Syk Interacts with and Phosphorylates Nucleolin To Stabilize Bcl-x_L mRNA and Promote Cell Survival

Abstract

The Syk protein tyrosine kinase, a well-characterized regulator of immune cell function, plays an increasingly recognized role in tumorigenesis as a promoter of cell survival in both hematological and nonhematological malignancies. We show here that the expression of Syk in MCF7 or MDA-MB-231 breast cancer cells or in DG75 B-lymphoma cells protects cells from apoptosis induced by oxidative or genotoxic stress by stabilizing the mRNA for Bcl-x_L, an antiapoptotic protein. Syk binds robustly to nucleolin and phosphorylates it on tyrosine, enhancing its ability to bind the Bcl-x_L mRNA. Consequently, reducing the level of nucleolin by RNA interference attenuates the ability of Syk to protect cells from stress-induced cell death.

INTRODUCTION

In addition to its well-studied role in normal immune cell biology (1, 2), the Syk protein tyrosine kinase plays an increasingly recognized, albeit poorly understood, role in tumorigenesis. Constitutively active Syk has been reported to promote the survival of non-Hodgkin's lymphoma, acute lymphoblastic leukemia, chronic lymphocytic leukemia, and Epstein-Barr virus-associated B-cell lymphoma (3–11). The inhibition of Syk also promotes the differentiation of acute myeloid leukemia (AML) and attenuates the growth of AML cell lines and primary blasts (12, 13). The formation of a Tel-Syk fusion protein that results from a chromosomal translocation results in a myeloid proliferative disorder, while Itk-Syk fusion proteins are found in some T-cell lymphomas (14, 15). The aberrant expression of Syk itself is also found in a variety of peripheral T-cell lymphomas (16).

Even in nonhematological malignancies, Syk can play an important prosurvival function. Lung and pancreatic carcinomas that are dependent on activated K-Ras for viability are distinguished from those not dependent on K-Ras by the expression of Syk (17). These K-Ras-dependent cells undergo apoptosis in response to the inhibition of Syk activity or knockdown of Syk expression. Retinoblastoma cells in which the expression of Syk is induced by changes in gene methylation also undergo apoptosis in response to reductions in the activity or level of the kinase (18). The survival of breast and ovarian cancer cells is promoted by the alternative splicing of SYK transcripts in response to epidermal growth factor, which enhances expression of the long form of the kinase (19). While the mechanisms by which Syk promotes cancer cell survival are incompletely understood, these observations have led to the exploration of Syk inhibitors as antitumor agents (e.g., see references 18 and 20 to 22).

The ability to evade cell death is one of the fundamental hallmarks of a cancer cell (23). Programmed cell death in eukaryotic cells is regulated through the intrinsic pathway by members of the Bcl-2 family of proteins (24). These proteins function to modulate outer mitochondrial membrane channel opening and the release of cytochrome c necessary for the formation of apoptosomes. The Bcl-2 family includes both pro- and antiapoptotic members. Among these are Bcl-x_L and Bcl-x_S, which are products of alternatively spliced transcripts of the BCL2L1 gene (25). The product of the longer transcript, Bcl-x_L, protects cells from apoptosis, while the smaller Bcl-x_S protein promotes apoptosis by negatively regulating Bcl-x_L and Bcl-2. The relative level of Bcl-x_L and Bcl-x_S in a cell is an important determinant of susceptibility to stress-induced cell death.

In this study, we explored the mechanism by which Syk enhances cell survival by examining its effect on the responses of cancer cells to induced stress. We found that the presence of Syk increases the resistance of several cancer cell types to H₂O₂-induced apoptosis by protecting Bcl-x_L mRNA from degradation by a mechanism that involves the interaction of both Syk and the Bcl-x_L mRNA with nucleolin (NCL). Reductions in the level of nucleolin destabilize the Bcl-x_L message and inhibit the ability of Syk to protect cells from apoptosis induced by both oxidative and genotoxic stress.

MATERIALS AND METHODS

Plasmids and DNA constructs.

For constructing the tetracycline (Tet)-inducible enhanced green fluorescent protein (EGFP)-tagged Syk (Syk-EGFP) lentiviral vectors, cDNAs for Syk-EGFP, Syk-EGFP(K396R), Syk-EGFP(Y317F), Syk-EGFP(Y342F), Syk-EGFP(Y346F), Syk-EGFP(Y342F/Y346F), and Syk-EGFP(Y317F/Y342F/Y346F) were amplified by PCR from the corresponding EGFP-N2 (Clontech) constructs described previously (26). These were then cloned into the Tet-inducible lentiviral vector pLVX-Tight-Puro (Clontech) between the MluI and EcoRI restriction sites. Lentiviral pGIPZ short hairpin RNA (shRNA) sets for the knockdown of nucleolin and Syk were purchased from Thermo Scientific. The Bcl-x_L expression plasmid pSFFV-neo Bcl-x_L (27) was obtained from Addgene (plasmid 8749).

Cell lines.

A line of MCF7 cells lacking endogenous Syk (MCF7-BD) was described previously, as were MCF7-BD cells stably expressing exogenous Syk-EGFP (MCF7-Syk) (28). Syk-deficient MCF7-BD cells with tetracycline-regulated Syk-EGFP expression (MCF7-TRS) were constructed previously using a T-REx system (Invitrogen) (26). These cells were treated with 1 μg/ml doxycycline to induce Syk-EGFP expression. MDA-MB-231 breast cancer cells were obtained from ATCC. A line of MDA-MB-231 cells (MDA-MB-231-TRS) with inducible expression of Syk-EGFP was described previously (26).

MDA-MB-231 cells expressing Syk-EGFP, Syk-EGFP(K396R), or EGFP were constructed using a Lenti-X Tet-On advanced inducible expression system (Clontech). To constitutively express the tetracycline-controlled transactivator rtTA in the Tet-On inducible system, cells were first infected with viral particles with the pLVX Tet-On advanced regulator. Lentiviral particles were generated by cotransfecting HEK293T cells with 4 μg of pLVX-Tet-On, 4 μg of pHR′-CMV-ΔR8.20 vpr, and 2 μg of pHR′-CMV-VSVG using Lipofectamine 2000 (Invitrogen). Supernatants containing viral particles were harvested at 48 h posttransfection and used to infect MDA-MB-231 cells. Two days after infection, cells were selected with 500 μg/ml G418 and screened for rtTA expression. Cells constitutively expressing rtTA protein were infected with lentiviral particles packaged with pLVX-Tight-Puro-Syk-EGFP [or with Syk-EGFP(K396R) or EGFP] as described above. After 48 h, cells were selected with 1 μg/ml puromycin and screened for expression by Western blotting.

To establish nucleolin knockdown cells, MCF7-TRS, MDA-MB-231-TRS, and DG75 cells were stably infected by one of a set of eight lentiviral particles containing an shRNA sequence for NCL. Knockdown of Syk in DG75 cells was generated by stably infecting cells with one of a set of six lentiviral particles containing an shRNA sequence for human Syk. Cell lines were isolated from each lentiviral infection by selection with puromycin (1.0 μg/ml) and screened by Western blotting with antibodies against nucleolin (Abcam) or Syk (Cell Signaling Technology).

All breast cancer cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 7.5% fetal bovine serum, 100 IU/ml penicillin, and 100 μg/ml streptomycin. Human DG75 B-lymphoma cells (ATCC) were grown in RPMI 1640 medium supplemented with 10% fetal calf serum, 50 μM 2-mercaptoethanol, 1 mM sodium pyruvate, 100 IU/ml penicillin G, and 100 μg/ml streptomycin.

DLBCL.

Fresh tumor tissue for ex vivo bioassays was obtained via surgical biopsy from pet dogs with naturally occurring diffuse large B-cell lymphoma (DLBCL). All dogs were seen as clinical patients at the Purdue University Veterinary Teaching Hospital. Written informed consent was obtained from the owners of all dogs, and the biopsy protocol was approved by the Purdue Animal Care and Use Committee. Dogs were anesthetized with propofol (10 mg/kg of body weight intravenously), and a surgical plane of anesthesia was maintained with inhaled isoflurane (1 to 5%). The surgical procedure consisted of an incisional wedge biopsy or complete extirpation of an affected peripheral lymph node. Postoperative analgesia was provided with hydromorphone (0.1 mg/kg subcutaneously). All dogs recovered from surgery uneventfully.

Lymph node biopsy specimens were handled aseptically and divided into sections. One section was placed into 10% neutral buffered formalin to be used for histopathologic diagnosis. Formalin-fixed, paraffin-embedded tissues were stained with hematoxylin-eosin and also prepared routinely for immunohistochemical analysis. The diagnosis of diffuse large B-cell lymphoma was based upon histomorphology and immunohistochemical detection of CD79a and/or CD20, as previously described (29). A second section was placed in RPMI 1640 and processed immediately for use in ex vivo bioassays. Cells were teased apart mechanically and cultured in RPMI 1640 containing 10% heat-inactivated fetal calf serum, 50 μM 2-mercaptoethanol, 1 mM sodium pyruvate, 100 IU/ml penicillin G, and 100 μg/ml streptomycin and supplemented with 10 ng/ml soluble CD40L (InvivoGen).

Antibodies and immunoprecipitation.

Antibodies against Syk, poly(ADP-ribose) polymerase (PARP), active caspase 3, and Bcl-x_L were purchased from Cell Signaling Technology. Antinucleolin and anti-γ-tubulin were from Abcam, antiphosphotyrosine (4G10) was from Millipore, anti-green fluorescent protein (anti-GFP) was from Santa Cruz, and anti-GAPDH (anti-glyceraldehyde-3-phosphate dehydrogenase) was from Ambion.

For Syk and nucleolin coimmunoprecipitation assays, 5 × 10⁶ MDA-MB-231 cells with inducible expression of Syk-EGFP, EGFP, or one of the site-directed Syk-EGFP mutants were treated with doxycycline (1 μg/ml) for 18 h. Cells were lysed in 20 mM Tris-HCl, pH 8, 137 mM NaCl, 2 mM EDTA, 1% NP-40, 10% glycerol, 1× protease inhibitor cocktail (Sigma), and 2 mM Na₃VO₄. EGFP or EGFP-tagged proteins were immunoprecipitated using GFP-Trap agarose beads (ChromoTek). Bound immune complexes were washed with lysis buffer, separated by SDS-PAGE, and analyzed by Western blotting with the indicated antibodies. Where indicated, immune complexes containing Syk-EGFP or Syk-EGFP(K396R) and associated nucleolin were incubated in 25 mM HEPES, pH 7.2, 5 mM MnCl₂, 0.5 mM Na₃VO₄, 0.02 mg/ml leupeptin, and 0.02 mg/ml aprotinin with or without 1 mM ATP at 37°C for 10 min. Immune complexes were separated by SDS-PAGE. Nucleolin was detected by Western blotting.

RT-PCR analyses of BCL2L1 transcripts.

Total cellular RNA was isolated from 5 ×10⁶ cells using the QIAzol reagent (Qiagen) according to the manufacturer's protocol. RNA (2 μg) from each sample was reverse transcribed and amplified using a SuperScript III Platinum One-Step quantitative reverse transcription-PCR (RT-PCR) system (Invitrogen) containing 0.1 μM each BCL2L1 forward and reverse primers in a 25-μl reaction mix. The primer sequences used were CATGGCAGCAGTAAAGCAAG and GCATTGTTCCCATAGAGTTCC for the human transcripts and GCATTGTTCCCGTAGAGTTCC and GATCATCTCGCGCTACTTG for the canine transcripts. The primers used to amplify the canine GAPDH transcripts were TGATTCTACCCACGGCAAATTC and TCATGGTTCACGCCCATCAC. The reverse transcription was performed at 50°C for 30 min. PCRs were optimized for semiquantification by activating Platinum Taq polymerase at 95°C for 2 min and carried out in 25 cycles of two-step conditions of 95°C for 15 s and 60°C for 30 s and finally ending with a 5-min cycle extension at 72°C. An aliquot (10 μl) of each reaction mixture was separated on a 2% agarose gel, visualized by ethidium bromide staining, and quantified with ImageJ software. Where indicated, cells were treated prior to lysis with H₂O₂ (5 mM).

RNA immunoprecipitation assay.

MDA-MB-231 cells were treated without or with 5 mM H₂O₂ and lysed in RNA immunoprecipitation (RIP) buffer (50 mM Tris HCl, pH 7.4, 150 mM NaCl, 0.5 mM dithiothreitol, 1% NP-40, 10 mM vanadyl ribonucleoside complex [NEB], 2 mM phenylmethylsulfonyl fluoride, 1× protease inhibitor cocktail [Sigma]) for 15 min on ice. After centrifugation at 14,000 × g for 10 min, antibody to nucleolin was added to the supernatant and the mixture was incubated at 4°C for 4 h. Prior to precipitation, protein A-agarose beads (40 μl) were added and the mixture was incubated for 1 h at 4°C with gentle rotation. The beads were then washed three times with RIP buffer and once in phosphate-buffered saline. One-fourth of the precipitate was used to detect nucleolin by Western blotting following SDS-PAGE. RNAs in the nucleolin precipitates were isolated with 1 ml of QIAzol reagent according to the manufacturer's instructions. RNAs were resuspended in 20 μl nuclease-free water and subjected to semiquantitative RT-PCR to detect BCL2L1 transcripts.

PARP cleavage assay.

To detect the cleavage of PARP, lysates of 5 × 10⁶ cells treated with 5 mM H₂O₂ or 1 μg/ml doxorubicin (Calbiochem) for the times indicated below were separated by SDS-PAGE and subjected to Western blot analysis using anti-PARP antibodies. Alternatively, activated caspase 3 was detected by Western blotting using antibodies specific for the cleaved caspase.

RESULTS

Syk stabilizes Bcl-x_L mRNA in MCF7 breast cancer cells treated with H₂O₂.

MCF7 breast cancer cells have an epithelial morphology and express endogenous Syk (30). We had previously identified a line of MCF7 cells (MCF7-BD) that lacks Syk and is able to survive in its absence (28). We generated from these cells a new line in which the expression of Syk (as Syk-EGFP) could be induced by treatment with doxycycline (26). We then compared the responses of the Syk-deficient cells to those of cells induced to express Syk-EGFP to an external stress: exposure to H₂O₂. The activation of executioner caspases was monitored by Western blotting of cleavage products of poly(ADP-ribose) polymerase (PARP) (31). Cells expressing Syk-EGFP were more resistant to H₂O₂-induced PARP cleavage than were cells lacking the kinase (Fig. 1A). Consistent with this observation, treatment of Syk-deficient MCF7-BD cells with H₂O₂ caused them to round up and detach from the tissue culture plate, an effect that was attenuated in Syk-EGFP-expressing cells (see Fig. S1 in the supplemental material).

FIG 1.

Syk expression protects MCF7 cells from oxidative stress-induced apoptosis and degradation of Bcl-x_L mRNA. (A) MCF7-BD cells lacking Syk (−) or stably expressing Syk-EGFP (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by SDS-PAGE and Western blotting with antibodies against PARP (top). The cleaved form of PARP is indicated by the arrow. The expression of Syk-EGFP was visualized by Western blotting of cell lysates (bottom). (B) MCF7-BD cells lacking Syk (−) or stably expressing Syk-EGFP (+) were exposed to 5 mM H₂O₂ for 30 min (pulse) and then moved to fresh medium for the indicated total incubation times or treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA (top) or by Western blotting to detect expressed Syk-EGFP (bottom). (C) Comparison of relative levels of Bcl-x_L mRNA. Changes in the ratio of Bcl-x_L mRNA to Bcl-x_S mRNA were normalized to their relative levels of expression in Syk-deficient cells at time zero, which was set equal to a value of 1.0. Bars represent means ± SEMs from three replicate experiments. Significant differences between pairs were determined using an unpaired, two-tailed Student's t test. *, P < 0.01; **, P < 0.005; ***, P < 0.001.

Bcl-x_L, the product of the longer BCL2L1 gene transcript, is a potent antiapoptotic factor in breast cancer cells (32, 33). To explore the mechanism by which Syk influences the response of MCF7 cells to oxidative stress, we measured the levels of mRNA for Bcl-x_L and Bcl-x_S in cells expressing or lacking Syk as a function of time of exposure to H₂O₂. The levels of each mRNA were equivalent in untreated cells whether they were lacking or expressing Syk (Fig. 1B and C). The treatment of MCF7-BD cells with H₂O₂ resulted in a striking reduction in the level of the longer Bcl-x_L transcript. In cells expressing Syk-EGFP, the decrease in the level of mRNA for Bcl-x_L in response to oxidative stress was markedly attenuated. This protective effect of Syk-EGFP was observed both in cells treated with a short 30-min pulse of H₂O₂ and then monitored over time and in cells in which H₂O₂ remained in the medium for 2 to 4 h. The protective effect of induced Syk-EGFP on Bcl-x_L mRNA was retained in cells pretreated with actinomycin D, an inhibitor of gene transcription (see Fig. S2 in the supplemental material). Thus, Syk functioned to stabilize preexisting Bcl-x_L mRNA rather than alter its rate of transcription.

Syk stabilizes Bcl-x_L mRNA in MDA-MB-231 cells treated with H₂O₂.

MDA-MB-231 breast cancer cells are highly invasive, have a mesenchymal phenotype, and lack endogenous Syk (30). A clone of cells transfected with the pcDNA6/TR regulatory plasmid was isolated (MDA-MB-231-TR), and from these we generated clones that contained a Tet-inducible plasmid expressing Syk-EGFP (MDA-MB-231-TRS) (26). The MDA-MB-231-TRS cells express low levels of Syk-EGFP in the absence of doxycycline due to leaky expression from the inducible plasmid but can be induced to express higher levels of Syk-EGFP upon treatment with doxycycline (Fig. 2A). The levels of Bcl-x_L and Bcl-x_S mRNA were comparable, regardless of the expression level of Syk. The treatment of MDA-MB-231-TR cells with H₂O₂ resulted in a reduction in the level of the longer transcript (Fig. 2A and B). The times of exposure to H₂O₂ required to produce this decreased level of Bcl-x_L mRNA were longer in MDA-MB-231-TR cells than in MCF7-BD cells. The decrease in Bcl-x_L mRNA level in Syk-deficient cells resulted in a concomitant reduction in the level of expression of Bcl-x_L protein (Fig. 2C). The induction of Syk-EGFP protected the MDA-MB-231 cells from the oxidative stress-induced decrease in Bcl-x_L mRNA (Fig. 2A and B). Even low levels of Syk-EGFP resulting from leaky transcription partially rescued the cells from the decrease in Bcl-x_L mRNA caused by oxidative stress. This protective effect of Syk was again retained in cells pretreated with actinomycin D (see Fig. S2 in the supplemental material). The induced expression of Syk also protected MDA-MB-231 cells from H₂O₂-induced activation of caspase activity, as visualized by a decreased cleavage of PARP (Fig. 2D).

FIG 2.

Syk expression protects MDA-MB-231 cells from oxidative stress-induced apoptosis and degradation of Bcl-x_L mRNA. (A) MDA-MB-231-TR (TR) cells lacking Syk or MDA-MB-231-TRS (TRS) cells either induced (+) or not induced (−) with doxycycline (Tet) to express Syk-EGFP were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA (top) or by Western blotting to detect expressed Syk-EGFP (bottom). (B) Comparison of relative levels of Bcl-x_L mRNA to Bcl-x_S mRNA. Ratios were normalized to a value of 1.0 for Syk-deficient cells at time zero. Bars represent means ± SEMs from three replicate experiments. *, P < 0.05; **, P < 0.01. (C) MDA-MB-231-TR cells (−) or doxycycline-induced MDA-MB-231-TRS cells (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by SDS-PAGE and Western blotting (WB) with antibodies against Bcl-x_L (top) and Syk-EGFP (bottom). (D) MDA-MB-231-TR cells lacking Syk (−) or MDA-MB-231-TRS cells induced with doxycycline to express Syk-EGFP (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by SDS-PAGE and Western blotting with antibodies against PARP. The cleaved form of PARP is indicated by the arrow.

Syk stabilizes Bcl-x_L mRNA in DG75 cells treated with H₂O₂.

To determine if Syk also plays a prosurvival role when expressed at normal endogenous levels, we examined the effects of a knockdown in the level of the kinase on Bcl-x_L mRNA expression in a human Burkitt's lymphoma. DG75 B cells were infected with a lentivirus expressing shRNA targeted against human Syk mRNA. Of six lentiviruses tested, one resulted in a greater than 80% reduction in the level of the endogenous protein (Fig. 3A). The level of Bcl-x_L mRNA was consistently lower in DG75 cells in which the expression of Syk was suppressed. Cells expressing either normal or reduced levels of Syk were treated with H₂O₂ for various periods of time and examined for changes in the levels of Bcl-x_L and Bcl-x_S mRNA. Cells with a reduced level of endogenous Syk exhibited a markedly reduced level of Bcl-x_L mRNA when treated with H₂O₂ compared to that exhibited by cells expressing a normal level of Syk (Fig. 3A and B). Again, this effect was retained in cells pretreated with actinomycin D (see Fig. S2 in the supplemental material). DG75 cells with a reduced level of Syk also exhibited an increased sensitivity to oxidative stress, as measured by the cleavage of PARP (Fig. 3C). To confirm that this increased sensitivity resulted from a reduced level of Bcl-x_L protein, we transfected the wild-type and Syk knockdown cells with a Bcl-x_L expression plasmid to increase the intracellular level of the protein. Increasing the level of Bcl-x_L reduced the sensitivity of the DG75 cells to oxidative stress-induced apoptosis, as measured by the cleavage of PARP in both cell types (Fig. 3D). Together, these results indicate that Syk functions to protect Bcl-x_L mRNA from degradation in response to oxidative stress in multiple cell lines, thus protecting them from apoptosis.

FIG 3.

The knockdown of Syk expression sensitizes DG75 B lymphoma cells to oxidative stress-induced apoptosis and degradation of Bcl-x_L mRNA. (A) DG75 B lymphoma cells (−) or DG75 cells stably expressing shRNA targeted against Syk (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA (top) or by Western blotting to detect endogenous Syk (bottom). (B) Comparison of relative levels of Bcl-x_L to Bcl-x_S mRNA. Ratios were normalized to a value of 1.0 for wild-type DG75 B cells at time zero. Bars represent means ± SEMs from three replicate experiments. *, P < 0.05; **, P < 0.01. (C) DG75 B lymphoma cells (−) or DG75 cells stably expressing shRNA targeted against Syk (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by SDS-PAGE and Western blotting with antibodies against PARP. The cleaved form of PARP is indicated by the arrow (top). The expression level of endogenous Syk was probed by Western blotting with antibodies against Syk (bottom). (D) DG75 B lymphoma cells (−) or DG75 cells stably expressing shRNA targeted against Syk (+) were transiently transfected with a Bcl-x_L expression plasmid (+) or empty vector (−) and then treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by SDS-PAGE and Western blotting with antibodies against PARP (top), Syk (middle), or Bcl-x_L (bottom). The cleaved form of PARP is indicated by the arrow.

Syk promotes the survival of DLBCL.

To determine if the protective effect of Syk on cell viability extended to clinical samples, we examined the effects of the Syk inhibitor R406 on biopsy specimens isolated from dogs suffering from naturally occurring diffuse large B-cell lymphoma (DLBCL). DLBCL cells die rapidly in culture, but their life span can be enhanced by the addition of CD40L (34). DLBCL cells expressed Syk, the activity of which could be inhibited by treatment with R406, as measured by Western blotting of whole-cell lysates with antibodies against phosphotyrosine (see Fig. S3A in the supplemental material). Treatment of DLBCL cells with R406 enhanced cell death, as judged by the approximately 3-fold increase in PARP cleavage measured 12 or 24 h after isolation (see Fig. S3C and D in the supplemental material), and enhanced the rate of loss of Bcl-x_L mRNA in the absence or presence of oxidative stress (see Fig. S3E in the supplemental material).

The catalytic activity of Syk is required for stabilization of Bcl-x_L mRNA.

To determine if Syk-catalyzed protein phosphorylation was important for the protective effect of Syk on Bcl-x_L mRNA, we generated two populations of MDA-MB-231 cells using a Lenti-X Tet-On inducible expression system in which either Syk-EGFP or a catalytically inactive mutant, Syk-EGFP(K396R), could be induced by treatment with doxycycline. This system provides more tightly controlled expression of inducible genes. We then compared the sensitivity of the Bcl-x_L transcript to H₂O₂-stimulated degradation in cells induced or not induced to express either Syk-EGFP or Syk-EGFP(K396R). The expression of Syk-EGFP, but not that of Syk-EGFP(K396R), protected the Bcl-x_L mRNA from stress-induced degradation (Fig. 4A). Thus, the ability of Syk to protect cells from the oxidative stress-induced loss of Bcl-x_L mRNA required its catalytic activity. This suggests the involvement of one or more substrates of Syk in the regulation of Bcl-x_L mRNA turnover.

FIG 4.

Syk interacts with nucleolin. (A) MDA-MB-231 cells expressing rtTA but not Syk or MDA-MB-231 cells with Tet-regulated expression of Syk-EGFP or Syk-EGFP(K396R) (Lenti-X Tet-On) pretreated with doxycycline (+) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA (top) or by Western blotting with anti-Syk antibodies to detect Syk-EGFP or Syk-EGFP(K396R) (bottom). (B) Tet-responsive MDA-MB-231 cells not induced (−) or induced with doxycycline to express Syk-EGFP (+) were treated with 5 mM H₂O₂ for the indicated times. Syk-EGFP was immunoprecipitated (IP) from cell lysates with GFP-nanotrap beads. Anti-GFP immune complexes were separated by SDS-PAGE and analyzed by Western blotting (WB) with antibodies against NCL, γ-tubulin, or GFP (to detect Syk-EGFP). Whole-cell lysates (WCL) were analyzed by Western blotting with antibodies against phosphotyrosine (pTyr) (bottom). The migration position of the 50-kDa molecular mass marker is indicated. (C) Syk-EGFP was immunoprecipitated with GFP-nanotrap beads from lysates of Tet-responsive MDA-MB-231 cells induced to express Syk-EGFP. Immune complexes and whole-cell lysates were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL or γ-tubulin. (D) Proteins were immunoprecipitated with GFP-nanotrap beads from lysates of Tet-responsive MDA-MB-231 cells induced to express Syk-EGFP (Syk) or Syk-EGFP(K396R) (KD) and treated with (+) or without (−) 5 mM H₂O₂. Immune complexes and whole-cell lysates were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL (top) and, to detect Syk-GFP, antibodies against GFP (bottom). (E) Proteins were immunoprecipitated with GFP-nanotrap beads from lysates of Tet-responsive MDA-MB-231 cells induced to express EGFP (lane GFP), Syk-EGFP (lane Syk), Syk-EGFP(Y342F/Y346F) (lane 2F), Syk-EGFP(Y317F/Y342F/Y346F) (lane 3F), Syk-EGFP(Y317F) (lane 317), Syk-EGFP(Y342F) (lane 342), or Syk-EGFP(Y346F) (lane 346) and treated with (+) or without (−) 5 mM H₂O₂. Immune complexes and whole-cell lysates were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL (top) or GFP (bottom).

Syk interacts with nucleolin.

In a previous mass spectrometry-based analysis of Syk-interacting proteins, we identified nucleolin as a Syk-binding protein (35). Interestingly, nucleolin has been shown to stabilize Bcl-x_L mRNA by binding directly to the AU-rich elements (AREs) in the 3′ untranslated region (UTR) that mediate its rapid turnover (36, 37). To confirm an interaction between Syk and nucleolin, we immunoprecipitated Syk-EGFP from the Tet-responsive MDA-MB-231 cells that had been induced with doxycycline to express the fusion protein and then searched the immune complex for associated nucleolin. Western blotting analyses of proteins present in the anti-Syk-EGFP immune complexes confirmed that nucleolin is a Syk-binding protein (Fig. 4B and C).

The treatment of cells that express Syk with H₂O₂ leads to its activation and to an increase in the phosphorylation of Syk substrates on tyrosine (38). As expected, treatment of MDA-MB-231 cells induced to express Syk-EGFP with H₂O₂ led to a transient increase in the level of proteins phosphorylated on tyrosine present in whole-cell lysates (Fig. 4B). The interaction between Syk-EGFP and nucleolin, as measured by coimmunoprecipitation, was the strongest in untreated cells and decreased as a function of time following the addition of H₂O₂. As a control, we also monitored the interaction of Syk with γ-tubulin, as the treatment of cells with H₂O₂ promotes the interaction of Syk with centrosomes (39). As expected, H₂O₂ treatment led to an increase in the amount of γ-tubulin in the anti-Syk immune complexes (Fig. 4B), indicating that H₂O₂ treatment did not, in general, block all protein-protein interactions involving Syk. Treatment of cells with H₂O₂ had no apparent effect on the overall level of expression of either nucleolin or γ-tubulin in the whole-cell lysate (Fig. 4C). These observations suggested that the interaction between Syk and nucleolin would be most robust when neither protein was phosphorylated on tyrosine. To test this, we immunoprecipitated either Syk-EGFP or the inactive (kinase-dead) Syk-EGFP(K396R) from the corresponding doxycycline-induced MDA-MB-231 cells treated with or without H₂O₂ and examined the resulting immune complexes for the presence of nucleolin. Indeed, the inactive version of Syk bound robustly to nucleolin, and this interaction was refractory to treatment with H₂O₂ compared to the interaction between Syk-EGFP and nucleolin (Fig. 4D).

The Syk-nucleolin interaction is disrupted by the phosphorylation of Syk in linker B.

The Syk molecule comprises an N-terminal tandem pair of SH2 domains separated by a linker (linker B) from the C-terminal catalytic domain. To explore further the mechanism of the H₂O₂-induced dissociation of Syk and nucleolin, we generated an additional series of MDA-MB-231 cells in which various EGFP-tagged variants of Syk could be inducibly expressed. These included a set of Syk mutants in which known sites of tyrosine phosphorylation in the linker B region (40) were replaced by phenylalanines: Syk-EGFP(Y317F), Syk-EGFP(Y342F), Syk-EGFP(Y346F), Syk-EGFP(Y342F/Y346F), and Syk-EGFP(Y317F/Y342F/Y346F). As a control, cells inducibly expressing EGFP alone were also generated. The various EGFP-tagged proteins were immunoprecipitated with an immobilized GFP-binding protein, and the presence of nucleolin in the immune complex was detected by Western blotting (Fig. 4E). No nucleolin coimmunoprecipitated with EGFP. The replacement of Y317, the binding site on Syk for Cbl-family proteins, had no significant effect on either the binding of Syk to nucleolin or its dissociation in response to H₂O₂. Forms of the kinase lacking single tyrosines at either position 342 or 346 bound nucleolin and largely dissociated from it in response to H₂O₂. Interestingly, the H₂O₂-induced dissociation from nucleolin of forms of Syk lacking both Y342 and Y346 or these two sites plus Y317 was compromised. Since the treatment of cells with H₂O₂ is known to result in the extensive phosphorylation of the linker B tyrosines on Syk (40), these results are consistent with the phosphorylation of Syk at both Y342 and Y346 disrupting the interaction between Syk and nucleolin.

Syk phosphorylates nucleolin.

The nucleolin associated with Syk appeared to undergo a change in its electrophoretic mobility when it was analyzed by SDS-PAGE (Fig. 4B and C) following the treatment of cells with H₂O₂, suggesting that it was covalently modified. To determine if nucleolin was phosphorylated on tyrosine in Syk-expressing cells, we treated MDA-MB-231 cells lacking or expressing Syk-EGFP with or without H₂O₂ and immunoprecipitated the tyrosine-phosphorylated proteins using antibodies against phosphotyrosine. Nucleolin could be detected in the antiphosphotyrosine immune complexes recovered from lysates of H₂O₂-treated cells that expressed Syk-EGFP (Fig. 5A) or endogenous Syk (see Fig. S3B in the supplemental material). Similarly, tyrosine-phosphorylated nucleolin could be isolated from lysates of DG75 B cells treated with H₂O₂ but not from cells also treated with the Syk inhibitor piceatannol or R406 (Fig. 5B and C). Nucleolin isolated from H₂O₂-treated MDA-MB-231-TRS cells expressing Syk-EGFP (Fig. 5D) or DG75 cells expressing normal levels of endogenous Syk (Fig. 5E) reacted on Western blots with antibodies against phosphotyrosine, indicating that nucleolin itself was directly phosphorylated on tyrosine in a Syk-dependent manner.

FIG 5.

Syk phosphorylates nucleolin and promotes its binding to Bcl-x_L mRNA. (A) Tet-responsive MDA-MB-231 cells pretreated without or with doxycycline to induce Syk-EGFP were treated without or with 5 mM H₂O₂ for 15 min. Tyrosine-phosphorylated proteins were immunoprecipitated from cell lysates with antibodies against phosphotyrosine. Immune complexes (top) and whole-cell lysates (WCL; bottom) were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL. (B) DG75 B cells were pretreated with 50 μM piceatannol (PIC; +) or dimethyl sulfoxide carrier alone (−) and then treated without or with 5 mM H₂O₂ for 15 min. Tyrosine-phosphorylated proteins were immunoprecipitated from cell lysates with antibodies against phosphotyrosine. Immune complexes (top) and whole-cell lysates (bottom) were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL. (C) DG75 cells were pretreated with the indicated concentrations of R406 and then treated without or with 5 mM H₂O₂ for 15 min. Tyrosine-phosphorylated proteins were immunoprecipitated from cell lysates with antibodies against phosphotyrosine. Immune complexes (top) and whole-cell lysates (bottom) were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL. (D) MDA-MB-231-TR (TR) or MDA-MB-231-TRS cells induced to express Syk-EGFP (TRS) were treated without or with 5 mM H₂O₂ for 15 min. Nucleolin was immunoprecipitated, and the resulting immune complexes were probed by Western blotting for phosphotyrosine (top) or NCL (bottom). (E) DG75 B cells (−) or DG75 cells stably expressing shRNA targeted against Syk (+) were treated without or with 5 mM H₂O₂ for 15 min. Nucleolin was immunoprecipitated, and the resulting immune complexes were probed by Western blotting for phosphotyrosine (top) or NCL (bottom). (F) Syk-EGFP (Syk) or Syk-EGFP(K396R) (KD) was immunoprecipitated from the corresponding doxycycline-induced lines of MDA-MB-231 cells using GFP-nanotrap beads. The resulting immune complexes were incubated with buffer containing (+) or lacking (−) ATP. The immune complexes and whole-cell lysates were separated by SDS-PAGE and analyzed by Western blotting with antibodies against NCL. (G) Nucleolin was immunoprecipitated from Tet-responsive MDA-MB-231 cells either uninduced (−) or induced (+) with doxycycline to express Syk-EGFP and either treated with 5 mM H₂O₂ for 3 h or not treated. Immune complexes were examined for the presence of Bcl-x_L mRNA by RT-PCR (top) and nucleolin by Western blotting (middle). The expression of Syk-EGFP was determined by Western blotting of whole-cell lysates with antibodies against Syk (bottom). (H) The relative amount of Bcl-x_L mRNA associated with nucleolin, analyzed as described in the legend to panel G, was quantified. The data represent means ± SEMs from three replicate experiments. The level of mRNA bound to nucleolin in Syk-EGFP-expressing cells not treated with H₂O₂ was set equal to a value of 1.0.

To confirm further that the associated Syk could directly phosphorylate nucleolin, resulting in a shift in its electrophoretic mobility, we immunoprecipitated either Syk-EGFP or Syk-EGFP(K396R) from each Tet-induced MDA-MB-231 cell line along with the associated nucleolin and incubated the resulting immune complexes in vitro in a kinase reaction buffer containing ATP. The immune complexes were then analyzed by SDS-PAGE and Western blotting for the presence of nucleolin. Incubation with ATP induced a shift in the mobility of nucleolin in the immune complex containing active Syk-EGFP but not in the immune complex containing Syk-EGFP(K396R) (Fig. 5F). Together, these results indicate that Syk interacts with and phosphorylates nucleolin on tyrosine.

Nucleolin binds the Bcl-x_L mRNA in a Syk-dependent manner.

Nucleolin has been reported to stabilize Bcl-x_L mRNA through direct binding (37). To determine how Syk might influence such an interaction, we immunoprecipitated nucleolin from untreated or doxycycline-induced, Syk-EGFP-expressing MDA-MB-231 cells with antinucleolin antibodies, extracted RNA from the immune complexes, and analyzed it by semiquantitative PCR for the presence of the Bcl-x_L mRNA. Interestingly, the association of nucleolin with the Bcl-x_L mRNA was largely dependent on the expression of Syk-EGFP (Fig. 5G and H). Since the treatment of cells with H₂O₂ resulted in the phosphorylation of nucleolin on tyrosine, we examined the effect of H₂O₂ on the association of nucleolin with the Bcl-x_L mRNA. The treatment of cells with H₂O₂ further enhanced the binding interaction between nucleolin and the Bcl-x_L transcript selectively in Syk-expressing cells (Fig. 5G and H).

Nucleolin is required for the Syk-dependent stabilization of Bcl-x_L mRNA.

To confirm that nucleolin plays a direct role in the ability of Syk to regulate Bcl-x_L mRNA stability, we infected Tet-responsive, Syk-EGFP-expressing MDA-MB-231 cells with a series of lentiviruses coding shRNAs directed against nucleolin and selected populations of cells with a reduced level of the protein (Fig. 6A). These cells were then treated with H₂O₂ and examined for the levels of Bcl-x_L mRNA. While the induced expression of Syk-EGFP protected Bcl-x_L mRNA from H₂O₂-induced degradation, this effect was essentially abrogated in cells that had a reduced level of nucleolin (Fig. 6B). Similarly, the knockdown of nucleolin in DG75 B cells, which express endogenous Syk, increased their sensitivity to the oxidative stress-induced loss of Bcl-x_L mRNA (Fig. 6C and D). The knockdown of nucleolin also reduced the ability of Syk-EGFP to protect Bcl-x_L mRNA from H₂O₂-induced degradation in Tet-responsive MCF7 cells (Fig. 6E and F). Thus, the ability of Syk to stabilize Bcl-x_L mRNA was dependent on a normal level of expression of nucleolin in all three cell types.

FIG 6.

Nucleolin is required for Syk-dependent stabilization of Bcl-x_L mRNA. (A) Tet-responsive MDA-MB-231 cells were untreated (control [Ctrl]) or infected with one of a set of lentiviruses encoding shRNAs for nucleolin (shNCL). Nucleolin levels were measured by Western blotting. The level of GAPDH was measured as a loading control. Results from three different populations of infected cells are shown. (B) Tet-responsive MDA-MB-231 cells and two of the three sets of Tet-responsive cells carrying the nucleolin shRNA (shNCL2 and shNCL3) were either uninduced (−) or induced with doxycycline to express Syk-EGFP (+) and then treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA. (C) DG75 B cells were untreated (control) or infected with a lentivirus encoding shRNA directed against nucleolin. Nucleolin levels were measured by Western blotting. The level of GAPDH was measured as a loading control. (D) DG75 cells either infected (+) or not infected (−) with the lentivirus carrying the nucleolin shRNA were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA. (E) Tet-responsive MCF7 cells were untreated (control) or infected with a set of lentiviruses encoding shRNAs for nucleolin. Nucleolin levels were measured by Western blotting. Results from two different populations of infected cells are shown. (F) Tet-responsive MCF7 cells and the two sets of Tet-responsive cells carrying the nucleolin shRNA (shNCL1 and shNCL2) were either uninduced (−) or induced with doxycycline to express Syk-EGFP (+) and then treated with 5 mM H₂O₂ for the indicated times. Cell lysates were analyzed by RT-PCR to measure the levels of Bcl-x_L and Bcl-x_S mRNA.

To determine if decreasing the level of nucleolin also increased the sensitivity of Syk-expressing cells to oxidative stress-induced apoptosis, we measured the cleavage of PARP in MCF7 cells in which the expression of Syk-EGFP was induced and the level of nucleolin was either normal or reduced by shRNA expression. The ability of expressed Syk-EGFP to inhibit H₂O₂-induced PARP cleavage was decreased by the knockdown of nucleolin (Fig. 7A). Syk-expressing cells with a reduced level of nucleolin also rounded up and lost cell-cell contacts in response to oxidative stress to a much greater extent than did cells with normal levels of the protein (see Fig. S1 in the supplemental material). A similar increased sensitivity to an apoptotic stimulus was seen for DG75 B cells, in which the knockdown of either Syk or nucleolin enhanced their susceptibility to PARP cleavage induced by exposure to H₂O₂ (Fig. 7B). For both cell types, the effects of Syk and nucleolin expression on PARP cleavage were observed in multiple biological replicates (Fig. 7E and F). Syk also protected cells from etoposide-induced cell death, as shown previously in MCF7 cells (41) and in MDA-MB-231 cells, as measured by the activation of caspase 3 (see Fig. S4 in the supplemental material).

FIG 7.

Nucleolin is required for Syk-dependent protection of cells from stress-induced apoptosis. (A) Tet-responsive MCF7 cells or Tet-responsive cells carrying the nucleolin shRNA were either uninduced or induced with doxycycline to express Syk-EGFP and then treated with 5 mM H₂O₂ for the indicated times. Cell lysates were separated by SDS-PAGE and probed with antibodies against PARP (top), nucleolin (middle), or Syk (bottom). The cleaved form of PARP is indicated by the arrow. (B) DG75 cells or DG75 cells expressing the shRNA targeting either Syk (Syk-shRNA) or nucleolin (NCL-shRNA) were treated with 5 mM H₂O₂ for the indicated times. Cell lysates were separated by SDS-PAGE and probed with antibodies against PARP (top), nucleolin (middle), or Syk (bottom). The cleaved form of PARP is indicated by the arrow. (C) Tet-responsive MCF7 cells or Tet-responsive cells carrying the nucleolin shRNA were either uninduced or induced with doxycycline to express Syk-EGFP and then treated with 1 μg/ml doxorubicin (Dox) for the indicated times. Cell lysates were separated by SDS-PAGE and probed with antibodies against PARP (top), nucleolin (middle), or Syk (bottom). The cleaved form of PARP is indicated by the arrow. (D) DG75 cells or DG75 cells expressing either the shRNA targeting Syk (Syk-shRNA) or nucleolin (NCL-shRNA) were treated with 1 μg/ml doxorubicin for the indicated times. Cell lysates were separated by SDS-PAGE and probed with antibodies against PARP (top), nucleolin (middle), or Syk (bottom). The cleaved form of PARP is indicated by the arrow. (E) The degree of PARP cleavage was quantified from Western blots of lysates of MCF7 cells lacking Syk (no Syk), expressing Syk-EGFP (SykGFP), or expressing Syk-EGFP and shRNA for nucleolin (shNCL) and treated for 24 h with 5 mM H₂O₂ (left) or 1 μg/ml doxorubicin (right). The data represent means ± SEMs from three replicate experiments. **, P < 0.005 (compared to Syk-EGFP-expressing cells); ***, P < 0.001 (compared to Syk-EGFP-expressing cells). (F) The degree of PARP cleavage was quantified from Western blots of lysates of DG75 cells expressing shRNA for Syk (shSyk), wild-type DG75 cells (WT), or wild-type cells expressing shRNA for nucleolin (shNCL) and treated for 24 h with 5 mM H₂O₂ (left) or 1 μg/ml doxorubicin (right). The data represent means ± SEMs from three replicate experiments. *, P < 0.05 (compared to wild-type cells); **, P < 0.005 (compared to wild-type cells); ***, P < 0.001 (compared to wild-type cells).

Nucleolin is required for Syk-dependent resistance to genotoxic stress.

We had shown previously that the expression of Syk in MCF7 cells protected them from apoptosis induced by exposure to the genotoxic agents doxorubicin and etoposide (41). To determine if nucleolin plays a role in the capacity of Syk to protect cells from genotoxic agents, we monitored the ability of induced Syk-EGFP to protect Syk-deficient MCF7 cells from doxorubicin-induced PARP cleavage in cells expressing either a normal or a reduced level of nucleolin. As shown in Fig. 7C, the ability of Syk to protect MCF7 cells from doxorubicin-induced apoptosis was attenuated in cells with a reduced level of nucleolin. This protective role for Syk was also seen in DG75 cells, in which the knockdown of either Syk or nucleolin by shRNA increased their susceptibility to the doxorubicin-induced cleavage of PARP (Fig. 7D). This effect was observed in multiple biological replicates (Fig. 7E). Thus, Syk stabilizes Bcl-x_L mRNA against the actions of genotoxic agents in a manner also dependent on the expression of normal levels of nucleolin.

DISCUSSION

The Syk protein tyrosine kinase is best known for its roles in hematopoietic cells, where it is required for signaling through immune recognition receptors that contain immunoreceptor tyrosine-based activation motifs (ITAMs) (1, 2). Examples include the antigen receptor on B cells (BCR), the high-affinity IgE receptor of mast cells, and IgG receptors on neutrophils and macrophages. Receptor clustering leads to the phosphorylation of a pair of ITAM tyrosines, creating a high-affinity docking site for the tandem pair of Syk SH2 domains. The binding of Syk to the receptor activates the enzyme, which becomes phosphorylated on tyrosine and functions as both a kinase and a scaffold to couple the receptor to downstream effectors of multiple signaling pathways. In certain malignancies of B-cell origin, it is tonic signaling from the BCR that is proposed to activate Syk to promote cell survival (4, 7, 8). Interestingly, the repertoire of cells in which Syk functions as a prosurvival factor extends to tumors of B-cell origin that have not yet rearranged immunoglobulin genes, hematological malignancies not of B-cell origin, and nonhematological cancers, such as retinoblastoma and certain carcinomas of the lung and pancreas (12–22). Whether ITAM-bearing receptors other than the BCR are involved in sending tonic signals via Syk in these cell types is as yet unclear.

In some cancer cell types, including AML, retinoblastoma, and Ras-addicted lung and pancreatic carcinomas, cell survival is dependent on the expression of Syk, such that inhibitors of the kinase alone are sufficient to induce cell death (13, 17, 18). It is clear, however, that Syk is not essential for the survival of all cancer cells. In fact, Syk is frequently absent from certain tumor types, such as highly metastatic breast cancer, hepatocellular carcinoma, and melanoma, as a result of gene silencing by promoter methylation (30, 42–48). The exogenous expression of Syk in such cells decreases cellular motility, invasion, and metastasis (but does not induce apoptosis [30]). In all cell types examined in our study, a nonaggressive breast carcinoma, a B-cell lymphoma, naturally occurring DLBCL, and a highly aggressive breast carcinoma that does not normally express the kinase, Syk functions as a prosurvival factor. This is true even for MDA-MB-231 breast cancer cells, in which the expression of Syk is silenced and in which Syk, when exogenously expressed, has been described to be a tumor suppressor (30). Syk is unlikely to be the driver of transformation in all of the cancer cell types in which it is expressed but instead offers protection from stress-induced apoptosis regardless of the molecular mechanism by which the cells became transformed. In cells lacking Syk, it is likely that other antiapoptotic proteins function to protect the cells from external stress. For example, the MCF7-BD cells that lack Syk express much higher levels of Bcl-2 than do MCF7 cells that express endogenous Syk (41). In addition to protecting cells from oxidative stress, Syk also protects cells from certain inducers of genotoxic stress (41). Thus, the expression of Syk in a tumor cell could lead to drug resistance. These results indicate that Syk inhibitors might be useful in combination with genotoxic agents for the treatment of cancers in which the kinase is expressed.

The enhanced expression of one or more members of the Bcl-2 family of antiapoptotic proteins is a frequent occurrence in malignant cells (49, 50). Our studies indicate that Syk functions as a prosurvival factor, at least in part, by stabilizing the mRNA for Bcl-x_L. The longer Bcl-x_L transcript has a decreased half-life compared to that of the shorter transcript that encodes Bcl-x_S due to the presence of AREs located in the 3′ UTR. It is here that nucleolin binds to stabilize the message. This was first demonstrated in HeLa cells, in which the downregulation of nucleolin was shown to shorten the half-life of the Bcl-x_L message, while overexpression lengthened the half-life of the Bcl-x_L message (37). In this study, nucleolin was shown to bind to an ARE on the Bcl-x_L transcript in conjunction with poly(A) binding protein (PABP). Interestingly, in a previous mass spectrometric analysis of Syk-binding proteins, we found that both nucleolin and PABP coimmunoprecipitated with Syk, suggesting that all three proteins may be part of the same ribonucleoprotein complex (35). Nucleolin also modulates Bcl-x_L expression in primary lung endothelial cells (51). In these cells, treatment with angiotensin II induces apoptosis through the activation of the tyrosine phosphatase SHP-2, which inhibits the interaction of nucleolin with the Bcl-x_L mRNA. Thus, a tyrosine kinase (Syk) and a tyrosine phosphatase (SHP-2) have opposing effects on the ability of nucleolin to stabilize the Bcl-x_L transcript. This suggests the existence of one or more critical substrates that modulate the turnover of Bcl-x_L mRNA, whose phosphorylation on tyrosine is regulated in a positive or negative fashion by Syk or SHP-2, respectively. It was suggested that nucleolin itself might be the SHP-2 target. Our studies indicate that nucleolin is, in fact, phosphorylated on tyrosine by Syk in a manner that correlates with its enhanced ability to bind the Bcl-x_L mRNA and protect it from degradation.

We propose a model by which Syk associates with nucleolin, which is an intriguingly pleiotropic molecule with functions ranging from ribosome biogenesis to chromatin remodeling, transcriptional regulation, and binding to lipoproteins at the cell surface (52, 53). This Syk-nucleolin association likely occurs prior to any activation of Syk or phosphorylation of either component of the complex since catalytically inactive Syk binds robustly. Treatment of cells with hydrogen peroxide causes dissociation of the Syk-nucleolin complex. Site-directed mutagenesis studies indicate that the dissociation of Syk from nucleolin occurs as a consequence of its phosphorylation on tyrosines 342 and 346, which are located within the linker B region of the kinase. We did not observe an ATP-dependent dissociation of nucleolin from Syk-EGFP in anti-EGFP immune complexes. Thus, it is likely that the phosphorylation of both Y342 and Y346 on Syk within an intact cell generates a high-affinity docking site for a protein or proteins that contain an SH2 domain that binds to this region to disrupt the interaction between Syk and nucleolin. This hypothesis is consistent with the fact that several SH2 domains found on Syk-interacting proteins contain two phosphotyrosine-binding pockets and these interact with Syk preferentially when both Y342 and Y346 are phosphorylated (54, 55).

Active Syk phosphorylates nucleolin on tyrosine to promote its binding to the Bcl-x_L mRNA. This interaction protects cells from the loss of Bcl-x_L resulting from exposure to inducers of oxidative or genotoxic stress. It is of interest to note that both Syk and Bcl-x_L are important prosurvival factors in pre-B cells that have undergone gene rearrangements necessary for expression of a pre-B-cell antigen receptor, an event essential for B-cell survival in vivo (56, 57). Bcl-x_L is also elevated in germinal center B cells undergoing class switching and somatic hypermutation (58), conditions under which DNA strand breaks are likely to occur (and which are events also induced in cells treated with doxorubicin or etoposide). Thus, the prosurvival role of Syk in cancer is perhaps an unintended manifestation of one of its normal functions in B-cell development and is exploited by the cancer cell to escape from signals that would otherwise lead to cell death.