Allograft Acceptance Despite Differential Strain-Specific Induction of TGF-β/IL-10-Mediated Immunoregulation

Authors

  • Alice A. Bickerstaff,

    1. The Ohio State University College of Medicine, Departments of aSurgery, bPathology, and Molecular Virology, Immunology, and cMedical Genetics, and the dComprehensive Cancer Center, Columbus, OH 43210, USA
    Search for more papers by this author
  • a Jiao-Jing Wang,

    1. The Ohio State University College of Medicine, Departments of aSurgery, bPathology, and Molecular Virology, Immunology, and cMedical Genetics, and the dComprehensive Cancer Center, Columbus, OH 43210, USA
    Search for more papers by this author
  • a Dongyuan Xia,

    1. The Ohio State University College of Medicine, Departments of aSurgery, bPathology, and Molecular Virology, Immunology, and cMedical Genetics, and the dComprehensive Cancer Center, Columbus, OH 43210, USA
    Search for more papers by this author
  • and a Charles G. Orosz ad

    1. The Ohio State University College of Medicine, Departments of aSurgery, bPathology, and Molecular Virology, Immunology, and cMedical Genetics, and the dComprehensive Cancer Center, Columbus, OH 43210, USA
    Search for more papers by this author

*Corresponding author: Alice A. Bickerstaff, bickerstaff-1@medctr.osu.edu

Abstract

We examined the immune approaches that C57BI/6 and BALB/c mice take when treated to accept cardiac allografts. C57Bl/6 mice accept DBA/2 cardiac allografts when treated with gallium nitrate (GN) or anti-CD40L mAb (MR1). These allograft acceptor mice fail to mount donor-reactive delayed type hypersensitivity (DTH) responses, and develop a donor-induced immunoregulatory mechanism that inhibits DTH responses. In contrast, BALB/c mice accept C57BI/6 cardiac allografts when treated with MR1 but not with GN. These allograft acceptor mice display modest donor-reactive DTH responses, and do not develop donor-induced immune regulation of DTH responses. Real-time PCR analysis of rejecting graft tissues demonstrated no strain-related skewing in the production of cytokines mRNAs. In related studies, C57Bl/6 recipients of cytokine and alloantigen educated syngeneic peritoneal exudate cells (PECs) failed to mount DTH responses to the alloantigens unless neutralizing antibodies to transforming growth factor-beta (TGF-β were present at the DTH site demonstrating regulation of cell-mediated alloimmune responses. In contrast, BALB/c recipients of cytokine-and alloantigen-educated PECs expressed strong DTH responses to alloantigens demonstrating a lack of regulated alloimmunity. In conclusion, C57BI/6 mice respond to immunosuppression by accepting cardiac allografts and generating TGF-β-related regulation of donor-reactive T cell responses, unlike BALB/c mice that do not generate these regulatory responses yet still can accept cardiac allografts.

Abbreviations
DTH

delayed-type hypersensitivity

IVC

inferior vena cava

TT

tetanus toxoid

GN

gallium nitrate

TGF-β

transforming growth factor-beta

PEC

peritoneal exudate cells

Introduction

Studies on the immunobiology of murine cardiac allografts have revealed that multiple mechanisms of acute graft rejection must exist. For example, both IFN-γ+/+ and IFN-γ–/– mice can rapidly reject cardiac allografts, but the relative availability of IFN-γ determines whether CD8+ T cells participate in the rejection response (1). In general, much remains to be learned regarding the panorama of immune response options that are available to the immune system in allograft recipients (2,3). Even less is known about the mechanisms by which specific immune response options are selected during a given response. These issues appear to have been best studied in infectious disease models, where it has been demonstrated that different strains of mice will engage different immune mechanisms to deal with the same infectious agent. Often, these differences involve the selective use of the TH1 or TH2 cytokines (4).

In transplant models, patterns of cytokine production have been determined in many different strain combinations, but there has been little direct evidence for a strain-related predisposition toward differential TH1 or TH2 cytokine production in response to allograft implantation. Instead, TH1-like cytokine production has been associated with acute allograft rejection (5), regardless of strain combination, while TH2-like cytokine production has been associated with the influence of various immunosuppressive agents that promote allograft acceptance (6,7). However, some strain-specific immune differences have been identified in murine transplant models. For example, the effectiveness of selected immunosuppressive agents can vary among different strains of mice (8). Further, there appear to be strain-specific differences in the spontaneous acceptance of murine renal allografts (9). When an immune mechanism is identified in mice, its penetrance among the murine population is rarely tested in confirmatory studies that employ either multiple inbred strains of mice or out-bred mouse strains. This would be helpful if only to identify immune situations in which different response options are selected. This would facilitate not only the study of the different immune response options, but also the conditions that lead to their expression.

Many studies currently being done in the area of experimental allograft tolerance have demonstrated that development of regulatory T cells is often associated with drug-induced allograft acceptance (10–13). Indeed, regulatory T cells have emerged as an important area of investigation in several disciplines (14–18). Again much of the definitive work in this area has been done in only a few strain combinations. Before clinical application of this work, it would be informative to determine if the employment of regulatory T cells for allograft acceptance is conditional (strain related) and necessary.

In this communication, we have examined drug-induced cardiac allograft acceptance in two different strain combinations. We provide evidence that this allograft acceptance results from two very different immune approaches in these two strain combinations, only one in which appears to involve regulatory T cells. These studies raise questions regarding the array of mechanisms that promote allograft acceptance, and how they relate to regulatory T cells.

Materials and Methods

Mice

C57BI/6 (H-2b) and DBA/2 (H-2d) mice were obtained from Taconic (Germantown, NY, USA). BALB/c (H-2d) mice were obtained from Harlan Bioproducts Inc. (Indianapolis, IN, USA). All mice were housed and treated in accordance with Animal Care Guidelines established by the National Institute of Health and The Ohio State University.

Murine cardiac transplantation

Heterotopic cardiac transplantation was performed as described by Corry et al. (19). In general, the native hearts from heparinized donor mice (DBA/2 or C57BI/6) were anastomosed to recipient (C57BI/6 or BALB/c) abdominal aorta and vena cava using microsurgical techniques. Graft survival was assessed by trans-abdominal palpation.

Immunosuppression with gallium nitrate

As described previously (20), gallium nitrate (GN, Ganite, Fujisawa, Deerfield, IL, USA) was administered as an initial subcutaneous bolus injection of 2.2 mg 24 h prior to graft implantation. This was followed by 28 d of continuous delivery via osmotic minipumps (model 2002, Alzet Inc., Palo Alto, CA, USA), which delivered 0.5 µL (12.5 µg GN) per hour. Circulating levels of GN fall to subtherapeutic levels within 7 d of pump removal (21).

Immunosuppression with anti-CD40L mAb (MR1)

Anti-CD40L mAb (MR1, purified by Ligocyte Pharmaceuticals, Bozeman, MT, USA) was administered as an intraperitoneal (i.p.) injection of 1.075 mg (0.2 mL/injection) after graft implantation and on day 1 post-transplantation, and 0.54 mg (0.10 mL/injection) on days 2, 4, 6, 8, and 10 post-transplantation.

PEC education

As previously described by Takeuchi (14), peritoneal exudate cells (PECs) are obtained from normal mice that receive an i.p. injection of 1 mL of thioglycollate (Sigma Chemical Co., St. Louis, MO, USA) 4 d earlier. PECs are washed and suspended in DMEM that contained 1.6 mm l-glutamine, 0.27 mm folic acid, 0.27 mm l-asparagine, 0.55 mm l-arginine, 10 mm HEPES, 1 mm sodium pyruvate solution, 100 U/mL penicillin–streptomycin solution (all obtained from Gibco, Long Island, NY, USA) plus 10% FCS (Atlanta Biologicals, Atlanta, GA, USA). The cells were then cultured overnight at 10% CO2 in 24-well plastic tissue culture plates (Falcon 3047) at a concentration of 1.5 × 106 cells/well with DBA/2 subcellular DBA/2 antigen (100 µg/well) ± TGFβ(5 ng/well) or IL-10 (10 ng/well). After 24 h, cells were collected, washed, resuspended in PBS at a concentration of 3 × 105 cells per 0.3 mL, and injected intravenously (i.v.) into naïve syngeneic recipients. After 4–8 weeks mice are tested for delayed-type hypersensitivity (DTH) responses to alloantigen.

Transfer DTH assay

For this assay, the pinnae of naïve mice were injected via a 30-gauge insulin syringe with a mixture of 8 × 106 splenocytes from PEC-educated mice or cardiac allograft recipients along with challenge antigen. Changes in thickness were measured both before injection and 24 h after injection using a dial thickness gauge (Swiss Precision Instruments, Carlstadt, NJ, USA). For reference, changes in the range of 0–30 × 10−4 inches represent background swelling due to injection trauma, changes in the range of 40–60 × 10−4 inches represent moderate DTH responses, and changes in the range of 70–100 × 10−4 inches represent strong DTH responses.

Cytokines and cytokine antibodies

Porcine transforming growth factor-beta (TGF-β, polyclonal rabbit anti-human TGF-β, polyclonal goat anti-murine IL-10 antibodies, control rabbit Ig, and control goat Ig were all obtained from R & D Systems (Minneapolis, MN, USA). Murine IL-10 was obtained from Gibco BRL Products (Grand Island, NY, USA).

Cellular alloantigen

Cellular alloantigen was prepared by depleting splenocytes of red blood cells and suspending 8 × 106 cells in 25 µL of PBS.

Subcellular alloantigen

Subcellular alloantigen was prepared according to the methods of Engers et al. (22). Briefly, fresh red blood cell (RBC)-depleted splenocytes suspended in PBS, were subjected to three rapid freeze–thaw cycles, using liquid nitrogen, and spun at 13 000 r.p.m. (300 g) for 30 min to remove residual debris. The supernatant was adjusted to 3–5 mg protein/mL and was used as the source of subcellular alloantigen. For DTH challenge, 25 µL (75–125 µg protein) of this solution is injected into murine pinnae.

Tetanus toxoid

Tetanus toxoid (TT, Wyeth-Ayerst, Marietta, PA, USA) was obtained at a concentration of 10 lf (limits of flocculation) per ml in PBS. To sensitize mice, 0.1 mL (1 lf) of TT was injected subcutaneously (s.c.) at least 14 d before DTH challenge. To challenge mice for DTH reactivity, 25 µL (0.25 lf) of TT was injected at the DTH site.

RNA isolation and RT-PCR

The total RNA was isolated using the RNeasy_, a RNA isolation kit (Qiagen Inc., Valencia, CA, USA). Normal hearts and transplanted hearts were collected and frozen in liquid nitrogen and then ground into powder with liquid nitrogen. Next, 25 mg of the tissue powder was transferred into in a 1.5-mL centrifuge tube containing 350 µL of the LTR lysis solution. The mixture was vortexed, centrifuged and mixed with an equal volume of 70% ethanol, before being loaded into the RNeasy_ filter. After washing and eluting, 35 mL of the total RNA sample was obtained. Three micrograms of the total RNA was then reverse transcribed (RT) into cDNA using the murine monkey leukemia virus reverse transcriptase, M-MLV, reverse transcriptase (Gibco BRL, Grand Island, NY). This involved mixing 30 µL (3 µg) of total RNA and 1 µL (150 ng) of the random primer (Boehringer Mannheim Co., Indianapolis, IN, USA) and denaturing it in 95 °C for 5min. It was then placed on ice for 10min. RT buffer, dNTP and M-MLV were added to the denatured mixture and the RT reaction was completed via incubation at 37 °C for 2 h.

Real-time PCR for mRNA quantification

Real-time PCR has been described in detail by other investigators (23–26). This technique uses an oligonucleotide hybridization probe that is labeled with a reporter fluorescent dye (6-carboxy-fluorescein) at the 5′ end and with a quencher fluorescent dye (6-carboxy-tetramethylrhodamine) at the 3′ end. Before the start of the PCR reaction when the probe is intact, the reporter fluorescent dye emission is quenched by the close physical proximity of the quencher. During the extension phase of the PCR cycle, the nucleolytic activity of the Taq DNA polymerase cleaves the hybridization probe and releases the reporter dye from the probe. The resulting relative increase in reporter fluorescent dye emission is monitored in real-time during PCR amplification using the Sequence Detection System (ABI PRISM 7700 Sequence Detection System and software, PE Applied Biosystems, Inc., Foster city, CA, USA). Sequences for all primers and probes used in these studies are listed in Table 1.

Table 1. : Sequences of the murine cytokine primers and probes
GenePrimerSequence
IL-2Forward5′-CTCCTGAGCAGGATGGAGAATT-3′
Reverse5′-CGCAGAGGTCCAAGTGTAGCT-3′
Probe5′ 6FAM-CTGAAACTCCCCAGGAGTCTCACCTTC-TAMRA-3′
IL-4Forward5′-GGCATTTTGAACGAGGTCACA-3′
Reverse5′-AGGACGTTTGGCACATCCAT-3′
Probe5′ 6FAM-CTCCGTGCATGGCGTCCCTTCT-TAMRA-3′
IL-10Forward5′-TTTGAATTCCCTGGGTGAGAA-3′
Reverse5′-ACAGGGGAGAAATCGATGACA-3′
Probe5′-TGAAGACCCTCAGGATGCGGCTG-TAMRA-3′
γ-INFForward5′-AGCAACAGCAAGGCGAAAA-3′
Reverse5′-CTGGACCTGTGGGTTGTTGA-3′
Probe5′ 6FAM-CCTCAAACTTGGCAATACTCATGAATGCATCC-TAMRA-3′
18SForward5′-CGGCTACCACATCCAAGGAA-3′
rRNAReverse5′-GCTGGAATTACCGCGGCT-3′
Probe5′ JOE-TGCTGGCACCAGACTTGCCCTC-TAMRA-3′

Real-time PCR was preformed in a 25-µL reaction volume containing 2.5 µL of cDNA from the RT reaction, primer/probe mixture (900 nm of each sense, antisense primers, 120 nm of probe with FAM labeled) and 12.5 µL of the 2MM buffer (Perkin Elmer, Norwalk, CT, USA). A series of target cDNA fragments with a predetermined copy number was run as the standard simultaneously with the samples for the quantitative analysis. The copy number of each cytokine in each reaction was calculated using the standard curve. The 18S ribosomal RNA (with VIC labeled probe) was chosen as an internal control to regulate the variability in amplification, which was due to different starting mRNA concentrations. The copy number of 18S rRNA in each reaction was calculated using the 18S standard curve. The relative expression level of the gene of interest was described as copies of cytokine mRNA/1010 copies of 18S rRNA.

Results

The DBA/2 (H-2d)→C57BI/6 (H-2b) strain combination is among the most potent of MHC barriers for cardiac allograft acceptance in mice, and a difficult challenge for immunosuppressive agents. As shown in Figure 1A, C57BI/6 mice reject DBA/2 cardiac allografts within 10 d, but transient treatment of these cardiac allograft recipients with experimental immunosuppressants such as GN or MR1 routinely subverts allograft rejection in 80–100% of the mice for more than 60 d. C57BI/6 and BALB/c mice are known to employ different immune approaches to the same antigenic stimuli (4). As shown in Figure 1B, BALB/c mice reject C57BI/6 cardiac allografts within 10–12 d. Transient, 28-d treatment C57BI/6→BALB/c cardiac allograft recipients with GN is completely ineffective at prolonging allograft survival, and allograft rejection occurs at control times in most mice well before the treatment with GN has been completed. In contrast, treatment of C57BI/6→BALB/c cardiac allograft recipients with MR1 delays allograft rejection for more than 60 days in about 80% of the mice, remarkably similar to its effect in the DBA/2→C57BI/6 strain combination. Thus, both GN and MR1 allow C57BI/6 mice to accept DBA/2 cardiac allografts, whereas only MR1, but not GN, allows BALB/c mice to accept C57BI/6 cardiac allografts.

Figure 1.

Cardiac allograft survival by GN- or MR1-treated C57BI/6 and BALB/c recipients. (A) C57BI/6 mice received heterotopic cardiac allografts (DBA/2) and were either left untreated (•, n = 4), treated with GN (▴, n = 10) or treated with MR1 (▾, n = 8). Graft survival was assessed by trans-abdominal palpation. (B) BALB/c mice received heterotopic cardiac allografts (C57BI/6) and were either left untreated (•, n = 3), treated with GN (▴, n = 4) or treated with MR1 (▾, n = 10). Graft survival was assessed by trans-abdominal palpation.

BALB/c and C57BI/6 mice are known to deploy different cytokine arrays in response to challenge with various infectious agents (27). We determined if a skewing of cytokine mRNA production occurred in DBA/2→C57BI/6 vs. C57BI/6→BALB/c cardiac allografts. For these studies, cardiac allografts were obtained 7–10 d post-transplant from nonsuppressed allograft recipients, and control hearts were obtained from normal, nontransplanted C57BI/6, DBA/2 and BALB/c mice. RNA was extracted from these tissues and real-time PCR was used to determine the amount of mRNA for IL-2, IFN-γ, IL-4 or IL-10 that was present. As shown in Figure 2, cardiac allografts in both strain combinations produced mRNAs for all four cytokines, demonstrating that there was no strain-related skewing in the production of these intragraft cytokines. We did note, however, that the C57BI/6→BALB/c strain combination actually produced more of each cytokine mRNA than did the DBA/2→C57BI/6 strain combination.

Figure 2.

Real-time PCR quantitation of cytokine mRNAs in cardiac allografts. RNA was extracted from normal C57BI/6, DBA/2, and BALB/c hearts and DBA/2→C57BI/6 and C57BI/6→BALB/c cardiac allografts harvested 7–10 d post-transplant. The RNA was reverse transcribed into cDNA and tested by real-time PCR for reactivity with primer pairs for the designated cytokines. The relative expression level of the gene of interest was computed with respect to the 18S rRNA level of the internal control (mean expression of n = 3 hearts per group ± 1 SD).

In previous studies (20), we demonstrated that DBA/2→C57BI/6 cardiac allograft rejectors display strong donor-reactive DTH responses, whereas GN-induced cardiac allograft acceptors fail to display donor-reactive DTH responses. As shown in Figure 3, most MR1-induced DBA/2→C57BI/6 cardiac allograft acceptors similarly fail to mount donor-reactive DTH responses, whether the mice are challenged with intact, viable donor splenocytes or with freeze–thawed subcellular donor splenocytes. In contrast, BALB/c recipients of C57BI/6 cardiac allografts display modest donor-reactive DTH responses to both cellular and subcellular alloantigens, whether they are nonsuppressed allograft rejectors, GN-treated allograft rejectors, or MR1-treated allograft acceptors (Figure 3).

Figure 3.

C57BI/6 cardiac allograft acceptors express a donor-reactive mechanism of DTH inhibition. Splenocytes from GN-treated C57BI/6 cardiac allograft acceptors (n = 4), MR1-treated C57BI/6 cardiac allograft acceptors (n = 6), and untreated C57BI/6 cardiac allograft rejectors (n = 4) were obtained between 60 and 90 d post-transplant and transferred into the pinnae of naïve C57BI/6 mice along with cellular and subcellular donor (DBA/2) alloantigens. DTH responses were measured after 24 h as change in ear thickness (mean ± 1 SD). The dashed line, background (BKG), represents the mean change in ear thickness when naïve mice are challenged with syngeneic splenocytes alone (n = 4). Splenocytes from GN-treated BALB/c cardiac allograft rejectors (n = 4), MR1-treated BALB/c cardiac allograft acceptors (n = 8) and untreated BALB/c cardiac allograft rejectors (n = 4) were obtained between 60 and 90 d post-transplant and transferred into the pinnae of naïve BALB/c mice along with cellular and subcellular donor (C57BI/6) alloantigens. DTH responses were measured after 24 h as the change in ear thickness (mean ± SD). The solid black circles represent the responses of individual mice.

The failure of DBA/2→C57BI/6 cardiac allograft acceptors to mount donor-reactive DTH responses is due to the development of a donor-induced immunoregulatory mechanism. Although this mechanism is antigen specific in induction, it is not antigen specific in effect. Thus, it can inhibit DTH responses to third-party antigens, such as tetanus toxoid, that have been colocalized with donor alloantigens (28). To illustrate this, cardiac allograft acceptors were immunized with tetanus toxoid; their splenocytes were transferred to the pinnae of naïve, syngeneic mice. These transferred splenocytes were challenged with tetanus toxoid, donor alloantigens, or both together. Shown in Figure 4 are the resulting DTH responses of DBA/2→C57BI/6 cardiac allograft acceptors, and C57BI/6→BALB/c cardiac allograft acceptors. Both the GN-treated and the MR1-treated DBA/2→C57BI/6 cardiac allograft recipients mounted tetanus-reactive, but not donor-reactive DTH responses. However, neither mounted tetanus-reactive DTH responses when tetanus was colocalized with donor alloantigens. Thus, GN and MR1 therapy both permit the development of DTH-detectable, donor-induced immunoregulation in the DBA/2→C57BI/6 strain combination. In GN-treated mice, these regulatory mechanisms are mediated by TGF-β and IL-10 (29). In the tetanus-sensitized BALB/c mice, however, the MR1-treated cardiac allograft acceptors mounted DTH responses to tetanus weak DTH responses to donor alloantigens, and strong DTH responses were expressed when both antigens were mixed together. Thus, a mechanism for donor-induced inhibition of DTH responses does not develop in the C57BI/6→BALB/c strain combination.

Figure 4.

BALB/c cardiac allograft acceptors lack a donor-reactive mechanism of DTH inhibition. Splenocytes from GN-treated C57BI/6 cardiac allograft acceptors (n = 6), MR1-treated C57BL/6 cardiac allograft acceptors (n = 6), and MR1-treated BALB/c cardiac allograft acceptors (n = 8) were obtained between 60 and 90 d post-transplant and transferred into the pinnae of naïve syngeneic mice along with subcellular donor alloantigens, TT, or a combination of both. DTH responses were measured as described in Figure 3.

Streilein (15) and Neiderkorn (16) have demonstrated that PECs can be educated ex vivo to antigens, after which they can promote development the antigen-specific DTH sensitization when transferred into naïve, syngeneic mice. The outcome of this sensitization process changes when TGF-β is included during ex vivo PEC education to antigens. Recipients of these TGF-β-influenced PEC fail to mount DTH responses when challenged with antigen. We used this experimental system to examine strain-related differences in the development of allosensitization that occurs in the absence of engraftment and immunosuppression.

Thioglycollate-induced C57BI/6 PECs were cultured for 24 h with subcellular DBA/2 alloantigens, washed, and transferred i.v. into naïve C57BI/6 mice. Figure 5A demonstrates that these mice mount strong DBA/2-reactive DTH responses when challenged 30–60 days later with subcellular DBA/2 alloantigens. However, inclusion of TGF-β or IL-10 in the PEC cultures during their ex vivo alloeducation changes this outcome. Recipients of these cytokine-influenced PECs fail to mount DTH responses when challenged with subcellular DBA/2 alloantigens 30–60 d later. The same approach was used to educate BALB/c PECs to C57BI/6 alloantigens, but the recipients of these educated PECs generated strong DTH responses when challenged 30–60 d later with C57BI/6 alloantigens, whether the PECs were alloeducated in the presence or absence of TGF-β or IL-10 (Figure 5B).

Figure 5.

(A ) C57BI/6 mice infused with PEC that were educated to DBA/2 alloantigens in the presence of TGF-β or IL-10 express a DBA/2-reactive mechanism of DTH inhibition. C57BI/6 mice infused with PEC that were educated to DBA/2 alone (•), in the presence of TGF-β (▾), or in the presence of IL-10 (█) were challenged in the pinnae with subcellular DBA/2 antigen at days 30 and 60. DTH responses were measured as described in Figure 6. (B) BALB/c mice infused with PEC that were educated to C57BI/6 alloantigens in the presence of TGF-β or IL-10 fail to express a C57BI/6-reactive mechanism of DTH inhibition. BALB/c mice infused with PEC that were educated to C57BI/6 alone (•), in the presence of TGF-β (▾), or in the presence of IL-10 (█) were challenged in the pinnae with subcellular C57BI/6 antigen at days 30 and 60. DTH responses were measured as described in Figure 6.

We next determined if the DTH nonreactive recipients of ex vivo educated PECs fail to mount DTH responses because a TGF-β-mediated regulatory mechanism had been induced. Sixty days after naïve C57BI/6 mice were given C57BI/6 PECs that had been educated to DBA/2 alloantigens in the presence of TGF-β or IL-10, their splenocytes were transferred to the pinnae of naïve C57BI/6 mice and challenged with subcellular DBA/2 alloantigen plus either anti-TGF-β antibodies or control sera. As shown in Figure 6, these transferred splenocytes mounted DBA/2-reactive DTH responses only when antibodies to TGF-β are included at the DTH challenge site. The DTH responses made under these conditions were similar in magnitude to the DTH responses made by recipients of control PECs that had been educated to alloantigens in the absence ofTGF-β or IL-10. Thus, ex vivo alloeducation of C57BI/6 PECs in the presence of TGF-β or IL-10 allows them to induce DTH-detectable immunoregulatory mechanisms similar to those that develop in cardiac allograft acceptor mice (29). In contrast, the transferred splenocytes from BALB/c recipients of BALB/c PECs that were ex vivo educated to C57BI/6 alloantigens mounted C57BI/6-reactive DTH responses, whether or not the PECs had been alloeducated in the presence of TGF-β or IL-10 (Figure 5). Further, the inclusion of antibodies to TGF-β at the DTH challenge site did not influence the magnitude of the DTH responses made by these mice. Thus, ex vivo education of BALB/c PECs to C57BI/6 alloantigens in the presence of TGF-β or IL-10 does not allow the BALB/c PECs to induce DTH-detectable immunoregulatory mechanisms. Indeed, these mice generated DTH responses that were similar to those observed in GN-treated or MR1-treated C57BI/6→BALB/c cardiac allograft recipients.

Figure 6.

C57BI/6 mice infused with PEC that were educated to DBA/2 alloantigens in the presence of TGF-β or IL-10 express a DBA/2-reactive mechanism of DTH inhibition that involves TGF-β. Splenocytes from C57BLI/6 mice infused with PEC that were educated to DBA/2 alone, in the presence of TGF-β, or in the presence of IL-10 were transferred into the pinnae of naïve syngeneic mice along with subcellular DBA/2 antigen alone (open bars) or with neutralizing antibodies to TGF-β (hatched bars). A similar experiment was performed using splenocytes from BALB/c mice infused with PEC that were educated to C57BI/6 in the absence or presence of cytokines. DTH responses were measured as described in Figure 3.

Discussion

This manuscript documents a major difference between C57BI/6 mice and BALB/c mice in their relative abilities togenerate alloreactive, TGF-β/IL-10-dependent mechanisms of immune regulation. In C57BI/6 allograft acceptor mice, this regulatory activity is displayed during DTH responses and, presumably, during other manifestations of cell-mediated immunity, such as acute allograft rejection. DTH regulatory mechanisms have been appreciated for more than 15 years (30). More recently, similar immune regulatory mechanisms, described by their immunologic symptoms (‘linked nonresponsiveness’ and ‘infectious tolerance’), have emerged in studies with various transplant models (31). During the recognition of graft alloantigens, the immune system can either choose to eradicate or accept the alloantigens. Usually eradication is selected, and the immune system mobilizes potent pro-inflammatory mechanisms to facilitate this. However, under the conditions promoted by certain immunosuppressants, acceptance of the graft alloantigens is selected, and the immune system mobilizes potent anti-inflammatory mechanisms to facilitate this. Much is known about the pro-inflammatory mechanisms, including the facts that they involve the expression of the B7 costimulatory molecules on antigen-presenting cells (APCs) (32), and the IL-12-facilitated production of IFN-γ by graft-reactive T cells(33). In contrast, relatively little is known about the anti-inflammatory mechanisms associated with allograft acceptance.

We have studied a set of anti-inflammatory mechanisms that develops in DBA/2→C57BI/6 cardiac allograft recipients that have been immunosuppressed with depleting anti-CD4 mAb (28), depleting anti-CD40L mAb (Figure 1), and nondepleting GN (29). Transient treatment of cardiac allograft recipients with any of these agents permits long-term allograft acceptance (Figure 1) associated (except for anti-CD40L mAb) with continued donor-reactive alloantibody production (28). This suggests that the grafts are accepted despite some form of ongoing alloimmunity. Indeed, these mice display active, anti-inflammatory, donor-reactive immune responses. Thus, they are unable to mount donor-reactive DTH responses, due to the donor alloantigen-induced production of TGF-β and IL-10 by T cells at the DTH site (29). Both TGF-β and IL-10 are known to inhibit DTH responses (34,35). In allograft acceptor mice, both pro-inflammatory and anti-inflammatory mechanisms apparently colocalize to the DTH challenge site, but the anti-inflammatory mechanisms are dominant. Thus, allograft acceptor mice can express donor-reactive DTH responses when antibodies to TGF-β or IL-10 are included at the DTH challenge site to neutralize the anti-inflammatory response (29). Further, the anti-inflammatory responses are donor-specific in their induction, but antigen nonspecific in their effects. Thus, the anti-inflammatory responses will impair DTH responses to third-party recall antigens if they are colocalized with donor alloantigens (28)(Figure 6). This demonstrates and explains the phenomena of ‘linked nonresponsiveness’ that develops under these experimental conditions. Interestingly, very similar anti-inflammatory mechanisms apparently develop in these mice, despite the dissimilarity of these three immunosuppressive agents.

We have observed that BALB/c mice are unable to accept cardiac allografts when treated with GN (Figure 1). We do not know the mechanism by which transient treatment with GN inhibits acute rejection and promotes long-term cardiac allograft rejection in C57BI/6 mice. We have observed that C3H mice are similar to BALB/c mice in their nonresponsiveness to GN (E. Wakely, Department of Surgery/Transplant, Ohio State University, June 1998, unpublished observation). However, out-bred CD1 mice are C57BI/6-like, in that three of five will accept BALB/c cardiac allografts after transient treatment with GN (A. A. Bickerstaff, Department of Surgery/Transplant, Ohio State University, November 2000, unpublished observation), suggesting that responsiveness to GN may be the rule, rather than the exception in mice. Because most of our studies utilize GN, we initially failed to appreciate the depth of the differences between C57BI/6 and BALB/c mice. Much has been reported about the differences in the immunologic approach of these strains to selected infectious agents, such as Leishmania (4).

Having observed that BALB/c mice will accept C57BI/6 cardiac allografts after treatment with anti-CD40L mAb (Figure 1), we used DTH responses to evaluate whether this acceptance was associated with the expression of donor-induced anti-inflammatory responses. In contrast to C57BI/6 mice, which clearly develop donor-induced anti-inflammatory responses after treatment with anti-CD40L mAb, BALB/c mice displayed no evidence of anti-inflammatory responses after treatment with anti-CD40L mAb (Figures 3 and 6). This was surprising, as we have observed the expression of donor-reactive anti-inflammatory responses in virtually every other experimental system of allograft acceptance studied by us to date, including spontaneous murine renal allograft acceptance (36).

The inability of BALB/c mice to develop anti-inflammatory alloreactive responses was not due to their skewed responses to graft alloantigens or immunosuppressive agents. Rather, it appears to result from a fundamentally different approach of APC toward alloantigens. Using the experimental system of ex vivo education reported by Streilein (37) and Neiderkorn (16), we briefly educated PECs to alloantigens in the presence or absence of TGF-β or IL-10, transferred them in to naïve, syngeneic mice, and used DTH to evaluate their allosensitization status 60 d later. In related studies, we have successfully used PECs from SCID mice in similar studies (A. A. Bickerstaff, June 2001, unpublished observations), indicating that the critical PEC agent in this system is not a T cell, but presumably a macrophage. When naïve C57BI/6 mice were educated to DBA/2 alloantigens with antigen-experienced APC, they displayed the same DTH phenotypes as allograft recipients (Figures 5 and 6), i.e. alloeducation in the absence of TGF-β/IL-10 resulted in pro-inflammatory, alloreactive DTH responses (allograft-rejector-like), whereas alloeducation in the presence of TGF-β/IL-10 resulted in anti-inflammatory, alloreactive DTH responses that were neutralized when TGF-β was serologically neutralized at the challenge site (allograft acceptor-like). This pattern of responses did not develop when BALB/c PECs were educated ex vivo to C57BI/6 alloantigens (Figures 5 and 6), i.e. alloeducation in either the presence or absence of TGF-β/IL-10 resulted in pro-inflammatory, alloreactive DTH responses, with no evidence for anti-inflammatory immune responses. This could be due to differential PEC trafficking, differential PEC reactivity to TGF-β or IL-10, etc.

The fact that the development of anti-inflammatory donor-reactive immunity has been commonly associated with murine allograft acceptance has fostered the hypothesis that anti-inflammatory donor reactivity may be causal for allograft acceptance. If so, it is neither essential, nor the only mechanism of allograft acceptance, as BALB/c mice accept cardiac allografts under the influence of anti-CD40L mAb therapy, but fail to exhibit evidence of anti-inflammatory, donor-reactive alloimmunity (Figures 1, 3 and 4). Indeed, the anti-CD40L mAb-treated, C57BI6→Balb/c cardiac allograft system has emerged as an important model for the study of alternative, causal mechanisms of allograft acceptance. In general, the TGF-β/1L-10-mediated, anti-inflammatory alloreactive mechanisms can be considered as surrogate markers for the development of a nonaggressive immune disposition toward graft alloantigens, but caution should be exercised when considering it as a causal mechanism of allograft acceptance.

We have recently developed evidence that pro-inflammatory and anti-inflammatory responses can also develop in human allograft recipients (38). In our transplant program, approximately 50% of renal transplant patients develop pro-inflammatory donor-reactive DTH responses, while about 20% develop anti-inflammatory donor-reactive DTH inhibition (Orosz, personal communication). Thus, the expression of anti-inflammatory immune mechanisms is not species specific, but appears to be shared with humans. It is not known why some patients develop a pro-inflammatory immune disposition towards their grafts, while others develop an anti-inflammatory immune disposition. In this manuscript, we provide evidence for a similar differential development of these two polar immune dispositions among different mouse strains. Further, the disparity between C57BI/6 mice, which adopt the anti-inflammatory disposition under the influence of immunosuppression, and BALB/c mice, which do not and maintain a pro-inflammatory disposition, can now be explored experimentally.

Acknowledgments

The authors wish to thank Kate Orosz, Jake Jansen, Jennifer Yee and Kimberlee Rudisill for their technical assistance with these experiments, Elaine Wakely for helpful discussions and administrative assistance, and Marsha Stalker for her assistance in the preparation of this manuscript.

This is manuscript no. 149 from the Transplant Sciences Program of The Ohio State University College of Medicine. This study was supported by National Institutes of Health grants PO1-AI/HL40150, RO1-HL61966, PO1-HL70294, and in part by grant P30-CA16058 (cgo), National Cancer Institute, Bethesda MD.

Ancillary