Metformin Improves Immunosuppressant Induced Hyperglycemia and Exocrine Apoptosis in Rats

Abstract

Background.

Immunosuppressants are an important cause of posttransplantation diabetes mellitus. We have shown that tacrolimus and sirolimus induce hyperglycemia and hyperinsulinemia in normal rats. We hypothesized that metformin, given concurrently with tacrolimus and/or sirolimus, prevents disturbances in glucose and insulin metabolism.

Methods.

Eight groups (n=6) of normal Sprague-Dawley rats were studied: four groups received tacrolimus, sirolimus, tacrolimus/sirolimus, or control for 14 days, and four more groups received similar treatments along with metformin. Daily glucoses were measured. All rats were administered an oral glucose challenge before sacrifice. Pancreata were analyzed by terminal deoxynucleotide tranferase-mediated dUTP nick-end labeling staining and immunohistochemistry.

Results.

Tacrolimus, sirolimus, and tacrolimus/sirolimus impaired glucose tolerance compared to control. Sirolimus and tacrolimus/sirolimus also increased random blood glucose levels. Sirolimus alone resulted in hyperinsulinemia after oral glucose challenge compared to control. In the sirolimus/metformin and tacrolimus/sirolimus/metformin groups, mean daily random glucose was no longer increased, although the response to glucose challenge was still impaired. Metformin decreased pancreatic exocrine and trended to decrease endocrine apoptosis in tacrolimus/sirolimus group and reduced islet insulin content in sirolimus group.

Conclusions.

This is the first study to show that metformin can improve immunosuppressant-induced hyperglycemia, when administered concurrently, and reduces exocrine apoptosis (reducing the impact on potential islet progenitor cells).

A significant consequence of solid organ transplantation is the development of new-onset diabetes after transplantation or posttransplantation diabetes mellitus (PTDM). Studies have shown reduced graft and host survival and increased fatal and nonfatal cardiovascular disease in kidney transplant recipients with PTDM (1, 2). In 2003, the American Diabetes Association criteria for diagnosis of diabetes in patients who have no transplants were adopted to define PTDM (1). Using these criteria, the incidence of PTDM in a retrospective analysis of consecutive nondiabetic kidney transplant recipients ranged from 56% to 74% (3). The incidence of PTDM varies depending on the organ transplanted, the criteria for definition of PTDM used, the study design, and the time of testing.

Although the etiology of PTDM is multifactorial, immunosuppressant use remains a distinct contributing factor. Supporting this statement, a retrospective study from 2005 showed that immunosuppressant use was the strongest risk factor for development of PTDM (3). Calcineurin inhibitors (CNIs) were clearly associated with increased risk of PTDM in this and many other studies (2–5). Tacrolimus (TAC) has been generally believed to exhibit more diabetogenicity as compared to cyclosporine (6–9). Impaired insulin secretion has been implicated as one of the main causes of CNI-associated PTDM (10, 11). Decreased glucokinase activity and reduced insulin gene expression have been demonstrated in animal models of CNI treatment (12–14). Short-term (24 hr) treatment of human islets with TAC has been shown to increase apoptosis through up-regulation of caspase-3 cleavage and activity (15).

The other commonly used immunosuppressant in solid organ transplantation, sirolimus (SIR), a mammalian target of rapamycin (mTOR) inhibitor, has not been consistently associated with PTDM. Previous retrospective studies of kidney transplant recipients have shown that SIR does not increase risk of PTDM (16, 17). However, more recent data from the U.S. Renal Data system suggest that SIR is independently associated with increased onset of PTDM in patients with kidney transplant (18). Sirolimus has been demonstrated to have effects on insulin secretion as well as insulin sensitivity (19–23). Sirolimus exerts antiproliferative effects on multiple cell types including pancreatic ductal cells and endothelial cells in vitro (20). Sirolimus has been shown to impair in vivo proliferation of islet beta cells in a pregnant murine model and to induce apoptosis in rat and human islets (19, 23). Yet, SIR causes severe hypertriglyceridemia in many individuals, much like that observed with insulin resistance states, and has been shown to worsen hyperglycemia and hyperinsulinemia in diabetic Psammomys obesus rats, suggesting that it also affects insulin action (21, 22).

While much of the focus has been on impaired insulin secretion, we have previously shown that both TAC and SIR can dose-dependently cause hyperinsulinemia (24). This suggests that both CNIs and mTOR inhibitors may contribute to insulin resistance.

Metformin (MET), an oral hypoglycemic agent, is often prescribed to prevent or treat type 2 diabetes in the population without transplant (25). There is increasing interest in the use of MET after solid organ transplantation (26). The safety and efficacy of using MET in the population who underwent transplantation with preexisting diabetes or PTDM has been demonstrated (27). However, the effect of MET on preventing PTDM in transplant recipients has not been well studied. Therefore, we aimed to study whether MET prevents hyperglycemia caused by immunosuppressants, specifically TAC and SIR, in normal male rats.

RESULTS

Treatment with SIR, TAC, or TAC+SIR caused elevated random glucose levels compared to controls (Fig. 1). Metformin treatment significantly reduced mean random glucose levels in both the SIR and TAC+SIR groups (Fig. 1). All immunosuppressant-treated groups experienced reduced weight gain compared to controls (see Materials and Methods and Figure S1, SDC, http://links.lww.com/TP/A740). Weight gain was similar in MET-treated groups compared to their respective non–MET-treated groups (see Materials and Methods; and Figure S1, SDC, http://links.lww.com/TP/A740).

FIGURE 1.

Random blood glucose over time. Daily glucose concentration for the groups treated for 2 weeks are shown. Metformin significantly reduced the random glucoses in rats treated with SIR and TAC+SIR compared to SIR alone or TAC+SIR (A and C, respectively). Data are presented as mean±SEM, n=6. *P<0.05 versus respective time points in TAC, SIR, and TAC+SIR groups. Black triangle=CTRL, white triangle=CTRL+MET, black square=SIR, white square=SIR+MET, black circle=TAC, white circle=TAC+MET, black diamond=TAC+SIR, white diamond=TAC+SIR+MET. TAC, SIR, and TAC+SIR increased daily glucoses compared to control (CTRL) (A, B, and C, respectively).

Treatment with SIR, TAC, or TAC+SIR caused impaired glucose tolerance (see Materials and Methods and Figure S2, SDC, http://links.lww.com/TP/A740). Metformin did not significantly improve glucose tolerance for any treatment group (see Materials and Methods and Figure S2, SDC, http://links.lww.com/TP/A740). Insulin levels in response to oral glucose load in SIR-treated animals were significantly higher than controls (see Materials and Methods and Figure S3a, SDC, http://links.lww.com/TP/A740). Both TAC and TAC+SIR seemed to decrease the insulin response to glucose challenge but did not reach statistical significance (see Materials and Methods and Figure S3a and 3b, SDC, http://links.lww.com/TP/A740). Metformin did not alter the glucose-stimulated insulin response to SIR, TAC, or TAC+SIR.

Immunosuppressant concentrations were measured after 2 weeks of treatment. Sirolimus concentrations were 36±3.8 ng/mL (SIR), 25±1.3 ng/mL (SIR+MET), 28±1.8 ng/mL (TAC+SIR), and 19±2.5 ng/mL (TAC+SIR+MET). Tacrolimus concentrations were 21±1 ng/mL (TAC), 15±0.9 ng/mL (TAC+ MET), 15±2.5 ng/mL (TAC+SIR), and 6T0.9 ng/mL (TAC+SIR+MET). The TAC+SIR treatment induced significantly higher islet and exocrine apoptosis than SIR, TAC, and CTRL (Fig. 2A and B). Metformin treatment seemed to reduce islet apoptosis in the TAC+SIR group but did not achieve statistical significance (Fig. 2A); it did significantly reduce exocrine apoptosis in the TAC+SIR–treated groups (Fig. 2B).

FIGURE 2.

Endocrine and exocrine apoptosis in various groups. White bars=placebo (water), black bars=MET. A, Islet apoptosis in all groups. Apoptotic cells per islet were significantly higher in the TAC+SIR group compared to those in the TAC-treated, SIR-treated, or control (CTRL) group (*P<0.05). Metformin treatment seemed to reduce islet apoptosis but did not reach statistical significance. B, Exocrine apoptosis in all groups. Apoptotic cells per field were significantly higher in the TAC+SIR group compared to those in the TAC-treated, SIR-treated, CTRL, or TAC+SIR+MET–treated group (*P<0.05). C and D, Representative micrograph of exocrine and endocrine apoptosis in the TUNEL-stained section of pancreas from TAC+SIR– and TAC+SIR+MET–treated groups (C) and a micrograph for CTRL and CTRL+MET groups (D).

Islet insulin content trended higher in SIR-treated rats compared to controls (P=0.09) (see Materials and Methods and Figure S4, SDC, http://links.lww.com/TP/A740). Metformin significantly lowered the islet insulin content compared to SIR treatment alone (see Figure S4, SDC, http://links.lww.com/TP/A740). Insulin content was lower in the TAC+SIR group compared to the SIR group, but the TAC group was unchanged compared to controls. The glucagon content, islet area, and islet number were not significantly different between treatment groups or between MET-treated and non–MET-treated animals (data not shown).

DISCUSSION

In this experimental study, we aimed to examine the effects of MET on immunosuppressant-altered glucose metabolism. Consistent with our prior studies, we demonstrated hyperglycemia induced by SIR, TAC, and TAC+SIR in normal rats in this study. In addition, SIR induced hyperinsulinemia, whereas TAC tended to decrease insulin concentration. This suggests that the primary effect of TAC was on insulin secretion, whereas insulin resistance is a key component to SIR-induced hyperglycemia. When MET was coadministered with SIR, there was a significant decrease in random blood glucose levels. However, the responses of blood insulin and glucose levels after a glucose challenge were unchanged. This suggests that SIR-induced whole-body insulin resistance was unchanged by MET, but hepatic glucose output was reduced. Metformin had no effect in TAC-treated animals, consistent with the concept that TAC does not contribute significantly to insulin resistance.

When the immunosuppressants were combined, random blood glucose levels were elevated, slightly higher than SIR alone, and considerably greater than TAC alone. Metformin improved this condition, but not to the degree seen in SIR alone. Importantly, the insulin response to a glucose challenge seemed to show a decrease similar to that with TAC alone and substantially less than with SIR alone. This suggests that the insulin resistance in these animals was similar to SIR alone but that the insulin secretion was decreased by TAC administration.

In our examination of pancreatic islets, SIR alone slightly increased insulin content, which was significantly decreased with MET coadministration, suggesting that MET may have decreased the need for increased insulin production. When TAC was coadministered with SIR, insulin content was significantly decreased, with or without MET, suggesting that TAC had a suppressive effect on insulin production.

We also studied apoptosis, both endocrine and exocrine, because pancreatic ductal cells have been thought to contain islet precursor cells (28–30). Alone, SIR and TAC had minimal effects, whereas TAC+SIR treatment increased both islet and exocrine apoptosis as compared to other groups. Metformin treatment tended to reduce islet apoptosis and significantly reduced exocrine apoptosis in the TAC+ SIR–treated rats.

In conclusion, immunosuppressants can contribute to the development of PTDM through mechanisms that involve decreasing insulin sensitivity (SIR) and suppressing islet function (TAC). Although MET did not seem to appreciably reverse the primary effects of SIR or TAC, its suppression of hepatic glucose output did decrease random blood glucose levels. Metformin may be useful a useful therapeutic agent for decreasing the effects of immunosuppressants in humans, but it will likely not prevent PTDM.

MATERIALS AND METHODS

Animal Care

All animal procedures were approved by the Institutional Animal Care and Use Committee in accordance with The Guide for the Care and Use of Laboratory Animals. Principles of laboratory animal care (NIH publication no. 85–23, revised 1985; http://grants1.nih.gov/grants/olaw/references/phspol.htm) were followed. Adult (176–200 g, 46–50 days old), male Sprague-Dawley rats (Charles River, Wilmington, MA) were kept under 12-hr light and dark periods and fed standard laboratory chow. Before oral glucose challenge at 2 weeks, rats were fasted for 12 hr by removing food but not water.

Experimental Design

Animals (n=6 per group) were given one of eight daily treatment regimens that consisted of two subcutaneous injections and one oral gavage. Two groups received daily injections of TAC (2 mg/kg per day; LC Laboratories, Woburn, MA) and diluent alone (10% ethanol in sunflower oil) subcutaneously for 2 weeks. Two groups received daily SIR (1 mg/kg per day; LC Laboratories, Woburn, MA) and diluent alone subcutaneously for 2 weeks. Two groups received a daily subcutaneous injection of both TAC (2 mg/kg per day) and SIR (1 mg/kg per day) for 2 weeks. Finally, two groups received two injections of diluent daily and served as the control group. These drug doses were chosen based on our previously published studies (24). Of the two groups receiving the same daily injections, one group received a daily oral gavage of MET (200 mg/kg; Medisca, Inc, Plattsburgh, NY), whereas the other group received a daily oral gavage of water. Daily weights were obtained, and nonfasting blood glucose levels were measured by glucometer (FreeStyle Flash; Abbott Diabetes Care, Inc, Alameda, CA) every other day.

After 2 weeks of treatment, an oral glucose challenge was performed on each group (1.5 g/kg by gavage) before being killed. To obtain fasting samples, food was removed from the animals 12 hr before the oral glucose challenge. Baseline glucose and insulin levels were measured from blood taken by tail vein before administration of glucose. Blood was removed from the tail vein at 15, 30, 60, and 120 min after glucose challenge to measure glucose and insulin levels. The groups of rats were then killed under anesthesia, and cardiac blood, pancreas, and liver were harvested.

Insulin and Immunosuppressant Concentrations

Plasma insulin concentration was measured using a high sensitivity radioimmunoassay (31, 32) (Linco Research, St. Charles, MO; limit of sensitivity=0.02 ng/mL). The interassay variations were 3.8% to 10.8% and intra-assay variations were 2.7% to 5.8% for this assay. Tacrolimus and SIR levels were measured by liquid chromatography–tandem mass spectrometry on blood samples taken 24 hr after the last subcutaneous injections at the Clinical Laboratory of The Nebraska Medical Center (33). Intra-assay variations were 3.5% and 3.9% and interassay variations were 2.7% and 5.0% for TAC and SIR, respectively.

Pancreas Pathology

The pancreata were fixed in 10% formalin and set in paraffin blocks. Sections were cut at the Tissue Sciences Facility at the University of Nebraska Medical Center and were stained by terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling of DNA fragmentation (TUNEL) method (DeadEnd Colorimetric TUNEL System; Promega, Madison, WI) for cell apoptosis. Light microscopy was used to study TUNEL-stained sections for the exocrine and islet apoptosis. Hematoxylin-eosin–stained sections were used to study the number of islets of Langerhans and mean islet area. ImageJ software was used to quantify mean islet area and represented as the integrated density (34). The analysts were blinded to the slides before evaluation. Area values were calculated in square pixels and converted to square micrometers.

Each pancreatic section was photographed, and islets were analyzed at ×20 magnification. Islets were evaluated for apoptosis; mean apoptotic cells per field were compared between groups. Exocrine cell apoptosis was quantified by counting TUNEL+ cells in 10 representative randomized fields under ×10 magnification. The mean number of apoptotic cells per field was compared between groups.

Statistical Analysis

Change in weight between groups over time was compared by two-way analysis of variance (ANOVA). Bonferroni posttest was used for analysis. Mean random glucoses were compared by two-way ANOVA with Bonferroni tests for post hoc differences. Glucose and insulin concentration responses to oral glucose load were calculated as area under the curve and were compared by one-way ANOVA with Tukey multiple comparison tests for post hoc differences. Two-way ANOVA was used to compare islet cell apoptosis and islet area between treatment groups and P values less than 0.05 were considered significant for posttest differences. All data are represented by mean-SEM, unless otherwise specified. GraphPad Prism software was used for all the statistical analyses. There were six rats in each group unless otherwise specified.