Introduction
Intravenous mesenchymal stem cell (MSC) transplantation has been demonstrated to be a useful therapeutic strategy for functional recovery in a variety of diseases such as therapy-resistant graft-versus-host disease (
28,
42), acute myocardiac infarction (
20), stroke (
3), and multiple systemic atrophy (
33) by various preclinical and clinical trials. In our previously reported preclinical study, intravenous transplantation of MSCs was effective in rescuing lethal fulminant hepatic failure (
27). We found that single dose of intravenous MSC transplantation reduced oxidative stress, promoted repopulation of host hepatocytes, and replenished host liver cells by differentiation of transplanted MSCs into functional hepatocytes, suggesting that both reactive oxygen species (ROS)-scavenging and paracrine effects of MSCs contributed to the survival of mice with lethal hepatic failure besides direct hepatocyte differentiation (
27).
Depletion of β-cells in the pancreatic islets results in type 1 diabetes mellitus (T1DM) due to insufficient insulin production. Insulin replenishment is essential to treat insulin-dependent diabetes (
11,
30,
38). Exogenous insulin supply fails to respond to plasma glucose level and may result in asymptomatic hypoglycemia, which pushes the therapeutic strategy towards cell transplantation (
16,
30). Islet cell transplantation has gained some clinical success (
5,
36,
37,
44); however, limited donor source is the major hurdle for extensive clinical application. Stem cells, which possess the potentials of self-renewal and multilineage differentiation, may be an ideal source of β-cells (
2,
39,
40). MSCs, which can be isolated from many tissues and can be extensively culture expanded (
6,
23,
50,
53), possess the multipotency in vivo (
9,
39), and immunomodulation abilities because of their low immunogenicity as well as the ability to inhibit T-cell proliferation and to prevent cytotoxic T-cell development (
1,
4,
26,
29), indicating the possibility of allogenic MSC transplantation.
It has been reported that in streptozocin (STZ)-induced diabetic animal models, a single dose of intravenous MSC injection diminished hyperglycemia and attenuated diabetic complications in some short-term animal studies (
14,
45,
52). It has been described that single dose of human multipotent marrow stromal cell injection preserved mouse islet cells in nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice (
34). In experimental autoimmune DM model, allogeneic MSCs resulted in temporary reversal of hyperglycemia through immunomodulation (
17). Besides, the immunosuppressive ability of MSCs not only prevented the rejection of allogeneic islet grafts (
12) but also prolonged the survival of semiallogeneic heart (
7).
However, DM is a chronic, progressive disease, and long-term follow-up is indispensable to evaluate the therapeutic effects of MSC transplantation. Besides, multiple MSC transplantations may be necessary. However, the preclinical studies of multiple MSC transplantations on DM are lacking. The aim of this study is to investigate the effects and the mechanisms of systemic MSC transplantation via intravenous injection on blood glucose regulation on a long-term basis. In this study, serial intravenous MSC transplantations were carried out in STZ-induced diabetic mice for 6 months at the interval of 2 weeks. Blood sugar, insulin, renal function, cholesterol, and triglyceride (TG) levels were continuously monitored and histopathology studies were performed at the end of the study.
Materials and Methods
Animals
Eight-week-old BALB/c male mice were purchased from the National Laboratory Animal Center (NLAC, Taipei, Taiwan). All the animals were housed following the animal care guidelines of NLAC, and the experimental protocols had been reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at the Taipei Veteran General Hospital. Body weight of the experimental animals was recorded weekly.
Isolation and Expansion of MSCs
MSCs were isolated from human bone marrow following the previously reported protocol (
31) that had been approved by the Institutional Review Board at the Taipei Veteran General Hospital. Informed consent had been obtained prior to the aspiration of bone marrow from patients. Briefly, after negative immune selection and limiting dilution from the mononuclear fraction of the bone marrow aspirates, colonies obtained from single cells were culture expanded in T75 flasks with MesenPro medium (Invitrogen, Carlsbad, CA, USA). When the cell density reached 60—70% of confluence, they were detached from one flask and reseeded to three new flasks after two gentle phosphate-buffered saline (PBS, Gibco BRL, Grand Island, NY, USA) washes. The surface immune phenotype and multipotency of MSCs were confirmed as previously described (
22,
31,
32). An approval from the Institutional Review Board had been obtained prior to the commencement of this study.
Induction of Diabetes and MSC Transplantation
T1DM was induced by intraperitoneal injection of STZ (Sigma-Aldrich, St. Louis, MO, USA) in 25 mM sodium citrate (pH 4.5), with the dosage of 200 mg/kg body weight, into 10-week-old BALB/c mice with the body weight of 23—25 g. Onset of DM was defined as elevated blood sugar values of higher than 200 mg/dl found at two consecutive tests at 3-day intervals. Same volume of phosphate-buffered saline (PBS) was injected into BALB/c mice intraperitoneally as controls.
Diabetic mice were further divided into two groups, STZ group and STZ with MSC transplantation group. In the MSC transplantation group, we performed single and multiple MSC transplantations. For single MSC transplantation, 4.2 × 107 cells/kg body weight in 0.2 ml PBS were injected into diabetic mice via tail vein 7 weeks after STZ injection. For multiple MSC transplantation, the same dosage of cells (4.2 × 107 MSCs/kg body weight in 0.2 ml PBS) were injected into diabetic mice via the tail vein from 1 week after the onset of diabetes, and a series of MSC transplantations were performed with the interval of 2 weeks. In STZ group, 0.2 ml of PBS was injected into diabetic mice via tail vein at the same time points of MSC transplantation. In this study, insulin was not given to any experimental animals.
Measurement of Blood Sugar, Insulin, Cytokines, Growth Factors, and Systemic Oxidative Stress Level
Mice were under starvation for 6 h before the measurement of fasting blood sugar level. Tail capillary blood sugar levels were measured twice a week using One Touch Ultra (LifeScan Inc., Milpitas, CA, USA). Whole blood was collected from facial vein weekly, and the plasma was collected after centrifugation at 7000 rpm for 15 min. Serum interferon-γ (INF-γ) and tissue necrosis factor-α (TNF-α) was detected by ELISA kits (eBioscience, Inc., San Diego, CA, USA). Serum level of mouse insulin was measured by Mercodia Mouse Insulin ELISA Enzyme immunoassay (Mercodia AB, Uppsala, Sweden) and human insulin was detected by Mercodia Ultrasensitive Insulin ELISA (Mercodia AB) as per the manufacturer's instructions. All the ELISA data were analyzed by an ELISA reader (Spectra MAX 250, Spectra Devices, Sunnyvale, CA, USA).
Tissue growth factors (TGF-α, TGF-β1, TGF-β2, and TGF-β3) and hepatocyte growth factor (HGF) produced by MSCs were measured by RayBio human growth factor antibody array (RayBiotech, Inc., Norcross, GA, USA). MSCs from three independent donors were maintained in MesenPro medium (Invitrogen) for 2 days, and then the conditioned medium was collected for growth factor array analysis according to the manufacturer's instructions. MesenPro medium served as the control.
Systemic oxidative stress was determined by the ratio of glutathione versus oxidative glutathione (GSH/ GSSG) in whole blood using a commercially available kit (Bioxytech GSH/GSSG-412TM kit, OxisResearch, Portland, OR, USA). Higher GSH/GSSG ratios represented lower systemic reactive oxidative stress levels in the circulation.
Histopathology and Immunohistochemical Staining
Tissues/organs were harvested, fixed in formalin, and prepared in paraffin-embedded blocks for sectioning at the thickness of 3—4 μm. Tissue sections were stained with hematoxylin and eosin (H&E) (Sigma-Aldrich). Besides, tissue sections were incubated with rabbit antibodies against human IgG (1:400, Abcam Inc., Cambridge, MA, USA), rabbit antibodies against mouse insulin (1:400, Cell Signaling, Danvers, MA, USA), mouse antibodies against human insulin (1:100, Abcam), or rabbit antibodies against human c-peptide (1:2000, Abcam) at the room temperature for 1 h, followed by adding goat antibodies against rabbit IgG (Dako Cytomation, Glostrup, Denmark), or goat anti-mouse IgG (Dako Cytomation), for another 40—60 min. Tissue sections were assessed by fluorescent microscopy (Leitz, Germany). Image acquisition was performed with SPOT RT Imaging system (Diagnostic Instruments, Sterling Heights, MI, USA).
Western Blot Analysis
MSCs were cultured in MesenPro medium for 2 days. After removal of the medium, MSCs were washed by PBS twice, and cell lysates were prepared. Three different dosages of protein extracted from MSCs lysate (10, 20, and 30 μg) as well as 20 μg of human insulin (Actrapid HM, Novo Nordisk Inc., Princeton, NJ, USA) were separated on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted onto polyvinylidene fluoride (PVDF) membrane (Amersham Biosciences, Uppsala, Sweden). After blocking, the membrane was then blotted with indicated primary antibodies: rabbit against human insulin (1:200, Cell Signaling) or mouse against human β-actin (1:10000, Sigma-Aldrich). After washes, the membrane was incubated with the horse radish peroxidase (HRP)-conjugated secondary antibody: goat against rabbit IgG (1: 8000, Epitomics Inc., Burlingame, CA, USA) or rabbit against mouse IgG (1:10000, Sigma-Aldrich) at room temperature for 1 h and the protein intensity was determined by ECL chemiluminescence reagent (PerkinElmer Life Sciences, Inc. Waltham, MA, USA), and their intensities were quantitatively measured by a densitometry (LabWorks, UVP Inc., Upland, CA, USA).
Statistical Analysis
Statistical analyses were performed using the Statistical Package for Social Science-12 software (SPSS Inc., Chicago, IL, USA) or GraphPad Prism 5 (GraphPad Software, Inc., La Jolla, CA, USA). Data of INF-γ and TNF-α were analyzed by ANOVA tests with Tukey's post hoc tests at 95% confidence intervals. Data of blood sugar was analyzed at the same time points by ANOVA tests with Bonferroni's post hoc tests at 95% confidence intervals. Different characters represented different levels of significance. Results of secreted growth factors was analyzed by two-tail, nonpaired t-tests, and values of p < 0.05 were considered statistically significant.
Discussion
For the first time, we demonstrated the importance of multiple transplantations of MSCs to regulate blood glucose homeostasis. Multiple MSC transplantations via intravenous injection restored glucose homeostasis in STZ-induced insulin-deficient diabetic mice, but a single MSC transplantation only exhibited a transient effect on blood sugar reduction. MSCs gradually decreased hyperglycemia after three doses of transplantation, which was associated with reduced systemic oxidative stress. From the 11th week, multiple MSC transplantations stabilized blood sugar by the engraftment of MSCs and subsequent differentiation into insulin-producing cells in the liver.
STZ-induced murine diabetic model has been extensively used (
21,
25,
43,
51). In our study, STZ-induced diabetes was indeed a consequence of progressive insulin exhaustion (
Fig. 1). However, we noted that upon STZ administration, the damage of insulin-producing cells was not associated with marked local inflammatory cell infiltration in pancreatic islets (
Fig. 1E—H); STZ administration did not result in alteration of systemic inflammatory cytokine levels such as TNF-α and INF-γ either (
Fig. 2A—D). These observations suggest that STZ-induced diabetic model is not the equivalent of T1DM. Also, whether MSC transplantation works on autoimmune-induced DM requires further investigation.
Unsurprisingly, human MSCs did not induce an inflammatory response in recipient mice (
Fig. 2), indicating that MSCs possessed low immunogenicity and the transplanted MSCs were tolerated in this xenotransplantation model. For the DM patients who are not suitable for harvesting autologous MSCs, allogenic MSCs may therefore be an alternative choice.
In this study, blood glucose homeostasis could not be maintained by a single MSC transplantation and blood sugar level rebounded 1 month after a single MSC transplantation (
Fig. 3A). Multiple transplantations gradually stabilized blood glucose homeostasis (
Fig. 3B), and the change of sugar level was in line with that of systemic oxidative stress in the first 15 weeks (
Fig. 3C). In the first 10 weeks after STZ injection, both mouse and human plasma insulin levels were low (
Fig. 4B), but sugar level was gradually reduced (
Fig. 4C) after multiple MSC transplantation, suggesting that reduction of systemic oxidative stress may contribute the blood sugar homeostasis in the early stage when the production of insulin of transplanted cells is still insufficient (
Fig. 3B, C). Moreover, gradual increase in human insulin levels from the 11th week after a series of human MSC transplantations indicated that insulin production from MSC-differentiated cells restored blood glucose homeostasis in STZ-induced diabetic mice at this later stage (
Fig. 4B, C). Although serum levels of TNF-α and INF-γ were not changed by multiple MSC transplantations (
Fig. 2C, D), we still could not rule out the possibility of anti-human antibodies produced by BALB/c mice in the absence of TNF-α or INF-γ increase in this xenotransplanted model. Consequently, high dosage of MSCs and multiple transplantations were used in this study. To elucidate the role of multiple transplantations in regulation of blood glucose homeostasis, we discontinued MSC transplantation after the sugar level stabilized. As shown in
Figure 3, when MSC transplantation was discontinued, a decrease in human insulin (
Fig. 4B) and a sharp rise in blood sugar (
Fig. 4C) within 1 month, illustrated that continuous systemic MSC transplantation was indispensible. After further transplantation, blood sugar could be restabilized and human insulin level was detectable and remaining relatively low, suggesting the possibility of increased insulin sensitivity in recipients after a series of MSC transplantations. Since insulin-producing cells differentiated from MSCs play a crucial role for regaining blood glucose homeostasis, the niche that allows long-term engraftment of MSCs is equally important. It is known that stem cells can find their own niches in vivo (
46). In many animal models of injury, MSCs can home and engraft into damaged tissues (
8,
10,
19,
41). In this study, we have detected the existence of human β2 microglobulin DNA to distinguish human cells from mouse cells by PCR. However, human cells were barely detectable in the pancreas (data not shown), showing that the microenvironment in the pancreas was not suitable for MSCs after STZ treatment. This was further confirmed by histopathology since no insulin-producing cells could be found in the diseased pancreas after STZ treatment (
Fig. 5B, C).
Besides the pancreas, we found that the liver has provided a favorable environment for MSCs to engraft (
Fig. 6), evidenced by multiple human IgG-positive cells in the murine liver (
Fig. 6A, C, E). The finding of human insulin- as well as human c-peptide-positive cells in the recipient liver indicated that the insulin-producing cells were of donor origin (
Figs. 6B, D, F, and
7A—C). Moreover, the finding that MSCs did not contain human insulin before transplantation has ruled out the possibility of insulin ingestion in culture followed by subsequent release after transplantation (
Fig. 7D). Clinically, liver parenchyma is the niche for pancreatic islets to engraft in islet cell transplantation (
44,
47,
48,
49). In our study, we demonstrated that 51% of the transplanted human cells in the recipient liver parenchyma expressed insulin (
Fig. 6B, G), illustrating that half of MSCs in diabetic liver could differentiate into insulin-producing cells. These human insulin-producing cells aggregated near the central vein (
Figs. 6D and
7B) rather than the portal vein (
Figs. 6F and
7C), suggesting that the environmental cues in the central vein areas were crucial for insulin-producing cell differentiation. However, the differentiation of human MSCs into fully functional β-islet cells has yet to be proven.
Fusion of donor and host cells is always an important issue for adult stem cell plasticity (
24). It has been reported that hematopoietic stem cells develop into hepatocytes through cell fusion (
18). In this study, the frequency of human- and mouse-cell fusion was very low (
Fig. 6H, I), showing that cell fusion was not the main mechanism when MSCs engrafted into liver. Besides, the percentage of cell fusion in human IgG-expressing cells (
Fig. 6H) and human insulin-producing cells (
Fig. 6I) were the same, indicating that the role of cell fusion for MSCs is negligible during differentiation into insulin-producing cells in liver.
In summary, we concluded that multiple transplantations of MSCs were essential to effectively restore and maintain blood glucose homeostasis in the preclinical model of insulin-insufficient diabetes due to β-cell depletion. Liver offered a favorable microenvironment to support long-term MSC engraftment and differentiation into insulin-producing cells. Multiple intravenous MSC transplantations may therefore serve as a new therapeutic strategy for DM patients.
Acknowledgments
The authors acknowledge financial support from the Wan Fang Hospital, Taipei Medical University (100swf03, to J.H.H.), as well as the support of research grants from the Taipei Veterans General Hospital (V99E1-014, V99C1-097, and V99S4-001, to O.K.L.), and the National Science Council (NSC-99-2120-M-010-001, to O.K.L. and J.H.H.; NSC98-2314-B-010-001, NSC98-2627-B-010-004, and NSC98-3111-B-010-003, to O.K.L.; and NSC98-2314-B-038-010, to J.H.H.). The authors also acknowledge the research grant support from Steminent Biotherapeutics Inc. (Grant number: W184 to JHH). This study was under the support of the Ministry of Education, Aim for the Top University Plan. The animal experiments were assisted in part by the Division of Experimental Surgery of the Department of Surgery, Taipei Veterans General Hospital. The authors declare no conflict of interest.