Roles of Deletion and Regulation in Creating Mixed Chimerism and Allograft Tolerance Using a Nonlymphoablative Irradiation-Free Protocol
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
The induction of mixed chimerism (MC) is a powerful and effective means to achieve transplantation tolerance in rodent models. Host conditioning with irradiation or cytotoxic drugs has been used in many protocols for chimeric induction across allogeneic barriers. The deletion of alloreactive T cell clones has been described as the main mechanism responsible for the induction of a stable MC. In this study, we demonstrate that a stable MC and skin allograft tolerance can be established across MHC barriers by a noncytotoxic, irradiation-free approach using costimulation blockade plus rapamycin treatment. By using an adoptive transfer model of skin allograft and using specific V TCR probes, we demonstrated that deletion of donor-reactive cytopathic T cell clones is indeed profound in tolerant hosts. Nonetheless, the challenge of tolerant mixed chimeras with 5 million mononuclear leukocytes (MNL) from naive syngeneic mice was neither able to abolish the stable MC nor to trigger skin allograft rejection, a hallmark of peripheral, not central tolerance. Furthermore, in an adoptive transfer model, MNLs harvested from tolerant hosts significantly inhibited the capacity of naive MNLs to reject same donor, but not third-party, skin allografts. Moreover, when we transplanted skin allografts from stable tolerant chimeras onto syngeneic immune-incompetent mice, graft-infiltrating T cells migrated from the graft site, expanded in the new host, and protected allografts from acute rejection by naive syngeneic MNLs. In this model, both deletional and immunoregulatory mechanisms are active during the induction and/or maintenance of allograft tolerance through creation of MC using a potentially clinically applicable regimen.
Introduction
The induction of mixed chimerism (MC) 2 is a powerful and effective means to achieve transplantation tolerance in rodent models (1, 2, 3). Host conditioning with irradiation or cytotoxic drugs has been used in many protocols that have succeeded in creating a long lasting and stable state of hemopoietic MC (1, 2, 4, 5, 6, 7). The deletion of alloreactive T cell clones has been described as the main mechanism responsible for the induction and maintenance of the mixed chimeric state and indefinite survival of the transplanted allografts (8, 9). Nonetheless, concerns about the safety of cytotoxic drugs and/or irradiation have impeded the widespread clinical application of these strategies for organ transplantation (1, 2, 4, 5, 6, 7).
Costimulation blockade using anti-CD154 and CTLA4Ig treatment, without bone marrow transplantation (BMT), has been shown to prolong allograft survival without inducing tolerance when the most stringent test, fully MHC-mismatched skin grafting, is applied (10, 11, 12, 13). In contrast, the use of this combination of costimulatory blockers along with BMT has proven to be potent in producing MC and skin allograft tolerance in rodent models (6, 14, 15). To induce MC in the complete absence of host conditioning with cytotoxic drugs and/or low-dose total body irradiation (TBI), repetitive or single high-dose infusions of donor bone marrow cells (BMCs) and administration of anti-CD154 mAb have been used with or without CTLA4Ig (14, 16). Combined treatment with high-dose BMT and costimulation blockade can induce MC in 65% of C57BL/6 mice receiving fully allogeneic BMCs (14). However, this result is highly dependent on the specific CTLA4Ig preparation used. 3
Because a rapamycin (RPM)-sensitive pathway appears responsible for the failure of costimulation blockade to produce tolerance in certain murine cardiac and skin allograft models (12, 17), we hypothesized that adjunctive administration of RPM to a costimulation-based therapeutic protocol may provide a noncytotoxic, irradiation-free approach that induces MC and skin allograft tolerance.
In this study, we report that a stable MC state and skin allograft tolerance can be established in 93% of MHC-mismatched recipients by a noncytotoxic, irradiation-free approach using costimulation blockade plus RPM treatment. Moreover, we vigorously test the balance of deletion and regulation mechanisms in this model. We provide evidence that both profound deletional mechanism and CD4+CD25+ regulatory T cell-dependent immunoregulatory network are active in this skin allograft tolerance model through creation of MC, using costimulation blockade plus RPM treatment.
Materials and Methods
Mice
Eight- to 12-wk-old male B10.A (H-2a) and C57BL/6 (H-2b) mice were used as bone marrow and skin allograft donors and recipients. Male DBA/1 (H-2q) mice were used as third-party strain donors. Male C57BL/6J-Rag knockout (KO) mice were used as recipients for adoptive lymphocyte transfer experiments. All animal studies were approved by our Institutional Review Board. All animals were purchased from The Jackson Laboratory and maintained under standard conditions.
BMT protocol
Age- and gender-matched C57BL/6 recipients received injections with 2 x 108 unseparated BMCs from tibial and femoral bones of fully MHC-mismatched B10.A (H-2a) donors via tail vein injection.
Skin transplantation
Skin transplantation was performed 1 day after BMT. Full thickness tail skin (1 x 1 cm in size) from B10.A (donor-specific) or DBA/1 (third-party) mice was transplanted onto the lateral thoracic wall of C57BL/6 recipients and secured with 6.0 Prolene sutures. Skin graft survival was monitored daily. Rejection was defined as complete necrosis of the skin graft. Skin allografts from tolerant hosts were removed from the lateral thoracic wall of recipient mice and grafted onto the flank of C57BL/6J-Rag KO mice.
Treatment protocols
Mouse CTLA4Ig was designed, constructed, and expressed in our laboratory as described previously (18). This CTLA4Ig preparation, which is different from the one used in previous experiments involving high doses of allogeneic bone marrow (14), has nonetheless proved to be effective, even as monotherapy, in several transplant models (12, 19, 20). A hybridoma-producing hamster mAbs against mouse CD154 (MR1, IgG2a, ATCC HB11048) was purchased from American Type Culture Collection. RPM was provided by S. Sehgal (Wyeth-Ayerst Pharmaceuticals). Treatment of BMT and skin allograft recipients was initiated on the day of BMT and continued for 4 wk thereafter. Recipients were treated with a single dose of purified anti- CD154 (0.5 mg i.p. on the day of BMT) and with a single dose of purified CTLA4Ig (0.5 mg i.p. on day 2 after BMT). RPM was given i.p. at a dose of 3 mg/kg bodyweight for the first 7 days, and every other day thereafter for a total treatment period of 4 wk. To deplete CD25+ T cells, recipients were treated with 3 doses of anti-CD25 mAb (PC 61) at days –7, –5, and –3 before BMT and skin transplantation, and the depletion determined by FACS analysis on the day of BMT was >99% (21). For comparison purposes, in some mice MC was induced by administering 1.5–2.5 x 108 BMC, anti-CD154 (0.5 mg on day 0), depleting anti-CD8 mAb 2.43 (1.44 mg on day –1), and nonmyeloablative (3 Gy) TBI as described previously (22).
Analysis of multilineage chimerism
The presence of donor hemopoietic cell lineages in the peripheral blood of BMT recipients was serially and quantitatively evaluated using lineage specific surface markers and CellQuest software via a FACScan flow cytometer (BD Biosciences). Two-color flow cytometry was used to distinguish donor and host mononuclear leukocytes (MNLs) present in the circulating blood of the recipient mice. The percentage of donor cells circulating in host peripheral blood was calculated as described previously (23). FITC-conjugated or biotinylated mAbs directed against H-2d (clone, 34-2-12) and H-2b (clone, KH95) and mouse IgG2a and IgG2b; PE-conjugated mAbs against CD3, B220, MAC1, and anti rat-IgG2a were used. All Abs were obtained from BD Pharmingen. To block non-Ag-specific FcR-related binding of Abs to peripheral MNLs, lymphoid cells were preincubated with a mAb against the mouse FcR (clone, 2.4G2; BD Pharmingen).
Analysis of TCR V families
PBLs were obtained at different time points after transplantation from C57BL/6 skin allograft recipients treated with B10.A BMT, RPM, and costimulation blockade. Samples were stained with specific fluorochrome-conjugated Abs against CD4, V5.1/5.2, V8, and V11, or control Abs, and the proportion of CD4 T cells expressing each V was determined. As controls, the same analyses were performed in naive B10.A and C57BL/6 mice. B10.A mice express I-E, which is required to present superantigens derived from endogenous retroviruses encoded in the B10.A genome. Thus, thymocytes expressing T cell receptors containing V5.1/5.2 or V11, which can bind to these superantigens, are deleted in I-E-positive B10.A mice (or in chimeric C57BL/6 receiving B10.A BMT), but not in mice lacking I-E expression such as naive C57BL/6. In contrast, the percentage of T cells expressing irrelevant TCR V families, such as V8, should be similar in both chimeric and naive C57BL/6 hosts (5, 6, 8, 9, 14, 16).
Magnetic cell separation and adoptive cell transfer
Single-cell suspensions of splenic and lymph node MNLs from skin-tolerant mixed chimeras or naive C57BL/6 mice were prepared, and RBC were lysed by hypotonic shock. Magnetic beads coated with mAbs (Dynal Biotech) were used to separate tolerant MNLs into CD4+, CD4–, CD8+, CD8–, and CD4+CD25– subsets by exposure of bead/cell mixtures to a magnetic field as described previously (24). The purity of the resultant populations was determined by flow cytometry, and was >95% in all experiments. Various mixtures of MNLs or subpopulations thereof from C57BL/6 naive and/or tolerant mixed chimeric hosts were adoptively transferred into C57BL/6J-Rag KO hosts via tail vein injection. Mice were then transplanted with B10.A or third-party strain (DBA/1) skin allografts. For all adoptive transfer experiments, tolerant MC C57BL/6 mice were used at >90 days post-BMT and skin grafting.
CFSE labeling and in vivo quantification
Single-cell suspensions were prepared from pooled spleen and lymph node of naive C57BL/6 mice and labeled with the tracking dye CFSE (Molecular Probes) as described previously (25). B10.A mice were irradiated with 1000 RAD using a Gammacell irradiator (Nordion, Kanata, Ontario, Canada) and then received 7 x 107 CFSE-labeled C57BL/6 MNLs by tail vein injection. These hosts were treated with the following: 1) anti-CD154 mAb and CTLA4Ig; or 2) combined RPM, anti-CD154 mAb, and CTLA4Ig. The anti-CD154 mAb was given at a single dose of 0.5 mg i.p. on the day of adoptive transfer. The CTLA4Ig was given at a single dose of 0.5 mg i.p. on day 1 after the adoptive transfer. RPM was given i.p. at a dose of 3 mg/kg body weight for 3 days. The B10.A hosts were sacrificed 3 days following the adoptive transfer. The harvested splenic and lymph node MNLs were stained with PE-conjugated anti-CD4 and anti-CD8 (clone, GK1.5 and 53-6.7, respectively; BD Pharmingen). The frequencies of CFSE-labeled CD4+ or CD8+ T cells proliferating in response to alloantigen in vivo were analyzed by flow cytometry as reported previously (25).
Statistical analysis
Skin graft survival was analyzed using the Kaplan-Meier method; for comparisons, the log rank test was used. Continuous variables were compared using the Student’s t test. For all analyses, a p value <0.05 was considered significant.
Results
Combined costimulation blockade and RPM treatment enables the induction of stable MC and skin allograft tolerance
All C57BL/6 recipients that did not receive BMT, but were treated with RPM alone, costimulation blockade, or combined costimulation blockade with RPM, rejected their skin allograft at median survival time (MST) of 14, 37, and 67 days, respectively (Fig. 1a).
All four C57BL/6 recipients of 2 x 108 B10.A BMT alone failed to develop MC (data not shown) and rejected their skin allografts at a MST of 13 days (Fig. 1b). All six recipients treated with anti-CD154 and CTLA4Ig 3 failed to develop MC (data not shown) and rejected their skin allografts after a MST of 16.5 days (Fig. 1b). Recipient mice treated with RPM alone developed transient low-level MC for 5 wk (data not shown), but rejected their skin allografts (MST, 73 days). Prolonged engraftment was noted with RPM monotherapy treatment as compared with the duration of engraftment in recipients treated with anti-CD154 and CTLA4Ig costimulation blockade (p = 0.0005) (Fig. 1b). In contrast, 14 of 15 recipients (93%) treated with combined costimulation blockade and RPM developed stable multilineage mixed hemopoietic chimerism (Fig. 2) and accepted donor-specific skin allografts (Fig. 1b) throughout the follow-up period of 30 wk. Donor strain-specific skin allograft tolerance was demonstrated by placing the second B10.A or third-party skin allografts 90 days after BMT and skin transplantation. Although the second B10.A skin grafts were accepted without further immunosuppression (n = 7), third-party (DBA/1) skin grafts were promptly rejected (MST, 17 days; n = 7; data not shown). These data indicate that BMT synergizes with combined costimulation blockade and RPM treatment to induce allograft tolerance.
In chimeric recipients of combined costimulation blockade and RPM, MC persisted throughout the follow-up period of 30 wk (Fig. 2). In mice with stable multilineage MC, donor cells (B10.A) contributed up to 7.7% of the circulating white blood cells 21 wk after BMT (see Fig. 5). The proportion of donor CD3+ T cells among peripheral T cells remained stable at 2–3% in the circulating pool of BMT recipient mice (Fig. 2). The percentage of donor-derived B cells and MAC1-positive cells present in the peripheral blood of BMT recipients remained stable in the range of 5–9% (Fig. 2). The persistence of significant multilineage donor hematopoiesis for up to 30 wk after BMT demonstrates that engraftment of donor hemopoietic progenitor cells occurred (26).
Combined treatment with costimulation blockade and RPM inhibits the proliferation of allo-Ag-triggered CD4+ and CD8+ T cells
To evaluate the effect of a therapeutic regimen consisting of costimulation blockade and RPM on allo-Ag-triggered T cell immune response, we used a graft-vs-host disease-like model in which CFSE-labeled C57BL/6 (H-2b) MNLs were transferred into lethally irradiated allogeneic B10.A (H-2a) hosts. Three days after the adoptive transfer, 42% fewer CFSE-labeled C57BL/6 T cells were recovered from B10.A recipients treated with costimulation blockade and RPM as compared with untreated recipients (data not shown). Costimulation blockade and RPM administered alone were associated with mean reductions of 16 and 21% in the proportion of cells that had undergone proliferation, respectively, whereas the combination of both treatments led to an average 35% reduction. Moreover, the proportions of CFSE-labeled CD4+ and CD8+ T cells entering the cell cycle was reduced by 35% and 34%, respectively, in recipients of combined therapy compared with untreated recipients (Fig. 3). Thus, the combination of costimulation blockade and RPM had additive effects that were evident in the inhibition of CD4+ T cell proliferation, as compared with either treatment alone. In contrast, maximal inhibition of CD8+ T cell proliferation was achieved with RPM alone, and the addition of costimulation blockade did not enhance this effect. These results were consistent with our previous report using a different MHC-mismatched strain combination (12). Fig. 3 represents five independent experiments.
Administration of BMT, RPM, and costimulation blockade leads to deletion of T cell clones expressing specific TCR V families
To study whether clonal deletion was involved in the tolerizing effect promoted by combined BMT, RPM, and costimulation blockade treatment, we serially analyzed PBLs obtained from treated skin allograft recipients, or from control C57BL/6 and B10.A naive mice. Two weeks after BMT, there was already a marked reduction of CD4+ T cells expressing V5.1/5.2+ and V11+ (Fig. 4), as compared with naive C57BL/6 controls. In contrast, there was no reduction in the proportion of irrelevant V8+CD4+ T cells. Throughout a follow-up of >16 wk, the deletion of V5.1/5.2+ and V11+CD4+ T cells was further amplified (Fig. 4). These data suggest that clonal deletion of donor-reactive T cells is a mechanism implicated in the induction of allograft tolerance via combined BMT, RPM, and costimulation blockade treatment.
Challenge of tolerant chimeras with 5 x 106 MNLs from naive syngeneic donors does not abolish the stable mixed chimeric state or trigger skin allograft rejection
To test whether immunoregulatory mechanisms may be present in the maintenance phase of a stable mixed chimeric state and skin allograft tolerance, we challenged three skin allograft-tolerant MC hosts (>120 days post BMT) with 5 x 106 splenic MNLs from naive syngeneic mice. Following the challenge with 5 x 106 naive syngeneic MNLs, the proportion of donor cells present in peripheral blood remained stable, albeit with slightly, but not statistically significant, decrease throughout an 8-wk assessment period (Fig. 5), and donor skin allografts were retained. These results suggest that immunoregulatory mechanisms are involved in the maintenance of the stable mixed chimeric state and skin allograft tolerance (27).
Donor-specific immunoregulatory CD4+ T cells can be detected in secondary lymph organs of tolerant mixed chimeras
To further analyze the nature of the host antidonor response during the maintenance phase of tolerance in MC hosts, we used a passive transfer model using male RAG1-deficient, C57BL/6J-Rag KO mice (H-2b). These mice lack the RAG1, which is critical for the generation of a functioning TCR (they have no CD3+, or TCR--positive cells); thus, RAG1-deficient recipients are unable to elicit a sufficient immune response to reject allografts. These mice are missing CD4+ and CD8+ T cells and B cells; therefore, no CD4+ T cells can be detected in PBLs of naive RAG1-deficient mice.
As shown in Fig. 6a, adoptive transfer of as few as 0.5 x 106 MNLs harvested from naive C57BL/6 mice into immunoincompetent syngeneic Rag KO recipients triggered rapid rejection of B10.A skin allografts (MST of 15 days). In contrast, adoptive transfer of 0.5 x 106 (data not shown) or 10 x 106 MNLs from tolerant MC skin allograft recipients obtained >120 days after BMT into syngeneic Rag KO mice receiving a fresh B10.A fresh skin allograft did not elicit allograft rejection (Fig. 6a). To explore the possibility that regulatory cells might be present in tolerant long-term mixed chimeras, we cotransferred 0.5 x 106 naive C57BL/6 MNLs with a relative excess (10 x 106) of MNLs harvested from long-term C57BL/6 chimeras (120 days after BMT) into syngeneic Rag KO mice receiving a B10.A skin allograft. As shown in Fig. 6a, C57BL/6 MNLs harvested from tolerant hosts significantly inhibited the capacity of naive syngeneic MNLs to reject B10.A skin allografts (MST, 62 days; p = 0.0027). Moreover, these regulatory effects were donor Ag-specific, because the immunosuppressive effect was abolished when comixed MNLs from naive and tolerant mice were passively transferred into syngeneic recipients of third-party DBA/1 strain allografts (MST, 10.5 days).
To characterize the T cell subsets responsible for this suppressive effect, we repeated the cell transfer experiments using distinct subpopulations of tolerant cells. As shown in Fig. 6b, passive cotransfer of 5 x 106 tolerant CD4+ T cells, but not 5 x 106 CD8+ T cells or CD4+CD8+-depleted cells, with 0.5 x 106 naive syngeneic MNLs resulted in a significant delay of skin allograft rejection (Fig. 6b). Moreover, when CD25+ T cells were selectively deleted from tolerant CD4+ T cells, the cotransfer of 5 x 105 CD4+CD25– T cells from tolerant host with 0.5 x 106 naive MNLs resulted in acute allograft rejection of B10.A skin grafts (MST, 17 days; p = 0.0027). Thus, the immunoregulatory effects observed in this adoptive transfer model of skin transplantation are CD4+CD25+ T cell dependent (Fig. 6c).
To determine whether immunoregulatory T cells were also present in tolerant MC transplant recipients prepared with treatment of TBI, depleting anti-CD8 mAb, and anti-CD154 other than RPM plus costimulation blockade regimen, we repeated the adoptive transfer experiments, administering 0.5 x 106 naive MNLs together with 10 x 106 MNLs (naive:tolerant at 1:20 ratio) harvested from tolerant MC hosts treated with anti-CD154, depleting anti-CD8, and 3 Gy TBI or RPM plus costimulation blockade. Inconsistent with results reported previously (22), no immunoregulation could be detected in the adoptive transfer experiments using the mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, and naive hosts at 20:1 ratio (tolerant:naive; Fig. 6d). These results also suggest that the immunoregulatory effects observed in the transfer model using MNLs from tolerant RPM and costimulation blockade-treated MC hosts are not due to inhibition of homeostatic T cell proliferation.
To vigorously test whether there is immunoregulation, if any, present in the tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, we repeated the adoptive transfer experiment using mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment and naive hosts at 20:1, 30:1, and 40:1 ratio (tolerant:naive; Fig. 6e). Indeed, slight but significant inhibition of allografts rejection was evident when C57BL/6J-Rag KO skin graft recipients received mixture of MNLs from tolerant chimera and naive hosts at 40:1 ratio (tolerant:naive; Fig. 6e).
Functionally active immunoregulatory T cells reside in donor skin allografts from tolerant mixed chimeras treated with RPM and costimulation blockade
Because these experiments demonstrated that immunoregulatory CD4+ T cells were present in the secondary lymphoid organs of tolerant mixed chimeras (Fig. 6, a–c), we next investigated whether regulatory cells home to tolerated B10.A skin allografts and are active in protecting the allografts from rejection. B10.A skin allografts (n = 4) were obtained from tolerant C57BL/6 chimeras prepared with RPM and costimulatory blockade (>120 days after BMT and skin transplantation). As a control, B10.A skin autografts (n = 3) were harvested from B10.A recipients (>120 posttransplantation). Tolerated B10.A allografts or control B10.A autografts were then transplanted onto C57BL/6-Rag KO hosts. To determine whether graft-infiltrating T cells from tolerant skin allografts migrate and expand in the peripheral blood of Rag KO recipients, PBL samples were collected 30 days post skin transplantation. Because Rag KO mice are missing CD4+ and CD8+ T cells and B cells, no CD4+ T cells can be detected in PBL of naive RAG1-deficient mice. As depicted in Fig. 7a, CD4+ T cells could be detected readily in PBL of C57BL/6-Rag KO recipients of tolerated allografts, which otherwise lack T and B cells, but not in recipients of syngeneic grafts 30 days after graft transfer (6.78 ± 2.037% vs 0.98 ± 0.456%; p = 0.033). Because in an adoptive transfer model of skin allografts in C57BL/6-Rag KO recipients the CD4+ T cells, but not CD8+ T cells or B220+ B cells, from tolerant chimeras inhibit the ability of naive MNL to reject same donor strain allograft (Fig. 6b), we focus on the homing of CD4+ T cells in the tolerant grafts in this study. It would be of interest to test other types of cells, such as CD8+ T cells and B cells, as well in future studies.
To further test the immunoregulatory function of these graft-homing T cells from tolerant chimeras, 0.5 x 106 MNLs from naive C57BL/6 mice were adoptively transferred into the C57BL/6-Rag KO hosts bearing the tolerated B10.A allografts or B10.A autografts 30 days following skin transplantation. Following a challenge with 0.5 x 106 MNLs from naive C57BL/6 mice, survival of tolerated B10.A skin allografts transplanted onto C57BL/6-Rag KO mice was significantly prolonged (MST, 48 days) in comparison with the survival of B10.A skin autografts (MST, 16 days) (Fig. 7b; p = 0.01).
These data clearly indicate that the regulatory T cells reside within tolerated B10.A skin allografts of C57BL/6 chimeras and are functionally active in protecting the allografts from rejection in an adoptive transferred model of skin allografts.
CD4+CD25+-dependent immunoregulation is not required for tolerance induction after administration of BMT, RPM, and costimulation blockade
The thymus-derived CD4+CD25+ T cells, identified as suppressor cells in 1990 (28, 29), have emerged as critical effectors in both the control of autoimmunity (30, 31, 32) and the maintenance of peripheral allograft tolerance (24, 33, 34). In several allograft models in which tolerance is induced by either donor-specific blood transfusion (DST) plus anti-CD154 (24) or combined RPM plus lytic IL-2/Fc and antagonist mutant IL-15/Fc (35), the deletion of CD4+CD25+ Treg cells before transplantation results in acute allograft rejection, indicating that CD4+CD25+-dependent regulatory networks are critical for the induction of tolerance. Therefore, we tested the role of CD4+CD25+-dependent regulatory networks in the induction of MC and tolerance in hosts treated with combined BMT, RPM, and costimulation blockade. The CD4+CD25+ Treg cells were deleted in the recipient mice before BMT, and the effect of the treatment with PC61 mAb was evaluated by staining for CD25 in peripheral blood and analysis by flow cytometry.
Interestingly, the deletion of CD4+CD25+ Treg cells before BMT and skin transplantation did not interfere with the establishment of stable MC and skin allograft tolerance in the recipients treated with BMT, RPM, and costimulation blockade (data not shown). Thus, CD4+CD25+-dependent regulatory networks are not required for the tolerance induction in this model.
CD4+CD25– T cells from tolerant chimeras do not reject donor strain skin allografts after adoptive transfer
To determine the effects of BMT plus costimulation blockade and RPM treatment on CD4+CD25– effector T cells, we conducted additional adoptive transfer experiments in which we compared the capacity of CD25– MNLs harvested from naive or from tolerant chimeras to mediate rejection. The adoptive transfer of 10 x 106 CD25– MNLs from tolerant MC C57BL/6 hosts into C57BL/6J-Rag KO recipients did not trigger the rejection of same donor strain B10.A skin allografts throughout a follow-up period of 100 days. In contrast, naive MNLs selectively depleted of the CD25+ T cell subset were able to elicit acute rejection of B10.A skin allografts with a MST of 12.5 days (Fig. 8). Moreover, the adoptive transfer of 10 x 106 CD25– MNLs from tolerant chimeras resulted in acute rejection of third-party strain DBA/1 allografts at a MST of 10 days (Fig. 8), suggesting that profound selective deletion of anti-B10.A effector CD25– T cells in the tolerant chimeras had taken place.
Discussion
In this study, we demonstrate that combined treatment with costimulation blockade and RPM, in the absence of any host preconditioning, enables the induction of stable MC and skin allograft tolerance in fully MHC-mismatched recipients (B10.A into C57BL/6) (Figs. 1 and 2). In skin-tolerant MCs, the percentage of donor cells present among circulating MNLs was stable throughout the follow-up period of 30 wk. In parallel with the development of stable multilineage mixed hemopoietic chimerism, donor-specific skin allograft tolerance was achieved (Fig. 1b). In this experimental setting, combined high-dose BMT with costimulation blockade treatment was unable to induce stable MC and prevent acute skin allograft rejection (Fig. 1b). The difference from previous experiments (14) most likely reflects differences in the CTLA4Ig preparations used (see footnote).
Because a RPM-sensitive mechanism is responsible for the failure of costimulation blockade to create peripheral tolerance in many models (12, 13, 17), and RPM does not inhibit the development of chimerism or tolerance induced by TBI, BMT, and costimulation blockade (36), we studied the effect of RPM as an adjunct to costimulation blockade in our attempt to induce stable MC and allograft tolerance. Indeed, the adjunctive administration of RPM to this costimulation blockade-based protocol provides a noncytotoxic, irradiation- free approach that induces MC and skin allograft tolerance.
There were numerous studies focused on the deletion mechanism in MC induced by both lymphoablative conditioning or nonlyphoablative regimens. In hosts in which MC and transplant tolerance is induced with regimens using lymphoablative conditioning (1, 2, 4, 5, 6, 7), ongoing alloreactive T cell deletion occurring in the thymus is the dominant, if not exclusive, mechanism responsible for the induction and maintenance of organ allograft tolerance (8, 9, 37). In contrast, in regimens using costimulatory blockade instead of peripheral T cell depletion to induce MC and allograft tolerance, specific peripheral deletion of donor-reactive T cells without global depletion takes place (6, 14). These reported mechanisms (i.e., intrathymic and peripheral deletion of donor-reactive T cells after BMT) appear to differ from those noted in models of peripheral tolerance, in which permanent engraftment of organ allografts is achieved via costimulation blockade without BMT. In these latter models, although a substantial depletion of alloreactive T cell clones is required for tolerance induction, the long-term maintenance of the tolerant state is dependent upon self-perpetuating immunoregulatory networks that constrain the remaining cytopathic T cells (11, 12, 21, 24, 27, 38, 39, 40). However, the role of regulatory mechanism in MC has not been fully investigated. According to the pool size model of allograft tolerance, the allograft outcome, rejection or tolerance, often depends on the balance between cytopathic and regulatory T cells. Although both deletion and regulation play important roles in allograft tolerance, recent studies showed that the quantitative details for each mechanism differ from model to model. Therefore, we hypothesize that there is a delicate balance between deletion and regulation in allograft tolerance. In the current study, we sought to vigorously test the balance of deletional and regulatory mechanisms in this model.
Our data suggest that deletion of donor-reactive cytopathic T cell clones is indeed profound in tolerant chimeras receiving BMT plus a nonlymphoablative, irradiation-free protocol. This is supported by the failure of CD4+CD25– T cells from tolerant chimeras to reject same donor, but not third-party, strain allografts upon adoptive transfer into syngeneic recipients (Fig. 8). In addition, using V5, V11, and V8 TCR probes as reported previously (5, 6, 8, 9, 14, 16), we have demonstrated specific deletion of donor superantigen-reactive CD4 T cell clones in skin allograft-tolerant MC hosts (Fig. 4). Taken together, these data strongly support the notion that deletional mechanisms play an important role in this model of tolerance. This is sharply contrasted with the allograft tolerance model induced by DST plus MR1, in which the capacity of CD4+CD25– effector T cells from the tolerant hosts to trigger skin allograft rejection on a per cell-to-cell basis is identical with that from naive mice, suggesting that the deletion is not nearly complete in this model (24).
In addition, in the current study, we sought to determine whether immunoregulatory networks were also present in allograft-tolerant MC hosts receiving costimulation blockade and RPM treatment, and if so, to ascertain the precise balance between immunoregulatory and deletional mechanisms in the acquisition of allograft tolerance with this therapy.
Indeed, several lines of evidence indicate that immunoregulatory mechanisms are also present in the maintenance phase of tolerance in mixed chimeric hosts treated with RPM and costimulation blockade. First, the challenge of tolerant MC skin allograft recipients with 5 x 106 MNLs from naive syngeneic mice was neither able to abolish the mixed chimeric state (Fig. 5) nor to trigger skin allograft rejection (data not shown), a hallmark of peripheral, not central tolerance. Furthermore, in an adoptive transfer model, MNLs harvested from the secondary lymphoid organs of RPM and costimulation blockade-treated MC skin-tolerant hosts significantly inhibited the capacity of naive MNLs to reject B10.A skin allografts (Fig. 6a). These T cell-dependent regulatory effects were donor specific and dependent on the presence of CD4+CD25+ T cells among the tolerant MNLs. This was supported by the observation that selective depletion of CD4+CD25+ regulatory T cells abolished the ability of tolerant MNLs cells to prevent naive MNLs from rejecting allografts (Fig. 6c), and is consistent with the results obtained in costimulation blockade-based models of peripheral tolerance (21, 24, 34). Finally, when we transplanted skin allografts from stable tolerant chimeras onto syngeneic immune-incompetent mice, graft-infiltrating T cells migrated from the graft site, expanded in the new host, and protected test allografts from acute rejection after transfer of naive syngeneic MNLs (Fig. 7). A similar finding has been reported recently by Graca et al. (41) in a model of peripheral tolerance to minor histocompatibility Ag-mismatched skin allografts. Thus, in this model in which skin allograft tolerance is achieved via BMT, RPM, and costimulation blockade, regulatory T cells are present in both secondary lymphoid organs and in the allograft itself, and are functionally active in protecting the grafts from rejection (Figs. 6 and 7).
In short, the use of BMT, RPM, and costimulation blockade treatment to induce transplantation tolerance is associated with both deletional and active immunoregulatory phenomena. However, these results are in contrast to those obtained in mixed chimeras in which transplantation tolerance was established by BMT, anti-CD154, depleting anti-CD8, and low-dose TBI, because adoptive transfer of 10 x 106 MNLs obtained from tolerant hosts treated with RPM and combined costimulation blockade, but not MNLs harvested from anti-CD8, anti-CD154, and TBI-treated tolerant MC hosts, inhibit the capacity of 0.5 x 105 naive MNLs to reject B10.A skin allografts (Fig. 6d) (22). A noteworthy difference between these strategies, using TBI and anti-CD8 mAb treatment and our current treatment, is the lower levels of multilineage chimerism achieved with combined RPM and costimulation blockade treatment (9, 22, 42). To vigorously test whether there is immunoregulation, if any, present in the tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, we repeated the adoptive transfer experiment using mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, and naive hosts at 20:1, 30:1, and 40:1 ratio (tolerant:naive; Fig. 6e). Indeed, a slight but significant inhibition of allograft rejection was evident only when C57BL/6J-Rag KO skin graft recipients received a mixture of MNLs from tolerant chimera and naive hosts at 40:1 ratio (tolerant:na?ve; Fig. 6e). CD4+CD25+ T cells are known to be prone to thymic deletion (43), although their specific susceptibility to deletion, as compared with nonregulatory T cell populations, has not been extensively addressed. We speculate that conventional alloreactive, but not regulatory, T cells are depleted in hosts with modest levels of hemopoietic MC, whereas both conventional and regulatory donor-reactive T cells are centrally eliminated in hosts with more robust MC. A recent report indirectly supports this hypothesis (44). Alternatively, the presence of different levels of immunoregulation in MC models with different levels of hemopoietic MC might relate to the persistence of residual donor-reactive T cells, which may be required to maintain effective immunoregulatory networks. Hence, with more complete long-term donor-reactive CD4+ and CD8+ T cell deletion, regulation would be less necessary to maintain tolerance.
Is immunoregulation truly necessary to achieve tolerance in this costimulation blockade-based model of MC? Unlike protocols inducing "pure" peripheral tolerance (24, 35), the depletion of CD4+CD25+ T cells before BMT did not preclude the induction of tolerance in our model (data not shown). Hence, our data suggest that in MC, CD4+CD25+ T cell-dependent immunoregulation is not as important for the induction and maintenance of transplantation tolerance as seen in more conventional models of peripheral tolerance (24, 35, 45). These results further support our clone-size hypothesis that the outcome of the allograft response, rejection or tolerance, depends on the balance between cytopathic and regulatory T cells (46, 47). Although both deletion and regulation play important roles in allograft tolerance, our recent studies showed that the quantitative details for each mechanism differ from model to model. Therefore, we hypothesize that there is a delicate balance between deletion and regulation in allograft tolerance. As in a model of allograft tolerance in which the deletional mechanisms play a dominant role (e.g., tolerance produced via creation of mixed chimeras), the regulatory mechanisms, albeit sometimes present, are far less important (6, 8, 37). Although, in a model in which the regulation mechanisms play critical role (e.g., DST plus MR1-induced allograft tolerance), deletional mechanisms lower the threshold for effective regulatory T cell action (Fig. 9) (24, 35, 45).
In short, the presence of both T cell depletion and immunoregulation has been noted in a model in which skin allograft tolerance is achieved through the induction of MC. Hence, the classic definitions of central and peripheral tolerance may be somewhat imprecise. Altogether, tolerance can be achieved through subtly differing mixtures of clonal deletion and immunoregulation.
Acknowledgments
We thank Yan Tian for excellent technical assistance, and Dr. Thomas Fehr and Raquel Oliviera for providing bone marrow chimeras for some skin transfer experiments.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 Address correspondence and reprint requests to Dr. Xin Xiao Zheng and Dr. Terry B. Strom, Transplantation Research Center, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215; E-mail address: xzheng@bidmc.harvard.edu and tstrom@bidmc.harvard.edu
2 Abbreviations used in this paper: MC, mixed chimerism; BMT, bone marrow transplantation; TBI, total body irradiation; BMC, bone marrow cell; RPM, rapamycin; KO, knockout; MNL, mononuclear leukocyte; MST, median survival time; DST, donor-specific blood transfusion.
3 When a CTLA4lg preparation different than the one employed by Wekerle et al. (14 ) was used to generate tolerant MC host together with BMT, anti-CD154, and 3 Gy TBI, we observed that CD8 cells could no longer be rendered tolerant (48 ). In constrast, long-term MC and long-term systemic donor-specific tolerance of both CD4+ and CD8+ T cells (48 ) could be reliably induced in mice receiving anti-CD8 depleting-mAb and a single injection of anti-CD154 with 3 Gy TBI (48 ).
Received for publication June 9, 2004. Accepted for publication March 25, 2005.
References
Ildstad, S. T., D. H. Sachs. 1984. Reconstitution with syngeneic plus allogeneic or xenogeneic bone marrow leads to specific acceptance of allografts or xenografts. Nature 307: 168-170.
Sharabi, Y., D. H. Sachs. 1989. Mixed chimerism and permanent specific transplantation tolerance induced by a nonlethal preparative regimen. J. Exp. Med. 169: 493-502.
Colson, Y. L., H. Li, S. S. Boggs, K. D. Patrene, P. C. Johnson, S. T. Ildstad. 1996. Durable mixed allogeneic chimerism and tolerance by a nonlethal radiation-based cytoreductive approach. J. Immunol. 157: 2820-2829.
Nomoto, K., K. Yung-Yun, K. Omoto, M. Umesue, Y. Murakami, G. Matsuzaki. 1995. Tolerance induction in a fully allogeneic combination using anti-T cell receptor- monoclonal antibody, low dose irradiation, and donor bone marrow transfusion. Transplantation 59: 395-401.
Sykes, M., G. L. Szot, K. A. Swenson, D. A. Pearson. 1997. Induction of high levels of allogeneic hematopoietic reconstitution and donor-specific tolerance without myelosuppressive conditioning. Nat. Med. 3: 783-787.
Wekerle, T., M. H. Sayegh, J. Hill, Y. Zhao, A. Chandraker, K. G. Swenson, G. Zhao, M. Sykes. 1998. Extrathymic T cell deletion and allogeneic stem cell engraftment induced with costimulatory blockade is followed by central T cell tolerance. J. Exp. Med. 187: 2037-2044.
Sharabi, Y., I. Aksentijevich, T. M. Sundt, 3rd, D. H. Sachs, M. Sykes. 1990. Specific tolerance induction across a xenogeneic barrier: production of mixed rat/mouse lymphohematopoietic chimeras using a nonlethal preparative regimen. J. Exp. Med. 172: 195-202.
Tomita, Y., A. Khan, M. Sykes. 1994. Role of intrathymic clonal deletion and peripheral anergy in transplantation tolerance induced by bone marrow transplantation in mice conditioned with a nonmyeloablative regimen. J. Immunol. 153: 1087-1098.(Christoph Domenig*, Alber)
The induction of mixed chimerism (MC) is a powerful and effective means to achieve transplantation tolerance in rodent models. Host conditioning with irradiation or cytotoxic drugs has been used in many protocols for chimeric induction across allogeneic barriers. The deletion of alloreactive T cell clones has been described as the main mechanism responsible for the induction of a stable MC. In this study, we demonstrate that a stable MC and skin allograft tolerance can be established across MHC barriers by a noncytotoxic, irradiation-free approach using costimulation blockade plus rapamycin treatment. By using an adoptive transfer model of skin allograft and using specific V TCR probes, we demonstrated that deletion of donor-reactive cytopathic T cell clones is indeed profound in tolerant hosts. Nonetheless, the challenge of tolerant mixed chimeras with 5 million mononuclear leukocytes (MNL) from naive syngeneic mice was neither able to abolish the stable MC nor to trigger skin allograft rejection, a hallmark of peripheral, not central tolerance. Furthermore, in an adoptive transfer model, MNLs harvested from tolerant hosts significantly inhibited the capacity of naive MNLs to reject same donor, but not third-party, skin allografts. Moreover, when we transplanted skin allografts from stable tolerant chimeras onto syngeneic immune-incompetent mice, graft-infiltrating T cells migrated from the graft site, expanded in the new host, and protected allografts from acute rejection by naive syngeneic MNLs. In this model, both deletional and immunoregulatory mechanisms are active during the induction and/or maintenance of allograft tolerance through creation of MC using a potentially clinically applicable regimen.
Introduction
The induction of mixed chimerism (MC) 2 is a powerful and effective means to achieve transplantation tolerance in rodent models (1, 2, 3). Host conditioning with irradiation or cytotoxic drugs has been used in many protocols that have succeeded in creating a long lasting and stable state of hemopoietic MC (1, 2, 4, 5, 6, 7). The deletion of alloreactive T cell clones has been described as the main mechanism responsible for the induction and maintenance of the mixed chimeric state and indefinite survival of the transplanted allografts (8, 9). Nonetheless, concerns about the safety of cytotoxic drugs and/or irradiation have impeded the widespread clinical application of these strategies for organ transplantation (1, 2, 4, 5, 6, 7).
Costimulation blockade using anti-CD154 and CTLA4Ig treatment, without bone marrow transplantation (BMT), has been shown to prolong allograft survival without inducing tolerance when the most stringent test, fully MHC-mismatched skin grafting, is applied (10, 11, 12, 13). In contrast, the use of this combination of costimulatory blockers along with BMT has proven to be potent in producing MC and skin allograft tolerance in rodent models (6, 14, 15). To induce MC in the complete absence of host conditioning with cytotoxic drugs and/or low-dose total body irradiation (TBI), repetitive or single high-dose infusions of donor bone marrow cells (BMCs) and administration of anti-CD154 mAb have been used with or without CTLA4Ig (14, 16). Combined treatment with high-dose BMT and costimulation blockade can induce MC in 65% of C57BL/6 mice receiving fully allogeneic BMCs (14). However, this result is highly dependent on the specific CTLA4Ig preparation used. 3
Because a rapamycin (RPM)-sensitive pathway appears responsible for the failure of costimulation blockade to produce tolerance in certain murine cardiac and skin allograft models (12, 17), we hypothesized that adjunctive administration of RPM to a costimulation-based therapeutic protocol may provide a noncytotoxic, irradiation-free approach that induces MC and skin allograft tolerance.
In this study, we report that a stable MC state and skin allograft tolerance can be established in 93% of MHC-mismatched recipients by a noncytotoxic, irradiation-free approach using costimulation blockade plus RPM treatment. Moreover, we vigorously test the balance of deletion and regulation mechanisms in this model. We provide evidence that both profound deletional mechanism and CD4+CD25+ regulatory T cell-dependent immunoregulatory network are active in this skin allograft tolerance model through creation of MC, using costimulation blockade plus RPM treatment.
Materials and Methods
Mice
Eight- to 12-wk-old male B10.A (H-2a) and C57BL/6 (H-2b) mice were used as bone marrow and skin allograft donors and recipients. Male DBA/1 (H-2q) mice were used as third-party strain donors. Male C57BL/6J-Rag knockout (KO) mice were used as recipients for adoptive lymphocyte transfer experiments. All animal studies were approved by our Institutional Review Board. All animals were purchased from The Jackson Laboratory and maintained under standard conditions.
BMT protocol
Age- and gender-matched C57BL/6 recipients received injections with 2 x 108 unseparated BMCs from tibial and femoral bones of fully MHC-mismatched B10.A (H-2a) donors via tail vein injection.
Skin transplantation
Skin transplantation was performed 1 day after BMT. Full thickness tail skin (1 x 1 cm in size) from B10.A (donor-specific) or DBA/1 (third-party) mice was transplanted onto the lateral thoracic wall of C57BL/6 recipients and secured with 6.0 Prolene sutures. Skin graft survival was monitored daily. Rejection was defined as complete necrosis of the skin graft. Skin allografts from tolerant hosts were removed from the lateral thoracic wall of recipient mice and grafted onto the flank of C57BL/6J-Rag KO mice.
Treatment protocols
Mouse CTLA4Ig was designed, constructed, and expressed in our laboratory as described previously (18). This CTLA4Ig preparation, which is different from the one used in previous experiments involving high doses of allogeneic bone marrow (14), has nonetheless proved to be effective, even as monotherapy, in several transplant models (12, 19, 20). A hybridoma-producing hamster mAbs against mouse CD154 (MR1, IgG2a, ATCC HB11048) was purchased from American Type Culture Collection. RPM was provided by S. Sehgal (Wyeth-Ayerst Pharmaceuticals). Treatment of BMT and skin allograft recipients was initiated on the day of BMT and continued for 4 wk thereafter. Recipients were treated with a single dose of purified anti- CD154 (0.5 mg i.p. on the day of BMT) and with a single dose of purified CTLA4Ig (0.5 mg i.p. on day 2 after BMT). RPM was given i.p. at a dose of 3 mg/kg bodyweight for the first 7 days, and every other day thereafter for a total treatment period of 4 wk. To deplete CD25+ T cells, recipients were treated with 3 doses of anti-CD25 mAb (PC 61) at days –7, –5, and –3 before BMT and skin transplantation, and the depletion determined by FACS analysis on the day of BMT was >99% (21). For comparison purposes, in some mice MC was induced by administering 1.5–2.5 x 108 BMC, anti-CD154 (0.5 mg on day 0), depleting anti-CD8 mAb 2.43 (1.44 mg on day –1), and nonmyeloablative (3 Gy) TBI as described previously (22).
Analysis of multilineage chimerism
The presence of donor hemopoietic cell lineages in the peripheral blood of BMT recipients was serially and quantitatively evaluated using lineage specific surface markers and CellQuest software via a FACScan flow cytometer (BD Biosciences). Two-color flow cytometry was used to distinguish donor and host mononuclear leukocytes (MNLs) present in the circulating blood of the recipient mice. The percentage of donor cells circulating in host peripheral blood was calculated as described previously (23). FITC-conjugated or biotinylated mAbs directed against H-2d (clone, 34-2-12) and H-2b (clone, KH95) and mouse IgG2a and IgG2b; PE-conjugated mAbs against CD3, B220, MAC1, and anti rat-IgG2a were used. All Abs were obtained from BD Pharmingen. To block non-Ag-specific FcR-related binding of Abs to peripheral MNLs, lymphoid cells were preincubated with a mAb against the mouse FcR (clone, 2.4G2; BD Pharmingen).
Analysis of TCR V families
PBLs were obtained at different time points after transplantation from C57BL/6 skin allograft recipients treated with B10.A BMT, RPM, and costimulation blockade. Samples were stained with specific fluorochrome-conjugated Abs against CD4, V5.1/5.2, V8, and V11, or control Abs, and the proportion of CD4 T cells expressing each V was determined. As controls, the same analyses were performed in naive B10.A and C57BL/6 mice. B10.A mice express I-E, which is required to present superantigens derived from endogenous retroviruses encoded in the B10.A genome. Thus, thymocytes expressing T cell receptors containing V5.1/5.2 or V11, which can bind to these superantigens, are deleted in I-E-positive B10.A mice (or in chimeric C57BL/6 receiving B10.A BMT), but not in mice lacking I-E expression such as naive C57BL/6. In contrast, the percentage of T cells expressing irrelevant TCR V families, such as V8, should be similar in both chimeric and naive C57BL/6 hosts (5, 6, 8, 9, 14, 16).
Magnetic cell separation and adoptive cell transfer
Single-cell suspensions of splenic and lymph node MNLs from skin-tolerant mixed chimeras or naive C57BL/6 mice were prepared, and RBC were lysed by hypotonic shock. Magnetic beads coated with mAbs (Dynal Biotech) were used to separate tolerant MNLs into CD4+, CD4–, CD8+, CD8–, and CD4+CD25– subsets by exposure of bead/cell mixtures to a magnetic field as described previously (24). The purity of the resultant populations was determined by flow cytometry, and was >95% in all experiments. Various mixtures of MNLs or subpopulations thereof from C57BL/6 naive and/or tolerant mixed chimeric hosts were adoptively transferred into C57BL/6J-Rag KO hosts via tail vein injection. Mice were then transplanted with B10.A or third-party strain (DBA/1) skin allografts. For all adoptive transfer experiments, tolerant MC C57BL/6 mice were used at >90 days post-BMT and skin grafting.
CFSE labeling and in vivo quantification
Single-cell suspensions were prepared from pooled spleen and lymph node of naive C57BL/6 mice and labeled with the tracking dye CFSE (Molecular Probes) as described previously (25). B10.A mice were irradiated with 1000 RAD using a Gammacell irradiator (Nordion, Kanata, Ontario, Canada) and then received 7 x 107 CFSE-labeled C57BL/6 MNLs by tail vein injection. These hosts were treated with the following: 1) anti-CD154 mAb and CTLA4Ig; or 2) combined RPM, anti-CD154 mAb, and CTLA4Ig. The anti-CD154 mAb was given at a single dose of 0.5 mg i.p. on the day of adoptive transfer. The CTLA4Ig was given at a single dose of 0.5 mg i.p. on day 1 after the adoptive transfer. RPM was given i.p. at a dose of 3 mg/kg body weight for 3 days. The B10.A hosts were sacrificed 3 days following the adoptive transfer. The harvested splenic and lymph node MNLs were stained with PE-conjugated anti-CD4 and anti-CD8 (clone, GK1.5 and 53-6.7, respectively; BD Pharmingen). The frequencies of CFSE-labeled CD4+ or CD8+ T cells proliferating in response to alloantigen in vivo were analyzed by flow cytometry as reported previously (25).
Statistical analysis
Skin graft survival was analyzed using the Kaplan-Meier method; for comparisons, the log rank test was used. Continuous variables were compared using the Student’s t test. For all analyses, a p value <0.05 was considered significant.
Results
Combined costimulation blockade and RPM treatment enables the induction of stable MC and skin allograft tolerance
All C57BL/6 recipients that did not receive BMT, but were treated with RPM alone, costimulation blockade, or combined costimulation blockade with RPM, rejected their skin allograft at median survival time (MST) of 14, 37, and 67 days, respectively (Fig. 1a).
All four C57BL/6 recipients of 2 x 108 B10.A BMT alone failed to develop MC (data not shown) and rejected their skin allografts at a MST of 13 days (Fig. 1b). All six recipients treated with anti-CD154 and CTLA4Ig 3 failed to develop MC (data not shown) and rejected their skin allografts after a MST of 16.5 days (Fig. 1b). Recipient mice treated with RPM alone developed transient low-level MC for 5 wk (data not shown), but rejected their skin allografts (MST, 73 days). Prolonged engraftment was noted with RPM monotherapy treatment as compared with the duration of engraftment in recipients treated with anti-CD154 and CTLA4Ig costimulation blockade (p = 0.0005) (Fig. 1b). In contrast, 14 of 15 recipients (93%) treated with combined costimulation blockade and RPM developed stable multilineage mixed hemopoietic chimerism (Fig. 2) and accepted donor-specific skin allografts (Fig. 1b) throughout the follow-up period of 30 wk. Donor strain-specific skin allograft tolerance was demonstrated by placing the second B10.A or third-party skin allografts 90 days after BMT and skin transplantation. Although the second B10.A skin grafts were accepted without further immunosuppression (n = 7), third-party (DBA/1) skin grafts were promptly rejected (MST, 17 days; n = 7; data not shown). These data indicate that BMT synergizes with combined costimulation blockade and RPM treatment to induce allograft tolerance.
In chimeric recipients of combined costimulation blockade and RPM, MC persisted throughout the follow-up period of 30 wk (Fig. 2). In mice with stable multilineage MC, donor cells (B10.A) contributed up to 7.7% of the circulating white blood cells 21 wk after BMT (see Fig. 5). The proportion of donor CD3+ T cells among peripheral T cells remained stable at 2–3% in the circulating pool of BMT recipient mice (Fig. 2). The percentage of donor-derived B cells and MAC1-positive cells present in the peripheral blood of BMT recipients remained stable in the range of 5–9% (Fig. 2). The persistence of significant multilineage donor hematopoiesis for up to 30 wk after BMT demonstrates that engraftment of donor hemopoietic progenitor cells occurred (26).
Combined treatment with costimulation blockade and RPM inhibits the proliferation of allo-Ag-triggered CD4+ and CD8+ T cells
To evaluate the effect of a therapeutic regimen consisting of costimulation blockade and RPM on allo-Ag-triggered T cell immune response, we used a graft-vs-host disease-like model in which CFSE-labeled C57BL/6 (H-2b) MNLs were transferred into lethally irradiated allogeneic B10.A (H-2a) hosts. Three days after the adoptive transfer, 42% fewer CFSE-labeled C57BL/6 T cells were recovered from B10.A recipients treated with costimulation blockade and RPM as compared with untreated recipients (data not shown). Costimulation blockade and RPM administered alone were associated with mean reductions of 16 and 21% in the proportion of cells that had undergone proliferation, respectively, whereas the combination of both treatments led to an average 35% reduction. Moreover, the proportions of CFSE-labeled CD4+ and CD8+ T cells entering the cell cycle was reduced by 35% and 34%, respectively, in recipients of combined therapy compared with untreated recipients (Fig. 3). Thus, the combination of costimulation blockade and RPM had additive effects that were evident in the inhibition of CD4+ T cell proliferation, as compared with either treatment alone. In contrast, maximal inhibition of CD8+ T cell proliferation was achieved with RPM alone, and the addition of costimulation blockade did not enhance this effect. These results were consistent with our previous report using a different MHC-mismatched strain combination (12). Fig. 3 represents five independent experiments.
Administration of BMT, RPM, and costimulation blockade leads to deletion of T cell clones expressing specific TCR V families
To study whether clonal deletion was involved in the tolerizing effect promoted by combined BMT, RPM, and costimulation blockade treatment, we serially analyzed PBLs obtained from treated skin allograft recipients, or from control C57BL/6 and B10.A naive mice. Two weeks after BMT, there was already a marked reduction of CD4+ T cells expressing V5.1/5.2+ and V11+ (Fig. 4), as compared with naive C57BL/6 controls. In contrast, there was no reduction in the proportion of irrelevant V8+CD4+ T cells. Throughout a follow-up of >16 wk, the deletion of V5.1/5.2+ and V11+CD4+ T cells was further amplified (Fig. 4). These data suggest that clonal deletion of donor-reactive T cells is a mechanism implicated in the induction of allograft tolerance via combined BMT, RPM, and costimulation blockade treatment.
Challenge of tolerant chimeras with 5 x 106 MNLs from naive syngeneic donors does not abolish the stable mixed chimeric state or trigger skin allograft rejection
To test whether immunoregulatory mechanisms may be present in the maintenance phase of a stable mixed chimeric state and skin allograft tolerance, we challenged three skin allograft-tolerant MC hosts (>120 days post BMT) with 5 x 106 splenic MNLs from naive syngeneic mice. Following the challenge with 5 x 106 naive syngeneic MNLs, the proportion of donor cells present in peripheral blood remained stable, albeit with slightly, but not statistically significant, decrease throughout an 8-wk assessment period (Fig. 5), and donor skin allografts were retained. These results suggest that immunoregulatory mechanisms are involved in the maintenance of the stable mixed chimeric state and skin allograft tolerance (27).
Donor-specific immunoregulatory CD4+ T cells can be detected in secondary lymph organs of tolerant mixed chimeras
To further analyze the nature of the host antidonor response during the maintenance phase of tolerance in MC hosts, we used a passive transfer model using male RAG1-deficient, C57BL/6J-Rag KO mice (H-2b). These mice lack the RAG1, which is critical for the generation of a functioning TCR (they have no CD3+, or TCR--positive cells); thus, RAG1-deficient recipients are unable to elicit a sufficient immune response to reject allografts. These mice are missing CD4+ and CD8+ T cells and B cells; therefore, no CD4+ T cells can be detected in PBLs of naive RAG1-deficient mice.
As shown in Fig. 6a, adoptive transfer of as few as 0.5 x 106 MNLs harvested from naive C57BL/6 mice into immunoincompetent syngeneic Rag KO recipients triggered rapid rejection of B10.A skin allografts (MST of 15 days). In contrast, adoptive transfer of 0.5 x 106 (data not shown) or 10 x 106 MNLs from tolerant MC skin allograft recipients obtained >120 days after BMT into syngeneic Rag KO mice receiving a fresh B10.A fresh skin allograft did not elicit allograft rejection (Fig. 6a). To explore the possibility that regulatory cells might be present in tolerant long-term mixed chimeras, we cotransferred 0.5 x 106 naive C57BL/6 MNLs with a relative excess (10 x 106) of MNLs harvested from long-term C57BL/6 chimeras (120 days after BMT) into syngeneic Rag KO mice receiving a B10.A skin allograft. As shown in Fig. 6a, C57BL/6 MNLs harvested from tolerant hosts significantly inhibited the capacity of naive syngeneic MNLs to reject B10.A skin allografts (MST, 62 days; p = 0.0027). Moreover, these regulatory effects were donor Ag-specific, because the immunosuppressive effect was abolished when comixed MNLs from naive and tolerant mice were passively transferred into syngeneic recipients of third-party DBA/1 strain allografts (MST, 10.5 days).
To characterize the T cell subsets responsible for this suppressive effect, we repeated the cell transfer experiments using distinct subpopulations of tolerant cells. As shown in Fig. 6b, passive cotransfer of 5 x 106 tolerant CD4+ T cells, but not 5 x 106 CD8+ T cells or CD4+CD8+-depleted cells, with 0.5 x 106 naive syngeneic MNLs resulted in a significant delay of skin allograft rejection (Fig. 6b). Moreover, when CD25+ T cells were selectively deleted from tolerant CD4+ T cells, the cotransfer of 5 x 105 CD4+CD25– T cells from tolerant host with 0.5 x 106 naive MNLs resulted in acute allograft rejection of B10.A skin grafts (MST, 17 days; p = 0.0027). Thus, the immunoregulatory effects observed in this adoptive transfer model of skin transplantation are CD4+CD25+ T cell dependent (Fig. 6c).
To determine whether immunoregulatory T cells were also present in tolerant MC transplant recipients prepared with treatment of TBI, depleting anti-CD8 mAb, and anti-CD154 other than RPM plus costimulation blockade regimen, we repeated the adoptive transfer experiments, administering 0.5 x 106 naive MNLs together with 10 x 106 MNLs (naive:tolerant at 1:20 ratio) harvested from tolerant MC hosts treated with anti-CD154, depleting anti-CD8, and 3 Gy TBI or RPM plus costimulation blockade. Inconsistent with results reported previously (22), no immunoregulation could be detected in the adoptive transfer experiments using the mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, and naive hosts at 20:1 ratio (tolerant:naive; Fig. 6d). These results also suggest that the immunoregulatory effects observed in the transfer model using MNLs from tolerant RPM and costimulation blockade-treated MC hosts are not due to inhibition of homeostatic T cell proliferation.
To vigorously test whether there is immunoregulation, if any, present in the tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, we repeated the adoptive transfer experiment using mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment and naive hosts at 20:1, 30:1, and 40:1 ratio (tolerant:naive; Fig. 6e). Indeed, slight but significant inhibition of allografts rejection was evident when C57BL/6J-Rag KO skin graft recipients received mixture of MNLs from tolerant chimera and naive hosts at 40:1 ratio (tolerant:naive; Fig. 6e).
Functionally active immunoregulatory T cells reside in donor skin allografts from tolerant mixed chimeras treated with RPM and costimulation blockade
Because these experiments demonstrated that immunoregulatory CD4+ T cells were present in the secondary lymphoid organs of tolerant mixed chimeras (Fig. 6, a–c), we next investigated whether regulatory cells home to tolerated B10.A skin allografts and are active in protecting the allografts from rejection. B10.A skin allografts (n = 4) were obtained from tolerant C57BL/6 chimeras prepared with RPM and costimulatory blockade (>120 days after BMT and skin transplantation). As a control, B10.A skin autografts (n = 3) were harvested from B10.A recipients (>120 posttransplantation). Tolerated B10.A allografts or control B10.A autografts were then transplanted onto C57BL/6-Rag KO hosts. To determine whether graft-infiltrating T cells from tolerant skin allografts migrate and expand in the peripheral blood of Rag KO recipients, PBL samples were collected 30 days post skin transplantation. Because Rag KO mice are missing CD4+ and CD8+ T cells and B cells, no CD4+ T cells can be detected in PBL of naive RAG1-deficient mice. As depicted in Fig. 7a, CD4+ T cells could be detected readily in PBL of C57BL/6-Rag KO recipients of tolerated allografts, which otherwise lack T and B cells, but not in recipients of syngeneic grafts 30 days after graft transfer (6.78 ± 2.037% vs 0.98 ± 0.456%; p = 0.033). Because in an adoptive transfer model of skin allografts in C57BL/6-Rag KO recipients the CD4+ T cells, but not CD8+ T cells or B220+ B cells, from tolerant chimeras inhibit the ability of naive MNL to reject same donor strain allograft (Fig. 6b), we focus on the homing of CD4+ T cells in the tolerant grafts in this study. It would be of interest to test other types of cells, such as CD8+ T cells and B cells, as well in future studies.
To further test the immunoregulatory function of these graft-homing T cells from tolerant chimeras, 0.5 x 106 MNLs from naive C57BL/6 mice were adoptively transferred into the C57BL/6-Rag KO hosts bearing the tolerated B10.A allografts or B10.A autografts 30 days following skin transplantation. Following a challenge with 0.5 x 106 MNLs from naive C57BL/6 mice, survival of tolerated B10.A skin allografts transplanted onto C57BL/6-Rag KO mice was significantly prolonged (MST, 48 days) in comparison with the survival of B10.A skin autografts (MST, 16 days) (Fig. 7b; p = 0.01).
These data clearly indicate that the regulatory T cells reside within tolerated B10.A skin allografts of C57BL/6 chimeras and are functionally active in protecting the allografts from rejection in an adoptive transferred model of skin allografts.
CD4+CD25+-dependent immunoregulation is not required for tolerance induction after administration of BMT, RPM, and costimulation blockade
The thymus-derived CD4+CD25+ T cells, identified as suppressor cells in 1990 (28, 29), have emerged as critical effectors in both the control of autoimmunity (30, 31, 32) and the maintenance of peripheral allograft tolerance (24, 33, 34). In several allograft models in which tolerance is induced by either donor-specific blood transfusion (DST) plus anti-CD154 (24) or combined RPM plus lytic IL-2/Fc and antagonist mutant IL-15/Fc (35), the deletion of CD4+CD25+ Treg cells before transplantation results in acute allograft rejection, indicating that CD4+CD25+-dependent regulatory networks are critical for the induction of tolerance. Therefore, we tested the role of CD4+CD25+-dependent regulatory networks in the induction of MC and tolerance in hosts treated with combined BMT, RPM, and costimulation blockade. The CD4+CD25+ Treg cells were deleted in the recipient mice before BMT, and the effect of the treatment with PC61 mAb was evaluated by staining for CD25 in peripheral blood and analysis by flow cytometry.
Interestingly, the deletion of CD4+CD25+ Treg cells before BMT and skin transplantation did not interfere with the establishment of stable MC and skin allograft tolerance in the recipients treated with BMT, RPM, and costimulation blockade (data not shown). Thus, CD4+CD25+-dependent regulatory networks are not required for the tolerance induction in this model.
CD4+CD25– T cells from tolerant chimeras do not reject donor strain skin allografts after adoptive transfer
To determine the effects of BMT plus costimulation blockade and RPM treatment on CD4+CD25– effector T cells, we conducted additional adoptive transfer experiments in which we compared the capacity of CD25– MNLs harvested from naive or from tolerant chimeras to mediate rejection. The adoptive transfer of 10 x 106 CD25– MNLs from tolerant MC C57BL/6 hosts into C57BL/6J-Rag KO recipients did not trigger the rejection of same donor strain B10.A skin allografts throughout a follow-up period of 100 days. In contrast, naive MNLs selectively depleted of the CD25+ T cell subset were able to elicit acute rejection of B10.A skin allografts with a MST of 12.5 days (Fig. 8). Moreover, the adoptive transfer of 10 x 106 CD25– MNLs from tolerant chimeras resulted in acute rejection of third-party strain DBA/1 allografts at a MST of 10 days (Fig. 8), suggesting that profound selective deletion of anti-B10.A effector CD25– T cells in the tolerant chimeras had taken place.
Discussion
In this study, we demonstrate that combined treatment with costimulation blockade and RPM, in the absence of any host preconditioning, enables the induction of stable MC and skin allograft tolerance in fully MHC-mismatched recipients (B10.A into C57BL/6) (Figs. 1 and 2). In skin-tolerant MCs, the percentage of donor cells present among circulating MNLs was stable throughout the follow-up period of 30 wk. In parallel with the development of stable multilineage mixed hemopoietic chimerism, donor-specific skin allograft tolerance was achieved (Fig. 1b). In this experimental setting, combined high-dose BMT with costimulation blockade treatment was unable to induce stable MC and prevent acute skin allograft rejection (Fig. 1b). The difference from previous experiments (14) most likely reflects differences in the CTLA4Ig preparations used (see footnote).
Because a RPM-sensitive mechanism is responsible for the failure of costimulation blockade to create peripheral tolerance in many models (12, 13, 17), and RPM does not inhibit the development of chimerism or tolerance induced by TBI, BMT, and costimulation blockade (36), we studied the effect of RPM as an adjunct to costimulation blockade in our attempt to induce stable MC and allograft tolerance. Indeed, the adjunctive administration of RPM to this costimulation blockade-based protocol provides a noncytotoxic, irradiation- free approach that induces MC and skin allograft tolerance.
There were numerous studies focused on the deletion mechanism in MC induced by both lymphoablative conditioning or nonlyphoablative regimens. In hosts in which MC and transplant tolerance is induced with regimens using lymphoablative conditioning (1, 2, 4, 5, 6, 7), ongoing alloreactive T cell deletion occurring in the thymus is the dominant, if not exclusive, mechanism responsible for the induction and maintenance of organ allograft tolerance (8, 9, 37). In contrast, in regimens using costimulatory blockade instead of peripheral T cell depletion to induce MC and allograft tolerance, specific peripheral deletion of donor-reactive T cells without global depletion takes place (6, 14). These reported mechanisms (i.e., intrathymic and peripheral deletion of donor-reactive T cells after BMT) appear to differ from those noted in models of peripheral tolerance, in which permanent engraftment of organ allografts is achieved via costimulation blockade without BMT. In these latter models, although a substantial depletion of alloreactive T cell clones is required for tolerance induction, the long-term maintenance of the tolerant state is dependent upon self-perpetuating immunoregulatory networks that constrain the remaining cytopathic T cells (11, 12, 21, 24, 27, 38, 39, 40). However, the role of regulatory mechanism in MC has not been fully investigated. According to the pool size model of allograft tolerance, the allograft outcome, rejection or tolerance, often depends on the balance between cytopathic and regulatory T cells. Although both deletion and regulation play important roles in allograft tolerance, recent studies showed that the quantitative details for each mechanism differ from model to model. Therefore, we hypothesize that there is a delicate balance between deletion and regulation in allograft tolerance. In the current study, we sought to vigorously test the balance of deletional and regulatory mechanisms in this model.
Our data suggest that deletion of donor-reactive cytopathic T cell clones is indeed profound in tolerant chimeras receiving BMT plus a nonlymphoablative, irradiation-free protocol. This is supported by the failure of CD4+CD25– T cells from tolerant chimeras to reject same donor, but not third-party, strain allografts upon adoptive transfer into syngeneic recipients (Fig. 8). In addition, using V5, V11, and V8 TCR probes as reported previously (5, 6, 8, 9, 14, 16), we have demonstrated specific deletion of donor superantigen-reactive CD4 T cell clones in skin allograft-tolerant MC hosts (Fig. 4). Taken together, these data strongly support the notion that deletional mechanisms play an important role in this model of tolerance. This is sharply contrasted with the allograft tolerance model induced by DST plus MR1, in which the capacity of CD4+CD25– effector T cells from the tolerant hosts to trigger skin allograft rejection on a per cell-to-cell basis is identical with that from naive mice, suggesting that the deletion is not nearly complete in this model (24).
In addition, in the current study, we sought to determine whether immunoregulatory networks were also present in allograft-tolerant MC hosts receiving costimulation blockade and RPM treatment, and if so, to ascertain the precise balance between immunoregulatory and deletional mechanisms in the acquisition of allograft tolerance with this therapy.
Indeed, several lines of evidence indicate that immunoregulatory mechanisms are also present in the maintenance phase of tolerance in mixed chimeric hosts treated with RPM and costimulation blockade. First, the challenge of tolerant MC skin allograft recipients with 5 x 106 MNLs from naive syngeneic mice was neither able to abolish the mixed chimeric state (Fig. 5) nor to trigger skin allograft rejection (data not shown), a hallmark of peripheral, not central tolerance. Furthermore, in an adoptive transfer model, MNLs harvested from the secondary lymphoid organs of RPM and costimulation blockade-treated MC skin-tolerant hosts significantly inhibited the capacity of naive MNLs to reject B10.A skin allografts (Fig. 6a). These T cell-dependent regulatory effects were donor specific and dependent on the presence of CD4+CD25+ T cells among the tolerant MNLs. This was supported by the observation that selective depletion of CD4+CD25+ regulatory T cells abolished the ability of tolerant MNLs cells to prevent naive MNLs from rejecting allografts (Fig. 6c), and is consistent with the results obtained in costimulation blockade-based models of peripheral tolerance (21, 24, 34). Finally, when we transplanted skin allografts from stable tolerant chimeras onto syngeneic immune-incompetent mice, graft-infiltrating T cells migrated from the graft site, expanded in the new host, and protected test allografts from acute rejection after transfer of naive syngeneic MNLs (Fig. 7). A similar finding has been reported recently by Graca et al. (41) in a model of peripheral tolerance to minor histocompatibility Ag-mismatched skin allografts. Thus, in this model in which skin allograft tolerance is achieved via BMT, RPM, and costimulation blockade, regulatory T cells are present in both secondary lymphoid organs and in the allograft itself, and are functionally active in protecting the grafts from rejection (Figs. 6 and 7).
In short, the use of BMT, RPM, and costimulation blockade treatment to induce transplantation tolerance is associated with both deletional and active immunoregulatory phenomena. However, these results are in contrast to those obtained in mixed chimeras in which transplantation tolerance was established by BMT, anti-CD154, depleting anti-CD8, and low-dose TBI, because adoptive transfer of 10 x 106 MNLs obtained from tolerant hosts treated with RPM and combined costimulation blockade, but not MNLs harvested from anti-CD8, anti-CD154, and TBI-treated tolerant MC hosts, inhibit the capacity of 0.5 x 105 naive MNLs to reject B10.A skin allografts (Fig. 6d) (22). A noteworthy difference between these strategies, using TBI and anti-CD8 mAb treatment and our current treatment, is the lower levels of multilineage chimerism achieved with combined RPM and costimulation blockade treatment (9, 22, 42). To vigorously test whether there is immunoregulation, if any, present in the tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, we repeated the adoptive transfer experiment using mixture of MNLs from tolerant chimera receiving anti-CD154, depleting anti-CD8, and 3 Gy TBI treatment, and naive hosts at 20:1, 30:1, and 40:1 ratio (tolerant:naive; Fig. 6e). Indeed, a slight but significant inhibition of allograft rejection was evident only when C57BL/6J-Rag KO skin graft recipients received a mixture of MNLs from tolerant chimera and naive hosts at 40:1 ratio (tolerant:na?ve; Fig. 6e). CD4+CD25+ T cells are known to be prone to thymic deletion (43), although their specific susceptibility to deletion, as compared with nonregulatory T cell populations, has not been extensively addressed. We speculate that conventional alloreactive, but not regulatory, T cells are depleted in hosts with modest levels of hemopoietic MC, whereas both conventional and regulatory donor-reactive T cells are centrally eliminated in hosts with more robust MC. A recent report indirectly supports this hypothesis (44). Alternatively, the presence of different levels of immunoregulation in MC models with different levels of hemopoietic MC might relate to the persistence of residual donor-reactive T cells, which may be required to maintain effective immunoregulatory networks. Hence, with more complete long-term donor-reactive CD4+ and CD8+ T cell deletion, regulation would be less necessary to maintain tolerance.
Is immunoregulation truly necessary to achieve tolerance in this costimulation blockade-based model of MC? Unlike protocols inducing "pure" peripheral tolerance (24, 35), the depletion of CD4+CD25+ T cells before BMT did not preclude the induction of tolerance in our model (data not shown). Hence, our data suggest that in MC, CD4+CD25+ T cell-dependent immunoregulation is not as important for the induction and maintenance of transplantation tolerance as seen in more conventional models of peripheral tolerance (24, 35, 45). These results further support our clone-size hypothesis that the outcome of the allograft response, rejection or tolerance, depends on the balance between cytopathic and regulatory T cells (46, 47). Although both deletion and regulation play important roles in allograft tolerance, our recent studies showed that the quantitative details for each mechanism differ from model to model. Therefore, we hypothesize that there is a delicate balance between deletion and regulation in allograft tolerance. As in a model of allograft tolerance in which the deletional mechanisms play a dominant role (e.g., tolerance produced via creation of mixed chimeras), the regulatory mechanisms, albeit sometimes present, are far less important (6, 8, 37). Although, in a model in which the regulation mechanisms play critical role (e.g., DST plus MR1-induced allograft tolerance), deletional mechanisms lower the threshold for effective regulatory T cell action (Fig. 9) (24, 35, 45).
In short, the presence of both T cell depletion and immunoregulation has been noted in a model in which skin allograft tolerance is achieved through the induction of MC. Hence, the classic definitions of central and peripheral tolerance may be somewhat imprecise. Altogether, tolerance can be achieved through subtly differing mixtures of clonal deletion and immunoregulation.
Acknowledgments
We thank Yan Tian for excellent technical assistance, and Dr. Thomas Fehr and Raquel Oliviera for providing bone marrow chimeras for some skin transfer experiments.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 Address correspondence and reprint requests to Dr. Xin Xiao Zheng and Dr. Terry B. Strom, Transplantation Research Center, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215; E-mail address: xzheng@bidmc.harvard.edu and tstrom@bidmc.harvard.edu
2 Abbreviations used in this paper: MC, mixed chimerism; BMT, bone marrow transplantation; TBI, total body irradiation; BMC, bone marrow cell; RPM, rapamycin; KO, knockout; MNL, mononuclear leukocyte; MST, median survival time; DST, donor-specific blood transfusion.
3 When a CTLA4lg preparation different than the one employed by Wekerle et al. (14 ) was used to generate tolerant MC host together with BMT, anti-CD154, and 3 Gy TBI, we observed that CD8 cells could no longer be rendered tolerant (48 ). In constrast, long-term MC and long-term systemic donor-specific tolerance of both CD4+ and CD8+ T cells (48 ) could be reliably induced in mice receiving anti-CD8 depleting-mAb and a single injection of anti-CD154 with 3 Gy TBI (48 ).
Received for publication June 9, 2004. Accepted for publication March 25, 2005.
References
Ildstad, S. T., D. H. Sachs. 1984. Reconstitution with syngeneic plus allogeneic or xenogeneic bone marrow leads to specific acceptance of allografts or xenografts. Nature 307: 168-170.
Sharabi, Y., D. H. Sachs. 1989. Mixed chimerism and permanent specific transplantation tolerance induced by a nonlethal preparative regimen. J. Exp. Med. 169: 493-502.
Colson, Y. L., H. Li, S. S. Boggs, K. D. Patrene, P. C. Johnson, S. T. Ildstad. 1996. Durable mixed allogeneic chimerism and tolerance by a nonlethal radiation-based cytoreductive approach. J. Immunol. 157: 2820-2829.
Nomoto, K., K. Yung-Yun, K. Omoto, M. Umesue, Y. Murakami, G. Matsuzaki. 1995. Tolerance induction in a fully allogeneic combination using anti-T cell receptor- monoclonal antibody, low dose irradiation, and donor bone marrow transfusion. Transplantation 59: 395-401.
Sykes, M., G. L. Szot, K. A. Swenson, D. A. Pearson. 1997. Induction of high levels of allogeneic hematopoietic reconstitution and donor-specific tolerance without myelosuppressive conditioning. Nat. Med. 3: 783-787.
Wekerle, T., M. H. Sayegh, J. Hill, Y. Zhao, A. Chandraker, K. G. Swenson, G. Zhao, M. Sykes. 1998. Extrathymic T cell deletion and allogeneic stem cell engraftment induced with costimulatory blockade is followed by central T cell tolerance. J. Exp. Med. 187: 2037-2044.
Sharabi, Y., I. Aksentijevich, T. M. Sundt, 3rd, D. H. Sachs, M. Sykes. 1990. Specific tolerance induction across a xenogeneic barrier: production of mixed rat/mouse lymphohematopoietic chimeras using a nonlethal preparative regimen. J. Exp. Med. 172: 195-202.
Tomita, Y., A. Khan, M. Sykes. 1994. Role of intrathymic clonal deletion and peripheral anergy in transplantation tolerance induced by bone marrow transplantation in mice conditioned with a nonmyeloablative regimen. J. Immunol. 153: 1087-1098.(Christoph Domenig*, Alber)