Monomeric and Oligomeric Complexes of the B Cell Antigen Receptor

Immunity, Vol. 13, 5–14, July, 2000, Copyright 2000 by Cell Press

Monomeric and Oligomeric Complexes of the B Cell Antigen Receptor Wolfgang W. A. Schamel and Michael Reth* Department of Molecular Immunology Biology III University of Freiburg and Max Planck Institute for Immunobiology Stu¨beweg 51 79108 Freiburg Germany

Summary The current structural model of the B cell antigen receptor (BCR) describes it as a symmetric protein complex in which one membrane-bound immunoglobulin molecule (mIg) is noncovalently bound on each side by an Ig-␣/Ig-␤ heterodimer. Using peptide-tagged Ig-␣ proteins, blue native polyacrylamide gel electrophoresis (BN–PAGE), and biosynthetical labeling of B cells, we find that the mIg:Ig-␣/Ig-␤ complex has a stoichiometry of 1:1 and not 1:2. An anti-Flag stimulation of B cells coexpressing Flag-tagged and wild-type Ig-␣ proteins results in the phosphorylation of both Ig-␣ proteins, suggesting that on the surface of living B cells, several BCR monomers are in contact with each other. A BN–PAGE analysis after limited detergent lysis provides further evidence for an oligomeric BCR structure. Introduction The B cell antigen receptor (BCR) plays a central role in the development, survival, and activation of B lymphocytes. The BCR is a multiprotein complex consisting of the membrane-bound immunoglobulin (mIg) molecule and the Ig-␣/Ig-␤ heterodimer (CD79a,b). Antigen is bound by the variable Ig domains of the mIg heavy and light chains (HC, LC), whereas coupling of the receptor to intracellular signaling proteins is achieved by the Ig␣/Ig-␤ heterodimer (for review see DeFranco, 1997; Kurosaki, 1997; Reth and Wienands, 1997; Benschop and Cambier, 1999). The Ig-␣ and Ig-␤ molecules are disulfide-bonded type I transmembrane proteins. They carry an extracellular Ig domain, a transmembrane (TM) region with three polar amino acids, and a cytoplasmic tail including an immunoreceptor tyrosine-based activation motif (ITAM) (Reth, 1989). Apart from glycosylation differences in the Ig-␣ protein, all mIg classes are noncovalently coupled with the same Ig-␣/Ig-␤ heterodimer (Campbell et al., 1991; Venkitaraman et al., 1991; Wienands and Reth, 1991). The membrane-proximal CH domain of the mIg molecule is required for stable binding to the Ig-␣/Ig-␤ heterodimer (Hombach et al., 1990; Reth et al., 1991). In addition, the TM regions of the mIg molecule and the Ig-␣/Ig-␤ heterodimer are critically involved in BCR complex formation (Shaw et al., 1990; Grupp et al., 1993; Sanchez et al., 1993; Campbell et * To whom correspondence should be addressed (e-mail: reth@ immunbio.mpg.de).

al., 1994; Stevens et al., 1994). The 25 amino acid–long TM region of the mIg HC most likely crosses the membrane as an ␣ helix. This sequence contains several polar amino acids (nine in the case of mIgM and seven in the case of mIgD), which are distributed on all sides of the ␣ helix. One side of the helix is conserved between the different mIg classes (Reth, 1992; Campbell et al., 1994) and in the case of mIgM has been shown to bind to Ig-␣/Ig-␤ (Shaw et al., 1990; Grupp et al., 1993; Sanchez et al., 1993; Stevens et al., 1994). Mature B lymphocytes coexpress an IgD-BCR and an IgM-BCR on their surface. Treatment of these cells with anti-␦ antibodies induces phosphorylation of only the mIgD-associated Ig-␣, whereas anti-␮ antibodies induce phosphorylation of the mIgM-associated Ig-␣ (Gold et al., 1991). The two different BCR classes can transmit distinct signals (Ales et al., 1988; Chen et al., 1990; Kim and Reth, 1995). The expression of the BCR is required for the positive selection of immature B cells and the maintenance or survival of mature B cells (Lam et al., 1997; Neuberger, 1997). This may indicate that already in the absence of an antigen, the BCR transmits a signal. The nature of this selection or maintenance signal is not known at present but may involve a preformed BCR signaling complex (Wienands et al., 1996). Upon antigen binding, the survival signal is transformed into an activation signal. An early step in signal transduction is the activation of protein tyrosine kinases that phosphorylate several substrate proteins including the Ig-␣/Ig-␤ heterodimer (DeFranco, 1997; Kurosaki, 1997; Reth and Wienands, 1997; Benschop and Cambier, 1999; Wienands, 1999). The phosphorylated ITAMs of Ig-␣ and Ig-␤ serve as docking sites for SH2 domain– containing proteins like the tyrosine kinase Syk. For BCR activation, multivalent antigens or anti-BCR antibodies are required (Dintzis et al., 1982, 1989). It was therefore proposed that upon antigen binding, several monomeric receptors are brought into close proximity, thereby allowing cross-wise phosphorylation of BCR-associated kinases. For a better understanding of the signaling function of the BCR, it is important to learn more about its structure. The mIg molecule is a symmetric molecule (two HC and two LC), and we previously suggested that it is noncovalently bound on each side by an Ig-␣/Ig-␤ heterodimer (Hombach et al., 1990). However, this 1:2 stoichiometry of the mIg:Ig-␣/Ig-␤ complex has never been experimentally tested. Here we show that the BCR isolated from 1% digitonin lysates is a stable complex of one mIg molecule and only one Ig-␣/Ig-␤ heterodimer and thus has a 1:1 stoichiometry. In the presence of limiting detergent concentrations, however, we find higher molecular complexes, suggesting that on the surface of living B cells, the BCR has an oligomeric structure. Results BN–PAGE Analysis of the Digitonin-Solubilized BCR Complex We previously generated transfectants of J558L B cells expressing on their surface either an IgM-BCR or an

Immunity 6

dissociates into the mIg molecule and the Ig-␣/Ig-␤ heterodimer (Figure 1A, lanes 2, 4, 6, and 8). The composition of purified BCR complexes was analyzed by two-dimensional (BN–PAGE/reducing SDS–PAGE) gel electrophoresis (Figure 1B). Total proteins were visualized using the biotinylation-in-gel (BIG) protocol (Schamel, 1999). Apart from a minor protein band running above the ␦m HC, this analysis did not reveal any new proteins. Thus, the purified BCR complex consists of a HC (␮m or ␦m), a LC (␭), and Ig-␣ and Ig-␤.

Figure 1. BN Gel Analysis of the Size and Composition of AffinityPurified BCR Complexes (A) The IgM-BCR and IgD-BCR were size separated either in the absence (minus) or the presence (plus) of SDS by BN–PAGE and detected on the Western blot with either anti-␮/anti-␦ or anti-Ig-␣ antibodies. Size markers were dimeric and monomeric ferritin (880 and 440 kDa) and catalase (230 kDa). (B) Two-dimensional (BN–PAGE/reducing SDS–PAGE) gel analysis of the IgM-BCR and IgD-BCR components. In-gel biotinylated proteins were detected according to the BIG method, and the identity of these proteins was determined by immunoblotting. The fastest migrating protein is a degradation product of ␭, since it is stained with anti-␭ antibodies.

IgD-BCR with specificity for the hapten 3-nitro-4-hydroxyphenylacetate (NP) (Hombach et al., 1988; Kim et al., 1994). The BCR complexes were isolated from digitonin lysates of pervanadate-stimulated J558L Ig transfectants by a two-step affinity purification procedure. First, all tyrosine-phosphorylated proteins were purified by anti-phosphotyrosine-coupled agarose and eluted with phenylphosphate. Second, the BCR complex in the eluate was purified over NP-conjugated-Sepharose and eluted with the free hapten 5-iodo-NP (NIP). This procedure separates the BCR complex from the free mIg molecules in the lysate. The purified BCR complex was then size separated by BN–PAGE (Scha¨gger and Jagow, 1991; Scha¨gger et al., 1994). In contrast to SDS–PAGE, BN–PAGE does not disrupt transmembrane protein complexes and thus allows the direct analysis of different size forms of the BCR complex. However, the exact molecular weight of a protein complex cannot be determined by this method (Scha¨gger et al., 1994; Moro et al., 1999). In the BN gel, the purified IgM-BCR as well as the IgD-BCR migrates as a distinct protein complex detected by immunoblotting with both anti-Ig and antiIg-␣ antibodies (Figure 1A, lanes 1, 3, 5, and 7). In the presence of a stringent detergent like SDS, this complex

Composition of the Digitonin-Solubilized BCR Complex The mIg tetramer is a symmetric molecule and it was assumed that it binds the Ig-␣/Ig-␤ heterodimer symmetrically, forming an mIg:Ig-␣/Ig-␤ complex with a 1:2 stoichiometry (Figure 2A). To test this model, we coexpressed wild-type Ig-␣ and an Ig-␣ Flag-tagged at the C terminus (Ig-␣flC) in J558L␮m B cells. If the symmetric BCR model is correct, an anti-Flag antibody should coprecipitate the wild-type Ig-␣ from a purified BCR preparation. This should not be the case if the BCR complex contains only one Ig-␣/Ig-␤ heterodimer (Figure 2B). The IgM-BCR complex was affinity purified with NP-Sepharose from the digitonin lysates of two nonstimulated J558L␮m/Ig-␣/Ig-␣flC transfectants and analyzed by SDS–PAGE (Figure 2C). In the NIP eluates, both forms of Ig-␣ were present and could be distinguished by their sizes (Figure 2C, lanes 1 and 4). The purified BCR complexes were precipitated sequentially by anti-Flag and anti-␮ antibodies. The anti-Flag precipitate contained Ig-␣flC but not wild-type Ig-␣ (Figure 2C, lanes 2 and 5). The BCR complexes remaining in the supernatant of the anti-Flag precipitation were subsequently precipitated with anti-␮ antibodies and contained all the wildtype Ig-␣ (Figure 2C, lanes 3 and 6). The exclusive presence of Flag-tagged or wild-type Ig-␣ also holds true for the IgD-BCR complex (Figure 2D), although here the analysis was complicated by the two glycosylation forms of the mIgD-associated Ig-␣. Note that the IgDBCR precipitated with the anti-Flag antibody contains only the mature and not the immature glycosylation form of the ␦m heavy chain (Figure 2D, upper panel, lanes 2 and 5). We also analyzed J558L␮m B cells coexpressing wild-type and N-terminally flagged Ig-␣ (Ig-␣flN). Again, the results suggest that the detergent-solubilized BCR complex contains only one Ig-␣/Ig-␤ heterodimer (Figure 2E) and thus forms an asymmetric mIg:Ig-␣/Ig-␤ complex. It is possible that the above result is due to a selective loss of one Ig-␣/Ig-␤ heterodimer during the purification of a symmetric 1:2 BCR complex. To rule out this possibility, we analyzed the glycosylation form of Ig-␣flN molecules. The affinity-purified BCR contained only the mature glycosylation form of Ig-␣flN, which is partially resistant to endoglycosidase H (EndoH) digestion (Figure 2F, lanes 1 and 2) (Campbell et al., 1991; Pogue and Goodnow, 1994). The Ig-␣flN proteins remaining in the supernatant were all of the immature, EndoH-sensitive form (Figure 2F, lanes 5 and 6). The absence of a mature Ig-␣ molecule in the supernatant demonstrates that during the lysis and purification procedure the BCR complex is stable and does not lose an Ig-␣/Ig-␤ heterodimer.

Oligomeric BCR Complexes 7

Figure 3. Exclusive Presence of Only One Tagged Ig-␣ Protein in BCR Complexes Isolated from B Cells Coexpressing Ig-␣flN and Ig␣HAN

Figure 2. The Digitonin-Solubilized BCR Complex Contains Only One Ig-␣/Ig-␤ Heterodimer (A) Current structural model of a symmetric 1:2 mIg:Ig-␣/Ig-␤ complex. (B) Alternative model of a 1:1 mIg:Ig-␣/Ig-␤ complex. The location of the N- and C-terminal Flag tag in Ig-␣ is indicated. (C–E) SDS–PAGE and Western blot analysis of NP-specific BCR complexes purified from digitonin lysates of unstimulated J558L transfectants coexpressing wild-type and Flag-tagged Ig-␣. The BCR complexes were eluted from NP-Sepharose affinity columns with NIP-cap (lanes 1 and 4) and subsequently precipitated with anti-Flag (lanes 2 and 5). The remaining complexes in the Flagdepleted eluates were then precipitated using anti-HC antibodies (lanes 3 and 6). Proteins were detected with anti-HC (upper panel) or anti-Ig-␣ (lower panel) antisera. (F) The mIg molecule and the Ig-␣/Ig-␤ heterodimer do not dissociate during affinity purification of the BCR. Unstimulated J558L␮m/Ig␣flN cells were lysed with digitonin, and BCR complexes were purified using NP-Sepharose (lanes 1 and 2). NP affinity chromatography was repeated (lanes 3 and 4) and Ig-␣ molecules remaining in the supernatant were purified with an anti-Flag antibody (lanes 5 and 6). Proteins were incubated in the absence (minus) or presence (plus) of EndoH and detected with anti-HC (upper panel) or antiIg-␣ (lower panel) antisera.

To exclude the possibility that the restricted Ig-␣ usage is due to a heterogeneous expression of the two Ig-␣ proteins in the transfected B cell population, we coexpressed hemagglutinin (HA)-tagged (Ig-␣HAN) and Flag-tagged (Ig-␣flN) Ig-␣ in J558L␮m cells. A FACScan analysis shows that the majority of cells of the double transfectants coexpress Ig-␣flN and Ig-␣HAN molecules on their surface (Figure 3A). The BCR complexes of these transfectants were first affinity purified with NPSepharose and the proteins in the NIP eluate were further purified in parallel with anti-␮, anti-Flag, or anti-HA antibodies. The purified BCR complexes showed the exclusive presence of only one of the two different tagged Ig-␣ proteins (Figure 3B, lanes 8, 9, 11, and 12).

(A) FACScan analysis of tagged Ig-␣ expression on a Ig-␣flN and a Ig-␣HAN single transfectant and two Ig-␣flN/Ig-␣HAN double transfectants of J558L␮m. Cells were stained with anti-Flag (PE) and anti-HA (FITC) antibodies. (B) SDS–PAGE and Western blot analysis of affinity-purified BCR complexes from unstimulated J558L␮m transfectants. The NP-purified BCR complexes in the NIP eluate were precipitated with either anti-␮, anti-Flag, or anti-HA antibodies. Proteins were detected on the Western blot with anti-␮ (upper panel), anti-Flag (middle panel), or anti-HA (lower panel) antibodies.

The specificity of the antibodies used was controlled with BCR complexes from single Ig-␣ transfectants (Figure 3B, lanes 1–6). In conclusion, the digitonin-solubilized BCR contains only one Ig-␣/Ig-␤ heterodimer. A 1:1 Stoichiometry of the BCR Complex To measure the amount of each BCR component directly, we metabolically labeled J558L␮m/Ig-␣ and J558L␦m/Ig-␣ cells for 3–4 days with [35S]methionine. The long labeling period should ensure that most BCR proteins had incorporated the [35S]methionine. The BCR complexes from the digitonin lysates of these cells were isolated by the two-step affinity purification protocol. The BCR components were size separated by SDS– PAGE and their radioactivity determined (Figure 4A). Considering the total number of methionine residues in each IgM-BCR and IgD-BCR component, we calculated the molar ratio of HC:␭:Ig-␣:Ig-␤ to be 1.0:1.0 (⫾ 0.1):0.4 (⫾ 0.1):0.5 (⫾ 0.1). This stoichiometric analysis supports the model of a 1:1 association of the mIg molecule and the Ig-␣/Ig-␤ heterodimer. Next, we compared the size of three different BCR classes (IgG, IgD, and IgM) by BN–PAGE (Figure 4B). In this analysis, the IgG-BCR is the smallest and the IgMBCR is the largest protein complex (Figure 4B, lanes 1–3). The small size difference between the three BCR complexes is well explained by the increasing size of

Immunity 8

Figure 4. Stoichiometry and Composition of Different Classes of the BCR (A) Metabolic labeling and phosphoimager analysis. J558L␦m/Ig-␣ and J558L␮m/Ig-␣ cells were grown in medium containing [35S]methionine. Subsequently, the BCRs were purified by two-step affinity chromatography and separated by 10% reducing SDS–PAGE. Proteins were visualized by autoradiography. (B) BN–PAGE analysis of different classes of the BCR. The BCR were purified from the indicated Ig-transfectants of J558L or from splenic B lymphocytes (lane 8) of an Ig-transgenic mouse strain whose B cell receptor is NP specific. The eluted material was separated by BN–PAGE, transfered to a membrane, and blotted with anti-Ig-␣ or anti-␭ light chain antisera. Size markers were dimeric and monomeric ferritin (880 and 440 kDa).

their respective HC. It is therefore likely that the IgG2aBCR also has a 1:1 stoichiometry. By BN–PAGE, we analyzed whether the incorporation of a Flag-tagged Ig-␣ changes the composition of the BCR complex. The size of IgM-BCR complexes with Flag-tagged Ig-␣ proteins was the same as that of the BCR complex carrying a wild-type Ig-␣ (Figure 4B, lanes 4–7) and the same holds true for the IgD-BCR (Figure 4B, lanes 9–11). In a further control experiment, we found that BCR complexes isolated from the digitonin lysate of either unstimulated or pervanadate-stimulated cells have the same size (data not shown). Thus, the stimulation and phosphorylation of the BCR does not alter its stoichiometry. In order to analyze the BCR complex of normal murine B cells, we prepared a digitonin lysate of splenic B cells from a transgenic mouse strain that expresses an NPspecific BCR on the majority of its B cells (Pelanda et al., 1997). We then affinity purified the BCR and found that on a BN–PAGE gel, it has the same size as the IgDBCR isolated from J558L␦m/Ig-␣ cells (Figure 4B, lanes 8 and 9). This is consistent with the fact that IgD is the predominant class of BCR on splenic B cells. In summary, our studies show that in digitonin lysates, the BCR of different classes isolated from different sources is a stable complex with a ratio of mIg:Ig-␣/Ig-␤ of 1:1. Transphosphorylation of Two Different IgM-BCRs The TM sequence of the ␮m chain contains nine hydrophilic amino acids, and it was previously thought that the two Ig-␣/Ig-␤ heterodimers shield these sequences from the hydrophobic environment of the lipid bilayer (Williams et al., 1993). In the 1:1 BCR complex, the mIg molecule is partially unsheathed. It is, however, possible that on the surface of living B cells, the mIgM:Ig-␣/Ig-␤ complex forms a higher ordered structure in which the nine hydrophilic amino acids are sheathed. To obtain

Figure 5. Western Blot Analysis of the Tyrosine Phosphorylation of Ig-␣ Proteins in Stimulated B Cells Coexpressing Two Different BCR Complexes (A) IgM-BCR/IgD-BCR-coexpressing J558L cells were incubated for 2 min at 37⬚C with either control (c, lane 1), anti-␮ (lane 2), or anti-␦ (lane 3) antisera. (B) J558L␮m/Ig-␣ and J558L␮m/Ig-␣flN single transfectant and two J558L␮m/Ig-␣/Ig-␣flN double transfectants (4 and 3) were incubated for 2 min at 37⬚C with either the control anti-Ezrin antibody (c), with anti-␭ antiserum, or with the anti-Flag antibody. After cell lysis, phosphotyrosine-containing proteins were precipitated, separated by SDS–PAGE, and analyzed with an anti-Ig-␣ antiserum.

some evidence for such an oligomeric BCR structure, we monitored Ig-␣ phosphorylation in J558L transfectants exposed to different antibodies. Stimulation of normal B lymphocytes coexpressing IgM-BCR and IgD-BCR with anti-␮ antibodies induces phosphorylation of only the mIgM-associated Ig-␣, whereas anti-␦ antibodies induce phosphorylation of the mIgD-associated Ig-␣ (Gold et al., 1991). We found the same exclusive Ig-␣ phosphorylation in J558L cells coexpressing the two BCR classes (Figure 5A). Apparently, the activated protein tyrosine kinases phosphorylate only the engaged and not the nonligated BCR complex. We next monitored Ig-␣ phosphorylation in either single Ig-␣ transfectants (Figure 5B, lanes 1–6) or in J558L␮m/Ig-␣/Ig-␣flN cells coexpressing IgM-BCRs with wild-type and N-terminal-flagged Ig-␣ proteins. Cross-linking of the BCR with anti-␭ resulted in strong Ig-␣ phosphorylation, whereas phosphorylation induced by the anti-Flag antibody was less intense but clearly above background (Figure 5B, lanes 5 and 6). Interestingly, anti-Flag stimulation of the Ig-␣/Ig-␣flN double transfectant resulted in phosphorylation of both the mIg:Ig-␣flN/Ig-␤ and the mIg:Ig-␣/Ig-␤ complex (Figure 5B, lanes 9 and 12). Thus, the two IgM-BCR complexes, which are isolated separately in digitonin lysates, could be in contact with each other on the cell surface. Detection of Oligomeric BCR Complexes by BN–PAGE To obtain more direct evidence for the existence of an oligomeric BCR structure, we employed a technique recently developed by Scha¨gger and colleagues (Arnold et al., 1998). This group used a low detergent to protein

Oligomeric BCR Complexes 9

Figure 6. Analysis of Oligomeric BCR Complexes IgM-BCR (A), IgD-BCR (B), and IgD-hSbap-BCR (C) expressing cell lines were lysed with the indicated concentrations of thesit. After NP purification, the BCRs were eluted in 0.3% digitonin, size separated by two-dimensional BN–PAGE/SDS–PAGE, and detected on the Western blot with an anti-Ig-␤ antibody. For each transfectant, the anti-Ig-␤ reactive bands of the SDS–PAGE gels (second dimension) of the dilution series are combined into one panel. (D) Comparison of the putative ␣-helical TM regions of mIgD and of mIgD-hSbap. Amino acids are numbered from the N terminus to the C terminus, and the introduced mutations are marked by an asterisk. (E) Surface biotinylated J558L␦m/Ig-␣ cells were lysed with a buffer containing the indicated concentrations of thesit. NP-purified complexes were separated by reducing SDS–PAGE and detected on the Western blot with streptavidin. (F) IgM-BCR/IgD-BCR-coexpressing J558L cells (J558L␮m/␦m/Ig-␣) were lysed with the indicated concentrations of thesit. The BCR complexes were purified using an anti-␦ (lanes 1–6) or an anti-␮ antiserum (lanes 7–12). Proteins were separated by reducing SDS–PAGE, and the HC was detected with anti-␦ and anti-␮ antisera.

ratio for the solubilization of membranes and established the dimeric nature of a transmembrane protein previously thought to be monomeric. We first used low concentrations of digitonin that failed to solubilize the BCR (data not shown). The detergent thesit, however, efficiently solubilized the BCR even if used at a 100-fold reduced concentration (Figure 6E). We therefore lysed J558L␮m/Ig-␣ and J558L␦m/Ig-␣ cells with decreasing amounts of thesit (0.5%, 0.06%, 0.03%, 0.02%, and

0.015%) and purified the BCR complexes from these lysates with NP-Sepharose. After elution with NIP in 0.3% digitonin, the BCR complexes were size separated by twodimensional BN–PAGE/SDS–PAGE and detected with an anti-Ig-␤ antibody (Figures 6A–6C). In contrast to the IgD-BCR, the IgM-BCR is not stable in 0.5% or 0.06% thesit (Schamel and Reth, 2000), and thus the Ig-␣/ Ig-␤ heterodimer is not copurified. From 0.03% or 0.02% thesit lysates, however, the complete BCR is purified

Immunity 10

as a high molecular weight complex which barely enters the BN gel (Figure 6A). In 0.5% thesit, the IgD-BCR is mainly monomeric (Figure 6B) and has the same size as the IgD-BCR we previously isolated from 1% digitonin lysates. In 0.06% thesit, the analysis reveals dimeric and tetrameric BCR complexes. With decreasing detergent concentrations, the size of the IgD-BCR increases (Figure 6B). The TM region of mIg molecules has two sides. The sequence of one side is conserved between all Ig classes and is involved in the binding of the Ig-␣/Ig-␤ heterodimer (Figure 6D). The sequence of the opposite side (Sbap) is class specific but evolutionarily conserved (Adachi et al., 1996). This side is expected to be involved in the formation of the class-specific oligomeric BCR complex. Therefore, we mutated all hydrophilic and aromatic amino acids on the Sbap side of the ␦m TM region to small hydrophobic amino acids (Figure 6D). This mutant ␦m molecule (␦m-hSbap) was expressed in J558L/ Ig-␣ cells, where it was assembled with the Ig-␣/Ig-␤ heterodimer into an IgD-BCR and transported on the cell surface (data not shown). Lysis of J558L␦m-hSbap/ Ig-␣ cells in the presence of limiting concentrations of thesit revealed a monomeric BCR complex, which in low detergent concentrations shifted only to a dimeric complex (Figure 6C). The formation of higher molecular weight complexes of the IgD-BCR is thus critically dependent on the Sbap side of the ␦m TM helix. To exclude that the high molecular weight BCR complexes are formed by nonspecific aggregation with other membrane proteins, J558L␦m/Ig-␣ cells were surface biotinylated and lysed in limiting concentrations of thesit. After affinity purification, the proteins were separated by reducing SDS–PAGE and surface proteins were detected with streptavidin (Figure 6E). No membrane proteins other than the BCR components (␦m, ␭, Ig-␣, and Ig-␤) were detected. Thus, the low detergent concentrations that allow the detection of oligomeric IgDBCR complexes do not result in a nonspecific aggregation of the BCR with other membrane proteins. We cannot exclude, however, that an intracellular protein is attached to the IgD-BCR under these detergent conditions. Next, we examined J558L␮m/␦m/Ig-␣ cells that coexpress the IgD-and IgM-BCR on the same cell (Kim and Reth, 1995). The cells were lysed with decreasing amounts of thesit, and the BCR was purified with either anti-␦ or anti-␮ antisera (Figure 6F). We did not observe any copurification of the two BCR classes even at low thesit concentrations. This suggests that the oligomers formed by the IgM-BCR and the IgD-BCR differ from each other and only contain BCR monomers of the same class.

Discussion Transmembrane receptors have to convert an extracellular signal into an intracellular signal. In other words, ligand binding has to be translated into a structural alteration of the receptor that is transduced across the membrane and results in the activation or deactivation of intracellular signaling molecules. To understand the signaling function of a receptor, it is important to know its

exact structure before and after ligand binding. Unfortunately, this information is hard to obtain for most receptors because their purification requires detergent lysis, which can alter their structure and in most cases prevent their crystallization. To increase our knowledge of the basic structure of the BCR, we have analyzed its size and composition under different lysis conditions. By using BN–PAGE (Scha¨gger and Jagow, 1991; Scha¨gger et al., 1994) and the BIG method (Schamel, 1999), we show that the digitonin-solubilized BCR is a discrete complex composed of only HC, LC, Ig-␣, and Ig-␤. These components are also necessary and sufficient for the expression of the BCR in nonlymphoid mammalian cells (Venkitaraman et al., 1991; Matsuuchi et al., 1992) and in Drosophila cells (W. A. S. and M. R., unpublished data). By coexpressing wild-type and peptide-tagged Ig-␣ molecules, we demonstrate that the digitonin-solubilized BCR complex possesses only one Ig-␣/Ig-␤ heterodimer. This observation and the specific purification of the BCR from the digitonin lysate of biosynthetically labeled cells allowed us to determine a 1:1 stoichiometry for the mIg:Ig-␣/Ig-␤ complexes. This finding corrects the current structural model of the BCR. In BN gels, the IgD-, IgM-, and IgG2a-BCR complexes have a similar size and it is thus likely that the BCRs of all five IgH classes have a 1:1 stoichiometry. This ratio is not changed after stimulation of B cells, and no Ig-␣/Ig-␤ was released during the purification of the BCR. The 1:1 stoichiometry of the BCR suggests that the mIg molecule is not completely symmetrical. Indeed, if the two HC TM regions do not bind to each other with the same side, the symmetry of the mIg molecule is broken. It is possible that both HC TM regions are involved in the binding of the Ig-␣/Ig-␤ heterodimer, as suggested by the fact that the Ig-␣/Ig-␤ heterodimer can only be copurified with the complete mIg molecule (2HC/2LC) and not with a single HC/LC pair (Benlagha et al., 1999). For the stoichiometric analysis, the BCR was solubilized with digitonin because this detergent can preserve the structure of multicomponent receptors. However, it has recently been found that at low concentrations, detergents like thesit or Triton X-100 provide much milder lysis conditions than digitonin (Arnold et al., 1998). This observation allows the detection of higher organized forms of membrane proteins (Scha¨gger and Pfeiffer, 2000). We applied these methods to the BCR and found that the BCR can be purified from low percentage thesit lysates as a high molecular weight complex containing several BCR monomers. Although the stoichiometry of these oligomeric BCR complexes remains unknown, our study suggests that the IgM-BCR and IgD-BCR form distinct oligomeric complexes and that the HC TM region is critically involved in the formation and/or stability of the oligomer. It is possible that in the oligomeric BCR structure, all polar amino acids of the HC TM region are shielded from the hydrophobic environment of the lipid bilayer (Williams et al., 1993). The CH domains may also be involved in the formation of an oligomeric BCR complex, and we are currently testing this possibility. The secreted form of the IgM molecule is a pentamer. Oligomerization of secreted IgM is not controlled by the J chain alone, since it also occurs in the absence of the J chain (Niles et al., 1995; Cattaneo

Oligomeric BCR Complexes 11

and Neuberger, 1987). Indeed, the ␮s-specific C terminus, the C␮3 domain, and the C␮4 domain cooperate to mediate the oligomerization of IgM (Wiersma et al., 1998). The C␮3 and C␮4 domains are also present in the mIgM molecule, and it is thus possible that on the surface of B cells, the IgM-BCR has a pentameric or hexameric structure. The secreted IgM and mIgM oligomers should, however, be distinct from each other, as the interdomain disulfide bonds are not formed in the latter case (Figure 1A). Recent biophysical studies indicate that receptors which, according to their migration in SDS–PAGE, were previously thought to be monomers, form dimeric or multimeric structures in the absence of their ligand. This is the case for the bacterial chemotaxis receptor (Liu et al., 1997; Kim, 1999), the EPO-R (Syed et al., 1998; Livnah et al., 1999; Remy et al., 1999), the EGF-R (Gadella and Jovin, 1995), and the PDGF-R (Wiseman et al., 1997). Furthermore, in many cases the dimerization is necessary but not sufficient for receptor activation (reviewed in Jiang and Hunter, 1999). The binding of a ligand to these receptors does not result in further aggregation but rather in a conformational change. Often it is the ligand that fixes the receptor molecules in a precise orientation required for signal transduction (Chi et al., 1997; Burke and Stern, 1998; Syed et al., 1998; Livnah et al., 1999; Ottemann et al., 1999; Remy et al., 1999). In this context, the BCR must accomplish a very special task: to reliably translate the binding of thousands of different antigenic structures into an activation signal. How can all these antigens bring the BCR in the same active conformation? An oligomeric BCR structure may provide a solution to this problem. Within the BCR oligomer, transducer elements may be bound in an inactive form, and the Ig-␣/Ig-␤ heterodimer may not be accessible for phosphorylation. Binding to antigen may result in a disturbance or disruption of the oligomeric complexes. This could lead to the accessibility and phosphorylation of the Ig-␣/Ig-␤ heterodimer and the activation of the BCR-transducing elements like Syk. It is easy to imagine that unlike the precise dimerization, the disruption of a preformed BCR complex does not require a specific ligand structure. It requires, however, a multivalent antigen or bivalent antibodies, as a monovalent ligand would not efficiently disturb the oligomer. The model of an oligomeric BCR can also explain the finding of the class-specific ITAM phosphorylation. DeFranco and Gold (Gold et al., 1991) found that anti-␮ antibodies induce phosphorylation of only the mIgMassociated Ig-␣, whereas anti-␦ antibodies induce phosphorylation of the mIgD-associated Ig-␣. Our data suggest that the IgM-BCR and the IgD-BCR form distinct oligomeric complexes. Thus, the engagement of the IgDBCR should not result in the disturbance of the IgMBCR. The coexpressed wild-type and Flag-tagged Ig-␣ proteins, however, can be incorporated in the same oligomeric BCR structure, and that would explain why the anti-Flag antibody resulted in phosphorylation of both forms of Ig-␣ (Figure 6B). When immature B cells have succeeded to produce a LC that binds to the preexisting HC, they express an IgM-BCR on their surface. At that stage, the cells receive a signal from the BCR that prevents further VL gene assembly and allows their export from the bone marrow

(Rolink et al., 1999; Carsetti, 2000). It was proposed that this signal involves the cross-linking of the BCR by a B cell–autonomous ligand or by self-antigens (Pillai, 1999). However, BCR cross-linking at that B cell–developmental stage should result in negative rather than positive selection of the B cells. According to our model, it is the formation of the oligomeric BCR structure that gives the signal for positive selection of immature B cells and also for the survival of normal B cells (Lam et al., 1997; Neuberger, 1997). The nature of that signal is not yet clear, but the signal should be active in all B cells expressing an oligomeric BCR on their cell surface. Autoreactive B cells would not receive the positive selection signal but rather an activation signal, as their oligomeric BCR structure would be disturbed upon contact with the self-antigen. The oligomeric BCR could provide a working model for studies on the T cell antigen receptor (TCR) complex. Using T cells coexpressing a murine and a human CD3⑀, it has been shown that the TCR contains more than one CD3⑀ subunit (de la Hera et al., 1991). Similarly, T cells coexpressing two different TCR␤ chains were used to demonstrate that more than one TCR␤ chain is present in the TCR complex (Fernandez-Miguel et al., 1999). The exact stoichiometry of the TCR under different conditions of detergent lysis, however, has not been determined so far. A better structural description of the BCR and TCR would greatly help to elucidate the signaling mechanism operating during positive and negative selection of lymphocytes. Experimental Procedures Expression Vectors To generate the expression vectors for tagged Ig-␣ proteins, oligonucleotides coding for the Flag tag or HA tag were inserted into the pEVmb-1neo vector. The insertion occurred either before the stop codon of the Ig-␣ coding sequence (pEVmb-1FlagCneo, a kind gift of Manfred Kraus and Klaus Rajewsky) or behind the sequence coding for the signal peptidase cleavage point (pEVmb-1FlagNneo and pEVmb-1HANneo). The expression vector pSV␦m-hSbap was constructed essentially as described (Adachi et al., 1996). Briefly, the sequence coding for the TM part of pSV␦m-Xho1 was replaced by the following three double strand oligonucleotides utilizing the Xho1 site: 5⬘-TCGAGGAGGAGAACGGCCTGGTGGTCACAATGGC CCT-3⬘, 5⬘-U␦m-␣BAP; 5⬘-GAAGAGGGCCATTGTGACCACCAGGCC GTTCTCCTCC-3⬘, 5⬘-L␦m-␣BAP; 5⬘-CTTCGTGGCCCTCTTCCTGCT CACACTGCTC-3⬘, M-U␦m; 5⬘-TGTAGAGCAGTGTGAGCAGGAAG AGGGCCAC-3⬘, M-L␦m; 5⬘-TACAGTGGCGTCGTCACCGTCATC AAGGTAAAG-3⬘, 3⬘-U␦m-␣BAP; 5⬘-CTTTACCTTGATGACGGTGA CGACGCCAC-3⬘, 3⬘-L␦m-␣BAP. Underlined sequences represent the Xho1 site and bold sequences represent the TM region. The sequence of the vector pSV␦m-hSbap was confirmed by sequencing. Cell Culture, Mouse Strain, and Flow Cytometry The cell lines J558L␮m/Ig-␣, J558L␦m/Ig-␣, and J558L/␮m/␦m/Ig␣7–6 have been described previously (Hombach et al., 1988; Kim et al., 1994; Kim and Reth, 1995) (named J558L␮m3, J558L␦␦m/mb-1, and J558L/␮m3/␦m7–6 therein). To obtain J558L lines with tagged Ig-␣, cells were transfected with the corresponding expression vectors and grown as described (Adachi et al., 1996). The B1-8 mouse (a gift of Pelanda and Rajewsky) was derived from a cross of B18Hi, 3-83Ki, and 3-83Hi; 3-83Ki animals having a MHC class I H-2b background (Pelanda et al., 1997). Due to intensive receptor editing, approximately 50% of the B1-8Hi B cells expressed a ␭ light chain (R. Pelanda, personal communication). For FACScan analysis, cells were stained sequentially with anti-Flag antibodies (M2, 1:100; Kodak), a phycoerythrin (PE)-labeled goat anti-mouse IgG antiserum

Immunity 12

(1:50; Southern Biotechnology), and fluorescein (FITC)-labeled antiHA antibodies (12CA5, 1:25; Boehringer). Cells were washed, analyzed with a FACScan (Becton Dickinson), and 104 living cells were plotted on a double logarithmic scale.

in 1ml lysis buffer (Adachi et al., 1996) including 1% Triton X-100 and 0.5 mM sodium orthovanadate. Anti-phosphotyrosine affinity chromatography and Western blotting are described above. Acknowledgments

Affinity Purification and Western Blotting For the two-step affinity purification of the BCR, 2 ⫻ 107 cells were stimulated with 50 ␮M pervanadate for 3 min as described in Wienands et al. (1996) and lysed in 1 ml lysis buffer (Adachi et al., 1996) containing 0.75% digitonin and 0.5 mM sodium orthovanadate. Phosphorylated proteins were first purified with 5 ␮l anti-phosphotyrosine agarose (PT66, 6 hr, 4⬚C; Sigma). After elution of bound proteins in 1 ml lysis buffer containing 50 mM phenylphosphate, BCRs were further purified using NP-conjugated Sepharose and subsequently eluted with 0.5 mM NIP-cap (5-iodo-NP coupled to ⑀-aminocaproic acid) in 50 mM Bis-Tris, 10% glycerol, 750 mM ⑀-aminocaproic acid, and 50 mM NaCl. Samples (material from 1 ⫻ 107 cells) with or without 1% SDS were loaded onto a BN–PAGE gel (5.5%–16%) and run as described (Scha¨gger et al., 1994). To purify the tagged Ig-␣ proteins, we used either the monoclonal anti-Flag antibody M2 (Kodak) or the monoclonal anti-HA antibody 12CA5 and protein G–coupled Sepharose (Pharmacia). For deglycosylation, proteins were incubated in 50 mM sodium citrate (pH 5.5), 0.05% SDS, and 50 mM ␤-mercaptoethanol containing 1 mU EndoH for 10 hr at 37⬚C. SDS–PAGE and Western blotting are described in Adachi et al. (1996). Western blots were developed either with anti-Flag (M2) or biotinylated anti-HA (12CA5) antibodies, anti-Ig-␣, horseradish peroxidase (HRP)-conjugated anti-␦, or HRP-conjugated anti-␮ antisera (Southern Biotechnology). The rabbit anti-Ig-␣ serum was a generous gift of Dr. J. Wienands. Secondary antibodies were used as described (Kim et al., 1994; Adachi et al., 1996). Molecular weight standards were from Biorad (Kaleidoscope Prestained Standards). Lysis with Limiting Detergent Concentrations Cells (8 ⫻ 106) were lysed for 15 min on ice in 1 ml lysis buffer (Adachi et al., 1996) containing different amounts of thesit (Boehringer). The lysates were incubated with 2 ␮l NP-Sepharose for 4 hr, and the depleted supernatant was removed completely from the beads. For SDS–PAGE, the beads were washed once in 20 ␮l of 0.3% digitonin, 40 mM Bis-Tris (pH 7.0), 10% glycerol, 500 mM ⑀-aminocaproic acid, 2 mM EDTA, and 50 mM NaCl (BN-elution buffer) for 20 min. After boiling of the beads in Laemmli buffer, the proteins were separated by 10% reducing SDS–PAGE, and Western blotting was performed as described above. For BN–PAGE, the bound proteins were eluted with 0.5 mM NIP in the BN-elution buffer for 20 min. After separation of the samples (material from 1 ⫻ 107 cells) by BN–PAGE (5.5%– 12%), a second dimension-reducing SDS–PAGE (10%) was applied. The proteins were transferred to nitrocellulose membranes and detected with anti-Ig-␤ monoclonal antibodies (MAB29T-04). Metabolic Labeling J558L␮m/Ig-␣ and J558L␦m/Ig-␣ cells were grown for 3 days in complete RPMI 1640 medium in the presence of 0.15 ␮Ci/ml of [35S]methionine (revidue L-[35S]methionine; Amersham). The BCR complex was isolated from 5 ⫻ 108 cells using the two-step affinity purification protocol, and the BCR components were separated by reducing SDS–PAGE. Gels were fixed in 30% ethanol, 10% acetic acid for 30 min, dried, and exposed to Fuji BasIIIs phosphorimaging plates. The radioactivity of each individual protein was measured (Fujix Bas1000 and Fujifilm Bas1500, Fuji) and divided by the number of methionines present in the corresponding mature protein (6 for ␮m, 12 for ␦m, 2 for ␭1, 3 for Ig-␣, and 2 for Ig-␤). Values were normalized to those obtained for the HCs. B Cell Stimulation Cells (1.5 ⫻ 107) were first incubated for 30 min at 37⬚C and 5% CO2 in serum-free RPMI medium (Adachi et al., 1996) and then placed on ice for 20 min with 2.5 ␮g of control or BCR-binding antibodies in phosphate-buffered saline (PBS). An isotype-matched anti-Ezrin antibody was used as control for the anti-Flag antibody. After two washes in PBS, cells were resuspended in 100 ␮l of the same medium containing 200 ␮M sodium orthovanadate (RPMIV). Cells were then incubated for 2 min in 1ml RPMIV (37⬚C and 5% CO2) and lysed

We thank Dr. J. Wienands for the anti-Ig-␣ antiserum and advice, Drs. M. Kraus and K. Rajewsky for the expression vector pEVmb1flgCneo, Drs. H. Scha¨gger and N. Pfanner for help with the BN– PAGE, K. Gimborn for technical help, S. Kuppig for discussions, Dr. R. Pelanda for the B1-8 transgenic mice, and Drs. J. Wienands, R. Pelanda, P. J. Nielsen and L. Leclercq for reading of the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft through SFB 388 and the Leibniz Prize. Received April 21, 2000; revised June 13, 2000. References Adachi, T., Schamel, W.W.A., Kim, K.M., Watanabe, T., Becker, B., Nielsen, P.J., and Reth, M. (1996). The specificity of association of the IgD molecule with the accessory proteins BAP31/BAP29 lies in the IgD transmembrane sequence. EMBO J. 15, 1534–1541. Ales, M.J.E., Warner, G.L., and Scott, D.W. (1988). Immunoglobulins D and M mediate signals that are qualitatively different in B cells with an immature phenotype. Proc. Natl. Acad. Sci. USA 85, 6919–6923. Arnold, I., Pfeiffer, K., Neupert, W., Stuart, R.A., and Scha¨gger, H. (1998). Yeast mitochondrial F1F0-ATP synthase exists as a dimer: identification of three dimer-specific subunits. EMBO J. 17, 7170– 7178. Benlagha, K., Guglielmi, P., Cooper, M.D., and Lassoued, K. (1999). Modifications of Iga and Igb expression as a function of B lineage differentiation. J. Biol. Chem. 274, 19389–19396. Benschop, R.J., and Cambier, J.C. (1999). B cell development: signal transduction by antigen receptors and their surrogates. Curr. Opin. Immunol. 11, 143–151. Burke, C.L., and Stern, D.F. (1998). Activation of Neu (ErbB-2) mediated by disulfide bond-induced dimerization reveals a receptor tyrosine kinase dimer interface. Mol. Cell. Biol. 18, 5371–5379. Campbell, K.S., Hager, E.J., and Cambier, J.C. (1991). Alpha-chains of IgM and IgD antigen receptor complexes are differentially N-glycosylated MB-1-related molecules. J. Immunol. 147, 1575– 1580. Campbell, K.S., Ba¨ckstro¨m, B.T., Tiefenthaler, G., and Palmer, E. (1994). CART: a conserved antigen receptor transmembrane motif. Semin. Immunol. 6, 393–410. Carsetti, R. (2000). The development of B cells in the bone marrow is controlled by the balance between cell-autonomous mechanisms and signals from the microenvironment. J. Exp. Med. 191, 5–8. Cattaneo, A., and Neuberger, M.S. (1987). Polymeric immunoglobulin M is secreted by transfectants of non-lymphoid cells in the absence of immunoglobulin J chain. EMBO J. 6, 2753–2758. Chen, J.Z., Stall, A.M., Herzenberg, L.A., and Herzenberg, L.A. (1990). Differences in glycoprotein complexes associated with IgM and IgD on normal murine B cells potentially enable transduction of different signals. EMBO J. 9, 2117–2124. Chi, Y.I., Yokota, H., and Kim, S.H. (1997). Apo structure of the ligand-binding domain of aspartate receptor from E. coli and its comparison with ligand-bound or pseudoligand-bound structures. FEBS Lett. 414, 327–332. DeFranco, A.L. (1997). The complexity of signaling pathways activated by the BCR. Curr. Opin. Immunol. 9, 296–308. de la Hera, A., Muller, U., Olsson, C., Isaaz, S., and Tunnacliffe, A. (1991). Structure of the T cell antigen receptor (TCR): two CD3 epsilon subunits in a functional TCR/CD3 complex. J. Exp. Med. 173, 7–17. Dintzis, R.Z., Vogelstein, B., and Dintzis, H.M. (1982). Specific cellular stimulation in the primary immune response: experimental test of a quantized model. Proc. Natl. Acad. Sci. USA 79, 884–888. Dintzis, R.Z., Okajima, M., Middleton, M.H., Greene, G., and Dintzis,

Oligomeric BCR Complexes 13

H.M. (1989). The immunogenicity of soluble haptenated polymers is determined by molecular mass and hapten valence. J. Immunol. 143, 1239–1244.

and Rajewsky, K. (1997). Receptor editing in a transgenic mouse model: site, efficiency, and role in B cell tolerance and antibody diversification. Immunity 7, 765–775.

Fernandez-Miguel, G., Alarcon, B., Iglesias, A., Bluethmann, H., Alvarez-Mon, M., Sanz, E., and de la Hera, A. (1999). Multivalent structure of an alphabeta T cell receptor. Proc. Natl. Acad. Sci. USA 96, 1547–1552.

Pillai, S. (1999). The chosen few? Positive selection and the generation of naive B lymphocytes. Immunity 10, 493–502.

Gadella, T.W.J., Jr., and Jovin, T.M. (1995). Oligomerization of epidermal growth factor receptors on A432 cells studied by timeresolved fluorescence imaging microscopy. A stereochemical model for tyrosine kinase receptor activation. J. Cell Biol. 129, 1543–1558. Gold, M.R., Matsuuchi, L., Kelly, R.B., and DeFranco, A.L. (1991). Tyrosine phosphorylation of components of the B-cell antigen receptors following receptor crosslinking. Proc. Natl. Acad. Sci. USA 88, 3436–3440. Grupp, S.A., Campbell, K., Mitchell, R.N., Cambier, J.C., and Abbas, A.K. (1993). Signaling-defective mutants of the B lymphocyte antigen receptor fail to associate with Ig-alpha and Ig-beta/gamma. J. Biol. Chem. 268, 25776–25779. Hombach, J., Leclercq, L., Radbruch, A., Rajewsky, K., and Reth, M. (1988). A novel 34-kd protein coisolated with the IgM molecule in surface IgM-expressing cells. EMBO J. 7, 3451–3456. Hombach, J., Lottspeich, F., and Reth, M. (1990). Identification of the genes encoding the IgM-alpha and Ig-beta components of the IgM antigen receptor complex by amino-terminal sequencing. Eur. J. Immunol. 20, 2795–2799. Jiang, G., and Hunter, T. (1999). Receptor signaling: when dimerization is not enough. Curr. Biol. 9, R568–R571. Kim, K.K. (1999). Four-helical-bundle structure of the cytoplasmic domain of a serine chemotaxis receptor. Nature 400, 787–792. Kim, K.M., and Reth, M. (1995). The B cell antigen receptor of class IgD induces a stronger and more prolonged protein tyrosine phosphorylation than that of class IgM. J. Exp. Med. 181, 1005–1014. Kim, K.M., Adachi, T., Nielsen, P.J., Terashima, M., Lamers, M.C., Kohler, G., and Reth, M. (1994). Two new proteins preferentially associated with membrane immunoglobulin D. EMBO J. 13, 3793– 3800.

Pogue, S.L., and Goodnow, C.C. (1994). Ig heavy chain extracellular spacer confers unique glycosylation of the Mb-1 component of the B cell antigen receptor complex. J. Immunol. 152, 3925–3934. Remy, I., Wilson, I.A., and Michnick, S.W. (1999). Erythropoietin receptor activation by a ligand-induced conformation change. Science 283, 990–993. Reth, M. (1989). Antigen receptor tail clue. Nature 338, 383–4. Reth, M. (1992). Antigen receptors on B lymphocytes. Annu. Rev. Immunol. 10, 97–121. Reth, M., and Wienands, J. (1997). Initiation and processing of signals from the B cell antigen receptor. Annu. Rev. Immunol. 15, 453–479. Reth, M., Wienands, J., Tsubata, T., and Hombach, J. (1991). Identification of components of the B cell antigen receptor complex. Adv. Exp. Med. Biol. 292, 207–214. Rolink, A.G., Melchers, F., and Andersson, J. (1999). The transition from immature to mature B cells. Curr. Top. Microbiol. Immunol. 246, 39–43. Sanchez, M., Misulovin, Z., Burkhardt, A.L., Mahajan, S., Costa, T., Bolen, J.B., and Nussenzweig, M. (1993). Signal transduction by immunoglobulin is mediated through Ig-␣ and Ig-␤. J. Exp. Med. 178, 1049–1055. Scha¨gger, H., and Jagow, G. (1991). Blue native electrophoresis for isolation of membrane protein complexes in enzymatically active form. Anal. Biochem. 199, 223–231. Scha¨gger, H., and Pfeiffer, K. (2000). Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J. 19, 1777–1783.

Kurosaki, T. (1997). Molecular mechanisms in B cell antigen receptor signaling. Curr. Opin. Immunol. 9, 309–318.

Scha¨gger, H., Cramer, W.A., and von Jagow, G. (1994). Analysis of molecular masses and oligomeric states of protein complexes by blue native electrophoresis and isolation of membrane protein complexes by two-dimensional native electrophoresis. Anal. Biochem. 217, 220–230.

Lam, K.P., Kuhn, R., and Rajewsky, K. (1997). In vivo ablation of surface immunoglobulin on mature B cells by inducible gene targeting results in rapid cell death. Cell 90, 1073–1083.

Schamel, W.W.A. (1999). Detection of proteins after biotinylation within polyacrylamide gels (BIG Method). Anal. Biochem. 274, 144–146.

Liu, Y., Levit, M., Lurz, R., Surette, M.G., and Stock, J.B. (1997). Receptor-mediated protein kinase activation and the mechanism of transmembrane signaling in bacterial chemotaxis. EMBO J. 16, 7231–7240.

Schamel, W.W.A., and Reth, M. (2000). Stability of the B cell antigen receptor complex. Mol. Immunol. 37, 253–259.

Livnah, O., Stura, E.A., Middleton, S.A., Johnson, D.L., Jolliffe, L.K., and Wilson, I.A. (1999). Crystallographic evidence for preformed dimers of erythropoietin receptor before ligand activation. Science 283, 987–990. Matsuuchi, L., Gold, M.R., Travis, A., Grosschedl, R., DeFranco, A.L., and Kelly, R.B. (1992). The membrane IgM-associated proteins MB-1 and Ig-beta are sufficient to promote surface expression of a partially functional B-cell antigen receptor in a nonlymphoid cell line. Proc. Natl. Acad. Sci. USA 89, 3404–3408. Moro, F., Sirrenberg, C., Schneider, H.-C., Neupert, W., and Brunner, M. (1999). The TIM17.23 preprotein translocase of mitochondria: composition and function in protein transport into the matrix. EMBO J. 18, 3667–3675. Neuberger, M.S. (1997). Antigen receptor signaling gives lymphocytes a long life. Cell 90, 971–973. Niles, M.J., Matsuuchi, L., and Koshland, M.E. (1995). Polymer IgM assembly and secretion in lymphoid and nonlymphoid cell lines: evidence that J chain is required for pentamer IgM synthesis. Proc. Natl. Acad. Sci. USA 92, 2884–2888. Ottemann, K.M., Xiao, W., Shin, Y.K., and Koshland, D.E., Jr. (1999). A piston model for transmembrane signaling of the aspartate receptor. Science 285, 1751–1754. Pelanda, R., Schwers, S., Sonoda, E., Torres, R.M., Nemazee, D.,

Shaw, A.C., Mitchell, R.N., Weaver, Y.K., Campos, T.J., Abbas, A.K., and Leder, P. (1990). Mutations of immunoglobulin transmembrane and cytoplasmic domains: effects on intracellular signaling and antigen presentation. Cell 63, 381–392. Stevens, T.L., Blum, J.H., Foy, S.P., Matsuuchi, L., and DeFranco, A.L. (1994). A mutation of the mu transmembrane that disrupts endoplasmic reticulum retention. Effects on association with accessory proteins and signal transduction. J. Immunol. 152, 4397–4406. Syed, R.S., Reid, S.W., Li, C., Cheetham, J.C., Aoki, K.H., Liu, B., Zhan, H., Osslud, T.D., Chirino, A.J., Zhang, J., et al. (1998). Efficiency of signalling through cytokine receptors depends critically on receptor orientation. Nature 395, 511–516. Venkitaraman, A.R., Williams, G.T., Dariavach, P., and Neuberger, M.S. (1991). The B-cell antigen receptor of the five immunoglobulin classes. Nature 352, 777–781. Wienands, J. (1999). The B-Cell Antigen Receptor: Formation of Signaling Complexes and the Function of Adaptor Proteins (Springer-Verlag: Berlin). Wienands, J., and Reth, M. (1991). The B cell antigen receptor of class IgD can be expressed on the cell surface in two different forms. Eur. J. Immunol. 21, 2373–2378. Wienands, J., Larbolette, O., and Reth, M. (1996). Evidence for a preformed transducer complex organized by the B cell antigen receptor. Proc. Natl. Acad. Sci. USA 93, 7865–7870.

Immunity 14

Wiersma, E.J., Collins, C., Fazel, S., and Shulman, M.J. (1998). Structural and functional analysis of J chain-deficient IgM. J. Immunol. 160, 5979–5989. Williams, G.T., Dariavach, P., Venkitaraman, A.R., Gilmore, D.J., and Neuberger, M.S. (1993). Membrane immunoglobulin without sheath or anchor. Mol. Immunol. 30, 1427–1432. Wiseman, P.W., Ho¨ddelius, P., Petersen, N.O., and Magnusson, K.-E. (1997). Aggregation of PDGF-beta receptors in human skin fibroblasts: characterization by image correlation spectroscopy (ICS). FEBS Lett. 401, 43–48.