Antigen presentation by mouse CD4⁺ T cells involving acquired MHC class II:peptide complexes: another mechanism to limit clonal expansion?

Major histocompatibility complex (MHC) class II molecules are constitutively expressed on specialized antigen-presenting cells (APCs) such as dendritic cells (DCs) and macrophages. Antigen presentation by such cells to T cells results in T-cell proliferation and differentiation toward effector function. However, MHC class II expression can be induced on multiple cell types under the influence of cytokines such as gamma interferon.1-3 The consequences of antigen presentation by such nonspecialized APCs remain unclear. Most of the cell types on which MHC class II expression can be induced do not express the major costimulatory molecules CD80 and CD86. Using gamma interferon–treated primary cultures of human epithelial cells, we have previously reported the induction of antigen-specific unresponsiveness in human CD4⁺ T cells.2 In most species, T cells themselves express MHC class II molecules following activation.4-6 Antigen presentation by activated T cells to activated T cells has been reported by several groups to induce either T-cell anergy or apoptosis.7-12 Indeed, the first report of antigen-induced unresponsiveness resulted from the culture of a human T-cell clone with its cognate peptide in the absence of added APCs.7 Mechanistically, this is a conundrum in that activated T cells also express substantial levels of B7 family costimulatory molecules.13-15 However, the regulatory effects of T:T antigen presentation may have a biologic role in limiting clonal expansion at the later stages of an immune response. One species in which T cells do not synthesize MHC class II molecules is the mouse.16,17 It has been assumed therefore that CD4⁺ T:T presentation does not apply in this species. However, there have been several recent reports describing the acquisition of molecules by T cells from APCs. These have included the transfer of MHC class I molecules,18,19 B7 family molecules,20,21 and MHC class II molecules in a series of rat experiments.22-24 The earliest description of MHC class II molecule transfer came from electron microscopy studies of thymocytes in irradiation chimeras25 and studies with T-cell clones in vitro.26 

In this study we have examined the capacity of T-cell receptor–transgenic and T-hybridoma cells to acquire MHC II:peptide complexes from APCs and subsequently present them to other T cells. The functional consequences of these events suggest that T:T antigen presentation may serve to limit T-cell clonal expansion. The relevance of these findings to alloimmunity following tissue and stem cell transplantation is discussed.

T-cell receptor (TCR)–transgenic DO11.10 mice were bred in the Biological Services Unit of the Imperial College Faculty of Medicine. BALB/c and CBA mice were obtained from Harlan.

The 3A9 murine CD4⁺ T-cell hybridoma is specific for hen-egg white lysozyme (HEL) peptide_46-61restricted by H2-A^k. A DO11.10 cell line was generated from DO11.10 TCR-transgenic mice and is specific for ovalbumin (OVA) peptide_323-339 restricted by H2-A^d. DAP.3-H2-A^k transfectants have been generated in our laboratory. CTLL2 is a murine interleukin-2 (IL-2)–dependent CD8⁺ cell line for the detection of IL-2 production. T cells were maintained in RPMI 1640 supplemented with 10% fetal calf serum (FCS), 2 mM l-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 5 × 10⁻⁵ M 2-ME. DAP.3 transfectants were maintained in Dulbecco modified Eagle medium supplemented with 10% FCS, 2 mM l-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 5 × 10⁻⁵ M 2-ME with appropriate drug selection to maintain expression of the transfected genes.

Antibodies used for the T-cell analysis are listed: fluorescein isothiocyanate (FITC)–conjugated anti–H2-A^k(clone 14V.18; Serotec, Oxford, United Kingdom); cy-chrome–conjugated DO11.10 TCR clonotypic antibody (clone KJ1-26; Caltag Lab, Burlingame, CA); FITC-conjugated annexin V, FITC-conjugated CD44 (clone Ly-24), FITC-conjugated CD3 (clone 145-2C11), phycoerythrin (PE)-conjugated anti–H2-A^k (clone 10.2.16), PE-conjugated anti–H2-A^d (clone AMS-32.1), PE-conjugated anti–TCR Vβ 8.2 (clone MR5-2), PE-conjugated anti-CD4 (clone GK1.5), and Cy-conjugated anti-CD4 (clone GK1.5) were obtained from Pharmingen (Heidelberg, Germany). FITC-conjugated sheep anti–mouse IgG was from Sigma (Dorset, United Kingdom). Anti-HEL (46-61)-H2-A^k complex (clone C4H3, kindly provided by Caetano Reis e Souza, Immunobiology Laboratory, Cancer Research UK, London Research Institute, United Kingdom) was purified from hybridoma culture.27 All flow cytometric analyses were conducted on a Becton Dickinson FACScalibur running CellQuest software (Becton Dickinson, Heidelberg, Germany).

The protocol of Inaba et al28 was used to generate DCs from bone marrow. Briefly, bone marrow cells (5 × 10⁵/mL) were cultured with 10% FCS RPMI 1640 containing 5% (vol/vol) supernatant from a granulocyte-macrophage colony-stimulating factor (GM-CSF)–secreting transfected cell line. On day 3, nonadherent cells were removed. Fresh medium with GM-CSF was added to the adherent fraction for continuing culture. Cells were used from day 8 to day 10.

Lymph node and spleen cells were treated with a mixture of anti-CD8 and anti–H2-A^d supernatants (M5/114, YTS169, and YTS191) for 30 minutes. After antibody treatment, the cells were washed and incubated with goat anti-mouse IgG and goat anti-rat IgG dynal beads (Dynals, Oslo, Norway) for 30 minutes. MHC class II positive cells and CD8⁺ T cells bound to the dynal beads were removed with a magnet. Purified CD4⁺ T cells were recovered from the unbound cell suspension. To purify naive CD4⁺ T cells, anti-CD44 supernatant (Ly-24) was added to the antibody supernatant cocktail.

5 × 10⁶ DO11.10 CD4⁺ T cells were transferred into BALB/c mice. After 3 days, the mice were immunized either with complete Freund adjuvant (CFA; Sigma) or CFA plus 300 μg OVA peptide via footpad injection. The DO11.10 CD4⁺ T cells were recovered from draining lymph node 1 day after injection.

APCs (DAP.3-H2-A^k transfectants or BALB/c DCs) were incubated with 5 μM CFSE (5-carboxy-fluoroscein diacetate succinimidyl ester) in phosphate-buffered saline (PBS) for 15 minutes in 37°C and then washed with PBS 3 times. T-cell responders (3A9 cells or DO11.10) were cocultured with CFSE-labeled APCs either pulsed with or without peptide overnight. CD4⁺cells were isolated from APCs by cell sorting and fixed with 1% (wt/vol) paraformaldehyde or γ-irradiated before being used as APCs in T-cell hybridoma or proliferation assays.

1 × 10⁵ fresh responder 3A9 cells were cocultured with 2 × 10⁵ APC-preincubated 3A9 cells in 96 flat-bottom micrometer plates. After 24 hours 50 μL supernatant was harvested and IL-2 activity determined by addition to the IL-2–dependent CTLL-2 cell line. CTLL-2 proliferation was measured by pulsing [³H]thymidine after a further 24 hours. Cells were harvested 18 hours later, followed by liquid scintillation spectroscopy. As a positive control, 5 × 10⁴DAP.3-H2-A^k transfectants pulsed with HEL_46-61were used instead of fixed 3A9 APCs.

Responder CD4⁺ T cells (1 × 10⁴cells/well) were stimulated with irradiated (30 Gray) APCs (DC- or APC-preincubated T cells) in 96-well plates for 3 days. Proliferation was assessed by [³H]thymidine incorporation during the last 18 hours of 72-hour assays.

DO11.10 CD4⁺ T cells were cocultured with CFSE-labeled BALB/c DCs with or without OVA peptide at a ratio of 4:1 for 4 hours to allow acquisition of MHC II or MHC II:peptide complexes from DCs. Afterward, the T cells were purified from the DCs by cell sorting. The purified DO11.10 CD4⁺ T cells that had either acquired H2-A^d:OVA complexes or H2-A^d alone were cultured in fresh medium overnight to allow T:T presentation to occur. DO11.10 CD4⁺ T cells without coculture served as a control. After the second stage culture, the cells were washed 3 times with PBS and then rested in fresh medium containing 10 μg/mL of anti–H2-A^d (M5/114) for 5 days to prevent any further cognate T:T interaction. After the rest culture, the cells were tested for proliferative responses to antigen restimulation by splenic APCs.

To explore MHC class II molecule acquisition by mouse CD4⁺ T cells, TCR-transgenic DO11.10 T cells were cocultured with BALB/c dendritic cells with increasing concentrations of cognate peptide. After overnight culture the cells were harvested and stained with FITC-conjugated anti-CD4 and PE-conjugated anti–H2-A^d antibodies. The level of H2-A^dexpression on gated CD4⁺ T cells is shown in Figure1A. An antigen dose–dependent increase in MHC class II expression on the T cells was seen such that the majority of cells were MHC class II positive at the highest peptide concentration. These data indicate that the acquisition of MHC class II molecules was influenced by cognate recognition of the APCs. However, significant MHC class II expression also was seen on the T cells cocultured with DCs in the absence of added peptide. The kinetics of MHC class II acquisition is illustrated in Figure 1B and Table1. As can be seen, readily detectable expression of H2-A^d on T cells was detectable within 3 hours of coculture with the peptide-pulsed dendritic cells.

To further define the nature of the acquired MHC class II molecules, we made use of a monoclonal antibody specific for a defined MHC II:peptide complex. This antibody, C4H3, is specific for a complex comprising H2-A^k:HEL_46-61.The 3A9 murine CD4⁺ T-cell hybridoma is specific for the same class II peptide complex. DAP.3-H2-A^k transfectants were pulsed with HEL peptide leading to expression of high levels of the C4H3 epitope (Figure 1C). When the peptide-pulsed transfectants were incubated with 3A9 cells overnight, the majority of 3A9 cells stained positively for H2-A^k expression, and a significant proportion also displayed the C4H3 epitope (Figure 1C, bottom panels). In contrast, when the 3A9 cells were cultured with the transfectants in the absence of peptide, low-level H2-A^k expression was detected and no staining was observed with C4H3. It is interesting to note that the fraction of H2-A^k molecules occupied by the HEL peptide was significantly increased on the 3A9 T cells, compared with the transfectant APCs, as judged by the ratio of H2-A^kto C4H3 fluorescence (7.6 on the APCs versus 3.7 on the T cells). This suggests 1 of 2 possibilities, either the TCR contributes to the transfer by the physical capture of MHC II:peptide complexes or the transfer involves molecules that are enriched in the molecular synapse between T cells and APCs such as the MHC II:peptide complexes for which the T cell is specific.

If the acquired MHC II:peptide complexes were to be of any functional significance in vivo, they would need to be stably expressed. The rate of decay of serologically detectable MHC class II molecules was assessed by culturing the sorted CD4⁺ T cells after MHC class II acquisition for varying time periods. As shown in Table 2, the percentage of positive cells and the mean fluorescence intensity did decay over time. However, readily detectable MHC class II expression was observed as much as 18 hours after in vitro culture.

When staining freshly isolated ex vivo CD4⁺ T cells from DO11.10 mice, we had consistently noticed a low level of staining with anti–MHC class II antibody, which we interpreted as nonspecific background activity. However, having observed the above results, we tested the possibility that this was genuine MHC II expression on the T cells due to acquisition in vivo. In order to address this, the T cells were analyzed either immediately following purification or after 24 hours of culture in the absence of APCs. As shown in Figure2A, the percentage of positive cells and the mean fluorescence intensity both diminished after 24 hours of culture, suggesting that this was genuine MHC class II expression due to transfer of molecules in vivo.

To further characterize the freshly isolated ex vivo MHC II–positive T cells, whole CD4⁺ T cells were purified from BALB/c mice and compared with the naive CD4⁺ cells prepared by depleting of CD44^hi cells. The cells were then triple-stained with anti-CD4, anti–H2-A^d, and other markers. The H2-A^d+ CD4⁺ cells were found only in the whole CD4⁺ population and not in the naive population. They were also CD3⁺, confirming that they are CD4⁺ T cells. Moreover, they were confined to the activated/memory population as shown by the anti-CD44 staining (Figure2B).

To examine whether activation can amplify the acquisition event in vivo, DO11.10 CD4⁺ T cells were transferred into BALB/c mice. After 3 days, the mice were immunized with CFA plus OVA peptide. Mice immunized with CFA or without any immunization served as controls. For mice immunized with CFA plus peptide, DO11.10 CD4⁺ T cells showed higher expression of MHC class II as compared with the nonimmunized animal, or animals immunized with CFA alone (Figure 2C), indicating that transfer of MHC class II can occur in vivo and is augmented by T-cell activation.

The mechanisms whereby T cells acquire molecules from APCs are unclear. The levels of acquired complexes were markedly augmented following cognate recognition of MHC class II–presented antigen by APCs, implying a role for the TCR in the capture of MHC II:peptide complexes. To distinguish between a direct involvement of TCR in the capture of complexes and TCR-dependent T-cell activation, DO11.10 CD4⁺ T cells were cocultured with BALB/c DCs or CBA DCs in the presence or absence of OVA peptide or in the presence of anti-CD3 antibody. Low levels of MHC class II molecules were acquired by DO11.10 T cells irrespective of the types of DCs used (Figure3, top panels). This suggests that the cognate recognition independent mechanism of MHC class II capture is MHC unrestricted. However, addition of OVA peptide markedly increased acquisition of H2-A^d from BALB/c DCs, but made no difference to the levels of acquired H2-A^k, again suggesting the involvement of a cognate interaction. To determine whether the influence of cognate interaction reflected physical capture of complexes by the TCR or merely the need for T-cell activation, anti-CD3 antibody was added to cocultures of T cells with APCs in the absence of antigen, or even with APCs expressing irrelevant MHC class II molecules. In the presence of 1 μg/mL anti-CD3, DO11.10 T cells acquired comparable levels of MHC class II molecules from CBA DCs, as from BALB/c DCs in the presence of either peptide or anti-CD3. This makes it less likely that the TCR is physically involved in the capture of complexes. It is not clear whether the effect of T-cell activation is mediated by some alteration in the properties of the T-cell membrane that make it more acquisitive, or that T-cell activation merely leads to increased intimacy of T-cell:APC conjugate formation. The enrichment of cognate MHC II:peptide complexes on the T-cell surface following antigen-dependent activation, as seen with the 3A9 hybridoma, probably reflects the fact that transfer of molecules is concentrated at the interface between the T cell and the APC where cognate MHC II:peptide complexes are concentrated within the “molecular synapse” that is instrumental in T-cell activation.

To investigate whether the acquired MHC II:peptide complexes on T cells can be presented to fresh responder T cells with the same antigen specificity, T cells were isolated from the T-APC cocultures and used as APCs in T-cell hybridoma or T-cell proliferation assays. More than 99% purity of isolated T cells was obtained after depletion of the CFSE-labeled APCs by cell sorting as shown for DO11.10 T cells in Figure 4D.

3A9 cells that acquired MHC II:peptide complexes (3A9^IIp) were fixed and cocultured with fresh responder 3A9 cells. The ability of 3A9 cells to present the acquired MHC II:peptide complexes to another T cell with the same antigen specificity was determined by their ability to induce IL-2 production by the fresh responder cells using the CTLL-2 bioassay. Figure 4A shows the positive control results of culturing 3A9 cells with the DAP.3 H2-A^k transfectant as APC. This led to peptide-dependent IL-2 production. When 3A9^IIp were used, comparable levels of IL-2 secretion were induced, albeit at a higher APC/T-cell ratio (Figure 4B). This demonstrates that 3A9 cells cannot only acquire MHC II:peptide complexes from APC, but the acquired complexes can be presented to other T cells with the same antigen specificity.

Parallel observations were made with T-cell lines established from DO11.10 mice (specific for OVA_323-339:H2-A^d complexes) and peptide-pulsed BALB/c DCs. DO11.10 CD4⁺ T cells were isolated from the T-DC coculture using the method described in “Materials and methods” and used as APCs. The cells were γ-irradiated and used to stimulate resting DO11.10 T cells. DO11.10 T cells that acquired the MHC II:peptide (DO11.10^IIp) isolated from cocultures with peptide-pulsed BALB/c DCs stimulated proliferation of resting DO11.10 T cells (Figure 4C), and the presentation could be blocked by adding anti–H2-A^d antibody (M5/114) (data not shown). Given the potency of dendritic cells as APCs, it was important to exclude the possibility that contaminating DCs accounted for the observed proliferation. The level of DC contamination following T-cell purification is shown in panel D of Figure 4. A dose response was performed with DO11.10 T cells and DC numbers corresponding to 1% of the number of DO11.10^IIp cells used in the same experiment. The lower level of proliferation induced by this number of DCs indicates that contamination is highly unlikely to account for the response induced by DO11.10^IIp T cells.

Human and rat data have demonstrated that T:T antigen presentation, involving activated T cells, induces anergy. To determine whether the presentation of acquired MHC II:peptide complexes by activated mouse T cells to activated T cells would have a similar outcome, the DO11.10 T-cell line (composed of activated T cells) was cultured with peptide-pulsed BALB/c DCs to allow MHC II acquisition, and the DCs were then removed and the T cells allowed to engage in T:T interactions. The T cells were then rested, by the addition of a blocking anti–H2-A^d antibody for 5 days. Finally, the T cells were rechallenged with peptide-pulsed BALB/c DCs. Annexin V staining of the T cells immediately after the period of T:T interaction revealed an increase in the fraction of the T cells undergoing apoptosis (15%) compared with the T cells exposed to DC antigen presentation (2%) (Figure 5A). As shown in Figure 5B, the cells that had engaged in T:T interactions were hyporesponsive compared with those that had been cultured with DCs, T cells that had acquired only MHC class II, or medium. The reduced proliferation could not be due to TCR:CD3 down-regulation or premature T-cell stimulation, because after 5 days resting, the levels of CD3:TCR had fully returned to normal, and T cells that had been stimulated with DCs had comparable proliferation to T cells in medium alone. The reduced proliferation also was accompanied by reduced IL-2 production (Figure 5C) and hyperresponsiveness to exogenous IL-2 (Figure 5D), a cardinal feature of T-cell anergy.

Taken together, these data suggest that mouse CD4⁺ T cells acquire MHC II:peptide complexes from APCs and can present these to each other, and that such T:T antigen presentation can induce both apoptosis and hyporesponsiveness.

The first observation of T-cell anergy was made when a human T-cell clone was incubated with cognate peptide in the absence of APCs. The induction of anergy required the presentation of antigen by activated T cells to activated T cells. If activated human T cells were used as APCs for resting human T cells, anergy did not occur, and strong proliferation was observed (G. Lombardi and R.L., unpublished observations, 1994). Two considerations have cast doubt as to the likely significance of these findings. First, T cells are inefficient at antigen internalization and processing, and second, mouse T cells do not synthesize MHC class II molecules, thereby challenging the generality of this phenomenon. The observations made here address these 2 issues.

There have been numerous reports of the transfer of molecules from APCs to T cells in mouse and rat systems. The first such observations were made using immunoelectron microscopic analysis of mouse thymic sections, in which murine thymocytes were noted to express MHC class II molecules, presumably acquired from thymic epithelium.25More recently, the capture of MHC class I and class II molecules by mouse and rat T cells from APCs has been described.18,19,22-24 Additional reports of the transfer of costimulatory molecules from APCs to T cells have been made.20,21 Indeed, such acquired MHC:peptide complexes appeared to be functional, in that CD8⁺ T cells became sensitive to peptide-specific lysis by neighboring T cells after acquisition of MHC I:peptide complexes from APCs.18,19Also, naive T cells that had acquired CD80 from APCs were capable of inducing IL-2 production by responder T cells.21 

The results described extend these findings by showing that the presentation of acquired MHC II:peptide complexes has the same functional effects that have been observed using human T cells that express endogenously synthesized MHC class II molecules. In previous studies in a human system we have described heterogeneity among T-cell clones, such that T:T antigen presentation by some clones induces apoptosis, whereas for other clones the dominant effect is the induction of anergy. Using a T-cell line derived from DO11.10 TCR-transgenic mice, the presentation of acquired MHC class II:peptide complexes induced both apoptosis and anergy.

The mechanisms whereby T cells acquire molecules from APCs are unclear. The levels of acquired complexes were markedly augmented by T-cell activation. This was seen following cognate recognition of MHC class II–presented antigen by APCs, implying a role for the TCR in the capture of MHC II:peptide complexes. However, the need for cognate recognition could be circumvented by the addition of anti-CD3 antibody to cocultures of T cells with APCs in the absence of antigen, or even with APCs expressing irrelevant MHC class II molecules. This makes it less likely that the TCR is physically involved in the capture of complexes. It is not clear whether the effect of T-cell activation is mediated by some alteration in the properties of the T-cell membrane that make it more acquisitive, or that T-cell activation merely leads to increased intimacy of T-cell:APC conjugate formation. Although the levels of MHC class II on T cells were increased in the context of activation, significant MHC class II acquisition was observed in the absence of activation. This activation-independent acquisition of MHC molecules appeared to be influenced by B7 family molecules, in that the acquisition of H2-A^d by DO11.10 T cells from Chinese hamster ovary transfectants was observed only, in the absence of antigen, when the transfectants coexpressed CD86 (data not shown). This result is consistent with the findings described previously by Hwang et al20 and may reflect the capacity of B7:CD28 interactions to induce intracellular signaling events independently of TCR:CD3-transduced signals.29 

The in vivo significance of these observations remains a matter of speculation. We have argued previously that cognate interactions between T cells may provide an additional mechanism to limit T-cell clonal expansion during the later stages of an immune response in lymphoid tissue. As T cells clonally expand and move away from the DCs that initiated their activation, they may engage in T:T interactions of the kind described here, thereby limiting the final clone size. The involvement of acquired complexes from the APCs makes this a more credible mechanism of immunoregulation in that it obviates the need for T cells to internalize and process soluble antigen that is likely to be in very limited supply in the lymph node. The fact that acquired complexes were readily detectable 18 hours after removal of the “donor” APCs suggests that the acquired complexes may be sufficiently stable for T:T interactions to occur long after the T cell has moved away from the local APCs. The detection of low-level expression of MHC class II molecules on freshly isolated T cells ex vivo that decayed after 24 hours in culture and the amplification of these events in vivo following antigen-induced activation suggest that transfer of molecules from APCs to T cell may indeed occur in vivo. The effects of T:T antigen presentation of acquired MHC:peptide complexes may play an important role in diminishing alloimmunity following hematologic stem cell transplantation. Donor T cells specific for minor or major histocompatibility antigens are likely to acquire MHC:antigen complexes from DCs. Presentation of these complexes to recently activated antirecipient T cells may serve to diminish the antirecipient alloresponse.

An alternative interpretation of these findings in an MHC class I–restricted system has been offered recently in the context of investigating competition between T-cell populations with high and low T-cell receptor affinity in vivo.30 In this study it was argued that the greater ability of the higher affinity T cells to remove MHC:peptide complexes from APCs deprived the lower affinity T cells of adequate levels of cognate ligand to be activated. This was proposed as a mechanism of affinity maturation of T cells. Whichever interpretation is correct, the transfer of MHC:peptide complexes from APCs to T cells appears to provide a further mechanism of immunoregulation, the nature of which will require elucidation by carefully designed in vivo experiments.

We thank A. George and H. Stauss for critical reading of the manuscript.