Neurons in the rostral medullary raphé/parapyramidal region regulate cutaneous sympathetic nerve discharge. Using focal electrical stimulation at different dorsoventral raphé/parapyramidal sites in anesthetized rabbits, we have now demonstrated that increases in ear pinna cutaneous sympathetic nerve discharge can be elicited only from sites within 1 mm of the ventral surface of the medulla. By comparing the latency to sympathetic discharge following stimulation at the ventral raphé site with the corresponding latency following stimulation of the spinal cord [third thoracic (T3) dorsolateral funiculus] we determined that the axonal conduction velocity of raphé-spinal neurons exciting ear pinna sympathetic vasomotor nerves is 0.8 ± 0.1 m/s (n = 6, range 0.6–1.1 m/s). Applications of the 5-hydroxytryptamine (HT)2A antagonist trans-4-((3Z)3-[(2-dimethylaminoethyl)oxyimino]-3-(2-fluorophenyl)propen-1-yl)-phenol, hemifumarate (SR-46349B, 80 μg/kg in 0.8 ml) to the cerebrospinal fluid above thoracic spinal cord (T1-T7), but not the lumbar spinal cord (L2-L4), reduced raphé-evoked increases in ear pinna sympathetic vasomotor discharge from 43 ± 9 to 16 ± 6% (P < 0.01, n = 8). Subsequent application of the excitatory amino acid (EAA) antagonist kynurenic acid (25 μmol in 0.5 ml) substantially reduced the remaining evoked discharge (22 ± 8 to 6 ± 6%, P < 0.05, n = 5). Our conduction velocity data demonstrate that only slowly conducting raphé-spinal axons, in the unmyelinated range, contribute to sympathetic cutaneous vasomotor discharge evoked by electrical stimulation of the medullary raphé/parapyramidal region. Our pharmacological data provide evidence that raphé-spinal neurons using 5-HT as a neurotransmitter contribute to excitation of sympathetic preganglionic neurons regulating cutaneous vasomotor discharge. Raphé-spinal neurons using an EAA, perhaps glutamate, make a substantial contribution to the ear sympathetic nerve discharge evoked by raphé stimulation.
- 5-hydroxytryptamine2A receptors
- kynurenic acid
- sympathetic preganglionic neuron
- cutaneous blood flow
neurons in the rostral medullary raphé/parapyramidal region regulate cutaneous vasomotor tone by controlling cutaneous sympathetic preganglionic neuronal activity (7, 39, 42, 45). Identification and characterization of the relevant neurons is an important task. Activity of these neurons contributes to the heat exchange aspect of thermoregulation as well as to the integrated bodily response to nociceptive or salient environmental events (7, 21, 35, 36, 39, 42, 45, 49). Neuroanatomic studies using transneuronal transport of live virus injected into the tail vasculature indicate that a proportion of sympathetic premotor raphé-spinal neurons synthesize serotonin (5-hydroxytryptamine, 5-HT) (32, 44).
Although 5-HT is a neurotransmitter traditionally considered important for central nervous system (CNS) control of thermoregulation (24), attention has usually focused on effects of 5-HT in the upper brain stem and basal forebrain, with body temperature considered as an integral variable. Much less attention has been paid to the possibility that 5-HT might contribute to temperature regulation via control of heat dissipation through the cutaneous circulation. Even when temperature studies have focused on 5-HT neurons in the medullary raphé region, interpretation of the results has emphasized possible ascending projections of the cells (4, 13), rather than descending projections to cutaneous sympathetic preganglionic neurons in the spinal cord. Results of some recent studies investigating possible thermoregulatory roles of raphé-spinal neurons have been interpreted as providing evidence both for (30) and against (32, 33) involvement of 5-HT cells.
Axonal conduction velocities of raphé-spinal neurons controlling brown adipose tissue (BAT) activity are <1 m/s, in the unmyelinated range (31) and serotonergic axons in the dorsolateral funiculus of the spinal cord are unmyelinated (2). A study of cold-activated raphé-spinal neurons identified by antidromic activation from the upper lumbar spinal cord of the rat (cutaneous vasomotor sympathetic outflow) demonstrated a higher conduction velocity (∼6 m/s), although the authors suggest a cautious interpretation in view of the technical difficulty of antidromically activating unmyelinated fibers (42).
Studies of sympathetic control of cutaneous blood flow from our laboratory, using systemically administered drugs that interact with 5-HT1A or 5-HT2A receptors, strongly support a role for raphé-spinal 5-HT neurons in regulation of sympathetic cutaneous vasomotor discharge (see discussion). In the present study we investigated this hypothesis using electrophysiological and pharmacological techniques. We recorded multiunit discharge in postganglionic cutaneous sympathetic vasomotor axons in the ear pinna of the anesthetized rabbit. Using focal electrical stimulation in the rostral medullary raphé/parapyramidal region, we mapped the dorsoventral location of sites from which it was possible to increase peripheral sympathetic cutaneous vasomotor discharge. We then determined the conduction velocity in the descending raphé-spinal pathway by comparing the latency to peripheral sympathetic activation following electrical stimulation of the rostral medullary raphé with the latency following stimulation in the dorsolateral funiculus at the third thoracic (T3) spinal level. We also determined the latency to peripheral sympathetic activation following electrical stimulation in the spinal tract of the trigeminal nerve. Finally, we used direct spinal application of a 5-HT2A receptor antagonist, trans-4-((3Z)3-[(2-dimethylaminoethyl)oxyimino]-3-(2-fluorophenyl)propen-1-yl)-phenol, hemifumarate (43), or a broad-spectrum excitatory amino acid (EAA) receptor antagonist (kynurenic acid) and determined whether these agents reduced activation of cutaneous sympathetic vasomotor discharge elicited by electrical stimulation of the rostral medullary raphé.
MATERIALS AND METHODS
Male New Zealand White rabbits (n = 25) weighing 2.6–3.6 kg were used. All animal procedures were approved by the Flinders University Animal Welfare Committee.
Surgical preparation for experiments with anesthetized rabbits.
Rabbits were anesthetized with urethane (ethyl carbamate, Sigma, 1.5 g/kg iv infused over 20 min). The level of anesthesia was maintained at a depth sufficient to abolish corneal and withdrawal reflexes. After the trunk and limbs were shaved, an endotracheal tube was inserted via a tracheostomy. The left femoral artery and vein were cannulated for measurement of systemic arterial pressure and intravenous drug administration, respectively. The rabbits were mounted prone in a stereotaxic frame (Kopf Instruments), with an electrical heating pad to maintain rectal (body) temperature. The body temperature was continuously measured by a thermocouple probe and maintained between 38° and 40°C by heating pad and lamp. The medulla oblongata was exposed by incision of the atlanto-occipital membrane. The head angle was fixed so that the dorsal surface of medulla oblongata was horizontal. Rabbits were paralyzed with vecuronium bromide (1–1.5 mg/kg iv) and mechanically ventilated with 100% oxygen, using a ventilator (Harvard model 681, Harvard Apparatus). Deep anesthesia level was established before the paralyzing, and the animal was allowed to recover between doses so that adequate anesthesia could be confirmed before paralysis was reestablished. End-tidal CO2 concentration was monitored continuously and was maintained at 30–40 mmHg. The dorsal surface of the thoracic spinal cord (T1-T7) or the lumber spinal cord (L2-L4) was exposed by laminectomy.
Electrical stimulation of the medullary raphé, parapyramidal region, spinal tract of trigeminal nerve, dorsolateral funiculus of the spinal cord and cervical sympathetic trunk.
The medullary raphé and dorsolateral funiculus of the spinal cord at the third thoracic vertebral segment were stimulated with a glass-coated tungsten electrode insulated to within 50 μm of the tip. To approach the medullary raphé and the parapyramidal region, the electrode was inclined 10 degrees to the vertical (tip rostral). The insertion point was 0.5–1.0 mm rostral from the rostral-midline edge of the area postrema in the midline. Electrode insertions were made in the midline (raphé stimulation) and either 0.75 or 1.5 mm lateral to the midline (parapyramidal stimulation).
To stimulate the spinal cord, electrodes were inserted vertically near the entry of the dorsal rootlets at the third segment of the thoracic spinal cord and fixed at the point where electrical stimulation evoked maximal excitation of the ear pinna sympathetic nerve (1.0–1.5 mm ventral to the dorsal surface of the spinal cord). Electrical stimulation of the spinal trigeminal tract and cervical sympathetic trunk was performed as described in our previous papers (7, 40). Single or triple cathodal pulses (0.5 ms, 100 Hz) were delivered with a Grass S88 stimulator and a Grass PSIU6 constant current unit (Grass Telefactor, West Warwick, RI). Single-pulse stimulation was used in the mapping study to obtain fine spatial resolution. Triple-pulse stimulation was used in other studies, because a larger response was required to obtain good signal-to-noise ratio so that amplitude of nerve response was easily compared between stimulus intensities.
Measurement of ear pinna sympathetic vasomotor nerve discharge.
The proximal end of a cut sympathetic nerve supplying an arterial vessel in ear pinna was placed across a pair of silver-wire electrodes and covered with a mixture of paraffin oil and vaseline (9, 37, 39, 41). Multifiber nerve activity was recorded differentially using a Neurolog NL 100 preamplifier (gain, ×1), NL104 amplifier (gain, ×20,000), and NL125 filter (band pass between 100 Hz and 1 kHz) (Digitmer, Hertfordshire, UK). At the end of the experiments, hexamethonium bromide (50 mg/kg iv) was administered to confirm loss of nerve activity, thereby confirming the sympathetic nature of the recording. The intact ipsilateral cervical sympathetic nerve trunk was exposed and placed across a pair of silver-wire electrodes.
SR-46349B was kindly supplied by Sanofi-Recherche (Montpellier, France). Kynurenic acid, hexamethonium bromide, and DMSO were obtained from Sigma Chemical. Kynurenic acid was dissolved in alkalinized ringer with 80 mM NaOH. SR-46349B was dissolved in DMSO and then diluted with Ringer so that the final concentration of DMSO was 10%.
Application of drugs to the cerebrospinal fluid above spinal cord segments.
The spinal cord was exposed via laminectomy from T1-T7 because preganglionic neurons projecting to the superior cervical ganglion are distributed in T1-T8 and concentrated in T3-T5 (19). The rostral and caudal ends of the exposed spinal cord were sealed with agar to prevent spread of drug solution beyond the confines of the laminectomy. Solutions of drugs, warmed to body temperature before application, were added to the cerebrospinal fluid (CSF) bathing the dorsal surface of the exposed spinal cord. The volume of SR-46349B and kynurenic acid solution was 0.8 and 0.5 ml, respectively.
In different rabbits, to exclude an effect of SR-46349B via spread from the CSF to the systemic circulation, and to establish regional specificity in the spinal cord for the effect of SR-46349B, the cord was exposed via laminectomy from L2-L4 and the effect of raphé stimulation was determined before and after this lumbar region was bathed in SR-46349B (80 μg/kg in 0.8 ml of 10% DMSO in Ringer).
The onset latency of the evoked discharge was determined by cumulative sum analysis (18). A program custom-written in IgorPro was used to rectify the peristimulus averages of nerve discharge (Fig. 1B) and calculate the baseline value of the rectified nerve discharge in the 200-ms control period before stimulation and then to calculate the cumulative sum of poststimulus nerve discharge data relative to this baseline value. Nerve discharge elicited by stimulation was considered to have increased significantly when the cumulative sum exceeded the critical limit at the P < 0.01 level of confidence (Fig. 1C). Onset and termination of the evoked response were taken as the beginning and the end of the rising phase of the cumulatively summed data (Fig. 1C). Amplitude of the evoked response was determined as an integral value of the rectified data between the onset and the termination of the response.
Linear regression analysis was used to show the relationship between the amplitude of raphé-elicited nerve discharge and the logarithm of stimulation intensity (expressed in multiples of the minimum stimulus intensity required to evoke nerve discharge in a particular animal). The slope of the regression line was compared before and after drug application. Fisher’s protected t-test was used to determine significant differences with the significant threshold of the initial ANOVA set at the 0.05 level.
Stimulation sites were marked by passing negative DC lesioning current (150 μA for 30–60 s). On completion of the experiments, under deep anesthesia, the brain was fixed by transcardial perfusion of formaldehyde solution. The brain was removed, the medulla was sectioned (60 μm) on a freezing microtome, and the lesion site was examined after Nissl stain.
Rostral midline medullary sites from which sympathetic discharge was evoked.
In nine rabbits, the midline medulla was electrically stimulated (single pulse, 0.5-ms duration, 100–200 μA, 0.1 Hz). In a screening study, stimulation was carried out with the electrode tip positioned 4.0 mm below the dorsal surface of the medulla. The electrode was first inserted 1.0 mm rostral to the rostral-midline edge of the area postrema (obex), and then 0.5 or 1.0 mm rostral and 0.5 or 1.0 mm caudal to the obex. The rostrocaudal level at which the ventral site yielded the largest increase in ear pinna sympathetic discharge was then chosen for more detailed analysis. At this level, stimulation was carried out from 1.5–2.0 mm below the dorsal surface of the fourth ventricle to the ventral surface, in 1.0-mm steps.
Figure 2A shows an example from one rabbit with midline raphé stimulation. The electrical stimulation did not evoke ear pinna sympathetic nerve until the tip was ∼4.0 mm from the dorsal surface of the medulla. The tip was advanced a few more steps (in 0.5-mm and 0.7-mm steps) to decide the stimulation site evoking maximal response in ear sympathetic nerve. The effective sites were concentrated between the ventral surface of the medulla and 1.5 mm dorsal to this (Fig. 2C). The site of maximum sensitivity was found to be within 0.5 mm of the ventral surface of the medulla, as confirmed by histological analysis following an electrical lesion made at this site at the end of experiments (Fig. 3A). A similar pattern of ventrally located parapyramidal effective stimulation sites was observed when the electrode was inserted 0.75 or 1.5 mm lateral to the midline (Fig. 2C and 3A).
Latencies to onset of ear pinna sympathetic nerve discharge after stimulation of raphé, parapyramidal region or trigeminal tract, and conduction velocity of raphé-spinal axons activating ear pinna sympathetic nerve discharge.
Stimulation of the midline medullary raphé (3 pulses, 10-ms interval, 0.5-ms duration, 100–200 μA) evoked discharge in the ear pinna sympathetic nerve with an onset latency of 259 ± 6 ms [n = 32 rabbits, including 21 animals used in the experiments reported in our previous papers (37, 39, 40) (Fig. 4A)]. Stimulation of the parapyramidal region 0.75 mm and 1.5 mm lateral to the midline (3 pulses, 10-ms interval, 0.5-ms duration, 100–200 μA) evoked discharge in the ear pinna sympathetic nerve with an onset latency of 275 ± 16 and 271 ± 16 ms, respectively (n = 6 rabbits) (Fig. 4, B and C). There was no difference between onset latencies from the three different ventral medullary stimulation sites (P > 0.05). Stimulation of the trigeminal tract (3 pulses, 10-ms interval, 0.5-ms duration, 100 μA-1.5 mA) evoked discharge in the ear pinna sympathetic nerve with an onset latency of 321 ± 27 ms [n = 27 rabbits, including 21 animals used in the experiment reported in our previous paper (37, 39, 40)] (Fig. 4D). The onset latency from the trigeminal tract stimulation was slightly longer than onset latencies from the raphé/parapyramidal stimulation sites (P < 0.05).
In six rabbits, the midline medullary raphé and the dorsolateral funiculus in the spinal cord at the third thoracic vertebral segment (DLF-T3) were electrically stimulated in the same animal, and the latencies to onset of ear pinna sympathetic discharge were compared (Fig. 3E). Stimulation of the medullary raphé (3 pulses, 10-ms interval, 0.5-ms duration, 100–200 μA) evoked discharge in the ear pinna sympathetic nerve with an onset latency of 267 ± 14 ms and stimulation of the DLF-T3 (3 pulses, 10-ms interval, 0.5-ms duration, 100 μA-1 mA) evoked ear pinna sympathetic discharge with a latency of 136 ± 10 ms (Fig. 4, A and E). The difference between these two latencies was 131 ± 10 ms and the distance from the medullary raphé to the DLF-T3 stimulation site was 100 ± 6 mm. Therefore, the conduction velocity of the descending pathway (from the raphé to the DLF-T3) was 0.8 ± 0.1 m/s (n = 6, range 0.6–1.1 m/s).
At the end of the experiment, before the raphé stimulation site was lesioned, in four of six rabbits, the thoracic spinal cord was transected between T4 and T5. This procedure did not change the ear pinna sympathetic nerve discharge evoked by raphé stimulation (Fig. 5, A and B). Subsequent transverse section between C8 and T1 abolished the raphé-elicited ear pinna sympathetic nerve discharge (Fig. 5C), without affecting discharge evoked by electrical stimulation of DLF at T3 (Fig. 5D). The DLF-T3-elicited nerve discharge was eliminated by subsequent administration of hexamethonium bromide (50 mg/kg iv).
Effect of 5-HT2A antagonists applied to spinal cord on raphé-elicited cutaneous nerve discharge.
In eight rabbits, vehicle and then SR-46349B were administered into a pool of cerebrospinal fluid on the dorsal aspect of the spinal cord extending between the T1 and the T7 laminectomy sites. In these animals baseline ear pinna nerve activity was absent or very low before any treatment, so that it was not possible to determine the effect of the drug on resting nerve discharge. Topical application of vehicle (10% DMSO in Ringer) did not change the amplitude of ear pinna nerve discharge elicited by electrical stimulation of the medullary raphé. Topical application of SR-46349B (80 μg/kg in 0.8 ml) substantially reduced raphé-elicited ear pinna nerve discharge (Fig. 6B). In six of eight rabbits, topical applications of 0.8 μg/kg (lowest dose) and 8 μg/kg (middle dose) of SR-46349B were performed after vehicle but before the highest dose of SR-46349B (80 μg/kg). After either the lowest or the middle dose of SR-46349B, the raphé-elicited ear pinna sympathetic nerve discharge at each stimulus intensity was not significantly different from those after vehicle (P > 0.05, n = 6) (Fig. 6D).
In six rabbits, SR-46349B (80 μg/kg) was applied to a pool of CSF on the dorsal aspect of the spinal cord extending between the L2 and the L4 laminectomy sites. This did not affect the amplitude of raphé-elicited ear pinna nerve discharge (Fig. 7, A, B, and G). In these rabbits, subsequent intravenous administration of SR-46349B (100 μg/kg iv) did significantly reduce the increase in ear pinna nerve discharge elicited by raphé stimulation (Fig. 7, C and G) without affecting the increase elicited by stimulation of preganglionic sympathetic axons in the ipsilateral cervical sympathetic nerve trunk (Fig. 7, D, E, and H). Intravenous administration of hexamethonium (50 mg/kg) completely abolished the resting discharge in the ear pinna sympathetic nerve, as well as the increase in discharge normally elicited by stimulation of the cervical sympathetic trunk (Fig. 7F).
Effect of kynurenic acid applied to spinal cord on raphé-elicited cutaneous nerve discharge.
In five rabbits, kynurenic acid (25 μmol in 0.5 ml, 50 mM) was applied to the spinal cord after completion of the experiments involving application SR-46349B (80 μg/kg). This additional application of kynurenic acid substantially reduced or abolished nerve discharge elicited by raphé stimulation (Fig. 6, C and D).
The present study demonstrates the location and the axonal conduction velocity of the raphé-spinal pathway responsible for excitation of sympathetic preganglionic neurons regulating cutaneous vasomotor discharge. In addition, our study provides evidence that serotonin, by activation of spinal 5-HT2A receptors, is a potential neurotransmitter in this pathway.
Medullary region from which it is possible to activate sympathetic cutaneous vasomotor discharge.
Low-amplitude (usually <100 μA) electrical stimulation evoked clear excitatory sympathetic cutaneous vasomotor responses only when the electrode tip was positioned within 1 mm of the ventral surface of the medulla oblongata, either in the midline raphé or in the nearby parapyramidal region, with maximal responses at the rostral level of the caudal border of the facial nucleus. Since electrical stimulation activates fibers as well as cell bodies, the present data do not show specifically the location of neurons mediating the excitatory sympathetic cutaneous vasomotor responses. However, taken together with relevant neuroanatomic studies (see below), it is likely that the dominant effect of the stimulation in the medullary raphé/parapyramidal region is excitation of raphé/parapyramidal neurons regulating cutaneous vasomotor discharge. The raphé and parapyramidal region is at the center of the medullary region containing 5-HT neurons (6, 14, 17, 29). There are direct projections from these sites to the spinal cord, and projection targets include the intermediolateral column (6, 15). In rats, transneuronal tracing studies demonstrate that medullary raphé and parapyramidal neurons, including 5-HT cells in the region, are labeled prominently and early after injection of pseudorabies virus into the tail (32, 44), and our preliminary data using injection of pseudorabies virus into the ear pinna give similar results in rabbits (10).
Axonal conduction velocity of raphé-spinal neurons activating sympathetic cutaneous vasomotor discharge is very low, within the range for unmyelinated axons.
Mean conduction velocity of the pathway from the medullary raphé to the T3 level of the spinal cord was 0.8 m/s and the fastest conduction velocity was 1.1 m/s. In a total of 32 rabbits subjected to raphé, parapyramidal, or trigeminal tract stimulation, we never observed latencies suggesting a faster conducting descending pathway. In rats, similar orthodromic activation techniques have also yielded conduction velocities <1 m/s for raphé-spinal neurons controlling the sympathetic regulation of brown adipose tissues (BAT) (31). Since electrical stimulation of the raphé is likely to activate all classes of sympathetic premotor neurons, as well as axons near the tip of the electrode, whatever their function or their neurotransmitters, axonal conduction velocities of >1 m/s are most unlikely for any raphé-spinal axon with a strong excitatory input to the sympathetic preganglionic neurons regulating the ear pinna cutaneous vascular bed.
The exclusively low conduction velocities documented in our present study are at variance with our previous conclusion (34) that fast-conducting raphé-spinal axons contribute to excitation of cutaneous sympathetic vasomotor discharge. In that study, we antidromically activated medullary raphé-spinal neurons and examined the subclass of cells activated by electrical stimulation of the trigeminal tract (a procedure that vigorously constricts the ear pinna bed). The discharge of raphé-spinal neurons with axonal conduction velocities as high as 30 m/s was substantially increased by this procedure (34). The pathway from the trigeminal tract to the spinal cord is presumably exclusively via the raphé, since inactivation of this region with muscimol entirely abolishes trigeminally elicited increases in sympathetic cutaneous vasomotor discharge (7, 40). In the present study, we confirmed that electrical stimulation of the trigeminal tract does vigorously activate cutaneous sympathetic vasomotor discharge. However, our present orthodromic activation experiments indicate that fast-conducting raphé-spinal neurons do not excite the sympathetic preganglionic neurons regulating the ear pinna cutaneous vascular bed. Perhaps these raphé-spinal neurons project to the spinal dorsal horn and play a role in pain modulation there (25).
The antidromic activation procedure was also used in rats to determine the axonal conduction velocity of cold-activated raphé-spinal neurons projecting to the lumbar spinal cord (42). With respect to neurons activated by cooling the skin of the trunk, no slowly conducting (<1 m/s) raphé-spinal neurons were detected possibly because, as the authors note, the stimulation parameters used were not optimal for activating unmyelinated axons. Another possibility is that slowly conducting raphé-spinal neurons are too small and/or too close to the ventral surface of medulla to be identified in single-cell recording studies. This issue requires further investigation.
At present, there are no “gold standard” studies establishing the conduction velocity of descending serotonergic raphé-spinal axons. Raphé-spinal neurons whose axonal conduction velocity has been established have been identified functionally rather than by appropriate immunohistochemical techniques combined with intracellular or juxtacellular labeling. In the present study, conduction velocities are in the range of 0.6 to 1.1 m/s, indicating that premotor raphé-spinal neurons regulating sympathetic outflow to the ear pinna vascular bed have unmyelinated or thinly myelinated fibers. Axonal conduction velocities in the range for unmyelinated axons would be consistent with participation of serotonergic raphé-spinal neurons in the regulation of cutaneous vasomotor tone (26, 46, 47).
Medullary raphé-spinal cells regulating cutaneous sympathetic vasomotor discharge are likely to include neurons that synthesize 5-HT.
We previously demonstrated that systemically administered SR-46349B, a 5-HT2A receptor antagonist that prevents the cutaneous vasoconstricting effect of the 5-HT2A agonist (+/−)−/−(2,5-dimethoxy-4-iodophenyl)-2-aminopropane) (DOI), reduces both resting sympathetic cutaneous vasomotor discharge and raphé-induced vasoconstriction of the cutaneous vascular bed (40). The present study demonstrates similar findings after local application of the same 5-HT2A receptor blocking agent to the thoracic spinal cord where ear pinna sympathetic vasomotor preganglionic neurons are located, but not after application of the same drug to the lumbar spinal cord.
In the present study, we also established that intravenous administration of SR-46349B does not reduce the increase in ear pinna sympathetic vasomotor discharge elicited by electrical stimulation of preganglionic sympathetic axons in the cervical sympathetic trunk. Taken together with our demonstration in a previous study (8) that the vigorous ear pinna-vasoconstricting effect of DOI is entirely prevented by section of the cervical sympathetic trunk, the present findings provide strong evidence that raphé-spinal excitation of cutaneous vasomotor sympathetic preganglionic neurons is partially dependent on excitation of 5-HT2A receptors in the region of the spinal cord containing the relevant sympathetic preganglionic neurons.
5-HT2A receptors are excitatory postsynaptic receptors normally innervated by axons of 5-HT neurons (1). The 5-HT2A receptors distributed in the intermediolateral nucleus in the spinal cord (12) are therefore also likely to be postsynaptic excitatory receptors normally innervated by serotonergic neurons in the medullary raphé region. Although there have been no studies of the effects of local application of specific 5-HT2A receptor agonists in the spinal cord, it is well established that microiontophoresis of 5-HT excites sympathetic preganglionic neurons (11, 27).
Studies of sympathetic control of cutaneous blood flow from our laboratory using systemically administered drugs that interact with 5-HT1A receptors also support a role for central 5-HT neurons in regulation of sympathetic cutaneous vasomotor discharge. Administration of 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT), a 5-HT1A receptor agonist, markedly inhibits cold-activated cutaneous sympathetic vasomotor nerve discharge in rabbits, impairing the animal’s ability to maintain body temperature in the cold (37). Similarly, systemically administered 8-OH-DPAT reverses cutaneous vasoconstriction occurring as part of the lipopolysaccharide-induced febrile response (5). These effects may at least partially reflect direct inhibition of raphé-spinal 5-HT neurons via stimulation of 5-HT1A autoreceptors (38), because classes of 5-HT1A receptors include inhibitory autoreceptors located on the perikarya of serotonin-synthesizing neurons, as well as on nearby cells that do not synthesize 5-HT (16) (20). Intramedullary injections of 8-OH-DPAT also profoundly inhibit BAT activation by stimuli acting at the hypothalamic level, consistent with a role for serotonergic neurons in regulating BAT metabolism (30).
Major role for glutamatergic raphé-parapyramidal neurons in regulation of sympathetic cutaneous vasomotor discharges.
Blockade of 5-HT2A receptors only moderately reduced cutaneous vasomotor sympathetic discharge elicited by raphé stimulation. Subsequent additional blockade of EAA receptors entirely abolished raphé-elicited sympathoexcitation. This emphasizes the importance of an EAA, perhaps glutamate, as well as 5-HT as a neurotransmitter in raphé-spinal neurons regulating sympathetic outflow to the cutaneous vascular bed. Glutamate and 5-HT could be colocalized and/or glutamate could be the transmitter in non-5-HT neurons. The interaction between, and relative roles of, 5-HT and glutamate in the spinal cord remain to be established. 5-HT, via 5-HT2A receptors, could have a modulatory excitatory action on the relevant sympathetic preganglionic neurons, so that SR-46349B-mediated reduction in the net tonic excitation of the neuron effectively blocks glutamatergic raphé-spinal neurotransmission.
Glutamatergic premotor raphé-spinal neurons also appear to have a major role in regulating BAT activity since blockade of EAA (NMDA-subtype) receptors entirely abolishes raphé-elicited increases in BAT temperature in the anesthetized rat (32). Glutamate receptors are expressed on sympathetic preganglionic neurons (22, 23, 28), and microiontophoresis of glutamate excites sympathetic preganglionic neurons in the spinal cord (11, 27, 48). Pharmacological blockade of spinal NMDA receptors also attenuates lumbar sympathetic nerve discharge evoked by activation of bulbospinal neurons (3).
Nakamura et al. (32) demonstrated that many raphé and parapyramidal neurons transneuronally labeled after injection of pseudorabies virus into the rat tail express vesicular glutamate transporter 3 (VGLUT3), suggesting an important neurotransmitter role for glutamate in raphé-spinal control of the cutaneous circulation. The relatively small proportion of neurons demonstrated to be doubly-labeled for VGLUT3 and 5-HT (approximately 10–20%), led Nakamura and colleagues to minimize any possible role for raphé-spinal 5-HT neurons either in regulation of BAT activity or in regulation of cutaneous sympathetic vasomotor nerve discharge. However, Nakamura and colleagues did not carry out colocalization immunohistochemical studies directly examining the proportion of virus-containing neurons also positive for markers of 5-HT synthesis. When Smith and colleagues (44) used this colocalization procedure, 5-HT neurons constituted ∼25% of raphé/parapyramidal virus-containing neurons after injection of pseudorabies virus into the rat tail. Our preliminary studies suggest a similar conclusion applies to the proportion of virus-containing raphé/parapyramidal neurons after injection of pseudorabies virus into the rabbit ear pinna (10).
In conclusion, previous studies using intramedullary injections of pharmacological agents that excite or inhibit neuronal cell bodies have established that the raphé-parapyramidal region contains neurons whose activity increases sympathetically mediated cutaneous vasoconstriction. Our present study in rabbits shows that axons descending from the raphé/parapyramidal region to excite cutaneous sympathetic vasomotor preganglionic have slowly conducting (unmyelinated) axons. Blockade of 5-HT2A receptors in the spinal region containing the relevant sympathetic preganglionic neurons moderately inhibits increases in ear pinna sympathetic vasomotor discharge elicited by electrical stimulation of the medullary raphé. These 5-HT2A receptors are likely to be postsynaptic receptors normally innervated by axons of raphé-parapyramidal neurons that use 5-HT as a neurotransmitter. Subsequent blockade of EAA receptors in the same spinal region entirely abolishes raphé-elicited increases in ear pinna sympathetic vasomotor discharge, suggesting a major role for glutamatergic neurons in raphé-spinal control of sympathetic cutaneous vasomotor discharge. Further experiments are required before the role of 5-HT in the natural physiological regulation of cutaneous sympathetic vasomotor discharge by raphé-parapyramidal neurons is established.
Our research was supported by the National Health and Medical Research Council of Australia.
We thank M. Blair, C. Morgan, and R. Flook for technical assistance.
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.
- Copyright © 2005 the American Physiological Society