| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROSCIENCE |
1 Department of Physiology and Pathophysiology, Friedrich-Alexander-University Erlangen-Nürnberg, Universitätsstrasse 17, D-91054 Erlangen, Germany
2 Department of Anatomy I, Friedrich-Alexander-University Erlangen-Nürnberg, Krankenhausstasse 9, 91054, Erlangen, Germany
3 Department of Otorhinolaryngology, University of Dresden Medical School, Fetscherstrasse 74, 01307, Dresden, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
5.7 mN). No difference in response characteristics of C and A
fibres was encountered. Mucosal proton stimulation (pH 5.4 for 3 min), mimicking gastro-oesophageal reflux disease (GORD), revealed in 31% of units a desensitizing response that peaked around 20 s and faded within 60 s. Cold stimulation (15°C) was proportionally encoded but the response showed slow adaptation. In contrast, the noxious heat (48°C) response showed no obvious adaptation with discharge rates reflecting the temperature's time course. Polymodal (69%) mucosal units, > 30% proton sensitive, were found in each fibre category and were considered nociceptors; they are tentatively attributed to vagal nerve endings type I, IV and V, previously morphologically described. All receptive fields were mapped and the distribution indicates that the posterior upper oesophagus may serve as a ‘cutbank’, detecting noxious matters, ingested or regurgitated, and triggering nocifensive reflexes such as bronchoconstriction in GORD.
(Received 3 March 2007;
accepted after revision 26 April 2007;
first published online 3 May 2007)
Corresponding author P. W. Reeh: Department of Physiology and Pathophysiology, Friedrich-Alexander-University Erlangen-Nürnberg, Universitätsstrasse 17, 91054 Erlangen, Germany. Email: reeh{at}physiologie1.uni-erlangen.de
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Behavioural studies showed that balloon distension with presumed noxious intensity resulted in greater pseudo-affective responses when the cervical as compared to the lower thoracic oesophagus was stimulated. These differences were even more pronounced with acid stimulation of the mucosa, and neurectomy of the SLN greatly reduced the response, as determined by neck muscle electromyography (Hummel et al. 2003). Other functional studies focused on sensors encoding proton stimuli (Kollarik & Undem, 2002; Page et al. 2002), mucosal mechanical sensitivity (Andrew, 1957; Page & Blackshaw, 1998), or both (Barthelemy et al. 1996; Sekizawa et al. 1999). Thus far, single-unit recordings in the uppermost oesophagus have not been performed with the focus on nociceptors.
From anatomical studies we know that the upper oesophagus receives a distinct dual vagal innervation (Wank & Neuhuber, 2001). While endings in the muscular layer represented by intraganglionic laminar endings (IGLEs) are evenly distributed along the organ and derive from axons travelling through the RLN and cervical vagus, the mucosa of the uppermost oesophagus receives a very dense network of axons passing through the SLN. Anterograde tracing studies showed abundant axons in the submucosa and mucosa with some fibres terminating in the epithelium (Neuhuber, 1987), and the combination of morphological and immunocytochemical techniques was used to distinguish five types of mucosal endings (Dutsch et al. 1998; Wank & Neuhuber, 2001). To date, functional characteristics of these mucosal endings are missing.
Considering the functional significance of the upper oesophagus for the integrity of the organism and the incomplete functional knowledge on its afferent innervation, we thought that a detailed electrophysiological investigation, taking anatomical pecularities into account, should be mandatory. In particular, we asked if there are vagal nociceptors in the upper oesophagus. The anatomical distribution, with mucosal afferents travelling through the SLN, offers the possibility of gaining selective insight into the spectrum of mucosal sensitivities. To address these questions, we developed novel ex vivo preparations of the isolated rat oesophagus with adjacent vagal nerve branches. The electrophysiological characterization of SLN and RLN single fibres aimed at attributing functional features to the five known morphological subtypes of mucosal endings (Dutsch et al. 1998; Wank & Neuhuber, 2001) and to compare the sensory response characteristics with those of cutaneous afferents (Reeh, 1988; Kress et al. 1992). Thereby, we were able to demonstrate that a significant proportion of mucosal SLN afferents in the upper oesophagus show properties similar to those reported from cutaneous nociceptors. The findings illustrate the spectrum of vagal sensitivities in the upper oesophagus and emphasize the role of vagal afferents in encoding potentially noxious mucosal stimuli.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Experiments were conducted using male Wistar rats (Charles River Laboratories, Sulzfeld, Germany), weighing 350–600 g. Animals were killed in a pure CO2 atmosphere. The European Communities Council Directive and animal welfare protocols approved by the local government were followed. All preparations were performed, where necessary, under a binocular microscope (Stemi 200 with KL1500-electronic cold light source, both from Zeiss, Jena, Germany).
Oesophagus preparations
Isolated SLN–oesophagus preparations. After a ventral midline cervical skin incision, sternocleidomastoid muscles were pulled aside and infrahyoid muscles were resected. The SLN beneath the carotid artery was tracked medially and the laryngeal branches were cut, leaving the branches to the oesophagus intact. The RLN including the communicating branch (Howell & Huber, 1891; Andrew, 1954) to the SLN was removed. The vagus nerve was transected at the cranial pole of the nodose ganglion to achieve a maximal length of 9–14 mm. Later, the cervical vagus was cut 1 cm distal to the nodose ganglion (handling site). After removal of the trachea the cervical oesophagus was opened in situ through either a ventral midline or paramedian incision and resected together with the caudal pharynx (total specimen length: 20–30 mm) and in continuity with the nerve. The oesophagus was placed in the specimen compartment (150 x 80 x 16 mm) of a custom made Perspex chamber (manufactured at the Department of Physiology and Pathophysiology, University of Erlangen-Nürnberg, Germany). In some SLN–oesophagus preparations, both SLNs were dissected and used for recording. In 55 of the SLN single-unit recordings the specimen was pinned with the adventitia facing down; these recordings were termed ‘inside-up’ (IU). In some experiments, the SLN recordings were performed using three modifications of the preparation: (1) after pinning the specimen down (IU) we removed the entire oesophageal mucosa from the underlying tissue (‘mucosa-off preparation’ n = 4); (2) after a single-fibre recording (IU) was completed, the mucosa at the receptive field site was removed (‘local mucosa-off preparation’ n = 8); (3) the SLN–oesophagus specimen was pinned with the mucosa down (n = 4), ‘inside-down’ (ID). The ‘mucosa-off’ preparation did not provide any recordings; the ‘local mucosa-off’ and ‘inside-down’ preparations could be used for further recordings.
Isolated RLN–oesophagus preparations. Using at first the same approach as for SLN, the RLN was tracked along the cervical trachea and oesophagus. The tracheal branches were exposed on the left side and cut leaving the RLN branches to the oesophagus intact. To receive maximal length, the left RLN was tracked around the aorta to its outlet from the cervical vagus. Approximately 16 mm cranial and caudal from this point the adjacent cervical vagus was cut. The caudal end was used to handle the nerves. The trachea was removed, the oesophagus was opened in situ (midline cut just anterior to the left RLN) and the excised RLN–oesophagus preparation (total specimen length 27–40 mm) was placed in the specimen chamber. In contrast to the SLN–oesophagus preparation, no additional mucosal preparations or preparations of the right RLN were performed.
Chamber
The specimen chamber was perfused with carbogen-saturated (14 ml min–1, 95% O2–5% CO2, Linde, Höllriegelskreuth, Germany) synthetic interstitial fluid (SIF) containing (mM): 108 NaCl, 3.48 KCl, 3.5 MgSO4, 26 NaHCO3, 11.7 NaH2PO4, 1.5 CaCl2, 9.6 sodium gluconate, 5.55 glucose and 7.6 sucrose (Bretag, 1969), resulting in a pH around 7.4. The fluid was heated in a thermostatically controlled heat exchanger (Haake D8, Karlsruhe, Germany) providing a chamber-entry temperature of 36°C.
Recording
The vagus nerve was threaded from the specimen chamber through a hole into a separate recording chamber (30 mm x 20 mm x 10 mm). The handling site was placed on a mirror plate (10 mm x 7 mm) and freshly cut to expose the nerves for recording, SLN or RLN, respectively. The vagus nerve was subsequently desheated. The free length of the SLN and RLN varied between 8 and 12 mm and 19 and 27 mm, respectively. A layer of paraffin oil (paraffinum liquidum, 110–230 mPas, DAB10, Bufa B.V., Uitgeest, Holland) was placed on top of the aqueous solution in the recording chamber. The mirror plate was adjusted slightly above the level of the interface of the two liquid phases. Fifteen to forty-five minutes were allowed after preparation before recording was started.
Fine filaments were teased out of the nerve and further subdivided until single unit activity could be recorded (Reeh, 1986, 1988). The gold-wire recording electrode (0.1 mm diameter) was advanced through the paraffin oil to the nerve filaments. A similar reference and a ground electrode were placed in the SIF in the recording and specimen chamber, respectively. For teasing nerve filaments we used custom-ground Dumont forceps No. 5 (Strauss, Nürnberg, Germany) and ophthalmosurgical microscissors (Riede, Emmingen-Liptingen, Germany).
Experimental protocol
Location and shape of the receptive fields on the mucosa were assessed manually by probing with different sized coloured blunt glass rods (tip diameter: 0.2–1 mm). Shape and (greatest) diameter of the receptive field were documented using a pair of compasses and a grid drawing (square: 0.5 mm). The aboral edge of the palatopharyngeal eminence (Nakano & Muto, 1985) provided a reference point to define the exact position using a coordinate system (y-axis: oral-aboral distance, x-axis: lateral distance) that was normalized to the circumference of the opened and pinned down specimen (Fig. 4). The mechanical sensitivity of the afferent unit in its receptive field was tested with hand-held calibrated von Frey hairs (manufactured at the Department of Physiology and Pathophysiology, University of Erlangen-Nürnberg, Germany). The parameters of the von Frey hairs are given in Fig. 1B. The von Frey threshold was defined as the probe with the lowest bending force that could repeatedly evoke activity from within the receptive field. Mechanical adaptation was checked by applying constant force of > 10x threshold value. We categorized fibres as slowly adapting if this constant mechanical stimulation produced a sustained response (Fig. 4). To evaluate the conduction velocity we used an electric stimulator (model DS7, Digitimer, Welwyn Garden City, UK), connected to a microneurography electrode (0.25 mm diameter) guided by a manipulator (Bachofer, Reutlingen, Germany), and positioned in the centre of the receptive field. The conduction velocity was calculated from time and distance between stimulating and recording electrode. To determine the action potential threshold, we used the stimulator in the constant-voltage mode. After increasing the voltage to twice threshold we determined the latency. To reassure that no hidden fibres were in the teased filament we temporarily increased the voltage 100-fold.
|
|
In order to apply the thermal or acidic stimuli in a controlled way to the receptive field, we used prewarmed hollow metal cylinders (stainless steel). The diameter fitting the entire receptive field was chosen and the cylinder placed over it. The height of the ring was minimally 1 mm greater than the SIF surface level. No leakage of stimulus solution out of the cylinder occurred due to the sealing properties of the mucosa. The metal cylinder was removed to rinse the SLN–oesophagus preparation between stimulations (heat, cold or acid). The interstimulus interval was 8–12 min with stimulus duration of 1–3 min. Acid sensitivity was tested only once per receptive field to avoid bias due to potential sensitization. After the thermal and chemical stimuli we rechecked the mechanical sensitivity of the single units.
Acid and thermal stimulation was applied after mechanical threshold and conduction velocity had been established. Thermal stimuli consisted of SIF superfusion at 50°C or 4°C. To assess, the true temperature in the receptive field we used a thermometer (Voltcraft VC303 Digital Multimeter, Taiwan) with a fine Type K temperature probe (Ni–Cr–Ni). The thermal inertia of the probe was 
= 0.53 s. The true specimen temperature reached by heat stimulation was 48°C (
= 5064 s), by cold stimulation only 15°C (
= 6.33 s) were reached (Fig. 4) because of the 36°C bath temperature surrounding the ring. We tested the acid sensitivity with 36°C phosphate-buffered SIF at pH 5.4 (NaHCO3 replaced by appropriate amounts of NaH2PO4 and Na2HPO4).
Amplification, recording and statistical analysis
The recording electrode signal was passed to a DAM 80 differential amplifier (World Precision Instruments, Sarasota, FL, USA). Amplification factor was 10 000 with the low frequency filter set to 300 Hz and high frequency filter to 1 kHz. To fine-adjust the signal amplitude we used an attenuator (0–100%). Discrimination of single units was achieved on the basis of shape, duration and amplitude of the action potentials during mechanical and electrical stimulation. To that end, the spikes were displayed and identified on a two-channel digital oscilloscope (Tektronix TDS210, Beaverton, OR, USA). For acoustic monitoring we used a Grass AM8 audio monitor (Grass Medical Instruments, Quincy, MA, USA). Bypassing the audio monitor, we digitized the signal using a DAP3000a-Card (Microstar Laboratory Inc., Bellevue, WA, USA) slot into a personal computer. The sampling rate of the card was 20 000 s–1. For data acquisition and subsequent off-line analysis, additional discrimination of the single units we used the Spike-Spidi program (Forster & Handwerker, 1990).
Statistical comparisons of means were made using Student's two-tailed t test. For all statistical tests P < 0.05 was used as the criterion for significance and mean values are given ± S.E.M. or ± S.D. when appropriate, as noted.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We characterized 55 single fibres in 23 SLN–oesophagus preparations isolated from 23 rats; 23 units were recorded from the left and 22 from the right SLN (Table 1). Shape and duration of the action potentials remained constant throughout the duration of each experiment while the amplitude showed a variable degree of run-down. Irregular spontaneous activity prior to any physical stimulation was encountered in 22 fibres (range: 0.013–6.3 spikes s–1, mean rate 1.1 ± 1.71 s–1 (± S.E.M.)).
|
-fibres (n
= 16) with a mean conduction velocity of 2.84 ± 2 m s–1 (±
S.D.). For the fibres conducting 1–1.9 m s–1 we abstained from deciding whether they were unmyelinated or thinly myelinated (n
= 14); in the subsequent figures these units are subsumed in a class of fibres conducting faster than 1 m s–1. A
-fibres were not encountered during these experiments. There were no differences in the distribution of response properties between A and C fibres (Table 1). Probing of the oesophagus at the mucosal side revealed 50 mechanoreceptive fields; the von Frey thresholds ranged from 1.4 to 128 mN. More than half of the units had thresholds beyond 5.7 mN (Fig. 1B) which, in the skin, would be considered ‘high mechanical thresholds’ (n = 27). Seven units with low mechanical thresholds (< 5.7 mN) did not respond to heat or cold stimuli, as well as six exclusively mechanosensitive fibres of high threshold; both groups were subsumed as a merely mechanosensitive subpopulation (M in Table 1 and Fig. 2). All responses to mechanical stimulation were slowly adapting and increased with the strength of von Frey stimulation (Fig. 3A). The distribution of von Frey thresholds was about the same in fibres conducting faster and slower than 1 m s–1 (Fig. 2). A regression analysis of von Frey threshold versus conduction velocity showed no significant correlation (r = 0.126).
|
|
Mechanically insensitive fibres (‘other’ in Table 1 and Fig. 2) were detected by spontaneous activity and/or by their response to thermal stimuli aimed at other previously characterized units. This small group of fibres (n = 5) showed spontaneous activity in three cases; four units responded to heat stimuli – one of them to acid as well – and one fibre to both heat and cold stimuli. MH fibres in the isolated rat skin–nerve preparation often show a peculiar ‘rewarming response’ after strong cold stimulation, a discharge activity that starts around 18°C and fades away around 24°C (P. W. Reeh, unpublished observation). Very similar responses were observed in (10 of 23) oesophageal heat-sensitive fibres. Vice versa, three (of 19) MC units showed some discharge activity during the cooling (recovery) phase following noxious heat stimulation.
The population response to acid stimulation (Fig. 3B) started around 10 s after buffer administration, reached a maximum within about 11 s (S.D.: 5 s) and then slowly faded away in further 50 s, although the acid exposure lasted for another 110 s. Indeed, all units showed desensitization (or inactivation) during sustained proton stimulation. In contrast, the averaged heat response (Fig. 3C) did not exhibit obvious adaptation; the mean discharge activity truly reflected the time course of the temperature change with a threshold around 40°C and a peak discharge coinciding with the peak temperature. With cold stimulation, the population response (Fig. 3D) also reached peak discharge close to the lowest temperature achieved, but then strong adaptation occurred, decreasing the mean discharge rate in spite of relatively constant low temperature (15°C). The threshold of the averaged cold response, and of many individual fibres, was just below the organ bath temperature of around 35°C.
Topographic distribution of the receptive fields recorded from the SLN–oesophagus preparation showed that the majority of fibres were located ipsilateral to the SLN (Fig. 4). In those SLN–oesophagus preparations that were opened along a paramedian line, we found 11 receptive fields to be located contralateral (Fig. 4). Multiple receptive fields extended over the posterior (often) or anterior midline (seldom), and three receptive fields were located completely on the opposite side (with regard to the recording side, Fig. 4). In addition, we found receptive fields beyond the pole of the palatopharyngeal eminence (n = 5). Four of these units recorded cranial to the palatopharyngeal eminence had conduction velocities that were faster than 2 m s–1. Three had a high mechanical threshold and two were acid sensitive. We found no cold-sensitive fibres above the palatopharyngeal eminence. However, most of the SLN units had receptive fields below the palatopharyngeal eminence (n = 50). Therefore, our findings suggest, that the SLN fibres do not only course up and down, but also to a minor extent circular around the oesophagus (red fibres in Supplemental Fig. 1).
The size and shape of the SLN receptive fields varied, with greatest (usually craniocaudal) diameters ranging from 0.3 to 3.5 mm. Based on the outlines of the receptive fields, we divided the SLN units into two groups, those with symmetric outlines and those with irregular shapes (n = 25). We also performed several modified SLN–oesophagus preparations (n = 16) in an attempt to further localize the receptive spot within the oesophageal wall (see Methods). In the IU–’mucosa-off preparation’ (n = 4) no activity in the SLN was found. In the preparations that were pinned mucosa down (ID, n = 4), no activity could be elicited by gentle adventitial probing, even though in one case irregular spontaneous activity was noted. The ID preparations were subsequently turned around and used for IU recordings. In addition to these more categorical approaches we modified a previously described elegant approach (Zagorodnyuk et al. 2003) and applied the ‘local mucosa-off’ dissection (n = 8). After complete characterization of an SLN unit in an IU preparation, the very focal removal of the mucosa abolished the responsiveness of this unit, indicating the location of the endings in the superficial tissue. In all eight units, the sensitivity vanished when the ‘local mucosa-off preparation’ was performed. In some cases, we verified that this approach was valid, by excising a random piece of mucosa located between the receptive field and the SLN. This did not abolish the responsivity of intact receptive fields. We were also able to further use these preparations for recordings, finding other units in the same oesophagus, but never within the area freed of mucosa.
Isolated RLN–oesophagus preparation
Only 14 RLN fibres were identified and this figure may reflect the smaller number of vagal afferents travelling from the oesophagus through the RLN. The units had conduction velocities from 0.4 to 9.6 m s–1. Eight fibres had CV below 1 m s–1 and were classified as C-fibres. Three fibres conducted between 1 and 2 m s–1, and three above 2 m s–1 and were therefore classified as A-fibres. A-fibres were not encountered during these experiments. In some of the RLN fibres, heat and cold stimulation by jet superfusion was applied, and von Frey thresholds of the four oralmost fibres were obtained (range: 22.6–90.5 mN). Notably, all RLN units had spot-like receptive fields with a maximal diameter of 1 mm (Fig. 4 and Supplemental Fig. 1). In consequence of our search stimulus, all recorded fibres were mechanosensitive and two fibres had irregular spontaneous discharge (0.08 s–1; 0.01 s–1). Two A (22.6 mN; 32 mN) and two C fibres (90.5 mN; 32 mN) were mechano-heat sensitive and therefore classified as polymodal nociceptors. The position of all receptive fields, stimulated by mucosal (n
= 11) or adventitial (n
= 3) probing, was transcribed into the normalized coordinate system. We found no receptive field located more cranial than 4.5 mm to the palatopharyngeal eminence (Fig. 4 and Supplemental Fig. 1).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The preparation, stimuli, and recording technique were largely adopted from the established rat skin–nerve preparation (Reeh, 1986, 1988). A comparison with data from this preparation therefore appears justified. Most sensory categories, as found in isolated rat skin, were identified: M, MH, MC and, rarely, H (= mechanically insensitive, heat-activated). The von Frey threshold distribution appears different between oesophagus and skin, although the overall range was the same. In skin, there is a clear gap between low- (< 2.8 mN) and high-threshold (> 5.7 mN) mechanoreceptive fibres (Kress et al. 1992), whereas the distribution is unimodal in the oesophagus with most frequently (82%) found values between 1.4 and 16 mN (Fig. 1B). This continuous threshold distribution appears typical for visceral mechanosensory afferents (Sengupta et al. 1990; Cervero, 1994; Page et al. 2002). The mechanically insensitive, low-threshold cold-activated C-C fibre type was also missing, because all cold-sensitive fibres were also mechanosensitive, and this MC fibre type was more often found than in the skin, in fact as frequently as the classical polymodal nociceptor (MH). These MC fibres were also more sensitive to cooling than in the rat skin, delivering most of their discharge at temperatures between 30°C and 15°C. This difference, however, may partly be due to the higher ambient temperature of the oesophageal versus cutaneous preparation (36°C versus 32°C) causing differential adaptation. Another slowly conducting cutaneous fibre type missing in the oesophagus was the rapidly adapting low-threshold mechanoreceptive A fibre, in the skin associated with down hairs (Kress et al. 1992). In fact, all mechanosensory fibres in the oesophagus appeared slowly adapting to punctate mechanical stimulation of sufficient strength, in contrast to the nearby trachea where vagal rapidly adapting low-threshold mechanoreceptive A fibres are found and suggested to be responsible for the mechanically evoked cough reflex (Sant'Ambrogio & Widdicombe, 2001; Widdicombe, 2001; Kollarik & Undem, 2002). However, afferent fibres from the oesophagus generally exhibit slow adaptation to distension (Andrew, 1956a,b; Clerc & Mei, 1983b; Satchell, 1984; Sekizawa et al. 1999; Dong et al. 2001). It is noteworthy that similar stimulus–response functions were found in ‘sympathetic afferents’ of the opossum oesophagus, considered as mechanosensitive nociceptors (Sengupta et al. 1990).
With respect to conduction velocities, we found fibres in the range of C and slowly conducting A fibres, but no A fibres. The absence of A fibres has been reported from primary afferents terminating in the oesophagus and stomach (Page & Blackshaw, 1998, 1999). This is in contrast to vagal afferents from the lungs, where rapidly adapting chemosensitive fibres with conduction velocities of up to 30 m s–1 have been reported (Ho et al. 2001). A phenotypic switch of somatic low-threshold A afferents with implications for inflammatory pain hypersensitivity has been described (Neumann et al. 1996); however, the available vagal studies, characterizing nociceptors in oesophageal and cardiac afferents, did not report units conducting in this velocity range (Yu et al. 2005; Hisata et al. 2006). While the formal definition of A fibres varies considerably (10–25 m s–1) among species (Ricco et al. 1996; Kajekar et al. 1999; Page & Blackshaw, 1999; Hisata et al. 2006), the appropriate cut-off in the rat is approximately 10 m s–1 (Koltzenburg et al. 1992). All SLN fibres were found below this limit and the fastest polymodal nociceptors in the rat oral mucosa are documented in the upper A range with 9.1 ± 2.0 m s–1 (Toda et al. 1997). The conduction speed in the A and C-fibre group reported here are similar to oesophageal afferents recorded in other species. In the isolated skin preparation, no C-fibre has been reported to conduct faster than 1.2 m s–1 (Kress et al. 1992), although at an organ bath temperature of 32°C in contrast to 36°C in the oesophagus preparation. The transition range in Fig. 1A (open columns) may, thus, include some fast conducting C-fibres as well as ultra-slow conducting A-fibres.
Another difference between oesophagus and skin innervation appears to be the proportion of fibres with spontaneous activity, as found in almost half of the oesophageal units. While this may be attributed, at least in low-threshold mechanosensitive fibres, to stretch exerted by pinning down the elastic preparation, spontaneous activity was also observed in high-threshold and mechanically insensitive units. Spontaneous activity is, in contrast to skin, a common observation in visceral preparations demonstrated in the stomach (Ozaki et al. 1999) or mediastinal pleura (Wedekind, 1997; Ozaki et al. 1999). Further than the aforementioned in vitro studies and studies on the oesophagus and proximal stomach (Partosoedarso & Blackshaw, 1997), vagal spontaneous activity has also been reported from in vivo (Seagard et al. 1999) and is traditionally described as typical of mucosal endings (Clerc & Mei, 1981). While in the majority of vagal muscular afferents a relatively high rate of spontaneous firing (5–40 s–1) was documented, the impulse rate reported here is in the range of the rhythmic firing during cardiorespiratory cycles (0.28–18 s–1), as noted in spinal afferents (Clerc & Mei, 1983a; Sengupta et al. 1990, 1992). In contrast, mucosal afferents recorded from the lower oesophagus showed similar impulse rates of spontaneous discharge as in our preparation (Falempin et al. 1978; Blackshaw et al. 2000).
Acid sensitivity in the upper oesophagus was found in a considerable proportion of units (31%) and among all sensory fibre types. This has also been reported from skin (Steen et al. 1992). However, in contrast to the skin, oesophageal afferents did not show sustained but transient discharge activity during the 3 min period of acid stimulation. This apparent desensitization may be due to the drastic proton concentration applied (pH 5.4) in order to mimic gastro-oesophageal reflux. Sustained discharge in cutaneous nociceptors is achieved at pH 6.1; although higher proton concentrations induced increasingly more overall discharge activity, up to a maximum at pH 5.2, the acid response was more and more shifted towards the onset of proton stimulation with a discharge profile reminiscent of adaptation or desensitization. In fact, the desensitization may reflect the increasing, time-dependent proton block of action potential generation that finally prevails at pH values below 5.2, decreasing the overall acid response (Steen et al. 1992, 1999). The voltage-gated sodium currents in sensory neurons, TTX-sensitive as well as TTX-insensitive ones, are reduced at pH 5.2 to one fifth of their original amplitude (M. Gautam & P. W. Reeh, unpublished observations). In cutaneous nerve fibres, this leads to a major and time-dependent decrease of conduction velocity and electrical excitability, both signs of imminent block (Sauer et al. 2005).
The desensitizing acid response of oesophageal afferents could be taken as an argument for involvement of the rapidly inactivating ASIC channels. Recently, it has been shown that proton-induced cationic currents pass through ASIC3 channels in vagal sensory ganglia of the rat (Sutherland et al. 2001; Naves & McCleskey, 2005). However, the inactivation time constant of ASIC3 channels in the rat is < 500 ms (Escoubas et al. 2000; Diochot et al. 2004), and therefore the time course of the acid response (Fig. 3B) makes the involvement of the slowly activating and inactivating capsaicin receptor channel TRPV1 in sustained proton-induced currents more probable, in particular at body temperature (Tominaga et al. 1998). Based on the time course alone, however, we cannot make inferences about the molecular transducers of proton stimulation. Mucosal data from the lower oesophagus suggest that 10% of mucosal afferents respond to intraluminal acid (Page & Blackshaw, 1998; Page et al. 2002). The significantly higher rate of acid-sensitive fibres (> 30%) in the SLN provides evidence for the function of this site, namely the encoding and triggering of reflexes necessary to initiate clearing mechanisms (Andrews & Sanger, 2002), constriction of the upper oesophageal sphincter as well as bronchoconstriction, typical responses described in the context of gastro-oesophageal reflux (Heatley et al. 1980; Harding, 2001).
Electrophysiological investigations classified oesophageal afferents (in the ferret) on the basis of their response pattern (to intraluminal distension) and termination site of their ending: muscular tension-sensitive, mucosal mechano/chemosensitive, and tension-sensitive mucosal sensors (Page & Blackshaw, 1998). While recent studies identified IGLEs as a transducer of mechanical stimuli in the muscle layer (Zagorodnyuk et al. 2003), an attribution for mucosal sensors is missing. The inferred term ‘mucosal sensor’ was conventionally defined as describing a spontaneously active, slowly adapting unit responsive to gentle stroking (i.e. mechanically low threshold; Clerc & Mei, 1983b). In general, mucosal afferents are insensitive to peristaltic contractions and exquisitely sensitive to light touch on the mucosal surface. Typically, mechanical probing produces a rapidly adapting response, but these afferents adapt slowly to chemical stimuli (Mei, 1983; Page et al. 2002). We were able to demonstrate individual mechanosensitive fibres that exhibited a slowly adapting response. We did not, however, test all mechanosensitive fibres for their mechanical adaptation and have to abstain from further speculations. We did test all heat-, cold- and chemosensitive fibres, and found sustained discharge in each of the three fibre groups (Fig. 3). The finding of many polymodal and some high-threshold mechanosensitve fibres is unusual in comparison to earlier studies that did not focus on nociception (Andrew, 1957; Clerc, 1984; Sengupta, 2000; Page et al. 2002). Most of the existing studies have been done on the vagus nerve or its branches, and the knowledge on mucosal afferents in spinal pathways is limited. Nevertheless, evidence for similar properties of mucosal sensory terminals of vagal and spinal pathways is growing (Berthoud et al. 2001; Grundy, 2002; Holzer, 2002).
Thermosensitivity is for two reasons interesting in the context of sensory vagal functions. First, the conscious sensation, e.g. from drinking cold water on an empty stomach, leads to a distinct emotional response and functional change in the upper gastrointestinal tract (Webber et al. 1980). Second, thermosensors are known to trigger autonomous reflex circuits from outside and inside the gastrointestinal tract (Fall et al. 1990; Villanova et al. 1997). Afferents sensitive to thermal stimuli have been identified projecting to the oesophagus and other regions in the gastrointestinal tract (El Ouazzani & Mei, 1982; Villanova et al. 1997). Originally, visceral thermosensors were reported to show no basal activity at body temperature and are almost all unmyelinated fibres (Mei, 1983). However, more recent reports on visceral organs (e.g. colon), using similar stimuli to those in the present study, revised this notion. We found similar proportions of thermal responses and polymodal fibres within the A and C fibre groups as previously described in the rat colon (Su & Gebhart, 1998). Furthermore, the same laboratory evaluated the mechanosensitive pelvic nerve (spinal) afferents for their coresponsiveness to thermo- and chemo-stimulation and found 92% of fibres to be polymodal; applying the same criteria, we found 84% in the oesophagus. While most thermoreceptive fibres showed higher resting discharge rates, similar thermal thresholds to cold (
30°C) and heat stimuli (
42°C) as well as sustained response functions were described (Su & Gebhart, 1998). The SLN mechano-cold sensors showed proportional encoding with highest mean discharge rate at lowest temperature (15°C). The receptor–channel proteins TRPM8 and TRPA1 are potential molecular transducers encoding these temperatures. In the mouse, however, it has been shown that many cold-sensitive neurons do not express these channel proteins (Babes et al. 2006; Munns et al. 2006). In contrast, TRPM8 was identified in urinary bladder afferents found in the S1 dorsal root ganglion in guinea pig, mediating the bladder-cooling reflex (Tsukimi et al. 2005). The noxious heat-sensitive SLN fibre population could be expected to show acid sensitivity. This assumption is based on the finding that TRPV1 shows heat and acid sensing capacities (Tominaga et al. 1998), and both sensitivities are highly coexpressed in MH fibres in isolated rat skin (Steen et al. 1992). As in the skin, however, acid sensitivity was also found among MC, mechanosensitive fibres as well as in two mechano-insensitive but heat-sensitive units. The association of heat and acid responsiveness is therefore by far not exclusive, and TRPV1 may not be the only acid transducer in vagal neurons projecting to the oesophageal mucosa (Kollarik et al. 2006). Acid sensitivity certainly depends on the permeability of the epithelium overlying the nerve endings. While the data on oesophageal epithelial paracellular permeability are inconclusive (Orlando et al. 1992), there is detrimental alteration in the expression of tight junction proteins in a rat model with chronic acid reflux oesophagitis (Asaoka et al. 2005), increasing the permeability of the oesophageal epithelium and thereby impairing the defence mechanism of this epithelium (Miwa et al. 2004).
Functional properties and morphological characteristics of mucosal endings type I–V
Morphological studies of the mucosal innervation in the oesophagus have been carried out for a while (Cecio & Califano, 1967; Rodrigo et al. 1975a,b, 1980) using anterograde neuronal tracing in the lower (Clerc & Condamin, 1987) and a combination of tracing and immunohistochemistry for calcium binding proteins in the upper oesophagus (Dutsch et al. 1998). Retrograde tracing experiments in this region showed that 80% of the petrosal and jugular neurons express calcitonin gene-related peptide (CGRP), a neuropeptide typically found in small diameter spinal nociceptors (Rosenfeld et al. 1983; Traub et al. 1990). While CGRP shows the same density in the mucosa throughout the length of the oesophagus, the vagal afferent marker calretinin (Dutsch et al. 1998) highlights a dense plexus in the uppermost part, stretching predominantly in the long axis of the oesophagus (red fibres in Supplemental Fig. 1). This network gives rise to four distinct terminal structures, types I–IV, in the upper third of the oesophagus (Supplemental Fig. 2), and SLN neurectomy almost abolished the ipsilateral plexus (Dutsch et al. 1998; Wank & Neuhuber, 2001). Furthermore, SLN neurectomy produced also a significant decrease in thin CGRP positive calretinin negative fibre endings; here type V (Supplemental Fig. 2). The other thin fibre ending, the calretinin positive varicose type IV that costains frequently for CGRP, branches subepithelial in a ‘fingerlike’ manner. We postulate that slowly conducting polymodal units are those with endings type I, IV or V (Supplemental Fig. 2). In the same oesophageal segment, 87.5% (21/24) mechanosensitive unmyelinated fibres (< 1 m s–1) were either high-threshold mechanosensitive or polymodal and thus nociceptive in phenotype. Further morphological findings on the oesophageal innervation are reviewed in the Supplemental Information (online) and illustrated in Supplemental Figs 1 and 2.
Behavioural experiments showed vigorous pseudoaffective reactions, measured by neck muscle electromyography (EMG), upon noxious stimulation of the upper oesophagus while responses were less pronounced when the aboral oesophagus was distended. Instillation of hydrochloric acid (0.1 N) elicited even more vigorous responses, again more pronounced with stimulation of the upper oesophagus. The contribution of vagal afferents was ascertained by transection of the cervical vagus and the SLN, which led to a gradual diminution of the EMG responses (Hummel et al. 2003). Our results, accordingly, suggest that the upper oesophageal mucosa is a favourite site (‘cutbank’) of mechano- and chemo-nociception. In contrast to the behavioural experiments, the electrophysiological study was not biased by the potential contribution of spinal afferents in encoding the noxious stimuli. Nevertheless, we are reluctant to assign the sensory-discriminative properties found in SLN afferents directly to a role in transmitting pain sensations. This reluctance is based on the absence of evidence on the pseudoaffective responses after selective spinal afferent denervation of the oesophagus. However, we concede that vagal afferents possibly contribute to the affective and emotional connotations (Webber et al. 1980) of painful sensation (e.g. nausea). This is evidenced by reports on patients with elective proximal gastric vagotomy that had fewer complaints of abdominal pain (Lindsetmo et al. 1998), and in reverse, patients with high spinal cord lesions that continued to sense oesophageal distension (DeVault et al. 1996). These perceptions and symptoms are, however, described as vague and vagrant, present in the term vagus nerve. In combination with the above-mentioned behavioural studies, we conclude that the SLN fibres characterized in this study (1) reveal the location of the sensory terminals to be mucosal, (2) are probably originating from distinct morphological types of mucosal sensors, and (3) are possibly important for the affective component of pain (Traub et al. 1996).
Functional and clinical implications
The proximal migration of gastric acid with a pH drop < 4, has been shown to elicit classic symptoms of gastro-oesophageal reflux disease (GORD) (Weusten et al. 1994, 1995; Cicala et al. 2003; Bredenoord et al. 2006), which are more likely to be evoked when the pH drop is large and if pure gas reflux (‘acid vapour’) is causing it (Bredenoord et al. 2006). Our data highlight the upper oesophagus as a possible trigger zone for the initiation of protective reflexes like bronchoconstriction and cough reflex as in the clinical setting of GORD (Stein, 2003). This assumption is supported by recent evidence, that tracheal and laryngeal polymodal -fibres, travelling in the SLN and RLN, regulate the cough reflex in guinea pigs (Canning et al. 2004). The authors point out, however, that C fibre-facilitated cough requires coactivation of other, faster conducting, afferents for the full reflex to be initiated. Given the innervation by the same nerves and the clinical association with GORD (Sifrim et al. 2005), our data provide evidence for the importance of vagal afferents in this context, reviewed by Kollarik et al. (2006). The nociceptive/nocifensive function may not be restricted to low pH as it has been shown that gastric and bile acids form a particularly noxious combination when they interact with the mucosa. A critical range is pH 3–6, in which bile acids exist in their soluble, non-ionized form, can penetrate cell membranes and accumulate within the mucosa (Kauer & Stein, 2005). Accordingly, all mucosal sensors reported from the distal oesophagus in ferret responded to diluted bile (Page et al. 2002).
In conclusion, the projections through different vagal branches (SLN/RLN) and distinct mucosal distribution of SLN afferents in the upper oesophagus were utilized in new isolated oesophagus–nerve preparations. Thereby, we were able to show that mucosal endings, conveyed by the SLN, have mechano-, thermo-, acid sensing and, in general, nociceptive properties. This spectrum of mucosal sensory capacities integrates well with neurochemical, morphological, and behavioural findings and highlights the upper oesophagus and its vagal innervation as critical for eliciting aversive and nocifensive responses. Nevertheless, all available evidence implies that the conscious discriminative sensation of pain is mediated via spinal pathways; a good reason not to equate vagal polymodal sensors with ‘pain sensors’, but to acknowledge their mediation of nocifensive reflexes and possible contribution to the affective component of pain (e.g. nausea) with the term vagal nociceptor.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Andrew BL (1956a). A functional analysis of the myelinated fibres of the superior laryngeal nerve of the rat. J Physiol 133, 420–432.
Andrew BL (1956b). The nervous control of the cervical oesophagus of the rat during swallowing. J Physiol 134, 729–740.
Andrew BL (1957). Activity in afferent nerve fibres from the cervical oesophagus. J Physiol 135, 54–55P.
Andrews PL & Sanger GJ (2002). Abdominal vagal afferent neurones: an important target for the treatment of gastrointestinal dysfunction. Curr Opin Pharmacol 2, 650–656.[CrossRef][Medline]
Asaoka D, Miwa H, Hirai S, Ohkawa A, Kurosawa A, Kawabe M, Hojo M, Nagahara A, Minoo T, Ohkura R, Ohkusa T & Sato N (2005). Altered localization and expression of tight-junction proteins in a rat model with chronic acid reflux esophagitis. J Gastroenterol 40, 781–790.[CrossRef][Medline]
Babes A, Zorzon D & Reid G (2006). A novel type of cold-sensitive neuron in rat dorsal root ganglia with rapid adaptation to cooling stimuli. Eur J Neurosci 24, 691–698.[CrossRef][Medline]
Barthelemy P, Sabeur G & Jammes Y (1996). Reflex relationships between the cervical esophagus and the respiratory system in cats. Neurosci Lett 217, 81–84.[Medline]
Berthoud HR, Blackshaw LA, Brookes SJ & Grundy D (2004). Neuroanatomy of extrinsic afferents supplying the gastrointestinal tract. Neurogastroenterol Motil 16 (Suppl. 1), 28–33.[CrossRef][Medline]
Berthoud HR, Lynn PA & Blackshaw LA (2001). Vagal and spinal mechanosensors in the rat stomach and colon have multiple receptive fields. Am J Physiol Regul Integr Comp Physiol 280, R1371–R1381.
Berthoud HR & Neuhuber WL (2000). Functional and chemical anatomy of the afferent vagal system. Auton Neurosci 85, 1–17.[CrossRef][Medline]
Blackshaw LA, Page AJ & Partosoedarso ER (2000). Acute effects of capsaicin on gastrointestinal vagal afferents. Neuroscience 96, 407–416.[CrossRef][Medline]
Bredenoord AJ, Weusten BL, Curvers WL, Timmer R & Smout AJ (2006). Determinants of perception of heartburn and regurgitation. Gut 55, 313–318.
Bretag AH (1969). Synthetic interstitial fluid for isolated mammalian tissue. Life Sci 8, 319–329.[Medline]
Canning BJ, Mazzone SB, Meeker SN, Mori N, Reynolds SM & Undem BJ (2004). Identification of the tracheal and laryngeal afferent neurones mediating cough in anaesthetized guinea-pigs. J Physiol 557, 543–558.
Cecio A & Califano G (1967). Neurohistological observations on the oesophageal innervation of rabbit. Z Zellforsch Mikrosk Anat 83, 30–39.[CrossRef][Medline]
Cervero F (1994). Sensory innervation of the viscera: peripheral basis of visceral pain. Physiol Rev 74, 95–138.
Chahal PS & Rao SS (2005). Functional chest pain: nociception and visceral hyperalgesia. J Clin Gastroenterol 39, S204–S209.[CrossRef][Medline]
Cicala M, Emerenziani S, Caviglia R, Guarino MP, Vavassori P, Ribolsi M, Carotti S, Petitti T & Pallone F (2003). Intra-oesophageal distribution and perception of acid reflux in patients with non-erosive gastro-oesophageal reflux disease. Aliment Pharmacol Ther 18, 605–613.[CrossRef][Medline]
Clerc N (1984). Afferent innervation of the lower esophageal sphincter of the cat. Pathways and functional characteristics. J Auton Nerv Syst 10, 213–216.[CrossRef][Medline]
Clerc N & Condamin M (1987). Selective labeling of vagal sensory nerve fibers in the lower esophageal sphincter with anterogradely transported WGA-HRP. Brain Res 424, 216–224.[CrossRef][Medline]
Clerc N & Mei N (1981). Demonstration of mucous mechanoreceptors in the lower esophageal sphincter. Comparison with muscle mechanoreceptors. C R Seances Soc Biol Fil 175, 352–356.[Medline]
Clerc N & Mei N (1983a). Thoracic esophageal mechanoreceptors connected with fibers following sympathetic pathways. Brain Res Bull 10, 1–7.[CrossRef][Medline]
Clerc N & Mei N (1983b). Vagal mechanoreceptors located in the lower oesophageal sphincter of the cat. J Physiol 336, 487–498.
Cunningham ET Jr, Ravich WJ, Jones B & Donner MW (1992). Vagal reflexes referred from the upper aerodigestive tract: an infrequently recognized cause of common cardiorespiratory responses. Ann Intern Med 116, 575–582.[Medline]
DeVault KR, Beacham S, Castell DO, Streletz LJ & Ditunno JF (1996). Esophageal sensation in spinal cord-injured patients: balloon distension and cerebral evoked potential recording. Am J Physiol Gastrointest Liver Physiol 271, G937–G941.
Diochot S, Baron A, Rash LD, Deval E, Escoubas P, Scarzello S, Salinas M & Lazdunski M (2004). A new sea anemone peptide, APETx2, inhibits ASIC3, a major acid-sensitive channel in sensory neurons. EMBO J 23, 1516–1525.[CrossRef][Medline]
Dong H, Loomis CW & Bieger D (2001). Vagal afferent input determines the volume dependence of rat esophageal motility patterns. Am J Physiol Gastrointest Liver Physiol 281, G44–G53.
Dutsch M, Eichhorn U, Worl J, Wank M, Berthoud HR & Neuhuber WL (1998). Vagal and spinal afferent innervation of the rat esophagus: a combined retrograde tracing and immunocytochemical study with special emphasis on calcium-binding proteins. J Comp Neurol 398, 289–307.[CrossRef][Medline]
El Ouazzani T & Mei N (1982). Electrophysiologic properties and role of the vagal thermoreceptors of lower esophagus and stomach of cat. Gastroenterology 83, 995–1001.[Medline]
Escoubas P, De Weille JR, Lecoq A, Diochot S, Waldmann R, Champigny G, Moinier D, Menez A & Lazdunski M (2000). Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels. J Biol Chem 275, 25116–25121.
Euchner-Wamser I, Sengupta JN, Gebhart GF & Meller ST (1993). Characterization of responses of T2–T4 spinal cord neurons to esophageal distension in the rat. J Neurophysiol 69, 868–883.
Falempin M, Mei N & Rousseau JP (1978). Vagal mechanoreceptors of the inferior thoracic oesophagus, the lower oesophageal sphincter and the stomach in the sheep. Pflugers Arch 373, 25–30.[CrossRef][Medline]
Fall M, Lindstrom S & Mazieres L (1990). A bladder-to-bladder cooling reflex in the cat. J Physiol 427, 281–300.
Forster C & Handwerker HO (1990). Automatic classification and analysis of microneurographic spike data using a PC/AT. J Neurosci Meth 31, 109–118.[CrossRef][Medline]
Furusawa K, Yasuda K, Okuda D, Tanaka M & Yamaoka M (1996). Central distribution and peripheral functional properties of afferent and efferent components of the superior laryngeal nerve: morphological and electrophysiological studies in the rat. J Comp Neurol 375, 147–156.[CrossRef][Medline]
Gebhart GF (2000a). Pathobiology of visceral pain: molecular mechanisms and therapeutic implications IV. Visceral afferent contributions to the pathobiology of visceral pain. Am J Physiol Gastrointest Liver Physiol 278, G834–G838.
Gebhart GF (2000b). Visceral pain-peripheral sensitisation. Gut 47 (Suppl. 4), iv54–55; discussion iv58.
Grundy D (2002). Neuroanatomy of visceral nociception: vagal and splanchnic afferent. Gut 51 (Suppl. 1), i2–5.
Harding SM (2001). Gastroesophageal reflux, asthma, and mechanisms of interaction. Am J Med 111 (Suppl. 8A), 8S–12S.[CrossRef][Medline]
Heatley RV, Collins RJ, James PD & Atkinson M (1980). Vagal function in relation to gastro-oesophageal reflux and associated motility changes. Br Med J 280, 755–757.[Medline]
Hisata Y, Zeredo JL, Eishi K & Toda K (2006). Cardiac nociceptors innervated by vagal afferents in rats. Auton Neurosci 126–127, 174–178.
Ho CY, Gu Q, Lin YS & Lee LY (2001). Sensitivity of vagal afferent endings to chemical irritants in the rat lung. Respir Physiol 127, 113–124.[CrossRef][Medline]
Holzer P (2002). Sensory neurone responses to mucosal noxae in the upper gut: relevance to mucosal integrity and gastrointestinal pain. Neurogastroenterol Motil 14, 459–475.[CrossRef][Medline]
Howell WH & Huber GC (1891). Physiology of the communicating branch between the superior and the inferior laryngeal nerves. J Physiol 12, 5–11.
Hummel T, Barz S, Holscher T & Neuhuber WL (2003). Differences in responses to nociceptive stimulation of the oral and aboral oesophagus. J Clin Neurosci 10, 223–225.[CrossRef][Medline]
Ishikawa T, Sekizawa SI, Sant'Ambrogio FB & Sant'Ambrogio G (1999). Larynx vs. esophagus as reflexogenic sites for acid-induced bronchoconstriction in dogs. J Appl Physiol 86, 1226–1230.
Kajekar R, Proud D, Myers AC, Meeker SN & Undem BJ (1999). Characterization of vagal afferent subtypes stimulated by bradykinin in guinea pig trachea. J Pharmacol Exp Ther 289, 682–687.
Kauer WK & Stein HJ (2005). Bile reflux in the constellation of gastroesophageal reflux disease. Thorac Surg Clin 15, 335–340.[CrossRef][Medline]
Kenshalo DR & Duclaux R (1977). Response characteristics of cutaneous cold receptors in the monkey. J Neurophysiol 40, 319–332.
Kollarik M, Ru F & Undem BJ (2006). Acid-sensitive vagal sensory pathways and cough. Pulm Pharmacol Ther 20, 402–411.[Medline]
Kollarik M & Undem BJ (2002). Mechanisms of acid-induced activation of airway afferent nerve fibres in guinea-pig. J Physiol 543, 591–600.
Koltzenburg M, Kress M & Reeh PW (1992). The nociceptor sensitization by bradykinin does not depend on sympathetic neurons. Neuroscience 46, 465–473.[CrossRef][Medline]
Kress M, Koltzenburg M, Reeh PW & Handwerker HO (1992). Responsiveness and functional attributes of electrically localized terminals of cutaneous C-fibers in vivo and in vitro. J Neurophysiol 68, 581–595.
Lang IM, Medda BK, Ren J & Shaker R (1998). Characterization and mechanisms of the pharyngoesophageal inhibitory reflex. Am J Physiol Gastrointest Liver Physiol 275, G1127–G1136.
Lindsetmo RO, Johnsen R & Revhaug A (1998). Abdominal and dyspeptic symptoms in patients with peptic ulcer treated medically or surgically. Br J Surg 85, 845–849.[CrossRef][Medline]
Loomis CW, Yao D & Bieger D (1997). Characterization of an esophagocardiovascular reflex in the rat. Am J Physiol Regul Integr Comp Physiol 272, R1783–R1791.
Mei N (1983). Sensory Structures of the Viscera, vol. 4. Springer, Berlin.
Miwa H, Asaoka D, Hojo M, Iijima K & Sato N (2004). GERD and tight junction proteins of the esophageal mucosa. Nippon Rinsho 62, 1441–1446.[Medline]
Munns C, Alqatari M & Koltzenburg M (2006). Many cold sensitive peripheral neurons of the mouse do not express TRPM8 or TRPA1. Cell Calcium 41, 331–342.[CrossRef][Medline]
Nakano T & Muto H (1985). Anatomical observations in the pharynx of the mouse with special reference to the nasopharyngeal hiatus (Wood Jones). Acta Anat (Basel) 121, 147–152.[Medline]
Naves LA & McCleskey EW (2005). An acid-sensing ion channel that detects ischemic pain. Braz J Med Biol Res 38, 1561–1569.[Medline]
Ness TJ & Gebhart GF (1990). Visceral pain: a review of experimental studies. Pain 41, 167–234.[CrossRef][Medline]
Neuhuber WL (1987). Sensory vagal innervation of the rat esophagus and cardia: a light and electron microscopic anterograde tracing study. J Auton Nerv Syst 20, 243–255.[CrossRef][Medline]
Neumann S, Doubell TP, Leslie T & Woolf CJ (1996). Inflammatory pain hypersensitivity mediated by phenotypic switch in myelinated primary sensory neurons. Nature 384, 360–364.[CrossRef][Medline]
Orlando RC (2004). Esophageal perception and noncardiac chest pain. Gastroenterol Clin North Am 33, 25–33.[Medline]
Orlando RC, Lacy ER, Tobey NA & Cowart K (1992). Barriers to paracellular permeability in rabbit esophageal epithelium. Gastroenterology 102, 910–923.[Medline]
Ozaki N, Sengupta JN & Gebhart GF (1999). Mechanosensitive properties of gastric vagal afferent fibers in the rat. J Neurophysiol 82, 2210–2220.
= M, 
= MH, 
= MC, 
= other). One symbol of each category and conduction velocity group (
) was found in the uppermost part of the RLN–oesophagus preparation and a broken-line circle shows its receptive field. The numbers inside the diagram represent conduction velocity in m s–1 and punctate mechanical threshold in mN.
