Induction of immunological tolerance is highly desirable for the treatment and prevention of autoimmunity, allergy, and organ transplant rejection. Adoptive transfer of MHC class I disparate mature T cells at the time of reconstitution of mice with syngeneic bone marrow resulted in specific tolerance to allogeneic skin grafts that were matched to the T cell donor strain. Mature allogeneic T cells survived long-term in reconstituted hosts and were able to re-enter the thymus. Analysis of T cell development using transgenic mice expressing an alloantigen-reactive TCR revealed that expression of allogeneic MHC class I on adoptively transferred mature T cells mediated negative selection of developing alloreactive T cells in the thymus. Thus, mature allogeneic T cells are able to mediate central deletion of alloreactive cells and induce transplantation tolerance without the requirement for any other alloantigen-expressing cell type.
The ability to induce immunological tolerance has the significant potential to allow for organ replacement without the need for lifelong immunosuppression, as well as to overcome autoimmunity and allergy. Based on studies in bone marrow irradiation chimeras, it is clear that bone marrow-derived hemopoietic cells are able to induce transplantation tolerance (1, 2). It has been suggested for a number of years that among bone marrow-derived hemopoietic cell lineages, bone marrow-derived APCs are perhaps most critical for inducing tolerance to MHC Ags (3, 4). However, thymocyte precursors have also been reported to be able to induce tolerance to MHC class I Ags (5). It has also been reported that in transgenic mice, expression of allogeneic MHC class I in CD2+ cells was not sufficient to induce complete T cell tolerance and acceptance of allogeneic skin grafts (6). More recently, we have suggested that in bone marrow irradiation chimeras, expression of alloantigen on mature lymphocytes is required to induce T cell tolerance, and that expression on bone marrow-derived APCs was not sufficient to induce transplantation tolerance (7).
Insofar as the capacity of distinct hemopoietic cell lineages to induce long-term tolerance is controversial, we set out to further examine which lymphocyte subsets are able to induce immunological tolerance. Based on the observation that expression of alloantigen on mature lymphocytes is required to induce transplantation tolerance (7), we examined the ability of mature B and T cells expressing an allogeneic MHC class I Ag to induce tolerance when adoptively transferred into hosts undergoing transplantation with syngeneic bone marrow. Mature allogeneic T cells were able to survive long-term when adoptively transferred into conditioned hosts undergoing reconstitution with syngeneic bone marrow and induce transplantation tolerance. In contrast, adoptive transfer of mature allogeneic B cells failed to result in transplantation tolerance. Induction of tolerance following adoptive transfer of mature allogeneic T cells appeared to occur via negative selection of alloreactive T cells in the thymus. Together, our data demonstrate that mature peripheral T cells are able to actively participate in the induction and maintenance of central tolerance. We suggest that delivery of Ags to the thymus by T cells may be of particular importance for overcoming transplant rejection and re-establishing self-nonself recognition in autoimmunity.
Materials and Methods
CBA/CaJ (H-2k) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). CBK (H-2Kb transgenic mice on the CBA/Ca background (8)) and BM3.3 mice (H-2k, CBA/Ca background (9)) were kindly provided by A. Mellor (Medical College of Georgia, Augusta, GA). CBK mice were backcrossed to the CBA/CaJ strain for six generations. Resulting offspring carrying the transgenic H-2Kb allele were selected based on analysis of peripheral blood by cell surface staining and flow cytometry and backcrossed to CBA/CaJ mice. At the sixth generation, mice carrying the H-2Kb transgene were identified and intercrossed to establish a colony of CBK mice. These CBK mice did not reject skin grafts from CBA/CaJ mice, indicating that no minor antigenic differences exist between these strains. Mice were housed using microisolator conditions in autoclaved cages and maintained on irradiated feed and autoclaved acidified drinking water. All sentinel mice housed in the same colony were viral Ab free. Six- to 12-wk-old mice were used in all experiments. All experiments were conducted in accordance with institutional guidelines.
mAbs specific for H-2Kb (AF6-88.5), CD3 (2C11), CD4 (RM4-5), CD8 (53-6.7), CD19 (1D3), B220 (RA3-6B2), CD11b (Mac-1, M1/70), NK cells (DX5), CD11c (HL3), and Ly-6G (Gr-1, RB6-8C5) were obtained from BD Pharmingen (San Diego, CA). Anti-mouse neutrophil mAb (7/4) was obtained from Caltag Laboratories (Burlingame, CA).
Spleen harvest, magnetic cell sorting (MACS), and flow cytometry sorting (FACS)
Spleens from donor mice were harvested, and a single-cell suspension was prepared. CD4+ and CD8+ T cells as well as B220+ B cells were positively selected using magnetic beads (Miltenyi Biotec, Auburn, CA), according to the manufacturer’s instructions. The cells were then further purified by flow cytometry-based cell sorting based on expression of either CD4 and CD8 for T cells or B220 for B cells. Double-purified T cells, B cells, or T and B cells were transplanted together with bone marrow into recipients, as described in the text.
Bone marrow harvest and transplantation
Bone marrow cells were harvested from CBA/CaJ mice or BM3.3 mice treated 7 days prior with 5-fluorouracil (150 mg/kg). A depleting dose of 2 mg of anti-CD8 (116-13.1 (10)) and 1 mg of anti-CD4 (GK1.5 (11)) was given 4 days before bone marrow harvest to eliminate alloreactive T cells in vivo. Bone marrow was then harvested and transplanted into CBA/CaJ recipients, as described previously (12). We confirmed that bone marrow was >99% T cell depleted by cell surface staining and flow cytometry at the time of harvest. At the time of bone marrow transplantation, mice also received either purified CBA/CaJ T cells, CBK T cells, CBK B cells, or CBK T and B cells. In all experiments, recipient mice were treated with 10.25 Gy whole body irradiation 1 day before bone marrow transplantation.
Polymerase chain reaction
DNA was purified from blood leukocytes using a QIAamp blood DNA mini kit, according to the manufacturer’s instructions (Qiagen, Valencia, CA). Primer sequences used are as follows: Y-chromosome forward primer, 5′-CTCCTGATGGACAAACTTTACG-3′; Y-chromosome reverse primer, 5′-TGAGTGCTGATGGGTGACGG-3′; VβBM3.3 forward primer, 5′-GCAACTACAGTGGCTGTTCAC-3′; VβBM3.3 reverse primer, 5′-CGTATTTCCAACCCTGTCTGC-3′; β-actin forward primer, 5′-AACCCCAAGGCCAACCGCGAGAAGATGACC-3′; β-actin reverse primer, 5′-GGTGATGACCTGGCCGTCAGGCAGCTCGTA-3′. Semiquantitative PCR was performed using standard techniques and quantitated by comparison with standards containing a defined frequency of male to female cells. PCR results were quantitative by densitometry using Molecular Analyst Version 1.4.1 (Bio-Rad, Hercules, CA).
Tail skin grafting was performed and evaluated as previously described (13).
Long-term survival of allogeneic mature T cells after adoptive cell transfer into conditioned mice
To examine the ability of mature B and T cells to induce tolerance, we made use of MHC class I transgenic mice that express the H-2Kb gene under the control of its autologous promoter on the CBA/CaJ (H-2k) background (CBK mice (8)). We reasoned that by adoptively transferring purified CBK B or T cells into conditioned CBA/CaJ mice at the time of reconstitution with syngeneic bone marrow, it would be possible to specifically deliver the alloantigen, Kb, to developing T cells on defined lymphocyte subsets and examine their ability to induce tolerance. Furthermore, the use of purified CBK cells would allow us to conduct these experiments without complications associated with graft-vs-host disease.
CD4+ and CD8+ T cells as well as CD45R+ (B220) B cells were purified from the spleens of CBK mice. Following purification, T and B cell preparations were greater than 99% pure based on cell surface staining and flow cytometry (Fig. 1,a). To examine the ability of T and B cells to induce tolerance, CBA/CaJ mice were lethally irradiated and reconstituted with 106 T cell-depleted syngeneic CBA/CaJ bone marrow cells. At the time of bone marrow transplantation, the mice also received 107 purified CBK T or B cells. Mice that received CBK T cells together with syngeneic bone marrow contained CBK-derived T cells expressing Kb in their blood at 1, 2, and 24 wk after reconstitution (Fig. 1,b). No other cell types expressing Kb were detected (data not shown). In contrast, while B cells expressing Kb were detected in the blood of mice that received CBK B cells immediately after reconstitution, the frequency of B cells expressing Kb in these mice fell to undetectable levels by 2 wk after bone marrow transplantation (Fig. 1 b). No other cell type expressing Kb was detected in these mice at either early or late time points after reconstitution (data not shown). These data suggest that alloantigen-expressing T cells, but not B cells, are able to survive long-term after adoptive cell transfer into conditioned mice.
MHC-matched mature B cells survive long-term in irradiated recipients
The inability to detect Kb-expressing cells long-term in mice that received CBK B cells together with syngeneic bone marrow could have been the result of either rejection, or of a poor survival capacity of adoptively transferred B cells. If rejection of MHC class I disparate B cells resulted in elimination of the transferred cells, then adoptively transferred MHC-matched B cells should survive long-term. Alternatively, if B cells are simply unable to survive long-term after adoptive cell transfer, then MHC-matched B cells should also exhibit short-term survival. To address this issue, female CBA/CaJ recipients were reconstituted with 106 syngeneic female bone marrow cells, as described above, and received 107 purified B cells isolated from the spleens of male CBA/CaJ mice. Survival of male B cells was then analyzed in blood by PCR using Y chromosome-specific primers. Male CBA/CaJ B cells were readily detectable in the blood of these mice over a 26-wk follow-up period (Fig. 2,a). Approximately 2% of cells in the blood of these mice were of male origin and therefore derived from the transferred B cells (Fig. 2,b). The frequency of male B cells remained relatively stable over time (Fig. 2). Thus, in the absence of an MHC disparity, mature B cells are able to survive long-term in irradiated recipients.
MHC-mismatched B cells survive long-term in irradiated recipients when adoptively transferred with T cells
We hypothesized that allogeneic CBK T cells might be able to survive long-term upon adoptive cell transfer because, unlike B cells, they are able to induce tolerance to Kb. To test this hypothesis, CBA/CaJ mice were reconstituted with 106 syngeneic bone marrow cells and received a mixture of 5 × 106 T and 5 × 106 B cells from CBK mice. CBK-derived B and T cells expressing Kb on their surface were detectable in the blood of these mice at all time points analyzed over a 24-wk follow-up period (Fig. 1,b). The frequency of CBK-derived B cells was lower than the frequency of CBK T cells in these mice; however, the frequency of B cells expressing Kb remained stable over the 24-wk follow-up period (∼0.4 ± 0.1%, n = 5 in blood; Fig. 1,b). Furthermore, the frequency of B cells expressing Kb in mice receiving a mixture of CBK B and T cells at the time of bone marrow reconstitution was similar to that observed for MHC-matched male B cells (Fig. 2). Therefore, MHC class I disparate B cells are able to survive long-term in mice reconstituted with syngeneic bone marrow when transferred together with alloantigen-expressing T cells. These data suggest that allogeneic T cells protected the CBK-derived B cells from alloantigen-mediated rejection.
Expression of alloantigen on adoptively transferred T cells is sufficient to induce transplantation tolerance
We next examined whether protection mediated by T cells resulted from induction of tolerance to Kb. CBA/CaJ mice were reconstituted with T cell-depleted syngeneic bone marrow and received either 107 purified syngeneic CBA/CaJ T cells, CBK T cells, CBK B cells, or a mixture of both 5 × 106 CBK T and 5 × 106 CBK B cells. Ten weeks after reconstitution, mice received both CBK and B10.AKM/SnJ third party skin grafts. Mice that received CBK T cells (Fig. 3,a) or a mixture of allogeneic T and B cells (Fig. 3,b) accepted their CBK skin allografts long-term (>107 days), but were able to rapidly reject third party B10.AKM/SnJ skin. In contrast, mice that received CBK B cells (Fig. 3,c) or syngeneic mature CBA/CaJ T cells (Fig. 3 d) rapidly rejected both CBK as well as third party B10.AKM/SnJ skin grafts. Note that CBA/CaJ mice that received T cell-depleted syngeneic bone marrow transplantation alone also rapidly rejected both CBK and third party skin grafts (data not shown). These data demonstrate that expression of alloantigen on T cells alone is sufficient to induce transplantation tolerance, and suggest that adoptive cell transfer of CBK T cells protected CBK B cells by inducing tolerance to Kb.
Expression of Kb on adoptively transferred T cells prevents the development of alloreactive CD8 T cells
Transplantation tolerance is generally thought to result from either negative selection of alloreactive T cells in the thymus, central tolerance, or inactivation of alloreactive T cells in the periphery. To determine possible mechanisms by which allogeneic T cells are able to induce tolerance, we used BM3.3 TCR transgenic mice. BM3.3 mice (CBA/Ca background) express a transgenic TCR on CD8+ T cells that recognize the alloantigen Kb (9). The BM3.3 TCR can be detected by cell surface staining and flow cytometry using the anti-clonotypic Ab Ti98 (14). CBA/CaJ mice, conditioned as described above, were reconstituted with a mixture of T cell-depleted syngeneic CBA/CaJ and BM3.3 bone marrow and received either 107 CBK T cells, syngeneic CBA/CaJ T cells, or CBK B cells. Five weeks after reconstitution, blood cells were harvested, and development of clonotype-positive CD8+ T cells was analyzed by cell surface staining and flow cytometry. CD8+ T cells expressing the BM3.3 TCR were readily detectable in the blood of mice receiving either syngeneic CBA/CaJ T cells (n = 4) or CBK B cells (n = 5) (Fig. 4). In mice receiving syngeneic CBA/CaJ T cells, ∼7.9 ± 0.4% (n = 4) of CD8+ T cells were BM3.3 clonotype positive, while ∼8.8 ± 1.1% (n = 5) were clonotype positive in mice receiving CBK B cells. In contrast, we were unable to detect clonotype-positive CD8+ T cells in the blood of mice receiving CBK T cells (Fig. 4, n = 5, p < 0.001 between groups). These data suggest that adoptive cell transfer of CBK T cells at the time of bone marrow reconstitution prevented the development of Kb-specific BM3.3-expressing CD8 T cells in the blood.
Adoptively transferred T cells re-enter the thymus and mediate negative selection of newly developing alloreactive T cells
Analysis of T cells within the thymus of CBA/CaJ mice that were reconstituted with syngeneic bone marrow and received CBK T cells or a mixture of CBK T and B cells revealed the presence of Kb-expressing CD3+ T cells 1 wk after bone marrow transplantation (Fig. 5,a). We were unable to detect cells expressing Kb in the thymus of mice that received CBK B cells 1 wk after reconstitution (Fig. 5,a), even though at this time point CBK B cells are detectable in the blood (Fig. 1,a). Kb-expressing T cells persisted long-term in the thymus after adoptive transfer and were readily detectable 24 wk after reconstitution in mice receiving either CBK T or a mixture of CBK T and B cells (Fig. 5,b). We were unable to detect any other Kb-expressing lineages in these mice at either early or late time points (Fig. 5). Together, these data suggest that mature T cells are able to re-enter the thymus and deliver alloantigen to newly developing T cells.
To examine whether the presence of T cells expressing Kb affected development of alloreactive T cells, we analyzed the development of Kb-reactive BM3.3-derived CD8 T cells in the thymus of mice reconstituted with a mixture of CBA/CaJ and BM3.3 bone marrow that had received either syngeneic CBA/CaJ T cells, CBK B cells, or CBK T cells at the time of bone marrow transplantation. Five weeks after bone marrow transplantation, BM3.3 clonotype-positive CD4+, CD8+ double-positive T cells as well as CD8+ single-positive T cells were readily detectable in the thymus of mice receiving either syngeneic CBA/CaJ T cells or CBK B cells (Fig. 5,c). The thymi of mice that received CBK T cells contained a reduced number of BM3.3 clonotype-positive CD4+, CD8+ double-positive T cells when compared with controls (Fig. 5,c). These data suggest that BM3.3 bone marrow was able to engraft in mice receiving CBK T cells and develop into CD4+, CD8+ double-positive T cells. However, we were unable to detect BM3.3 clonotype-positive CD8+ single-positive T cells in the thymi of CBA/CaJ mice receiving CBK T cells (Fig. 5 c). These data suggest that the presence of CBK T cells in the thymus was sufficient to induce negative selection of developing alloreactive T cells.
Based on studies in bone marrow irradiation chimeras, it is clear that bone marrow-derived hemopoietic cells are potent inducers of tolerance (1, 2). Harnessing the ability of bone marrow-derived cells to induce tolerance has the potential to allow organ replacement without the need for lifelong immunosuppression, as well as to overcome allergy and autoimmunity. The use of bone marrow to induce tolerance leads to robust tolerance. However, transplantation of allogeneic bone marrow to induce tolerance to allogeneic transplants requires significant host conditioning to allow engraftment (15), often leads to graft-vs-host disease (16), and can lead to immunoincompetence (17). To overcome these problems, it would be desirable to identify a long-lived hemopoietic cell lineage capable of inducing tolerance upon adoptive transfer, thereby avoiding problems associated with transplantation of allogeneic bone marrow.
Using an adoptive cell transfer system, we show that mature MHC class I disparate T cells are able to survive long-term when adoptively transferred into conditioned hosts. Mature T cells re-entered the thymus and mediated negative selection of newly developing alloreactive CD8 T cells. Tolerance induced by alloantigen-expressing T cells permitted long-term survival of skin allografts that were matched to the T cell donor strain. Adoptive cell transfer of allogeneic B cells was not sufficient to induce tolerance. The inability of mature B cells to induce tolerance was not related to poor survival capacity, because MHC-matched B cells exhibited long-term survival upon adoptive cell transfer. Furthermore, MHC-mismatched B cells exhibited long-term survival when adoptively transferred with allogeneic T cells. However, expression of Kb on T cells was necessary to induce tolerance.
The ability of mature T cells harvested from peripheral lymphoid tissue to circulate to the thymus has been reported previously (18, 19, 20, 21); however, the function of these cells has not previously been determined. We suggest that the ability of T cells to induce tolerance may be related to their ability to efficiently deliver Ag to the thymus. Mature B cells, as shown in this study, and APCs, as shown in other studies (7), do not efficiently circulate to the thymus, and therefore may not deliver Ag to newly developing T cells. It has been suggested that resting mature T cells recirculate into the thymus inefficiently when compared with activated T cells (21, 22). In our experiments, we were able to detect a relatively high frequency of mature allogeneic T cells in the thymus. It is therefore possible that the CBK T cells became activated either during cell sorting, or in the irradiated host as a result of homeostatic expansion.
Mature T cells were the only cell type required to express alloantigen to induce tolerance; however, other cell types may be involved in tolerance induction. It has previously been shown that purified splenic T cells are potent inducers of neonatal tolerance to endogenous mouse mammary tumor virus superantigens (vSAgs).3 In these experiments, mature T cells expressing vSAgs were able to enter the thymus and transfer the soluble vSAgs to APC-expressing MHC class II, which then mediated negative selection of vSAg-reactive Vβ6+ T cells (23). Unlike vSAgs, however, Kb is a transmembrane cell surface protein, and therefore would have to be transferred to other cell types in the thymus upon being shed from the cell surface. Alternatively, it is possible that a previously unreported mechanism may allow other cell types, such as APCs, to pick up Kb from the surface of mature T cells and present intact Kb molecules to newly developing T cells. Insofar as we were able to detect Kb only on the surface of T cells, we suggest that if intact Kb is transferred to other cell types, it occurs at a low efficiency. Nevertheless, we have previously shown that Kb disparate skin grafts can be rejected by alloreactive CD4 T cells that recognize peptides derived from Kb presented on donor or host MHC class II (13). To prevent rejection through this pathway, T cells that recognize Kb-derived peptides in the context of MHC class II must also be tolerized by the transferred alloantigen-expressing T cells. Because mouse T cells do not express MHC class II, this would imply the involvement of MHC class II-positive cells to induce full tolerance and acceptance of allogeneic skin.
It has been reported that in transgenic mice, expression of allogeneic MHC class I under the control of the human CD2 promoter was not sufficient to induce complete T cell tolerance and acceptance of allogeneic skin grafts (6). Although alloreactive CD8+ T cells were deleted by the expression of alloantigen on lymphoid cell types, skin allografts were still rejected in a CD4+ T cell-dependent fashion. It is possible that the differences observed between these two models are due to nonphysiologic expression of the alloantigen in transgenic mice, or to insertional mutation by the transgene generating minor Ag disparities.
Our data suggest that the delivery of Ag to the thymus by mature T cells can reshape the immune repertoire. This observation opens up the possibility that modification of syngeneic T cells to express either transplantation Ags, or perhaps self Ags in patients suffering from autoimmunity, by genetic engineering could represent a novel approach to tolerance induction, and has the potential to provide clinically relevant therapies. In addition, these findings may have significant implications for infectious diseases such as HIV, in which the expression of viral Ags on the T cells of chronically infected individuals may prevent the development of effective T cell responses to the virus.
We thank Drs. Henry Winn and Shiv Pillai for critical review of the manuscript, the members of the Iacomini laboratory for helpful discussions, and Dr. Andrew L. Mellor for generously providing BM3.3 and CBK transgenic mice.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Grant RO1 AI43619 from the National Institutes of Health (to J.I.). J.B. is supported in part by a grant from the Children’s A-T Project. D.F. is supported in part by National Institutes of Health Training Grant T32 AI07529.
Abbreviation used in this paper: vSAg, virus superantigen.