Activation of the EBV/C3d Receptor (CR2, CD21) on Human B Lymphocyte Surface Triggers Tyrosine Phosphorylation of the 95-kDa Nucleolin and Its Interaction with Phosphatidylinositol 3 Kinase

Cells

Raji cells are from a Burkitt B lymphoma cell line that express CR2. K562A cells express CR2, but not CD19, as derived from K562, wild-type cells transfected with the CR2 cDNA (38). Cells were grown in RPMI 1640 medium supplemented with 10% FCS, 2 mM l-glutamine, 1000 U/ml penicillin, and 100 μg/ml streptomycin at 37°C in 5% CO₂. Neomycin (200 μg/ml) were added for K562A cells. Normal B lymphocytes purified from tonsils of normal human donors were purified as previously described (35)_.

Abs and reagents

The following Abs and reagents were purchased from the indicated companies: anti-CR2 mAb: BL13 and anti-CD19 mAb (B4) from Immunotech (Westbrook, ME) and OKB-7 from Ortho Diagnostics (Raritan, NJ), goat anti-human IgM F(ab′)₂ from Cappel (Durham, NC), and goat anti-mouse Ig (GAM) from Dako (Carpenteria, CA); polyclonal Ab anti-p85 subunit of PI 3 kinase, anti-phosphotyrosine residue (anti-Ptyr) mAb (4G10), and GST fusion proteins of the N-terminal SH2 domain of p85 (SH2-Nt-p85) from Upstate Biotechnology (Lake Placid, NY); Grb2, Fyn, and GAP fusion proteins were obtained from PharMingen (Becton Dickinson, France). Anti-nucleolin mAb were generous gifts from Maria Lopez-Trascasa (Hospital La Paz, Madrid, Spain), Nancy Maizel (Yale University, CT), and Ning-Hsing Yeh (National Yang-Ming University, Taipei, Taiwan). Anti-Grb2 mAb was purchased from Transduction Laboratory (Lexington, KY).

Construction of recombinant SH2 domain of 3BP2

PCR samples containing 500 ng PSport-3BP2 kindly provided by Margaret Knowles (Marie Curie Research Institute, Oxted, U.K.) were heated for 5 min at 94°C, followed by 30 cycles of 1 min at 94°C, 1 min at 55°C, 1.30 min at 72°C, followed by a final extension of 7 min at 72°C. 5′ primer was 5′-CGGGATCCATTACCTTATACGTGATA-3′ (nt 3141–3156 with a BamHI site) and 3′ primer was 5′-TAGAATTCGCTGGCTGGGTTGTATGG-3′ (nt 3246–3261 with a EcoRI site). The PCR fragment obtained was purified after electrophoresis and excision of an agarose gel. Then, SH2 domain of 3BP2 was ligated into PGEX4T3 vector (Amersham, France). After sequencing the insert, fusion protein was produced in BL21 bacteria.

Cell activation

Cells (2 × 10⁷) were washed with RPMI 1640 medium and incubated with the indicated mAb (5 μg/ml) in a final volume of 500 μl at 4°C for 30 min. Then, GAM (20 μg/ml) was added at 37°C for the indicated times to cross-link the first Abs only when BL13 or anti-CD19 was used and not with OKB7. Indeed, it was already well established that among the identified anti-CR2 mAb some, as OKB-7, induced CR2 activation without cross-linking by a secondary Ab, whereas some others, as BL13, needed to be cross-linked by a secondary Ab to activate CR2 (6). Cells were collected by centrifugation and lysed for 30 min at 4°C in buffer containing 50 mM Tris (pH 7.6), 150 mM NaCl, 5 mM EDTA, 1% Nonidet P-40, phosphatase inhibitors as 50 mM NaF, 1 mM Na₃VO₄, 30 mM Na₄P₂O₇, and protease inhibitors as 1 mM PMSF, 60 mM EACA, 12.5 mM benzamidine, and 12.5 mM iodoacetamide. Solubilized proteins were collected by centrifugation at 15,000 × g for 15 min and were stored at −80°C. Protein content was normalized after a bicinchoninic acid assay.

Immunoprecipitation and immunoblotting

One percent Nonidet P-40 solubilized proteins were incubated for 2 h at 4°C with indicated Abs as recommended by the manufacturer. Immune complexes were bound to protein A-Sepharose (15 mg; Pharmacia LKB, Piscataway, NJ) or protein A/G-agarose beads (20 μl; Sigma, St. Louis, MO) for 2 h at 4°C. For each mAb used, an irrelevant mAb of the same isotype has been used in preliminary experiments to verify the specificity of immunoprecipitation assays (data not shown). Immunobeads were washed five times in lysis buffer. Then, proteins were eluted in sample buffer, divided in two or three samples depending on the experiments and submitted to 7% SDS-PAGE under reducing conditions. After SDS-PAGE, separated proteins were electrotransferred on Hybond-ECL nitrocellulose sheet (Amersham, Arlington Heights, IL). The membranes were incubated for 1 h in PBS containing 3% powdered milk and then washed. Specific Abs to appropriate molecules were added and incubated for 2 h at room temperature. After washing in PBS/0.05% Tween 20, peroxidase-linked secondary Abs were added for 1 h at room temperature. Nitrocellulose sheets were then washed in PBS/0.05% Tween 20 and binding of second Abs was detected using the enhanced chemiluminescence kit (Amersham).

In vitro binding assay

GST fusion proteins (5 μg) were bound to glutathione Sepharose 4B beads (Pharmacia LKB) for 1 h at 4°C and then incubated for 2 h at 4°C with 1% Nonidet P-40 solubilized proteins prepared from nonactivated or activated cells as described above. After this incubation, beads were washed five times with lysis buffer. Bound proteins were eluted, submitted to 7.5% SDS-PAGE, then analyzed by immunoblotting using the indicated Abs.

Purification of p95 tyrosine phosphorylated protein

Raji cells (10⁹) were activated by BL13, an anti-CR2 mAb, that was cross-linked with GAM at 37°C, for 2 min, as previously described (39). Then, total cellular components were solubilized in 1% Nonidet P-40 and immunoprecipitated on anti-PTyr mAb covalently linked to Sepharose 4B beads. After washing, proteins were eluted in sample buffer then submitted to 7.5% SDS-PAGE under reducing conditions and electrotransferred on polyvinylidene difluoride (PVDF) membranes (Amersham). Proteins present on PVDF membranes were stained with Ponceau Red. The membrane part that contained the 95-kDa protein detected in samples prepared from CR2-activated cells and absent in samples prepared from nonactivated cells was cut and submitted to high mass accuracy tryptic peptide mapping using matrix-assisted laser desorption/ionization mass spectrometry (Protana A/S; Odense, Denmark). Briefly, the extracted peptides were purified using micro column filled with Poros R2, 50-μm beads. The column was packed in a borosilicate purification needle. After packing, the microcolumn was equilibrated in 5% formic acid and the digest, dissolved in 5% formic acid, was loaded. Several column volumes of 5% formic acid were used for washing. The peptides were eluted with 1 μl of 5% formic acid/50% HPLC-grade methanol in to a nanoES needle. ESMS was performed on a prototype quadrupole time-of-flight mass spectrometer (Sciex, Thornhill, Toronto, Canada) equipped with a nanoES source from Protana A/S (Odense, Denmark). Matrix-assisted laser desorption/ionization mass spectrometry was performed on a Voyager-DE STR from PerSeptive Biosystems. Analysis of the ms/ms data was performed using BioMultiView version 1.4b (Sciex). Database searches were performed with PepSea Client 1.1.

Identification of the p95 protein whose tyrosine phosphorylation was induced by CR2 activation

To identify the p95 tyrosine phosphorylated component, total solubilized proteins were immunoprecipitated on anti-phosphotyrosine (P-Tyr) mAb covalently linked to Sepharose 4B beads. After washing, eluted proteins were submitted to SDS-PAGE and electrotransferred on PVDF membranes that were stained with Ponceau Red. As shown in Fig. 1⇓A, a 95-kDa cellular protein was detected in samples prepared from CR2-activated cells (lane 2) and was absent in samples prepared from nonactivated cells (lane 1). In control, this p95-kDa component was tyrosine phosphorylated in CR2-activated cells (lane 4) and was not in nonactivated cells (lane 3). The PVDF membrane part, which contained this 95-kDa protein, was cut and submitted to sequencing by mass spectrometry method developed by Protana (as described in Materials and Methods). From this analysis (Fig. 1⇓B), two different peptides were chosen for the collision dissociation experiments (Protana) and characterized by the following sequences EVFEDAAEIR and GFGFVDFNSEEDAK. These two peptide sequences were 100% identical with two sequences of human nucleolin and were localized in two different RNA-binding domains (or RRM) of nucleolin, between aa 411–420 and 611–624, respectively. Subsequently, solubilized cellular components from CR2 activated or nonactivated Raji cells were immunoprecipitated on anti-nucleolin mAb. As shown in Fig. 2⇓, although nucleolin was present (B) in both CR2-activated (lanes 2–5) and nonactivated cells, incubated only with buffer (lane 1) or second Ab alone (lane 6), nucleolin was tyrosine phosphorylated (A) only after CR2 activation. The maximum of phosphorylation was reached at 2 min, i.e., a kinetics identical with that previously described for the tyrosine phosphorylated p95 (39). Identical results were obtained when CR2 was activated on Raji cell surface by OKB-7, another anti-CR2 mAb (data not shown). Furthermore, immunodepletion on anti-nucleolin mAb of cell extracts, prepared from CR2 activated cells, led to selective loss of the p95 tyrosine phosphorylated nucleolin (Fig. 2⇓, A and B, lanes 7–9), All these data demonstrated that the p95-kDa protein whose phosphorylation on tyrosine was induced by CR2 activation was nucleolin.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Tyrosine-phosphorylated p95 protein is nucleolin. Total cellular components from 10⁹ Raji cells nonactivated (lanes 1 and 3) or activated (lanes 2 and 4) by BL13, an anti-CR2 mAb, cross-linked with GAM at 37°C, for 2 min, were solubilized in 1% Nonidet P-40 and immunoprecipitated on anti-PTyr mAb covalently bound to Sepharose 4B beads. After washing, eluted proteins were submitted to 7.5% SDS-PAGE under reducing conditions and electrotransferred on PVDF membranes. Proteins present on PVDF membranes were either stained with Ponceau Red (A, lanes 1 and 2) or immunodetected with anti-Ptyr mAb (A, lanes 3 and 4). The part of the membrane that contained the 95-kDa protein detected in samples prepared from CR2-activated cells and absent in samples prepared from nonactivated cells was cut and submitted to high mass accuracy tryptic peptide mapping using matrix-assisted laser desorption/ionization mass spectrometry. Analysis of the ms/ms data was performed using BioMultiView, Version 1.4b. Database searches gave two different peptide sequences, which corresponded to different parts of human nucleolin (B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Activated CR2 triggers tyrosine phosphorylation of human nucleolin. Raji cells (2 × 10⁷) were incubated with 5 μg/ml BL13 in a final volume of 500 μl at 4°C for 30 min (lanes 1–9), then without (lanes 1 and 7) or with 20 μg/ml GAM at 37°C for 2 min (lanes 2, 7, and 8), 4 min (lane 3), 6 min (lane 4), 8 min (lane 5) to cross-link the first Ab. Another control was made by incubating cells only with 20 μg/ml GAM at 37°C for 2 min (lane 6). One milligram of total solubilized proteins (lanes 1-8) or solubilized proteins preabsorbed four times on anti-nucleolin mAb (lane 9) were incubated for 2 h at 4°C with 5 μg anti-nucleolin mAb. Immune complexes were bound to 20-μl protein A/G-agarose beads for 2 h at 4°C. Immunobeads were washed five times in lysis buffer, then proteins were eluted in sample buffer, divided in two identical samples, and submitted to two 7% SDS-PAGE under reducing conditions. After SDS-PAGE and electrotransfer to nitrocellulose sheet, immunodetection was performed with 1 μg/ml anti-phosphotyrosine mAb (A) or anti-nucleolin mAb diluted 1/1000 (B).

CR2 activation triggers binding of tyrosine phosphorylated nucleolin on the p85 subunit of the PI 3 kinase

Our previous demonstration that CR2 activation on Raji cell surface induced the interaction of the p95 tyrosine phosphorylated protein with the p85 subunit of PI 3 kinase (39) led us to analyze the binding properties of nucleolin. Therefore (Fig. 3⇓), solubilized components, prepared from CR2-activated (lane 2) or nonactivated (lane 1) cells, were immunoprecipitated either on anti-p85 Ab or anti-nucleolin mAb and analyzed with different specific Abs. First, analysis with anti-nucleolin (A) or anti-phosphotyrosine (B) Abs of cellular components immunoprecipitated on anti-p85 subunit Abs demonstrated that after 2 min activation of CR2, nucleolin bound on the p85 subunit of PI 3 kinase (A). The demonstration that the tyrosine-phosphorylated p95 component bound on the p85 subunit (B) was nucleolin was completed by analyzing the solubilized components immunoprecipitated on anti-nucleolin mAb with anti-p85 subunit Ab (C) or anti-phosphotyrosine mAb (D). Results demonstrated that although nucleolin was always immunoprecipitated on anti-nucleolin, as already shown in Fig. 2⇑, the p85 subunit interacted with nucleolin (C) only when this latter was phosphorylated (D). In addition, as previously shown (39), the p85 subunit of P I3 kinase coprecipitated with nucleolin was not tyrosine phosphorylated (B and D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Activated CR2 triggers binding of tyrosine-phosphorylated nucleolin on the p85 subunit of the PI 3 kinase. Raji cells (2 × 10⁷) were incubated with 5 μg/ml OKB7 in a final volume of 500 μl at 4°C for 30 min, then activated at 37°C for 2 min (lane 2) or not activated (lane 1). One milligram of solubilized proteins of each assay was incubated for 2 h at 4°C either with 5-μg rabbit anti-p85 subunit of PI 3 kinase (A and B) or with 5-μg anti-nucleolin mAb (C and D). Immune complexes were bound to 20-μl protein A/G-agarose beads for 2 h at 4°C. Immunobeads were washed five times in lysis buffer, then proteins were eluted in sample buffer, divided in two identical samples, and submitted to two 7% SDS-PAGE under reducing conditions. After SDS-PAGE and electrotransfer to nitrocellulose sheet, immunodetection was performed with 1 μg/ml anti-phosphotyrosine mAb (B and D), anti-nucleolin mAb diluted 1/1000 (A) or anti-p85 subunit Ab diluted 1/2000 (C).

Study of the interaction of tyrosine phosphorylated nucleolin with the p85 subunit induced by CR2 activation was completed by using different SH2 domains fused to GST proteins (Fig. 4⇓, I). Solubilized components prepared from CR2-activated (lane 2) or nonactivated (lane 1) Raji cells were incubated with SH2-containing proteins, as the p85 subunit of PI 3 kinase, 3BP2, Grb2, Fyn, or Gap, then bound components were analyzed by immunoblotting with anti-nucleolin mAb. Data demonstrated that nucleolin whose tyrosine phosphorylation was induced only when Raji cells were activated interacted with SH2 domains of p85 subunit (A), 3BP2 (B), or Grb2 (C) but not with SH2 domains of Fyn (D) or Gap (E) or on GST alone (data not shown). The specificity of the interaction of tyrosine-phosphorylated nucleolin with SH2 domains of p85 subunit, 3BP2 and Grb2 but not with SH2 domains of Fyn and Gap is in good agreement with our previous demonstration of the interaction of the tyrosine phosphorylated p95 with these same SH2 domains (39). In addition, taking into account that anti-Grb2 mAb was available (but not anti-3BP2 mAb), herein we also demonstrated that nucleolin coprecipitated with intracellular stores of Grb2 molecules (Fig. 4⇓, II).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Binding of tyrosine-phosphorylated nucleolin on different SH2 domains. Raji cells (2 × 10⁷) were incubated with 5 μg/ml OKB7 in a final volume of 500 μl at 4°C for 30 min, then activated at 37°C for 2 min (lane 2) or not activated (lane 1). One milligram of solubilized proteins of each assay was incubated for 2 h at 4°C: in I, with 5 μg GST-SH2 domain of p85 subunit of PI 3 kinase (A), 3BP2 (B), Grb2 (C), Fyn (D), or Gap (E); in II, with 5 μg anti-nucleolin mAb and immune complexes were bound to 20-μl protein A/G-agarose beads for 2 h at 4°C. Then, beads were washed five times in lysis buffer, then proteins were eluted in sample buffer and submitted to 7% SDS-PAGE under reducing conditions. After SDS-PAGE and electrotransfer to nitrocellulose sheet, immunodetection was performed in I, with anti-nucleolin mAb diluted 1/1000 and in II, either with anti-nucleolin mAb (diluted 1/1000) or anti-Grb2 mAb (diluted 1/5000).

Thus, CR2 activation on Raji cells induced interaction of tyrosine-phosphorylated nucleolin with SH2 domains of p85 subunit of PI 3 kinase and also with SH2 domains of 3BP2 and Grb2.

Tyrosine phosphorylation of nucleolin is induced by CR2 activation and not by CD19 or sIg activation

We previously demonstrated that the 95-kDa protein whose tyrosine phosphorylation was triggered by activated CR2 was not CD19 and that this phosphorylation was independent of CD19 activation (39). Thus, the following experiments were performed. First, CR2 was activated by OKB-7 on K562A cells, which are K562 CD19-negative and CR2-positive cells (as transfected with CR2 cDNA), in the same conditions used for Raji cells (Fig. 5⇓A). In these conditions, although expression of nucleolin was the same both in activated or nonactivated K562A cells (lanes 1 and 2), tyrosine phosphorylation of nucleolin was induced only after CR2 activation (lane 4) and not when K562A cells were not activated (lane 3). Second, because some molecules of CR2 were described associated to sIg or CD19 (22, 23, 24, 25), we also used normal B lymphocytes and analyzed whether cross-linking of sIg or CD19 on normal B lymphocytes surface could also trigger tyrosine phosphorylation of nucleolin. Using cross-linking conditions of CD19 or sIg widely described by others (22, 23, 24, 25), we demonstrated (Fig. 5⇓B) that nucleolin, expressed in normal B lymphocytes (lane 1), was tyrosine phosphorylated after CR2 activation (lane 5). In addition, in the same conditions, tyrosine phosphorylation of nucleolin was neither induced by sIg (lane 3) nor CD19 (lane 4) cross-linking. Nonactivated normal B lymphocytes were used in control (lane 2). All these data supported that tyrosine phosphorylation of nucleolin triggered by activated CR2 occurred in normal B lymphocytes as well as in lymphoma B cells. In addition, this regulation was specific of CR2 as not induced by activated CD19 or sIg.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Tyrosine phosphorylation of nucleolin is induced by CR2 activation and not by CD19 or sIg activation. A, K562A, K562 cells transfected with CR2 cDNA, (2 × 10⁷) were incubated with 5 μg/ml OKB7 in a final volume of 500 μl at 4°C for 30 min, then nonactivated (lanes 1 and 3) or activated at 37°C for 2 min (lane 2 or 4). One milligram of solubilized proteins of each assay was incubated for 2 h at 4°C with 5 μg anti-nucleolin mAb. Immune complexes were bound to 20-μl protein A/G-agarose beads for 2 h at 4°C. Immunobeads were washed five times in lysis buffer, then proteins were eluted in sample buffer, divided in two identical samples, and submitted to 7% SDS-PAGE under reducing conditions. After SDS-PAGE and electrotransfer to nitrocellulose sheet, immunodetection was performed either with 1 μg/ml anti-phosphotyrosine mAb (lanes 3 and 4) or anti-nucleolin mAb diluted 1/1000 (lanes 1 and 2). B, Normal human B lymphocytes (2 × 10⁷) were incubated with 20 μg/ml F(ab′) ₂ of anti-human IgM (lane 3), 4 μg/ml anti-CD19 mAb (lane 4), or 5 μg/ml OKB7 (lane 5), in a final volume of 500 μl at 4°C for 30 min then activated at 37°C for 2 min (lanes 3–5), as described in Material and Methods. In control, cells were incubated, in the same conditions, with GAM alone (lanes 1 and 2). One milligram of solubilized proteins of each assay was immunoprecipitated on anti-nucleolin mAb and processed as mentioned in A, until detection of components on nitrocellulose sheet: immunodetection was performed either with 1 μg/ml anti-phosphotyrosine mAb (lanes 2–5) or anti-nucleolin mAb diluted 1/1000 (lane 1).