Role of 5-HT3 receptors in activation of abdominal sympathetic C fibre afferents during ischaemia in cats

Authors

  • Liang-Wu Fu,

    1. Division of Cardiovascular Medicine, Departments of Internal Medicine and Human Physiology, University of California School of Medicine, Davis, CA 95616, USA
    Search for more papers by this author
  • John C. Longhurst

    1. Division of Cardiovascular Medicine, Departments of Internal Medicine and Human Physiology, University of California School of Medicine, Davis, CA 95616, USA
    Search for more papers by this author

Corresponding author L.-W. Fu: Division of Cardiovascular Medicine, TB 172, University of California at Davis, Davis, CA 95616, USA. Email: lwfu@ucdavis.edu

Abstract

  • 1Activation of abdominal sympathetic afferents during ischaemia reflexly excites the cardiovascular system. We have shown previously that exogenous 5-hydroxytryptamine (5-HT, i.e. serotonin) stimulates abdominal sympathetic afferent nerve endings, and recently have documented increased concentrations of 5-HT in intestinal lymph and portal venous plasma during brief abdominal ischaemia. The present investigation evaluated the role of endogenously produced 5-HT in activation of ischaemically sensitive abdominal sympathetic afferents.
  • 2Nerve activity of single-unit C fibre afferents innervating duodenum, mesentery, pancreas, portal hepatis, bile duct, gall bladder and jejunum was recorded from the right thoracic sympathetic chain of anaesthetized cats. Ischaemically sensitive C fibre afferents were identified according to their response to 5-10 min of abdominal ischaemia.
  • 3Intra-arterial injection of 5-HT (20 μg kg−1) increased discharge activity of twelve afferents from 0.23 ± 0.05 to 0.96 ± 0.09 impulses s−1 after an onset latency of 5.7 ± 1.4 s. Also, 2-methylserotonin (100 μg kg−1, i.a.), a 5-HT3 receptor agonist, stimulated eleven of twelve afferents to significantly increase their discharge activity from 0.25 ± 0.05 to 0.90 ± 0.10 impulses s−1 after a latency of 3.3 ± 0.4 s. Furthermore, intravenous injection of tropisetron (200 μg kg−1), a 5-HT3 receptor antagonist, significantly attenuated the increase in activity of twelve other C fibre afferents during 10 min of abdominal ischaemia from 1.62 ± 0.18 to 0.94 ± 0.22 impulses s−1, and eliminated the response of eleven other afferents to 5-HT.
  • 4Both the 5-HT2 receptor agonist, α-methylserotonin (100 μg kg−1, i.a.), and the 5-HT1 receptor agonist, 5-carboxamidotryptamine (100 μg kg−1, i.a.), did not alter the impulse activity of these twelve afferents (0.29 ± 0.05 to 0.31 ± 0.06, and 0.26 ± 0.06 to 0.29 ± 0.06 impulses s−1, respectively).
  • 5Treatment with indomethacin (5 mg kg−1, i.v.) in eight different cats did not alter the response of nine C fibre afferents to exogenous 5-HT (0.91 ± 0.17 vs. 1.19 ± 0.25 impulses s−1, P > 0.05).
  • 6The results suggest that, during mesenteric ischaemia, endogenous 5-HT contributes to the activation of abdominal sympathetic afferents, mainly through direct stimulation of 5-HT3 receptors and that the action of 5-HT on these afferents appears to be independent of the cyclo-oxygenase pathway.

Activation of abdominal sympathetic visceral afferents reflexly excites the cardiovascular system (Longhurst, 1995; Pan, Zeisse & Longhurst, 1996). Abdominal ischaemia, resulting from mesenteric artery occlusion, activates abdominal sympathetic visceral afferents to cause strong reflex responses that can increase arterial blood pressure by more than 50 mmHg (Huang & Longhurst, 1994; Rendig, Chahal & Longhurst, 1997). Ischaemic metabolites produced in the abdominal visceral region are believed to be responsible for activation of ischaemically sensitive sympathetic visceral afferents. We have demonstrated previously that local tissue hypoxia does not directly stimulate these afferents during ischaemia (Fu, Pan, Pitsillides & Longhurst, 1996). However, hypoxia and ischaemia lead to metabolic changes with the production of a number of chemical mediators that potentially could stimulate these afferent endings. For instance, our laboratory demonstrated previously that bradykinin (Pan, Stahl, Rendig, Carretero & Longhurst, 1994), prostaglandins (Longhurst, Rotto, Kaufman & Stahl, 1991), lactic acid (Stahl & Longhurst, 1992), oxygen-derived free radicals (Stahl, Pan & Longhurst, 1993) and histamine (Fu, Pan & Longhurst, 1997b) contribute to activation of ischaemically sensitive abdominal visceral afferents; conversely, other metabolic products such as leukotriene B4 (LTB4) modulate the activity of these nerve endings (Pan, Stahl & Longhurst, 1995). We believe, however, that other ischaemic metabolites are responsible for activation or sensitization of abdominal visceral afferents during ischaemia, because none of the previously investigated ischaemic metabolites appears to be responsible solely for activation of abdominal afferents during ischaemia.

Recently, we have begun to consider the possibility that endogenously produced 5-hydroxytryptamine (5-HT) contributes to activation of these afferent endings during mesenteric ischaemia. In this regard, in mammals, the main source of 5-HT is the enterochromaffin cells in the gastrointestinal tract (Gaginella, 1995). Another source of 5-HT is platelets that contain most of the circulating 5-HT through active uptake of 5-HT from blood (Meyers, Holsmen & Seachord, 1982). It has been shown that prolonged mesenteric ischaemia (30 min) followed by reperfusion only slightly increases the concentration of 5-HT in portal venous plasma in dogs (Strauss, Rubin, Newman, Strauss & Stuckey, 1973). Using a 10 min period of abdominal ischaemia in cats, we have demonstrated an increased 5-HT concentration in intestinal lymph fluid (Fu, O'Neill & Longhurst, 1997a), which reflects concentrations in the interstitial compartment, where most nerve endings exist (Mei, 1985). Although our laboratory has shown that exogenous 5-HT stimulates abdominal sympathetic afferents (Lew & Longhurst, 1986), there is no direct evidence regarding the role of endogenously produced 5-HT during brief ischaemia with regard to activation of ischaemically sensitive abdominal sympathetic afferents.

The physiological effects of 5-HT in tissue are mediated by a large family of receptors, which have been divided into at least seven subtypes according to a recent classification by Hoyer et al. (1994). Four major receptor subtypes designated 5-HT1, 5-HT2, 5-HT3 and 5-HT4 receptors have been shown to to be present in the spinal cord and peripheral neurons, which are coupled to a ligand-gated ion channel, intracellular adenyl cyclase or phospholipase C and inositol triphosphates (Cornfield & Nelson, 1991; Hoyer et al. 1994). Recently selective agonists and antagonists to 5-HT receptors have become available. For instance, hyperalgesia is produced by direct action of the 5-HT1A receptor on the primary afferents (Taiwo & Levine, 1992). Also, activation of 5-HT2 receptors stimulates somatic C fibres (Peroutka, 1994). Furthermore, Grundy, Blackshaw & Hillsley (1994) observed that exogenous 5-HT stimulates vagal mucosal chemosensitive afferents through 5-HT3 receptors. The 5-HT3 receptors have also been assumed to mediate an excitatory action on nociceptor units (Richardson, Engel, Donatsch & Stadler, 1985). However, the role of any of the 5-HT receptor subtypes in activation of sympathetic afferents during ischaemia has not been examined.

Previous studies have demonstrated that 5-HT induces synthesis of PGE2 and PGI2 in mesangial cells (Knauss & Abboud, 1986), while activation of thromboxane A2 (TxA2) receptors initiates the release of 5-HT from activated platelets (Hamberg, Svensson & Samuelsson, 1975). Cyclo-oxygenase products can sensitize the afferent endings to respond to bradykinin and histamine (Pan et al. 1994; Fu et al. 1997b). Although the effect of prostaglandins on 5-HT activation of ischaemically sensitive afferents is uncertain, it is possible that prostaglandins may sensitize these afferents to the action of 5-HT, since it is known that cyclo-oxygenase products are produced during brief mesenteric ischaemia (Rendig, Pan & Longhurst, 1994).

Therefore, the general aim of this study was to determine whether endogenously produced 5-HT stimulates abdominal visceral afferents during ischaemia. We hypothesized that: (1) 5-HT, through a 5-HT3 receptor mechanism, contributes to activation of abdominal visceral afferents during ischaemia; and (2) the stimulating effect of 5-HT on ischaemically sensitive visceral afferent endings is dependent upon production of prostaglandins.

METHODS

Surgical preparation

Experiments were performed on fifty-four fasted adult cats of either sex (2.9 ± 0.8 kg, mean ±s.d.). Surgical and experimental protocols used in this study were approved by the Animal Use and Care Committee at the University of California at Davis. The studies conformed to American Physiological Society Guideline and Principles Involving Animals. Anaesthesia was induced with ketamine (20-30 mg kg−1, i.m., Phoenix Scientific, Inc., St Joseph, MO, USA) and maintained with α-chloralose (40-50 mg kg−1, i.v.). Additional injections of α-chloralose (5-10 mg kg−1, i.v.) were given as needed to maintain an adequate depth of anaesthesia that was assessed by observing the absence of the conjunctival reflex. The trachea was intubated and respiration maintained artificially (Harvard ventilator, model 661, Ealing, South Natick, MA, USA). The cat was ventilated with 100 % O2 through the respirator. A femoral vein was cannulated for administration of drugs and fluid. A femoral arterial catheter was positioned with its tip in the thoracic descending aorta for measurement of pressure and administration of drugs. Systemic arterial blood pressure was measured by a pressure transducer (Statham P 23 ID, Gould) connected to the femoral arterial catheter. We frequently assessed arterial blood gases with a blood gas analyser (Model ABL-3, Radiometer) and maintained them within physiological limits (PO2 > 100 mmHg, PCO2= 28-35 mmHg; pH 7.35-7.45) by adjusting the respirator rate or tidal volume, or by administering NaHCO3 (1 M i.v.). Body temperature was monitored by a rectal thermistor and maintained at 36-38°C with a circulating water heating pad and a heat lamp. At the end of the experiment, the animals were killed by administration of saturated potassium chloride (i.v.).

Afferent recording

The surgical preparation used for recording single-unit activity of abdominal sympathetic C fibre afferents has been described previously (Fu et al. 1996). In brief, a mid-line sternotomy was performed. The third-eleventh right ribs and the middle and caudal lobes of the right lung were removed. Both phrenic nerves were isolated and cut. The fascia overlying the right paravertebral sympathetic chain was removed. The chain was then draped over a Plexiglass platform and covered with warm mineral oil. Small nerve filaments were dissected gently from the sympathetic chain or rami between T6 and T10 with the use of an operating microscope (Zeiss, Germany) and the caudal ends were placed across a recording electrode. One pole of the recording electrode was earthed with cotton thread to the animal. The recording electrode was attached to a high impedance probe (Grass Instruments, model HIP511, Quince, MA, USA), and the signal was amplified (model P511 Preamplifier, Grass) and processed through an audioamplifier (AM8B, Audiomonitor, Grass), an oscilloscope (Tektronix, model 2201, Beavertown, OR, USA), and then recorded on a chart recorder (model TA 4000B, Gould, Cleveland, OH, USA). The neurogram also was fed into an IBM compatible 486DX-based computer through an analog-to-digital interface card (R.C. Electronics Inc., Santa Barbara, CA, USA) for subsequent off-line analysis. The discharge frequency of afferents was analysed using data acquisition and analysis software (EGAA, version 3.02, R.C. Electronics Inc., Santa Barbara, CA, USA).

An inflatable occlusion cuff was placed around the thoracic descending aorta just above the diaphragm. A ventral mid-line incision was used to expose abdominal visceral organs. We closed the abdominal incision with towel clamps and covered the viscera with warm saline-soaked gauze to prevent fluid and heat losses. Receptive fields of afferents were located precisely using a fine-tipped glass rod and a stimulating electrode to evoke the action potential of the afferent. We determined conduction time by measuring the interval from stimulation to the afferent's action potential on the recording electrode. Conduction distance was estimated with a thread placed from the receptive field along the supposed afferent pathway through the prevertebral ganglion along the course of the major splanchnic nerve to the sympathetic chain and the recording electrode. C fibre afferents were classified as those with a conduction velocity (CV) of < 2.5 m s−1. Afferents included in this study had a range of CVs from 0.26 to 1.14 m s−1. Each had a receptive field that could be located precisely. Afferents were considered to be ischaemically sensitive if their discharge activity during 10 min of abdominal ischaemia was increased at least twofold above baseline activity (Fu et al. 1996).

Experimental protocols

Effect of 5-HT, 5-carboxamidotryptamine (5-CT), α-methylserotonin (α-M-5-HT), and 2-methylserotonin (2-M-5-HT) on afferent discharge activity

This protocol consisted of ten cats subjected to 5-10 min of abdominal ischaemia followed by 2-3 min of reperfusion. After identification of an ischaemically sensitive C fibre unit, we opened the abdomen, located the receptive field of the nerve ending and measured the conduction velocity, as noted previously. Warm, moist gauze sponges were placed over the viscera, and the abdomen was closed with towel clamps. Subsequently, 5-HT (20 μg kg−1), 5-CT (100 μg kg−1, 5-HT1 receptor agonist), α-M-5-HT (100 μg kg−1, 5-HT2 receptor agonist) or 2-M-5-HT (100 μg kg−1, 5-HT3 receptor agonist) were injected intra-arterially. These drugs were purchased from Research Biochemicals International (RBI, Natick, MA, USA), were dissolved in physiological saline and prepared fresh daily. The four agonists were injected randomly, maintaining at least 25 min of recovery between injections. The afferent activity was recorded as we have described previously (Fu et al. 1996). The doses of these drugs have been shown to effectively stimulate vagal and cutaneous afferents (Taiwo & Levine, 1992; Grundy et al. 1994).

Effect of 3-tropanyl-indole-3-carboxylate (ICS-205-930 or tropisetron) on the response of afferents to ischaemia

The effect of 5-HT3 receptor antagonism with tropisetron (200 μg kg−1, i.v.) on the afferent's response to 10 min of ischaemia was studied in ten animals. Tropisetron (RBI) was dissolved in 0.9 % (w/v) NaCl to a concentration of 2 mg ml−1 and was prepared fresh daily. We repeated ischaemia 35-45 min after the first period of ischaemia, including at least 15 min after treatment with tropisetron. In the occasional circumstance, where afferent activity was suppressed completely, we mechanically manipulated the receptive field or stimulated it with an electrode to establish viability of the nerve ending.

To differentiate between variations in afferent response related to drug effect and time-related effects, since the elimination half-life following intravenous administraton of tropisetron is 7.3 h (Lee, Plosker & Mctavish, 1993), six additional C fibre afferents were utilized to determine repeatability of the afferent's response. In this protocol, after identification, each C fibre was treated identically except that 0.9 % NaCl (2-3 ml, i.v.) was used in place of tropisetron.

Effect of tropisetron on the response of afferents to 5-HT

The effect of 5-HT3 receptor blockade with tropisetron (200 μg kg−1, i.v.) on the response of eleven ischaemically sensitive C fibres to 5-HT was investigated in ten cats. After identification of an ischaemically sensitive unit, we injected 5-HT (20 μg kg−1) into the thoracic descending aorta while recording afferent activity. We repeated the injection of 5-HT (20 μg kg−1, i.a.) 25-30 min after its initial injection including at least 15 min following treatment with tropisetron. After treatment with tropisetron, we administered bradykinin (10 μg) into the thoracic descending aorta to establish responsiveness of the afferent nerve ending.

To determine the reproducibility of afferent responses to 5-HT, eight additional animals were studied as time control. After identification of an ischaemically sensitive unit, each animal was treated in an identical manner but was not subjected to tropisetron.

Effect of indomethacin on the response of afferents to 5-HT stimulation

Eight additional animals were subjected to 5-10 min of abdominal ischaemia followed by 2-3 min of reperfusion. After identification of an ischaemically sensitive unit, 5-HT (20 μg kg−1) was injected into the thoracic descending aorta. Indomethacin (Sigma) was dissolved in sodium carbonate (100 mM), diluted by 0.9 % NaCl to a concentration of 10 mg ml−1 and administered (5 mg kg−1) into the femoral vein. This dose effectively inhibits cyclooxygenase activity in cats (Longhurst et al. 1991). Subsquently, we repeated injection of 5-HT (20 μg kg−1) 30 min after its initial injection, including at least 15 min after intravenous administration of indomethacin.

Data analysis

Peak discharge rates of ischaemically sensitive afferents were measured over 60 s during 3 to 5 min of control, and 10 min of ischaemia, when the greatest number of spikes occurred (Fu et al. 1996). We measured the afferent response to each 5-HT receptor agonist by averaging discharge rates of the afferent during the entire period of response. We assessed the latency of afferent response to ischaemia and each 5-HT receptor agonist from the time of arterial occlusion or intra-arterial injection of the chemicals to the point when sustained discharge activity of afferents exceeded a 50 % increase over baseline. If an afferent did not respond to ischaemia after treatment with drugs, an onset latency equal in length to the maximum period of observation was assigned.

Data are expressed as means ±s.e.m. The effects of repeated injection of 5-HT and tropisetron and repeat ischaemia on the responses of the afferents were compared using one-way repeated-measures analysis of variance with a post hoc Bonferroni t test. If the data were not normally distributed, as determined by the Kolmogorov-Smirnov test, they were compared with the Friedman repeated-measures analysis of variance on ranks with Dunnett's test. We examined the afferent discharge activity response to each 5-HT receptor agonist with Student's paired t test. Student's paired t test also was used to evaluate the effects of tropisetron and indomethacin on the 5-HT-induced increases in discharge activity of the afferents. We used the Wilcoxon signed rank test to compare data, if the data were not normally distributed. Statistical calculations were performed with SigmaStat software (Jandel Scientific Software, San Rafael, CA, USA). Values were considered to be significantly different when P < 0.05.

RESULTS

Effects of 5-HT, 5-CT, α-M-5-HT and 2-M-5-HT on afferent activity

Figure 1 displays original tracings of an ischaemically sensitive C fibre (CV = 0.33 m s−1) innervating the bile duct during injection of 5-HT (Fig. 1A), 5-CT (Fig. 1B), α-M-5-HT (Fig. 1C) and 2-M-5-HT (Fig. 1D). The discharge activity of this C fibre afferent increased from 0.03 to 0.78 impulses s−1 during ischaemia. After release of aortic occlusion, the discharge activity of this fibre gradually returned to control levels. Injection of 5-HT (Fig. 1A) and 2-M-5-HT (Fig. 1D) resulted in an immediate burst of afferent activity after 4 s. In contrast, injection of 5-CT (Fig. 1B) and α-M-5-HT (Fig. 1C) did not alter the impulse activity of this afferent.

Figure 1.

Response of an ischaemically sensitive C fibre afferent to 5-HT, 5-CT, α-M-5-HT or 2-M-5-HT

Original tracings of an ischaemically sensitive C fibre (CV = 0.33 m s−1) innervating the bile duct. Abdominal ischaemia increased the baseline discharge activity of this afferent from 0.03 to a peak activity of 0.78 impulses s−1 after an onset latency of 170 s. A-Dshow representive tracings of discharge activity of this afferent and phasic aortic pressure during injection of 5-HT (A), 5-CT (B), α-M-5-HT (C) or 2-M-5-HT (D).

The effect of 5-HT, 5-CT, α-M-5-HT and 2-M-5-HT treatment on the entire group of twelve ischaemically sensitive C fibres (CV = 0.61 ± 0.05 m s−1) is summarized in Fig. 2. The location of each of the twelve afferent nerve endings is shown in Table 1. Aortic occlusion decreased distal mean arterial pressure from 89 ± 10 to 9 ± 1 mmHg (P < 0.05). We have shown previously that this degree of arterial occlusion is associated with a significant increase in portal venous blood and mesenteric lymph lactate concentration within 5 min (Longhurst et al. 1991; Longhurst, Benham & Rendig, 1992), following a significant decrease in portal venous blood and tissue PO2 (Longhurst, 1995; Fu et al. 1996). Injection of 5-HT (20 μg kg−1) into the thoracic descending aorta stimulated all twelve fibres, significantly increasing their discharge activity from 0.23 ± 0.05 to 0.96 ± 0.09 impulses s−1, after an onset latency of 5.7 ± 1.4 s. Intra-arterial injection of 2-M-5-HT (100 μg kg−1) stimulated eleven of twelve C fibres, significantly increasing their discharge activity from 0.25 ± 0.05 to 0.90 ± 0.10 impulses s−1, after an onset latency of 3.3 ± 0.4 s. In contrast, injection of α-M-5-HT (100 μg kg−1, i.a.) stimulated only two of twelve fibres, and thus did not significantly alter impulse activity of the group of fibres (0.29 ± 0.05 to 0.31 ± 0.06 impulses s−1); injection of 5-CT (100 μg kg−1, i.a.) did not stimulate any of the twelve fibres tested (0.26 ± 0.06 to 0.29 ± 0.06 impulses s−1).

Figure 2.

Mean responses of ischaemically sensitive C fibre afferents to 5-HT, 5-CT, α-M-5-HT or 2-M-5-HT

Bar graphs showing peak impulse activity of twelve ischaemically sensitive abdominal C fibre afferents before (open columns) and after (filled columns) intra-arterial injection of 5-HT (20 μg kg−1), 5-CT (100 μg kg−1), α-M-5-HT (100 μg kg−1) or 2-M-5-HT (100 μg kg−1). See text for explanation of abbreviations. Columns and error bars are means ±s.e.m.*P < 0.05 compared with respective control value.

Table 1. Location of ischaemically sensitive abdominal visceral C fibre afferent endings
 Protocol
  1. Values reflect numbers of afferent endings.

Location5-HT, 5-CT,α-M-5-HT, 2-M-5-HTRepeat ischaemiaTropisetronRepeat 5-HTTropisetron on 5-HTIndomethacin on 5-HT
Duodenum113
Mesentery21212
Pancreas312242
Porta hepatis325353
Bile duct1111
Gall bladder211
Jejunum11
Total126128119

Effect of tropisetron on the response of afferents to ischaemia

Figure 3 shows a neurohistogram and original recordings of an ischaemically sensitive C fibre with a conduction velocity of 0.43 m s−1 that innervated the pancreas. Ischaemia increased the baseline activity of this afferent from 0.02 to 2.06 impulses s−1 after an onset latency of 136 s (Fig. 3A). Antagonism of the 5-HT3 receptor with tropisetron (200 μg kg−1, i.v.) attenuated the increase in discharge activity of this afferent (0.03 to 0.63 impulses s−1) during repeat ischaemia after an onset latency of 204 s (Fig. 3B).

Figure 3.

Response of an ischaemically sensitive C fibre afferent to 10 min of abdominal ischaemia before and after treatment with tropisetron

Neurohistograms showing responses of an abdominal visceral C fibre (CV = 0.43 m s−1) innervating the pancreas to 10 min of ischaemia. Ischaemia increased the baseline activity from 0.02 to a peak activity of 2.06 impulses s−1 during ischaemia after an onset latency of 136 s (A). Tropisetron (200 μg kg−1, i.v.) treatment attenuated the increase in discharge activity of this afferent (0.03 to 0.63 impulses s−1) during repeated ischaemia after an onset latency of 204 s (B). Records labelled 1-4 are representative tracings of the discharge activity of the C fibre and phasic aortic pressure at the times indicated by arrows above the histogram.

Figure 4 B summarizes the effect of treatment with tropisetron on the peak 60 s impulse activity of twelve (2 C fibres were recorded in 2/10 animals) ischaemically sensitive C fibres (CV = 0.56 ± 0.06 m s−1) located in the duodenum, pancreas, porta hepatis, bile duct or gall bladder (Table 1). Aortic occlusion significantly decreased mean arterial pressure from 87 ± 11 to 10 ± 2 mmHg (P < 0.05). Ischaemia significantly increased baseline discharge activity of these afferents from 0.09 ± 0.03 to 1.62 ± 0.18 impulses s−1 after an onset latency of 174 ± 19 s. Tropisetron treatment did not alter distal arterial pressure during ischaemia (10 ± 2 vs. 11 ± 1 mmHg) or the pre-occlusion mean arterial pressure (87 ± 11 vs. 85 ± 8 mmHg), comparing before versus after treatment. However, 5-HT3 receptor blockade with tropisetron significantly attenuated the peak discharge activities of the C fibre during ischaemia (Fig. 4B), and increased the onset latency (274 ± 30 s, P < 0.05). This attenuation in activity of afferents during the second period of ischaemia was not due to a general decrease in reactivity over time since six additional ischaemically sensitive afferents (CV = 0.43 ± 0.04 m s−1) consistently were responsive to repeated 10 min periods of abdominal ischaemia (Fig. 4A), an observation that was consistent with our study (Fu et al. 1997b).

Figure 4.

Mean and summated responses of ischaemically sensitive C fibre afferents to 10 min of abdominal ischaemia before and after the 5-HT3 receptor blockade

A, bar graph summarizing the effect of repeat ischaemia on peak (60 s) impulse activity of six ischaemically sensitive C fibre afferents before and during 10 min of ischaemia. B, bar graph summarizing the effect of tropisetron (200 μg kg−1, i.v.) on peak (60 s) impulse activity of twelve abdominal visceral C fibre afferents during 10 min of ischaemia. C, neurohistograms showing summated 5 s discharge activity of twelve ischaemically sensitive C fibre afferents during 10 min of ischaemia before (a) and after (b) treatment with tropisetron (200 μg kg−1, i.v.). Data are presented as means ±s.e.m.*P < 0.05 compared with pre-ischaemia value; †P < 0.05, post-tropisetron (+ Tropisetron) vs. pre-tropisetron (- Tropisetron) control.

Figure 4 C is a neurohistogram showing summated 5 s discharge activity during 10 min of ischaemia in all twelve C fibre afferents before and after application of tropisetron. Similar to the changes in peak 60 s discharge activity, summated impulse activity during the entire 10 min of ischaemia was attenuated by 44 % after application of tropisetron.

Effect of tropisetron on the response of afferents to 5-HT

The effect of 5-HT3 receptor blockade with tropisetron on the response of eleven other (2 C fibres were recorded in 1/10 cats) ischaemically sensitive afferents (CV = 0.52 ± 0.04 m s−1) to 5-HT was examined (Fig. 5B). We found that 5-HT (20 μg kg−1) injected intra-arterially significantly increased the discharge activity of all eleven C fibres after a mean onset latency of 3.5 ± 1 s. Tropisetron (200 μg kg−1, i.v.) did not alter mean arterial pressure (86 ± 10 before vs. 88 ± 9 mmHg after), but virtually eliminated the afferents’ responses to 5-HT (Fig. 5B). This decrease in responsiveness of afferents to 5-HT was not due to a general decrease in reactivity over time because eight other ischaemically sensitive afferents (CV = 0.62 ± 0.05 m s−1) responded consistently to repeated injection of 5-HT (20 μg kg−1) over the same time frame (Fig. 5A). Also, each of the eleven afferents still responded to intra-arterial injection of 10 μg of bradykinin (0.07 ± 0.02 to 1.21 ± 0.18 impulses s−1, P < 0.05) after tropisetron.

Figure 5.

Effect of 5-HT3 receptor blockade on the responses of ischaemically sensitive C fibre afferents to 5-HT

A, effect of repeat injection of 5-HT (20 μg kg−1) on impulse activity of eight ischaemcally sensitive C fibre afferents. B,effect of injection of 5-HT (20 μg kg−1) on impulse activity of eleven ischaemcally sensitive C fibre afferents before and after treatment with tropisetron (200 μg kg−1, i.v.). Columns and error bars are means ±s.e.m.*P < 0.05 compared with respective control value; †P < 0.05, post-tropisetron (+ Tropisetron) vs. pre-tropisetron (- Tropisetron) control.

Effect of indomethacin on afferent response to 5-HT

Figure 6 summarizes the effect of indomethacin on the discharge activity of nine (2 C fibres were recorded in 1/8 cats) ischaemically sensitive afferents (CV = 0.62 ± 0.05 m s−1) during intra-arterial injection of 5-HT (20 μg kg−1). We observed that indomethacin (5 mg kg−1, i.v.) did not alter the responses of these afferents to exogenous 5-HT or their latency of response (3.4 ± 0.9 vs. 3.7 ± 1.2 s, initial vs. repeat, respectively). Table 1 shows the location of these afferent endings.

Figure 6.

Effect of indomethacin on the responses of ischaemically sensitive C fibre afferents to 5-HT

Effect of 5-HT (20 μg kg−1, i.a.) on impulse activity of nine ischaemia-sensitive C fibre afferents before and after treatment with indomethacin (5 mg kg−1, i.v.). Columns and error bars are means ±s.e.m.*P < 0.05 compared with respective control value.

DISCUSSION

Three important observations were made in this study. First, endogenously produced 5-HT contributes to activation of abdominal sympathetic C fibre afferent nerve endings during brief mesenteric ischaemia. Second, activation of 5-HT3 receptors, but not 5-HT1 or 5-HT2 receptors, is responsible for 5-HT-induced activation of ischaemically sensitive abdominal sympathetic afferent nerve endings. In this regard, a 5-HT3 receptor agonist, but not 5-HT1 or 5-HT2 receptor agonists, activated ischaemically sensitive abdominal sympathetic afferents. Furthermore, the response of abdominal visceral afferents to ischaemia and exogenous 5-HT was attenuated significantly after treatment with tropisetron, a 5-HT3 receptor antagonist. Third, inhibition of endogenous prostaglandin production with indomethacin did not affect the response of ischaemically sensitive C fibre afferents to 5-HT. Data from the present study strongly suggest that 5-HT produced during ischaemia contributes to activation of ischaemically sensitive abdominal sympathetic C fibre afferents through stimulation of 5-HT3 receptors.

Abdominal ischaemia activates abdominal sympathetic afferents and evokes reflex excitation of the cardiovascular system (Huang & Longhurst, 1994; Pan et al. 1996; Rendig et al. 1997). Visceral ischaemia and reperfusion lead to metabolic changes that result in the production of a number of substances, including lactic acid, prostaglandins, LTB4, bradykinin, histamine and reactive oxygen species (Longhurst, 1995; Fu et al. 1997a). These metabolites each individually stimulate, sensitize or modulate the activity of ischaemically sensitive abdominal visceral afferent endings (Longhurst, 1995; Fu et al. 1997b). The concentration of 5-HT is particularly high in the gastrointestinal tract, which contains large numbers of enterochromaffin cells, the principal source of 5-HT. Platelets also store and can release 5-HT when activated and aggregated at the site of vascular injury (Van den Berg, Schmitz, Benedict, Malloy, Willerson & Dehmer, 1989). In this regard, previous studies have demonstrated that the plasma concentration of 5-HT is elevated markedly (18- to 27-fold) at the site of coronary arterial stenosis and in coronary sinus blood just prior to complete occlusion of coronary artery when a thrombus forms (Benedict, Mathew, Rex, Cartwright & Sordahl, 1986). Presumably a similar mechanism can occur during thrombosis of the mesenteric vasculature. Strauss et al. (1973) have shown that the concentration of 5-HT in portal venous plasma is increased during prolonged ischaemia (30-75 min) and early reperfusion. Of more direct relevance to the current study, we have shown that 5-HT levels are significantly increased in both portal venous plasma and intestinal lymph during brief (10 min) periods of abdominal ischaemia (Fu et al. 1997a). The present investigation documents that increased 5-HT in the local tissue is capable of contributing to activation of ischaemically sensitive C fibre afferents.

According to a recent classification, at least seven subtypes of 5-HT receptors probably exist (Hoyer et al. 1994). However, based on the pharmacological data four distinct receptor subtypes, 5-HT1, 5-HT2, 5-HT3 and 5-HT4, have been located in the spinal cord and peripheral neurons (Hoyer et al. 1994). In this regard, studies from other laboratories have documented that 5-HT produces hyperalgesia by a direct action on cutaneous primary afferent neurons in rats through an action on the 5-HT1A receptors (Taiwo & Levine, 1992). Activation of 5-HT2 receptors can potentiate pain produced by inflammatory mediators (Abbott, Hong & Blier, 1996) and can increase transmission of nociception at the spinal level through stimulation of somatic C fibre afferents (Peroutka, 1994). Moreover, other investigators have demonstrated that 5-HT induces a short-latency, dose-dependent membrane depolarization in rabbit superior cervical ganglion cells through an action on 5-HT3 receptors (Wallis & North, 1978; Round & Wallis, 1986). However, the direct role of 5-HT receptor subtypes in activating ischaemically sensitive sympathetic afferent nerve endings is unknown. Since, in the present study, injection of the 5-HT1 receptor agonist, 5-CT, and the 5-HT2 receptor agonist, α-M-5-HT, into the mesenteric arterial circulation failed to stimulate ischaemically sensitive C fibre afferents that were responsive to 2-M-5-HT, a specific 5-HT3 receptor agonist, we suggest that 5-HT3, but not 5-HT1 or 5-HT2, receptors are responsible for the action of 5-HT on ischaemically sensitive abdominal sympathetic afferent nerve endings. This conclusion is confirmed by our observation that the response of these afferents to 5-HT was abolished completely by blockade of 5-HT3 receptors with tropisetron. Importantly also, the increased impulse activity of abdominal sympathetic C fibre afferents during ischaemia was significantly attenuated by tropisetron, an 5-HT3 receptor antagonist. Therefore, we have provided consistent neurophysiological data to document that, during ischaemia, the stimulating effect of endogenous 5-HT on abdominal sympathetic C fibre afferents is mediated by activation of 5-HT3 receptors, which we speculate are located on the sensory endings of the afferent neurons like receptors on visceral ganglion cell bodies (Mei, 1985).

The importance of interaction between prostaglandins and 5-HT is controversial. Some studies have shown that TxA2 enhances the release of 5-HT from platelets (Hamberg et al. 1975), and that aggregation of platelets by 5-HT in cat is primarily dependent upon formation of TxA2 (Meyers, Seachord, Holmson, Smith & Prieur, 1979). Chester, Allen, Tadjkarimi & Yacoub (1993) have found that activation of TxA2 receptors increases the response of 5-HT in human coronary arteries. However, other studies have shown that 5-HT does not stimulate PGI2 production in cultured endothelial cells (Coughlin, Moskowitz, Antoniades & Levine, 1981). Also, Yao et al. (1991) have observed that blockade of the cyclo-oxygenase pathway with aspirin does not prevent accumulation of 5-HT at sites of coronary endothelial injury. There is no information on the interaction between 5-HT and prostaglandins in hyperalgesia and activation of visceral afferents. We have shown previously that prostaglandins sensitize abdominal sympathetic afferents to ischaemia and that they enhance the response of these afferents to ischaemic metabolites such as histamine (Fu et al. 1997b) and bradykinin (Pan et al. 1994). We therefore wondered whether all or part of the action of 5-HT on ischaemically sensitive sympathetic afferents is dependent upon the production of prostaglandins. We observed that blockade of the cyclo-oxygenase pathway with indomethacin did not affect the response of the afferents to 5-HT. Thus, unlike histamine and bradykinin, 5-HT stimulates ischaemically sensitive abdominal sympathetic C fibre afferents by a mechanism that is independent of the cyclo-oxygenase pathway. It is more likely that 5-HT stimulates ischaemically sensitive sympathetic afferents through direct activation of 5-HT3 receptors, which is dependent upon a large conductance cation channel complex (Cornfield & Nelson, 1991). In this regard, intracellular microelectrode recordings show that the rapid depolarization produced directly by 5-HT is the result of opening Na+ channels in the nerve (Peters, Malone & Lambert, 1993).

An important issue is whether the concentration range of 5-HT used in the present study was in the physiological/pathophysiological range. Evidence from our previous study demonstrates that the concentration of 5-HT increases to 4.6 nmol ml−1 (i.e. ∼1 μg ml−1) in portal venous plasma or mesenteric lymph during brief abdominal ischaemia and reperfusion (Fu et al. 1997). It is likely that the concentrations of 5-HT present at the sites of release were considerably higher. Furthermore, it should be noted that 5-HT probably was diluted during the process of diffusion into the interstitium. In this regard, we have shown previously that aortic flow in cat is 200-230 ml min−1 (Huang, Stahl & Longhurst, 1995). We used approximately 5 s (equivalent to 18-20 ml of blood) for injection of 1 ml of 5-HT (∼58 μg ml−1, i.e. 20 μg kg−1× 2.9 kg body weight = 58 μg) into thoracic descending aorta in the present study. Thus, the concentration of 5-HT in celiac or superior mesenteric arterial blood was in the range 2.9-3.2 μg ml−1 since the 5-HT concentration was diluted with aortic blood during the period of injection (i.e. 58 μg (18-20 ml)−1= 2.9-3.2 μg ml−1). We believe therefore that the concentration of 5-HT present at afferent endings was within the pathophysiological range present during abdominal ischaemia and reperfusion.

In addition, although the effects of tropisetron are reversible, previous studies have shown that the mean elimination half-life following intravenous administraton of tropisetron is 7.3 h (Lee et al. 1993). Abdominal visceral C fibre afferents generally can be studied for about 3-4 h. Thus, it was not possible to examine for reversibility after administering tropisetron. Rather, two separate groups of animals were required to test the effect of tropisetron on the response of abdominal afferents to exogenous and endogenous 5-HT, respectively.

In conclusion, the results of this study have shown that activation of 5-HT3 receptors with any of several 5-HT3 receptor agonists can stimulate ischaemically sensitive abdominal sympathetic afferents in cats; conversely, 5-HT1 or 5-HT2 receptors do not play a role in this process. Blockade of 5-HT3 receptors during ischaemia attenuates the response of these afferents. Inhibition of cyclo-oxygenase activity does not affect the response of these afferents to 5-HT. Thus, these data demonstrate clearly that endogenously produced 5-HT contributes to the activation of ischaemically sensitive abdominal sympathetic afferents during ischaemia through an action on 5-HT3 receptors.

Acknowledgements

We gratefully acknowledge the expert technical assistance of Dr H.-L. Pan in the conduct of part of these experiments. We also thank Mr Stephen Rendig, Ms Roberta Holt and Mr Koullis Pitsillides for their technical assistance, and Ms Debbie Chase for her secretarial assistance. This study was supported by grants HL-36527 and HL-51428 from the National Institutes of Health and a grant 96-84 from the American Heart Association, California Affiliate. L.-W. Fu is a recipient of the Research Fellowship Award from the American Heart Association, California Affiliate. Part of this work was presented at the Neuroscience Meeting in Washington DC in 1996.

Ancillary