Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 1998 Jun 1;509(Pt 2):541–550. doi: 10.1111/j.1469-7793.1998.541bn.x

Mediation by 5-hydroxytryptamine of the femoral vasoconstriction induced by acid challenge of the rat gastric mucosa

Christof H Wachter *, Ákos Heinemann *, Josef Donnerer *, Maria A Pabst *, Peter Holzer *
PMCID: PMC2230965  PMID: 9575302

Abstract

  1. Gastric mucosal barrier disruption in the presence of luminal acid causes femoral vasoconstriction via a pathway that appears to be stimulated by messengers generated in the injured gastric mucosa. This study was undertaken to analyse the gastric factors that are responsible for the femoral vasoconstrictor response.

  2. Gastric mucosal barrier disruption in the presence of luminal acid was induced by perfusing the stomach of urethane-anaesthetized rats with ethanol (15 %) in 0.01-0.15 M HCl. Blood flow in the left gastric and right femoral artery was estimated by the ultrasonic transit time shift technique.

  3. Gastric perfusion of ethanol in HCl caused loss of H+ ions from the gastric lumen, decreased the HCO3− concentration in hepatic portal vein blood, induced macroscopic histological damage to the gastric mucosa, dilated the left gastric artery and constricted the femoral artery. These responses were related to the HCl concentration in the ethanol-containing perfusion medium.

  4. The femoral vasoconstriction was also seen when, instead of ethanol, taurocholate (20 mM) was used to disrupt the gastric mucosal barrier in the presence of 0.15 M HCl.

  5. The femoral vasoconstriction evoked by gastric perfusion of ethanol in HCl was left unaltered by pharmacological blockade of gastrin and histamine receptors. In contrast, the 5-hydroxytryptamine 5-HT1/2 receptor antagonist methiothepin, but not the 5-HT2A receptor antagonist ketanserin or the 5-HT3 receptor antagonist granisetron, inhibited the ability of both 5-hydroxytryptamine and gastric acid back-diffusion to constrict the femoral artery.

  6. Gastric acid back-diffusion caused release of 5-hydroxytryptamine into the gastric lumen, which was related to the HCl concentration in the ethanol-containing perfusion medium.

  7. These data show that femoral vasoconstriction evoked by gastric mucosal barrier disruption depends on back-diffusion of acid into the mucosa. The acid-induced damage results in release of 5-hydroxytryptamine from the gastric mucosa, and the pathway leading to constriction of the femoral artery involves 5-hydroxytryptamine acting via 5-HT1/2 receptors as a messenger molecule.

Chemical disruption of the gastric mucosal barrier in the presence of excess luminal acid causes back-diffusion of H+ ions into the mucosal tissue. The resulting tissue damage is limited to the surface of the mucosa by virtue of endogenous defence mechanisms among which a prompt and marked rise of mucosal blood flow plays a prominent role (Whittle, 1977; Bruggeman, Wood & Davenport, 1979; Holzer, Livingston & Guth, 1991). Gastric hyperaemia, however, is only part of the overall cardiovascular response to gastric acid challenge, since the increase in blood flow to the stomach is supported by somatic vasoconstriction as measured by a reduction of blood flow through the femoral artery (Wachter, Heinemann, Jocic & Holzer, 1995). Importantly, the mechanisms which underlie the gastric vasodilatation and femoral vasoconstriction are grossly divergent.

The increase in gastric blood flow is due to a neural mechanism which involves stimulation of capsaicin-sensitive spinal afferent neurones (Holzer et al. 1991; Raybould, Sternini, Eysselein, Yoneda & Holzer, 1992; Holzer, Wachter, Jocic & Heinemann, 1994), release of calcitonin gene-related peptide (Li, Raybould, Quintero & Guth, 1992; Holzer et al. 1994) and formation of nitric oxide (Lippe & Holzer, 1992; Holzer et al. 1994). In contrast, the reduction of femoral blood flow seems to be brought about by a pathway that depends on the extrinsic innervation of the stomach and on humoral vasoconstrictor messengers (Wachter et al. 1995). Because the gastric acid-evoked constriction of the femoral arterial bed is the result of acid injury to the gastric mucosa rather than a sequel of the gastric hyperaemia, it has been hypothesized that the pathway leading to femoral vasoconstriction is stimulated by factors that are generated in the damaged gastric mucosa (Wachter et al. 1995).

With this background, the present study set out to clarify two major aspects of the femoral vasoconstrictor response to gastric acid back-diffusion. The first aim was to characterize which factors in our experimental model of gastric acid injury have a particular bearing on the femoral vasoconstriction. This question was addressed by examining whether the femoral vasoconstrictor response is related to the concentration of acid in the gastric lumen, the type of gastric mucosal barrier breaker (ethanol or taurocholate), the extent of gastric mucosal damage, the loss of H+ ions from the gastric lumen and the blood bicarbonate (HCO3−) concentration in the hepatic portal vein.

The second aim was to extend our search for the factors which when released from the acid-injured gastric mucosa may stimulate the pathway that ultimately leads to constriction of the femoral arterial bed. In this context the possible participation of gastrin, histamine and 5-hydroxytryptamine (5-HT) was examined. Since a role for 5-HT was envisaged from the pharmacological experiments, the release of this amine from the gastric mucosa under conditions of acid challenge was also tested.

METHODS

Basic animal preparation

The animal experiments of this study were approved by an ethical committee at the Austrian Ministry of Science, Traffic and Arts. Female Sprague-Dawley rats, weighing 180-230 g, were fasted for 20 h but had free access to water. After the induction of anaesthesia with urethane (1.5 g kg−1s.c.) the rats were placed on a heated table, to maintain their rectal temperature at 37°C, and fitted with a tracheal cannula, to facilitate spontaneous breathing. The arterial blood pressure was recorded from a cannula in the right carotid artery by means of a pressure transducer connected to a bridge amplifier. The blood pressure signal was fed into a personal computer via an analog-digital converter, mean arterial blood pressure (MABP) and heart rate (HR) being simultaneously calculated from the blood pressure curve by the acquisition software.

A catheter in the left jugular vein was used for the continuous infusion of physiological saline (0.15 mM NaCl, 1.5 ml h−1), to avoid dehydration, and for the intravenous administration of drugs. The stomach, exposed by a mid-line laparotomy, was fitted with an inflow cannula placed in the forestomach and an outflow cannula inserted through the pylorus (Holzer et al. 1991). The experimental media (for composition see Experimental protocol) were kept at room temperature (20-22°C) and perfused through the stomach at a rate of 0.7-0.8 ml min−1, unless stated otherwise.

Haemodynamic measurements

Blood flow in the left gastric and femoral artery (GBF and FBF, respectively) was recorded continuously with the ultrasonic transit time shift technique which measures the net volume of blood flow with a factory-calibrated sensor (Barnes, Comline, Dobson & Drost, 1983; Holzer et al. 1994). To this end, the terminal segment of the left gastric artery close to its entry into the stomach and the right femoral artery proximal to the superficial epigastric artery were separated from the surrounding tissue over a length of 4-5 mm. A perivascular ultrasonic flow probe (model 1RB; Transonic, Ithaca, NY, USA) was placed around the artery under study and connected to a small animal flowmeter (model T106; Transonic) which calculated the volume of blood flow (in ml min−1) and fed the signal into a computer for analysis. Since some of the experimental manipulations changed MABP, the haemodynamic properties and changes observed in the left gastric and femoral artery are presented as changes in vascular conductance (GVC and FVC, respectively) rather than blood flow (Wachter et al. 1995). Vascular conductance, defined as blood flow divided by MABP and expressed in μl min−1 mmHg−1, was calculated on-line by the data acquisition programme.

Determination of H+ ion loss from the gastric perfusate

The gastric perfusate was collected in 15 min samples, and their H+ ion content was determined by titration of the samples to pH 7.0 with 0.01 M NaOH (Lippe & Holzer, 1992). Blank samples of the perfusion media were treated in the same manner as the gastric perfusate samples, and the loss of H+ ions was calculated by subtracting the acidity of the perfusate samples from the acidity of the respective blank samples.

Determination of blood HCO3− concentration in hepatic portal vein

After the mesenteric origin of the hepatic portal vein had been isolated, a metal cannula (o.d., 0.9 mm) connected to a polyethylene tubing (PE 100) was inserted in the portal vein such that the tip of the cannula lay proximal to the gastroduodenal vein. The catheter was fixed with cyanacrylate glue and the system filled with heparinized saline so that repeated sampling of blood was possible. During the experiments two samples of blood (each 0.1 ml) were taken and the HCO3− concentration measured with a blood gas analyser (model Compact 1; AVL, Graz, Austria).

Determination of 5-HT

5-HT concentration in the gastric perfusate and content in the gastric corpus wall was determined by high performance liquid chromatography (HPLC) according to the method of Holzer & Skofitsch (1985) with some modifications. Gastric perfusion samples (about 1 ml) were collected, placed on ice, centrifuged (5000 g, 10 min) and passed through a filter (type Millex-HV; Millipore, Molsheim, France). Aliquots (0.1 ml) of the clear supernatants were directly applied to the HPLC system. To extract 5-HT from the gastric wall, pieces of the gastric corpus (100-200 mg) were rapidly frozen in liquid nitrogen and kept at -70°C until assay. Immediately before assay, the frozen specimens were dropped into five volumes (w/v) of 0.2 M perchloric acid and homogenized by ultrasonication. The homogenates were centrifuged (5000 g, 15 min) and 0.1 ml aliquots of the filtered supernatants analysed.

The HPLC system (Beckman Gold; Beckman, San Ramon, CA, USA) was connected to an electrochemical detector (Bioanalytical Systems, West Lafayette, IN, USA) which was set at a voltage of +500 mV versus an Ag-AgCl reference electrode. 5-HT in the samples was separated with a reverse-phase column (LiChroCART, 125 × 4 mm) filled with LiChrospher 100 RP-18 (particle size, 5 μm), both of which were obtained from Merck (Darmstadt, Germany). The degassed mobile phase consisted of NaH2PO4 (0.1 M), heptanesulphonic acid (4 mM), methanol (14 % w/w) and ethylenediaminetetraacetic acid (1 mM) and was delivered at a constant flow rate of 1.25 ml min−1 at a pressure of about 20 000 Pa. All recordings and calculations were performed with a computer. The detection limit of the assay, which was defined as trace peak twice the background noise, was 0.1 ng ml−1 5-HT, and the dose-detector response relationship was linear for the range of 0.1-20 ng ml−1 5-HT. To account for any changes in the retention times or in the detector responses to the standards, which might occur with time, a run of standards was included after every eighth run of samples.

Examination of gross and deep mucosal injury

At the end of the experiments the stomachs were rapidly excised, opened along the greater curvature, pinned flat on a board, placed in fixative (2.5 % glutaraldehyde, 2 % paraformaldehyde in 0.1 M cacodylic acid buffer pH 7.2; Merck) and processed for histology as previously described (Pabst, Wachter & Holzer, 1996). Gross and histological injury of the gastric mucosa were evaluated by an observer who was unaware of the experimental treatment.

Gross damage of the gastric glandular mucosa was assessed by computerized planimetry. The excised fixed stomachs were photographed, and the area of the mucosa covered by visible haemorrhagic lesions was determined with a digitizing tablet and SigmaScan software (Jandel Scientific, Corte Madera, CA, USA). The area occupied by gross injury was expressed as a percentage of the total area of the glandular mucosa of the rat stomach (Holzer et al. 1991).

For histological examination, pieces of the fixed stomachs were embedded in Historesin (LKB, Bromma, Sweden) and cut to obtain 4 μm sections which were stained with a mixture of Methylene Blue-Azur II and basic Fuchsin (Pabst et al. 1996). The sections were taken randomly from the gastric corpus and included areas of haemorrhagic damage, if present. Deep mucosal injury was quantified by dividing the sections lengthwise into 10 μm segments and determining the section length that was occupied by damage involving more than 25 % of the mucosal depth. The extent of deep histological injury was expressed as a percentage of the total section length (Pabst et al. 1996).

Experimental protocol

The basic experimental protocol to study the effects of gastric acid challenge has been evaluated in previous studies (Holzer et al. 1991; Wachter et al. 1995). After setting up of the preparation an equilibration period of 15 min was allowed during which the stomach was perfused with saline (time 0-15 min). Thereafter the stomach was perfused with 0.05 or 0.15 M HCl for a period of 45 min (time 15-60 min). Finally, the acidic perfusion medium (0.05 or 15 M HCl) was replaced by a medium containing the same concentration of HCl plus 15 % ethanol (w/w) or 20 mM taurocholate, and the gastric perfusion continued for another 45 min (time 60-105 min). Ethanol and taurocholate were used to disrupt the gastric mucosal barrier and to allow luminal acid to diffuse into the mucosal tissue.

Three studies were carried out. In study 1, the effect of gastric acid challenge on mucosal lesion formation, loss of H+ ions from the gastric lumen, decrease in blood HCO3− concentration in the portal vein and blood flow in the left gastric and femoral artery was examined in four different sets of experiments. Macroscopic and histological damage to the gastric mucosa were determined in the first set of experiments. The stomachs were excised and examined for lesions at the end of the experiments (at 105 min), after they had been perfused (during the period 60-105 min) with 0.15 M HCl alone, 15 % ethanol in 0.05 M HCl or 15 % ethanol in 0.15 M HCl. When the influence of gastric acid challenge on the blood HCO3− concentration in the portal vein was analysed in the second set of experiments, blood samples were taken from the portal vein at 60 and 105 min after the start of the gastric perfusion. In the same experiments, the loss of H+ ions from the gastric perfusate was measured during the period 45-60 min (for gastric perfusion with 0.05 or 0.15 M HCl alone) or 90-105 min (for gastric perfusion with 15 % ethanol in 0.05 or 0.15 M HCl). Blood flow changes in the left gastric and femoral artery in response to six different gastric perfusions were analysed in the third set of experiments in which either saline alone, 0.15 M HCl alone, 15 % ethanol in saline, 15 % ethanol in 0.01 M HCl, 15 % ethanol in 0.05 M HCl or 15 % ethanol in 0.15 M HCl was perfused through the stomach during the period 60-105 min. The fourth set of experiments tested the influence of 20 mM taurocholate in saline or 0.15 M HCl (perfused through the stomach during the period 60-105 min) on the blood HCO3− concentration in the portal vein and on blood flow in the left gastric and femoral artery.

Study 2 was undertaken to identify potential mediators of the femoral vasoconstrictor response to gastric acid back-diffusion which was induced by perfusion of the stomach with 15 % ethanol in 0.15 M HCl during the period 60-105 min. The implication of 5-HT, histamine, and gastrin-cholecystokinin (CCK) was probed by pretreating rats with appropriate receptor antagonists. Thus, rats received either methiothepin (0.1 mg kg−1i.v.), ketanserin (0.1 mg kg−1i.v.), granisetron (0.5 mg kg−1i.v.), a combination of pyrilamine (2 mg kg−1i.v.) plus cimetidine (20 mg kg−1i.p.) or a combination of 4-{[2-[[3-(1H-indol-3-yl)-2-methyl-1-oxo-2-[[[1.7.7-trimethylbicyclo(2.2.1)hept-2-yl]-oxy]carbonyl]amino]propyl]amino]-1-phenylethyl}amino-4-oxo-{1S-1α.2β[S*(S)4α]}-butanoate N-methyl-D-glucamine (CAM-1481) and N-[α-methyl-N-[(tricyclo{3.3.1.13,7]dec-2-yloxy)carbonyl]-L-tryptophanyl]-D-3-(phenylmethyl)-β-alanine (CAM-1028) (each drug at 1 mg kg−1i.v.). These drugs were administered by bolus injection 15 min (CAM-1481 and CAM-1028) or 30 min (methiothepin, ketanserin, granisetron, pyrilamine and cimetidine) before exposure of the gastric mucosa to ethanol plus acid. The ability of methiothepin, ketanserin and granisetron to inhibit the constrictor action of exogenous 5-HT in the femoral artery was tested by recording the vasoconstrictor response to 5-HT (100 nmol kg−1i.v.) 15 min after the i.v. administration of vehicle or one of the 5-HT receptor antagonists.

Study 3 was carried out to examine whether gastric acid back-diffusion causes release of 5-HT from the gastric mucosa into the gastric lumen. To this end, the gastric perfusion rate was reduced to 0.35 ml min−1, 3 min perfusate samples were collected at 40 and 60 min (for gastric perfusion with 0.05 or 0.15 M HCl alone) or 80 and 100 min (for gastric perfusion with 0.15 M HCl alone, 15 % ethanol in 0.05 M HCl, or 15 % ethanol in 0.15 M HCl), and the 5-HT content estimated by HPLC. The 5-HT content remaining in the gastric corpus wall after the end of the experiments (105 min) was also determined in some rats.

Substances

The saline solution used here was 0.15 M NaCl. Urethane (250 g l−1; Fluka) and ketanserin (1 mg ml−1; Tocris Cookson) were dissolved in water. Sodium taurocholate (20 mM; Sigma) was dissolved in saline or 0.15 M HCl. CAM-1028 and CAM-1481 (each 1 mg ml−1; Parke-Davis, Cambridge, UK), granisetron (0.5 mg ml−1; SmithKline Beecham), 5-HT (100 μM; Sigma), methiothepin maleate (1 mg ml−1; Tocris Cookson) and pyrilamine (2 mg ml−1; Sigma) were dissolved in saline. Cimetidine (SmithKline Beecham) was dissolved in 0.1 M HCl, the solution being neutralized with 0.1 M NaOH and phosphate buffer of pH 7.4 (Merck), to a final concentration of 10 mg ml−1.

Statistics

Some of the experimentally induced alterations of the test parameters are given as increments or decrements (experimental minus baseline values) or as percentage changes relative to the baseline values. All data are expressed as means ±s.e.m. of n rats. Statistical evaluation of the results was performed with the Wilcoxon matched-pairs test, Mann-Whitney U test or Kruskal- Wallis H test followed by the Mann-Whitney U test, as appropriate. Probability values P < 0.05 were regarded as significant.

RESULTS

Effect of gastric perfusion with ethanol and/or acid on FBF

After a 60 min period of gastric perfusion with saline or HCl, MABP was 85 ± 1.5 mmHg, HR 408 ± 7.3 beats min−1, FBF 3.2 ± 0.18 ml min−1 and FVC 38 ± 2.2 μl min−1 mmHg−1 (n= 40 for each parameter). When the perfusion of the stomach was continued with saline for another 45 min, MABP and HR did not change significantly (data not shown), whereas FBF and FVC decreased slowly during the course of the experiment. The tendency of FVC to decrease with time did not differ significantly whether the stomach was perfused with saline, ethanol (15 % in saline) or HCl (0.15 M). FVC declined by 17 ± 2.8 % (n= 6), 17 ± 2.7 % (n= 8) and 12 ± 6.0 % (n= 8), respectively, during the 45 min period of gastric perfusion. The ethanol- and acid-induced decreases in FVC relative to baseline FVC (ΔFVC, in μl min−1 mmHg−1) are illustrated in Fig. 1.

Figure 1. Effect of gastric perfusion with ethanol, acid, and ethanol plus acid to decrease the vascular conductance in the femoral artery.

Figure 1

Open in a new tab

The graph shows the decrease in FVC (ΔFVC, experimental - baseline values), which was seen after gastric perfusion of saline (NaCl) plus ethanol (15 %), 0.15 M HCl, 0.01 M HCl plus ethanol, 0.05 M HCl plus ethanol, and 0.15 M HCl plus ethanol. *P < 0.05vs. 0.15 M HCl, †P < 0.05 vs. 0.01 M HCl plus ethanol (Kruskal-Wallis H test followed by Mann-Whitney U test), n= 8. Here and in subsequent figures, unless otherwise stated, the bars (±s.e.m.) represent the maximal changes that occurred during the 45 min perfusion of the stomach with the experimental media.

When ethanol (15 %) and HCl (0.15 M) were combined in the gastric perfusion medium, FVC decreased to a significantly larger extent than with either medium alone (Fig. 1). In the presence of ethanol, the decrease in FVC was related to the concentration of HCl in the gastric lumen. Figure 1 shows that 0.05 M HCl constricted the femoral artery to the same extent as 0.15 M HCl, whereas 0.01 M HCl was less effective. MABP and HR did not change during gastric perfusion with any of the media (data not shown), with the exception of ethanol in 0.15 M HCl, which caused the HR to increase by 23 ± 9.3 beats min−1 (n= 8, P < 0.05) over the 45 min observation period.

Relationship between gastric H+ ion loss, portal vein blood HCO3− concentration, gastric injury, gastric vasodilatation and femoral vasoconstriction

It has previously been shown that the gastric vasodilatation and damage caused by ethanol-induced disruption of the gastric mucosal barrier depends on the concentration of HCl in the gastric lumen (Holzer et al. 1991; Pethö, Jocic & Holzer, 1994; Pabst et al. 1996). Figure 2 depicts the relationship between gastric pathophysiology and femoral vasoconstriction, which was examined by gastric perfusion of 15 % ethanol in 0.05 or 0.15 M HCl as compared with perfusion of 0.15 M HCl alone. The FVC data in Fig. 2 are identical to the values presented in Fig. 1. While the ethanol plus acid-induced decrease in FVC was the same for 0.05 and 0.15 M HCl, the increase of vascular conductance in the left gastric artery (GVC) was related to the intragastric concentration of HCl (Fig. 2E and F) as has been reported earlier (Holzer et al. 1991; Pethöet al. 1994). Basal GBF and GVC were 0.59 ± 0.13 ml min−1 and 7.7 ± 1.4 μl min−1 mmHg−1, respectively (n= 12). The effect of gastric perfusion with 0.15 M HCl alone on GVC was not tested here because previous studies had shown this treatment to be ineffective (Holzer et al. 1991).

Figure 2. Effect of gastric perfusion with acid or ethanol plus acid in reducing the portal vein blood HCO3− concentration, causing H+ ion loss from the gastric lumen, inducing gross and deep injury to the mucosa, and evoking gastric vasodilatation and femoral vasoconstriction.

Figure 2

Open in a new tab

The graphs show the decrease in portal blood HCO3− concentration (Δ[HCO3−], A), the decrease in the H+ ion concentration of the gastric perfusate (Δ[H+], B), the extent of gross (macroscopic, C) and deep (histological, D) injury to the gastric mucosa (see Methods), the increase in GVC (ΔGVC, E) and the decrease in FVC (ΔFVC, F), which were seen after gastric perfusion of 0.15 M HCl, 0.05 M HCl plus ethanol (15 %) and 0.15 M HCl plus ethanol. *P < 0.05vs. 0.15 M HCl, †P < 0.05vs. 0.05 M HCl plus ethanol (Mann-Whitney U test or Kruskal-Wallis H test followed by Mann-Whitney U test), n= 5-8.

Figure 2A-D demonstrates that the effect of ethanol-induced disruption of the gastric mucosal barrier in causing loss of H+ ions from the gastric lumen, decreasing the HCO3− concentration in the portal venous blood and causing macroscopic and histological injury to the gastric mucosa was proportional to the HCl concentration in the gastric perfusion medium. The effect of HCl (0.15 M) alone was not examined in the experiments of Fig. 2A and B, since a comparison of the H+ ion loss induced by 15 % ethanol in 0.05 M HCl (this study) with that caused by 0.15 M HCl alone in another study (Holzer et al. 1991) indicated that the two treatments were equieffective. Gastric perfusion of ethanol (15 %) or HCl (0.15 M) alone had been shown to cause only little mucosal injury (Holzer et al. 1991; Pabst et al. 1996), which was confirmed in the present study as regards the injurious effect of 0.15 M HCl (Fig. 2C and D). However, combined perfusion of ethanol and acid through the stomach led to the formation of extensive lesions when the HCl concentration in the perfusion medium was increased from 0.05 to 0.15 M (Fig. 2C and D). This was true not only for the area of the gastric corpus mucosa which was covered by haemorrhagic lesions (gross injury) but also for the histological injury which involved more than 25 % of the mucosal height (deep damage).

Effect of mucosal barrier disruption with taurocholate in the presence of luminal acid

The question as to whether the type of gastric mucosal barrier breaker has any bearing on the gastric acid-induced vascular effects was examined by replacing ethanol in the acidic (0.15 M HCl) perfusion medium with taurocholate (20 mM). As illustrated in Fig. 3, taurocholate in saline failed to alter GVC, FVC and the portal venous concentration of HCO3−. In contrast, exposure of the gastric mucosa to taurocholate in 0.15 M HCl reduced the portal venous concentration of HCO3−, increased GVC and decreased FVC. MABP increased by 7.3 ± 1.4 mmHg (n= 12, P < 0.01) and HR increased by 51 ± 9.2 beats min−1 (n= 12, P < 0.01) during gastric perfusion of taurocholate-HCl, whereas under perfusion of taurocholate-saline MABP did not change significantly (+1.0 ± 1.3 mmHg, n= 12) and HR increased by 17 ± 4.3 beats min−1 (n= 12, P < 0.01) only. The taurocholate-HCl-induced peak changes of GVC and FVC were of similar magnitude to those evoked by ethanol in 0.15 M HCl (compare Fig. 2E and F with Fig. 3B and C). The latency of the maximal FVC decrease was similar whether ethanol or taurocholate was used to break the gastric mucosal barrier in the presence of 0.15 M HCl (41 ± 2.9 min (n= 8) compared with 38 ± 3.6 min (n= 6), respectively), whereas the increase in GVC peaked 20 ± 2.1 min (n= 7) after exposure to acidified ethanol, but 12 ± 2.8 min (n= 6, P < 0.05) after exposure to acidified taurocholate.

Figure 3. Effect of gastric perfusion with saline plus taurocholate or acid plus taurocholate in reducing the portal vein blood HCO3− concentration and evoking gastric vasodilatation and femoral vasoconstriction.

Figure 3

Open in a new tab

The graphs show the decrease in portal vein blood HCO3− concentration (Δ[HCO3−], A), the increase in GVC (ΔGVC, B) and the decrease in FVC (ΔFVC, C), which were seen after gastric perfusion of saline (NaCl) plus taurocholate (TC, 20 mM) and acid (HCl, 0.15 M) plus TC. *P < 0.01vs. NaCl plus TC (Mann-Whitney U test), n= 5-6.

Effect of 5-HT, histamine and gastrin-CCK receptor antagonists on the gastric ethanol-acid-induced femoral vasoconstriction

Rats were pretreated with 5-HT, histamine or gastrin- CCK receptor antagonists to investigate whether the respective endogenous receptor agonists play a role in the reduction of FVC that ensues in response to gastric perfusion of ethanol (15 %) in HCl (0.15 M). Intravenous injection of the 5-HT1/2 receptor antagonist methiothepin (0.1 mg kg−1) suppressed the femoral vasoconstrictor response to gastric acid challenge. As shown in Fig. 4A, perfusion of ethanol- HCl through the stomach of methiothepin-pretreated rats was no longer able to diminish FVC to a degree larger than that achieved by perfusion of the stomach of vehicle-pretreated rats with HCl alone. In contrast, neither the 5-HT2A receptor antagonist ketanserin (0.1 mg kg−1i.v.), the 5-HT3 receptor antagonist granisetron (0.5 mg kg−1i.v.), the combination of the CCKA receptor antagonist CAM-1481 with the CCKB receptor antagonist CAM-1028 (each at 1 mg kg−1i.v.) nor the combination of the histamine H1 receptor antagonist pyrilamine (2 mg kg−1i.v.) with the histamine H2 receptor antagonist cimetidine (20 mg kg−1i.p.) was able to attenuate the femoral vasoconstrictor response to gastric challenge with ethanol-HCl (Fig. 4A). None of the receptor antagonists had any significant influence on baseline MABP, HR, FBF and FVC as measured immediately before exposure of the gastric mucosa to ethanol-HCl (data not shown), except methiothepin which decreased MABP by 39 ± 7.8 mmHg, HR by 92 ± 25 beats min−1 and FBF by 0.90 ± 0.31 ml min−1 (n= 6, P < 0.05). Baseline FVC, however, did not significantly change in rats pretreated with methiothepin.

Figure 4. Pharmacological evidence for 5-HT being a mediator of the femoral vasoconstriction evoked by gastric perfusion of ethanol plus acid.

Figure 4

Open in a new tab

The graphs show the effects of vehicle (saline, 1 ml kg−1i.v.), methiothepin (0.1 mg kg−1i.v.), ketanserin (0.1 mg kg−1i.v.), granisetron (0.5 mg kg−1i.v.), CAM-1028 plus CAM-1481 (each 1 mg kg−1i.v.), and pyrilamine (2 mg kg−1i.v.) plus cimetidine (20 mg kg−1i.p.) on the decrease in the vascular conductance of the femoral artery (ΔFVC) evoked by gastric perfusion with acid (HCl, 0.15 M, A), ethanol (15 %) plus acid (HCl + ethanol, A) or i.v. injection of 5-HT (100 nmol kg−1, B). The bars in B represent the maximal changes that occurred immediately after the i.v. injection of 5-HT. *P < 0.05vs. vehicle (Mann-Whitney U test), n= 6-8. Note that all values from vehicle-treated rats receiving intragastric HCl + ethanol or i.v. 5-HT were pooled (n= 16 in A, n= 25 in B) to yield an overall vehicle response, although statistically the treatment groups were compared with the respective vehicle groups only.

Further experiments were carried out to test whether exogenous 5-HT constricts the femoral arterial bed and whether methiothepin, ketanserin or granisetron are able to antagonize this action of the amine. Intravenous injection of 5-HT (100 nmol kg−1) reduced FVC (Fig. 4B) to a similar extent as a 45 min perfusion of the stomach with 15 % ethanol in 0.15 M HCl (Fig. 4A). The vasoconstrictor response to 5-HT lasted 1-2 min only and was accompanied by a decrease in MABP and HR (data not shown). Methiothepin (0.1 mg kg−1i.v.) inhibited the ability of 5-HT to reduce FVC whereas ketanserin (0.1 mg kg−1i.v.) and granisetron (0.5 mg kg−1i.v.) were without effect (Fig. 4B).

Effect of gastric perfusion with ethanol-acid to release 5-HT from the gastric mucosa

Perfusion of the stomach with HCl (0.15 M) alone failed to enhance the 5-HT concentration in the gastric perfusate, which was close to the detection limit of the assay (Fig. 5A). When, however, ethanol (15 %) was added to the acidic perfusion medium, the 5-HT concentration in the gastric perfusate was markedly increased, the amounts of released 5-HT being related to the concentration of HCl in the gastric lumen (Fig. 5A). The 5-HT concentration of the gastric perfusate reached a maximum 20 min after the perfusion with ethanol in HCl had begun and stayed elevated for the rest of the experiment (Fig. 5A). Extraction of the gastric corpus wall after the end of the experiment revealed a 25 % depletion of 5-HT from stomachs perfused with ethanol in 0.15 M HCl compared with those treated with 0.15 M HCl alone (Fig. 5B).

Figure 5. Effect of gastric perfusion with ethanol plus acid in releasing 5-HT from the stomach.

Figure 5

Open in a new tab

The graphs show the 5-HT concentration in the gastric perfusate (A) and the content in the gastric corpus wall (B) of rats whose stomachs were perfused with 0.15 M HCl (▪), 0.05 M HCl plus ethanol (15 %, ▴) and/or 0.15 M HCl plus ethanol (•). The 5-HT content of the gastric corpus wall was determined after the end of the experiments. The values represent means ±s.e.m. (n= 6-8). *P < 0.05vs. 0.15 M HCl, †P < 0.05vs. 0.05 M HCl plus ethanol (Mann-Whitney U test or Kruskal-Wallis H test followed by Mann-Whitney U test).

DISCUSSION

Gastric back-diffusion of acid as a stimulus for femoral vasoconstriction

It has previously been demonstrated that the gastric hyperaemic response to ethanol-HCl perfusion is related to ‘back-diffusion’ of acid through the ethanol-disrupted mucosal barrier, because the response is absent when the stomach is perfused with either ethanol or HCl alone (Holzer et al. 1991). As shown here, the same is true for the concomitant constriction of the femoral arterial bed, the magnitude of which was directly related to the HCl concentration (0.01-0.15 M) in the ethanol-containing perfusion medium (Fig. 1). Direct evidence for acid influx into the mucosa comes from the ionic and histological consequences of gastric perfusion with ethanol-HCl. Exposure of the mucosa to ethanol-HCl led to a significant loss of H+ ions from the gastric perfusate, which coincided with a proportional decrease in the HCO3− concentration in portal vein blood collected proximally from the gastroduodenal vein which is the major venous vessel of the stomach. These complementary ionic changes indicate that acid back-diffusion into the gastric mucosa and circulation does in fact take place once ethanol has disrupted the mucosal barrier in the presence of luminal acid. One consequence of this process is that some of the circulating HCO3− is used up in buffering the influxing acid. Another consequence is that H+ ions enter the mucosa in amounts sufficient to cause tissue damage. We have found previously (Pabst et al. 1996) that deep mucosal injuries, which macroscopically manifest themselves as haemorrhagic erosions, develop only if substantial acid back-diffusion is initiated by exposing the mucosa to combined ethanol-HCl.

The current results signify that gastric H+ ion loss, HCO3− depletion in portal venous blood, formation of mucosal erosions and hyperaemia in the left gastric artery are similarly dependent on the HCl concentration (0.05-0.15 M) in the ethanol-containing perfusion medium. In contrast, the relationship between gastric HCl concentration and femoral vasoconstriction is different, since the constriction of the femoral artery was maximal at 0.05 M HCl while gastric pathology and hyperaemia increased further when the luminal HCl concentration was raised to 0.15 M (Figs 1 and 2). Femoral vasoconstriction hence seems more sensitive to perturbations of gastric mucosal homeostasis than gastric vasodilatation. This finding supports our working hypothesis that the gastric hyperaemic and femoral vasoconstrictor responses to gastric ethanol-HCl perfusion are mediated by diverse mechanisms (Wachter et al. 1995). Gastric back-diffusion of acid, rather than the specific agent causing disruption to the gastric mucosal barrier, appears to be the common primary stimulus for both vascular responses, because neither gastric vasodilatation nor femoral vasoconstriction was dependent on the agent being ethanol. The bile salt taurocholate, which is known to injure the mucosal surface (Whittle, 1977), was able to replace ethanol in the acidic perfusion medium in terms of portal venous HCO3− depletion, rise of GVC and fall of FVC (Fig. 3).

5-HT as mediator of the gastric acid-induced constriction of the femoral arterial bed

Although noradrenergic neurones and adrenal gland-derived mediators do not participate in the gastric acid-evoked decrease in FVC, it has been hypothesized that femoral vasoconstriction is brought about by a pathway that depends on the extrinsic innervation of the stomach and involves humoral vasoconstrictor messengers (Wachter et al. 1995). The vasoconstrictor pathway seems to be stimulated by factors that are generated in the damaged stomach because intragastric capsaicin, which causes gastric hyperaemia without injuring the mucosa, fails to constrict the femoral artery (Wachter et al. 1995). After a contribution by vasopressin, angiotensin II, endothelin and prostanoids had been ruled out (Wachter et al. 1995), the current study set out to test the implication of three factors that may be released from the acid-injured gastric mucosa: gastrin, histamine and 5-HT. Gastrin is expressed in endocrine gastrin (G) cells of the gastric antrum (Larsson & Rehfeld, 1977), while the major cellular source of histamine in the stomach is enterochromaffin-like cells (Håkanson et al. 1986; Prinz, Kajimura, Scott, Mercier, Helander & Sachs, 1993), and 5-HT is primarily contained in enterochromaffin cells of the oxyntic and antral mucosa of the rat stomach (Facer, Polak, Jaffe & Pearse, 1979; Nilsson, Ericson, Dahlström, Ekholm, Steinbusch & Ahlman, 1985; Oomori, Iuchi, Ishikawa, Satoh & Ono, 1992).

Gastrin has been ruled out as playing a role in the gastric acid-induced decrease in FVC because a combination of the gastrin-CCK CCKA receptor antagonist CAM-1481 and the CCKB receptor antagonist CAM-1028 failed to alter the vascular response under study. The doses of CAM-1481 and CAM-1028 used here have previously been shown to be effective and selective in antagonizing CCKA and CCKB receptors, respectively (Heinemann et al. 1995). Histamine, which plays a role in the gastric vasodilator response to acid back-diffusion in the cat (Gislason, Guttu, Sørbye, Schifter, Waldum & Svanes, 1995) but not in the rat (Holzer et al. 1991), likewise does not contribute to the femoral vasoconstriction due to gastric acid challenge, since effective doses of the histamine H1 receptor antagonist pyrilamine (Hahn, 1978) plus the histamine H2 receptor antagonist cimetidine (Brimblecombe, Duncan, Durant, Emmett, Ganellin & Parsons, 1975) were unable to attenuate the gastric acid-evoked constriction of the femoral artery. Both H1 and H2 receptor antagonists were employed since histamine-induced cardiovascular changes involve both histamine receptors (Black, Owen & Parsons, 1975).

In contrast, the 5-HT1/2 receptor antagonist methiothepin inhibited the decrease in FVC evoked by gastric perfusion with ethanol-HCl, which proves that this amine participates in the femoral vasoconstrictor response to gastric acid back-diffusion. The implication of 5-HT1/2 receptors was corroborated by the observation that neither the 5-HT2A receptor antagonist ketanserin nor the 5-HT3 receptor antagonist granisetron was able to attenuate femoral vasoconstriction caused by 5-HT or gastric acid back-diffusion. It should be emphasized in this context that the doses of methiothepin, ketanserin and granisetron used here have been established to be fully effective in antagonizing the respective 5-HT receptors (Connor, Feniuk, Humphrey & Perren, 1986; Sanger & Nelson, 1989; Banner, Carter & Sanger, 1995). Antagonists for 5-HT1, 5-HT2 and 5-HT3 receptors were tested because the 5-HT receptors expressed on blood vessels belong to the 5-HT1 and 5-HT2 receptor type (Ullmer, Schmuck, Kalkman & Lübbert, 1995; Sanders-Bush & Meyer, 1996) and 5-HT3 receptors, which are expressed on neurones but not blood vessels (Sanders-Bush & Meyer, 1996), can modulate regional blood flow (Vanner, Jiang & Surprenant, 1993; Cho, Koo & Ko, 1994). Since methiothepin reduced the decrease in FVC evoked by intragastric ethanol-HCl to a level that was indistinguishable from the baseline decrease in FVC over time (Fig. 4A), we conclude that 5-HT acting via 5-HT1/2 receptors is an essential mediator of the femoral vasoconstriction evoked by gastric acid back-diffusion. We think this conclusion to be valid because baseline FVC was not altered by methiothepin despite the drug's known effect on MABP and HR (Connor et al. 1986).

Consistent with the participation of 5-HT in the femoral vasoconstriction elicited by gastric acid challenge (Fig. 4) is the finding that exposure of the gastric mucosa to ethanol-HCl released substantial amounts of 5-HT into the gastric lumen, which resulted in depletion of the amine from the gastric wall (Fig. 5). Since the relationship between HCl concentration and 5-HT release (Fig. 5A) was similar to that between HCl concentration and gastric pathology (Fig. 2C and D) it would seem that 5-HT release is a manifestation of acid-induced injury to the mucosa and reflects disruption of enterochromaffin cells and possibly other cells containing 5-HT such as platelets (Sanders-Bush & Meyer, 1996). It is worth noting in this context that gastric lesion formation provoked by intracisternal administration of a thyrotropin- releasing hormone analogue is also accompanied by intraluminal release of 5-HT (Stephens & Taché, 1989).

Conclusions

It has previously been hypothesized that the gastric acid-evoked constriction of the femoral artery is due to a pathway that is stimulated by factors released from the injured gastric mucosa, that depends on the extrinsic innervation of the stomach and that brings about vasoconstriction via humoral messengers (Wachter et al. 1995). The present study demonstrates that the femoral vasoconstrictor response to gastric mucosal barrier disruption in the presence of luminal acid arises from back-diffusion of acid into the mucosa and involves 5-HT acting via 5-HT1/2 receptors as essential mediator. This inference is supported by the observation that acid influx into the gastric mucosa releases 5-HT into the gastric lumen. It would hence appear that 5-HT released from the injured gastric mucosa stimulates a neural and/or endocrine pathway that ultimately leads to femoral vasoconstriction, although it cannot be ruled out that 5-HT itself may also act as a vasoconstrictor messenger.

Acknowledgments

This work was supported by the Jubiläumsfonds of the Austrian National Bank (grant 4207) and the Austrian Science Foundation (grants P9473-MED and P11834-MED). The authors thank Dr K. Sabin (Department of Surgery, University of Graz) for providing the AVL blood gas analyser, Dr D. C. Horwell (Parke-Davis Neuroscience Research Centre, Cambridge, UK) for providing samples of CAM-1028 and CAM-1481, Dr G. J. Sanger (SmithKline Beecham, Harlow, UK) for providing a sample of granisetron, and Mr W. Schluet for organizational assistance.

References

  1. Banner SE, Carter M, Sanger GJ. 5-Hydroxytryptamine3 receptor antagonism modulates a noxious visceral pseudoaffective reflex. Neuropharmacology. 1995;34:263–267. doi: 10.1016/0028-3908(94)00159-p. [DOI] [PubMed] [Google Scholar]
  2. Barnes RJ, Comline RS, Dobson A, Drost CJ. An implantable transit time ultrasonic blood flow meter. The Journal of Physiology. 1983;345:2–3. P. [Google Scholar]
  3. Black JW, Owen DAA, Parsons ME. An analysis of the depressor responses to histamine in the cat and dog: involvement of both H1- and H2-receptors. British Journal of Pharmacology. 1975;54:319–324. doi: 10.1111/j.1476-5381.1975.tb07571.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brimblecombe RW, Duncan WAM, Durant GJ, Emmett JC, Ganellin CR, Parsons ME. Cimetidine - a non-thiourea H2-receptor antagonist. Journal of International Medical Research. 1975;3:86–92. [Google Scholar]
  5. Bruggeman TM, Wood JG, Davenport HW. Local control of blood flow in the dog's stomach: vasodilatation caused by acid back-diffusion following topical application of salicylic acid. Gastroenterology. 1979;77:736–744. [PubMed] [Google Scholar]
  6. Cho CH, Koo MWL, Ko JKS. Modulatory role of 5-HT3 receptors in gastric function and ethanol-induced mucosal damage in rat stomachs. Pharmacology. 1994;49:137–143. doi: 10.1159/000139227. [DOI] [PubMed] [Google Scholar]
  7. Connor HE, Feniuk W, Humphrey PPA, Perren MJ. 5-Carboxamidotryptamine is a selective agonist at 5-hydroxytryptamine receptors mediating vasodilatation and tachycardia in anaesthetized cats. British Journal of Pharmacology. 1986;87:417–426. doi: 10.1111/j.1476-5381.1986.tb10832.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Facer P, Polak JM, Jaffe BM, Pearse AGE. Immunocytochemical demonstration of 5-hydroxytryptamine in gastrointestinal endocrine cells. Histochemical Journal. 1979;11:117–121. doi: 10.1007/BF01041271. [DOI] [PubMed] [Google Scholar]
  9. Gislason H, Guttu K, Sørbye H, Schifter S, Waldum HL, Svanes K. Role of histamine and calcitonin gene-related peptide in the hyperemic response to hypertonic saline and H+ back-diffusion in the gastric mucosa of cats. Scandinavian Journal of Gastroenterology. 1995;30:300–310. doi: 10.3109/00365529509093281. [DOI] [PubMed] [Google Scholar]
  10. Hahn F. Antianaphylactic and antiallergic effects. In: Rochee E, Silva M, editors. Histamine II and Anti-Histaminics Handbook of Experimental Pharmacology. Vol. 18. Berlin: Springer; 1978. pp. 440–504. part 2. [Google Scholar]
  11. Håkanson R, Boettcher G, Ekblad F, Panula P, Simonsson M, Dohlsten M, Hallberg T, Sundler F. Histamine in endocrine cells in the stomach. Histochemistry. 1986;86:5–17. doi: 10.1007/BF00492340. [DOI] [PubMed] [Google Scholar]
  12. Heinemann A, Jocic M, Holzer-Petsche U, Pethö G, Peskar BM, Horwell DC, Holzer P. Mediation by CCKB receptors of the CCK-evoked hyperaemia in rat gastric mucosa. British Journal of Pharmacology. 1995;116:2274–2278. doi: 10.1111/j.1476-5381.1995.tb15064.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Holzer P, Livingston EH, Guth PH. Sensory neurons signal for an increase in rat gastric mucosal blood flow in the face of pending acid injury. Gastroenterology. 1991;101:416–423. doi: 10.1016/0016-5085(91)90020-l. [DOI] [PubMed] [Google Scholar]
  14. Holzer P, Skofitsch G. Release of endogenous 5-hydroxytryptamine from the myenteric plexus of the guinea-pig isolated small intestine. British Journal of Pharmacology. 1984;81:381–386. doi: 10.1111/j.1476-5381.1984.tb10089.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Holzer P, Wachter C, Jocic M, Heinemann A. Vascular bed-dependent roles of the peptide CGRP and nitric oxide in acid-evoked hyperaemia of the rat stomach. The Journal of Physiology. 1994;480:575–585. doi: 10.1113/jphysiol.1994.sp020385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Larsson LI, Rehfeld JF. Characterization of antral gastrin cells with region-specific antisera. Journal of Histochemistry and Cytochemistry. 1977;25:1317–1321. doi: 10.1177/25.12.925341. [DOI] [PubMed] [Google Scholar]
  17. Li D-S, Raybould HE, Quintero E, Guth PH. Calcitonin gene-related peptide mediates the gastric hyperemic response to acid back-diffusion. Gastroenterology. 1992;102:1124–1128. [PubMed] [Google Scholar]
  18. Lippe IT, Holzer P. Participation of endothelium-derived nitric oxide but not prostacyclin in the gastric mucosal hyperaemia due to acid back-diffusion. British Journal of Pharmacology. 1992;105:708–714. doi: 10.1111/j.1476-5381.1992.tb09043.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Nilsson O, Ericson LE, Dahlström A, Ekholm R, Steinbusch HWM, Ahlman H. Subcellular localization of serotonin immunoreactivity in rat enterochromaffin cells. Histochemistry. 1985;82:351–355. doi: 10.1007/BF00494064. [DOI] [PubMed] [Google Scholar]
  20. Oomori Y, Iuchi H, Ishikawa K, Satoh Y, Ono K. Gamma-aminobutyric acid immunoreactivity in the enterochromaffin cells of the rat stomach. Acta Anatomica. 1992;145:127–131. doi: 10.1159/000147353. [DOI] [PubMed] [Google Scholar]
  21. Pabst MA, Wachter C, Holzer P. Morphologic basis of the functional gastric acid barrier. Laboratory Investigation. 1996;74:78–85. [PubMed] [Google Scholar]
  22. Pethö G, Jocic M, Holzer P. Role of bradykinin in the hyperaemia following acid challenge of the rat gastric mucosa. British Journal of Pharmacology. 1994;113:1036–1042. doi: 10.1111/j.1476-5381.1994.tb17097.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Prinz C, Kajimura M, Scott DR, Mercier F, Helander HF, Sachs G. Histamine secretion from rat enterochromaffin-like cells. Gastroenterology. 1993;105:449–461. doi: 10.1016/0016-5085(93)90719-s. [DOI] [PubMed] [Google Scholar]
  24. Raybould HE, Sternini C, Eysselein VE, Yoneda M, Holzer P. Selective ablation of spinal afferent neurons containing CGRP attenuates gastric hyperemic response to acid. Peptides. 1992;13:249–254. doi: 10.1016/0196-9781(92)90104-b. 10.1016/0196-9781(92)90104-B. [DOI] [PubMed] [Google Scholar]
  25. Sanders-Bush E, Mayer SE. 5-Hydroxytryptamine (serotonin) receptor agonists and antagonists. In: Hardman JG, Limbird LE, Molinoff PB, Ruddon RW, Gilman AG, editors. Goodman & Gilman's The Pharmacological Basis of Therapeutics. 9. New York: McGraw-Hill; 1996. pp. 249–263. [Google Scholar]
  26. Sanger GJ, Nelson DR. Selective and functional 5-hydroxytryptamine3 receptor antagonism by BRL 43694 (granisetron) European Journal of Pharmacology. 1989;159:113–124. doi: 10.1016/0014-2999(89)90695-x. 10.1016/0014-2999(89)90695-X. [DOI] [PubMed] [Google Scholar]
  27. Stephens RL, Taché Y. Intracisternal injection of a TRH analogue stimulates gastric luminal serotonin release in rats. American Journal of Physiology. 1989;256:G377–383. doi: 10.1152/ajpgi.1989.256.2.G377. [DOI] [PubMed] [Google Scholar]
  28. Ullmer C, Schmuck K, Kalkman HO, Lübbert H. Expression of serotonin receptor mRNAs in blood vessels. FEBS Letters. 1995;370:215–221. doi: 10.1016/0014-5793(95)00828-w. 10.1016/0014-5793(95)00828-W. [DOI] [PubMed] [Google Scholar]
  29. Vanner S, Jiang M-M, Surprenant A. Mucosal stimulation evokes vasodilation in submucosal arterioles by neuronal and nonneuronal mechanisms. American Journal of Physiology. 1993;264:G202–212. doi: 10.1152/ajpgi.1993.264.2.G202. [DOI] [PubMed] [Google Scholar]
  30. Wachter CH, Heinemann Á, Jocic M, Holzer P. Visceral vasodilatation and somatic vasoconstriction evoked by acid challenge of the rat gastric mucosa: diversity of mechanisms. The Journal of Physiology. 1995;486:505–516. doi: 10.1113/jphysiol.1995.sp020829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Whittle BJR. Mechanisms underlying gastric mucosal damage induced by indomethacin and bile salts, and the actions of prostaglandins. British Journal of Pharmacology. 1977;60:455–460. doi: 10.1111/j.1476-5381.1977.tb07522.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES