B cell Ag receptor (BCR) signaling changes dramatically during B cell development, resulting in activation in mature B cells and apoptosis, receptor editing, or anergy in immature B cells. BCR signaling in mature B cells was shown to be initiated by the translocation of the BCR into cholesterol- and sphingolipid-enriched membrane microdomains that include the Src family kinase Lyn and exclude the phosphatase CD45. Subsequently the BCR is rapidly internalized into the cell. Here we show that the BCR in the immature B cell line, WEHI-231, does not translocate into lipid rafts following cross-linking nor is the BCR rapidly internalized. The immature BCR initiates signaling from outside lipid rafts as evidenced by the immediate induction of an array of phosphoproteins and subsequent apoptosis. The failure of the BCR in immature B cells to enter lipid rafts may contribute to the dramatic difference in the outcome of signaling in mature and immature B cells.
One of the hallmarks of the immune system is the ability to discriminate self from nonself. To a large extent this is accomplished for B cells by the elimination of self-reactive cells during development (1). An underlying premise of the selection process is that the engagement of immune receptors by Ags at different developmental stages has radically different outcomes; signaling at an immature stage leads to elimination vs activation at a mature stage.
Although the discreet developmental stages of B cells are well defined (2, 3, 4), the molecular mechanisms that underlie the differential response to Ags in mature and immature cells remain to be elucidated (5). Recent studies of B cell Ag receptor (BCR)3 function in mature B cells have revealed a previously unappreciated step in BCR-induced activation following Ag encounter (reviewed in 6). Following cross-linking, the BCR was observed to rapidly translocate into specialized cholesterol- and sphingolipid-rich membrane microdomains, termed lipid rafts (7). The lipid rafts appear to serve as platforms for BCR signaling and concentrate the Src family kinase Lyn, as well as regulators of Lyn function (7, 8), and exclude the phosphatase CD45 (7). Following cross-linking, the BCR and Lyn in the lipid rafts were phosphorylated, and the BCR was subsequently rapidly targeted into the cell (7). Thus, lipid rafts appear to serve as a platform for both BCR signaling and trafficking in mature B cells.
Here we characterize the location of the BCR in the immature B cell line, WEHI-231, following cross-linking and provide evidence that the translocation of the BCR into lipid rafts defines an early step in the BCR signaling pathway that discriminates the response to Ags in immature and mature B cells.
Materials and Methods
Cell lines and Abs
The phosphotyrosine-specific mAb containing HRP, RC20H, was purchased from BD Transduction Laboratories (San Diego, CA). F(ab′)2 goat Abs specific for mouse IgG + IgM (F(ab′)2 anti-Ig) or mouse IgM (F(ab′)2 anti-IgM), fluorescein-conjugated whole and F(ab′)2 goat Abs specific for mouse IgM (FL-anti-IgM and FL-F(ab′)2 anti-IgM), and nonspecific, isotype control F(ab′)2 goat IgG (F(ab′)2 Ig) were purchased from Jackson ImmunoResearch (West Grove, PA). Fab goat Abs specific for mouse IgM (Fab anti-IgM) or mouse IgG (Fab anti-IgG) and HRP-conjugated goat Abs specific for mouse IgG + IgM (HRP-anti-Ig) or rabbit IgG (HRP-anti-rabbit) were also purchased from Jackson ImmunoResearch. WS-2, affinity-purified rabbit polyclonal Abs specific for the cytoplasmic domain of Igα, were previously described (11). HRP-conjugated cholera toxin B subunit (HRP-CTB) was purchased from Sigma (St. Louis, MO). Fab anti-IgM and Fab anti-IgG were iodinated as described (12).
Raft isolation and fluorometry
Membrane rafts from 1 × 107 cells were isolated on OptiPrep gradients as described (13). Aliquots of fractions from the OptiPrep gradient were diluted 1:4 in 10 mM Tris, pH 9.0, and analyzed on an ISS-PC1 spectrofluorometer ISS (Champagne, IL).
Cells were incubated in 100 μl binding buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2) containing 5 ml FITC-annexin V (BD PharMingen, San Diego, CA) and propidium iodide (5 μg/ml) (BD PharMingen) at 25οC in the dark for 15 min. Binding buffer (400 μl) was added to dilute the samples before analysis on a flow cytometer. Cells positively labeled for both annexin V and propidium iodide were considered apoptotic.
Results and Discussion
The position of the BCR on the plasma membrane of WEHI-231 and CH27 cells was compared following cross-linking. Cells were incubated with FL-F(ab′)2 anti-IgM at 4°C for 60 min to label and cross-link the BCR and warmed to 37°C for increasing lengths of time up to 60 min. The cells were washed and lysed at 4°C in buffer containing 1% Triton X-100 detergent, conditions under which the lipid rafts are insoluble. Cell lysates were subjected to discontinuous density gradient centrifugation, gradient fractions were collected, and the amount of fluoresceinated Ab in the fractions was determined (Fig. 1). For CH27 cells immediately following cross-linking, a significant portion ∼40% of the BCR was present in the lipid raft fraction of the gradient, fraction 2. The remainder of the FL-F(ab′)2 anti-IgM was found in the soluble membrane fraction 4 and in the intermediate soluble fraction, fraction 3. The amount of the BCR in the raft fractions decreased after 10–20 min at 37°C such that by 60 min <15% of the BCR was in the rafts.
The WEHI-231 cells showed a strikingly different response to BCR cross-linking. The presence of the cross-linking reagent altered only slightly the position of the BCR on the plasma membrane with ∼3% of the BCR localized to the raft region 10–30 min after warming to 37°C. The vast majority of the FL-F(ab′)2 anti-IgM remained in the soluble fraction, fraction 4. These results indicate that the BCR on WEHI-231 cells does not significantly translocate into Triton X-100 insoluble lipid rafts upon cross-linking.
To investigate the position of the BCR in the earliest times after BCR cross-linking the cells were incubated with 125I-labeled Fab anti-Ig for 15 min at 4°C to label the BCR. The BCR was cross-linked using anti-Ig or an isotype-matched control Ig, and the cells were incubated at 37°C for 1, 10, or 30 min. The rafts were isolated, and the amount of 125I-Fab anti-Ig was determined. No translocation of the BCR into lipid rafts was detected in WEHI-231 cells, while a significant fraction of the BCR in CH27 cells was detected in the raft fraction at 1 min following cross-linking, which increased over the 30-min incubation at 37°C (Fig. 2). Thus, translocation of the BCR is not observed at any time following BCR cross-linking in WEHI-231 cells.
The gradient fractions in Fig. 2 were further analyzed by SDS-PAGE and immunoblotting. Fraction 2 in CH27 cells and WEHI-231 cells contained the ganglioside GM1, a raft component, detected using the GM1-specific ligand CTB (Fig. 3). GM1 was also detected in the soluble fraction, which likely reflects the efficiency of raft membrane isolation under the present conditions. Cross-linking the BCR did not alter the distribution of GM1 in the density gradients, indicating that raft isolation was similar under the different cross-linking conditions (data not shown). The position of the BCR μ-chain and Igα in the gradient fractions, detected by immunoblotting, correlated with the presence of the 125I-Fab anti-IgM shown in Fig. 2. In untreated CH27 cells, the BCR μ-chain and Igα were found exclusively in the soluble membrane fraction (Fig. 3). The addition of the cross-linking reagent resulted in the movement of the BCR μ-chain and Igα into the raft fractions. The maximum movement was detected 10 and 30 min after cross-linking. In contrast, the position of the BCR μ-chain and Igα in WEHI-231 cells in the soluble membrane fraction was not altered by the addition of the cross-linking reagent. Both the μ-chain and Igα were detected only in the soluble fractions. Taken together, these results indicate that the plasma membrane of WEHI-231 contains GM1-positive, Triton X-100 insoluble, membrane microdomains, but these do not stably accommodate the BCR upon cross-linking.
Cross-linking the BCR in mature B cells leads to the phosphorylation of a number of lipid raft-associated proteins including Lyn and Igα (7). The response to BCR cross-linking in WEHI-231, CH27, and a second mature IgG-expressing cell line, A20, was determined. To do so, cells were treated as described in Fig. 2 to cross-link the BCR, and the raft and soluble fractions were analyzed by immunoblotting, probing with the phosphotyrosine-specific mAb RC20H.
In the absence of cross-linking, there was a constitutive low level of phosphotyrosine-containing proteins in the raft membrane fractions of the WEHI-231, CH27, and A20 cells (Fig. 4). Immediately upon cross-linking, the number and intensity of the phosphotyrosine containing proteins increased dramatically in the raft fractions of the CH27 and A20 cells. The phosphoprotein pattern was most intense immediately upon cross-linking and decreased upon warming to 37°C for 1 and 10 min. Significantly, the pattern of tyrosine-phosphorylated proteins in the WEHI-231 raft fractions did not change upon BCR cross-linking. However, the number and intensity of the phosphorylated proteins increased dramatically in the soluble fractions of WEHI-231 cells following treatment with the BCR cross-linking reagent (Fig. 4), indicating that the BCR cross-linking triggered a response in the WEHI-231 cells. A similar pattern was observed in the soluble fractions of both the A20 and CH27 cells. In addition to the induction of protein phosphorylation, treatment of WEHI-231 cells with F(ab′)2-anti-IgM induced apoptosis in ∼15% of the cells by 24 h (Table I), which increased to ∼50% by 48 h (data not shown) but had no effect on CH27 cells, as previously described (14, 15). Taken together, these results provide evidence that cross-linking the BCR on WEHI-231 cells induces signaling resulting in protein tyrosine phosphorylation and apoptosis.
|Treatment .||Percent Apoptosis .||.|
|.||WEHI-231 .||CH27 .|
|Control||3.4 ± 0.1||3.6 ± 0.7|
|Anti-BCR||14.7 ± 0.3||3.1 ± 0.7|
|Treatment .||Percent Apoptosis .||.|
|.||WEHI-231 .||CH27 .|
|Control||3.4 ± 0.1||3.6 ± 0.7|
|Anti-BCR||14.7 ± 0.3||3.1 ± 0.7|
Cells (1–3 × 105 per ml) were incubated in 10% culture medium containing anti-IgM (10 μg/ml) or no Ab (control) at 37°C, in 5% CO2 for 24 h. The cells were labeled with FITC-annexin V and propidium iodide and subjected to flow cytometry. Cells that labeled positively for both annexin V and propidium iodide were considered to be apoptotic. The data shown is an average of three experiments.
In mature B cells, the BCR that translocates into the lipid rafts following cross-linking is subsequently internalized from the rafts into the cell, a process that plays a role both in receptor down-regulation and Ag targeting (7). To follow BCR internalization, cells were incubated with 125I-Fab anti-Ig at 4°C in the presence or absence of the cross-linking reagent F(ab′)2 anti-Ig. The cells were washed and warmed to 37°C for increasing lengths of time. At the end of each time point, the radioactivity released from the cells, on the cell surface and internalized, was determined as previously described (16). In the absence of the cross-linking reagent, the CH27 and A20 cells show a constitutive level of internalization that reaches ∼40% by 15 min for A20 cells and 25% by 40 min for CH27 cells (Fig. 5). Addition of the F(ab′)2 anti-Ig results in a more rapid internalization of the BCR, reaching maximal levels of nearly 60% within 10 min for A20 cells and 35–40% for CH27 cells. In contrast, in WEHI-231 cells the internalization of 125I-Fab anti-Ig is unaffected by the addition of the cross-linking reagent. The WEHI-231 cells show a constitutive internalization of the 125I-Fab-anti-Ig of ∼25% by 40 min, and the rate and amount of internalization are unchanged by BCR cross-linking. Thus, the BCR expressed on WEHI-231 cell surfaces does not undergo accelerated internalization upon cross-linking, suggesting that raft-dependent BCR internalization may not function in WEHI-231 cells.
Taken together, the results presented provide evidence that the behavior of the BCR expressed by the immature B cell line WEHI-231 is strikingly different from that of mature B cells. The divergent behavior of the BCR in mature and immature B cells as represented by the WEHI-231 cells may contribute in part to the mechanisms underlying the differences in signaling outcomes. At present the mechanisms that might account for the failure of the immature BCR to translocate into lipid rafts is not known and may await a better understanding of the forces that drive translocation of the BCR into lipid rafts in mature B cells. Our recent characterization of the translocation of the mature BCR into rafts showed that translocation is independent of receptor phosphorylation (P. C. Cheng and S. K. Pierce, unpublished observations). Similar results were previously obtained by Field et al. (17) for the IgE receptor, which lead these authors to propose that the translocation of the IgE receptor into rafts was the initiating event in the signal transduction cascade. Thus, the lipid rafts in mature and immature B cells may not be identical in their ability to allow translocation or stable residency of the BCR within lipid rafts.
Weintraub et al. (18) recently provided evidence that the BCR in tolerant B cells is not efficiently translocated into lipid rafts following cross-linking. The tolerant B cells were obtained from double-transgenic mice expressing a hen egg lysozyme (HEL)-specific BCR and soluble HEL. The HEL-specific B cells in these transgenic mice presumably encountered HEL at an immature stage, which resulted in induction of tolerance. The results presented here suggest the possibility that the exclusion from the lipid rafts is not a repercussion of tolerance induction but rather a reflection of the position of the BCR in the plasma membrane of immature B cells at the time of Ag contact. It will be of interest to determine whether the lipid rafts in B cells of other discrete developmental and differentiated stages function similarly.
This work supported by grants from the National Institute of Allergy and Infectious Diseases (AI 27957, AI18939, and AI40309; to S.K.P.).
Abbreviations used in this paper: BCR, B cell Ag receptor; CTB, cholera toxin B subunit; HEL, hen egg lysozyme.