Avexitide – Enoblock inhibitor

Chronic n-3 fatty acid intake enhances insulin response to oral glucose and elevates GLP-1 in high-fat diet-fed obese mice†

ABSTRACT
n-3 polyunsaturated fatty acids (PUFA) can exert beneﬁcial effects on glucose homeostasis, especially in obese rodents. Gut incretin hormones regulate glucose and lipid homeostasis, but their involvement in the above effects is not entirely clear. This study aims to assess the effects of chronic n-3 PUFA adminis- tration on the insulin and incretin responses in C57BL/6N obese male mice subjected to oral glucose tolerance test (oGTT) after 8 weeks of feeding a corn-oil-based high-fat diet (cHF). The weight gain and adi- posity were partially reduced in mice fed cHF in which some of the corn oil was replaced with n-3 PUFA concentrate containing ∼60% DHA and EPA in a 3 : 1 ratio. In addition, these mice had improved glucose tolerance, which was consistent with an increased insulin response to oral glucose and plasma glucagon-like peptide-1 (GLP-1) levels. While the stimulatory effects of n-3 PUFA on GLP-1 levels could not be attributed to changes in intestinal or plasma dipeptidyl peptidase-4 activity, their beneﬁcial effects on glucose tolerance were abolished when mice were pretreated with the GLP-1 receptor antagonist exendin 9–39. Moreover, chronic n-3 PUFA intake prevented the detrimental effects of cHF feeding on glucose-stimulated insulin secretion in the pancreatic islets. Collectively, our data suggest that n-3 PUFA may modulate postprandial glucose metabolism in obese mice through a GLP-1-based mechanism. The signiﬁcance of these ﬁndings in terms of the effective DHA and EPA ratio of the n-3 PUFA concentrate as well as the effect of n-3 PUFA in humans requires further research.

Introduction
Overweight and obesity are frequently associated with disturb- ances of lipid and glucose homeostasis, which predispose to the development of type 2 diabetes mellitus (T2DM), cardio- vascular disease and other diseases.1 Although impaired secretory functions and metabolism of hypertrophic white adipose tissue (WAT) in obesity are central to the development of ectopic lipid storage, inflammation, insulin resistance and T2DM,2 the role of the gastrointestinal tract in the pathophy- siology of obesity-linked metabolic abnormalities is beginning to emerge, also in connection with the beneficial metabolic eﬀects of bariatric surgery in obese T2DM patients.3 Therefore, nutrient sensing in the gastrointestinal tract as well as the regulation of gut hormones such as glucose-dependent insuli- notropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1), i.e. incretins that potentiate insulin secretion after food ingestion and regulate glucose and lipid metabolism, rep- resent attractive targets especially for the treatment of glucose metabolism abnormalities in obesity and T2DM.3,4 Lifestyle changes such as increased physical activity and healthy nutrition could reduce the incidence of T2DM in pre- diabetic subjects by up to 60%.5 Regarding nutrition, dietary fatty acids (FA) can have a significant impact on body metab- olism, including the regulation of inflammatory responses and insulin sensitivity,6–8 and ectopic lipid storage.9,10 Among the diﬀerent types of FA, n-3 polyunsaturated FA (PUFA), such as docosahexaenoic acid (DHA; 22:6n-3) and eicosapentaenoic acid (EPA; 20:5n-3), which are contained in marine fish oils, exert many positive eﬀects on the organism, including eﬀects on inflammation,11 and higher circulating levels of n-3 PUFA are associated with a reduction in overall mortality in older adults.12 In contrast, saturated FAs and n-6 PUFA are more pro-inflammatory;8,13 furthermore, n-6 PUFA in the diet appears to increase weight gain, in part due to stimulation of endocannabinoid system activity.14–16 In the context of obesity, beneficial eﬀects of n-3 PUFA could be mediated by (i) the changes in hepatic gene expression induced mainly by peroxi-some proliferator-activated receptor α,17 (ii) increased production of lipid mediators with anti-inflammatory and insulin- sensitizing properties,18 (iii) secretion of insulin-sensitizing hormone adiponectin from WAT,19 and (iv) modulation of endocannabinoid levels in tissues such as the liver and WAT.16,20,21 However, other mechanisms may be also involved, including the eﬀect of n-3 PUFA on the gastrointestinal tract22 and/or gut microbiota.23

While dietary n-3 PUFA can reduce circulating lipid levels in both obese mice10,24 and humans,25 their favourable eﬀects on glucose metabolism and hepatic insulin sensitivity were observed only in obese insulin-resistant rodents fed a high-fat diet,19,26–28 but not in obese insulin-resistant subjects with or without T2DM.29,30 An acceptable explanation of the above divergent eﬀects of n-3 PUFA on glucose homeostasis in obese mice and humans is still missing, and it could involve diﬀer- ences in the regulation of gut incretin hormones including GLP-1 and/or GIP. However, studies investigating the eﬀects of n-3 PUFA on incretin responses to glucose, particularly in the context of chronic n-3 PUFA supplementation, are lacking.This study aimed to evaluate the eﬀects of chronic intake of n-3 PUFA on the insulin response to oral glucose adminis- tration in dietary obese mice and to determine whether the eﬀect of chronic n-3 PUFA intake on insulin response could be related to changes in plasma levels of incretins.Before the experiments, male C57BL/6N mice (Taconic Tornberg; Lille Skensved, Denmark) were maintained for 2 weeks in a 12 h light/dark cycle (light from 6:00 a.m.) at 22 °C, with ad libitum access to drinking water and low-fat standard diet (Chow; Rat/Mouse – Maintenance extrudate; Ssniﬀ Spezialdieten GmbH, Soest, Germany; metabolizable energy optionally also with n-3 PUFA concentrate (see above; 52.5 g kg−1 diet), which replaced a proportional amount of corn oil in the diet. Both cHF and cHF + F diets had a pasty consistency and were stored at −20 °C, in sealed plastic bags filled with nitrogen. The animals received a fresh aliquot of the diet every other day. For the composition of macronutrients in high-fat diets – see Table S1,† and for the composition of FAs – see ref.Chow-fed mice served as lean controls. All animal pro- cedures were performed in accordance with the Federation of European Laboratory Animal Science Associations and the European Communities Council Directive (Directive 2010/63/ EU) and approved by the Animal Care and Use Committee of the Institute of Physiology CAS (Approval no. 81/2016).

To study the eﬀects of chronic n-3 PUFA intake on glucose tol- erance, the “prevention of obesity” approach was used, in which the cHF + F diet was administered during obesity devel- opment. Dietary interventions lasted for 8 weeks. Body weight was monitored weekly, while the calculation of energy intake was based on weekly food consumption measurements made over 24 hours. During the 8th week of the experiment, mice were subjected to oral glucose tolerance test (oGTT), and fasting blood glucose (FBG) and plasma insulin levels were used to quantify insulin resistance based on the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR). Mice were killed after being deprived of food for 6 hours (6:00 a.m.– 12:00 p.m.). Thirty minutes before killing, half of the mice (n = 7–8) in each dietary group received 0.5 mL of saline (i.e. the “Saline” group) by gavage, while the other half received 0.5 mL of 30% D-glucose (“Glucose”). Wet weights of selected tissues including the liver and WAT from the epididymal, sub- cutaneous, and mesenteric fat depot were recorded, and the adiposity index was calculated as the sum of weights of all above fat depots expressed relative to body weight. Aliquots of the liver, epididymal WAT, proximal (i.e. duodenum) and distal (i.e. ileum) small intestine, as well as colon, were snap-frozen in liquid nitrogen and stored at −80 °C for subsequent analyses.

During in vivo tests, blood was obtained by the tail incision and collected into EDTA-containing tubes. In the experiment where saline or D-glucose was administered by gavage before dissection (see above), blood was obtained by the cannulation of the portal vein and collected into EDTA-containing tubes that additionally contained the dipeptidyl peptidase-4 (DPP4) inhibitor (10 µL mL−1; Sigma-Aldrich, St Louis, MO, USA). Glucose tolerance testing and quantification of insulin resistance Mice (n = 7–8) fasting for 6 hours (6:00 a.m.–12:00 p.m.) were subjected to oGTT using a 0.5 mL-bolus of 30% D-glucose (i.e. 150 mg per mouse). Glucose levels were measured using a glucometer Contour Plus (Bayer, Leverkusen, Germany) in a drop of blood collected from the tail tip before and 15, 30, 60,Plasma levels of insulin, total GIP and active GLP-1 were assessed by xMAP technology using the MILLIPLEX MAP Mouse Metabolic and Gut Hormone Panels (i.e. MMHMAG-44 K and MGTMAG-78 K; Merck) and Luminex 100 instrument (Luminex Corp., USA). Plasma levels of active GIP were measured by the Mouse Active GIP ELISA kit (cat. # 81511; Crystal Chem, Downers Grove, Il, USA) and used to cal- culate the levels of inactive GIP, expressed as a diﬀerence in the levels of total and active forms of GIP. Lipid metabolites in plasma including total triacylglycerols (TAG) and cholesterol, as well as non-esterified FA, were assessed using enzymatic kits and colorimetric methods as before.10 Isolation of pancreatic islets and measurement of insulin secretion ex vivo Mice from the Chow, cHF and cHF + F group (n = 4) were used to isolate the pancreatic islets from collagenase-perfused pan- creata, using the Ficoll gradient as before.31 The yield was 100 to 200 islets per mouse. Pancreatic islets were seeded at 100 islets per well that was coated with Biolaminin 521 (BioLamina, Sweden), followed by overnight culture in CMRL medium (PAN-Biotech, Germany). Glucose-stimulated insulin secretion (GSIS; 22 mM glucose) was quantified from glucose- free Krebs-Ringer HEPES (KRH) buﬀer (135 mM NaCl, 3.6 mM KCl, 10 mM HEPES, 0.5 mM MgCl2, 1.5 mM CaCl2, 0.5NaH2PO4, 0.1% bovine serum albumin, pH 7.4) by Mouse Insulin High Sensitivity ELISA kit (BioVendor, Brno, Czech Republic). The islets were then lysed and the DNA content was quantified using the PicoGreen Assay (Thermo Scientific).

In vivo administration of a GLP-1 receptor antagonist Mice fasted for 6 hours (6:00 a.m.–12:00 p.m.) were injected with a GLP-1 receptor antagonist exendin 9–39 (Exendin Fragment 9–39; Sigma-Aldrich; cat. no. # E7269) at a dose of 0.1 mg per kg body weight, and then 30 min later subjected to oGTT as described above. Measurements of the enzymatic activity of DPP4 Enzymatic activity of DPP4 was assessed ex vivo in plasma and tissue homogenates (mesenteric WAT and ileum) as described earlier.32 Briefly, tissues (∼100 mg) were homogenized in 300 µL
of PBS/Triton-X100 (100 : 1) solution containing Aprotinin (100 KIU mL−1; MilliporeSigma, St Louis, MO, USA). The hom- ogenate was centrifuged at 100g, for 10 min at 4 °C, and the resulting supernatant was further cleared by double centrifu- gation at 20 000g, for 10 min at 4 °C. Aliquots (15–50 µL) of cleared supernatants or plasma were mixed with 35 µL of Tris- buﬀer and 50 µL of DPP4 substrate solution (H-Gly-Pro-pNA·p- tosylate; Bachem AG, Bubendorf, Switzerland), which was prepared by mixing DPP4 substrate (10 mmol L−1; dissolved in DMSO) with Tris-buﬀer (1 : 50). The kinetics of DPP4 substrate degradation was measured spectrophotometrically at n =
405 nm, for 30 min at 37 °C, and quantified as the slope of the linear segment of the respective kinetic curves (expressed as arbitrary units; AU), which was then recalculated per 1 mL of plasma or normalized to the tissue sample weight.Total RNA was isolated using TRI Reagent (MilliporeSigma, St Louis, MO, USA). Samples of the intestinal tissue were obtained from fasted mice killed 30 min after administration of saline. The levels of mRNA for preproglucagon (Gcg; Forward: ACTTCCCAGAAGAAGTCGCC, Reverse: GAAGTCCCTG GTGGCAAGAT), GIP (Gip; Forward: CACAGGAGAGCTCTTTG CCC, Reverse: AGGAGCCAAGCAAGCTAAGG), and DPP4 (Dpp4; Forward: TCTAGTGCGGCTCCCATCCAAATC, Reverse: GGTTCT
GCGGGCCTAAATCTTCC) in diﬀerent intestinal segments were analyzed by real-time quantitative PCR using the LightCycler
480 Instrument (Roche Diagnostics; Mannheim, Germany) and primers (shown in 5′–3′ orientation) designed using Lasergene software (DNASTAR, Madison, WI, USA). The mRNA levels in the gut were normalized to the geometric mean of mRNA levels of reference genes, i.e. glyceraldehyde-3-phos- phate dehydrogenase (Forward: CCCGGCATCGAAGGTGGAA GAGT, Reverse: CTGACGTGCCGCCTGGAGAAAC) and hypoxanthine phosphoribosyltransferase 1 (Forward: GGATACAGGC CAGACTTTG, Reverse: GCAGATGGCCACAGGACTA).Data are expressed as means ± SEM. Statistical analysis was performed by SigmaStat software, using one-way ANOVA fol- lowed by Student–Newman–Keuls post-hoc test for pairwise multiple comparisons among the groups (Chow, cHF, cHF + F). In a situation where only two groups were compared (cHF vs. cHF + F), e.g. measurement of GLP-1 in an unstimulated (saline) or stimulated (glucose) state, or in the experiment investigating the eﬀects of GLP-1 receptor blockade, Student’s t-test was used. Logarithmic transformation was used to stabil- ise variance in cells when necessary. Comparisons were judged to be significant at p < 0.05. Results The eﬀect of n-3 PUFA intake on energy balance and basic parameters of glucose and lipid metabolism in dietary obese mice We first examined the eﬀect of chronic n-3 PUFA adminis- tration on glycemic and insulin responses using mice with diet-induced obesity. To induce obesity, mice were fed for 8 weeks a corn oil-based high-fat diet (i.e. cHF) enriched with n-6 PUFA, which are known to promote weight gain (see Introduction). Feeding cHF diet markedly induced body weight and weight gain (∼4.5-fold increase vs. Chow-fed mice; Table 1). In contrast, body weight gain of C57BL/6N mice fed the n-3 PUFA-supplemented high-fat diet (cHF + F group) was Table 1 Parameters of energy balance, adiposity, ectopic fat accumu- lation, and glucose and lipid homeostasis in mice fed a high-fat diet with or without n-3 PUFA supplementation the glycemic response to orally administered glucose, possibly in association with elevated incretin levels. In oGTT performed at week 8 (Fig. 1A), an improvement in glucose tolerance (i.e., a decrease in AUC) was observed in the cHF + F group, as well as in lean Chow-fed mice, compared to obese controls fed cHF.Data are means ± SEM (Chow, n = 14; cHF and cHF + F, n = 16); for plasma levels of lipid metabolites (Chow, n = 7; cHF and cHF + F, n = 8). Male C57BL/6N mice were fed for 8 weeks a high-fat diet alone (cHF) or supplemented with the EPA + DHA concentrate (cHF + F). Mice were killed after 6 hours of fasting (6:00 a.m.–12:00 p.m.), with half of the mice given saline and the other half glucose solution (30 min before killing). Plasma levels of lipid metabolites were measured in plasma of mice given saline before dissection. Energy intake was calculated using values for the metabolizable energy of the respective diet. a Based on food consumption data obtained during the first 6 weeks of experiment. b Food eﬃciency calculated as the average of weekly measurements (week 1–6). *, significant diﬀerence vs. cHF;†, significant diﬀerence vs. Chow (one-way ANOVA)(Fig. 1B). To be able to determine changes in plasma levels of insulin as well as incretin hormones following oral glucose administration in dietary obese mice, we used an alternative experimental approach based on oral administration of saline or D-glucose to fasted mice, followed by portal vein cannula- tion 30 min later. Glucose-stimulated insulin levels in DPP4 inhibitor-treated plasma, obtained from blood collected 30 min after glucose administration, showed a ∼7.5- and ∼4.8- fold elevation (vs. saline-treated mice; p < 0.01) in the cHF + F and Chow groups, respectively, as compared to obese cHF-fed mice showing only a ∼3.3-fold increase (Fig. 1C). Furthermore, glucose-induced stimulation of insulin secretion in cHF + F mice was associated with increased plasma levels of active GLP-1, which was not observed in cHF mice (Fig. 1D). Glucose- stimulated levels of active GIP increased in cHF + F mice com- pared to Chow-fed mice, while they did not diﬀer from the cHF mice (Fig. 1E). Although total GIP levels were markedly elevated ( p < 0.001) in response to glucose administration in all groups, they were significantly lower in lean Chow-fed mice than in the cHF or cHF + F group (Fig. 1F). However, plasma levels of inactive GIP were reduced in both the Chow and cHF + F groups as compared to obese cHF mice (Fig. 1G). These data suggest that increased glucose-stimulated insulin secretion observed in dietary obese mice with chronic n-3 PUFA intake can be primarily linked to elevated plasma levels of active GLP-1. Reduced expression of the GIP precursor in the intestine but no changes in DPP4 activity in obese mice in response to chronic n-3 PUFA intake reduced by ∼16% as compared to obese cHF-fed controls, despite the absence of significant changes in daily food or energy intake assessed during the first 6 weeks of the study(Table 1). However, food eﬃciency was reduced in the cHF + F mice (Table 1). Consistent with reduced weight gain, cHF + F mice had reduced adiposity, mainly due to the lower weight of epididymal and mesenteric WAT (Table 1). Liver TAG content and plasma levels of non-esterified FA and cholesterol, measured after 6 hours of fasting, were also reduced in cHF + F mice, compared to cHF-fed controls (Table 1). Regarding the state of glucose homeostasis characterized in mice after 6 hours of fasting, FBG and HOMA-IR levels were significantly increased in cHF mice (vs. Chow), while n-3 PUFA supplemen- tation in cHF + F animals tended to reduce HOMA-IR ( p = 0.06) (Table 1), indicating a possible partial improvement in insulin sensitivity.Increased glucose-stimulated insulin levels are linked to elevated plasma levels of active GLP-1 in dietary obese mice with chronic n-3 PUFA supplementation The next step was to determine whether in obese mice fed a high-fat diet, chronic intake of n-3 PUFA may positively aﬀect.Given the diﬀerences between the cHF and cHF + F mice regarding glucose-stimulated plasma levels of active GLP-1 and inactive GIP (see above), gene expression of their precursors was measured in several intestinal segments of the corres- ponding mice (Fig. 2A). The expression of GIP (Gip) in the segment 1 and 2 (i.e. in the duodenum and distal ileum, respectively) was the highest in obese cHF mice as compared to lean Chow-fed controls, while it was reduced in the cHF + F mice; no eﬀect of n-3 PUFA supplementation was observed in terms of GLP-1 precursor preproglucagon (Gcg) expression (Fig. 2A). The expression of Dpp4, which encodes the incretin- inactivating enzyme DPP4, was lower in the intestinal segment 1 and 3 (i.e. in the duodenum and colon) of the Chow as com- pared to either cHF or cHF + F mice (Fig. 2A), and was consist- ent with the data on tissue activity of DPP4 (Fig. 2B and C). However, the opposite was observed in mesenteric WAT (Fig. 2D and E), where DPP4 activity was much higher in the Chow as compared to the other two groups. On the other hand, when expressed per weight of the whole fat depot, it was similar between all groups (Chow, 0.45 ± 0.07; cHF, 0.62 ± 0.07; and cHF + F, 0.59 ± 0.03 AU per depot). Furthermore, plasma DPP4 activity showed no diﬀerences among the groups Fig. 1 The effect of chronic n-3 PUFA intake on glucose tolerance and insulin and incretin responses to oral glucose in dietary obese mice. Glucose tolerance was analysed during week 8 of dietary interventions in mice fed the Chow (grey circles or columns), cHF (white circles or columns), and cHF + F (i.e. with n-3 PUFA supplementation; black circles or columns) diets. After a 6-hour fast (6:00 a.m.–12:00 p.m.), mice were subjected to oGTT [(A); corresponding AUC values for glucose shown in (B)]. For analysis of hormonal responses, mice in each dietary group were divided into two subgroups, left without food for 6 hours (6:00 a.m.–12:00 p.m.), and then one subgroup was given saline (white columns in C, D and F) while the other subgroup received D-glucose (black columns in C, D and F) by oral gavage. Using portal vein cannulation, blood was collected 30 min later and plasma levels of insulin (C), active GLP-1 (D), and total GIP (F) were determined after saline or glucose administration. Plasma levels of active GIP (E) were determined only in glucose-stimulated state (Chow, grey columns; cHF, white columns; cHF + F, black columns) and used to calculate the levels of inactive GIP (G; see Methods). Data are means ± SEM (Chow, n = 7; cHF and cHF + F, n = 8). *, signiﬁcant difference vs. cHF; †, signiﬁcant difference vs. Chow (one-way ANOVA); under the same conditions in C, D and F (i.e., after saline or glucose)(not shown). Thus, the elevated GLP-1 levels observed in cHF + F mice after oral glucose administration do not appear to be associated with changes in the gene expression of its precursor or the plasma/tissue activity of the degrading enzyme. In vivo eﬀects of GLP-1 receptor antagonism and ex vivo analysis of GSIS using isolated pancreatic islets.To provide further evidence for the role of GLP-1 in the increased insulin response observed in cHF + F mice after oral glucose administration, we injected the GLP-1 receptor antag- onist exendin 9–39 into cHF-fed control mice as well as cHF + F animals fed n-3 PUFA, and then repeated oGTT (Fig. 3). Compared to cHF-fed controls (Fig. 3A), exendin 9–39 injection in the cHF + F group caused a significant upward shift in the glycemic curve (Fig. 3B), indicating impaired glucose tolerance due to GLP-1 receptor blockade. Indeed, incremental AUC values derived from glycemic curves for the respective groups of mice showed significantly reduced glucose tolerance (i.e. an increase in AUC) in response to exendin 9–39 injections in cHF + F mice but not in cHF-fed controls when compared to their respective saline-treated controls (Fig. 3C). Furthermore, when the degree of response to exendin 9–39 administration was evaluated as the ratio of AUC values obtained from exendin-treated vs. saline-treated animals (Fig. 3D). Fig. 2 Gene expression of the incretin precursors in the intestine and tissue DPP4 activity. Gene expression (A) of GIP (Gip), GLP-1 precursor pre- proglucagon (Gcg), and Dpp4 was analyzed in the intestinal segment “1” (duodenum), segment “2” (distal ileum), and segment “3” (colon) of mice in the Chow (grey columns), cHF (white columns), and cHF + F (black columns) group. Progress curves for an enzyme reaction (B, D) and the quantiﬁcation (C, E) of DPP4 activity in the ileum (B, C) and mesenteric WAT (D, E) of mice in the Chow (grey columns or circles), cHF (white columns or circles), and cHF + F (black columns or circles) group. The linear segments of the kinetic curve for samples of the ileum (B; 1–6 min) and mesenteric WAT (D; 3–12 min) were used to calculate respective DPP4 speciﬁc activities expressed per gram of tissue (C and E). Data are means ± SEM (Chow, n= 7; cHF and cHF + F, n = 8). *, signiﬁcant difference vs. cHF; †, signiﬁcant difference vs. Chow (one-way ANOVA)∼1.8-fold greater in cHF + F mice compared to the cHF group. These results suggest that dietary n-3 PUFA improved the impaired glucose tolerance induced by cHF feeding and suggest that GLP-1-induced insulin secretion may be a major mechanism involved in this eﬀect. Since the improved glucose tolerance in n-3 PUFA-sup- plemented mice can be explained only partially by increased secretion of GLP-1, we further investigated how cHF feeding aﬀected the intrinsic properties of insulin-producing β-cells, and whether n-3 PUFA supplementation was able to aﬀect GSIS under these conditions. Therefore, pancreatic islets were isolated from collagenase-perfused pancreata of mice from all dietary groups (n = 4), and the GSIS level was determined under conditions where the medium contained either 3 mM or 22 mM glucose (Fig. 3E). When the islets were incubated in medium containing low glucose (i.e. 3 mM), GSIS levels were low and no significant diﬀerences were observed between groups. However, GSIS increased ∼5- and 10-fold in the islets of Chow and cHF + F mice, respectively, when a glucose con- centration of 22 mM was used (Fig. 3E); in contrast, GSIS decreased in the islets of cHF mice under conditions of increased glucose concentration in the medium. Fig. 3 In vivo effects of GLP-1 receptor antagonism and ex vivo analysis of GSIS using isolated pancreatic islets. Mice fed for 8 weeks the obeso- genic cHF diet or cHF + F diet supplemented with n-3 PUFA were subjected to oGTT following the i.p. injection of saline (black circles) or GLP-1 receptor antagonist exendin 9–39 (white circles); corresponding glycemic curves for saline- or exendin-treated mice in the cHF (A) and cHF + F (B) groups. The corresponding levels of glucose tolerance determined as AUC values (C) and the degree of response to exendin 9–39 administration expressed as the ratio of AUC values obtained from exendin-treated vs. saline-treated animals within each dietary group (D). Insulin secretion over 24 hours was determined in pancreatic islets isolated from mice fed the respective diets (i.e. Chow, cHF and cHF + F) for 8 weeks; GSIS was measured at two different glucose concentrations, i.e. 3 mM (white columns) or 22 mM (black columns; E). Data are means ± SEM (Chow, n = 6; cHF and cHF + F, n = 8). *, signiﬁcant difference vs. Saline (t-test; in A, B and C) or cHF (t-test; in D). *, signiﬁcant difference vs. cHF; †, signiﬁcant difference vs. Chow; under the same conditions, i.e. after 3 mM or 22 mM glucose (one-way ANOVA; in E)suggest that chronic supplementation of n-3 PUFA prevents the detrimental eﬀects of cHF feeding on GSIS and that the pancreatic islets of these mice can produce increased amounts of insulin in the presence of high glucose concentrations. Discussion This study examined in dietary obese C57BL/6N mice how chronic intake of n-3 PUFA aﬀected the insulin response to oral glucose, and whether this response depended on changes in plasma incretin levels.The insulin response to oral glucose was increased in obese mice fed the cHF diet supplemented with n-3 PUFA in the form of DHA-enriched TAG-based concentrate EPAX 1050 TG (DHA : EPA ratio ∼3 : 1). This eﬀect was greater than that observed in the control mice fed the cHF diet alone, which contained predominantly n-6 PUFA (i.e. linoleic acid; 18:2n-6) and essentially no n-3 PUFA.10,27 Previous studies suggested that diets high in linoleic acid can induce obesity and impair glucose tolerance while elevating the endocannabinoid system activity;14,16,21 on the other hand, n-3 PUFA supplementation using the same type of TAG-based EPA and DHA concentrate was shown to partially improve insulin sensitivity (HOMA-IR) in obese cHF-fed mice of the same strain (i.e. C57BL/6N10). Interestingly, in terms of changes in glucose tolerance follow- ing i.p. injection of glucose, the eﬀects of n-3 PUFA were mostly mild10,27 or absent.21,33 This has led us to investigate the impact of chronic n-3 PUFA intake on insulin response to oral glucose. We now demonstrate that, in contrast to our pre- vious results,10,21,27,33 the beneficial eﬀects of n-3 PUFA on glucose tolerance can be “unmasked” when glucose is given orally, which was associated with an increased insulin response in cHF + F mice. Dietary n-3 PUFA have been shown to preserve the first and especially the second phase of GSIS after previous short-term exposure to palmitate in perifused islets from rats fed a sucrose-rich diet.34 Moreover, transgenic mice overexpressing a C. elegans desaturase fat-1 have elevated endogenous levels of n-3 PUFA and were shown to have increased GSIS in isolated pancreatic islets35 while being protected against streptozotocin-induced β-cell damage and hypoinsulinaemia.36 Accordingly, our study now shows that chronic n-3 PUFA supplementation is not only able to prevent the harmful eﬀects of n-6 PUFA-enriched high-fat diet on GSIS in isolated pancreatic islets but even stimulates GSIS above the level observed in lean Chow-fed mice. This result suggests modulation of pancreatic structure and/or function due to chronic n-3 PUFA supplementation. Enrichment of cell mem- branes with n-3 PUFA may contribute to increased insulin secretion by suppressing the negative eﬀects of various pro- inflammatory cytokines and lipid mediators.11 Likely, n-3 PUFA will also modulate the composition and function of lipid rafts in pancreatic β-cells,37 which in turn would lead to increased insulin secretion. In contrast, previous study in rats fed a diet rich in saturated FA and supplemented with n-3 PUFA38 demonstrated lower plasma insulin levels over the 30 min period after i.v. injection of glucose. It is, therefore, possible that in diﬀerent models of diet-induced obesity and/ or insulin resistance16,27,34,38 the ability of n-3 PUFA to potenti- ate insulin response to glucose and to modulate glucose homeostasis may depend on the type of FA in the diet. Indeed, our group has shown10 that mice fed the lard-based high-fat diet (rich in saturated FA) supplemented with n-3 PUFA for 6 weeks had significantly higher FBG and HOMA-IR values com- pared to mice, in which the same amount of n-3 PUFA was supplemented as part of the n-6 PUFA-rich cHF diet. The intake of n-3 PUFA in dietary obese mice could modu- late insulin secretion from β-cells by a variety of direct and indirect mechanisms.39,40 However, given the fact that the oral (current study) but not i.p. route (e.g. in ref. 10, 21, 27 and 33) of glucose administration was associated with improved glucose tolerance in cHF + F mice with n-3 PUFA supplemen- tation indicated a potential involvement of incretin hormones (reviewed in ref. 4). Accordingly, in cHF + F mice, oral glucose administration resulted in an ∼8-fold increase in insulin levels in plasma isolated from portal vein blood, which was associated with a significant increase in plasma GLP-1 levels, not observed in obese cHF-fed controls. This is in line with the notion that GLP-1 response in obesity and/or T2DM is usually impaired (ref. 41, 42, and reviewed in ref. 43), and suggests that chronic intake of n-3 PUFA either stimulates some mecha- nisms that allow increased GLP-1 secretion or prevents its deterioration due to cHF feeding. The involvement of GIP in the above potentiation of glucose-stimulated insulin levels in the cHF + F mice is questionable because these mice showed only a tendency to increased levels of active GIP after oral glucose administration. However, n-3 PUFA could also stimu- late the activity of the incretin system by alleviating GIP resis- tance that is known to occur in obesity/T2DM and is character- ized by increased GIP mRNA expression in K-cells and GIP hypersecretion,44 unlike retained sensitivity to GLP-1.45 Indeed, a recent study in calorie-restricted obese subjects demonstrated reduced serum GIP levels in response to chronic n-3 PUFA supplementation,46 whereas in our study, the pattern of GIP expression in the small intestine of cHF + F mice and their tendency to have reduced total plasma GIP levels appear to be consistent with a partial alleviation of GIP resistance otherwise induced by cHF feeding. As already mentioned, the involvement of GLP-1 in poten- tiating the insulin response in mice with chronic n-3 PUFA intake is more likely. It has previously been shown47 that fol- lowing i.p. injection of glucose in non-obese mice, colon- specific delivery of EPA and in particular DHA markedly increased plasma levels of GLP-1 and insulin and subsequently reduced glycemia. Our current findings, which show that chronic dietary intake of n-3 PUFA leads to potentiation of insulin secretion associated with an increased GLP-1 response in obese mice stimulated by oral glucose, are therefore consist- ent with these results. n-3 PUFA are known to stimulate GLP-1 release from enteroendocrine L-cells by the activation of free fatty acid receptor 4 (also known as G-protein coupled receptor 120), which was observed both in vitro and in vivo,39,48 and chronic n-3 PUFA intake was able to alter the expression of G-protein coupled receptor 120 in the rat colon.49 In the present study, chronic intake of n-6 PUFA in the cHF diet resulted in an attenuated insulin response compared to the eﬀects of n-3 PUFA supplementation. Impaired GLP-1 responses and/or the deleterious eﬀects of cHF feeding on GSIS may provide a plausible explanation for this phenom- enon. In the case of n-3 PUFA, the mechanism(s) involved in their stimulatory eﬀects on GLP-1 release are unknown. However, the involvement of GLP-1 signaling in the eﬀect of n-3 PUFA on the insulin response was unequivocally demon- strated in an experiment with exendin 9–39, an established GLP-1 receptor antagonist.50 The involvement of interleukin-6 pathway51 is less likely since in our previous study plasma interleukin-6 levels were similar between cHF + F and cHF mice.33 On the other hand, n-3 PUFA may modulate the func- tion of intestinal sodium-dependent glucose cotransporter 1 that mediates glucose-induced incretin secretion and the inhi- bition of which is associated with increased GLP-1 response due to delayed glucose absorption.52,53 Further research is needed to demonstrate the involvement of the above mecha- nisms in stimulating the GLP-1 response due to chronic intake of n-3 PUFA. There are several other aspects of the current study that deserve comment. Firstly, the average daily intake of EPA + DHA in our mouse model was approximately 90 mg per animal. The maximum dose used in human studies was 15 g per day,54 which is approximately 10 times less in relation to body weight compared to our study. However, it is not entirely possible to compare the doses of n-3 PUFA used in mouse and human studies simply because one of the decisive factors for the biological eﬀects of n-3 PUFA is their bioavailability. For instance, in our recent study55 an identical daily dose of EPA + DHA administered via the same type of TAG concentrate (i.e. ∼90 mg of EPA and DHA in the form of EPAX 1050 TG) increased the Omega-3 index (i.e., a marker of the bio- availability of n-3 PUFA) to 7.5% compared to 2.5% found in the control mice fed a high-fat diet without n-3 PUFA. In con- trast, in humans, the Omega-3 index in the range of 8–12% is known to indicate better overall health,56,57 while the rec- ommended dose of n-3 PUFA, which is suﬃcient to achieve such Omega-3 index values, corresponds to an intake of 0.5–1.2 g EPA + DHA per day. Clearly, there are diﬀerences between rodents and humans in the dose of n-3 PUFA required to achieve a certain level of Omega-3 index and the incorpor- ation of these FAs into the body. Second, the cHF + F diet in our present study contained n-3 PUFA admixed in the form of a DHA-enriched concentrate with a DHA : EPA ratio of approximately 3 : 1. This particular DHA : EPA ratio is by no means exceptional, as the same ratio is commonly found in many marine fish, such as salmon, sea brass and yellowfin tuna.58 Moreover, the reason for using this particular type of n-3 PUFA concentrate (i.e. EPAX 1050 TG) in this work was mainly based on our previous studies, which show that (i) a low ratio of EPA to DHA may promote adiposity reduction,24 and (ii) the same type of n-3 PUFA concentrate produced a number of beneficial eﬀects in obese cHF-fed mice, including improved insulin sensitivity, alleviation of hepatic steatosis, as well as decreased adipocyte hypertrophy and endocannabinoid levels in adipose tissue.10,16,27,28 Consistent with these observations, a recent double-blind, ran- domized controlled trial in adults with abdominal obesity showed a reduction in insulin resistance due to administration of DHA-rich fish oil (DHA : EPA ratio 7 : 1) at a dose of 2 g daily for 12 weeks.59 While in our current study, chronic consump- tion of the cHF + F diet led to improved glucose tolerance in obese mice in conjunction with an increased insulin response to oral glucose, it is unclear whether this eﬀect is attributable to DHA alone or whether it also depended on the presence of EPA. While EPA is known to exert protective eﬀects against insulin resistance and stimulate insulin secretion,60,61 DHA could protect β-cells from apoptosis while increasing GLP-1 receptor expression and circulating insulin levels.62 There are not many studies that directly compare the insulinotropic eﬀects of DHA and EPA side by side. However, it appears that DHA may be slightly more eﬀective in inducing GLP-1 and insulin secretion compared to EPA.47 Therefore, it is tempting to speculate that in our study, the stimulatory eﬀects of n-3 PUFA supplementation on GLP-1 and the insulin response were primarily due to DHA. However, the relative contribution of DHA and EPA regarding the eﬀects of n-3 PUFA concentrate used in our study has yet to be clarified. Conclusion In summary, chronic n-3 PUFA intake in dietary obese mice fed the corn oil-based high-fat diet increased insulin levels stimulated by oral glucose administration. Glucose tolerance in mice with chronic n-3 PUFA intake was improved in line with the increased insulin response to oral glucose, which in turn was associated with elevated GLP-1 levels. This eﬀect of n-3 PUFA supplementation could not be attributed to changes in the DPP4 activity either in the intestine or plasma but was observed in a situation where the β-cell function was maintained and GSIS in the pancreatic islets improved. The precise mechanism responsible for the stimulation and/or preser- vation of GLP-1 response in dietary obese mice with chronic n-3 PUFA supplementation remains to be determined. The potential diﬀerences in the ability of n-3 PUFA to activate these mechanisms in obese mice, in contrast to obese patients with Avexitide T2DM, could help explain the absence of beneficial eﬀects of n-3 PUFA on glucose homeostasis in patients.