The Cell Surface Expression of SAP-binding Receptor CD229 Is Regulated via Its Interaction with Clathrin-associated Adaptor Complex 2 (AP-2)*

CD229 (Ly9) is a cell surface receptor selectively expressed on T and B lymphocytes, and it belongs to the CD150 receptor family. Like other receptors of this family, CD229 interacts with SAP/SH2D1a protein, mutation of which is responsible for the fatal X-linked lymphoproliferative disease. Receptors of the CD150 family function as costimulatory molecules, regulating cytokine production and cytotoxicity. Thus, their signaling and regulation in lymphocytes may be critical to an under-standing of the pathogenesis of the X-linked lymphoproliferative disease. Here we show that CD229 interacts with the (cid:1) 2 chain of the AP-2 adaptor complex that links transmembrane proteins to clathrin-coated pits. CD229 was the only member of the CD150 family associated with AP-2. We also show that the (cid:1) 2 chain interacts with the Y 470 EKL motif of CD229. The integrity of this site was necessary for CD229 internalization, but it was not involved in SAP recruitment. Moreover, CD229 binds to the AP-2 complex in T and B cell lines, and it is internalized rapidly from the cell surface on T cells after antibody ligation. In contrast, cross-linking of CD229 receptors with intact antibody inhibited CD229 internalization on B cells. However, when F(ab (cid:1) ) 2

CD229 (also termed Ly9) is a cell surface glycoprotein of relative mass 120 kDa found on T and B cells (1). It is a member of the immunoglobulin superfamily, and it belongs to the CD150 family of receptors (2,3). Six members of the CD150 family (CD150, CD229, CD84, CD244, NTB-A, and CS1) have one or more cytoplasmic tyrosine motifs with the consensus sequence Thr-(Ile/Val)-Tyr(P)-X-X-(Ile/Val). This motif functions as a docking site for the SAP 1 /SH2D1a protein (4 -9). Mutations on the SAP gene, sh2d1a, are the responsible for the fatal X-linked lymphoproliferative disease after Epstein-Barr virus infection, familial hemophagocytic lymphohistiocytosis, and some cases of combined variable immunodeficiency (2, 4, 10 -12). These disease phenotypes point to distinct roles for the six CD150 family/SAP signaling pathways in the control of T, B, and natural killer cell activation and homeostasis. Studies on SAP knock-out mice reveal that T cells have an impaired ability to differentiate into T helper 2 cells, resulting in increased resistance to T helper 2-mediated disease such as infection with Leishmania major. In addition, SAP Ϫ/Ϫ mice fail to resolve the lymphocytic choriomeningitis virus, showing an increased number of interferon-␥-producing cells in the spleen and liver (13) and a nearly complete absence of virus-specific long lived plasma cells and memory B cells (14). Altogether, these data indicate that SAP controls several distinct key T cell signal transduction pathways, which may be controlled differentially by the various CD150 receptors. It is crucial to elucidate the contribution of each member to these immune functions. Unlike other CD150 family members, CD229 has four rather than two extracellular Ig domains. Domains 1 and 3 are similar, as are domains 2 and 4, suggesting that CD229 arose from a progenitor with two domains (15). Although the function of CD229 is unknown, other members of the subfamily such as CD84, CD244, and CD150 are known to activate lymphocytes (16 -22), suggesting that the CD229 molecule is also involved in leukocyte activation. Recently, NTB-A and CS1 have also been shown to regulate cytotoxicity in natural killer cells (23,24).
In this study, we have investigated the biochemical mechanism involved in CD229 signaling and intracellular trafficking by identifying proteins that interact with the CD229 cytoplasmic domain. Two clones encoding human AP-2 2 chain were isolated using CD229 as bait in a three-hybrid screen of a human B cell library. 2 , also termed AP50, is the 50-kDa subunit of the AP-2 complex. AP-2 is a key component of the endocytic machinery that links cargo membrane proteins to the clathrin lattice. It is a heterotetramer consisting of two large (100 -115 kDa) ␣ and ␤ subunits or adaptins, one medium 2 (50 kDa), and one small (17 kDa) 2 subunit. The ␣-adaptin appears to contain the major membrane binding interface. The hinge domain of the ␤ subunit binds to the clathrin heavy chain, providing a mechanism for the formation of polyhedral lattices. In addition to its structural role in coat assembly, AP-2 is also involved in the transport of proteins from the plasma membrane to the endosomal and lysosomal compartments. For effective cargo into coated pits, receptors should contain specific internalization signals, namely tyrosine-based motifs such as NPXY or YXX (where X stands for any amino acid and for a bulky hydrophobic residue), dileucine motifs, and acidic clusters (25,26). The structural analysis of the AP-2 core reveals that the binding site for YXX endocytic motifs is buried, indicating that a conformational change, probably triggered by phosphorylation in the disordered 2 linker, is necessary to allow YXX motif binding (27). Here, we demonstrate that the only member of the CD150 family which specifically binds 2 is CD229. We mapped the CD229-2 interaction in Tyr 470 of the CD229 cytoplasmic domain. Moreover, we show that TCR and BCR signaling regulates CD229 expression.
Anti-human CD229 (clone HLy-9.1.25) was generated at our laboratory (1). Biotinylated anti-CD229 was generated using EZ-Link biotinylation kit from Pierce. F(abЈ) 2 anti-CD229 was prepared using immobilized ficin from Pierce. The purity of the F(abЈ) 2 preparation was confirmed by SDS-PAGE under nonreducing conditions and reducing conditions followed by silver staining.
Yeast Three-hybrid Screen-The cytoplasmic tail of human CD229, cloned in pBridge and transformed in the yeast strain CG1945, was used as bait to screen a human B cell cDNA library in pACT2 vector (Clontech, Palo Alto, CA). CD229 and mutated Fyn were inserted in the bicistronic vector pBridge. In mutant Fyn, the regulatory tyrosines 420 and 531 were replaced by phenylalanine, and arginine was replaced by glutamine (Fyn Y420F,Y531F,R176Q ) as described elsewhere (7).
For the three-hybrid screen in the presence of Fyn Y420F,Y531F,R176Q, the yeast strain CG1945 was cotransformed sequentially with the vector (pBridge), containing both CD229 and the Fyn mutant, and these transformants were selected on SD medium lacking tryptophan and transformed further with 1 mg of the human B cell cDNA library. Double transformants were then plated in SD medium lacking tryptophan, leucine, histidine, and methionine in the presence of 3 mM 3-aminotriazole.
Yeast clones that grew in these restrictive conditions were then tested by the ␤-galactosidase assay. DNA was extracted from the ␤-galactosidase-positive clones and then expanded after transformation of HB101 bacteria and selection in M9 medium, lacking leucine, to isolate the GAL4 activation domain plasmid pACT2. Purified plasmids were sequenced.
The ␤-galactosidase colony-lift filter assay and liquid culture assay using o-nitrophenyl ␤-D-galactopyranoside as a substrate were carried out as described in the Clontech yeast protocols handbook. The vector pBridge containing CD229, in the absence of catalytic activity of mutated Fyn, was generated after BglII digestion, ligation, and purification.
To generate 2 -EGFP fusion protein, 2 was amplified by PCR using a cDNA peripheral blood lymphocyte library as a template with sense primer 5Ј-AAACTGCAGGCCGCCATGATTGGAGGCTTATTC-3Ј and antisense primer 5Ј-TTCCCCGGGGCGAGTTTCATAAATGCCACTGC-GG-3Ј and cloned in-frame into the pEGFP-N2 vector (Clontech) using the PstI/SmaI sites. The presence of desired nucleotide changes was established by bidirectional nucleotide sequencing using dye terminator chemistry (PE Biosystems, Foster City, CA).
Transfections-COS-7 cells (1 ϫ 10 6 ) were seeded into 100-mm dishes and 24 h later transfected with 4 g of each plasmid DNA/dish using the LipofectAMINE Plus kit (Roche Applied Science) following the manufacturer's instructions. Cells were incubated for 48 h, and the experiments were performed.
Immunoprecipitation and Immunoblotting-Transfected COS-7 cells were biotinylated and lysed with 1% CHAPS (Pierce), and immunoprecipitation was carried out using antibodies against CD229. Likewise, previously biotinylated Jurkat and Daudi cells (100 ϫ 10 6 ) were lysed with 1% Triton X-100 as described elsewhere (7). Briefly, cell lysates were precleared for 1 h with 50 l of protein A-Sepharose beads (Amersham Biosciences). Immunoprecipitations were performed using 2 g of each indicated antibody, and 40 l of protein A-Sepharose beads was added for 2 h at 4°C. Beads were then washed, and samples were boiled. All of the precipitates were separated by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes (Millipore, Bedford, MA). Filters were blocked for 1 h with 5% skim milk (or 3% bovine serum albumin) and then probed with the indicated antibodies. Bound

FIG. 1. CD229 interacts with 2 in a three-hybrid screen in yeast.
A, the sequence obtained in the two positive clones overlaps (in bold) with the amino acidic sequence of the 2 (AP50) subunit of the human AP-2 complex. B, ␤-galactosidase semiquantitative assay after cotransfection of pBridge CD229 (ϩFyn) or (ϪFyn) and clone 1.
antibody was detected using horseradish peroxide-conjugated secondary antibodies and enhanced chemiluminescence (Pierce).
Internalization Assay-Cells (1 ϫ 10 6 /sample) were incubated with anti-human CD229 mAb (1 g/ml ϭ 1 ϫ 10 6 ) at 37°C. At various time points (0 -90 min), cells were chilled rapidly and acid stripped to remove cell surface-bound mAb as described elsewhere (28). Samples were washed twice in cold complete culture medium, and they were incubated for 25 min with biotinylated anti-CD229 followed by avidin-PE incubation (1). For recycling experiments, cells were incubated at 37°C with anti-CD229 in the presence or absence of BFA at 10 g/ml in complete culture medium for the indicated time periods. Samples were stripped and stained with biotin-conjugated anti-human CD229 followed by streptavidin-PE. For stimulation, cells were treated for 5 min with anti-CD3 or anti-IgM. Intact anti-CD229 or F(abЈ) 2 was added for an additional 30 min. Thereafter, cells were harvested, chilled, and acid stripped as described above. The samples were analyzed in a flow cytometer (FACS Caliber; BD Biosciences, Mountain View, CA) to detect PE (FL2) fluorescence (mean fluorescence intensity; MFI). The percentage of endocytosis was calculated using MFI values of PE as follows: 100 Ϫ [(MFI at 37°C/MFI at 4°C) ϫ 100].
Immunofluorescence Microscopy-COS-7 cells were cultured on coverslips, transfected with 2 -EGFP cDNA, fixed, and permeabilized with Ϫ20°C methanol for 15 min. After two washes, they were incubated for 30 min at room temperature with blocking buffer (phosphatebuffered saline containing 0.2% skim milk, 2% fetal bovine serum, 1% bovine serum albumin, 0.1 mM glycine). Cells were then washed and incubated with 2 g/ml rabbit anti-rat AP-2␣-C for 1 h at 4°C. After three washes, they were incubated with a biotinylated anti-rabbit (Biogenex, San Ramon, CA) at 4°C for 30 min. After several additional washes, cells were incubated with streptavidin-Cy3 (Jackson Immunoresearch, West Grove, PA) at 4°C for 20 min. Samples were washed and mounted in Fluoromount-G (Southern Biotechnology, Birmingham, AL). For endocytosis experiments, COS-7 cells were cultured on coverslips and cotransfected with human CD229 (WT or Y470F or Y470A mutants) with 2 -EGFP cDNAs in duplicate. After 48 h they were labeled with 1 g/ml biotinylated mAb CD229 at 4°C for 30 min. After two washes with ice-cold phosphate-buffered saline, samples were incubated with streptavidin-Cy3 at 37°C (to induce endocytosis) for 30 min. Cells were then washed and fixed in Ϫ20°C methanol for 15 min. After washing them twice, cells were mounted and visualized as described above. Fluorescence images were obtained using a confocal microscope (TCS NT; Leica, Heidelberg, Germany).

CD229 Interacts with 2 in a Yeast Three-hybrid Screen-
The cytoplasmic tail of human CD229 was used as bait to screen a human B cell library. The CD229 cytoplasmic tail was cloned in the pBridge vector together with Fyn (7). The screening of ϳ2 ϫ 10 6 clones yielded two clones that encoded portions of human 2 , the medium chain of clathrin-associated coated pit adaptor protein complex AP-2 (Fig. 1A). Because the 2 interaction with proteins is independent of phosphorylation, we assayed the binding by depleting the catalytic activity of Fyn. CD229 binding to 2 was similar in the presence and absence of kinase activity (Fig. 1B).
CD229 Is the Only Member of the CD150 Family of Receptors That Interacts with 2 -We assayed the interaction of 2 with all of the CD150 members using a semiquantitative ␤-galactosidase assay. All bound to SAP/SH2D1a when Fyn was active, with the exception of CD150, which bound to SAP regardless of its phosphorylation status, as reported elsewhere (29) (Fig. 2A). Surprisingly, only CD229 was able to bind to 2 in conditions where Fyn was present or absent, whereas the rest of members failed to do so (Fig. 2B).
CD229 Binds to 2 -EGFP Chimera and Integrates in AP-2 Endogenous Complex in COS-7 Cells-To test the interaction CD229-2 in a mammal system, we made the 2 -EGFP construct and transiently transfected COS-7 cells alone or together with CD229. To determine whether CD229 and the 2 -EGFP chimera integrated in the endogenous AP-2 complex in COS-7 cells, we performed an immunoprecipitation with anti-␣ chain of AP-2 (Fig. 3A). 2 -EGFP and CD229 coprecipitated with AP-2 endogenous complex (second and third lanes, respectively). Moreover, CD229⅐ 2 -EGFP and AP-2 were in a complex (fourth lane). To test whether most 2 -EGFP was free or complexed with the endogenous AP-2, we performed 2 -EGFP transfection in COS-7 cells and immunochemical staining with anti-␣ chain AP-2. The majority of the 2 -EGFP protein colocalized with the endogenous AP-2 (Fig. 3B).
Tyrosine 470 Is Required for CD229 Binding to 2 -Analysis of the CD229 sequence revealed several putative motifs for 2 binding (Y-X-X-) (where X stands for any amino acid and for a bulky hydrophobic residue) (Fig. 4A). A more restricted binding motif has been proposed for 2 : YPP, where P represents a polar amino acid (25). Following these criteria we found a suitable motif for the binding (Fig. 4A). The Y470F mutation abolished the interaction between CD229 and 2 in a yeast system (Fig. 4B). COS cells were transfected with CD229 and 2 -EGFP ϩ/Ϫ Fyn. Samples were biotinylated and divided. One set was immunoprecipitated with anti-CD229 and the second with anti-GFP. CD229 was found to interact with 2 in the absence of phosphorylation, although some binding was found in the triple transfection, maybe because of residual nonphosphorylated CD229 in this sample (Fig. 4C). Moreover, only nonphosphorylated CD229 was found to bind 2 (anti-GFP immunoprecipitation) in the samples where Fyn was present. Altogether, these data indicate that CD229 interacts with 2 when the receptor is not phosphorylated, and CD229 phosphorylation by Fyn abrogates the binding of the CD229 to the 2 subunit. Tyrosine 470 Is Essential for CD229 Endocytosis-To determine the relevance of CD229-2 interaction for receptor internalization, we generated CD229 mutant constructs in which the Tyr 470 residue was replaced by phenylalanine or alanine. We transiently transfected CD229 wild-type and mutants encoding constructs into COS-7 and measured their internalization after ligation with anti-CD229 mAb at 37°C. Thereafter, cells were stripped and stained with CD229 biotinylated anti- FIG. 4. CD229 Y470 is critical for 2 binding. A, CD229 cytoplasmic tail showing binding sites for the 2 subunit of the AP-2 (underlined). B, ␤-galactosidase assay measuring the interaction between pBridge CD229 WT or pBridge CD229 Y470F with pACT2-2 in yeast cells in the presence or absence of Fyn activity. The binding of CD229 WT and CD229 Y470F to SAP was also performed. These data are the mean of duplicates and are representative of three experiments. C, CD229 was cotransfected with 2 ϩ/Ϫ Fyn in COS cells. Samples were divided and immunoprecipitated (Ip) with anti-GFP or anti-CD229 antibody. Avidin-peroxidase, anti-Tyr(P), and anti-GFP blots were performed.

FIG. 5. CD229 internalization in COS cells transfected with CD229 Y470 mutants is impaired.
A, internalization was assessed 30 min after incubation with anti-CD229 at 37°C and was measured by flow cytometry. These data are the mean of three experiments. B, endocytosis was carried out in CD229 WT and CD229 mutants and 2 -EGFP-transfected COS cells. Cells were stained with biotinylated anti-CD229, and avidin-Cy3 was added at 37°C for 30 min, inducing endocytosis of the receptor. Cells were washed and mounted. body plus avidin-PE labeling, and samples were measured by flow cytometry (Fig. 5A). Within 30 min, 30 -40% of wild-type CD229 surface receptors had been internalized. In contrast, internalization was completely abrogated, and CD229 surface expression even increased slightly, in CD229 mutants, which Tyr 470 had been replaced by phenylalanine or alanine. We also transfected COS cells with the CD229 WT and CD229 Y470 mutants plus 2 -EGFP. After 24 h, cells were stained with biotinylated antibody at 4°C followed by avidin-Cy3 at 37°C (Fig.  5B). In cells transfected with CD229 WT , the receptor was internalized and colocalized with the 2 -EGFP, whereas CD229 Y470F and CD229 Y470A internalization was impaired, and no colocalization with 2 was found. Both results indicate that Tyr 470 is needed for CD229 binding to 2 and further endocytosis.
The Cytoplasmic Tail of CD229 Is Associated with the AP-2 Adaptor Complex in T and B Lymphocytes-Because CD229 is expressed only in cells of the immune system, we next determined whether the association of the AP-2 adaptor complex with CD229 occurred in lymphocytes. To this end, we performed coimmunoprecipitation experiments in two lymphocytic cell lines: Jurkat (T cell line) and Daudi (B cell line). Both cell lines were immunoprecipitated using anti-CD229 mAb and an anti-CD84 as a control mAb. Western blot was probed with anti-AP-2␣ mAb (Fig. 6). A band representing AP-2␣ was found only in the lane in which CD229 was immunoprecipitated. Taken together, these data demonstrate that the AP-2 adaptor complex specifically interacts with the cytoplasmic tail of CD229 in T and B lymphocytes.
Anti-CD229 mAbs Trigger CD229 Internalization-Anti-CD229 mAbs induce the internalization of CD229 in COS cells after 30 min of mAb coligation. Because CD229 is expressed on lymphocytes and associates with the AP-2 complex, we assayed CD229 internalization in these cells. Endocytosis was assayed in the Jurkat T cell line and in the Daudi B cell line. CD229 was endocytosed rapidly in Jurkat cells, reaching a plateau within 15 min. The internalization of CD229 after incubation with intact anti-CD229 was comparable with that obtained with F(abЈ) 2 in Jurkat cells (Fig. 7A). In contrast, internalization of CD229 with intact mAb was impaired in B cells. However, when F(abЈ) 2 was used to coligate, the effect on CD229 internalization resembled that observed in T cells, indicating that CD229 endocytosis may be regulated through Fc␥ receptors (Fig. 7B). In agreement with this observation, a reduction of intact antibody concentration used for ligation induced some CD229 endocytosis in B cells, indicating that a limiting mAb concentration elicits coligation with CD229 with minimal Fc␥ receptor binding (data not shown). The internalization of CD229 after antibody coligation may represent a constitutive turnover rate of CD229 from the cell surface. To distinguish these two events (endocytosis versus turnover), we used cycloheximide and BFA, which block the expression of newly synthesized protein. BFA blockage inhibits the transport of proteins from the endoplasmic reticulum to the Golgi apparatus and also recycling of endocytosed proteins (30 -32). Treatment with cycloheximide did not alter CD229 surface expression on T or B cells within the time period assayed for endocytosis (data not shown). Treatment with BFA for 2 h slightly reduced CD229 expression in T cells after 60 min (20 -25%) but did not significantly reduce the surface expression of CD229 in B cells (Fig. 7C). The transferrin receptor (CD71), which is rapidly recycled (33), was used as a control. After 15 min, CD71 surface expression was reduced more than 50% in BFA-treated cells (Fig. 7D). Altogether, these results indicate that CD229 expression on the cell surface is more stable on B than on T cells, and in both cases, the rate of turnover is lower than that for CD71.
Endocytosed CD229 Mainly Follows the Lysosomal Degradation Pathway-We next analyzed the events following CD229 ligation in lymphocytes. CD229, once internalized, may traffic to lysosomal vesicles for degradation or be recycled back to the cell surface. To distinguish these two events we assessed CD229 endocytosis in the presence or absence of BFA. If most CD229 were recycled back to the cell surface, the loss of CD229 after mAb ligation would be much greater in the presence of BFA. On the other hand, if most endocytosed CD229 were transferred to lysosomes, we would obtain identical results in

AP-2-mediated CD229 Endocytosis
the presence or absence of BFA. CD229 surface levels decreased slightly on T cells induced by F(abЈ) 2 anti-CD229 in the presence of BFA, suggesting that CD229 was partially recycled back to the cell surface (20 -25%) but mostly trafficked to lysosomal degradation (Fig. 8A). In B cells, CD229 surface levels were similar in both cases. Thus, endocytosed CD229 follows mainly the lysosomal degradation pathway in these cells (Fig. 8B). Taken together, these results suggest that most CD229 is transported to the lysosomal compartment rather than recycled back to the cell surface.
CD229 Internalization Is Regulated after TCR or BCR Crosslinking-We next investigated whether antigen receptor crosslinking modulates CD229 internalization in T and B lymphocytes. When we assayed CD229 expression after TCR ligation at 30 -60 min, the expression of CD229 did not vary significantly. In contrast, CD229 ligation did not affect CD3 expression in this period (data not shown). Unexpectedly, the coligation of CD229 and TCR enhanced CD229 internalization (Fig.  9A). Daudi cells were incubated with anti-IgM F(abЈ) 2 and anti-CD229 for 30 min, and CD229 endocytosis was studied. CD229 internalization after IgM coligation was enhanced (Fig. 9B).
Taken together, these data suggest that signals delivered by the TCR and BCR regulate CD229 cell surface expression. DISCUSSION In this study, we have shown that CD229 binds to the 2 chain of the AP-2 complex. It is well established that the 2 subunit binds to the consensus sorting signal NPXY or YPP (25). Although it has been suggested that 2 interacts with dileucine motifs (34), peptides containing dileucine motifs bind only to the ␤ subunit of AP1 and AP-2 complexes and not to the other subunits (35,36). Analysis of CD229 revealed only one sequence that follows the YPP consensus (Y 470 EKL), which is conserved in mouse (37). Using site-directed mutagenesis, we mapped the 2 interaction to CD229 in this motif. The single replacement of tyrosine by phenylalanine abrogates 2 binding to CD229 in yeast. It has been proposed that the hydroxyl group of tyrosine contributes to a network of interactions with 2 residues that form the hydrophobic pocket, explaining why phenylalanine does not efficiently replace tyrosine in YPP signals (38,39). As expected, the recruitment of SAP in the mutant is not altered because both molecules do not compete for the same motif (7). CD229 binds to 2 when Fyn is active, or kinase activity of Fyn is completely abrogated by depletion of its catalytic region in the yeast system. This suggests various possibilities: CD229-2 interaction occurs when Tyr 470 is phosphorylated, or phosphorylation of the CD229 tail is partial, allowing the screening of interactions dependent and independent of phosphorylation at the same time. Our results point to this last possibility because the CD229-2 interaction only occurred when the receptor was not phosphorylated. This agrees with literature data, which show that this type of interactions does not depend on phosphorylation. The crystal structure of 2 indicates that phosphotyrosine does not fit into its hydrophobic pocket (40).
We have also demonstrated that the cytoplasmic tail of CD229 but not that of any of the other members of CD150 family can interact with the AP-2 complex via its 2 chain. This intriguing difference suggests that the expression and internalization of CD229 are tightly and differentially regulated. The structure of CD229 differs from that of the rest of the CD150 family of receptors. It is the only member that has four extracellular Ig-like domains, and it has the longest cytoplasmic tail, comprising 180 amino acids, 16 threonines, 17 serines, and 8 tyrosines. Two unique tyrosine-based motifs (with the consensus amino acid sequence TV/I YXXV/I) are critical for the binding to the SAP/SH2D1a protein and the Eat-2 SH2 domain protein homolog to SAP in B cells (41).
Furthermore, CD229 can bind to endogenous AP-2 in COS cells, 2 -EGFP chimera can efficiently integrate in the endogenous AP-2 complex, and CD229 can coprecipitate with 2 -EGFP and the AP-2 endogenous complex. The complex is functional because CD229 internalization and CD229⅐ 2 -EGFP colocalization were detected in COS cells after antibody-induced endocytosis. In agreement with our mapping data, CD229 Y470F and CD229 Y470A endocytosis was severely impaired. Although in our model a phenylalanine substitution abrogates the binding to 2 , the FXX motif can fit into the hydrophobic pocket of 2 (42) but leads to a weaker interaction and slower endocytosis. Thus, Tyr 470 is essential for CD229 endocytosis and may inhibit the internalization of this receptor after its phosphorylation in vivo. CD229 can be phosphorylated by Src-kinase Fyn in COS cells, and its possible interaction with a SH2 domain protein may further regulate the signaling and trafficking of the molecule, as proposed for CTLA-4 (43).
Because CD229 is expressed in T and B lymphocytes, we have defined the mechanism of CD229 internalization and demonstrated that it can be regulated differentially in T and B cell lines. CD229 was endocytosed rapidly in response to antibody ligation in Jurkat T cells, whereas no internalization was observed in B cells. The same results were obtained in lymphocytes from peripheral blood. CD229 in B cells was inhibited by the Fc␥ receptor coligation because F(abЈ) 2 antibodies induce CD229 endocytosis in B cells. The CD229 uptake rate was similar using the intact and F(abЈ) 2 antibodies in Jurkat cells lacking Fc␥ receptors. In agreement with our results, the coligation of Fc␥ receptors and BCR has been suggested to inhibit endocytosis by blocking the processing and presentation of the BCR-bound antigen (44). Using BFA and cycloheximide, we have determined that CD229, unlike transferrin receptor (CD71), exhibits a low constitutive rate of turnover, and therefore, the rapid internalization of CD229 represents ligandinduced endocytosis. After endocytosis, CD229 was segregated into two distinguishable pools, one that is shuttled to a nonrecycling degradation pathway and another that is recycled to the cell surface. The results suggest that ϳ75-80% of CD229 in T cells become degraded via lysosomal compartments, and the remaining 20 -25% of complexes are recycled to the cell surface.
In contrast, most CD229 is trafficked to the lysosomal degradation pathway in B cells.
Moreover, TCR and anti-IgM coligation enhances CD229 endocytosis, suggesting that the antigen receptor signaling regulates the availability of CD229 at the plasma membrane. Thus, endocytosis may represent a critical mechanism that modulates CD229 cell surface expression rapidly. After TCR activation, phosphatidylinositol 3-kinase may become activated and catalyze the conversion of phosphatidylinositol 4,5biphosphate to phosphatidylinositol 3,4,5-trisphosphate. The pleckstrin homology domain in the ␣ 2 subunit of the AP-2 adaptor complex binds with high affinity to phosphatidylinositol 3,4,5-trisphosphate. This binding mediates two effects: localization of AP-2 in the membrane and increase in the affinity of 2 binding to endocytosis sorting signals (45,46). Because CD229 endocytosis is impaired when Fc␥ receptors are triggered and enhanced when the BCR is activated, this molecule may have a differential role in these different events. The ability to regulate CD229 internalization in T and B cells differentially may allow for different net effects of CD229 signaling which may be cell-specific. Recent studies indicate that CD229, unlike CD150 and CD84, inhibits interferon-␥ secretion after TCR activation in T cells. 2 TCR and BCR may increase the rate of CD229 endocytosis to prevent CD229 from delivering negative signals to the cell and Fc␥ receptor ligation retains the CD229 on the cell surface, thus favoring signaling through this receptor and the maintenance of an inhibitory response. The development of CD229 knock-out mice may be a key step to define the physiological role of CD229.