Monounsaturated fatty acids, such as oleic acid (OA), and certain milk proteins, especially whey protein (WP), have insulinotropic effects and can reduce postprandial glycemia. This effect may involve the incretin hormones glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1). To explore this, we examined the release and inactivation of GIP and GLP-1 after administration of glucose with or without OA or WP through gastric gavage in anesthetized C57BL/6J mice. Insulin responses to glucose (75 mg) were 3-fold augmented by addition of WP (75 mg; P < 0.01), which was associated with enhanced oral glucose tolerance (P < 0.01). The insulin response to glucose was also augmented by addition of OA (34 mg; P < 0.05) although only 1.5-fold and with no associated increase in glucose elimination. The slope of the glucose-insulin curve was increased by OA (1.7-fold; P < 0.05) and by WP (4-fold; P < 0.01) compared with glucose alone, suggesting potentiation of glucose-stimulated insulin release. WP increased GLP-1 secretion (P < 0.01), whereas GIP secretion was unaffected. OA did not affect GIP or GLP-1 secretion. Nevertheless, WP increased the levels of both intact GIP and intact GLP-1 (both P < 0.01), and OA increased the levels of intact GLP-1 (P < 0.05). WP inhibited dipeptidyl peptidase IV activity in the proximal small intestine by 50% (P < 0.05), suggesting that luminal degradation of WP generates small fragments, which are substrates for dipeptidyl peptidase IV and act as competitive inhibitors. We therefore conclude that fat and protein may serve as exogenous regulators of secretion and inactivation of the incretin hormones with beneficial influences on glucose metabolism.
ORAL GLUCOSE STIMULATES the secretion of gut-derived incretins, the two most important being glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1) (1). These incretins potentiate glucose-stimulated insulin release, and findings in incretin receptor gene-deleted (knockout) mice demonstrate their importance for normal glucose tolerance (2, 3). In humans, it has been shown that fat and protein also stimulate GLP-1 secretion, whereas GIP secretion seems stimulated mainly by oral fat and carbohydrate (4–6), suggesting that incretin hormones may contribute to changes in glucose tolerance also after ingestion of fat and protein. Experimental evidence for this was recently presented in studies showing that diets rich in monounsaturated fatty acids, such as oleic acid (OA), or certain proteins, such as whey protein (WP), improve glycemic control and result in a greater insulin release compared with a control meal with the same glucose content in Zucker rats (7) and in humans (8).
GIP and GLP-1 are inactivated by the ubiquitously expressed serine protease dipeptidyl peptidase-IV (DPP-4), which rapidly cleaves the N-terminal dipeptide from the hormones (9, 10). Because of this rapid inactivation, circulating GLP-1 and GIP consist mainly of the inactive form of the incretins, and therefore, measurements of the total concentration of the hormones (using C-terminally directed antisera) do not necessarily reflect levels of bioactive hormones. For this reason, measurements of both intact and total concentrations of the hormones are required to make accurate conclusions regarding the influence of nutrients on incretin hormone-mediated effects. Because most previous studies have determined incretin hormone concentrations using assays that do not distinguish the intact hormones from their primary metabolites, conclusions on secretion of incretins after nutrient ingestion and their possible contribution to glucose homeostasis should be viewed cautiously. Furthermore, measurements of both intact and total GLP-1 and GIP levels after nutrient ingestion may also explore the physiological regulation of rapid incretin degradation. This is an important aspect of incretin endocrinology during food ingestion, which is currently poorly explored, although it has been demonstrated that DPP-4 activity is increased in subjects with diabetes (11). Therefore, the aim of this study was to examine the release and circulating forms of the two incretins GIP and GLP-1 after oral administration of glucose together with fat or protein in mice. We undertook standardized oral administration of physiological quantities of nutrients, and we measured glucose and insulin in serial samples obtained after ingestion. Importantly, we determined the concentrations of circulating intact GIP and GLP-1 and their respective metabolites in mice using specific assays detecting N- and C-terminal portions of the hormones (12, 13), and we assessed DPP-4 activity in plasma and in different regions of the gut. This approach allows conclusions on both the secretion and inactivation of GIP and GLP-1 and the importance of the incretins for insulin secretion and glucose homeostasis after ingestion of nutrients.
Materials and Methods
Animals
Female C57BL/6J mice (Taconic, Skensved, Denmark), weighing 17.1–26.1 g (mean, 20.9 g) were housed six to eight per cage in a temperature-controlled (22 C) room with a 12-h light, 12-h dark cycle with lights on at 0600 h. Mice were fed a standard chow consisting of 18.0 kJ/kg, with 11.4% fat, 62.8% carbohydrate, and 25.8% protein (Lactamin, Stockholm, Sweden). Food and water were provided ad libitum. All animals were housed for 1–2 wk before their use. Approximately 100 mice were used in this study. The study was approved by the Animal Ethics Committee in Lund, Sweden.
Experiments
All reagents for experiments, as well as for analytical procedures, were from Sigma-Aldrich (St. Louis, Mo), unless otherwise stated. Overnight fasted mice were anesthetized with an ip injection of midazolam (Dormicum,10 mg/kg; Hoffmann-La Roche, Basel, Switzerland) and a combination of fluanison (20 mg/kg) and fentanyl (Hypnorm, 0.8 mg/kg; Janssen, Beerse, Belgium). After 30 min, paracetamol [acetaminophen (AAP), 2 mg/mouse] was administered through a gastric tube (outer diameter, 1.2 mm) alone or together with d-glucose (75 mg/mouse), with or without addition of OA (34 mg/mouse) and/or WP (100% Anywhey, 75 mg/mouse; Optimum Nutrition, Lindesberg, Sweden) in saline (500 μl total volume). Taurodeoxycholate (4 mmol/liter final concentration) was included in all test solutions containing oil, and these were subsequently sonicated for 10 sec to attain stable emulsions. Control mice received saline. Gastric emptying was assessed using the AAP assay, which is based on the poor absorption of AAP from the stomach and the rapid and almost complete absorption from the small intestine, resulting in an estimate of gastric emptying as the area under the plasma AAP concentration curve, which correlates with measurements of gastric emptying performed with isotopic techniques (14, 15). Blood samples were collected from the retrobulbar capillary plexus immediately before gastric gavage and at times 15, 30, 60, and 120 min using a 100-μl pipette that had been prerinsed in EDTA (0.5 mol/liter). After centrifugation, plasma was divided into appropriate aliquots and stored at −20 C until analysis. For incretin analyses, blood was collected 15 min after gavage with the specific DPP-4 inhibitor valine-pyrrolidide (0.01 mmol/liter, final concentration) (Novo Nordisk A/S, Bagsvaerd, Denmark). Also, 125-μl plasma samples obtained from four mice were pooled, and the pool was used for determination of intact GIP, total GIP, intact GLP-1, and total GLP-1. A total of 36–44 mice were used in each group, resulting in n = 9–11 for each group. For measurements of DPP-4 activity, blood was collected 15 min after gavage, and proximal and distal segments of the small intestine (each approximately 5 cm long) were removed immediately after mice were killed. The segments were washed in ice-cold saline and subsequently frozen at −70 C. For analysis, tissue segments were homogenized in buffer containing 20 mmol/liter Tris-HCl, 150 mmol/liter NaCl, 2 mmol/liter EDTA, and 1% Triton X-100 with pH 7.5. Cellular debris was removed by centrifugation and supernatants were immediately analyzed for DPP-4 activity.
Plasma and tissue analyses
Plasma glucose was determined with the glucose oxidase method using 2,2′-azino-bis(3-ethyl-benzothialozine-6-sulfonate) as substrate and absorbance measurements at 405 nm with Fluostar Galaxy spectrophotometer (BMG Labtechnologies GmbG, Offenburg, Germany). Insulin was determined by a RIA with a guinea pig antirat antibody, 125I-labeled porcine insulin as a tracer, and rat insulin as standard (Linco Research, St. Charles, MO). Free fatty acids (FFAs) were assessed with the NEFA C assay (Wako Chemicals GmbH, Neuss, Germany). Plasma AAP levels were determined with the acetaminophen assay (Cambridge Life Science, Ely, Cambridgeshire, UK) with absorbance measurements at 620 nm. For measurements of GIP (total and intact) and GLP-1 (total), each plasma pool was extracted once with 70% ethanol (final concentration). After vacuum centrifugation, the residue was reconstituted in assay buffer and immediately analyzed in RIAs. Total GIP was measured using the C-terminally directed antiserum R65 (10), which reacts with intact GIP and the N-terminally truncated metabolite GIP(3–42). Intact, biologically active GIP was measured using antiserum 98171, which was raised against the synthetic sequence, GIP(1–10)-cys [Genosys Biotechnologies (Europe) Ltd., Cambridge, UK] as described previously (12). The assay is specific for the intact N terminus of GIP and cross-reacts less than 0.1% with GIP(3–42) or with the structurally related peptides GLP-1(7–36)amide, GLP-1(9–36)amide, GLP-2(1–33), GLP-2(3–33), or glucagon at concentrations of up to 100 nmol/liter. It has a detection limit of 5 pmol/liter. Total GLP-1 was measured using two different C-terminally directed antisera recognizing amidated and nonamidated glycine-extended GLP-1, respectively (13). Thus, all intact and total GIP and total GLP-1 measurements were determined in extracts derived from the same pools and assayed in the same assay, thereby eliminating intraassay variation. Intact GLP-1 was measured in an ELISA with N-terminally directed antiserum (Linco Research) (16). DPP-4 activity was assessed kinetically using Gly-Pro-p-nitroaniline (1 mmol/liter) as substrate by monitoring the release of p-nitroaniline at 405 nm (17).
Statistical analysis
Values are means ± sem. Glucose tolerance, glucose-stimulated insulinemia, and gastric emptying were assessed by calculating the incremental area under the curve (AUC), calculated using the trapezoid rule, from 0–120 min for plasma glucose and insulin and from 0–60 min for AAP. The glucose elimination constant (KG) was estimated as the glucose elimination rate between 15 and 60 min. The early insulin response to glucose was calculated as the suprabasal plasma insulin concentration at 15 min. Pearson′s product moment correlation coefficients were obtained to estimate linear correlation between glucose and insulin levels. Statistical comparisons between groups were performed with a one-way ANOVA followed by Dunnett’s test using GraphPad Prism (version 4.01; GraphPad Software Inc., San Diego, CA).
Results
Glucose and insulin responses
After saline administration, plasma levels of glucose and insulin remained stable throughout the 120-min study period. Glucose administration increased plasma insulin levels at 15 min, and thereafter insulin levels gradually declined, reaching baseline after 60 min (Fig. 1A). Glucose levels peaked at 15–30 min after glucose administration and thereafter declined (Fig. 1B). When WP was added to glucose, the insulin response was markedly augmented (approximately 3-fold increase; P < 0.01), which was associated with a 31% augmentation of glucose disposal; KG was enhanced by WP plus glucose (1.4 ± 0.1%/min; n = 17) vs. glucose alone (0.9 ± 0.1%/min; n = 36; P < 0.01) (Table 1). When OA was administered together with glucose, plasma insulin responses were modestly augmented vs. glucose alone (∼1.5-fold increase; P < 0.05 for AUCinsulin; n = 19), whereas plasma glucose concentrations were not significantly different from those after glucose alone (Fig. 1, A and B). Figure 1C shows the glucose-insulin relationship in mice receiving glucose alone or glucose with WP or OA. The slope of the curve was steeper when glucose was combined with OA (167 ± 16 pmol insulin/mmol glucose; P < 0.05 vs. glucose alone) or WP (387 ± 58 pmol insulin/mmol glucose; P < 0.01 vs. glucose alone) compared with glucose alone (96 ± 8 pmol insulin/mmol glucose), suggesting potentiation of glucose-stimulated insulin release.
Plasma concentrations of insulin (A) and glucose (B) and the relationship between plasma glucose and plasma insulin (C) immediately before and 15, 30, 60, and 120 min after administration of saline (○) or glucose (75 mg; •) alone or together with WP (75 mg; ▪) or OA (34 mg; ▴) through gastric gavage in overnight fasted anesthetized female C57BL/6J mice. Data are expressed as means ± sem; n = 17–36 in each group.
Incremental AUC for insulin and glucose, the 15-min insulin response (EIR), and the 15- to 60-min glucose elimination rate (KG) after administration of glucose (75 mg) alone or together with WP (75 mg) or OA (34 mg) through gastric gavage in anesthetized C57BL/6J mice
| AUCinsulin (nmol/liter × 120 min) | EIR (nmol/liter) | AUCglucose (mmol/liter × 120 min) | KG (%/min) | |
|---|---|---|---|---|
| Glucose | 49.0 ± 3.9 | 1.65 ± 0.12 | 770.4 ± 51.9 | 0.86 ± 0.09 |
| Glucose + OA | 78.8 ± 12.0a | 2.24 ± 0.25 | 696.8 ± 36.7 | 0.91 ± 0.11 |
| Glucose + WP | 144.6 ± 18.8b | 4.71 ± 0.70b | 415.6 ± 42.b | 1.38 ± 0.14b |
| AUCinsulin (nmol/liter × 120 min) | EIR (nmol/liter) | AUCglucose (mmol/liter × 120 min) | KG (%/min) | |
|---|---|---|---|---|
| Glucose | 49.0 ± 3.9 | 1.65 ± 0.12 | 770.4 ± 51.9 | 0.86 ± 0.09 |
| Glucose + OA | 78.8 ± 12.0a | 2.24 ± 0.25 | 696.8 ± 36.7 | 0.91 ± 0.11 |
| Glucose + WP | 144.6 ± 18.8b | 4.71 ± 0.70b | 415.6 ± 42.b | 1.38 ± 0.14b |
Means ± sem are shown. There were 17–36 mice in each group.
Letters indicate probability level of random difference vs. results with glucose alone:
P < 0.05
P < 0.01.
Incremental AUC for insulin and glucose, the 15-min insulin response (EIR), and the 15- to 60-min glucose elimination rate (KG) after administration of glucose (75 mg) alone or together with WP (75 mg) or OA (34 mg) through gastric gavage in anesthetized C57BL/6J mice
| AUCinsulin (nmol/liter × 120 min) | EIR (nmol/liter) | AUCglucose (mmol/liter × 120 min) | KG (%/min) | |
|---|---|---|---|---|
| Glucose | 49.0 ± 3.9 | 1.65 ± 0.12 | 770.4 ± 51.9 | 0.86 ± 0.09 |
| Glucose + OA | 78.8 ± 12.0a | 2.24 ± 0.25 | 696.8 ± 36.7 | 0.91 ± 0.11 |
| Glucose + WP | 144.6 ± 18.8b | 4.71 ± 0.70b | 415.6 ± 42.b | 1.38 ± 0.14b |
| AUCinsulin (nmol/liter × 120 min) | EIR (nmol/liter) | AUCglucose (mmol/liter × 120 min) | KG (%/min) | |
|---|---|---|---|---|
| Glucose | 49.0 ± 3.9 | 1.65 ± 0.12 | 770.4 ± 51.9 | 0.86 ± 0.09 |
| Glucose + OA | 78.8 ± 12.0a | 2.24 ± 0.25 | 696.8 ± 36.7 | 0.91 ± 0.11 |
| Glucose + WP | 144.6 ± 18.8b | 4.71 ± 0.70b | 415.6 ± 42.b | 1.38 ± 0.14b |
Means ± sem are shown. There were 17–36 mice in each group.
Letters indicate probability level of random difference vs. results with glucose alone:
P < 0.05
P < 0.01.
Incretin responses
Figure 2, A–D, shows the plasma concentrations of total and intact GLP-1 and GIP 15 min after administration of glucose with or without WP or OA. Compared with glucose alone, addition of WP to glucose increased both total and intact GLP-1 concentrations (both P < 0.01). Total GIP concentrations were not altered by WP, whereas intact GIP levels increased (P < 0.01). OA in combination with glucose did not change total GLP-1 concentrations, although levels of intact GLP-1 were increased (P < 0.05). In contrast, levels of intact GIP were unaltered by OA, but total GIP concentrations were reduced (P < 0.01). The plasma levels of intact GLP-1 were linearly related to the slope of the glucose and insulin curve, whereas this relationship was not apparent with circulating intact GIP (Fig. 3, A and B).
Plasma concentrations of total GLP-1 (A), intact GLP-1 (B), total GIP (C), and intact GIP (D) after administration of glucose (75 mg) alone or together with WP (75 mg) or OA (34 mg) through gastric gavage in anesthetized female C57BL/6J mice. Data are expressed as means ± sem; n = 9–11 in each group. Asterisks indicate probability level of random difference between groups: *, P < 0.05; **, P < 0.01.
Relationship between plasma levels of intact GLP-1 (A) and intact GIP (B) and the slope of the relationship between plasma glucose and plasma insulin after administration of glucose (75 mg; •) alone or together with WP (75 mg; ▪) or OA (34 mg; ▴) by gastric gavage in anesthetized C57BL/6J mice. Data are expressed as means ± sem.
DPP-4 activity
Figure 4, A–C, shows DPP-4 activity in the proximal and distal small intestine and in plasma 15 min after administration of saline or glucose with or without WP or OA. Compared with glucose alone, addition of WP to glucose reduced DPP-4 activity in the proximal small intestine by approximately 50% (P < 0.05; Fig. 4A). In contrast, DPP-4 activity in the distal small intestine and in plasma was unaltered by nutrient administration (Fig. 4, B and C).
DPP-4 activity in proximal small intestine (A), distal small intestine (B), and plasma (C) after administration of saline or glucose (75 mg) alone or together with WP (75 mg) or OA (34 mg) through gastric gavage in anesthetized female C57BL/6J mice. Data are expressed as means ± sem; n = 4 in each group. Asterisks indicate probability level of random difference between groups: *, P < 0.05.
Gastric emptying
Figure 5A shows plasma concentrations of AAP in the different groups. Gastric emptying, as judged by incremental 0–60 min AUC of plasma AAP concentrations, was 8.4 ± 0.6 mmol/liter·60 min after administration of glucose alone. This was reduced by addition of WP to 4.0 ± 0.9 mmol/liter·60 min, i.e. by 52% (P < 0.01), and by addition of OA to 6.0 ± 0.6 mmol/liter·60 min, i.e. by 28% (P < 0.05). An inverse linear relationship was observed between intact GLP-1 and AAP (r = 0.98; P = 0.029; Fig. 5B).
A, Plasma concentrations of AAP immediately before and 15, 30, 60, and 120 min after coadministration of AAP (2 mg) with saline (○) or glucose (75 mg; •) alone or together with WP (75 mg; ▪) or OA (34 mg; ▴) through gastric gavage in overnight fasted anesthetized female C57BL/6J mice. B, Relationship between plasma levels of intact GLP-1 and AAP 15 min after administration of saline or glucose alone (75 mg) or glucose with WP (75 mg) or OA (34 mg). Data are expressed as means ± sem; n = 12–13 in each group.
Lipid responses
After saline administration, plasma levels of FFAs remained stable throughout the 120-min study period. Plasma FFA concentrations were reduced 15 min after administration of glucose with or without OA (P < 0.001 vs. controls; Fig. 6A). This reduction was maintained also at 60 min, but plasma FFAs had returned to control levels after 120 min. The reduction in FFA levels was inversely and nonlinearly related to plasma insulin levels at 60 min (Fig. 6B).
A, Plasma levels of FFAs immediately before and 15, 30, 60, and 120 min after administration of saline (○) or glucose (75 mg; •) alone or together with OA (34 mg; ▴) through gastric gavage in overnight fasted anesthetized female C57BL/6J mice. B, Relationship between plasma levels of FFAs and insulin 60 min after administration of saline or glucose (75 mg) alone or together with OA (34 mg) in C57BL/6J mice. Data are expressed as means ± sem; n = 12–13 in each group.
Discussion
We show that oral ingestion of WP in combination with glucose augments the secretion of GLP-1, as judged by measurements of GLP-1 using C-terminally reactive antibodies in mice. This is consistent with previous studies in humans showing that protein meal ingestion stimulates the secretion of GLP-1 (4–6). Of more importance, however, is the novel finding that also the concentration of active GLP-1, measured with N-terminally reactive antibodies, increased after administration of protein in combination with glucose vs. glucose alone. In contrast, the secretion of GIP, determined with a C-terminal assay, was not augmented by addition of protein to glucose, supporting earlier studies indicating that a protein-rich meal does not further (or only weakly) stimulate GIP secretion to a greater extent than glucose alone (4–6). However, despite the lack of change in overall GIP secretion, this study presents the novel finding that the active form of GIP, determined with N-terminally directed antibodies, is clearly increased by WP. This raises the intriguing possibility that protein ingestion reduces the inactivation of GIP leading to an increased concentration of active GIP, whereas total GIP levels are unchanged. Hence, the results of the present study suggest that oral protein ingestion differentially affects the two main incretins, while still leading to increased intact incretin hormone levels, possibly indicating that protein ingestion may be a physiological mediator of incretin degradation in mice. Thus, overall GLP-1 secretion is augmented, whereas it appears that there is a preferential inhibition of N-terminal GIP degradation by protein ingestion.
The degradation of GIP, but not GLP-1 appeared inhibited by oral protein administration, indicating selectivity toward GIP degradation after protein ingestion. Interestingly, protein ingestion inhibited DPP-4 activity in the proximal small intestine, suggesting that luminal digestion of WP generates small fragments (di- and tripeptides), which are substrates for DPP-4 and act as competitive inhibitors. It is well known that tripeptides, e.g. diprotin A (Ile-Pro-Ile) and diprotin B (Val-Pro-Leu), can behave as DPP-4 inhibitors. It is also known that small peptides are actively transported across the brush-border membrane (at a rate greater than that for single amino acids), that the jejunum is more active in this respect than the ileum, and that small, but significant amounts of di- and tripeptides escape intracellular degradation and are transported over the basolateral membrane by a basolateral peptide transporter (18). It is therefore not unreasonable to think that some of these di- or tripeptides could come in contact with DPP-4 within the intestinal wall and inhibit its activity. Ingested nutrients may reach GIP-producing K cells in the upper small intestine within the 15-min study period, whereas a longer period of time is required to affect L cells. The difference between inhibition of proximal intestinal DPP-4 activity and the lack of inhibition more distally may explain the differential findings regarding intact GIP and intact GLP-1 concentrations. Moreover, because plasma activity was not affected, the data suggest that DPP-4-mediated degradation may occur immediately after secretion in mice (i.e. probably within the intestine, very close to the site of secretion) (19). Finally, it cannot be excluded that selective incretin clearance may also contribute to the observed results.
Protein administration in combination with glucose augmented insulin levels and enhanced glucose elimination compared with glucose alone, and the slope relating plasma glucose to plasma insulin was augmented. This is consistent with a previous report showing that dietary supplementation with WP enhances glucose tolerance in healthy volunteers (8). This would suggest that an increased insulinotropic effect arising from the elevated intact GIP and GLP-1 concentrations found after protein ingestion results in augmented glucose disposal, which is supported by the significant correlation between plasma levels of intact GLP-1 and the slope relating glucose to insulin levels. It is also conceivable that the enhanced insulinemia is partially related to augmented amino-acid-induced insulin secretion, either directly or through stimulation of glucagon secretion, as inferred from human studies (20–23).
In contrast to the marked increase in insulinemia and glucose disposal after ingestion of WP in combination with glucose vs. glucose alone, administration of OA with glucose did not change glucose disposal and only slightly increased plasma insulin levels. Furthermore, OA administration did not further stimulate GLP-1 secretion beyond that achieved by glucose alone, as assessed from total GLP-1 levels. These findings seem to be different from previous reports demonstrating that dietary supplementation with medium-chain monounsaturated fatty acids enhances glucose tolerance because of augmented insulin secretion in rats (7), increased incretin hormones after fat-rich meal ingestion (4–6), and stimulation of GLP-1 secretion by fatty acids through activation of G protein-coupled receptor 120 (24). Similarly, previous in vitro studies have shown that OA stimulates GLP-1 release in primary rat L cells and in a mouse L-cell line (7, 25). Our failure to detect OA-mediated GLP-1 release in vivo may be because of a low concentration of OA at the location of the L cells after nutrient intake. It is also possible that OA-induced GLP-1 secretion is a slow phenomenon in vivo, perhaps because of the slowing of gastric emptying by fat (27), and therefore, no increase was seen during the 15-min period. Nevertheless, a novel finding in the present study is that although administration of OA together with glucose did not augment GLP-1 secretion, as judged from total GLP-1 concentrations, an increase in the active GLP-1 concentration was evident. This may suggest that OA inhibits GLP-1 inactivation processes and that the increased levels of active GLP-1 would contribute to the augmented insulin response to glucose, which may be too small to result in augmented glucose disposal.
It is known that gastric emptying contributes to postprandial glucose concentrations in healthy subjects and in patients with type 2 diabetes (28, 29). In this study, we determined the rate of gastric emptying with the AAP assay, which has previously been validated and used in human studies (30, 31). We found that glucose administration reduced gastric emptying compared with saline and that both OA and WP further delayed gastric emptying compared with glucose alone. This is consistent with previous findings that carbohydrates empty faster than fat from the stomach in rats (27). The delayed gastric emptying may be mediated by GLP-1, along its proposed function as an ileal brake hormone (32, 33). Interestingly, we found a clear inverse relationship between plasma levels of active GLP-1 and AAP 15 min after nutrient administration, which would support that a feedback inhibition of gastric emptying exists in mice. It is, however, unlikely that the delayed gastric emptying is responsible for the acute changes in insulin secretion after ingestion of OA or WP observed in this study, because the 15-min glucose concentration did not differ between experimental groups. However, the AUC for glucose and insulin may be affected by the reduced gastric emptying in the later phase of the study period. Importantly, other gut-derived hormones and neurotransmitters including cholecystokinin, islet amyloid polypeptide, and peptide YY are known to participate in the regulation of gastric emptying (34–36) and may thus also contribute to the delayed gastric emptying observed after nutrient ingestion in this study.
In this study, we also determined plasma FFA levels, which were rapidly reduced after administration of glucose alone and glucose with OA. Because this reduction was related to the insulin levels, we suggest that the major mechanism for the reduced FFA concentrations after OA administration is the inhibition of lipolysis by insulin. Whether, in addition, GLP-1 influences lipolysis is not clear. Previous studies in human adipocytes have shown that GLP-1 is lipogenic at physiological concentrations (ED50, 1 pmol/liter) and lipolytic at doses 10–100 times higher (37). Other studies have concluded that GLP-1 does not affect the rate of lipolysis in human adipose tissue or skeletal muscle (38). Thus, it is not clear whether GLP-1 contributed directly to the reduced lipolysis observed in this study after OA administration.
In conclusion, we present evidence that WP administration in combination with glucose increases GLP-1 but not GIP secretion compared with glucose alone. This conclusion is limited by the fact that GLP-1 release has been shown to be directly proportional to calorie intake (39, 40). Thus, the higher calorie content in test meals containing both glucose and protein compared with test meals with glucose alone may contribute to the superior GLP-1 release. In contrast to WP, OA administration had no further effect on incretin hormone secretion compared with glucose alone. Nevertheless, administration of WP increased circulating levels of both intact GIP and intact GLP-1, and OA administration increased circulating levels of intact GLP-1, suggesting that dietary factors, such as protein and fat, can modulate bioactive incretin hormone levels. This results in augmented insulinemia, improved glucose tolerance, and inhibited lipolysis. The results suggest that nutrients may serve as exogenous physiological regulators of secretion and inactivation of the incretin hormones with beneficial influences on glucose and lipid homeostasis. This would be of importance for further development of incretin-based therapy for the treatment of type 2 diabetes, which up to now heavily depends on GLP-1 receptor agonists and DPP-4 inhibitors (26, 41). However, it should be born in mind that these conclusions are based on incretin hormone measurements taken at a single time point (15 min after nutrient administration), which may not reflect the overall response of both incretins to nutrient ingestion. Limitations on the volume of blood that one can withdraw in mice necessitate the choice of a single sampling point. Nevertheless, because this time point corresponds to timing of the maximal insulin response, the conclusions are likely to reflect nutrient-mediated modulation of the incretin hormones as it affects β-cell secretion. However, additional sampling over a longer period after nutrient administration would further illustrate the apparent differential effect of the individual nutrients on GLP-1 and GIP secretion. Therefore, the conclusions of this study would be further extended if more detailed measurements of incretin hormone kinetics in relation to glucose and lipid homeostasis were undertaken in larger animals and man, in which larger blood volumes may be taken without disrupting physiological well-being.
Acknowledgments
We are grateful to Lilian Bengtsson, Lena Kvist, and Sophie Pilgaard for expert technical assistance.
This study was supported by the Swedish Research Council (Grant 6834), Albert Påhlsson Foundation, Swedish Diabetes Association, Novo Nordisk A/S, The Velux Foundation, Region Skåne, and the Faculty of Medicine, Lund University.
P.T.G., M.S.W., and C.F.D. have nothing to declare. M.O.L., K.J., and R.D.C. are employed by Novo Nordisk A/S and have equity interest in Novo Nordisk A/S. B.A. consults for Novo Nordisk A/S and Novartis and has received lecture fees from Novartis and grant support (2004–2005) from Novo Nordisk A/S.
Abbreviations:
- AAP
Acetaminophen
- AUC
area under the curve
- DPP-4
dipeptidyl peptidase-IV
- FFA
free fatty acid
- GIP
glucose-dependent insulinotropic polypeptide
- GLP-1
glucagon-like peptide-1
- OA
oleic acid
- WP
whey protein
