JBC Advanced Glycation Endproducts

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Benlagha, K.
Right arrow Articles by Lassoued, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Benlagha, K.
Right arrow Articles by Lassoued, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

J Biol Chem, Vol. 274, Issue 27, 19389-19396, July 2, 1999


Modifications of Igalpha and Igbeta Expression as a Function of B Lineage Differentiation*

Kamel BenlaghaDagger , Paul Guglielmi§, Max D. Cooperparallel , and Kaïss LassouedDagger **

From the Dagger  Laboratoire d'Immunopathologie, Institut d'Hématologie, Hôpital Saint-Louis, 75475 Paris Cédex 10, France, § Institut de Génétique Moléculaire, 34293 Montpellier cedex 05, France, and  Developmental and Clinical Immunology, Departments of Medicine, Pediatrics, Pathology and Microbiology, Howard Hughes Medical Institute, University of Alabama at Birmingham, Birmingham, Alabama 35294

    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Transcription of the mb1 and B29 genes is initiated when lymphoid progenitors enter the B cell differentiation pathway, and their transmembrane Igalpha and Igbeta products constitute essential signaling components of pre-B and B cell antigen receptors. We analyzed Igalpha /Igbeta biosynthesis, heterogeneity, and molecular interactions as a function of human B lineage differentiation in cell lines representative of the pro-B, pre-B, and B cell stages. All B lineage representatives produced a 36-kDa Igbeta form and three principal Igalpha forms, transient 33/40-kDa species and a mature 44-kDa glycoprotein. Deglycosylation revealed a major Igalpha core protein of 25 kDa and a minor 21-kDa Igalpha protein, apparently the product of an alternatively spliced mRNA. In pro-B cells, the Igalpha and Igbeta molecules existed primarily in separate unassembled pools, exhibited an immature glycosylation pattern, did not associate with surrogate light chain proteins, and were retained intracellularly. Their unanticipated association with the Lyn protein-tyrosine kinase nevertheless suggests functional potential for the Igalpha /Igbeta molecules in pro-B cells. Greater heterogeneity of the Igalpha and Igbeta molecules in pre-B and B cell lines was attributable to increased glycosylation complexity. Finally, the Igalpha /Igbeta heterodimers associated with fully assembled IgM molecules as a terminal event in B cell receptor assembly.

    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Membrane-bound immunoglobulin (Ig) molecules are non-covalently bound to the transmembrane Igalpha (CD79a) and Igbeta (CD79b) proteins, respectively the products of the mb1 and B29 genes, to form the B cell antigen receptor (BCR)1 complex (1). Igalpha /Igbeta heterodimers are also integral components of pre-BCR complexes composed of surrogate light chain (psi LC) and µ heavy chains (HC) on the surface of pre-B cells (2-8). Ligation of BCR and pre-BCR initiates cytoplasmic signals via the Igalpha and Igbeta molecules whose cytoplasmic domains contain immunoregulatory tyrosine-based activation motifs. BCR aggregation thereby promotes interaction with protein-tyrosine kinases and resultant immunoregulatory tyrosine-based activation motif phosphorylation, hydrolysis of phosphatidylinositol, sustained intracellular calcium elevation (9-12), and the activation of multiple signaling pathways (13-15).

Expression of Igalpha and Igbeta transcripts begins very early in B lineage differentiation prior to the onset of DH-JH rearrangements in the µHC locus (16, 17), and Igbeta -deficient mice are unable to generate µHC-producing pre-B cells (18). Surprisingly, B cell development in Igbeta -/- mice appears to be compromised as early as the pro-B stage when VH-DJH rearrangements are occurring, thereby suggesting an Igbeta role in B lymphopoiesis even prior to µHC synthesis. Although pro-B cell lines from humans also produce Igalpha and Igbeta , expression of these as components of cell surface receptors has not been demonstrable (4). Correspondingly, the psi LC in human pro-B cells were found exclusively in the endoplasmic reticulum and early Golgi compartments, where they transiently associated with 40-, 60-, and 98-kDa proteins before undergoing intracellular degradation (7, 8). In murine pro-B cells, however, Igalpha /Igbeta heterodimers have recently been found on the cell surface, perhaps in association with calnexin (19, 20).

The physiological role of Igalpha and Igbeta during the earliest stages in B lineage differentiation thus remains unclear, and may differ in mice and humans. In this analysis of human B lineage cells, we have compared Igalpha and Igbeta expression, heterogeneity, and molecular association in pro-B cells versus their more mature pre-B and B cell offspring. The results reveal a remarkable progressive complexity of the Igalpha and Igbeta glycoproteins during B lineage differentiation, an unanticipated intracellular association with a Src family protein-tyrosine kinase in pro-B cells, and late stage Igalpha /Igbeta union with assembled IgM molecules to form the BCR on B cells.

    EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Antibodies-- Mouse monoclonal antibodies (mAbs) included SA-DA 4;4 (gamma 1kappa ) anti-human µH chain (21), CB3-1 (gamma 1kappa ), and CB3-2 (gamma 1kappa ) anti-human Igbeta (4), HM57 (gamma 1kappa ) anti-human Igalpha (22), SLC1 (gamma 1kappa ) anti-human psi LC (8), 4G10 (gamma 1kappa ) anti-phosphotyrosine (Upstate Biotechnology Inc., Lake Placid, NY), 4D10 (gamma 2a) anti-human Syk (Santa Cruz Biotechnology, Inc. CA), and Fyn (gamma 1kappa ) anti-human Fyn (Santa Cruz Biotechnology). The JH3 (gamma 1kappa ) anti-human Ig idiotype (23) and CT4 (gamma 1kappa ) anti-chicken CD4 (24) mAbs were used as controls. In immunoprecipitation experiments, mouse mAbs were either directly coupled to Sepharose 4B beads or incubated with rat anti-mouse kappa LC-coupled beads (Interchim, France). Rabbit antibodies to human Syk (a kind gift of Urlich Blanc, Institut Pasteur, Paris, France) and Lyn (Upstate Biotechnology Inc.) were also employed, with rabbit gamma -globulins (Pentex, Miles Laboratory, Kankakee) serving as a control. Phycoerythrin-conjugated goat anti-mouse Ig was obtained from Southern Biotechnology Associates (Birmingham, AL).

Cells-- Human cell lines included the RS4;11 (25) and Nalm16 (26) pro-B cells, the 697 (27) and Nalm6 (26) pre-B cells, and Ramos (28) B cells. These were maintained in stationary culture in RPMI 1640 medium supplemented with L-glutamine (2 mM), penicillin (100 units/ml), streptomycin (100 mg/ml), and 10% fetal calf serum at 37 °C in 5% CO2.

Cell Surface Biotinylation and Biosynthetic Protein Labeling-- Viable cells (5 × 107) washed twice in PBS were incubated with 1 mg/ml biotin (NHS-sulfonyl biotin, Pierce) in PBS for 1 h at 4 °C. Cells with biotinylated cell surface proteins were washed once in chilled RPMI 1640 and twice in PBS before lysis in 1% digitonin or 0.5% Triton X-100 lysis buffer. Cells (1.5 × 108) were also metabolically labeled with [35S]Met and [35S]Cys (400 µCi each) for 6 h, then washed and lysed. For pulse labeling and chase analysis, cells (1-2 × 108) were preincubated in Met- and Cys-free RPMI 1640 for 2 h to deplete internal pools, and then labeled with 300-500 µCi of both [35S]Met and [35S]Cys for 15 min. Labeling was terminated by addition of 100-fold excess of cold Met and Cys. The pulsed cells were incubated for various intervals before harvesting.

Immunoprecipitation-- Labeled cells were harvested and lysed in 1% Nonidet P-40, 1% Triton X-100, or 1% digitonin lysis buffer. Nuclei were sedimented at 10,000 × g for 20 min, and the supernatants used for immunoprecipitation. After successive incubations with bovine serum albumin and IgG-coupled Sepharose 4B beads, the precleared lysates were incubated with beads coupled with test or control antibodies. Bound materials were eluted with Laemmli's sample buffer (29) and resolved by SDS-PAGE, using 10% or 12% acrylamide. In reprecipitation experiments, digitonin-treated cell lysates were incubated with anti-Igbeta or anti- Igalpha mAbs, and bound materials resuspended in 500 µl of 1% Nonidet P-40 lysis buffer were incubated with anti-Lyn or anti-Syk antibodies. Likewise, materials bound by anti-Lyn antibodies were eluted and immunoprecipitated with anti-Igbeta antibodies. For two-dimensional gel electrophoretic analysis, immunoprecipitates were separated initially on nonreducing SDS-polyacrylamide (9-10%) gels. The lanes were then excised, equilibrated for 45 min in SDS sample buffer containing beta -mercaptoethanol, and rotated 90° before electrophoresis in the second dimension on SDS-polyacrylamide (10%) gels under reducing conditions.

Western Blots-- Anti-Igalpha , anti-Igbeta , and anti-µ immunoprecipitates were resolved on one-dimensional SDS-PAGE and two-dimensional diagonal gels, blotted onto nitrocellulose membrane (Schleicher & Schuell), before incubation with anti-µ, anti-Igbeta , or anti-Igalpha mAbs. In some experiments, the anti-Igalpha and anti-Igbeta precipitates were submitted to a second immunoprecipitation with an anti-phosphotyrosine mAb; the eluted material was electrophoresed, blotted, and reincubated with the anti-phosphotyrosine mAb. The blots were developed with the ECL chemiluminescence system (Amersham Pharmacia Biotech) using a horseradish peroxidase-conjugated goat anti-mouse Ig (Sigma). In cell surface biotinylation experiments, immunoprecipitates were resolved by SDS-PAGE, blotted as described above, then incubated with a horseradish peroxidase-conjugated streptavidin (Sigma) and revealed with the ECL method.

Deglycosylation-- Immunoadsorbed proteins were treated with endoglycosidase H (endo H), endoglysosidase F (endo F), N-glycanase, O-glycanase, or combinations of these enzymes for 18 h at 37 °C, according to the manufacturer's instructions (Roche Molecular Biochemicals), before elution with Laemmli's buffer. Fetuin protein was used as a control for optimal digestion.

In Vitro Kinase Assay-- Nalm16 cells were washed twice in PBS and lysed in 1% digitonin, 25 mM Hepes, pH 7.2, 150 mM NaCl, 5 mM KCl, 5 mM EDTA, 1 mM orthovanadate, and protease inhibitors before lysate immunoprecipitation with anti-Igalpha , anti-Igbeta , or control mAbs. Bound materials were washed three times with lysis buffer and once with kinase buffer (25 mM Hepes, pH 7.2, 150 mM NaCl, 5 mM KCl, 5 mM MnCl2, 5 mM MgCl2, 1% digitonin) before resuspension in 50 µl of kinase buffer containing 1 mCi of [gamma -32P]ATP (>4000 Ci/mmol, ICN Biomedicals, Orsay, France). After a 20-min incubation at room temperature, beads were washed three times with lysis buffer, then either eluted in 50 µl of SDS sample buffer and boiled for 5 min or submitted to reprecipitation as follows: elution in 50 µl of 0.5% SDS lysis buffer instead of digitonin, incubation for 10 min at 60 °C, 10-fold dilution in 500 µl of 1% Triton X-100, then immunoprecipitation. In one series of experiments, Nalm16 cells were incubated with the CB3-1 anti-Igbeta antibody (30 µg/ml) or a control mAb (gamma 1kappa isotype) for various intervals (0, 1, and 5 min) before exposure to 1% Triton X-100 lysis buffer containing 5 mM EDTA, 1 mM orthovanadate, and protease inhibitors. Cell lysates were electrophoresed, blotted, and probed with an anti-phosphotyrosine mAb. Alternatively, lysates of the anti-Igbeta -stimulated pro-B cells were immunoprecipitated with an anti-phosphotyrosine mAb, and the immunoprecipitate was electrophoresed, blotted, and probed with the anti-phosphotyrosine mAb.

Northern Blots, Reverse Transcription-Polymerase Chain Reactions (RT-PCR), and DNA Sequencing-- Total RNA was prepared using the guanidinium isothiocyanate method, electrophoresed on 7% formaldehyde, 1% agarose gels, and blotted onto nitrocellulose filters (Schleicher & Schuell). Northern blots were hybridized with random-primed 32P-labeled DNA fragments. Igalpha and Igbeta transcripts were further analyzed by RT-PCR. The first strand cDNA synthesis employed Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.) for 50 min at 42 °C in the presence of 5 µg of total RNA and 100 ng of primers complementary to the coding sequence of Igalpha mRNA (Igalpha -R, 5'- CTGGACATCTCCTATGTTGA-3') or Igbeta mRNA (Igbeta -R1, 5'-CTCCTGGCCTGGGTGCTCAC-3'). The cDNAs were amplified in the presence of both the 5' and 3' primers with Taq polymerase (Cetus, Emeryville, CA) through 28 cycles involving denaturation at 92 °C for 1 min, annealing at 55 °C for 1 min, and elongation at 72 °C for 1 min. Igalpha forward and reverse primers were Igalpha -F (5'-ATGCCTGGGGGTCCAGGAGTC-3') and Igalpha -R; Igbeta primers were Igbeta -F (5'-ATGGCCAGGCTGGCGTTGTC-3') and Igbeta -R2 (5'- AATGTCCAGGCCCTCGTAAGG-3'). The RT-PCR protocol was essentially that of Hashimoto et al. (30). Amplified PCR products were analyzed on 2.5% agarose gels. DNA fragments corresponding to the different PCR products were eluded from the agarose gel, cloned in the PCR 2.1 vector (Invitrogen, Carlsbad, CA), sequenced using dye terminator chemistry, and analyzed with an automated ABI DNA sequencer (Perkin Elmer, Foster City, CA).

Densitometric Analysis-- Radiolabeled protein intensities were determined relative to cell surface µHC levels using a densitometer (UltraScan XL) and a software program (GelScan XL; Amersham Pharmacia Biotech, Uppsala, Sweden). Background readings of the autoradiographic films were designated zero.

    RESULTS
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

Characterization of Igalpha and Igbeta Molecules in Pro-B Cell Lines-- Four molecules with apparent molecular masses of 44, 40, 36, and 33 kDa were immunoprecipitated with anti-Igalpha antibodies from the metabolically labeled Nalm16 and RS4;11 pro-B cell lines. In contrast, two molecules of 44 and 36 kDa were precipitated by the anti-Igbeta mAb from both pro-B cell lines (Fig. 1A). Analysis of the anti-Igalpha precipitates by two-dimensional gel electrophoresis indicated the covalent linkage of a minor fraction of the available pools of the 44- and 36-kDa species (Fig. 1C, and data not shown), with the major portion of the 44-kDa molecules remaining on the diagonal. Similarly, two-dimensional gel analysis of anti-Igbeta immunoprecipitates indicated that most of the 36-kDa molecules migrated on the diagonal, whereas small fractions were found in off-diagonal positions indicative of their formation of homodimers or heterodimers with the 44-kDa protein (Fig. 1D). Western blot analysis confirmed the Igbeta identity of the 36-kDa molecules (Fig. 1B, lanes 3 and 4). It also indicated reactivity of the 44-, 40-, and 33-kDa molecules with the anti-Igalpha antibody, thus confirming the molecular heterogeneity of Igalpha proteins (Fig. 1B, lanes 1 and 2). Western blot analysis of anti-Igbeta immunoprecipitates resolved on two-dimensional gels indicated that most of the Igbeta molecules exist in a free pool, whereas minor fractions exist as covalently linked heterodimers or homodimers in the Nalm16 pro-B cell line (Fig. 1, D and F). A similar two-dimensional analysis of the anti-Igalpha precipitates showed that only a fraction of the 44-kDa Igalpha species ran off the diagonal (Fig. 1, C and E), thus indicating that the 40- and 33-kDa Igalpha forms do not form covalent bonds with Igbeta .


View larger version (47K):
[in this window]
[in a new window]
  		Fig. 1.   Analysis of Igalpha and Igbeta expression in pro-B cell lines. A, Anti-Igalpha (lanes 1 and 3) and anti-Igbeta immunoprecipitates (lanes 2 and 4) of metabolically labeled pro-B cells. Cells were lysed in 1% digitonin lysis buffer and the immunoprecipitates separated under reducing conditions on a 12% SDS-polyacrylamide gel. Relative molecular masses in kilodaltons are indicated. Western blot analysis (see below) suggests that the higher molecular mass band seen in the anti-Igalpha precipitates primarily represents actin. B, Western blot analysis of Igalpha and Igbeta molecules. Pro-B cells were lysed in a 1% Nonidet P-40 solution, and the cell lysates incubated with anti-Igalpha or anti-Igbeta mAbs. Immunoprecipitates resolved by SDS-PAGE under reducing conditions were transferred to nitrocellulose membranes before Western blot analysis with anti-Igalpha (lanes 1 and 2) and anti-Igbeta mAbs (lanes 3 and 4). Western blots were developed with horseradish peroxidase-conjugated goat anti-mouse Ig antibodies using a chemiluminescence system. C and D, two-dimensional gel analysis of Igalpha and Igbeta molecules produced by Nalm16 pro-B cells. Digitonin lysates of metabolically labeled cells were incubated with monoclonal antibodies against Igalpha (C) or Igbeta (D). Immunoprecipitates were resolved in SDS-PAGE under nonreducing conditions in the first dimension and under reducing conditions in the second dimension. E and F, Igalpha and Igbeta identification by Western blot analysis of pro-B cell lysates in two-dimensional gels. Anti-Igalpha or anti-Igbeta precipitates from digitonin lysates of unlabeled RS4;11 pro-B cells were separated by two-dimensional gel electrophoresis, transferred to nitrocellulose membranes, and developed with anti-Igalpha (E) or anti-Igbeta (F) antibodies as indicated above. Covalent association of a minor fraction of the 44-kDa Igalpha molecules and the 36-kDa Igbeta molecules in the pro-B cells is indicated by their off-diagonal position. The arrow indicates the trace amounts of Igbeta components of homodimeric units. Note that the photographic magnification of E and F differs slightly from that of C and D.

Pulse-chase analysis of the RS4;11 and Nalm16 pro-B cells indicated that newly synthesized Igalpha and Igbeta proteins exist initially as completely separate pools with limited Igalpha and Igbeta association occurring thereafter (Fig. 2A, and data not shown). The levels of newly synthesized 40- and 33-kDa Igalpha variants progressively declined with estimated half-lives of less than 30 min, whereas the levels of 44-kDa Igalpha molecules were maintained over the 3-h observation period (Fig. 2B). Notably, neither the anti-Igalpha nor the anti-Igbeta immunoprecipitates of pro-B cell lysates contained psi LC proteins, and the Igalpha and Igbeta molecules could not be detected on the surface of the Nalm16 pro-B cells (data not shown).


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 2.   Analysis of Igalpha and Igbeta biosynthesis and glycosylation status in pro-B cells. Pulse-chase analysis of Igalpha and Igbeta in RS4;11 pro-B cells (A and B). Cells were pulse-labeled with [35S]Met and [35S]Cys, chased with cold amino acids for various time intervals, and then lysed before incubation with antibodies against Igalpha or Igbeta . Immunoprecipitates were analyzed by SDS-PAGE under reducing conditions. A, Igalpha and Igbeta molecules are unassociated at time 0. B, Analysis of the biosynthetically labeled 44-, 40-, and 33-kDa Igalpha molecules at later time points, by determination of relative autoradiographic intensities, indicates the relative stability of the 44-kDa Igalpha form and the transient nature of the 33/40-kDa species. C, 24.5-kDa Igbeta core protein and 21.5/25-kDa Igalpha core proteins are revealed by deglycosylation. Nonlabeled RS4;11 cell lysates were subjected to immunoprecipitation with anti-Igalpha or anti-Igbeta anti-bodies, the immunoprecipitates treated with endo F, and the deglycosylated Igalpha and Igbeta proteins identified by immunoblots as described in Fig. 1.

These results indicate that the Igalpha and Igbeta proteins produced by human pro-B cell lines exist largely in unassembled free pools, although a limited fraction of these molecules form Igalpha /Igbeta heterodimers and trace amounts of Igbeta homodimers. While Igbeta appears to be expressed as a single species, Igalpha is expressed in three sizes, the smaller two of which apparently represent immature forms that are relatively transient in nature and do not form disulfide-linked dimers.

Igalpha Protein Heterogeneity in Pro-B Cell Lines-- Igalpha and Igbeta heterogeneity may vary as a function of B lineage differentiation stage and the expressed immunoglobulin isotype, in part due to differential glycosylation (4, 31-40). Experiments were therefore conducted to determine whether the three Igalpha species in pro-B cells reflect variably glycosylated forms or other modifications of protein structure. When lysates of metabolically labeled RS4;11 cells were precipitated with anti-Igalpha or anti-Igbeta antibodies and the bound material treated with endo F or with N-glycanase and endo F, anti-Igalpha precipitates were resolved as two bands of 25 and 21.5 kDa, whereas anti-Igbeta precipitates yielded a single band of 24.5 kDa (Fig. 2C, and data not shown). Western blot analysis of the anti-Igalpha or anti-Igbeta precipitates confirmed the Igalpha nature of the 25- and 21.5-kDa bands, and the single band of 24.5-kDa Igbeta forms (Fig. 2C). When the anti-Igalpha and anti-Igbeta immunoprecipitates of unlabeled RS4;11 and Nalm16 pro-B cell lysates were treated with endo H, immunoblot analysis again indicated 25- and 21.5-kDa Igalpha species and a single 24.5-kDa Igbeta band (data not shown). This endo H sensitivity pattern suggests restriction of the Igalpha and Igbeta glycoproteins to the endoplasmic reticulum and early Golgi in human pro-B cells.

Igalpha and Igbeta Transcript Heterogeneity-- The fact that the molecular mass of the Igalpha core protein predicted from complete transcription of the human mb1 gene is 24.5 kDa suggested that the 21.5-kDa deglycosylated molecule could be generated by post-transcriptional modification. Mb-1 (Igalpha ) and B29 (Igbeta ) transcripts in the Nalm16 and RS4;11 pro-B cell lines were therefore analyzed by RT-PCR. The Igalpha reactions yielded a major 660-bp DNA fragment and a minor 550-bp DNA fragment (Fig. 3A). Sequence analysis indicated that the 660-bp PCR product corresponds to the full-length Igalpha transcript, whereas the 550-bp fragment reflects the amplification of an alternatively spliced Igalpha mRNA. This spliced form could be attributed to the use of a cryptic splice site located 186 bp after the start of the second exon, which is joined to the normal splice acceptor site of the third exon. The resulting transcript, which is devoid of 114 bp of the normal sequence, would maintain the same reading frame as the full-length transcript, and therefore could encode an Igalpha protein lacking part of the extracellular domain. This truncated protein would not be expected to contain the cysteine residue involved in covalent association with Igbeta , but would maintain the transmembrane and cytoplasmic portions as well as four N-linked glycosylation sites.


View larger version (51K):
[in this window]
[in a new window]
  		Fig. 3.   Analysis of mb1 (Igalpha ) and B29 (Igbeta ) transcripts in cell lines representative of different stages in human B lineage differentiation. Igalpha (A) and Igbeta (B) RT-PCR products separated on a 2.5% agarose gel. The sequences determined for the indicated 660- and 550-bp Igalpha cDNAs and for the 600- and 290-bp Igbeta cDNAs matched the mRNA regions indicated by closed bars in the schematics and described under "Experimental Procedures." Hybridization experiments and sequence analysis indicated that the 500-bp band observed in some Igbeta reactions did not correspond with an Igbeta gene product.

Igbeta reactions produced major 600-bp and minor 290-bp DNA amplification products (Fig. 3B). The sequence of the larger fragment contained all six CD79b exons, whereas the smaller product reflected a complete loss of the third exon. The 105 codon truncation maintains a correct reading frame in the 3' portion of the short Igbeta transcripts that could encode for an 11-kDa protein. The predicted truncated Igbeta protein would lack all N-linked glycosylation sites and the ability to form disulfide-bonded heterodimers with Igalpha . Antibodies to the C-terminal portion of Igbeta are needed to determine whether this truncated Igbeta is expressed in pro-B cells. The present RT-PCR and sequence analyses of pro-B cell lines thus agree with those previously reported for human pre-B and B cell lines (41-45).

Association of Igalpha and Igbeta Molecules with Protein-tyrosine Kinases in Pro-B Cell Lines-- Antigen ligation of the BCR complex on B cells leads to interaction of the Igalpha /Igbeta subunits with Src family kinases (Lyn, Fyn, Blk, Hck, and Lck) and Syk kinase to initiate cell activation (46-56). Pre-BCR interaction with Src kinases has also been demonstrated (12, 57). It was therefore of interest to determine whether interaction of the Igalpha /Igbeta molecules with these protein-tyrosine kinases may occur at the pro-B cell stage. Both Lyn (53/56 kDa) and Syk (70 kDa) tyrosine kinases were identified in digitonin lysates of metabolically labeled RS4;11 pro-B cells, although the level of background radioactivity did not permit clear resolution of their association status with Igalpha and Igbeta molecules. When the anti-Lyn and anti-Syk eluates were submitted to a second immunoprecipitation with the anti-Igbeta antibody, Igbeta was detected in the anti-Lyn immunoprecipitate (Fig. 4A, lane 1) but not in the anti-Syk immunoprecipitate (data not shown). A band of 44 kDa, likely Igalpha , was also coprecipitated with anti-Lyn. Conversely, p53/56 Lyn was identified in anti-Igbeta precipitates (Fig. 7A, lane 2), whereas Syk was not (data not shown).


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4.   Igalpha /Igbeta association with Lyn in pro-B cells. A, digitonin lysates of metabolically labeled RS4;11 pro-B cells were incubated with anti-Lyn antibody and the eluate submitted to a second precipitation with anti-Igbeta (lane 1). Conversely, the anti-Igbeta eluate was reprecipitated with anti-Lyn (lane 2). B, kinase activity of pro-B cells phosphoproteins associated with Igalpha . Pervanadate-treated (P+) and unstimulated (P-) Nalm16 cells were examined for kinase activity; anti-Igalpha (lanes 3 and 4) and anti-Lyn (lanes 1 and 2) precipitates of digitonin cell lysates were subjected to an in vitro kinase reaction in the presence of [gamma -32P]ATP. C, analysis of the tyrosine phosphorylation status of Igalpha -associated Lyn in pro-B cells. Whole cell lysates (L, lanes 1 and 2) and anti-Igalpha precipitates (lanes 3 and 4) of pervanadate (P)-stimulated (+) and unstimulated (-) Nalm16 cells were separated by SDS-PAGE before immunoblotting with an anti-phosphotyrosine antibody.

An in vitro kinase assay was conducted using unstimulated and pervanadate-treated Nalm16 pro-B cells. The anti-Igalpha antibody coprecipitated a faint 53/56-kDa doublet with kinase activity in unstimulated Nalm16 cells, and pervanadate treatment, employed to alter the kinase/phosphatase equilibrium to favor activation of the tyrosine kinases, strongly enhanced the Lyn 53/56 signal (Fig. 7B). The 53/56 doublet was also revealed by an anti-phosphotyrosine mAb in the anti-Igalpha immunoprecipitates of Nalm16 pro-B cells preincubated with pervanadate (Fig. 4C). An additional 44-kDa band, likely Igalpha , was also seen. In contrast, Syk could not be identified in anti-Igalpha /Igbeta precipitates of pervanadate-treated pro-B cells. Similarly, Fyn, which was also abundant in pro-B cells, was not detected in association with Igalpha /Igbeta .

The above data suggest that Lyn preferentially associates with the minor population of Igalpha /Igbeta heterodimers within pro-B cells. In order to examine the possibility that an otherwise undetectable level of cell surface Igalpha /Igbeta molecules accounted for the Lyn association, we examined the tyrosine phosphorylation status of proteins in pro-B cells before and after treatment with a known stimulatory antibody against an extracellular Igbeta epitope. No differences were observed in the tyrosine phosphorylation status of Lyn or other proteins were observed within 5 min following the anti-Igbeta treatment of the pro-B cells (data not shown), thereby mitigating against the possibility of Lyn association with pro-B cell surface Igalpha /Igbeta .

Comparison of Igalpha and Igbeta Expression in Pre-B and B Cell Lines-- Two-dimensional gel analysis of anti-Igalpha and anti-Igbeta immunoprecipitates indicated a progressive increase in heterogeneity of the Igalpha /Igbeta heterodimers in pre-B and B cells (Figs. 5 and 6). Additional minor Igalpha and Igbeta forms of lower molecular weights were also revealed by this Western blot analysis. Igalpha and Igbeta existed in major and minor heterodimeric forms in pre-B cell lines as well as in a B cell line. Comparative two-dimensional gel analysis of the Ramos B cell line also revealed prominent fan-like extensions of the major Igalpha and Igbeta units (Fig. 5, C-E), the identity of which was confirmed by Western blot analysis (Fig. 6, C and D). In both the pre-B and B cells, µHC were coprecipitated with the Igalpha and Igbeta molecules. Comparative analysis of the anti-µHC precipitate (Fig. 5E) with the anti-Igalpha and anti-Igbeta precipitates (Fig. 5, C and D) indicates that the (µHC)2/(lambda LC)2 complexes could be coprecipitated with antibodies against the Igalpha and Igbeta molecules, while the µHC/lambda LC subunits could not.


View larger version (73K):
[in this window]
[in a new window]
  		Fig. 5.   Analysis of the Igalpha and Igbeta molecules produced by pre-B and B cell lines. Anti-Igalpha (A and C), anti-Igbeta (B and D), and anti-µ (E) immunoprecipitates of metabolically labeled 697 pre-B cells (A and B) and Ramos B cells (C-E) were assessed by two-dimensional gel electrophoresis. Note that the µHC and LC components representing the unassembled µHC/LC subunits (small arrows) are seen in the anti-µHC immunoprecipitate (E), whereas only components of the assembled (µHC)2/(LC)2 units (large arrows) are evident in the anti-Igalpha and anti-Igbeta immunoprecipitates of pre-B and B cells. The surrogate light chain components (16/18-kDa Vpre-B, and 22-kDa lambda 5/14.1) in the pre-B cells are not shown in A and B.


View larger version (64K):
[in this window]
[in a new window]
  		Fig. 6.   Analysis of Igalpha /Igbeta association and heterogeneity in human pre-B and B cells. Anti-Igalpha (A and C) and anti-Igbeta (B and D) immunoprecipitates of lysed, unlabeled 697 pre-B (A and B) and Ramos B cells (C and D) were resolved by two-dimensional SDS-PAGE and blotted with antibodies against Igalpha and Igbeta as described in Fig. 1. Note the greater heterogeneity of Igalpha (C) and Igbeta (D) molecules in the B cells, and the subfractions of smaller Igalpha and Igbeta species present in both the pre-B and B cells.

EndoF digestion of the anti-Igalpha precipitates from pre-B and B cells revealed core proteins similar to those observed for pro-B cells, and truncated mb1 and B29 transcripts were likewise identified in pre-B and B cells (Fig. 3). A minor population of Igbeta homodimers was also detected in pre-B and B cells, but these Igbeta homodimers were not seen in the anti-µ precipitates (Fig. 5E), anti-Igalpha precipitates (Fig. 5, A and C), or on the cell surface (Fig. 7C). Pulse-chase analysis of anti-Igalpha precipitates from metabolically labeled cells indicated that Igalpha /Igbeta association is initiated immediately after their biosynthesis in pre-B and B cells (Fig. 7A). However, completion of the complex Igalpha and Igbeta glycosylation process required more than 2 h after biosynthesis, as illustrated in pre-B cells in Fig. 7B. The final heterogeneity of cell surface Igalpha /Igbeta components was found to be indistinguishable in the pre-BCR on pre-B cells and BCR on B cells (Fig. 7C).


View larger version (29K):
[in this window]
[in a new window]
  		Fig. 7.   Pulse-chase analysis of Igalpha and Igbeta biosynthesis and association in pre-B and B cell lines and comparative analysis of Igalpha /Igbeta heterogeneity in the pre-BCR and BCR. A, Pre-B and B cell lines, 697 and Ramos, were pulse-labeled with [35S]Met and [35S]Cys and chased with cold Met and Cys for varying time intervals before cell lysis and analysis of anti-Igalpha or anti-Igbeta immunoprecipitates by SDS-PAGE under reducing conditions. The different molecular forms of Igalpha /Igbeta seen in pro-B cells (Figs. 1 and 2) are also present in the pre-B and B cells, in which Igalpha and Igbeta association is already evident at time 0. B, two-dimensional gel analysis of anti-Igalpha immunoprecipitates at various intervals following pulse labeling of the 697 pre-B cells. Note the relative homogeneity of Igalpha and Igbeta molecules as late as 2 h after biosynthesis. C, comparable heterogeneity of Igalpha /Igbeta molecules on the surface of 697 pre-B cells and Ramos B cells. Cells were surface-biotinylated, lysed in digitonin, then incubated with an anti-µ antibody. Immunoprecipitates were resolved by two-dimensional SDS-PAGE.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES

In these studies, we observed a remarkable diversity of the Igalpha and Igbeta glycoproteins, the nature of which is altered as a function of B lineage progression. Igalpha species of 44, 40, and 33 kDa were produced by pro-B, pre-B, and B cell lines, whereas Igbeta was found in a single 36-kDa form. The 40- and 30-kDa Igalpha species represented transient immature forms that did not associate with Igbeta , whereas the 44-kDa Igalpha and 36-kDa Igbeta were found to be relatively stable and to associate with each other to form heterodimers. The formation of Igalpha /Igbeta heterodimers and the complexity of their glycosylation patterns were shown to increase dramatically as a function of B cell differentiation.

Two Igalpha protein backbones of approximately 25 and 21.5 kDa were revealed in deglycosylation studies, and two types of mb1 (Igalpha ) gene transcripts were identified that could account for these. The smaller, less-abundant variant, which lacked 114 bp as a consequence of alternative splicing of the Igalpha mRNA, maintains the same reading frame as the full-length transcript. It has the potential to give rise to a 4-kDa deleted protein that likely corresponds to the truncated Igalpha detected in B lineage cells. The predicted product of this small mb1 transcript would lack the cysteine residue involved in forming disulfide-linked heterodimers with Igbeta , but would maintain many N-linked glycosylation sites. Similarly, two B29 (Igbeta ) transcripts were identified. The smaller one has the potential to encode a 11-kDa protein that could not covalently bind Igalpha and would lack glycosylation sites. All of the representative pro-B, pre-B, and B cell lines expressed identical patterns of Igalpha and Igbeta transcripts, and deglycosylation with N-glycanase revealed the truncated 21-kDa Igalpha molecule in cells representative of each differentiation stage. In mature human B cells, a post-transcriptional regulation of mb1 and B29 gene expression has been suggested. In particular, activation of mature B cells with anti-IgM antibody, interleukin 4, or lipopolysaccharide was shown to induce alternative splicing of mb1 and B29 (45).

The molecular interactions and functional potential of the truncated Igalpha and Igbeta molecules are poorly understood. The predicted amino acid sequences of the truncated human Igalpha and Igbeta proteins suggest they would not covalently associate with each other, and therefore would not be incorporated into either the BCR or pre-BCR complexes. Accordingly, fibroblast co-transfection of human µHC and kappa LC genes with mb1 and B29 variants failed to reconstitute the BCR (45). In the mouse, a C-terminally truncated Igbeta product was detectable only in activated B cells (38). A truncated murine Igalpha product has been found to be preferentially associated with IgD, whereas the predicted full-length Igalpha product associated primarily with IgM (39). In late stage murine pre-B cells, a truncated 23 kDa protein has been identified with a monoclonal antibody against an extracellular Igalpha epitope (5). This suggests that truncated Igalpha and Igbeta proteins may be differentially expressed according to the stage of B lineage differentiation in the mouse. Truncated Igalpha variants may also be secreted by bovine B lymphocytes (58).

A functional potential for the Igbeta homodimers that we observed in human B lineage cells was not revealed in these studies. The Igbeta homodimers were not found to associate with either µHC or Igalpha , nor were they detected on the surface of pre-B and B cells. It is theoretically possible that Igbeta homodimers play an undefined role inside the cell or, less likely, on the cell surface in levels below our detection threshold.

The Igalpha and Igbeta interactions and glycosylation heterogeneity increased progressively in representative pro-B, pre-B, and B cell lines. The Igalpha and Igbeta molecules in pro-B cells were found largely in separate pools, with only a minor fraction forming disulfide-bonded Igalpha /Igbeta heterodimers. In contrast to the Igalpha /Igbeta status in pro-B cells, Igalpha readily associated with Igbeta in pre-B cells and in B cells. Variable glycosylation of the Igalpha /Igbeta heterodimers occurred during their progression through the Golgi to reach the cell surface in association with µHC to form pre-B receptors (pre-BCR) and B cell receptors (BCR), respectively. While the final Igalpha /Igbeta glycosylation spectra in the cell surface pre-BCR and BCR were indistinguishable, a restricted fraction of mature glycosylated molecules was observed in the Igalpha /Igbeta pool in pre-B cells relative to that seen in B cell lines. This striking feature, which is also evident in data obtained in prior studies (4, 6), reflects the relative inefficiency of pre-BCR assembly in pre-B cells compared with BCR assembly in B cells.

Igalpha /Igbeta heterodimers are essential elements in pre-B and B receptor signaling (1, 9-15). They mediate B cell activation by interaction with Syk and Src family tyrosine kinases and also serve as pre-BCR signal-transducing components to promote pre-B cell differentiation and allelic exclusion (59-62). Much less is known about functional Igalpha /Igbeta potential before VDJH gene rearrangement. One important clue, however, is provided by the demonstration that mice lacking in Igbeta exhibit a block at the pro-B cell stage in differentiation prior to the completion of V-DJH rearrangements (18). In contrast, V-DJH rearrangement proceeds normally in mice that have a cµ mutation that prevents membrane-bound µHC expression (63). These observations suggest that the Igalpha /Igbeta molecules may play an important biological role during the pro-B cell stage in differentiation before µHCs are expressed. A recent study suggests that Igalpha /Igbeta heterodimers may be expressed with calnexin on the surface of pro-B cells from RAG-2-deficient mice (20). Ligation of this Igalpha /Igbeta cell surface complex induced rapid, transient phosphorylation of Igalpha and associated tyrosine kinase to promote pro-B differentiation. However, these findings in mice may not be directly applicable to humans, given that Igalpha /Igbeta heterodimers apparently do not reach the cell surface of human pro-B cells. Instead, our analysis of human pro-B cells indicates the intracellular association of Igalpha /Igbeta with Lyn, a member of the Src tyrosine kinase family. The possibility that the association of Igalpha /Igbeta with Lyn is artifactual, occurring after pro-B cell lysis, is unlikely since Igalpha /Igbeta binding to other phosphoproteins was not observed in the pro-B cells. Specifically, phosphorylated Fyn and Syk, which are also abundant in pro-B cell lines, were not found to be associated with Igalpha and Igbeta . Our findings in human pro-B cells thus indicate that, although Igalpha and Igbeta molecules exist primarily in separate pools in the ER/early Golgi compartments, do not associate with surrogate light chains, and fail to reach the cell surface, a minor population of intracellular Igalpha and Igbeta molecules associate with the Lyn protein-tyrosine kinase with potentially functional consequences.

The Igalpha /Igbeta heterodimers are known to be essential for transport of the pre-BCR and BCR components to the cell surface, as well as for their subsequent mediation of signal transduction. Our observation that Igalpha /Igbeta heterodimers in B cells were associated with the (µHC)2/(LC)2 units, but not with the precursor µHC/LC subunits, indicates that the Igalpha /Igbeta -IgM association represents one of the terminal events in BCR assembly. In view of the possibility that the BCR complex includes two Igalpha /Igbeta heterodimers, one Igalpha /Igbeta heterodimer associated with each µHC in the BCR complex (1), it may be important to determine whether when the Igalpha /Igbeta -µHC association occurs immediately before or after assembly of the membrane-bound IgM subunits.

    ACKNOWLEDGEMENTS

We thank Drs. Peter Burrows and Jean-Claude Brouet for review of the manuscript and helpful suggestions, and Ann Brookshire and Muriel Bargis-Touchard for help in preparing the manuscript.

    FOOTNOTES

* This work was supported in part by Grant 6661 from the Association pour la Recherche sur le Cancer; Grant 94043 from the Société Française d'Hématologie, Contrat de Recherche Clinique AP-HP; and Grant AI39816 from NIAID, National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

parallel Investigator of the Howard Hughes Medical Institute.

** To whom correspondence and reprint requests should be addressed: Laboratoire d'Immunopathologie, Institut d'Hématologie, Hôpital Saint-Louis, 1, Avenue Claude Vellefaux, 75475 Paris Cédex 10, France. Tel.: 33-1-53-72-21-64; Fax: 33-1-53-72-21-58; E-mail: kaiss.lassoued{at}sa.u-picardie.fr.

    ABBREVIATIONS

The abbreviations used are: BCR, B cell receptor; LC, light chain; HC, heavy chain; mAb, monoclonal antibody; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; endo, endoglycosidase; RT, reverse transcription; PCR, polymerase chain reaction; bp, base pair(s).

    REFERENCES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
  1. Reth, M. (1992) Annu. Rev. Immunol. 10, 97-121[CrossRef][Medline] [Order article via Infotrieve]
  2. Iglesias, A., Kopf, M., Williams, G. S., Bühler, B., and Köhler, G. (1991) EMBO J. 8, 2147-2156
  3. Ishihara, K., Wood, W. J., Jr., Damore, M., Hermanson, G., Wall, R., and Kincade, P. W. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 633-637[Abstract/Free Full Text]
  4. Nakamura, T., Kubagawa, H., and Cooper, M. D. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 8522-8526[Abstract/Free Full Text]
  5. Ishihara, K., Wood, W. J., Jr., Wall, R., Sakaguchi, N., Michnoff, C., Tucker, P. W., and Kincade, P. W. (1993) J. Immunol. 150, 2253-2262[Abstract]
  6. Brouns, G. S., de Vries, E., van Noesel, C. J. M., Mason, D. Y., van Lier, R. A. W., and Borst, J. (1993) Eur. J. Immunol. 23, 1088-1097[Medline] [Order article via Infotrieve]
  7. Lassoued, K., Illges, H., Benlagha, K., and Cooper, M. D. (1996) J. Exp. Med. 183, 421-429[Abstract/Free Full Text]
  8. Lassoued, K., Nunez, C. A., Billips, L., Kubagawa, H., Monteiro, R. C., LeBien, T. W., and Cooper, M. D. (1993) Cell 73, 73-86[CrossRef][Medline] [Order article via Infotrieve]
  9. Bossy, D., Salamero, J., Olive, D., Fougereau, M., and Schiff, C. (1993) Int. Immunol. 5, 467-478[Abstract/Free Full Text]
  10. Cambier, J. C., Pleiman, C. M., and Clark, C. M. (1994) Annu. Rev. Immunol. 12, 457-468[CrossRef][Medline] [Order article via Infotrieve]
  11. Gold, M. R., and DeFranco, A. L. (1994) Adv. Immunol. 55, 221-295[Medline] [Order article via Infotrieve]
  12. Kuwahara, K., Kawai, T., Mitsuyoshi, S., Matuso, Y., Kikuchi, H., Imajoh-Ohmi, S., Hashimoto, E., Inui, S., Cooper, M. D., and Sakaguchi, N. (1996) Int. Immunol. 8, 1273-1285[Abstract/Free Full Text]
  13. Clark, M. R., Campbell, K. S., Kazalauskas, A., Johnson, S. A., Hertz, M., Potter, T. A., Pleiman, C., and Cambier, J. C. (1992) Science 258, 123-126[Abstract/Free Full Text]
  14. Kim, K., Alber, G., Weiser, P., and Reth, M. (1993) Eur. J. Immunol. 23, 911-916[Medline] [Order article via Infotrieve]
  15. Choquet, D., Ku, G., Cassard, S., Malissen, B., Korn, H., Fridman, W. H., and Bonnerot, C. (1994) J. Biol. Chem. 269, 6491-6497[Abstract/Free Full Text]
  16. Li, Y. S., Wasserman, R., Hayakawa, K., and Hardy, R. R. (1996) Immunity 5, 527-535[CrossRef][Medline] [Order article via Infotrieve]
  17. Boekel, E. T., Melchers, F., and Rolink, A. (1995) Int. Immunol. 7, 1013-1019[Abstract/Free Full Text]
  18. Gong, S., and Nussenszweig, M. C. (1996) Science 272, 411-414[Abstract]
  19. Koyama, M., Ishihara, K., Karasuyama, H., Cordell, J. L., Iwamoto, A., and Nakamura, T. (1997) Int. Immunol. 9, 1767-1772[Abstract/Free Full Text]
  20. Nagata, K., Nakamura, T., Kitamura, F., Kuramochi, S., Taki, S., Campbell, K. S., and Karasuyama, H. (1997) Immunity 7, 559-570[CrossRef][Medline] [Order article via Infotrieve]
  21. Maruyama, S., Kubagawa, H., and Cooper, M. D. (1985) J. Immunol. 135, 192-198[Abstract]
  22. Mason, D. Y., Cordell, J. L., Tse, A. G. D., van Dongen, J. M. M., van Noesel, C. J. M., Micklem, K., Pulford, K. A. F., Valensi, F., Comans-Bitter, W. M., Borst, J., and Gatter, K. C. (1991) J. Immunol. 147, 2474-2482
  23. Kiyotaki, M., Cooper, M. D., Bertoli, L. F., Kearney, J. F., and Kubagawa, H. (1987) J. Immunol. 138, 4150-4158[Abstract]
  24. Chen, C. H., Pickel, J. M., Lahti, J. M., and Cooper, M. D. (1990) in Avian Cellular Immunology (Sharma, J. M., ed) , pp. 1-22, CRC Press, Orlando, FL
  25. Stong, R. C., Korsmeyer, S. J., Parkin, J. L., Arthur, D. C., and Kersey, J. H. (1985) Blood 65, 21-31[Abstract/Free Full Text]
  26. Korsmeyer, S. J., Arnold, A., Bakhshi, A., Ravetch, J. V., Siebenlist, U., Heiter, P. A., Sharrow, S. O., LeBien, T. W., Kersey, J. H., Poplack, D., Leder, P., and Waldmann, T. A. (1983) J. Clin. Invest. 71, 301-313
  27. Finley, H. W., Cooper, M. D., Kim, J. H., Alvarado, C., and Ragab, A. H. (1982) Blood 60, 1305-1309[Abstract/Free Full Text]
  28. Pulvertaft, R. J. V. (1964) Lancet 1, 238-240
  29. Laemmli, U. K. (1970) Nature 227, 680-685[CrossRef][Medline] [Order article via Infotrieve]
  30. Hashimoto, S., Chiorazzi, N., and Gregersen, P. K. (1995) Mol. Immunol. 32, 651-659[CrossRef][Medline] [Order article via Infotrieve]
  31. Campbell, K. S., and Cambier, J. C. (1990) EMBO J. 9, 441-448[Medline] [Order article via Infotrieve]
  32. Chen, J., Stall, A. M., Herzenberg, L. A., and Herzenberg, L. A. (1990) EMBO J. 9, 2117-2124