig. 3). Figure 3F demonstrates that, after a single administration of DOI (0.2 mg/kg), CPNA is maintained up to 25 min later. In fact, DOI-induced CPNA in the left phrenic nerve persisted for up to 2 h after a single administration (data not shown).
Effects of DOI and 5-HT2A/2C-receptor antagonists on respiratory recovery. Figures 4 and 5 demonstrate the results of DOI-induced CPNA under conditions of maintained CO2 in the presence or absence of antagonist compounds for both 5-HT2 (ketanserin) and 5-HT2C receptors (RS-102221). Figure 4 qualitatively demonstrates neurogram tracings for each group, whereas a quantitative summary of the results is shown in Fig. 5. Predrug analysis demonstrates a complete absence of respiratory-related activity in the left phrenic nerve, once again indicative of a functionally complete cervical spinal hemisection (Fig. 4A). Five minutes after a single intravenous injection of DOI (1.0 mg/kg), CPNA is detected in the left phrenic nerve (Fig. 4B). Subsequent administration of the 5-HT2C-receptor antagonist RS-102221 (2.0 mg/kg) appeared to decrease qualitatively respiratory burst rate induced by DOI but did not quantitatively alter the DOI-induced CPNA significantly (Fig. 4C and Table 1). Conversely, 5 min after a single dose of the 5-HT2-receptor antagonist ketanserin (2.0 mg/kg), the DOI-mediated CPNA was significantly attenuated but not completely abolished (Fig. 4D). These burst-amplitude data are consistent with the effects of DOI and antagonists on burst rate (bursts/min; Table 1). Specifically, DOI at all doses significantly increased respiratory rate (range: 63.3 ± 4.5 to 74.8 ± 6.2 bursts/min) compared with predrug controls (44.3 ± 4.8 bursts/min; Table 1). Interestingly, injection of RS-102221 did not significantly decrease the DOI-mediated increases in respiratory rate (64.1 ± 6.7 bursts/min); however, ketanserin reduced the DOI-evoked increases in respiratory rate to levels not significantly different from that of predrug controls (52.4 ± 4.4 bursts/min). It should also be noted that, whereas RS-102221 appeared to reduce blood pressure in Fig. 4C, Table 1 demonstrates no significant quantitative changes in blood pressure. Conversely, because ketanserin alone significantly lowers blood pressure (70.3 ± 4.6 mmHg; Table 1), epinephrine (1:5,000) was also given to rule out the possible effects of blood pressure on the attenuation of DOI-mediated CPNA. Figure 4E demonstrates that the addition of epinephrine, while normalizing blood pressure, did not significantly alter the ketanserin-mediated attenuation of the DOI-induced CPNA from that of ketanserin alone. For comparative purposes, Fig. 4F demonstrates the CPNA induced by asphyxia before any drug administration. It should also be noted that ketanserin, when administered in animals before DOI treatment, did not alter predrug neurogram profiles in either nerve yet attenuated the DOI-mediated respiratory recovery after cervical spinal cord hemisection (data not shown).
Figure 5 quantitatively summarizes the dose-dependent effects of DOI as well as the influences of 5- HT2A/2C-receptor antagonism on respiratory activity in both phrenic nerves. Figure 5A summarizes the data as a percentage of predrug R-PNA under conditions in which CO2 levels were maintained. Data are expressed as a percentage of the intact contralateral (right) nerve, because predrug respiratory activity in the ipsilateral (left) phrenic nerve is zero. Asterisks show significantly different levels compared with predrug R-PNA. Left phrenic nerve activity (L-PNA) in predrug controls showed a complete absence of functional activity (0.0% of R-PNA). However, DOI significantly increased L-PNA compared with predrug R-PNA at doses of 0.05 (8.6 ± 1.5%), 0.1 (11.0 ± 1.9%), 0.2 (23.4 ± 3.8%), 0.5 (27.3 ± 4.8%), 1.0 (33.6 ± 8.5%), and 2.0 mg/kg (30.3 ± 12.3% of predrug R-PNA), but all DOI-treated group levels, including antagonist treated and asphyxia induced, remained significantly less than R-PNA levels. RS-102221 did not significantly alter DOI-induced CPNA (25.2 ± 3.7%), whereas ketanserin significantly attenuated the DOI-evoked recovery at doses of 0.5-2.0 mg/kg (significantly different from DOI + ketanserin) but did not completely abolish the DOI-mediated CPNA (6.1 ± 1.7%). Interestingly, at higher doses (0.2-2.0 mg/kg), DOI-induced increases in CPNA were not significantly different from predrug asphyxia-induced levels in the left phrenic nerve. R-PNA was not significantly altered under any drug combination compared with predrug right phrenic nerve controls, with the exception of asphyxia-induced increases (198.4 ± 30.9%; Fig. 5A).
Figure 5B summarizes the drug-induced data derived from each nerve under conditions in which CO2 levels were maintained and presented as a percentage of the predrug asphyxic response from each respective nerve. Data were compared with the respective asphyxic control, representing maximal respiratory response in each nerve to assess what percentage of the maximal effect is elicited after 5-HT2A/2C-receptor stimulation. By setting the asphyxic response as maximum (100%) CPNA, activity in the left nerve predrug (1.0 ± 0.9%) or after DOI doses at 0.05 mg/kg (24.1 ± 2.8%) and at 0.1 mg/kg (31.9 ± 6.7%) was significantly less than asphyxia-induced activity in the respective left control nerve as indicated. However, L-PNA, after DOI at 0.2, 0.5, 1.0, and 2.0 mg/kg, was not significantly different from asphyxia-induced activity in the respective left control nerve. Subsequent application of RS-102221, although trending above DOI-induced levels, did not significantly alter DOI-induced L-PNA from maximal asphyxic control levels. However, after ketanserin, DOI-induced normalization (in 0.2-2.0 mg/kg DOI groups) was significantly different and/or attenuated (17.7 ± 8.1%; compared with DOI + ketanserin) but not completely abolished. R-PNA was significantly less in all experimental and predrug (range: 33.7 ± 9.7 to 61.2 ± 8.6%) groups compared with respective right phrenic asphyxia-response controls.
5-HT2C-receptor stimulation temporarily depresses respiratory activity. Figure 6 demonstrates the qualitative effects of 5-HT2C-receptor stimulation on respiratory activity at various time points after a single injection of the 5-HT2C-receptor agonist MK-212 (0.1 mg/kg). Burst amplitude in the intact, right phrenic nerve is apparently depressed at 3, 5, and 10 min after a single injection yet appeared to return to predrug levels within 50 min. However, burst rate 50 min after a single injection of MK-212 was qualitatively increased (Fig. 6). The effects of MK-212 on the quiescent left phrenic nerve after cervical hemisection are not shown because no noticeable effects could be detected.
DISCUSSION
Serotonin2A, but not serotonin2C, receptors are likely involved in respiratory recovery after cervical spinal cord hemisection. The present study demonstrates that the 5-HT2A/2C-receptor agonist DOI elicits CPNA in the ipsilateral phrenic nerve rendered quiescent after a C2 spinal cord hemisection. DOI-mediated increases in burst amplitude and respiratory rate (bursts/min) were attenuated with ketanserin and not with RS-102221. Although not absolutely conclusive, these results strongly implicate 5-HT2A- and not 5-HT2C-receptor linkage to respiratory recovery in this model. Elevated amplitude in phrenic nerve burst activity is thought to be related to an increase in tidal volume (6) and is a normal response to asphyxia (36). The present study demonstrated an increase in integrated phrenic nerve waveforms recorded from both contralateral and ipsilateral nerves during asphyxia. Qualitatively, DOI augmented burst amplitudes in the left nerve above those of asphyxia-induced levels alone, replicating previous data that suggested that increases in 5-HT neurotransmission may also augment asphyxia-induced CPNA (40). Functionally, the results suggest that PMNs ipsilateral to hemisection likely do not reach maximum output capacity during asphyxia, and thus there is an opportunity for 5-HT2 receptors to enhance the CPNA.
Stimulation of 5-HT2 receptors in the present study also activated CPNA under conditions in which CO2 levels were maintained. This suggests that 5-HT2 receptors do not require asphyxia-induced cues to activate the CPP. In fact, under these conditions, DOI induced CPNA <1 min after injection and maintained burst activity for up to 2 h. The DOI-mediated increases in CPNA under the above conditions matched levels induced by asphyxia, indicating that 5-HT2 receptors alone are capable of completely activating the CPP. DOI-induced CPNA is dose dependent and leads to a gradual increase in burst amplitude and frequency, both of which are attenuated by ketanserin and not RS-102221. Interestingly, at higher doses, DOI evoked prolonged inspiratory burst duration. Because the medullary raphe nuclei have postinspiratory and expiratory-related bulbospinal projections (7) to the PMNs (18), it is possible that 5-HT2 receptors associated with these bulbospinal pathways were activated. Alternatively, it is possible that 5-HT2 receptors produced a long-lasting depolarization of the PMN membrane potential, because 5-HT has been shown to induce this excitatory effect at the spinal level (23). This is very likely because our laboratory recently demonstrated that PMNs colocalize 5-HT2A- and not 5-HT2C-receptor mRNA expression (1), suggesting that PMNs are capable of 5-HT2A-receptor biosynthesis, which may be focally linked with the postsynaptic regulation of PMN activity. Therefore, it is possible that higher doses of DOI may lead to overstimulation of 5-HT2 receptors and PMN excitation, leading to prolonged burst activity. The notion that 5-HT2 receptors may be regulating CPNA at the level of the spinal cord is supported by studies showing that microapplication of 5-HT into the phrenic nucleus of the rabbit increased the peak integrated amplitude of phrenic nerve activity, an effect prevented by the 5-HT-receptor antagonist methysergide (33). Moreover, excitation of PMNs, after electrical stimulation of the raphe obscuris (13-15), is blocked with methysergide (14) and more specifically with the 5-HT2-receptor antagonists ketanserin (23) or SR-46349B (12). Together, these observations, combined with the present results, point to the likely possibility that 5-HT2A receptors may directly regulate the excitation of PMNs at the level of the spinal cord. Stimulation of these receptors leads to an enhancement of phrenic burst amplitude.
In the present study, DOI also mediated increases in respiratory rate (bursts/min), under conditions of both maintained CO2 levels and asphyxia. An increase in respiratory frequency during asphyxia in rats is a normal response to hypercapnia (35, 36). DOI-mediated increases in respiratory rate under conditions of maintained CO2 and asphyxia were attenuated with ketanserin but not RS-102221. This finding also points to the possibility that 5-HT2A receptors may also regulate CPNA at medullary respiratory centers. The likelihood is supported by previous work in which application of 5-HT (27) or 5-HTP (28) to an in vitro bath containing rat brain stem-spinal cord explant preparations from newborn rats increased respiratory-related discharge frequencies recorded from the cervical ventral roots. The increases were prevented either by reduction in 5-HT with the 5-HT synthesis inhibitor p-chlorophenylalanine, or by general 5-HT-receptor antagonism with methysergide (28). Moreover, 5-HT microinjection into the rostral ventrolateral medulla also increased the frequency of respiratory-related activity in the cervical ventral roots (5). Electrical stimulation of the nucleus raphé obscuris increased the frequency of respiratory activity recorded from the ipsilateral phrenic nerve, implicating the influence of 5-HT afferents on medullary respiratory centers that regulate respiratory rhythm (13, 25). The above observations, combined with the present results, suggest that 5-HT2 receptors, most likely 2A, may also mediate effects at the level of the medulla after C2 spinal cord hemisection.
It should be noted that ketanserin produces hypotension after intravenous administration. To rule out the possibility that hypotension caused the attenuation of DOI-induced CPNA, epinephrine was coadministered with ketanserin in the present study. The administration of epinephrine normalized blood pressure to levels seen before ketanserin but did not alter the depression of the DOI-induced CPNA compared with ketanserin alone. It is generally accepted that ketanserin is more selective for 5-HT2A receptors but is a less effective blocker of 5-HT2C receptors (21, 22, 32, 38). Thus 5-HT2A- rather than 5-HT2C-receptor activation is most likely involved in the induction of the CPNA after C2 spinal cord hemisection. It is noteworthy to mention that, although ketanserin is a poor 5-HT2C-receptor antagonist, it also shows only moderate binding affinity for histamine H1 and 1-adrenergic-receptor binding sites, binds very weakly to dopamine receptors, and is inactive in other known neurotransmitter-receptor binding assays (21). Therefore, given the binding characteristics of ketanserin, the results of the present study strongly implicate 5-HT2A-receptor-mediated mechanisms yet do not fully rule out the involvement of 5-HT2C receptors. However, with the growing availability of more specific 5-HT2A-receptor antagonists (MDL-100907), future studies may be able to discriminate more selectively 5-HT2A-receptor-mediated mechanisms in the spinal cord.
In the present study, RS-102221, a 5-HT2C-receptor antagonist, did not significantly attenuate either burst amplitude or rate after DOI-induced CPNA. This result, combined with a lack of colocalization of 5-HT2C-receptor mRNA expression with PMNs (1), suggests that 5-HT2C receptors may not be functionally linked to PMNs at the level of the spinal cord. Interestingly, specific stimulation of 5-HT2C receptors with MK-212 temporarily depressed phrenic nerve burst amplitudes in the intact right phrenic nerve. Similar observations have been reported in which 5-HT2C-receptor stimulation depressed respiratory rate in cats (20) and in vitro brain stem preparations (30). In addition, it has recently been shown that MK-212 inhibited neurons of the nucleus tractus solitarius, an effect blocked by RS-102221 (34). Together, these observations suggest that the influence of 5-HT2C receptors on respiratory activity may be isolated to brain stem respiratory centers with little or no direct regulation over PMNs.
Functional importance of serotonin2-receptor-mediated respiratory recovery. The present study provides strong evidence that 5-HT2, most likely 2A and not 2C, receptors may be functionally linked to the activation of respiratory pathways that become disrupted after cervical spinal cord injury. These findings are in direct agreement with previous work that implicated 5-HT in the activation of latent CPPs (24). Electrical stimulation of the lateral funiculus, contralateral to cervical hemisection, prompted short-latency responses in the ipsilateral phrenic nerve (24). Administration of 5-HTP induced tonic activity in the phrenic nerve, as well as two long latency responses. Similar findings were recently reported from our laboratory in which 5-HTP induced CPNA ipsilateral to cervical hemisection. This effect was prevented with methysergide (40, 41). Collectively, these data suggest that 5-HT may convert ineffective CPPs in the spinal cord to effective pathways capable of inducing CPNA ipsilateral to cervical hemisection. The present data, showing increases in burst rate and amplitude, implicate excitatory 5-HT2-receptor mechanisms, likely mediated at both medullary and spinal cord levels. Whereas the pharmacological outcomes of the present study strongly point to a 5-HT2A-receptor-mediated mechanism, additional studies will be required with more specific 5-HT2A-receptor antagonists to prove that the 5-HT2A receptor alone is responsible for these effects.
ACKNOWLEDGEMENTS
This work was supported by National Institute of Child Health and Human Development Grant HD-31550 (to H. G. Goshgarian).
FOOTNOTES
Address for reprint requests and other correspondence: H. G. Goshgarian, Dept. of Anatomy and Cell Biology, Wayne State Univ.-School of Medicine, 540 East Canfield Ave., Detroit, MI 48201 (E-mail: hgoshgar@med.wayne.edu).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 24 May 2001; accepted in final form 8 August 2001.
REFERENCES
1.Basura, GJ, Zhou SY, Walker PD, and Goshgarian HG. Distribution of serotonin 2A and 2C receptor mRNA expression in the cervical ventral horn and phrenic motoneurons following acute spinal cord hemisection. Exp Neurol 169: 255-263, 2001.
2.Bonhaus, DW, Weinhardt KK, Taylor M, Desouza A, McNeely PM, Szczepanski K, Fontana DJ, Trinh J, Rocha CL, Dawson MW, Flippin LA, and Eglen RM. RS-102221: a novel high affinity and selective 5-HT2C receptor antagonist. Neuropharmacology 36: 621-629, 1997.
3.Bowker, RM, Westlund KN, and Coulter JD. Origins of serotonergic projections to the spinal cord in rat: an immunocytochemical-retrograde transport study. Brain Res 226: 187-199, 1981.
4.DeVries, KL, and Goshgarian HG. Spinal cord localization and characterization of the neurons which give rise to the accessory phrenic nerve in the adult rat. Exp Neurol 104: 88-90, 1989.
5.Di Pasquale, E, Morin D, Monteau R, and Hilaire G. Serotonergic modulation of the respiratory rhythm generator at birth: an in vitro study in the rat. Neurosci Lett 143: 91-95, 1992.
6.Eldridge, FL. Quantification of electrical activity in the phrenic nerve in the study of ventilatory control. Chest 70: 154-157, 1976.
7.Gilbey, PM, Futuro-Neto HA, and Zhou SY. Respiratory-related discharge patterns of caudal raphe neurones projecting to the upper thoracic spinal cord in the rat. J Auton Nerv Syst 50: 263-273, 1995.
8.Goshgarian, HG, Ellenberger HH, and Feldman JL. Decussation of bulbospinal respiratory axons at the level of the phrenic nuclei in adult rats: a possible substrate for the crossed phrenic phenomenon. Exp Neurol 111: 135-139, 1991.
9.Goshgarian, HG, and Guth L. Demonstration of functionally ineffective synapses in the guinea pig spinal cord. Exp Neurol 57: 613-621, 1977.
10.Goshgarian, HG, and Rafols JA. The phrenic nucleus of the albino rat: a correlative HRP and Golgi study. J Comp Neurol 201: 441-456, 1981.
11.Hadley, SD, Walker PD, and Goshgarian HG. Effects of the serotonin synthesis inhibitor p-CPA on the expression of the crossed phrenic phenomenon 4 h following C2 spinal cord hemisection. Exp Neurol 160: 446-453, 1998.
12.Hilaire, G, Bou C, and Monteau R. Serotonergic modulation of central respiratory activity in the neonatal mouse: an in vitro study. Eur J Pharmacol 329: 115-120, 1997.
13.Holtman, JR, Anastasi NC, Norman WP, and Dretchen KL. Effect of electrical and chemical stimulation of the raphe obscurus on phrenic nerve activity in the cat. Brain Res 362: 214-220, 1986.
14.Holtman, JR, Dick TE, and Berger AJ. Involvement of serotonin in the excitation of phrenic motoneurons evoked by stimulation of the raphe obscurus. J Neurosci 6: 1185-1193, 1986.
15.Holtman, JR, Dick TE, and Berger AJ. Serotonin-mediated excitation of recurrent laryngeal and phrenic motoneurons evoked by stimulation of the raphe obscurus. Brain Res 417: 12-20, 1987.
16.Holtman, JR, Norman WP, and Gillis RA. Projections from the raphe nuclei to the phrenic motor nucleus in the cat. Neurosci Lett 44: 105-111, 1984.
17.Holtman, JR, Norman WP, Skirboll L, Dretchen KL, Cuello C, Visser TJ, H?kfelt T, and Gillis RA. Evidence for 5-hydroxytryptamine, substance P, and thyrotropin-releasing hormone in neurons innervating the phrenic motor nucleus. J Neurosci 4: 1064-1071, 1984.
18.Hosogai, M, Matsuo S, Sibahara T, and Kawai Y. Projection of respiratory neurons in rat medullary raphe nuclei to the phrenic nucleus. Respir Physiol 112: 37-50, 1998.
19.Kinkead, R, and Mitchell GS. Time-dependent hypoxic ventilatory responses in rats: effects of ketanserin and 5-carboxamidotryptamine. Am J Physiol Regulatory Integrative Comp Physiol 277: R658-R666, 1999.
20.Lalley, PM. The excitability and rhythm of medullary respiratory neurons in the cat are altered by the serotonin receptor agonist 5-methoxy-N,N,dimethyltryptamine. Brain Res 648: 87-98, 1994.
21.Leysen, JE, Awouters F, Kennis L, Laduron PM, Vandenberk J, and Janssen PAJ Receptor binding profile of R 41 468, a novel antagonist at 5-HT2-receptors. Life Sci 28: 1015-1022, 1981.
22.Leysen, JE, Niemegeer CJE, Van Neuten JM, and Laduron PM. [3H]ketanserin (R 41 468), as selective 3H-ligand for serotonin2 receptor binding sites: binding properties, brain distribution and functional role. Mol Pharmacol 21: 304-314, 1982.
23.Lindsay, AD, and Feldman JL. Modulation of respiratory activity of neonatal rat phrenic motoneurons by serotonin. J Physiol 461: 213-233, 1993.
24.Ling, L, Bach KB, and Mitchell GS. Serotonin reveals ineffective spinal pathways to contralateral phrenic motoneurons in spinally hemisected rats. Exp Brain Res 101: 35-43, 1994.
25.Millhorn, DE. Stimulation of raphe (obscurus) nucleus causes long-term potentiation of phrenic nerve activity in cat. J Physiol 381: 169-179, 1986.
26.Moreno, DE, Yu X, and Goshgarian HG. Identification of the axon pathways which mediate functional recovery of a paralyzed hemidiaphragm following spinal cord hemisection in the adult rat. Exp Neurol 116: 219-228, 1992.
27.Morin, D, Hennequin S, Monteau R, and Hilaire G. Serotonergic influences of central respiratory activity: an in vitro study in the newborn rat. Brain Res 535: 281-287, 1990.
28.Morin, D, Monteau R, and Hilaire G. 5-Hydroxytryptamine modulates central respiratory activity in the newborn rat: an in vitro study. Eur J Pharmacol 192: 89-95, 1991.
29.O'Hara, TE, and Goshgarian HG. Quantitative assessment of phrenic nerve functional recovery mediated by the crossed phrenic reflex at various time intervals after spinal cord injury. Exp Neurol 111: 244-250, 1990.
30.Onimaru, H, Shamoto A, and Homma I. Modulation of respiratory rhythm by 5-HT in the brainstem spinal cord preparation from newborn rat. Pflügers Arch 435: 485-494, 1998.
31.Pilowsky, PM, Castro D, Llewellyn-Smith I, Lipski J, and Voss MD. Serotonin immunoreactive boutons make synapses with feline phrenic motoneurons. J Neurosci 10: 1091-1098, 1990.
32.Sanders-Bush, E, and Breeding M. Putative selective 5-HT2 antagonists block serotonin 5-HT1C receptors in the choroid plexus. J Pharmacol Exp Ther 247: 169-173, 1988.
33.Schmid, K, Bohmer G, and Merkelbach S. Serotonergic control of phrenic motoneuronal activity at the level of the spinal cord of the rabbit. Neurosci Lett 116: 204-209, 1990.
34.Sevoz-Couche, C, Spyer KM, and Jordan D. Inhibition of rat nucleus tractus solitarius neurones by activation of 5-HT2C receptors. Neuroreport 11: 1785-1790, 2000.
35.St. John, WM, and Bianchi AL. Responses of bulbospinal and laryngeal respiratory neurons to hypercapnia and hypoxia. J Appl Physiol 59: 1201-1207, 1985.
36.St. John, WM, and Wang SC. Response of medullary respiratory neurons to hypercapnia and isocapnic hypoxia. J Appl Physiol 43: 812-821, 1977.
37.Tai, Q, Palazzolo KL, and Goshgarian HG. Synaptic plasticity of 5-hydroxytryptamine-immunoreactive terminals in the phrenic nucleus following spinal cord injury. A quantitative electron microscopic analysis. J Comp Neurol 386: 613-624, 1997.
38.Wolf, WA, and Schutz LJ. The serotonin 5-HT2C receptor is a prominent serotonin receptor in basal ganglia: evidence from functional studies on serotonin-mediated phosphoinositide hydrolysis. J Neurochem 69: 1449-1458, 1997.
39.Zhan, WZ, Ellenberger HH, and Feldman JL. Monoaminergic and GABAergic terminations in phrenic nucleus of rat identified by immunohistochemical labeling. Neuroscience 31: 105-113, 1989.
40.Zhou, SY, and Goshgarian HG. Effects of serotonin on crossed phrenic nerve activity in cervical spinal cord hemisected rats. Exp Neurol 160: 446-453, 1999.
41.Zhou, SY, and Goshgarian HG. 5-Hydroxytryptophan-induced respiratory recovery after cervical spinal cord hemisection in rats. J Appl Physiol 89: 1528-1536, 2000.(Shi-Yi Zhou, Gregory J. B)