Calcineurin Inhibitors Block MHC-Restricted Antigen Presentation In Vivo¹

The T lymphocytes recognize Ags in the form of short peptides presented by MHC molecules. In the classical paradigm of Ag presentation by professional APCs, endogenous Ags are presented via class I MHC molecules to CD8⁺ T cells, whereas exogenous Ags are presented via class II MHC molecules to CD4⁺ T cells (1). Exceptions abound, however, and even predominate in certain situations. Exogenous Ags can be presented on class I MHC molecules via several mechanisms. This process, termed cross-presentation, was first demonstrated in the generation of CTL responses to minor histocompatibility Ags (2). It is now recognized as a more general mechanism for the generation of CTL responses to tumor cells, transplanted cells, bacteria, and even viruses (3, 4, 5, 6, 7, 8).

Although Ag presentation represents half of the process of tissue rejection, it has received little attention as a target for therapeutic immunosuppression. The principal drugs used to prevent transplantation rejection are the calcineurin inhibitors cyclosporin A (CsA)³ and tacrolimus. Even though chemically unrelated, CsA and tacrolimus bind to and inhibit the same protein, calcineurin, in a calcium-dependent signaling pathway after the formation of a complex with cyclophilin A and FK-506 binding proteins, respectively (9, 10). These drugs are generally believed to act strictly by their effects on T cells (11). However, we recently showed that CsA and tacrolimus, but not rapamycin, inhibit class I and class II MHC-restricted Ag presentation pathways in DCs in vitro (12). Because T cells can only recognize Ags presented on MHC molecules, the impact of the inhibition of MHC-restricted Ag presentation is far-reaching.

In the present study, we examined the in vivo relevance of the inhibition of the MHC-restricted presentation of exogenous Ags by calcineurin inhibitors. We found that these inhibitors efficiently block class I and class II MHC-restricted exogenous Ag presentation in mice.

T cell hybridomas, B3Z86/90.14 (B3Z) and DOBW, were provided by Dr. Nilabh Shastri (University of California, Berkeley, CA) and by Dr. Clifford V. Harding (Case Western Reserve University, Cleveland, OH), respectively (13, 14). The DC cell line (DC2.4) was obtained from the Dana-Farber Cancer Institute, Boston, MA (15).

Male 8- to 12-wk-old C57BL/6 and BALB/c mice were purchased from Orient. Mice were used according to the protocols approved by the Animal Care Committee of Chungbuk National University.

Tacrolimus (Prograf) was injected (s.c.) into C57BL/6 mice once at 0.15 mg/Kg, which was followed by maintenance doses of 0.06 mg/Kg every 6 h. OVA, which was used as an exogenous Ag, was injected (i.v., 10 mg/mouse) 2 h after the initial tacrolimus injection.

Lymph nodes (popliteal, inguinal, mesenteric, and axillary) and spleens were collected and pooled, from ≥10 mice per treatment group. Single cell suspensions were prepared by gentle disruption after treating organs with 2 mg/ml collagenase (Sigma-Aldrich) for 1 h at 37°C, and were then layered on a 62% Percoll (Sigma-Aldrich) step gradient, centrifuged for 20 min at 1750 g (3,000 rpm), and DC-enriched layers were collected. Further purification was performed using a negative selection method using a mouse DC enrichment kit (StemSep) as instructed by the manufacturer. DC purity was 78–84% as determined by cell percentages expressing both CD11c and I-A^b by flow cytometry.

LacZ T cell activation assays were used to assess the amounts of cross-presented OVA peptides, as previously described (12, 13). In brief, APCs (as indicated in the appropriate figure legends) were added to a 96-well microtiter plate (1 × 10⁵/well), and the plate was incubated for 1 h at 37°C. The plate was then washed twice with 300 μl/well of prewarmed PBS, and then fixed with 100 μl/well of ice-cold 1.0% paraformaldehyde for 5 min at room temperature. The plate was washed three times with 300 μl/well of PBS, and B3Z cells were added (2 × 10⁵/well). After incubating for 4 h at 37°C, lacZ activity was measured either by X-gal (5-bromo-4-chloro-3-indoyl-β-d-galactopyranoside; Sigma-Aldrich) staining (12), or by colorimetric analysis after incubating freeze-thaw lysed cells with β-galactosidase substrate, chlorophenol red β-d-galactopyranoside (Calbiochem) as described previously (12).

Class II MHC-complexed OVA peptides quantities were assessed as previously described (12). In brief, APCs (as indicated in the appropriate Fig. legends) were added to a 96-well microtiter plate (1 × 10⁵/well), and the plate was incubated for 1 h at 37°C. The plate was then washed twice with 300 μl/well of prewarmed PBS, and then fixed with 100 μl/well of ice-cold 1.0% paraformaldehyde for 5 min at room temperature. The plate was washed three times with 300 μl/well of PBS, and DOBW cells were added (1 × 10⁵/well). After incubating for 48 h at 37°C, the plate was centrifuged at 400 g, and the culture supernatants were collected and assayed for IL-2 content using IL-2 ELISA kits (BD Biosciences).

Microspheres were prepared using a solvent-evaporation method, as described previously (12). Microspheres containing OVA and CsA were prepared by adding CsA (0.1 - 1 mg/ml; Calbiochem) to the poly(DL-lactide-coglycolide) (Sigma-Aldrich) solution. The concentration of OVA was determined by microbicinchoninic acid assay kit (Pierce) according to the manufacturer’s instructions after lysing the microspheres in a lysis buffer containing 0.1% SDS and 0.1 NaOH. CsA concentrations were determined by high performance liquid chromatography using a Nucleosil C18 column (2.5 × 250 mm, 10 μm) after lysing the microspheres in methanol.

C57BL/6 mice were immunized (i.v.) with microspheres containing OVA, CsA, or both CsA and OVA. In some experiments, mice were also immunized (i.v.) with DCs (5 × 10⁵/mouse) that had been initially treated with mitomycin C and then pulsed with OVA_257–264 peptide. One week later, splenic lymphocytes (3 × 10⁶/well) were restimulated in vitro using mitomycin C-treated DC2.4 cells (3 × 10⁵/well) that had been incubated with OVA-microspheres (50 μg OVA/ml) for 2 h. After incubating for 4 days at 37°C, the stimulated cells were harvested, and cytotoxicity was measured using ⁵¹Cr-release assays. EL-4 cells (1 × 10⁴/well) that had been preincubated for 2 h with OVA_257–264 peptide (1 μM) and then labeled with ⁵¹Cr (100 μCi/1 × 10⁷ cells) were used as targets.

C57BL/6 mice were immunized (i.v.) with microspheres containing OVA, CsA, or both CsA and OVA. In some experiments, mice were also immunized (i.v.) with DCs (5 × 10⁵/mouse) that had been initially treated with mitomycin C and then pulsed with OVA_257–264 peptide. Target cells for in vivo evaluation of cytotoxic activity were prepared as described previously (16). In brief, erythrocytes were removed from naive C57BL/6 spleen and lymph node cell suspensions by osmotic lysis. The cells were then washed and split into two populations. One population was pulsed with 10⁻⁶ M OVA_257–264 peptide for 1 h at 37°C, and labeled with a high concentration of CFSE (5 μM). The other control target population was labeled with a low concentration of CFSE (1 μM). For i.v. injection, an equal number of cells from each population was mixed together, and the cells were injected into immunized recipient mice (1 × 10⁷/mouse). Specific in vivo cytotoxicity was determined for the lymph node and spleen cells isolated from the recipient mice 18 h after i.v. injection by flow cytometry. The ratio between the percentages of uncoated vs OVA_257–264-coated (CFSE^low/CFSE^high) was calculated to obtain a numerical value of cytotoxicity.

The effects of the calcineurin inhibitor tacrolimus on MHC-restricted Ag presentation were examined in mice injected subcutaneously (s.c.) with clinical doses of tacrolimus (initial dose of 0.15 mg/kg, followed by maintenance doses of 0.06 mg/Kg every 6 h), as used in human kidney transplantation. OVA was injected (i.v., 10 mg/mouse) 2 h after the initial tacrolimus injection. At the indicated time points, DCs were isolated from lymph nodes and spleens of mice using a negative selection method. DC purity was 78–84% as determined by percentages of cells expressing both CD11c and I-A^b by flow cytometry. A representative histogram of the DCs isolated from lymph nodes and spleens is shown in Fig. 1,A. The efficacy of class K^b-restricted-OVA peptide presentation was evaluated using OVA-specific CD8 T cell hybridoma cells, B3Z cells, which express β-galactosidase when activated by OVA_257–264-H-2K^b complexes (13). Tacrolimus was found to inhibit the class I MHC-restricted presentation of OVA efficiently in both lymph node DCs (Fig. 1,B) and spleen DCs (Fig. 1 C).

The effects of tacrolimus on the class II MHC-restricted presentation of OVA were also examined in mice injected with tacrolimus and OVA using the regimen described in Fig. 1. In this experiment, DCs were isolated from lymph nodes and spleens 20 h after injecting OVA, and the efficacy of OVA peptide presentation by class II MHC was evaluated using OVA-specific CD4 T cell hybridoma cells, DOBW cells, which recognize OVA_323–339-I-A^d complexes and express IL-2 (14). Tacrolimus also efficiently inhibited the class II MHC-restricted presentation of OVA in lymph node DCs (Fig. 2,A) and spleen DCs (Fig. 2 B).

The effects of tacrolimus on the MHC-restricted presentation of OVA were also examined in thioglycollate-elicited mouse peritoneal macrophages. In this experiment, mice were injected once with tacrolimus (s.c., 0.15 mg/Kg or 1.0 mg/Kg) followed by OVA (i.p., 2 mg/mouse) 2 h after the tacrolimus injection, and peritoneal macrophages were isolated 6 h after OVA injection. Amounts of OVA peptides complexed with MHC were assessed using OVA-specific T cell hybridomas, B3Z cells, or DOBW cells. Again, tacrolimus was found to efficiently inhibit both the class I (Fig. 3,A) and the class II (Fig. 3,B) MHC-restricted presentation of OVA in peritoneal macrophages. To show the specificity of action of tacrolimus, rapamycin, which has been shown not to inhibit MHC-restricted exogenous Ag presentation in vitro (12), was injected in the same experimental protocol (s.c., 0.15 mg/Kg followed by OVA injection), and the amounts of OVA peptides complexed with MHC were assessed using B3Z cells, or DOBW cells. As expected, rapamycin did not inhibit both the class I (Fig. 3,C) and the class II (Fig. 3 D) MHC-restricted presentation of OVA.

To determine whether the inability of macrophages isolated from tacrolimus-injected mice to activate OVA-specific B3Z cells is due to inhibition of K^b-OVA peptide complex generation and not other factors, peritoneal macrophages isolated from tacrolimus-injected mice were incubated with the OVA_257–264 peptide, SIINFEKL, and subsequently examined in terms of their ability to activate B3Z cells. It was found that macrophages from tacrolimus-treated mice performed identically to those from tacrolimus-untreated mice using SIINFEKL from limiting to saturating concentration (Fig. 4). We also examined the effects of tacrolimus on the expression of major costimulatory molecules B7-1, B7-2 and CD40. Treating mice for up to 20 h in vivo with tacrolimus had no discernible effect on the surface expression of these molecules as determined by flow cytometry (data not shown).

To bolster the evidence that calcineurin inhibitors block cross-presentation in vivo, we examined the CTL-inducing activity of calcineurin inhibitor-exposed APCs in mice. Because calcineurin inhibitors are indeed highly potent at inhibiting T cell responses, we prepared microspheres containing OVA and CsA using the biocompatible/biodegradable polymer, poly(d,l-lactide-coglycolide), to target the delivery of CsA to phagocytes. The amount of CsA required in microspheres to inhibit the class I MHC-restricted presentation of OVA was first determined using a DC cell line, DC2.4 cells (15). We found that DC2.4 cells that phagocytosed microspheres containing 85 μg of CsA per mg OVA were unable to present OVA_257–264 peptide on class I MHC molecule (Fig. 5,A). Subsequently, microspheres containing 85 μg of CsA per mg OVA were injected i.v. into mice tail veins (20 μg as OVA/mouse). OVA-specific CTL activities were then assessed one week later using in vitro restimulated splenic lymphocytes. It was found addition of CsA to microspheres completely abrogated their immunogenicity (Fig. 5,B). An in vivo CTL assay was also performed in mice immunized with microspheres containing OVA, CsA, or both CsA and OVA. Representative histograms were shown in Fig. 6,A. Consistent with the results obtained from the in vitro restimulation experiment, addition of CsA to microspheres completely abrogated their ability to induce OVA-specific CTLs in the spleens (Fig. 6,B) and lymph nodes (Fig. 6,C). The inability of these microspheres to induce CTL responses was not due to systemic effects caused by CsA release from microspheres, because injection of OVA_257–264 peptide-pulsed DCs (5 × 10⁵/mouse) together with microspheres containing OVA and CsA induced normal OVA-specific CTL responses (Fig. 5,B and Fig. 6).

The studies presented here show that calcineurin inhibitors efficiently block class I and class II MHC-restricted exogenous Ag presentation in vivo. We propose that this effect accounts for the reported ability of calcineurin inhibiting drugs to block tolerance induction following transplantation in humans, monkeys, and mice (17, 18, 19, 20, 21, 22, 23). CsA blocked tolerance induction in mouse skin and heart graft models when administered together with costimulation blocking Abs (17, 18, 19). Long-term chimerism and tolerance induction were also inhibited by CsA and tacrolimus in a mouse model using nonmyeloablative total body irradiation, standard-dose bone marrow transplantation, and costimulation blockade (20). Tacrolimus was also found to be detrimental in a nonhuman primate renal allograft model using anti-CD154 mAbs (21). Follow-up studies on the results of renal cadaver transplants performed between 1984 and 1994 also revealed that coadministration of CsA and anti-CD3 mAb at the time of transplantation resulted in shorter-term graft survival than when CsA therapy was delayed until transplanted kidney achieved satisfactory function (23). Thus, so-called “calcineurin-sparing” immunosuppressive regimens that entail delayed administration following transplantation have been designed to circumvent this phenomenon (24, 25).

In contrast, rapamycin, which inhibits immune responses by a mechanism distinct from calcineurin inhibitors, was shown to be highly compatible with costimulation blockade in tolerance induction (17, 18, 19, 20). Rapamycin was found to fully preserve the tolerogenic effect of costimulation blockade in a mouse cardiac allograft model (18). In the mouse cardiac graft model, the combination of CsA and rapamycin together with costimulation blockade, however, was found to be highly detrimental to long-term cardiac allograft survival, showing that CsA antagonized the tolerizing effects of the nonlymphoablative protocol, i.e., rapamycin plus costimulation blockade (18).

The differential effects of calcineurin inhibitors and rapamycin on tolerance induction has been explained in several ways. One mechanism that accounts for the differential effects is that calcineurin inhibitors, by inhibiting the secretion of IL-2, prevent the deletion of alloantigen-reactive T cells (26). Consistent with this hypothesis, IL-2 knock out mice which have profound defects in activation-induced cell death were shown to be resistant to the induction of tolerance to islet and cardiac allografts by costimulation blockade (27). IL-2 has also been shown to be necessary for long-term allograft survival (28). In contrast to calcineurin inhibitors, rapamycin, which blocks IL-2-induced signal transduction, was shown to permit alloreactive T cell apoptosis (26). A second possibility advanced is that calcineurin inhibitors, but not rapamycin, prevent the development of CD4⁺CD25⁺ T_reg which have been demonstrated to play an important role in many models of tolerance induction (29). Indeed, numerous studies have shown that calcineurin inhibitors impair the development and function of T_reg, whereas rapamycin fostered the development and function of highly suppressive T_reg in vitro and in vivo (30, 31, 32, 33, 34). It has also been proposed that IL-2 signaling is crucial for the functional activity of T_reg (35, 36). Nevertheless, the differential effects of calcineurin inhibitors and rapamycin on tolerance induction require clarification. Calcineurin inhibitors inhibit IL-2 production, whereas rapamycin inhibits IL-2-induced signaling. Thus, in theory, both calcineurin inhibitors and rapamycin could interfere with processes requiring IL-2, such as the activation-induced cell death of activated T cells and the function of T_reg.

We recently showed that calcineurin inhibitors, but not rapamycin, inhibit MHC-restricted exogenous Ag presentation in vitro (12). We found that addition of CsA or tacrolimus, but not rapamycin, to cultures of DCs blocked intracellular processing events leading to the class I and class II MHC-restricted presentation of the exogenously added model Ag, OVA. The present study also shows that calcineurin inhibitors efficiently block class I and class II MHC-restricted exogenous Ag presentation in vivo. Moreover, injection of calcineurin inhibitors followed by OVA inhibited MHC-restricted exogenous Ag presentation not only in DCs isolated from lymph nodes and spleens, but also in macrophages isolated from thioglycollate-elicited mouse peritoneal cavity. In addition, we found that biodegradable microspheres containing both OVA and CsA were defective at inducing OVA-specific CTL responses when injected into mice. Based on the current findings, we propose that the inhibitory effects of calcineurin inhibitors on tolerance induction may be due to the inhibition of alloantigen presentation by host APCs.

The activation of T cells and subsequent generation of effector function are dependent on at least two signals delivered by APCs, an Ag-specific signal delivered via an MHC/peptide structure and a costimulatory signal delivered via costimulatory ligands such as B7-1 and B7-2. Antigenic stimulation of T cells in the absence of costimulation leads to the development of functional unresponsiveness or clonal anergy of T cells (37, 38, 39). The present study as well as our previous study shows that calcineurin inhibitors, but not rapamycin, inhibit MHC-restricted presentation pathways of exogenous Ag. In the absence of MHC-restricted Ag presentation (signal-1), blockade of costimulation (signal-2) would be meaningless in terms of induction of Ag-specific T cell tolerance or apoptosis. Likewise, in a situation that blocks MHC-restricted Ag presentation completely, T_reg cells cannot be induced because T cells are unaware of the Ag’s presence.

The findings of the present study may also explain the reduced T cell stimulatory capacity of calcineurin inhibitor-exposed DCs. DCs cultured in the presence of calcineurin inhibitors were found to have a reduced capacity to stimulate allogeneic T cell response, although the expression of costimulatory molecules were unaffected (40, 41, 42). During allorecognition, DCs can sensitize alloreactive T cells via two pathways, a direct pathway initiated by donor DCs presenting intact donor MHC molecules, and via an indirect pathway that results from recipient DCs processing and presenting donor MHC as peptides (43). The importance of indirect presentation pathway becomes increasingly clear. Cross-presentation was originally viewed as an obscure phenomenon associated with transplantation immunity. However, it is now appreciated to be a major mechanism by which the immune system monitors tissues for the presence of foreign Ags (3, 4, 5, 6, 7, 8, 44). The inhibition of the indirect presentation pathway of alloantigens may mechanistically underlie the suppressed allostimulatory capacity of calcineurin inhibitor-exposed DCs.

It is noteworthy that in assessing the amounts of OVA peptides complexed with MHC molecules, we used OVA-specific T cell hybridomas such as B3Z cells and DOBW cells. Thus, the readouts are not quantitative measures of the amounts of OVA-derived peptide-MHC complexes. In fact, we tried to measure the number of OVA_257–264-H-2K^b complexes on APCs using the mAb 25-D1.16 (45). However, the surface expression of OVA_257–264-H-2K^b complexes on APCs was below the detection limit. Nevertheless, the current study clearly demonstrates that calcineurin inhibitors block the expression of OVA peptide-MHC complexes at high enough levels on APCs to provoke OVA-specific T cell responses in mice following OVA injection.

The authors have no financial conflict of interest.