Gastro-oesophageal afferent and serotonergic inputs to vagal efferent neurones

Journalof the

ELSEVIER

Autonomic Nervous System Journal of the Autonomic Nervous System 49 (1994) 93-103

Gastro-oesophageal afferent and serotonergic inputs to vagal efferent neurones L.

Ashley Blackshaw

*

Gastroenterology Unit, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5000, Australia (Received 16 August 1993; revision received 17 November 1993; accepted 10 December 1993)

Abstract Peripheral 5-HT 3 receptor mechanisms are involved in activation of gastrointestinal (GI) mucosal vagal afferent fibres. 5-HT 3 receptor mechanisms in the central nervous system (CNS) may be involved in behavioural and reflex motility responses. This study investigates the processing of different sensory inputs in the CNS and the involvement of 5-HT 3 receptors at these different levels. In Urethane (1.5 g/kg, i.p.) anaesthetized, splanchnectomized ferrets, the jugular vein was cannulated for intravenous (i.v.) drug injection, and the coeliac axis for intraarterial (i.a.) injection close to the upper GI tract. The carotid artery was intubated with a T-cannula for CNS-directed intracarotid (i.c.) injections. An intragastric cannula was used for fluid distension (40-50 ml), and an oesophageal catheter for balloon distension (2 ml). Efferent fibres were dissected from the right cervical vagus for single-unit recording. Nineteen single vagal efferent fibres were selected, with low frequency resting discharge (2.5_+ 0.3 impulses/s), but no respiratory or cardiovascular phasic input. All responded rapidly ( < 2.5 s) to gastric distension (532 _+ 230% change in firing rate) and oesophageal distension (300 _+ 170%). Gastric distension caused excitation in 14 fibres, inhibition in 4 fibres, and a biphasic response in 1. Oesophageal distension excited 16 and inhibited 3. Discharge was also influenced by i.a. injection of 5-HT or the 5-HT 3 receptor agonist 2-methyl 5-HT (10-100/.~g) in all fibres tested. These responses consisted of rapid ( < 2.5 s) and powerful changes in firing rate, with excitation, inhibition or biphasic responses. 65% of responses to i.c. or i.v. injection were opposite in direction to those after close i.a. injection, indicating the activation of a different population of receptors. No differences were seen between effects of i.c. and i.v. injections. The 5-HT 3 receptor antagonist granisetron (100 ~ g / k g , i.v.) blocked or reduced efferent responses to 5-HT receptor agonists, whereas responses to gastric and oesophageal distension were unchanged. Thus there is extensive convergence of inputs from gastric and oesophageal mechanoreceptors onto vagal motorneurones. These central effects of mechanical stimuli do not involve 5-HT 3 receptor mechanisms. Other 5-HT 3 receptor inputs are evident, probably peripherally from GI mucosal afferent fibres and from within the CNS. Key words: Ferret; Oesophagus; Stomach, 5-HT receptor; Vagus nerve

1, Introduction

* Tel.:+61 8 224 5207; Fax:+61 8 224 0989.

T h e vagus nerves are i m p o r t a n t in m e d i a t i n g reflexes which control u p p e r g a s t r o i n t e s t i n a l function, for example, o e s o p h a g e a l peristalsis,

0165-1838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0165-1838(94)00010-H

94

L.A. B l a c k s h a w / Journal o/' the A u t o m ) m i c Nercous System 49 (1994) 9.L . 105

gastric and lower oesophageal sphincter relaxation, and coordination of propulsive activity in the stomach and small intestine. Reflexes controlling these functions may have both afferent and efferent pathways in the vagus nerves. For example, gastric or oesophageal distension causes relaxation of the stomach and lower oesophageal sphincter (LOS) via a vago-vagal reflex [1,19,22, 27]. In these reflexes synaptic transmission via release of neurotransmitters in the central nervous system is of obvious importance. In this respect, a high density of 5-HT 3 receptors are found in the dorsal medulla oblongata on vagal afferent terminations, which implies that 5-HT 3 mechanisms are involved in central control of vagal reflexes [5,25,29]. Otherwise little is known of neurotransmitter mechanisms in these reflexes. In addition to the CNS, 5-HT may be involved in the control of gastrointestinal function at two other levels: stimulation of sensory neurones in the mucosa after its release from enterochromaffin (EC) cells [3,13], and in neurotransmission in the myenteric and submucous plexuses of the gut wall [20,29,30]. There are a number of different subtypes of the 5-HT receptor which may be activated at each of these different locations; however, it is the 5-HT 3 subtype which mainly mediates rapid neuronal depolarisation [16,20,30]. 5-HT 3 receptor activation excites vagal afferent endings in the gastrointestinal mucosa [13]. On reaching the central nervous system, these mucosal afferent signals may be manifested as vagal reflexes, and at intense levels as nausea and vomiting [3]. A vago-vagal reflex is responsible for inhibition of gastric tone and motility after peripheral 5-HT administration [2]. However, reflexes via the CNS are not involved in 5-HT3-induced relaxation of the lower oesophageal sphincter (LOS) [10]. It is therefore possible that the organization of reflexes to the proximal stomach and the LOS and their sensitivity to 5-HT is different. The aim of this study was to elucidate the organization of vagal reflexes triggered from the stomach and distal oesophagus, and their sensitivity to 5-HT peripherally and centrally. A preliminary account of this work has been published in abstract form [11].

2. Methods

Experiments were performed on 7 male and 12 female ferrets (weight range 0.5-1.1 kg) anaesthetized with a single intraperitoneal dose of Urethane (1.5 g/kg). They were fed a standard carnivore diet with free access to water but were deprived of food for 18 h prior to experimentation. A tracheal cannula was introduced via the mouth, and the right jugular vein was cannulated for administration of further anaesthetic and drugs. The vagi were dissected free of the carotid arteries, and the left carotid artery was intubated with a T-cannula to allow monitoring of blood pressure, which was above 100 mmHg (mean) for the duration of experiments. This cannula was also used for injection of drugs directly into the CNS arterial supply. The oesophagus, which is composed entirely of striated muscle in the ferret [18], was intubated with a balloon catheter, which was positioned accurately during abdominal surgery. This catheter comprised an 8-lumen PVC extrusion 6 mm in diameter, around which silicon membrane was fitted to form a balloon which was distended via a sidehole to a maximum capacity of 5 ml. This was located within the distal oesophagus approximately 4 cm above the lower oesophageal sphincter. The greater splanchnic nerves were isolated at the crura of the diaphragm and sectioned. A fine polyethylene catheter was introduced via the aorta at the iliac bifurcation and passed such that the tip lay at the coeliac axis. This was used for rapid close-intraarterial injections to the upper gastrointestinal tract. Catheter position was confirmed at post mortem. A cannula was inserted into the stomach via the pylorus for drainage or distension with up to 50 ml of warm (37°C) 0.9% NaCI. Rectal temperature was maintained at 38°C + 0.5°C with a warming pad.

2.1. Neural recordings A paraffin pool was made in the neck by suturing skin and muscle to a steel ring. The right vagus nerve was mobilized and placed intact on a small perspex recording platform. Under a dissecting microscope (Olympus SZ60), the nerve

L.A. Blackshaw/ Journal of the Autonomic Neruous System 49 (1994) 93-103 sheath was split with a sharp blade over a length of approx. 5 ram. Fine filaments of efferent fibres were dissected from the main nerve trunk, and placed on a platinum hook recording electrode. Perineural connective tissue was attached to a reference electrode. Nerve activity was amplified and filtered (JRak, Melbourne BA1 and F1), and displayed on an oscilloscope (Tektronix 5111). Single unit action potentials were discriminated electronically on the basis of shape, duration and amplitude (JRak WD1). Filaments containing more than three active fibres were discarded because it was not normally possible to discriminate single units in these cases. A spike counter (Amalgamated Instruments, Sydney) was used to generate integrated output of action potential frequency, which was displayed on a 2-channel Rikadenki chart recorder, along with blood pressure. Neural activity was stored on a digital tape recorder (Sony PCM2300), and raw traces were displayed off-line using Macintosh-based software (National Instruments LabView). 2. 2. Experimental protocol U p o n confirmation of single unit recording, the spontaneous activity of the fibre was carefully observed. Those with respiratory or cardiovascular phasic input were discarded on the grounds that they were unlikely to receive a predominant gastrointestinal input. In order to evaluate mechanosensitive inputs from the stomach and oesophagus, these regions were distended. Gastric distension was achieved by infusion of 0.9% NaCI at 37°C over 8 - 1 0 s via the pyloric cannula. 40 ml were infused for animals with body weight below 0.75 kg, and 50 ml for animals over 0.75 kg. Distal oesophageal distension was p e r f o r m e d by inflation of the balloon with 2 ml air over 1-2 s. Fibres not responding with a > 50% peak change in firing rate upon gastric and oesophageal distension were not studied further. Fibres satisfying these criteria were subsequently tested with injections of 5-HT a n d / o r 2-methyl 5-HT at doses of 10 /xg, 50 /xg and 100 /xg where possible. Data are given for 50 txg injections. These injections were given towards the CNS into the carotid artery, close arterially to the u p p e r G I tract via

96

the coeliac axis and systemically via the jugular vein in r a n d o m order. This was to direct drugs towards different populations of receptor. The time allowed between injections was always over 2 min. No desensitization was seen where the same stimulus was given twice at this interval. The possibility of inputs from carotid body chemoreceptors in neural responses to intracarotid drug injection was rendered unlikely, as carotid body stimulation by intracarotid injection of CO2 enriched saline had no obvious effects on units included for analysis. The 5-HT 3 receptor antagonist granisetron was given via the jugular vein at doses of 100-200 g g / k g following recording of responses to distension and 5-HT a n d / o r 2-methyl 5-HT. 2.3. Drugs Granisetron and 2-methyl 5-HT were obtained from SmithKline Beecham Pharmaceuticals, Harlow, UK. 5-Hydroxytryptamine creatinine sulphate was obtained from Sigma. All drugs were dissolved in isotonic NaCI, which served as a vehicle control. Injection volumes were below 0.5 ml, and were standardised for each stimulus. 2.4. Data analysis Resting discharge of fibres was measured over 2 min. Responses to distension are expressed as the mean discharge rate during the stimulus (over 30 s to 1 rain). Responses to drug injection are shown as peak or nadir discharge rate. Statistical significance of differences between responses was assessed by the Wilcoxon signed rank test unless otherwise stated, and data expressed as mean _+ S.E.M., with n = number of observations.

3. Results

3.1. Resting vagal efferent discharge All fibres recorded showed a low frequency, irregular pattern of resting discharge (2.5 +_ 0.3 impulses/s, range 0.1-14.0, n = 19). This bore no obvious relationship to respiratory, cardiovascular

90

L.A. Blackshaw / Journal of the Autonomic Nerl:ous System 49 (1994) 93--103

o r g a s t r o i n t e s t i n a l c o n t r a c t i o n r h y t h m s , a n d rem a i n e d at t h e s a m e l e v e l t h r o u g h o u t e x p e r i m e n t s when unassociated with a particular stimulus. G r a n i s e t r o n (100 / z g / k g ) h a d n o o b s e r v a b l e eff e c t o n t h e p a t t e r n o r f r e q u e n c y o f r e s t i n g disc h a r g e in e f f e r e n t f i b r e s i m m e d i a t e l y a f t e r injection. W h e n d i s c h a r g e r a t e was a s s e s s e d 5 m i n after granisetron administration, discharge freq u e n c y w a s n o t s i g n i f i c a n t l y d i f f e r e n t to t h a t b e f o r e t r e a t m e n t (2.0 __+0.6 i m p / s b e f o r e , 1.6 + 0.6 i m p / s a f t e r ; P = 0.1, p a i r e d t-test).

c h a r g e was r a p i d l y c h a n g e d in all fibres, w i t h a t h r e s h o l d v o l u m e u s u a l l y b e t w e e n 10 a n d 20 ml. 14.fibres s h o w e d e x c i t a t i o n o f d i s c h a r g e r a t e (657 + 2 8 4 % o f basal), 4 s h o w e d i n h i b i t i o n (61 + 13%), and 1 fibre showed a biphasic response (see Table 1), w h i c h c o m p r i s e d e a r l y i n h i b i t i o n a n d l a t e r excitation. Responses were of short latency ( < t s u p o n r e a c h i n g t h r e s h o l d , e.g., Fig. 1) a n d s h o w e d an e a r l y d y n a m i c a n d a l a t e r static c o m p o n e n t (Fig. 2), as s e e n in g a s t r i c v a g a l m e c h a n o r e c e p t o r s [4,9,12]. D u r i n g e x c i t a t o r y r e s p o n s e s , a p h a sic m o d u l a t i o n o f d i s c h a r g e f r e q u e n c y w a s evid e n t in 4 fibres. T h i s c o r r e s p o n d e d to t h e exp e c t e d o c c u r r e n c e o f a n t r a l c o n t r a c t i o n s at 9 / m i n . T h i s w a s c o n f i r m e d in o n e e x p e r i m e n t by c o n n e c t i n g t h e g a s t r i c o u t l e t t u b e to a p r e s s u r e t r a n s d u c e r , a n d is as s e e n in o t h e r s t u d i e s o f v a g a l a f f e r e n t a n d e f f e r e n t d i s c h a r g e [4,6,8,12,15,

3. 2. Responses to gastric and oesophageal distension Upon distension of the stomach with 40-50 ml i s o t o n i c i s o t h e r m a l 0 . 9 % NaC1 in a s i n g l e s t e p o v e r 8 - 1 0 s via t h e p y l o r i c c a n n u l a , r e s t i n g dis-

Table 1 Direction of responses in all fibres to all stimuli Unit No.

2-Methyl 5-HT

5-HT

i.a.

i.v.

i.c.

1

--

±

±

2 3 4 5

++

+ -

-

6

-

-

±

7 8 9 10 11 12 13 14 15 16 17 18 19 Total+ TotalTotal ±

i.a.

i.v.

i.c.

±

±

++ -

± ++ ++ -

±

++

++ ++ ++ ++

++ ++ ++ ++

++ 6 3 2

++ 6 1 2

GD

++ + + + ++

++ ++ ++ ++

.

±

-

-

++

++

++

. ++ + ± . ++ 4 7 1

OD

3 3 2

.

.

.

. ++ ± ++ . 4 3 2

++ ++ ++ .

. 4 3 1

.

.

.

.

.

+ ++ ++ + ++ + + ++ + ++

++ ++ ++ ++ ++ ++ + --± + +

++ 16 3 0

++ 14 4 l

.

+ + denotes > 100% excitation of baseline firing rate; + denotes < 100% excitation; - - denotes complete inhibition; and denotes partial inhibition. Where no symbol is present, it was not possible to test the stimulus. Oesophageal (OD, 2 ml) and gastric ( G D , 40-50 ml) distension caused mainly excitation of discharge. In response to 5-HT and 2-methyl 5-HT, inhibitory and biphasic responses were more frequent, i.a., close-intraarterial; i.v, intravenous; i.c., intracarotid. Note the frequent occurrence of similar responses to i.v. and i.c. injections and the lack of relationship between responses to 5-HT agonists and those to gastro-oesophageal distension.

L.A. Blackshaw /Journal of the Autonomic NerL,ous System 49 (1994) 93-103

23]. After distension for 1 min, free drainage of the stomach was allowed, which resulted in a rapid return of discharge rate to predistension level (Figs. 1 and 2). Oesophageal balloon distension was performed by injecting 2 ml air rapidly via the oesophageal catheter over 1-2 s. Of 19 fibres tested, 16 efferent fibres showed excitation and 3 showed inhibition in response to oesophageal distension (Table 1). The threshold for efferent responses was < 1 ml, however this may reflect that there was no deadspace in the lumen because the catheter was designed to occupy the full inner luminal diameter before distension commenced. Responses were of short latency ( < 1 s, e.g., Fig. 1), and showed a dynamic and static component, although the differences between dynamic and static components were relatively smaller than those in responses to gastric distension (e.g., Fig. 2). This is presumably owing to the lack of ability of the oesophagus to accommodate like the stomach, so the stimulus to mechanoreceptors is better maintained. After 30 s of oesophageal disten-

Oesophageal Distension 2 ml

97

sion, the balloon was allowed to deflate, which occurred over < 1 s, at which point efferent discharge rapidly resumed its predistension level (e.g., Fig. 2). In 3 fibres whose discharge was powerfully excited by oesophageal distension, discharge dropped below predistension level for 1-2 s before returning to baseline frequency, presumably due to transient offioading of the mechanoreceptor afferent input. In order to establish the involvement of 5-HT 3 receptor mechanisms at any point along the reflex pathway, responses to gastric or oesophageal distension were tested after treatment with granisetron up to a dose of 200 /zg/kg. Granisetron had no effect on the amplitude or profile of these responses (Figs. 2 and 3).

3.3. Responses to peripheral close systemic administration of drugs (a) 2-Methyl 5-Hydroxytryptamine Selective activation of vagal mucosal endings in the upper gastrointestinal tract by close in-

i

Deflate

T Gastric Distension 40 ml

2Methyl 5-HT 50pg IA

Drain

30 sec

Fig. 1. Raw record of action potentials in a vagal efferent fibre. The low frequency resting discharge was rapidly inhibited by gastric distension, esophageal distension and close intraarterial injection of 2-methyl 5-HT. The response to distension was maintained for the duration of the stimulus, and returned to basal levels rapidly, whereas the response to 2-methyl 5-HT lasted approximately 2 rain after injection.

98

L.A. Blackshaw /.Iournal t~f the Autonomic Nercous System 49 (1994) ¢3-103

CONTROL

imp/5

sec

0 OBD 2ml

GD 5Oral

251

t 2 M 5HT 50pg IA

~ 2M 5HT 50lug IV

GRANISETRON

imp/5see

0~ 1

GO 50ml

o e o 2ml

T 2M

5HT

soug IA

12M

SliT

so,g tv 2 rain

Fig. 2. Integrated record of firing rate in a vagal efferent fibre, 5 s reset. The low frequency resting discharge was rapidly excited gastric distension (GD 50 ml), oesophageal distension (OBD 2 ml) and by close intraarterial injection of 2-methyl 5-HT (2M5HT ~g, i.a. (IA)). Intravenous (i.v. (IV)) injection of 2-methyl 5-HT evoked the opposite response, inhibiting this unit, Responses gastric and oesophageal distension were unchanged after treatment with granisetron (100/zg/kg, i.v.), whereas both responses 2-methyl 5-HT were abolished.

n=19

[ ] Inhibited ,1 Control •

17s0

Inhibited ) Granlmetron1 0 ~ g

lsoo % Chmlge

1250 lO00

10

~

~

c=:i==

m

Gastric

Olmopllogeal

II~;zermlon 40-50rnl

D,IMenldon 2ml

Fig. 3. Excitatory and inhibitory responses of efferent fibres to gastric (GD, 40-50 ml) or oesophageal (OD, 2 ml) distension were not significantly affected by 5-HT 3 receptor antagonism with granisetron.

by 50 to to

traarterial injection of the selective 5-HT 3 receptor agonist 2-methyl 5-HT (10-100 /zg) into the coeliac axis had three different types of effect on vagal efferent discharge: 7/12 fibres tested showed inhibition, 4 fibres showed excitation, and one showed a biphasic response (Table 1). The latency of these responses was rapid in all cases (1.5-5 s, e.g., Figs. 1 and 2), and the short latency that was observed probably reflected the sum of the time taken for blood flow from the cannula tip to the receptive endings in the gastrointestinal mucosa plus the vagal afferent and efferent conduction time (see Refs. 7,9,13). Both inhibitory and excitatory responses showed powerful changes in discharge (Fig. 4), and were maintained for 74 + 21 s before returning to preinjection level. No phasic modulation of firing was seen during responses, and no long-term changes

L.A. Blackshaw /Journal of the Autonomic Nerl,ous System 49 (1994) 93-103

in resting discharge were observed. Following granisetron treatment, responses to 2-methyl 5-HT were completely blocked (Fig. 4).

99 n=7

•

Excited

[]

Inhibited J*

[]

Excited

•

Inhibited [

/

Control

Granisetron 100pg/kg

• p<0.05 vs Control

1400

(b) 5-Hydroxytryptamine

)

t200 •

Efferent responses to peripheral activation of all 5-HT receptor subtypes by 5-HT showed similar patterns to those after 2-methyl 5-HT (Table 1). The duration of response was 81 + 38 s, and powerful changes in discharge were consistently observed (Fig. 5). The only major difference that was observed between 5-HT and 2-methyl 5-HT was that whereas the response to 2-methyl 5-HT was completely blocked by granisetron treatment, the amplitude of response to 5-HT was unchanged• However, the latency of responses was significantly longer after granisetron in response to 5-HT (1.3 s vs. 11.2 s, P < 0.05 paired t-test) suggesting an indirect effect•

3.4. Responses to intrauenous and intracarotid administration of drugs In 65% of fibres, responses to close intraarterial injection of 5-HT or 2-methyl 5-HT (described above) and intravenous (i.v.) or intracarotid (i.c.) 5-HT receptor agonists (described below) were opposite in direction (Table 1; e.g., n=l 2

1~

•

Excited

[]

Inhlblted

)

--

Granlsetron 100pg/l(g

Control

• p<0,05 vs Control

~400

10c~% Change

Discharge Rate

800 soo • 4OO 2OO 0 -100 Close IA

Intra-carotld

intravenous

Fig. 5. Both inhibitory and excitatory responses to intracarotid and intravenous 5-HT were reduced or abolished by granisetron, but responses to close i.a. 5-HT were relatively unaffected. Latency of responses was, however significantly changed (see text), showing that the 5-HT 3 receptor is the predominant subtype involved in all direct responses. Paired data from inhibitory and excitatory responses together were used for analysis by Wilcoxon signed-rank test.

Fig. 2). Different mechanisms are therefore activated depending upon the site of 5-HT receptor activation.

(a) 2-Methyl 5-hydroxytryptamine Responses of efferent fibres to i.v. or i.e. 2methyl 5-HT showed excitatory, inhibitory and biphasic responses (Table 1). Latency of response was not normally different from that after i.a. injection, however the amplitude of responses was significantly greater after i.v. and i.c. injection ( P < 0•05). Both inhibitory and excitatory responses to i.v. and i.c. 2-methyl 5-HT were completely blocked by granisetron ( P < 0.01, Fig. 4).

lOe~

% Change

Dim:barge Rate

(b) 5-Hydroxytryptamine

S~

Close IA

Intra-carotid

~ntravenous

Fig. 4. Both inhibitory and excitatory responses of efferent fibres to 5-HT 3 receptor activation with 50 p,g 2-methyl 5-HT after close-intraarterial (IA), intracarotid and intravenous injection were all completely blocked by granisetron (100 g g / k g ) . Responses to i.v. and i.c. injection of 2-methyl 5-HT were larger than those to i.a. injections. This relationship was statistically significant ( P < 0.05). Paired data from inhibitory and excitatory responses together were used for analysis by Wilcoxon signed-rank test.

Efferent responses to i.v. or i.c. 5-HT were not significantly different to those to 2-methyl 5-HT in direction, latency, size or duration• Granisetron blocked responses to i.v. and i.c. 5-HT (Fig. 6). whereas responses to i.a. 5-HT were affected only slightly in amplitude (see earlier).

4. Discussion

This study is the first to show that there is extensive convergence of inputs from gastric and

101)

L.A. Blackshaw /.lournal Of the Autonomic Nervous System 49 (1994) 93-103

oesophageal mechanoreceptors onto vagal efferent fibres. The patterns of excitation and inhibition of efferents may correspond to their targets within the enteric nervous system. For example, if smooth muscle relaxation is the endpoint of a vagal reflex, then efferent fibres showing excitation would be targeted toward activation of inhibitory enteric motorneurones, whereas fibres showing inhibition may be targeted toward inhibition of excitatory enteric pathways. This type of reciprocal control may serve to amplify and finetune the effector response (see also Refs. 6,8,15,23). This example represents only one of several possible consequences of efferent responses seen in this study, but would appear to be one of the most likely. The patterns of convergence infer that the destination of efferent fibres is in the region of the gastro-oesophageal junction, although this is not possible to confirm as fibres were truncated at the recording point. Nevertheless, the mechanical stimuli used in this study are known to cause relaxation of the LOS and proximal stomach via vagal reflexes [1,2,19,22,27]. The responses of efferent fibres closely reflect the gastric or oesophageal vagal mechanoreceptor input [4,9,12,31] although efferent fibres showed either excitation or inhibition of discharge, whereas mechanoreceptors show excitation only. Efferent responses to gastric distension are comparable with those seen in previous studies of vagal efferent fibres [6,8,15,23] which together indicate strong mechanoreceptor afferent inputs to vagal reflexes. In addition to vagal mechanoreceptor inputs, spinal afferent inputs to vagal reflexes also exist [7,23]. In this study, the greater splanchnic nerves were sectioned in order to focus on vagal inputs alone, although spinal inputs via the lesser and least splanchnic nerves from more distal regions of intestine and oesophageal spinal afferents would have remained intact. The data from this study show that granisetron, a potent and selective 5-HT 3 receptor antagonist which crosses the blood-brain barrier [28] did not affect efferent responses to distension. Afferent inputs from gastric and oesophageal mechanoreceptors to vagal efferent fibres do not therefore involve 5-HT 3 receptor mechanisms. This obser-

vation is consistent with evidence from studies in the ferret indicating that vagal gastrointestinal mechanoreceptor endings are not directly sensitive to 5-HT administered peripherally [12]. The present study presents evidence 5-HT 3 receptors also do not appear to be involved further centrally in transmission of mechanoreceptor signals through vago-vagal reflexes. This is in contrast to positive findings of 5-HT 3 receptor mechanisms involved in central control of the baroreceptor reflex in rat [26], and is despite the existence of a high density of presynaptic 5-HT 3 receptors on central vagal afferent terminations in ferret [5,25]. Therefore, either a population of vagal afferents other than gastric and oesophageal mechanoreceptors possesses central 5-HT 3 receptors, or presynaptic inputs from other afferents or other nuclei must be significantly activated to affect transmission through the reflex pathway. Vagal mucosal afferent endings are well-positioned to respond to release of 5-HT from mucosal enterochromaffin (EC) cells because of their close proximity within the GI tract. This may account for the effects of mucosal 5-HT release in pathophysiological situations. Degranulation of EC ceils occurs in response to luminal nutrients and toxins [24], and has been shown to be the basis of vomiting in many cases. These include side-effects of cancer chemotherapy and radiotherapy, which may be alleviated by vagotomy or 5-HT 3 antagonism [3,28]. Electrophysiological data further indicate that vagal mucosal afferents are powerfully activated by close i.a. injection of 5-HT via a 5-HT 3 receptor mechanism [13], whereas smooth muscle mechanoreceptors are only indirectly affected due to changes in motility [12]. Activation of mucosal afferents by 5-HT may cause vagal reflex effects on gastric motility, such as gastric relaxation following local injection of 5-HT, which is abolished by 5-HT 3 antagonism or vagotomy [2]. The responses of efferent fibres to peripheral close i.a. 5-HT in the present study represent the neurophysiological correlate of reflexes such as this. This study shows that the same fibres mediating reflex effects of 5-HT receptor agonists are also responsible for mediating reflex effects of gastric and oesophageal distension. This repre-

L.A. Blackshaw / Journal of the Autonomic Nert~ous System 49 (1994) 93-103

sents an economical means of achieving control with the small number of vagal efferent fibres available. Convergence of information from mechanoreceptors and mucosal receptors centrally onto vagal efferents has previously been shown, where responses to gastric and duodenal distension occurred in fibres which responded to mucosal receptor activation by intraluminal chemical stimuli [6,8]. These responses to chemical stimuli were generally small in amplitude and poorly reproducible compared to those following distension, suggesting that mucosal receptor inputs to vagal reflexes were only minor. However, a later study in the same species showed that responses in mucosal afferent endings to intraluminal chemical stimuli were smaller than those to 5-HT [13]. In the present study changes in efferent discharge after peripheral 5-HT agonists and distension were similar in amplitude. Mucosal receptors must therefore have a considerable potential input to vagal reflexes, particularly when large numbers of fibres from different regions are powerfully activated. This situation would occur after close i.a. injection of 5-HT 3 receptor agonists to the upper GI tract as a whole but not after chemical application to a small area of mucosa. The duration of responses to 5-HT receptor agonists was often 1-2 min in efferent fibres in this study, whereas the duration of individual mucosal afferent fibre responses is normally less than 1 min [13]. This is likely to reflect a persistent central effect of mucosal receptor input. It is unlikely to reflect changes in mechanoreceptor input secondary to motility, as 5-HT receptor induced inhibition of motility in an undistended gut causes only minor changes in mechanoreceptor firing [9,12]. Where excitation of motility may have occurred, giving rise to increased mechanoreceptor input, this may have influenced longer latency efferent responses to 5-HT (see later). After peripheral 5-HT injection close to the GI tract, it is inevitable that some of the compound will reach the systemic circulation, as not all is metabolized on the first pass through the liver [24]. This may be in sufficient concentration to activate bronchial and cardiovascular afferents [14,17,32], and possibly afferent endings in the

101

oesophageal body not reached by injections at the coeliac axis. These inputs may in turn prolong the period of central efferent activation. In order to account for this type of effect, intravenous injections were also given. Small quantities of systemic 5-HT may also reach the CNS to activate vagal reflex pathways more directly, particularly in paraventricular regions where the blood-brain barrier is weak. To investigate potential effects of activation of central 5-HT 3 receptors, intracarotid arterial injections were given via a T-cannula to administer directly into the central carotid arterial blood flow. The responses to intravenous and intracarotid injection were generally similar in direction and profile, which suggests that both were ultimately acting at the same site. Although it was not possible to resolve latencies of response less than within 0.5-1 s, it was often qualitatively observed that responses to i.v. injection had a distinct delay of at least 1 s, whereas responses to i.c. injection were practically instantaneous. This implies that the CNS was the common site of action, rather than pulmonary or thoracic receptors, although the region of activation within the CNS cannot be concluded from these data. A possible involvement of carotid body chemoreceptor inputs activated by i.c. and i.v. injections e n r o u t e to the CNS in efferent responses was excluded, since i.c. injection of CO z enriched saline had no effect on efferent fibre discharge in 3 experiments whereas i.c. injection of 5-HT agonists had powerful effects. The frequency of occurrence of responses to i.c. 5-HT agonists and their size indicates that 5-HT~ receptors are present functionally along central pathways which affect vagal efferent outflow, although the data show that they are not directly involved in mechanosensitive reflexes. In 65% of cases, efferent responses to i.v. or i.c. injections were in completely the opposite direction to those after peripheral administration, and were therefore unrelated. Whether there are functional or structural differences between 5-HT 3 receptor populations peripherally and centrally remains to be determined. Not all efferent responses to 5-HT injection were 5-HT 3 receptor-mediated. When the effects of 5-HT and 2-methyl 5-HT arc compared, it

102

L.A. Blackshaw /Journal of the Autonomic Nen'ous System 49 (1994) 93-103

becomes apparent that there is a non-5-HT 3 component to efferent responses after peripheral i.a. 5-HT injection (Fig. 6). This component was significantly longer in latency than the response prior to granisetron. The time course of these non-5-HT 3 mediated responses was similar to smooth muscle contractile responses seen in the stomach and lower oesophageal sphincter (LOS) after 5-HT3 antagonism [2,10]. The LOS response has been shown to be 5-HT 2 receptor mediated, and probably reflects a direct action of 5-HT on smooth muscle [10]. Thus, the granisetron-resistant response is probably secondary to mechanoreceptor activation following smooth muscle contraction. In conclusion, there is the potential for extensive convergence of peripheral mechanoreceptor and mucosal receptor inputs onto vagal reflexes. Central 5-HT 3 receptor activation probably has strong effects on the vagal motor outflow, but involvement of central 5-HT 3 receptors in normal reflex control of motility is unlikely. Involvement of central 5-HT 3 receptors in reflexes activated from the mucosa is the subject of further study.

Acknowledgements SmithKline Beecham Pharmaceuticals are acknowledged for their generous financial support of this project. I am grateful to Mrs. Voula Nisyrios for preparation of catheters and artwork, and to Prof. John Dent for his advice and use of Gastroenterology Unit facilities. Dr. Charles Malbert is thanked for his valuable computer expertise.

References [1] Abrahamsson, H., Studies on the inhibitory nervous control of gastric motility, Acta Physiol. Scand., 38 (1973) Suppl. 390. [2] Andrews, P.L.R., Bingham, S. and Davidson, H.t.M., Vagotomy and 5-HT 3 receptor antagonism modify the motility response to 5-HT in the gastric antrum of the anaesthetized ferret, J. Physiol. (Lond.) 422 (1990) 93P (Abstract). [3] Andrews, P.L.R., Davis, C.J., Bingham, S., Davidson,

H.I.M., Hawthorn, J. and Maskell, L., The abdominal visceral innervation and the emetic reflex: pathways, pharmacology, and plasticity, Can. J. Physiol. Pharmacol., 68 (1990) 325-345. [4] Andrews, P.L.R., Grundy, D., and Scratcherd, T., Vagal afferent discharge from mechanoreceptors in different regions of the ferret stomach, J. Physiol. (Lond.), 298 (1980) 513-524. [5] Barnes, J.M., Barnes, N.M., Costall, B., Naylor, 1.L., Naylor, R.J. and Rudd. J.A., Topographical distribution of 5-HT 3 receptor recognition sites in the ferret brain stem, Naunyn-Schmiedeberg's Arch. Pharmacol., 342 (1990) 17-21. [6] Blackshaw, L.A., D. Grundy and Scratcherd, T., Involvement of gastrointestinal mechano- and intestinal chemoreceptors in vagal reflexes: an electrophysiotogical study, J. Auton. Nerv. Syst., 18 (1987) 225-234. [7] Blackshaw, L.A. and Grundy, D., Splanchno-vagal convergence on to gastrointestinally modulated vagal efferent units in the ferret, J. Physiol. (Lond) 396 (1988) 35P (Abstract). [8] Blackshaw, L.A. and Grundy, D., Responses of vagat efferent fibres to stimulation of gastric mechano- and chemoreceptors in the anaesthetized ferret, J. Auton. Nerv. Syst., 27 (1989) 39-45. [9] Blackshaw, L.A. and Grundy, D., Effects of cholecystokinin (CCK-8) on two classes of gastroduodenal vagal afferent fibre, J. Auton. Nerv. Syst., 31 (1990) 191-202. [10] Blackshaw, L.A., Nisyrios, V. and Dent, J., Pathway of serotonin (5-HT) induced reduction of lower esophageal sphincter pressure in the ferret, Gastroenterology 104 (1993) A478 (abstract). [11] Blackshaw, L.A. Gastroesophageal vagal afferent and serotonergic inputs to ferret central vagal neurons, Gastroenterology 104 (1993) A478 (abstract). [12l Blackshaw, L.A. and D. Grundy. Effects of 5-Hydroxytryptamine (5-HT) on the discharge of vagal mechanoreceptors and motility in the upper gastrointestinal tract in the ferret, J. Auton. Nerv. Syst., 45 (1993) 51-59. [13] Blackshaw, L.A. and Grundy, D., Effects of 5-Hydroxytryptamine on discharge of vagal mucosal receptors from the upper gastrointestinal tract of the ferret, J. Auton. Nerv. Syst., 45 (1993) 41-50. [14] Coleridge, H.M. and Coleridge, J.C.G., Impulse activity in afferent vagal C-fibres with endings in the intrapulmonary airways of dogs, Respir. Physiol., 29 (1977) 125142. [15] Davison, J.S. and Grundy, D., Modulation of single vagal efferent fibre discharge by gastrointestinal afferents in the rat, J. Physiol. (Lond.), 284 (1978) 69-82, [16] Derkach, V., Surprenant, A. and North, R.A., 5-HT 3 receptors are membrane ion channels, Nature, 339 (1989) 706-709. [17] Dixon, M., Jackson, D.M. and Richards, I.M., The effects of histamine, acetylcholine and 5-hydroxytryptamine on lung mechanics and irritant receptors in dogs, J. Physiol. (Lond.), 287 (1979) 393-403.

L.A. Blackshaw / Journal of the Autonomic Nereous System 49 (19941 93-103 [18] Fox, J.G., The Biology and Diseases of the Ferret, Lea & Febiger, Philadelphia, 1988. [19] Franzi, S.J., Martin, CJ., Cox, M.R. and Dent, J., Response of canine lower esophageal sphincter to gastric distension, Am. J. Physiol., 259 (1990) G380-G385. [2(I] Furness, J.B. and Costa, M., The enteric nervous system, 1987, Churchill Livingstone, Edinburgh [21] Glaum, S.R., Brooks, P.A., Spyer, K.M. and Miller, R.J., 5-Hydroxytryptamine-3 receptors modulate synaptic activity in the rat nucleus tractus solitarius in vitro, Brain Res., 589 (1992) 62-68. [22] Goyal, R.K. and Paterson, W.G., Control of esophageal motility, In: Handbook of Physiology - The Gastrointestinal System I. Washington DC: American Physiological Society, 1989, pp. 865-908. [23] Grundy, D., Salih, A.A. and Scratcherd, T., Modulation of vagal efferent fibre discharge by mechanoreceptors in the stomach, duodenum and colon of the ferret, J. Physiol. (Lond,), 319 (1981) 43-52. [24] Larsson, I., Studies on the extrinsic neural control of serotonin release from the small intestine, Acta Physiol. Scand., (1981) Suppl. 499. [25] Leslie, R.A., Reynolds, D.J.M., Andrews, P.L.R., Grahame-Smith, D.G., Davis, C.J. and Harvey J.M., Evidence for presynaptic 5-hydroxytryptamine-3 recognition sites on vagal afferent terminals in the brainstem of the ferret, Neuroscience, 38 (1990) 667-673. [26] Merahi, N., Orer, H.8., Laporte, A-M., Gozlan, H., Ha-

[27]

[28]

[29]

[30]

[31]

[32]

103

mon, M. and Laguzzi, R., Baroreceptor reflex inhibition induced by the stimulation of serotonin-3 receptors in the nucleus tractus solitarius of the rat, Neuroscience, 46 (1992) 91-100. Nguyen, C.C., Rossiter, C.D., Benjamin, S.B. and Gillis, R,A., Blockade of gastric distention-induced lower esophageal sphincter (LES) relaxations by lesioning the caudal portion of the dorsal motor nucleus (DMV) of the vagus in cats, Gastroenterology, 100 (19911 A476 (Abstract). Plosker, G.L. and Goa, K.L., Granisetron: a review of its pharmacological properties and therapeutic use as an antiemetic, Drugs, 42 (19911 805-824. Richardson, B.P. and Buchheit, K-H., The pharmacology, distribution and function of 5-HT 3 receptors, In: Osborne, N.N., Hamon, M. (Eds.), Neuronal Serotonin, Wiley, Chichester, 1988, pp. 465-506. Schemann, M., 5-Hydroxytryptamine-rnediated responses in myenteric neurons of the guinea pig gastric corpus: effect of ICS 205-930 and cisapride. J. Gastrointest. Motil., 3 (1991) 255-262. Sengupta J.N., Kauvar D. and Goyal R.K., Characteristics of vagal esophageal tension sensitive afferent fibers in the opossum, J. Neurophysiol, 61 (1989) 1001-1010. Yoshioka, M., Effect of a novel 5-hydroxytryptamine antagonist, GR38032F, on the 5-hydroxytryptamine-induced increase in carotid sinus nerve activity in rats, J. Pharmacol. Exp. Ther., 250 (1989) 637-641.