Abstract

The renin-angiotensin system (RAS) is present in the gastrointestinal (GI) tract but remains to be fully characterized, particularly in man. The duodenum plays a role in both the upper and lower GI regulation, as well as in distant organs. The present study investigates the presence and functional potential of RAS in the human duodenal mucosa of healthy individuals. Endoscopically acquired mucosal biopsies from healthy volunteers were examined using western blot, immunohistochemistry, and ELISA. Functionality was examined by using Ussing chambers and recording duodenal transmucosal potential difference (PD) and motility in vivo. Angiotensinogen, Angiotensin II (AngII) and its receptors (AT1R, AT2R) as well as to the RAS associated enzymes renin, ACE, and neprylisin were detected in all samples of duodenal mucosa. Migrating motility complex induced elevations of transmucosal PD were significantly larger after per-oral administration of the AT1R receptor antagonist candesartan. Fasting duodenal motility per se was not influenced by candesartan. The epithelial current produced by duodenal mucosae mounted in Ussing chambers increased significantly after addition of AngII to specimens where the AT1R was blocked using losartan. The epithelial current also increased after addition of the AT2R-selective agonist C21. Immunostaining and pharmacological data demonstrate the presence of a local RAS in the human duodenal mucosa with capacity to influence epithelial ion transport by way of particulary the AT2R.

Introduction

The duodenum is the most proximal part of the small intestine and its functional characteristics differ markedly from the rest of the intestine. The duodenum is regulating gastric emptying and delivers its contents to the rest of the intestine with controlled action on hepatobiliary secretions. With its central position, the duodenum has recently been claimed a role in functional dyspepsia or as a regulatory center for metabolic controls [1,2]. Both these locations involving immunomodulatory actions due to microbial profile in the luminal contents. The renin-angiotensin system (RAS) is mainly known for its role in body fluid homeostasis and circulation. The RAS has been linked to a multitude of actions like Na+ and fluid transport, inflammatory mechanism, fibrosis/antifibrosis, including immunomodulatory actions [3]. In addition to its classical endocrine character, the RAS has been recognized also as a locally expressed signaling system in many tissues [3]. Even though the gastrointestinal (GI) tract is one of the major players in fluid and electrolyte intake and excretion, the role of RAS in the GI-tract remains to be fully characterized.

The duodenal mucosa protects itself from the highly acidic luminal content disposed from the stomach by secreting bicarbonate ions (HCO3) resulting in a neutralizing zone in the pre-epithelial mucus layer [4]. An insufficient ability to adapt this secretion to oncoming, or ongoing, luminal acidification contributes to, for example, duodenal peptic ulcer disease [4]. Interestingly, in rats the duodenal mucosal anion (HCO3) transport is influenced by a mucosal RAS. The principal mediator Angiotensin II (AngII) can reduce this secretion via activation of the AT1R and increase the secretion via the AT2R [5,6].

The aim of the present study was to determine if RAS is present and influences the functional state of the normal human duodenum. First, the presence and cellular location of classical RAS components in human mucosal biopsies of healthy volunteers were assessed using western blotting and immunohistochemistry. Second, effects on duodenal secretion by interdigestive motility were assessed in vivo by administration of a selective AT1R-antagonist (the antihypertensive drug candesartan). Third, epithelial transport was assessed in human duodenal mucosae in vitro using mini-Ussing chambers. The present study in man shows that AngII may be synthetized in the duodenum and that it influences mucosal functions, but not the fasting muscular activity of the intestinal wall.

Materials and methods

General

The present study was approved by the Ethical committee of the University of Gothenburg (193-02) and the Regional Ethical Review Board in Gothenburg (647-05, 007-09) and was carried out in accordance with the World Medical Association’s Declaration of Helsinki. All study subjects received verbal and written information about the study and signed a written consent form prior to participation.

Exploration of RAS components in the duodenal mucosa

A total of 13 overnight fasted healthy volunteers (mean: age 47 years; range: 20–61 years, one female) underwent duodenoscopy with sampling of mucosal biopsies. All subjects tested negative for Helicobacter pylori (13C-urea breath test). One subject was diagnosed as having Barrett’s esophagus indicating chronic gastroesophageal reflux whereas the rest of the examined subjects had normal endoscopic esophago-gastric appearances. The duodenal mucosa was carefully inspected during the endoscopy and none of the subjects showed any signs of pathology. Six to eight mucosal biopsies were harvested from the descendo-horisontal duodenum in each individual and were either snap-frozen in liquid nitrogen for later ELISA or western blot analysis, or fixated in 4% formaldehyde for subsequent immunohistochemistry.

ELISA and western blotting

To optimize protein content three mucosal biopsies from each individual were sonicated in a PE buffer (10 mM potassium phosphate buffer, pH 6.8, and 1 mM EDTA) containing 10 mM 3-([3-cholamidopropyl] dimethylammonio)-1-propane sulfonate (CHAPS: Boehringer Mannheim, Mannheim, Germany) and protease inhibitor cocktail tablet Complete (Roche Diagnostics, Stockholm, Sweden). The homogenate was then centrifuged (10,000 g for 10 min at 4°C) and the supernatant was analyzed for protein content according to the Bradford method [7] and stored at −80°C until further analysis. A small amount of the homogenates was used for determinations of AngII in relation to the actual amount of protein. The proteins were solubilized in SDS buffer. The analysis was performed according to the manufacturer’s instructions (AngII enzyme immunoassay kit A05880-96 wells, Spi Bio Bertin Pharma, Montigny le Bretonneux, France). The antibody–AngII complex forms a yellow compound together with a tracer (acetyl-cholinesterase) and a chromogen (Ellman’s reagent). Each sample’s concentration is colorimetrically determined (absorbance at 405 nm, TECAN, Salzburg, Austria) and subsequently compared with a prepared standard curve on the same ELISA plate. The limit of detectability was 0.09 pg/mg tissue protein and CV was 5%. The rest of homogenates were used for western blotting of the following components of RAS: AT1R and AT2R, renin, angiotensin-converting enzyme (ACE), neprilysin (NEP), and angiotensinogen (AGT) (Supplementary Table S1). Samples were diluted in SDS buffer and heated at 70°C for 10 min before being loaded on a NuPAGE® 10% BisTris gel (Invitrogen AB, Lidingö, Sweden). One lane was loaded with pre-stained molecular weight standards (SeeBlue®, Invitrogen AB), and whole cell lysates were used as positive controls. Following electrophoresis, the proteins were transferred on to a PVDF membrane (Amersham, Buckinghamshire, U.K.) using an iBlot® (Invitrogen AB). The membrane was incubated with the primary antibody at 4°C overnight and then an alkaline phosphatase-conjugated secondary antibody was added with CDP-Star® as a substrate (Tropix, Bedford, MA, U.S.A.) to identify immunoreactivity by chemiluminescense. Images were captured by a ChemiDoc™ XRS cooled charge coupled device camera. After image capturing, PVDF membranes were stripped of antibodies using an antibody stripping solution (Re-Blot Plus Mild Solution (10X), Millipore, Temecula, CA, U.S.A.), washed in wash buffer and blocking buffer, and then incubated with a new primary antibody.

Table 1
Fasting duodenal motor activity in healthy male volunteers
Contraction frequency (number/min)Contraction amplitude (cmH2O)Mean intraluminal pressure (cmH2O)
No treatmentCandesartanNo treatmentCandesartanNo treatmentCandesartan
Phase II 1.9 (0.3) 2.4 (0.6) 14.5 (2.0) 15.1 (3.3) 14.0 (1.7) 17.3 (1.4) 
Phase III 11 (0.5) 10.4 (0.4) 19.4 (2.3) 23.1 (2.7) 25.9 (2.3 24.4 (1.5) 
Phase I 0.9 (0.4) 0.6 (0.2) 4.1 (0.5) 4.4 (1.2) 15.3 (1.9) 17 (1.7) 
Erythro. 11 (0.4) 11.3 (0.4) 22.3 (2.0) 22.7 (2.5) 27.2 (2.5) 27.3 (2.5) 
Contraction frequency (number/min)Contraction amplitude (cmH2O)Mean intraluminal pressure (cmH2O)
No treatmentCandesartanNo treatmentCandesartanNo treatmentCandesartan
Phase II 1.9 (0.3) 2.4 (0.6) 14.5 (2.0) 15.1 (3.3) 14.0 (1.7) 17.3 (1.4) 
Phase III 11 (0.5) 10.4 (0.4) 19.4 (2.3) 23.1 (2.7) 25.9 (2.3 24.4 (1.5) 
Phase I 0.9 (0.4) 0.6 (0.2) 4.1 (0.5) 4.4 (1.2) 15.3 (1.9) 17 (1.7) 
Erythro. 11 (0.4) 11.3 (0.4) 22.3 (2.0) 22.7 (2.5) 27.2 (2.5) 27.3 (2.5) 

Data from two separate study-days either without treatment or with the AT1R antagonist candesartan 16 mg p.o. One manometric recording site in the mid-duodenum was used in the analysis. Phase II: motor activity during 15 min before onset of phase III. Phase I: motor activity 15 min after ending of phase III; and Phase III: the propagating high activity motor complex. Data are given as mean (S.E.M.), n=7.

Erythro.: high activity motor complex induced by erythromycin (50 mg intravenously).

Immunohistochemistry

Mucosal biopsies were immediately fixed in 4% formaldehyde (Histolab products AB, Sweden) at room temperature and subsequently dehydrated, embedded in paraffin, cut into 4 μm thick sections, and mounted on glass slides. Before staining, sections were deparaffinized and boiled in citric acid buffer (10 mM, pH 6.0, 0.05% TWEEN®) for 20 min for antigen retrieval. Afterward, the samples were left to cool for 2 h in the citrate buffer and then incubated for 2 h in blocking solution (5% goat-serum and PBS, 0.3% Triton® X-100). Incubation with primary antibody against AGT, AT1R, AT2R, or S100 (Supplementary Table S1) diluted in block-solution was carried out overnight. After washing in PBS-0.3% triton, the slides were incubated with secondary antibody (Alexa Fluor 488 cat.no: A11008, Invitrogen corporation, Oregon, U.S.A.) in 2 h and then with Hoechst solution (Sigma–Aldrich Sweden AB, Stockholm, Sweden) for 10 min. In a separate set of analyses immunostaining for AGT, S-100 and AT1R or AT2R were run on samples cut in series using a peroxidase based method and the colour was developed using 3,3′-diaminobenzidine (DAB). Hematoxylin (Gill’s formulation #2, sc-24973, Santa Cruz) was used for counterstaining.

Duodenal motor activity and transmucosal PD in vivo

After an overnight fast, seven healthy male volunteers (n=7, median: age 26 years, range: 24–28 years; median BMI: 24, 1 kg/m2, range: 19.4–28.9) were supplied with a nasogastroduodenal catheter, containing ten separate channels with sideholes at 2 cm intervals (modified Zinectics tube, Synmed AB, Stockholm, Sweden). The catheter was fixed at one nostril and positioned with the two most proximal side-holes in the stomach and one in the pyloric canal. The remaining seven sideholes were located to the duodenum (Supplementary Figure S1). The catheter was equipped with radio-opaque markers and its position was assessed by fluoroscopy at the beginning and end of the investigation as well as during phase III of the migrating motility complex (MMC). Each sidehole was connected to a pressure transducer that was separately fed with 150 mmol/l NaCl at a low flow rate (3 ml/h). In addition, each pressure-recording line (thus electrically isolated from each other) was used as an electrode for electrical potential recording via an Ag–AgCl bridge connected to a high-impedance voltmeter, with a reference electrode positioned subcutaneously. Both the intraluminal hydrostatic pressure and transmucosal electrical potential difference (PD) at each sidehole were displayed on-line and stored for later analysis using a specially designed software (Labview; National instruments, Austin, TX, U.S.A.).

Definition of fasting motor activity

Phase I was defined as motor quiescence with two contractions or less per 10 min after a phase III period, phase II was a period with contractions of a frequency of between 2 and 8 per minute and with irregular propagation (i.e., not fulfilling the properties of either phase I or III), phase III was a distinct more than 2 min long period of coordinated powerful contractions at a frequency over 10 per min, propagating in the aboral direction.

Study protocol

The protocol was designed as a randomized crossover study where the subjects acted as their own controls and participated on two separate study days. The AT1R antagonist candesartan (16 mg) was given orally at 10 pm the night before one of the study days, the other study day served as time control. All experiments were run at the same time in the morning and consisted of the recording of fasted motor activity with at least one MMC phase III complex followed by a pharmacologically triggered phase III-like motility complex. An erythromycin infusion, 5 mg/ml given intravenously at a rate of 60 ml/h during 10 min, resulting in a total dose of 50 mg was started 20 min after the end of the spontaneous phase III, thus in the beginning of phase II of the MMC, to trigger a phase III-like motility complex [8].

Data analysis

The variable of primary interest was transmucosal PD in the descending duodenum. The recording channel positioned immediately before the fluoroscopically identified pars horizantalis (corresponding to side hole three Supplementary Figure S1) was used as a rule. To simplify interindividual comparisons, the following definitions were made: PD during phase I: the mean PD during 10 min after ending of phase III, PD during phase III: maximal PD recorded during phase III motor activity. The phase III associated PD elevation was defined as Δ value between PD phase III and PD phase I (see results). PD during pharmacologically triggered phase III-like motility: the maximal PD obtained during erythromycin induced motility. The phase III-like PD elevation was defined as Δ value between PD during the phase III-like motility and the preceding phase I as defined above to calculate the PD elevation.

Western blots and immunohistchemistry of human duodenal mucosa.
Figure 1
Western blots and immunohistchemistry of human duodenal mucosa.

Panel A images of the western blots demonstrating immunoreactivity to AT1R, AT2R, AGT, renin, ACE, and NEP in the human duodenal mucosa of three study subjects. Panel B (axial, along villi) and C (cross-sectional, crypt region); small cells in the duodenal mucosa with immunreactivity toward angioteninsinogen.

Figure 1
Western blots and immunohistchemistry of human duodenal mucosa.

Panel A images of the western blots demonstrating immunoreactivity to AT1R, AT2R, AGT, renin, ACE, and NEP in the human duodenal mucosa of three study subjects. Panel B (axial, along villi) and C (cross-sectional, crypt region); small cells in the duodenal mucosa with immunreactivity toward angioteninsinogen.

Ussing chamber experiments

A total of 17, overnight fasted, healthy volunteers (mean age: 25 years; range: 19–31 years; nine females) underwent duodenoscopy with sampling of mucosal biopsies that were immediately immersed in ice-cold oxygenated (95% O2 and 5% CO2) Krebs solution (118 mM NaCl, 4.69 mM KCl, 2.52 mM CaCl2, 1.16 mM MgSO4, 1.01 mM NaH2PO4, 25 mM NaHCO3, and 11.1 mM glucose) and transported to the laboratory. The biopsied duodenal mucosa was mounted in mini-Ussing chambers with an insert with a square area of 0.031 cm2 (Warner instruments, Hamden, CT, U.S.A.). After mounting, each half chamber was filled with 5 ml Krebs solution, bathing both the mucosal and serosal side of the specimen. The Krebs solution was maintained at 37°C and continuously oxygenated with 95% O2 and 5% CO2 and stirred by gas flow in the chambers. Three to six preparations could be retrieved from each individual. PD was measured with a pair of matched calomel electrodes (REF401, Radiometer analytical, Denmark). Square wave analysis was used to determine epithelial electrical resistance (Rep) as described in detail by Björkman et al. [9]. The epithelial ion current (Iep) was then calculated using Ohm’s law, where I = U/R (i.e., Iep = PD/Rep). Briefly, this technique is based on the concept that the epithelium acts as a capacitor and resistor coupled in parallel. Separate trains of short current pulses induce a voltage response in the tissue and charge the epithelial capacitor which, when the current ends, is gradually discharged. The epithelial voltage response obtained from the discharge curve and the known magnitude of the applied current are used to calculate Rep. The data were collected using specially constructed amplifiers and a software developed in LabView (National Instruments, Austin, TX, U.S.A.).

Study protocol

Following a 20-min equilibration period, baseline values were recorded. After serosal addition of 10−6 M of the AT1R selective antagonists losartan, and/or the AT2R selective antagonists PD123319, or the AT2R agonist C21 (10−6 M) or vehicle, an additional stabilization period of 10 min was allowed before the second values were recorded. AngII was then added to the serosal solution at a concentration of 10−6 M and a third recording was done after another 20 min. At least one Ussing-preparation from each study individual served as a time control. Effect was expressed as percent change from the individual baseline. The concentrations of the losartan, PD123319 and C21 were based on the results from a previous study performed on human esophageal preparations [10].

Drugs and chemicals

The in vivo study used the AT1R antagonist candesartan (Atacand®, AstraZeneca, Mölndal, Sweden) and erythromycin (Abboticin®, Abbott, IL, U.S.A.). In the in vitro study, the following chemicals used, losartan (Merck, NJ, U.S.A.) and AngII and PD123319 (Sigma Chem. Inc., St Louis, MO, U.S.A.) as well as C21 (Department of Medicinal Chemistry, BMC, Uppsala University, Sweden) all chemicals in the in vitro study, were dissolved in Krebs solution.

Data analysis and statistics

Statistics were analyzed using the SPSS statistical software (SPSS Inc, Chicago, IL, U.S.A.). Statistical significance was set at P<0.05. Analyses in the in vivo experiment were conducted using Wilcoxon’s singed rank test and analyses of the in vitro experiment was done using Kruskal Wallis to assert potential group differences and Mann-Whitney U-test to contrast between groups.

Results

RAS components were abundantly distributed in the human duodenal mucosa

The proteins of AT1R, AT2R, AGT, renin, ACE, and NEP were identified by western blotting in all samples of duodenal mucosa (Figure 1, panel A). A small number of epithelial cells with AGT-immunoreactivity were surprisingly found in the crypts and along the villi (Figure 1, panel B and C). AT1R was shown in the apical area of the enterocytes by immunofluorescence staining (Figure 2A), whereas AT2R was shown both apically and intracellulary in enterocytes (Figure 2B). In a separate set of in serial sections immunostaining for AT1R or AT2R were combined with the nerve specific marker S-100. These slides showed that AT1R staining was associated with submucosal nerve plexa (Figure 2C,D), whereas AT2R was not (data not shown). The presence of the hormone AngII in the duodenal biopsies was determined to be 4.4 ± 0.8 pg/mg tissue protein (n=9).

Immunoreactivity to AT1R and AT2R of the RAS in the human duodenal mucosa

Figure 2
Immunoreactivity to AT1R and AT2R of the RAS in the human duodenal mucosa

Panel A immunostaining for AT1R and panel B for AT2R. Panel C shows immunostaining for the nerve specific S-100. Panel D shows a section cut in series with panel C with immunostaining for AT1R.

Figure 2
Immunoreactivity to AT1R and AT2R of the RAS in the human duodenal mucosa

Panel A immunostaining for AT1R and panel B for AT2R. Panel C shows immunostaining for the nerve specific S-100. Panel D shows a section cut in series with panel C with immunostaining for AT1R.

AT1R blockade reinforced the mucosal electrogenic response to endogenous motor complexes, but not to erythromycin induced motility

The interdigestive GI motility pattern (i.e., the migrating motility complex, MMC) is associated with changes in mucosal ion transport as assessed by recording of the transmucosal PD. It was confirmed that PD increased (i.e., the lumen got more negative) during phase III of MMC in all individuals (illustrated in one subject in Figure 3). Following a single peroral dose of the AT1R antagonist candesartan, the phase III associated PD elevation was significantly larger (P-value = 0.028) than during the study day without treatment (Figure 4). Erythromycin (total dose 50 mg i.v.) was administered intravenously 20 min after the end of the spontaneous phase III, thus usually late in phase I or in the beginning of phase II of the MMC. Within 10 min this erythromycin infusion triggered a propulsive duodenal motor complex with characteristics resembling endogenous MMC phase III. These pharmacologically triggered motor complexes were also associated with increased transmucosal PD with an order of magnitude similar to those associated with MMC phase III. In contrast with the latter, the erythromycin-induced PD responses were not altered after candesartan intake (Figure 4).

A 60-min original recording in one healthy volunteer

Figure 3
A 60-min original recording in one healthy volunteer

Intraluminal pressures (upper panel) and transmucosal PD (middle panel) along the duodenum (recording positions according to Supplementary Figure S1). The lower panel is a closeup of the MMC phase III-associated elevation of transmucosal PD (higher values corresponds to increased lumen negativity) in recording 6 with Δ value indicated between maximal PD during high motor activity and the preceding baseline PD.

Figure 3
A 60-min original recording in one healthy volunteer

Intraluminal pressures (upper panel) and transmucosal PD (middle panel) along the duodenum (recording positions according to Supplementary Figure S1). The lower panel is a closeup of the MMC phase III-associated elevation of transmucosal PD (higher values corresponds to increased lumen negativity) in recording 6 with Δ value indicated between maximal PD during high motor activity and the preceding baseline PD.

Net differences in transmucosal potential difference (∆TMPD) in the duodenum of healthy male volunteers (n=7) between phase I and the maximal PD obtained during the propagating high motor activity complex either occurring spontaneously (=phase III of the MMC, left) or following the prokinetic erythromycin (50 mg intravenously, right)

Figure 4
Net differences in transmucosal potential difference (∆TMPD) in the duodenum of healthy male volunteers (n=7) between phase I and the maximal PD obtained during the propagating high motor activity complex either occurring spontaneously (=phase III of the MMC, left) or following the prokinetic erythromycin (50 mg intravenously, right)

For each participant recordings were performed on two separate study days with or without intake of the AT1-receptor antagonist candesartan (16 mg single dose per os). Box whiskers showing median, quartiles, min, and max values. Wilcoxon signed rank test.

Figure 4
Net differences in transmucosal potential difference (∆TMPD) in the duodenum of healthy male volunteers (n=7) between phase I and the maximal PD obtained during the propagating high motor activity complex either occurring spontaneously (=phase III of the MMC, left) or following the prokinetic erythromycin (50 mg intravenously, right)

For each participant recordings were performed on two separate study days with or without intake of the AT1-receptor antagonist candesartan (16 mg single dose per os). Box whiskers showing median, quartiles, min, and max values. Wilcoxon signed rank test.

Fasting duodenal motility per se was not influenced by AT1R blockade

The differential effect by the AT1R antagonist on PD following endogenous and pharmacologically induced motor complexes prompted a detailed analysis of the manometric recordings. Contraction frequency and amplitude, as well as mean intraluminal pressure of each of the three MMC phases (for definitions see methods) did not differ between study days with and without candesartan. Furthermore, neither the endogenous phase III of MMC nor the erythromycin induced high motor activity complexes were significantly influenced by the presence of absence of candesartan (Table 1).

AngII stimulated duodenal mucosal electrogenic transport in vitro via activation of AT2R

After mounting of the duodenal mucosa and leaving a 20-min equilibration period, baseline values of Rep (7.43 ± 0.58 Ohm cm2) and Iep (327 ± 21 µA/cm2) were recorded (both given as mean ± S.E.M.; 60 Ussing-preparations with mucosae from 17 individuals). There were no differences between groups during baseline. In time control preparations, these variables never changed significantly during the course of the experiment. Furthermore, addition of AngII alone did not change any of the recorded values. However, addition of AngII in the presence of the AT1R antagonist losartan raised Iep significantly and this response was sensitive to the AT2R selective antagonist PD123319 suggesting an effect via the unopposed AT2R (Figure 5A). AngII in presence of only PD123319 had no effect. Further support for an involvement of AT2R was obtained by use of the AT2R selective agonist C21. Addition of C21 significantly increased Iep compared with time controls, as well as compared with preparations treated with the AT2R antagonist PD123319 alone or in combination with losartan (Figure 5A). Neither losartan nor PD123319 per se changed baseline Iep (data not shown in figure). Furthermore, none of the drugs in any combination influenced Rep (Figure 5B).

Human duodenal mucosal samples mounted in Ussing chambers

Figure 5
Human duodenal mucosal samples mounted in Ussing chambers

Data are given as net differences in epithelial current (Iep, panel A) and resistance (rep, panel B) as percent of baseline. Data are recorded 30 min after administration of compounds at a concentration of 10−6 M. * denotes significant difference between the time controls and the Losartan + AngII or C21 treated samples (P=0.008 and P=0.012, respectively). #Indicate significant difference (P<0.05) between the samples marked with lines. The boxplots show median, interquartiles, and min/max values of at least six preparations. The results were analyzed with Kruskal–Wallis test and differences were contrasted by use of Mann–Whitney U-test.

Figure 5
Human duodenal mucosal samples mounted in Ussing chambers

Data are given as net differences in epithelial current (Iep, panel A) and resistance (rep, panel B) as percent of baseline. Data are recorded 30 min after administration of compounds at a concentration of 10−6 M. * denotes significant difference between the time controls and the Losartan + AngII or C21 treated samples (P=0.008 and P=0.012, respectively). #Indicate significant difference (P<0.05) between the samples marked with lines. The boxplots show median, interquartiles, and min/max values of at least six preparations. The results were analyzed with Kruskal–Wallis test and differences were contrasted by use of Mann–Whitney U-test.

Discussion

The present study confirms the presence and activity of a mucosal RAS in the duodenum of man. Signs of a local AGT-store, the angiotensin-generating enzymes renin, ACE and NEP, as well as the hormone AngII, were all detected in the duodenal mucosa. The classical view is that the angiotensin precursor AGT, which is essential for the functionality of the RAS, is produced by the liver and delivered by the blood stream. Hepatic origin is still regarded to be quantitatively most important, but AGT can be produced locally in many tissues [3]. The picture with a local AGT store and angiotensin-generating enzymes indicates capacity for local generation of the principal RAS-mediator AngII, and possibly also other bioactive ATG-fragments like Ang III, Ang IV, and Angiotensin(1–7). These features, together with the expression of both subtypes of AngII-receptors, suggest the presence of a complete local RAS in the human duodenal mucosa. This location has not been shown previously in man but is supported by animal studies [6].

Immunohistochemistry showed that the AT1R were distinctly present at the apical border of enterocytes, whereas the AT2R were distributed both on the apical border and in the intracellular space. Because this picture is a snapshot of the situation after an overnight fast, it is probable that the picture can change following, for example, food intake, stress reactions, etc. It raises possibility of translocation of AT1R and AT2R in a manner like the one described regarding rat internal anal sphincter 11]. It remains to be investigated if such an AngII receptor replacement is the case also in human enterocytes.

Immunoreactivity to AT1R was also associated with nerve structures (S100-sensitive) in the stroma and the submucosa, some of those penetrating into the villi. We found no functional consequence of this in the fasting state. The endoscopic biopsies reached the muscularis mucosa only so it was not possible to judge the innervation pattern in deeper levels.

To test the physiological relevance of a duodenal mucosal RAS in man in vivo, we used MMC phase-III-associated elevation of duodenal transmucosal PD. The MMC is a motility pattern occurring in the fasting state and phase III of the MMC is a period of coordinated propulsive contractions, moving as an activity complex from the stomach or the duodenum toward the distal small intestine [12]. This is contrasted with the fed state that occurs immediately after food intake and per definition involves continuous high-pressure contractions. We considered that the fed state with intubation and food intake as a potentially risky situation for the participant. We therefore restrict the findings to the fasting state and effects on mucosal functions.

The function of the MMC propagating ‘motility train’ is very probably to clean the GI lumen between meals. In addition to wall motor activity several, GI secretory processes are simultaneously and shortly activated, for example, gastric acid, bile and pancreatic secretions [13]. The MMC and the associated secretory phenomena are integrated by the enteric nervous system into a complete secretomotor programme [13]. The duodenal MMC phase III associated elevation of transmucosal PD depends on an increased lumen electrical negativity, probably due to an accelerated mucosal anion secretion (further discussed below) [14]. Furthermore, the magnitude of the increased PD change is closely related to intraluminal pressure oscillations and frequency of the wall contractions of interdigestive MMC [15,16]. In the present study, the elevation of the PD following MMC phase III was significantly larger after a per-oral dose of candesartan with confirmed AT1R blockade [17,18].

The manometric properties of the various phases of the MMC and pharmacological simulation of phase III motor activity did not differ following the AT1R blockade. The failure of AT1R antagonist to show effect on PD increase by erythromycin is intriguing as effect on motility was as prominent as by the endogenous MMC phase III (see Table 1). Erythromycin-activated motilin receptors have, compared with the endogenous MMC phase III, probably a different point of action [19]. Details await further exploration, but the findings indicate that activation of motor activity may be dissociated from secretion-promoting signals targetting the mucosa. One reasonable interpretation is that the AT1R antagonist selectively removes an AngII mediated inhibition on the mucosal electrogenic transport following MMC phase III.

Transmucosal PD in vivo consists of several physiologically relevant components. The MMC-phase III associated elevation of the lumen-negative transmural PD can be explained either as an increased electrogenic secretion of anions (mainly CFTR mediated Cl or HCO3 secretion), or as an increased absorption of cations (e.g., SGLT-1 mediated co-transport of Na+), or as an increased mucosal electrical resistance (i.e., reduced epithelial paracellular conductivity), as well as combinations thereof. An in vitro approach was applied to allow separation of these factors, having also the advantage of avoiding influences from non-GI organs. The absolute values of the electrophysiological variables recorded from duodenal mucosae in Ussing chambers vary considerably in the literature. For example, reported baseline epithelial current (Iep) varies between <10 and 350 uA/cm2 [20–25]. The main reason for this variability is methodological, caused by technical set up, the selection of study subjects, biopsy sites as well as measuring principles [9,16]. Using each individual as its own control is therefore the most reliable way of measuring differences. In the present study on duodenal mucosae from healthy volunteers, the main finding was that AngII in the presence of the AT1R selective antagonist losartan increased the epithelial current (Iep). Because this response was sensitive to the AT2R selective antagonist PD123319, the results indicate that the effect was due to activation of the unopposed AT2R. It is a well-known pharmacological fact that blockade of AT1R uncovers AT2R mediated effects in many tissues, for example, in the esophageal mucosa [10]. In the present study the AT2R selective agonist C21 also increased the epithelial current, thus adding support for an AT2R-mediated action. Interestingly, the epithelial electrical resistance was not influenced by any of these drugs strongly indicating that AngII targets the epithelial secretory processes, either a net cation-absorption or an anion secretion, rather than altering paracellular conductance [9].

The molecular background to duodenal electrogenic transport is complex and can be expected to involve ions channels like CFTR (Cl or HCO3 secretion) and/or sodium co-transporters like SGLT1 (Na+-glucose absorption). Both these structures are present in human duodenal mucosa (see example, Supplementary Figure S2) and have been reported to be regulated in relation to the RAS in animal studies. Duodenal mucosal alkaline secretion has been inhibited following experimental bleeding-induced hypovolemia in anesthetised rats [5,6,26,27]. In these experiments, it was found that AngII via AT1R and a sympathoadrenergic mechanisms reduced a lumen-directed net HCO3 secretion [5]. Activation of AT2R, on the other hand, increased this secretion thus very similar to what we now report [6,29]. In addition, SGLT1-dependent glucose-Na+ uptake across the rat small jejunal brush border appears to be inhibited through a local RAS involving AT1R activation [29]. Such an AT1R inhibitory mechanism appears to exist in the human jejunum as well, but in this tissue it occurs together with an opposing (stimulatory) effect via AT2R [30]. AngII modulation of SGLT1-mediated electrogenic sodium-glucose absorption also in the human duodenum thus seems probable.

In conclusion, the present study demonstrates immunostaining as well as pharmacological data speaking strongly in favor of the presence of a local RAS in the human duodenal mucosa. This system has capacity to influence mucosal transporting capacity involving AT2R, but exact organization and physiological relevance remains to be elucidated.

Clinical perspectives

  • AngII is a well-known blood borne signaling molecule (hormone) regulating blood pressure and kidney functions. AngII can also be formed locally in many tissues, but it is unclear if this is the case in the human gut.

  • The study shows that AngII is synthetized in the duodenum and that it may influence mucosal functions via AT2R, but not the interdigestive muscular activity of the intestinal wall.

  • It seems that duodenal AngII acts on the upper gut in man and thereby can contribute to disease states both in the gut and in the homeostatic principles of the body.

Acknowledgements

Excellent technical assistance from Sören Lundberg, Eva Lotta Een, and My Engström is gratefully acknowledged. Some of the here presented in vivo results have previously been shown in a preliminary form [31].

Funding

The present study received financial support by the Swedish Research Council [grant number K2010-55X-21432-01-2], the University of Gothenburg and Sahlgrenska University Hospital [ALF-grant number 147231], the Gothenburg Medical Association (to E.S.).

Author contribution

L.F. and E.S. designed the study, analyzed as well as interpreted data and prepared the manuscript. E.S. and A.C. performed the experiments. A.E. and P.H. contributed with endoscopies and histopathological analyses. All authors critically revised and approved the final manuscript.

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Abbreviations

     
  • ACE

    angiotensin-converting enzyme

  •  
  • AGT

    angiotensinogen

  •  
  • AngII

    Angiotensin II

  •  
  • AT1R

    Angiotensin II type 1 receptor

  •  
  • AT2R

    Angiotensin II type 2 receptor

  •  
  • GI

    gastrointestinal

  •  
  • Iep

    epithelial ion current

  •  
  • MMC

    migrating motility complex

  •  
  • NEP

    neprilysin

  •  
  • PD

    potential difference

  •  
  • RAS

    renin-angiotensin system

  •  
  • Rep

    epithelial electrical resistance

References

References
1.
Walker
M.M.
,
Potter
M.
and
Talley
N.J.
(
2018
)
Eosinophilic gastroenteritis and other eosinophilic gut diseases distal to the oesophagus
.
Lancet Gastroenterol. Hepatol.
3
,
271
280
[PubMed]
2.
van Baar
A.C.G.
,
Nieuwdorp
M.
,
Holleman
F.
,
Soeters
M.F.
,
Groen
A.K.
and
Bergman
J.J.G.H.M.
(
2018
)
The duodenum harbors a broad untapped therapeutic potential
.
Gastroenterology
154
,
773
777
[PubMed]
3.
Paul
M.
,
Mehr
A.P.
and
Kreutz
R.
(
2006
)
Physiology of local renin-angiotensin systems
.
Physiol. Rev.
86
,
747
803
[PubMed]
4.
Allen
A.
and
Flemström
G.
(
2005
)
Gastroduodenal mucus bicarbonate barrier: protection against acid and pepsin
.
Am. J. Physiol. Cell Physiol.
288
,
C1
C19
[PubMed]
5.
Johansson
B.
,
Holm
M.
,
Chen
L.
,
Pettersson
A.
,
Jönson
C.
and
Fändriks
L.
(
1997
)
ANG II prolongs splanchnic nerve-mediated inhibition of duodenal mucosal alkaline secretion in the rat
.
Am. J. Physiol.
273
,
R942
R946
[PubMed]
6.
Johansson
B.
,
Holm
M.
,
Ewert
S.
,
Casselbrant
A.
,
Pettersson
A.
and
Fändriks
L.
(
2001
)
Angiotensin II type 2 receptor-mediated duodenal mucosal alkaline secretion in the rat
.
Am. J. Physiol. Gastrointest. Liver Physiol.
280
,
G1254
G1260
[PubMed]
7.
Bradford
M.M.
(
1976
)
A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding
.
Anal. Biochem.
72
,
248
254
[PubMed]
8.
Medhus
A.W.
,
Bondi
J.
,
Gaustad
P.
and
Husebye
E.
(
2000
)
Low-dose intravenous erythromycin: effects on postprandial and fasting motility of the small bowel
.
Aliment. Pharmacol. Ther.
14
,
233
240
[PubMed]
9.
Björkman
E.
,
Casselbrant
A.
,
Lundberg
S.
and
Fändriks
L.
(
2012
)
In vitro assessment of epithelial electrical resistance in human esophageal and jejunal mucosae and in Caco-2 cell layers
.
Scand J. Gastroenterol.
47
,
1321
1333
[PubMed]
10.
Casselbrant
A.
,
Edebo
A.
,
Wennerblom
J.
,
Lonroth
H.
,
Helander
H. F.
and
Vieth
M.
(
2007
)
Actions by angiotensin II on esophageal contractility in humans
.
Gastroenterology
132
,
249
260
[PubMed]
11.
de Godoy
M.A.F.
and
Rattan
S.
(
2006
)
Translocation of AT1- and AT2-receptors by higher concentrations of Angiotensin II in the smooth muscle cells of rat internal anal sphincter
.
J. Pharmacol. Exp. Therap.
319
,
1088
1095
12.
Sarna
S.K
(
1985
)
Cyclic motor activity; migrating motor complex
.
Gastroenterology
89
,
894
913
[PubMed]
13.
Sjövall
H.
(
2011
)
Meaningful or redundant complexity - mechanisms behind cyclic changes in gastroduodenal pH in the fasting state
.
Acta. Physiol. (Oxf.)
201
,
127
131
[PubMed]
14.
Mellander
A.
,
Abrahamsson
H.
and
Sjövall
H.
(
1995
)
The migrating motor complex–the motor component of a cholinergic enteric secretomotor programme
.
Acta Physiol. Scand.
154
,
329
341
[PubMed]
15.
Mellander
A.
,
Järbur
K.
and
Sjövall
H.
(
2000
)
Pressure and frequency dependent linkage between motility and epithelial secretion in human proximal small intestine
.
Gut
46
,
376
384
[PubMed]
16.
Larsson
M.H.
,
Simrén
M.
,
Thomas
E.A.
,
Bornstein
J.C.
,
Lindström
E.
and
Sjövall
H.
(
2007
)
Elevated motility-related transmucosal potential difference in the upper small intestine in the irritable bowel syndrome
.
Neurogastroenterol. Motil.
19
,
812
820
[PubMed]
17.
Azizi
M.
,
Bissery
A.
,
Lamarre-Cliche
M.
and
Ménard
J.
(
2004
)
Integrating drug pharmacokinetics for phenotyping individual renin response to angiotensin II blockade in humans
.
Hypertension
43
,
785
790
[PubMed]
18.
Gleiter
C.H.
and
Morike
K.E.
(
2002
)
Clinical pharmacokinetics of candesartan
.
Clin. Pharmacokinet.
41
,
7
17
[PubMed]
19.
Sanger
G.J.
and
Furness
J.B.
(
2016
)
Ghrelin and motilin receptors as drug targets for gastrointestinal disorders
.
Nat. Rev. Gastroenterol. Hepatol.
13
,
38
48
[PubMed]
20.
Pratha
V.S.
,
Thompson
S.M.
,
Hogan
D.L.
,
Paulus
P.
,
Dreilinger
A.D.
,
Barrett
K.E.
et al.
(
1998
)
Utility of endoscopic biopsy samples to quantitate human duodenal ion transport
.
J. Lab. Clin. Med.
132
,
512
518
[PubMed]
21.
Hosokawa
M.
,
Tsukade
H.
,
Fukuda
K.
,
Oya
M.
,
Onomura
M.
,
Nakamura
H.
et al.
(
1999
)
Troglitazone inhibits bicarbonate secretion in rat and human duodenum
.
J. Pharmacol. Exp. Ther.
290
,
1080
1084
[PubMed]
22.
Larsen
R.
,
Mertz-Nielsen
A.
,
Hansen
M.B.
,
Poulsen
S.S.
and
Bindslev
N.
(
2001
)
Novel modified Ussing chamber for the study of absorption and secretion in human endoscopic biopsies
.
Acta. Physiol. Scand.
173
,
213
222
[PubMed]
23.
Mellander
A.
,
Järbur
K.
,
Hemlin
M.
and
Sjövall
H.
(
2001
)
Effects of motility on epithelial transport in the human descending duodenum
.
Acta. Physiol. Scand.
172
,
69
80
[PubMed]
24.
Matysiak-Budnik
T.
,
Candalh
C.
,
Dugave
C.
,
Namane
A.
,
Cellier
C.
,
Cerf-Bensussan
N.
et al.
(
2003
)
Alterations of the intestinal transport and processing of gliadin peptides in celiac disease
.
Gastroenterology
125
,
696
707
[PubMed]
25.
Troeger
H.
,
Epple
H.J.
,
Schneider
T.
,
Wahnschaffe
U.
,
Ullrich
R.
,
Burchard
G.D.
et al.
(
2007
)
Effect of chronic Giardia lamblia infection on epithelial transport and barrier function in human duodenum
.
Gut
56
,
328
335
[PubMed]
26.
Åneman
A.
,
Pettersson
A.
,
Eisenhofer
G.
,
Friberg
P.
,
Holm
M.
,
von Bothmer
C.
et al.
(
1997
)
Sympathetic and renin-angiotensin activation during graded hypovolemia in pigs: impact on mesenteric perfusion and duodenal mucosal function
.
Shock
8
,
378
384
[PubMed]
27.
Chen
L.
,
Holm
M.
,
Fändriks
L.
,
Pettersson
A.
and
Johansson
B.
(
1997
)
ACE inhibition by enalaprilate stimulates duodenal mucosal alkaline secretion via a bradykinin pathway in the rat
.
Dig. Dis. Sci.
42
,
1908
1913
[PubMed]
28.
Aihara
E.
,
Kagawa
S.
,
Hayashi
M.
and
Takeuchi
K.
(
2005
)
ACE inhibitor and AT1 antagonist stimulate duodenal HCO3- secretion mediated by a common pathway - involvement of PG, NO and bradykinin
.
J. Physiol. Pharmacol.
56
,
391
406
[PubMed]
29.
Wong
T.P.
,
Debnam
E.S.
and
Leung
P.S.
(
2009
)
Diabetes mellitus and expression of the enterocyte renin-angiotensin system: implications for control of glucose transport across the brush border membrane
.
Am. J. Physiol. Cell Physiol.
297
,
C601
C610
[PubMed]
30.
Casselbrant
A.
,
Malinauskas
M.
,
Marschall
H.U.
,
Wallenius
V.
and
Fändriks
L.
(
2015
)
Angiotensin II exerts dual actions on sodium-glucose transporter 1-mediated transport in the human jejunal mucosa
.
Scand. J. Gastroenterol.
50
,
1068
1075
[PubMed]
31.
Fändriks
L.
(
2011
)
The renin-angiotensin system and the gastrointestinal mucosa
.
Acta. Physiol. (Oxf.)
201
,
157
167
[PubMed]