|Home | About | Journals | Submit | Contact Us | Français|
Involvement of Treg in transplant tolerance has been demonstrated in multiple models. During the active process of graft rejection, these regulatory cells are themselves regulated and inactivated, a process termed counter-regulation. We hypothesize that ligation of the costimulatory molecule glucocorticoid-induced TNF receptor-related protein (GITR) on Treg inhibits their ability to promote graft survival, and by blocking GITR ligation graft survival can be prolonged. To this aim, we have designed a soluble GITR fusion protein (GITR-Fc), which binds GITR ligand and inhibits activation of GITR. Here, we show that GITR-Fc prolonged mouse skin graft survival, and this prolongation is dependent on Treg. In a full MHC-mismatched skin graft setting, GITR-Fc significantly improved graft survival when used in combination with MR1, anti-CD40L, while GITR-Fc alone did not demonstrate graft prolongation. These results demonstrate that disruption of binding of GITR with GITR ligand may be an important strategy in prolonging allograft survival.
Sub-optimal long-term graft survival rates and the need for chronic immunosuppression to sustain survival in virtually all cases of solid organ allotransplantation continue to undermine success in clinical organ transplantation. To improve graft survival rates and eliminate the need for immunosuppression , donor-specific transplant tolerance must be achieved.
The application of Treg is an area of intense focus in the field of transplant biology. The naturally occurring population of CD4+CD25+ Treg maintains both self-tolerance and transplant tolerance by suppressing T-cell responses both in vivo and in vitro . Phenotypically, Treg constitutively express the cell surface molecules CTLA-4, certain toll-like receptors, CD103 (αE integrin), and glucocorticoid-induced TNF receptor-related protein (GITR) at high levels [3-7], and they specifically express the transcription factor Foxp3 [8, 9].
GITR is a type I transmembrane protein with high homology to other members of the TNFR family, especially to 4-1BB and CD27 [4, 7]. GITR is preferentially expressed by Treg; however, like CD25, GITR is also upregulated in conventional effector T cells (Teff) upon activation . In a screen to identify proteins specific to CD25+ Treg, Sakaguchi’s group identified a GITR-specific antibody that could block the ability of Treg to suppress other cells . When this agonistic antibody, DTA-1, is added to a suppression assay in which Treg are co-cultured with CD25− T cells, the Treg no longer have suppressor ability, whereas addition of antibodies to CD40 and to other TNFR family members such as CD27 and OX40 did not have this effect [4, 5]. Isolated GITR+ T cells, regardless of expression of CD25, could regulate the mucosal immune responses and prevent development of colitis in an adoptive transfer model, whereas administration of anti-GITR antibodies led to the induction of inflammatory bowel disease . In a mouse transplant model, DTA-1 reverses Treg-mediated regulation resulting in acute allograft rejection .
GITR ligand (GITRL) is expressed on endothelial cells, B cells, macrophages, and bone-marrow-derived dendritic cells [4, 11]. GITRL expression is upregulated on APC by toll-like receptor (TLR) stimulation but is downregulated within 48 h in vitro [4, 11]. We have previously demonstrated that GITRL is upregulated by dendritic cells in the draining lymph node within 1 wk of a skin graft procedure, and we have suggested that this upregulation results in the inactivation of Treg .
In an in vivo transplantation model, adoptive transfer of antigen-specific Treg can suppress skin graft rejection by Teff when co-transferred in the absence of inflammation. Established skin grafts are protected by Treg. In contrast, the acute inflammatory setting surrounding the skin graft procedure can inactivate the function of these Treg, and this inactivation results in rejection of the skin graft . Our data demonstrate in the period immediately following the skin transplant procedure, blockade of GITR–GITRL interaction maintains this Treg-dependent tolerance which allows the skin graft to survive indefinitely in most cases.
GITRL is expressed on APC, is upregulated with APC activation, and disrupts Treg function through its binding to GITR [4, 5]. As GITR–GITRL ligation prevents the ability of Treg to promote graft survival, we targeted this interaction by generating and purifying a soluble GITR-Fc fusion construct, as soluble TNF family members have been demonstrated to disrupt receptor–ligand interaction [4, 5]. One hundred and fifty-three amino acids of the extracellular domain of mouse GITR were cloned into the pMT/ Bip/V5 vector containing the human IgG Fc fragment (Fig. 1A). This construct was stably transfected into Drosophila Schneider S2 cells then purified from culture supernatants on protein A-sepharose beads (Fig. 1B).
To demonstrate that GITR-Fc binds to GITRL, 293 cells were stably transfected with GITRL and then incubated with increasing doses of GITR-Fc. Binding was analyzed by flow cytometry using anti-human IgG secondary antibody. GITR-Fc demonstrates a dose-dependent increase in binding to transfected cells while analysis with control human IgG and secondary antibody shows no specificity in binding (Fig. 1C).
In a previously described adoptive transfer model of graft rejection, HA-specific TS1 TCR transgenic lymphocytes reject HA-bearing skin grafts . When Teff and Treg are adoptively transferred 30 days after the skin graft procedure, Treg suppress graft rejection by Teff in the absence of inflammation resulting from the procedure. However, when adoptive transfer occurs shortly before or during the acute period after the graft procedure, Treg are counter-regulated by the inflammation surrounding the procedure (Fig. 2A, control). We hypothesize that GITR–GITRL interaction is involved in this counter-regulatory process .
To evaluate whether GITR-Fc blocks GITR–GITRL binding and promotes prolongation of acute skin grafts, transgene-negative BALB/c hosts bearing HA104 skin grafts received 5 × 105 (TS1xHA28)F1 lymph node cells (LNC) 1 day after transplant. These LNC contain a roughly 1:1 mixture of TCR transgenic CD4+CD25−Foxp3− Teff: CD4+CD25+Foxp3+ Treg. Recipient mice then received three doses of GITR-Fc or control human IgG antibody (100 μg at days 0, 4, and 10 post-adoptive transfer) intraperitoneally to block the GITR–GITRL interaction in vivo. In contrast to control mice, all mice injected with 100 μg GITR-Fc, grafts survived longer than 90 days (Fig. 2A). Mice injected with 500 μg GITR-Fc did not demonstrate significant prolongation of graft survival (data not shown). Data suggest that GITR-Fc is able to maintain Treg function and prolong graft survival.
To determine the cellular target of GITR-Fc, we repeated these experiments instead using 5 × 105 TS1 LNC in the absence of antigen-specific Treg. In mice with acute skin grafts and injected with TS1 cells, GITR-Fc treatment resulted in no prolongation when compared with mice injected with control IgG antibody (Fig. 2B). These data demonstrate that the ability of GITR-Fc to prolong graft survival is dependent on antigen-specific Treg and suggest that GITR-Fc acts on this Treg population. Moreover, this data also demonstrate that GITR-Fc is not a general immunosuppressant as mice are still able to reject their grafts. Our data do not completely exclude the possibility that GITR-Fc has a different effect on Teff from TS1 than from (TS1xHA28)F1 mice; however, detailed comparison of these cells has shown no difference in phenotype thus far.
In a highly stringent skin graft model, it was recently demonstrated that donor-specific transfusion in combination with anti-CD40L and anti-CD45RB uniformly achieves >90-day survival . This combination therapy was demonstrated to generate Treg and to be dependent on Treg. To demonstrate that GITR-Fc can prolong skin grafts in an aggressive allograft rejection response, we used the reagent in combination with these other costimulatory blockade antibodies. C57BL/6 mice were transplanted with BALB/c skin and treated with GITR-Fc (100 μg days 0, 4, and 10) with or without anti-CD40L (250 μg i.p. days 0, 2, 4, 6, and 8) and anti-CD45RB (100 μg i.p. days 0, 1, 3, 5, and 7). Neither anti-CD40L nor GITR-Fc treatment alone prolongs graft survival. When anti-CD40L and GITR-Fc are used together, graft survival is significantly prolonged (MST = 19 days treated versus 13 days control; p<0.0001) compared with mice injected with control antibodies, and graft survival is significantly prolonged when compared with mice treated with anti-CD40L alone (MST = 14.5 days; p<0.005) (Fig. 3). The median survival time of anti-CD40L+anti-CD45RB treatment was 87 days (n = 9), which is similar to that previously reported ; however, GITR-Fc in combination with those two antibodies did not significantly extend skin graft survival (MST = 94 days, n = 13, data not shown). GITR-Fc treatment did not significantly extend graft survival of anti-CD45RB-treated mice (data not shown). In combination with anti-CD40L, GITR-Fc significantly prolongs skin graft survival.
Skin, lung, and intestine, unlike heart, pancreas, and kidney allografts, are organs considered more resistant to tolerance induction. One common feature of these resistant organs is their continual exposure to environmental antigens and commensal microbes, which are believed to activate the innate immune system. This activation of the innate immune system both may promote graft rejection and prevent tolerance induction. With this in mind, inhibition of the innate immune system should promote tolerance to transplants such as skin allografts. For example, inhibition of TLR signaling while at the same time blocking costimulation (anti-CD154 with or without CTLA4Ig) led to successful skin allograft acceptance and also correlated with intragraft recruitment of Foxp3 Treg .
GITRL is transiently upregulated upon inflammation and innate immune signaling such as TLR activation [4, 11, 12, 16]. Binding of GITRL on APC to GITR on newly activated T cells breaks suppression and allows the Teff to resist the effects of Treg. In vivo administration of the agonist GITR antibody, DTA-1, results in autoimmune disease in otherwise normal mice and exacerbates autoimmune disease in susceptible mice.
Descriptions of several GITR–GITRL ligand-blocking reagents have been published [11, 17]. The in vivo therapeutic effects of blocking GITR–GITRL interaction have been striking. For example, treating animals with GITR-Fc fusion protein has been shown to ameliorate inflammation [18, 19], and injection into an experimental inflammatory bowel disease model prevented the disease .
Transduction of a signal by rGITR through GITRL has been described, and this “reverse signaling” has been demonstrated with other family members, namely OX40, CD40, CD30, and 4-1BB . However, data from administration of our reagent in vivo are consistent with the idea that it blocks GITR signaling. Nocentini et al. demonstrate that in vivo GITR-Fc has an effect on ameliorating inflammation in a spinal cord injury model in GITR+/+ mice, but the benefit of this reagent was absent when used in GITR−/− mice . This evidence suggests that GITR-Fc functions by blocking GITR triggering and not due to GITRL activation. While it is possible GITR-Fc activates GITRL, published data suggest that such activation results in a proinflammatory response, which is contrary to our results .
Treg-mediated control of immunity remains an attractive alternative to current immunosuppressive therapies for prolongation of allograft survival, but application of these cells to a more complex transplantation setting will require further systematic characterization. These data contribute to the growing knowledge that interfering with GITR–GITRL interaction mitigates many deleterious effects of the immune response by acting on APC or Teff. Our data suggest that tolerance induction by natural Treg can also be fostered by blockade of GITR signaling. The elucidation of factors that regulate Treg function in vivo is crucial given the extensive inflammatory response accompanying activation of graft-specific lymphocytes. We provide evidence that innate immune signals induced early in the response to an allograft may interfere with Treg-mediated suppression and thereby abrogate tolerance. This finding has relevant implications in light of the interest in transferring Treg to gain tolerance in the clinical setting as our results suggest that this strategy may be thwarted by natural process of counter-regulation that accompanies initiation of immunity.
The cDNA encoding the extracellular domain amino acids 1–153 was subcloned into the Pmt/Bip/V5 vector containing the Fc fragment of human IgG1. The plasmid pMT-GITR-Fc was stably transfected into Drosophila Schneider S2 cells according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA), and supernatants were collected. GITR-Fc was affinity-purified on protein A-Sepharaose (Pharmacia, Piscataway, NJ, USA) and aliquots were stored at −70°C.
TS1 transgenic mice possess a high frequency of CD4+ T cells specific for the immunodominant epitope of the influenza PR8 virus HA protein in the context of MHC class II I-Ed . HA28 mice and HA104 mice are previously described . Briefly, HA104 mice provide a source of HA-expressing grafts as they carry the full-length HA transgene controlled by the SV40 early region promoter/enhancer which results in ubiquitous transgene expression . HA28 mice carry a truncated HA (amino acids 1–237) also controlled by the SV40 early region promoter/ enhancer, which results in ubiquitous expression. (TS1xHA28)F1 mice provide a source of HA-specific Treg . TS1, HA28, and HA104 transgenic lines are maintained as hemizygotes back-crossed onto BALB/c background (Jackson Laboratory, Bar Harbor, ME, USA). All animals are maintained in a pathogen-free environment under IACUC approved protocols. C57BL/6 mice were purchased from Jackson Laboratory.
293T cells were stably transfected with pMTF-GITRL. 293-MTF-GITRL cells were incubated with GITR-Fc and binding was assessed by flow cytometry using a PE-labeled anti-human IgG1 antibody (CALTAG, Burlingame, CA, USA).
Lymph node single cell suspensions were prepared in FACS buffer (PBS+3% HI-FBS+0.1% NaN3), and flow cytometric analysis was performed on a Becton Dickinson (San Jose) FACSCalibur cytometer.
For FACS, cells were stained with anti-CD25 FITC, anti-CD45RB PE, and anti-CD4 APC and sorted on a BD FACSVantage SE high-speed cell sorter. Sorted populations were gated on CD4+, CD25+, and CD45RB-intermediate. Purity checks on the sorted population ranged from 94 to 99%.
Skin transplantation was performed as described previously . Bandages were removed on the tenth day. Grafts were scored as rejected when more than 75% of the grafted tissue area had been lost. Described antibodies were purchased from Bio-X (New Lebanon).
Survival data were compared with the Kaplan–Meier method and analyzed by the log-rank test. For normally distributed data, Student’s t-test was applied. p-values less than 0.05 were considered significant.
This research was made possible by NIH grants: R01 AI-048820 (J. F. M.), K01 DK079207-02 (J. I. K.), and T32DK007006-33 (S. B. S.).
Conflict of interest: The authors declare no financial or commercial conflict of interest.