Objective:

Glucagon-like peptide-1 (GLP-1) (7–36) amide is a glucoregulatory hormone with insulinotropic and insulinomimetic actions. We determined whether the insulinomimetic effects of GLP-1 are mediated through its principal metabolite, GLP-1 (9–36) amide (GLP-1m).

Methods and Procedures:

Glucose turnover during two, 2-h, euglycemic clamps was measured in 12 lean and 12 obese (BMI <25 or >30 kg/m²) male and female subject volunteers with normal oral glucose tolerance test. Saline or GLP-1m were infused from 0 to 60 min in each study. Additionally, seven lean and six obese subjects underwent a third clamp in which the GLP-1 receptor (GLP-1R) antagonist, exendin (9–39) amide was infused from -60 to 60 min with GLP-1m from 0 to 60 min.

Results:

No glucose infusion was required in lean subjects to sustain euglycemia (glucose clamp) during saline or GLP-1m infusions. However, in obese subjects glucose infusion was necessary during GLP-1m infusion alone in order to compensate for a marked (>50%) inhibition of hepatic glucose production (HGP). Plasma insulin levels remained constant in lean subjects but rose significantly in obese subjects after termination of the peptide infusions. During GLP-1R blockade, infusion of glucose was immediately required upon starting GLP-1m infusions in all subjects due to a more dramatic reduction in HGP, as well as a delayed and modest insulinotropic response.

Discussion:

We conclude that GLP-1m potently inhibits HGP and is a weak insulinotropic agent. These properties are particularly apparent and pronounced in obese but only become apparent in lean subjects during GLP-1 (7–36) receptor blockade. These previously unrecognized antidiabetogenic actions of GLP-1m may have therapeutic usefulness.

Introduction

Glucagon-like peptide-1 (7–36) (GLP-1) amide, an incretin hormone produced in enteroendocrine cells of the gut, has potent insulinotropic activities (1). In addition to modulating insulin secretion (2), GLP-1 increases insulin biosynthesis and gene expression (3,4). It also increases the expression of glucose transporter 2 (GLUT2) and glucokinase (3). In rodents, GLP-1 regulates -cell mass by decreasing apoptosis (5) and increasing the proliferation rate of -cells (6,7). Other actions of GLP-1 are the inhibition of glucagon secretion (8), the inhibition of gastrointestinal secretion and motility (9), and reduction in appetite (10) (for recent reviews of the effects of GLP-1, see references (11,12,13,14)). A salient feature of GLP-1 is that its insulinotropic action is glucose-dependent, and unlike the incretin hormone gastric inhibitory polypeptide, is preserved in the diabetic state (15,16). Therefore, GLP-1 is an attractive agent for treating type 2 diabetes. Circulating GLP-1 is inactivated within 1–2 min by the enzyme dipeptidyl peptidase-4 (DPP-4) (17,18) that cleaves the first two amino acids at the N-terminus, resulting in the metabolite GLP-1 (9–36) amide (GLP-1m). GLP-1m is known not to be insulinotropic and has thus been considered to be biologically inactive (19).

We demonstrated earlier that total peripheral glucose utilization during the infusion of GLP-1 (7–36) amide could not be fully explained by its insulinotropic actions alone (20). We attributed the increased glucose utilization to one or more insulinomimetic actions of the peptide, distinct from its insulinotropic effects. It has been shown that infusion of GLP-1 (7–36) amide results in high levels of GLP-1m due to its cleavage by DPP-4 in plasma (18). Indeed, we showed that when steady-state levels were achieved during infusion of full-length peptide, 80% of the circulating plasma levels of peptide were in the form of GLP-1m (20). We hypothesized that the insulinomimetic action of GLP-1 may be due to GLP-1m formation. Therefore, we undertook studies (61 glucose clamp studies) in human subjects with the aim of elucidating the effects of GLP-1m.

Methods and Procedures

Selection of subject volunteers

Twelve lean (BMI 20.6–24.9 kg/m²; %body fat 9.6–30.3, age 19–41 years) and twelve obese (BMI 30.7–47.1 kg/m²; %body fat 31.8–51.4, age 26–50 years) volunteers were enrolled in the study. There were six men and six women in each group; all were white except for one African-American man in each group. The volunteers' clinical characteristics are presented in Table 1. Normal serum electrolytes, liver and renal function, a fasting plasma glucose <6.1 mmol/l (110 mg/dl), oral 75-glucose tolerance test, and hematocrit of 38 in men and 36 in women were requirements for entry. In addition to BMI in each volunteer, percent body fat was determined using dual-energy X-ray absorptiometry (model Prodigy; General Electric, Madison, WI) on the day of the first study. Volunteers' exclusion criteria included any active hepatic, renal, pulmonary, cardiovascular, gastrointestinal, or endocrine disease, any evidence of autonomic insufficiency or ingestion of any medication that might influence carbohydrate metabolism. All methods and procedures were approved by the Massachusetts General Hospital Institutional Review Board. We obtained an investigator initiated new drug application from the Food and Drug Administration for GLP-1m and the GLP-1 receptor (GLP-1R) antagonist, exendin (9–39) amide. All volunteers provided written informed consent in accordance with the Helsinki II Declaration.

Table 1 - Subject characteristics.

Full table (42K)

Euglycemic clamps

All volunteers were weight and activity stable and consumed 200 g of carbohydrates for at least 3 days before the studies. All volunteers had two euglycemic clamps performed: one during saline only infusion and one during GLP-1m infusion. The two studies were carried out no less than 4 weeks apart. Additionally, seven of the lean and six of the obese volunteers had a third euglycemic clamp performed within 12 months in which both GLP-1m and exendin (9–39) were infused. Thus, a total of 61 clamps were performed.

Two hours before the start of each clamp (-120 min), a priming dose of sterile and pyrogen-free [3-³H] glucose (8.7 kBq/kg, 0.236 nCi/kg) was administered, followed by a continuous infusion of [3-³H] glucose (87 Bq/kg/min, 2.36 Ci/kg/min) for the duration of the experiment. Steady-state glucose specific activity was achieved within -30 min. To confirm this and to assess fasting glucose and hormone/substrate levels, four arterialized blood samples were obtained at 10-min intervals starting at -30 min. With the start of the GLP-1m clamp, at time 0, plasma GLP-1m levels were rapidly raised and then maintained stable with a falling primed (0–10 min, with a change at 2-min intervals) followed by a constant (10–60 min, 1.5 pmol/kg/min, 5 ng/kg/min) infusion rate, as described previously, for the infusion of GLP-1 (7–36) amide (21). With the start of the clamp during saline infusion, at time 0, saline was administered by a falling primed constant infusion as described for the GLP-1m infusion.

After all volunteers had undergone both the saline and the GLP-1m studies, we repeated the GLP-1m study in seven of the lean and six of the obese volunteers during blockade of the GLP-1R with exendin (9–39) within 12 months. These subsets were selected because they had not altered their body weight in the preceding year and were willing to participate. Exendin (9–39) was infused at a constant dose of 300 pmol/kg/min (101 ng/kg/min) from -60 to 60 min. This dose has been demonstrated to suppress the insulinotropic actions of GLP-1 (7–36) amide completely (22). GLP-1m and exendin (9–39) were synthesized by the Protein/Peptide Core Facility at Massachusetts General Hospital. The peptides were >99% pure, displayed a single peak on high performance liquid chromatography (HPLC), and the peptide contents were 82 and 85%, respectively. The peptides were subsequently lyophilized in vials under sterile conditions for single use and certified to be both pyrogen free and sterile.

During the combined GLP-1m and exendin (9–39) study, a blood sample was obtained just before the start of the exendin (9–39) infusion at -60 min and thereafter at 10-min intervals until the start of GLP-1m infusion at time 0. Additionally, two muscle biopsies were obtained from the vastus lateralis from each volunteer 10 min before the start of the exendin (9–39) infusion and 10 min before the end of the exendin (9–39)/GLP-1m infusions, as described previously (23). The samples were immediately stored in liquid nitrogen for analysis of specific proteins, the samples were placed in buffer containing 20 mmol/l 4-(2-hydroxyethyl)-1-piperazineethansulfonic acid, pH 7.2, 1 mmol/l EDTA, 250 mmol/l sucrose and Complete protease inhibitor cocktail tablets (Roche Diagnostics, Indianapolis, IN), homogenized with a Brinkman polytron for 45 s at setting six and membranes prepared for western blot analysis (23).

We employed the euglycemic clamp technique (24) to maintain the prevailing plasma glucose levels within 5 mg/dl of the fasting levels of each subject with an infusion of 20% glucose solution (Travenol, Deerfield, IL) whenever necessary. The glucose solution was spiked with [3-³H] glucose to maintain a stable specific activity (hot Ginf) (25). At the start of GLP-1m or saline infusions, blood samples were obtained every 5 min for 120 min for the determination of plasma glucose and every 10 min for the determination of hormones, nonesterified fatty acids (NEFAs), and plasma glucose specific activity.

Analytical techniques

Blood samples were collected using heparinized syringes. Plasma glucose was immediately analyzed using the glucose oxidize method (Beckman Glucose Analyzer II; Beckman Instruments, Fullerton, CA). The remaining blood samples were placed in prechilled test tubes containing peptidase inhibitors as described previously (21). Plasma insulin, C-peptide, glucagon, total GLP-1 (both GLP-1 and GLP-1m), leptin, adiponectin, resistin, NEFA, and specific activity of glucose were analyzed using methodologies described previously (16). The GLP assay uses an antibody to the C-terminus. Therefore, both endogenous and GLP-1m are equally recognized. Glut-4, total Akt-1, and phosphoAKT-1 protein levels were determined from the muscle membranes using western blot analysis (23). The antibodies used were goat antihuman GLUT4 antibody (1:10,000) from Santa Cruz Biotechnology (Santa Cruz, CA), mouse antihuman AKT-1 antibody (1:2,000) from BD Transduction Laboratories (San Diego, CA) and rabbit antihuman phosphoAKT-1(Ser 473) (1:3,000) from Cell Signaling Technology (Beverly, MA).

The rates of total appearance (R_a) and disappearance (R_d) of glucose were calculated according to the nonsteady-state equations of Steele (26), as modified for the use of hot Ginf. The volume of distribution of glucose was assumed to be 210 ml/kg (27). During the basal period, glucose R_a is equal to hepatic glucose production (HGP) because the liver is the principal source of glucose. When glucose was infused (i.e., during the clamp), endogenous glucose production was estimated as the difference between the calculated total glucose R_a and the exogenous glucose infusion rate for the appropriate time interval. Glucose turnover rates were calculated at 10-min intervals from -60 or from -30 to 120 min. The trapezoidal rule was used to calculate the integrated responses over 30-min intervals. The integrated responses were divided by the time interval, which resulted in mean concentration or rates. All data were analyzed using Statistical Analysis System version 9.1 (Cary, NC). Standard methods were used to compute means, s.e.m., and Pearson correlation coefficients. Mixed-model analysis for repeated-measures design was used to analyze hormone and metabolite responses. When a significant effect was observed, changes between clamps in the same group were evaluated using a paired t-test and differences between groups were evaluated using unpaired t-test, using the Bonferroni adjustment. All statistical tests were two tailed. Except where otherwise stated, data are means s.e., and P < 0.05 was regarded as statistically significant.

Results

We first analyzed our data to ascertain the requirement of glucose infusion to maintain stable fasting levels during the euglycemic clamps under the various experimental paradigms. In none of the lean subjects either during the saline or the GLP-1m infusions, glucose was required or plasma insulin levels were altered, but there was a gradual decline in R_a. There was a significant drug-time interaction in R_a between the two studies; however, the decline in R_a during saline or GLP-1m infusions was not sufficient to affect plasma glucose levels significantly, and consequently no glucose infusion was required (Figure 1). In the obese subjects, during the saline infusions, glucose was required in a single subject and then only for the last 15 min of the study (105–120 min, 4.2 mol/kg/min) (Figure 1). In sharp contrast to the lean group, however, glucose infusion was necessary in each and every obese volunteer between 10 and 50 min after the start of the GLP-1m infusion, and the requirement continued for 40–60 min following termination of the infusion (Figure 1). Plasma glucose levels following termination of GLP-1m were slightly higher for the whole of the follow-up period of 60 min compared to the same period following saline infusion and became significantly higher in the obese group, from 90 to 120 min. Plasma insulin levels increased in this group toward the end of the infusion, but had returned to fasting levels by 30 min after discontinuation of the infusion; therefore, the brief insulin elevation cannot be due to the sustained glucose elevation as insulin levels were actually falling as glucose levels were significantly elevated compared to the saline study. C-peptide levels also began to increase after termination of the GLP-1m (Figure 1). There was a significant drug-time interaction in C-peptide levels between the saline and the GLP-1m studies in the obese group (Table 2, P < 0.05). Glucagon, ghrelin, cytokine (leptin, adiponectin, resistin), and NEFA plasma levels were not altered in either group while saline or GLP-1m were being infused (0–60 min) (Table 2). GLP-1 (reflecting endogenous production and exogenous GLP-1m administration) plasma levels were 145–182 pmol/l during the GLP-1m infusion (Table 2, Figure 3). These levels are 1.5-fold those seen after an oral glucose tolerance test. The areas under the curves for GLP-1m in the lean and the obese groups were not significantly different (P = 0.408).

Figure 1.

Plasma glucose levels (top panel), glucose infusion rates (second panel), plasma insulin levels (third panel), C-peptide levels (fourth panel), and rates of appearance of glucose (R_a, bottom panel) in 12 lean (left side) and 12 obese (right side) volunteers who received GLP-1 (9–36) amide or saline from 0 to 60 min (mean s.e.). *Significant difference between the two studies at indicated times. GLP-1, glucagon-like peptide-1.

Full figure and legend (83K)

Figure 3.

Plasma glucagon-like peptide-1 (GLP-1) levels during the GLP-1 (9–36) amide (GLP-1m) and saline infusions in (a) 12 lean and (b) 12 obese volunteers, and during exendin (9–39) amide and GLP-1m infusions in (c) 7 lean and (d) 6 obese volunteers (mean s.e.). *Significant difference between the two studies at indicated times.

Full figure and legend (33K)

Table 2 - Hormone levels in normal (n = 12) and obese (n = 12) groups during the saline or GLP-1m clamps (mean s.e.).

Full table (42K)

Basal R_a levels were similar in lean and obese subjects (13.8 0.69 vs. 12.8 0.57 (lean, not significantly different) and 14.2 0.54 vs. 13.5 0.56 (obese, not significantly different) mol/kg/min) before saline and GLP-1m infusions, respectively (Figure 1). R_a declined in both lean and obese subjects during both saline and GLP-1m infusions but the decline in R_a during GLP-1m infusion was significantly different from the decline during saline infusion (90–120 min, R_a = 12.8 1.0 vs. 9.5 0.3 (lean) and 12.9 0.4 vs. 9.3 0.9 (obese) mol/kg/min, saline vs. GLP-1m, respectively) (Figure 1). Additionally, the difference in the decline of R_a from basal was also significantly different between the two studies in both lean and obese subjects (Figure 1). The greater fall in R_a of the lean group during GLP-1m infusion occurred without any significant changes in plasma insulin levels. It also should be noted that the fall in R_a of the obese subjects precedes the increase in insulin and C-peptide by at least 30 min (Figure 1), indicating that the fall is not a consequence of any changes in plasma insulin levels, and the fall in R_a in the lean subjects occurred even though there was no change in insulin and C-peptide levels.

To ascertain whether the observed decline in R_a during GLP-1m infusion was mediated through the known GLP-1R, we repeated the GLP-1m infusion in half of the lean and obese volunteers during GLP-1R blockade, accomplished by the administration of exendin (9–39) (22). For this study, we infused exendin (9–39) for 60 min before beginning, and continued during, GLP-1m infusion for another 60 min (Figure 2). In both lean and obese subjects, plasma glucose during the entire clamps were similar, and of the 48 paired t-test comparisons, only one point was statistically different between GLP-1m and GLP-1m/exendin (9–39), in the lean group. In the lean volunteers, there was a nonsignificant decline in insulin levels during the first 60 min of the exendin (9–39) infusion compared to during the GLP-1m infusion but on addition of GLP-1m to exendin (9–39), insulin levels began to increase. The time course of C-peptide levels was very similar to that of insulin (Figure 2).

Figure 2.

Plasma glucose levels (top panel), glucose infusion rates (second panel), plasma insulin levels (third panel), C-peptide levels (fourth panel), and rates of appearance of glucose (bottom panel) in seven lean (left side) and six obese (right side) volunteers who received GLP-1 (9–36) amide from 0 to 60 min. Exendin (9–39) amide was infused from -60 to 60 min (mean s.e.). *Significant difference between the two studies at indicated times. GLP-1, glucagon-like peptide-1.

Full figure and legend (67K)

Glucose infusion was immediately necessary during combined exendin (9–39) and GLP-1m infusions in order to maintain euglycemia (Figure 2) and the need for the glucose infusion lasted until the end of the study. In the obese volunteers, insulin levels began to increase within 10 min of initiation of GLP-1m superimposed on the on-going exendin (9–39), and by 30-min insulin levels were statistically different from the insulin levels attained during GLP-1m alone (Figure 2). C-peptide levels also began to increase 30 min after the initiation of the GLP-1m. The pattern of C-peptide response between GLP-1m alone and exendin (9–39)/GLP-1m was similar (Table 3). In both studies, C-peptide levels remained elevated for the 60-min recovery period. Glucose infusion was also necessary immediately after beginning GLP-1m in addition to exendin (9–39) (Figure 2); lasting until the end of the study. Basal Ras were similar with GLP-1m alone as with GLP-1m/exendin (9–39) in both lean and obese subjects, that is exendin (9–39) alone did not have an effect on R_a. However, R_a began to fall rapidly in both groups with infusion of GLP-1m/exendin (9–39). Consequently, there was a significant drug-time interaction between GLP-1m vs. GLP-1m/exendin (9–39) (P < 0.05) for both groups. In the lean group, the fall in R_a was significantly different during the GLP-1m/exendin(9–39) compared to the GLP-1m studies from 25 to 95 min and in the obese group from the 25- to 55-min period (P < 0.05). The R_a fell by at least 50% in both groups from basal levels (P < 0.05): endogenous GLP-1 did not influence R_a during fasting, when the known GLP-1R was blocked by the exendin(9–39) antagonist.

Table 3 - Hormone levels in the normal (n = 7) and obese (n = 6) groups during the GLP-1m and exendin (9–39) clamps (mean s.e.).

Full table (67K)

We also compared the fall of R_a during the three studies to the basal fasting rate. In the lean group, during the saline infusion none of the four consecutive 30-min internal Ras (e.g., 0–30, 30–60, etc.) were significantly different from the basal R_a. However, during GLP-1m infusion Ras for the 60–90 and 90–120 min intervals were significantly suppressed compared to the basal R_a (P < 0.05). In the obese group, during the saline infusion, the Ras for the 30–60 and the 60–90 min were also significantly suppressed compared to the basal R_a (P < 0.05). During GLP-1m infusion, the Ras for 30–60, 60–90, and 90–120 min were suppressed compared to basal (P < 0.05) and the suppression was greater than the corresponding intervals for the saline infusion (P < 0.05). In the lean group, during GLP-1m infusion during GLP-1R blockade, the Ras for the last three consecutive 30-min intervals were significantly suppressed compared to basal (P < 0.05) and in the obese group the Ras for all four consecutive 30-min intervals were significantly suppressed compared to basal R_a (P < 0.05). Furthermore, when GLP-1m was infused, the suppression during the 30-min intervals was significantly greater when the GLP-1R was blocked (P < 0.05). Finally, we compared the changes in plasma insulin levels during the three studies to the basal fasting insulin levels in each group. None of the consecutive 30-min interval insulin levels in any of the studies were significantly different from the basal insulin levels.

Except for the rise in GLP levels from 0 to 60 min, all other hormones and cytokine levels were stable in both lean and obese groups (Table 3). However, as expected, lean and obese subjects showed differences in insulin and leptin values. Exendin (9–39) did not change the GLP-1m clearance rate as the plasma GLP-1 levels during GLP-1m infusion alone and GLP-1m/exendin (9–39) infusions were similar in both studies (Figure 3). Basal R_d in the lean and obese groups during the saline infusions were 13.7 0.55 and 14.5 0.65 mol/kg/min, respectively and was not altered significantly during any studies. Akt-1 total protein levels, and phosphoAkt-1 (Ser 473) levels, measured from the muscle biopsies before starting exendin (9–39) and during the last 10 min of GLP-1m/exendin (9–39) did not change in either group (data not shown). GLUT4 protein levels were significantly higher in the obese group compared to the lean group (P < 0.05) in the basal state. By the last 10 min of the GLP-1m/exendin (9–39) infusion, GLUT4 levels were lower in both groups compared to basal levels; however, the decrease was significant only in the lean group (pre, 583 106: post, 368 61 optical density units/1.5 g homogenized protein, P < 0.05).

Discussion

In an earlier study, we infused GLP-1 (7–36) amide in fasted obese, insulin-resistant subjects and observed an increase in peripheral glucose uptake that was not attributable solely to increased endogenous insulin secretion (20). Notably, in this study of the effects of the insulinotropic GLP-1(7–36) amide, no inhibition of HGP (R_a) was observed beyond that attributable to the increase in plasma insulin. In other studies, the administration of GLP-1 (7–36) amide to fasted dogs (28) or pigs (29) resulted in increased glucose utilization, independent of changes in insulin (28). Furthermore, the direct infusion of GLP-1 (9–36) amide into insulin-resistant dogs with dilated cardiomyopathy enhances myocardial glucose uptake without changes in plasma insulin levels (30). We hypothesized that these extrapancreatic effects of GLP-1 (7–36) amide shown by us in humans, and by others in dogs and pigs is due to GLP-1's primary metabolite, GLP-1m. This study allowed us to evaluate the effects of GLP-1m alone, and to block any effects mediated through the GLP-1R. Plasma levels of GLP-1m attained in this study were similar to those attained during GLP-1 (7–36) amide infusion in our previous study (20).

Here, we show that GLP-1m has insulinomimetic properties in humans that are particularly evident in the insulin-resistant state of obesity and in lean subjects when GLP-1R is blocked by exendin (9–39), a known GLP-1R antagonist (22). The effects of GLP-1m appear to be primarily due to the direct suppression of endogenous (hepatic) glucose production (R_a), and not as a consequence of increased insulin or decreased glucagon secretion. In the obese group, there was a modest but statistically significant increase in insulin and C-peptide which became evident after 30–40 min of GLP-1m infusion and remained evident for some time after its termination. This insulinotropic response resulted in a further suppression of R_a. To investigate the mechanism underlying suppression of R_a, we neutralized the effects of the known GLP-1R by blocking it with exendin (9–39). We expected that GLP-1R blockade would not affect the response to GLP-1m alone, and would establish the presence of a distinct mechanism, separate from any interaction with the GLP-1R. To our surprise, we found that the effect of GLP-1m on R_a was actually augmented in both lean and obese subjects through GLP-1R blockade. This augmentation might be due, in part, to the slight rise in insulin secretion which occurred after the introduction of the GLP-1m during GLP-1R blockade. It is, therefore, likely that GLP-1m is indeed insulinotropic, but this is only evident when the more potent insulinotropic effect of GLP-1 (7–36) amide is negated by GLP-1R blockade (31). In lean subjects, on introduction of GLP-1m during receptor blockade, insulin levels increase and R_a decreased quickly and more robustly that when GLP-1m alone was infused. In our study, we found no effect of GLP-1m on rates of glucose disappearance (R_d) in lean or obese subjects, with or without GLP-1R blockade, which suggests that GLP-1m has no effect on peripheral glucose uptake. The levels of adipokines, ghrelin, glucagon, and NEFA do not change and do not explain why the actions of GLP-1m predominate in the obese state.

Our findings are consistent with a prior study in pigs in which GLP-1m was found to reduce blood glucose, although R_a was not studied (29). In addition, a recent article showed that GLP-1m treatment is associated with a reduction in glycemia after a solid meal, although the mechanism was not determined (32). GLP-1m has been shown to antagonize human GLP-1R expressed in baby hamster kidney cells (33) and in vivo studies in pigs have demonstrated that GLP-1m antagonizes the inhibitory effects of GLP-1 (7–36) amide on antral motility (29). In mice, GLP-1m has no effect on glucose disposal or insulin secretion. This was demonstrated in transgenic mice that had a complete disruption of GLP-1R as well as in normal mice with suppression of insulin secretion with diazoxide (34). Although species differences probably exist, these data in toto suggest that GLP-1m may have a physiological role.

The mechanisms underlying the suppression of R_a by GLP-1m are unknown. The effect of GLP-1m on hepatic glucose metabolism may be due to a novel GLP-1m receptor (GLP-1mR) present on hepatocytes, on the hepatoportal sensor as suggested by Burcelin et al. (35), on -cells, and/or on vagal afferent fibers (36). It is also possible that GLP-1m is interacting with pathways in the central nervous system leading to increased glucose uptake in the liver, as it is now understood that there is a neural component to insulin regulation of HGP, the so-called indirect effect (37,38). Because GLP-1 (7–36) amide has been shown to cross the blood–brain barrier (39) so GLP-1m is similarly likely to do so. Therefore, the effects of GLP-1m on the regulation of HGP may have both an indirect and direct component. That GLP-1m directly affects hepatocytes is the subject of a patent (40). The patent proposes that GLP-1m inhibits the formation of hyperglycemia-induced reactive oxygen species in isolated cultured hepatocytes. The insulinotropic GLP-1 (7–36) amide similarly inhibits hyperglycemia-induced reactive oxygen species formation in hepatocytes but its actions are completely blocked by DPP-4 and neutral endoprotease inhibitors indicating that GLP-1 (9–36) amide, and not GLP-1 (7–36) amide, is the active peptide responsible for the suppression of reactive oxygen species formation in hepatocytes (40). Therefore, there is a direct effect of GLP-1m on hepatocytes, distinct and separate from the hepatoportal sensor, vagus, and brain.

The effects of GLP-1m were previously examined in healthy, normal-weight humans in the presence of GLP-1 (7–36) amide (19). Those investigations found that the insulin secretory response to GLP-1 (7–36) amide was the same whether or not GLP-1m was coinfused. However, HGP was not assessed, and it is likely that the effects of GLP-1m on insulin secretion would not be appreciated due to the more potent insulinotropic properties of GLP-1 (7–36) amide. In contradiction to our results, Zander et al. examined the effect of GLP-1m generated from exogenous GLP-1 (7–36) amide by inhibiting its formation using DPP-4 inhibition (41). There was no significant insulin response evident and glucagon levels did not decrease even with a dose of GLP-1 as high as 9.6 pmol/kg/min. Additionally, although GLP-1 and DPP-4 inhibition resulted in lower blood glucose than GLP-1 alone, this does not necessarily mean that the glucose-lowering effects were from insulinotropic augmentation by GLP-1. It is also unclear whether the exogenously administered GLP-1 recapitulated the physiologic formulation of GLP-1 because plasma levels of intact (N-terminal) GLP-1 and its metabolite GLP-1m were not in the usual ratios observed during infusion studies quoted in the literature (20,32).

The cleavage of histadine and alanine from the N-terminus of GLP-1 (7–36) amide by DPP-4, which results in GLP-1m, has generally been regarded as the limiting factor in the therapeutic usefulness of GLP-1 because it necessitates continuous infusion of the peptide in order to have any sustained effect on blood glucose (42,43). To overcome this deficiency, two approaches are currently under investigation. GLP-1 analogs that are resistant to DPP-4 such as exenatide have been developed, and DPP-4 inhibitors such as sitagliptin and vildagliptin, are now either approved for human use or being tested in humans (44,45). DPP-4 inhibitors have been found in obese diabetic subjects, to decrease postprandial glucose levels, and increase postprandial GLP-1 levels, but not change gastric emptying, plasma insulin levels, glucose disposal, or HGP, when compared to placebo (46). This is surprising because one would expect that the increase in plasma GLP-1 levels, accompanied by hyperglycemia, should have resulted in decreased glucose production and increased glucose disposal. We hypothesize that the lack of effects of the DPP-4 inhibitors on HGP are because of prevention of formation of GLP-1m. Moreover, the small but very statistically significant fall in postprandial glucose levels may be due to effects on plasma levels of other gut hormones and neuropeptides similarly degraded by DPP-4.

We speculate that the GLP-1 (7–36) amide has a dual function. First, it binds to the known GLP-1R on the -cell and stimulates insulin release. Increased insulin levels result in well-known metabolic effects, including reduction of NEFAs, suppression of R_a and decreases in glucagon levels. GLP-1 (7–36) amide is then cleaved by DPP-4, rendering GLP-1 (9–36) amide. This peptide then binds to a second receptor (GLP-1mR) and initiates a series of events that result in further reductions in R_a, independent of plasma glucose levels and the actions of insulin. In the postabsorptive state glucose oxidation is reduced and NEFA oxidation is activated, even in lean subjects. Because the effects of GLP-1m were indeed more pronounced in the obese state we can speculate that GLP-1m is acting in some manner to suppress fatty acid oxidation and thereby inhibiting production of glucose by the liver, through effects on the glucose–fatty acid oxidation cycle (47). It can be further hypothesized that the activity of the putative GLP-1mR can be upregulated in insulin-resistant states. In obesity, the putative receptor is upregulated to compensate for insulin resistance. During blockade of GLP-1-induced insulin release by exendin (9–39), this response is augmented even further. As it has been previously shown that GLP-1 introduced into the brains of mice results in insulin resistance (48), the effects and location of the putative GLP-1mR may be either central, splanchnic, or both. Therefore, exendin 9–39 is a facilitator or sensitizer of GLP-1m effects, both on insulin secretion from the -cell and reduction of R_a in the liver. Finally, it is possible that there may be several GLP-1R subtypes and that the binding affinity to a putative GLP-1mR subtype is different in the obese state. An analogy for different actions of GLP-1 and GLP-1m with respect to action and affinity to receptor subtype is the well described physiological functions of peptide YY(1–36) (PYY (1–36)) and the second biologically active form of PYY, PYY(3–36), which is formed by the removal of amino-terminal dipeptide Tyr-Pro by DPP-4 (49). PYY (3–36) has different biological action than intact PPY (1–36), mediated by different receptors (50).

We therefore propose a working hypothesis for the effects of GLP-1m (Figure 4). The liver is the major site of glucose formation that is inhibited by GLP-1m because of its role in glucose metabolism. We show here that the actions of GLP-1m are not mediated by the GLP-1R as we know it, which suggests that the actions are mediated by an as yet unidentified "novel" receptor. We propose that the GLP-1mR signals mitochondrial events in hepatocytes by inhibiting fatty acid oxidation. It may also be possible, however, that the GLP-1R is differentially expressed in the liver under special circumstances, such as obesity (51). If the GLP-1R is expressed differentially in the livers of obese individuals we suggest that one mechanism for its unique recognition of GLP-1m is its recruitment of a special receptor activity modifying proteins. Such a mechanism of receptor activity modifying protein recruitment is known to change the hierarchy of ligand recognition of the calcitonin gene–related receptor between the ligands, calcitonin, and adrenomedullin (52). The calcitonin gene–related receptor and the GLP-1R are both members of the structurally highly related Family B of seven-membrane spanning G-protein coupled receptors.

Figure 4.

Hypothetical model of putative actions of GLP-1 (9–36) amide (GLP-1m) on the inhibition of glucose formation in the liver of obese human volunteers. Two possible receptor-mediated mechanisms are proposed. The known GLP-1 receptor (GLP-1R) expressed in steatotic livers of obese individuals recruits a specific receptor activity modifying protein (RAMP) that switches ligand specificity of the receptor from that of the insulinotropic hormone GLP-1 (7–36) to the insulinomimetic hormone, GLP-1 (9–36), GLP-1m. Alternatively, there may exist in the liver an as yet unidentified "novel" receptor that recognizes GLP-1m. Both putative receptor signaling pathways would result in the inhibition of glucose formation. GLP-1, glucagon-like peptide-1.

Full figure and legend (9K)

There are some limitations to our study. We did not measure portal venous insulin levels and it is likely that the liver was exposed to higher insulin levels than that measured in the plasma. Therefore, it is possible that part of the observed suppression of HGP is due to a weak and inapparent insulinotropic action of GLP-1m. The GLP-1R antagonist, exendin (9–39), inhibits the insulinotropic action of GLP-1 (7–36) amide by 76% during hyperglycemia (22). Although unlikely, the possibility exists that the insulinotropic effect observed with the infusion of GLP-1m during euglycemia was mediated by the stimulation of the known GLP-1R. We assume that the suppression of endogenous glucose production is due solely to an effect on the liver, but contributions of the kidney, albeit very small, may have also occurred. Regardless of the mechanism(s) responsible, our data demonstrate that GLP-1m is biologically active. This major circulating form of GLP-1 has distinct and potent effects on the inhibition of HGP in the insulin-resistant state of obesity and it is weakly insulinotropic in humans. We therefore conclude that the glucoregulatory effects of GLP-1 are actually mediated, in part, by its metabolite, GLP-1m. These findings corroborate the observations that GLP-1m has cardiac and vascular effects that are equal to the effects of full-length GLP-1 (30,40), and suggest that GLP-1m may have a therapeutic role in the regulation of glucose levels in the diabetic state.

Our findings have implications for the use of DPP-4 inhibitors in humans with type 2 diabetes. By preventing formation of GLP-1m, the effects of GLP-1 on glucose uptake and R_a are abrogated. A very recent study has shown that DPP-4 inhibition resulted in elevated postprandial GLP-1 levels, a very mild decrease in glucose levels and no change in glucose disposal or uptake (46), supporting the observation that the effects of GLP-1m are lost.

Disclosure

The authors declared no conflict of interest.

References

REFERENCES

1. Kreymann B, Williams G, Ghatei MA, Bloom SR. Glucagon-like peptide-1 7-36: a physiological incretin in man. Lancet 1987;2:1300–1304. | Article | PubMed | ISI | ChemPort |
2. Qualmann C, Nauck MA, Holst JJ, Orskov C, Creutzfeldt W. Insulinotropic actions of intravenous glucagon-like peptide-1 (GLP-1) [7-36 amide] in the fasting state in healthy subjects. Acta Diabetol 1995;32:13–16. | Article | PubMed | ChemPort |
3. Wang Y, Perfetti R, Greig NH et al. Glucagon-like peptide-1 can reverse the age-related decline in glucose tolerance in rats. J Clin Invest 1997;99:2883–2889. | Article | PubMed | ChemPort |
4. Fehmann HC, Habener JF. Insulinotropic hormone glucagon-like peptide-I(7-37) stimulation of proinsulin gene expression and proinsulin biosynthesis in insulinoma beta TC-1 cells. Endocrinology 1992;130:159–166. | Article | PubMed | ChemPort |
5. Farilla L, Bulotta A, Hirshberg B et al. Glucagon-like peptide 1 inhibits cell apoptosis and improves glucose responsiveness of freshly isolated human islets. Endocrinology 2003;144:5149–5158. | Article | PubMed | ISI | ChemPort |
6. Stoffers DA, Kieffer TJ, Hussain MA et al. Insulinotropic glucagon-like peptide-1 agonists stimulate expression of homeodomain protein IDX-1 and increase islet size in mouse pancreas. Diabetes 2000;49:741–748. | Article | PubMed | ISI | ChemPort |
7. Perfetti R, Zhou J, Doyle ME, Egan JM. Glucagon-like peptide-1 induces cell proliferation and pancreatic-duodenum homeobox-1 expression and increases endocrine cell mass in the pancreas of old, glucose-intolerant rats. Endocrinology 2000;141:4600–4605. | Article | PubMed | ISI | ChemPort |
8. Creutzfeldt WO, Kleine N, Willms B, Orskov C, Holst JJ, Nauck MA. Glucagonostatic actions and reduction of fasting hyperglycemia by exogenous glucagon-like peptide I(7-36) amide in type I diabetic patients. Diabetes Care 1996;19:580–586. | Article | PubMed | ChemPort |
9. Wettergren A, Schjoldager B, Mortensen PE et al. Truncated GLP-1 (proglucagon 78-107 amide) inhibits gastric and pancreatic functions in man. Dig Dis Sci 1993;38:665–673. | Article | PubMed | ChemPort |
10. Flint A, Raben A, Astrup A, Holst J. Glucagon-like peptide-1 promotes satiety and suppresses energy intake in humans. J Clin Invest 1998;101:515–520. | Article | PubMed | ISI | ChemPort |
11. Amori RE, Lau J, Pittas AG. Efficacy and safety of incretin therapy in type 2 diabetes: systematic review and meta-analysis. JAMA 2007;298:194–206. | Article | PubMed | ChemPort |
12. Vilsboll T, Holst JJ. Incretins, insulin secretion and Type 2 diabetes mellitus. Diabetologia 2004;47:357–366. | Article | PubMed | ChemPort |
13. Riddle MC, Drucker DJ. Emerging therapies mimicking the effects of amylin and glucagon-like peptide 1. Diabetes Care 2006;29:435–449. | Article | PubMed | ChemPort |
14. Holst JJ, Deacon CF. Glucagon-like peptide-1 mediates the therapeutic actions of DPP-IV inhibitors. Diabetologia 2005;48:612–615. | Article | PubMed | ISI | ChemPort |
15. Nauck MA, Heimesaat MM, Orskov C et al. Preserved incretin activity of glucagon-like peptide 1[7-36 amide] but not of synthetic human gastric inhibitory polypeptide in patients with type-2 diabetes mellitus. J Clin Invest 1993;91:301–307. | Article | PubMed | ISI | ChemPort |
16. Elahi D, McAloon-Dyke M, Fukagawa N et al. The insulinotropic actions of glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (7-37) in normal and diabetic subjects. Regul Pept 1994;51:63–74. | Article | PubMed | ChemPort |
17. Deacon CF, Johnsen AH, Holst JJ. Degradation of glucagon-like peptide-1 by human plasma in vitro yields an N-terminally truncated peptide that is a major endogenous metabolite in vivo. J Clin Endocrinol Metab 1995;80:952–957. | Article | PubMed | ChemPort |
18. Kieffer TJ, McIntosh CH, Pederson RA. Degradation of glucose-dependent insulinotropic polypeptide and truncated glucagon-like peptide 1 in vitro and in vivo by dipeptidyl peptidase IV. Endocrinology 1995;136:3585–3596. | Article | PubMed | ISI | ChemPort |
19. Vahl TP, Paty BW, Fuller BD, Prigeon RL, D'Alessio DA. Effects of GLP-1-(7-36)NH2, GLP-1-(7-37), and GLP-1-(9-36)NH2 on intravenous glucose tolerance and glucose-induced insulin secretion in healthy humans. J Clin Endocrinol Metab 2003;88:1772–1779. | Article | PubMed | ChemPort |
20. Egan JM, Meneilly GS, Habener JF, Elahi D. Glucagon-like peptide-1 augments insulin-mediated glucose uptake in the obese state. J Clin Endocrinol Metab 2002;87:3768–3773. | Article | PubMed | ChemPort |
21. Ryan AS, Egan JM, Habener JF, Elahi D. Insulinotropic hormone glucagon-like peptide-1-(7-37) appears not to augment insulin-mediated glucose uptake in young men during euglycemia. J Clin Endocrinol Metab 1998;83:2399–2404. | Article | PubMed | ChemPort |
22. Schirra J, Sturm K, Leicht P et al. Exendin(9-39)amide is an antagonist of glucagon-like peptide-1(7-36)amide in humans. J Clin Invest 1998;101:1421–1430. | Article | PubMed | ChemPort |
23. Hughes V, Fiatarone M, Fielding R et al. Exercise increases muscle GLUT-4 levels and insulin action in subjects with impaired glucose tolerance. American J Physiol 1993;264:E855–E862. | ChemPort |
24. Andres R, Swerdloff R, Pozefsky T, Coleman D. Manual feedback technique for control of glucose concentration. In: Skeggs L Jr (ed). Automation in Analytic Chemistry Medaid: New York, 1966, pp 486–501.
25. Hother-Nielsen O, Beck-Nielsen H. On the determination of basal glucose production rate in patients with type 2 (non-insulin-dependent) diabetes mellitus using primed-continuous 3-3H-glucose infusion. Diabetologia 1990;33:603–610. | Article | PubMed | ChemPort |
26. Steele R. Influences of glucose loading and of injected insulin on hepatic glucose output. Ann NY Acad Sci 1959;82:420–430. | Article | PubMed | ChemPort |
27. Insel PA, Liljenquist JE, Tobin JD et al. Insulin control of glucose metabolism in man: a new kinetic analysis. J Clin Invest 1975;55:1057–1066. | Article | PubMed | ChemPort |
28. Dardevet D, Moore MC, Neal D et al. Insulin-independent effects of GLP-1 on canine liver glucose metabolism: duration of infusion and involvement of hepatoportal region. Am J Physiol Endocrinol Metab 2004;287:E75–E81. | Article | PubMed | ChemPort |
29. Deacon CF, Plamboeck A, Moller S, Holst JJ. GLP-1-(9-36) amide reduces blood glucose in anesthetized pigs by a mechanism that does not involve insulin secretion. Am J Physiol Endocrinol Metab 2002;282:E873–E879. | PubMed | ChemPort |
30. Nikolaidis LA, Elahi D, Shen YT, Shannon RP. Active metabolite of GLP-1 mediates myocardial glucose uptake and improves left ventricular performance in conscious dogs with dilated cardiomyopathy. Am J Physiol Heart Circ Physiol 2005;289:H2401–H2408. | Article | PubMed | ChemPort |
31. Serre V, Dolci W, Schaerer E et al. Exendin-(9-39) is an inverse agonist of the murine glucagon-like peptide-1 receptor: implications for basal intracellular cyclic adenosine 3',5'-monophosphate levels and beta-cell glucose competence. Endocrinology 1998;139:4448–4454. | Article | PubMed | ISI | ChemPort |
32. Meier JJ, Gethmann A, Nauck MA et al. The glucagon-like peptide-1 metabolite GLP-1-(9-36) amide reduces postprandial glycemia independently of gastric emptying and insulin secretion in humans. Am J Physiol Endocrinol Metab 2006;290:E1118–E1123. | Article | PubMed | ChemPort |
33. Knudsen LB, Pridal L. Glucagon-like peptide-1-(9-36) amide is a major metabolite of glucagon-like peptide-1-(7-36) amide after in vivo administration to dogs, and it acts as an antagonist on the pancreatic receptor. Eur J Pharmacol 1996;318:429–435. | Article | PubMed | ISI | ChemPort |
34. Rolin B, Deacon CF, Carr RD, Ahren B. The major glucagon-like peptide-1 metabolite, GLP-1-(9-36)-amide, does not affect glucose or insulin levels in mice. Eur J Pharmacol 2004;494:283–288. | Article | PubMed | ChemPort |
35. Burcelin R, Da Costa A, Drucker D, Thorens B. Glucose competence of the hepatoportal vein sensor requires the presence of an activated glucagon-like peptide-1 receptor. Diabetes 2001;50:1720–1728. | Article | PubMed | ChemPort |
36. Nakabayashi H, Nishizawa M, Nakagawa A, Takeda R, Niijima A. Vagal hepatopancreatic reflex effect evoked by intraportal appearance of tGLP-1. Am J Physiol 1996;271:E808–E813. | PubMed | ChemPort |
37. Buettner C, Patel R, Muse ED et al. Severe impairment in liver insulin signaling fails to alter hepatic insulin action in conscious mice. J Clin Invest 2005;115:1306–1313. | PubMed | ChemPort |
38. Edgerton DS, Lautz M, Scott M et al. Insulin's direct effects on the liver dominate the control of hepatic glucose production. J Clin Invest 2006;116:521–527. | Article | PubMed | ChemPort |
39. Kastin AJ, Akerstrom V, Pan W. Interactions of glucagon-like peptide-1 (GLP-1) with the blood-brain barrier. J Mol Neurosci 2002;18:7–14. | Article | PubMed | ISI | ChemPort |
40. Brownlee M. GLP-1 (9-36) Methods and Compositions. Application number: PCT/US2004/040852. <http://www.wipo.int/pctdb/en/wo.jsp?WO=2005%2F060986&IA=WO2005%2F060986&DISPLAY=STATUS> (Accessed 16 July 2007.)
41. Zander M, Madsbad S, Deacon CF, Holst JJ. The metabolite generated by dipeptidyl-peptidase 4 metabolism of glucagon-like peptide-1 has no influence on plasma glucose levels in patients with type 2 diabetes. Diabetologia 2006;49:369–374. | Article | PubMed | ChemPort |
42. Zander M, Madsbad S, Madsen JL, Holst JJ. Effect of 6-week course of glucagon-like peptide 1 on glycaemic control, insulin sensitivity, and beta-cell function in type 2 diabetes: a parallel-group study. Lancet 2002;359:824–830. | Article | PubMed | ISI | ChemPort |
43. Meneilly GS, Greig N, Tildesley H et al. Effects of 3 months of continuous subcutaneous administration of glucagon-like peptide 1 in elderly patients with type 2 diabetes. Diabetes Care 2003;26:2835–2841. | Article | PubMed | ISI | ChemPort |
44. Ahren B, Pacini G, Foley JE, Schweizer A. Improved meal-related beta-cell function and insulin sensitivity by the dipeptidyl peptidase-IV inhibitor vildagliptin in metformin-treated patients with type 2 diabetes over 1 year. Diabetes Care 2005;28:1936–1940. | Article | PubMed | ISI | ChemPort |
45. Ahren B, Gomis R, Standl E, Mills D, Schweizer A. Twelve- and 52-week efficacy of the dipeptidyl peptidase IV inhibitor LAF237 in metformin-treated patients with type 2 diabetes. Diabetes Care 2004;27:2874–2880. | Article | PubMed | ISI | ChemPort |
46. Vella A, Bock G, Giesler PD et al. Effects of dipeptidyl peptidase-4 inhibition on gastrointestinal function, meal appearance, and glucose metabolism in type 2 diabetes. Diabetes 2007;56:1475–1480. | Article | PubMed | ChemPort |
47. Randle PJ. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle after 35 years. Diabetes Metab Rev 1998;14:263–283. | Article | PubMed | ISI | ChemPort |
48. Knauf C, Cani PD, Perrin C et al. Brain glucagon-like peptide-1 increases insulin secretion and muscle insulin resistance to favor hepatic glycogen storage. J Clin Invest 2005;115:3554–3563. | Article | PubMed | ChemPort |
49. Batterham RL, Bloom SR. The gut hormone peptide YY regulates appetite. Ann NY Acad Sci 2003;994:162–168. | PubMed | ChemPort |
50. Unniappan S, McIntosh CH, Demuth HU, Heiser U, Wolf R, Kieffer TJ. Effects of dipeptidyl peptidase IV on the satiety actions of peptide YY. Diabetologia 2006;49:1915–1923. | Article | PubMed | ChemPort |
51. Ding X, Saxena NK, Lin S, Gupta NA, Anania FA. Exendin-4, a glucagon-like protein-1 (GLP-1) receptor agonist, reverses hepatic steatosis in ob/ob mice. Hepatology 2006;43:173–181. | Article | PubMed | ISI | ChemPort |
52. Sexton PM, Morfis M, Tilakaratne N et al. Complexing receptor pharmacology: modulation of family B G protein-coupled receptor function by RAMPs. Ann NY Acad Sci 2006;1070:90–104. | Article | PubMed | ChemPort |

Acknowledgments

Our appreciation is extended to the men and women who participated in this study. We thank Gail Chin, Benjamin Weirick, Linda Agnello, Christine Lockhart, Carol Stolarski, Teresa Hentosz, Karen McManus for technical assistance, and Melissa Scudder for assistance in the preparation of this article. This work has been supported in part by the National Institutes of Health (NIH) Grant AG 00599, NIH Grant AG 623175, NIH Grant DK 30834, and by the Intramural Research Program of NIA. We gratefully acknowledge the support of the Alan McGavin Geriatric Medicine Endowment of the University of British Columbia.