The survival of memory T cells is critical to vaccination strategies for infectious diseases and cancer, whereas their elimination may be crucial for treatment of autoimmune states. We examined the consequences of gamma-irradiation, which induces apoptosis of memory T cells in vitro, on the memory response to MHC class I alloantigen in vivo. Sublethal gamma-irradiation of primed mice eliminated accelerated rejection of skin allografts but failed to induce tolerance. Accelerated rejection was restored in irradiated mice by infusion of bone marrow cells expressing the priming alloantigen on immunostimulatory APCs (dendritic cells), whereas the memory response was not restored by infusion of bone marrow cells expressing the priming alloantigen on nonstimulatory APCs (B cells). Strikingly, irradiated mice infused with nonstimulatory bone marrow APCs exhibited long-term survival or tolerance to skin grafts expressing the priming MHC class I alloantigen. The mechanism of tolerance in this setting is explored.

The survival of memory T cells mediating immunity to infectious agents and potentially to tumor Ags is crucial to the success of vaccination programs. Reciprocally, the elimination of memory T cells mediating autoimmune disease is the paramount goal of tolerance-inducing strategies. The susceptibility of T cell memory populations to apoptosis in response to stresses, including gamma-irradiation, has been elucidated primarily using T cell clones or lines in vitro. In the resting state, such memory T cells are highly susceptible to induction of apoptosis, which is attributable to down-regulation of the anti-apoptotic molecule bcl-2, up-regulation of the pro-apoptotic molecule fas, and failure to generate IL-2 (1, 2, 3, 4, 5). It was further demonstrated that memory T cells can be rescued from radiation-induced apoptosis by Ab-mediated ligation of CD3 and CD28 molecules, mimicking Ag stimulation (4), or by the addition of IL-2 (3). The relevance of these findings to T cell-mediated memory responses in vivo is not known.

To assess the fate of the memory T cell response in vivo to apoptosis-inducing regimens, we used an allogeneic skin transplantation model to generate memory T cells and to assess the memory response after sublethal gamma-irradiation. The effects of Ag presentation by stimulatory vs nonstimulatory APCs, after irradiation, were also studied.

CD2-Dd mice (line 4906), in which the Dd cDNA coding sequence (a gift of Dr. Randy Ribaudo, Molecular Applications Group, Palo Alto, CA) was ligated into the human CD2 promoter and enhancer expression cassette p29Δ2(Sal−) (6, 7), and MHC-Dd mice (line 3604), in which the genomic Dd gene, including the MHC class I promoter, was isolated from pDd1 (a gift of Dr. Gilbert Jay, Origene Technologies, Rockville, MD), (8) have been described previously (9).

FVB/N (FVB)2 mice, an inbred H-2q strain, were bred in-house or purchased (Taconic Farms, Germantown, NY). Mice were primed by a single injection (i.v. or i.p.) of spleen and/or bone marrow cells (BMC) or by rejection of skin from MHC-Dd mice.

A minimum of 2 wk after priming by i.p. or i.v. injection of MHC-Dd spleen cells or BMC or 2 wk after the rejection of skin allografts from MHC-Dd mice, primed mice were irradiated and injected with BMC from CD2-Dd mice (generating CD2-Dd BMT mice) or MHC-Dd mice (generating MHC-Dd BMT mice). BMC were harvested from posterior limbs and cervical vertebrae of donor mice and washed three times in HBSS containing 5% FCS and 1% HEPES buffer. A total of 30 × 106 BMC (in 0.5 ml sterile PBS, calcium and magnesium free) were infused into hosts at 1–3 h after irradiation (500–600 rad, as specified) in a Cs137 source (Gammacell 40 irradiator, Nordion, Ontario, Canada).

Spleens, thymi, and bone marrow of tolerant mice were harvested and lysed with ACK (ammonium chloride, postassium carbonate; Quality Biological, Gaithersburg, MD) and dually stained with mAbs to Dd (PE, 06134D, PharMingen, San Diego, CA) and CD4 (FITC, PharMingen), CD4 and CD8 (FITC, PharMingen), or CD34 (FITC, 09434D, PharMingen). Cells were run on a FACScan (Becton Dickinson, Mountain View, CA), and data were analyzed using the CellQuest software program (Becton Dickinson). Ab-coated cells were evaluated by staining with goat anti-mouse (GAM) FITC (12064D, PharMingen). Triple staining was performed on thymocytes of tolerant (CD2-Dd BMT) mice, using CD4-PE (PharMingen), CD8-biotin(PharMingen) plus streptavidin-Quantum Red (Sigma, St. Louis, MO), and Dd-FITC (PharMingen).

Bone marrow was dually stained for Dd (PE, 06135A, PharMingen, or biotin, 0613D, with streptavidin-Quantum Red conjugate) and either c-kit (FITC, 01904D, PharMingen), an Ag expressed on short-term and self-renewing hemopoietic stem cells, or CD34 (09434D, PharMingen), an Ag expressed on 7–10% of BMC identified as short-term multilineage progenitor cells (10).

Mice were engrafted on the flank with tail skin from donor mice according to published methods (11). Grafts were scored daily or every other day until rejection, which was considered to be a loss of >50% of graft tissue.

BMC were harvested from tolerant mice, and 4 × 107 of these cells were infused into naive mice that had been irradiated with 600 rad several hours previously. Mice were bled at 7 wk after BMT and engrafted with Dd skin 16 wk after BMT.

FVB mice were grafted with skin from transgenic mice (FVB) that express Dd under the control of the MHC promoter (MHC-Dd mice) (9). Previously, we demonstrated that in MHC-Dd mice, the MHC promoter induces ubiquitous transgene expression, and that bone marrow-derived dendritic cells (DC), which are the APCs that trigger rejection responses to tissue grafts (12, 13), express low to moderate levels of Dd (9). Thus, skin from MHC-Dd mice, which contains a significant population of DC, is immunogenic: MHC-Dd tail skin grafts are rejected by naive FVB mice in 9–16 days (median survival time (MST) = 12 days, n = 57, Table I) and generate memory T cells that reject second MHC-Dd skin grafts in an accelerated time frame (7–10 days, MST = 8 days, n = 29, Table I and Fig. 1,A). Interestingly, sublethal irradiation of primed mice abrogated the memory response and induced delayed rejection of Dd skin grafts (Fig. 1 A, MST 24.5 days, n = 10). Thus, signals known to induce apoptosis of memory T cells in vitro disrupt memory T cells in vivo.

Table I.

Rejection of MHC-Dd skin grafts in resting primed micea

HostSkin DonorTreatmentRejected No.MST (days)
FVB MHC-Dd None 57 /57 12 
FVB 1° Dd MHC-Dd None 29 /29 
FVB 1° Dd MHC-Dd 500 rad 19 /19 13 
FVB 1° Dd MHC-Dd 500 rad plus CD2-Dd BMC 9 /21 >109 
FVB 1° Dd MHC-Dd 500 rad plus MHC-Dd BMC 10 /10 7.5 
HostSkin DonorTreatmentRejected No.MST (days)
FVB MHC-Dd None 57 /57 12 
FVB 1° Dd MHC-Dd None 29 /29 
FVB 1° Dd MHC-Dd 500 rad 19 /19 13 
FVB 1° Dd MHC-Dd 500 rad plus CD2-Dd BMC 9 /21 >109 
FVB 1° Dd MHC-Dd 500 rad plus MHC-Dd BMC 10 /10 7.5 
a

Primed mice were exposed to 500 rad and were either not treated or infused with 3 × 107 MHC-Dd BMC or CD2-Dd BMC within several hours. Mice were bled 3 wk after BMT to assess chimerism and lymphocyte populations and engrafted with MHC-Dd skin a minimum of 4 wk after BMT. Naive significantly differs from primed (untreated) (p < 0.0001 by log-rank); primed mice treated with 500 rad do not statistically differ from naive mice (p = 0.0764) but differ from primed mice (p < 0.0001); 500 rad plus CD2-Dd BMC-treated mice differ significantly from all other groups (p ≤ 0.0012), and 500 rad plus MHC-Dd BMC-treated mice do not differ from primed mice (p = 0.8544), but differ significantly from all other groups (p ≤ 0.0008).

FIGURE 1.

Fate of the T memory response after sublethal γ-irradiation of primed mice. A, Rejection of Dd skin grafts in resting primed mice. Primed mice were exposed to 550 rad and infused with 3 × 107 BMC within several hours or not infused. Mice were bled 3 wk after BMT to assess chimerism and lymphocyte populations and engrafted with Dd skin a minimum of 4 wk after BMT. This figure was compiled from three experiments. B, Rejection of Dd skin grafts in “activated” primed mice. Primed mice were regrafted with MHC-Dd skin 3 days before irradiation (600 rad) and BMT. MST of skin grafts: 600 rad = 10 days; 600 rad plus MHC-Dd BMC = 9 days; 600 rad plus CD2-Dd BMC = 8 days.

FIGURE 1.

Fate of the T memory response after sublethal γ-irradiation of primed mice. A, Rejection of Dd skin grafts in resting primed mice. Primed mice were exposed to 550 rad and infused with 3 × 107 BMC within several hours or not infused. Mice were bled 3 wk after BMT to assess chimerism and lymphocyte populations and engrafted with Dd skin a minimum of 4 wk after BMT. This figure was compiled from three experiments. B, Rejection of Dd skin grafts in “activated” primed mice. Primed mice were regrafted with MHC-Dd skin 3 days before irradiation (600 rad) and BMT. MST of skin grafts: 600 rad = 10 days; 600 rad plus MHC-Dd BMC = 9 days; 600 rad plus CD2-Dd BMC = 8 days.

Close modal

To examine signals that might rescue the function of memory T cells, we stimulated memory T cells after irradiation with cell populations that express the priming alloantigen on immunostimulatory or nonstimulatory APCs. As a source of cells containing stimulatory APCs, we used MHC-Dd BMC, because they prime naive FVB mice to Dd, as shown by accelerated skin graft rejection (MST = 8 days, n = 9) in mice previously infused with such cells. For nonstimulatory APCs, we used BMC from transgenic mice in which the Dd transgene was regulated by the CD2 promoter, and therefore expressed solely on lymphocytes (CD2-Dd mice) (9). In previous work, we showed that BMC of CD2-Dd mice express the transgene on relatively mature B cells (B220+) but not on B progenitors (6C3+), cells of granulocyte/macrophage (Gr-1+) lineage, or DC (CD11c+) (9). As predicted from the tolerogenicity of resting B cells (14, 15), injection of CD2-Dd BMC induced hyporesponsiveness to Dd in naive recipients, as shown by delayed skin graft rejection (MST = 21.5 days, n = 6). Further evidence of the tolerogenicity of CD2-Dd BMC was that they engrafted in naive FVB mice at a low level of conditioning (300 rad), generated lymphocyte chimerism, and induced Ag-specific hyporesponsiveness (9). In contrast, using the same conditioning regimen, transplantation of MHC-Dd BMC did not produce engraftment, and instead provoked immunity as shown by accelerated rejection of MHC-Dd skin grafts (9). Consistent with these findings, studies in vitro using MHC-Dd and CD2-Dd spleen cells as stimulators in MLR assays revealed that CD2-Dd spleen cells poorly elicited Dd-specific CTLs (data not shown).

FVB mice primed to Dd were irradiated with 550 rad and infused with MHC-Dd or CD2-Dd BMC within several hours. Remarkably, accelerated rejection was preserved in irradiated mice infused with MHC-Dd BMC (Fig. 1 A, MST = 8, n = 9; p = 0.0002, by log-rank); the kinetics of rejection did not differ statistically from those of unmanipulated primed animals (p = 0.12). This finding demonstrates that the T memory response, which is abrogated in vivo by sublethal gamma-irradiation, is protected by presentation of the priming Ag on immunostimulatory APCs.

Infusion of CD2-Dd BMC into irradiated primed mice (CD2-Dd BMT mice) not only failed to preserve the T memory response but also led to long-term survival or tolerance of MHC-Dd skin grafts (Fig. 1,A, MST > 111 days, n = 15; p = 0.0006, by log-rank). Similar results were observed in mice conditioned with 500 rad (Table I). We do not think that the failure of CD2-Dd BMC to preserve the memory response after irradiation is attributable to their lower Dd expression level compared with MHC-Dd cells (both on a per cell basis as well as on total cells), for two reasons: first, infusion of CD2-Dd BMC into naive mice dramatically altered the rejection response to Dd skin grafts (9), indicating that although low, the level of Ag was sufficient to mediate effects on Ag-specific cell populations in vivo; second, infusion of whole spleen or T cell-depleted CD2-Dd spleen cells into primed irradiated mice failed to preserve the memory response (data not shown and Table II), despite expression of Dd on virtually all cells (Fig. 6, upper left panel, PBLs of CD2-Dd mice). Thus, presentation of Ag in the absence of effective costimulatory interactions failed to protect T cell-mediated memory from irradiation. Furthermore, although spleen cells induced hyporesponsiveness to Dd, tolerance was not evident in most spleen cell-injected mice. This finding may pertain to the limited life span of spleen cells, and thus to the eventual loss of exposure to alloantigen; in addition, this finding is in contrast to the tolerance observed in mice treated with BMC that are self-renewing and thus provide continual exposure to alloantigen. Taken together, these results indicate that the presence of Ag is requisite for maintaining tolerance.

Table II.

Persistence of Ag is required for tolerance inductiona

HostTreatmentRejected No.MST (days)
FVB 1° Dd None 5 /5 
FVB 1° Dd 500 rad 9 /9 11 
FVB 1° Dd 500 rad plus CD2-Dd T- depleted spleen cells 7 /10 22 
FVB 1° Dd 500 rad plus CD2-Dd BMC 2 /10 >97 
HostTreatmentRejected No.MST (days)
FVB 1° Dd None 5 /5 
FVB 1° Dd 500 rad 9 /9 11 
FVB 1° Dd 500 rad plus CD2-Dd T- depleted spleen cells 7 /10 22 
FVB 1° Dd 500 rad plus CD2-Dd BMC 2 /10 >97 
a

Primed mice were exposed to 500 rad and either untreated or infused with 3 × 107 CD2-Dd T depleted spleen cells or BMC within several hours. Mice were engrafted with MHC-Dd skin a minimum of 4 wk after BMT or adoptive transfer of spleen cells.

FIGURE 6.

BMC of tolerized mice (CD2-Dd BMT mice) contain CD2-Dd progenitor cells that give rise to mature lymphocytes in secondarily transplanted hosts. BMC were harvested from tolerant CD2-Dd BMT mice, and 4 × 107 of these cells were infused into naive recipients conditioned with 600 rad. Naive recipients were bled at 7 wk posttransplant, and PBLs were assessed for expression of Dd and lymphocyte markers. The PBL profiles of two tolerant donor mice (which lack lymphocyte chimerism) and two naive recipients of tolerant BMC (of four) are displayed. The mean percentage of CD3+ Dd+ T cells in four naive recipients was 20%; the mean percentage of B220+ Dd+ B cells was 5%.

FIGURE 6.

BMC of tolerized mice (CD2-Dd BMT mice) contain CD2-Dd progenitor cells that give rise to mature lymphocytes in secondarily transplanted hosts. BMC were harvested from tolerant CD2-Dd BMT mice, and 4 × 107 of these cells were infused into naive recipients conditioned with 600 rad. Naive recipients were bled at 7 wk posttransplant, and PBLs were assessed for expression of Dd and lymphocyte markers. The PBL profiles of two tolerant donor mice (which lack lymphocyte chimerism) and two naive recipients of tolerant BMC (of four) are displayed. The mean percentage of CD3+ Dd+ T cells in four naive recipients was 20%; the mean percentage of B220+ Dd+ B cells was 5%.

Close modal

Because Ag stimulation rescued the T memory response after gamma-irradiation, we wished to ascertain whether activation of memory cells before irradiation protected these cells. Primed mice were therefore engrafted with MHC-Dd skin 3 days before irradiation and BMT. Regardless of whether Ag was subsequently presented or the form in which it was presented, all groups rejected their skin grafts in an accelerated fashion (Fig. 1 B), indicating that once activated, the memory response is resistant to disruption by irradiation.

The basis for tolerance in primed mice transplanted with CD2-Dd BMC was sought. Transgene-containing cells were detected in the spleens, bone marrow, and thymi of tolerant mice by a DNA PCR-based assay (data not shown), indicating that tolerant mice were chimeric. This finding was not surprising in view of observations that the rejection of BMC is Ag-specific, requiring expression of a target alloantigen (16), and that CD2-Dd BMC progenitors do not express target Dd alloantigen (Fig. 2). Despite the detection of transgene-positive cells in peripheral lymphoid tissues by molecular means, Dd-expressing cells were detected by FACS analysis only in the thymi of tolerant mice (Fig. 3,A). Failure to detect Dd-expressing cells in the PBLs or secondary lymphoid organs of tolerant mice by FACS was not attributable to a block in T cell development, as single positive thymocytes (CD4+CD8 or CD4CD8+) expressing Dd were generated in tolerant mice (Fig. 4); rather, this failure was attributable to the presence of Dd-specific Abs, which were detected in sera (Fig. 5). Analogous to the mechanism by which cell populations are eliminated in vivo by the administration of Ab directed to a cell surface determinant, endogenously produced Dd Abs could coat Dd-expressing cells and target them for sequestration and destruction in the reticuloendothelial system (RES). In support of such a possibility was the finding that a substantial population of thymocytes in tolerant mice was coated with Ab before egress (i.e., they were Ig-positive, as detected by an anti-mouse Ig (Fig. 3 B), but were not B cells, as B220+ cells comprised <5% of total thymocytes (data not shown)). Although primed mice transplanted with MHC-Dd BMC possessed titers and isotypes of Dd-specific Abs in serum (IgG1 and IgG2, data not shown) that were similar to those seen in CD2-Dd transplanted mice, thymic chimerism was not detected in MHC-Dd transplanted mice (data not shown), suggesting that the establishment of chimerism was necessary for tolerance. Interestingly, thymic chimerism was not always sufficient to induce tolerance, as CD2-Dd BMT mice that rejected Dd skin grafts also displayed thymic chimerism (data not shown).

FIGURE 2.

Dd expression on bone marrow progenitor cells of transgenic mice. BMC were isolated as described previously and stained in a two-color FACS analysis. A gate was placed on the CD34+, c-kit+, or B220+ cell population, and Dd expression was assessed within the gate.

FIGURE 2.

Dd expression on bone marrow progenitor cells of transgenic mice. BMC were isolated as described previously and stained in a two-color FACS analysis. A gate was placed on the CD34+, c-kit+, or B220+ cell population, and Dd expression was assessed within the gate.

Close modal
FIGURE 3.

Chimerism in tolerant mice. A, Tolerant mice were sacrificed a minimum of 7 mo after BMT, and harvested tissues were examined for Dd expression in a two-color FACS analysis. Data are shown for a primed control and two tolerant mice. Data in tolerant mice are representative of 27 examined. The percentage of Dd+ cells detected in the thymus ranged from 0% to 89%, with a median of 44%. B, Assessment of Ab-coated cells in the thymi of tolerant mice. The thymi of tolerant mice or control primed mice were stained in a single-color analysis using GAM Ig-FITC. The percentage of B cells (detected by the B220 marker) did not exceed 4%.

FIGURE 3.

Chimerism in tolerant mice. A, Tolerant mice were sacrificed a minimum of 7 mo after BMT, and harvested tissues were examined for Dd expression in a two-color FACS analysis. Data are shown for a primed control and two tolerant mice. Data in tolerant mice are representative of 27 examined. The percentage of Dd+ cells detected in the thymus ranged from 0% to 89%, with a median of 44%. B, Assessment of Ab-coated cells in the thymi of tolerant mice. The thymi of tolerant mice or control primed mice were stained in a single-color analysis using GAM Ig-FITC. The percentage of B cells (detected by the B220 marker) did not exceed 4%.

Close modal
FIGURE 4.

Dd+ T cells develop normally in the thymi of tolerant mice. A three-color analysis was performed on thymus cells of tolerant mice (CD2-Dd BMT mice). Thymocytes were triply stained for CD4, CD8, and Dd. A gate was placed around Dd+ cells only, Dd− cells only, or on all thymocytes, and the CD4 vs CD8 profile was examined.

FIGURE 4.

Dd+ T cells develop normally in the thymi of tolerant mice. A three-color analysis was performed on thymus cells of tolerant mice (CD2-Dd BMT mice). Thymocytes were triply stained for CD4, CD8, and Dd. A gate was placed around Dd+ cells only, Dd− cells only, or on all thymocytes, and the CD4 vs CD8 profile was examined.

Close modal
FIGURE 5.

Presence of Dd-specific Ab in the sera of tolerant mice, but not in the sera of naive mice transplanted with CD2-Dd BMC. Sera were obtained from primed mice transplanted with CD2-Dd BMC, which were tolerant to Dd skin grafts, and naive mice transplanted with CD2-Dd BMC, which were hyporesponsive but not tolerant to Dd skin grafts; serial dilutions were performed. Sera were incubated with thymocytes from MHC-Dd transgenic mice as a first step. After washing, cells were incubated with GAM FITC mAb. After extensive washing, a FACS analysis was performed. Control GAM staining is superimposed on each dilution. Reactivity of all sera on control parental (FVB) thymocytes was undetectable at all dilutions.

FIGURE 5.

Presence of Dd-specific Ab in the sera of tolerant mice, but not in the sera of naive mice transplanted with CD2-Dd BMC. Sera were obtained from primed mice transplanted with CD2-Dd BMC, which were tolerant to Dd skin grafts, and naive mice transplanted with CD2-Dd BMC, which were hyporesponsive but not tolerant to Dd skin grafts; serial dilutions were performed. Sera were incubated with thymocytes from MHC-Dd transgenic mice as a first step. After washing, cells were incubated with GAM FITC mAb. After extensive washing, a FACS analysis was performed. Control GAM staining is superimposed on each dilution. Reactivity of all sera on control parental (FVB) thymocytes was undetectable at all dilutions.

Close modal

Additional support for an Ab-mediated loss of Dd+ lymphocytes is the finding that peripheral chimerism is readily apparent in transplanted mice that lack Ab (i.e., in naive mice transplanted with CD2-Dd BMC (Ref. 9 and Fig. 5) and more persuasively, in the blood of naive mice transplanted with BMC from tolerant mice (Fig. 6)). Thus, transplant of CD2-Dd BMC from tolerant mice, which contain Dd Ab and lack peripheral chimerism, into naive recipients that lack Ab regenerates peripheral chimerism. Interestingly, unlike the tolerant BMC donors, naive recipients of tolerant BMC were not tolerized to Dd (MST of MHC-Dd skin = 29 days, n = 3). Taken together with the observation that naive mice transplanted with CD2-Dd BMC were also not tolerized by the same regimen that induced tolerance in primed mice, these findings establish a correlation between induction of tolerance and the presence of Dd-specific Abs.

The presence of Dd-specific IgG Abs in tolerant mice and the finding that Dd-specific CTLs were readily elicited from such mice (data not shown) indicated that Dd-specific Th cells, CTLs, and B cells had not been tolerized. Furthermore, tolerance was Ag-specific, in that third-party DBA/1 skin grafts were rejected by tolerant mice in a time course that was comparable with that seen for control mice (data not shown). Taken together, these results imply that skin graft tolerance in this setting is mediated by the suppression of Ag-specific effector cells, and that Ab may have a role in the induction and/or maintenance of tolerance. These possibilities are under investigation.

This study demonstrates that the memory response to an MHC class I alloantigen is abrogated by sublethal gamma-irradiation, and that within this setting, presentation of Ag by immunostimulatory APCs preserves memory, whereas presentation of Ag by nonstimulatory APCs fails to preserve T cell memory. This study further reveals that activation of the memory response in vivo protects it against the disruptive effects of irradiation, regardless of the context in which Ag was subsequently presented. These results are consistent with induction of apoptosis of memory T cells by sublethal γ-irradiation and rescue by signaling through TCR and costimulatory molecules, as has been observed in vitro (4). Such interactions are thought to confer survival by up-regulating anti-apoptotic molecules such as bclX-L, the induction of cytokine production, and the stimulation of proliferation (3, 4, 17, 18). Alternatively, memory cells may not be deleted but rather functionally inactivated by irradiation such that they fail to produce IL-2 and therefore fail to proliferate. To study these possibilities, Dd tetramers are being constructed to follow the fate of Dd-specific memory T cells, and 2C TCR transgenic mice will further be used to follow the fate of Ld-specific memory cells.

The failure of CD2-Dd BMC to rescue the T memory response was puzzling, as many studies have supported less stringent costimulatory requirements in the activation of T memory vs naive T cell populations (19, 20, 21, 22). Interestingly, the level of proliferation and cytokine production of memory CD4+ T cells prompted by B cell APCs, which are deficient in costimulatory activity (23), was substantially lower than that produced by DC or activated B cell APCs (21), which provide potent costimulatory activity (23), suggesting that insufficient cytokine activity may be elicited by B cell APCs to rescue memory cells from radiation lethality in vivo. Other studies have challenged the notion that memory T cells are less dependent upon costimulatory interactions (24, 25). Although an argument could be made that differences in the level of Dd transgene expression on MHC-Dd vs CD2-Dd BMC (Fig. 2) may factor into the observed pattern of rescue vs ablation of the T memory response, we do not think that this is the case. Previous studies in several lines of CD2-Dd transgenic mice, which were distinguished from each other on the basis of the level of Dd expression on lymphocytes, revealed that the degree of nonresponsiveness to Dd correlated precisely with the level of transgene expression (9), so that the higher the expression level, the more profound the degree of nonresponsiveness. This finding is consistent with that of other transgenic systems, in which low levels of transgene expression incurred positive thymic selection, whereas higher levels incurred negative selection (26). Thus, our prediction is that boosting the expression level of the transgene in CD2 transgenic mice should enhance the tolerizing capacity of its BMC populations.

Finally, these findings have interesting implications for treatment strategies of autoimmune disease and cancer, in which commonly used therapeutics (steroids, irradiation, and chemotherapy) may delete memory T cells (3). Thus, after such treatments, persistence of Ag in an immunogenic form may rescue autoimmune T cells mediating disease, an unfavorable outcome. The data further suggest that the optimal time for definitive treatment of diseases mediated by autoimmune T memory cells is when the disease is in remission, because of the heightened susceptibility to apoptosis of resting vs activated memory T cells. Reciprocally, in the case of cancer therapies that induce apoptosis of memory T cells, persistence of Ag on tumor cells, the vast majority of which lack costimulatory capacity, might have the adverse outcome of inducing or maintaining tolerance to putative tumor Ags.

Regarding the association of Ab with tolerance in primed mice transplanted with CD2-Dd BMC, Ab-mediated immune suppression (enhancement) is a phenomenon that has been observed as a result of passive administration of graft-specific Ab or as a result of active induction by prior administration of Ag alone or Ag in conjunction with Ab (27, 28, 29, 30, 31). Suppression of graft rejection by this means has been most effective in the setting of directly vascularized renal allografts in naive mice; enhancement of skin grafts has been observed, but the effects were relatively mild, and no enhancement was observed in primed mice (32). Nonetheless, tolerance in our mice is consistent with many aspects of Ab-mediated immune suppression based on the following similarities: the Abs observed in our tolerant mice were of the “enhancing” IgG1 and IgG2 phenotypes, although such Abs failed to enhance skin graft survival when infused into naive FVB mice engrafted with Dd skin (data not shown); immune reactivity to Dd was preserved, as levels of Dd-specific CTLs in tolerant mice were equivalent to those of primed mice, implying no significant deletion or anergizing of Dd-specific T cells; maintenance of tolerance required the continuous presence of Ag; and finally, suppression was Ag-specific. The failure of thymically mediated mechanisms of tolerance to reduce CD8+ Dd-specific CTLs in tolerant mice, which lysed MHC class II-negative Dd+ P815 cells equivalently to primed mice, was puzzling given the abundance of Dd-expressing T cells in the thymus and previous evidence that expression of class I alloantigen on T cells deletes allospecific CD8+ thymocytes (33). This raises the intriguing possibility that Dd-specific Ab diminished Dd expression on thymocytes to the point at which Dd allospecific T cells were actually positively selected, as has been observed in other systems (26).

Suppressor T cells, which have been found to be associated with tolerance in some models of Ab-mediated suppression, were assessed in our tolerant mice by adoptive transfer of negatively selected T cells into thymectomized sublethally irradiated hosts. Although Dd-specific (but not third-party) rejection responses were suppressed in the adoptively transferred hosts (data not shown), these results were confounded by the finding that a substantial population of non-B, non-T cells were included in the negatively selected T cell population, the identity and role of which are not known, and by the development of a B cell malignancy in many of the adoptively transferred animals. Because plasma cells secreting Ab are radiation-resistant (34, 35), can be long-lived (36), and do not express B220 and MHC class II (37), it is possible that tolerance in the adoptively transferred mice was also mediated to some extent by Ab. This possibility is being investigated.

We think that tolerance in CD2-Dd BMT mice is most likely mediated by a mechanism involving opsonization of Dd-specific T cells (Ag reactive cell opsonization (ARCO)) (30), potentially abetted by suppressor T cells. ARCO could occur in our model by the generation of a tripartite complex consisting of a Dd+ lymphocyte to which is bound a Dd-specific T cell and Dd-specific Abs. Opsonization of this complex in the RES could be mediated through FcR binding and uptake. We think that radiation generates the mechanics of ARCO in our model by eliminating or inhibiting Dd-specific memory and naive T cells, establishing marrow engraftment and thymic chimerism, and leaving unchecked the production of Dd-specific Ab by plasma cells. Thus, as new Dd alloreactive cells develop in and exit the thymus, they do so in a milieu that is laden with Dd-expressing T cells and anti-Dd Ab, leading to the formation of an Ab-bound cellular conjugate. The presence of Dd+ cells in the thymus, some of which are bound by Ab, their absence in the periphery, and the presence of Dd-specific Ab in the sera support this possibility. Such a mechanism would prevent the mobilization of Dd-specific T cells to the skin allograft by diverting cells to the RES, where they may be trapped and destroyed. The generation of Dd-specific CTLs from the spleens of tolerant mice suggests that Dd-reactive cells are sequestered but not destroyed. Dd-specific T cells may be further blocked from engaging Ag on skin grafts by Ab coating of graft cells. Thus, it is possible that tolerance in our model represents a form of immune “diversion” rather than a truly tolerant state.

We thank Dr. J. Tiwari for statistical analyses; Karen Mason, Muffin Brand, and Anthony Ferrine for expert care of experimental animals; and Drs. Wendy Shores, Michael Norcross, Dennis Klinman, Melanie Vacchio, and Al Singer for critical review of the manuscript.

2

Abbreviations used in this paper: FVB, FVB/N; BMC, bone marrow cell(s); BMT, bone marrow transplantation; GAM, goat anti-mouse; DC, dendritic cell(s); MST, median survival time; RES, reticuloendothelial system; ARCO, Ag reactive cell opsonization.

1
Akbar, A. N., N. Borthwick, M. Salmon, W. Gombert, M. Bofill, N. Shamsadeen, D. Pilling, S. Pett, J. E. Grundy, G. Janossy.
1993
. The significance of low bcl-2 expression by CD45RO T cells in normal individuals and patients with acute viral infections: the role of apoptosis in T cell memory.
J. Exp. Med.
178
:
427
2
Tuosto, L., E. Cundari, E. S. G. Montani, E. Piccolella.
1994
. Analysis of susceptibility of mature human T lymphocytes to dexamethasone-induced apoptosis.
Eur. J. Immunol.
24
:
1061
3
Mor, F., I. R. Cohen.
1996
. IL-2 rescues antigen-specific T cells from radiation or dexamethasone-induced apoptosis.
J. Immunol.
156
:
515
4
Mueller, D. L., S. Seiffert, W. Fang, T. W. Behrens.
1996
. Differential regulation of bcl-2 and bcl-x by CD3, CD28, and the IL-2 receptor in cloned CD4+ Th cells.
J. Immunol.
156
:
1764
5
Sprent, J..
1997
. Immunological memory.
Curr. Opin. Immunol.
9
:
371
6
Greaves, D. R., F. D. Wilson, G. Lang, D. Kioussis.
1989
. Human CD2 3′-flanking sequences confer high-level, T cell-specific, position-independent gene expression in transgenic mice.
Cell
56
:
979
7
Love, P. E., E. W. Shores, E. J. Lee, A. Grinberg, T. I. Munitz, H. Westphal, A. Singer.
1994
. Differential effects of ζ and η transgenes on early α/β T cell development.
J. Exp. Med.
179
:
1485
8
Evans, G. A., D. H. Margulies, B. Shykind, J. G. Seidman, K. Ozato.
1982
. Exon shuffling: mapping polymorphic determinants on hybrid mouse transplantation antigens.
Nature
300
:
755
9
Hansal, S. A., D. Morris, J. M. G. Sechler, P. Love, A. S. Rosenberg.
1998
. Cutting edge: Induction of antigen-specific hyporesponsiveness by transplantation of hemopoietic cells containing an MHC class I transgene regulated by a lymphocyte-specific promoter.
J. Immunol.
61
:
1063
10
Osawa, M., K. Hanada, H. Hamada, H. Nakauchi.
1996
. Long term lymphohematopoietic reconstitution by a single CD34 low/negative hematopoietic stem cell.
Science
273
:
242
11
Rosenberg, A. S..
1994
. Skin allograft rejection. J. Coligan, and A. M. Kruisbeek, and D. H. Margulies, and E. M. Shevach, and W. Strober, eds.
Current Protocols in Immunology
4.4.
John Wiley & Sons, New York.
12
Lafferty, K. J., S. J. Prowse, C. J. Simeonovic, H. S. Warren.
1983
. Immunobiology of tissue transplantation: a return to the passenger leukocyte concept.
Annu. Rev. Immunol.
1
:
143
13
Banchereau, J., R. M. Steinman.
1998
. Dendritic cells and the control of immunity.
Nature
392
:
245
14
Matzinger, P..
1994
. Tolerance, danger, and the extended family.
Annu. Rev. Immunol.
12
:
991
15
Eynon, E. E., D. C. Parker.
1993
. Parameters of tolerance induction by antigen targeted to B lymphocytes.
J. Immunol.
151
:
2958
16
Nakamura, H., R. E. Gress.
1990
. Graft rejection by cytolytic T cells: specificity of the effector mechanism in the rejection of allogeneic marrow.
Transplantation
49
:
453
17
Boussiotis, V. A., B. J. Lee, G. J. Freeman, J. G. Gribben, L. M. Nadler.
1997
. Induction of T cell clonal anergy results in resistance whereas CD28-mediated costimulation primes for susceptibility to fas- and bax-mediated programmed cell death.
J. Immunol.
159
:
3156
18
Boise, L. H., P. J. Noel, C. B. Thompson.
1995
. CD28 and apoptosis.
Curr. Opin. Immunol.
7
:
620
19
Luqman, M., K. Bottomly.
1992
. Activation requirements for CD4+ T cells differing in CD45R expression.
J. Immunol.
149
:
2300
20
Ronchese, F., B. Hausmann.
1993
. B lymphocytes in vivo fail to prime naive T cells but can stimulate antigen experienced T lymphocytes.
J. Exp. Med.
177
:
679
21
Croft, M., L. M. Bradley, S. L. Swain.
1994
. Naive versus memory CD4 T cell response to antigen.
J. Immunol.
152
:
2675
22
Croft, M., S. B. Joseph, K. T. Miner.
1997
. Partial activation of naive CD4 T cells and tolerance induction in response to peptide presented by resting B cells.
J. Immunol.
159
:
3257
23
Lenschow, D. J., T. L. Walunas, J. A. Bluestone.
1996
. CD28/B7 system of T cell costimulation.
Annu. Rev. Immunol.
14
:
233
24
Davis, L. S., P. E. Lipsky.
1993
. Tolerance induction of human CD4+ T cells: markedly enhanced sensitivity of memory versus naive T cells to peripheral anergy.
Cell. Immunol.
146
:
351
-361.
25
Chai, J.-G., R. I. Lechler.
1997
. Immobilized anti-CD3 mAb induces anergy in murine naive and memory CD4+ T cells in vitro.
Int. Immunol.
9
:
935
26
Cook, J. R., E. M. Wormstall, T. Hornell, J. Russell, J. M. Conolly, T. H. Hansen.
1997
. Quantification of the cell surface level of Ld resulting in positive versus negative selection of the 2C transgenic T cell receptor in vivo.
Immunity
7
:
233
27
Morris, P. J..
1980
. Suppression of rejection of organ allografts by alloantibody.
Immunol. Rev.
49
:
93
28
Batchelor, J. R...
1981
. Immune mechanisms responsible for prolonged kidney allograft survival in immunologic enhancement.
Transplant. Proc.
13
:
562
29
Stuart, F. P., T. J. McKearn, A. Weiss, F. W. Fitch.
1980
. Suppression of rat renal allograft rejection by antigen and antibody.
Immunol. Rev.
49
:
127
30
Hutchinson, I. V..
1980
. Antigen-reactive cell opsonization (ARCO) and its role in antibody-mediated immune suppression.
Immunol. Rev.
49
:
167
31
Winn, H. J..
1986
. Antibody-mediated rejection. G. M. Williams, and J. F. Burdick, and K. Solez, eds.
Kidney Transplantation
17
Marcel Dekker, New York.
32
Jeekel, J. J., I. F. C. McKenzie, H. J. Winn.
1972
. Immunological enhancement of skin grafts in the mouse.
J. Immunol.
108
:
1017
33
Schulz, R., A. L. Mellor.
1996
. Self major histocompatibility complex class I antigens expressed solely in lymphoid cells do not induce tolerance in the CD4+ T cell compartment.
J. Exp. Med.
184
:
1573
34
Miller, J. J., L. J. Cole.
1967
. The radiation resistance of long-lived lymphocytes and plasma cells in mouse and rat lymph nodes.
J. Immunol.
98
:
982
35
Dixon, F. J., D. W. Talmage, P. H. Maurer.
1952
. Radiosensitive and radioresistant phases in the antibody response.
J. Immunol.
68
:
693
36
Slifka, M. K., R. Antia, J. K. Whitmire, R. Ahmed.
1998
. Humoral immunity due to long-lived plasma cells.
Immunity
8
:
363
37
Latron, F., M. Jotterand-Bellomo, A. Maffei, L. Scarpellino, M. Bernard, J. L. Strominger, R. S. Accolla.
1988
. Active suppression of MHC class II gene expression during differentiation from B cells to plasma cells.
Proc. Natl. Acad. Sci. USA
85
:
2229