of March 26, 2018.This information is current as

SynapseImmunologicalLipid Rafts Accumulate at the

MHC Class II-Peptide Complexes and APC

Elizabeth M. Hiltbold, Neil J. Poloso and Paul A. Roche

http://www.jimmunol.org/content/170/3/1329doi: 10.4049/jimmunol.170.3.1329

2003; 170:1329-1338; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/170/3/1329.full#ref-list-1

, 21 of which you can access for free at: cites 44 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2003 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

MHC Class II-Peptide Complexes and APC Lipid RaftsAccumulate at the Immunological Synapse

Elizabeth M. Hiltbold, Neil J. Poloso, and Paul A. Roche1

Activation of CD4� Th cells requires their cognate interaction with APCs bearing specific relevant MHC class II-peptide com-plexes. This cognate interaction culminates in the formation of an immunological synapse that contains the various proteins andlipids required for efficient T cell activation. We now show that APC lipid raft membrane microdomains contain specific classII-peptide complexes and serve as platforms that deliver these raft-associated class II molecules to the immunological synapse.APC rafts are required for T cell:APC conjugate formation and T cell activation at low densities of relevant class II-peptidecomplexes, a requirement that can be overcome at high class II-peptide density. Analysis of confocal microscopy images revealedthat over time APC lipid rafts, raft-associated relevant class II-peptide complexes, and even immunologically irrelevant class IImolecules accumulate at the immunological synapse. As the immunological synapse matures, relevant class II-peptide complexesare sorted to a central region of the interface, while irrelevant class II molecules are excluded from this site. We propose that Tcell activation is facilitated by recruitment of MHC class II-peptide complexes to the immunological synapse by virtue of theirconstitutive association with lipid raft microdomains. The Journal of Immunology, 2003, 170: 1329–1338.

C D4� Th cells are activated by specific antigenic peptidespresented by class II MHC molecules on professionalAPCs. Although each T cell bears a clonotypic TCR spe-

cific for one particular MHC-peptide complex, this relevant MHC-peptide complex is likely to represent an extremely small fractionof the total pool of class II-peptide complexes on the surface of anAPC. Although several studies have established that activation ofT cells requires less than 500 specific MHC-peptide complexes perAPC (1–3), the molecular mechanisms by which such a smallnumber of class II molecules are able to stimulate T cells remainto be determined.

Activation of Ag-specific T cells by APCs requires the orches-tration of a complex series of biochemical and morphologicalchanges in T cells. Before interaction with an APC, the TCR isbroadly distributed over the entire T cell surface. Encounter of anAg-specific T cell with Ag-loaded APC halts its migration (4),results in polarization of the T cell cytoskeleton toward the APC(5–7), and initiates the movement of the TCR and CD3 to the Tcell:APC interface (7, 8). This early and critical step in T cellactivation ultimately leads to the formation of an immunologicalsynapse, a tight, highly organized cognate interaction between theT cell and APC (7, 9, 10).

Although the presence of distinct membrane-anchored proteinsat the immunological synapse is well documented (9), the parti-tioning of membrane lipids and membrane microdomains in T cell:APC conjugates has not been extensively investigated. Lipid raftsare dynamic membrane microdomains enriched in cholesterol, gly-cosphingolipids, and GPI-linked proteins (11–13). Association ofproteins with cholesterol-rich raft microdomains is thought to fa-cilitate signal transduction, protein transport, and membrane fusion

by acting as platforms (or rafts), thereby effectively concentratingthe molecules involved in these diverse processes (14–17). Wehave recently shown that MHC class II molecules constitutivelyreside in plasma membrane lipid raft microdomains on resting Bcells (18), and have suggested that raft association increases theprobability for TCR cross-linking by MHC class II-peptidecomplexes.

To evaluate the importance of APC lipid rafts during the initialevents in T cell activation, we have now examined their role in theformation of conjugates between Ag-specific T cells and Ag-loaded APCs. We find that the integrity of APC lipid rafts is crit-ical for T cell:APC conjugate formation and T cell activation, butonly under conditions of low relevant class II-peptide density. Wehave also used confocal immunofluorescence microscopy to ex-amine the distribution of APC lipid rafts and distinct MHC classII-peptide complexes at the immunological synapse. Lipid raftsand class II-peptide complexes accumulate at the immunologicalsynapse in conjugates formed between T cells and APCs possess-ing limiting amounts of specific class II-peptide complexes. At theimmunological synapse, class II molecules segregate from eachother based on their relevance to the T cell. We propose that underphysiologically relevant conditions, the association of class II mol-ecules with lipid rafts results in the delivery of TCR-ligated andnonligated relevant class II-peptide complexes to the immunolog-ical synapse, thereby facilitating T cell activation.

Materials and MethodsCell lines, Abs, and reagents

The mouse B lymphoma cell line, CH27 (H-2k), and the 3A9 T cell hy-bridoma (specific for hen egg lysozyme (HEL)2 (1) residues 46–61 pre-sented by I-Ak) were gifts from S. Pierce (National Institute of Allergy andInfectious Diseases (NIAID), National Institutes of Health). As indicated,splenic primary T cells from 3A9 TCR transgenic mice (The Jackson Lab-oratory, Bar Harbor, ME) were purified by negative selection and usedimmediately. B cells were maintained in DMEM containing 15% FCS,with 50 �M 2-ME and 50 �g/ml gentamicin. T cells were maintained in

Experimental Immunology Branch, National Cancer Institute, National Institutes ofHealth, Bethesda, MD 20892

Received for publication September 6, 2002. Accepted for publication December3, 2002.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 Address correspondence and reprint requests to Dr. Paul A. Roche, Building 10,Room 4B36, Bethesda, MD 20892. E-mail address: paul.roche@nih.gov

2 Abbreviations used in this paper: HEL, hen egg lysozyme; MCD, methyl-�-cyclo-dextrin; MFI, mean fluorescence intensity; DiIC16, 1,1�dihexadecyl-3,3,3�,3�-tetra-methylindocarbocyanine perchlorate.

The Journal of Immunology

Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

RPMI containing 10% FCS, 50 �M 2-ME, and 50 �g/ml gentamicin.DCEK (I-Ek transfectant) and RT7.eH3.B45 (I-Ak transfectant) mouse fi-broblast cell lines were a gift from R. Germain (NIAID, National Institutesof Health) and were maintained as previously described (19).

mAb 14-4-4S, a mouse IgG2a specific for I-Ek, was purchased fromAmerican Type Culture Collection (Manassas, VA), and C4H3, a rat IgG2bspecific for the complex of I-Ak-HEL(46–61) (20), was a gift from R. Ger-main. Both Abs were used as culture supernatants from hybridoma celllines. Transferrin receptor was detected using a mAb (clone C2); CD3� wasdetected using mAb 500A2; and B7.2 was detected using mAb GL1, allpurchased from BD PharMingen (San Diego, CA). Fluorescently labeledsecondary Abs used included Alexa 488- and Alexa 568-labeled anti-ratand anti-mouse Abs purchased from Molecular Probes (Eugene, OR). Anti-mouse and anti-rat Cy5-labeled Abs were purchased from Jackson Immu-noResearch (West Grove, PA). HEL protein was purchased from Sigma-Aldrich (St. Louis, MO). A peptide encompassing HEL aa 46–61 wassynthesized by the Experimental Immunology Branch Laboratory Re-sources section. Methyl-�-cyclodextrin (MCD) and cholera toxin B subunitconjugated to HRP were purchased from Sigma-Aldrich. Cholera toxin Bsubunit labeled with Alexa 488- and Alexa 633-labeled phalloidin werepurchased from Molecular Probes.

Biochemical isolation of detergent-insoluble membranes

Lipid rafts were isolated as previously described (18). Briefly, 108 CH27 Bcells were treated with 5 mM MCD at 37°C for 10 min (or mock treated),washed, and lysed in 1 ml of 1% Triton X-100 on ice for 1 h. The lysatewas mixed with an equal volume of 90% sucrose, and overlaid with 6 mlof a 30% sucrose solution and 4 ml of a 5% sucrose solution. The sampleswere subjected to ultracentrifugation at 35,000 rpm at 4°C for 18 h. One-milliliter fractions were collected and solubilized in 1% octylglucoside for1 h at 4°C. The fractions were solubilized in SDS sample buffer, but notboiled before SDS-PAGE analysis to preserve peptide-loaded, MHC classII-stable dimers. I-Ak-HEL(46–61) complexes were detected by Westernblot analysis using the C4H3 mAb, lipid rafts using cholera toxin B-HRP,and B7.2 using mAb GL1.

Measurement of T cell activation

IL-2 production was measured as previously described (18). Briefly, CH27B cells were loaded with the indicated doses of HEL peptide for 4 h,washed, and treated (or not) with 5 mM MCD at 37°C for 10 min. Fol-lowing treatment, the B cells were washed and fixed with 2% paraformal-dehyde for 30 min at room temperature, quenched with 50 mM NH4Cl,washed, and combined at a 1:1 ratio with 3A9 T cells in culture overnight(105 total cells/well in a 96-well culture plate). Secreted IL-2 was quanti-tated by sandwich ELISA. The amount of IL-2 secreted from T cells in-cubated with MCD-treated APCs at each dose of Ag was expressed as apercentage of the amount secreted when T cells were incubated withmock-treated APCs.

Quantitation of T:B cell conjugates by flow cytometry

Conjugate formation between T and B cells was measured using a modifica-tion of a previously described technique (21). Briefly, 1 � 106 3A9 T cellswere stained with 50 nM CellTracker Green (Molecular Probes) in serum-freeHBSS for 30 min at 37°C. B cells (1 � 106) loaded with the indicated amountof Ag were stained with 1 �g/ml of the lipophilic dye 1,1�-dihexa-decyl-3,3,3�,3�-tetramethylindocarbocyanine perchlorate (DiIC16; MolecularProbes) in DMEM on ice for 30 min. Following staining, the cells werewashed three times in PBS containing 5% FCS. Prestained T and B cells werethen combined, pelleted at 1200 rpm for 5 min, and incubated for 10 min at37°C. In some experiments, stained B cells were pretreated with 5 mM MCDfor 10 min at 37°C, pelleted, and washed before interaction with T cells. Thecells were then gently resuspended in cold PBS/FCS, and 2 � 105 cells werecollected and analyzed immediately on a BD Biosciences (San Diego, CA)FACScan. For reversibility studies, after pretreatment with 5 mM MCD for 10min at 37°C, stained B cells were pelleted, washed, and incubated in serum-containing medium for various times before interaction with T cells. The per-centage of conjugates formed at each time point was expressed as a percentageof net conjugates in MCD-treated cells (conjugates formed in the presence ofAg minus conjugates formed in the absence of Ag) to net conjugates in mock-treated cells.

Conjugate formation and immunofluorescence staining

CH27 B cells (1 � 106) were pulsed with the indicated Ags for the indi-cated times and washed. The 3A9 T cells (1 � 106) were then combinedwith the B cells in 2 ml warm medium in 15-ml round-bottom tubes (Fal-con) and pelleted by centrifugation at 1200 rpm for 5 min at room tem-

perature. The pellets were incubated at 37°C for either 10 or 30 min, afterwhich the medium was aspirated and the cells were gently resuspended in5 ml PBS. An equal volume of 4% paraformaldehyde was added for 30 minat room temperature to fix the cells. The fixed cells were then washed twicefor 10 min each with 50 mM NH4Cl in PBS, then once in PBS alone. Thecells were plated on poly(L-lysine)-coated coverslips for 1 h at room tem-perature. Before staining, nonspecific protein binding was blocked by in-cubating coverslips for 30 min in 5% goat serum diluted in PBS. Thisbuffer was also used as the washing and staining buffer for all the subse-quent steps. Cell conjugates were then stained with the indicated primaryAb (purified Abs were used at a final concentration of 10 �g/ml, andculture supernatants were used undiluted), washed, stained with the appro-priate fluorescently tagged secondary Ab (10 �g/ml), and washed. Succes-sive rounds of staining with other primary and secondary Abs were per-formed as indicated. Staining with cholera toxin (1 �g/ml) or phalloidin (2U/ml) was performed as the final step of staining. The specificity of stain-ing was controlled by using isotype-control Abs, followed by the appro-priate fluorescent secondary Abs. Background staining with control Abswas routinely compared with positively stained cells and was not visibleusing identical acquisition settings. The coverslips were mounted in Pro-long mounting medium (Molecular Probes) and examined using a Zeiss(Oberkochen, Germany) 410 confocal laser-scanning microscope. Eachconjugate was analyzed by acquiring thin (�0.3-�m) sections through thex-y-axis. Images were collected both simultaneously and sequentially torule out bleed-through from one fluorescence channel to another. Sectionsin the xz plane were also acquired through the interface of the cell:cellconjugates.

Quantitation of fluorescence and colocalization in microscopicimages

The percentage of conjugates displaying APC raft recruitment was deter-mined by allowing a blinded observer to score individual conjugates aspositive or negative for raft recruitment. To quantitate the recruitment oflipid rafts or class II molecules to the immunological synapse, boxes (ormasks) of equal area were drawn around the area at the immune synapse,the opposite pole of the cell, and a background area outside of the cell. Themean fluorescence intensity (MFI) for each area was measured using theImage J version of National Institutes of Health Image. The relative re-cruitment index was calculated as indicated: (MFI at synapse � back-ground)/(MFI at opposite pole � background). Twenty-five conjugatesformed under low Ag conditions and twenty-five conjugates formed underhigh Ag dose conditions were examined quantitatively, and the relativerecruitment of APC lipid rafts, relevant I-Ak HEL(46–61) complexes, andirrelevant I-Ek molecules was calculated. The data are plotted with average�/� SEM for these conjugates, and statistical significance was calculatedusing a Student’s t test.

The overlap or colocalization in the fluorescent staining patterns for theindicated molecules was determined using fluorescence intensity profileanalysis acquired from the Zeiss LSM 510 confocal image analysis soft-ware. The profile of fluorescence intensity (expressed in arbitrary units)was measured along a line taken through the mid-plane of the cell in twochannels.

Cross-linking and immunofluorescence

CH27 B cells were incubated in the presence of 10 �g/ml primary Ab for30 min on ice. Unbound primary Ab was removed with two washes in coldPBS. The B cells were resuspended in warm PBS containing 1 �g/mlfluorescently tagged cross-linking secondary Ab. The cells were cross-linked at 37°C for 10 min, pelleted, and resuspended in ice-cold PBS. Anequal volume of 4% paraformaldehyde was then added to fix the cells for30 min. Following fixation, the cells were quenched with two washes in 5mM NH4Cl, resuspended in PBS, and plated on poly(L-lysine)-coated cov-erslips. The cells were stained with the appropriate Abs or cholera toxin,mounted, and analyzed by confocal microscopy. To determine colocaliza-tion of cholera toxin and DiIC16 staining patterns, DiIC16-stained cellswere incubated with Alexa 488-tagged cholera toxin (1 �g/ml) for 30 minon ice. The cells were washed and warmed to 37°C for 10 min to locallypatch lipid rafts. The cells were then fixed, mounted on poly(L-lysine)-coated coverslips, and analyzed by confocal microscopy. The imagesshown are three-dimensional reconstructions of stacked (x-y) sections ac-quired from the top of the cell through the mid-plane.

1330 CLASS II/RAFT ACCUMULATION AT THE IMMUNE SYNAPSE

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

ResultsClass II I-Ak-HEL(46–61) complexes reside in lipid raftmembrane microdomains

We have previously reported that a significant portion of surfaceMHC class II molecules resides in lipid raft microdomains and thatthe inclusion of class II-peptide complexes in APC lipid rafts isimportant for the activation of T cells (18). We now set out toexamine the association of specific class II-peptide complexes withthese lipid rafts and determine the importance of this associationfor the activation of Ag-specific T cells. For these studies, weexamined the lipid raft association of I-Ak-HEL(46–61) complexeson HEL-pulsed mouse CH27 B cells (H-2k) using a class II-pep-tide complex-specific mAb while examining the role APC raftsplay in the activation of I-Ak-HEL(46–61)-restricted 3A9 T cells.

The kinetics of I-Ak-HEL(46–61) complex formation on CH27 Bcells was examined using mAb C4H3 (Fig. 1A), a reagent thatspecifically recognizes SDS-stable I-Ak-HEL(46–61) complexes(20). To determine whether I-Ak-HEL(46–61) complexes werepresent in lipid rafts, Triton X-100 lysates of HEL(46–61)-pulsed Bcells were analyzed by sucrose density gradient centrifugation andimmunoblotting using mAb C4H3. Approximately 26% of all I-Ak-HEL(46–61) complexes partitioned into the Triton X-100-insol-uble, low buoyant density lipid raft fractions of the gradient (Fig.1B). The positions of the raft and nonraft fractions in the gradientwere revealed by probing the blot with HRP-conjugated choleratoxin B subunit (which binds to ganglioside GM1) and anti-B7.2mAb, respectively. To confirm that the low buoyant density ofthese class II-peptide complexes was a consequence of their asso-ciation with lipid raft microdomains, Ag-loaded B cells were pre-treated with MCD, a reagent that extracts plasma membrane cho-lesterol and perturbs lipid raft integrity (22, 23). Pretreatment ofthe cells with MCD completely prevented I-Ak-HEL(46–61) asso-ciation with these microdomains (Fig. 1B). Quantitative analysis ofthe immunoblots confirmed that MCD did not alter the totalamount of I-Ak-HEL(46–61) complexes on the cells.

APC lipid rafts are required for T cell activation atlimiting Ag dose

To determine whether raft association of I-Ak-HEL(46–61) com-plexes was required for T cell activation, we examined the effect ofAPC raft disruption on the activation of I-Ak-HEL(46–61)-specific3A9 T cells. Treatment of peptide-loaded CH27 B cells with MCDinhibited 3A9 T cell activation when the cells were pulsed with asuboptimal dose of HEL(46–61) peptide; however, raft disruptionhad little effect when the APCs were preloaded with a high dose ofAg (Fig. 1C). These results reveal that a significant fraction ofI-Ak-HEL(46–61) complexes constitutively resides in lipid raft mi-crodomains and that B cell lipid rafts are important for 3A9 T cellactivation when APCs possess suboptimal amounts of relevant I-Ak-HEL(46–61) complexes.

APC lipid rafts are required for T:B cell conjugate formation atlimiting Ag dose

Whereas APC rafts are predominantly important for T cell activa-tion under conditions of limiting Ag dose, it remains to be deter-mined at which point in the complex T cell activation cascade APCrafts exert their stimulatory effect. Because the formation of a Tcell:APC conjugate is the earliest event in T cell activation, wehave examined the requirement for APC rafts in the formation ofT:B cell conjugates using a dynamic, live cell system. Conjugatesbetween 3A9 T cells and CH27 B cells were examined using aflow cytometric assay in which cells were stained with distinctfluorescent tracers before interaction (21).

To address the importance of APC lipid rafts in conjugate for-mation, Ag-loaded B cells were pretreated with MCD, washed, andimmediately cultured with T cells before FACS analysis. Asshown in Fig. 1B, MCD treatment does not perturb pre-existingI-Ak-HEL(46–61) complexes. In the absence of Ag, �3% of T cellswere present in T:B cell conjugates representing nonspecific orbackground adhesion (Fig. 2A). MCD-mediated extraction ofplasma membrane cholesterol had no effect on background adhe-sion in this system, demonstrating that APC rafts are not involvedin Ag-independent adhesive interactions between T and B cells(Fig. 2B). When APCs were pulsed with various doses of antigenicpeptide overnight, we observed a dose-dependent increase in thepercentage of T cells in conjugates up to a maximum of �15%. Aswas observed in our studies examining IL-2 production, APC lipidraft disruption had a minimal effect on conjugate formation when

FIGURE 1. A substantial fraction of I-Ak-HEL(46–61) complexes is lipidraft associated. A, CH27 B cells were loaded with HEL (1 mg/ml) forvarious times, washed, and lysed in Triton X-100. Aliquots of the lysateswere incubated in SDS-PAGE sample buffer at 37°C for 1 h and analyzedby SDS-PAGE and immunoblotting with mAb C4H3. B, CH27 B cellswere loaded with HEL (1 mg/ml) overnight, washed, then treated withMCD (or not). The cells were then lysed in Triton X-100, and subsequentlyfractionated by sucrose density gradient centrifugation. The fractions werecollected and analyzed by SDS-PAGE and immunoblotted with mAbC4H3 to detect I-Ak-HEL(46–61) complexes, HRP-cholera toxin B subunitto detect GM1 in lipid rafts, and mAb GL1 to detect B7.2 in nonraft frac-tions. C, CH27 B cells were pulsed with the indicated dose of HEL peptide,treated with MCD to disrupt lipid rafts, and then fixed with paraformal-dehyde. The 3A9 T cells were then cultured with the fixed B cells, and Tcell activation was measured 18 h later by IL-2 production.

1331The Journal of Immunology

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

the B cells were pulsed with a high dose of peptide. By contrast,APC lipid raft disruption essentially eliminated Ag-dependent con-jugate formation with B cells pulsed with low doses of peptidedown to the level of nonspecific cell adhesion (Fig. 2B). This dis-ruption of conjugate formation was statistically significant onlywhen B cells were pulsed with 0.1 �g/ml HEL(46–61) ( p � 0.05)and 1 �g/ml HEL(46–61) ( p � 0.01). These results demonstratethat disruption of lipid rafts on B cells preferentially inhibits Ag-

specific T:B cell conjugate formation under conditions of low classII-peptide density.

The effects of APC raft disruption are reversible

An obvious concern in studies using agents that perturb cellularcholesterol levels is that these treatments can irreparably damagecell integrity and function. To determine whether the effects ofAPC raft disruption were reversible, we pretreated peptide-pulsedB cells with MCD for 10 min, washed away the drug, and allowedthe B cells to recover in cholesterol-containing medium for varioustimes before the addition of T cells. Although the initial treatmentwith MCD almost completely prevented conjugate formation, theability of T cells to form conjugates with these B cells was re-gained in a time-dependent manner, leading to a 90% recovery ina 2-h culture period (Fig. 2C). Similarly, when MCD-treated Bcells were allowed to recover, they regained their capacity to elicitIL-2 production by Ag-specific T cells (data not shown). Thesedata show that raft perturbation does not irreparably damage the Bcells and does not selectively remove preformed I-Ak-HEL(46–61)

complexes from the B cell membrane.

APC lipid rafts accumulate at the immunological synapse

We next set out to examine the distribution of class II moleculesand APC lipid rafts at the site of contact between T and B cells byimmunofluorescence microscopy. To restrict our analysis to pro-ductive Ag-dependent T:B cell interactions, we routinely costainedconjugates with fluorescent phalloidin, a marker for filamentousactin that polarizes to the immunological synapse during T cellsignaling (7). Analysis of lipid raft localization in T:B conjugatesusing fluorescent cholera toxin (which binds to ganglioside GM1)revealed lipid raft concentration at the immunological synapse(Fig. 3A). Unfortunately, due to the ability of cholera toxin to bindto GM1 on both cells and the close apposition of APC and T cellmembranes in these conjugates, it was impossible to exclusivelyexamine APC lipid rafts with this technique. Therefore, to specif-ically monitor the behavior of B cell lipid rafts at the immunolog-ical synapse, we chose to prelabel the B cell rafts with DiIC16, afluorescent lipid analog that preferentially inserts into liquid-or-dered (raft-like) domains on the plasma membrane (24–27). Toconfirm that this fluorescent lipid faithfully tagged lipid raft mi-crodomains, we assessed the colocalization of DiIC16 with fluo-rescent cholera toxin on raft-aggregated B cells. We found that thetwo markers exhibited almost complete colocalization (Fig. 3A),demonstrating that they labeled the same microdomains.

Because APC lipid rafts are most important for T:B cell conju-gate formation at low doses of Ag, we examined the distribution ofB cell lipid rafts at the immune synapse using APCs loaded witheither low or high doses of Ag. In the absence of Ag, there was noaccumulation of B cell rafts at the synapse of nonspecific T:B cellconjugates (data not shown), while at low Ag doses B cell lipidrafts accumulated at the T:B cell interface (Fig. 3B). As expected,in the absence of Ag there was no accumulation of B cell rafts atthe synapse of nonspecific T:B cell conjugates (data not shown).Interestingly, under high Ag dose conditions, we also observed nodetectable enrichment of B cell lipid rafts at the synapse relative tothe rest of the cell. To address the prevalence of this dose-depen-dent lipid raft recruitment in a population of individual conjugates,raft recruitment to the synapse was scored by a blinded observer.Whereas significant raft accumulation was observed in 88% ofconjugates formed under low Ag dose conditions, raft accumula-tion was observed in less than 10% of conjugates formed underhigh dose Ag conditions (Fig. 3B). Similar results were obtainedwhen the B cell lipid rafts were prelabeled with fluorescent cholera

FIGURE 2. T:B cell conjugate formation at low Ag dose is dependenton B cell lipid rafts. A, CH27 B cells were pulsed with no Ag or with 100�g/ml HEL(46–61) peptide for 4 h. Live B cells stained with DiIC16 (ab-scissa) were combined with live 3A9 T cells stained with CellTrackerGreen (ordinate) in culture for 10 min, and were analyzed by flow cytom-etry. T:B cell conjugates migrate to the upper right double-positive quad-rant. B, CH27 B cells were pulsed for 4 h with the indicated dose ofHEL(46–61) peptide before treatment with MCD for 10 min. The B cellswere washed and combined with 3A9 T cells and cultured for 10 min at37°C. The percentage of T cells in T:B cell conjugates was determined byflow cytometry and dot-plot analysis. Data shown are the mean �/� SEMof four independent experiments. Values of p for each Ag dose (�, p �0.05; ��, p � 0.01) were determined using a Student’s t test. C, CH27 Bcells were pulsed with 1 �g/ml HEL(46–61) peptide, stained with DiIC16,washed, and treated with MCD (5 mM, 10 min, 37°C). The cells were thenallowed to recover in complete medium for the indicated periods andwere analyzed for their ability to form conjugates with 3A9 T cells, asabove. Data are expressed as percentage of control (mock treated) foreach time point.

1332 CLASS II/RAFT ACCUMULATION AT THE IMMUNE SYNAPSE

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

toxin B subunit (data not shown), demonstrating that these resultswere not restricted to our use of DiIC16 as a lipid raft marker.

MHC class II molecules and APC lipid rafts accumulate at theimmunological synapse

Because APC lipid rafts are critical to conjugate formation only atlow Ag dose, and APC lipid rafts are differentially recruited to the

immunological synapse under these same conditions, we set out toaddress the localization of class II-peptide complexes recognizedby the T cell at the immunological synapse at high and low Agdoses. To specifically visualize the class II peptide complexes rec-ognized by the 3A9 T cell, we used mAb C4H3, a reagent thatspecifically stains I-Ak-HEL(46–61) complexes on living cells (20).Because several types of class II MHC molecules reside in lipidraft microdomains (18), we also wanted to determine whetherother types of class II molecules (even those irrelevant class IImolecules not recognized by the T cell) could also be recruited tothe immunological synapse. To detect these irrelevant class II mol-ecules, we used the mAb, 14-4-4S, which detects I-Ek molecules.The exquisite specificity of these Abs was confirmed using classII-transfected mouse fibroblasts expressing only I-Ak or I-Ek mol-ecules or using CH27 B cells in the absence or presence ofHEL(46–61) peptide (Fig. 4).

Under conditions in which B cell lipid rafts were recruited to theimmunological synapse (low Ag dose), we observed an enrichmentof relevant I-Ak HEL(46–61) complexes at the immunological syn-apse (Fig. 5A). This accumulation was not observed when B cellswere not pulsed with HEL peptide (data not shown). Like lipid raftrecruitment, the concentration of relevant class II-peptide com-plexes was not observed at high Ag dose. Surprisingly, under lowAg dose conditions, we also observed a moderate enrichment ofirrelevant (I-Ek) class II molecules at the T:B cell interface (Fig.

FIGURE 3. APC lipid rafts accumulate at the immunological synapse atlow Ag doses. CH27 B cells were pulsed with either 100 �g/ml HEL(46–61)

peptide (high dose) or 1 �g/ml HEL(46–61) peptide (low dose) overnightat 37°C. A (upper two panels), The B cells were then cultured with 3A9 Tcells, fixed, and stained with fluorescent phalloidin (green) and choleratoxin (red). A (lower two panels), CH27 B cells were stained live withDiIC16 (red), and fluorescent cholera toxin was used to locally patch lipidrafts (green). Confocal sections were acquired in the x-y plane. Areas ofcolocalization of GM1 and DiIC16 are indicated by arrows. B, The B cellswere stained live with DiIC16 (red), washed, and combined with 3A9 Tcells in culture for 10 min at 37°C. The cells were fixed and stained withfluorescent phalloidin (green) to label filamentous actin. Single confocalsections through the center of the conjugates are shown. Scale bars repre-sent 3 �m. Fifty T:B cell conjugates were examined by a blinded observerand scored for lipid raft recruitment to the T:B cell interface. Each conju-gate examined was counterstained with phalloidin to ensure that these wereproductive T:B cell conjugates. The percentage of conjugates in each groupthat displayed marked raft recruitment to the immunological synapse wasthen calculated.

FIGURE 4. Specificity of class II MHC Abs. A, Mouse fibroblast trans-fectants expressing either I-Ak (RT-7) or I-Ek (DCEK) were incubatedalone or with 100 �g/ml HEL(46–61) peptide for 4 h before staining withmAb C4H3 (recognizing the I-Ak-HEL(46–61) complexes) or mAb 14-4-4S(recognizing I-Ek molecules). Bright field images illustrate the averagedensity of cells per field. B, CH27 B cells were incubated with 100 �g/mlHEL(46–61) peptide for various times, fixed, and stained with mAb C4H3 todetect I-Ak-HEL(46–61) complexes.

1333The Journal of Immunology

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

5A). Again, this enrichment was not observed under conditions ofhigh Ag dose.

To address the extent of accumulation of lipid rafts, relevantI-Ak HEL(46–61) complexes, and irrelevant I-Ek molecules to theimmune synapse, we quantitatively compared the total pixel in-tensity of the area encompassing the T:B contact site with the totalpixel intensity of an equivalent area on the opposite pole of the Bcell. Thus, with no relative enrichment at the immunological syn-apse, the recruitment index would be 1:1. In conjugates formedbetween 3A9 T cells and B cells loaded with a high dose of HELpeptide, there was no significant enrichment of either B cell lipidrafts (1.1:1), relevant I-Ak HEL(46–61) complexes (1.0:1), or irrel-evant I-Ek molecules (0.9:1) at the immunological synapse (Fig.5B). In contrast, at low Ag dose there was a striking increase in therecruitment index of B cell lipid rafts (3.8:1), relevant I-Ak

HEL(46–61) complexes (1.9:1), and even irrelevant I-Ek molecules(1.5:1). These values comparing recruitment at low Ag dose vshigh Ag dose were significantly different for each marker evalu-ated ( p � 0.0001). These data, together with those describedabove, demonstrate that relevant class II-peptide complexes andlipid rafts are preferentially recruited to (or retained at) the immu-nological synapse under low Ag dose conditions, and that the re-

cruitment of relevant class II molecules and B cell rafts to thesynapse brings a limited amount of irrelevant class II moleculesinto the synapse as well.

Distinct class II MHC molecules are present in the same lipidraft microdomains

One could account for the coaccumulation of relevant I-Ak-HEL(46–61) complexes and irrelevant I-Ek molecules to the immu-nological synapse if these two distinct types of class II moleculesoccupied the same membrane microdomain. To test this experi-mentally, we used an Ab-patching approach (11, 28). In mockcross-linked cells, relevant I-Ak-HEL(46–61) and irrelevant I-Ek

molecules were uniformly distributed over the surface of CH27 Bcells (Fig. 6A). When I-Ek molecules were aggregated using amAb, we observed large surface clusters of I-Ek. In many of theseclusters, coaggregation of I-Ak-HEL(46–61) complexes was clearlyapparent (Fig. 6A). Costaining of fixed cells with fluorescent chol-era toxin revealed a similar copatching of a fraction of the plasmamembrane lipid rafts with the aggregated I-Ek molecules. Fluores-cence intensity profile analysis confirmed that there were dramaticcoconcentrations of I-Ek and I-Ak-HEL(46–61) complexes as wellas I-Ek and lipid rafts after I-Ek cross-linking. Similarly, patchingof I-Ak-HEL(46–61) complexes resulted in copatching of some I-Ek

molecules, but with less dramatic effects. This is most likely due tosmaller numbers of molecules patched with the I-Ak-HEL(46–61)

complex-specific mAb as compared with the I-Ek-specific mAb(which recognizes a large fraction of the total class II pool). As acontrol, we examined the distribution of the transferrin receptor, anonraft protein, on I-Ek cross-linked cells. Patching of I-Ek did notalter the uniform distribution of transferrin receptor on these cells,an observation that was confirmed by the lack of correlation be-tween the clustered intensity profile of the I-Ek patches with therelatively smooth intensity profile of the transferrin receptor (Fig.6B). These data confirm that the I-Ek patches did not nonspecifi-cally sequester all plasma membrane proteins.

To determine whether copatching of the two distinct class IImolecules was dependent on the integrity of lipid rafts, we treatedthe B cells with MCD before Ab-mediated patching of I-Ek. Al-though patching of I-Ek molecules was still observed in MCD-treated B cells, copatching of both I-Ak-HEL(46–61) complexes andB cell lipid rafts with the cross-linked I-Ek molecules was com-pletely prevented by raft disruption of the B cell (Fig. 6C, and datanot shown). These results demonstrate that distinct types of classII molecules can occupy the same membrane microdomains andthat their coaggregation is lipid raft dependent.

Relevant and irrelevant class II molecules segregate at theimmunological synapse

Because we found the accumulation of both relevant and irrelevantclass II molecules at the immune synapse, we set out to examinethe distribution of these distinct types of class II molecules at thissite. The 3A9 T cells were allowed to interact with HEL(46–61)-loaded B cells for either 10 or 30 min, and the conjugates wereanalyzed by confocal microscopy. Fig. 7A shows a three-dimen-sional reconstruction of a typical conjugate displaying prominent Tcell actin polarization (green) and MHC class II staining (red) onthe B cell. To examine the distribution of class II molecules at theimmunological synapse, images were acquired in the xz plane (rep-resented by the yellow plane in Fig. 2A), generating an en faceview of the T:B cell interface. Upon short periods of incubation(�10 min), we observed a diffuse distribution of both relevant I-Ak

HEL(46–61) complexes and irrelevant I-Ek molecules at the immu-

FIGURE 5. Class II-peptide complexes are corecruited to the immuno-logical synapse under low Ag dose conditions. A, CH27 B cells werepulsed with either 100 �g/ml HEL(46–61) peptide (high dose) or 1 �g/mlHEL(46–61) peptide (low dose) overnight at 37°C, stained with DiIC16

(red), and incubated with 3A9 T cells to form T:B cell conjugates. The cellswere fixed and stained with mAb C4H3 to label relevant I-Ak-HEL(46–61)

complexes (green) and mAb 14-4-4S to label irrelevant I-Ek molecules(blue). Because the T cells were not stained in these studies, the outline ofthe T cells (visualized by bright field imaging) is indicated by a dashed line.Scale bars represent 3 �m. B, Recruitment of lipid rafts and class II mol-ecules to the immunological synapse was quantitated. Conjugates formedat either low or high Ag dose were stained independently with Abs foreither I-Ak-HEL(46–61) or I-Ek along with the lipid raft dye DiIC16 andfluorescent phalloidin. Relative recruitment index for 25 conjugates in eachcondition was calculated as described in Materials and Methods.

1334 CLASS II/RAFT ACCUMULATION AT THE IMMUNE SYNAPSE

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

nological synapse (Fig. 7B). After 30 min of interaction, a periodreported by others to be sufficient to form a mature synapse (29,30), we observed the sorting of the relevant I-Ak-HEL(46–61) com-plexes toward the center of the synapse and the exclusion of theI-Ek molecules from this region. The sorting of relevant and irrel-evant class II molecules was observed in conjugates formed underboth high and low Ag dose conditions (data not shown). Stainingof conjugates with the pan I-Ak mAb 10.2.16 failed to reveal anypartitioning at the immunological synapse (E. M. Hiltbold, unpub-lished observations), reflecting the fact that this mAb recognizesboth relevant I-Ak-HEL(46–61) complexes as well as irrelevant I-Ak-peptide complexes.

Because the distinction between relevant and irrelevant class IImolecules is defined by the specificity of the TCR, we examinedthe distribution of CD3� with relevant and irrelevant class II mol-ecules at the immunological synapse. In the mature synapse, therewere dramatic concentration and marked colocalization of CD3�

and relevant I-Ak-HEL(46–61) complexes in a tight cluster near thecenter of the immunological synapse (Fig. 7C). By contrast, weconsistently observed a complete exclusion of irrelevant I-Ek mol-ecules from CD3 in the mature immunological synapse.

To determine the frequency of segregation of relevant and ir-relevant class II molecules at a mature (30-min) immunologicalsynapse, we scored the segregation of I-Ak-HEL(46–61) complexesand irrelevant I-Ek molecules in 45 conjugates (Fig. 7C). We ob-served complete segregation of relevant and irrelevant class IImolecules in 71% of the conjugates and partial segregation in anadditional 15% of the conjugates, and no detectable segregation ofthese proteins at the immunological synapse was observed in only14% of the conjugates examined. These results reveal that although aT cell may encounter a mixed population of relevant and irrelevantclass II molecules upon initial interaction with an APC, these mole-cules, based on their relevance to the T cell, are sorted into either thecenter or the periphery of the interaction site over time.

FIGURE 6. Distinct MHC class II molecules coaggregate on the plasma membrane in a lipid raft-dependent manner. A, CH27 B cells were loaded with1 mg/ml HEL protein overnight, washed, and incubated with the following Abs: rat IgG (Control), or mAb 14-4-4S (I-Ek X-linked) under cross-linkingconditions (as described in Materials and Methods). The cells were then fixed and counterstained with the indicated Abs to stain bystander molecules (listedabove pictures). Fluorescently tagged cholera toxin (blue) was used to label lipid rafts on the fixed cells. The fluorescence profile histogram for each imageillustrates the fluorescence intensity in each channel along a line traversing the membrane along the mid-plane of the cell. B, I-Ek molecules on the B cellswere cross-linked as above (red). The cells were then fixed and counterstained with mAb C2 to stain transferrin receptor (green). C, CH27 cells were treatedwith 5 mM MCD or mock treated for 10 min. The cells were then washed and cross-linked with 14-4-4S mAb, as above (red). The cells were then fixedand counterstained with the C4H3 Ab specific for I-Ak-HEL(46–61) complexes, as above (green). Scale bars represent 3 �m.

1335The Journal of Immunology

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

DiscussionAlthough the precise physical nature of lipid raft microdomains isthe topic of intense investigation and vigorous debate, there is littledoubt that the concentration of specific membrane proteins withinthese microdomains on the plasma membrane of T cells is essentialfor TCR-dependent T cell activation (28, 31, 32). In this study, wehave investigated the role of APC rafts in lymphocyte biology byexamining the behavior of a particular class II-peptide complexinvolved in T cell activation. Approximately one-quarter of these

I-Ak-HEL(46–61) complexes reside in lipid rafts on B cells and,most importantly, cholesterol depletion disrupted class II-raft as-sociations and inhibited the activation of I-Ak-HEL(46–61)-re-stricted T cells when APCs possessed limiting amounts of relevantclass II-peptide complexes. Because lipid raft isolation proceduresare likely to perturb some protein-lipid interactions, our biochem-ical determination of class II association with lipid rafts may ac-tually underestimate the in vivo association of I-Ak-HEL(46–61)

complexes with these microdomains. This is supported by our

FIGURE 7. Segregation of MHCclass II molecules at the immunolog-ical synapse. A, CH27 B cells werepulsed with 100 �g/ml HEL(46–61)

overnight at 37°C, washed, and cul-tured with 3A9 T cells for 10 min.Conjugates were fixed and stainedwith fluorescent phalloidin to visual-ize filamentous actin (green) and mAb14-4-4S to visualize MHC class IImolecules (red). A three-dimensionalreconstruction of the conjugate isshown on the left panel, and a depic-tion of a typical xz plane used in con-focal analysis of the immunologicalsynapse is shown on the right panel.B, 3A9 T cells were cultured with Ag-loaded (100 �g/ml HEL(46–61)) CH27B cells for either 10 or 30 min. Thecells were then fixed and stained withmAb C4H3 to label relevant I-Ak

HEL(46–61) complexes (red) or mAb14-4-4S to label irrelevant I-Ek mole-cules (green). The organization ofeach type of class II molecule at theimmunological synapse was analyzedby confocal analysis of z sections ac-quired through the T:B cell interface.Regions of colocalization are shownin yellow in the overlay. Scale barsrepresent 3 �m. C, 3A9 T cells andHEL-loaded (100 �g/ml HEL(46–61))CH27 B cells were combined for 30min at 37°C, fixed, and stained with amAb recognizing mouse CD3�(green), mAb C4H3 to label relevantI-Ak HEL(46–61) complexes (red), andmAb 14-4-4S to label irrelevant I-Ek

molecules (blue). Confocal sectionswere collected in the x-y plane (twoleft panels) and through the z plane ofT:B cell interface (three right panels).Scale bars represent 3 �m. D, Fre-quency of class II sorting was scoredby a blinded observer. Forty-five con-jugates like those shown above wereexamined in the xz plane, and the seg-regation of I-Ak-HEL(46–61) com-plexes from I-Ek class II moleculeswas scored as being either completelysegregated, partially segregated, ornot differentially sorted, as describedin the text.

1336 CLASS II/RAFT ACCUMULATION AT THE IMMUNE SYNAPSE

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

observation that class II clustering leads to a profound redistribu-tion of rafts on live B cells. Overall, these data are in excellentagreement with our previous study using I-Ek-pigeon cytochromec(88–103)-restricted T cells (18) and point to an important role oflipid rafts in concentrating specific class II-peptide complexes inmembrane microdomains.

Because T cell activation (as measured by IL-2 release) is arelatively late indicator of Ag presentation, we set out to determinewhy disrupting APC rafts was detrimental to T cell activation onlywhen APCs possessed limiting amounts of Ag. By examining theearliest event in T cell activation, T:B cell conjugate formation, wefound that disrupting APC rafts also preferentially inhibited T:Bcell conjugate formation under conditions of low Ag dose, therebyproviding a mechanism to account for diminished T cell activationunder these conditions. Because raft association is thought to in-crease the local density of plasma membrane proteins (17), it islikely that under conditions of high Ag dose the density of classII-peptide complexes is sufficiently great so that effective T:B con-jugate formation and T cell activation occur even when APC raftsare perturbed.

The significance of the constitutive association of class II mol-ecules with APC lipid rafts was revealed by our surprising discov-ery that even irrelevant or bystander class II I-Ek molecules wererecruited to and retained at a synapse formed by an APC and anI-Ak-HEL(46–61)-restricted T cell. The corecruitment of relevantand irrelevant class II molecules to the immunological synapsecould only occur if these two types of class II molecules werephysically associated. By coaggregation analysis, we have shownthat these distinct class II molecules can be associated with thesame membrane microdomain and that perturbation of plasmamembrane rafts prevented their association, suggesting that coas-sociation of relevant and irrelevant class II molecule with lipidrafts is responsible for their corecruitment to the immunologicalsynapse. Unfortunately, because class II concentration at the syn-apse only occurs at low Ag dose conditions and because underthese conditions raft perturbation prevents T:B cell conjugate for-mation, we are unable to test this hypothesis directly. Assumingthat class II ligation by mAb is analogous to class II ligation by theTCR, the failure of distinct class II molecules to coaggregate inraft-depleted cells strongly supports this idea.

The concentration of relevant class II-peptide complexes to theimmunological synapse is likely to be a TCR-driven process. Dur-ing T cell activation, there is a profound reorganization of the Tcell cytoskeleton to the immunological synapse (33), and althoughdisruption of the T cell actin cytoskeleton prevents class II con-centration and T cell activation, disruption of the APC actin cy-toskeleton has no effect on either ICAM-1 or class II concentrationat the synapse (34, 35). Additional evidence in support of thishypothesis are data showing that lipid-anchored MHC-peptidecomplexes in artificial bilayers move to the center of the immunesynapse upon encounter with an Ag-specific T cell (29). Based onthese observations, we propose that the concentration of class IImolecules and APC lipid rafts at the immunological synapse is dueto TCR-mediated trapping of raft-associated relevant class II-pep-tide complexes that have passively diffused into the T:B cell in-terface. The rate of lateral diffusion of class II molecules on the Bcell plasma membrane (�1 � 10�10 cm2 s�1) (36) is sufficientlyhigh so as to allow significant movement of class II molecules onthe B cell to the T:B cell interface under the time frame of immu-nological synapse formation examined in this study; however, thisdoes not necessarily rule out a novel active process of class IImovement to the synapse.

Our study suggests that for APC lipid rafts to functionally con-centrate class II molecules, there must be at least two of these

molecules per raft at any given time. Whether this occurs throughthe fusion of pre-existing rafts, the preferential association ofnewly formed class II-peptide complexes in rafts during class IItransports to the cell surface, or the association of class II mole-cules with pre-existing plasma membrane rafts at the cell surfaceremains to be determined. The delivery of class II molecules to theplasma membrane via tubules emanating from lysosomal Ag-pro-cessing compartments has recently been visualized in stimulateddendritic cells (37, 38), and it is intriguing to speculate that such aprocess could allow the delivery of large numbers of relevant classII-peptide complexes to plasma membrane microdomains. Further-more, our demonstration that distinct class II molecules can residein the same membrane microdomain provides a potential mecha-nism to retain not only TCR-engaged class II molecules, but alsothose potentially relevant class II-peptide complexes that have notyet engaged the TCR. It is possible that this is the physiologicalsignificance of class II association with lipid rafts.

The TCR-dependent segregation of immunologically relevantand irrelevant class II molecules at the synapse is perhaps the moststriking finding of our study. Because distinct types of class IImolecules can reside in the same lipid raft and because lipid raftsare thought to be less than 200 nm in diameter (12, 39), there mustbe some mechanism to remove a constraint to class II diffusion atthe synapse. It is possible that fusion of many individual rafts at thesynapse forms a giant raft that allows the diffusion of irrelevantclass II molecules from the centrally located TCR in a raft-richmembrane, an idea that is supported by data showing that the im-mune synapse is lipid raft rich (38, and this study). It is also pos-sible that at the synapse relevant and irrelevant class II moleculesdissociate from rafts and that irrelevant class II molecules thendiffuse from the centrally located TCR in a nonraft membrane. Theidea that individual rafts exist only transiently is consistent with thislatter possibility (40, 41); however, the precise mechanism behindclass II segregation remains to be determined experimentally.

Our finding of differential recruitment and sorting of class II-peptide complexes may in part explain discrepancies in the liter-ature regarding the extent of class II concentration at the immu-nological synapse (35, 42, 43). As we found by examining thedistribution of relevant I-Ak-HEL(46–61) complexes at the immunesynapse, the amount of recruitment observed will depend primarilyon the amount of relevant class II molecules bound to a givenamount of TCR and, obviously, whether one is examining all classII molecules on the APC or only immunologically relevant classII-peptide complexes. It is interesting to note that when APCs pos-sess extremely large amounts of relevant class II-peptide com-plexes, accumulation at the synapse is also not observed, a resultthat is most likely due to the fact that the number of relevant classII-peptide complexes far exceeds the number of TCRs availablefor ligation at the synapse.

It is interesting to note that like irrelevant class II molecules,CD4 sorts from a central to a peripheral location in the maturesynapse (44). The possibility exists, therefore, that as CD4 mole-cules dissociate from the TCR and migrate to the periphery of thesynapse, they bind class II molecules that are not engaged withTCR, resulting in their segregation into the periphery. The pres-ence of these CD4-irrelevant class II complexes may serve to aug-ment adhesive interactions between T cells and APCs that alreadyexist in the ICAM-1/LFA-1-rich adhesive zone of the immunolog-ical synapse.

A recent study revealed the clustering of both agonist peptide-MHC complexes as well as nonstimulatory null class II-peptidecomplexes at the immune synapse (35). In contrast to our findingsin which irrelevant class II molecules were excluded from the cen-tral region of the synapse, they observed these null complexes

1337The Journal of Immunology

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

centrally concentrated at the immunological synapse. These nullpeptide-class II complexes are distinct from the immunologicallyirrelevant class II alleles examined in our study in that they consistof a version of the relevant agonist peptide (lacking two of threeTCR contact residues) bound to a relevant class II allele. Wulfinget al. (35) proposed that these complexes accumulate at the syn-apse in part due to their immeasurably low direct binding to theTCR. This hypothesis is entirely consistent with our data that dem-onstrate that class II-peptide complexes that are in any way rele-vant to the T cell sort into the center of the synapse along withthe TCR.

AcknowledgmentsWe thank Dr. Ronald Germain for the gift of the C4H3, DCEK, and RT7cell lines, and Drs. Susan Pierce and Anu Cherukuri for the gift of the 3A9and CH27 cell lines. We are also grateful to Drs. Tilmann Brotz, MartinBrown, and Michael Kruhlak for technical advice and instruction in mi-croscopic studies, and Susan Sharrow for advice in flow cytometric anal-yses. Thanks also to David Winkler for synthesis of HEL peptide. We alsoappreciate the critical review of this manuscript by Drs. Alfred Singer,Richard Hodes, and David Wiest.

References1. Demotz, S., H. M. Grey, and A. Sette. 1990. The minimal number of class II

MHC-antigen complexes needed for T cell activation. Science 249:1028.2. Harding, C. V., and E. R. Unanue. 1990. Quantitation of antigen-presenting cell

MHC class II/peptide complexes necessary for T-cell stimulation. Nature 346:574.

3. Reay, P. A., K. Matsui, K. Haase, C. Wulfing, Y. H. Chien, and M. M. Davis.2000. Determination of the relationship between T cell responsiveness and thenumber of MHC-peptide complexes using specific monoclonal antibodies. J. Im-munol. 164:5626.

4. Dustin, M. L., S. K. Bromley, Z. Kan, D. A. Peterson, and E. R. Unanue. 1997.Antigen receptor engagement delivers a stop signal to migrating T lymphocytes.Proc. Natl. Acad. Sci. USA 94:3909.

5. Anton van der Merwe, P., S. J. Davis, A. S. Shaw, and M. L. Dustin. 2000.Cytoskeletal polarization and redistribution of cell-surface molecules during Tcell antigen recognition. Semin. Immunol. 12:5.

6. Acuto, O., and D. Cantrell. 2000. T cell activation and the cytoskeleton. Annu.Rev. Immunol. 18:165.

7. Dustin, M. L., and J. A. Cooper. 2000. The immunological synapse and the actincytoskeleton: molecular hardware for T cell signaling. Nat. Immun. 1:23.

8. Krummel, M. F., and M. M. Davis. 2002. Dynamics of the immunological syn-apse: finding, establishing and solidifying a connection. Curr. Opin. Immunol.14:66.

9. Bromley, S. K., W. R. Burack, K. G. Johnson, K. Somersalo, T. N. Sims,C. Sumen, M. M. Davis, A. S. Shaw, P. M. Allen, and M. L. Dustin. 2001. Theimmunological synapse. Annu. Rev. Immunol. 19:375.

10. Delon, J., and R. N. Germain. 2000. Information transfer at the immunologicalsynapse. Curr. Biol. 10:R923.

11. Harder, T., P. Scheiffele, P. Verkade, and K. Simons. 1998. Lipid domain struc-ture of the plasma membrane revealed by patching of membrane components.J. Cell Biol. 141:929.

12. Pralle, A., P. Keller, E. Florin, K. Simons, and J. K. Horber. 2000. Sphingolipid-cholesterol rafts diffuse as small entities in the plasma membrane of mammaliancells. J. Cell Biol. 148:997.

13. Simons, K., and E. Ikonen. 1997. Functional rafts in cell membranes. Nature387:569.

14. Miceli, M. C., M. Moran, C. D. Chung, V. P. Patel, T. Low, and W. Zinnanti.2001. Co-stimulation and counter-stimulation: lipid raft clustering controls TCRsignaling and functional outcomes. Semin. Immunol. 13:115.

15. Cherukuri, A., M. Dykstra, and S. K. Pierce. 2001. Floating the raft hypothesis:lipid rafts play a role in immune cell activation. Immunity 14:657.

16. Ikonen, E. 2001. Roles of lipid rafts in membrane transport. Curr. Opin. CellBiol. 13:470.

17. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nat. Rev.Mol. Cell Biol. 1:31.

18. Anderson, H. A., E. M. Hiltbold, and P. A. Roche. 2000. Concentration of MHCclass II molecules in lipid rafts facilitates antigen presentation. Nat. Immun.1:156.

19. Ronchese, F., R. H. Schwartz, and R. N. Germain. 1987. Functionally distinctsubsites on a class II major histocompatibility complex molecule. Nature 329:254.

20. Zhong, G., C. Reis e Sousa, and R. N. Germain. 1997. Production, specificity, andfunctionality of monoclonal antibodies to specific peptide-major histocompati-bility complex class II complexes formed by processing of exogenous protein.Proc. Natl. Acad. Sci. USA 94:13856.

21. Perez, P., J. A. Bluestone, D. A. Stephany, and D. M. Segal. 1985. Quantitativemeasurements of the specificity and kinetics of conjugate formation betweencloned cytotoxic T lymphocytes and splenic target cells by dual parameter flowcytometry. J. Immunol. 134:478.

22. Keller, P., and K. Simons. 1998. Cholesterol is required for surface transport ofinfluenza virus hemagglutinin. J. Cell Biol. 140:1357.

23. Ilangumaran, S., and D. C. Hoessli. 1998. Effects of cholesterol depletion bycyclodextrin on the sphingolipid microdomains of the plasma membrane. Bio-chem. J. 335:433.

24. Sheets, E. D., D. Holowka, and B. Baird. 1999. Critical role for cholesterol inLyn-mediated tyrosine phosphorylation of Fc�RI and their association with de-tergent-resistant membranes. J. Cell Biol. 145:877.

25. Thomas, J. L., D. Holowka, B. Baird, and W. W. Webb. 1994. Large-scale co-aggregation of fluorescent lipid probes with cell surface proteins. J. Cell Biol.125:795.

26. Mukherjee, S., T. T. Soe, and F. R. Maxfield. 1999. Endocytic sorting of lipidanalogues differing solely in the chemistry of their hydrophobic tails. J. Cell Biol.144:1271.

27. Harder, T., and M. Kuhn. 2000. Selective accumulation of raft-associated mem-brane protein LAT in T cell receptor signaling assemblies. J. Cell Biol. 151:199.

28. Janes, P. W., S. C. Ley, and A. I. Magee. 1999. Aggregation of lipid rafts ac-companies signaling via the T cell antigen receptor. J. Cell Biol. 147:447.

29. Grakoui, A., S. K. Bromley, C. Sumen, M. M. Davis, A. S. Shaw, P. M. Allen,and M. L. Dustin. 1999. The immunological synapse: a molecular machine con-trolling T cell activation. Science 285:221.

30. Lee, K. H., A. D. Holdorf, M. L. Dustin, A. C. Chan, P. M. Allen, and A. S. Shaw.2002. T cell receptor signaling precedes immunological synapse formation. Sci-ence 295:1539.

31. Xavier, R., T. Brennan, Q. Li, C. McCormack, and B. Seed. 1998. Membranecompartmentation is required for efficient T cell activation. Immunity 8:723.

32. Montixi, C., C. Langlet, A. M. Bernard, J. Thimonier, C. Dubois, M. A. Wurbel,J. P. Chauvin, M. Pierres, and H. T. He. 1998. Engagement of T cell receptortriggers its recruitment to low-density detergent-insoluble membrane domains.EMBO J. 17:5334.

33. Wulfing, C., and M. M. Davis. 1998. A receptor/cytoskeletal movement triggeredby costimulation during T cell activation. Science 282:2266.

34. Wulfing, C., M. D. Sjaastad, and M. M. Davis. 1998. Visualizing the dynamicsof T cell activation: intracellular adhesion molecule 1 migrates rapidly to the Tcell/B cell interface and acts to sustain calcium levels. Proc. Natl. Acad. Sci. USA95:6302.

35. Wulfing, C., C. Sumen, M. D. Sjaastad, L. C. Wu, M. L. Dustin, andM. M. Davis. 2002. Costimulation and endogenous MHC ligands contribute to Tcell recognition. Nat. Immun. 3:42.

36. Barisas, B. G., W. F. Wade, T. M. Jovin, D. Arndt-Jovin, and D. A. Roess. 1999.Dynamics of molecules involved in antigen presentation: effects of fixation. Mol.Immunol. 36:701.

37. Boes, M., J. Cerny, R. Massol, M. Op den Brouw, T. Kirchhausen, J. Chen, andH. L. Ploegh. 2002. T-cell engagement of dendritic cells rapidly rearranges MHCclass II transport. Nature 418:983.

38. Chow, A., D. Toomre, W. Garrett, and I. Mellman. 2002. Dendritic cell matu-ration triggers retrograde MHC class II transport from lysosomes to the plasmamembrane. Nature 418:988.

39. Varma, R., and S. Mayor. 1998. GPI-anchored proteins are organized in submi-cron domains at the cell surface. Nature 394:798.

40. Harder, T., and K. Simons. 1997. Caveolae, DIGs, and the dynamics of sphin-golipid-cholesterol microdomains. Curr. Opin. Cell Biol. 9:534.

41. Kenworthy, A. K., N. Petranova, and M. Edidin. 2000. High-resolution FRETmicroscopy of cholera toxin B-subunit and GPI-anchored proteins in cell plasmamembranes. Mol. Biol. Cell 11:1645.

42. Monks, C. R., B. A. Freiberg, H. Kupfer, N. Sciaky, and A. Kupfer. 1998. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature395:82.

43. Balamuth, F., D. Leitenberg, J. Unternaehrer, I. Mellman, and K. Bottomly. 2001.Distinct patterns of membrane microdomain partitioning in Th1 and Th2 cells.Immunity 15:729.

44. Krummel, M. F., M. D. Sjaastad, C. Wulfing, and M. M. Davis. 2000. Differentialclustering of CD4 and CD3� during T cell recognition. Science 289:1349.

1338 CLASS II/RAFT ACCUMULATION AT THE IMMUNE SYNAPSE

by guest on March 26, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from