Dendritic cells control T cell tonic signaling required for responsiveness to foreign antigen

Abstract

Dendritic cells (DCs) are key components of the adaptive immune system contributing to initiation and regulation of T cell responses. T cells continuously scan DCs in lymphoid organs for the presence of foreign antigen. However, little is known about the functional consequences of these frequent T cell–DC interactions without cognate antigen. Here we demonstrate that these contacts in the absence of foreign antigen serve an important function, namely, induction of a basal activation level in T cells required for responsiveness to subsequent encounters with foreign antigens. This basal activation is provided by self-recognition of MHC molecules on DCs. Following DC depletion in mice, T cells became impaired in TCR signaling and immune synapse formation, and consequently were hyporesponsive to antigen. This process was reversible, as T cells quickly recovered when the number of DCs returned to a normal level. The extent of T cell reactivity correlated with the degree of DC depletion in lymphoid organs, suggesting that a full DC compartment guarantees optimal T cell responsiveness. These findings indicate that DCs are specialized cells that not only present foreign antigen, but also promote a “tonic” state in T cells for antigen responsiveness.

T lymphocytes protect the body against invading microorganisms while avoiding autoaggression toward self tissues. The contribution of dendritic cells (DCs) to this process is essential (1, 2). DCs are rare hematopoietic cells, contributing only 1%–3% of cellularity in lymphoid organs (3). There are several subpopulations of DCs with distinct properties scattered throughout the body, but overall one of DCs’ main tasks is to capture antigen for transport into secondary lymphoid organs and presentation to CD4 and CD8 T lymphocytes (4, 5). Specific recognition of cognate antigen results in activation of T cells, proliferation, and acquisition of effector functions, provided that the DCs are activated, by, for example, pathogen-derived stimulatory compounds, such as TLR ligands (1, 6, 7). Alternatively, recognition may result in induction of tolerance in situations where antigens, such as self antigens derived from the body’s own tissues, are presented by steady-state DCs in the absence of an inflammatory signal (6). Tolerance can manifest itself in the deletion of antigen-specific T cells (8, 9) or generation of regulatory T cells that down-modulate immune responses (10, 11).

Elegant two-photon microscopy studies have revealed frequent and dynamic contacts between T cells and DCs in the T cell zone of lymph nodes (12–14). T cells move at high speed (11 μm/s) through the three-dimensional network formed by the DCs’ dendrites, scanning DCs for cognate antigen (15, 16). On recognition of a cognate antigen, the transient interactions formed with DCs are prolonged to more than 1 h, resulting in immune synapse formation, activation, and proliferation of T cells (17, 18). The importance of DC contact for T cell activation during presentation of foreign antigen is thus well established; however, gaps remain in our knowledge of the consequences of T–DC interactions in the absence of foreign antigen. It has been assumed that the scanning process remains inconsequential unless DCs present foreign antigen.

In the present study, we investigated whether the frequent contacts between DCs and T cells in the steady state in the absence of foreign antigen really have no consequences, or whether they indeed affect T cell function. Our results demonstrate a previously unrealized role for DCs in regulating immune reactivity. The intense scanning of DCs by T cells in the absence of foreign antigen results in the induction and maintenance of a basal activation level that enables T cells to rapidly respond to foreign antigen. This is achieved by continuous recognition of self-peptide/MHC complexes on DCs.

Results

Dendritic Cells Support the Antigen Sensitivity of T Cells.

To examine the role of DCs in maintaining the intrinsic T cell reactivity toward foreign antigen, we used CD11c.DOG mice that express the human diphtheria toxin (DT) receptor on CD11c⁺ cells and thus allow depletion of about 90%–97% of the so-called “conventional” DCs (cDCs) (CD11c^hi MHC-II⁺) after application of DT with no signs of toxicity (Fig. 1A) (19). We then investigated the sensitivity toward foreign antigen of T cells after 2 days of cDC depletion. For the quantification of monoclonal responses, CD4 T cells were purified from DC-depleted (CD11c.DOG X 2D2) F₁ mice, which harbor a transgenic TCR specific for I-A^b/MOG33-55, and challenged with MOG33-55 peptide presented by irradiated splenocytes. These cells displayed a severely impaired proliferative capacity compared with CD4 T cells from DC-containing (B6 × 2D2) F₁ mice treated equally with DT (Fig. 1B).

Fig. 1.

cDCs control the sensitivity of CD4 and CD8 T cells to foreign antigen. (A) Flow cytometry analysis of splenocytes from WT (B6) and CD11c.DOG mice 1 day after two daily DT injections. Numbers next to the gate indicate the percentage of cDCs. (B–E) Proliferation of splenic CD4 T cells (B and C) and CD8 T cells (D and E) isolated from mice with normal (B6) or reduced (CD11c.DOG and CD11c.DTR) numbers of cDCs. After 2 days of DT administration, purified splenic CD4 T cells were stimulated with APCs with titrated amounts of MOG35-55 peptide (B) or SEA (C). Purified CD8 T cells were stimulated with APCs with titrated amounts of SIINFEKL peptide (D) or SEA (E). Data in C–F are presented as mean ± SEM and are representative of five experiments.

To investigate whether CD4 T cells bearing TCRs of other specificities are also dependent on DCs for responsiveness to foreign antigen, we stimulated polyclonal CD4 T cells with irradiated splenocytes and staphylococcal enterotoxin A (SEA). Again, the proliferation of CD4 T cells isolated from DC-depleted CD11c.DOG mice was drastically reduced (Fig. 1C). Similar results were obtained when cDCs were depleted for 24 h instead of 48 h before assessing T cell proliferation (Fig. S1) or when using purified DCs instead of whole spleen as antigen-presenting cells (APCs) (Fig. S2). As for CD4 T cells, both the monoclonal CD8 T cell proliferative response to the SIINFEKL peptide and the polyclonal CD8 T cell response to SEA were strongly impaired after a 2-day depletion of DCs in vivo (Fig. 1 D and E). For the monoclonal CD8 T cell response, OT-I mice were crossed to CD11c.DTR mice (20) instead of CD11c.DOG mice, because the latter also express ovalbumin (OVA) under the CD11c promoter, resulting in OT-I T cell deletion in the thymus. These data suggest that the frequent interactions of naïve CD4 and CD8 T cells with DCs in lymphoid organs during steady state serve to keep the T cells in a basal state of activation required for antigen responsiveness.

We next investigated the number of cDCs required in vivo to maintain T cell sensitivity. For this purpose, we generated mixed bone marrow (BM) chimeras with 0%, 50%, 80%, and 100% of CD11c.DOG BM (Fig. S3). Depletion of 50% of the splenic cDCs resulted in a small but statistically significant decrease in the proliferative response of T cells, whereas 80% depletion of splenic cDCs led to a marked loss of T cell responsiveness (Fig. 2 A and B). These findings suggest that optimal maintenance of basal T cell reactivity depends on a largely complete cDC compartment.

Fig. 2.

Titration of T cell antigen sensitivity depending on the size of the cDC compartment. Mixed BM chimeras were generated as depicted in Fig. S3 to obtain ~100%, 50%, 20%, and 0% of cDCs in the spleen of mice treated with two daily DT injections before isolation of splenic T cells and stimulation with APCs/SEA. (A and B) Proliferation of CD4 (A) and CD8 (B) T cells isolated from the spleen of the indicated BM chimeras in response to APCs/SEA. Numbers on the right of each graph indicate the number of cDCs in the spleen after two daily DT injections. Data are presented as mean ± SEM (n = 3).

Hyporesponsive T Cells Regain Their Antigen Sensitivity Once DC Numbers Recover.

We next investigated whether hyporesponsive T cells in cDC-deficient mice can recover their normal level of responsiveness once the cDC compartment has returned to its normal size. Full recovery of cDC numbers was observed about 4 days after termination of DT administration (Fig. 3A). CD4 and CD8 T cells from mice with a recovered DC population were able to fully respond to antigen (Fig. 3B and Fig. S4), indicating that the hyporesponsive T cells regained their antigen sensitivity once they interacted again with cDCs in the absence of cognate antigen. To more precisely investigate the interaction time with cDCs required for the hyporesponsive T cells to recover their antigen sensitivity, we preincubated CD4 T cells from normal and cDC-depleted mice with WT cDCs before adding APCs with SEA. Preincubation of hyporesponsive T cells with cDCs for only 30 min resulted in recovery of their antigen sensitivity (Fig. 3C). These results show that hyporesponsive T cells are able to quickly regain their antigen sensitivity on reencounters with cDCs.

Fig. 3.

Rescue of T cell sensitivity after a short interaction with cDCs in the absence of foreign antigen. (A) Numbers of cDCs (CD11c^hi MHC-II⁺) in the spleen of CD11c.DOG mice treated with a single DT injection on day 0. Three mice were analyzed at each time point. (B) Proliferation of polyclonal CD4 T cells purified from mice with normal (B6; two daily DT injections), depleted (CD11c.DOG; two daily DT injections), and recovered (CD11c.DOG; 4 days after DT administration) cDC levels in response to APCs and 0.5 μg/mL of SEA. (C) Proliferation of splenic CD4 T cells isolated from cDC-depleted and normal mice and preincubated with cDCs at a ratio of 5:1 immediately before (ex vivo; Left) or 30 min before (Right) stimulation with APCs and SEA. Data are representative of two or three independent experiments (mean ± SEM).

The lymphokines IL-2 and IL-7 are known to promote T cell proliferation and/or survival (21, 22). We investigated whether IL-2 and IL-7 are able to restore the antigen-specific proliferation of T cells derived from cDC-depleted mice. Stimulation of T cells from cDC-depleted mice with antigen and APCs in the presence of IL-2 or IL-7 did not result in reversal of hyporesponsiveness (Fig. S5).

Impaired TCR Signaling in Hyporesponsive T Cells from cDC-Depleted Mice.

We next investigated whether the defect in T cell proliferation was specific to TCR-mediated stimulation or due to a more general defect in cell cycle entry. Toward this end, T cells isolated from normal or cDC-depleted mice were activated with phorbol myristate acetate and ionomycin, which target signaling events downstream the TCR. We found no effect of cDC depletion on the TCR-independent proliferative capacity of CD4 or CD8 T cells (Fig. S6 A and B), suggesting that the proliferative defect of T cells from cDC-depleted mice was specific to antigen recognition.

To determine whether TCR signaling was defective in T cells re-ndered hyporesponsive by the absence of cDCs, we first quantified the level of surface TCR expression by flow cytometry. We found no differences in CD4 and CD8 T cells derived from normal and cDC-depleted mice (Fig. S6C). Furthermore, we detected no bias in the frequency or expression intensity of different TCR variable regions, including V_β11, a main chain recognizing SEA (23) (Fig. S6D). Sim-ilarly, cell surface expression of CD3, TCR coreceptors (CD4 and CD8), and activatory/inhibitory molecules (CD28, ICOS, OX40, RANKL, CTLA-4, and PD-1) was unaltered on T cells from cDC-depleted mice (Fig. S6 E and F).

We next assessed for any biochemical changes in TCR signaling induced by cDC depletion. For this, we quantified the phosphorylation status of the TCR-associated ζ chain, which indicates early events in the TCR signaling cascade. Indeed, cDC depletion led to a ~50% decline in ZAP-70–associated basal TCRζ phosphorylation in T cells compared with T cells from cDC-sufficient mice (Fig. 4A). The observed phosphorylated TCRζ chain most likely corresponds to p21 rather than to p23, because the former is the dominant ζ chain phosphorylated in murine lymphocytes not activated with cognate antigen (24, 25). These data indicate that cDCs sustain a basal level of TCR signaling in the absence of foreign antigen, consistent with the decreased T cell responsiveness seen in the absence of cDCs.

Fig. 4.

Recognition of self-MHC class I and II on cDCs is required for maintenance of T cell sensitivity to foreign antigen. (A) Biochemical analysis of TCRζ chain phosphorylation in CD4 T cells from the spleen of mice with normal (DC⁺) or reduced (DC⁻) cDC levels. Anti–ZAP-70 immunoprecipitates were blotted for ZAP-70 and phosphotyrosine content (Left), and bands were semiquantified by densitometry (Right). The density of the phosphorylated TCRζ band in the DC⁺ sample was set at 100%. IP, immunoprecipitation. (B) Proliferation of CD4 (Left) and CD8 (Right) T cells purified from unmanipulated mice and preincubated at a ratio of 5:1 with cDCs from WT (B6 DC) or A_β^−/− β₂m^−/− (MHC KO DC) mice or without cDCs for 18 h before stimulation with APCs with titrated amounts of SEA. Data are representative of three independent experiments (mean ± SEM).

Self-MHC Recognition on cDCs Is Required to Maintain T Cell Antigen Sensitivity.

Because TCRs recognize MHC–peptide complexes, we asked whether TCR recognition of self-MHC molecules on cDCs is required for maintenance of a basal activation state in T cells. Toward this end, we used an in vitro system in which polyclonal T cells from normal mice were preincubated with those from WT mice, MHC-deficient mice, or mice without cDCs and then tested for their ability to respond to SEA presented by functional APCs. Both CD4 and CD8 T cells preincubated with WT cDCs responded to SEA, but T cells preincubated in the absence of cDCs did not (Fig. 4B), mimicking the results obtained from cDC depletion in vivo (Fig. 1 C and E). Importantly, preincubation with MHC-negative cDCs isolated from mice lacking both MHC-I and MHC-II molecules (β₂m^−/− A_β^−/− mice) impaired the subsequent proliferative response of both CD4 and CD8 T cells to SEA presented by APCs (Fig.4B). T cell viability was decreased in the absence of cDCs, but was comparable in cultures with WT cDCs and MHC-negative cDCs (Fig. S7A), indicating that the differences in proliferation were not due to differences in viability. In these experiments, SEA and APCs were added directly to the preincubation culture of T cells with MHC-positive or -negative cDCs. To exclude the possibility that the SEA was taken up and presented mainly by the MHC-positive cDCs present in the culture, T cells were repurified via magnetic-activated cell sorting (MACS) after preincubation with cDCs and then challenged with SEA plus APCs. This repurification process resulted in CD3⁺CD4⁺ and CD3⁺CD8⁺ T cell purities of >90%. T cells remained hyporesponsive after preincubation without cDCs or preincubation with MHC-deficient DCs (Fig. S7B).

These results indicate that the “tonic” TCR signals required for antigen responsiveness are mediated by continuous stimulation with self-peptide/MHC complexes on cDCs. Preincubation of T cells with MHC-deficient cDCs did not result in complete lack of proliferative capacity (Fig. 4B), suggesting a non-MHC contribution to promotion of T cell sensitivity. Alternatively, it is possible that the A_β^−/−β₂m^−/− cells used as MHC-deficient cDCs express A_αE_β hybrid molecules that can be recognized by T cells, as has been suggested previously (26), as well as low levels of β₂m-independent MHC class I molecules.

Impaired Immune Synapse Formation.

On TCR signaling in response to cognate peptide/MHC complexes, T cells accumulate surface receptors and scaffolding proteins in the contact zone between T cells and APCs, resulting in the formation of the immune synapse (IS). The IS is characterized by a central cluster of TCR/CD3 mol-ecules known as a central supramolecular activation cluster (cSMAC), which is surrounded by a second cluster containing lym-phocyte function–associated antigen 1 (LFA-1) known as a peripheral SMAC (pSMAC) (27). LFA-1 binds to ICAM-1 molecules on the APC, thereby mediating firm adhesion of the T cell to the APC. Formation of a mature IS takes about 30 min and is be-lieved to be crucial for complete T cell activation (28). Thus, we investigated whether T cells from DC-depleted mice were defective in IS formation. For this, we performed high-throughput multispectral imaging flow cytometry in conjugates consisting of 2D2 CD4 T cells and peptide-loaded B cells as APCs. Because TCR and LFA-1 are major components of the IS (27), we analyzed the enrichment of these two receptors in the intercellular contact zone. As expected, we found no enrichment of TCR and LFA-1 in the intercellular contact zone in the absence of MOG33-55 peptide (Fig. 5A and Fig. S8). In the presence of peptide, T cells from mice containing normal numbers of cDCs specifically rearranged TCR and LFA-1, resulting in IS generation. TCR and LFA-1 recruitment peaked at 20–30 min after initiation of T cell–APC contact (Fig. 5B); however, T cells from DC-depleted mice failed to reorganize TCR and LFA-1, resulting in lack of IS maturation even in the presence of cognate peptide (Fig. 5A and B). These findings indicate that T cells from DC-depleted mice are defective in IS maturation in response to peptide presented by APCs. This effect is directly correlated with these cells’ decreased antigen-specific proliferative capacity (Fig. 1B).

Fig. 5.

Impaired immune synapse formation. (A) CD4 T cells isolated from 2D2 mice with a normal (Left) or deficient (Right) cDC compartment were incubated with B cells with (Lower) or without (Upper) cognate MOG35-55 peptide. Fluorescence images of T cell–APC conjugates were acquired using the high-throughput ImageStream system and analyzed with IDEAS software. Shown is an exemplary conjugate for each sample 20 min after initiation of cellular contacts. BF, bright field. (B) Quantification of TCRβ and LFA-1 enrichment in the contact area between T cells and APCs induced by recognition of cognate peptide at the indicated times after initiation of cellular contacts. TCRβ and LFA-1 enrichment at each time point in the absence of peptide was set at a value of 1. Data are presented as mean ± SEM from two pooled experiments.

Dendritic cells Are Not Required for T Cell Viability.

MHC recognition is believed to be required for naive T cell survival in the long term (>3 weeks) (29–33). Whether a single specific cell type is required to provide self-MHC for this purpose is unknown, however. Thus, we investigated whether the decreased antigen sensitivity of T cells in mice depleted of cDCs for 2 days was the result of decreased T cell viability. For this, we extended the period of DC depletion to 10 consecutive days of DT administration, to amplify a potential effect of DC depletion of T cell survival, and then quantified the number of T cells and their sensitivity toward antigen. As expected, DC depletion of ~95% was seen after 10 days of DT treatment (Fig. 6A) (34). Both the frequency and the numbers of splenic CD4 and CD8 T cells were not altered (Fig. 6 B and C), indicating no compromise of T cell viability in mice with significantly reduced DC levels. Consistent with the results obtained after 2 days of DT treatment, CD4 and CD8 T cells lost their antigen sensitivity after 10 days of DC depletion (Fig. 6D).

Fig. 6.

Normal numbers of T cells in DC-depleted mice. WT (B6) and CD11c.DOG mice were treated with DT for 10 consecutive days and analyzed the next day. (A) FACS dot plots of total spleen cells. Numbers next to the gate indicate the percentage of cDCs. (B and C) Frequency (B) and number (C) of splenic CD4 and CD8 T cells of the indicated mice. Data are representative of three independent experiments (mean ± SEM; n = 3). (D) Proliferative response of splenic CD4 T cells in response to SEA. Data are presented as mean ± SEM (n = 4).

Discussion

T cells continuously scan cDCs in lymphoid organs in the absence of infection (12–14). It is generally accepted that these dynamic interactions serve to increase the likelihood of rare cognate interactions between T cells and cDCs. Our results indicate that cDCs serve an important physiological function in the absence of antigen as well, promoting the responsiveness of T cells toward foreign antigen. CD4 and CD8 T cells isolated from cDC-depleted CD11c.DOG mice lost their capacity to proliferate in response to foreign antigen presented by functional APCs (Fig. 1). Possibly because of the low cDC frequency (~3% in spleen), the large majority of the cDC compartment is required to maintain the normal level of T cell sensitivity (Fig. 2). Furthermore, our experiments using MHC-deficient cDCs show that self-peptide/MHC recognition on cDCs is crucial to maintaining this tonic state of T cell responsiveness toward foreign antigen (Fig. 4B and Fig. S6B). This self recognition in the periphery is likely a consequence of positive selection in the thymus, where T lymphocytes engaging in low-affinity interactions with self-peptide/MHC survive and exit into the periphery. Once there, T cells not only maintain their ability to recognize self-peptide/MHC complexes, but actually require self-stimulation for full responsiveness to foreign antigen (35). Whether nonself allogeneic MHC recognition also can maintain the high an-tigen sensitivity of T cells remains to be resolved.

Previous investigations have shown that adhesion signals resulting from the interaction between T cells and DCs in the absence of cognate antigen lower the threshold for antigen-specific activation of T cells (36–38). The relationship between this adhesion-induced T cell priming and the DC/self-peptide–induced basal T cell signaling reported here is not yet clear, but both processes likely contribute to increased T cell response.

Because formation of a mature IS is crucial for optimal T cell ac-tivation and acquisition of effector functions (28), we investigated whether the decreased antigen-specific proliferative capacity of T cells from DC-depleted mice might be the result of impaired IS formation. Interestingly, T cells isolated from mice with a reduced cDC compartment were not able to form a mature IS (Fig. 5). This finding suggests that self-peptide/MHC recognition on cDCs maintains T cells in a tonic state that licenses them for IS generation and efficient response to foreign antigen.

Mature lymphocytes are constantly transiting from blood to peripheral lymphoid organs, where they have close encounters with cDCs, and back to the blood via the lymphatic system. Collectively, more T cells pass through the spleen than through all of the lymph nodes together (see ref. 39 for a review). The mean residence time of a lymphocyte in the blood of ~20–30 min suggests that most T cells are in contact with cDCs at any given time (39). Our results demonstrate that the loss of antigen sensitivity in T cells from cDC-depleted mice is not an end stage, but rather a reversible process. T cells quickly recover their antigen reactivity after the reinstatement of cDC–T cell interactions. Our experiments showed that a contact time of 30 min between hyporesponsive T cells and MHC⁺ cDCs is sufficient for full recovery (Fig. 3C).

Our finding that MHC recognition on cDCs induces basal TCR signals that increase T cell sensitivity to foreign antigen is in agreement the results of a previous study in which in vivo blockade of MHC class II molecules with antibody resulted in CD4 T cell hyporesponsiveness; that study did not investigate the cell type mediating this effect, however (35). Recently, mice with constitutive depletion of cDCs were generated through the expression of DTα in CD11c⁺ cells (40, 41). The authors of that study concluded that T cells from “cDC-less” mice were able to proliferate normally to foreign antigen, as demonstrated by adoptive transfers of OVA-specific OT-II CD4 T cells isolated from normal or cDC-less mice into unmanipulated mice, followed by OVA immunization (40, 41). These findings are in apparent contrast with our results showing that T cells require cDCs to maintain their antigenic sensitivity, as well as with those of Ingulli et al. (42) showing that OT-II CD4 T cells transferred into MHC-II–deficient hosts lose their proliferative capacity when challenged with OVA-expressing cDCs. One likely reason for this apparent discrepancy is that the cDC-less T cells were transferred into mice 24 h before OVA immunization, and thus frequent and intense T–DC interactions occurred before antigenic challenge (40, 41). As shown in the present study, the low antigen sensitivity of T cells resulting from lack of self-peptide/MHC recognition on cDCs is a transient state that can be quickly reverted within 20–30 min (Fig. 3C).

We believe that the enhanced antigen sensitivity of T cells after the interaction of self-MHC and cDCs is not due to improved T cell viability. The role of TCR signaling in promoting naïve T cell survival remains controversial (43–46). It has been suggested that MHC recognition is required for naive T cell survival in the long term (>3 weeks) (29–33). The survival of T cells was not compromised in our study, however; neither the number of splenic T cells (Fig. 6 B and C) nor these cells’ responsiveness to TCR-independent stimulus (Fig. S6 A and B) was altered after cDC depletion. Furthermore, adult mice that permanently lack cDCs have normal numbers of T cells in the periphery (40, 41). Thus, the decreased antigen sensitivity of T cells from mice lacking cDCs is due not to loss of cell viability, but rather to defective TCR signaling. This is further supported in the present study by the finding of a reduce basal level of TCRζ phosphorylation in naïve splenic CD4 T cells in mice lacking cDCs (Fig. 4A). The precise mechanisms through which the tonic TCR signals following MHC recognition on cDCs result in IS maturation and enhanced T cell responsiveness to foreign antigen remain to be investigated.

Our findings in the present study indicate a previously unrealized role for cDCs in the induction and maintenance of a tonic state of T cell antigen sensitivity by providing self-MHC ligands for T cell recognition in the absence of foreign antigen. A previous study found that doubling the number of cDCs resulted in T cell hyperactivation and autoimmunity (47). Here we show that a 50% reduction in the number of cDCs results in decreased T cell antigen sensitivity (Fig. 2). Thus, cDC homeostasis emerges as a key immunologic regulator by providing the proper level of cDCs that favors efficient T cell awareness of foreign antigen but avoids hyperactivation and autoimmunity.

Materials and Methods

Mice.

C57BL/6N mice (B6; CD45.2) and congenic B6.SJL-Ptprca Pep3b/BoyJ (CD45.1) mice were purchased from Charles River Laboratories. 2D2 mice were obtained from V.K. Kuchroo (Harvard Medical School, Boston) (48). Mice deficient in both I-A_β^b and β₂-microglobulin were obtained from T. Schüler (Charité University, Berlin). OT-I mice were obtained from A. Limmer (University of Bonn). BAC transgenic CD11c.DOG mice carry the human diphtheria toxin receptor (DTR) under the control of the CD11c promoter (19). CD11c.DTR mice (20) were obtained from S. Jung (Rehovot, Israel). All mice were bred and maintained at the German Cancer Research Center under specific pathogen-free conditions. All mouse experiments were conducted according to institutional guidelines and regulations of the German Cancer Research Center.

Dendritic Cell Depletion in Vivo.

CD11c.DOG mice and CD11c.DTR mice were injected intraperitoneally with 8 ng/g body weight (gbw) and 4 ng/gbw diphtheria toxin (DT; Sigma-Aldrich) in PBS, respectively (19, 20). cDC depletion in spleen was typically 90%–97%. cDC depletion by daily DT injection was carried out for 2 consecutive days, unless noted otherwise.

Cell Sorting.

Splenic CD4 and CD8 T cells were positively isolated using magnetic microbeads (Miltenyi Biotech) at 4 °C. Purity was consistently ≥90%. Purification of other cell populations was performed as detailed in SI Materials and Methods.

T Cell Proliferation Assays.

CD4 and CD8 T cells were magnetically purified as described above. T cells were immediately stimulated in 96-U microtiter plates with titrated doses of SEA (Sigma-Aldrich), MOG35-55 peptide, or SIINFEKL peptide in the presence of irradiated splenocytes or cDCs isolated from unmanipulated B6 mice as APCs at a ratio of 10:1. Cell proliferation was measured by [³H]-thymidine incorporation as described in SI Materials and Methods. To investigate the T cell–cDC interaction time required for recovery of sensitivity in hyporesponsive T cells, T cells from mice with a normal or reduced number of cDCs were preincubated with cDCs isolated from unmanipulated B6 at a ratio of 5:1 before the addition of APCs and SEA.

For experiments investigating the maintenance of T cell antigen sensitivity in vitro, splenic T cells from untreated B6 mice were labeled with 1 μM carboxyfluorescein succinimidyl ester (Molecular Probes) and incubated with various cDC populations, macrophages, or B cells at a ratio of 5:1 for 18 h. The T cells were then activated with irradiated splenocytes plus SEA for 18 h. After 72 h, T cell proliferation was quantified using flow cytometry.

Assessment of TCRζ Chain Phosphorylation.

In brief, Nonidet P-40 cell lysates of MACS-sorted CD4 T cells were subjected to coimmunoprecipitation using anti–ZAP-70 antibody (clone 1E7.2; Upstate Biologicals), after which the phosphorylation status was evaluated by immunoblotting with anti-phospho tyrosine antibody (clone 4G10; Upstate Biologicals). Further details are provided in SI Materials and Methods.

Generation of Mixed BM Chimeras.

Mixed BM chimera mice were generated as described previously (19) and detailed in SI Materials and Methods.

Flow Cytometric Analysis.

The analysis and antibodies used are described in detail in SI Materials and Methods. Propidium iodide was used as a viability dye.

Analysis of Immune Synapses by High-Throughput Imaging Flow Cytometry.

Recruitment of TCRβ and LFA-1 to the immune synapse between T cells and APCs was analyzed as reported previously (49). In brief, 2D2 CD4 T cells were mixed at a 1:1 ratio with B cells loaded or not loaded with MOG35-55 peptide. Then 2 × 10⁶ cells were centrifuged and incubated at 37 °C for 5 min to allow for T–B cell conjugate formation. Pellets were carefully resuspended in 50 μL RPMI supplemented with 10% FSC and further incubated at 37 °C for the specified times before being fixed in 2% paraformaldehyde and then stained with biotinylated anti-TCRβ antibody (H57-597) plus streptavidin–Alexa Fluor 488, phycoerythrin-labeled anti–LFA-1 antibody (2D7), and 7-AAD. A many as 10,000 images per sample were acquired with the ImageStream system (Amnis). Automated image analysis was performed with IDEAS software (Amnis) as described previously (49) and detailed in SI Materials and Methods. Values of 1 correspond to TCRβ and LFA-1 distribution in conjugates generated in the absence of peptide, whereas values >1 indicate peptide/MHC-induced accumulation in the area of contact with the APC.