FTY720 Prevents Anti-CD4 mAb-Induced Tolerance but Cannot Reverse Established Tolerance in a Rat Kidney Transplantation Model


*Corresponding author: Grit Schroeder, grit.schroeder@med.uni-rostock.de


FTY720 is highly effective in various models of transplantation and autoimmunity. In order to find drugs that act synergistically with a tolerance-inducing nondepleting anti-CD4 mAb we studied this combination in a strong DA to LEW kidney transplantation model. Rats were treated with 0.3 mg/kg of FTY720 for 14 days and anti-CD4 mAb RIB5/2, alone or in combination. After kidney transplantation serum creatinine and blood lymphocyte counts were monitored. Immunohistology, ELISPOT and TaqMan™-PCR analysis of biopsies were performed.

Short-term application of RIB5/2 but not FTY720 induced long-term survival of kidney transplants. Moreover, the combination of FTY720 + RIB5/2 prevented tolerance induction. In the combination group serum creatinine levels increased 1 week after cessation of therapy and all rats died from uremia within 72 days. Intragraft immunohistology, ELISPOT and real-time RT-PCR analysis at day 21 demonstrated an enhanced T-cell infiltration and activation but a diminished up-regulation of protective genes in the grafts from recipients receiving the combination therapy. In contrast, delayed application of FTY720 to RIB5/2-treated rats did not interact with RIB5/2-induced tolerance.

In summary, FTY720 is powerful in preventing intragraft infiltration by naive T cells but this might also affect the early development of graft-protecting regulatory T cells and tolerance induction.


The novel immunomodulator FTY720 is a structural analog of myriocin, the metabolite of the ascomycete Isaria sinclarii (1). This drug has pronounced immunosuppressive efficacy in a variety of models of allotransplantation and autoimmune diseases (2) and has been tested in human renal transplantation (3,4). The mode of its immunomodulatory action is related to the reversible redistribution of naive peripheral lymphocytes from the circulation into the lymph nodes and Peyer's patches (5). Recent work demonstrated that FTY720 is phosphorylated in vivo. FTY720-P acts as an agonist at S1P receptors expressed on lymphocytes and endothelial cells which accelerate homing to lymph nodes and Peyer's patches and empty lymphoid sinuses by retention of lymphocytes on the abluminal side of the sinus-lining endothelium (6,7). In contrast to other immunosuppressive drugs, FTY720 does not attenuate lymphocyte function regarding cytokine synthesis and proliferation. Pinschewer et al. (5) showed that FTY720 did not impair humoral immune responses, the generation of primary virus-specific cytotoxic T lymphocytes in lymph nodes and memory T and B cell responses against virus. Several studies demonstrated the reduced inflammatory cellular infiltration of allogeneic grafts under the cover of FTY720 treatment (8,9). FTY720 synergistically acts in combination with calcineurin inhibitors or rapamycin, preventing more efficiently acute rejection and development of chronic graft dysfunction (10,11). Based on these data, the beneficial effects of FTY720 in combination with several immunosuppressive protocols are now under clinical investigation.

The usefulness of FTY720 for tolerance induction protocols is less well studied. Recently, Chueh et al. demonstrated that the combination of FTY720 and allochimeric class I MHC proteins bearing donor-type amino-acid epitope substitutions induced long-term acceptance of WF heart allografts in ACI recipients (12). Recent studies by Bai et al. indicated that lymph nodes or other lymphoid tissues are an important site for peripheral tolerization to alloantigen. Blockade of homing receptors like CD62L prevented CD2/CD3-mediated tolerance in a cardiac mouse model, while permanent FTY720 treatment reversed this effect, suggesting that T lymphocytes use CD62L-dependent migration for alloantigen-specific tolerance (13).

The tolerance-inducing power of nondepleting anti-CD4 monoclonal antibodies (mAb) is well established in various murine transplantation models. However, the cumulative dose that is required for tolerance induction by nondepleting anti-CD4 mAb is quite high (>50 mg/kg). Application of anti-CD4 mAb for clinical tolerance-inducing trials requires more knowledge about the synergistic or antagonistic action of conventional immunosuppressive drugs with the anti-CD4 mAb approach.

The aim of our study was to analyze the influence of FTY720 treatment on anti-CD4-induced tolerance using the nondepleting anti-rat CD4 mAb RIB5/2 and the model of MHC-different high-responder strain combination DA to LEW.

Materials and Methods


Male inbred rats of the Lewis (LEW, RT1l) and the Dark Agouti (DA, RT1av1) genetic background, aged 8–12 weeks, were purchased from Møllegaard Breeding Center Ltd, Ejby, Denmark, and were used for orthotopic DA to LEW kidney transplantation.

Immunosuppressive treatment and kidney transplantation

RIB5/2 (a nondepleting mouse anti-rat CD4 mAb) was purified from ascites by protein A affinity chromatography and the concentration was determined by enzyme-linked immunosorbent assay using IgG2a standards (Sigma, Deisenhofen, Germany) (14,15). FTY720 was kindly provided by Novartis Pharma GmbH, Nürnberg, Germany. For analyzing the effects of FTY720 on the distribution of lymphocyte population in the peripheral blood naive rats were treated with 0.3 mg/kg/d of FTY720 orally for 4 days. In the transplant setting, rats were treated with 0.3 mg/kg/d of FTY720 from days −2 to 14 (or until rejection) or alternatively from days 7 to 21 (or until rejection) and with 10 mg/kg/d (tolerance-inducing protocol) of the anti-CD4 mAb RIB5/2 from days −1 to 3 (i.p.), alone or in combination. Orthotopic kidney transplantation was performed in bilaterally nephrectomized recipients using standard method (16). For analyzing the effect of FTY720 during the stable phase of anti-CD4-induced tolerance this drug was administered in long-term survivors on day 170 after kidney transplantation for 14 days. Serum creatinine and peripheral blood lymphocyte counts were monitored.

Flow cytometry and immunohistology

For flow cytometric analysis of circulating immune cells, heparinized peripheral blood was collected at day 4 and lymphocyte counts were determined. For double-color immunofluorescence we incubated 1 × 106 peripheral blood leukocytes together with the appropriate concentration of the mouse anti-rat T cell receptor (TCR)-α/β-FITC mAb (R73; Caltag, Hamburg, Germany) and the respective R-PE stained mAb (hamster anti-rat-CD62L HRL1, Pharmingen, San Diego, CA; mouse anti-rat-CD45RC mAb OX22, DPC Biermann, Bad Nauheim, Germany) for 30 min at 4 °C. As controls we used a mouse IgG1 FITC + mouse IgG1 R-PE isotype control (Caltag) and a mouse IgG1 FITC isotype control in combination with a Hamster IgG R-PE isotype control (Caltag). The samples were measured by using a FACScan (Becton Dickinson, Heidelberg, Germany). At day 7 after transplantation CD4+CD25+ peripheral blood lymphocytes, peripheral lymph node cells and splenocytes were labeled using an anti-CD4-FITC mAb (RIB5/2, own production) and an anti-CD25 R-PE mAb (OX-39, Caltag) at the appropriate concentration.

At days 4 and 21 animals receiving RIB5/2 and FTY720 alone or in combination were sacrificed for analysis of graft biopsies. An alkaline phosphatase anti-alkaline phosphatase (APAAP) technique was used to analyze cellular graft infiltration as previously described (17). Sections were incubated with mouse antibodies against rat TCR-α/β (R73), CD8-α-chain (Ox8), CD8-β-chain (341), p55 chain of IL-2R (CD25, ART18), granulocytes (RK4), tissue macrophages (ED2), NK cells (10178) and macrophages, monocytes, and dendritic cells (ED1) for 30 min. Hybridoma cells or antibodies were kindly provided by T. Hünig/Würzburg (R73, 341, 10178), T. Diamantstein/Berlin (ART-18, OX8) and BMA Biomedicals AG, Augst, Switzerland (ED1, RK4, ED2).

Quantitative real-time RT-PCR

IL-2, IL-4, IFNγ, TNFα, CD25, bcl-2, bag-1 and HO-1 mRNA levels in kidney biopsies from animals sacrified at day 21 were analyzed using quantitative ‘real-time’ RT-PCR using the TaqMan™ system (Perkin-Elmer Applied Biosystems, Moersheim, Germany) as described in detail, elsewhere (18) (primer sequences Foxp3: forward: 5′-TGGCAAACGGAGTCTGCAA-3′, reverse: 5′-TCTCATCCAAGAGGTGATCTGCTT-3′, probe: 5′-AGCCGGGAGAGTTTCT CAAGCACTGC-3′; CD25 forward: 5′-CACAGTCTGTGTACCAGGAGAACCT-3′, reverse: 5′-CCACGAAGTGGTAGATTCTCTTGG-3′, probe: 5′-CAGGTCACTGCAGGGAGCCCCC-3′). Foxp3 mRNA levels were measured in kidney samples from animals sacrified at day 7 after transplantation. As negative controls we prepared samples without DNA. In order to exclude contaminations with genomic DNA we applied additional negative controls using RT-negative control RNA's amplificating with a cross-reactive primer pair (reacting with both genomic and cDNA). Although all specific primer/probe pairs used are not cross-reactive with genomic DNA, samples with a CT of <35 by using the cross-reactive panel in the RT-negative control were excluded from the analysis. The cycle number at which the reporter fluorescence crosses the threshold (CT value) was used for quantitative measurement. Values are given as 2−delta  CT. The gene of interest expression level is given in relation to the house-keeping gene expression (β-actin) (17).

Enzyme-linked immunosorbent spot assay (ELISPOT) of spleen cells and intragraft lymphocytes

Spleen cells and peripheral lymph node cells were isolated by passing the spleen and the lymph nodes through a metal strainer. The spleen mononuclear cells were obtained after performing standard Ficoll density-gradient centrifugation (Ficoll 1083, Sigma, St. Louis, MO). Intragraft lymphocytes were obtained by homogenizing the grafts. After centrifuging and washing, the pellet was resuspended with 10 mg/mL of collagenase Type IV (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) in RPMI1640 (Biochrom, Berlin, Germany) and incubated for 30 min at 37 °C. The suspension was washed, filtered through a 100-μm strainer and cells were obtained after Ficoll density separation. The ELISPOT was performed as previously described (19) using the Diaclone Research (Besançon, France) rat-IFNγ ELISPOT kit. Cells were cultivated with 3 × 105-irradiated DA spleen cells for 24 h at 37 °C. The results were calculated as IFNγ-producing cells per 300 000 cells.

Statistical analysis

Values are reported as mean ± standard deviation (SD). Group comparisons were performed by the parameter-free U-test and Wilcoxon's test using SPSS software. Survival analysis was obtained by the method of Kaplan-Meier. Differences were considered significant at p < 0.05.


Although FTY720 induces marked lymphopenia and prevents more efficiently graft-infiltration by T cells than anti-CD4 mAb, short-term anti-CD4 mAb but not FTY720 treatment induces long-term graft survival

In naive animals, 0.3 mg/kg/d of FTY720 treatment significantly reduced the number of peripheral blood lymphocytes from 10.9 × 106± 2.6 × 106/mL to 4.2 × 106± 0.7 × 106/mL 1 day after treatment. FACS analysis of the distribution of lymphocyte populations in the peripheral blood of naive rats showed that FTY720 treatment had a marked effect on T cells. However, it reduced the TCR+/CD62L+ and TCR+/CD45RC+ cells to a greater extent than TCR+/CD62L and TCR+/CD45RC cells. This suggests that in the peripheral blood, naive T cells are strongly reduced while T cells of memory/effector type are less targeted (Figure 1A). In contrast, in agreement with recent data, treatment with the nondepleting anti-CD4 mAb did not induce significant lymphopenia. It only modulated the CD4 molecule (data not shown).

Figure 1.

(A) FACS-analysis: Proportion of TCR+, TCR+/CD62L+, TCR+/CD62L, TCR+/CD45RC+ and TCR+/CD45RC cell populations of peripheral lymphocytes after daily treatment of naive rats with 0.3 mg/kg of FTY720 at days 1–3 (n = 5), *p < 0.05 in comparison with the values of day 0 (Wilcoxon-test). (B) Peripheral lymphocyte count after DA to LEW transplantation (each n = 5), *p < 0.05 in comparison with the values before treatment (Wilcoxon's test).

In the transplant situation, peripheral lymphocyte counts were also significantly reduced in animals treated with FTY720 (<25% vs. pretransplantation levels), whereas lymphopenia was hardly detectable (70–85% vs. pretransplantation levels, no significance) following RIB5/2 monotherapy (Figure 1B). Transplantation by itself also moderately affected peripheral lymphocyte counts (80–90% vs. pretransplantation levels, data not shown). Thus FTY720 but not anti-CD4 mAb induced a marked lymphopenia in transplant recipients.

Despite this different action on circulating T lymphocytes, the number of graft-infiltrating T cells was strongly reduced in both groups at day 4 following transplantation (Figure 2). Comparing both therapies, FTY720 had a slightly stronger influence on infiltration by α/β-TCR+, CD8+ and CD8-β-chain+ cells (Figure 2) while RIB5/2 therapy more intensively decreased CD25+ (IL2-R+) cells, ED2+ macrophages and granulocytes. The effects on graft infiltration by ED1+ macrophages and NK cells was absolutely comparable in both groups. These data suggest that anti-CD4 mAb, although a marginally less efficient in preventing graft infiltration by T cells, is more efficient than FTY720 in preventing intragraft immune activation (less granulocytes, ED1+ macrophages, and CD25+ activated T cells). After combination of RIB5/2 and FTY720 we found strongly reduced leukocyte infiltrations within this group. Importantly, the infiltration by T cells and ED1+ macrophages was even lower than in the group treated with FTY720 or RIB5/2 alone. RK4+ granulocytes, ED2+ tissue macrophages, NK cells and CD25+ cells were reduced to the same extent as that observed after single RIB5/2 treatment.

Figure 2.

Cellular infiltration of transplant biopsies at day 4 after DA to LEW transplantation and 0.3 mg/kg of FTY720 monotherapy, 10 mg/kg of RIB5/2 monotherapy and 0.3 mg/kg of FTY720 + 10 mg/kg of RIB5/2 combination therapy (each n = 3). Six-μm tissue sections were incubated with R73 (anti-α/β TCR), OX8 (anti-CD8 α-chain), 341 (anti-CD8 β-chain), RK4 (anti-granulocyte-membrane protein), ED2 (anti-tissue macrophage membrane protein), ED1 (anti-macrophage membrane protein), 10178 (anti-NK cell membrane protein) and ART18 (anti-IL-2R/CD25). MAb binding was visualized by the alkaline phosphatase anti-alkaline phosphatase (APAAP) technique.

The differences in early intragraft events after RIB5/2 and FTY720 monotherapy are reflected by the distinct effects on graft survival. RIB5/2 at 5 × 10 mg/kg given alone resulted in seven out of nine rats in permanent graft survival (>100 d) while FTY720 therapy prolonged graft survival in this high responder strain combination for 1 day only [mean survival time (MST) 7.0 d, n = 7] compared with untreated controls (MST 6.2 d, n = 5) (Figure 3A).

Figure 3.

(A) RIB5/2 therapy given alone resulted in seven out of nine rats in permanent graft survival (>100 d) while FTY720 therapy prolonged graft survival for 1 day [mean survival time (MST) 7.0 d, n = 7] compared with untreated controls (MST 6.2 d, n = 5). The group receiving combination therapy survived for 45.2 d only (n = 5). (B) Serum creatinine values were significantly increased from day 21 after combination therapy FTY720 + RIB5/2 (n = 5) in comparison with controls receiving 10 mg/kg of RIB5/2 therapy (n = 9) (U-test, p < 0.05).

FTY720 antagonizes the tolerance-inducing capacity of anti-CD4 mAb RIB5/2

Moreover, FTY720 prevented tolerance induction by RIB5/2 mAb. All graft recipients receiving combination therapy developed increased serum creatinine levels after day 21 (Figure 3B) and died by uremia within 72 days (mean 45.2 days) (Figure 3A).

Next we wondered whether different intragraft activation might explain the graft rejection in the combination group. As serum creatinine levels increased after day 21 in the FTY720 + RIB5/2 group, we sacrified some rats of the two RIB5/2 mAb groups (with/without FTY720) on day 21 for intragraft analysis.

Histological analysis of these biopsies revealed obvious signs of acute rejection in the FTY720 + anti-CD4 mAb combination group only, with significantly higher numbers of α/β-TCR+ cells, CD8+ cells, CD8-β-chain+ cells, CD25+ cells, ED1+ and ED2+ macrophages, and NK cells compared with RIB5/2 therapy alone (Figure 4). Additionally, TaqMan™-PCR examinations at day 21 after transplantation showed significantly elevated IL-2, IL-4 and IFNγ and a moderately enhanced TNFα and CD25 expression within the grafts of the group receiving combination therapy (Figure 5). Even if taking into account the higher T-cell infiltration in this group, the levels of IL-2, IFN-γ, and IL-4 were also significantly up-regulated if corrected by T-cell infiltration, suggesting more T-cell activation in general as well as per infiltrating cell. Interestingly, the expression of protective genes such as bcl-2 and bag-1 was inhibited in the combination group (p < 0.01 and p = 0.07, respectively).

Figure 4.

At day 21 after transplantation histological evaluation of the graft revealed significantly higher numbers of α/β-TCR+ cells, CD8+ cells, CD8-β chain+ cells, IL2-R+ cells, macrophages, tissue macrophages and NK cells in the group receiving FTY720 + RIB5/2 combination therapy (n = 4) compared with RIB5/2 therapy alone (n = 4) (U-test, p < 0.05).

Figure 5.

TaqMan™-PCR examinations at day 21 after transplantation showed significantly elevated IL-2, IL-4 and IFNγ but decreased bcl-2 expression within the grafts of the group receiving FTY720 + RIB5/2 combination therapy (n = 4) compared with RIB5/2 therapy alone (n = 4) (U-test, p < 0.05).

Moreover, ELISPOT analysis of intragraft lymphocytes revealed approximately threefold higher frequencies of spontaneous IFNγ secreting T cells in the combination group compared with the RIB5/2 group (p < 0.05) (Figure 6B). Following donor-antigen stimulation, the frequencies of IFNγ secreting T cells further increased, reaching 278 ± 33 and 168 ± 57 IFNγ secreting T cells per 300 000 intragraft cells in the FTY720 + anti-CD4 mAb group and anti-CD4 mAb group, respectively (p < 0.05). The enhanced frequencies of intragraft spontaneously active and donor-reactive T cells in the FTY720 group at day 21 are comparable to the frequencies seen in untreated allograft recipients at day 4: the time point of maximal graft infiltration in untreated animals (235 and 336 IFNγ secreting T cells/300 000 cells, respectively).

Figure 6.

ELISPOT assay for IFNγ-producing cells: (A) splenocytes, peripheral lymph node cells at day 7 and (B) splenocytes and graft-infiltrating cells at day 21 of rats treated with FTY720 + RIB5/2 (n = 4), RIB5/2 monotherapy (n = 4) were isolated. Control wells contained responder cells plus medium alone. Cells were tested against donor antigen by adding 3 × 105 irradiated DA spleen cells, and incubated for 24 h. Results are represented as IFNγ-producing cells/300 000 leukocytes ± SD (*U-test, p < 0.05). For comparison: nontransplanted rats showed <30/300 000 spontaneous IFNγ producers in the spleen and lymph nodes; untreated transplanted rats expressed 250 and <30 spontaneous IFNγ producers and 336 and 273 donor antigen-induced IFNγ producers/300 000 intragraft cells and splenocytes, respectively.

Taking into account the higher T-cell infiltration in the combination group (FTY720 + anti-CD4 mAb), the data demonstrate a sixfold and fourfold higher frequency of spontaneous and donor-reactive IFNγ secreting T cells, respectively, in the grafts of these rats compared with the anti-CD4 mAb treatment alone (p < 0.01). Interestingly, in the spleen taken at day 21 similar frequencies of IFNγ-producing cells were detectable in both groups (Figure 6B).

FTY720 interacts with very early events of anti-CD4 mAb-mediated tolerance induction

Next we wondered whether a delayed application of FTY720 might also interact with tolerance induction in our model. If FTY720 therapy (0.3 mg/kg/d) is not started perioperatively but delayed at day +7 after transplantation and maintained for 14 days, all rats survived, showed no increase in serum creatinine and tolerance was stable during the 90-day follow up (data not shown).

In agreement with these observations, oral application of 14 × 0.3 mg/kg/d of FTY720 into anti-CD4 mAb-mediated tolerant animals at day 170 after transplantation did not break tolerance despite reversible decreased peripheral blood lymphocyte numbers (day 0: 8.95 ± 1.63 × 106/mL, day 2: 1.83 ± 0.24 × 106/mL, day 50: 6.68 ± 0.8 × 106/mL, n = 4) in comparison with the control group without FTY720 treatment (p < 0.01). We observed no influence on survival and serum creatinine of these recipients neither during the treatment phase nor after withdrawal of the drug within the 100-day follow-up period (data not shown).

These data emphasize the role of early events leading to tolerance induction and support the hypothesis that FTY720 interferes with early events of anti-CD4-mediated tolerance induction. In order to prove this hypothesis, we studied the influence of FTY720 given in parallel with anti-CD4 mAb on early immunological events after transplantation.

First, we analyzed the frequencies of donor-specific T cells in the spleen and in the peripheral lymph nodes by ELISPOT (Figure 6A). At day 7 after transplantation spontaneous IFNγ producing T cells were significantly increased in both lymph nodes (35–60/300 000 cells) and particularly spleen (90–100/300 000 cells) compared with nontransplanted controls (<30/300 000), however, the two anti-CD4 mAb groups with or without FTY720 were comparable (Figure 6A). Stimulation with donor-antigen further increased the T-cell response. Interestingly, in the spleen of animals treated with RIB5/2 + FTY720 combination therapy higher frequencies of donor-reactive T cells were detectable in comparison with RIB5/2 controls (p < 0.05). In contrast, there were no differences in the number of IFNγ-producing alloreactive T cells within the peripheral lymph nodes of these groups.

These data show that both protocols do not effectively deplete donor-reactive T cells from secondary immune organs, suggesting a role of regulatory T cells (Treg) in maintaining/inducing immune tolerance of anti-CD4 mAb-treated recipients. Very recently, we were able to demonstrate the generation of intragraft Treg within 7 days after anti-CD4 mAb treatment while detection of these cells in the spleen required 3–4 weeks (30). Therefore, we speculated that FTY720 interacts with the generation of early Treg.

To prove this hypothesis, we analyzed the relative and absolute number of CD4+CD25+ T cells within the blood (Figure 7C), spleen and peripheral lymph nodes (Figure 7B) by flowcytometry at day 7 after transplantation. Comparing both groups we detected no significant differences within the blood and the spleen. Interestingly, the peripheral lymph nodes of RIB5/2 + FTY720-treated animals showed a significantly higher relative and absolute number of CD4+CD25+ T cells than the control group (p < 0.05). Thus the frequencies of donor-reactive T cells and CD4+25+ T cells were inversely influenced by FTY720 in spleen and lymph nodes from anti-CD4 mAb-treated recipients (Figures 7B and 6A). If FTY720 redirects Treg to lymph nodes, less Treg should be detectable in the graft. CD25 expression is less useful for intragraft analysis of Treg, as it cannot discriminate between recently activated alloreactive-effector T cells and Treg. Foxp3 is a novel transcription factor which is specifically expressed by Treg. Therefore, we quantified the intragraft Foxp3 mRNA expression by using Taqman™-PCR at day 7 after transplantation. Indeed, we found significantly higher copy numbers of Foxp3 in kidney allografts from animals treated with RIB5/2 alone (p < 0.05). After combination therapy Foxp3 mRNA expression was approximately fourfold lower than in the anti-CD4 mAb control group (Figure 7A). These results show that the generation within or migration of Treg to the allograft is impaired after RIB5/2 + FTY720 combination therapy.

Figure 7.

(A) TaqMan™-PCR examinations at day 7 after transplantation showed significantly elevated levels of Foxp3 as a marker of Treg within allograft biopsies of RIB5/2 (n = 5)-treated animals in comparison with the group receiving RIB5/2 + FTY720 combination therapy (n = 7) (U-test, p < 0.05). (B) FACS-analysis of CD4+CD25+ T cells within the spleen, peripheral lymph nodes and C: blood revealed a significantly higher percentage of these cells in the peripheral lymph nodes of RIB5/2 + FTY720 (n = 4)-treated animals in comparison with animals treated with RIB5/2 alone (n = 5) (U-test, p < 0.05).


FTY720 is a new and potent immunosuppressive drug that causes in various animal models a rapid, reversible reduction of peripheral blood lymphocytes by inducing their migration to secondary lymphoid organs. It prevents acute rejection in several transplant models (8,20). Tolerance induction is the holy grail of transplant research. During recent years several promising protocols for tolerance induction have been developed in animal models, inducing macrochimerism, blocking signal 2, modifying signal 1, and re-educating T cells in the presence of the graft following extensive perioperative T cell depletion. Modifying signal 1 by nondepleting anti-CD4 mAb has been shown to be powerful in various transplant models in rodents. However, the cumulative dose that is required for anti-CD4 mAb-induced tolerance is quite high (>50 mg/kg). So the search for drugs that synergize with anti-CD4 mAb induction is important for the successful application of this principle in patients. It has been shown by several groups that calcineurin inhibitors and antiproliferative drugs (21–23) that interact with T-cell activation and clonal expansion prevent tolerance induction in several models. In contrast, FTY720 does not interact with T cell activation but only with homing. So we speculated that short-term treatment of transplant recipients with FTY720 during the early postoperative phase where priming of alloreactive T cells occurs might improve anti-CD4 mAb-based tolerance-inducing protocols. Unfortunately, the opposite is the case. In contrast to anti-CD4 mAb, short-term treatment with FTY720 only marginally prolongs graft survival in a high-responder rat kidney transplant model. Although FTY720 but not RIB5/2 induced marked lymphopenia in peripheral blood, intragraft analysis at day 4 showed similar infiltration by immune cells during treatment. Moreover, increased levels of CD25+ lymphocytes, ED2+ macrophages, and granulocytes suggest more intragraft inflammation in the FTY720 group compared with the RIB5/2 group. These observations on enhanced early graft infiltration are surprising as other reports have shown improvement of ischemia/reperfusion injury by FTY720 (24). This might be owing to the high-responder strain combination we used in our model. Most importantly, 14-day FTY720 treatment completely prevented anti-CD4 mAb induced tolerance. Within 1 week following cessation of FTY720 serum creatinine levels increased (after day 21), and all animals receiving the combined induction by FTY720 + RIB5/2 mAb, died within 72 days by uremia, whereas RIB5/2 alone induced long-term survival in 7 of 9 rats.

What are the mechanisms for antagonizing tolerance induction? First, we wondered whether FTY720 supports priming of alloreactive T cells in secondary immune organs. Indeed, there was a significantly higher frequency of donor-specific IFNγ-producing splenocytes of the combination group compared with the anti-CD4 mAb monotherapy at day 7 but the difference was not longer significant at day 21 (Figure 6). In contrast, there was no accumulation of donor-reactive T cells in the lymph nodes of FTY720 + anti-CD4 mAb vs. anti-CD4 mAb-treated rats. This ‘split’ phenomenon might be explainable by the reversed accumulation of CD4+25+ T cells in the lymph nodes but not in the spleen of FTY720-treated recipients that can down-regulate the effector alloresponse. After cessation of FTY720 (and anti-CD4 mAb) the alloreactive T cells are no longer trapped in the immune organs and can migrate into the graft. Intragraft analysis at day 21, the day of ongoing rise of serum creatinine, revealed significant differences between the two groups: anti-CD4 mAb-treated rats showed both relatively and absolutely less intragraft donor-reactive T cells than untreated rejecting controls, whereas addition of FTY720 abolished this effect. Immunohistology confirmed the enhanced immune cell infiltration in the combination group, demonstrating an intragraft accumulation of all immune cell types. In addition, the enhanced expression of many cytokine genes, particularly of IL-2, IL-4, and IFNγ, supports the strong intragraft immune activation of the combination group. Moreover, graft protective genes, like bcl-2, were up-regulated in the anti-CD4 mAb treatment group only. Up-regulation of these protective genes is seen in association with infiltration by Th2 and regulatory T cells (Treg) (25).

The dramatic intragraft differences between the combination group and the anti-CD4 mAb monotherapy group at day 21 is difficult to explain by the higher frequency of donor-reactive splenocytes alone. Recent data demonstrate that in RIB5/2 mAb-induced tolerance nondeletional, anergy-like regulatory mechanisms may operate (26). The moderate effects on the frequency of peripheral alloreactive T cells by RIB5/2 mAb, as shown here, support this view. As FTY720 does not interact with T cell signaling, prevention of anti-CD4 mAb-induced tolerance might be owing to interaction with the generation of Treg. Natural CD4+CD25+ Treg are able to prevent autoimmunity but cannot prevent allograft rejection following adoptive transfer. In contrast, antigen-activated CD4+ Treg generated under the umbrella of anti-CD4 mAb treatment (or other drugs like CTLA4Ig) are able to adoptively transfer tolerance even in high-responder strain combinations (27–29).

Very recently, we observed that the graft by itself plays a dominant role in inducing antigen-driven Treg. Even from rejecting grafts, tolerance-mediating Treg can be isolated (29,30; Tullius et al., submitted) suggesting that direct contact between graft and regulatory precursor T cells is required for their effective activation. Thymectomy before but not after transplantation also interacts with tolerance induction and generation of Treg in RIB5/2 mAb-treated rats, suggesting that these precursor Treg are derived from recent thymus emigrants (31).

As FTY720 preferentially targets circulation of naive T cells (32; Bold et al., in preparation), it might interact with the early graft-infiltration and activation of these precursor Treg. If tolerance is established and antigen-driven Treg are generated, FTY720 can no longer interact with the tolerance maintenance process as shown here. ‘Effector’ Treg express the phenotype of memory/effector T cells and FTY720 does only moderately interact with the homing of these cells (Bold et al., in preparation). The lack of tolerance prevention following delayed application of FTY720 starting as early as at day 7 after transplantation demonstrates that FTY720 interacts with very early events of tolerance induction. Interestingly, FTY720 + anti-CD4 mAb-treated rats showed an accumulation of CD4+25+ T cells in the lymph nodes compared with the anti-CD4 mAb monotherapy, suggesting a redirection of Treg from the graft to the lymph nodes. The fourfold lower expression of Foxp3, a transcription factor associated with Treg differentiation (33,34), in the FTY720 + anti-CD4 mAb vs. anti-CD4 mAb group support this view.

In summary, FTY720 interacts with very early events of tolerance-induction, at least if the CD4 molecule is targeted. The prevention of early T cell infiltration and intragraft generation and/or migration of Treg by FTY720 might explain the tolerance-preventing effect.

This effect is in contrast to recent reports suggesting an usefulness of FTY720 for tolerance induction in other models (12,13). However, these models use other principles of tolerance induction (targeting the T cell receptor by modified MHC molecules or CD2/3 mAb). Depletion of alloreactive T cells might play a greater role than regulation in these models. On the other hand, these two groups used a heart transplantation model in mice and we studied rat kidney transplantation. Very recently, Koshiba et al. showed that FTY720 interacts with tolerance induction following intestinal but not heart transplantation, suggesting organ differences (35). Our data show that putative interactions of immunosuppressive drugs with tolerance induction protocols have to be carefully proved for new combinations and applications.


We thank Annett Kott and Sigrid Giers for excellent technical support. This project was partly supported by Novartis Pharma AG, Basel, Switzerland, and by BMBF, NBL3, FKZ 01ZZ0108, Germany.