Gut hormones are considered to be locally produced in specific regions of the gastrointestinal tract (1, 2). Thereby, it has been also considered that secretin, cholecystokinin (CCK), and glucose dependent-insulinotropic polypeptide (GIP) are immediately released from the proximal small intestine, while glucagon like-peptide-1 (GLP-1), glucagon like-peptide-2 (GLP-2), and peptide-YY (PYY) are released from the distal intestine later, in response to meal ingestion. However, GLP-1 and PYY production in the proximal small intestine (3–5) could contribute to a rapid response of these hormones (6, 7) after a meal. There is no study to prove this hypothesis under conscious conditions in an animal model.

The primary purpose of the present study was to demonstrate regional gut hormone responses to voluntary meal ingestion under physiological (conscious and unrestrained) conditions, in a rat model. For this purpose, we developed surgical model rats by inserting a catheter into the mesenteric vein of the distal small intestine (ileal mesenteric vein, ILMV). We firstly compared gut hormone concentrations in the plasma samples collected from the ILMV to those from the portal vein (PV) of another group of rats with a portal catheter (8, 9).

In our previous studies using rats, oral administration of dietary peptides (2 g/kg dose) significantly increased plasma GLP-1 concentrations (10, 11), while the same dose of glucose did not (10, 12). These observations raised a hypothesis that proteins in the meal have a critical role in postprandial GLP-1 secretion. To test this hypothesis, we further examined the role of dietary protein in postprandial GLP-1 responses using the PV- or ILMV-cannulated rats, under a voluntary feeding condition.

In experiment 1, rats consumed more than 95% of the diet (2.80 ± 0.05 g in the PV group and 2.75 ± 0.04 g in the ILMV group) provided at the dose of 10 g/kg, within 15 min. As both groups consumed almost equivalent amount of the diet, the following results could not be attributed to a difference in the amount of food intake.

Plasma glucose concentrations ( Figure 2A ) elevated immediately in both groups, glucose concentrations at 15, 60, and 120 min were significantly higher in the PV group than those in the ILMV group, accordingly, the PM group had higher AUC of glucose than the ILMV group ( Figure 2B ). The concentration of active GLP-1 increased in the ILMV group immediately (15 min) after ingesting the meal, whereas it showed no increment in the PV group ( Figure 2C ). The AUC of GLP-1 was higher in the ILMV group than in the PV group ( Figure 2D ). GIP concentrations continuously increased over time in both groups, but the increment and the AUC were higher in the PV than in the ILMV group ( Figures 2E, F ). The concentration of PYY transiently increased only in the ILMV group, and a significant difference between the regions was detected by two-way ANOVA ( Figure 2G ). The AUC of PYY was not significantly different between two groups ( Figure 2H ). Insulin concentrations were similarly elevated both in the PV and the ILMV, and not significantly different between the regions ( Supplementary Figures 1A, B ).

In experiment 2, rats with the PV catheter consumed more than 95% of test diets (2.63 ± 0.05 g in the control group, 2.51 ± 0.05 g in the protein-free group) provided at a dose of 10 g/kg, within 15 min. Rats with the ILMV catheter consumed more than 92% of test diets (2.48 ± 0.05 g in control group, 2.38 ± 0.11 g in the protein-free group) within 15 min, and more than 97% of the respective diets within 60 min.

Basal (0 min) glucose concentrations (mg/dL) were similar in all the groups; 86.5 ± 2.5, 84.5 ± 4.0, 83.3 ± 4.4, and 90.9 ± 4.8 in Control/PV, Protein-free/PV (Pro-free/PV), Control/ILMV, and Protein-free/ILMV (Pro-free/ILMV), respectively. Glycemic response to both of test diets were higher in the PV group than in the ILMV group ( Figures 3A, B ).

Because basal values of gut hormone concentrations had unexpectedly high variances, results are presented as changes from the basal values ( Figure 3 ). Active GLP-1 concentrations (pM) were 28.4 ± 7.1, 60.3 ± 17.3, 239.0 ± 146.1, and 292.1 ± 150.5; total GLP-1 concentrations (pM) were 25.7 ± 4.8, 54.8 ± 11.4, 46.8 ± 17.5, and 75.2 ± 27.0; total GIP concentrations (pM) were 9.6 ± 1.3, 8.8 ± 0.8, 13.9 ± 1.6, and 15.4 ± 2.5; PYY concentrations (pM) were 28.0 ± 6.0, 44.3 ± 6.6, 94.8 ± 55.8, and 106.1 ± 44.2, in the Control/PV, Pro-free/PV, Control/ILMV, and Pro-free/ILMV groups, respectively.

In the PV, plasma glucose was higher in the Pro-free group than the control group ( Figure 3A ). No significant differences were observed between two groups in the ILMV ( Figure 3B ). Active GLP-1 concentration was transiently (15 min) elevated only in the Control/ILMV group with a significant difference as compared to the value at the same time point in the Pro-free/ILMV group ( Figures 3C, D ). We additionally measured the total GLP-1 concentration, as it reflects GLP-1 release rather than biologically active GLP-1. Because plasma samples in two rats (one in the Control/PV group and another in the Pro-free/ILMV group) did not remain sufficient for total GLP-1 measurement, the data were obtained respectively from 4 and 5 rats in these groups. Only in the ILMV group, total GLP-1 concentration sharply increased (P < 0.05 vs 0 min within the group, and, vs Pro-free group at the same time point) 15 min after ingestion of the control diet ( Figures 3E, F ). The Protein-free diet failed to increase total GLP-1 concentrations in the ILMV group. Total GIP concentrations increased similarly by both of diets in the PV and ILMV, respectively ( Figures 3G, H ). The Protein-free diet failed to increase total GLP-1 concentrations in the ILMV group. Postprandial responses of PYY ( Figures 3I, J ) were overall similar to those of GLP-1 both in the PV and ILMV groups. Incremental AUC ( Supplemental Figure 2 ) of glucose was higher in PV of the Pro-free group than in PV of the control group, and incremental AUC of total GLP-1was higher in ILMV of the control group than the Pro-free group.

It has been widely accepted that postprandial glucose absorption occurs in the proximal small intestine based on old and valuable studies (14–16). In the present study, differences in glycemia between the PV and ILMV groups probably reflects avid glucose absorption in the proximal small intestine. This is also supported by the result in experiment 2, in which the glycemic response was higher to the Pro-free diet than to the control diet in the PV group, but the difference was not observed in the ILMV group. To the best of our knowledge, this is possibly the first study demonstrating postprandial glycemia in veins of different regions of the intestine under physiological (conscious, unrestrained, and voluntary feeding) conditions.

Although we had expected similar profiles of GLP-1 responses, possibly at different degrees between the PV and ILMV groups, GLP-1 responses to the normal diet apparently differed. These results strongly suggest that GLP-1 is initially released from the ileum in response to meal ingestion. Such an interpretation agrees with the abundant distribution of GLP-1 producing cells in the distal than in the proximal small intestine (3–5). The immediate response (at 15 min) appears discrepant when considering the transit of luminal nutrients after meal ingestion. This could be explained by a “proximal-distal enteroendocrine loop” (17–19), in which certain proximal signals and the vagus nerve mediate GLP-1 release from the distal intestine. However, it is still possible that some of nutritional components immediately reached the middle-distal part of small intestine so that they directly stimulated L cells locally.

In contrast to GLP-1, postprandial GIP concentrations were higher in the PV than in the ILMV, which could be explained by the localization of GIP-producing K cells in the proximal small intestine (3–5). Such region-dependent responses in glucose, GLP-1, and GIP levels help to validate that our cannulation models (especially ILMV) are useful for distinguishing the events that occur in the proximal and distal intestine.

The protein-free diet failed to increase GLP-1 concentrations both in the PV and ILMV, indicating that protein is an essential component for the induction of GLP-1 release in response to the “meal”. This is partially supported by previous studies demonstrating that dietary proteins had a potent effect on GLP-1 secretion (20–23) as compared to that of carbohydrate and fat. As expected, profiles of PYY response were largely similar to that of GLP-1 responses in both experiments. This is consistent with the fact that these hormones are co-produced in L cells of the distal intestine, and with previous studies demonstrating the potent effect of dietary proteins on PYY secretion (21, 23, 24).

GLP-1 and PYY-producing cells exist abundantly in the ileum and colon, nevertheless these responses appeared within 15 min postprandially. Such quick responses might be mediated by neurohumoral reflexes triggered by sensing nutrients in the proximal small intestine (17–19, 25, 26).

Dietary fat is known to stimulate PYY in animal models and humans (2). Although all of test diet contained soybean oil at 7% (wt/wt), PYY concentrations did not increase in the ILMV after Pro-free diet ingestion ( Figure 3H ). In previous rat (approx. 300 g body weight) experiments, ileal infusion of 100 mM (28.2 mg/mL, 200 µL/min) oleic acid for 30 min (totally 169 mg oleic acid) induced much smaller PYY response than that of 5% peptone (totally 300 mg peptone) (27), and duodenal administration of 3 ml of 150 mM (42.3 mg/mL) oleic acid (totally 127 mg oleic acid = 1.14 kcal) induced much smaller PYY response than that of a liquid meal (totally 21 kJ = 5 kcal) (28). In the present study, rats ingested approximately 200 mg (70 mg fat/g diet x 2.5-2.8 g diet) of fat in MTTs. Because fat content was apparently lower than carbohydrate and protein contents in the diets, and these nutrients need to be emptied from the stomach into the small intestine to exert stimulatory effect on GIP/GLP-1/PYY secretions, we speculate that the amount of fat presented in the small intestine was not sufficient to trigger PYY secretion.

Insulin is not secreted from the intestine, and its secretion is stimulated by multiple factors such as GLP-1, GIP, glucose, amino acids, and fatty acid in the plasma, therefore, we are not able to provide relevant reasons for similar insulin responses observed in the present study regardless of the regions of blood sampling and of the diets given ( Supplementary Figure 1 ).

Concentrations of plasma gut hormones could be affected by the rate of blood flow and by merging with the blood outside from the intestine such as the pancreas and kidney (29). Portal blood flow is increased by meal ingestion and plasma glucose (30–33). Decrements in GLP-1 concentrations observed in PV of rats fed the protein free diet may be due to dilution by these physiological factors. Blood sampling from a mesenteric vein tributary from the proximal small intestine would be desirable to exclude these effects, though it is technically challenging for repeated blood sampling in awake rats.

Liberated peptides (10, 11, 34–36) and/or amino acids (37, 38) from the dietary proteins likely stimulated GLP-1 release. We speculate that dipeptidyl peptidase-IV (DPP-IV) inhibitory peptides liberated from the dietary proteins (10, 34, 39–43) are absorbed in the small intestine, wherein they protect GLP-1 from degradation by DPP-IV in the plasma. Measurements of active GLP-1, total GLP-1 concentrations, their ratio, and DPP-IV activity in systemic bloods collected without full protease protection will be needed to examine the hypothesis in future studies. A previous study demonstrated greater effect of carbohydrates than proteins on GLP-1 release in conscious rats (44) using lymph fistula and duodenal nutrient infusion. The opposite conclusion to our study would be attributed to the differential experimental conditions.

It seemed surprising that the protein-free diet did not increase GLP-1 concentrations even though the diet mainly (more than 80% wt/wt) consisted of digestible carbohydrates. This could be because carbohydrates were rapidly digested and absorbed as monosaccharides in the proximal small intestine, as suggested by the data obtained for glycemic and GIP responses (experiment 1). The dose of the test diet at 10 g/kg was possibly insufficient to detect an elevation of GLP-1 concentrations for the protein-free (high carbohydrate) diet, even in the mesenteric and the portal veins. Indeed, we could not detect any increment in GLP-1 levels by oral administration of glucose or dextrin at the dose of 2-3 g/kg (45, 46), whereas oral administration of protein hydrolysates at 2 g/kg sufficiently enhanced GLP-1 concentrations in the peripheral vein of normal rats (10, 11).

It is unclear why the values of active GLP-1 concentration were higher than that of total GLP-1 concentration in the present study. The discrepancy might come from the difference in methodology (a multiplex assay system and a conventional ELISA kit). However, the postprandial responses were almost parallel between both forms of GLP-1s. Therefore, our results seem to appropriately represent GLP-1 secretory response at the postprandial state. Another limitation in the present study is that it is not possible to separate circulated blood from the ILMV and PV blood samples. In addition, comparing with postprandial GLP-1 concentrations in systemic blood samples would help to distinguish secreted, circulated, and degraded GLP-1s. If we can successfully cannulate into the cecal vein or mesenteric artery, it will be possible to evaluate the region-specific hormone responses or nutrient/drug absorption.

In summary, using novel ILMV cannulated model rats and comparing them to PV cannulated model rats, it was demonstrated that postprandial GLP-1 is released from the distal intestine under physiological conditions. Furthermore, by comparing postprandial GLP-1 responses to the diets with or without protein, it was revealed that protein in the meal has a critical role in triggering postprandial GLP-1 responses from the distal intestine.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by The Hokkaido University Animal Committee.

TH, MS, and HH designed research. MS, and TH conducted research and analyzed data. TH and MS wrote the paper. TH had primary responsibility for final content. All authors contributed to the article and approved the submitted version.

This work was supported by JSPS KAKENHI Grant Number 26252016 and 18K19158.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.