New Insights into Mechanisms of Spontaneous Liver Transplant Tolerance: The Role of Foxp3-Expressing CD25+CD4+ Regulatory T Cells

Authors


  • This study was presented in part at the American Transplant Congress 2004, May 15–19, Boston, MA, the XX International Congress of the Transplantation Society, 5–10 September 2004, Vienna, Austria and at the World Transplant Congress 2006, July 22–27, Boston, MA.

* Corresponding author: Wei Li, weili8@u.washington.edu

Abstract

Liver allografts in mice are accepted across MHC barriers without requirement for immunosuppressive therapy. The mechanisms underlying this phenomenon remain largely undefined. In this study, we investigated the role of Foxp3-expressing CD25+CD4+ regulatory T cells (Treg) in the induction of murine liver transplant tolerance. Foxp3+CD25+CD4+ T cells were increased in liver grafts and recipient spleens from day 5 to day 100 posttransplantation, associated with enhanced CTLA4 and TGF-β expression and IL-4 production. Depletion of recipient CD25+CD4+ T cells using anti-CD25 mAb (250 μg/day) induced acute liver allograft rejection. This was associated with a decreased ratio of Foxp3+ Treg: T effector cells, decreased IL-4 and elevated IL-10 and IL-2 production by graft-infiltrating T cells, and reduced apoptotic activity of graft-infiltrating CD4+ and CD8+ T cells in anti-CD25-mAb-treated recipients. Thus, the data suggest that Foxp3+CD25+CD4+Treg are involved in spontaneous acceptance of liver allografts in mice. The ratio of Treg to T effector cells appears to determine liver transplant outcome. CTLA4, IL-4, TGF-β and apoptosis of graft-infiltrating T cells are also associated with liver transplant tolerance and may contribute, at least in part, to the mechanisms of Treg-mediated immune regulation in this model.

Abbreviations: 
B10

C57BL/10

DC

dendritic cell

GIC

graft-infiltrating cell

NPC

nonparenchymal cell

OLTx

orthotopic liver transplantation

SC

spleen cell

Treg

T regulatory cell

Introduction

In mice, liver allografts are accepted across MHC barriers without requirement for immunosuppressive therapy (1,2). The underlying mechanisms remain unclear. Recently, we and others have shown that liver dendritic cells (DCs), apoptosis of liver graft-infiltrating cells (GICs) and recipient spleen cells (SCs), especially CD8+ T cells, contribute to the development of liver transplant tolerance (3–5). Blockade of CTLA4, a negative signal to effector T cells and an effector mechanism of Foxp3+CD25+CD4+ regulatory T cells (Treg), or depletion of CD25+CD4+ T cells prevents spontaneous liver acceptance (6–8). These findings suggest that the induction of liver transplant tolerance relies upon active suppression of immune reactivity, possibly mediated by Treg.

CD25+CD4+ T cells, a naturally occurring subset of CD4+ T cells, represent a unique population of immunoregulatory cells that play an important role in maintaining self-tolerance, regulating graft rejection and graft-versus-host disease (9–13). Foxp3, a member of the forkhead winged helix protein family of transcription factors, was first identified as a specific molecular marker for CD25+CD4+ Treg that control the development and function of these cells (14,15). A majority of Foxp3+ cells are both CD4+ and CD25+; only a very small population of CD25CD4+ cells and CD8+ cells express Foxp3 (15,16). Other Treg-associated molecules, such as CTLA4 are also dominantly expressed by CD25+CD4+ Treg and upregulated by Foxp3 transduction (14). Transfer of CD25+CD4+ Treg to naive mice can abrogate autoimmune disease and promote allograft survival (17–19). Depletion of CD25+CD4+ T cells can induce autoimmune disease and antitumor immunity (11,20). Recent reports indicate that depletion of CD25+CD4+ T cells also induces MHC class II-mismatched heart allograft rejection and prevents islet allograft tolerance in mice (21,22). Importantly, CD25+CD4+ Treg are not only generated in the thymus, but can also be generated in the periphery (18,23,24), holding considerable promise for therapy of autoimmune disease and allograft rejection.

To explore the mechanisms underlying liver tolerogenicity and the induction of donor-specific tolerance, we characterized Foxp3-expressing CD25+CD4+ Treg in vivo in a murine model of orthotopic liver transplantation (OLTx). We examined whether depleting this population of Treg could prevent ‘spontaneous’ liver allograft tolerance, and if so, which source(s) of Treg, donor or recipient, might play a role, and by what mechanism(s). Our study demonstrates, for the first time, that Foxp3+CD25+CD4+ Treg are increased in murine liver allograft recipients. Novel data also show that this increase correlates with enhanced CTLA4 and transforming growth factor-beta (TGF-β) expression, and with increased interleukin-4 (IL-4) production. Depletion of recipient Foxp3+CD25+CD4+ Treg in OLTx resulted in acute allograft rejection. This effect was associated with a decreased ratio of Treg to effector T cells. It was also linked to increased IL-2 and IL-10, and decreased IL-4 production and to reduced apoptosis of graft-infiltrating T cells. Thus, in this model, recipient Foxp3+CD25+CD4+ Treg appear necessary for ‘spontaneous’ acceptance of liver allografts and contribute to regulation of peripheral T-cell tolerance. The mechanisms underlying Foxp3+CD25+CD4+ Treg-mediated liver transplant tolerance appear to involve CTLA4 and IL-4 signaling and the apoptosis of graft-infiltrating T cells.

Materials and Methods

Animals

Male C57BL/10 (B10; H2b) and C3H/HeJ (C3H; H2k) mice, 10–12 weeks of age, were purchased from The Jackson Laboratory, Bar Harbor, ME, and maintained in a specific pathogen-free facility of the University of Washington Medical Center. The mice were provided with Purina Rodent Chow and tap water ad libitum. Animal care was in compliance with an institutional animal care and use committee-approved protocol and with the ‘Guide for the Care and the Use of Laboratory Animals’ published by the National Institutes of Health.

Liver and heart transplantation

OLTx with revascularization accomplished with a combination of suture and cuff techniques, was performed as described (from B10 donors to C3H recipients) (1). Allograft survival was determined by recipient survival, and rejection was confirmed histologically. Heterotopic heart transplantation was performed with an end-to-side anastomosis of donor aorta and pulmonary artery to the recipient abdominal aorta and inferior vena cava, as described previously (25). Heart graft survival was monitored by abnormal palpation, and rejection was defined as total cessation of cardiac muscle contraction.

Liver nonparenchymal cell (NPC), GIC, SC and T-cell preparation

NPCs or liver GICs were isolated as described (26). In brief, liver tissue was mashed between two glass slides, ground with a 1-mL syringe pestle, then filtered through a 70-μm cell strainer before Percoll (Sigma, Milwaukee, WI) gradient centrifugation. SC suspensions were filtered through nylon mesh. Erythrocytes were lysed in red cell lysing buffer (BD PharMingen, San Diego, CA). T cells from liver and SC suspensions were purified by immunomagnetic bead isolation (negative selection) using a MACS® T-cell isolation kit, according to the manufacturer's instructions (Miltenyi Biotec Inc., Auburn, CA). The purity of CD3+ T cells was consistently > 95%, as determined by flow cytometry.

Antibody (Ab) administration

Rat anti-mouse CD25 monoclonal antibody (mAb) (PC61; Chimerigen, Allston, MA) was administered on alternative days (250 μg/day; i.p.) to either liver donors or recipients on days −6, −4, −2 pre- or days 0, 1, 2, 3 posttransplantation, respectively. Control animals received rat IgG (250 μg) (Sigma, St. Louis, MO) using an identical regimen.

Flow cytometry

To characterize cell phenotypes, 4-color flow cytometry was performed. GICs and host SCs were stained with fluorochrome-conjugated mAbs against mouse CD3, CD4, CD8, CD25, CTLA4 (BD PharMingen, San Diego, CA), or Foxp3 (eBioScience, Inc., San Diego, CA). Data were analyzed with an LSR1 flow cytometer (Becton Dickinson, San Jose, CA) and Flowjo software (Tree Star, Inc., Ashland, OR).

ELISPOT assay

To determine the frequencies of cytokine-producing cells in liver grafts and host spleens, enzyme-linked immunosorbent spot (ELISPOT) assay was performed as described (6). T cells were isolated as described above from naive C3H mice or liver transplant recipients on day 5 posttransplant following treatment either with anti-CD25 mAb, or control rat IgG. They were stimulated in vitro in precoated ELISPOT plates (Millipore, Bedford, MA); 105 cells were plated per well in 100 μL RPMI-1640 complete medium in triplicate. An equal amount of γ-irradiated B10 SCs were used as stimulators. The plates were cultured for 2 days for IL-2 and IFN-γ and 3 days for IL-4 and IL-10 detection. Capture and detection Abs for IL-2, IFN-γ, IL-4 and IL-10 were purchased from BD Biosciences (San Diego, CA). Spots were counted using a CTL series 3A analyzer (CTL Analyzers LLC, Cleveland, OH). Results were expressed as mean number of spots (cytokine-producing cells) per 106 cells.

Immunohistochemistry

To examine Foxp3 and TGF-β protein expression, cryostat sections (4 μm) of liver grafts and host spleens were stained with anti-mouse Foxp3 and TGF-β mAbs (eBioScience, San Diego, CA), respectively, using an avidin-biotin-peroxidase complex (ABC) method, as described (6). Sections were fixed in acetone at −20°C. Endogenous peroxidase activity was quenched in 1% H2O2 before the addition of biotin-mouse Foxp3 Ab. Rat IgG2a was used as an isotype control. ABC (Vector Laboratories, Inc., Burlingame, CA) was then added. Aminoethylcarbazole (AEC) (Vector laboratories, Inc.) was used as the substrate and sections were counterstained with hematoxylin.

Apoptosis

Apoptotic cells in frozen sections (4 μm) were identified using the ApopTag Peroxidase In Situ Apoptosis Detection kit (Intergen, Purchase, NY) and followed with the manufacturer's instruction, as described (3–5). Apoptotic cells in GIC and SC suspension were detected by the Fluorescein In Situ Cell Death Detection Kit (Roche Diagnostics Corporation, Indianapolis, IN) as described (4,5).

Statistical analysis

Graft survival was analyzed using Kaplan–Meier log rank test. Other comparisons were made using Student's t-test. A value of p < 0.05 was considered significant.

Results

Foxp3+CD25+CD4+ T cells are increased in liver allografts and host spleens posttransplantation

We first examined CD25+CD4+ T cells in liver GIC and host SC populations at different times posttransplant using 4-color flow cytometry. As shown in Figure 1, naturally occurring CD25+CD4+ T cells were detected in both normal liver NPC and SC populations. The incidence of CD25+CD4+ T cells within the CD3 cell population increased in both liver grafts and recipient spleens following transplantation. Levels peaked within 2 weeks of transplantation, and declined thereafter (Figure 1B). These CD25+CD4+ T cells expressed intracellular Foxp3 transcripts (>95%) and CTLA4 (>80% in GIC and >60% in SC) throughout the time course of the experiment (Figure 1C), indicating that the CD25+CD4+ T cells detected after liver transplantation expressed a Treg phenotype. The average incidence of Foxp3+CD25+CD4+ cells showed statistically significant increases at day 5 and day 7 posttransplantation in the liver grafts and recipient spleens, at day 14 posttransplantation in the liver grafts and at day 100 posttransplantation in the recipient spleens (Figure 1D). The incidence of Foxp3+CD25+CD4+ cells did not change significantly in either the livers or spleens of syngeneic liver graft recipients (data not shown).

Figure 1.

Foxp3+CD25+CD4+ T cells within grafts and host spleens are increased following orthotopic liver transplantation (OLTx). Liver GICs and recipient SCs were stained for PerCp-CD3, APC-CD4, FITC-CD25, PE-Foxp3 and CTLA4, and analyzed by flow cytometry (A). Gating strategy: CD25+CD4+ T cells were analyzed within the CD3 positive gate; Foxp3 and CTLA4 were analyzed on CD25+ CD4+ double positive cells (B) and (C). The percentage of CD25+CD4+ T cells at various times posttransplant from day 5 up to day 100 (B), and the percentage of Foxp3- and CTLA4-expressing CD25+CD4+ T cells (C), the results are representative of 2–3 separate experiments (D). Average incidence of Foxp3+CD25+CD4+ T cells in the GIC and SC population throughout the 100 days post liver transplantation. The cells were analyzed within CD3 positive gate and were pooled from 2–3 mice in each group, and 2–3 separate experiments were performed at different time points. * p < 0.01, †p < 0.05, compared to normal mice.

Adoptive transfer of the cells from long-term survivors of liver allograft recipients prolong heart allograft survival

To examine whether the cells from liver allograft recipients have regulatory function in vivo, we adoptively transferred the cells, either GIC, SC or CD4+ T cells, isolated from long-surviving liver allograft recipients to the heart allograft recipients 1 day before heart transplantation. The heart allograft survival was significantly prolonged from the same liver donors in a dose-dependent manner. This effect was enhanced by the transfer of purified CD4+ T cells from SC sources of liver graft recipients, but not from the CD4+ T cells of naive mice (Figure 2). However, the heart grafts from the third-party BALB/c mice were rejected acutely (data not shown), further suggesting an antigen specific regulatory capacity of liver GIC and recipient SC, associated with an increased incidence of Foxp3+CD25+CD4+ cells in these sites.

Figure 2.

Adoptive transfer of GIC or SC from long-term liver allograft recipients protects heart allografts from the same liver donors from acute rejection. The C3H heart allograft recipients received GIC (4 × 106), SC (20 × 106) or spleen CD4 T cells (4 × 106) isolated from long-term liver allograft recipients (B10 → C3H, >70 days posttransplantation) 1 day before B10 heart transplantation, respectively. Heart allograft survival was prolonged in the recipients that received the cell transfers from long-term liver graft recipients. MST were 32.4 ± 37.8 days in GIC treated group, 51.3 ± 34.4 days in SC-treated group, and 80.7 ± 33.5 days in CD4 T-cell-treated group, p < 0.05 in SC- and CD4-treated groups versus naive CD4 T-cell-treated (11.3 ± 1.3 days) and nontreated (10.8 ± 0.9 days) controls.

Foxp3 and TGF-β expression are elevated in liver allografts and host spleens

We next examined Foxp3 and TGF-β expression in the livers and spleens of allograft recipients by immunohistochemistry. Foxp3 has been implicated as a specific marker of CD25+CD4+ Treg, while TGF-β is also an important inhibitory cytokine produced by Treg. In agreement with the flow analysis, the incidence of Foxp3+ cells detected in frozen sections of liver grafts and host spleens was increased posttransplant, for up to 100 days. Foxp3+ cells were detected occasionally in the portal areas of normal livers and also scattered throughout the interfollicular areas of normal spleens (Figure 3). Foxp3+ cells were not detected on the sections of acutely rejected heart allografts at day 7 posttransplantation in the same strain combination in liver transplantation, and were markedly reduced in the recipient spleens (data not shown). TGF-β+ cells were also detected at a significantly higher level in the liver grafts and recipient spleens throughout the observation period, and with similar kinetics to Foxp3 expression (Figure 4). However, it is uncertain that the TGF-β+ cells are Treg since TGF-β is potentially expressed by a variety of cell types. Our results suggest that TGF-β may be a key cytokine in promoting liver allograft tolerance.

Figure 3.

Foxp3+ cells are increased in liver allografts and recipient spleens following transplantation. (A) Foxp3 staining of cryostat sections of liver allografts or recipient spleens. The results are representative of 2–3 separate experiments with 2–3 mice in each group. Magnification, ×200 (insert, ×400). (B) Quantitation of Foxp3+ cells in sections of liver allografts and recipient spleens. Foxp3+ cells (number per 40 high power fields [hpf]) were enumerated in five sections per tissue per mouse, with three mice in each group. The results are means ± 1 SD. *p < 0.001, #p < 0.01 compared to naive control.

Figure 4.

TGF-β-producing cells are increased in liver allografts and recipient spleens following transplantation. (A) TGF-β staining of cryostat sections of liver allografts or recipient spleens. The results are representative of 2–3 separate experiments, with 2–3 mice in each group. Magnification, ×200 (insert, ×400). (B) Quantitation of TGF-β+ cells in sections of liver allografts and recipient spleens. TGF-β+ cells (number per 40 high power fields [hpf]) were enumerated in 5 sections per tissue per mouse, with three mice in each group. The results are means ±1 SD. p < 0.01 in all OLTx groups compared to naive control.

Depletion of host CD4+CD25+T cells prevents the induction of liver transplant tolerance

To ascertain the role of CD25+CD4+ T cells in spontaneous liver tolerance induction, we employed anti-CD25 mAb to deplete CD25+CD4+ T cells in liver donors or recipients prior to transplantation. First, we confirmed that three doses of anti-CD25 mAb administration (PC61; 250 μg, on alternate days), led to >70% depletion of CD25+CD4+ T cells in livers and spleens in naive mice (Figure 5A), consistent with reports by others (27–29). When anti-CD25 mAb was administered to recipient mice prior to liver transplant (days −4, −2, 0) or following liver transplantation (days 0, +2, +4), grafts were acutely rejected. The mean survival time [MST] was 14.9 ± 9.8 days (n = 8) in pretreated recipients, or 16.8 ± 12.6 days (n = 6) in posttransplant-treated recipients (Figure 5B). Control (rat IgG)-treated mice, either pre- or postoperatively, did not alter liver allograft survival (MST >100 days; n = 6, 3 for each treatment regimen). Livers from anti-CD25 mAb pretreated donor mice experienced long-term survival (MST = 71.5 ± 34.5 days; n = 8); 5 out of 8 mice survived >100 days. These data suggest that recipient, but not donor CD25+CD4+ T cells are necessary for murine liver transplant tolerance. Depletion of recipient CD25+ cells resulted in acute liver rejection.

Figure 5.

(A) Anti-CD25 mAb administration efficiently depletes CD25+CD4+ T cells in the liver and spleen. Normal C3H mice were treated with three doses of rat anti-mouse CD25 mAb (PC61), 250 μg i.p., every other day. Two days after the last treatment, CD25+CD4+ T cells within the liver NPC and SC population were analyzed by 3-color flow cytometry (Cyc-CD3, FITC-CD25 and PE-CD4). The results are representative of three separate experiments. (B) Depletion of CD25+CD4+ T cells prevents liver transplant tolerance. Anti-CD25 mAb was given to either B10 donors or C3H liver recipients (250 μg i.p.; d 6, –4, –2 or d 0, 1, 2, 3). The graft survival times were: (1), 72 ± 34 days in recipients of livers from B10 donors treated with anti-CD25 mAb (n = 8); 2), (2); 15 ± 10 days in recipients pretreated with anti-CD25 mAb (n = 6); 3), 17 ± 12 days in recipients treated posttransplant with anti-CD25 mAb (n = 8); and 4), >100 days in recipients that received control rat IgG (n = 6; 3 pretreated and 3 post-transplant-treated). Groups 2 and 3 versus group 4, p < 0.01.

Administration of anti-CD25 mAb reduces Foxp3+CD25+CD4+ Treg in liver allografts and host spleens

To further address the role of Foxp3+CD25+CD4+ Treg in liver transplant tolerance, we examined the incidence of Foxp3+CD25+CD4+ Treg in liver grafts and host spleens posttransplant by flow cytometry. On day 5 posttransplant, CD25+CD4+ Treg were still detectable and were maintained at higher levels in livers and spleens of graft recipients, despite acute rejection resulting from anti-CD25 mAb treatment. The incidence of CD25+CD4+ Treg within the liver GIC population of anti-CD25 mAb pre- or posttreated recipients was similar to OLTx control mice (Figure 6A); it was reduced moderately in SC of anti-CD25 pre- or posttreated recipients in comparison with cells from untreated OLTx recipients and normal mice (Figure 6B); over 95% of these CD25+CD4+ Treg expressed intracellular Foxp3 and CTLA4 (Figure 6A, B). However, the total percentages of CD3 T cells were increased in the GICs and SCs of anti-CD25 mAb-treated recipients. Therefore, the ratios of Foxp3+CD25+CD4+ Treg to total CD3+ T cells were reduced significantly in both GIC and SC of anti-CD25 mAb pre- or posttreated recipients, but not donor pretreated recipients (Figure 6C). Immunohistochemistry staining on sections of liver graft recipients revealed a marked increase in GICs in the liver grafts and significantly reduced Foxp3+ cells in both livers and spleens of anti-CD25 pre- or posttreated recipients on day 5 post-transplant compared to the OLTx control group (Figure 7A, B). These results indicate that administration of anti-CD25 mAb to liver graft recipients decreases the absolute number of Foxp3+CD25+CD4+ Treg. Although CD25 may be potentially expressed by other cell types, including activated T cells, CD25+CD4+ T cells detected posttransplant in anti-CD25 mAb-treated recipients likely equate with Foxp3+ Treg in this model. The ratio of Foxp3+CD25+CD4+ Treg to effector T cells may serve as a key indicator of the outcome of liver allografts.

Figure 6.

Figure 6.

Administration of anti-CD25 mAb reduces Foxp3+ CD25+CD4+ T cells in liver allografts and host spleens. Foxp3-expressing CD25+CD4+ T cells within GIC and SC populations of liver allograft recipients were analyzed by 4-color flow cytometry. The gating strategy was as described in Figure 1A. The percentage of CD25+CD4+ T cells in either GIC or SC of the anti-CD25 mAb-treated donors or recipients was detected at higher levels in the liver grafts (A) and at reduced levels in the spleens (B) compared to the control OLTx group. These T cells also expressed intracellular Foxp3 and CTLA4 (>95%). The ratio of CD25+CD4+ Treg to total CD3+ T cells in the liver GIC and SC of anti-CD25 mAb-treated recipients was reduced significantly (C). The results are representative of three separate experiments, and the cells were pooled from two mice in each group.

Figure 6.

Figure 6.

Administration of anti-CD25 mAb reduces Foxp3+ CD25+CD4+ T cells in liver allografts and host spleens. Foxp3-expressing CD25+CD4+ T cells within GIC and SC populations of liver allograft recipients were analyzed by 4-color flow cytometry. The gating strategy was as described in Figure 1A. The percentage of CD25+CD4+ T cells in either GIC or SC of the anti-CD25 mAb-treated donors or recipients was detected at higher levels in the liver grafts (A) and at reduced levels in the spleens (B) compared to the control OLTx group. These T cells also expressed intracellular Foxp3 and CTLA4 (>95%). The ratio of CD25+CD4+ Treg to total CD3+ T cells in the liver GIC and SC of anti-CD25 mAb-treated recipients was reduced significantly (C). The results are representative of three separate experiments, and the cells were pooled from two mice in each group.

Figure 7.

Foxp3+ cells are reduced in liver allografts and spleens of anti-CD25 mAb-treated recipients. (A) Foxp3 staining on cryostat sections of liver allografts or recipient spleens at day 5 posttransplantation. The results are representative of 2–3 separate experiments, with 2–3 mice in each group. Magnification, ×200 (insert, ×400). (B) Quantitation of Foxp3+ cells in sections of liver allografts and recipient spleens. Foxp3+ cells (number per 40 high power fields [hpf]) were enumerated in five sections per tissue per mouse, with three mice in each group. The results are means ± 1 SD. *p < 0.001 compared to OLTx control.

Cytokine profiles of graft-infiltrating and spleen T cells of anti-CD25 mAb-treated recipients

To determine the cytokine profiles of anti-CD25-treated liver graft recipients and to evaluate cytokine-producing cells and clone sizes from liver grafts and host spleens, T cells from liver GICs and recipient SCs were isolated 5 days posttransplant and ELISPOT assays conducted. The results revealed a significant increase in IL-4-producing cells in the accepted liver grafts (OLTx controls). IL-4-producing cells were reduced significantly in the livers of anti-CD25-treated recipients (both pre- and posttransplant treated) compared with OLTx controls. Conversely, there was a significant increase in IL-2- and IL-10-producing T cells in the livers and of IFNγ-producing cells in the spleens of the anti-CD25-treated recipients (Figure 8). There was no significant difference in cytokine-producing cells between OLTx controls and those given grafts from anti-CD25 pretreated donors. These results indicate that IL-4 may play an important role within liver grafts in promoting liver transplant tolerance. Depletion of recipient CD25+CD4+ T cells by anti-CD25 mAb administration, pre- or postoperatively, decreased intrahepatic and splenic IL-4-producing T cells and promoted increased numbers of IL-2- and IL-10-producing cells in the graft. The reduction in IL-4 production in liver grafts that were rejected acutely as the result of anti-CD25 mAb treatment suggested that IL-4 might contribute significantly to the regulation of alloimmune responses by Foxp3+CD25+CD4+ T cells.

Figure 8.

Anti-CD25 mAb administration decreases IL-4 production in liver grafts and spleens but increases IL-2 and IL-10 production in the grafts and IFN-γ production in host spleens. Purified T cells from either liver GIC or SC populations were isolated at day 5 posttransplant from naive C3H, OLTx control, OLTx with anti-CD25 pretreated donor, or OLTx with anti-CD25 pre- or posttreated recipients and were restimulated in vitro with γ-irradiated donor SCs at a 1:1 ratio, then cultured for 2 days (for IL-2 and IFN-γ quantitation) or 3 days (for IL-4 and IL-10 quantitation) before analysis by ELISPOT. Each positive spot was counted as one cytokine-producing cell. The results were expressed as number of spots per 106 cells. The cells were pooled from two mice in each group and the data are representative of two separate experiments. *p < 0.01, †p < 0.05, compared to OLTx controls.

Administration of anti-CD25 mAb protects graft-infiltrating T cells from apoptotic death

We have shown previously (3–5) that spontaneous liver allograft acceptance in mice is associated with a relatively high incidence of apoptosis of GICs that is linked to downregulation of the cytotoxic activity of these cells. To determine the influence of CD25+CD4+ Treg on apoptotic death of GICs and SCs, we performed in situ TUNEL staining on tissues harvested 5 days posttransplant. Histological examination of control Ig-treated liver graft sections revealed GICs located mainly in the periportal areas. Many of those cells were undergoing apoptosis (Figure 9A, top left). By contrast, liver allograft recipients treated with anti-CD25 mAb, either pre- or posttransplant, showed substantially increased lymphocyte infiltration in the portal and parenchymal areas, with a reduction in the frequency of apoptotic cells (Figure 9A, top right 2 panels). Similarly, anti-CD25 mAb-treated host spleens showed an expanded T-cell area and reduced T-cell apoptosis compared with control Ig-treated recipient spleens (Figure 9A, lower panels), while the apoptotic cell frequency and histological appearance of the grafts and spleens in recipients of anti-CD25 pretreated donor livers was similar to the Ig control group. Quantification of apoptotic cells in sections of liver grafts and recipient spleens confirmed these observations (Figure 9B). TUNEL staining of GICs confirmed that significant numbers of CD4 and CD8 T cells underwent apoptosis in the GICs from control Ig-treated allograft recipients, while the incidence of apoptotic CD4 and particularly CD8 T cells was decreased in the GICs of anti-CD25 mAb-treated recipients (Figure 9C). Thus, CD25+CD4+ T-cell depletion, using anti-CD25 mAb, increases graft-infiltrating T-cell numbers and survival and reduces CD4+ and CD8+ T-cell apoptosis compared with control (OLTx) mice.

Figure 9.

CD4+CD25+ Treg depletion protects alloreactive T cells from apoptotic death. (A) Cryostat sections obtained 5 days after liver transplantation together with anti-CD25 mAb or control rat IgG administration (250 μg i.p.; day 0, 1, 2, 3 or day −6, −4, −2) were stained (red) by TUNEL using antidigoxigenin-peroxidase. AEC was used for color development, and the sections were counterstained with hematoxylin. The results are representative of three animals in each experimental group. Magnification ×200. (B) Quantitation of the numbers of apoptotic cells in cryostat sections of liver allografts and recipient spleens. Apoptotic cells (number per 40 high power fields [hpf]) were enumerated in five sections per tissue per mouse, with three mice in each group. The results are means ± 1 SD. Comparisons of apoptotic cells in liver grafts and recipient spleens between anti-CD25-treated and rat IgG-treated controls, *p < 0.01, Δp < 0.05. (C) Quantitation of the numbers of apoptotic liver GIC by flow cytometry. GICs from control IgG and anti-CD25-treated recipients were isolated, cultured overnight in complete RPMI-1640 medium, and then stained by TUNEL. The percentages of apoptotic cells were analyzed in the whole lymphocyte gate.

Discussion

Our findings demonstrate that Foxp3+CD25+CD4+ T cells are markedly increased in the graft and recipient spleen following allogeneic OLTx. Elevated levels peaked during the first 2 weeks after transplantation and were maintained in the spleen up to 100 days after grafting. These Foxp3+CD25+CD4+ cells also expressed high levels of CTLA4, a functional marker of CD25+CD4+ Treg, as well as the transcription factor Foxp3, which programs their development and function (14,15). Therefore, the number of Foxp3+ cells usually reflects an equal number of CD25+CD4+ Treg. Our data clearly demonstrate that Foxp3+CD25+CD4+ Treg are expanded post liver transplantation, suggesting that they play a key role in inducing or maintaining ‘spontaneous’ liver transplant tolerance. Depletion of recipient CD25+CD4+ T cells by administration of anti-CD25 mAb, either pre- or postoperatively, significantly reduced the survival of liver allografts. Indeed, all the grafts were rejected within 30 days, whereas those in control Ig-treated mice achieved long-term survival (>100 days). Notably, depletion of donor CD25+CD4+ T cells did not induce acute liver rejection. The findings are consistent with Jiang et al. (8) who showed that CD25+CD4+ Treg in the recipient but not the donor, were important in mouse liver transplant tolerance.

The present studies extend the observations of Jiang et al. (8) in several ways. First, we have performed precise quantification of Treg (as opposed to Foxp3 transcripts) and enumerated these cells at multiple time points (as opposed to a single time point) after liver transplantation. Second, we have analyzed the expression of molecules (CTLA4, IL-4, IL-10, TGF-β) associated with Treg function and immune regulation, and produced by graft-infiltrating cells. Compared with Jiang et al. (8), who reported that anti-CD25 mAb (PC61) administration (x1) depleted CD25+CD4+ T cells to 0% for 1 month, we observed >70% depletion, similar to levels (>80%) reported by others (28) after four injections of the same dose of this mAb. A further difference between our findings and those of Jiang et al. is that while we report reduced apoptosis of CD4+ and CD8+ T cells after depletion of CD25+ cells, the former authors show increased proliferation of both the CD4+ and CD8+ T cells. Taken together, these findings demonstrate the importance of recipient CD25+CD4+ Treg for liver transplant tolerance, and the role of specific cytokines and apoptosis in the regulation of graft outcome. Although it has been suggested that anti-CD25 mAb treatment may inactivate but not preferentially deplete CD25+CD4+ Treg, there is strong evidence that PC61 does indeed deplete a large fraction of CD25+ Treg (30). Possibly, anti-CD25 mAb administration may deplete other types of CD25+ cells in our model, including activated T cells. The different mechanisms remain unclear. Depletion of CD25+CD4+ T cells by anti-CD25 mAb treatment, resulting in the prevention of allograft tolerance, has also been reported in murine heart and pancreatic islet transplantation (21,22). Studies in human heart transplant patients have demonstrated high Foxp3 gene transcript levels during allogeneic responses in vivo or in vitro, suggesting that regulatory activities of CD25hi T cells or the generation of these cells, is an intrinsic part of immune activation. Interference in these events by anti-CD25 mAb inhibits tolerance induction after organ transplantation (31).

In our studies, Foxp3+CD25+CD4+ T cells could still be detected in the rejecting liver grafts and recipient spleens following anti-CD25mAb administration. The percentage of CD25+CD4+ T cells in the liver grafts of anti-CD25 mAb-treated hosts was as high as in OLTx controls, while it was reduced in the SC of anti-CD25 mAb-treated recipients. This suggested that the liver allograft might be a site of CD25+CD4+ Treg induction/expansion/recruitment, with the depleting effect of anti-CD25 mAb more effective in peripheral lymphoid organs than in the liver, since Foxp3+CD25+CD4+ cells continued to increase in the liver after OLTx. However, the ratio of Foxp3+CD25+CD4+ T cells to CD3+ T cells decreased in the grafts and recipient spleens, indicating that, although Foxp3+CD25+CD4+ T cells were induced or expanded after liver transplantation, they were not as active as T-cell expansion due to anti-CD25 mAb treatment. The balance between Foxp3+CD25+CD4+ T cells and effector T cells appears to determine the outcome of liver transplantation. However, the mechanisms that control induction or expansion of Foxp3+CD25+CD4+ Treg after liver transplantation are largely unknown. We propose that these cells may be induced by liver-derived DCs and their expansion/function affected by immunoregulatory cytokines and coregulatory molecules. Our previous studies have suggested that liver DCs play an important role in determining the outcome of liver transplantation (4,5). Liver-derived DCs, lacking adequate levels of costimulatory molecules, are poor allogeneic T-cell stimulators and can induce donor-specific T-cell hyporesponsiveness (32,33). It has been reported recently that DCs can direct T-cell differentiation to antigen-specific Treg (24,34–38).

The mechanism by which CD4+CD25+ Treg downregulate T-cell responses in vivo remains to be determined. It has been shown that Treg suppress immune responses by affecting APCs, effector T-cell activation and proliferation, Ab production and cytokine secretion (18,19,39) by means of direct cell–cell interaction (39,40), or via IL-10, TGF-β or CTLA4-mediated effects (17,18,41). In the present study, we have detected a large amount of TGF-β+ cells in the liver grafts and recipient spleens, indicating that TGF-β may, at least in part, contribute to the mechanism of Treg, although it could be produced by other types of cells. We have also demonstrated that Foxp3+ T cells, which expressed higher CTLA4, were increased in the liver grafts and recipient spleens, suggesting that CTLA4 may be involved in the immunosuppressive function of CD25+CD4+ Treg. We have reported recently that CTLA4 blockade using anti-CTLA4 mAb prevents mouse ‘spontaneous’ liver transplant tolerance (6), further suggesting that CTLA4 not only provides a negative signal to activated T cells, but may be also involved in the regulatory mechanism of CD25+CD4+ Treg. Studies by others in a murine colitis model (17,41) have demonstrated that CTLA4 signaling is required for the function of CD25+CD4+ Treg, and that blockade of CTLA4 signaling by anti-CTLA4 mAb blocks the role of CD25+CD4+ Treg in the control of intestinal inflammation. Ligation of CD80/CD86 on DCs by CTLA4 on CD25+CD4+ Treg may trigger production of the tryptophan-catabolizing enzyme indoleamine 2,3 deoxygenase that inhibits T-cell proliferation and enhances apoptotic cell death (42,43).

We have shown herein that liver allograft acceptance is associated with strikingly increased numbers of IL-4-producing (Th2) cells in the liver and spleen. Depletion of recipient CD25+CD4+ T cells reduced IL-4-producing T cells and promoted IL-2- and IL-10- secreting T-cell expansion in the liver and spleen. This finding suggests that liver transplant tolerance in mice may be mediated in part by Th2-like cells that secrete IL-4. However, it should also be noted with caution that anti-CD25 mAb administration may activate some Th1 cells that produce IL-2, and possibly IFN-γ, which may cause rejection. The role of IL-4 in transplant tolerance has also been demonstrated in a mouse cardiac allograft model. Anti-IL-4 mAb administration abrogates heart allograft tolerance induced by anti-CD2 mAb plus anti-CD3 mAb that does not develop in IL-4−/- mice. Restoration of IL-4 to IL-4−/− mice by gene transfer prolongs heart allograft survival (44). Whether IL-4 signaling is involved in CD25+CD4+ Treg development and function is controversial. Several studies have suggested that IL-4 may be involved in CD25+CD4+ Treg-mediated suppression. Thus, intragraft IL-4 gene transfer enhances the tolerogenic effects of systemic infusion of CD25+CD4+ Treg (45), whereas IL-4 potentiates the suppressive activity of CD25+CD4+ Treg in vitro (46). On the other hand, IL-10 has been reported to promote CD25+CD4+ Treg differentiation and function (17,18,47), but not in the mouse liver transplant model. In the present study, the increase in IL-10-producing cells within the liver grafts of anti-CD25-treated recipients, together with IL-2, likely contributes to the elevated anti-donor immune responses and acute liver graft rejection (5,6,48). Together, those and other data suggest that IL-4 may contribute to liver tolerance induction by enhancing CD25+CD4+ Treg function.

Depletion of CD25+CD4+ T cells reduced apoptosis of GICs and recipient SCs, particularly liver graft CD8+ T cells, and enhanced both lymphocyte accumulation in the graft and T-cell expansion in the spleen. This suggests that one mechanism by which CD25+CD4+ Treg may act in this model is via induction of activated T-cell apoptosis. We have shown previously that apoptosis of liver GICs and SCs may contribute to liver transplant tolerance (3–5). Other reports indicate that CD25+CD4+ Treg can promote activated CD8+ T-cell apoptosis and suppress skin allograft rejection (49), whereas coculture of CD25+CD4+ Treg and effector T cells induces apoptosis of the effector cells (50). Recently, CD25+CD4+ Treg have been shown to induce apoptosis of effector CD4+ T cells through a cytokine-dependent pathway in a murine inflammatory bowel disease model (51).

In summary, our findings argue that Foxp3+CD25+CD4+ Treg may contribute to murine liver transplant tolerance. The regulatory effects of CD25+CD4+ Treg in vivo may be mediated, at least in part, by CTLA4, TGF-β, IL-4 and apoptosis of graft-infiltrating T cells. Depletion of host CD25+CD4+ T cells induces acute liver allograft rejection, associated with a decreased Treg: T effector cell ratio, reduced IL-4 production and diminished apoptotic death of alloreactive GICs and SCs. These new insights into naturally occurring ‘immune tolerance’ provide a platform upon which to better understand how the mammalian immune system may be manipulated through regulatory mechanisms to optimize their effect in clinical organ and cell transplantation.

Acknowledgments

The authors thank Dr. Paul Warner of Puget Sound Blood Center, Seattle, WA for his help in ELISPOT analysis, Ms. Zeila Schmidt for statistical analysis and Ms. Marilyn Carlson for administrative support. This work was supported by funding from the Division of Transplantation, Department of Surgery and the Provost Bridge Fund of University of Washington, Seattle, WA.

Ancillary