Intragastric infusion of the bitter tastant quinine suppresses hormone release and antral motility during the fasting state in healthy female volunteers

Intragastric administration of the bitter tastant denatonium benzoate inhibits the increase of motilin plasma levels and antral contractility. While these findings suggest that gastrointestinal bitter taste receptors could be new targets to modulate gastrointestinal motility and hormone release, they need confirmation with other bitter receptor agonists. The primary aim was to evaluate the effect of intragastric administration of the bitter tastant quinine‐hydrochloride (QHCl) on motilin and ghrelin plasma levels. Secondly, we studied the effect on interdigestive motility.


| INTRODUCTION
In 1916 Carlson first reported the inhibitory effect of intragastric bitter administration on fasting gastric contractility in man. 1 Eighty years later, Höfer et al. provided the first evidence for gastrointestinal taste signaling by reporting the expression of α-gustducin in the gut of rats. 2 Further research showed that the bitter taste receptors, taste 2 receptors (TAS2Rs), are expressed in the human gastrointestinal tract. 3,4 Their presence has stimulated research to modulate gastrointestinal physiology via the administration of bitter tastants.
In rats and mice, direct intraluminal administration of bitter agonists inhibited ongoing ingestive behavior and gastric emptying. [4][5][6] Bitter sham feeding significantly reduced gastric emptying of a liquid meal in humans. 7 Gastric emptying of a solid meal was, however, not affected when a bitter tastant was administered directly into the stomach of healthy volunteers. 8 Intragastric administration of denatonium benzoate (DB) was shown to inhibit gastric accommodation and subsequently food intake. 4 A similar response was observed after the intragastric administration of quinine-hydrochloride (QHCl). 9 In vitro studies have shown that DB increases the release of cholecystokinine (CCK), glucagon-like peptide-1 (GLP-1), peptide YY (PYY), and insulin. [10][11][12] A similar result has been described for quinine on the release of GLP-1. 11 However, results obtained in vivo report contradictory results. Van Avesaat et al. 13 reported no significant change in CCK, GLP-1, and PYY after intraduodenal infusion of quinine, while another study did report an increase in postprandial CCK release after intradoudenal quinine administration. 14 Ghrelin release was not altered after intragastric administration of DB in humans. 8 A similar result was obtained for ghrelin secretion after intraduodenal administration of quinine. 14 Our group has recently shown that intragastric administration of DB inhibited the increase in motilin plasma levels and reduced the occurrence of gastric phase III contractions of the migrating motor complex (MMC) in man. 8 While these findings suggest that gastrointestinal bitter taste receptors could be new targets to modulate gastrointestinal motility and hormone release, they need confirmation with other bitter receptor agonists. Therefore, the aim of this study was to evaluate the effect of intragastric administration of QHCl on gastrointestinal hormone release and motility during the interdigestive state.

| Ethical approval
This study was approved by the Medical Ethics Committee of the Leuven University Hospital, Leuven, Belgium, and performed in full accordance with the Declaration of Helsinki. This trial was registered at clinicaltrials.gov as NCT02759926.

| Subjects
Volunteers were eligible to participate if they were healthy, aged between 18 and 60 y old, had a BMI (in kg/m²) ranging from 19.0 to 25.0, and were recruited from an existing volunteer database in our group. Exclusion criteria were gastrointestinal diseases, abdominal surgery (appendectomy allowed), psychiatric illnesses, and usage of drugs affecting the gastrointestinal tract or central nervous system.
Only female participants were included because it has been reported that sensitivity to intragastrically administered bitter tastants differs between men and women. 8 Written informed consent was obtained from all volunteers before the start of the study. All subjects were studied after an overnight fast of 12 h and were asked to refrain from smoking at least 1 h before the start of the study.
Solutions of QHCl were prepared in Milli-Q water. The stock concentration for intragastric administration was 100 mM. A volume of 0.1 ml/kg bodyweight was administered. The dosage of QHCl (10 μmol/kg) was chosen based on its inhibitory effect on gastric accommodation in healthy volunteers. 9 Vehicle was given during the placebo condition in a volume of 0.1 ml/kg bodyweight.

| Study design
After an overnight fast of 12 h, antroduodenal motility was measured until the occurrence of the first phase III contraction. Twenty minutes after the end of phase III, either placebo or QHCl (10 μmol/kg bodyweight) was administered intragastrically in a single-blind randomized fashion. Intragastric administration was done through a nasogastric feeding tube (Flocare, Nutricia, Bornem, Belgium). The feeding tube was placed with its tip in the upper part of the stomach and the position was checked with fluoroscopy prior to the start of the study.
The measurement continued for another 2 h after the administration.
Blood samples were collected every 10 min with the baseline sample taken 10 min prior to administration.

| Antroduodenal manometry
Activity of the MMC was measured using a high-resolution solid- • Gastric bitter taste receptors could be new targets to modulate gastrointestinal motility in health and disease. analysis software v2.0) as described previously. 15 Phases of the MMC were identified based on standardized definitions. 16,17 Motility index (MI) was calculated as follows: (number of contractions*average amplitude contractions*average duration contractions)/5 min. 15,18,19 An average MI was calculated for the antrum and duodenum by averaging the MI of six consecutive channels.

| Blood collection
Blood samples were collected via an intravenous catheter. Samples for motilin and ghrelin were collected in lithium heparin and EDTA tubes each containing 500 kIU/ml aprotinin (Roche Applied Science, Penzberg, Germany). Samples were centrifuged at 4°C for 10 min at The time interval until the occurrence of phase III was compared between the two conditions using Wilcoxon signed rank test (GraphPad). The percentage origin of phase III was compared between placebo and QHCl with McNemar's test (GraphPad). For both hormone plasma concentrations and antral motility, the peak values were determined to correlate the time of peak occurrence between antral motility and hormone concentrations. Spearman's rank correlation coefficient (one-tailed) was used for both conditions (GraphPad).
None of the subjects reported any adverse events after the administration of QHCl.

| Change in hormone plasma levels after quininehydrochloride administration
The effect of intragastric QHCl administration on the change of plasma concentrations of motilin and ghrelin (total and octanoylated) was

| Change in antroduodenal motility after quininehydrochloride administration
Antroduodenal motility was measured using high-resolution manometry ( Figure 2). The average duration until the occurrence of phase III prior to administration did not differ between placebo and QHCl (63

| Correlation between motility and peptide plasma levels
There was a significant positive correlation (r=.58; P=.04) between the occurrence of antral peak activity and the peak concentration of motilin during the placebo administration ( Figure 4A)  administration ( Figure 4B). There was however a significant positive correlation between the occurrence of antral peak activity and the peak concentration of total ghrelin (r=.58; P=.04).

| DISCUSSION
Our study showed that intragastric administration of QHCl suppresses interdigestive gastrointestinal motor function and release of motilin and ghrelin. These results confirm our previous findings that intragastric bitter administration alters interdigestive functions of the gastrointestinal tract. 8 Quinine binds to nine different TAS2Rs (TAS2R4, TAS2R7,   TAS2R10, TAS2R14, TAS2R39, TAS2R40, TAS2R43, TAS2R44,   TAS2R46). Five of these receptors also bind DB (TAS2R4, TAS2R10, TAS2R39, TAS2R43, TAS2R46). 20 The difference in receptor activation might be an explanation for the different effect of QHCl and DB on ghrelin release. We previously reported that intragastric administration of DB inhibited the release of motilin, but had no effect on ghrelin. 8 Intragastric administration of QHCl however inhibited the release of both motilin and ghrelin. A discrepancy between intragastric DB and QHCl administration on ghrelin release has already been described in mice. 6 However, oral gavage of QHCl had no effect on ghrelin release in mice, while DB caused an increase. These results show that bitter agonists might have different effects in different species.
Our study showed that intragastric administration of 10 μmol/kg QHCl inhibited the release of total (desoctanoyl and octanoyl) ghrelin and reduced the fluctuation of octanoylated ghrelin. Two cell types have been described that secrete ghrelin, a cell type containing both octanoylated and desoctanoylated ghrelin and another cell type containing only desoctanoylated ghrelin. 6 The octanoylation of ghrelin is caused by a post-translational esterification of the third serine residue by ghrelin O-acyltransferase (GOAT). 21  and TAS2R10 has been shown in mucosal tissue from the fundus and small intestine in humans. 22 However, co-expression of TAS2Rs and ghrelin-secreting cells has not been shown to date. Another possibility could be that the transduction pathway after QHCl administration also affects the expression of GOAT causing a discrepancy between total and octanoylated ghrelin. Further research is needed to specify the mechanism by which QHCl influences ghrelin secretion.
A previous study, where quinine (18 mg) was administered intraduodenally, showed that there was no effect on ghrelin release. 14 The dosage given in the current study is higher, which may account for some of the different effects. Moreover, ghrelin-secreting cells are mainly located in the oxyntic mucosa of the stomach, so bypassing the stomach via intraduodenal administration could abolish the effects of quinine on ghrelin secretion. 23 The inhibition of motilin release after intragastric bitter administration is seen after both DB and QHCl administration. Moreover, the correlation between motilin and antral motility is lost after both bitter administrations. 8 Interdigestive antral motility in man is regulated by motilin. 15 The inhibitory effect of bitter administration on antral motility might therefore be mediated by its effect on motilin release. However, in the present study we also found a significant correlation between peak activity of antral motility and octanoylated ghrelin, which was lost after QHCl administration. Although endogenous ghrelin plasma concentrations were not found to fluctuate with the different phases of the MMC, exogenous administration of octanoylated ghrelin is known to induce a gastric phase III contraction. 24 Nevertheless, the effect of desoctanoylated ghrelin was not investigated. In rats, it has been shown that desacyl ghrelin disrupts fasting gastrointestinal motility. 25 In suncus murinus, a small laboratory animal model, ghrelin has been identified as an essential factor for the motilin-induced gastric contractions. 26 A similar synergistic effect between motilin and ghrelin might therefore be present in humans.
After QHCl administration, antral motility seems to be controlled by total ghrelin as the positive correlation with motilin and octanoylated ghrelin was abolished. Although this correlation shows that the timing of total ghrelin correlates with the timing of antral peak activity, it does not say anything about the strength of antral contractility. Moreover, MMC cycle length varies significantly between subjects. 16 In the current study, antral MI varied significantly between subjects, as seen by the high error flags in the placebo condition. This is due to the different onset of phase III contractions between subjects. The variability in antral MI after QHCl administration was diminished compared to placebo. This decrease in variability is probably due to a delay in the occurrence of phase III. However, with the current protocol, we did not consistently continue to measure until the next phase III after administration, so this cannot be confirmed just on the basis of the current study.
In conclusion, our study confirms the inhibitory effect of in-