We and others previously observed that IgM and CD40 stimulation in murine B cells resulted in activation of extracellular signal-regulated kinase (ERK), a subfamily of mitogen-activated protein kinase. The present study demonstrated that ERK was rapidly phosphorylated and translocated to the nucleus in murine B cells upon stimulation with CD40, whereas it was preferentially localized within the cytosol after stimulation with IgM, suggesting that signaling through CD40 and IgM differentially regulates ERK subcellular localization. Costimulation with CD40 and IgM (CD40/IgM) resulted in subcellular localization of ERK within the cytosol, supporting the notion that stimulation with IgM delivers the signal responsible for inhibition of ERK nuclear transport. Consistent with these observations, IgM and CD40/IgM stimulation resulted in activation of ribosomal S6 kinase, which is a cytoplasmic substrate for ERK, whereas CD40 stimulation had little effect on its activity. Disruption of the microtubule by colchicine in WEHI231 cells resulted in reduction of ERK activity in IgM signaling, but not in CD40 signaling, compatible with the notion that the microtubule network may hold cytoplasmic ERK activity mediated by IgM stimulation. These results support the notion that ERK could mediate different effector functions in B cells upon stimulation with IgM and CD40.
The B cell Ag receptor (BCR)4 complex and CD40 play a pivotal role in B cell maturation and activation (for review see Ref. 1 and 2). Despite a difference in structure between BCR- and CD40-associated signaling molecules (2, 3, 4, 5, 6, 7, 8), it has been observed previously that B cell stimulation through BCR and CD40 results in activation of the Src and Syk protein tyrosine kinases coupled with an increase in the activity of the substrates of these kinases, including phosphatidylinositol-3 kinase and phospholipase C-γ, resulting in regulation of membrane-associated and soluble inositol polyphosphates (2, 9, 10). In addition to activation of protein tyrosine kinases, BCR and CD40 stimulation caused activation of NF-κB, NF-AT, and mitogen-activated protein kinase (MAPK) subfamilies, including extracellular signal-regulated kinase (ERK) (11, 12, 13, 14, 15, 16, 17, 18, 19). However, how these signaling pathways mediate biological effects that are unique to each receptor remains largely unknown. We have previously observed that IgM stimulation activated the ERK isoforms ERK1 and ERK2, whereas CD40 preferentially activated ERK2 (17). The ERK activity was sustained with IgM stimulation, but was transient with CD40 stimulation (17). Although the Ras-Raf-mediated pathway is involved as a common cascade in ERK activation by BCR and CD40, each receptor also uses a distinct signaling cascade in its activation (19, 20), which may contribute to the difference in the dynamics of activated ERK in IgM and CD40 signaling.
ERK is known to play an important role in growth and differentiation in several mammalian cells (21). In the immune system, it has been suggested that ERK activation along the Ras-Raf-MAPK/ERK activating kinase (MEK)-mediated pathway plays an important role in T cell selection and commitment in the thymus and in early B cell development in the bone marrow (22, 23, 24). ERK effector function is mediated through several substrates, which have been identified in the cytosol and the nucleus (21). Resting ERK enters the nucleus of unstimulated cells, where it returns to the cytoplasm in association with MEK, which exposes a nuclear export signal (25). A sizeable portion of ERK is associated with microtubules, suggesting a role of the protein as a cytoplasmic anchor to retain ERK within the cytosol in fibroblast cell lines (26). However, ERK translocates to the nucleus upon stimulation with serum and α-thrombin (26, 27, 28). ERK translocation to and retention in the nucleus is tightly associated with the phosphorylation state (29), compatible with the observation that active ERK phosphorylates nuclear substrates such as the transcription factor ELK-1/p62TCF, c-Jun, c-Myc, NF-IL-6, TAL1, RNA polymerase II, and STAT5 (21, 30, 31, 32, 33, 34, 35, 36, 37, 38). A part of ERK remains within the cytoplasm after stimulation and activates the substrates in a variety of cytoplasmic proteins, such as neuronal microtubule-associated proteins, MAP2 and tau, ribosomal S6 kinase (p90rsk or RSK), cytosolic phospholipase A2, Raf-1, MEK, and epidermal growth factor (EGF) receptor (21, 39, 40, 41, 42).
Here we examined the subcellular localization of ERK and its substrate specificity in B cells after stimulation with IgM and CD40. We observed that ERK was rapidly translocated to the nucleus upon CD40 stimulation, whereas it dominantly localized within the cytosol after stimulation with IgM, in which the microtubule network may in part support ERK activity. Costimulation with IgM and CD40 (IgM/CD40) resulted in ERK subcellular localization in the cytosol, supporting the notion that IgM stimulation may mediate the signal responsible for inhibition of ERK nuclear transport. Consistent with these observations, IgM and IgM/CD40 stimulation resulted in activation of ribosomal S6 kinase (p90rsk or RSK) in WEHI231 cells, which is known to be a cytoplasmic substrate for ERK. In contrast, CD40 stimulation had little effect on the activity of p90rsk. These results suggest that IgM and CD40 differentially regulate the subcellular localization of ERK and support the notion that the ERK signaling pathway may mediate distinct biological effects in B cells upon stimulation with IgM and CD40.
Materials and Methods
Single cell preparation and cell culture
A single-cell suspension was prepared from the spleen of specific pathogen-free C57BL/6 mice at 12–20 wk old, which were purchased from Japan SLC (Shizuoka, Japan). After preparation, cells were treated with 0.83% NH4Cl lysis buffer for removal of RBC. A B cell lymphoma, WEHI231, was maintained in RPMI 1640 medium supplemented with 10% heat-inactivated FCS, 55 μM 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin.
Cell staining for ERK subcellular localization
To investigate the subcellular localization of activated ERK in WEHI231 cells, we cultured the cells in the absence of FCS for 4 h, thus reducing the basal activity of ERK. Subsequently, cells were stimulated with either affinity-purified goat anti-μ Ab (25 μg/ml; Cappel, Aurora, OH) or anti-CD40 mAb (25 μg/ml; HM40-3; Ref. 43), or with both Abs. To ensure staining specificity, cells were incubated with the synthetic MEK inhibitor PD98059 (New England Biolabs, Beverly, MA) for 30 min at 37°C before stimulation (44). After stimulation, cells were fixed with 2% paraformaldehyde in PBS for 5 min at 37°C and quenched in 50 mM NH4Cl, pH7.0, in PBS. Thereafter, the cells were permealized with 0.1% Triton X-100 in TBS for 10 min and washed with TBS followed by blocking with 20% goat serum (Life Technologies, Rockville, MD) in TBS for 1 h at room temperature. The cells were stained with biotinylated wheat germ agglutinin (WGA; Cosmo Bio, Tokyo, Japan) at 1:100 dilution or anti-tubulin Ab (Calbiochem, La Jolla, CA) at 1:50 dilution and rabbit anti-phospho ERK Ab (New England Biolabs) at 1:50 dilution or anti-pan ERK Ab (New England Biolabs) at 1:50 dilution in TBS containing 3% BSA (Sigma, St. Louis, MO) at 4°C overnight. Anti-phospho ERK Ab was prepared against phospho-tyrosine peptide corresponding to ERK1 amino acid residues at positions 196–206, but the Ab cross-reacts with phosphorylated ERK2. After staining in the first step, cells were washed twice and incubated with Cy3-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) at 1:300 dilution and DTAF-conjugated streptavidin (EY Laboratories, San Mateo, CA) at 1:100 dilution or Alexa 350-conjugated goat anti-mouse IgG (Molecular Probes, Eugene, OR) at 1: 50 dilution in TBS containing 3% BSA and 0.1% Triton X-100 for 1 h at room temperature. For nuclear staining, cells were incubated with 1 μg/ml 4′,6′-diamidino-2-phenylindole (DAPI) (Molecular Probes) or 5 μM TO-PRO3 (Molecular Probes). As a control, cells were stained with Cy3-conjugated goat anti-rabbit Ig without preincubation with rabbit anti-ERK Ab. In the case of stimulated or unstimulated splenocytes, cells were stained with allophycocyanin-coupled anti-mouse B220 (RA3-6B2; PharMingen, San Diego, CA) at 1:100 dilution instead of WGA for identification of B cells in the population. After staining, cells were washed extensively and mounted in 0.1% p-phenylenediamine, 50 mM carbonate buffer, pH 9.0, and 80% glycerol. Specimens were examined under a confocal laser scanning microscope (LSM410 or LSM510; Carl Zeiss, Jena, Germany). Confocal images from each specimen, including staining controls, were subjected to equivalent contrast enhancement.
Subcellular fractions of WEHI231 were prepared as described previously (45). Briefly, quiescent WEHI231 cells (2 × 107), stimulated with or without anti-μ Ab, anti-CD40 mAb, or both, were washed two times with cold PBS at 4°C. All of the subsequent manipulations were performed at 4°C. These cells were then suspended in hypotonic buffer (20 mM HEPES-NaOH, pH 7.4, 10 mM EDTA, 2 mM DTT, 1 mM PMSF, 5 mM benzamidine, and 10 μg/ml aprotinin) and incubated for 15 min. After cells were homogenized with 50 strokes in a tight Dounce homoginizer, the rupture of >98% of the cells and the presence of apparently intact nuclei were microscopically confirmed. Nuclei were separated by centrifugation for 1 min at 800 × g, washed twice with sodium phosphate buffer containing 0.5% Nonidet P-40, and solubilized by boiling in Laemmli SDS-PAGE sample buffer. The postnuclear supernatant was separated by centrifugation at 30,000 × g for 30 min into membrane (pellet) and cytoplasmic (supernatant) fractions. The membrane fraction was washed with sodium-phosphate buffer and solubilized by boiling in Laemmli SDS-PAGE sample buffer. The cytoplasmic fraction was mixed with 6× sample buffer and boiled.
In vitro protein kinase assay and Western blot analysis
ERK activity was measured by in vitro kinase assay with myelin basic protein (MBP; Sigma) as an exogenous substrate according to a method previously described (17). Kinase activity of p90rsk was measured by in vitro kinase assay with use of histone H3 as an exogenous substrate (46) as follows: cell lysates were prepared from stimulated or unstimulated cells in lysis buffer (1% Triton X-100, 10 mM Tris, pH 7.6, 50 mM NaCl, 50 mM NaF, 1 mM EGTA, 1 mM Na2VO4, 30 mM sodium pyrophoasphate, 1 mM PMSF, 5 mM benzamidine, and 10 μg/ml aprotinin) and precleared by incubation with protein G-Sepharose 4B at 4°C overnight. After centrifugation, the supernatant was incubated with anti-p90rsk Ab (Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C for 1 h, followed by incubation with protein G-Sepharose 4B at 4°C for 1 h. After washing with lysis buffer, immunoprecipitates immobilized on Sepharose 4B were resuspended in 30 μl of kinase reaction buffer (30 mM HEPES pH 8.0, 10 mM MgCl2, 1 mM DTT, 20 μM ATP, 5 mM benzamidine, 2 μg Histone H3 (Boehringer Mannheim, Mannheim, Germany), 20 μCi [γ-32P]ATP) and incubated at 30°C for 5 min. The protein was separated by SDS-PAGE on 12% gels, and incorporated radioactivity was quantitated with an Image Analyzer (BAS2000; Fuji Photo Film, Tokyo, Japan). A quarter of the immunoprecipitates immobilized on protein G-coupled Sepharose 4B was provided for immunoblotting. Proteins were separated by SDS-PAGE on 7.5% gels, transferred onto Immobilon-P membrane (Millipore, Bedford, MA), and incubated with anti-p90rsk Ab (Santa Cruz Biotechnology), followed by visualization with use of the enhanced chemiluminescence system, as described previously (17).
In some experiments, WEHI231 cells were incubated with colchicine (Sigma) for 60 min or with cytochalasin D (Sigma) for 30 min at 37°C as described previously (47, 48) and provided for an in vitro kinase assay after stimulation with anti-μ Ab, anti-CD40 mAb, or both. All stock solutions of these reagents were prepared with DMSO at a concentration of 100 mM.
For analysis of subcellular distribution of activated ERK, the same amount of protein from the cytosol, nuclear, and membrane fractions of WEHI231 cells, respectively, were resolved by SDS-PAGE on 10% gels, transferred onto Immobilon-P membrane. The membranes were incubated for 1 h at room temperature with anti-phospho ERK Ab (New England Biolabs) at 1:200 dilution, anti-IκB kinase γ (IKKγ; see Refs. 49 and 50) Ab (Santa Cruz Biotechnology) at 1:500 dilution, and anti-TFIID Ab (Sigma; see Refs. 51 and 52) at 1:500 dilution. After washing, filters were incubated with anti-rabbit IgG conjugated with HRP, followed by visualization with use of the enhanced chemiluminescence system.
Distinct subcellular localization of ERK in murine B cells after stimulation with IgM and CD40
ERK is activated in B cells by IgM and CD40 stimulation (17, 18, 19), although with different kinetics (17). As shown in Fig. 1, we confirmed that ERK was rapidly and transiently activated in WEHI231 cells after cross-linking of CD40, whereas ERK activity was sustained above the basal level after maximal activation with IgM. Costimulation of IgM and CD40 (IgM/CD40) resulted in a response with kinetics similar to that observed in IgM stimulation, suggesting that signal through IgM is preponderant over that delivered by CD40 in this cell line. Targets of the ERK signaling pathway are located within several cytoplasmic compartments (21). Therefore, to investigate whether signal transduction mediated by IgM and CD40 requires the localization of ERK in each subcellular compartment, we cultured WEHI231 cells in serum-free medium to reduce the basal ERK activity and stimulated them with anti-μ Ab, anti-CD40 mAb, or both. After stimulation, cells were fixed and stained with WGA (Fig. 2, A and C) or DAPI (Fig. B) and with anti-phospho ERK or anti-pan ERK Ab in indirect immunofluorescence, followed by examination under a confocal laser microscope. As shown in Fig. 2, A and B, ERK was barely detected by anti-phospho ERK Ab in unstimulated WEHI231 cells (b and c), whereas it was detected dominantly within the cytoplasm in WEHI231 cells 5 min after cross-linking of IgM (e and f). However, we could not exclude the possibility that ERK was also associated with the plasma membrane. Pretreatment with the synthetic MEK inhibitor PD98059 (44) at 100 μM significantly inhibited ERK kinase activity in WEHI231 cells after stimulation with IgM (Fig. 2,C, upper panel) and abrogated the staining with anti-phospho ERK Ab in the IgM-stimulated cells (Fig. 2 C, lower panel), confirming that the staining observed was specific.
In contrast to the effect of IgM stimulation, ERK was barely detected by anti-phospho ERK Ab within the cytosol and/or in the plasma membrane in WEHI231 cells 1 min (data not shown) and 5 min after stimulation with CD40 (Fig. 2, A and B, h and i). As shown in Fig. 2,A, WGA staining supported the nuclear localization of ERK in CD40-stimulated cells. In addition, a nuclear counter-staining with DAPI and the nuclear image in differential interference micrographs (Fig. 2,B, g and h, respectively) supported the view that anti-phospho ERK Ab stained ERK within the nucleus in CD40-stimulated cells. As shown in Fig. 2, A and B, k and l, IgM/CD40 stimulation resulted in cytoplasmic localization of activated ERK, suggesting that IgM stimulation dominates CD40 stimulation in the regulation of the subcellular localization of ERK.
By indirect immunofluorescence using anti-pan ERK Ab, we detected ERK within the cytosol and the nucleus in unstimulated WEHI231 cells (data not shown), consistent with previous observations in fibroblast cell lines (27, 28). The finding that the subcellular localization of ERK detected by anti-pan ERK Ab was consistent with that observed by staining with anti-phospho ERK Ab in IgM- and CD40-stimulated WEHI231 cells (data not shown) led us to speculate that phosphorylated and unphosphorylated ERK may colocalize within the same subcellular compartment upon stimulation with IgM and CD40.
To investigate whether the difference in subcellular localization of ERK after IgM and CD40 stimulation is a common feature in normal B cells, we stimulated splenic B cells in vitro with anti-μ Ab, anti-CD40 mAb, or their mixture. After fixation and permialization, cells were stained with anti-B220 mAb and anti-phospho ERK Ab and then subjected to microscopic examination. Under this experimental condition, anti-B220 mAb could detect membrane-associated and cytosolic B220 molecules. As shown in Fig. 3, although ERK was barely detected in unstimulated splenic B cells (b and c), it appeared to colocalize with B220 molecules in B cells after IgM (e and f) and IgM/CD40 stimulation (k and l). In contrast, the results in Fig. 3, h and i, supported the possibility that CD40 stimulation resulted in subcellular localization of ERK predominantly in the nucleus.
To investigate further the subcellular localization of ERK in IgM and CD40 stimulation, WEHI231 cells were fractionated into nuclear, membrane, and cytosol fractions, as previously reported (45). As shown in Fig. 4, immunoblotting with anti-phspho ERK Ab was used to detect activated ERK in those fractions. The purity of cytosol and nucleus fractions was monitored by immunoblots with use of Abs against IKKγ, which is an essential regulatory component of the IκB kinase complex (49, 50), and basal transcriptional factor TFIID (51, 52), respectively. ERK was barely detected in quiescent WEHI231 cells (Fig. 4, lanes 1, 5, and 9), whereas phosphorylated ERKs with different molecular masses were significantly detected in the cytosolic fractions of IgM- or IgM/CD40-stimulated WEHI231 cells (Fig. 4, lanes 6 and 8). These two proteins most likely correspond to p44 ERK1 and p42 ERK2 (17, 21). Although the level of cytosolic ERK in CD40-stimulated WEHI231 cells was almost comparable to the basal level (Fig. 4, lanes 5 and 7), a substantial amount of ERK, probably corresponding to ERK2, was detected in the nuclear fraction of CD40-stimulated WEHI231 cells (Fig. 4, lane 11), comparable to the level of that detected in the cytosolic fraction of IgM-stimulated cells (Fig. 4, lane 6). However, ERK was barely detected in the nuclear fraction of IgM- and IgM/CD40-stimulated WEHI231 cells (Fig. 4, lanes 10 and 12). We observed that the level of ERK in the membrane fractions of stimulated WEHI231 cells was extremely low compared with that of the cytosol or nuclear fractions (Fig. 4, lanes 2, 3, and 4).
Taken together, these results suggest that ERK dominantly translocates to the nucleus in murine B cells upon stimulation with CD40, whereas it preferentially localizes within the cytoplasmic compartment after stimulation with IgM.
p90rsk, a cytoplasmic substrate for ERK, is activated in IgM signaling, but not in CD40 signaling
The results illustrated in Figs. 2–4 led us to speculate that ERK could efficiently activate substrates in the cytoplasm after stimulation with IgM, but not CD40. Therefore, we analyzed the activation of ribosomal S6 kinase (p90rsk or RSK) in WEHI231 cells after stimulation with IgM, CD40, and IgM/CD40. Members of the p90rsk family are regulated by phosphorylation on Ser and Thr residues (53, 54, 55), and several lines of evidence suggest that ERK isoforms are upstream activator of p90rsk (46, 55, 56, 57, 58, 59, 60). The p90rsk localizes dominantly in the cytosol in a resting state, but, once activated, translocates to the nucleus and phosphorylates several transcription factors (46). As shown in Fig. 5,A, cell lysates from stimulated or unstimulated cells were immunprecipitated with anti-p90rsk Ab, and the immunoprecipitates were provided for an in vitro kinase assay with Histone H3 as an exogenous substrate (46). We observed that p90rsk was activated 20- to 30-fold above the level of unstimulated cells within 3 min after cross-linking of IgM or IgM/CD40, whereas CD40 stimulation had little effect on its activity. Comparable amounts of immunoprecipitated p90rsk in each lane were confirmed by immunoblotting with anti- p90rsk Ab (data not shown). As shown in Fig. 5,B, to examine the link of ERK to p90rsk in IgM or IgM/CD40 signaling, WEHI231 cells were pretreated with the synthetic MEK inhibitor PD98059 and stimulated with anti-μ Ab, anti-CD40 mAb, or both. After stimulation, cell lysates were immunprecipitated with anti-p90rsk Ab, followed by an in vitro kinase assay with Histone H3. Pretreatment with PD98059 at 50–100 μM caused inhibition of ERK activity in WEHI231 cells after stimulation with IgM (Fig. 2,C), and this inhibition was accompanied by a parallel inhibition of p90rsk kinase activity (Fig. 5,B). Consistent with the result illustrated in Fig. 5 A, kinase activity of p90rsk was barely detected after CD40 stimulation. Each lane contained comparable amounts of immunoprecipitated p90rsk (RSK).
These results suggest that p90rsk could be an effective substrate for ERK activated by IgM or IgM/CD40 stimulation, but not by CD40. This is compatible with the notion that activated ERK may localize preferentially within the cytoplasm in WHEI231 cells after cross-linking of IgM and IgM/CD40, but not CD40.
Pretreatment with a microtubule-disrupting agent affected ERK activity by IgM, but not by CD40
It has been reported that ERK is associated with microtubules in several types of cells, including Xenopus oocytes, neural cells, fibroblasts, and macrophages (26, 61, 62, 63). It has been observed previously that ERK in association with microtubules displayed kinase activity in vitro (63), leading us to speculate that the microtubule network might be important in ERK activity caused by IgM stimulation. Therefore, we measured ERK activity induced by IgM, CD40, or IgM/CD40 stimulation in WEHI231 cells that were pretreated with colchicine, which prevents microtuble polymerization (64), and cytochalasin D, which inhibits actin polymerization (48). As shown in Fig. 6,A, pretreatment with colchicine, but not with cytochalasin D, reduced ERK activity in IgM signaling in a dose-dependent manner to a maximum of 50–60% of inhibition at a dose of 5 μM. As shown in Fig. 6,B, pretreatment with 5 μM colchicine reduced the ERK activity mediated by IgM/CD40 stimulation as well as IgM stimulation, whereas the treatment did not affect ERK activity by CD40 (Fig. 6,B, left panel). In contrast, pretreatment with cytochalasin D had no effect on ERK activity caused by these stimuli (Fig. 6,B, right panel). Consistent with these results, ERK was dominantly detected by anti-phospho ERK Ab within the cytoplasm in WEHI231 cells 5 min after cross-linking of IgM (Fig. 6,C, f and g), whereas pretreatment with colchicine at 5 μM reduced the cytoplasmic staining with anti-phospho ERK Ab in the IgM-stimulated cells (Fig. 6,C, n and o) and mostly abrogated the staining with anti-microtuble Ab (Fig. 6 C, l and p). In addition, pretreatment with colchicine did not affect subcellular localization of ERK in IgM-stimulated cells. Taken together, these results suggest that microtubules may function to hold a part of cytoplasmic ERK activity in IgM and IgM/CD40 signaling.
The present study demonstrated that ERK rapidly translocates to the nucleus in murine B cells upon stimulation with CD40, whereas it dominantly localizes within the cytosol after stimulation with IgM, suggesting that signaling through CD40 and IgM differentially regulates ERK subcellular localization in B cells. Costimulation with IgM and CD40 resulted in subcellular localization of ERK within the cytosol, supporting the notion that IgM stimulation may deliver the signal responsible for inhibition of ERK nuclear transport, although other possibilities could not be excluded. Inhibition of microtubule polymerization caused a partial reduction in ERK activity in WEHI231 cells after stimulation with IgM, but not with CD40, compatible with the notion that the microtubule network may play a role as an anchoring protein to hold activated ERK by IgM stimulation. Our previous results indicated that ERK1 and ERK2 are activated to a similar extent in murine B cells after stimulation with IgM, in contrast to the predominant activation of ERK2 after stimulation with CD40 (17), suggesting that both ERK isoforms localize within the same subcellular compartment of B cells after stimulation with IgM. In agreement with this view, previous reports indicated colocalization of ERK isoforms within the cytosol or the nucleus in fibroblasts (26, 28). However, further analysis is needed for clarification of this issue.
IgM and CD40 associates with distinct signal-transducing molecules (2, 3, 4, 5, 6, 7, 8). Stimulation with CD40 and IgM activates ERK by a Ras-dependent pathway (15, 19, 20), although both receptors may use distinct adaptor/guanine nucleotide-exchange factors to couple to Ras activation (19). In addition, activation of ERK by IgM and CD40 is mediated by a Ras-independent pathway in which distinct signal-transducing molecules are involved in both receptor-mediated stimulations (19, 20). Therefore, the distinct subcellular localization of ERK could be affected by upstream elements of ERK in Ras-dependent or -independent pathways or by multiple signaling cascades delivered by CD40 and IgM stimulation. In PC12 cells, stimulation with nerve growth factor (NGF) sustains the activation of ERK, accompanying its nuclear translocation, whereas treatment with EGF transiently activates ERK, which mostly localizes within the cytosol (for review, see Ref. 65). NGF activates ERK through Ras- and Rap1-dependent pathways in which Ras is required for initial activation of ERK, whereas ERK activation is sustained by Rap1 (66). Thus, the duration of ERK activation is regulated by distinct upstream elements of ERK, although whether the same pathways affect ERK subcellular localization in PC12 cells remains unknown.
The requirement for ERK nuclear transport has been studied extensively in fibroblast cell lines by expression of exogenous ERK with transfection and microinjection techniques (25, 26, 27, 28, 29). The results suggest that ERK nuclear transport occurs independently of ERK kinase activity and activation of upstream elements of ERK (27, 28, 29). A recent report suggests that dimerization of phosphorylated ERK with either phosphorylated or unphosphorylated ERK is sufficient for ERK nuclear transport in unstimulated fibroblast cell lines (29). In this context, our staining data indicated that subcellular localization of ERK detected by anti-pan ERK Ab was consistent with that observed by staining with anti-phospho ERK Ab in IgM- and CD40-stimulated WEHI231 cells. This raises the possibility that phosphorylated ERK also forms dimers with unphosphorylated ERK in B cells after stimulation with IgM and CD40.
Nuclear transport is initiated by the binding of the import substrates containing a nuclear localization signal (NLS) to the importin αβ heterodimer (for review, see Ref. 67). In addition, a novel pathway for the import substrate comprising nonclassical NLS has been reported (68). Signal-transducing molecules such as NF-κB and NF-AT contain NLS, whereas the NLS is masked by a different mechanism in a resting state (69, 70, 71). These molecules translocate to the nucleus when the NLS is unmasked by an active process through activation of signal cascades mediated by several external stimuli (69, 70, 71). ERK does not contain a sequence homologous to consensus NLS (29), and the ERK dimer could not be accessible for the nuclear import by diffusion through nuclear pores, with respect to a molecular mass larger than 60 kDa (67). If the ERK dimer translocates to the nucleus in the context of a transport protein containing a classical or nonclassical NLS, as yet to be defined, the present study supports the view that the activity of the transporter would be masked by signaling events linked to IgM. Further analysis is needed for clarification of the molecular mechanism for the dynamics of ERK subcellular localization in IgM and CD40 signaling.
Consistent with the subcellular localization of phosphorylated ERK, which is different in IgM and CD40 signaling, we observed that stimulation of WEHI231 cells with IgM, but not with CD40, significantly activated p90rsk, which localizes dominantly within the cytoplasm in a resting state (46). It has been reported that p90rsk activates immediate-early genes through activation of the transcription factor, cAMP response element-binding protein by phosphorylation (72), and probably participates in activation of NFκB (73, 74) and the down-regulation of the Ras-mediated pathway (75). Although the role of p90rsk in B cell function remains unknown, the present result led us to speculate that effector functions mediated by ERK activation could be different in B cells after stimulation with IgM and CD40, in association with the distinct subcellular localization of ERK caused by these stimuli. In addition, observations that microtubules are associated with ERK substrates, including microtubule-associated proteins (see Ref. 63), led us to speculate that ERK activity in association with the microtubule network would also contribute to the effector function of ERK activated by IgM. The association between ERK activity and its subcellular localization was also postulated previously for stimulation of PC12 cells with NGF and EGF; stimulation with NGF caused ERK translocation to the nucleus and triggered differentiation into sympathetic-like neurons, whereas treatment of the cells with EGF promotes proliferation without any noticeable effect on ERK nuclear translocation (65, 66).
It has been suggested that ERK plays a role in cell migration, proliferation, differentiation, and survival (21, 76, 77, 78, 79, 80). In the immune system, the ERK-signaling pathway may participate in T cell commitment and selection in the thymus (22, 23) and in early B cell development in the bone marrow (24). In addition, it has been proposed that the activation of ERK and p90rsk through BCR in tolerant B cells may block terminal differentiation into autoantibody-secreting plasma cells (81). We could not observe any significant synergistic effects in ERK activation by CD40 and IgM costimulation in vitro, but this condition might not correspond well to the events in B cell maturation mediated by BCR and CD40 in vivo. In the T cell dependent immune response, the first encounter between BCR on naive B cells and cognate Ag activates B cells, leading to uptake, processing, and presentation of the Ag (82). Ag-activated B cells may move at the interphase between B cell and T cell areas in the secondary lymphoid organs and make contact with T cells, which have been activated by professional APCs such as dendritic cells (83). At that step, the interaction between CD40 and CD40 ligand plays a pivotal role in B cell proliferation, isotype switching, and differentiation into memory B cells through formation of germinal centers, accompanying inhibition of the differentiation pathway into Ab-secreting cells (1, 84). Further analysis is needed for clarification of the role of ERK in B cell activation and maturation in the immune response.
We thank Dr. Y. Takahashi (National Institute of Infectious Diseases, Tokyo, Japan) for reading of this manuscript, Dr. H. Nishizumi (University of Tokyo, Tokyo, Japan) for technical advice, and Drs. T. Tsubata (Tokyo Medical Dental University, Tokyo, Japan) and M. Kashiwada and H. Nagaoka (NIID, Tokyo, Japan) for helpful discussion.
This work was supported by a grant from the Agency of Technology and Science of Japan (to T.T.).
Abbreviations used in this paper: BCR, B cell Ag receptor; ERK, extracellular signal-regulated kinase; MAPK, mitogen-activated protein kinase; MBP, myelin basic protein; MEK, MAPK/ERK activating kinase; NLS, nuclear localization signal; p90rsk/RSK, ribosomal S6 kinase; EGF, epidermal growth factor; WGA, wheat germ agglutinin; DAPI, 4′,6′-diamidino-2-phenylindole; IKKγ, IκB kinase γ; NGF, nerve growth factor.