Encounter with Antigen-Specific Primed CD4 T Cells Promotes MHC Class II Degradation in Dendritic Cells

Encounter with antigen-speciﬁc primed CD4 T cells promotes MHC class II degradation in dendritic cells

Kazuyuki Furutaa, Satoshi Ishidob, and Paul A. Rochea,1

aExperimental Immunology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892; and bLaboratory of Infectious Immunity, RIKEN Research Center for Allergy and Immunology, Yokohama, Kanagawa 230-0045, Japan

Edited by Peter Cresswell, Yale University School of Medicine, New Haven, CT, and approved October 12, 2012 (received for review August 9, 2012)

Major histocompatibility complex class II molecules (MHC-II) on murine APCs and acquired by mouse CD4 T cells (14); however, antigen presenting cells (APCs) engage the TCR on antigen-specific this acquisition is antigen independent. Huang et al. have shown CD4 T cells, thereby providing the specificity required for T cell that MHC-I can be transferred to antigen-specific CD8 T cells priming and the induction of an effective immune response. In this after conjugate formation (15); however, this process merely study, we have asked whether antigen-loaded dendritic cells (DCs) transfers intact MHC protein from the APC to the antigen- fi that have been in contact with antigen-specific CD4 T cells retain the speci c T-cell and promotes T-cell fratricide (15). Prolonged interactions of antigen-loaded DCs with antigen- ability to stimulate additional naïve T cells. We show that encounter fi with antigen-specific primed CD4 T cells induces the degradation of speci c T cells are required to initiate naïve CD4 T-cell activa- surface MHC-II in antigen-loaded DCs and inhibits the ability of tion and proliferation (16, 17). This process is tightly regulated to allow the effective initiation of an immune response. However, these DCs to stimulate additional naïve CD4 T cells. Cross-linking once this response is established, there could be mechanisms with MHC-II mAb as a surrogate for T-cell engagement also inhibits that limit overstimulation of additional antigen-specificTcells, APC function and induces MHC-II degradation by promoting the a process that could initiate a pathological inflammatory cascade. clustering of MHC-II present in lipid raft membrane microdomains, In this study, we have investigated the function of DCs after a process that leads to MHC-II endocytosis and degradation in lyso- engagement with antigen-specific CD4 T cells. We find that an fi somes. Encounter of DCs with antigen-speci c primed T cells or encounter of antigen-experienced T cells with antigen-bearing engagement of MHC-II with antibodies promotes the degradation APCs promotes the down-regulation of surface MHC-II ex- of both immunologically relevant and irrelevant MHC-II molecules. pression in DCs and renders these DCs incapable of stimulating IMMUNOLOGY These data demonstrate that engagement of MHC-II on DCs after additional naïve CD4 T cells. Essentially identical results are encounter with antigen-specific primed CD4 T cells promotes the obtained when MHC-II was cross-linked with MHC-II mAb, down-regulation of cell surface MHC-II in DCs, thereby attenuating a process that results in lipid raft-dependent MHC-II clustering, additional activation of naïve CD4 T cells by these APCs. endocytosis, and lysosomal degradation. Finally, we show that the down-regulation of MHC-II by T-cell engagement or MHC- T-cell activation | protein aggregation II cross-linking severely limits the ability of the DC to stimulate even unrelated naïve CD4 T cells. Our data thus demonstrate endritic cells (DCs) are professional antigen presenting cells that one consequence of CD4 T-cell recognition of antigen- (APCs) that function to prime naïve CD4 T cells through bearing APCs is down-regulation of surface pMHC-II by endo- D cytosis and lysosomal degradation, thereby limiting the ability of cell surface major histocompatibility complex class II (MHC-II) fi molecules. Immature DCs have high potency of antigen capture these APCs to interact with additional antigen-speci c T cells. and processing. After capturing antigen, DCs migrate to draining Results and Discussion lymph nodes, where they mature and express peptide-loaded MHC-II (pMHC-II) on their cell surface (1, 2). The pMHC-II on Encounter with Antigen-Specific CD4 T Cells Inhibits Subsequent MHC- the surface of APCs then “presents” the expressed antigenic II–Restricted APC Function of DCs. Although it is well known that peptide to naïve antigen-specific CD4 T cells, a process that DCs can efficiently prime naïve CD4 T cells, it remains to be results in the proliferation and activation of CD4 T cells (3). seen whether these DCs are capable of activating additional In APCs, cell surface MHC-II expression is tightly regulated naïve CD4 T cells of the same (or differing) specificity after they have completed their task of T-cell priming. To examine this and is important for the generation and propagation of an im- k mune response (4). Newly synthesized MHC-II forms a complex question, we have pretreated HEL-pulsed DCs with I-A - – with a chaperone protein termed the invariant chain (Ii) in the HEL46–61 restricted 3A9 CD4 T cells before adding additional fl endoplasmic reticulum. The MHC-II–Ii complex is transported carboxy uorescein succinimidyl ester (CFSE)-tagged naïve 3A9 to cell surface, internalized, and targeted to the antigen pre- T cells to the culture. Preincubation of the DCs with naïve senting compartment where Ii is degraded and MHC-II loads control or 3A9 T cells had no effect on the ability of the DCs to A with antigenic peptides (5). Peptide-loaded MHC-II is then simulate additional naïve 3A9 T cells (Fig. 1 ). Surprisingly, transported to the cell surface to present antigenic peptides to preincubation of the DCs with previously primed 3A9 CD4 CD4 T cells. Once at the cell surface, pMHC can internalize and T cells, but not control T cells, almost completely prevented the enter the endocytic pathway. Some fraction of internalized subsequent proliferation of CFSE-labeled naïve 3A9 T cells by A A pMHC-II is then recycled back to cell surface (6–8) while an- these DCs (Fig. 1 and Fig. S1 ). To rule out the possibility that other fraction is ubiquitinated by the E3 ubiquitin ligase March-I the proliferation of the T cells in the preculture affected naïve and targeted for degradation in lysosomes (9, 10). CD4 T-cell proliferation, the entire preculture was irradiated There are many reports demonstrating that upon conjugate formation between antigen-specific CD4 T cells with antigen- bearing APCs, the TCR internalizes and is degraded in the T Author contributions: K.F. and P.A.R. designed research; K.F. performed research; S.I. con- cell, a process that is thought to limit T-cell activation by APCs tributed new reagents/analytic tools; K.F. and P.A.R. analyzed data; and K.F. and P.A.R. (11, 12). This process can be mimicked by TCR cross-linking wrote the paper. with mAb (13), leading to the widely held belief that in vivo TCR The authors declare no conflict of interest. cross-linking by APC ligands promotes TCR endocytosis and This article is a PNAS Direct Submission. degradation. By contrast, the fate of the stimulatory pMHC-II 1To whom correspondence should be addressed. E-mail: [email protected]. complexes on APCs after an encounter with antigen-specific This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. CD4 T cells has not been followed. MHC-II can be shed from 1073/pnas.1213868109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1213868109 PNAS Early Edition | 1of6 Downloaded by guest on September 24, 2021 Fig. 1. Preculture with antigen-speciﬁc T cells inhibits subsequent naïve T-cell proliferation by DCs. (A) HEL-loaded DCs were pretreated alone (no T cells), with naïve or primed control CD4 T cells − k (3A9 tg ), or with naïve or primed I-A -HEL46–61– speciﬁc 3A9 CD4 T cells (3A9 tg+), for 4 h at 1:1 ratio. After various treatments, naïve CFSE-labeled 3A9 CD4 T cells were added to the culture at a 1:10 ratio (DC:naïve T-cell), and naïve 3A9 CD4 T-cell proliferation was measured 48 h later by FACS analysis. (B) HEL-loaded DCs were preincubated alone (no T cells) or with primed 3A9 tg− or primed 3A9 tg+ T cells for 4 h. The entire pretreated culture was irradiated with 3,000 rad, washed, and then added directly to CFSE-labeled 3A9 T cells. In each experiment the division index was calculated by using FlowJo software. The division index under each condition was expressed relative to that culture condition in which CD4 T cells were not added. The data shown are the mean ± SD from three independent experiments. *P < 0.05 (relative to no T cells control).

before the addition of CFSE-labeled naïve 3A9 T cells. Once Cross-Linking of MHC-II Promotes Endocytosis and Lysosomal again, preculture with antigen-specific T cells prevented sub- Degradation of Surface MHC-II in DCs. The APC-mediated down- sequent naïve CD4 T-cell proliferation, whereas preculture with regulation of the TCR upon T-cell:APC conjugate formation can nonspecific T cells did not (Fig. 1B). Lastly, we purified the DCs be mimicked by TCR cross-linking using anti-TCR mAb (13). To after the preculture period with nonspecific or antigen-specificT mimic the engagement of MHC-II with ligand(s) on T cells, we cells, and even these purified DCs were unable to stimulate naïve cross-linked MHC-II on the surface of DCs by using MHC-II 3A9 T-cell proliferation (Fig. S1B), demonstrating that the en- mAb. Like preincubation of DCs with antigen-specific primed T counter of DCs bearing specific MHC-II–peptide complexes with cells, cross-linking of I-A molecules on the surface of HEL- antigen-specific primed CD4 T cells inhibits the ability of the DCs pulsed DCs for only 4 h dramatically reduced the ability of the to serve as APCs for subsequent naïve CD4 T-cell activation. DCs to stimulate antigen-specific 3A9 T cells (Fig. 3A). By using antibody cross-linking as a surrogate for T-cell en- Antigen-Specific T Cells Stimulate the Loss of MHC-II from the DC gagement, we were able to study the mechanism of MHC-II Surface. In an attempt to identify the mechanism of APC in- down-regulation after MHC-II ligation. Approximately 40% of activation by antigen-specific T-cell interactions, we monitored cell surface I-A was spontaneously internalized after 30 min in expression of MHC-II–peptide complexes on these DCs. By using mock–cross-linked DCs, and cross-linking dramatically enhanced k B the I-A -HEL46–61 complex-specific mAb Aw3.18.14 (18), we the extent of MHC-II endocytosis in DCs (Fig. 3 ). By following found that pretreatment of HEL-pulsed DCs with primed, but not the fate of surface biotinylated MHC-II, we found that ∼50% of naïve, I-Ak-HEL–restricted T cells dramatically reduced expression cell surface MHC-II I-A was degraded in DC after 8 h of culture k C of I-A -HEL46–61 complexes on these DCs (Fig. 2A) and inhibited in medium alone (Fig. 3 ). When cell surface MHC-II was cross- CD69 up-regulation of naïve 3A9 T cells by these DCs (Fig. 2B). linked, however, the extent of MHC-II degradation was dra- The reduced APC function observed by T-cell pretreatment was matically enhanced. The effect of cross-linking on MHC-II ex- fi not irreversible, because addition of HEL46–61 peptide restores I- pression did not depend on the speci c mAb used to cross-link k A –HEL46–61 complex expression and APC function by these DCs. MHC-II, because identical results were obtained when MHC-II Like the loss of APC function, the loss of MHC-II from the surface I-A was cross-linked by using the I-Ak-specific mAbs 11-5.2 or of DCs was antigen-specific, because preincubation of DCs with 10-3.6 (Fig. S3A) or when MHC-II I-E was cross-linked by using k k fi B control CD4 T cells did not affect expression of I-A -HEL46–61 the I-E -speci c mAbs 14-4-4S or 17-3-3 (Fig. S3 ). Control complexes or CD69 up-regulation of naïve 3A9 T cells. studies using anti–MHC-I mAb revealed that cross-linking of To investigate the fate of MHC-II after interaction with T cells surface MHC-I had no effect on the turnover of MHC-II (Fig. by using a different TCR transgenic mouse system, we monitored S3A). The disappearance of MHC-II after cross-linking was due the expression of biotinylated cell surface MHC-II after in- to lysosomal proteolysis, because neutralizing lysosomal protei- + cubating OVA-pulsed DCs with either naïve or primed OVA- nases activity by treating cells with the vacuolar H ATPase in- specific OT-II T cells. Pretreatment with primed, but not naïve, hibitor Bafilomycin A1 or the weak-base chloroquine completely OT-II T cells resulted in the loss of MHC-II from the DC surface blocked the cross-linking–induced loss of MHC-II (Fig. 3D), (Fig. 2C). In these experiments, even the OT-II T cells present in demonstrating that cross-linking promotes lysosomal degradation the pretreatment culture were present during the cell lysis pro- of internalized MHC-II. Curiously, the cross-linking–induced loss cedure, ruling out the possibility that intact surface MHC-II was of MHC-II did not depend on MHC-II ubiquitination, because merely transferred from the DCs to T cells. The loss of MHC-II cross-linking stimulated endocytosis and degradation of MHC-II from the surface of DCs was antigen dependent, because OT-II similarly in wild-type DCs, March-I deficient DCs, and MHC-II I- b CD4 T cells did not affect expression of cell surface MHC-II when A K225R ubiquitination-deficient DCs (Fig. S4). the DCs were not loaded with OVA323–339 peptide. The ability of primed, but not naïve, CD4 T cells to reduce MHC-II expression Cholesterol-Dependent MHC-II Clustering Is Necessary for Cross- in DCs is likely due to enhanced conjugate formation between Linking–Induced MHC-II Endocytosis and Degradation. Cross-linking DCs and primed, but not naïve, T cells (Fig. S2). This result is of surface proteins with antibodies often leads to their clustering attributable to increased expression of adhesion molecules on on the cell surface (20, 21) and, for some proteins, cross-linking primed T cells that enhance even antigen-independent T-cell can stimulate endocytosis (21) or prevent membrane recycling adhesion (19). These results demonstrate that primed T cells (22). Quantitation of surface MHC-II clustering revealed that induce the degradation of MHC-II from the DC surface in an MHC-II was distributed relatively uniformly over the surface of antigen-dependent manner. DCs and very few “clusters” could be observed; however, MHC-

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1213868109 Furuta et al. Downloaded by guest on September 24, 2021 Fig. 2. Preculture with antigen-specific T cells reduces specific MHC-II–peptide expression on the surface of DCs. HEL-loaded DCs were pretreated alone for 16 h (no T cells) or at a 1:1 ratio with − control CD4 T cells for 16 h (3A9 tg ), with I-Ak- + HEL46–61–specific 3A9 CD4 T cells for 16 h (3A9 tg ), k or for 4 h with I-A -HEL46–61–specific 3A9 CD4 T cells before the addition of 50 μM HEL46–61 peptide for an + additional 12 h (3A9 tg + HEL46–61). (A)Theex- k pression of surface I-A –HEL46–61 complexes in + CD11c population was measured by FACS analysis k using the I-A –HEL46–61-specific mAb Aw3.18.14 by FACS analysis. The mean fluorescent intensity of each sample was expressed relative to that of the no T-cell pretreatment DCs. The data shown are the mean ± SD from three independent experiments. *P < 0.05 (relative to control). (B) The ability of each DC population to stimulate CD69 expression on naïve 3A9 T cells was measured by FACS analysis. The APC capacity of each DC sample was expressed relative to that of the no T-cell pretreatment DCs. The data shown are the mean ± SD from three independent experiments. *P < 0.05 (relative to control). (C) Sur-

face proteins of OVA323–339 peptide-pulsed DCs (Left) or control (unpulsed) DCs (Right) were biotinylated on ice with sulfo-NHS-biotin. The cells were then cultured with naïve or previously primed OT-II CD4 T cells for 4 h. Biotinylated proteins in the entire culture were isolated by using avidin-agarose beads and analyzed by immunoblotting with the indicated IMMUNOLOGY antibodies. The amount of biotinylated I-A present in each sample was normalized to the amount of CD9 present in the same sample. The data shown are the mean ± SD of three independent experiments. *P < 0.05 (relative to no T cells control).

k II cross-linking resulted in profound MHC-II clustering in nearly inhibited their ability to stimulate I-A -HEL46–61–specific3A9 all cells examined (Fig. 4A). Because MHC-II has been found to T cells. This result is in agreement with our observation that either be associated with cholesterol-dependent “lipid raft” membrane I-A or I-E cross-linking promotes the down-regulation of I-A and microdomains in both human and mouse APCs (23, 24), we demonstrates that cross-linking–mediated down-regulation of explored the possibility that the integrity of these domains is MHC-II results in decreased antigen presentation by DCs. important for cross-linking–induced MHC-II clustering and endocytosis. Cholesterol depletion (and lipid raft disruption) using Antigen-Specific T Cells Promote Loss of Relevant and Irrelevant MHC- the cyclic carbohydrate MβCD almost completely prevented the II from DCs. To determine whether exposure of an APC to an cross-linking–induced clustering of MHC-II on the surface of antigen-specific T-cell can inhibit the ability of the APC to DCs (Fig. 4A). Although cholesterol depletion had no effect on stimulate unrelated T cells, we simultaneously pulsed DCs with the spontaneous internalization of surface-tagged MHC-II, preprocessed HEL and PCC peptides. In agreement with our cross-linking–enhanced MHC-II endocytosis was nearly blocked previous results (Fig. 1), we found that preincubation with β B k by M CD treatment (Fig. 4 ). These results demonstrate that primed I-A -HEL46–61–specific 3A9 T cells almost completely MHC-II cross-linking leads to MHC-II clustering and that cross- prevented the DCs from stimulating additional naïve 3A9 T cells linking–induced MHC-II clustering and endocytosis are choles- (Fig. 6A). We also found that pretreatment with primed 3A9 T k terol dependent. Unfortunately, disruption of lipid rafts on living cells prevented the DCs from stimulating naïve I-E -PCC81–104– cells prevents APC:T-cell interactions (20), so extending this specific AND T cells (Fig. 6B). The inhibition of AND T-cell approach to investigate whether APC raft integrity is required activation by DCs was antigen-specific, because pretreatment of for T-cell–induced down-regulation of MHC-II is not possible. HEL/PCC-pulsed DCs with control T cells had no effect on AND T-cell proliferation. Taken together with the results of Cross-Linking of MHC-II Induces Internalization of both Relevant and antibody cross-linking experiments, these data demonstrate that Irrelevant MHC-II in DCs. If anti–MHC-II mAb were able to physi- the interaction of antigen-specific primed T cells with DCs leads cally cross-link surface MHC-II molecules present in discrete to the internalization and degradation of both immunologically membrane microdomains, we would predict that cross-linking I-A “relevant” and “irrelevant” MHC-II from the DC surface. would also promote the endocytosis and degradation of different In this study, we found that the encounter of antigen-loaded MHC-II molecules present in these same domains. We tested this DCs with previously primed antigen-specific CD4 T cells leads to prediction by cross-linking surface MHC-II on DCs using either I- MHC-II loss from the APC that limits its ability to activate either A–specificorI-E–specific mAb and F(ab’)2 antibodies. Cross- immunologically relevant or irrelevant naïve CD4 T cells. In our linking either surface I-A or surface I-E molecules each promoted attempt to identify a molecular mechanism for this phenomenon, I-A degradation (Fig. 5A). To examine the effect of MHC-II we examined the importance of MHC-II organization on the cross-linking on APC function, we pretreated HEL-pulsed DCs APC surface in this process. MHC-II molecules on the surface of with anti–I-A or anti–I-E mAb under cross-linking (F(ab’)2)and APCs are present in cholesterol-rich membrane microdomains noncross-linking conditions. Mock cross-linking did not affect the termed lipid rafts, and the integrity of APCs rafts is essential for proliferation of naïve 3A9 T cells by these DCs (Fig. 5B). By efficient CD4 T-cell activation (23). We found that cross-linking contrast, cross-linking of DC surface bound I-A or I-E mAb of MHC-II promotes the aggregation of surface MHC-II into

Furuta et al. PNAS Early Edition | 3of6 Downloaded by guest on September 24, 2021 MHC-II encounter with ligand(s) on antigen-specific primed T cells cross-links MHC-II molecules in distinct lipid rafts, thereby clustering these rafts and their constituent relevant and irrelevant MHC-II molecules. This hypothesis is supported by studies in B cells in which MHC-II cross-linking resulted in the visual aggregation of lipid rafts that colocalized with clustered MHC-II (20). The enhanced endocytosis observed in our current study likely represents an increased amount of MHC-II internalization after aggregation and not necessarily an increase in the kinetics of endocytosis of MHC-II molecules per se, as if cross-linking of raft-associated relevant MHC-II molecules “drags” raft-associated irrelevant MHC-II molecules into the endocytic pathway for degradation. Our finding that cross-linking of either I-A or I-E molecules promotes I-A degradation, together with our previous report that distinct forms of MHC-II (I- A and I-E) can coimmunoprecipitate under conditions in which lipid raft integrity is maintained (25) also fits with this model. This hypothesis also provides a molecular mechanism for studies of Kruisbeek et al. in which in vivo treatment with anti–I-A mAb suppressed not only I-A function but also I-E function in spleen APCs (26). Our results showing that an encounter of I-Ak–restricted T cells with antigen-pulsed DCs prevents subsequent activation of additional naïve I-Ak–restricted T cells as well as naïve I-Ek–restricted T cells is in excellent agreement with a model in which both immunologically relevant and irrelevant MHC-II is internalized and degraded after an encounter of DCs with primed antigen-specific T cells. It should be emphasized Fig. 3. MHC-II cross-linking induces internalization and lysosomal degra- that down-regulation of APC function is not observed when DCs dation of cell surface MHC-II in DCs. (A) HEL-loaded DCs were incubated with biotinylated anti-mouse IgG or anti-mouse MHC-II I-A mAb 11-5.2 on ice, washed, and incubated without (squares) or with (circles) anti-mouse IgG F

(ab’)2 at 37 °C for 4 h. The DCs were washed and CFSE-labeled 3A9 CD4 T cells were added to the culture at a 1:10 ratio (DC:naïve T-cell), and naïve 3A9 CD4 T-cell proliferation was measured 48 h later by FACS analysis. (B) DCs were incubated with biotinylated anti-mouse MHC-II I-A mAb 11-5.2 for 30 min on ice. The cells were washed and incubated without (squares) or with

(circles) anti-mouse IgG F(ab’)2 at 37 °C for the indicated times. The remaining cell-surface MHC-II antibodies were determined by incubating with ﬂuorescently labeled streptavidin on ice, and the cells were analyzed by FACS analysis. The mean ﬂuorescent intensity of each time point was expressed as a fraction of the percentage of amount of tagged MHC-II present on the cell surface at time 0. The data shown are the mean ± SD from three independent experiments. *P < 0.05 (relative to control). (C) Surface proteins of DCs were biotinylated on ice by using sulfo-NHS-biotin, and one aliquot of cells was immediately harvested (control, t = 0). The remaining cells were incubated with an isotype control mouse mAb (mock) or MHC-II I-A mouse mAb 11-5.2 for 30 min at 37 °C before the addition of

anti-mouse IgG F(ab’)2 at 37 °C for the indicated times. The cells were harvested, lysed, and biotinylated proteins were isolated by using streptavidin- Sepharose beads. The amount of biotinylated MHC-II present in the mock cross-linked samples (squares) or MHC-II cross-linked samples (circles) was analyzed by immunoblotting using the indicated antibodies and quantita- tive densitometry of the blots. The amount of surface MHC-II present at each Fig. 4. Cholesterol-dependent MHC-II clustering is necessary for cross-link- time point was expressed as a percentage of the total amount of surface ing–induced MHC-II endocytosis and degradation. (A) DCs were treated (or MHC-II present on biotinylated cells at control time 0. The data shown are not) with MβCD (10 mM) for 20 min at 37 °C. The cells were incubated with the mean ± SD of three independent experiments. *P < 0.05 (relative to biotinylated MHC-II I-A mouse mAb 11-5.2 for 30 min on ice, washed, and ’ μ mock). (D) Surface proteins of DCs were biotinylated on ice by using sulfo- incubated with or without goat anti-mouse IgG F(ab )2 (10 g/mL) for 5 min NHS-biotin. The cells were preincubated with medium alone, Baﬁlomycin A1 at 37 °C. The cells were ﬁxed, and immunolabeled cell surface MHC-II was (200 nM), or chloroquine (100 μM) for 30 min before incubation in the in- visualized by confocal microscopy using Alexa Fluor 546-conjugated strep- dicated medium containing mock Ab (mouse IgG) or mouse anti-MHC-II mAb tavidin. A representative image of I-A distribution on mock cross-linked or

11-5.2 for 30 min at 37 °C and cross-linking with anti-mouse IgG F(ab’)2 for 2 MHC-II cross-linked DCs is shown. At least 20 individual cells in each condition h at 37 °C. The cells were lysed, biotinylated proteins were isolated with were examined in each experiment. The percentage of the cells that have streptavidin-Sepharose beads, and the amount of biotinylated MHC-II pres- clustered cell surface MHC-II were quantified. The data shown are the mean ± ent in each sample as well as MHC-II survival was determined as described SD from three independent experiments. *P < 0.05 (relative to control). (B) above. The data shown are the mean ± SD of three independent experi- DCs were treated (or not) with MβCD (10 mM) for 20 min at 37 °C. The cells ments. *P < 0.05 (relative to control). were incubated with MHC-II I-A mouse mAb 11-5.2 for 30 min on ice, washed, and incubated with or without anti-mouse IgG F(ab’)2 for 30 min at 37 °C. The remaining cell-surface MHC-II antibodies were detected by incubating with fl discrete clusters and that cholesterol depletion (and lipid raft uorescently labeled streptavidin on ice, and the cells were analyzed by FACS analysis. The mean fluorescent intensity after 30 min of mock- or MHC-II disruption) prevented both mAb-induced MHC-II clustering and cross-linking was expressed as a percentage of the amount of tagged MHC-II cross-linking–enhanced MHC-II endocytosis. By analogy with present on the cell surface at time 0. The data shown are the mean ± SD from our results obtained using mAb cross-linking, we propose that three independent experiments. *P < 0.05 (relative to control).

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1213868109 Furuta et al. Downloaded by guest on September 24, 2021 MHC-II–peptide dependent, it does not unambiguously identify the T-cell ligand that engages MHC-II on the DC. Because the TCR on antigen-specific T cells is actually down-regulated after activation (27), it is possible that other MHC-II interacting molecules are involved in the MHC-II cross-linking. Surface expression of the MHC-II binding protein LAG-3 is dramatically enhanced during CD4 T-cell activation (28, 29); however, preliminary experiments revealed that LAG-3–deficient OT-II T cells also promoted MHC- II degradation in DCs. The TCR and CD4 are also likely candi- dates for MHC-II ligands; however, analysis of conjugates between MHC-II antigen-specific T cells lacking CD4 and antigen-loaded DCs is not technically possible. It is also possible that signaling in the DC, either by MHC-II itself or by an additional DC receptor, leads to T-cell–dependent, antigen-specificMHC-IIclusteringand endocytosis. Although determining the precise mechanism leading to pMHC-II loss after primed T-cell engagement will require additional study, our data showing that antigen-loaded DCs can be rendered nonfunctional by antigen-specific primed T cells provides a cellular mechanism to limit naïve T-cell activation by APCs during the course of an immune response. Fig. 5. Cross-linking of MHC-II induces degradation of both relevant and irrelevant MHC-II in DCs. (A) Surface proteins of DCs were biotinylated on ice by Materials and Methods using sulfo-NHS-biotin, and the cells were incubated with isotype control mouse Mice. B10.BR mice (H-2k), C57BL/6 mice (H-2b), 3A9 TCR transgenic mice, OT-II TCR mAb (mock), MHC-II I-E mouse mAbs 14–4-4S, or MHC-II I-A mouse mAbs 11-5.2 transgenic mice, and AND TCR transgenic mice were purchased from Jackson for 30 min at 37 °C and cross-linked by using anti-mouse IgG F(ab’)2 at 37 °C for 4 h. The cells were harvested, lysed, and biotinylated proteins were isolated by Laboratory. March-I knock-out mice and MHC-II K225R ubiquitination mutant using streptavidin-Sepharose beads. The amount of biotinylated MHC-II present knock-in mice on the C57BL/6 background have been described (9, 30). All mice in each sample was analyzed by immunoblotting using the indicated antibody. were cared for in accordance with National Institutes of Health guidelines with The amount of surface MHC-II remaining after cross-linking was expressed rel- the approval of the National Cancer Institute Animal Care and Use Committee. IMMUNOLOGY ative to that present in mock cross-linked cells. The data shown are the mean ± SD of three independent experiments. *P < 0.05 (relative to mock). (B)HEL- Antibodies and Reagents. Anti-mouse I-A mAb (clone 11-5.2) and anti-mouse I- loaded DCs were incubated with isotype control mouse mAb, MHC-II I-E mAb 14– E mAb (clone 14-4-4S) was from BioLegend. Anti-FcγRII/RIII (clone 2.4G2), 4-4S, or MHC-II I-A mAb 11-5.2 for 30 min at 37 °C. The cells were incubated alone anti-TCRβ (clone H57-597), anti-CD28 (clone 37.51), anti-CD9 (KMC8), and k (mock cross-link) or with anti-mouse IgG F(ab’)2 (F(ab’)2 cross-link) at 37 °C for 4 h. anti-CD11c-PE (HL3) antibody were obtained from BD Biosciences. The I-A -

The pretreated DCs were incubated with naïve CFSE-labeled 3A9 CD4 T cells at HEL46–61–specific mAb Aw3.18.14 (18) was a gift from Emil Unanue (Wash- 1:10 ratio. CFSE dilution was examined 48 h later by FACS analysis. The naïve CD4 ington University School of Medicine, St. Louis, MO). The anti–I-A β-chain T-cell division index was calculated by using FlowJo software. The division index rabbit serum has been described (31), and rabbit anti-mouse MHC-I serum under each condition was determined and was expressed relative to that of cells was a gift from Jon Yewdell (National Institute of Allergy and Infectious ± dividing under mock cross-linking conditions. The data shown are the mean SD Diseases, National Institutes of Health, Bethesda, MD). Sulfo-NHS-biotin was < from three independent experiments. *P 0.05 (relative to isotype control). from Thermo Scientific. Streptavidin-Sepharose beads, methyl β-cyclodextrin (MβCD), Bafilomycin A1, chloroquine, and HEL protein were purchased from fi Sigma-Aldrich. Alexa Fluor-conjugated antibodies, CFSE, CellTracker Green, are pretreated with naïve antigen-speci c T cells and that and CellTracker Red were obtained from Invitrogen. HRP-conjugated anti-

nothing in this model precludes additional DCs (that have not bodies were obtained from Southern Biotech. Goat anti-mouse IgG F(ab’)2 yet encountered antigen-speciﬁc T cells) from acquiring antigen antibody was purchased from Jackson ImmunoResearch. CD4 T-cell Isolation and stimulating additional relevant (or irrelevant) T cells. Kits, CD11c MicroBeads, CD4 MicroBeads, and CD90.2 MicroBeads were – Although our study clearly shows that the primed CD4 T-cell purchased from Miltenyi Biotech. Peptides corresponding to HEL(46-61),

dependent down-regulation of MHC-II on DCs is both TCR and OVA(323-339), and PCC(81-104) were synthesized at the Center for Biologics

Fig. 6. Preculture of HEL/PCC-loaded DCs with HEL- specific T cells inhibits activation of both HEL- and PCC-specific T cells. DCs loaded with both HEL and PCC were preincubated alone (no T cells) or with primed 3A9 tg− or primed 3A9 tg+ T cells for 4 h at 1:1 ratio. The pretreated culture was irradiated, k washed, and either CFSE-labeled naïve I-A -HEL46– k 61-specific 3A9 T cells (A) or CFSE-labeled naïve I-E - PCC81–104–specific AND T cells (B) were added to the culture at a 1:10 ratio (DC:naïve T-cell). CD4 T-cell proliferation was measured 48 h later by FACS analysis. The division index under each condition was expressed relative to that culture condition in which CD4 T cells were not added. The data shown are the mean ± SD from three independent experiments. *P < 0.05 (relative to no T cells control).

Furuta et al. PNAS Early Edition | 5of6 Downloaded by guest on September 24, 2021 Evaluation and Research, Food and Drug Administration, and purified by Purified naive CD4 T cells were labeled with CFSE (5 μM) in PBS for 8 min at reverse-phase HPLC. room temperature and washed with medium. DCs (4 × 104 cells) and CFSE- labeled naïve T cells (4 × 105 cells) were cocultured for 2 d, and CFSE dye dilution DC and T-Cell Isolation. Dendritic cells were prepared by differentiating was assayed by FACS. The division index was calculated by FlowJo software. mouse bone marrow cells in medium containing 20 ng/mL GM-CSF for 7 d by + + using standard protocols (32). The cells were routinely 90% CD11c , MHC-II , FACS-Based MHC-II Internalization Assay. DCs were incubated with bio- CD86low, CD40low after 7 d of culture. Naïve CD4 T cells were obtained from tinylated anti–MHC-II antibody (5 μg/mL) in the presence of the Fc receptor isolated lymph nodes by negative selection using a MACS CD4 T Cell Isolation blocking antibody 2.4G2 antibody (5 μg/mL) on ice 30 min and washed twice Kit according to the manufacturer’s specifications. To generate primed CD4 with FACS staining buffer [PBS containing 2% (vol/vol) FBS]. Cell surface μ T cells, naïve CD4 T cells were cultured in 24-well plates coated with 5 g/mL MHC-II antibodies were cross-linked with goat anti-mouse IgG F(ab’)2 (10 μg/ anti-TCRβ and 5 μg/mL anti-CD28 mAb for 3 d. The activation status of T cells mL) in complete media at 37 °C for various times. After two washes with was confirmed by FACS analysis of cell scatter plots, increased expression of FACS buffer, the amount of primary antibody remaining on the cell surface the activation marker CD69, and antibody-induced TCR down-regulation. was identified by staining using Alexa-conjugated streptavidin on ice. The cells were washed in ice-cold FACS staining buffer and were then fixed in Cell Surface Biotinylation. Plasma membrane proteins of DCs were bio- 1% PFA in PBS. Expression of each antibody was determined by flow cy- 6 tinylated by incubating ∼10 × 10 cells per mL with the membrane-imper- tometry using a FACSCalibur (Becton Dickinson). The mean fluorescence meable biotinylation reagent sulfo-NHS-biotin (1 mg/mL) in HBSS for 30 min intensity was determined for each FACS profile and expressed as a per- on ice according to the manufacturer’s protocol. After surface biotinylation, centage of the value present on cells kept on ice for the duration of the free biotin was quenched by washing the cells twice with ice-cold HBSS internalization assay. containing 50 mM glycine. The cells were extensively washed in ice cold HBSS and resuspended in complete medium before use. At the conclusion of each Cross-Linking and Cell Surface MHC-II Staining. DCs were incubated with the · indicated incubation, the cells were solubilized in lysis buffer (10 mM Tris HCl biotinylated anti–MHC-II mAb 11-5.2 (5 μg/mL) on ice for 30 min and washed at pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mg/mL BSA, 50 mM PMSF, 0.1 twice with FACS staining buffer. Cell surface MHC-II antibodies were cross- mM TLCK, and 5 mM iodacetamide) and biotin-labeled proteins were iso- linked with goat anti-mouse IgG F(ab′)2 (10 μg/mL) in complete media at lated by using streptavidin-Sepharose beads. Biotinylated proteins were 37 °C for 5 min. The cells were fixed in PBS containing 4% PFA and plated on detected by SDS/PAGE and immunoblotting using protein-specific antibodies poly(L-lysine)-coated coverslips. Surface MHC-II was visualized by confocal by using protocols described (33). microscopy using Alexa Fluor 546-conjugated streptavidin. All cells were imaged by using a Zeiss LSM 510 META confocal microscope using a 63× oil- DC Pretreatment and T-Cell Proliferation. Day 6 DCs were pulsed overnight immersion objective lens (N.A. 1.4), and the cells were scored as having μ with or without 5 M antigenic peptide and, in some experiments, cell clustered or “nonclustered” MHC-II by a blind observer. surface proteins were biotinylated as described above. Antigen-pulsed DCs were precultured at a 1:1 ratio with naïve or primed CD4 T cells for 4 h. For – ACKNOWLEDGMENTS. We thank our many colleagues for the gifts of anti- MHC-II cross-linking studies, DCs were incubated with anti MHC-II antibody bodies used in this study. This work was supported by the Ministry of Edu- μ (5 g/mL) in the presence of the Fc receptor blocking antibody 2.4G2 anti- cation, Culture, Sports, Science and Technology of Japan (S.I.), the Japan body (5 μg/mL) at 37 °C for 30 min and cross-linked with goat anti-mouse IgG Society for the Promotion of Science (S.I.), and the Intramural Research Pro- F(ab’)2 (10 μg/mL) at 37 °C for indicated times. gram of the National Institutes of Health (P.A.R.).

1. Inaba K, et al. (2000) The formation of immunogenic major histocompatibility com- 19. Dustin ML, Springer TA (1989) T-cell receptor cross-linking transiently stimulates ad- plex class II-peptide ligands in lysosomal compartments of dendritic cells is regulated hesiveness through LFA-1. Nature 341(6243):619–624. by inflammatory stimuli. J Exp Med 191(6):927–936. 20. Hiltbold EM, Poloso NJ, Roche PA (2003) MHC class II-peptide complexes and APC lipid 2. Turley SJ, et al. (2000) Transport of peptide-MHC class II complexes in developing rafts accumulate at the immunological synapse. J Immunol 170(3):1329–1338. dendritic cells. Science 288(5465):522–527. 21. Bléry M, Tze L, Miosge LA, Jun JE, Goodnow CC (2006) Essential role of membrane 3. Banchereau J, Steinman RM (1998) Dendritic cells and the control of immunity. Nature cholesterol in accelerated BCR internalization and uncoupling from NF-kappa B in B – 392(6673):245 252. cell clonal anergy. J Exp Med 203(7):1773–1783. 4. Nepom GT, Erlich H (1991) MHC class-II molecules and autoimmunity. Annu Rev Im- 22. Marsh EW, Leopold PL, Jones NL, Maxfield FR (1995) Oligomerized transferrin re- munol 9:493–525. ceptors are selectively retained by a lumenal sorting signal in a long-lived endocytic 5. Trombetta ES, Mellman I (2005) Cell biology of antigen processing in vitro and in vivo. recycling compartment. J Cell Biol 129(6):1509–1522. Annu Rev Immunol 23:975–1028. 23. Anderson HA, Hiltbold EM, Roche PA (2000) Concentration of MHC class II molecules 6. Reid PA, Watts C (1990) Cycling of cell-surface MHC glycoproteins through in lipid rafts facilitates antigen presentation. Nat Immunol 1(2):156–162. primaquine-sensitive intracellular compartments. Nature 346(6285):655– 24. Eren E, et al. (2006) Location of major histocompatibility complex class II molecules in 657. fi 7. Reid PA, Watts C (1992) Constitutive endocytosis and recycling of major histocom- rafts on dendritic cells enhances the ef ciency of T-cell activation and proliferation. – patibility complex class II glycoproteins in human B-lymphoblastoid cells. Immunology Scand J Immunol 63(1):7 16. 77(4):539–542. 25. Khandelwal S, Roche PA (2010) Distinct MHC class II molecules are associated on the 8. Pinet V, Vergelli M, Martin R, Bakke O, Long EO (1995) Antigen presentation medi- dendritic cell surface in cholesterol-dependent membrane microdomains. J Biol Chem ated by recycling of surface HLA-DR molecules. Nature 375(6532):603–606. 285(46):35303–35310. 9. Matsuki Y, et al. (2007) Novel regulation of MHC class II function in B cells. EMBO J 26 26. Kruisbeek AM, Titus JA, Stephany DA, Gause BL, Longo DL (1985) In vivo treatment (3):846–854. with monoclonal anti-I-A antibodies: Disappearance of splenic antigen-presenting cell 10. Walseng E, et al. (2010) Ubiquitination regulates MHC class II-peptide complex re- function concomitant with modulation of splenic cell surface I-A and I-E antigens. J tention and degradation in dendritic cells. Proc Natl Acad Sci USA 107(47):20465–20470. Immunol 134(6):3605–3614. 11. Zanders ED, Lamb JR, Feldmann M, Green N, Beverley PC (1983) Tolerance of T-cell 27. Liu H, Rhodes M, Wiest DL, Vignali DA (2000) On the dynamics of TCR:CD3 complex clones is associated with membrane antigen changes. Nature 303(5918):625–627. cell surface expression and downmodulation. Immunity 13(5):665–675. 12. Valitutti S, Müller S, Cella M, Padovan E, Lanzavecchia A (1995) Serial triggering of 28. Triebel F, et al. (1990) LAG-3, a novel lymphocyte activation gene closely related to – many T-cell receptors by a few peptide-MHC complexes. Nature 375(6527):148 151. CD4. J Exp Med 171(5):1393–1405. 13. San José E, Borroto A, Niedergang F, Alcover A, Alarcón B (2000) Triggering the TCR 29. Workman CJ, Rice DS, Dugger KJ, Kurschner C, Vignali DA (2002) Phenotypic analysis complex causes the downregulation of nonengaged receptors by a signal trans- of the murine CD4-related glycoprotein, CD223 (LAG-3). Eur J Immunol 32(8): duction-dependent mechanism. Immunity 12(2):161–170. 2255–2263. 14. Sharrow SO, Mathieson BJ, Singer A (1981) Cell surface appearance of unexpected 30. Ohmura-Hoshino M, et al. (2009) Cutting edge: Requirement of MARCH-I-mediated host MHC determinants on thymocytes from radiation bone marrow chimeras. JIm- MHC II ubiquitination for the maintenance of conventional dendritic cells. J Immunol munol 126(4):1327–1335. 183(11):6893–6897. 15. Huang JF, et al. (1999) TCR-mediated internalization of peptide-MHC complexes ac- 31. Muntasell A, Berger AC, Roche PA (2007) T cell-induced secretion of MHC class II- quired by T cells. Science 286(5441):952–954. – 16. Iezzi G, Karjalainen K, Lanzavecchia A (1998) The duration of antigenic stimulation peptide complexes on B cell exosomes. EMBO J 26(19):4263 4272. determines the fate of naive and effector T cells. Immunity 8(1):89–95. 32. Inaba K, et al. (1992) Generation of large numbers of dendritic cells from mouse bone 17. Stoll S, Delon J, Brotz TM, Germain RN (2002) Dynamic imaging of T cell-dendritic cell marrow cultures supplemented with granulocyte/macrophage colony-stimulating interactions in lymph nodes. Science 296(5574):1873–1876. factor. J Exp Med 176(6):1693–1702. 18. Dadaglio G, Nelson CA, Deck MB, Petzold SJ, Unanue ER (1997) Characterization and 33. Walseng E, Bakke O, Roche PA (2008) Major histocompatibility complex class II-pep- quantitation of peptide-MHC complexes produced from hen egg lysozyme using tide complexes internalize using a clathrin- and dynamin-independent endocytosis a monoclonal antibody. Immunity 6(6):727–738. pathway. J Biol Chem 283(21):14717–14727.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1213868109 Furuta et al. Downloaded by guest on September 24, 2021