Natural Killer T-Cells Participate in Rejection of Islet Allografts in the Liver of Mice

Male C57BL/6 (H-2^b, 12–15 weeks old), BALB/c (H-2^d, 9–12 weeks old), and C3H/HeN (H-2^k, 9–12 weeks old) mice were purchased from Charles River Japan (Kanagawa, Japan) and used as recipients, donors, and third-party donors, respectively. CD1d-deficient (CD1d^−/−) mice and Vα14NKT cell–deficient (Jα281^−/−) mice (15) with a C57BL/6 background were also used as recipients. Vα14NKT cells are a major component of NKT cells reactive to CD1d (23). CD1d^−/− mice, originally developed by Dr. Luc Van Kaer (Vanderbilt University), were provided by Dr. Moriya Tsuji (Rockefeller University). All experiments were performed in accordance with the institutional animal care and use committee.

Diabetes was induced in recipient mice by an intravenous injection of streptozotocin (STZ; 180 mg/kg; Sigma, St. Louis, MO). Plasma glucose levels were measured with a glucose analyzer (Beckman Instruments, Tokyo). After the STZ injection, nonfasting plasma glucose levels exceeded 400 mg/dl by day 2, and mice remained hyperglycemic at the time of islet transplantation.

Islets were isolated using static collagenase (type S-1; Nitta Gelatin, Osaka, Japan) digestion (24) and Ficoll-Conray gradient separation (25). Isolated islets were incubated overnight at 24°C in a 5% CO₂ incubator and then handpicked using a Pasteur pipette with the aid of a dissecting microscope. We transplanted 600 islets beneath the left kidney capsule (26) or into the liver via the portal vein (27) of diabetic mice 3 days after STZ injection. Nonfasting plasma glucose levels and body weight were monitored three times a week, and rejection was defined when two consecutive plasma glucose levels exceeded 200 mg/dl after islet transplantation.

Rapamycin was generously provided by Wyeth-Ayerst Research (Princeton, NJ). A solution of 1.0 mg/ml rapamycin in 0.2% carboxymethylcellulose (Sigma) was dispersed after ultrasonic disruption by a Bioruptor UCD-200T (Cosmo Bio, Tokyo) for 5 min and then diluted to the specified concentrations for the experimental groups. Rapamycin was administered intraperitoneally for 7 consecutive days, starting on the day of transplantation.

For light microscopic examination, the livers were removed and fixed with Bouin’s solution, processed, and embedded in paraffin. The sections were stained with hematoxylin and eosin or aldehyde and fuchsin.

Hepatic mononuclear cells were prepared as described previously (28). In brief, an excised liver was pressed through a stainless steel mesh, and the resulting dissociated liver tissues were suspended in Dulbecco’s modified Eagle medium (DMEM/F-12; Life Technologies, Tokyo) and washed twice. The mixture was resuspended in an isotonic 33% Percoll solution containing heparin (67 units/ml) and centrifuged at 2,000g at 4°C for 15 min. The resulting pellet was suspended in 0.83% ammonium chloride solution to lyse erythrocytes. After counting, these hepatic mononuclear cells were washed twice in PBS and used for further analysis.

The antibodies used for flow cytometry analysis were as follows: Fc block (anti-mouse FcRγII/III monoclonal antibody [mAb], 2.4G2), fluorescein isothiocyanate–conjugated anti-CD3ε (clone 145-2C11, American hamster IgG1κ), allophycocyanin-conjugated anti–IFN-γ mAb (clone XMG1.2, rat IgG1), anti–interleukin (IL)-4 mAb (clone 11B11, rat IgG1), anti–IL-2 receptor α-chain (CD25; clone PC61, rat IgG1), and their isotype control (clone R3-34, rat IgG1) were purchased from PharMingen (San Diego, CA). Phycoerythrin-conjugated α-galactosylceramide (α-GalCer)-loaded CD1d tetramers were prepared as described previously (29).

The stained cells were analyzed on a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA), and the data were processed by CELLQuest software (Becton Dickinson). We collected 10,000 viable hepatic mononuclear cells for dot plot to analyze each cell population and 1,000 for histogram plot for each fraction.

Hepatic mononuclear cells were prepared from wild-type or CD1d^−/− mice in the same way as flow cytometry analysis under aseptic conditions. We transferred 5 × 10⁶ hepatic mononuclear cells in 200 μl Hanks’ balanced salt solution via the portal vein of CD1d^−/− mice at the time of islet transplantation.

Statistical significance of graft survival was determined by log rank test or Fisher’s exact probability test, and that of flow cytometry data was examined by unpaired Student^’s t test. Results were considered statistically different at P < 0.05.

First, we determined whether NKT cells are involved in acute rejection of islet allografts with the use of CD1d^−/− mice and Vα14 NKT cell^−/− mice as recipients. Furthermore, the effect of transplant sites, namely the renal subcapsular space versus the liver, on survival of islet allografts was examined because the liver contains relatively abundant NKT cells (8). When BALB/c islets were grafted beneath the kidney capsule of diabetic C57BL/6 mice, all islet allografts were rejected within 3 weeks after transplantation irrespective of the presence or absence of NKT cells (Table 1). Rejection was confirmed morphologically, in which only accumulations of mononuclear cells were seen at the site of islet transplantation, namely beneath the kidney capsule (histology not shown). In contrast, when BALB/c islets were grafted into the liver of diabetic mice, there was a significant prolongation of islet allograft survival in CD1d^−/− and Vα14 NKT cell^−/− mice compared with wild-type mice (Table 1). The difference in survival between wild-type and CD1d^−/− mice and between wild-type and Vα14 NKT cell^−/− mice was statistically significant (P < 0.05, log rank test) (Table 1). Morphologically, islet grafts infiltrated with mononuclear cells in the liver of recipient mice were seen at the time of rejection (Fig. 1, right panels).

To evaluate a role of NKT cells in the liver of mice rejecting islet allografts, hepatic mononuclear cells of mice were isolated for examination by flow cytometry. The yield of mononuclear cells from the liver of naïve untreated wild-type mice was 25.8 ± 1.2 ×10⁵ (n = 3), and that from the liver of wild-type mice that had received isografts or allografts was 40.3 ± 1.3 (n = 3) or 56.7 ± 2.7 ×10⁵ (n = 3), respectively, at 8 days after transplantation, when islet allografts are rejected. The difference in the yield of hepatic mononuclear cells between wild-type mice receiving isografts and those receiving allografts was statistically significant (P < 0.05, Student’s t test). The yield of hepatic mononuclear cells from CD1d^−/− mice was 27.3 ± 1.0 ×10⁵ (n = 3) and 54.2 ± 4.6 ×10⁵ (n = 3) at 8 and 16 days, respectively, after islet transplantation, when islet allografts maintained or ceased to function without or with rejection, respectively.

We then determined whether NKT cells in the liver of wild-type mice are activated in association with rejection. To detect NKT cells, α-GalCer–loaded CD1d tetramers (29) were used for fluorescence-activated cell sorting analysis. The percentage of tetramer⁺ CD3⁺ NKT cells in the liver of naïve wild-type mice was 10.2 ± 1.6% (n = 3), and that in the liver of wild-type mice receiving isografts and allografts was 1.5 ± 0.3% (n = 3) and 5.9 ± 0.6% (n = 3), respectively, at 8 days after transplantation. Tetramer⁺ CD3⁺ NKT cells and tetramer⁻ CD3⁺ T-cells were further gated and examined with respect to the production of IFN-γ and IL-4 as well as CD25 expression. NKT cells are known to secrete IFN-γ and IL-4 on activation. The percentage of IFN-γ–producing tetramer⁺ CD3⁺ NKT cells in the liver of naïve wild-type mice was 2.5 ± 0.2% (n = 3), and that of wild-type mice receiving isografts or allografts was 6.6 ± 2.5% (n = 3) or 20.6 ± 2.9% (n = 3), respectively, at 8 days after islet transplantation. In contrast, there was no upregulation of IL-4 production in NKT cells in association with rejection. The CD25 expression of NKT cells in naïve wild-type mice was 15.2 ± 1.1% (n = 3), and that of wild-type mice receiving isografts or allografts was 3.4 ± 1.4% (n = 3) or 42.8 ± 3.7% (n = 3), respectively, at 8 days after transplantation. Representative data of three experiments is shown in Fig. 2A.

Regarding CD3⁺ T-cells, an analysis similar to that of NKT cells was performed. It was found that IFN-γ production and CD25 expression of T-cells in the liver of wild-type mice receiving islet allografts became upregulated from 3.6 ± 0.2% (n = 3) to 21.9 ± 2.8% (n = 3) and from 7.4 ± 0.5% (n = 3) to 23.1 ± 1.7% (n = 3), respectively, at 8 days after transplantation, and that those of mice receiving isografts were from 3.6 ± 0.2% (n = 3) to 8.4 ± 3.7% (n = 3) and from 7.4 ± 0.5% (n = 3) to 8.5 ± 1.6% (n = 3), respectively. In marked contrast, the upregulation of IFN-γ production and CD25 expression of T-cells in the liver of CD1d^−/− mice receiving islet allografts was not seen at 8 days after transplantation, when the islet grafts were rejected in wild-type mice. Representative data of three experiments is shown in Fig. 2A and B.

We next determined the effect of NKT cells on rejection of islet allografts in the liver of mice treated with rapamycin. The survival rate of islet allografts in wild-type recipient mice treated with 0.2, 1.0, or 3.0 mg/kg rapamycin once a day for 6 days after transplantation was found to be 0% (0 of 9), 12.5% (1 of 8), and 71.7% (5 of 7), respectively, at 90 days after transplantation (Table 2). In marked contrast to wild-type mice, 91.0% (10 of 11) of CD1d^−/− mice and 60% (3 of 5) of Vα14NKT cell^−/− mice receiving islet allografts and treated with 1.0 mg/kg rapamycin were normoglycemic at 90 days after transplantation (Table 2 and Fig. 3, left panels). There was a statistical significance in the survival rate between islet allografts in CD1d^−/− mice treated with 1.0 mg/kg rapamycin and those in wild-type mice treated similarly (P < 0.01, Fisher’s exact probability test). Morphologically intact islets with well-granulated β-cells were seen in the liver of CD1d^−/− mice at 120 days after islet transplantation (Fig. 3, right panels).

To further confirm the role of NKT cells in rejection of islet allografts in the liver of mice treated with rapamycin, cell transfer experiments were performed. As described above, rejection of islet allografts is prevented in CD1d^−/− mice treated with 1 mg/kg rapamycin, and therefore CD1d^−/− mice were used for the following study. When 5 × 10⁶ hepatic mononuclear cells from wild-type mice were transferred via the portal vein into the liver of CD1d^−/− mice receiving islets and treated with rapamycin at the time of islet transplantation, all recipient mice (n = 3) became hyperglycemic again after transplantation (Fig. 4A). In marked contrast, when the same number of hepatic mononuclear cells from CD1d^−/− mice was transferred, recipient CD1d^−/− mice (n = 3) remained normoglycemic (Fig. 4B). Morphologically intact or infiltrated islets were seen in the livers of CD1d^−/− mice receiving the transfer of cells from CD1^−/− or wild-type mice, respectively (histology not shown).

Finally, we determined whether donor-specific unresponsiveness is induced in CD1d^−/− mice treated with rapamycin and accepting islet allografts. For this purpose, third-party or donor-specific islets were grafted into the liver of CD1d^−/− mice that had been made diabetic again with intravenous injection of STZ at 120 days after the initial islet transplantation. No immunosuppressive agent was administered at the time of the second transplantation. CD1d^−/− mice receiving third-party C3H islets (n = 3) became hyperglycemic at 22, 26, and 36 days (Fig. 5A) because of rejection, whereas CD1d^−/− recipient mice (n = 3) with donor-specific BALB/c islets remained normoglycemic after the second transplantation (Fig. 5B).

These findings clearly demonstrate that NKT cells play a significant role in rejection of islet allografts in the liver of mice without immunosuppressive treatment and with immunosuppression by rapamycin, which is currently used in clinical islet transplantation. In the current study, two kinds of NKT cell–deficient mice, namely CD1d^−/− and Vα14 NKT cell^−/− mice, were used as recipients to elucidate a role of NKT cells in islet allograft rejection. Vα14 NKT cell^−/− mice lack only Vα14NKT cells, which are a major population of CD1d-reactive NKT cells in mice (23). The important finding is that survival of islet allografts is prolonged in CD1d^−/− and Vα14NKT cell^−/− mice without immunosuppression compared with wild-type mice when the liver is the site of transplantation versus the renal subcapsular space. The finding suggests that NKT cells play a significant role in acute rejection of islet allografts in the liver of mice. Flow cytometry analysis revealed that not only NKT cells but also T-cells in the liver of wild-type mice became activated at the time of rejection. In marked contrast, CD3⁺ T-cells in the liver of CD1d^−/− mice at 8 days after transplantation, when islet allografts were rejected in wild-type mice but not in CD1d^−/− mice, was not activated. Consequently, CD1d^−/− mice rejected intrahepatic islet allografts with the upregulation of IFN-γ production and CD25 expression in CD3⁺ T-cells in the liver. These findings indicate that T-cell activation is delayed in the absence of NKT cells after islet transplantation, which may reflect the prolonged survival of islets in the liver of CD1d^−/− mice, although the precise role of NKT cells in T-cell activation in response to islet allografts remains unclear.

Regarding a role of NKT cells in islet allograft rejection under the immunosuppressive agent rapamycin, rejection was prevented in the absence of NKT cells with a subtherapeutic dose of rapamycin. The frequency of rejection appears to be high in Vα14 NKT cell^−/− mice compared with CD1d^−/− mice, although the difference has no statistical significance. The potential difference may rely on the disparity between CD1d^−/− and Vα14 NKT cell^−/− mice (23), in that CD1d^−/− mice lack not only Vα14NKT cells, a major component of NKT cells, but also the other CD1d-reactive cells, which are less well characterized and whose role in transplant immunity remains unknown. Furthermore, the exact mechanisms involved in the response of NKT cells to donor islets leading to rejection, in relation to cell components other than NKT cells, remain undetermined, which is a matter of interest for future studies.

The current finding that the dosage of rapamycin can be reduced to prevent rejection of islet allografts in the absence of NKT cells raises a possibility that the downregulation of NKT cells may help prevent islet allograft rejection in wild-type mice with reduced dosage of rapamycin. Currently, there is limited number of reports with respect to suppression of NKT cell function. NKT cells secrete large amounts of IFN-γ on activation (11,12). IFN-γ production from NKT cells is modulated by the use of their synthetic ligand α-GalCer (30), in that single administration of α-GalCer upregulates IFN-γ production, whereas repeated administration downregulates IFN-γ production (31). Recently, natural ligands of NKT cells have been reported (32–34). Thus, it is interesting to learn whether procedures targeting NKT cells with the use of their synthetic or natural ligands have any beneficial effects on prevention of islet allograft rejection.

In the current study, it was found that donor-specific unresponsiveness is induced and maintained in CD1d^−/− mice treated with rapamycin and accepting islet allografts. This finding indicates that NKT cells are not required for induction of donor-specific unresponsiveness in this model. Previously, we have shown that CD4⁺ Vα14 NKT cells are essential for acceptance of rat islet xenografts in mice treated with anti-CD4 mAb (35). In a cardiac allograft model, tolerance induction after the blockade of CD28/B7 interaction in conjunction with that of LFA-1/ICAM-1 (lymphocyte function–associated antigen 1/intercellular adhesion molecule 1) interaction is dependent on NKT cells (36). NKT cells are reported to be required for induction of allospecific T-regulatory cells and are essential for survival of cornea allografts (37). In contrast to these findings, human CD1d (hCD1d) has been reported to function as a transplantation antigen, and skin grafts are rejected by expression of human CD1d in the major histocompatibility complex–matched mouse model (38). Survival of porcine neural xenografts is shown to be enhanced in the brain of CD1d^−/− mice compared with wild-type controls (39). Recently, Oh et al. (40) reported that rejection of H-Y mismatched skin grafts is accelerated in CD1d^−/− mice as recipients compared with wild-type mice. Thus, these findings indicate that NKT cells may have a diverse effect as regulatory and/or effector cells on the outcome of transplants, depending on the type of graft (allograft versus xenograft, organ versus tissue [cell]) and/or the type of immunosuppression.

In summary, the current study affords evidence that NKT cells play a significant role in islet allograft rejection in the liver of mice without immunosuppressive treatment and with immunosuppression by rapamycin, providing a new insight into cellular immune response in association with rejection. Currently, a major obstacle facing clinical islet transplantation is that the rate of insulin independence in patients with type 1 diabetes after islet transplantation declines with time after transplantation (3). Current immunosuppressive regimes include FK506 and rapamycin, which are known to be toxic to β-cells of islets and are considered as a major cause of islet graft failure after transplantation. Thus, the current study raises the possibility that introduction of a novel immunosuppressive regime targeting NKT cells may lead to a beneficial outcome of clinical islet transplantation with reduction and/or replacement of FK506 and rapamycin.

This work was supported by funds (0211001) from the Central Research Institute of Fukuoka University, by health science research grants from the Ministry of Health, Labor, and Welfare, Japan, and a grant-in-aid for scientific research (B) from the Japan Society for the Promotion of Science.

We thank Dr. Moriya Tsuji, Rockefeller University, for providing CD1d^−/− mice.