Journal Logo

Original Basic Science

Coaggregates of Regulatory T Cells and Islet Cells Allow Long-term Graft Survival in Liver Without Immunosuppression

Takemoto, Naohiro1; Konagaya, Shuhei1; Kuwabara, Rei1; Iwata, Hiroo1

Author Information
doi: 10.1097/TP.0000000000000579

Transplantation of islets of Langerhans (islets) is a promising to treat patients with insulin-dependent diabetes mellitus. More than 300 insulin-dependent diabetes mellitus patients are treated with allogeneic islet transplantation each year. Immunosuppressive drugs are administered to the patients to inhibit rejection of the islet graft. Although the introduction of a steroid-free immunosuppressive protocol has greatly improved the outcomes of islet transplantation,1 lifelong administration of immunosuppressive drugs is expected to cause various side effects, such as infection, malignant tumor formation, and impairment of islet functions.2 Various studies have sought alternatives to immunosuppressive drugs, including the bioartificial pancreas3 in which islets are enclosed within a semipermeable membrane,4,5 radiation of islets with ultraviolet light,6 islet culture under low-temperature conditions before transplantation,7 and systemic infusion of regulatory T (Treg) cells.8

The Treg cells are indispensable for the maintenance of dominant self-tolerance and immune homeostasis.9 Their dysfunction causes various autoimmune diseases, immunopathology, and allergy.10 The Treg cells (most of which are CD4+ T cells that express CD25) can suppress the activation, proliferation, and effector functions of a wide range of immune cells in vitro and in vivo, including CD4+ and CD8+ T cells, natural killer and natural killer T cells, B cells, and antigen-presenting cells.10 Some groups have reported that Treg cells can control alloimmunity and autoimmunity and can also be used successfully for adoptive transfer in many animal models.11–14 The ability of systemic administration of Treg cells to control graft rejection in islet transplantation has been investigated, and results have demonstrated that Treg cells might be able to prevent allograft rejection.8 However, this procedure requires the cotransfusion of 1 to 5 Tregs per effector T cell,15 which is not a feasible ratio in a normal clinical setting.

In this study, coaggregates of CD4+CD25+ Treg cells and islet cells were prepared, and the coaggregates were infused into liver through the portal vein. The Treg cells in coaggregates were expected to coexist with islet cells and become actively involved in the creation of an immune-privileged site. The protective effects of Treg cells in coaggregates of BALB/c islet cells and C57BL/6 Treg cells were examined using streptozotocin (STZ)-induced diabetic C57BL/6 mice as recipients.


Detailed information is provided in the Supplemental Materials and Methods (SDC,

Preparation of CD4+CD25+ Treg Cells and Islet Cells

The Kyoto University Animal Care Committee approved all animal experiments. CD4+CD25+ Treg cells were isolated from C57BL/6 mice (8-week-old male mice; Japan SLC, Inc, Shizuoka, Japan) using magnetic-activated cell sorting (MACS) (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany).16 The procedures and antibodies used for Treg isolation are presented in the supplementary information (SDC, Islets were isolated from BALB/c mice (7-week-old male mice; Japan SLC, Inc) using the collagenase digestion method.17

Islets were incubated in 1 mL of a trypsin solution (500 μg/mL in Hanks’ balanced salt solution) at 37 °C for 2 minutes, followed by the addition of 1 mL of medium. The islets were mechanically disintegrated into single cells by pipetting with a Pasteur pipette. The single cells were resuspended in RPMI 1640 medium with 10% fetal bovine serum (FBS), 100 units/mL penicillin, 100 μg/mL streptomycin, and 50 μM 2-mercaptethanol.

Coaggregation of Treg Cells and Islet Cells

Coaggregates of Treg cells and islet cells were prepared on agarose hydrogel with small round-bottomed wells. A mold (Microtissues Inc., Providence, RI) was used to create the hydrogel with 256 (16 × 16) wells, each 250 μm in diameter. A hot 2.5% agarose solution in phosphate-buffered saline (50-90 °C) was injected into the mold and chilled to form a gel on ice. The hydrogel was equilibrated in RPMI 1640 medium supplemented with 10% FBS, 100 units/mL penicillin, 100 μg/mL streptomycin, and 50 μM 2-mercaptethanol. Immediately after isolation, Treg cells were suspended in 100 μL of RPMI 1640 medium at a density of 3000 cells per well. Cells that were obtained by the disintegration of 256 islets were suspended in 100 μL of RPMI 1640 medium with 10% FBS. These suspensions of Treg cells and islet cells were mixed together. The 200-μL cell suspension was applied to the hydrogel to achieve a final concentration of approximately 3000 Treg cells and islet cells equivalent to 1 islet per well. The cells were cultured on the hydrogel in an incubator for 4 days at 37 °C in a humidified atmosphere of 5% CO2/95% air. We analyzed the Foxp3+ cell content in the coaggregates at 4 days of culture by separating the coaggregates into single cells and then using a fluorescence-activated cell analyzer (Guava EasyCyte Mini; Millipore, Billerica, MA). Thin sections of the aggregates were also examined by Foxp3 and insulin immunostaining.

In vitro Glucose Stimulation Test for Coaggregates

Fifty coaggregates (each coaggregate was composed of 3000 Treg cells and islet cells equivalent to one islet) were exposed sequentially to solutions of 0.1 g/dL, 0.3 g/dL, and 0.1 g/dL glucose in Krebs-Ringer buffer. Coaggregates were incubated for 1 hour at 37 °C in each solution. The supernatants were collected, and the insulin concentration of each solution was determined using enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instructions. The glucose stimulation test was also performed using 50 unmodified islets as a reference.

Intraportal Transplantation of Coaggregates

Diabetes was induced in C57BL/6 mice (8-week-old male mice) with a single intraperitoneal injection of STZ (120 mg/kg body weight in citrate buffer, pH 4.2). After 1 week, mice whose plasma glucose levels exceeded 450 mg/dL in 2 consecutive measurements were used as diabetic recipients. A suspension of 400 coaggregates was infused into the portal vein and the coaggregates plugged small vessels in the liver. As a control experiment, 400 unmodified islets were also infused through the portal vein. Plasma glucose levels were monitored every day during the initial 30 posttransplantation days and were subsequently monitored every 2 days. The graft survival periods were defined from the day of transplantation to the first day of 2 consecutive plasma glucose levels over 250 mg/dL. Blood was collected from recipients and centrifuged to obtain plasma. We determined plasma insulin levels by ELISA according to the manufacturer’s instructions. Recipients were also subjected to intraperitoneal glucose tolerance tests 90 days after transplantation to evaluate glucose tolerance.

Histochemical Analysis

Some recipient mice were sacrificed at predetermined timepoints after the transplantation of coaggregates or unmodified islets for histochemical examinations.

Statistical Analysis

Data are shown as mean ± SD from at least 3 independent samples. The data were compared using Student t test, the Tukey-Kramer honestly significant difference test, or a log-rank test. All statistical calculations were performed using the software JMP (SAS Institute Inc., Cary, NC). A P value less than 0.05 was considered statistically significant.


Coaggregate Formation and Morphology

The Treg cells were isolated from the spleens and lymph nodes of C57BL/6 mice using MACS (Miltenyi Biotec GmbH). A fluorescence-activated cell analyzer (Guava EasyCyte Mini; Millipore) was used to evaluate cell purity. Of the MACS-sorted cells, 76.3% ± 6.9% were CD4-positive and CD25-positive (Figure 1A). Islets were isolated from BALB/c mice (Figure 1B). The islets were separated into single cells by treatment with trypsin. Their viability after trypsin treatment was 94.6% ± 1.5% as estimated by the trypan blue exclusion method. Figure 1C1 shows a phase contrast microphotograph of a mixed suspension of CD4+CD25+ cells and islet cells, and Figure 1C2 shows a fluorescent microphotograph of CD25+ cells. Compared to the islet cells, the CD25+ cells are much smaller in size.

Figure 1:
Treg cells and islet cells used for coaggregate preparation. A, Magnetic-activated cell-sorted Treg cells were isolated from C57BL/6 mice and immediately analyzed using a fluorescence-activated cell analyzer. B, A phase contrast microscopic image of islets that were isolated from BALB/c mice and then cultured for 2 days. C, (1) Phase contrast and (2) fluorescent microscopic images of a mixed suspension of Treg cells stained with phycoerythrin (PE) and islet cells.

A mixed suspension of Treg cells and islet cells was applied to a hydrogel with small wells and cultured. Each well contained 3000 Treg cells and 2400 islet cells, the equivalent of 1 islet. Coaggregates formed after 4 days in culture were collected for use in transplantation (Figure 2A). Coaggregates were subjected to a glucose stimulation test to determine whether they could control insulin secretion in response to changes in glucose level. When the glucose concentration was increased from 0.1 g/dL to 0.3 g/dL, insulin release increased from basal levels in both coaggregates and unmodified islets. After exposure to 0.3 g/dL glucose, the glucose concentration was returned to 0.1 g/dL, and the insulin release returned to basal levels in both cultures (Figure 2B). Although the islet cells in the coaggregates maintained their ability to regulate insulin release in response to glucose level changes, at 0.3 g/dL glucose, they released only about 73% the level of insulin released by the unmodified islets. This finding suggests that about 73% of the initially applied islet cells were incorporated into coaggregates. It is possible that some islet cells deteriorated during coaggregate preparation, or that some cells remained as single cells during coaggregate collection and thus were not incorporated into the coaggregates. After the separation of coaggregates into single cells, a fluorescence-activated cell analyzer was used to determine the proportion of Foxp3+ cells, which was found to be 12.42% after 4 days of culture (Figure 2C). Figure 2D shows the Foxp3 and insulin immunostaining of thin sections of aggregates. Immunofluorescence examination revealed an estimated 7.2% ± 3.5% content of Foxp3+ cells in coaggregates in a 4-day culture.

Figure 2:
Coaggregates of Treg cells and islet cells. A, A phase contrast microscopic image of coaggregates of Treg cells and islet cells after 4 days of culture. B, In vitro glucose stimulation test, with released insulin values expressed as mean ± SD for n = 3 cultures. White bars, islets; Black bars, coaggregates. Comparisons using Student’s t-tests revealed no significant differences. C, A fluorescence-activated cell analysis of single cells prepared from coaggregates formed after 4 days of culture. D, A thin section of a coaggregate immunostained for (1) Foxp3 and (2) insulin, along with the (3) superimposed image. Arrows, Foxp3; Scale bar = 50 μm.

Transplantation of Coaggregates Into the Livers of STZ-Induced Diabetic Mice

We transplanted 400 unmodified islets or 400 coaggregates that contained 1.2 × 106 Treg cells and islet cells from 400 islets into each STZ-induced diabetic C57BL/6 mouse (n = 9 per group) to ensure that the grafts contained equal numbers of islet cells. The recipient mice were not treated with any immunosuppressive therapy. Nonfasting blood glucose levels of the recipient mice before and after transplantation are summarized in Figure 3. When 400 unmodified BALB/c islets were transplanted, the blood glucose levels of 8 of the 9 control recipients were transiently normalized, but returned to the preoperative high level approximately 10 days after transplantation (Figure 3A). In contrast, when 400 coaggregates were transplanted, 6 of the 9 experimental recipients demonstrated normoglycemia just 1 day after transplantation, and their blood glucose levels were maintained at levels below the preoperative levels for more than 100 days (Figure 3B). The graft survival periods were defined from the day of transplantation to the day on which the first of 2 consecutive plasma glucose levels greater than 250 mg/dL were recorded. Graft survival was plotted versus the number of posttransplantation days (Figure 3C). These data suggest that Treg cells in the aggregates are able to effectively protect the islet cells from rejection.

Figure 3:
Nonfasting blood glucose levels of STZ-induced diabetic C57BL/6 mice after intraportal transplantation. A, After transplantation of 400 unmodified BALB/c mouse islets (n = 9 mice). B, After transplantation of 400 coaggregates, each prepared from 1.2 × 106 Treg cells isolated from C57BL/6 mice and islet cells from 400 BALB/c mouse islets (n = 9 mice). C, Kaplan-Meier survival curves for mice transplanted with unmodified islets (dashed line) and coaggregates (solid line). Comparisons were made using the log-rank test (P < 0.0015). Graft rejection was defined as the day on which the first of 2 consecutive plasma glucose measurements of greater than 250 mg/dL was recorded.

Plasma Insulin Levels and Glucose Tolerance in Recipients

Blood samples were obtained from recipient mice at predetermined days after the intraportal transplantation of unmodified islets or coaggregates, and their plasma insulin levels were determined using an ELISA. The plasma insulin levels of wild-type C57BL/6 mice and STZ-induced diabetic mice were also determined as references (Figure 4A). The recipients of unmodified islets at 14 days after transplantation (the population whose blood glucose levels had returned to the preoperative high levels) showed plasma insulin levels as low as the low plasma insulin levels observed in STZ-induced diabetic mice, indicating loss of the graft. At 120 days after transplantation, plasma insulin levels of the coaggregate recipients were approximately one half of the levels observed in wild-type mice, but were significantly higher than the levels observed in STZ-induced diabetic mice and in the recipients of unmodified islets at 14 days after transplantation. These results indicate that islet cells in the coaggregates were still functioning and secreting insulin.

Figure 4:
Effects of transplantation of coaggregates. A, Plasma insulin levels of wild-type C57BL/6 mice (Control), STZ-diabetic mice (STZ), and recipients of 400 unmodified islets (Islets) or 400 coaggregates (Coaggregates). Comparisons were made using the Tukey-Kramer honestly significant difference test. *P < 0.05. B, Intraperitoneal glucose tolerance tests performed 90 days after intraportal transplantation (n = 4 mice). Solid line, recipients of 400 coaggregates; dashed line, wild type C57BL/6 mice. C, Areas under the curves from the intraperitoneal glucose tolerance tests. White bar, wild type C57BL/6 mice (Control); Black bar, recipients of 400 coaggregates (Coaggregates). Comparisons made using Student t tests showed no significant differences. D, Immunohistochemical examination of islet grafts in C57BL/6 mouse livers (1) at 7 days after intraportal transplantation of unmodified BALB/c mouse islets and (2) at 120 days after intraportal transplantation of coaggregates of Treg cells isolated from C57BL/6 mice and islet cells from BALB/c mice. Thin tissue sections were stained with H&E or Alexa488-labeled anti-insulin antibodies (green) and Hoechst 33258 (blue). Scale bar = 100 μm. Areas enclosed by a dashed line indicate sites of lymphocyte accumulation. Areas enclosed by a solid line indicate the existence of insulin-positive cells. H&E, hematoxylin-eosin.

Intraperitoneal glucose tolerance tests were performed 90 days after transplantation to assess the blood glucose tolerance of coaggregate recipients. The blood glucose levels of wild-type C57BL/6 mice and coaggregate recipients were highest 30 minutes after glucose injection (Figure 4B). Afterward, blood glucose levels declined gradually over the following 120 minutes. No significant differences were observed in the changes in blood glucose levels between wild-type C57BL/6 mice and coaggregate recipients (Figure 4C). These data indicate that the islet cells of the coaggregates allowed normal regulation of blood glucose levels even 90 days after transplantation.

Histological Analysis of Islet Grafts in Liver

Livers were retrieved from recipient mice at predetermined time points after transplantation and subjected to histological examination. Figure 4D (1 and 2) depicts microscopic images of thin sections of the liver with unmodified islets and coaggregates, respectively. These tissue sections were stained with Alexa488-labeled anti-insulin antibodies (green fluorescence) and Hoechst 33258 (blue). Figure 4D presents these tissue sections stained with hematoxylin-eosin. When unmodified islets were transplanted, insulin-positive cells could be seen; however, their numbers decreased, and they were surrounded by dark purple-stained cells, indicating many lymphocytes, at 7 days after transplantation (Figure 4D1). Insulin-positive cells were rarely detected at 14 days after transplantation. These findings indicate that the allogeneic unmodified islets were rejected by the host immune system. In contrast, when coaggregates were transplanted, a number of insulin-positive cells were observed in the livers as late as 120 days after transplantation (Figure 4D2). Furthermore, the dark purple-stained cells that were observed after the transplantation of unmodified islets were not observed surrounding the coaggregates (Figure 4D2). These results suggest that Treg cells in the aggregates exerted to some extent protected the graft from attack by the host immune system, and thus allowed the graft to survive for a long time in the liver even without administration of immunosuppressive drugs.


Our previous study18 examined the intrahepatic transplantation of coaggregates of islet cells and Sertoli cells, which play crucial roles in creating the immunoprivileged environment of the testis.19 In these coaggregates, Sertoli cells occupy the core, whereas islet cells occupy the periphery. In contrast, the aggregates in present investigation showed a random distribution of Treg cells and islet cells (Figure 2). Sertoli cells are adherent cells that are intimately connected to each other through tight junctions,20 while Treg cells are floating cells that do not spontaneously agglutinate with each other. Such cell-cell interaction determines the distribution of each cell type in aggregates. Coaggregates were prepared by applying 3000 Treg cells and 2400 islet cells to a well; however, after 4 days of culture, only 12% of cells were Foxp3+. This may have been due to contamination with non–T cells (e.g., fibroblasts and mesenchymal cells) as indicated by the purity of the CD25+CD4+ cells (76.3% ± 6.9%), as well as the possible instability of Foxp3+ Treg cells in in vitro culture.15

The coaggregation of these 2 cell types allowed long-term allogeneic islet graft survival in the liver without immunosuppression.18 However, the source of Sertoli cells (testis) is limited, and thus they are rarely obtained in the clinical setting. The strong immune-suppressive functions of Treg cells are known, and they can be isolated from peripheral blood in sufficient numbers.9,10 Some groups have reported the use of Treg cell transfusion to successfully effect adoptive transfer in many animal models.8,11–14 In islet transplantation, the transfusion of Treg cells promotes tolerance of allogeneic islet grafts in animal models.8 Battaglia et al8 reported long-term islet graft survival in STZ-induced diabetic BALB/c mice that received islets from C57BL/6 mice under the kidney capsule; a Treg cell–enriched T-cell population (5 × 106 Treg cells per graft) isolated from BALB/c mice had been injected the day before islet transplantation. However, the systemic infusion of a large number of Treg cells might cause a variety of side effects, including opportunistic infections and regeneration of neoplasm. To overcome these problems, we used coaggregates of Treg cells and islet cells to decrease the number of Treg cells to 1.2 × 106 cells per graft and to localize Treg cells near the islet cells even when the islet cells were transfused through the portal vein into the liver as in clinical islet transplantation. Although we used only one fourth as many Treg cells per graft as Battaglia et al, in our experience, they effectively protected allogeneic islets from graft rejection.

There have been several attempts to localize Treg cells near islets in the context of intrahepatic islet transplantation.21 Marek et al21 previously demonstrated the ability to bind Treg cells to islet surfaces using biotin-poly(ethylene glycol)-N-hydroxysuccinimide ester and streptavidin as binding molecules. They reported that islets maintain their viability and insulin secretion function after Treg immobilization, and that Treg cells exert some protective activity on the islets against immune attack in vitro. However, their in vivo efficacy has not yet been reported. To prepare coaggregates, we used 3000 Treg cells and 2400 islet cells, which is equivalent to 1 islet; therefore, our number of Treg cells per islet is much larger than that in Marek’s study. This difference in the number of Treg cells may be the reason why we observed a clear immune-protective effect of Treg cells on allogeneic islet cell grafts (Figures 3 and 4).

Our approach successfully colocalized Treg cells and islet cells in the liver and demonstrated long-term graft survival. We believe that our method will be useful to realize immunosuppressive drug-free islet transplantation.


The authors would like to thank Professor Shimon Sakaguchi, Dr. Naoto Sasaki, and Dr. Masahide Hamaguchi for advising about Treg cell isolation and providing the antibodies for Treg cell isolation, and Mr. Hayato Nishimura for his technical assistance.


1. Froud T, Ricordi C, Baidal DA, et al. Islet transplantation in type 1 diabetes mellitus using cultured islets and steroid-free immunosuppression: Miami experience. Am J Transplant. 2005; 5: 2037.
2. Luzi L. Metabolic strategies to predict and improve intrahepatic islet graft function. J Mol Med. 1999; 77: 49.
3. Teramura Y, Iwata H. Bioartificial pancreas microencapsulation and conformal coating of islet of Langerhans. Adv Drug Deliv Rev. 2010; 62: 827.
4. Iwata H, Takagi T, Amemiya H, et al. Agarose for a bioartificial pancreas. J Biomed Mater Res. 1992; 26: 967.
5. Lim F, Sun AM. Microencapsulated islets as bioartificial endocrine pancreas. Science. 1980; 210: 908.
6. Benhamou PY, Stein E, Hober C, et al. Ultraviolet light irradiation reduces human islet immunogenicity without altering islet function. Horm Metab Res. 1995; 27: 113.
7. Ricordi C, Lacy PE, Sterbenz K, Davie JM. Low-temperature culture of human islets or in vivo treatment with L3T4 antibody produces a marked prolongation of islet human-to-mouse xenograft survival. Proc Natl Acad Sci U S A. 1987; 84: 8080.
8. Battaglia M, Stabilini A, Roncarolo MG. Rapamycin selectively expands CD4+CD25+FoxP3+ regulatory T cells. Blood. 2005; 105: 4743.
9. Sakaguchi S, Miyara M, Costantino CM, Hafler DA. FOXP3+ regulatory T cells in the human immune system. Nat Rev Immunol. 2010; 10: 490.
10. Sakaguchi S, Yamaguchi T, Nomura T, Ono M. Regulatory T cells and immune tolerance. Cell. 2008; 133: 775.
11. Kim JI, O’connor MR, Duff PE, et al. Generation of adaptive regulatory T cells by alloantigen is required for some but not all transplant tolerance protocols. Transplantation. 2011; 91: 707.
12. Feng G, Wood KJ, Bushell A. Regulatory T cell enrichment by IFN-γ conditioning. Methods Mol Biol. 2011; 677: 281.
13. Karim M, Feng G, Wood KJ, Bushell AR. CD25+CD4+ regulatory T cells generated by exposure to a model protein antigen prevent allograft rejection: antigen-specific reactivation in vivo is critical for bystander regulation. Blood. 2005; 105: 4871.
14. Feng G, Wood KJ, Bushell A. Interferon-gamma conditioning ex vivo generates CD25+CD62L+Foxp3+ regulatory T cells that prevent allograft rejection: potential avenues for cellular therapy. Transplantation. 2008; 86: 578.
15. Tang Q, Bluestone JA, Kang SM. CD4+Foxp3+ regulatory T cell therapy in transplantation. J Mol Cell Biol. 2012; 4: 11.
16. Asano M, Toda M, Sakaguchi N, Sakaguchi S. Autoimmune disease as a consequence of developmental abnormality of a T cell subpopulation. J Exp Med. 1996; 184: 387.
17. Lacy PE, Kostianovky M. Method for the isolation of intact islets of Langerhans from the rat pancreas. Diabetes. 1967; 16: 35.
18. Takemoto N, Liu X, Takii K, Teramura Y, Iwata H. Transplantation of co-aggregates of Sertoli cells and islet cells into liver without immunosuppression. Transplantation. 2014; 97: 287.
19. Streilein JW. Unraveling immune privilege. Science. 1995; 270: 1158.
20. Bart J, Groen HJ, van der Graaf WT, et al. An oncological view on the blood-testis barrier. Lancet Oncol. 2002; 3: 357.
21. Marek N, Krzystyniak A, Ergenc I, et al. Coating human pancreatic islets with CD4+CD25highCD127 regulatory T cells as a novel approach for the local immunoprotection. Ann Surg. 2011; 254: 512.

Supplemental Digital Content

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.