Ebselen Treatment Prevents Islet Apoptosis, Maintains Intranuclear Pdx-1 and MafA Levels, and Preserves β-Cell Mass and Function in ZDF Rats

Male ZDF rats (ZDF-Lepr^fa/fa) were obtained from Charles River (Wilmington, MA) at 5 weeks of age and fed Purina 5008 chow. At 6 weeks of age, the rats were divided into two experimental groups (n = 12 per group) and treated with 128 mg/kg/day ebselen (Cayman Chemicals, Ann Arbor, MI) by oral gavage (64 mg/kg/day twice daily) for 8 weeks. Carboxymethyl cellulose was chosen instead of dimethyl sulfoxide, a frequently used alternative, because the latter has antioxidant properties. Earlier, we had performed ebselen-dose oral glucose tolerance test (OGTT)-response work over 6 weeks using ZDF animals and ebselen doses of 0, 4, 16, and 64 mg/kg b.i.d., which were associated with fasting glucose levels of 293 ± 21, 257 ± 12, 199 ± 11, and 120 ± 10 mg/dL, respectively (20). The 64 mg/kg b.i.d. dose was the optimal dose for reducing the glucose area under the curve during OGTTs. This is consistent with the work of Brodsky et al. who found no effect on hyperglycemia of gavaging 5 mg/kg twice daily in ZDF rats (12). The control group was treated with 0.5% carboxymethyl cellulose in ddH₂O. Ebselen was prepared by making a 20 mg/mL suspension in 0.5% carboxymethyl cellulose fresh daily with constant mixing during dosing. Nonfasting blood glucose and insulin levels, OGTT, insulin tolerance test (ITT), insulin sensitivity in cultured adipocytes, and morphometric analysis were studied previously (4,8) and described in figure legends. Morphometry of islets and immunohistochemistry studies of markers of oxidative stress, Ki67, Pdx-1, and MafA were performed as previously described (8). Results are expressed as mean ± SEM. Statistical analysis was carried out with InStat Biostatistics 3.0. Comparisons between groups were done by Student t test and ANOVA, where appropriate.

By the end of the study, ebselen treatment had significantly lowered the nonfasting blood glucose level compared with control rats (268 ± 33 vs. 374 ± 33 mg/dL; n = 10 each; P < 0.05) (Fig. 1A). Insulin levels in both groups were maximally increased by 8–9 weeks of age. The control rats thereafter had decreased insulin levels, whereas the ebselen-treated rats maintained higher, relatively constant insulin levels. There was a significant difference in the insulin levels between control and ebselen-treated rats from 10 to 15 weeks of age (P < 0.01) (Fig. 1B). The area under the curve for insulin levels at 10–15 weeks was significantly less in the controls versus the ebselen-treated rats (P < 0.01) (Fig. 1C). HbA_1c levels were greater in controls compared with ebselen-treated rats (11.3% or 100 mmol/mol vs. 8.3% or 67 mmol/mol; P < 0.01) (Fig. 1D). There were no significant differences in body weight between controls and ebselen-treated rats (363 ± 10 vs. 378 ± 12 g at 14 weeks; P = ns) (Fig. 1E).

Fasting blood glucose levels were significantly less in ebselen-treated versus untreated animals (155 ± 14 vs. 259 ± 16 mg/dL). Control rats exhibited higher glucose levels during OGTT (Fig. 2A). Data from OGTT, expressed as change above fasting glucose levels, demonstrated a significant difference when comparing control rats to ebselen-treated rats (149 ± 13 vs. 92 ± 10 mg/dL; delta glucose level at 2 h; P < 0.01) (Fig. 2B).

Blood glucose levels during insulin tolerance tests (ITTs) demonstrated no significant difference in rate of fall in glucose in the first 60 min nor was there a significant difference in the blood glucose nadirs (Fig. 2C). Cultured 3T3L1 adipocytes were used as an independent model to test the effect of ebselen on insulin sensitivity. Only a small tendency toward increased insulin sensitivity with ebselen treatment was observed (Fig. 2D).

Immunoperoxidase staining revealed islets in the control rats that had irregular shape with poor insulin staining [Fig. 3A(a)]. β-Cell mass was doubled in ebselen-treated compared with control rats (6.8 ± 1.25 vs. 13.6 ± 1.53 mg; n = 6; P < 0.05). Islets from the ebselen group exhibited regular shape with intense insulin staining [Fig. 3A(b)]. TUNEL-positive cells were easily detectable in untreated rats [Fig. 3B(a)], whereas β-cell apoptosis was rarely positive in ebselen-treated rats [Fig. 3B(b)]. Quantification of apoptotic cells revealed a statistically significant difference between the groups (control = 4.56% ± 0.31%; ebselen treated = 0.03% ± 0.18%; n = 4; P < 0.001). Ki67 immunostaining to estimate β-cell replication was minimal in both groups (Fig. 3C). The number of cells that costained for both insulin and Ki67 was divided by the total number of insulin-positive cells per each field to obtain the percentage of actively dividing β-cells (6-week ZDF rats = 1.8 ± 0.23% vs. 14-week control ZDF rats = 0.56 ± 0.19%; n = 4; P < 0.05). Ebselen-treated ZDF rat islets did not stain for Ki67 (n = 4). We also used 4-week db/db-GPx animals (8) as positive controls to be certain that the antibody works well. In addition, we preformed negative staining to be certain fluorescein isothiocyanate and CY3 did not stain tissues when omitting the antibody.

To determine the mechanisms by which ebselen preserved β-cell morphology and function, 10 pancreatic sections from each of four control and four ebselen-treated rats were stained for 4-hydroxy-2-nonenal (HNE; an oxidative stress marker for lipid peroxidation) and another oxidative stress marker, 8-hydroxydeoxyguanosine (8-OHdG), which binds to oxidized DNA. The untreated rats exhibited intense cytoplasmic staining of 4-HNE in β-cells [Fig. 4A(a)], which was virtually absent in ebselen-treated rats [Fig. 4A(b)], indicating that ebselen had reduced lipid peroxidation. The untreated rats exhibited 8-OHdG abundantly in nuclei of β-cells [Fig. 4B(a)], whereas the ebselen-treated rats did not (68 ± 3.1% vs. 1.7 ± 0.2%; P > 0.0001) [Fig. 4B(b)].

To examine the effect of ebselen on Pdx-1 and MafA levels, double-labeling immunofluorescence staining was performed using 10 pancreatic sections from each of four control and four ebselen-treated rats for each nuclear transcription factor. The untreated rats had cytoplasmic localization of Pdx-1 with fewer β-cells showing nuclear staining [Fig. 4C(a)] compared with the ebselen-treated rats whose Pdx-1 was more abundant (718 ± 11 vs. 431 ± 10; P < 0.001) and localized exclusively in the nucleus [Fig. 4C(b)]. Expression of MafA was present in the control rats, although fewer β-cells were observed [Fig. 4D(a)]. In ebselen-treated rats, MafA was clearly evident in the nuclei of β-cells and more abundant (849 ± 7 vs. 212 ± 6; P < 0.0001) [Fig. 4D(b)].

The current work was designed to document oxidative stress markers within β-cells of hyperglycemic ZDF rats and to determine whether these markers were associated with apoptosis and absence of two critical insulin transcription factors in the nucleus and, if so, whether these abnormalities could be prevented by the drug ebselen, an oral GPx mimetic. In this study, 8-week treatment of ZDF rats with oral ebselen alone and no other drugs significantly ameliorated fasting hyperglycemia, improved glucose tolerance and postprandial levels of glucose and insulin, and maintained lower levels of HbA_1c compared with nontreated animals. Ebselen prevented accumulation of 4-HNE and 8-OHdG (markers of oxidative stress) in the β-cell and increased β-cell mass twofold over that observed in age-matched untreated animals. The expansion of β-cell mass in ebselen-treated animals was accompanied by robust insulin immunostaining. Control animals had 50% fewer β-cells that stained poorly for insulin. Islet apoptosis was abundant in the untreated animals but rare in the ebselen-treated animals. Intranuclear Pdx-1 and MafA levels were enhanced in β-cells of ebselen-treated animals compared with the untreated animals. ITTs failed to detect effects of ebselen on insulin sensitivity in ZDF rats. Ebselen treatment of cultured adipocytes caused a small increase in insulin-stimulated glucose uptake consistent with a mild antioxidant effect (21). There were no differences in β-cell replication or body weight between the two groups. These favorable outcomes of ebselen therapy indicate that the primary mechanism of action of this drug on β-cells is prevention of oxidative stress. This led to prevention of apoptosis and maintenance of sufficient β-cell mass to compensate for the insulin resistance caused by obesity characteristic in ZDF rats (18) with an associated enhancement of intranuclear Pdx-1and MafA levels.

Several interesting points arise when comparing our ZDF data with those previously published (12,22–26). The fasting glucose levels in the ebselen group, although not normal by human standards, were similar to levels are found in Zucker lean control animals (12). The HbA_1c levels reached were not normalized by ebselen treatment but are consistent with the persistence of impaired glucose tolerance in the high-fat-fed state and the fact that chow was available in the cages 24 h/day during these 8 weeks. The low levels of Ki67 in 14-week-old control and ebselen-treated ZDF animals compared with 6-week-old control animals suggests that by 14 weeks ZDF rats have aged sufficiently to enter a period of low islet replication. This, combined with the impressively high rate of islet apoptosis in the control, but not the ebselen-treated, group contributes heavily to the natural history of severe hyperglycemia in ZDF animals fed a high-fat diet.

Type 2 diabetes (T2D) in humans is a disease that is typified by an inexorable decline over time in β-cell function despite optimal therapy with conventional antihyperglycemic therapeutic agents. The initial cause of hyperglycemia in T2D is considered to involve polygenic/epigenetic defect(s) in β-cell function, which are amplified by states of insulin resistance. Studies in cell lines and animal models have suggested that this relentless decline in β-cell function is secondary to chronic hyperglycemia, which causes chronic oxidative stress (3–5). This line of thinking is supported by the fact that β-cells have poor intrinsic antioxidant activity (9,10) and are therefore at major risk for oxidative damage. Our current study demonstrates for the first time that the oral antioxidant drug ebselen, approved for use in human trails (14–17), prevents not only oxidative stress and the nucleocytoplasmic dislocation of Pdx-1 and MafA in β-cells, but also β-cell apoptosis associated with hyperglycemia in a robust rodent model of T2D. These findings point to the important opportunity to evaluate the potential use of oral ebselen for prevention of continuing deterioration of β-cell function and mass in humans with impaired glucose tolerance.

This work was supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases ROI 38325 (R.P.R.), the U.S. Department of Defense (R.P.R.), the American Diabetes Association Mentor-Based Fellowship Award (R.P.R.), and National Institutes of Health DK-084669 (D.A.B.).

No potential conflicts of interest relevant to this article were reported.

J.M. wrote the manuscript and performed the imaging studies. S.P. and E.O. performed laboratory assays. A.V.H. performed adipocyte studies. D.A.B., S.N.V., and J.S.H. contributed to discussion. C.-q.L. and M.L. performed the animal gavages. R.P.R. wrote the manuscript. R.P.R. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

The authors thank Dr. Peter Butler (Center Director, Larry L. Hillblom Islet Research Center, University of California, Los Angeles) for technical advice in measuring β-cell mass.

There are changes between this version of the manuscript and the version initially posted online on 25 June 2013. Figure 3C here is a new version representing the Ki67 studies. The former was mistakenly chosen. The legend to the figure has been changed, as well as the corresponding text. The changes report that ebselen-treated rats did not stain for Ki67.