Differentially Expressed Proteins in the Pancreas of Diet-induced Diabetic Mice*

The pancreas is a heterogeneous organ mixed with both exocrine and endocrine cells. The pancreas is involved in metabolic activities with the endocrine cells participating in the regulation of blood glucose, while the exocrine portion provides a compatible environment for the pancreatic islets and is responsible for secretion of digestive enzymes. The purpose of this study was to identify pancreatic proteins that are differentially expressed in normal mice and those with diet-induced type 2 diabetes (T2DM). In this study, C57BL/6J male mice fed a high fat diet became obese and developed T2DM. The pancreatic protein profiles were compared between control and diabetic mice using two-dimensional gel electrophoresis. Differentially expressed protein “spots” were identified by mass spectrometry. REG1 and REG2 proteins, which may be involved in the proliferation of pancreatic beta cells, were up-regulated very early in the progression of obese mice to T2DM. Glutathione peroxidase, which functions in the clearance of reactive oxidative species, was found to be down-regulated in the diabetic mice at later stages. The RNA levels encoding REG2 and glutathione peroxidase were compared by Northern blot analysis and were consistent to the changes in protein levels between diabetic and control mice. The up-regulation of REG1 and REG2 suggests the effort of the pancreas in trying to ameliorate the hyperglycemic condition by stimulating the proliferation of pancreatic beta cells and enhancing the subsequent insulin secretion. The down-regulation of glutathione peroxidase in pancreas could contribute to the progressive deterioration of beta cell function due to the hyperglycemia-induced oxidative stress.

The pancreas is composed of exocrine and endocrine cells. The endocrine portion of the pancreas, also called the islets of Langerhans, is distributed throughout the whole tissue and regulates blood glucose levels through insulin and glucagon secreted by beta and alpha cells, respectively. Exocrine pancreatic cells have been shown to be important in providing a compatible environment for the islets as they are in direct contact. Many studies have shown that the exocrine part of the pancreas secretes proteins that are important in the function and regeneration/replication of beta cells (1)(2)(3)(4).
Type 2 diabetes mellitus (T2DM) 1 is a disease characterized by the inability of the body to properly regulate blood glucose, leading to high blood glucose levels and also disturbed metabolism of other substances such as fat and protein. Obesity and T2DM are closely linked (5). Obesity is caused by a combination of diet, sedentary lifestyle, and genetic factors. In most cases, obesity occurs before the onset of T2DM. Although obesity does not result in diabetes in normal individuals with compensated insulin secretion, it can cause insulin resistance, hyperinsulinemia, and subsequent diabetes in susceptible subjects with an unidentified genetic predisposition. The purpose of this study was to detect any differential pancreatic proteins that are associated with obesity/diabetes in a mouse model of diet-induced diabetes.
Several animal models of diabetes have been used to study obesity-induced diabetes. The commonly used genetic models, such as ob/ob (obese) and db/db (diabetic) mice, have mutations in the leptin structural gene (ob) and in the leptin receptor gene (db), respectively (6 -8). The Zucker diabetic fatty rats (fa/fa) also possess mutations in the leptin receptor gene (9). However, none of these models of diabetes with known genetic defects reflects the disease in humans because these gene mutations are rare in the general population. Experimental animal models of T2DM also can be induced by chemical destruction of a portion of the beta cells or surgical removal of part of the pancreas (10). But these models do not resemble T2DM in humans in which the disease is often preceded by obesity. Diet-induced diabetes in C57BL/6J mice is a model of T2DM developed by dietary manipulations in otherwise healthy animals (11,12). In a hope to detect pathological changes in the pancreas that may represent common conditions associated with T2DM in humans, C57BL/6J mice with diet-induced diabetes were studied.
The pathological changes involved in T2DM with respect to pancreatic islets include hormone secretion dysfunction, proliferation of islet cells during the early phase, and exhaustion and death of islet cells during the later phase of the disease (13,14). During the early phase of hyperglycemia, the islets undergo intensive proliferation from ductal progenitor cells and existing islets (15). Loss of an acute phase insulin response and progressive deterioration of beta cell function coupled with peripheral insulin resistance are common in diabetic subjects with chronic hyperglycemia. In T2DM, the toxicity of hyperglycemia, known as glucotoxicity, has both global and tissue-specific effects, including those on the pancreas. However, these functional studies have not provided information on molecular events that underlie the tissue-specific pathological changes.
In this study, we used a proteomic profiling technique; i.e. two-dimensional gel electrophoresis (2-DE) followed by protein identification with mass spectrometry analysis. The advantage of the proteomic approach over microarray or DNA chip analysis is obvious as proteins are detected in the former, whereas mRNAs encoding the proteins are detected in the latter. Use of the technique and the diet-induced diabetes model may help to identify genes and metabolic characteristics in the pancreas that are causal and/or associated with the disease.

EXPERIMENTAL PROCEDURES
Animals-Three-week-old male C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Upon arrival, 50 mice were weaned onto a high fat diet with 26.6% of calories from carbohydrates, 57.6% from fat, and 15.7% from protein (F1850, Bio-Serv, Frenchtown, NJ) (12), while 20 were weaned onto a standard rodent chow containing 60% of calories from carbohydrates, 14% from fat, and 26% from protein (Prolab RMH 3000, PMI Nutrition International, Inc., St. Louis, MO). Mice were housed, up to two per cage, in a temperature-controlled (22°C) room on a 14-h light, 10-h dark cycle. Food and water were supplied ad libitum. At each time point (2,4,8, and 16 weeks on diet), three mice from each group were sacrificed by cervical dislocation. Mice from each group were selected according to their weights and fasting blood glucose and plasma insulin levels. Mice from the high fat diet displaying the following symptoms: overweight, high fasting glucose levels, and high fasting plasma insulin levels, were selected as diabetic mice, while mice from the normal diet displaying normal profiles were considered as age-matched controls. The pancreata were removed and immediately snap frozen in liquid nitrogen and then stored at Ϫ80°C until further processing. Protocols were approved by the Ohio University Institutional Animal Care and Use Committee and followed federal, state, and local laws.
Mouse Weight Measurements-Mice were weighed biweekly starting at weaning (3 weeks of age) throughout the course of the study. Means for each group at each time point were calculated and plotted.
Fasting Glucose and Insulin Measurements-Fasted blood glucose and plasma insulin concentrations were determined at different time points (2,4,8, and 16 weeks on diet). At the day of measurement, mice were fasted for 8 h starting early in the morning. After fasting, the mice were placed under a heat lamp for about 15 s to vasodilate the tail vein before obtaining the blood. Fasting blood glucose levels were measured using a Lifescan OneTouch glucometer (Johnson & Johnson, New Brunswick, NJ) with a drop of blood from the tip of the tail. At the same time, blood was collected into a heparinized capillary tube. After blood samples were centrifuged at 7000 ϫ g for 10 min at 4°C, plasma insulin concentrations were determined with the Mercodia Ultrasensitive rat insulin ELISA kit (ALPCO, Windham, NH). Values for mouse insulin were adjusted by multiplication by a factor of 1.23 according to the manufacturer's recommendation. The intra-and interassay coefficients of variation were less than 10% and less than 5%, respectively. These data along with weight data were used as phenotypic parameters of the animals.
Protein and RNA Extractions-For protein analysis, mice were sacrificed, and pancreata were removed and homogenized in a protein solubilization buffer (16) containing 7 M urea, 2 M thiourea, 4% CHAPS, and 2.5 l/ml of 40% (w/v) Bio-Lytes (pH 3-10) with a glass homogenizer. Sonication was used to disrupt the tissue. Cell debris and insoluble substances were removed by ultracentrifugation at 150,000 ϫ g for 45 min. Protein concentrations were determined by the Bradford method. All chemicals for protein solubilization and 2-DE were purchased from Bio-Rad unless indicated otherwise. Total RNA was extracted using RNA STAT-60 total RNA/mRNA isolation reagent (Tel-Test, Friendswood, TX) (17). The quantity of extracted RNA was determined by spectrophotometry at an absorbance of 260 nm.
Two-dimensional Gel Electrophoresis-Protein samples were treated for 2 h at room temperature with 5 mM tributylphosphine and 20 l/ml 1 M Tris buffer (pH 8.8) to reduce disulfide bonds and with 15 mM iodoacetamide for 30 min for alkylation (18,19). 2-DE was performed as described previously with slight modifications (20,21). The first dimensional IEF was carried out after protein samples were passively rehydrated on 17-cm IPG strips with a broad pI range of 3-10 (Bio-Rad). The voltage was linearly increased from 0 to 6000 V during the first 10 h followed by 10 h at 6000 V with a current limit of 50 A/strip. Following IEF, the strips were equilibrated in a buffer containing 6 M urea, 2% (w/v) SDS, 0.375 M Tris/HCl (pH 8.8), 20% (v/v) glycerol for 15 min before loading for secondary SDS-PAGE analysis. The strips were sealed on the border of the SDS-PAGE gel using 0.5% low melting point agarose gel. Proteins were separated by size in 15% SDS-polyacrylamide gels (20 ϫ 20 cm) at 26 mA/gel with a maximum voltage of 300 V for 9 h at 4°C. After electrophoresis, the gels were fixed in a solution containing 40% EtOH, 2% acetic acid, 0.0005% SDS overnight followed by washing three times in a buffer of 2% acetic acid and 0.0005% SDS. The gels were stained using a fluorescent dye, SYPRO Orange (1:5000) (Molecular Probes, Eugene, OR), as described previously (21,22).
Quantitative Analysis of Gel Images-Gel images were captured by a laser-scanning device (FLA-3000G, Fuji). The densities of protein spots, which were normalized by the total densities of all valid and matched spots in a set of gels, were quantitatively compared between diabetic and normal control groups using the PDQuest 7.0.1 program (Bio-Rad). For each time point (2,4,8, and 16 weeks on diet), pancreatic protein samples from three diabetic mice and from three control mice were analyzed. Protein "spots" were considered to be differential if the difference between the averages of spot densities from the diabetic mice and the control mice was 2-fold or greater at any time point (23)(24)(25).
Spot Identification by Mass Spectrometry-Protein spots of interest were excised from the gels and transferred to a 96-well plate with addition of 40 l of H 2 O. These samples were delivered to the Proteome Mapping Laboratory at the University of Michigan (www.proteomeconsortium.org). Proteins in gel plugs were digested with trypsin. A fraction of the resulting solution was spotted onto a MALDI target plate for mass spectrometric analysis. If the concentration of sample was too low to obtain a usable spectrum, the solution was purified and concentrated using C 18 microcartridges. The concentrated sample was then utilized. The Applied Biosystems 4700 Proteomics Analyzer (TOF/TOF) was used to obtain mass spectra that were queried against the NCBI database using the Mascot program (www.matrixscience.com/) for identification. Prior to peak list generation, MS spectra were calibrated by trypsin autodigestion peaks and smoothed. The signal-to-noise criterion was set to 25 or greater. The monoisotopic masses were processed for identification. For MS/MS spectra, the peaks were calibrated by default and smoothed. All peaks were deisotoped. The Mascot program has the "Peptide Mass Northern Blot Analysis-Equal amounts of total RNA (15 g) from different samples were resolved by a 1% agarose gel in 3(N-morpholino)propanesulfonic acid/formaldehyde solution and transferred to a positively charged nylon membrane using the NorthernMax kit (Ambion, Austin, TX) according to the manufacturer's instructions. All reagents for probe labeling and Northern blot detection were purchased from Roche Applied Science unless stated otherwise. Probe preparation and signal detection were carried out using a DIG labeling and detection kit following the manufacturer's instructions. Primers (synthesized by Sigma-Genosys, The Woodlands, TX) for probe synthesis were derived from clones that contain mouse Reg2 and Gpx1 cDNAs (Clones BM053708 and BI145182, Open Biosystems, Huntsville, AL). Plasmid DNA was isolated using a miniprep kit (Qiagen, Valencia, CA) according to the instructions of the manufacturer. DIGlabeled probes were synthesized using the Roche Applied Science PCR DIG probe synthesis kit. Labeling and subsequent mRNA signal detection were performed according to the protocols of Roche Applied Science. 5Ј-AGCGGATAACAATTTCACACAGG (sense) and 5Ј-CCCAGTCACGACGTTGTAAAACG (antisense) were used for 700-bp Reg2 probe synthesis. 5Ј-TGGCAGGAGATCAGGCGTCT (sense) and 5Ј-GGCTCTGAACTTGCAGACAAAG (antisense) were used for 1000-bp Gpx1 probe synthesis. The membranes were hybridized with DIG-labeled probes (50 ng/ml hybridization solution) overnight at 42°C. The membranes were then washed twice with 2ϫ SSC, 0.1% SDS at room temperature followed by two washes with 0.1ϫ SSC, 0.1% SDS at 68°C. After treatment in 1ϫ blocking solution, the membranes were incubated with anti-digoxigenin-alkaline phosphatase Fab fragment for 30 min at room temperature and finally detected using CDP-Star chemiluminescent substrate. Images were obtained with VersaDoc (Bio-Rad), and densitometric analysis was then performed using Quantity One software (Bio-Rad).
Statistical Analysis-Results are presented as mean Ϯ S.E. Data were analyzed by single factor analysis of variance (Microsoft Excel). Differences were considered statistically significant if p Ͻ 0.05.

RESULTS
High Fat Diet Induces Obesity-The weights of the mice on the high fat (HF) diet and their control littermates on the normal, low fat (LF) diet were recorded biweekly from weaning at the age of 3 weeks throughout the course of the study (Fig.  1). Significant differences of body weights between these two groups of mice were observed at 2 weeks and throughout the study.
Glucose Homeostasis Is Altered by High Fat Diet-Fasting blood glucose levels in the HF group of mice were significantly higher compared with those in the LF controls at all time points (Fig. 2). At each time point, three obese/diabetic and three control mice were chosen in this study. At the time these mice were sacrificed, their fasting glucose concentrations were 216.1 Ϯ 12.3 and 133.6 Ϯ 4.0 mg/dl for the obese/ diabetic mice and the control mice, respectively. The fasting plasma insulin levels in the HF group were significantly higher than those of the control LF group and showed a significant increase throughout the time course of the study (Fig. 3).
Differential Protein Expression as a Consequence of High Fat Diet-Proteins from pancreata of high fat-fed diabetic and control mice were resolved by 2-DE. Of ϳ450 spots analyzed in the molecular mass range of 5-45 kDa, four protein spots showed a difference of 2-fold or greater when comparing spots from obese/hyperglycemic mice with those from normal controls. The locations of these four spots on a 2-DE gel are shown in Fig. 4. The identifications of these spots by MS analysis are listed in Table I and Table II. Spots A7 and B6 (which were identified as REG1 by MS analysis) had the same two tryptic peptide fragments that were identified by MS/MS analysis (both p Ͻ 0.05) (Table II). Spots A7 and B6 are predicted to be the same protein. This could be possible via different post-translational modification states. Only one spot (B7, GSHPX1) showed a 2-fold or greater decrease in response to high fat feeding. Table III summarizes these spots and their relative levels at the four different time points studied.
Differential Reg2 and Glutathione Peroxidase Expression by Northern Blot-REG2 and GSHPX1 were the two proteins that showed the most dramatic up-regulation and down-regulation, respectively, in the diabetic mice compared with those of the controls (Table III). The genes encoding these proteins were selected for RNA analysis to check the consistency between protein and RNA levels. Reg2 and Gpx1 RNA levels in the pancreata of diabetic and control mice were determined by Northern blot analysis (Fig. 5). The RNA samples were extracted from two animals that were different from the animals used for protein analysis but possessed similar phenotypic characteristics. Consistent with the protein levels, the Reg2 RNA levels were substantially higher, whereas those of Gpx1 were relatively lower in the pancreas of diabetic mice compared with those in the controls. The average ratios of diabetics/controls as determined by signal densities were 15 and 0.46 for Reg2 and Gpx1, respectively. DISCUSSION Diet-induced obesity and diabetes were successfully established in C57BL/6J mice fed a high fat diet. It was unexpectedly observed that some of the mice exhibited hyperglycemia after only 2 weeks on the diet. At each of the four different time points studied, three mice from each group were selected for proteomic analysis based on their weights and fasting glucose and insulin levels that were representative for either obese/diabetic or normal controls.
Several studies have shown that high fat diet and genetic predisposition are critical for the development of hyperinsulinemia and hyperglycemia in these C57BL/6J mice (11,26,27). Defects in insulin reaction to glucose stimulation in these mice suggest that they have a genetically determined impairment in their beta cell function that renders them vulnerable to  develop T2DM when environmental factors, such as high fat diet-induced obesity, exaggerate this predisposition (11). Thus, C57BL/6J mice can serve as a valuable animal model mimicking the human condition of obesity-induced diabetes.
To study altered gene expressions associated with a certain disease state, different methodologies can be used. Others have used a microarray-based approach to ascertain gene expression as a function of progression of an individual from a state of obesity to that of T2DM (28). In the results presented here, a proteomic profiling technique, 2-DE analysis coupled with mass spectrometry, was used. This approach enables direct qualitative and quantitative analysis of proteins as the disease develops (21). Also the documented lack of correlation between mRNA levels and protein levels, which may be due to post-transcriptional and translational regulations, ne-cessitates the need for direct protein profiles (21,25,29,30).
We reported three proteins (REG1, REG2, and GSHPX1) that were differentially expressed in the pancreas between normal and obese/diabetic mice. We also observed variations on the magnitude of the difference at various time points for each of the three proteins. We repeated the experiment with another set of mice receiving the same treatment, and similar results were observed. REG1 and REG2 were two pancreatic proteins that showed an increased level of expression in diabetic C57BL/6J male mice versus controls. Reg genes belong to a gene family that  a The ratios were obtained using the averages from the spot intensities for each group (n ϭ 3 for diabetic and n ϭ 3 for control). a MS/MS probability score as defined by Mascot (www.matrixscience.com). b Significant ion score (p Ͻ 0.05). c The TOF MS data were scored as highly significant in identifying the protein. Also the same protein identification was determined when another MS analysis was performed in a repeated experiment. Also the molecular mass of the spot in the two-dimensional gel that matched the predicted size of the protein was used as supporting data for protein identification.

TABLE II Peptides identified by MS and tandem MS/MS analysis
has four subclasses consisting of Reg1 to Reg4 (31,32). Reg2 and REG4 are found only in mouse and human, respectively. It is believed that pancreatic stone protein (2,(33)(34)(35) and pancreatic thread protein (36) are also encoded by the same human REG1 gene. In our study, two spots (B6 and A7) with a similar molecular mass but different pI values were identified as REG1. Because REG1 is a secreted protein, this pI difference could be due to different glycosylation or other post-translational modifications. However, no definitive post-translational modification could be derived from our mass spectrometry data. The mouse Reg genes are located in a contiguous region in chromosome 6 (37). The mouse has two Reg genes, Reg1 and Reg2, with the mouse Reg1 gene having slightly higher homology to the rat and human REG1 gene than Reg2 (38).
Several studies have been conducted to locate sites of gene expression for Reg1 and Reg2. Reg1 has been found to be expressed in the exocrine pancreas by acinar cells and secreted into pancreatic ducts (1,2,39). Reg1 is also expressed specifically in regenerating and hyperplastic islets but not in normal islets, liver, or brain (40,41). Low expression of Reg1 has also been found in gastric mucosa, kidney (33), and gallbladder (38). In C57BL/6J mice, Reg1 and Reg2 mRNAs have not been detected in islets of normal mice (42). The question whether the expression of Reg1 and Reg2 in islets of C57BL/6J mice can be induced by the obese/diabetic state remains to be answered. It is likely that the increased expression of Reg1 and Reg2 observed in this study is mainly, if not totally, due to their upregulated expression in the exocrine pancreas. Future histoimmunological studies will resolve the site(s) of expression.
The expression of Reg1 and Reg2 genes has been previously associated with diabetes. Overexpression of Reg1 and Reg2 in the pancreas of NOD mice at various degrees of diabetogenesis has been found (43)(44)(45). In this NOD mouse model, expressions of both Reg1 and Reg2 mRNAs are found to be restricted to acinar cells of the exocrine tissue (45). With Reg1-and Reg2-specific cDNA probes, it has been shown that the increased expression of Reg genes in NOD mice is mainly due to increased expression of Reg2 but not Reg1 (43). Interestingly in normal C57BL/6J mice, when Reg1-and Reg2-specific cDNA probes are used, a decline in the expression of Reg1 but unchanged expression of Reg2 has been observed during the normal aging process (42). These results suggest that the expression of these two non-allelic genes may be differential and that they may have different physiological functions. These results also suggest the importance of studies that discern these two genes as many studies have not separated them because of the cross-reactivity of the Reg RNA probes and the antibody used. In our study with 2-DE, we were able to determine the levels of REG1 and REG2 separately. Of four time points in this study, REG2 showed a 2-fold or greater increase in pancreas in diabetic mice at two time points (2 and 16 weeks on diet), whereas REG1 showed an increase at one time point (2 weeks on diet). It is interesting to note that REG2 expression in the pancreas also showed a 2-fold or greater increase in the diabetic state, suggesting the involvement of REG2 during disease progression.
Several lines of evidence support that REG1 protein is a stimulator of beta cell proliferation and neogenesis. Administration of REG protein in 90% depancreatized rats results in a remarkable decrease in blood glucose levels and an increase in beta cell mass (3). A similar proliferative effect has also been observed in NOD mice with administration of REG protein alone or in combination with linomide, an immunomodulatory molecule (4). Transgenic mice that overexpress Reg1 in beta cells exhibit a significant delay in developing diabetes as compared with non-transgenic mice (46). Although Reg1 gene knock-out mice appear phenotypically normal, the average size of islets is significantly smaller than that of wild-type littermates under chemically induced hyperplastic condition (46). It is worth noting that these studies only involve Reg1 not Reg2. However, in the NOD mice study described above, Reg2 but not Reg1 was found to be differentially expressed (43). Both REG1 and REG2 were found to be overexpressed in our study, suggesting the biological importance of REG2. Although REG2 shows high homology to REG1, whether REG2 plays a role similar to that of REG1 remains to be explored.
It has been suggested that REG1 protein plays an important role in both beta cell replication from existing islets and neogenesis from ductal cells (47,48). Beta cells regenerate slowly with a dynamic balance between beta cell replication/neogenesis and apoptosis (15). Beta cells can compensate for insulin resistance or pregnancy by replication and hypertrophy. However, in both type 1 and later stage T2DM, pancreatic beta cells are either destroyed completely or damaged substantially (13).
Islet cells are derived from the epithelial cells of early pancreatic ducts during embryogenesis. In the adult pancreas, neogenesis of islets from ductal cells occurs under both physiological and pathological conditions. Due to its growth-promoting effect on beta cells, the up-regulation of Reg1 gene during disease development suggests that islets undergo active proliferation as a response to hyperglycemia. Thus, the increased expression of REG1 proteins in diabetic mice may be considered as a defense mechanism and therefore a favorable response. However, the specific role of REG1 during the progression of the animal from a normal state to that of diabetic still needs to be firmly established.
Another important result was a decreased level of GSHPX1 in the diabetic mice at later stages of T2DM. GSHPX1, a selenium-containing enzyme, is considered the most abundant isoform in the glutathione peroxidase gene family, which consists of at least five genes. It functions as a GSH-dependent enzyme to remove hydrogen peroxide and fatty acid hydroperoxide. It is believed to be ubiquitously expressed in all mammalian tissues with cytosol and mitochondria subcellular localization (49,50). There is substantial evidence that shows chronic hyperglycemia in diabetes results in overproduction of reactive oxidative species (ROS) and subsequent adverse effects on major molecules and cellular structures, a process known as glucose toxicity (51)(52)(53). The enzymatic antioxidant system that cells utilize to minimize the cellular damages caused by toxic ROS includes various forms of superoxide dismutases, catalase, and glutathione peroxidases.
It has been shown that supraphysiological glucose concentrations result in high levels of intracellular peroxide concentrations in isolated islets damaging beta cell function (54). These adverse effects can be prevented by transient overexpression of Gpx1 gene in islets suggesting an important antioxidant role of GSHPX1. Overexpression of Gpx1 in the transgenic mice renders hearts more resistant to ischemia reperfusion injury compared with those of control mice (55,56). The Gpx1deficient mice are remarkably more sensitive to oxidative stress (57)(58)(59). These studies support the role of GPX1 in the protection against oxidative stress and in disease pathogenesis.
Of many tissues in which oxidative stress can cause damage, islets are especially vulnerable. Extremely low levels of gene expression of intrinsic antioxidant enzymes, especially catalase and glutathione peroxidase, in pancreatic islets compared with those in other tissues render islets especially susceptible to ROS-induced damage (60,61). Down-regulated expression of glutathione peroxidase in pancreas in diabetic mice may result in less capacity to clear ROS. Whereas the demand for antioxidative enzymes is high, this imbalance may contribute to the progressive deterioration of beta cell function in diabetic mice with chronic hyperglycemia.
Using the diet-induced obesity/diabetes mouse model, we discovered three proteins that may play a role in pancreatic dysfunction. Our results suggest that maintenance of the balance between beta cell proliferation and cell death is very important for beta cell function. Also amelioration of the oxidative stress on beta cells may have significant benefit for diabetics.
Acknowledgments-We thank Dr. Douglas T. Kohn for critical reading and comments with the manuscript, Gayle Matheny for conducting the glucose and insulin tests, Dr. Karen Coschigano for providing the pancreas RNA samples, and Dr. Shigeru Okada for collecting some of the tissues used in this study. * This work was supported in part by DiAthegen LLC, the State of Ohio's Eminent Scholar Program that includes a gift from Milton and Lawrence Goll, and the Ohio University Student Enhancement Award program. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.