SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

The neuropeptide galanin is found in the central and peripheral nervous systems. It may have excitatory or inhibitory actions via three subtypes of G-protein-coupled receptor, and it modulates the mechanosensitivity of somatic sensory fibres. We aimed to determine if galanin also modulates vagal afferent mechanosensitivity, and to localize endogenous sources. The responses of ferret and mouse gastro-oesophageal vagal afferents to graded mechanical stimuli were investigated in vitro. The effects of galanin and/or the galanin receptor antagonist galantide on these responses were quantified. Immunohistochemistry for galanin was performed in ferret and mouse proximal stomach and nodose ganglion. In ferrets, retrograde labelling of gastric afferents to the nodose ganglion was combined with immunohistochemistry. When exposed to galanin (1–10 nm), 18/31 ferret and 12/15 mouse gastro-oesophageal afferents (tension, mucosal and tension/mucosal receptors) showed inhibition of mechanosensitivity. Four of 31 ferret afferents showed potentiation of mechanosensitivity, and 9/31 were unaffected (2/15 and 1/15 in mouse, respectively). Galanin effects were reversed after washout or by galantide (10–30 nm). Galantide given alone increased mechanosensitivity. Galanin immunoreactivity was found in nodose neurones, including those innervating the stomach in ferret. Enteric neurones were also galanin immunoreactive, as were endings associated with myenteric ganglia and smooth muscle. We conclude that galanin potently modulates mechanosensitivity of gastro-oesophageal vagal afferents with either facilitatory or inhibitory actions on individual afferent fibres. Both intrinsic and extrinsic (vagal) neurones contain galanin and are therefore potential sources of endogenous galanin.

Therapies that reduce afferent signalling of mechanical stimuli from the upper gut to the central nervous system may prove to be effective in a number of gastrointestinal (GI) diseases. These include functional dyspepsia and gastro-oesophageal reflux disease (GORD), the rationale being that functional dyspepsia is characterized by increased perception of nonpainful stimuli, including gastric distension and contraction (Tack et al. 2004), and in GORD that gastric distension leads to triggering of transient lower oesophageal sphincter (LOS) relaxations, and thence GOR (Mittal et al. 1995; Blackshaw, 2001). We have developed methods to evaluate the modulatory effects of endogenous and exogenous compounds on the mechanosensitivity of extrinsic GI primary afferents. So far we have shown that metabotropic receptors to the amino acids γ-amino butyric acid (GABA) and glutamate are coupled to potent inhibition of vagal afferent mechanosensitivity (Page & Blackshaw, 1999; Page et al. 2005), and that these effects may be accompanied by inhibition of transient LOS relaxations and gastro-oesophageal reflux (Blackshaw, 2001; Zhang et al. 2002). Another group has demonstrated that κ-opioid receptors may also inhibit vagal afferent fibres (Ozaki et al. 2000), suggesting that peptidergic receptors may also represent targets on these neurones.

Galanin is a 29–30 amino acid peptide found throughout the central and enteric nervous systems. It may have either excitatory or inhibitory effects on motor function and neural excitability, depending on which of the three galanin receptors it binds to (Branchek et al. 2000). For example, it may have both pro- and antinociceptive actions in the spinal cord (Wynick et al. 2001), and may contract or relax gastrointestinal smooth muscle (see Liu et al. 2003). It has recently been shown to affect signalling in central gastric vagal pathways, with both inhibitory and excitatory effects being demonstrated (Yuan et al. 2002; Tan et al. 2004). Its direct effects on intrinsic neurones of the enteric nervous system are specific to sensory (AH) neurones (Liu et al. 2003). We considered the possibility that galanin may also have peripheral actions on extrinsic sensory neurones which may be of importance in modulation of visceral sensation and reflex control of GI function. The aims of this study were therefore to determine the effects of galanin on mechanosensitivity of vagal afferents in two species (rodent and nonrodent), and the distribution of the peptide peripherally and in vagal sensory ganglia.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

All studies were performed in accordance with the guidelines of the Animal Ethics Committees of the Royal Adelaide Hospital, and the Institute for Medical and Veterinary Science, Adelaide, and also the Animal Ethics Committee of the University of Adelaide.

In vitro mouse gastro-oesophageal afferent preparation

This preparation is described in detail elsewhere (Page et al. 2002). Briefly, female mice (C57/BL6, n= 40, 20–30 g body weight) were killed by CO2 inhalation and cervical dislocation. The stomach and oesophagus with attached vagal nerves, heart and lungs were removed, and placed in modified Krebs solution of the following composition (mm): NaCl, 118.1; KCl, 4.7; NaHCO3, 25.1; NaH2PO4, 1.3; MgSO4.7H2O, 1.2; CaCl2, 2.5; citric acid, 1.0; glucose, 11.1; bubbled with 95% O2 and 5% CO2. Dissection was performed at 4°C to prevent metabolic degradation. After clearing of adjacent tissue, the preparation was opened longitudinally along the oesophagus and greater curve of the stomach. One side of the stomach was removed completely to enable the tissue to be pinned flat in the organ bath. The tissue was then pinned, mucosa side up, in a perspex chamber (Danz Instruments, Adelaide, Australia), and perfused at a rate of 11–12 ml min−1 with Krebs bicarbonate buffer solution maintained at 34°C. A sliding wall with a small ‘mouse hole’ for the vagus nerves to pass through was moved into position so that the nerves extended into a second round chamber where they were laid on a mirror and bathed in paraffin oil. Under a dissecting microscope, the nerve sheath was gently peeled back to expose the nerve trunk. Using fine forceps, nerve fibres were teased apart into 8–12 bundles. One by one, the small nerve bundles were placed onto a platinum recording electrode. A reference electrode rested on the mirror in a small pool of Krebs solution.

Ferret gastro-oesophageal afferent preparation

Ferrets (0.5–1.0 kg body weight) were deeply anaesthetized with sodium pentobarbitone (60 mg kg−1, i.p.), and the thorax was opened by a midline incision. The ferret was killed by exsanguination under anaesthesia. The stomach and oesophagus were then placed in an organ bath in a similar manner to the mouse preparation. This preparation has also been described in detail previously (Page & Blackshaw, 1998, 1999).

Characterization of gastro-oesophageal vagal afferent properties

In the ferret, three distinct types of afferent were recorded: those responding to circular tension but not to low intensity mucosal stimuli (tension receptors), those responding only to mucosal stoking (mucosal receptors), and those responding to both mucosal stroking and circular tension (tension/mucosal receptors; Page & Blackshaw, 1998, 1999). In the mouse, two types of mechanosensitive afferent were studied: those responding to mucosal stroking but not circular tension (mucosal receptors), and those responding to mucosal stroking and circular tension (tension receptors), as reported previously (Page et al. 2002).

Location of receptive fields of all types of vagal afferent fibre was determined by mechanical stimulation with a brush throughout the preparation, then more accurately with a blunt glass rod. Accurate quantification of mechanical responses was performed differently according to the primary adequate stimulus for the type of fibre. Mechanical thresholds of all types of fibre were determined using calibrated von Frey hairs. Mucosal receptors showed rapidly adapting responses to maintained pressure on the receptive field with a von Frey hair. This type of responsiveness was also seen in responses of tension/mucosal (TM) receptors to low level mucosal mechanical stimulation. The most reproducible, stimulus-dependent responses of these afferents to mucosal stimuli were evoked when the probe was moved at a rate of approximately 5 mm s−1 across the receptive field. Because receptive fields are small (1–3 mm2), precision in probing their centre is difficult to achieve manually. Therefore to compensate for this, the mean response was measured to the middle 8 of 10 standard strokes given at 1 s intervals. The von Frey hair (10–1000 mg) was bent throughout the stroking stimulus to provide an even force. This protocol was found to give highly reproducible data, and was therefore used to assess effects of galanin on vagal afferents. Tension–response curves were also obtained for all afferent fibres, which were used in combination with von Frey thresholds to determine whether the receptive fields of fibres were located in the mucosa, the muscle layer or both. Tension stimuli were applied via a thread attached to the edge of the tissue nearest to the mechanoreceptive field. The thread was attached to a cantilever via a pulley close to the preparation. Reference standard weights were then placed on the opposite end of the cantilever for 1 min each, and the response was measured as the mean discharge evoked. The tension–response curves were produced by randomly applying weights to the cantilever system in the range of 0.5–7 g on the ferret and 0.5–5 g on the murine tissue. A recovery period of at least 1 min was allowed between each tension stimulus.

Effect of galanin on mechanical sensitivity of vagal afferents

After mechanical sensitivity of the gastro-oesophageal vagal afferents had been established, the effects of galanin on mechanical sensitivity were determined. The lowest dose of galanin (0.1–1 nm) was added to the superfusing Krebs solution, and was allowed to equilibrate for 20 min, after which the tension–response and stroke–response curves were redetermined. This equilibration period was observed to ensure total penetration of the drug into all the layers of the tissue. This procedure was repeated for galanin at increasingly higher doses (3–10 nm). Time control experiments (n= 6) were performed in which there was no significant change in the mechanical responses over a comparable duration – a maximum of 20% change from first trial. A change in response following drug treatment was scored only if there was a >50% increase or decrease. In a separate series of experiments in mice, the galanin antagonist galantide (3–30 nm) was used with the aim of reversing the effects of galanin (1–10 nm). After mechanical sensitivity of the vagal afferent fibre had been established, galanin (1 or 10 nm) was added to the superfusing Krebs solution, and allowed to equilibrate for a period of 20 min. Mechanical response curves were then reestablished. Galantide (3 or 30 nm) was then added to the superfusing Krebs solution along with the galanin, and this was again allowed to equilibrate for 20 min before mechanical response relationships were redetermined. In some experiments on ferret tissue, galantide was given alone without galanin pretreatment. Four experiments on tension and TM receptors were performed in the presence of nifedipine (1.0 μm) in order to reduce the occurrence of effects of galanin secondary to smooth muscle responses or extracellular Ca2+-dependent mediator release. Further controls were performed in gastric smooth muscle strip experiments, which confirmed that galanin at the concentrations used did not cause contractile responses (data not shown).

Immunohistochemistry

Ferret left nodose ganglia were obtained from animals in which gastric neurones had been retrogradely traced with 0.5% cholera toxin subunit B (CTB) conjugated to fluorescein isothiocyanate (FITC) (Smid et al. 2001). Anaesthesia was induced with 5% halothane, which was reduced to 2–3% during surgery sufficient to abolish the hindlimb pinch–withdrawal reflex. CTB-FITC (50 μl) was injected into the proximal stomach serosal layer in 5 μl aliquots spaced circumferentially around, and 1.5 cm from the LOS. Post-operative analgesia and anti-inflammatory treatment was provided (Meloxicam 0.3 mg kg−1, s.c.) and long-acting antibiotic treatment was administered (oxytetracycline 20 mg kg−1, s.c.). Animals were allowed to recover under close observation for 4 days, after which time they were anaesthetized with sodium pentobarbitone (60 mg kg−1, i.p.). They were perfused transcardially with 300 ml of warm heparinized saline, followed by 1 l of 10% neutral buffered formalin (NBF) or 4% paraformaldehyde (PFA). The ferret left nodose was then removed and postfixed overnight at 4°C or 4 h at room temperature (RT) in NBF or PFA, after which the tissue was cryoprotected with 30% sucrose for a minimum of 18 h. The nodose was then frozen, and 20 μm transverse sections cut. Ferret stomach was obtained from animals without tracer application; however, the tissue processing was the same as for the ferret nodose. Cross-sections of 20 μm were cut through the proximal stomach. Mouse nodose ganglia were treated similarly to those from ferret, except 10 μm transverse sections were cut. The immunohistochemistry protocol was the same for mouse nodose, ferret nodose and ferret stomach. Sections were air dried at RT for 10 min, after which the tissue was washed with phosphate-buffered saline (pH 7.4)/0.1% Triton X-100 (PBST). Tissue was then blocked with 10% bovine serum albumin (BSA) in PBST for 30 min at RT, and subsequently incubated with rabbit anti-galanin (1/1500; Serotec, UK) in 1% BSA in PBST overnight at 4°C. Tissue was then washed three times with PBST, then incubated with goat anti-rabbit Alexa Fluor 546 (10 μg ml−1; Molecular Probes, USA) for 45 min at RT. The tissue was then washed three times with PBST and mounted with ProLong Antifade (Molecular Probes). Slides were allowed to dry overnight before viewing with an Olympus BX51 epifluorescence microscope, and imaging with a Photometrics CoolSnap fx camera. Negative controls were prepared as above except that the rabbit primary antibody was omitted; no obvious fluorescence was seen in these specimens.

Mouse stomach was dissected and pinned out flat into a Sylgard-coated dish in cold PBS. The mucosal layer was then stripped away from the underlying muscle tissue, which was subsequently fixed in 10% NBF for 2 h at RT. The stomach was then cut into smaller pieces, and all incubations were carried out on tissue free floating in solution. The tissue was washed three times with PBST before being blocked with 10% BSA in PBST for 2 h at RT. It was then incubated overnight at RT in rabbit anti-galanin (1/1500). After the overnight incubation, the tissue was washed three times with PBST before incubation with goat anti-rabbit Alexa Fluor 546 (10 μg ml−1) for 2 h at RT. The tissue was then washed three times with PBST, and mounted with ProLong Antifade. Slides were allowed to dry overnight before microscopy and imaging.

Data recording and analysis

Afferent impulses were amplified with a biological amplifier (DAM 50; World Precision Instruments, Sarasota, FL, USA), filtered (band-pass filter-932; CWE, Inc., Ardmore, PA, USA) and monitored using an oscilloscope (DL 1200 A; Yokogawa, Tokyo). Single units were discriminated on the basis of action potential shape, duration and amplitude using Spike 2 software (Cambridge Electronic Design, Cambridge, UK). All data were recorded and analysed off-line using a personal computer (IBM Thinkpad). Peristimulus time histograms and discharge traces were displayed using Spike 2 software. Data are expressed as means ±s.e.m., with n equal to the number of individual afferents in all instances. The pharmacological protocol was performed on a maximum of one afferent fibre in each preparation. Differences between stimulus–response curves were evaluated using two-way ANOVA. Differences were considered significant if P < 0.05. A fibre was considered as being affected by galanin if there was a >50% change in mechanical response to a submaximal stimulus at the maximal concentration of galanin (5 g tension for tension and TM receptors in ferret, 3 g tension in mouse; 10 mg von Frey hair for all mucosal receptors).

Drugs

Stock solutions of all drugs were kept frozen and diluted to their final concentration in Krebs solution on the day of the experiment. Galanin and galantide were both obtained from AUSPEP Pty Ltd (Victoria, Australia).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

Electrophysiology

Effect of galanin on the mechanosensitivity of gastro-oesophageal vagal afferents The effect of galanin on ferret mucosal, tension and TM receptor sensitivity to mechanical stimulation is illustrated in Figs 1, 2 and 3. Galanin (10 nm) significantly reduced the response of mucosal receptors to mucosal stroking with calibrated von Frey hairs (P < 0.05, Fig. 3A; see also supplementary material for proportions of all fibres affected). Galanin (1–3 nm) significantly reduced the responses of 7/8 tension receptors to circumferential tension (P < 0.001, Fig. 3Bi). The one tension receptor not inhibited by galanin is illustrated in Fig. 3Bii. The response of this fibre was substantially and dose-dependently potentiated by galanin (0.1–3 nm). The responses of 10/15 TM receptors to circumferential tension were reduced in the presence of galanin (3–10 nm; Fig. 3Ci), e.g. Fig. 1A. The responses of the other five TM receptors to circumferential tension were potentiated by galanin (1–10 nm; Fig. 3Cii), e.g. Fig. 1B. The responses to mucosal stroking of TM receptors were not always affected in the same way as their response to tension. The fibres whose tension responses were inhibited by galanin are illustrated again in Fig. 3Ciii. Galanin also significantly reduced the responses of these afferents to mucosal stroking (10–1000 mg). In contrast, the fibres whose tension responses were potentiated by galanin (Fig. 3Cii) did not show any change in their response to mucosal stroking (Fig. 3Civ).

image

Figure 1. Galanin effects on ferret tension/mucosal (TM) receptors A, an example in which galanin (3 nm) reduces the response of a unit to mechanical stimuli. The responses to mucosal stroking (50 mg von Frey hair is shown) are shown in the left panels, and the responses to circumferential tension (1, 2 and 3 g are shown) in the right panels. B, an example in which galanin (3 nm) potentiates the response to mechanical stimulation. The responses were obtained in the absence of galanin (i) and in the presence of galanin (ii). The responses to mucosal stroking are expressed as instantaneous frequency (top trace) and raw trace recordings (bottom trace). The responses to circumferential tension are expressed as instantaneous frequency (top trace) and a bin count of action potentials recorded every 5 s (bottom trace). The horizontal bars below the traces indicate when the mechanical stimuli were applied.

Download figure to PowerPoint

image

Figure 2. Galanin affected mechanosensitivity differently in the population of vagal afferent fibres studied in the ferret Lines show the concentration-dependent effect of galanin on individual fibres responding to circumferential tension at a submaximal level (5 g tension). The range of variability in time controls is shown between dashed lines. Twelve fibres had decreased mechanosensitivity in the presence of galanin (1–10 nm), 4 fibres had increased mechanosensitivity, and 7 fibres were unaffected (<50% change in response at maximal concentration). Some fibres were tested up to 3 nm, but most were tested up to 10 nm. Fibres showing 100% inhibition are separated so that each can be seen (}–100).

Download figure to PowerPoint

image

Figure 3. Facilitatory and inhibitory effects of galanin on vagal afferent mechanosensitivity in the ferret Group data for responses of ferret mucosal (A), tension (B), and TM (C) receptors to mechanical stimulation. A, galanin inhibited slightly, but significantly, the responses to mucosal stroking (impulses per stroke, mean of 8 strokes). B, the effects of galanin are shown on the response to circumferential tension (average impulses s−1 over stimulus duration of 1 min). Bi, fibres with reduced mechanosensitivity after galanin (defined as below zero on Fig.1); Bii, one fibre that had increased mechanosensitivity. Ci and ii, effects of galanin on responses to tension; Ciii and iv, effects on responses to mucosal stroking. Ci and iii, TM receptors that had reduced mechanosensitivity; Cii and iv, those with increased or unaffected mechanosensitivity. The graphs show the means ±s.e.m. Significant differences between treated and control (assessed using two-way ANOVA) are shown nearest each treated response curve (*P < 0.05, **P < 0.01 and ***P < 0.001).

Download figure to PowerPoint

The effect of galanin was reversible by removing it from the Krebs superfusate in all experiments. Figure 4 illustrates the effect of galanin on the mechanical sensitivity of ferret mucosal (Fig. 4A) and tension (Fig. 4B) receptor afferents. In these cases a low dose of galanin (1 nm) completely abolished the response of mucosal and tension receptors to mucosal stroking and circumferential tension, respectively. When galanin (1 nm) was removed, the response was restored. In order to determine if galanin effects were due to changes in smooth muscle contractility, the L-type Ca2+ channel blocker nifedipine was present in four experiments, which was shown in other experiments to block smooth muscle contractions. Both positive and negative effects of galanin were unaffected by this treatment, indicating a direct action on the nerve endings.

image

Figure 4. Example responses of (A) a ferret mucosal receptor to mucosal stroking (200 mg von Frey hair), and (B) a ferret tension receptor to circumferential tension (3 and 4 g) The horizontal bars below the traces indicate when the mechanical stimuli were applied. The responses were obtained in the absence of galanin (i), the presence of galanin (1 nm) (ii), and 20 min after removal of galanin from the Krebs superfusate (iii). The response of both fibres to mechanical stimulation was reversibly reduced by galanin.

Download figure to PowerPoint

The effect of galanin on mouse mucosal and tension receptor afferent sensitivity to mechanical stimulation is illustrated in Fig. 5. Galanin (1–10 nm) significantly and dose-dependently reduced the response curve of mucosal receptors to mucosal stroking with calibrated von Frey hairs (P < 0.05–0.001, n= 6, Fig. 5A). Galanin (1–10 nm) also significantly reduced the response to circumferential tension of tension receptors (P < 0.001, n= 7).

image

Figure 5. Inhibitory effects of galanin on mouse vagal afferents. Group data for responses of (A) mouse mucosal and (B) tension receptors to mechanical stimulation The effects of galanin are shown on the responses to mucosal stroking (impulses per stroke, mean of eight strokes) of mucosal receptors. The effects of galanin are shown also on the response to circumferential tension (average impulses s−1 over stimulus duration of 1 min) of tension receptors. The graphs show the mean ±s.e.m. Significant differences between treated and control (assessed using two-way ANOVA) are shown nearest each treated response curve (*P < 0.05, **P < 0.01 and ***P < 0.001).

Download figure to PowerPoint

Galanin (0.1–10 nm) did not significantly affect the basal discharge of ferret vagal gastro-oesophageal mucosal (P= 0.98), tension (Fig. 3B, P= 0.16) or TM receptors (Fig. 3Ci and ii, P= 0.18). It also had no significant effect on basal discharge of mouse vagal gastro-oesophageal mucosal (P= 0.98) or tension receptor (Fig. 5B, P= 0.26) afferents.

Effect of the galanin receptor antagonist galantide The effect of galanin (1 or 10 nm) on mechanosensitivity was reversed by the nonselective galanin receptor antagonist galantide (3 or 30 nm) on both mouse mucosal and tension receptor afferents (n= 5 and 6, respectively, Fig. 6). During the galantide series of experiments, the responses of two mouse mucosal afferents to mucosal stoking were seen to be enhanced in the presence of galanin (10 nm), as seen in ferret studies with galanin. These data are not shown in Fig. 5 due to the difference in protocol. This potentiation was reversed with the addition of galantide (30 nm) (data not shown). Basal discharge of mouse mucosal (data not shown) and tension (Fig. 6B) receptor afferents was unaffected by galantide.

image

Figure 6. Reversal of galanin effect by a selective antagonist Group data for responses of mouse mucosal receptors to mucosal stroking (impulses per stroke, mean of 8 strokes) (A), and mouse tension receptors to circumferential tension (average impulses s−1 over stimulus time of 1 min) (B). The graphs show means ±s.e.m.***P < 0.001, significant difference from control data (assessed using two-way ANOVA). †††P < 0.001, significant from responses obtained in the presence of galanin alone (assessed using two-way ANOVA). Galantide reversed the effect of galanin such that there was no significant difference between control responses and responses obtained in the presence of galanin + galantide (assessed using two-way ANOVA).

Download figure to PowerPoint

The possible involvement of endogenous galanin in modulation of mechanosensitivity was investigated in the ferret by observing the effect of the antagonist galantide alone. Galantide (30 nm) increased the responses of TM receptors (n= 3) to circumferential tension (from 8.4 ± 0.3 to 13.0 ± 3.5 impulses s−1 at 7 g) but not their responses to mucosal stroking. Galantide also enhanced the response of mucosal receptor afferents (n= 2) to mucosal stroking (from 2.0 ± 1.5 to 4.6 ± 2.2 impulses per stroke at 1000 mg, e.g. Fig. 7). However, it did not affect the response to circumferential tension of one tension receptor tested.

image

Figure 7. Example response of a ferret mucosal receptor to mucosal stroking with a calibrated von Frey hair (1000 mg) The horizontal bars below the raw traces indicate when the mechanical stimuli were applied. The responses were obtained in the absence of galantide (A), the presence of galantide (30 nm) (B), and after removal of galantide from the Krebs superfusate (C). Note that the response of this fibre was reversibly enhanced by galantide.

Download figure to PowerPoint

Immunohistochemistry

Galanin immunoreactivity of ferret and mouse cell bodies of the nodose ganglia is illustrated in Fig. 8A, B and D. The majority of cell bodies in the ferret nodose ganglia showed strong immunoreactivity for galanin, whereas others were either negative or showed weak immunoreactivity (Fig. 8A). A proportion of the cell bodies that were retrogradely labelled from the stomach were also immunoreactive for galanin (Fig. 8B). Galanin-positive cell bodies were less frequently observed in mouse nodose (Fig. 8D).

image

Figure 8. Localization of galanin immunoreactivity A, ferret nodose ganglion showing galanin immunoreactivity in red. B, same as A, with retrogradely labelled gastric neurones overlaid in green. Arrows indicate galanin-positive vagal neurones projecting to the stomach with different levels of labelling intensity. Arrowhead indicates galanin-negative nodose neurone. C, section of ferret proximal stomach showing galanin-positive fibres throughout, particularly in the thicker circular muscle (cm), and a galanin-positive neurone in the myenteric plexus (arrow). Fibres are also evident in the longitudinal muscle (lm) and mucosa (mu). D, mouse nodose ganglion section, showing neurones with varying degrees of galanin immunoreactivity. E, whole mount of mouse proximal stomach showing network arrays of endings in circular muscle, which is the layer in focus. F, whole mount of mouse stomach showing abundant endings in and around 6 myenteric ganglia. Immunopositive nerve cell bodies were not seen in mouse stomach. Scale bar, 50 μm (A, B and D), 100 μm (C, E and F).

Download figure to PowerPoint

In sections of the ferret corpus (Fig. 8C), galanin immunoreactive fibres were seen in all layers, particularly the circular muscle layer. Some myenteric neurones were also strongly immunoreactive for galanin (Fig. 8C). Figure 8E and F illustrates immunoreactivity for galanin in the mouse proximal stomach whole-mount preparation. There were abundant fibres in the circular muscle layer immunoreactive for galanin (Fig. 8E), and there was strong evidence for immunoreactive fibres in and around myenteric ganglia (Fig. 8F). There were abundant fibres in the circular muscle layer immunoreactive for galanin in all preparations of ferret and mouse (e.g. Fig. 8E), and there were immunoreactive fibres in and around all observable myenteric ganglia (Fig. 8F).

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

Our data indicate that galanin potently influences responses to mechanical stimuli in several types of gastro-oesophageal vagal afferents in ferret and mouse. It may have either facilitatory or inhibitory actions, with facilitation seen mostly in TM receptors in the population studied. Actions were restricted to changes in mechanical responsiveness, as changes in spontaneous discharge were not seen. However, whether galanin affects mechanotransduction mechanisms separately from general excitability could not be determined. The effects of galanin are reversible upon washout or with a competitive galanin receptor antagonist. The effect of the antagonist alone is an indication that endogenous release of galanin may be able to modulate mechanosensitivity. This is not surprising as there are several peripheral sources of galanin. We have shown using immunohistochemistry in mouse and ferret that both intrinsic (myenteric) and extrinsic (vagal) neurones contain galanin, and importantly that vagal afferent galanin-containing neurones project to the stomach.

The presence of galanin in enteric nerves has been established in a number of mammalian species (Melander et al. 1985; Furness et al. 1987) including humans (Singaram et al. 1991, 1994), and its presence in vagal afferents has been shown in the rat (Calingasan & Ritter, 1992). Our data extend these observations, and confirm them in both locations in the ferret and mouse. Moreover, we have shown vagal afferents that project to the stomach contain galanin as do other vagal afferent cell bodies in the nodose ganglia. Both vagal and enteric sources are candidates for providing ongoing modulation of mechanosensitivity by endogenous release of galanin. Galanin is released from the isolated perfused gut by distension at a rate of 2.4 pmol min −1. This would most probably result in local concentrations equivalent to or higher than those we administered in this study (Harling et al. 1991). Circulating levels of galanin are approximately 18 pmol l−1 in humans, but would be several orders of magnitude higher at the site of release. These considerations implicate galanin as a strong candidate for an ongoing modulator of extrinsic mechanosensitive reflexes. Galanin has also been suggested to play a predominantly inhibitory role in intrinsic reflexes based on anatomical and functional evidence (Pham et al. 2002; Liu et al. 2003). Because galanin may have facilitatory or inhibitory effects on vagal afferents, the resulting change in the signal received by the central nervous system from the periphery would depend on the net effect on the whole population, or on the specific central targets of each population of facilitated and inhibited fibres.

The combined observations that galanin modulates mechanosensitivity in vagal afferents, and that it is contained within their cell bodies, indicate the possibility of an autoregulatory role of galanin, in addition to the role of galanin from other sources. Although we have shown evidence of galanin in the cell bodies of gastric afferents, an autoregulatory role would require galanin to be present at peripheral vagal endings from which it could be released. Endings of vagal tension receptors in the stomach and oesophagus have been identified as intraganglionic laminar endings, or IGLEs (Zagorodnyuk & Brookes, 2000; Zagorodnyuk et al. 2001), which have been shown to exist in mouse (Fox et al. 2000). These vagal endings have collaterals that form intramuscular arrays, or IMAs, in the longitudinal and circular muscle layers. It is possible that the abundant galanin positive endings we found in the mouse stomach may be IGLEs and IMAs; however, we did not see evidence for a laminar structure, and the endings were not identified by anterograde tracing.

In order to mediate both positive and negative effects on neuronal function, galanin acts on three different receptor subtypes: GALR1 and GALR3 are coupled positively to K+ channels and may also reduce cyclic AMP, giving rise to hyperpolarization (Branchek et al. 2000); GALR2, on the other hand, is coupled positively to phospholipase C, resulting in cellular excitation (Branchek et al. 2000). Therefore, it is highly likely that the inhibitory effects we observed are mediated via GALR1 and/or GALR3, and the potentiating effects are via GALR2. The divergent effects of galanin we observed on vagal afferents (see Fig. 2) could be accounted for by widely different expression of different GALR subtypes on different afferent fibres, although this remains to be demonstrated. So far there is only evidence for localization of GALR1 in the GI tract (Pham et al. 2002). Unfortunately, at the time of this study, there are no selective antagonists for any of these receptors, so more precise identification of the receptors mediating increases and decreases in mechanosensitivity is not possible. The potency of galanin was high in both positive and negative effects we observed, being within the range of affinity found for both human and rat GALR1 and GALR2. GALR3 has a slightly lower affinity for galanin and analogues (Branchek et al. 2000), but none of these analogues can as yet be used to distinguish the predominant receptor active in a system. Galantide is a nonselective GALR agonist, and the reasons for its use in this study were firstly to gain insight into the role of endogenous galanin, and secondly to determine if effects of galanin are reversible by an antagonist and therefore mediated via conventional galanin receptors. It was effective in achieving both of these aims. However, clearly there will have to be major advances in drug development before better tools are available, and the possibility remains that galantide may have as yet undiscovered actions other than as a GALR antagonist.

Although ours is the first report of the actions of galanin on vagal afferents, there are existing accounts of its actions on somatic afferents that are interesting to compare. Studies on rat knee joint afferents (Heppelmann et al. 2000) and on rat skin nociceptors (Flatters et al. 2003) both found subpopulations of afferents whose mechanical or thermal sensitivity was increased or decreased by galanin. Effects were observed both in vitro and in vivo, and proportions of fibres affected in each direction are comparable with our data. The antagonist galantide was similarly able to increase mechanosensitivity of knee joint afferents (Heppelmann et al. 2000). Interestingly, in a model of neuropathic pain, the proportion of afferents inhibited by galanin increased compared to controls (Flatters et al. 2003), suggesting that GALR1 and/or GALR3 may be upregulated in this condition. This has led to speculation that galanin receptors may be therapeutic targets in conditions involving somatic pain (Branchek et al. 2000; Wynick et al. 2001). We propose that they may also be targets in diseases that involve disordered afferent signalling from the GI tract. The relatively higher frequency at which we encountered inhibitory effects compared to excitatory effects suggests that agonists aimed at GALR1 or GALR3 may have more therapeutic potential for reducing mechanosensory function than antagonists for GALR2.

In conclusion, we have discovered a novel role for a peptide that was discovered in the gut some two decades ago. This new role is in the modulation of extrinsic afferent function of the upper GI tract, and fits alongside the roles already demonstrated for galanin in intrinsic control of gut secretion and motility. It shows parallels with the function of galanin in the somatic sensory innervation, and further exemplifies the sophisticated control of the extrinsic sensory innervation of the gut.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

Appendix

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

Acknowledgements

L. Ashley Blackshaw was supported by a National Health and Medical Research Council of Australia Senior Research Fellowship. Work was supported by NHMRC Australia grant number 104814.

Supplementary material

The online version of this paper can be accessed at:

http://jp.physoc.org/cgi/content/full/jphysiol.2004.075291/DC1 and contains supplementary material consisting of a table entitled: Proportion of mouse and ferret afferents affected by the presence of galanin (1–10 nm) in the Krebs superfusate. This material can also be found at: http://www.blackwellpublishing.com/products/journals/suppmat/tjp/tjp738/tjp738sm.htm

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  8. Supporting Information

Table S1.

FilenameFormatSizeDescription
TJP_738_sm_S1.doc24KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.