| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Medicine, Division of Oncology, Stanford University Medical Center;
Institut André Lwoff, Institut de la Santé et de la Recherche Médicale Unité 268, Hôpital Paul Brousse, Villejuif, France; and
Department of Immunology, Duke University Medical Center, Durham, NC
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mechanisms by which CD81 acts are still unclear. Direct signal transduction is not likely because the intracellular domains of the molecule are short and do not contain known phosphorylation motifs (9). However, CD81 and all other tetraspanins tend to form multimolecular complexes in which tetraspanins associate with specific integrins and with lineage-specific proteins (9). These associations have been coined the "tetraspanin web" (10). We have speculated that the diverse functions of CD81 are related to its ability to form complexes and thereby to act as a "molecular facilitator" (11) of the specific functions of its associated proteins.
Little is currently known about the proteins that CD81 associates with in brain glial cells, in eggs, or in hepatocytes. In contrast, in B cells, CD81 directly associates with CD19, taking part in the CD19-CD21-CD81 signaling complex (12, 13, 14). In addition, expression of CD19 in bone marrow, spleen, and peripheral B cells is reduced in three independently derived lines of CD81-null mice (15, 16, 17). Since the reduction in CD19 was shown to be uniform in the spleen of cd81-/- mice, yet quite heterogeneous in the bone marrow, the current study was first aimed at analyzing the effect of the CD81-null mutation on CD19 expression during B cell development in the bone marrow. In this study, we show that the reduction in CD19 expression in cd81-/- B cells is correlated with their developmental stage. The most affected population, small pre-BII cells, expresses negligible levels of CD19.
To the best of our knowledge, this is the first example in which a tetraspanin molecule has been shown to be involved in cell surface expression of its associated protein. To address a possible reciprocity, i.e., the role of CD19 in CD81 expression, we tested developing bone marrow B cells from cd19-/- mice for CD81 expression. There was no effect of CD19 deficiency on CD81 expression. We also tested whether this effect on CD19 expression was specific to CD81 and showed that the deficiency of another B cell-expressed tetraspanin, CD9, did not affect CD19 levels. The unique effect of CD81 was demonstrated by its reintroduction into cd81-/- B cells, which restored CD19 expression.
Tetraspanins form the tetraspanin web and have been shown to compartmentalize in distinct submembranal microdomains, e.g., in CDw78 microdomains (18) and exosomes (19). However, the cellular compartment where the web originates is currently unknown. We thus sought to locate the cellular compartment where CD19 becomes dependent on CD81. Because CD19 is a highly glycosylated molecule, it is possible to follow its maturation in the secretory pathway. In this study, we demonstrate that CD19 accumulates in two different glycoforms in B cells, which in turn, enabled studies aimed at defining the compartment in which CD19 becomes dependent on CD81. Our studies show that CD81 affects CD19 expression beginning at a postendoplasmic reticulum (ER) compartment. Resolving the mode by which CD81 contributes to the activity of its associated partner, CD19, is a first step in defining how CD81 may affect its partners in other cell types.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because of the infertility of female cd81-/- and cd9-/- mice, these mice were bred as heterozygotes, giving rise to knockout (KO) and WT littermates (6, 15). Male and female BALB/c or C57BL/6 cd81-/- (KO) mice and WT littermates were bred at the Department of Comparative Medicine, Stanford University (Stanford, CA). C57BL/6 cd9-/-, double cd9-/-/cd81-/- mice, and WT littermates were bred at the Institute André Lwoff (Villejuif, France). BALB/c cd19-/- mice were bred at the Duke University Medical Center (Durham, NC) (20). In all experiments, mice ages 6–20 wk old (age and sex matched) were used. All animal protocols were approved by the Stanford University Committee on Animal Welfare.
Cell lines
70Z/3 pre-B cells, 38C13 (C3H/HeN), and A20 (BALB/c) mature B cells were grown in RPMI 1640, 50 µM 2-ME, and 10% FCS. A20 cells were sorted under sterile conditions by FACSCalibur (BD Immunocytometry Systems, San Jose, CA) to separate A20 CD19+ from A20 CD19- cells (21). Sorted cells were further subcloned at limiting dilutions from which the single-cell-cloned A20 CD19+ and A20 CD19- cell lines were derived.
Cell preparation
Single-cell suspensions of spleens were prepared by homogenizing the spleen tissue with frosted glass slides in medium (RPMI 1640, 50 µM 2-ME, 10% FCS) and filtering through a 70-µm cell strainer. Single-cell suspensions of bone marrow from the femur and tibia were prepared by flushing the marrow cavities with medium (RPMI 1640, 50 µM 2-ME, 1% FCS). RBC from bone marrow and spleen single-cell suspensions were removed by hypotonic lysis in 0.144 M NH4Cl/0.017 M Tris-HCl (pH 7.2) for 1–2 min at room temperature.
Bone marrow B lymphopoietic cultures
Fresh single-cell suspensions from bone marrow were prepared, without removal of RBC, and seeded at 106 cells/ml RPMI 1640, 50 µM 2-ME, and 5% FCS, and cultured as previously described (22). At 2–3 wk after culture initiation, nonadherent cells were washed gently from the stroma layer and analyzed.
Flow cytometric analysis and cell sorting
Bone marrow or spleen single-cell suspensions were first incubated with anti-CD16/CD32 mAb to block Fc
III/IIRs for 20 min at 4°C. Following the blocking step, cells were incubated with various combinations of fluorochrome- or biotin-conjugated Abs at saturation for 30 min at 4°C. Biotin-conjugated reagents were counterstained for 30 min with streptavidin (SA)-allophycocyanin (BD Immunocytometry Systems). Stained samples were washed and resuspended in staining medium with calcium- and magnesium-free PBS containing 2% BSA (ICN Biomedicals, Aurora, OH) and 0.02% NaN3 (Sigma-Aldrich, St. Louis, MO). Finally, the stained cells were fixed with 2% paraformaldehyde in PBS (calcium and magnesium free) for three- to four-color flow cytometric analysis (FACSCalibur; BD Immunocytometry Systems). Unless indicated otherwise, mAbs used for immunofluorescent staining were purchased from BD PharMingen (San Diego, CA): unconjugated anti-CD16/CD32 (FcIII/IIR), unconjugated anti-CD19 (1D3), PE-anti-CD19 (1D3), biotinylated anti-CD19 (1D3), FITC- anti-CD43 (S7), CyChrome-anti-CD45R/B220 (RA3-6B2), PerCP-anti-CD45R/B220 (RA3-6B2), biotinylated anti-CD25 (7D4), FITC-anti-IgMa (DS-1), PE-anti-IgMa (DS-1), PE-anti-CD81 (Eat2), biotinylated anti-CD9 (KMC8), biotinylated anti-CD24 (M1/69), FITC-anti-IgD (11-26c.2a), and all matched isotype controls. An additional PE anti-CD19 mAb (clone 6D5) was purchased from Caltag Laboratories (Burlingame, CA). hCD81 was detected with biotinylated mouse anti-hCD81 mAb (5A6). Live cells were identified by staining with propidium iodide (Boehringer Mannheim, Mannheim, Germany).
B splenocyte purification
Viable lymphocytes were isolated from single-cell suspensions followed by a density gradient centrifugation using Lympholyte-M (CEDARLANE Laboratories, Hornby, Ontario, Canada) according to the manufacturer’s instructions. B splenocytes were magnetically labeled with anti-mouse CD45R (B220) MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany) after blocking FcIII/IIRs with anti-CD16/CD32 mAbs (BD PharMingen) for 20 min at 4°C. Labeled cells were positive-selected by passing through a separation column using the automated magnetic cell sorter autoMACS (Miltenyi Biotec), resulting in a final purity of 
95% B220-positive cells.
Bone marrow pre-BII cell purification
Bone marrow cell suspensions, blocked with the anti-CD16/CD32 mAbs, as above, were incubated with the primary mAb FITC-anti-IgMa (DS-1; BD PharMingen) at 1 µg/106 cells for 30 min at 4°C. Stained samples were washed and resuspended in calcium- and magnesium-free PBS containing 2 mM EDTA and 0.5% BSA (ICN Biomedicals) buffered at pH 8. Cells were then magnetically labeled with a mixture of anti-mouse CD43 (Ly-48) and anti-FITC MicroBeads. Labeled cells (all CD43+ non-B cells, CD43low Pro-B and Pre-BI, and IgM+ immature and mature B cells) (23) were depleted by passage through a separation column using the autoMACS. Unlabeled cells (CD43-IgM-) were collected and stained with anti-mouse CD45R (B220) MicroBeads. Finally, CD43-IgM-B220+-labeled cells were positive-selected on the autoMACS separation column. All MicroBead reagents and the autoMACS device were obtained from Miltenyi Biotec. Cells enrichment was confirmed by four-color flow cytometry analysis: staining with a mixture of FITC-anti-CD43 (S7), PerCP-anti-CD45R/B220 (RA3-6B2), PE-anti-IgMa (DS-1), and biotinylated anti-CD25 (7D4), then counterstaining with SA-allophycocyanin. Isolated WT bone marrow B220+ cells included 90% pre-BII (CD43-CD25+IgM-) (23) and 10% pro-B (CD43lowCD25-IgM-) (23) cells. Based on the same analysis, isolated cd81-/- bone marrow B220+ cells included 83% pre-BII and 17% pro-B cells. No IgM+ cells could be detected in either of the samples.
Retroviral constructs and ecotropic viral production
hCD81 cDNA was cloned into the retroviral plasmid pBabeMN IRESGFP (detailed in http://www.stanford.edu/group/nolan/retroviral_systems/phx.html). Briefly, the vector contains an internal ribosome entry site (IRES), derived from the encephalomyocarditis virus, which permits the translation of two open reading frames from one messenger RNA, and was provided by Dr. G. Nolan (Stanford University). The hCD81 was inserted 5' to the IRES, which was followed by green fluorescent protein (GFP) cDNA. Ecotropic retrovirus was produced in the Phoenix (
NX)-Eco packaging cell line transfected with pBabe hCD81/GFP or pBabe GFP plasmids by using Gene PORTER 2 transfection reagent (Gene Therapy Systems, San Diego, CA). Supernatants containing replication-defective retroviruses were harvested after 48 h, sterile filtered to remove nonadherent NX cells, and used to infect primary B splenocytes.
Infection of primary B splenocytes
Viable single-cell suspensions of primary spleen lymphocytes, isolated as described above, were plated at 2.5 x 106 cells/ml per well in 12-well plates in growth medium (RPMI 1640, 10% FCS, 50 µM 2-ME). Cells were activated for 24 h by 10 µg/ml rat anti-mouse CD40 (FGK45) mAb (24) (a gift from Dr. S. P. Schoenberger, Division of Immune Regulation, La Jolla Institute for Allergy and Immunology, San Diego, CA), and 2 ng/ml recombinant murine IL-4 (DNAX, Palo Alto, CA). Activated cells (5 x 105)/200 µl were harvested and added to 800 µl of retrovirus containing supernatant in 24-well plates (1 ml/well). Polybrene was added to a final concentration of 1 µg/ml and cells were spin-infected by centrifugation at 2500 rpm for 90 min at 32°C. Cells were incubated overnight at 32°C, washed, and incubated in fresh growth medium containing stimulants for an additional 48 h, then harvested, and analyzed by flow cytometry.
Infection of bone marrow B lineage cells from WT and cd81-/- B lymphopoietic cultures
Bone marrow B lymphopoietic cultures as described above were established in six-well plates. Two-week-old cultures, in which colonies were visualized, were infected with 1.5 ml of retrovirus containing supernatant plus 1 ml of growth medium/well. Polybrene was added to a final concentration of 1 µg/ml. Cultures were incubated overnight at 32°C, washed, and incubated in fresh growth medium for an additional 48 h. B lymphopoietic cells were gently washed from the stroma layer and analyzed by flow cytometry.
Semiquantitative PCR
One microgram of cDNA was amplified using 50 µM each of the forward (exon 4) and reverse (exon 15) primers (described below), 10 mM each dNTP, 50 mM MgCl, 5 µl of 10x buffer, and 0.5 µl of Taq enzyme (Invitrogen, Carlsbad, CA) in a final volume of 50 µl. PCR were conducted in a GeneAmp PCR system 9600 (PerkinElmer, Norwalk, CT) using the following program: 94°C for 5 min, 63°C for 5 min, 72°C for 5 min (1 cycle); 95°C for 30 s, 63°C for 45 s, 72°C for 45 s (35 cycles); 72°C for 8 min (1 cycle); and 4°C soak. PCR products were analyzed on 1% agarose gel.
Quantitative real-time PCR
Total RNA was prepared from cell lines (70Z/3, A20 CD19+, A20 CD19-), purified primary B splenocytes, and bone marrow pre-BII cells using the RNeasy mini kit (Qiagen, Valencia, CA) following the manufacturer’s instructions. Quantitative conversion of total RNA to single-strand cDNA was performed using the High-Capacity cDNA Archive kit (Applied Biosystems, Foster City CA) according to the manufacturer’s protocol. Real-time TaqMan PCR and data analysis were performed on a sequence detection system (ABI PRISM 7900 SDS; Applied Biosystems). Each PCR included 14 µl of a master mixture (containing 1x TaqMan Universal PCR master mix 1 (Applied Biosystems), 100 nM each of forward and reveres primers, 200 nM of the relevant probe (described below)) and 6 µl of cDNA. TaqMan Rodent GAPDH control reagents (Primers & Probe) were obtained from (Applied Biosystems). The CD19 primers and probes, were designed with the assistance of the Primer Express software (Applied Biosystems) and synthesized by that company. Exons 4/5 boundary as follows: probe, 5'-VIC-TCATTGCAAGGTCAGCAGTGTGGCTC-TAMRA-3'; forward primer (exon 4), 5'-CCATCGAGAGGCACGTGAA-3'; reverse primer (exon 5), 5'-TCCATCCACCAGTTCTCAACAG-3'. Exons 14/15 boundary as follows: probe, 5'-VIC-CGTGACTCCCAAGTGACTAGCCTGGACT-TAMRA-3'; forward primer (exon 14), 5'- GAAGGAGAGGGCCACATGG-3'; and reverse primer (exon 15), 5'-ATGTGGTTCTTGGGACCTAACG-3'. Thermal cycling conditions comprised AmpliTaq Gold activation at 95°C for 10 min, 40 cycles of denaturation at 95°C for 15 s, and annealing and extension at 60°C for 1 min. The level of GAPDH mRNA was used as the endogenous control for each of the samples. Standard curves for CD19- and GAPDH-specific reactions were generated by serially diluting the 70Z/3 cDNA sample to 100, 50, 10, 2, 0.4, and 0.08 ng cDNA per PCR. This analysis determined that both genes were efficiently amplified (slope values); the value for GAPDH was -3.41 and that for CD19 exon 4/5 and 14/15 were -3.48 and -3.47, respectively. The threshold cycle values for GAPDH and CD19 standard curves (for 100–0.08 ng) were within the detectable range of the assay, 15–25 and 19–29, respectively. Subsequent analyses utilized 50 ng cDNA per PCR and performed in triplicates, with no-template controls. Sample threshold cycle values for GAPDH- and CD19-specific products were translated to quantity values based on the relevant standard curves. To account for variation in sampling and RNA preparation, quantity values for CD19-specific products were normalized by the respective quantity values for the GAPDH-specific products.
Immunoprecipitation
Cells (2 x 108 cells/ml) were lysed in calcium- and magnesium-free PBS containing 1% Nonidet P-40 (BDH Laboratory, Poole, U.K.) and complete ETDA-free protease inhibitor mixture (Roche Diagnostics, Mannheim, Germany) for 30 min on ice, then clarified by centrifuging at 14,000 rpm for 15 min at 4°C. Cell lysates (from 2 x 107 cells) were precleared by rotation for 1–2 h at 4°C with 60 µl of 50% v/v slurry of goat anti-rat agarose beads (Sigma-Aldrich). Precleared lysates were incubated for at least 4 h at 4°C by rotation with 60 µl of goat anti-rat agarose beads prebound with 10 µg purified rat anti-mouse CD19 mAb (1D3) (25) or a matched isotype control mAb (BD PharMingen). Immobilized complexes were collected by centrifugation at 1000 rpm for 1 min at 4°C, washed three to four times in 1 ml of lysis buffer, and eluted by boiling in SDS-PAGE loading buffer or in denaturation buffer for enzymatic removal of carbohydrates, as detailed below.
Endoglycosidase treatment
Immunoprecipitated CD19 proteins were divided into three aliquots, one was treated with endoglycosidase H (endo-H), one with peptide N-glycosidase F (PNGase-F), and one was left untreated. endo-H cleavage of high mannose structures was performed using the PO703S kit (New England Biolabs, Beverly, MA). PNGase-F digestion of total N-glycan chains was performed using the 704S kit (New England Biolabs). SDS-PAGE sample buffer was then added and digested proteins were analyzed by 10% SDS-PAGE followed by Western blotting.
Gel electrophoresis and immunoblotting
Cell lysates and immunoprecipitated products were separated by 10% SDS-PAGE and electrotransferred onto polyvinylidene difluoride membranes. CD19 molecules were detected by incubating the membranes with rabbit polyclonal anti-mouse CD19 extracellular domain serum, a generous gift from J. C. Cambier (Department of Immunology, University of Colorado Health Sciences Center and National Jewish Medical Research Center, Denver, CO). Membranes were washed in TBS plus 0.1% Tween 20 and incubated with donkey anti-rabbit HRP-linked secondary Abs (1/10,000; Amersham Life Science, Piscataway, NJ). Blots were visualized by chemiluminescence detection (ECL; Amersham, Little Chalfont, U.K.) following instructions of the manufacturer.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine whether lack of CD81 reduced CD19 expression differentially in distinct subsets of early B cells, we triple-stained bone marrow cells from cd81-/- mice and their WT littermates with anti-B220, anti-IgM, and anti-CD19 mAbs. We then analyzed CD19 expression levels in the B220highIgM+ and B220+IgM- gated populations (R1 and R2, respectively, Fig. 1A). CD19 expression in the B220highIgM+ gated (R1) cells, which include mature B cells, was reduced in cd81-/- cells (Fig. 1B, upper panel). This reduction (3- to 4-fold) in CD19 expression was similar to that observed in the spleen (15, 16, 17). In contrast, B220+IgM- gated (R2) cells, which include early B cell precursors, show an even more pronounced reduction (8-fold) in CD19 expression in the cd81-/- B cells (Fig. 1B, lower panel).
|
10-fold reduction) with a staining pattern that is similar to the isotype control Ab. A trend of recovery of CD19 expression was seen in the transition to the immature and mature stages, which show 3-fold reduction. The use of a different set of mAbs including anti-CD24 (heat-stable Ag) and anti-IgD to identify B cell subsets yielded similar results (data not shown). In addition, similar results were obtained using a different anti-CD19 mAb, clone 6D5 (data not shown). CD81 expression is independent of CD19 expression
Because CD81 and CD19 are complexed in B cells and because CD81 depletion is detrimental for CD19 expression, we tested whether this effect is reciprocal. We analyzed cd19-/- bone marrow cells for CD81 expression. All major cd19-/- bone marrow B lineage subsets expressed CD81 equally or above normal levels in comparison to WT cells (Fig. 1D). In addition, CD81 was expressed equally on CD19+ and CD19- A20 cells (data not shown). Thus, expression of CD81 is independent of CD19.
Analysis of CD19 expression in differentiating cd81-/- and WT B lymphopoietic bone marrow cultures
To determine whether newly developing cd81-/- pre-B cell express reduced levels of CD19, we cultured murine bone marrow precursors of B lymphocytes by harvesting total bone marrow cells from the femur and tibia of cd81-/- or WT mice under the original conditions used by Whitlock et al. (22). Both WT and cd81-/- hemopoietic precursors were grown on their respective stromal cells and both gave rise in vitro to newly formed pre-B colonies. B cells grown in these cultures were then analyzed for expression of B lineage markers. As seen in vivo, large CD43+B220+ cells from cd81-/- cultures expressed low levels of CD19 by comparison to cultured WT counterparts (data not shown). Moreover, cd81-/- cells within the small pre-BII cells, gated as CD43-B220int cells (Fig. 2A, gate R1, upper panels) and thenanalyzed as CD25+ cells, show no expression of surface CD19 (Fig. 2A, lower panels). However, unlike the recovery seen in vivo, no recovery of CD19 expression was observed within the cd81-/- immature B cell subset, characterized as small CD43-B220intCD25- cells (Fig. 2A) or as small B220intIgM+IgD- cells (Fig. 2B), although the accumulation of surface IgM expression was normal (Fig. 2B, upper panels). Furthermore, in vitro-differentiated cd81-/- bone marrow B cells show reduced CD19 levels also when seeded on a pre-established cd81+/+ feeder stroma layer (14F1.1 cell line; data not shown). We therefore suggest that the reduction in CD19 expression is an intrinsic cd81-/- B cell defect and not a reflection of a CD81-deficient microenvironment.
|
Because of redundancy of tetraspanins, a possibility exists that loss of a member within the tetraspanin web (10, 11) could also produce a similar defect in CD19 expression. The tetraspanin CD9 was shown to associate through CD81 with the CD19-CD21-CD81 complex in human B cells (26). We therefore investigated whether CD9 depletion could affect CD19 expression in the developing bone marrow B cells.
Because little is known about CD9 expression in mouse bone marrow, we first analyzed its pattern of expression in developing bone marrow B cell subsets. B cell subsets’ distribution appeared to be normal in the bone marrow of cd9-/- and cd9-/-/cd81-/- mice (Fig. 3A, left panels; Fig. 4A, upper panels). The specificity of staining was verified by parallel staining of bone marrow B cells from WT and cd9-/- mice, as shown for B220highCD43- cells (Fig. 3A, right panels). Similarly, surface CD9 expression was detected in all bone marrow B cell subsets as summarized in Fig. 3B.
|
|
Human CD81 restored CD19 expression in cd81-/- B cells to WT levels
Previous studies have demonstrated that mouse CD19 associates with hCD81 (27). In addition, expression of hCD81 can be followed in mouse WT cells. We therefore chose to use the human version of this gene and inserted hCD81 cDNA into a retroviral vector, pBMN-IRES-GFP, which encodes a ribosomal entry site upstream of the GFP gene. This vector and a vector encoding GFP alone were transfected into the ecotropic NX packaging cell line for production of the respective retroviruses. Freshly harvested spleen lymphocytes were preactivated for 24 h in culture and then infected with the CD81 encoding or the control virus. The yield of infection was very similar between WT and cd81-/- cells for both retroviruses (as indicated in Fig. 5A, R1-gated cells). As expected, because retroviruses infect dividing cells, only activated cells (based on their forward and side scatter patterns) in both cultures were infected and expressed GFP (Fig. 5A, R1-gated cells). Infected cells (GFP+) were then analyzed for expression of hCD81 by staining with the directly conjugated mAb 5A6. All cells infected by retrovirus expressing hCD81/GFP, but not GFP alone, were positive for surface hCD81, and the level of hCD81 expression was identical in WT and cd81-/--infected cells (Fig. 5B). Analysis of uninfected cells showed that only WT cells express mouse CD81 (Fig. 5C, upper panel) and their level of CD19 expression was higher than that of the uninfected cd81-/- B cells (Fig. 5C, lower panel). This analysis demonstrated that the absence of CD81 and reduced CD19 levels did not affect the ability of cd81-/- B cells to be activated and infected and that infection by the virus encoding hCD81 induced identical cell surface hCD81 levels on both WT and cd81-/- B cells.
|
|
CD81 could exert its effect by controlling CD19 mRNA levels. To test this possibility, we measured the accumulated CD19 mRNA in bone marrow pre-BII and mature B cells, harvested from cd81-/- or WT littermates, using quantitative real-time RT-PCR. These measurements utilized two specific primer and probe sets. One set amplified the middle portion of the gene overlapping the border between exon 4 (encoding one of the two extracellular Ig-like domains) and exon 5 (encoding the putative transmembrane domain). The second primer set was designed to amplify the 3' portion of the gene overlapping the border between exon 14 (encoding the carboxyl terminus of the protein) and exon 15 (containing the majority of the 3' untranslated region including the poly(A) signal) (as detailed in Materials and Methods) (28). This quantitative analysis revealed that sorted cd81-/- pre-BII and mature spleen B cells expressed CD19 mRNA levels comparable to their WT counterparts (Fig. 7A). Identical results were obtained with the two primer and probe sets (Fig. 7A). Positive and negative references for CD19 expression included A20 CD19+ and A20 CD19- cell lines, which had a 40-fold difference in CD19 mRNA (Fig. 7A, right panel). A20 CD19- cells do not express the CD19 protein (Fig. 7B), but unlike cd81-/- B cells express lower levels of CD19 mRNA (Fig. 7, A and C). From this analysis we conclude that reduced CD19 expression in cd81-/- B cells does not result from alteration in CD19 mRNA accumulation and that there is a common mechanism in which CD81 regulates normal surface expression of CD19 in all B lineage cells.
|
Analysis of the CD19 protein in A20 cells revealed the presence of two specific CD19 bands, a major band, above 115 kDa, and a minor band of 100 kDa (Fig. 7C). To follow up on this observation, we tested the mature B cell line 38C13 (Fig. 8A), pre-B cells, 70Z/3 (Fig. 8A), the immature B cells, WEHI-231 (data not shown), and primary B splenocytes (Fig. 8B) all expressed the two CD19 bands. Both bands reacted with the anti-CD19 Abs, neither band was detected by the isotype control mAb or in A20 CD19- cells (Figs. 7C and 8).
|
As a surface membrane protein, CD19 is processed and transported through the secretory pathway. The two CD19 glycoforms could possibly represent mature (major, high molecular mass) and immature (minor, low molecular mass) forms, retained in different compartments of the secretory pathway. Proteins that exit the ER undergo trimming of the high mannose and acquisition of terminal carbohydrate structures that render them resistant to digestion by endo-H (29, 30). Thus, sensitivity to endo-H digestion can serve to locate glycoproteins in the secretory pathway. CD19 was immunoprecipitated from cell lysates derived from the pre-B cell line 70Z/3 and from the mature B cell line 38C13, both cell lines accumulate significant levels of the minor, low Mr, CD19 glycoform. This analysis revealed that only the low Mr minor CD19 glycoform (Fig. 8A, filled arrowhead) is sensitive to digestion by endo-H (Fig. 8A), indicating that this glycoform resides in the ER or in the early Golgi. In contrast, the high Mr major glycoform is resistant to endo-H (Fig. 8A, arrow) and therefore accumulates in a post-ER compartment.
Lack of CD81 affects the expression of the major endo-H-resistant CD19 glycoform, but not the minor endo-H-sensitive glycoform
Analysis of CD19 glycoforms from B splenocytes purified from cd81-/- mice or their WT littermates demonstrated that both primary B cells express the minor glycoform shown to be sensitive to endo-H digestion (Fig. 8B, filled arrowhead). Most importantly, the expression of the minor endo-H-sensitive CD19 glycoform was not reduced in cd81-/- compared with WT B cells, indicating that lack of CD81 does not inhibit the accumulation of this glycoform (Fig. 8B, filled arrowhead, and 8C). In contrast, the major CD19 glycoform is significantly less abundant in cd81-/- than in WT B cells (Fig. 8B, arrow, and 8C) and, therefore, is unable to accumulate in normal levels in the absence of CD81.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CD81 associates in mature B cells with CD19, taking part in the CD19-CD21-CD81 molecular complex that acts in concert with the B cell receptor (BCR) complex to reduce the threshold of B cell activation (13, 33, 34, 35). Consequently, B lymphocytes from mice that overexpress CD19 are hyperresponsive to Ag receptor cross-linking (20) and C57BL/6 mice that overexpress CD19 have 2- to 4-fold higher levels of anti-DNA autoantibodies and rheumatoid factor (33, 36). Thus, CD19 overexpression shifts the balance between tolerance and immunity, which suggests that inappropriate CD19 expression could affect selection processes in the periphery and/or in the lymphopoietic primary organs.
B cells from three independent lines of cd81-/- mice exhibit a distinct phenotype, a reduction in CD19 (15, 16, 17), suggesting that CD81 facilitates CD19 surface expression, which in turn is critical for the membrane stability of the CD19-CD21 signaling complex. Since CD19 is the only surface molecule known to be affected by its tetraspanin partner, CD81, understanding how the CD81-CD19 partnership is formed might contribute to the understanding of how CD81 functions through its specific associations in other cell types. In turn, this may further promote the general understanding of how the tetraspanin webs are formed.
The reduction in CD19 expression in mature cd81-/- B cells is uniform, but is heterogeneous and correlated with B cell development in the bone marrow (Fig. 1). All bone marrow cd81-/- B cell subsets show reduced CD19 levels, yet some subsets are more severely affected; cd81-/- small pre-BII cells are almost devoid of surface CD19 (Figs. 1, 2, and 4). Human CD81 associates with CD19 in mature B cells (12, 13, 27) and in the pre-B cell line NALM-6 (26). In mice, CD81 is detected on all bone marrow B lineage cells (Fig. 1D and data not shown), thus it is coexpressed with CD19 from the pro-B stage and is necessary for surface expression of CD19 in all of the subsequent developmental B cell stages (Fig. 1). We assume that, like in mature B cells, the two molecules associate with each other in the early B cell stages. The role that CD81 plays is still unclear, but because the intracellular domains of CD81 are very short and do not contain known phosphorylation motifs, it is probably not directly involved in signal transduction. It is more likely that CD81 facilitates the stability of CD19 on the cell surface.
The role of CD19 in early B cell development is still unknown, but is likely linked to its signaling ability, as shown for pre-B cells (27, 37, 38). Recently, CD19 was shown to play an important role in promoting positive selection and in establishing L chain allelic exclusion by regulating the BCR signaling threshold (39). Similarly, perturbation by expression of CD21 earlier than normal, during the pre-B cell stage, severely reduced the number of mature B cells in the periphery (40). Because CD21 is not a signaling molecule, this result is probably due to inappropriate signaling via its associated molecule, CD19. These experiments indicate that it is likely that CD19-dependent selection plays a role during the early stages of B cell development. Studies aimed at analyzing pre-B cell selection generally make use of BCR-transgenic systems. Such transgenic systems suffer from an inherent problem, i.e., the use of a BCR that has already been selected in vivo. In the current work, we demonstrate that cd81-/- B cells have a very unique feature, the virtual absence of CD19 in one particular developmental stage, small pre-BII cells. Ongoing studies in cd81-/- mice are aimed at revealing the role of signaling through CD19 in B cell development during the transition from small pre-BII to the immature B stage.
In vitro, cd81-/- maturing B cells go on to regain normal levels of BCR expression but fail to express CD19 (Fig. 2). Because surface IgM accumulation is normal in cd81-/- immature B cells without a parallel recovery in CD19 surface expression, we suggest that only in vivo the bone marrow microenvironment supports B cell selection that is related to the level of CD19 expression. Interestingly, in vitro, the development of WT cells was also affected; while a large proportion of the immature WT B cells expressed both IgM and CD19, some failed to express CD19 (Fig. 2B). This indicates that in vitro, where B cell maturation is less subject to selective pressure, the expression of CD19 is not necessarily linked to the expression of the BCR.
We investigated whether CD81 is unique in its ability to promote CD19 expression. We found no reciprocity regarding cell surface expression of CD81 because cd19-/- B cells express normal levels of CD81 (Fig. 1D). The difference between WT and cd19-/- in fractions (Fr.) D and E is not statistically significant. The other molecule that CD19 associates with is CD21; however, cd81-/- or cd19-/- B cells express normal levels of CD21 (17, 41) and cd21-/- B cells express somewhat elevated levels of CD19 (41). To determine whether a different tetraspanin web (10) molecule could have a possible effect on CD19 expression, we analyzed CD19 expression in CD9 KO mice. CD9 is a tetraspanin that was previously shown to associate indirectly with CD19 in human B cell lines and in purified bone marrow CD10+ early B cells (26). In this study, we show that CD9 is expressed during mouse early B cell development, as illustrated in Fig. 3. Further analysis of the various bone marrow subsets demonstrated that lack of CD9 does not influence CD19 expression (Fig. 4). Taken together, the expression of CD19 on the surface of B cells is specifically dependent on CD81.
Because in vitro differentiated cd81-/- bone marrow B cells show reduced CD19 levels (Fig. 2) also when seeded on a pre-established cd81+/+ feeder stroma layer (data not shown), we suggest that the reduction in CD19 expression is an intrinsic cd81-/- B cell defect and not a reflection of a CD81-deficient microenvironment. This was further confirmed by reintroduction of the hCD81 gene into cd81-/- primary B cells: hCD81 completely restored CD19 expression in cd81-/- primary B cells to WT level (Fig. 6). Furthermore, the latter prove that optimal expression of CD19 is dependent on CD81 and that hCD81 can substitute for the mouse molecule. Based on the high amino acid sequence homology between human and mouse CD81 (42), these results implicate and possibly rule out certain CD81 domains as the ones responsible for CD19 expression.
Quantitative real-time RT-PCR determination ruled out the possibility that reduced CD19 expression is a result of reduced CD19 mRNA, even in pre-BII cells (Fig. 7A). Mouse CD19 mRNA encodes a protein of 547 aa residues containing seven potential N-linked glycosylation sites (28). The reported mass of the glycosylated protein has been 120 kDa for murine B splenocytes and 115–155 kDa for a panel of murine B cell lines (27). These previous studies detected only cell surface-labeled CD19. Immunoprecipitation followed by detection with a Western blot reagent enabled the identification of both CD19 glycoforms. The size of the major glycoform (arrows in Fig. 8, A and B) is similar to that reported for cell surface-expressed CD19 (27). The size of the minor CD19 glycoform is around 100 kDa, higher than the predicted protein backbone. Both major and minor glycoforms collapsed to an identical Mr band upon PNGase-F treatment, denoting that the minor form differs in its N-glycan composition and is not a degradation product.
Newly synthesized N-linked glycoproteins originate in the ER and pass through the Golgi complex on their way to the cell surface (30). In the ER they acquire the "core" high mannose glycans. Core glycans have been suggested to facilitate protein folding in the ER by promoting proper association with soluble molecular chaperones and folding enzymes, all of which provide "quality control" (43, 44). ER and early Golgi core high mannose-decorated proteins are endo-H sensitive. The minor CD19 glycoform was digested by endo-H, locating it to an early compartment. Accumulation of this CD19 glycoform was not diminished in cd81-/- B cells (Fig. 8B).
Trimming of core glycans and subsequent addition of terminal N-linked glycans occur in the Golgi complex. The resulting terminally glycosylated proteins are tremendously diverse and unlike the more uniform ER quality control assembly may be guided by specific chaperons (30, 43). Accumulation of the endo-H-resistant CD19 glycoform is impaired in cd81-/- B cells (Fig. 8B), indicating that CD81 regulates CD19 in the secretory pathway at a post-ER compartment. We speculate that CD81 may act as a chaperon that stabilizes CD19 during its maturation and trafficking from the Golgi to the cell surface membrane.
As elaborated above, only the high molecular CD19 glycoform is affected in cd81-/- B splenocytes. Analysis of a panel of murine B cell lines (including pre-B, immature, and mature B cell lines) revealed that this CD19 glycoform ranges between 115 and 155 kDa in different cell lines. In contrast, the size of the minor low molecular CD19 glycoform was identical in all analyzed B cells. The variation in size of this endo-H-resistant glycoform was caused by alteration in terminal N-linked glycosylation, as evident by the single size band formed after treatment by PNGase-F (Fig. 8A). It is currently unknown whether CD19 glycoforms differ within developing bone marrow B cell subsets. In this study, we show that different cd81-/- B lineage subsets express different CD19 levels; one possible explanation for this phenotype is that the association between CD19 and CD81 is glycosylation dependent.
In summary, this study scrutinizes the mode by which CD81 affects CD19 surface expression. The dependence of CD19 on CD81 is intrinsic to B cells and is controlled at a post-ER compartment. We propose that CD81 acts as a CD19 chaperon in the secretory pathway and/or as a cell surface membrane stabilizer. We believe that revealing how the CD81-CD19 partnership is formed could contribute to resolving the CD81 partnership(s) and role(s) in other cell types. It should also lead to a better understanding of the mechanism of action of tetraspanins in general.
|
Acknowledgments |
|---|
NX retroviral infection system; and Stephen P. Schoenberger for the rat anti-CD40 (FGK45) Ab. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Shoshana Levy, Department of Medicine, Division of Oncology, CCSR Room 1105a, 269 Campus Drive, Stanford University Medical Center, Stanford, CA 94305-5151. E-mail address: levy{at}cmgm.stanford
3 Abbreviations used in this paper: WT, wild type; KO, knockout; MFI, mean fluorescence intensity; SA, streptavidin; FSC, forward scatter; SSC, size scatter; Fr., fraction; endo-H, endoglycosidase H; PNGase-F, peptide N-glycosidase F; hCD81, human CD81; ER, endoplasmic reticulum; GFP, green fluorescent protein; NX, Phoenix.
Received for publication December 4, 2002. Accepted for publication August 15, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
heavy chain class switching. Immunity 5:319.[Medline]
) and WT (
) littermates were stained with anti-CD43 vs anti-B220, anti-CD25, and anti-CD19 (1D3). D, CD81 surface expression in bone marrow cd19-/- B lineage cells is not decreased: bone marrow cells from three individual cd19-/- mice (
) and three individual WT mice (
), cd81-/- (gray squares), and cd9-/-/cd81-/- (gray triangles) were analyzed. The data represent two independent experiments.
) and WT counterparts (
2-microglobulin (