Effect of Glucagon-Like Peptide-1 on β- and α-Cell Function in Isolated Islet and Whole Pancreas Transplant Recipients

Abstract

Context: Glucose-dependent insulin secretion is often impaired after islet transplantation where reduced β-cell secretory capacity indicates a low functional β-cell mass.

Objective: We sought to determine whether glucagon-like peptide-1 (GLP-1) enhanced glucose-dependent insulin secretion and glucagon suppression in islet recipients, and whether GLP-1 effects were dependent on functional β-cell mass by simultaneously studying recipients of whole pancreas transplants.

Setting: The study was performed in a clinical and translational research center.

Participants: Five intraportal islet and six portally drained pancreas transplant recipients participated in the study.

Intervention: Subjects underwent glucose-potentiated arginine testing with GLP-1 (1.5 pmol · kg⁻¹ · min⁻¹) or placebo infused on alternate randomized occasions, with 5 g arginine injected under basal and hyperglycemic clamp conditions.

Results: Basal glucose was lower with increases in insulin and decreases in glucagon during GLP-1 vs. placebo in both groups. During the hyperglycemic clamp, a significantly greater glucose infusion rate was required with GLP-1 vs. placebo in both groups (P < 0.05), an effect more pronounced in the pancreas vs. islet group (P < 0.01). The increased glucose infusion rate was associated with significant increases in second-phase insulin secretion in both groups (P < 0.05) that also tended to be greater in the pancreas vs. islet group (P = 0.08), whereas glucagon was equivalently suppressed by the hyperglycemic clamp during GLP-1 and placebo infusions in both groups. The GLP-1-induced increase in second-phase insulin correlated with the β-cell secretory capacity (P < 0.001). The proinsulin secretory ratio (PISR) during glucose-potentiated arginine was significantly greater with GLP-1 vs. placebo infusion in both groups (P < 0.05).

Conclusions: GLP-1 induced enhancement of glucose-dependent insulin secretion, but not glucagon suppression, in islet and pancreas transplant recipients, an effect dependent on the functional β-cell mass that may be associated with depletion of mature β-cell secretory granules.

Glucagon-like peptide 1 infusion significantly increases second-phase insulin secretion in islet and pancreas transplant recipients, an effect that is dependent on the β-cell secretory capacity.

Islet transplantation is an emerging mode of β-cell replacement therapy for type 1 diabetes, of which pancreas transplantation is the only established alternative. The function of a whole pancreas graft will in most cases be superior to an islet graft (1), an outcome related to the difference between immediate revascularization of 100% of the β-cell mass contained in a whole pancreas graft and an estimated approximately 25% engrafted β-cell mass in islet transplant recipients (2,3,4). To address the limited β-cell mass that follows most islet transplants, there is presently great interest in the application of glucagon-like peptide-1 (GLP-1)-based therapies (5,6) to potentially improve β-cell mass and function for islet recipients. The incretin hormone GLP-1 is secreted by intestinal L cells in response to nutrient ingestion and binds a G protein-coupled receptor that activates adenyl cyclase, which in the β-cell enhances calcium-dependent pathways involved in insulin secretion, synthetic pathways involved in insulin production, and nuclear pathways involved in cell survival and proliferation (5,6).

GLP-1 normally functions to enhance glucose-dependent insulin secretion and glucagon suppression, effects thought to be dependent in part on GLP-1 receptors present on visceral afferent nerves and α-cells (5) and requiring intact islet innervation that is disrupted in islet and pancreas transplantation (7). Nevertheless, in a canine model of islet autotransplantation, exogenous infusion of GLP-1 increased insulin secretion during a hyperglycemic clamp (8), a result recently confirmed in human islet transplantation (9). However, no data are available on the effects of GLP-1 on islet function under basal conditions or with maximal stimulation by glucose potentiation of a nonglucose secretagogue such as arginine in transplant recipients.

This study aimed: 1) to determine whether GLP-1 enhanced glucose-dependent insulin secretion and glucagon suppression in islet and pancreas transplant recipients under basal and hyperglycemic clamp conditions before and after the injection of arginine; and 2) to evaluate whether GLP-1 effects were dependent on functional β-cell mass. We hypothesized that responses to GLP-1 would be impaired in the islet vs. pancreas group and that differences in responses would be related to functional β-cell mass, anticipating that functional β-cell mass would be lower in islet compared with pancreas transplant recipients (10). Functional β-cell mass was measured as the β-cell secretory capacity derived from a glucose-potentiated arginine test (10,11), and GLP-1 effects were determined by subjects undergoing testing during GLP-1 or placebo infusion on separate randomized occasions. Because differences in β-cell secretory capacity have been associated with inappropriate proinsulin secretion (12,13), we also examined the effect of GLP-1 on PISRs.

Subjects and Methods

Subjects

All subjects had long-standing type 1 diabetes and had undergone transplantation of either isolated islets via intraportal infusion or a whole pancreas graft that has previously been described (1). The pancreas recipients were selected for having portal rather than systemic venous drainage of their grafts to match the portal insulin delivery and hepatic insulin extraction of the islet recipients (14). The study protocol was approved by the Institutional Review Board of the University of Pennsylvania, and all subjects gave their written informed consent to participate.

Subjects were admitted to the Clinical and Translational Research Center (CTRC) on two occasions between 1 wk and 1 month apart. Subjects fasted overnight after 2000 h. Islet recipients who were not insulin-independent withheld any sc insulin. In one islet recipient, iv insulin was administered overnight to maintain normoglycemia and was discontinued before testing. By 0700 h each morning, one catheter was placed in an antecubital vein for infusions, and one catheter was placed retrograde in a hand vein for blood sampling, with the hand placed in a thermoregulated box (∼50 C) to arterialize the venous blood (15).

GLP-1 administration

Lyophilized GLP-1 (7–36 amide; Clinalfa-Merck, Oakland, CA) was reconstituted in 0.9% saline containing 0.25% human serum albumin in a 1 μg/ml solution the evening before study. On the morning of study, basal blood samples were taken at t = −35 and −30 min by 0730 h. Then, GLP-1 or matching placebo infusion was initiated in a randomized fashion, with the alternate condition performed on the subsequent admission. GLP-1 was infused at a rate of 1.5 pmol · kg⁻¹ · min⁻¹, with a double infusion rate for the first 10 min, until the completion of blood sampling at t = 60 min; this rate of administration has been demonstrated to produce supraphysiological concentrations of GLP-1 that augment insulin responses in type 2 diabetes (16,17,18).

Glucose-potentiated arginine test

Prestimulus blood samples were taken at −5 and −1 min before the injection of 5 g arginine over a 1-min period starting at t = 0. Additional blood samples were collected at t = 2, 3, 4, and 5 min after injection. Beginning at t = 10 min, a hyperglycemic clamp technique (19) using a variable rate infusion of 20% glucose was performed to achieve a plasma glucose concentration of approximately 230 mg/dl (13 mmol/liter). Blood samples were taken every 5 min, centrifuged, and measured at bedside with a portable glucose analyzer (YSI 1500 Sidekick; Yellow Springs Instruments, Yellow Springs, OH) to adjust the infusion rate and achieve the desired glucose concentration. After 45 min of the glucose infusion (at t = 55 min), a 5-g arginine pulse was injected again with identical blood sampling.

Biochemical analysis

All samples were collected on ice into tubes containing EDTA, trasylol, and leupeptin, centrifuged at 4 C, separated, and frozen at −80 C for subsequent analysis. Dipeptidyl-peptidase 4 inhibitor (Linco Research, St. Charles, MO) was added to the samples for GLP-1_{7–36 amide} analysis. Plasma glucose was measured in duplicate by the glucose oxidase method using an automated glucose analyzer (YSI 2300; Yellow Springs Instruments). Plasma insulin, glucagon, and proinsulin were measured in duplicate by double-antibody RIA (Linco Research). The insulin assay has a specificity of 100% for human insulin and does not cross-react with intact or des-31,32 human proinsulin (<0.2%). The proinsulin assay has a specificity of 100% for intact human proinsulin and 95% for des-31,32 human proinsulin, and it does not cross-react with human insulin (<0.1%). Plasma GLP-1_{7–36 amide} was measured in duplicate by Luminex immunoassay (Linco Research). All paired samples from the same subject were assayed simultaneously.

Calculations and statistics

The acute insulin, glucagon, and proinsulin responses to arginine (AIR_arg, AGR_arg, and APR_arg, respectively) were calculated as the mean of the 2-, 3-, 4-, and 5-min poststimulus values minus the mean of the prestimulus values (19). The AIR_arg during the 230 mg/dl glucose clamp enables determination of glucose potentiation of arginine-induced insulin release (AIR_pot), a measure of β-cell secretory capacity (20). The AGR_arg during the hyperglycemic clamp allows for determining the glucose inhibition of arginine-induced glucagon release (AGR_inh). The proinsulin-to-insulin (PI/I) ratio was calculated as the molar concentration of proinsulin divided by the molar concentration of insulin × 100. Because the elimination patterns of proinsulin and insulin are distinct (21), estimation of the PI/I ratio within the secretory granules of the β-cell are most reliable after acute stimulation of secretion (22). Thus, we examined the PISR under basal and hyperglycemic conditions from the respective acute PI/I responses to arginine (23).

All data are expressed as mean ± se. Comparisons of responses with GLP-1 vs. placebo infusions within each transplant group were performed with the Wilcoxon matched pairs test, and comparisons between the transplant groups were performed with the unpaired Student’s t test or Mann-Whitney U test as appropriate, using the computer software Statistica (Tulsa, OK). Correlations were performed by least squares linear regression using Origin software (Northampton, MA). Significance was considered at P < 0.05 (two-tailed).

Results

Subject characteristics

The islet and pancreas transplant groups were comparable in sex, age, body mass index, duration of diabetes, and serum creatinine (Table 1). The islet recipients had received 17,073 ± 1,487 islet equivalents (IEs) per kilogram body weight via two or three intraportal infusions between 12 and 24 months before study. Four of the islet recipients had received a prior kidney transplant; the fifth had undergone islet transplantation alone. The pancreas recipients had received a pancreaticoduodenal graft with portal venous drainage between 3 and 48 months before study. Five of the pancreas recipients had received simultaneous pancreas-kidney transplantation; the sixth had undergone pancreas-after-kidney transplantation. All transplant recipients received tacrolimus-based immunosuppression with 0–5 mg of daily prednisone, and in all but one subject in each group, either mycophenolate mofetil or rapamycin. β-Cell exposure to tacrolimus was comparable in the islet and pancreas groups (Table 1) (24). Three of the five islet recipients required exogenous insulin therapy to maintain a near-normal glycosylated hemoglobin (HbA_1c); none of the pancreas recipients required insulin, and they had normal values for HbA_1c that were significantly lower than in the islet group (P = 0.001; Table 1).

Table 1.

Subject characteristics

Subject	Sex	Age (yr)	BMI (kg/m2)	T1D (yr)a	SCr (mg/dl)b	Kidney (Y/N)c	Pred (mg/d)	MMF (mg/d)	Rapa (mg/d)	Tac (mg/d)	Tac (μg/liter)d	HbA1c (%)	Insulin (U · kg−1 · d−1)e	IE/kgf
Islet 1	M	56	19.5	40	0.9	N			6	2	6.5	6.4	0.06	13,087
Islet 2	F	43	21.6	34	1.1	Y	5			5	7.1	6.1	0.09	16,526
Islet 3	F	39	17.2	36	0.9	Y			4	5	7.2	6.5	0	20,081
Islet 4	M	52	21.7	27	1.9	Y	5	1500		3	9.4	5.6	0	14,849
Islet 5	M	50	23.2	39	1.0	Y	5	1000		4	11.4	6.5	0.42	20,820
Mean ± se	3/2	48 ± 3	20.6 ± 1.0	35 ± 2	1.2 ± 0.2	4/1					8.3 ± 0.9	6.2 ± 0.2	0.11 ± 0.08	17,073 ± 1,487
Pancreas 1	M	46	22.5	33	1.3	Y	5	500		3	5.5	4.3
Pancreas 2	M	39	26.4	24	1.8	Y	5	720		6	9.8	5.4
Pancreas 3	F	45	24.0	38	0.9	Y	5	1000		4	12	5.5
Pancreas 4	F	35	21.7	30	0.8	Y	5	1000		6	8.2	4.8
Pancreas 5	F	32	21.3	29	1.0	Y	5			8	8.3	4.9
Pancreas 6	M	45	23.6	40	1.4	Y	5	1000		5	9.7	5.2
Mean ± se	3/3	40 ± 2	23.3 ± 0.8	32 ± 2	1.2 ± 0.2	6/0					8.9 ± 0.9	5.0 ± 0.2g

Data are presented as means ± se. SCr, Serum creatinine; Pred, prednisone; Tac, tacrolimus; Rapa, rapamycin; MMF, mycophenolate mofetil; T1D, type 1 diabetes; BMI, body mass index; M, male; F, female. 

^aT1D duration since diagnosis. 

^bTo convert creatinine to μmol/liter, multiply by 88.40. 

^cPresence (Y) or absence (N) of a kidney transplant for diabetic nephropathy. 

^dTacrolimus 12-h blood trough level at the time of study corrected in the islet recipients based on a portal-to-systemic ratio of 1.54 (Ref. 24). 

^eExogenous insulin requirements at the time of study. 

^fIE transplanted per kilogram recipient body weight, where an IE approximates a standard islet diameter of 150 μm. 

^gP = 0.001 for comparison to the islet group. 

GLP-1 levels

In the islet group, levels of GLP-1_{7–36 amide} increased from 6.9 ± 0.3 to 70.1 ± 7.1 pmol/liter during GLP-1 infusion and did not change during placebo infusion (P < 0.05; Fig. 1). Similarly in the pancreas group, GLP-1_{7–36 amide} levels increased from 10.4 ± 1.7 to 60.8 ± 4.4 pmol/liter during GLP-1 infusion and remained constant during placebo infusion (P < 0.05; Fig. 1).

Figure 1.

Plasma levels of active GLP-1_{7–36 amide} in the islet transplant group (n = 5) during GLP-1_{7–36 amide} (▪) or placebo (□) infusion, and in the pancreas transplant group (n = 6) during GLP-1_{7–36 amide} (•) or placebo (○) infusion.

Plasma glucose and glucose infusion rates

Basal glucose levels were not different before GLP-1 or placebo infusion in the islet (99 ± 6 vs. 100 ± 5 mg/dl) or pancreas (88 ± 4 vs. 85 ± 4 mg/dl) groups. In the islet group, glucose was lower by trend after 30 min of GLP-1 compared to placebo infusion (91 ± 6 vs. 107 ± 6 mg/dl; P = 0.08; Fig. 2A). In the pancreas group, glucose was significantly lower after 30 min of GLP-1 compared to placebo infusion (72 ± 4 vs. 85 ± 3 mg/dl; P < 0.05; Fig. 2B). For the hyperglycemic clamp, the actual glucose plateau achieved was lower during GLP-1 compared to placebo infusion in both the islet (224 ± 3 vs. 238 ± 6 mg/dl; P < 0.05; Fig. 2A) and pancreas (194 ± 7 vs. 232 ± 6 mg/dl; P < 0.05; Fig. 2B) groups. Despite achieving lower glucose plateaus during GLP-1 vs. placebo infusion, the glucose infusion rates during the hyperglycemic clamp were greater during GLP-1 compared to placebo infusion in both the islet (9.8 ± 1.2 vs. 7.6 ± 0.9 mg · kg⁻¹ · min⁻¹; P < 0.05; Fig. 2A) and pancreas (16.1 ± 1.0 vs. 10.2 ± 0.9 mg/dl; P < 0.05; Fig. 2B) groups.

Figure 2.

Plasma glucose levels before and 30 min after GLP-1_{7–36 amide} (filled columns) or placebo (lined columns) infusion, and at the end of the approximately 230 mg/dl hyperglycemic clamp. Against the opposite y-axis, M represents the glucose infusion rate required during the hyperglycemic clamp. A, Results for the islet transplant group (n = 5). B, Results for the pancreas transplant group (n = 6). To convert glucose to mmol/liter, multiply by 0.05551.

Insulin responses to GLP-1

Basal insulin levels were not different before GLP-1 or placebo infusion in the islet (7.1 ± 0.9 vs. 7.2 ± 1.1 μU/ml) or pancreas (9.9 ± 1.1 vs. 11.8 ± 2.3 μU/m) group. In the islet group, insulin levels were significantly greater after 30 min of GLP-1 compared to placebo infusion (21.6 ± 7.8 vs. 7.4 ± 3.5 μU/ml; P < 0.05; Fig. 3A). In the pancreas group, insulin levels were greater by trend after 30 min of GLP-1 compared to placebo infusion (19.3 ± 4.8 vs. 10.2 ± 1.5 μU/ml; P = 0.1; Fig. 3B). The AIR_arg was not different during GLP-1 vs. placebo infusion in the islet group (6.0 ± 2.1 vs. 21.2 ± 12.4 μU/ml; Fig. 3A), and was lower during GLP-1 compared to placebo infusion in the pancreas group (23.8 ± 7.1 vs. 37.2 ± 6.3 μU/ml; P < 0.05; Fig. 3B). In response to the hyperglycemic clamp, second-phase insulin levels were greater during GLP-1 compared to placebo infusion in both the islet (129.9 ± 89.8 vs. 32.3 ± 24.2 μU/ml; P < 0.05; Fig. 3A) and pancreas (298 ± 51 vs. 45.5 ± 12.4 μU/ml; P < 0.05; Fig. 3B) groups. The AIR_pot, however, was similar during GLP-1 and placebo infusions in both the islet (50.1 ± 34.5 vs. 57.7 ± 37.5 μU/ml; Fig. 3A) and pancreas (180 ± 53 vs.153 ± 28 μU/ml; Fig. 3B) groups.

Figure 3.

Plasma insulin before and after bolus injections of 5 g arginine under both basal and approximately 230 mg/dl hyperglycemic clamp conditions. A, Results for the islet transplant group (n = 5) during GLP-1_{7–36 amide} (▪) or placebo (□) infusion. B, Results for the pancreas transplant group (n = 6) during GLP-1_{7–36 amide} (•) or placebo (○) infusion. To convert insulin to pmol/liter, multiply by 7.1750.

Glucagon responses to GLP-1

Basal glucagon levels were higher by trend before GLP-1 compared to placebo infusion in the islet group (78.7 ± 27.2 vs. 64.3 ± 18.9 pg/ml; P = 0.08), and not different in the pancreas group (66.2 ± 7.8 vs. 67.1 ± 11.5 pg/ml). After 30 min of GLP-1 vs. placebo infusion, glucagon levels were no longer different in the islet group (55.7 ± 18.2 vs. 60.8 ± 15.9 pg/ml; Fig. 4A), due to a decrease during GLP-1 infusion, and were lower in the pancreas group (45.0 ± 9.3 vs. 60.0 ± 11.3 pg/ml; P < 0.05; Fig. 4B). The AGR_arg was not different during GLP-1 vs. placebo infusion in the islet group (88.0 ± 17.5 vs. 108 ± 19.9 pg/ml; Fig. 4A), and was lower during GLP-1 compared to placebo infusion in the pancreas group (107 ± 16.2 vs. 129 ± 17.8 pg/ml; P < 0.05; Fig. 4B). In response to the hyperglycemic clamp, glucagon levels were equivalently suppressed during GLP-1 and placebo infusions in both the islet (42.8 ± 7.9 vs. 40.9 ± 6.8 pg/ml; Fig. 4A) and pancreas (38.1 ± 9.7 vs. 38.2 ± 9.9 pg/ml; Fig. 4B) groups. The AGR_inh was lower during GLP-1 compared to placebo infusion in both the islet (54.5 ± 12.9 vs. 86.6 ± 12.0 pg/ml; P < 0.05; Fig. 4A) and pancreas (52.7 ± 6.6 vs.76.2 ± 8.2 pg/ml; P < 0.05; Fig. 4B) groups.

Figure 4.

Plasma glucagon before and after bolus injections of 5 g arginine under both basal and approximately 230 mg/dl hyperglycemic clamp conditions. A, Results for the islet transplant group (n = 5) during GLP-1_{7–36 amide} (▪) or placebo (□) infusion. B, Results for the pancreas transplant group (n = 6) during GLP-1_{7–36 amide} (•) or placebo (○) infusion.

Proinsulin responses to GLP-1

Basal proinsulin levels were not different before GLP-1 or placebo infusion in the islet (7.7 ± 1.5 vs. 7.3 ± 1.5 pmol/liter) or pancreas (9.2 ± 1.1 vs. 10.7 ± 1.5 pmol/liter) group. After 30 min of GLP-1 compared to placebo infusion, proinsulin levels were greater in both the islet (10.7 ± 3.4 vs. 6.5 ± 1.3 pmol/liter; P < 0.05; Fig. 5A) and pancreas (13.0 ± 1.9 vs. 10.0 ± 1.2 pmol/liter; P < 0.05; Fig. 5B) groups. The APR_arg was not different during GLP-1 vs. placebo infusion in either the islet (0.9 ± 0.4 vs. 1.4 ± 0.7 pmol/liter; Fig. 5A) or pancreas (1.1 ± 0.3 vs. 2.5 ± 1.1 pmol/liter; Fig. 5B) group. In response to the hyperglycemic clamp, proinsulin levels were greater during GLP-1 compared to placebo infusion in both the islet (41.9 ± 28.1 vs. 12.9 ± 7.3 pmol/liter; P < 0.05; Fig. 5A) and pancreas (62.7 ± 8.8 vs. 20.2 ± 3.3 pmol/liter; P < 0.05; Fig. 5B) groups. The APR_pot was greater by trend during GLP-1 compared to placebo infusion in the islet group (9.0 ± 6.4 vs. 6.3 ± 5.1 pmol/liter; P = 0.08; Fig. 5A), and was significantly greater during GLP-1 vs. placebo infusion in the pancreas group (54.1 ± 25.5 vs. 20.0 ± 4.5 pmol/liter; P < 0.05; Fig. 5B).

Figure 5.

Plasma proinsulin levels before and after bolus injections of 5 g arginine under both basal and approximately 230 mg/dl hyperglycemic clamp conditions. A, Results for the islet transplant group (n = 5) during GLP-1_{7–36 amide} (▪) or placebo (□) infusion. B, Results for the pancreas transplant group (n = 6) during GLP-1_{7–36 amide} (•) or placebo (○) infusion.

Basal PI/I ratios were not different before GLP-1 or placebo infusion in the islet (15.1 ± 1.8 vs. 13.7 ± 0.8%) or pancreas (13.0 ± 1.2 vs. 13.2 ± 1.0%) group. The PISR during the first injection of arginine was not different during GLP-1 vs. placebo infusion in the islet (2.8 ± 1.4 vs. 1.3 ± 0.4%) or pancreas (0.8 ± 0.3 vs. 1.0 ± 0.4%) group; however, the PISR during the second injection of arginine was significantly greater during GLP-1 compared to placebo infusion in both the islet (3.3 ± 0.9 vs. 1.3 ± 0.3%; P < 0.05) and pancreas (3.9 ± 1.0 vs. 1.9 ± 0.3%; P < 0.05) groups.

Comparison of responses to GLP-1 between islet and pancreas groups

The increase in glucose infusion rate during GLP-1 vs. placebo infusion was greater in the pancreas compared with islet group (P < 0.01). The increase in second-phase insulin during GLP-1 vs. placebo infusion was greater by trend in the pancreas compared with islet group (P = 0.08). The change in response to GLP-1 of all other variables was not different between the islet and pancreas groups.

Correlation between β-cell secretory capacity and GLP-1 effects

The β-cell secretory capacity (AIR_pot on the day of placebo infusion) was significantly lower in the islet compared with the pancreas group (57.7 ± 37.5 vs. 153 ± 28 μU/ml; P < 0.05; Fig. 3), confirming a lower engrafted β-cell mass in recipients of islet rather than whole pancreas transplants. Across both groups, AIR_pot correlated significantly with the GLP-1-induced change in second-phase insulin levels during the hyperglycemic clamp (P < 0.001; Fig. 6A). The percentage increase in second-phase insulin during GLP-1 vs. placebo infusion correlated significantly with the percentage increase in the glucose infusion rate required during the hyperglycemic clamp on the day of GLP-1 compared to placebo infusion (P < 0.01; Fig. 6B).

Figure 6.

A, Correlation of AIR_pot, a measure of β-cell secretory capacity, to the GLP-1_{7–36 amide} induced change in second-phase insulin levels in both islet (○) and pancreas (•) transplant recipients. B, Correlation of the percentage change in second-phase insulin levels to the percentage change in the glucose infusion rate required during the hyperglycemic clamp in both islet (○) and pancreas (•) transplant recipients. To convert insulin to pmol/liter, multiply by 7.1750.

Discussion

These results demonstrate that islet and pancreas grafts respond to the incretin hormone GLP-1 by enhancing glucose-dependent insulin secretion and glucagon suppression. GLP-1 infusion was associated with a reduction in basal glucose and an increase in the glucose infusion rate required during the hyperglycemic clamp, effects indicating improved glucose tolerance. Although the reduction in basal glucose occurred with both an increase in insulin and a decrease in glucagon, the increase in glucose infusion rate was attributable solely to GLP-1-induced augmentation of second-phase insulin secretion because glucagon levels were equivalently suppressed during GLP-1 and placebo infusions by the hyperglycemic clamp. Thus, enhancement of glucose-dependent insulin secretion rather than glucagon suppression may be the primary mechanism by which GLP-1 can improve glucose tolerance for transplant recipients.

The islet transplant group differed from the pancreas transplant group in having a lower β-cell secretory capacity (AIR_pot) associated with less GLP-1-induced augmentation of second-phase insulin secretion and consequently glucose infusion rate during the hyperglycemic clamp. We (2) and subsequently others (4) have reported a β-cell secretory capacity in insulin-independent islet transplant recipients of approximately 25% normal, suggesting the presence of a low engrafted islet mass even in successful cases. A reduced islet β-cell mass can explain the impairment in glucose-dependent insulin secretion observed in islet transplant recipients (2,25), although contribution of some functional β-cell defect remains possible. In type 2 diabetes, impaired glucose potentiation of arginine-induced insulin secretion is improved during GLP-1 infusion (16), evidence that GLP-1 can ameliorate a functional β-cell defect. However, in the present study, GLP-1 did not increase AIR_pot, suggesting that the reduced β-cell secretory capacity in the islet group is best explained by decreased β-cell mass rather than an impairment in function. Importantly, the GLP-1-induced change in second-phase insulin secretion correlated with the β-cell secretory capacity, indicating that the ability of GLP-1 to augment glucose-dependent insulin secretion is dependent on the engrafted β-cell mass (Fig. 6A).

GLP-1 effects on the acute insulin and glucagon responses to arginine can be explained by the differences in prestimulus glucose and insulin levels. AIR_arg was lower with GLP-1 after the decrease in basal glucose, whereas AIR_pot was not different with GLP-1 during the hyperglycemic clamp. AGR_arg was lower with GLP-1 and the associated increase in basal insulin, and AGR_inh was even more suppressed with GLP-1 during the hyperglycemic clamp with markedly elevated second-phase insulin levels. These data are in agreement with those from the rat perfused pancreas and intact mice, where GLP-1 administration did not affect arginine-induced insulin secretion (26), and suggest that GLP-1 influences islet secretory responses to arginine indirectly through altered glucose potentiation of arginine-induced insulin release and insulin suppression of arginine-induced glucagon release.

Excessive secretion of proinsulin relative to insulin, resulting in an elevated molar ratio of PI/I, can accompany both type 2 diabetes and hemipancreatectomy where increased β-cell demand attributed to either reduced β-cell secretory capacity (12,13) or chronic hyperglycemia (27,28) results in β-cell recruitment of immature secretory granules containing an abundance of incompletely processed proinsulin. In the present study, PI/I ratios were not increased in islet recipients with decreased β-cell secretory capacity compared with whole pancreas recipients, a finding explained by the avoidance through low-dose insulin therapy of chronic hyperglycemia in the islet recipients, as evidenced by their near-normal HbA_1c levels. This agrees with data reported by McDonald et al. (29) where basal PI/I ratios were not increased in insulin-dependent islet recipients. Thus, insulin therapy to maintain near-normal glycemia appears to protect subjects with a reduced β-cell secretory capacity from secreting incompletely processed proinsulin.

We were interested in whether β-cell stimulation with GLP-1 would increase the PISRs in transplant recipients. Stimulation of β-cell secretion with arginine and glucose-potentiated arginine led to similarly low PISRs in the islet and pancreas groups during placebo infusion, results consistent with maximal β-cell stimulation in normal subjects where preferential proinsulin secretion does not occur (30). However, whereas GLP-1 infusion did not affect the PISR in response to arginine, the PISR in response to glucose-potentiated arginine was significantly increased by GLP-1 in both the islet and pancreas groups. These data suggest that under hyperglycemic conditions the addition of GLP-1 may lead to preferential proinsulin release through recruitment of immature secretory granules.

Whether overstimulation of partially functioning islet grafts might adversely affect β-cell function in the long term requires further consideration for the potential use of GLP-1 agonist (exenatide) therapy in islet transplantation. Three uncontrolled studies have used exenatide to reduce insulin requirements after islet transplantation (31,32) and to minimize the number of islet donors required to achieve insulin independence (33). Although early reductions in daily insulin dose and increased frequency of insulin independence after single islet donors were observed, increases in insulin dose and a requirement for additional islet infusions occurred in some patients (31,32,33), the latter effects possibly explained by GLP-1-mediated depletion of mature β-cell granules from a low engrafted β-cell mass.

In conclusion, in islet and pancreas transplant recipients, GLP-1 lowered basal glucose by enhancing glucose-dependent insulin secretion and glucagon suppression and increased the glucose infusion rate required to maintain hyperglycemia by augmenting second-phase insulin secretion. These effects were more pronounced in the pancreas compared with the islet group where the GLP-1-mediated increase in second-phase insulin secretion correlated significantly with the β-cell secretory capacity, thus pointing to engrafted β-cell mass as a critical determinant of the potential for GLP-1-based therapeutics for transplant recipients. GLP-1 also increased the PI/I ratio during maximal stimulation of β-cell secretion in the islet and pancreas groups, an effect that could be related to β-cell depletion of mature secretory granules. Thus, while augmenting insulin secretion, pharmacological GLP-1 effects in hyperglycemic islet transplant recipients might contribute to exhaustion of an already low engrafted β-cell mass. Randomized controlled trials of GLP-1-based therapeutics for islet transplantation will be necessary to determine whether these acute effects on β-cell function may prove beneficial for long-term islet graft function and survival.