Metabolic and Pancreatic Effects of Bone Marrow Mesenchymal Stem Cells Transplantation in Mice Fed High-Fat Diet

The purpose of this study was to investigate the effects of multiple infusions of allogeneic MSCs on glucose homeostasis and morphometry of pancreatic islets in high- fat diet (HFD) fed mice. Swiss mice were fed standard diet (C group) or HFD (HFD group). After 8 weeks, animals of HFD group received sterile phosphate-buffered saline infusions (HFD-PBS) or four infusions of MSCs one week apart (HFD-MSCs). Fasting glycemia (FG) was determined weekly and glucose (GTT) and insulin (ITT) tolerance tests were performed 4, 8, 12, and 16 weeks after the infusions of MSCs. The MSCs transplanted mice were classified as responder (FG < 180 mg/dL, 72.2% of transplanted mice) or non-responder (FG > 180mg/dL, 28.8%) Seven weeks after MSCs infusions, FG decreased in HFD-MSCs responder mice compared with the HFD-PBS group. Sixteen weeks post MSCs infusions, GTT and ITT areas under the curve (AUC) decreased in HFD-MSCs responder mice compared to HFD-PBS group. Serum insulin concentration was higher in HFD-PBS group than in control animals and was not different compared with the other groups. The relative volume of α-cells was significantly smaller in HFD-PBS group than in C group and significantly higher in HFD-MSCs-NR than in HFD-PBS and HFD-MSCs-R groups. Cell apoptosis in the islets was higher in HFD-PBS group than in C group, and lower in HFD-MSCs responder mice than in HFD-PBS group and non-responder animals. The results demonstrate the ability of multiple infusions of MSCs to promote prolonged decrease in hyperglycemia and apoptosis in pancreatic islets and increase in insulin sensitivity in HFD fed mice.

Introduction humidity and lighting (12-h dark/light cycle) and with free access to water and standard rodent chow. After 7 days, the animals were randomly assigned into 2 groups, Control group (C) fed standard rodent chow and High-fat diet group (HFD) fed 60% of Kcal as fat (PragSoluções, São Paulo, Brazil). After 8 weeks, animals of HFD group were randomly divided into 2 groups: HF mice, which received sterile phosphate-buffered saline (PBS) infusions (HFD-PBS) and HF mice, which received multiple MSCs infusions (HFD-MSCs), (Fig 1).
Animal protocols were approved by the Ethic Committee of the Federal University of São Carlos (Approval ID number 053/2008).

Fasting glycemia, serum insulin determinations and glucose (GTT) and insulin (ITT) tolerance tests
Fasting glycemia (FG) was determined after 8 weeks of the diet regimen. The animals from which FG was 180 mg/dL were considered hyperglycemic [26]. Fasting glycemia was also determined weekly after MSCs infusions.
GTT and ITT were performed after 8 weeks of the diet regimen and 4, 8, 12, and 16 weeks after the MSCs infusions. After overnight fasting, unanesthetized mice were injected with 1.5g of 50% glucose solution per kg of body weight (BW) by intraperitoneal (i.p) route. Blood samples were collected before glucose injection and 30, 60, 90, and 120 min after the injection. ITT was performed after overnight fasting, in unanesthetized mice injected i.p. with human insulin 0.75 U/kg BW. Blood samples were collected before injection and 15, 30, 60, and 90 min after insulin challenge.
Blood glucose samples were collected from the tail vein. Blood glucose levels were measured by Accu-Check glucose meter (Roche Diagnostic, Indianapolis, USA) and serum insulin concentration was determined by ELISA (Millipore Corporation, Billerica, MA, U.S.A.) according to the manufacturer's protocol.

Isolation, culture and characterization of bone marrow MSCs
The bone marrow cells were isolated from the tibias and femurs of male Wistar rats, aged 6-8 weeks. After washing and centrifugation, cells were resuspended in a α-minimum medium (Gibco, Auckland, New Zealand) supplemented with 15% of fetal bovine serum (Hyclone, Logan, UT, USA) and 100 μg/ml penicillin (Gibco), 100 μg/ml streptomycin (Gibco) and 2 mM L-glutamine (Gibco), and plated in a 150x20mm dish (density of 5 x 10 6 nucleated cells/ dish). The cells were incubated in a humidified atmosphere containing 95% air and 5% CO 2 at 37°C. The non-adherent cells were removed by changing the medium after 3-day culture. Confluent primary cultures were washed with PBS and lifted by incubation with trypsin (Sigma-Aldrich, Saint Louis, MO, USA) at 37°C for 5 minutes. RPMI1640 (Gibco), supplemented with 10% fetal bovine serum (HyClone, Logan, UT, Canada) was added to neutralize the excess trypsin. The cells were centrifuged and seeded into a 150x20mm dish (density of 5x10 6 nucleated cells/dish). Subsequent passages were performed similarly until the fifth passage.
To characterize MSCs, one aliquot of trypsinised MSCs from the fifth passage was stained with phycoerythrin (PE)-conjugated monoclonal antibodies against CD31, CD45 or fluorescein isothiocyanate (FITC)-conjugated monoclonal antibodies against CD 11b, CD44, and CD29 (Becton-Dickinson/BD, San Jose, CA, USA) for 15 minutes at room temperature. Stained cells were washed and analyzed immediately on a FACSort flow cytometer using CellQuest software (BD).

Multiple administration of MSCs
Bone marrow MSCs between fourth to fifth passages were used for multiple infusions in HFD mice. HFD-MSCs group received four i.p. infusions of 5 -8x10 6 MSCs resuspended in 200μL of PBS, with one week of interval. HFD-PBS group received 200μL of PBS i.p. (Fig 1).

Pancreas immunohistochemistry
Sixteen weeks after the last MSCs infusion, animals were euthanized and pancreas was rapidly removed, fixed in 10% neutral-buffered formalin, dehydrated through graded ethanol passages, cleared in xylene and embedded in paraffin wax. Five μm sections were deparaffinized and rehydrated in a graded series of ethanol washes. Subsequently, 3 discontinuous sections of 5μm pancreas sample were obtained and stained by hematoxylin and eosin (HE) and immunodetection of insulin, glucagon, Ki67 and caspase. Only for Ki67 and caspase detection, antigen retrieval was performed using citrate buffer. Endogenous peroxidase was quenched by hydrogen peroxide 3% (Peroxidase-Blocking reagent, DAKO Cytomation, Fort Collins, CO, USA). Then, sections were incubated with PBS/ BSA 1% to prevent unspecific staining. Next, primary antibodies were applied to the sections. In this study, four primary antibodies diluted in PBS/1% BSA were used: 1) rabbit polyclonal anti-insulin (dilution 1:100; Santa Cruz Biotechnology, CA, USA), 2) mouse monoclonal anti-glucagon (dilution 1:2000; Abcam, Cambridge, UK), 3) rabbit polyclonal anti-caspase (dilution 1:500; Abcam, Cambridge, UK) or 4) rabbit monoclonal anti-Ki67 (dilution 1:100; Abcam, Cambridge, UK). Sections were then incubated with LSAB TM + Kit/HRP (DAKO Cytomation, Fort Collins, CO, USA). The slides were stained with 3.3'diaminobenzidine (DAB-DAKO) according to the manufacturer's instructions. Finally the sections were and counterstained with Harrys hematoxylin stain. Images were captured and analyzed by a system composed of a video cam (Carl Zeiss AxionCam-MRc) coupled to an Axion Vision Vert A1 microscope (Car Zeiss GmbH, Jena, German), linked to a microcomputer with a Axiovision 4.8 software. Captures were performed using X20 objective magnification.

Determination of islet size and quantification of β and α cells volume
For morphometric pancreas analysis, fifteen islets from each mouse were randomly chosen. Each islet was evaluated to obtain the total islet area and the area positive for insulin or glucagon within each islet using Image-Pro Plus software (Media Cybernetics, Silver Spring, MD). Then, the relative volume of β-cell or α-cell was calculated as the percentage area positive for insulin or glucagon within the islets [25].

Proliferation and apoptosis
Pancreatic islet apoptosis and proliferation was expressed as caspase-3 or Ki67 positive cells, respectively, per islet area. Software Image-Pro Plus (Media Cybernetics, Silver Spring, MD) was used to measure areas and Image J (National Institutes of Health, Bethesda, MD, USA) was used to count the positive cells.

Statistical analysis
Data are presented as mean ± SEM. Statistical analysis was done by unpaired and paired Student's t tests and analysis of variance (ANOVA). Tukey's multiple comparisons test was used for post hoc analysis of between-group comparisons. Data are presented as means ± SEM. P values < 0.05 were considered statistically significant.
All statistical analyses were performed by GraphPad Prism 5.0 software (GraphPad software, Inc., California, USA)

Characterization of allogeneic bone marrow derived MSCs
Allogeneic BM-MSCs expressed typical mesenchymal cell markers (Fig 2A) as previously described [29]. MSCs stained positive for CD44 and CD29 while negative for CD31, CD45 and CD11b. MSCs were further characterized by their osteogenic and adipogenic differential potential (Fig 2B-2D).

Metabolic characteristic of the animals
Eight weeks after the HFD regimen, mice fed high-fat diet had significantly higher FG than that of control animals (210.5±6.8 vs 133.22±4.8 mg/dL, p <0.05) until the end of the experimental period. These animals had significantly higher glycemia response to both glucose and insulin injections than that of control mice (Fig 3A and 3B).

MSCs multiple injections decrease blood glucose levels and improve glucose and insulin response
Sixteen weeks after BM-MSCs fourth infusion, 72.2% (n = 13) of the animals exhibited fasting glycemia lower than 180 mg/dL (henceforth called responders or HFD-MSCs-R), while 27.8% (n = 5) did not respond to MSCs therapy-non-responders (or HFD-MSCs-NR). Body weight of HFD fed animals was significantly higher than control animals beginning at the 8 th week until the end of the experimental period. At the end, the body weight of HFD-PBS, HFD-MSCs-R and HFD-MSCs-NR was not different (Table 1).    Serum insulin levels were significantly higher in HFD-PBS than in C (Fig 6). However, there was no difference among HFD-PBS, HFD-MSCs-R and HFD-MSCs-NR groups.
Glycemia response to glucose injection (GTT) was significantly higher in HFD-PBS than in C in the fourth week post-MSCs or PBS infusions. However, there was no difference among   HFD-PBS, HFD-MSCs-R and HFD-MSCs-NR groups. The same pattern was observed in the 8 th and 12 th weeks post-MSCs or PBS infusions (data not shown). In the 16 th week post-MSCs or PBS infusions, the glycemia response to glucose injection was significantly lower in HFD-MSCs-R than in HFD-PBS group (Fig 7). Glycemia response to insulin injection (ITT) was significantly higher in HFD-PBS than in C at the fourth week post-MSCs or PBS infusions. Nevertheless, there was no difference among HFD-PBS, HFD-MSCs-R and HFD-MSCs-NR groups. The same pattern was observed in the 8 th and 12 th weeks post-MSCs or PBS infusions (data not shown). In the 16 th week post-MSCs or PBS infusions, the glycemia response to insulin was significantly lower in both HFD-MSCs-R and HFD-MSCs-NR than in HFD-PBS group (Fig 8).

Pancreatic islet analysis
Pancreatic islets showed no morphological changes, maintaining their typical round or oval shapes. No inflammatory cells (insulitis) were observed in the pancreatic islets (Fig 9) According to the morphometric analysis of the islets, the total islet areas and the relative volumes of β-cells were not different among the groups (Fig 10A-10E). However, the relative volumes of α-cells were significantly smaller in HFD-PBS group than in C and significantly higher in HFD-MSCs-NR than in HFD-PBS and HFD-MSCs-R groups (Fig 11A-11E). Apoptosis in islet cells was significantly higher in HFD-PBS group than in C and in HFD-MSCs-NR group compared with HFD-PBS and HFD-MSCs-R groups. However, the number of apoptotic cells was reduced in HFD-MSCs-R group when compared with HFD-PBS group (Fig 12A-12E).
Regarding islet cellular proliferation, it was significantly smaller in HFD-PBS group than in C. No other difference among the groups was observed (Fig 13A-13E).

Discussion
The results show that multiple infusions of allogeneic BM-MSCs are able to promote prolonged decrease in glucose intolerance and apoptosis in pancreatic islets and increase in insulin sensitivity in hyperglycemic HFD fed mice. To the best of our knowledge this is the first report of The HFD fed rodent is one of the most widely used models for studying the metabolic derangements caused by obesity in humans, including glucose intolerance, insulin resistance and type 2 diabetes mellitus. It had been previously described that Swiss mice fed HFD for 8 weeks develop obesity, insulin resistance and hyperglycemia as observed in the present study [30]. This animal model resembles more closely the human type 2 diabetes background since it has no transgenic manipulations or drug-induced β cell toxicity, such as caused by streptozotocin.
Previous studies have reported beneficial effects of stem cells transplantation on the metabolic status in both types 1 and 2 diabetic patients and in rodent models of diabetes [5,13,14,15,16,17,20,22,31,32,33,34,35,36,37,38,39]. The beneficial effects have been attributed to immunological and regenerative properties of different origin stem cells, including umbilical cord cells [40,41,42,43]. However, most pre-clinical studies use streptozotocin-induced diabetes models to evaluate the metabolic effects of single or multiple intravenous MSCs infusions. To the best of our knowledge this is the first report of the effects of BM-MSCs infusion in pancreatic and metabolic parameters in exclusively HFD fed hyperglycemic mice.
Several studies suggest that MSCs may be autologously, allogeinic and even xenogeinic transplanted since the infusion of MSCs in immunocompetent animal models did not lead to rejection of the same and provided good therapeutic results [17,38,42,44,45]. In the present study, the progressive decrease of fasting glycemia in HFD-MSCs-R animals was observed since the 7 th week post-MSCs infusions and maintained until the 16 th week post-MSCs infusions. At this time, there was no difference in glucose levels between C animals and HFD-MSC-R animals. It was previously demonstrated that MSCs infusion in only STZ or STZ +HFD diabetes animal models is accompanied by decrease of hyperglycemia as observed in the present study, however in most of the studies, the experimental period was shorter than the one in the present study, even though in most of the studies, the experimental period was shorter than in the present study [13,14,15,16,17,20,22,38] and there was increase of plasma insulin levels [14,15,20,22,38].
In spite of the consistent decrease of hyperglycemia promoted by the multiple MSCs infusions, morphometric analysis of the islets showed that the total islet area and the relative volume of β-cell did not change post-MSCs infusions, as opposed to previous reports using other animal models [13,14,15,16,22]. This result associated with the unchanged insulin levels and the decreased glycemia response to both, glucose and insulin injections suggest the increase of insulin sensitivity of peripheral target tissues in the animals treated with MSCs, as recently reported by others [14,38]. Indeed, it was demonstrated that MSC infusion increased expression of GLUT4 and elevated phosphorilation of insulin receptor 1 and protein kinase B in insulin target tissues [14,39]. The mechanisms, however, are poorly understood. It is speculated that the migration of MSCs to different organs [46], in special to damaged tissues [47], could lead to modulation of insulin resistance and local inflammation, which are typical mechanisms involved in pathogenesis of obesity and type 2 diabetes. Most evidences suggest that the beneficial effects of BM-MSCs transplantation are not associated with trans-differentiation of MSCs into pancreatic β cells but rather related to MSC-derived paracrine factors such as cytokines and growth factorsresponsible for immunossupression, differentiation, angiogenesis and stimulation of endogenous cells [48,49,50]. The migration and homing ability of MSCs to injured tissues enable the pacrine effects of MSCs [51]. Even though we did not evaluate MSCs migration and homing analysis in the present study, previous studies have shown that MSCs injected intraperitoneally migrate to various organs such as lung, spleen, pancreas, kidneys, brain, heart, thymus, liver, among others [52,53].
The data regarding the pattern of α-cells and glucagon in pathologic states such as obesity, insulin resistance and type 2 diabetes in both humans and rodents are highly controversial [54,55,56,57,58,59,60]. In the present study, the volume of α-cells decreased in HFD fed mice. This finding could be associated with increased insulin levels since insulin is a potent α-cell inhibitor [59,60,61,62]. In addition, recent evidences point to the conversion of pancreatic α-cells to β-cells in conditions of stress [63,64]. A substantial proportion of α-cells becoming β-cells could explain the decrease in α-cells volume in the presence of unaltered β-cell observed in HFD. Nevertheless, this hypothesis does not explain the increase in α-cells volume observed in HFD-MSCs-NR group in comparison with HFD-MSCs-R group.
Beta-cell apoptosis has been considered an important mechanism in the pathogenesis of type 2 diabetes and increased islet apoptosis in HFD fed mice was demonstrated [65]. In the present study, not only did apoptosis increase in HFD fed mice but also proliferation decreased in these animals. Although islet proliferation did not change after MSCs infusions, islet apoptosis decreased in HFD-MSCs-R group. This result is confirmed by an in vitro study that demonstrates decreased caspase 3 expression and new expression of Ki67 in islet cell nuclei by electrofusion between dispersed islet cells and MSCs [66]. In addition, it was recently demonstrated in vitro that MSCs co-cultured with pancreatic islets release trophic factors that increase the survival of the islets and lead to expression of Pdx1 [67]. Altogether these data emphasize the cytoprotective property of MSCs [11]. Interestingly, islet apoptosis was positively correlated with fasting glycemia and glucose response to both glucose and insulin injections and islet proliferation was negatively correlated with fasting glycemia. These findings demonstrated the importance of the entire islet cell population and metabolic modulation.
In conclusion, the results demonstrate the ability of MSCs to promote prolonged decrease in hyperglycemia and apoptosis in pancreatic islets and increase in insulin sensitivity in HFD fed mice.