Advertisement

Fatty acid chain length determines cholecystokinin secretion and effect on human gastric motility

      Abstract

      Background & Aims: Fatty acids induce cholecystokinin (CCK) secretion and modify gastric motility, but the chain length requirements for these effects are not known. Nor is it clear whether the effects of fatty acids on gastric motility in humans are CCK mediated or directly exerted. The aim of this study was to determine the role of fatty acyl chain length in CCK secretion and in influencing gastric motility. Methods: Fatty acids were infused into the upper gut in healthy volunteers; plasma CCK was determined by radioimmunoassay. Effects of fatty acids on antral contractility were determined by percutaneous ultrasonography; effects on proximal gastric tone were studied during fundal distention. Results: Plasma CCK concentration was consistently and similarly elevated by fatty acids with a chain of 12 carbon atoms or longer, whereas those of 11 or fewer carbon atoms failed to increase plasma CCK. A 12-carbon but not a 10-carbon-long chain fatty acid reduced antral contractile amplitude, an effect that was abolished by loxiglumide (a specific CCK-A receptor antagonist). The 12-carbon fatty acid also reduced proximal gastric tone more than the 10-carbon fatty acid. Conclusions: A highly specific, chain length–sensitive fatty acid recognition system exists in the proximal gut mediating CCK secretion and gastric motility. An additional, probably CCK-independent, effect of fatty acid also regulates proximal gastric tone.
      GASTROENTEROLOGY 1999;116:46-53
      Cholecystokinin (CCK) is released by specialized mucosal endocrine cells, found mainly in the proximal small intestine. It is now accepted that CCK exerts several regulatory functions throughout the gastrointestinal tract. In the upper gut, in addition to stimulating gallbladder emptying and pancreatic enzyme secretion, it also delays gastric emptying and inhibits food intake.
      • Liddle RA.
      Gut peptides: biochemistry and physiology.
      Together, these actions integrate postprandial proximal digestive tract function and serve to optimize nutrient digestion and absorption.
      In humans, a major stimulus for secretion of CCK is intraluminal lipid.
      • Liddle RA.
      Gut peptides: biochemistry and physiology.
      Most studies showing the effect of lipid have used triglyceride emulsions, but it is now clear that free fatty acids, rather than triglycerides per se, stimulate CCK secretion. Pancreatic lipase insufficiency is associated with a poor CCK response to triglyceride, but a normal response to oleic acid,
      • Guimbaud R
      • Moreau JA
      • Bouisson M
      • Durand S
      • Escorrou J
      • Vaysse N
      • Frexinos J.
      Intraduodenal free fatty acids rather than triglycerides are responsible for the release of CCK in humans.
      and lipase inhibitors prevent a CCK response to triglyceride in healthy subjects.
      • Hildebrand P
      • Petrig C
      • Ketterer S
      • Lengsfeld H
      • Beglinger C.
      Inactivation of food-stimulated lipases in humans inhibits exocrine pancreatic secretion by a CCK-dependent feedback mechanism.
      However, the precise molecular determinants for CCK secretion by fatty acids remain unknown; only oleic acid (C18:1) has been shown to be effective.
      • Guimbaud R
      • Moreau JA
      • Bouisson M
      • Durand S
      • Escorrou J
      • Vaysse N
      • Frexinos J.
      Intraduodenal free fatty acids rather than triglycerides are responsible for the release of CCK in humans.
      • Beardshall K
      • Frost G
      • Morarji Y
      • Domin J
      • Bloom SR
      • Calam J.
      Saturation of fat and cholecystokinin release.
      • Olsen O
      • Ainsworth M
      • Schaffalitzky de Muckadell OB
      • Cantor P.
      Effects of oleic acid and oleyl alcohol on cholecystokinin and secretin in plasma and pancreatobiliary secretion.
      In particular, although medium-chain triglycerides seem to be weaker stimuli to CCK secretion than long-chain triglycerides,
      • Hopman WPM
      • Jansen JBMJ
      • Rosenbusch G
      • Lamers CBHW
      Effect of equimolar amounts of long chain triglycerides and medium chain triglycerides on plasma cholecystokinin and gallbladder contraction.
      • Isaacs PET
      • Ladas MD
      • Forgacs IC
      • Dowling RH
      • Ellam SV
      • Adrian TE
      • Bloom SR.
      Comparison of effects of ingested medium- and long-chain triglyceride on gallbladder volume and release of cholecystokinin and other gut peptides.
      the precise fatty acid chain length determining CCK secretion is unknown.
      CCK is an important inhibitory regulator of gastric motility. However, the specific responses of the proximal and distal stomach to CCK remain unclear. In the distal stomach, the normal postprandial pattern is of regular peristaltic contractions at a frequency of 3 per minute.
      • Ahluwalia NK
      • Thompson DG
      • Mamtora H
      • Troncon L
      • Hindle J
      • Hollis S.
      Evaluation of human postprandial antral motor function using ultrasound.
      In the proximal stomach, tone falls after a meal, allowing intragastric volume to increase without a parallel increase in pressure.
      • Ahluwalia NK
      • Thompson DG
      • Barlow J
      • Troncon LEA
      • Hollis S.
      Relaxation responses of the human proximal stomach to distension during fasting and after food.
      Administration of long-chain triglycerides into the upper gut suppresses antral contractility and induces a vagally mediated proximal gastric relaxation.
      • Heddle R
      • Dent J
      • Read NW
      • Houghton LA
      • Toouli J
      • Horowitz M
      • Maddern GJ
      • Downton J.
      Antropyloroduodenal motor responses to introduodenal lipid infusion in healthy volunteers.
      • Feinle C
      • D'Amato A
      • Read NW.
      Cholecystokinin-A receptors modulate gastric sensory and motor responses to gastric distention and duodenal lipid.
      Because CCK infusion produces similar changes
      • Raybould HE
      • Roberts ME
      • Dockray GJ.
      Reflex decreases in intragastric pressure in response to cholecystokinin in rats.
      and because selective CCK-A receptor antagonism inhibits gastric emptying effects of lipids,
      • De Giorgio R
      • Stanghellini V
      • Ricci-Maccarini M
      • Morselli-Labate AM
      • Barbara G
      • Franzoso L
      • Rovati LC
      • Corinaldesi R
      • Barbara L
      • Go VLW.
      Effects of dietary fat on postprandial gastrointestinal motility are inhibited by a cholecystokinin type A receptor antagonist.
      it has been suggested that CCK is the key mediator for the effects of ingested lipids on the stomach. However, an alternative, non-CCK mediated vagal mechanism for lipid-induced effects on gastric motility has been proposed. Melone et al.
      • Melone J.
      Vagal receptors sensitive to lipids in the small intestine of the cat.
      showed a direct effect of fatty acids on duodenal vagal afferent neurons in a feline model. Importantly, this effect seemed to be dependent on fatty acid chain length
      • Melone J.
      Vagal receptors sensitive to lipids in the small intestine of the cat.
      • Melone J
      • Mei N.
      Intestinal effects of the products of lipid digestion on gastric electrical activity in the cat: possible involvement of vagal intestinal receptors sensitive to lipids.
      and was associated with inhibitory changes in gastric myoelectrical activity.
      In this study we sought to resolve two interconnected issues: the effect of fatty acid chain length on CCK secretion and the relationship between fatty acid chain length, CCK secretion, and proximal and distal gastric motor function.

      Materials and methods

       Materials

      Fatty acids were purchased from Sigma-Aldrich (Gillingham, Dorset, England). Saturated fatty acids used were butyric acid (C4), decanoic acid (C10), undecanoic acid (C11), dodecanoic acid (C12), tetradecanoic acid (C14), and octadecanoic acid (C18). The monounsaturate cis-9-octadecanoic acid (oleic, C18:1) was also used. Tween 80 (food grade) was a gift of Quest Industries (Zwijndrecht, Netherlands). Sep-Pak C18 cartridges were purchased from Waters Inc. (Milwaukee WI). Loxiglumide, a specific CCK-A receptor antagonist, was supplied by Rotta Research Laboratorium (Monza, Italy).

       Fatty acid emulsions

      Test emulsions were made as follows: 1.5% (vol/vol) Tween 80 was added to 250 mL phosphate-buffered saline (pH 7.3, 350 mOsm/kg) to form the vehicle solution, which was used either alone (control) or with fatty acid added (test). Vehicle and fatty acid were warmed above the melting point of the relevant fatty acid and added in liquid phase, the two being vigorously mixed while cooling to 37°C. The particle sizes of the emulsions formed in this manner were assessed by use of a Malvern Autosizer RR146 (Malvern Instruments, Malvern, England). Each emulsion comprised a range of particle sizes (20–1000 nm) that did not differ between fatty acids, except for butyric acid, which, because of its high water solubility, did not form particles.

       Subjects

      All subjects were healthy volunteers, recruited from the staff and student population of Hope Hospital and Manchester University. None had history of digestive system disease or surgery, and none were taking medication that might affect gastric motility. All subjects were studied in the morning after an overnight fast. All gave written consent for each study protocol, approval for which had been obtained from the Salford Health Authority Ethical Committee.

       CCK radioimmunoassay

      Venous blood samples were taken from a cannula in an antecubital vein (Venflon, 18-gauge; BOC Ohmeda AB, Helsingborg, Sweden), placed in heparinized glass tubes, and centrifuged (3000 rpm, 5 minutes). Plasma was frozen and stored at −20°C until assayed.
      Plasma CCK was determined after concentrating samples by Sep-Pak C18 cartridges as described previously.
      • Beardshall K
      • Deprez P
      • Playford RJ
      • Alexander M
      • Calam J.
      Effect of chymotrypsin on human cholecystokinin release: use of clostripain in the validation of a new radioimmunoassay.
      The assay used antibody Dino (1:1,500,000), which is specific for the sulfated C terminus of CCK-8 and does not react significantly with sulfated gastrin. Bolton-Hunter labeled CCK-8 (Amersham, Buckinghamshire, England) was used as label.

       Study I: CCK secretion in response to fatty acids of different chain length

      Fifteen healthy volunteers (9 men and 6 women; age range, 18–45) participated in these studies. Each received between 3 and 7 different test meals on separate days.

       Study protocols

      All 15 subjects received the control meal (vehicle) and a series of fatty acid emulsions, each of a single acyl chain length; only one study was performed on any subject on a single day. Fatty acid emulsion (250 mL of 0.05 mol/L) was administered over 2 minutes by orogastric intubation (via a 10-gauge feeding tube; Pennine Healthcare, Derby, England) to eliminate possible confounding effects of their differing tastes. Blood samples were taken fasted and then at 15-minute intervals until 1 hour after fatty acid delivery.
      To compare the fatty acid–induced CCK responses to those observed after a conventional mixed nutrient meal, 3 subjects ingested 2 slices of white bread, 20 g butter, 250 mL whole milk, and 2 hard boiled eggs (fat content at 37.5 g of which 97.5% of fatty acids have 16-carbon or longer acyl chains; in total ~600 kcal). Blood samples were taken fasted and at 15-minute intervals for 1 hour.
      To determine if any differences between fatty acids on CCK secretion detected in the preceding protocol were attributable either to their effects on gastric emptying or to preduodenal fatty acid absorption, fatty acid emulsions were intraduodenally infused to 5 subjects. A fine-bore feeding tube was sited using manometric activity patterns to ensure its tip was in the duodenum. Each subject received 0.05 mol/L C10 or C12, on separate occasions, infused over 15 minutes. Blood samples were taken fasted and at 15-minute intervals until 1 hour after fatty acid delivery.
      To determine the role of fatty acid concentration in CCK secretion, 8 subjects received varying concentrations of C10, C11, and C12; fatty acids were used at doses of 0.025–0.10 mol/L. Studies were performed as described above. A fasted blood sample was taken and a single further sample at 15 minutes after fatty acid administration for CCK measurement.
      To assess the effect of monounsaturation of a stimulatory fatty acid on plasma CCK, 0.05 mol/L stearic (C18:0) was compared with 0.05 mol/L oleic acid (C18:1) in 10 subjects as described above.

       Study II: Effects of fatty acid chain length on antral contractility

      Direct assessment of antral dimensions and contraction amplitude was made by percutaneous ultrasonography using a previously described technique.
      • Ahluwalia NK
      • Thompson DG
      • Mamtora H
      • Troncon L
      • Hindle J
      • Hollis S.
      Evaluation of human postprandial antral motor function using ultrasound.
      Nine healthy volunteers (5 men; median age, 25 years; range, 18–35) participated in this study.

       Test meals

      C10 or C12 was chosen on the basis of the preceding studies, to either induce a plasma CCK increase (C12) or not (C10). Because pilot studies showed that the fatty acids alone did not induce antral contractions, it was necessary to give an antral stimulus after fatty acid delivery. For this, 360 g of a semisolid nutritive meal was used (Cream of Chicken Soup; Heinz, Wigan, England). This meal was known to reliably stimulate antral contractions from previous studies of antral motility reported from our center.
      • Ahluwalia NK
      • Thompson DG
      • Mamtora H
      • Troncon L
      • Hindle J
      • Hollis S.
      Evaluation of human postprandial antral motor function using ultrasound.

       Definition of terms

      Relaxed antral circumference is the maximum circumference measured between the passage of 2 successive antral waves. Contracted antral circumference is the minimum circumference measured during passage of an antral wave. Antral contractile amplitude is the difference between relaxed and contracted antral circumference.

       Assessment of gallbladder contraction

      Because gallbladder contraction is known to be induced by lipid and mediated predominantly by circulating CCK, gallbladder ultrasonography was also performed to provide a positive control for fatty acid effects. The volume of the gallbladder was derived from the following equation: V = 1/3 π L (W/2)2.
      • Cano N
      • Cicero F
      • Ranieri F
      • Martin J
      • Di Costanzo J.
      Ultrasonographic study of gallbladder motility during total parenteral nutrition.

       Study protocol

      Studies were performed in all 9 subjects in the morning after an overnight fast. Eight subjects received all 3 test solutions, the other receiving C12 and vehicle but not C10. Blood samples were drawn fasted and at 15-minute intervals for 60 minutes.
      First, a fasted ultrasound image of antrum and gallbladder was obtained.
      • Ahluwalia NK
      • Thompson DG
      • Mamtora H
      • Troncon L
      • Hindle J
      • Hollis S.
      Evaluation of human postprandial antral motor function using ultrasound.
      Then the subject received 100 mL of either vehicle solution or fatty acid (0.10 mol/L C10 or C12 emulsion) on separate occasions and in randomized order. All solutions were administered intragastrically by a fine-bore orogastric tube to retain any inhibitory influences mediated by fatty acids from within the stomach and to avoid any possible confounding effects of their differing smell and taste. The tube was then removed, and antral and gallbladder scans were repeated. Fifteen minutes later, the soup (~40°C) was consumed by mouth. Antral and gallbladder scans were repeated immediately after soup ingestion and again at 15-minute intervals for 1 hour.
      In 3 subjects, the C12 administration study was repeated during intravenous infusion of loxiglumide, a specific CCK-A antagonist. Loxiglumide was given through an indwelling intravenous catheter at 30 mg · kg−1 · h−1 for 10 minutes' loading phase, starting 20 minutes before fatty acid administration, and then at 10 mg · kg−1 · h−1 until 1 hour after the antral stimulus. Otherwise the study was performed as in preceding protocol.

       Study III: Effect of fatty acid chain length on proximal gastric tone

      This study was performed using a gastric distention technique developed and reported in detail by us.
      • Ahluwalia NK
      • Thompson DG
      • Barlow J
      • Troncon LEA
      • Hollis S.
      Relaxation responses of the human proximal stomach to distension during fasting and after food.
      In summary, the method involves serial measurements of intragastric pressure during stepwise inflation of a polythene bag sited in the proximal stomach. Seven healthy asymptomatic volunteers (6 men and 1 woman; age range, 18–50) participated in this study.

       Test meals

      Two fatty acid emulsions were again chosen on the basis of the preceding studies, to either induce a plasma CCK level increase (C12) or not (C10).

       Study protocol

      Each meal was assessed in all 7 subjects in the morning after an overnight fast, rested in a supine position in a quiet room. First, the bag was swallowed and advanced to the body of the stomach, then unfolded by inflation with 400 mL of air and withdrawn until resistance was met, indicating that its proximal end was sited at the lower esophageal sphincter. The bag was advanced 2 cm and fully deflated, and the catheter was affixed to the subjects' cheek with adhesive tape to prevent displacement. After a 10-minute rest, the bag was progressively inflated in 50-mL increments at intervals of 1 minute. Because previous studies in our laboratory had shown that the first fasted distention series induces an initial relaxatory response, but that subsequent fasting distentions produce a consistent pressure-volume relationship, a second inflation series was performed immediately after the first; these data were used for baseline analysis. At the end of the second fasted inflation, the bag was left deflated for a further 10 minutes, after which 250 mL of one of the test meals was instilled into the stomach. Five minutes after the completion of meal delivery, the bag inflation procedure was repeated. Blood was drawn from a venous cannula at the end of each of the first 2 distention series, and then again 15 minutes after the test meal.
      The proximal gastric pressure-volume relationships were then derived for each subject for the fasted and fed distentions. An index of gastric tone (mm Hg) was obtained from these data by adding the values of intragastric pressure recorded for all the inflation volume steps from 50 mL during each series (using the lower maximum volume from the two distention series if the maximum volumes differed) and then dividing this by the number of inflation steps to produce an average value.

       Data and statistical analysis

      For all protocols, data were analyzed using Minitab software. Results are shown as mean ± SEM. Area under curve values to represent integrated CCK responses were generated for individual subjects. Estimation of probability of difference between paired data was undertaken by the Student paired t test or by Wilcoxon matched pair signed-rank test, whereas multiple comparisons were made by analysis of variance as indicated.

      Results

       Study I: CCK secretion in response to fatty acids of different chain length

      Vehicle alone produced a small transient increase in CCK concentration, which was evident after 15 minutes and maintained until 30 minutes (Figure 1A).
      Figure thumbnail gr1
      Fig. 1Fatty acid–induced CCK secretion. (A) Time course of plasma CCK responses to vehicle (●) and 0.05 mol/L fatty acid emulsions. C11 (2) did not stimulate CCK secretion, whereas C12 (▩) produced an increase in plasma CCK concentration, maximal at 15 minutes (*P < 0.01 vs. vehicle, P < 0.01 vs. fasted; n = 8–12). (B) Integrated plasma CCK responses to 0.05 mol/L fatty acid emulsions are shown as area under the individual CCK-time response curves (n = 6–12). ▨, Response to vehicle alone. Fatty acid chain lengths below 12 failed to produce an increase in CCK release, confirming that the critical chain length lies between C11 and C12 fatty acids. (*C12–18, P < 0.01 vs. vehicle). (C) Effects of intraduodenally infused C10 (○) and C12 (▩) on plasma CCK concentration, showing that the chain length dependence of CCK secretion seen with intragastric fatty acid administration is neither a consequence of preduodenal fatty acid absorption nor an artifact of effects on gastric emptying (*P < 0.01 vs. fasted or C10). (D) Concentration-response relationships of plasma CCK concentration to fatty acids. The shaded area represents the range of plasma CCK concentrations after vehicle. Increasing the concentration of C10 (○) or C11 (2) to 0.10 mol/L did not produce an effect on plasma CCK, whereas reducing the concentration of C12 (▩) to 0.025 mol/L did not remove its ability to promote CCK secretion (*P < 0.05 vs. vehicle). Higher CCK concentration was produced by 0.10 mol/L C12 than by 0.025 mol/L C12, indicating a weak concentration-response relationship, although a difference was not detectable on statistical probability testing (n = 5–12).
      Fatty acid emulsions up to and including C11 did not increase plasma CCK concentration compared with vehicle. In contrast, all fatty acid emulsions of C12 and longer chain length produced a consistent CCK response, with plasma CCK concentration remaining above basal at 60 minutes. Further increase in acyl chain length above 12 carbon atoms did not affect the plasma CCK response in magnitude or time course. The integrated CCK responses for each fatty acid (expressed as area under the plasma concentration-time curve) further showed that the chain length threshold required to elevate plasma CCK occurs was C12, and that chain lengths of C12 and longer were equipotent (Figure 1B). The mixed nutrient meal induced comparable CCK secretion to C12, with peak values also occurring at 15 minutes (5.3 ± 1.3 pmol/L).
      Delivery of C10 fatty acid directly into the duodenum produced a small increase in plasma CCK concentration, similar to that observed after intragastric C10. However, intraduodenal C12 fatty acid again induced a large and consistent rise in plasma CCK concentration, which did not differ in magnitude to that observed after intragastric C12 (Figure 1C). The time course of the CCK response to intraduodenal C12 was, however, more transient than when it was given intragastrically.
      Although all concentrations of C12 used (0.025–0.10 mol/L) increased plasma CCK concentration, neither C10 or C11 fatty acids elicited a rise in plasma CCK at any concentration used (0.05–0.10 mol/L) (Figure 1D).
      Monounsaturation of C18 fatty acid (oleic, C18:1) did not affect CCK secretion compared with the saturated fatty acid (stearic, C18:0). Both the peak responses (stearic, 4.7 ± 0.6 pmol/L; oleic, 6.2 ± 1.0 pmol/L) and integrated responses (stearic, 234 ± 31 pmol/L for 60 minutes; oleic, 270 ± 46 pmol/L for 60 minutes) were similar.

       Study II: Effects of intragastric fatty acids on antral contractility

      As expected, after C12 there was a significant increase in plasma CCK concentration at 15 minutes (4.0 ± 0.7 pmol/L; P < 0.05; C12 vs. vehicle). Again, no difference in plasma CCK concentration was found 15 minutes after C10 (2.0 ± 0.7 pmol/L) compared with vehicle (2.0 ± 0.6 pmol/L). CCK concentrations following C12 remained highest after the soup meal throughout the period of study (Figure 2A).
      Figure thumbnail gr2
      Fig. 2Interrelationships between fatty acid chain length, CCK secretion, and antral contractility. (A) Plasma CCK concentrations in antral contractility study. Fatty acid emulsion (100 mL), either C10 (○), C12 (▩), or vehicle (●) was administered intragastrically at −15 minutes. Fasted CCK concentrations were identical. Fifteen minutes later (time 0), C12 had produced a significant increase in plasma CCK (P = 0.02, paired t test vs. fasted; n = 8–9), unlike C10 or vehicle. Chicken soup was consumed at time 0; plasma CCK concentrations then increased in all subjects. For the duration of the study, plasma CCK concentrations remained highest after soup when C12 was given initially. (B) Antral circumferences were determined by percutaneous ultrasonography. In the fasted state, relaxed antral circumferences (solid lines) were comparable. Fatty acid emulsion, either C10 (○), C12 (▩), or vehicle (●) was given intragastrically at −15 minutes, and relaxed antral circumference increased. However, the magnitude of this increase was the same for each of these preloads. Soup was consumed at 0 minutes, and the relaxed antral circumference was further increased, again independently of the type of preload. Regular antral contractions commenced after soup ingestion, the circumference reducing at regular intervals to the minimum, contracted value (dotted lines). Contracted circumferences did not differ between vehicle and C10. However, contracted circumferences following C12 were greater at all recording intervals (P < 0.05 vs. C10 or vehicle; n = 8–9), indicating that contractions after C12 were of lower amplitude.
      Regular antral contractions were induced by the nutrient meal (chicken soup) in all subjects. Prior administration of C12 reduced the antral contractile amplitude in all subjects compared with vehicle (P < 0.05; Figure 2B), an inhibitory effect that persisted throughout the study. In contrast, antral contractile amplitudes following C10 did not differ from the amplitude after vehicle at any time point (Figure 2B).
      Gallbladder contraction was detectable after fatty acids. The reduction in gallbladder volume 15 minutes after C12 was 4.1 ± 1.1 mL, whereas C10 (−2.2 ± 0.9 mL) and vehicle (−1.3 ± 1.0 mL) had smaller but similar effects (P < 0.05; C12 vs. vehicle or C10).
      Because only C12 was associated with antral inhibition, the effect of the CCK-A receptor antagonist loxiglumide was tested with this fatty acid. In all subjects studied, loxiglumide reversed the reduction in antral excursion seen following C12. This occurred despite also producing larger increases in plasma CCK during loxiglumide infusion (Figure 3).
      Figure thumbnail gr3
      Fig. 3Antral contractile amplitude (2) was significantly inhibited by C12, but this was reversed by loxiglumide, a specific CCK-A receptor antagonist (loxiglumide vs. vehicle, P > 0.05). Loxiglumide was also associated with a marked increase in plasma CCK concentration (▩) (*P < 0.05 vs. vehicle; n = 3 subjects for each data point).
      Loxiglumide also abolished gallbladder contraction after C12, with gallbladder volume rising despite feeding (8.7 ± 0.7 mL after 60 minutes).

       Study III: Effects of intragastric fatty acids of different chain lengths on proximal gastric tone

      Plasma CCK was only consistently elevated by C12 (4.1 ± 1.3 pmol/L), but not C10 (2.7 ± 0.5 pmol/L) or vehicle (2.2 ± 0.3 pmol/L, P < 0.05; C12 vs. vehicle or C10).
      Vehicle had no detectable effect on proximal gastric pressure-volume relationships, and therefore did not influence tone (Figure 4).
      Figure thumbnail gr4
      Fig. 4Group mean curves are presented for proximal pressure-volume relationships. Vehicle solution had no significant effect, whereas both C10 and C12 shifted the postmeal curve downward (P < 0.05, Wilcoxon test; n = 7). However, the effect of C12 was greater than that for C10. Plasma CCK values were increased above vehicle responses by C12 but not by C10 (P < 0.05).
      Both C10 and C12, however, induced a shift in the postmeal pressure-volume relationship, and hence both fatty acids reduced proximal gastric tone (Figure 4). The magnitude of this effect differed consistently between the 2 fatty acids; in each subject a greater decrease in tone was observed after C12 (P < 0.05; Figure 5).
      Figure thumbnail gr5
      Fig. 5The change in calculated gastric tone in each subject after C10 and C12. The change was greater after C12 than C10 (P < 0.05, paired t test; n = 7).
      None of the test meals induced nausea in any subject.

      Discussion

      Our study has shown for the first time that the human proximal gut possesses a recognition system that differentiates between fatty acid molecules differing by only a single carbon atom in their acyl chain. As a result, fat ingestion can generate an entirely different pattern of enteroendocrine cell and target organ responses depending on chain length. This system may entirely explain the effect of fatty acids on gastric antral contractility, as a reflex inhibition mediated via CCK secretion. However, it cannot fully explain the effects of fatty acids on proximal gastric tone, indicating that one or more other fatty acid–sensitive mechanisms with different chain length specificity are also present.
      Although experiments using long- and medium-chain triglycerides have suggested a role for chain length in lipid-induced CCK secretion,
      • Hopman WPM
      • Jansen JBMJ
      • Rosenbusch G
      • Lamers CBHW
      Effect of equimolar amounts of long chain triglycerides and medium chain triglycerides on plasma cholecystokinin and gallbladder contraction.
      • Isaacs PET
      • Ladas MD
      • Forgacs IC
      • Dowling RH
      • Ellam SV
      • Adrian TE
      • Bloom SR.
      Comparison of effects of ingested medium- and long-chain triglyceride on gallbladder volume and release of cholecystokinin and other gut peptides.
      the current study supports the concept that this process is fatty acid rather than triglyceride mediated.
      • Guimbaud R
      • Moreau JA
      • Bouisson M
      • Durand S
      • Escorrou J
      • Vaysse N
      • Frexinos J.
      Intraduodenal free fatty acids rather than triglycerides are responsible for the release of CCK in humans.
      • Hildebrand P
      • Petrig C
      • Ketterer S
      • Lengsfeld H
      • Beglinger C.
      Inactivation of food-stimulated lipases in humans inhibits exocrine pancreatic secretion by a CCK-dependent feedback mechanism.
      Our results also show that once the chain length threshold of 12 carbon atoms is achieved or exceeded, all fatty acids are equipotent for CCK secretion. Importantly, the same chain length specificity remains when fatty acid is infused directly into the duodenum, excluding transgastric absorption of the shorter fatty acids as an explanation for their lack of effect on CCK secretion. Similarly, this observation excludes an inhibitory effect of CCK on gastric emptying as the explanation for the chain length effect, although the more transient CCK response to intraduodenal C12 in comparison to intragastric C12 may in part reflect autoregulatory control of gastric emptying by CCK.
      In performing and interpreting studies of fatty acid–induced effects on gut function, the situation is complicated by their physicochemical characteristics, which differ markedly with chain length. As chain length increases, fatty acids become progressively less water soluble and their melting point rises, thereby changing from soluble liquids to hydrophobic solids. A rigorous comparison of the effects of different fatty acids in vivo therefore requires their delivery in a manner that overcomes these differences in initial physical state, which can be achieved by emulsification.
      It is interesting, and possibly biologically important, to consider why the acyl chain length threshold for CCK secretion lies at C12. Although our data cannot provide a conclusive answer to this question, it seems highly likely that the threshold relates directly to the physicochemical properties of fatty acids, and so to the role played by CCK in lipid digestion. CCK stimulates delivery of bile to the duodenum to solubilize fats into micelles and increase the triglyceride droplet surface area at which pancreatic lipases can act. It is of particular interest to note then that the longest saturated fatty acid that is liquid at human body temperature is C11, which has a melting point of 35°C, whereas that of C12 is 44°C.
      C12 would therefore revert to the solid phase in the human gut unless there was emulsification by bile. In contrast, shorter fatty acid molecules, which behave as more water soluble liquids at body temperature, do not need bile salts for dispersal in the aqueous luminal environment. Hence saturated fatty acids with an acyl chain length of 12 or more carbon atoms are those that present the gut with the most difficulty in assimilation, and therefore require the presence of bile.
      • Westergaard H
      • Dietschy JM.
      Delineation of the dimensions and permeability characteristics of the two major diffusion barriers to passive mucosal uptake in the rabbit intestine.
      The control of CCK secretion by saturated fatty acids seems to provide a perfectly attuned stimulus for this. Our findings do, however, differ from a previous report that suggested that acyl chain unsaturation was critical in CCK secretion.
      • Beardshall K
      • Frost G
      • Morarji Y
      • Domin J
      • Bloom SR
      • Calam J.
      Saturation of fat and cholecystokinin release.
      Might fatty acid–induced effects on gastric motility share a similar rationale and chain length characteristics? Triggering a delay in the gastric emptying of less readily assimilated fats into the small intestine would clearly enhance the efficiency of their handling. However, the role of CCK in the gastric effects of lipid is not fully established, and no prior evidence exists that fatty acid chain length differentially determines their effect on proximal or distal gastric motility. Gastric inhibition by fat in humans has in fact only been previously shown in response to long chain triglyceride.
      • Heddle R
      • Dent J
      • Read NW
      • Houghton LA
      • Toouli J
      • Horowitz M
      • Maddern GJ
      • Downton J.
      Antropyloroduodenal motor responses to introduodenal lipid infusion in healthy volunteers.
      • Feinle C
      • D'Amato A
      • Read NW.
      Cholecystokinin-A receptors modulate gastric sensory and motor responses to gastric distention and duodenal lipid.
      • De Giorgio R
      • Stanghellini V
      • Ricci-Maccarini M
      • Morselli-Labate AM
      • Barbara G
      • Franzoso L
      • Rovati LC
      • Corinaldesi R
      • Barbara L
      • Go VLW.
      Effects of dietary fat on postprandial gastrointestinal motility are inhibited by a cholecystokinin type A receptor antagonist.
      Our discovery that fatty acid chain length determines CCK secretion generates a novel perspective from which to study fatty acid effects on the stomach, and thereby assess the role of CCK, because any effects observed in association with fatty acids shorter than C12 must be independent of circulating CCK.
      Interestingly, our data show that the proximal and distal stomach differ in their patterns of response to fatty acids of different chain length. An exact relationship occurred between the chain length of fatty acid required for CCK secretion and for the inhibition of antral contractile amplitude; furthermore, loxiglumide abolished the effect, so an entirely CCK-mediated reflex may be responsible. Conversely, the fatty acid–induced effects on the proximal stomach cannot be entirely linked to CCK secretion, because both C10 and C12 induced a fall in tone. However, because the effect of C12 was consistently greater than that of C10, it seems likely that the additional relaxation effect induced by C12 was mediated by CCK, but that an additional CCK-independent mechanism is also in operation. This new finding may explain why CCK-A receptor antagonists do not entirely prevent the relaxatory effects of lipid on proximal gastric tone.
      • Mesquita MA
      • Thompson DG
      • Troncon LEA
      • D'Amato M
      • Rovati LC
      • Barlow J.
      Effect of cholecystokinin-A receptor blockade on lipid-induced gastric relaxation in humans.
      However, the chain length–dependent effect shown herein for CCK secretion and target organ effect is consistent with the single earlier observation made 30 years ago by Hunt and Knox
      • Hunt JN
      • Knox MT.
      A relation between the chain length of fatty acids and the slowing of gastric emptying.
      about gastric emptying of fatty acids of different chain length. They reported that gastric emptying rate was maximally delayed only by fatty acids with at least 12 carbon atoms.
      Neural pathways traveling via the vagus nerves play a key role in the effects of nutrients and CCK on gastric motility. Vagotomy, or selective vagal afferent lesioning with the sensory neurotoxin capsaicin, prevents the delay in gastric emptying and the reduction of proximal gastric tone produced by CCK or feeding.
      • Raybould HE
      • Tache Y.
      Cholecystokinin inhibits gastric motility and emptying via a capsaicin sensitive vagal afferent pathway in rats.
      CCK-A receptors are present on vagal afferents and transported to peripheral terminals,
      • Moriarty P
      • Dimaline R
      • Thompson DG
      • Dockray GJ.
      Characterisation of cholecystokinin A and B receptors expressed by vagal afferent neurons.
      and are therefore the most likely target for CCK-mediated effects of lipid on gastric motility. However, it has also been proposed that vagal afferents may be a target for direct activation by fatty acids, as shown electrophysiologically in a feline model.
      • Melone J.
      Vagal receptors sensitive to lipids in the small intestine of the cat.
      • Melone J
      • Mei N.
      Intestinal effects of the products of lipid digestion on gastric electrical activity in the cat: possible involvement of vagal intestinal receptors sensitive to lipids.
      This effect was associated with a reduction in gastric myoelectrical activity and was activated by fatty acids with chain lengths below the CCK secretory threshold. It could be argued therefore that the C10-associated reduction in proximal gastric tone that we observed could be consequent to direct activation of vagal afferents in the duodenal submucosa after the transit of the water-soluble fatty acid across the duodenal epithelium. It is also possible to argue that shorter chain length fatty acids activate nonvagal pathways to induce their gastric effects, for example by a direct inhibitory effect on gastric smooth muscle, although data to substantiate such an assertion are lacking. Another possibility is that the effect occurs via a CCK-independent fatty acid–sensitive mechanism located elsewhere in the intestine. One candidate for this effect is the ileal brake phenomenon, whereby fat in the distal small intestine, including fatty acids shorter than the CCK-secretory threshold, releases peptide YY and delays gastric emptying.
      • Spiller RC
      • Trotman IF
      • Adrian TE
      • Bloom SR
      • Misiewicz JJ
      • Silk BD.
      Further characterisation of the ‘ileal brake’ reflex in man: effect of ileal infusion of partial digests of fat, protein and starch on jejunal motility and release of neurotensin, enteroglucagon and peptide YY.
      • Aponte GW
      • Fink AS
      • Meyer JH
      • Tatemoto K
      • Taylor IL.
      Regional distribution and release of peptide YY with fatty acids of different chain length.
      In conclusion, the current study provides the first evidence that CCK secretion in direct response to fatty acids is exquisitely sensitive to chain length. However, fatty acids also seem able to activate CCK-independent mechanisms that show different chain length characteristics. Together, these systems seem to serve to optimize the handling of fatty acids by the gastrointestinal tract.

      Acknowledgements

      The authors thank Professor J Calam, Royal Postgraduate Medical School, London, for supplying CCK antibody; Angela Jones, Anthony Hobson, Leslie Heggie, and Josephine Barlow for providing technical assistance; Kevin Daly for assisting in this work while a final year medical student; and Julie Young for help in preparing this manuscript.

      References

        • Liddle RA.
        Gut peptides: biochemistry and physiology.
        in: Walsh JH Dockray GJ : Raven, New York1994
        • Guimbaud R
        • Moreau JA
        • Bouisson M
        • Durand S
        • Escorrou J
        • Vaysse N
        • Frexinos J.
        Intraduodenal free fatty acids rather than triglycerides are responsible for the release of CCK in humans.
        Pancreas. 1997; 14: 76-82
        • Hildebrand P
        • Petrig C
        • Ketterer S
        • Lengsfeld H
        • Beglinger C.
        Inactivation of food-stimulated lipases in humans inhibits exocrine pancreatic secretion by a CCK-dependent feedback mechanism.
        Gastroenterology. 1997; 114: 123-129
        • Beardshall K
        • Frost G
        • Morarji Y
        • Domin J
        • Bloom SR
        • Calam J.
        Saturation of fat and cholecystokinin release.
        Lancet. 1989; 2: 1008-1010
        • Olsen O
        • Ainsworth M
        • Schaffalitzky de Muckadell OB
        • Cantor P.
        Effects of oleic acid and oleyl alcohol on cholecystokinin and secretin in plasma and pancreatobiliary secretion.
        Scand J Gastroenterol. 1989; 24: 529-532
        • Hopman WPM
        • Jansen JBMJ
        • Rosenbusch G
        • Lamers CBHW
        Effect of equimolar amounts of long chain triglycerides and medium chain triglycerides on plasma cholecystokinin and gallbladder contraction.
        Am J Clin Nutr. 1984; 39: 356-359
        • Isaacs PET
        • Ladas MD
        • Forgacs IC
        • Dowling RH
        • Ellam SV
        • Adrian TE
        • Bloom SR.
        Comparison of effects of ingested medium- and long-chain triglyceride on gallbladder volume and release of cholecystokinin and other gut peptides.
        Dig Dis Sci. 1987; 32: 481-486
        • Ahluwalia NK
        • Thompson DG
        • Mamtora H
        • Troncon L
        • Hindle J
        • Hollis S.
        Evaluation of human postprandial antral motor function using ultrasound.
        Am J Physiol. 1994; 266: G517-G522
        • Ahluwalia NK
        • Thompson DG
        • Barlow J
        • Troncon LEA
        • Hollis S.
        Relaxation responses of the human proximal stomach to distension during fasting and after food.
        Am J Physiol. 1993; 40: G166-G173
        • Heddle R
        • Dent J
        • Read NW
        • Houghton LA
        • Toouli J
        • Horowitz M
        • Maddern GJ
        • Downton J.
        Antropyloroduodenal motor responses to introduodenal lipid infusion in healthy volunteers.
        Am J Physiol. 1988; 254: G671-G679
        • Feinle C
        • D'Amato A
        • Read NW.
        Cholecystokinin-A receptors modulate gastric sensory and motor responses to gastric distention and duodenal lipid.
        Gastroenterology. 1996; 110: 1379-1385
        • Raybould HE
        • Roberts ME
        • Dockray GJ.
        Reflex decreases in intragastric pressure in response to cholecystokinin in rats.
        Am J Physiol. 1988; 255: G242-G246
        • De Giorgio R
        • Stanghellini V
        • Ricci-Maccarini M
        • Morselli-Labate AM
        • Barbara G
        • Franzoso L
        • Rovati LC
        • Corinaldesi R
        • Barbara L
        • Go VLW.
        Effects of dietary fat on postprandial gastrointestinal motility are inhibited by a cholecystokinin type A receptor antagonist.
        Ann NY Acad Sci. 1994; 713: 226-231
        • Melone J.
        Vagal receptors sensitive to lipids in the small intestine of the cat.
        J Auton Nerv Syst. 1986; 17: 231-241
        • Melone J
        • Mei N.
        Intestinal effects of the products of lipid digestion on gastric electrical activity in the cat: possible involvement of vagal intestinal receptors sensitive to lipids.
        Gastroenterology. 1991; 100: 380-387
        • Beardshall K
        • Deprez P
        • Playford RJ
        • Alexander M
        • Calam J.
        Effect of chymotrypsin on human cholecystokinin release: use of clostripain in the validation of a new radioimmunoassay.
        Regul Pept. 1992; 40: 1-12
        • Cano N
        • Cicero F
        • Ranieri F
        • Martin J
        • Di Costanzo J.
        Ultrasonographic study of gallbladder motility during total parenteral nutrition.
        Gastroenterology. 1986; 91: 313-317
      1. 57th ed. Handbook of chemistry and physics. : CRC, Boca Raton, FL1976
        • Westergaard H
        • Dietschy JM.
        Delineation of the dimensions and permeability characteristics of the two major diffusion barriers to passive mucosal uptake in the rabbit intestine.
        J Clin Invest. 1974; 43: 718-732
        • Mesquita MA
        • Thompson DG
        • Troncon LEA
        • D'Amato M
        • Rovati LC
        • Barlow J.
        Effect of cholecystokinin-A receptor blockade on lipid-induced gastric relaxation in humans.
        Am J Physiol. 1997; 36: G118-G123
        • Hunt JN
        • Knox MT.
        A relation between the chain length of fatty acids and the slowing of gastric emptying.
        J Physiol (Lond). 1968; 194: 327-336
        • Raybould HE
        • Tache Y.
        Cholecystokinin inhibits gastric motility and emptying via a capsaicin sensitive vagal afferent pathway in rats.
        Am J Physiol. 1988; 255: G242-G246
        • Moriarty P
        • Dimaline R
        • Thompson DG
        • Dockray GJ.
        Characterisation of cholecystokinin A and B receptors expressed by vagal afferent neurons.
        Neuroscience. 1997; 79: 905-913
        • Spiller RC
        • Trotman IF
        • Adrian TE
        • Bloom SR
        • Misiewicz JJ
        • Silk BD.
        Further characterisation of the ‘ileal brake’ reflex in man: effect of ileal infusion of partial digests of fat, protein and starch on jejunal motility and release of neurotensin, enteroglucagon and peptide YY.
        Gut. 1988; 29: 1042-1051
        • Aponte GW
        • Fink AS
        • Meyer JH
        • Tatemoto K
        • Taylor IL.
        Regional distribution and release of peptide YY with fatty acids of different chain length.
        Am J Physiol. 1985; 249: G745-G750