Conventional Immunosuppression is Compatible with Costimulation Blockade-Based, Mixed Chimerism Tolerance Induction


* Corresponding author: Christian P. Larsen/ Thomas C. Pearson,


T-cell costimulatory blockade has emerged as an effective strategy to prevent allograft rejection in experimental models. We and others have reported that the beneficial effects of costimulation blockade can be negated when combined with certain immunosuppressants. The current study evaluates the compatibility of various immunosuppressive agents in a costimulation blockade-based, mixed chimerism tolerance protocol.

The addition of conventional agents, including calcineurin inhibitors, did not interfere with tolerance induction. All mice developed multilineage macrochimerism and accepted donor allografts. Analysis of specific T-cell receptor utilization demonstrated selective deletion of donor-reactive T cells. Challenge with donor and third-party allografts confirmed donor-specific tolerance.

Clinical introduction of costimulation blockade-based strategies will likely incorporate currently approved immunosuppressive agents. While it has been reported that certain conventional agents are detrimental to costimulation blockade-based strategies, our results suggest that these agents could safely be combined in clinical trials when used as part of a nonmyelosuppressive, mixed chimerism-based tolerance strategy.


Naïve T cells require several cellular signals for optimal activation, proliferation, and cytokine production. The interaction between the T-cell receptor (TCR) and antigenic peptide bound to a major histocompatibility (MHC) molecule, ‘signal 1’, imparts specificity to the immune response and is necessary but not sufficient for T-cell activation. Additional costimulatory signals, provided by a variety of molecules present on antigen-presenting cells including CD80/86 and CD40, constitute ‘signal 2’ and are necessary for the initiation of effector function (1).

Initiation of T-cell activation without appropriate costimulation results in anergy and/or death and is a recognized mechanism of peripheral tolerance (2). Over the past decade a number of laboratories have exploited this observation by blocking costimulatory signals at the time of transplantation in the hope of specifically inactivating/eliminating allo-reactive T cells. Indeed, many groups have shown that regimens employing antibodies or fusion proteins, which prevent key costimulatory interactions, can achieve long-term allograft survival in experimental models of organ and tissue transplantation (3–6).

Modern immunosuppressive therapy relies heavily on calcineurin inhibitors [i.e. cyclosporine A (CsA) and tacrolimus] and corticosteroid therapy, both of which block allo-immune responses by directly inhibiting early T-cell activation (7). Interestingly, we and others have shown that these same immunosuppressants antagonize the beneficial effects of costimulation blockade in transplantation tolerance models, suggesting the mechanism of action of costimulation blockade requires intact signaling through the T-cell receptor (4,5,8). More recent work by Turka and Strom has demonstrated that activation-induced cell death is required for costimulation blockade-dependent peripheral tolerance induction, and the addition of CsA negates the beneficial effects of costimulation blockade therapy by inhibiting proliferation, apoptosis, and subsequent deletion of alloreactive T cells (9,10). In contrast to calcineurin inhibitors, sirolimus (rapamycin), an agent that does not alter early signaling through the TCR and permits cell cycle-dependent apoptosis, synergizes with costimulation blockade to promote long-term allograft survival (9). Unfortunately despite such elegant studies there are additional, seemingly conflicting, reports describing the use of calcineurin inhibitors or other conventional agents in combination with a variety of costimulation blockade reagents resulting in synergistic activity in both murine and nonhuman primate models (11–16). Additionally, the compatibility of these agents has not been fully evaluated in the setting of bone marrow transplant. This setting involves several unique characteristics, including the distribution and phenotype of the donor cells in a hematopoietic graft vs. that of a solid-organ transplant, and thus may provide a distinct outcome when conventional immunosuppressive agents are combined with costimulation blockade in a tolerance regimen.

A proven method to produce robust transplantation tolerance is the induction of donor hematopoietic chimerism. Despite the promise demonstrated in experimental models, concerns regarding toxicities associated with recipient preconditioning have precluded widespread clinical application in solid organ transplantation. Recent studies have shown that combined blockade of the CD28 and CD40 costimulatory pathways dramatically decreases the amount of recipient conditioning required to achieve hematopoietic macrochimerism after bone marrow infusion (17). In fact, treatment with anti-CD154 alone (18) or in combination with CTLA4-Ig (19) completely eliminates the need for recipient conditioning when very large doses of donor bone marrow are used. Unfortunately the large dose of bone marrow required by this strategy renders it clinically impractical. We have reported that the combined regimen of anti-CD154, CTLA4-Ig, and busulfan, an agent that is relatively selective for hematopoietic stem cells, promotes stable macrochimerism, using much lower numbers of donor bone marrow cells and with minimal myelosuppression (20). This regimen promotes robust tolerance to skin allografts placed at the initiation of therapy and prevents the occurrence of chronic rejection in a vascularized heterotopic heart transplant model (21).

Initial clinical application of costimulation blockade-based tolerance strategies will most likely require concomitant treatment with currently approved immunosuppressive agents for ethical and safety concerns. The compatibility of costimulation blockade and conventional immunsuppressants, including calcineurin inhibitors and corticosteroids, in the setting of a mixed chimerism regimen has not yet been fully determined. The current study evaluates the effects of various conventional immunosuppressive agents when combined with costimulation blockade (anti-CD154 and CTLA4-Ig) in a nonmyelosuppressive, mixed chimerism tolerance regimen.

Materials and Methods


Adult male 4–8-week-old C57BL/6 (H-2b), Balb/c (H-2d), and C3H/HeJ (H-2k) were obtained from Jackson Laboratories (Bar Harbor, ME). All mice were housed in specific pathogen-free conditions and in accordance with institutional guidelines.

Bone marrow preparation and treatment regimens

Bone marrow was flushed from the tibiae, femora, and humeri. Red cell lysis was performed using a Trizma base ammonium chloride solution. The bone marrow cells were resuspended at 2 × 107cells/500 µL of sterile saline and injected intravenously on d0 and d6. Hamster antimouse CD154 mAb (MR1, Bioexpress, Lebanon, NH) and CTLA4-Ig (Bristol-Myers Squibb, Princeton, NJ) were administered on days 0, 2, 4, and 6 (500 µg/dose i.p.). Busulfan was diluted 1 : 5 with normal saline and administered on day 5 (600 µg i.p., Busulfex, Orphan Medical, Minnetonka, MN). The depleting anti-CD4 mAb (GK1.5) was administered 100 µg i.p. on days − 2, − 1, and 0, and weekly for the duration of the experiment. Immunosuppressive regimens: cyclosporine A (CsA) (600 µg s.c. daily, d0–30), tacrolimus (60 µg s.c. daily, d0–30), rapamycin (12 µg i.p. daily, d0–30), solumedrol (5 mg i.p. daily, d0–30), mycophenolate mofetil (MMF) (450 µg i.p. daily, d0–30), FTY720, a kind gift from Volker Brinkmann, Novartis (0.3 mg/kg daily, d0–10), or anti-CD25 (250 µg i.p. d0, 2, 4, 6 (PC61, Bioexpress, Lebanon, NH)).

Skin grafting

Full-thickness skin grafts (∼1 cm2) were transplanted on the dorsal thorax of the recipient mice and secured with a Band-Aid® (J & J, Arlington, TX) for 7 days. Rejection was defined as the complete loss of viable epidermal graft tissue.

Flow cytometric analysis

Peripheral blood was analyzed by staining with fluorochrome-conjugated antibodies [anti-CD3, anti-CD5, anti-CD11b, anti-CD11c, anti-GR1, anti-B220, anti-H-2Kd, anti-I-Ad anti-H-2Kb, anti-Vβ11, anti-Vβ5.1/5.2, anti-Vβ8.1/8.2 (Pharmingen), anti-CD4, anti-CD8 (Caltag Laboratories, Burlingame, CA) or immunoglobulin isotype controls (Pharmingen, Caltag)], followed by red blood cell lysis and washing with a whole blood lysis kit (R + D Systems, Minneapolis, MN). Stained cells were analyzed using Cellquest software on a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA). Percent donor chimerism was determined as follows: (# H-2Kd+ cells/total cells) × 100.


Conventional immunosuppressive agents do not impair tolerance induction

We have previously described a nonirradiation-based strategy utilizing a single dose of busulfan, costimulation blockade, and a clinically relevant amount of donor bone marrow to promote high-level chimerism and reshaping of the T-cell repertoire (20). Chimeras exhibit robust donor-specific tolerance, as evidenced by acceptance of fully allogeneic skin grafts and failure to generate donor-specific proliferative responses in an in-vivo GvHD model of alloreactivity.

To determine the effects of various clinically relevant immunosuppressive agents on the costimulation blockade-based tolerance regimen, the mice were treated with the tolerance protocol alone or in combination with various compounds [cyclosporine A (CsA), tacrolimus, sirolimus (rapamycin), FTY720, mycophenolate mofetil (MMF), corticosteroid (solumderol), or a monoclonal antibody targeting CD25]. In brief, the recipient B6 (H-2b) mice received fully allogeneic bone marrow [Balb/c (H-2d) 2 × 107 cells i.v. d0 and d6], costimulation blockade consisting of CTLA4-Ig and anti-CD154 (500 µg d0,2,4,6), and a single nonmyelosuppressive dose of busulfan (d5). In addition to the tolerance protocol, the mice received either saline (control group) or one of the immunosuppressive agents listed earlier. Immunosuppressive drug dosing regimens were obtained from published works, including reports from our own laboratory (4,8,9,22). All the mice received donor-type skin grafts at the initiation of therapy (d0).

While calcineurin inhibitors (CsA and tacrolimus) and corticosteroids have been reported to interrupt the beneficial effects of costimulation blockade, their addition to the nonmyelosuppressive tolerance protocol did not interfere with the establishment of high-level mixed chimerism and allograft tolerance (CsA- 7/7, Tacrolimus-7/7, Control 11/12, MST > 100 days, Figure 1A,C). Similar results were obtained in three separate experiments. Next we tested rapamycin, a novel immunosuppressant that unlike calcineurin inhibitors and steroids permits early signaling events in T-cell activation and acts later to inhibit events downstream of the IL-2 receptor. Consistent with previous reports, rapamycin was fully compatible with the costimulation blockade in the mixed chimerism tolerance protocol and promoted stable donor chimerism (7/7) and indefinite skin graft survival (MST > 100 days, Figure 1A,C). To ensure adequate dosing, serum trough levels were sampled and analyzed at various time points (CsA- 24-h trough level ranged from 750 to 850 ng/mL, rapamycin 15–18 ng/mL).

Figure 1.

Compatibility of conventional immunosuppression. (A) Animals were treated with the tolerance protocol (n = 12) or the tolerance protocol and one additional immunosuppressive agent (n = 7/gp). Samples of recipient blood were analyzed for the presence of cells bearing donor major histocompatibility complexes (H-2 kDa). All groups developed stable mixed chimerism; d100 analysis is shown as an example. (B) In all the animals that became chimeric following treatment with the tolerance protocol, analysis of specific lineage markers (T cell CD3+, B cell B220+, granulocyte GR1+) revealed multilineage production of donor cells, indicating stable engraftment of donor hematopoietic stem cells. A representative histogram plot is shown. (C) Primary skin allograft survival for animals receiving no treatment (control), tolerance protocol alone (no additional treatment), or the tolerance protocol and an additional immunosuppressive agent (as listed) is shown.

Mycophenolate mofetil, an agent widely used in clinical organ transplantation, inhibits lymphocyte proliferation by blocking an enzyme (IMP-dehydrogenase) critical for the de-novo synthesis of guanosine nucleotides. As MMF has no effect on the production or release of cytokines (e.g. IL-2) associated with early T-cell signal transduction, we hypothesized that it would not interfere with the action of costimulation blockade in our mixed chimerism protocol. Like the previous agents tested, animals treated with MMF and the tolerance regimen developed high levels of donor chimerism (7/7) and indefinite skin graft survival (MST > 100 days, Figure 1A,C). Similar results were obtained in three separate experiments.

The impact of the novel immunosuppressant FTY720 was also evaluated when used in combination with the costimulation blockade-based mixed chimerism protocol. FTY720, an analog of the fungal metabolite myricin, elicits a profound lymphopenia resulting from a reversible sequestration of lymphocytes in secondary lymphoid organs. Recent work has shown that this occurs as a result of phosphorylated FTY720 binding agonistically to sphingosine 1-phosphate receptors and modulating chemotactic responses (23). We hypothesized that FTY720 and costimulation blockade may represent an ideal combination for tolerance, as treatment with FTY720 would induce the migration of alloreactive T cells to secondary lymphoid organs where interactions with donor APCs would be facilitated while treating with costimulation blockade to promote tolerance. As predicted the addition of FTY720 to the tolerance protocol did not affect the outcome as 6/7 recipients were chimeric and tolerant, similar to the control group (11/12, Figure 1A,C, MST > 100 days). In all the experimental groups engraftment of donor cells resulted in multilineage chimerism as defined by lineage-specific markers using flow cytometric analysis (Figure 1B). The kinetics of chimerism development were similar to our initial report (20).

As a test of donor-specific tolerance, animals in each experimental group were re-challenged approximately 100 days after initiation of therapy with donor (BALB/c) and third-party (C3H/HeJ) skin grafts. Control animals promptly rejected both BALB/c and C3H/HeJ skin grafts (MST 10d and 12d, respectively, data not shown). Mice receiving the tolerance protocol alone or in combination with an additional immunosuppressive agent uniformly accepted donor-type (BALB/c) skin grafts (MST > 100 days, data not shown) and rejected third-party grafts in similar time to controls (data not shown). Similar results were obtained in two separate experiments.

Anti-CD25 therapy is compatible despite the early requirement for regulation

Treatment with monoclonal antibodies targeting CD25 is used extensively in clinical organ transplantation, however, the impact on the action of costimulation blockade in a mixed chimerism regimen has not yet been examined. We have previously shown that tolerance induction is prevented if a depleting anti-CD4 mAb is administered at the initiation of the protocol, suggesting regulatory cells are necessary during the tolerance induction phase (20). Recent studies have shown that CD4+CD25+ T cells have important regulatory functions in both autoimmunity and transplantation (24,25). Although we have shown that anti-CD25 therapy is synergistic with costimulation blockade in a nonchimerism model (26), we thought it important to determine the compatibility of this agent in a regimen where CD4+ T cells are essential in establishing the tolerant state. Interestingly despite the need for a CD4+ cell population during the induction phase of the tolerance protocol, anti-CD25 therapy was compatible in promoting stable chimerism and tolerance (7/7, MST > 100 days, Figures 1A and 2A).

Addition of conventional immunsuppressants does not abrogate deletion of donor-reactive T cells

Because of a paucity of defined epitopes for alloreactivity it is difficult to analyze and follow allogeneic antigen-specific T-cell responses. However, a well-developed surrogate marker system has been established based on differential retroviral (mouse mammary tumor virus) superantigen expression. Balb/c mice delete Vβ11 and Vβ5 bearing T cells whereas B6 mice do not express I-E and utilize Vβ11 on ∼4–5% of CD4+ T cells and Vβ5.1/2 on ∼2–3% of CD4+ T cells (27). As anticipated, control groups not receiving the full protocol failed to delete donor-reactive Vβ11+ or Vβ5+CD4+ T cells (data not shown). In contrast, recipients of BALB/c bone marrow, busulfan, and costimulation blockade therapy deleted CD4+Vβ11+ and CD4+Vβ5+ T cells when tested at 100 days post-transplant (n = 7, Figure 2B). The percentage of Vβ8-bearing CD4+ T cells, which are expressed on approximately 15–20% of Balb/c and B6 CD4+ T cells, was similar in all groups, suggesting the T-cell deletion was donor specific in nature (Figure 2B).

Figure 2.

Conventional agents do not abrogate costimulation blockade induced deletion. (A) The costimulation blockade-based tolerance protocol is dependent on a CD4+ cell population during the induction phase (MST = 32 days). Treatment with an anti-CD25 mAb did not inhibit tolerance and allowed for indefinite skin allograft survival (MST > 100 days). (B) Peripheral blood samples at day 100 post-transplant indicating deletion of donor reactive CD4+ T cells following treatment with the tolerance protocol alone (control, n = 7) or the tolerance protocol and an additional immunosuppressive agent (n = 7/gp, as listed) are shown.

We hypothesized that the addition of calcineurin inhibitors and corticosteroids, which inhibit early T-cell activation, might interfere with deletion of alloreactive T cells, while other immunosuppressants, such as rapamycin and MMF, that target later signaling steps would be compatible and perhaps promote deletion in a synergistic manner. Surprisingly when Vβ utilization was analyzed none of the conventional immunosuppressive agents antagonized the effects of the costimulation blockade to promote deletion of donor-reactive T cells (n = 7/gp, Figure 2B). The existence of any synergistic interactions between the costimulation blockade regimen and certain immunosuppressive agents is difficult to analyze in this system, as the base protocol reliably induces tolerance in approximately 90–95% of the recipients. However, there were no appreciable differences in the kinetics of the deletion of alloreactive T cells, suggesting that none of the agents acted synergistically to promote deletion in a more rapid manner.


Clinical trials to test the safety and efficacy of costimulation-based strategies in clinical solid organ transplantation will likely involve the concomitant use of conventional immunosuppressive agents. For example, the clinical trial currently in progress to test blockade of the CD28/B7 pathway in renal transplant patients includes concurrent treatment with MMF, steroids, and an anti-CD25 mAb. The design of clinical trials to include conventional agents provides an adequate margin of safety against allograft rejection thus permitting the testing of new, exciting agents in an ethically acceptable manner. As a result, a more basic understanding of the interactions between these agents and costimulation blockade strategies is critical for the design of preclinical nonhuman primate studies and clinical trials.

Blockade of the CD40 and CD28 pathways effectively inhibits rejection responses in rodent and nonhuman primate allograft models. However, when this strategy is combined with calcineurin inhibitors, and in some cases corticosteroids, the salutary effect of costimulation blockade on allograft survival is paradoxically impaired (4,28,29). Subsequent studies have suggested that calcineurin inhibitors prevent the apoptosis of donor-reactive T cells induced by costimulation blockade, while sirolimus conversely potentiates the effects of costimulation blockade (9). Other important clinical agents such as anti-IL-2R mAbs and mycophenylate mofetil have not been extensively studied. One might hypothesize that MMF and other cell-cycle inhibitors may enhance apoptosis when combined with costimulation blockade. The effects of anti-IL-2R monoclonal antibodies are harder to predict because IL-2 plays a critical role in both T-cell expansion and T-cell death. Furthermore, recent reports have highlighted the capacity of CD4+CD25+ T cells to promote prolonged allograft survival through regulatory mechanisms (24,30). The influence of anti-CD25 therapy on the function of these cells in vivo in an allograft model is unknown.

The results of this study indicate that conventional immunosuppressive agents can be combined with costimulation blockade in a mixed chimerism tolerance protocol without adverse consequences. We tested a variety of immunosuppressive agents that are currently used in clinical practice or are in phase III trials (CsA, tacrolimus, solumedrol, rapamycin, MMF, FTY720, and anti-IL-2R) for their effects on a costimulation blockade-based tolerance protocol. The intention of these studies was to aid in the development of similar tolerance protocols for testing in a preclinical nonhuman primate model.

There was no appreciable difference in the levels of donor chimerism, skin allograft survival, or deletion of donor reactive T cells when comparing the group receiving the tolerance protocol alone with experimental groups that included an additional immunosuppressive agent. Despite previous reports from our own group and others (4,5,9), the addition of calcineurin inhibitors or corticosteroids did not alter the tolerizing effects of costimulation blockade. These contrasting results may be explained by fundamental differences in the tolerance strategies tested. The previous reports described models employing costimulation blockade alone as a means to promote allograft survival, while our current work evaluated the effects of conventional immunosuppressants in a bone marrow-based, mixed chimerism tolerance strategy. In this setting the unique characteristics and systemic distribution of the donor-cell type in a hematopoietic graft differs greatly from that of a solid-organ transplant and may, in part, explain the contrasting results.

Our current findings are consistent with the recent work of Taylor et al. using an irradiation-based bone marrow protocol (31). Similar to our results both sirolimus and CsA were found to be compatible with anti-CD154 therapy in promoting long-term multilineage engraftment. More recently an additional paper by Blaha et al. examined the compatibility of many of these same reagents in an irradiation-based mixed chimerism protocol (32). Contrary to our results they observed that calcineurin inhibitors inhibit the development of long-term chimerism and abrogate tolerance and deletion of donor-reactive T cells. These differences could be the result of concomitant administration of CTLA4-Ig starting at the initiation of the tolerance regimen (i.e. not waiting 2 days), the lack of irradiation, and/or the duration and dosing of the treatment regimens.

We also report the compatibility of an antibody against CD25, the high-affinity chain of the IL-2 receptor, in our costimulation blockade chimerism regimen. We and others have previously reported that tolerance following infusion of donor cells and treatment with costimulation blockade is dependent on CD4+ cells (20,33). Given the recent studies demonstrating the important regulatory function of CD4+CD25+ T cells, we thought it was important to test the compatibility of our tolerance regimen with anti-IL-2R treatment. This was also relevant given the increasing use of anti-IL-2R mAbs in clinical transplantation. Interestingly while anti-CD4 treatment inhibited tolerance, administration of anti-CD25 at the time of tolerance induction did not affect the outcome. The exact mechanism whereby treatment with anti-IL-2R mAb enhances allograft survival is not fully elucidated but it might not be via depletion of CD25+ cells. In our previous report we show that treatment with anti-IL-2 produces similar prolongation of skin graft survival when combined with costimulation blockade, suggesting that the mechanism of action may be interruption of the IL-2 signaling and does not require depletion of CD25+ T cells (26). Interestingly in the same report we show that the increase in graft survival attained when anti-IL-2R is combined with costimulation blockade is dependent on CD4+ T cells, as depletion of this subset with anti-CD4 mAb removes the effect.

Our studies indicate that the immunosuppressive agents tested can singly be combined with the costimulation blockade mixed chimerism protocol without adverse consequences. Further studies would be needed to evaluate the effects of simultaneous administration of more than one conventional agent in combination with the tolerance protocol, but this could easily be tested once a specific regimen has been identified. The intent of these compatibility studies was to aid in the development of a similar protocol for testing in a preclinical nonhuman primate model utilizing additional clinically approved immunosuppressive agents. The observation that tolerance was not abrogated when calcineurin inhibitors and steroids were introduced into the regimen increases the likelihood that a similar tolerance protocol will successfully evolve into clinical application.


This work was supported in part by research grants AI/DK040519 and AI044644 from the National Institutes of Health, and by the ERC Program of the National Science Foundation under Award Number EEC-9731643, as well as by the Carlos and Marguerite Mason Trust.