Monovalent ligation of the B cell receptor induces receptor activation ...

3 downloads 0 Views 2MB Size Report
Feb 28, 2006 - We explored the role of antigen valency in B cell receptor (BCR) activation ... highly efficient mode of antigen presentation is thought to be a.
Monovalent ligation of the B cell receptor induces receptor activation but fails to promote antigen presentation You-Me Kim*, Jennifer Yi-Jiun Pan*, Gregory A. Korbel*, Victor Peperzak*, Marianne Boes†, and Hidde L. Ploegh*‡ *Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142; and †Department of Dermatology, Brigham and Women’s Hospital, Harvard Institute of Medicine, 77 Avenue Louis Pasteur, Boston, MA 02115 Communicated by Herman N. Eisen, Massachusetts Institute of Technology, Cambridge, MA, December 30, 2005 (received for review December 13, 2005)

antigen valency 兩 MHC class II

A

ctivation of B cell receptors by specific antigen unleashes a cascade of downstream events which includes alteration of transcriptional programs that lead to cell proliferation and differentiation (1–3). B cells further use their B cell receptors (BCRs) to facilitate the capture of antigen, and in doing so, they improve antigen presentation via class II MHC molecules (4). The latter event is essential for B cells to receive help from antigen-specific CD4⫹ T cells and to differentiate eventually into memory cells or plasma cells that secrete high-affinity antibody (5). Antigen-specific B cells present peptides derived from cognate antigens to CD4⫹ T cells at concentrations far below those required for other B cells that lack an antigen-specific receptor (6). This highly efficient mode of antigen presentation is thought to be a consequence of the dual role of the BCR in the delivery of specific antigen to MHC class II-containing peptide-loading compartments and in signaling that leads to enhanced generation of immunogenic peptides (7, 8). BCR-transduced signals promote efficient delivery of antigens through activation of Syk and B cell linker protein (BLNK) (9, 10). In addition, BCR-transduced signals induce physical and chemical remodeling of intracellular class II MHC-compartments (11). In the A20兾IIA1.6 mouse B lymphoma cell line, BCR activation by receptor cross-linking induced reorganization, fusion, and acidification of Lamp1-positive late endosomes where class II MHC molecules and invariant chain accumulate (12). In A20 B cells that express an anti-DNP IgM, BCR stimulation led to the transient intracellular accumulation of MHC molecules in newly formed www.pnas.org兾cgi兾doi兾10.1073兾pnas.0511315103

multivesicular bodies, into which the peptide exchange catalyst H2-M was recruited (13). Although these studies were among the first to shed light on a role for BCR signaling in MHC class II-mediated antigen presentation, virtually all of the studies that describe behavior of MHC class II molecules in relationship to the activation of B cells have been performed in transformed cell lines. Moreover, stimulation of the BCR is usually accomplished either by cross-linking receptors with anti-Ig antibodies or by transfecting established cell lines to express BCRs of known specificity to enable their ligation in an antigen-specific manner, which then still involves extensive crosslinking of BCRs. The outcome of B cell stimulation with a monovalent protein antigen, and its consequences for T cell activation have so far not been addressed. Here we have used hen egg lysozyme (HEL)-specific BCR transgenic mice (14) crossed to MHC class II-GFP knock-in mice (15) to investigate the role of antigen valency on BCR activation and rearrangement of intracellular MHC class II compartments in primary naı¨ve B cells. Through addition of monomeric or oligomeric versions of HEL, we have explored the contribution of antigen valency to successful BCR signaling and the ensuing rearrangement of MHC class II-containing peptide loading compartments. In addition, using live cell videomicroscopy, we have examined uptake and delivery of specific antigen into MHC class II-containing antigen-processing compartments and the concurrent changes in the intracellular MHC class II compartments in primary B cells. We find that monovalent HEL is presented far less efficiently than dimeric or higher-order complexes of HEL, despite its ability to activate the BCR. Results and Discussion Preparation of Oligovalent HEL. To produce versions of HEL capable

of multivalent ligation of BCRs on B cells from MD4 mice (14), which express the Ig heavy and light chain transgenes required for synthesis of anti-HEL IgM and IgD, we crosslinked HEL using a low concentration of glutaraldehyde. The resulting HEL conjugates were purified by gel filtration and their purity was assessed by SDS兾PAGE. We separated monomeric, dimeric and trimeric HEL fractions (Fig. 1a). A fraction containing mostly the tetrameric form and higher order complexes of HEL was designated as a tetrameric fraction. These HEL conjugates retain the ability to bind the HEL-specific BCR, because they efficiently competed for binding of monomeric HEL to MD4 B cells (Fig. 7, which is published as supporting information on the PNAS web site). Signal Transduction via the B Cell Receptor by Monomeric and Oligomeric HEL. Upon ligation by cognate antigen, cell surface B cell

receptors transduce signals by rapid induction of tyrosine phosConflict of interest statement: No conflicts declared. Freely available online through the PNAS open access option. Abbreviations: BCR, B cell receptor; HEL, hen egg lysozyme. ‡To

whom correspondence should be addressed. E-mail: [email protected].

© 2006 by The National Academy of Sciences of the USA

PNAS 兩 February 28, 2006 兩 vol. 103 兩 no. 9 兩 3327–3332

IMMUNOLOGY

We explored the role of antigen valency in B cell receptor (BCR) activation and rearrangement of intracellular MHC class II compartments as factors that contribute to the efficacy of antigen presentation. Using primary B cells that express a hen egg lysozyme (HEL)-specific BCR, we found that oligomeric HEL more efficiently promoted both BCR activation and internalization than did monovalent HEL, although monovalent HEL, unlike monovalent Fab fragments of anti-Ig, readily triggered the BCR. Nonetheless, oligovalent ligation positions the BCR in a membrane microdomain that is distinct from one engaged in the course of monovalent ligation, as judged by detergent extraction of the BCR. Furthermore, oligovalent HEL induced more pronounced rearrangement of MHC class II-containing antigen-processing compartments. Using videomicroscopy we observed in real time the rearrangement of MHC class II compartments as well as delivery of antigen in primary B cells. The observed increase in rearrangement of MHC class II-positive compartments and the disposition of antigen-bound BCRs therein correlates with improved presentation of a HEL-derived epitope. Although monomeric HEL efficiently engages the BCR, presentation of HEL-derived epitopes is impaired compared to oligovalent antigens. This trait may help explain the known ability of soluble, disaggregated antigen to induce a state of B cell tolerance.

Fig. 1. Preparation of HEL conjugates and HEL-induced tyrosine phosphorylation. (a) HEL conjugates crosslinked by glutaraldehyde were purified by gel filtration and resolved on a 12.5% SDS gel. Proteins were stained with Coomassie blue. (b) B cells purified from MD4 transgenic or wild type mice were incubated with 5 ␮g兾ml of monomeric, dimeric, trimeric HEL or 40 ␮g兾ml of F(ab⬘)2 fragment of goat anti-mouse IgM for 2 min at 37°C. Tyrosinephosphorylated polypeptides in cell lysates were detected with antiphosphotyrosine antibody 4G10. (c) Purified MD4 B cells were incubated with 5 ␮g兾ml of untreated monomeric HEL, glutaraldehyde-treated monomeric, dimeric, trimeric, tetrameric HEL or 20 ␮g兾ml of F(ab) or F(ab⬘)2 fragment of goat anti-IgM for 2min at 37°C. Tyrosine phosphorylation in B cell lysates was detected as described above. (d) MD4 B cells were incubated with indicated concentration of monomeric, dimeric, trimeric, or tetrameric forms of HEL for 2 min at 37°C. Tyrosine phosphorylated proteins were detected as described above.

phorylation of BCR-associated Ig␣兾␤ and intracellular signaling molecules such as lyn, syk, and BLNK (2, 3). We compared the ability of monomeric and oligomeric versions of HEL to activate the BCR by examining tyrosine phosphorylation. Although there was little tyrosine-phosphorylated protein in cell lysates prepared from unstimulated MD4 B cells, multiple tyrosine-phosphorylated polypeptides were detected in lysates from MD4 B cells stimulated with monomeric and oligomeric forms of HEL (Fig. 1b). Tyrosinephosphorylation induced by HEL conjugates was BCR-dependent, because no tyrosine phosphorylation was observed when B cells from wild-type littermate controls were incubated with the same HEL conjugates. In contrast and as expected, BCR ligation by polyclonal anti-mouse IgM antibodies stimulated tyrosinephosphorylation in B cells from both MD4 and control mice. Profiles of tyrosine-phosphorylated proteins induced by monomeric and oligomeric HEL in MD4 B cell lysates appeared to be similar. However, oligomeric versions of HEL were better at promoting tyrosine phosphorylation of cellular proteins than monomeric HEL (Fig. 1 b and c). We further tested HEL-dependent phosphorylation of Ig␣兾␤ after immunoprecipitating the Ig␣兾␤ heterodimers and found that monomeric HEL is capable of inducing their tyrosine phosphorylation (Fig. 7). Again, induction of Ig␣兾␤ phosphorylation was more efficient when cells were activated with dimeric and trimeric HEL. To exclude the possibility that tyrosine phosphorylation induced by monomeric HEL is caused by an altered physical characteristic of monomeric HEL incurred by glutaraldehyde treatment during preparation of HEL conjugates, we also tested a monomeric HEL preparation not exposed to glutaraldehyde, and found that it also promoted tyrosine phosphorylation in MD4 B cells (Fig. 1c, lane 2). Unlike monomeric HEL, monovalent Fab fragments of anti-IgM were unable to activate the BCR and did not induce tyrosine phosphorylation. This result implies that conformational changes in 3328 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0511315103

Fig. 2. Calcium mobilization induced by HEL conjugates in MD4 B cells. Changes in intracellular calcium levels were monitored in purified MD4 B cells loaded with fluorescent calcium indicators Fluo-3 and Fura Red by flow cytometry. After an initial 60 s of baseline recording, indicated concentrations of HEL conjugates or goat anti-IgM antibody fragments were added (indicated by dashed lines). (a) Changes in the mean fluorescent ratio were compared after addition of 1 ␮g兾ml of monomeric, dimeric, or trimeric HEL (Upper) or 10 ␮g兾ml of Fab or F(ab⬘)2 fragments of anti-IgM antibodies (Lower). (b) Induction of calcium flux by monomeric, dimeric, or trimeric HEL was compared at two different concentrations (1 ␮g兾ml and 10 ␮g兾ml).

the BCR mediated by binding of anti-IgM antibody to its constant regions are distinct from those that result from ligation of the BCR by its cognate antigen through the antigen combining site. In addition, the observed stimulation of MD4 B cells with monomeric HEL is not attributable to nonspecific adsorption of HEL onto neighboring cells or culture vessels (Fig. 7). Next, we examined tyrosine phosphorylation with increasing concentrations of HEL conjugates. Over the entire range of ⬇0.5–50 ␮g兾ml, oligomeric HEL conjugates consistently evoked higher levels of tyrosine phosphorylation than did monomeric HEL (Fig. 1d). In addition, no significant increase of tyrosine phosphorylation was observed by further increasing the concentration of monomeric HEL, whereas tyrosine phosphorylation by the other HEL conjugates was augmented with increasing doses of antigen, the maximum response being reached at 5 ␮g兾ml. These data argue against the possibility that the observed BCR activation by monomeric HEL is due to transient formation of a higher-order complex of monomeric HEL in solution or at the B cell surface. We also observed that it took consistently longer for monomeric HEL to Kim et al.

Fig. 4. Redistribution of MHC class II upon activation of B cell receptor. The relative increase in intracellular MHC class II-containing compartments in B cells was measured after stimulation with various forms of HEL. Integrated intensity of intracellular MHC class II-GFP signal above background was measured and the percent increase (mean ⫾ SEM) relative to the media control was plotted for each condition.

achieve maximum tyrosine phosphorylation (Fig. 7). Taken together, monomeric HEL is able to induce tyrosine phosphorylation, even though less well than oligomeric HEL. Monomeric HEL Evokes a Blunted Calcium Response. Recruitment of

phospholipase C␥ to BLNK and its activation upon BCR ligation leads to production of inositol-1,4,5-triphosphates and an increase in cytosolic calcium levels (16, 17). To discern possible qualitative and quantitative differences in calcium mobilization, we measured changes in intracellular calcium levels elicited by monomeric and oligomeric HEL. As was seen for the induction of tyrosine phosphorylation, monomeric HEL was less efficient than dimeric or trimeric HEL at evoking calcium mobilization. It took longer for monomeric HEL to induce a rise in intracellular calcium levels and the peak calcium level was lower than that obtained with oligomeric HEL (Fig. 2a). In addition, although dimeric and trimeric HEL induced both a rapid rise as well as a sustained elevation of calcium levels, monomeric HEL failed to do so. Instead, the intracellular calcium returned to basal levels after the initial spike. Fab fragments of anti-IgM antibody, as expected, were unable to induce calcium mobilization. A higher dose (10 ␮g兾ml) of dimeric HEL not only shortened the lag time for calcium induction but also increased both peak and sustained calcium levels (Fig. 2b). In contrast, increasing the concentration of monomeric HEL failed to evoke a calcium response comparable to that seen for HEL oligomers. Oligomeric versions of HEL are thus superior to monomeric HEL for activating the BCR and yield a response that is qualitatively and quantitatively distinct from that elicited by monomeric HEL.

3a). There was no significant change in digitonin solubility of the intracellular forms of Ig ␮ after incubation with any of HEL constructs. This result shows that reduced detergent solubility of terminally glycosylated Ig ␮ is a direct result of oligomeric HEL binding to surface IgM. Although less pronounced than for the surface ␮ chain, less Ig␣ is solublized in digitonin after activation with dimeric or trimeric HEL than with monomeric HEL. Unlike ␮ chains, for Ig␣ we cannot distinguish the intracellular fraction from the surface fraction by a simple difference in mobility by SDS兾PAGE. Digitonin interacts with cholesterol and precipitates proteins such as CD9 and CD81 that are targeted to lipid rafts (18). Upon crosslinking, the BCR rapidly translocates into lipid rafts (19, 20). Pretreatment of the B cells with methyl-␤-cyclodextrin, a cholesterol-extracting reagent, significantly reduced this oligomeric HELinduced translocation of the surface-disposed BCR, but had no effect on the digitonin solubility of the intracelluar Ig ␮ (Fig. 3b). Solubility of Ig ␮ chains and Ig␣ in 1% Triton X-100 was equivalent for all conditions tested. Internalization of the B Cell Receptor. Upon activation, the majority of BCRs internalize through a clathrin-mediated endocytic pathway (21, 22). To test whether various HEL conjugates promote BCR internalization, the steady-state level of cell surface BCR was measured by flow cytometry before and after incubation with HEL conjugates. All forms of HEL conjugates promoted rapid internalization of IgM (Fig. 8, which is published as supporting information on the PNAS web site). However, the maximum level of IgM internalization induced by monomeric HEL after 30-min incubation was only 44.7 ⫾ 5.1% compared to 64.7 ⫾ 5.8% for dimer, 66.5 ⫾ 6.6% for trimer, and 70.0 ⫾ 7.3% for tetramer. A similar result was obtained when internalization of IgD was analyzed after incubation with HEL conjugates (data not shown). Therefore, monomeric HEL is less efficient than oligomeric forms of HEL in promoting not only BCR-mediated signaling but also receptor internalization in MD4 B cells. Redistribution of MHC Class II Molecules. The internalized BCR

Translocation of BCR Complex to a Cholesterol-Rich Microdomain.

When immunoblotting for Ig ␮ chains, three distinct polypeptides are detected in MD4 cell lysates (Fig. 3). The lower two polypeptides are sensitive to Endoglycosidase H, therefore carry high mannose oligosaccharides and have not passed beyond the Golgi apparatus (data not shown). The upper polypeptide is Endoglycosidase H-resistant, corresponding to the terminally glycosylated form of ␮, and is responsible for HEL binding at the cell surface. Upon incubation with dimeric or trimeric HEL, we established that the surface Ig ␮ chains become increasingly resistant to extraction in 1% digitonin, whereas monomeric HEL had no such effect (Fig. Kim et al.

travels to multivesicular bodies or late endosomes where, after proteolysis of its bound antigen, it contributes antigenic peptide to MHC class II molecules. BCR activation may be responsible for inducing rearrangement of MHC class II-containing compartments to form transient antigen-processing compartments in B cell lines (12, 13). We tested whether the strength of signal transduction by the BCR correlates with the level of class II MHC redistribution. In unstimulated B cells, class II MHC was present mostly on the plasma membrane and only a small fraction was detected intracellularly in small vesicles near the plasma membrane (Fig. 9, which is published as supporting information on the PNAS web site). In B PNAS 兩 February 28, 2006 兩 vol. 103 兩 no. 9 兩 3329

IMMUNOLOGY

Fig. 3. Detergent extraction of BCR complex. (a) MD4 B cells activated with 5 ␮g兾ml of monomeric, dimeric, or trimeric HEL at 37°C were lysed in a lysis buffer containing 1% digitonin and the lysate was run on a 10% SDS-PAGE gel. Ig ␮ and Ig ␣ were then detected by immunoblotting. (b) MD4 B cells were pretreated with 1% DMSO or 10 mM methyl-␤-cyclodextrin (M␤CD) for 15 min and activated with 5 ␮g兾ml of HEL for 2 min at 37°C. The cells were lysed in a lysis buffer containing either 1% digitonin or 1% Triton X-100 and Ig ␮ and Ig ␣ were detected by immunoblotting. The open arrow denotes surface Ig ␮ and the two filled arrows indicate intracellular Ig ␮.

Fig. 5. Time-lapse imaging of HEL uptake and rearrangement of MHC Class II-containing compartments in live B cells. (a) Splenic B cells purified form MD4兾MHC class II-GFP were imaged at 37°C. The cells were initially imaged without HEL. After capturing a few frames of images, Alexa Fluor 568-conjugated dimeric HEL (final concentration of 5 ␮g兾ml) was added to cells (indicated by an arrow). Changes in distribution of MHC class II-GFP and Alexa Fluor 568-conjugated HEL were sequentially captured and shown is a gallery view of selected time-lapse images of a cell at one focal plane. Elapsed time is shown at the lower right corner of each frame. (b) Localization of MHC class II-GFP and Alexa Fluor 568-labeled dimeric HEL was examined at 5 s and 40 min after addition of 5 ␮g兾ml of Alexa Fluor 568-labeled dimeric HEL. (c) Localization of MHC class II-GFP and Alexa Fluor 568-labeled monomeric HEL was examined at 5 s, 40 min, and 120 min after addition of 5 ␮g兾ml Alexa Fluor 568-labeled monomeric HEL.

cells stimulated with HEL conjugates, more class II MHC molecules were visible in intracellular compartments compared to unstimulated B cells. Monomeric HEL was the least able to induce rearrangement of intracellular class II MHC-compartments, whereas with increasing valency, the oligomeric forms of HEL were more effective (Fig. 4). Cells did not measurably change their volume when activated for 30 min (Fig. 9). On the other hand, the intracellular area where the fluorescent signal of class II MHC-GFP was higher than background level was increased after stimulation with HEL conjugates (Fig. 9). BCR activation thus induces an increase in the total volume of intracellular class II MHCcontaining endosomes. Because the observations were apparent within 30 min, any such increase in intracellular class II MHC compartments must result largely from a rapid redistribution of class II MHC molecules, and not from de novo synthesis. Taken together, the extent of rearrangement of intracellular class II MHC compartments directly correlates with valency of HEL used for BCR activation and is superior for oligomeric versions of HEL. Imaging of Antigen Uptake and Rearrangement of MHC Class IIContaining Compartments in Live B Cells. To monitor in real-time the

uptake of HEL, BCR trafficking, and rearrangement of MHC class II compartments in B cells, we labeled HEL with the Alexa Fluor dyes. Binding of Alexa Fluor-labeled HEL to MD4 B cells is BCR-specific, mediated via the HEL moiety, and is not affected by 3330 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0511315103

the identity of the added fluorophore (Fig. 10, which is published as supporting information on the PNAS web site). B cells purified from MD4兾MHC class II-GFP mice were first imaged in the absence of HEL. After capturing a few frames of images, Alexa Fluor 568-labeled dimeric HEL was added to cells and images from three roughly equatorial sections of the cells (1 ␮m apart) were sequentially captured in the green and red channels. Time-lapse movies of the maximum z projection and images at one focal plane are shown in Movies 1–6, which are published as supporting information on the PNAS web site. A gallery view of a selected time series of a cell at a single focal plane is also shown in Fig. 5a. Before the addition of HEL, the MHC class II-GFP signal was found mostly on the plasma membrane. Upon addition, HEL initially bound to the surface of B cells in a uniform manner. At later time points, we observed increased MHC class II-GFP signal inside the cell and the Alexa Fluor 568-dimeric HEL signal started to colocalize with intracellular MHC class II, as evidence that HEL-bound BCRs were delivered to class II-containing antigen processing compartments (For detailed analysis, see Supporting Text, which is published as supporting information on the PNAS web site). Those endosomes that contain both BCR and class II further congregate and concentrate near the center of cell. At 40 min after addition of HEL, internalized dimeric HEL extensively colocalizes with MHC class II in every cell imaged (Fig. 5b). Binding of dimeric antigens to the B cell receptor thus promotes Kim et al.

rearrangement of intracellular MHC class II compartments, to which internalized antigens are delivered and processed for peptide presentation. We also examined uptake of Alexa Fluor 568-conjugated monomeric HEL. Like dimeric HEL, Alexa-monomeric HEL initially bound to the surface of B cells uniformly (Fig. 5c). However, after 40 min, only a small proportion of monomeric HEL was found inside the cells and much of it remained at the surface. Even after 120 min, low levels of monomeric HEL persisted at the surface. Internalized monomeric HEL colocalizes with MHC class II-GFP, but endosomes containing both the monomeric HEL and MHC class II are scattered, instead of being congregated in the center of the cells as seen with dimeric HEL. We speculate that these differences in intracellular arrangement of endosomes could reflect more mature antigen processing compartments and hence contribute to functional differences of antigen presentation discussed below. Valency of HEL Correlates with T Cell Stimulating Potential When Using B Cells as Antigen-Presenting Cells (APCs). We next performed an

antigen presentation assay by measuring production of IL-2 by the I-Ab restricted, HEL-specific T cell hybridoma, BO4H9 after culturing them with the B cells and HEL conjugates. When using the wild-type B cells as APCs, we observed no T cell stimulation, regardless of type and concentration of HEL used (Fig. 6). In contrast, significant IL-2 production resulted when MD4 B cells were exposed to HEL. Although oligomeric HEL evoked IL-2 release of up to 800 pg兾ml, IL-2 production by monomeric HEL was minimal. We estimate that, on a molar basis, monovalent HEL is 50to 100-fold less stimulatory in the antigen presentation assay than oligovalent HEL. The failure of B cells to present monomeric HEL is not due to inefficient up-regulation of costimulatory molecules, because monomeric, dimeric, and trimeric HEL all promoted up-regulation of CD86 equally well in MD4 cells (Fig. 11, which is published as supporting information on the PNAS web site). In addition, both monomeric and oligomeric HEL induced expression of CD69, a B cell activation marker (Fig. 11). Additional stimulation of MD4 B cells with LPS or CpG further activated B cells (CD69) and promoted up-regulation of costimulatory molecules such as CD86 and CD40. However, the presence of CpG did not restore presentation of monomeric HEL to T cells (Fig. 11). Four types of extracellular inputs shape the response of lymphocytes to antigen: (i) the concentration of antigen; (ii) the avidity with which antigen is bound; (iii) the timing and duration of antigen encounter; and (v) the presence of costimuli from innate immune system or other lymphocytes (23). Here, we examined the role of antigen valency in BCR signaling, internalization and antigen presentation, by using primary B cells from anti-HEL BCR transgenic mice and carefully prepared HEL monomer and oligomers. Kim et al.

Methods Preparation of Oligomeric Forms of HEL. Hen egg lysozyme (Sigma)

was dissolved in 50 mM Na2B4O7 (pH 8) and cross-linked with glutaraldehyde (Sigma) for 20 min at room temperature. The glutaraldehyde conjugation reaction was quenched by addition of 1 M glycine and insoluble materials were removed from the reaction mixture by centrifugation. Proteins in the supernatant were precipitated by addition of (NH4)2SO4 and the precipitate was dissolved in a buffer containing 50 mM Tris (pH 7.4) and 8 M urea. The HEL conjugates were then separated on Superdex 75 preparation column (Amersham Pharmacia) equilibrated with 50 mM Tris (pH 7.4), 5 M urea and 300 mM NaCl. Similar results were obtained when crosslinked HEL was cleared by centrifugation and subjected to gel filtration in the absence of urea. Fractions corresponding to monomeric, dimeric, trimeric, or high order complex PNAS 兩 February 28, 2006 兩 vol. 103 兩 no. 9 兩 3331

IMMUNOLOGY

Fig. 6. HEL-peptide presentation by MD4 B cells. B cells purified from wild-type or MD4 mice were incubated with T cell hybridoma BO4H9 and the indicated concentrations of HEL conjugates for 24 h at 37°C. Secreted IL-2 in the culture supernatants was measured by ELISA.

To our knowledge, a direct comparison of monovalent and oligovalent ligation of the BCR via a specific protein antigen has not been reported previously. We found that oligomeric HEL more efficiently promoted both BCR activation and internalization than did monovalent HEL. In addition, oligomeric HEL induced a more pronounced rearrangement of intracellular MHC class II compartments where antigen-bound BCR is targeted. Presentation of monomeric HEL to a HEL-specific T cell was poor compared to that of the HEL oligomers, despite the fact that monomeric HEL engages the BCR rather efficiently. Because the receptor-proximal events appear superficially similar for monovalent and oligovalent BCR ligands, the exact manner in which these signals are processed downstream to yield such different outcomes remains an open question. There are several possible explanations for the failure of MD4 B cells to efficiently present monomeric HEL to HEL-specific T cells, but both quantitative and qualitative differences of signals induced by monomeric versus oligomeric HEL likely account for inefficient presentation of monomeric HEL. We propose that the ability of BCR-bound HEL to be converted into a suitable MHC class II-peptide complex requires deposition of the BCR-HEL complex into a specific membrane microdomains operationally defined as resistant to direct extraction with digitonin. The differences seen for mono- and oligovalent HEL in this partitioning are striking and correlate very well with the ability of BCR-bound antigen to be presented to T cells. The ability of disaggregated, soluble antigen, administered intravenously, to induce B cell tolerance is a well-established immunological observation (24, 25). This phenomenon implies that B cells indeed perceive signals generated by monovalent ligands. The lesser ability of antigen-specific B cells to present monomeric antigen correlates well with reduced translocation of the BCR into a membrane microdoamin and rearrangement of class II MHC compartments caused by monomeric antigen. Our observations may thus have identified factors that contribute to the tolerogenic properties of monovalent antigen preparations. Equally noteworthy is the distinction between a monovalent anti-IgM Fab fragment and monovalent HEL. Whereas the former BCR ligand is unable to evoke tyrosine phosphorylation, calcium mobilization, and BCR internalization, monovalent HEL readily does so, at levels that cannot be explained by contamination with oligomeric HEL or by transient formation of an oligovalent ligand on a surface by adsorption. We suggest that the consequences of ligating a BCR via its antigen binding site are not necessarily equivalent to those produced by a monomeric Fab that binds to the BCR, at least in the MD4 anti-HEL model. The Fab fragments obtained from polyclonal anti-Ig reagents are likely to recognize conserved epitopes in the Ig constant regions, whereas all versions of HEL will engage the antigen binding site of surface Ig. In accordance with our finding, BCR ligation by antigens, but not by anti-Ig antibodies, leads to efficient phosphorylation of the inhibitory coreceptors CD22 and CD72 (26).

forms of HEL conjugates were pooled, dialyzed twice against PBS. The purity of each fraction was assessed by SDS兾PAGE followed by Coomassie blue staining of gels. Tyrosine Phosphorylation Assay. Naı¨ve B cells from mouse spleens

were purified by using magnetic anti-CD43 microbeads (Miltenyi Biotech). CD43-negative cells were resuspended in DMEM supplemented with 25 mM Hepes and incubated with various concentrations of HEL conjugates at 37°C. In some cases, B cells were stimulated with Fab or F(ab⬘)2 fragments of goat anti-IgM antibody (Jackson ImmunoResearch). The cells were then immediately lysed by addition of ice-cold lysis buffer (50 mM Tris, pH 8.0兾150 mM NaCl兾1 mM EDTA兾1% Triton X-100兾4 mM sodium pyrophosphate兾2 mM sodium vanadate兾10 mM sodium fluoride兾1 ␮g/ml leupeptin兾1 ␮g/ml aprotinin兾5 mM benzamidine). The lysate was analyzed on a SDS兾PAGE gel and transferred to a nitrocellulose membrane. Tyrosine-phosphorylated proteins were detected with anti-phosphotyrosine antibody 4G10 (Upstate Biotechnology). The membrane was reprobed with anti-actin antibody (Sigma) to verify loading with equivalent amounts of lysate. Intracellular Calcium Measurements. Changes in intracellular calcium levels were assessed in purified B cells by flow cytometry. B cells (5 ⫻ 106 per ml) in RPMI medium 1640 containing 5% FCS were loaded with 2 ng兾ml Fluo-3AM and 5 ng兾ml Fura Red AM (Molecular Probes) in the presence of 0.02% Pluronic F127 for 30 min at 30°C. The cells were then washed twice and resuspended in Krebs Ringer solution supplemented with 1% FCS and 2.5 mM probenecid at a cell density of 2 ⫻ 106 per ml. Subsequently, the cells were equilibrated at 37°C for 10 min before the calcium measurement. The changes in fluorescence intensity of Fluo-3 and Fura Red were monitored on a FACSCalibur cytometer (BD Biosciences). Acquisition was first conducted for 60 s without stimulus, followed by the addition of HEL conjugates or goat anti-IgM antibody fragments. Data were collected for 512 s, and fluorescence ratios of Fluo-3 and Fura Red were analyzed by using FLOWJO (TriStar).

␣ was detected with a polyclonal serum (prepared from a rabbit immunized with Ig ␣ cytosolic peptides). Live Cell Imaging. B cells purified from MD4兾MHC class II-GFP mice were incubated with monomeric or oligomeric forms of HEL in phenol red-free DMEM media supplemented with 25 mM Hepes in eight-well chambered coverglass (Nalge Nunc) for 30 min at 37°C. The cells were then immediately imaged by using a spinning disk confocal microscope. The changes of intracellular MHC class II levels were analyzed by using METAMORPH software (Molecular Devices). The image acquisition and the quantitative image analysis are fully described in Supporting Text. For videomicroscopy, the cell suspension were seeded on poly(Llysine)-treated eight-well chambered coverglass and allowed to settle to the bottom of the chamber for ⬇5–10 min on the stage of the confocal microscope equilibrated to 37°C. The cells were initially imaged without HEL and Alexa Fluor 568-conjugated HEL was added to cells after capturing a few frames of images. Changes in distribution of MHC class II-GFP and Alexa Fluor 568-conjugated HEL in three middle sections of the cells (1 ␮m apart) were sequentially captured in green and red channels with two binnings and a 400-ms exposure time per channel for 30 min. Antigen Presentation Assays. Naı¨ve splenic B cells were purified by

depleting the splenocytes of CD43- or CD11c-positive cells using antibody-conjugated microbeads (Miltenyi Biotec). The purified B cells (5 ⫻ 105) were cultured with the I-Ab restricted and HEL peptide (amino acid 74–88)-specific T cell hybridoma, BO4H9 (1 ⫻ 105) and HEL conjugates in RPMI medium 1640 containing 10% FCS, 2 mM glutamine, 50 ␮M 2-mercaptoethanol, 1 mM sodium pyruvate, 1⫻ nonessential amino acids, and 25 mM Hepes in round-bottomed 96-well plates for 24 h at 37°C. For some experiments, 100 nM CpG (TIB Molbiol) was also included in the culture. The cells were then spun down by centrifugation, and the supernatants were transferred to new plates and frozen. IL-2 in the culture supernatants was measured by ELISA assay. Concentrations of IL-2 in individual samples were determined in triplicate by using recombinant IL-2 as standards.

Detergent Solubilization of BCR Complex. HEL-activated B cells were lysed in a lysis buffer (50 mM Tris兾150 mM NaCl兾5 mM EDTA兾1 ␮g/ml leupeptin兾1 ␮g/ml aprotinin兾5 mM benzamidine) containing either 1% digitonin or 1% Triton X-100. The lysate was run on a SDS兾PAGE gel and transferred to a nitrocellulose membrane. Ig ␮ was detected with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgM (Southern Biotechnology Associates), and Ig

We thank Dr. John Cambier (University of Colorado Health Sciences Center, Denver) for the anti-Ig␤ hybridoma, Drs. Christian Hirsch and Shahram Misaghi for help with preparation of HEL conjugates, Dr. H. Christine Patterson for help with calcium measurement, and the members of the Ploegh laboratory for discussions. This work was supported by grants from National Institutes of Health (to H.L.P.) and the Leukemia and Lymphoma Society (to Y.-M.K.).

1. Wang, L. D. & Clark, M. R. (2003) Immunology 110, 411–420. 2. Dal Porto, J. M., Gauld, S. B., Merrell, K. T., Mills, D., Pugh-Bernard, A. E. & Cambier, J. (2004) Mol. Immunol. 41, 599–613. 3. Campbell, K. S. (1999) Curr. Opin. Immunol. 11, 256–264. 4. Lanzavecchia, A. (1990) Annu. Rev. Immunol. 8, 773–793. 5. Parker, D. C. (1993) Annu. Rev. Immunol. 11, 331–360. 6. Lanzavecchia, A. (1985) Nature 314, 537–539. 7. Siemasko, K. & Clark, M. R. (2001) Curr. Opin. Immunol. 13, 32–36. 8. Clark, M. R., Massenburg, D., Zhang, M. & Siemasko, K. (2003) Ann. N.Y. Acad. Sci. 987, 26–37. 9. Lankar, D., Briken, V., Adler, K., Weiser, P., Cassard, S., Blank, U., Viguier, M. & Bonnerot, C. (1998) J. Exp. Med. 188, 819–831. 10. Siemasko, K., Skaggs, B. J., Kabak, S., Williamson, E., Brown, B. K., Song, W. & Clark, M. R. (2002) J. Immunol. 168, 2127–2138. 11. Clark, M. R., Massenburg, D., Siemasko, K., Hou, P. & Zhang, M. (2004) Curr. Opin. Immunol. 16, 382–387. 12. Siemasko, K., Eisfelder, B. J., Williamson, E., Kabak, S. & Clark, M. R. (1998) J. Immunol. 160, 5203–5208. 13. Lankar, D., Vincent-Schneider, H., Briken, V., Yokozeki, T., Raposo, G. & Bonnerot, C. (2002) J. Exp. Med. 195, 461–472. 14. Goodnow, C. C., Crosbie, J., Adelstein, S., Lavoie, T. B., Smith-Gill, S. J., Brink,

R. A., Pritchard-Briscoe, H., Wotherspoon, J. S., Loblay, R. H., Raphael, K., et al. (1988) Nature 334, 676–682. Boes, M., Cerny, J., Massol, R., Op den Brouw, M., Kirchhausen, T., Chen, J. & Ploegh, H. L. (2002) Nature 418, 983–988. Winslow, M. M., Neilson, J. R. & Crabtree, G. R. (2003) Curr. Opin. Immunol. 15, 299–307. Ishiai, M., Sugawara, H., Kurosaki, M. & Kurosaki, T. (1999) J. Immunol. 163, 1746–1749. Charrin, S., Manie, S., Thiele, C., Billard, M., Gerlier, D., Boucheix, C. & Rubinstein, E. (2003) Eur. J. Immunol. 33, 2479–2489. Dykstra, M., Cherukuri, A., Sohn, H. W., Tzeng, S. J. & Pierce, S. K. (2003) Annu. Rev. Immunol. 21, 457–481. Cheng, P. C., Dykstra, M. L., Mitchell, R. N. & Pierce, S. K. (1999) J. Exp. Med. 190, 1549–1560. Stoddart, A., Dykstra, M. L., Brown, B. K., Song, W., Pierce, S. K. & Brodsky, F. M. (2002) Immunity 17, 451–462. Stoddart, A., Jackson, A. P. & Brodsky, F. M. (2005) Mol. Biol. Cell 16, 2339–2348. Healy, J. I. & Goodnow, C. C. (1998) Annu. Rev. Immunol. 16, 645–670. Eynon, E. E. & Parker, D. C. (1992) J. Exp. Med. 175, 131–138. Gahring, L. C. & Weigle, W. O. (1989) J. Immunol. 143, 2094–2100. Hokazono, Y., Adachi, T., Wabl, M., Tada, N., Amagasa, T. & Tsubata, T. (2003) J. Immunol. 171, 1835–1843.

3332 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0511315103

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Kim et al.