The mechanisms by which the enteroinsular axis influences β-cell function have not been investigated in detail. We performed oral and isoglycemic intravenous (IV) glucose administration in subjects with normal (NGT; n = 11) or impaired glucose tolerance (IGT; n = 10), using C-peptide deconvolution to calculate insulin secretion rates and mathematical modeling to quantitate β-cell function. The incretin effect was taken to be the ratio of oral to IV responses. In NGT, incretin-mediated insulin release [oral glucose tolerance test (OGTT)/IV ratio = 1.59 ± 0.18, P = 0.004] amounted to 18 ± 2 nmol/m2 (32 ± 4% of oral response), and its time course matched that of total insulin secretion. The β-cell glucose sensitivity (OGTT/IV ratio = 1.52 ± 0.26, P = 0.02), rate sensitivity (response to glucose rate of change, OGTT/IV ratio = 2.22 ± 0.37, P = 0.06), and glucose-independent potentiation were markedly higher with oral than IV glucose. In IGT, β-cell glucose sensitivity (75 ± 14 vs. 156 ± 28 pmol·min−1·m−2·mM−1 of NGT, P = 0.01) and potentiation were impaired on the OGTT. The incretin effect was not significantly different from NGT in terms of plasma glucagon-like peptide 1 and glucose-dependent insulinotropic polypeptide responses, total insulin secretion, and enhancement of β-cell glucose sensitivity (OGTT/IV ratio = 1.73 ± 0.24, P = NS vs. NGT). However, the time courses of incretin-mediated insulin secretion and potentiation were altered, with a predominance of glucose-induced vs. incretin-mediated stimulation. We conclude that, under physiological circumstances, incretin-mediated stimulation of insulin secretion results from an enhancement of all dynamic aspects of β-cell function, particularly β-cell glucose sensitivity. In IGT, β-cell function is inherently impaired, whereas the incretin effect is only partially affected.
- glucagon-like peptide-1
- glucose-dependent insulinotropic polypeptide
- insulin secretion
the augmented insulin secretion observed after oral compared with intravenous glucose at similar plasma glucose concentrations is attributed to the influence of the enteroinsular axis, also termed the incretin effect. There is evidence that most of the incretin effect is due to the gut-derived incretin hormones glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP), secreted by L and K cells, respectively, in the jejunum, ileum, and colon, which enhance glucose-dependent insulin secretion by binding to specific receptors on the β-cell (14, 20, 21). These hormones are released during glucose or meal intake in proportion to nutrient transport across the intestinal epithelium (7), their effects seem to be additive, and they stimulate insulin secretion both at fasting and postprandial plasma glucose levels (42).
The physiological mechanisms by which incretins stimulate insulin release have not been investigated in much detail. In their classic experiments, Nauck et al. (30) demonstrated the incretin effect by comparing the β-cell response to an oral glucose load with the response to an intravenous infusion at matched glucose levels. However, the quantitative role of the glucose and incretin stimulus in these conditions has not been determined. The first aim of the present study, therefore, was to employ a mathematical model of β-cell function (11, 26) to quantitate the incretin effect. The β-cell model describes insulin secretion as the result of three different processes: insulin release in response to changes in plasma glucose levels (or insulin-glucose dose response, a measure of β-cell glucose sensitivity), insulin discharge in response to the rate of change of plasma glucose concentration (or rate sensitivity), and potentiation (31). The latter is a complex phenomenon encompassing glucose-induced (25, 31) and incretin-induced potentiation of insulin release (30).
GLP-1 secretion is reduced in patients with type 2 diabetes both after an oral glucose load (40) and during a meal test compared with lean or obese nondiabetic subjects (43), whereas stimulation of insulin secretion by acute or extended GLP-1 infusions is relatively well preserved in diabetic patients (18, 35). In first-degree relatives of type 2 diabetic patients, the incretin effect, calculated as the difference between oral glucose loading and isoglycemic intravenous glucose administration, on insulin secretion has been reported to be normal, and the plasma GLP-1 and GIP concentrations after the oral test have been found to be normal or even increased (23, 29, 32, 35). In subjects with impaired glucose tolerance (IGT), on the other hand, a slightly lower increment of plasma GLP-1 concentrations in the first 30 min of the oral glucose tolerance test (OGTT) (36) and a decreased acute insulin response and second-phase secretion in response to a primed-continuous GLP-1 infusion superimposed on a hyperglycemic clamp have been described (12). IGT subjects frequently are insulin resistant (19) and display similar defects in β-cell function as overtly diabetic patients, only of a lesser degree (9, 10, 34). Therefore, the second aim of this study was to investigate the incretin effect in IGT subjects in terms of both incretin release and its effects on β-cell function.
MATERIALS AND METHODS
Twenty-one subjects selected from the clinic volunteered for the study. None of them had lost weight or changed dietary habits during the 3 mo preceding the study, and none was receiving medications. All subjects had resting arterial blood pressure ≤140/90 mmHg (37), fasting glucose <6.1 mmol/l (3), and normal results for liver and renal function tests. Fat mass was estimated as the difference between body weight and fat-free mass, which was measured by electrical bioimpedance using a Body Composition Analyzer (model TB-300; Tanita, Tokyo, Japan) (16). On the OGTT (see below), 11 subjects had normal glucose tolerance (NGT) and 10 had IGT according to the American Diabetes Association criteria. The study was approved by the Institutional Review Board, and all subjects gave informed, written consent to the study.
Two studies were carried out after an overnight (12–14 h) fast at a 1-wk interval. In the first study, subjects received a 3-h OGTT (75 g) with measurements of plasma glucose concentrations at 10-min intervals. In the second study, the plasma glucose profile was reproduced by a variable intravenous (IV) glucose (20% dextrose) infusion by using an ad hoc developed algorithm. In both studies, venous blood was sampled at −30, 0, 10, 20, 30, 40, 60, 90, 120, 150, and 180 min for plasma insulin, C-peptide, GLP-1, GIP, and free fatty acid (FFA) measurements.
Plasma insulin was measured in duplicate by radioimmunoassay with the use of a kit for human insulin with negligible cross-reactivity with proinsulin and its split products (Linco Research, St. Louis, Missouri). Plasma glucose was measured using the glucose oxidase technique (Beckman glucose analyzers; Beckman Instruments, Fullerton, CA). Plasma triglyceride and serum high-density lipoprotein (HDL) cholesterol were assayed in duplicate by using standard spectrophotometric methods on a Synchron clinical system CX4 (Beckman Instruments). Plasma FFA was assayed using a spectrophotometric method (Wako Chemicals, Neuss, Germany). Total COOH-terminal amidated GLP-1 was assayed by RIA with the use of polyclonal antiserum 89390, raised in rabbits, which has an absolute requirement for the amidated COOH terminus of GLP-1 and does not cross-react with COOH-terminally truncated metabolites or the glycine-extended forms. The assay cross-reacts <0.01% with GLP-1-(7–35) and GLP-1-(7–37) amide, 83% with GLP-1-(9–36) amide, and 100% with GLP-1-(1–36) amide, GLP-1-(7–36) amide, and GLP-1-(8–36) amide. The assay has a detection limit of ∼1 pmol/l and a 50% effective dose (ED50) of 25 pmol/l. Intra- and interassay coefficients of variation are <6 and <15%, respectively (15, 33). The active (NH2 terminal) GIP was assayed by RIA with the use of polyclonal antiserum 98171, raised in rabbits, which is NH2-terminally directed and does not recognize NH2-terminally truncated peptides. It has a cross-reactivity of 100% with human GIP-(1–42) and <0.1% with human GIP-(3–42), GLP-1-(7–36) amide, GLP-1-(9–36) amide, GLP-2-(1–33), GLP-2-(3–33), and glucagon. The assay has a detection limit of ∼5 pmol/l and an ED50 of 48 pmol/l. Intra- and interassay coefficients of variation are <6 and <15%, respectively (6).
Insulin sensitivity was estimated from the plasma glucose and insulin responses to oral glucose loading by calculating the OGIS (oral glucose insulin sensitivity) index, which has been shown previously to be well correlated with the M value from the euglycemic hyperinsulinemic clamp (24). Insulin, C-peptide, GLP-1, GIP, and FFA area under the time-concentration curves were calculated using the trapezoidal rule. To estimate the size and time course of the incretin effect, we used the ratio of oral to IV measures (OS/IV) (30). This calculation cancels the impact of glucose levels per se, which were matched by design. To carry out cross-correlation analysis of pairs of time courses, we interpolated data using cubic hermite polynomials (17).
The model used to reconstruct insulin secretion and its control by glucose has been previously described (22, 25). In brief, the model consists of three blocks: 1) a model for fitting the glucose concentration profile, the purpose of which is to smooth and interpolate plasma glucose concentrations; 2) a model describing the dependence of insulin (or C-peptide) secretion on glucose concentration; and 3) a model of C-peptide kinetics, i.e., the two-exponential model proposed by Van Cauter et al. (41), in which the model parameters are individually adjusted to the subject's anthropometric data.
In particular, with regard to the insulin secretion block, the relationship between insulin release and plasma glucose concentrations is modeled as the sum of two components. 1) The first component is the relationship between insulin secretion and glucose concentration, i.e., a dose-response function. The dose-response function is modulated by a time-varying factor, expressing a potentiation effect on insulin secretion. The mean slope of the dose-response function is taken to represent β-cell glucose sensitivity. The potentiation factor encompasses glucose-induced potentiation and incretin potentiation and was set to have an integral equal to unity over the 3 h of each study. 2) The second insulin secretion component represents a dynamic dependence of insulin secretion on the rate of change of glucose concentration. This component, termed rate sensitivity, accounts for anticipation of insulin secretion as glucose levels rise. Total insulin secretion is the sum of these two components and is calculated every 10 min for the whole 3-h period.
Data are given as means ± SE. Group values were compared using the Mann-Whitney U-test (for continuous variables) or χ2 analysis (for nominal variables). Paired group values were compared using the Wilcoxon test. Time series were analyzed using ANOVA for repeated measures, with a group factor (NGT vs. IGT), treatment factor (time), and their interaction. Linear regression models were tested using standard techniques. Adjustment for covariates was carried out with analysis of covariance (ANCOVA). Cross-correlation was used on interpolated data to test for phasing of time courses (using SPSS 13.0 software). A P value ≤0.05 was considered statistically significant; when performing multiple post hoc comparisons, the P value was divided by the number of comparisons.
IGT subjects were heavier than NGT subjects but were matched to them by sex distribution, age, and lipid profile (Table 1). By design, plasma glucose concentrations were virtually identical during the OGTT and IV test in each individual; as expected, on either test the glycemic excursion was greater in IGT than in NGT subjects (Fig. 1). The insulin response to glucose was twice as large on the OGTT than IV and was significantly greater in IGT than NGT subjects on either test (Fig. 1). The ratio of the insulin response to oral vs. IV glucose, however, was similar in NGT and IGT subjects (Table 2). The plasma C-peptide responses were similar to the insulin responses. For both plasma insulin and C-peptide, the between-group differences were no longer statistically significant when adjusted (by ANCOVA) for body mass index (BMI). Plasma FFA concentrations were markedly suppressed by both oral and IV glucose in both groups (Fig. 1); FFA suppression, however, was slightly greater with oral than IV glucose in NGT. In both study groups, plasma GLP-1 levels were hardly affected by IV glucose, whereas oral glucose elicited a prompt and marked rise, which persisted throughout the test (Fig. 2). The total plasma GLP-1 response was 25%, on average, smaller in IGT compared with NGT subjects (Table 2), a difference that was canceled when controlled for BMI. Plasma GIP concentrations displayed a similar pattern (Fig. 2).
Fasting insulin secretion rates were modestly increased in IGT vs. NGT subjects (Table 3). In both groups, total insulin secretion was ∼60% higher with oral compared with IV glucose. Furthermore, the temporal pattern was different between the two groups, being delayed and more persistent in IGT compared with NGT subjects on both tests (Fig. 3). When insulin secretion was modeled on the time course of plasma glucose levels, the dose-response functions describing the dependency of secretory rates on glucose concentration differed markedly both between tests and groups, being steeper in response to oral than IV glucose and steeper in NGT than IGT subjects (Fig. 4). Consequently, the mean slope (or β-cell glucose sensitivity) was approximately double in NGT compared with IGT subjects and double in response to oral compared with IV glucose in both NGT and IGT subjects (Table 3). In contrast, rate sensitivity (or the dependency of insulin secretion on the rate of glucose change) was only higher with oral than IV glucose but not significantly different between the two groups. When controlling for BMI, the absolute rates of insulin secretion (basal and total) were not different between NGT and IGT subjects, whereas β-cell glucose sensitivity remained significantly (P = 0.04) reduced in IGT compared with NGT subjects. Potentiation in response to oral glucose showed a different time course in NGT and IGT subjects, with a significantly delayed rise in the latter compared with the former. Interestingly, the time course of potentiation in response to IV glucose also showed significant changes from baseline, which differed between NGT and IGT (Fig. 5).
The incretin effect on insulin secretion was also calculated as the difference in total (3 h) insulin release between the oral and IV test. This difference amounted to 18 ± 2 nmol/m2 (32 ± 4% of the oral response), was similar in NGT and IGT subjects, and was weakly (and not significantly) related to both plasma GLP-1 and GIP responses. The incretin effect was further assessed in terms of time course by calculating the ratio of the response to oral vs. IV glucose at each time-point. In NGT subjects, the incretin effect on insulin secretion (Fig. 6) was roughly superimposable on the overall secretion profile (Fig. 3), with a peak at 20–40 min and a subsequent smooth decline. In IGT subjects, in contrast, the incretin effect on insulin secretion was very different from the overall secretory profile, showing an early peak at 30 min followed by a return toward baseline in the face of persistently elevated overall secretion rates (Fig. 3). The incretin effect on potentiation also was different (P < 0.0001 by repeated-measures ANOVA) between NGT and IGT subjects, with a positive wave within the first 90 min in the former and an essentially flat response in the latter (Fig. 6). The incretin effect on potentiation was directly related to GLP-1 responses (Fig. 6C), in both NGT and IGT subjects (r = 0.32, P = 0.0007 and r = 0.22, P = 0.03, respectively), but not to GIP responses. When the time courses of incretin effect on insulin secretion, potentiation, and GLP-1 responses were analyzed using cross-correlation, in the NGT group GLP-1 release was in phase with the rise in potentiation (cross-correlation r = 0.32, P < 0.01, n = 110) and insulin secretion rate (cross-correlation r = 0.22, P < 0.01). In the IGT group, GLP-1 was rapidly switched off after 40–50 min, in phase with potentiation (cross-correlation r = 0.22, P < 0.05, n = 100) and insulin secretion (cross-correlation r = 0.38, P < 0.001).
Insulin sensitivity, as estimated using the OGIS index, averaged 386 ± 16 ml·min−1·m−2 in NGT subjects and 325 ± 14 ml·min−1·m−2 in IGT subjects (P = 0.007). In NGT subjects, the metabolic clearance rate of insulin averaged 0.9 ± 0.1 and 1.3 ± 0.2 l·min−1·m−2 on the OGTT and IV test, respectively (P < 0.02); the corresponding values in the IGT group were 0.7 ± 0.1 and 0.9 ± 0.1 l·min−1·m−2 (P < 0.03).
The ability of mathematical modeling to describe the main features of dynamic β-cell function over and above the estimation of absolute rates of insulin secretion has been previously documented by our group (9, 10, 24) and others (2). Oral and isoglycemic IV glucose administration has been successfully employed to quantitate incretin release and effects (30). In this study, we combined mathematical modeling of β-cell function with the experimental design of isoglycemic oral and IV glucose administration in subjects with normal glucose tolerance and in individuals with impaired glucose tolerance, which is a model of insulin resistance and β-cell dysfunction.
Normal glucose tolerance.
Stimulation of the enteroinsular axis by oral glucose had marked effects on all aspects of β-cell function. An average 60–75% increase in the circulating concentrations of GLP-1 and GIP was associated with a clear increment in total insulin secretion (+60% on average) over the amount released during IV glucose. This incretin-mediated insulin release, however, represented only one-third of the total insulin secretory response to oral glucose. Thus the incretin effect itself was relatively small compared with the stimulatory effect of glycemia on insulin secretion. Previous estimates of incretin-mediated insulin release have been larger (29), but they were based on plasma insulin concentrations rather than insulin secretion rates. In fact, in our NGT subjects, the incretin-mediated plasma insulin response averaged 52 ± 5% of total plasma insulin response (P < 0.0001 vs. 35 ± 6% of total insulin secretory response). This discrepancy results from differences in metabolic clearance rate of insulin: with oral vs. IV glucose, portal insulin levels exceed the liver extraction capacity (8) thereby lowering whole body insulin clearance. A direct incretin effect on insulin clearance has not been detected during acute GLP-1 or GIP infusions (4, 27). It is important to emphasize that the contribution of incretin-mediated secretion to total secretion is a function of the oral load: in NGT individuals, larger glucose loads (or mixed meals) will cause greater insulin release without altering the plasma glucose profile, thereby amplifying the role of incretin hormones (30).
The time course of incretin-mediated insulin secretion was essentially parallel to that of total insulin secretory response to oral glucose, indicating that the incretin effect on secretion normally is spread throughout the absorptive period. Furthermore, the incretin-mediated insulin secretion profile was in phase with the GLP-1 time course (Fig. 5) but weakly related to the actual GLP-1 concentrations and unrelated to the GIP concentrations. This finding may be explained by the fact that the incretin effect may involve actions of GLP-1 and GIP that are not reflected in their circulating levels. In fact, GLP-1 is extensively degraded before entering the systemic circulation and may interact with sensory nerve fibers before making contact with endothelial dipeptidyl-dipeptidase IV (1, 13). The importance of the neural pathway for the incretin function of GLP-1 is still unclear, because the incretin effect has been shown to be preserved in type 1 diabetic patients after combined pancreas and kidney transplantation (28). Parasympathetic efferences to the islets, on the other hand, seem to play a role not only in the early (cephalic) phase of insulin secretion, because (in rhesus monkeys) atropine infusion during a meal blunts both acute and second-phase insulin response without important changes in either GLP-1 and GIP plasma levels (5). The lack of quantitative correspondence between incretin-mediated insulin release and GLP-1 and GIP concentrations may also imply that other, unmeasured hormones/substances are involved in the incretin effect as measured under in vivo circumstances.
With regard to the dynamic aspects of β-cell secretion, the incretin effect was associated with a marked increase in β-cell glucose sensitivity (Fig. 4) as well as rate sensitivity (Table 3). Furthermore, potentiation (as the OS/IV ratio) also showed a significant enhancement during the first hour of glucose ingestion, in good synchrony with the time course of GLP-1 levels (Fig. 6). In a previous study utilizing the same β-cell model, GLP-1 infusion (reproducing physiological postmeal GLP-1 levels) during meal ingestion induced an increase in β-cell glucose sensitivity as well as an increment in the potentiation factor (2). An increase in β-cell glucose sensitivity was also demonstrated during combined graded glucose infusion and GLP-1 infusion (4, 18). Thus both our physiological protocol and acute low-dose exogenous GLP-1 infusions concur in demonstrating that the chief mechanism by which incretins enhance insulin secretion is sensitization of the β-cell to glucose. Although the molecular mechanisms by which incretins induce these changes in β-cell function have not been fully elucidated, our study further establishes that the acute recruitment of the enteroinsular axis (i.e., at each meal) is a sufficiently potent stimulus to enhance all modes of β-cell response (response to glucose change, glucose rate of change, and glucose-independent potentiation), each of which contributes to mount a prompt and fully competent insulin secretory response. Of interest is the finding that the potentiation factor showed a significant rise even just with IV glucose (Fig. 5), indicating that glycemia per se can potentiate glucose-induced insulin secretion. This “glucose memory” effect has been previously quantitated with repeated IV glucose infusions (38) but not with a glucose profile simulating the OGTT.
Impaired glucose tolerance.
In our IGT subjects, dysfunction was evident in all modes of β-cell response: the time course of insulin secretion was markedly different from normal (Fig. 3), the slope of the dose-response curve was flatter (Fig. 4), rate sensitivity tended to be lower (although not significantly; Table 3), and the time course of potentiation was distorted; namely, it was delayed in response to oral glucose and pushed up by the higher glucose levels during IV glucose (Fig. 5). This picture is coherent with the findings already reported in a larger group of IGT individuals (10). The GLP-1 response to oral glucose was slightly diminished compared with NGT subjects, but this difference could be accounted for by the difference in BMI. The finding of a mildly depressed GLP-1 response agrees with those of Rask et al. (36) using the OGTT and Toft-Nielsen et al. (39) using a mixed meal. The incretin effect, on the other hand, was largely preserved, because incretin-mediated insulin secretion was only slightly lower (Fig. 6) and β-cell sensitivity and rate sensitivity had similar OS/IV ratios compared with the NGT group (Table 3). Nonetheless, the incretin effect on potentiation was distinctly abnormal, with a rather flat time course in temporal correspondence with an early drop in GLP-1 levels (Fig. 6). In addition, the incretin-mediated secretory profile (Fig. 6) was very different from the overall secretory profile, showing an early peak at 30 min followed by a return toward baseline in the face of persistently elevated total secretion rates (Fig. 3). This indicates that the contribution of incretins to insulin secretion in IGT was exhausted within 1 h of glucose ingestion, after which time secretion was sustained predominantly by hyperglycemia. Thus IGT is characterized by a partial inability of the enteroinsular axis to enhance insulin release, which consequently is driven by glucose levels to a greater extent than in NGT. Whether these abnormalities are the consequence of the chronic, mild hyperglycemia of IGT, and can therefore be reversed by improving glucose tolerance, or whether they are an inherent feature of the IGT state, and thus the cause rather than the consequence of the hyperglycemia, cannot be decided. Also, we cannot rule out the possibility that the incretin effect may be even more different in IGT states when a different stimulus (e.g., a mixed meal) is used.
This work was supported by an EFSD-Novo Nordisk Type 2 Programme Focused Research Grant and funds from the Italian Ministry of University and Scientific Research (prot. 2001065883-001). J. J. Holst's laboratory received support from the EFSD and the Danish Medical Research Council. E. Muscelli is on leave of absence from the Department of Internal Medicine, University of Campinas, Campinas, Brazil.
We thank Sara Burchielli and Silvia Pinnola for technical assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by American Physiological Society