Activation of the extracellular signal-regulated kinase by complement C5b-9

【关键词】  kinase

    Department of Medicine, McGill University Health Centre, McGill University, Montreal, Quebec, Canada

    ABSTRACT

    Extracellular signals may be transmitted to nuclear or cytoplasmic effectors via the mitogen-activated protein kinases. In the passive Heymann nephritis (PHN) model of membranous nephropathy, complement C5b-9 induces glomerular epithelial cell (GEC) injury, proteinuria, and activation of phospholipases and protein kinases. This study addresses the complement-mediated activation of the extracellular signal-regulated kinase (ERK). C5b-9 induced ERK threonine202/tyrosine204 phosphorylation (which correlates with activation) in GEC in culture and PHN in vivo. Expression of a dominant-inhibitory mutant of Ras reduced complement-mediated activation of ERK, but activation was not affected significantly by downregulation of protein kinase C. Complement-induced ERK activation resulted in phosphorylation of cytosolic phospholipase A2 and was, in part, responsible for phosphorylation of mitogen-activated protein kinase-associated protein kinase-2, but did not induce phosphorylation of the transcription factor, Elk-1. Activation of ERK was attenuated by drugs that disassemble the actin cytoskeleton (cytochalasin D, latrunculin B), and these compounds interfered with the activation of ERK by mitogen-activated protein kinase kinase (MEK). Overexpression of a constitutively active RhoA as well as inhibition of Rho-associated kinase blocked complement-mediated ERK activation. Complement cytotoxicity was enhanced after disassembly of the actin cytoskeleton but was unaffected after inhibition of complement-induced ERK activation. However, complement cytotoxicity was enhanced in GEC that stably express constitutively active MEK. Thus complement-induced ERK activation depends on cytoskeletal remodelling and affects the regulation of distinct downstream substrates, while chronic, constitutive ERK activation exacerbates complement-mediated GEC injury.

    glomerular epithelial cell; inflammation; protein kinases; signal transduction

    EXTRACELLULAR SIGNALS MAY be transmitted to nuclear or cytoplasmic effectors via a series of serine/threonine protein kinases, known as the mitogen-activated protein kinases (24, 26, 31, 38). There are several parallel mitogen-activated protein kinase pathways, including three major pathways, the p44/p42 or extracellular signal-regulated kinase (ERK)-1/2 pathway, typically activated by growth factors, as well as the c-Jun NH2-terminal kinase (JNK) and the p38 kinase pathways, which may be activated by diverse stimuli including "stress," e.g., hyperosmolality. In the ERK pathway, activation of Ras leads to the activation of Raf-1, followed by mitogen-activated protein kinase kinase, or MAP/ERK kinase (MEK), which then activates ERKs via dual phosphorylation on threonine and tyrosine. Raf-1 and/or Ras may also be downstream targets of protein kinase C (PKC), the activation of which is often preceded by phospholipase C-induced hydrolysis of inositol phospholipids and production of 1,2-diacylglycerol (24). The ERKs have multiple potential actions in the nucleus and in the cytoplasm. At present, the biological response or specificity of a signal arising from the ERK pathway is poorly understood. Specificity may be determined by activation of combinations of ERK and non-ERK effectors, the "strength" of ERK activation, and compartmentalization of the components of the ERK cascade (24, 26, 31). Such compartmentalization, as well as the structure and shape of cells, is dependent on the actin cytoskeleton, particularly the cortical cytoskeleton. Organization of the actin cytoskeleton is complex. Activated Rho GTPases can influence reorganization of filamentous (F)-actin and cell structure, although phenotype may differ between fibroblasts and epithelia (4, 19). Actin polymerization also appears to be regulated by phosphatidylinositol 4,5-bisphosphate and by proteins that bind both actin monomers and this inositol lipid (e.g., gelsolin, cofilin) (4, 34). The ezrin-radixin-moesin family of proteins regulates cross-linking of cortical actin to the plasma membrane (46).

    Activation of the complement cascade near a cell surface results in assembly of terminal components and insertion of the C5b-9 membrane attack complex into the lipid bilayer of the plasma membrane (28, 29, 37, 41). C5b-9 assembly leads to formation of transmembrane channels or rearrangement of membrane lipids with loss of membrane integrity. Nucleated cells require multiple C5b-9 lesions for lysis, whereas at lower doses, C5b-9 induces sublethal (sublytic) injury and various metabolic effects (28, 29, 41). Sublytic C5b-9 induces visceral glomerular epithelial cell (GEC) injury in passive Heymann nephritis (PHN) in the rat, a widely accepted model of human membranous nephropathy (8, 17). In PHN, antibody binds to GEC antigens and leads to the in situ formation of subepithelial immune complexes (8). C5b-9 assembles in GEC plasma membranes, injures GEC, and leads to proteinuria (8). Based on studies in GEC culture and in vivo, assembly of C5b-9 induces transactivation of receptor tyrosine kinases, an increase in cytosolic free Ca2+ concentration, and activation of phospholipases C, protein kinase C, cytosolic phospholipase A2- (cPLA2), as well as the ERK pathway (9, 10, 13, 14, 30). cPLA2 (18, 22) is an important mediator of C5b-9-dependent GEC injury. First, arachidonic acid released by cPLA2 is metabolized in GEC via cyclooxygenases-1 and -2 to prostaglandin E2 and thromboxane A2, and inhibition of prostanoid production reduces proteinuria in PHN and in human membranous nephropathy (8). Second, cPLA2 may mediate GEC injury more directly by inducing cell membrane phospholipid hydrolysis and the endoplasmic reticulum stress response (12). In addition, cPLA2 is a target for phosphorylation by ERK (22). The functional consequences of C5b-9-induced activation of the ERK pathway have not been fully elucidated, although a recent study suggests that ERK may limit complement-induced DNA damage (33).

    Our earlier studies indicate that the stimulation of cPLA2 activity and arachidonic acid release is dependent on subcellular localization or compartmentalization, in part, regulated by the cytoskeleton, specifically the actin filament network (11). The purpose of the present study was to examine the activation of the ERK signaling cascade by the C5b-9 complex and define the potential role of the actin cytoskeleton in activation. We demonstrate that complement activates ERK in GEC culture and in PHN in vivo and that ERK activation proceeds mainly via Ras and is dependent on actin cytoskeleton remodelling. The cytoskeleton regulates MEK-dependent ERK phosphorylation. Activated ERK was able to phosphorylate distinct substrates.

    MATERIALS AND METHODS

    Materials. Tissue culture reagents were obtained from Invitrogen Canada (Burlington, ON). Purified C8, complement-deficient sera, cytochalasin D, latrunculin B, calyculin A, phorbol myristate acetate (PMA), PD-98059, SB-203580, and epidermal growth factor (EGF) were purchased from Sigma Canada (Mississauga, ON). Y-27632 was from EMD Biosciences (La Jolla, CA). Rabbit antibody IgGs to phospho-ERK threonine202/tyrosine204, phospho-JNK threonine183/tyrosine185 phospho-Elk-1 serine383, phospho-cPLA2 serine505, and phospho-mitogen-activated protein kinase-activated protein kinase-2 (MAPKAPK-2) threonine222 were purchased from New England Biolabs (Mississauga, ON). Rabbit anti-ERK antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Electrophoresis and immunoblotting reagents were from Bio-Rad Laboratories (Mississauga, ON). Rhodamine-conjugated phalloidin and Alexa-Fluor 488-conjugated deoxyribonuclease I (DNase I) were obtained from Molecular Probes (Eugene, OR). Male Sprague-Dawley rats (150 g) were purchased from Charles River Canada (St. Constant, PQ).

    Cell culture and transfection. Rat GEC culture and characterization have been published previously (5, 9, 30). GEC were cultured in K1 medium and studies were done with cells between passages 8 and 60. Transfection and production of GEC that stably overexpress cPLA2, J25 Ras, L63RhoA, or R4F-MEK proteins and Neo GEC were described previously (3, 9, 11). Experiments were carried out in Neo GEC, or as indicated.

    Incubation of GEC with complement. The standard protocol involved incubation of GEC in monolayer culture with rabbit anti-GEC or sheep anti-Fx1A antiserum (5% vol/vol) in modified Krebs-Henseleit buffer, containing 145 mM NaCl, 5 mM KCl, 0.5 mM MgSO4, 1 mM Na2HPO4, 0.5 mM CaCl2, 5 mM glucose, and 20 mM HEPES, pH 7.4, for 40 min at 22°C (9, 30). GEC were then incubated with normal human serum (NS; diluted in Krebs-Henseleit buffer), or heat-inactivated (decomplemented) human serum (HIS; 56°C, 30 min) in controls, for the indicated times at 37°C. In some experiments, antibody-sensitized GEC were incubated with C8-deficient human serum, or C8-deficient serum supplemented with purified C8 (80 μg/ml undiluted serum). As in previous studies, we generally used heterologous complement to minimize possible signaling via complement-regulatory proteins, although we demonstrated that homologous complement induces similar biological effects (9, 30). Except for studies of cytolysis, experiments were carried out at concentrations of complement that induced minimal or no lysis (NS or C8-deficient serum reconstituted with C8 at 2.5% vol/vol). Previous studies showed that in GEC, complement is not activated in the absence of antibody (9, 30).

    Induction of PHN in rats. PHN was induced by a single intravenous injection of 0.4 ml of sheep anti-Fx1A antiserum, as described previously (12, 14). Urine was collected on day 14, and rats were then killed and glomeruli were isolated by differential sieving. All studies were approved by the McGill University Animal Care Committee.

    Preparation of cell and glomerular lysates and immunoblotting. GEC were scraped from culture dishes into homogenization buffer, containing 50 mM HEPES, 0.25 mM sucrose, 1 mM EDTA, 1 mM EGTA, 20 μM leupeptin, 20 μM pepstatin, and 0.1 mM PMSF, pH 7.4 (4°C), and were centrifuged at 1,500 g for 3 min at 4°C. Cell pellets were solubilized in buffer, containing 1.0% Triton X-100, 125 mM NaCl, 20 mM Tris, 20 μM leupeptin, 20 μM pepstatin, 0.2 mM PMSF, 25 mM NaF, 2 mM Na3VO4, 5 mM Na4P2O7, 1 mM EDTA, 1 mM EGTA, pH 7.4 (4°C). The mixture was centrifuged at 14,000 g for 10 min, and the supernatant was then used for immunoblotting (9, 10, 13, 14). Glomeruli were isolated from rat kidney cortices by differential sieving (12, 14). Glomeruli were centrifuged and resuspended in buffer, as above. Samples were boiled in Laemmli sample buffer and subjected to SDS-PAGE under reducing conditions. Proteins were then electrophoretically transferred onto nitrocellulose paper, blocked with 3% BSA/2% ovalbumin, and incubated with primary antibody, and then with horseradish peroxidase-conjugated secondary antibody. The blots were developed using the enhanced chemiluminescence technique (ECL; Amersham Pharmacia Biotech). Protein content was quantified by scanning densitometry, using NIH Image software. Preliminary studies demonstrated that there was a linear relationship between densitometric measurements and the amounts of protein loaded onto gels.

    Immunofluorescence microscopy. Cryostat kidney sections (4 μm) were fixed with ether-ethanol (1:1, 10 min), followed by ethanol (20 min) at 4°C (3). Sections were incubated with rabbit anti-phospho-ERK IgG (2.5 μg/ml) overnight at 4°C. Nonimmune rabbit IgG was used in control incubations. After being washed, sections were incubated with fluorescein-conjugated goat anti-rabbit IgG for 1 h at 22°C. The immunofluorescence signals were evaluated using a Nikon Diaphot fluorescence microscope with visual output connected to a Nikon Coolpix 995 digital camera.

    To quantitate actin, GEC adherent to coverslips were fixed with 3% paraformaldehyde in PBS and were permeabilized with 0.5% Triton X-100. After being washed, cells were incubated with rhodamine-phalloidin (0.043 μg/ml) to visualize F-actin and Alexa-Fluor 488-conjugated DNase I (9 μg/ml, diluted in 3% BSA in PBS) to visualize G-actin for 20 min. Coverslips were mounted onto glass slides and were examined using both rhodamine and fluorescein filters in a Nikon Diaphot fluorescence microscope. Slides were photographed using a Nikon Coolpix 995 digital camera locked at a constant exposure. Fluorescence intensity was then quantified using Adobe Photoshop software.

    Measurement of complement-dependent cytotoxicity. Complement-mediated cytolysis was determined by measuring release of lactate dehydrogenase (LDH), similarly to the method described previously (12, 32). Specific release of LDH was calculated as [NS HIS]/[100 HIS], where NS represents the percent of total LDH released into cell supernatants in incubations with NS, and HIS is the percent of total LDH released into cell supernatants in incubations with HIS.

    Statistics. Data are presented as means ± SE. The t-statistic was used to determine significant differences between two groups. One-way ANOVA was used to determine significant differences among groups. Where significant differences were found, individual comparisons were made between groups using the t-statistic and adjusting the critical value according to the Bonferroni method. Two-way ANOVA was used to determine significant differences in multiple measurements among groups.

    RESULTS

    C5b-9 induces ERK phosphorylation in GEC in culture and in PHN. In previous studies, we demonstrated that complement induces transactivation of receptor tyrosine kinases, including the EGF receptor (R), and activates ERK (10, 13). Activation of ERK was determined by use of an immune complex kinase assay, which correlated with induction of ERK threonine202/tyrosine204 phosphorylation (10). The present study confirms that incubation of GEC with antibody and NS stimulates ERK1 and 2 phosphorylation, whereas decomplemented HIS does not (Fig. 1A). Furthermore, C8-deficient serum reconstituted with purified C8 increased ERK phosphorylation, compared with C8-deficient serum alone (Fig. 1A), confirming that ERK activation is dependent on assembly of C5b-9.

    It is important to verify that the observations in GEC culture are relevant to in vivo pathophysiology. Thus we assessed ERK phosphorylation in PHN, where C5b-9 induces GEC injury and proteinuria. By immunoblotting, ERK phosphorylation was increased in glomeruli of rats with PHN, compared with normal controls (Fig. 1, B and C). Immunofluorescence staining of kidney sections from rats with PHN demonstrated specific glomerular localization of phospho-ERK (Fig. 2). The staining was focal, and involved glomerular capillary loops, consistent with the localization of phospho-ERK in GEC. Phospho-ERK staining in glomeruli of normal rats was faint or not detected.

    Complement-induced ERK activation occurs via Ras. In the next series of experiments, we addressed the pathways involved in ERK activation. Following C5b-9 assembly, transactivated EGF-R was shown earlier to bind a glutathione S-transferase-Grb2 SH2 domain fusion protein, suggesting linkage with the Ras-Raf-MEK-ERK pathway (13). In addition, EGF-R bound and activated phospholipase C-1, leading to activation of PKC (11). To determine whether complement-induced ERK activation occurred via Ras, we employed GEC that stably overexpress J25 Ras, a Ras mutant that constitutively activates the phosphatidylinositol 3-kinase pathway, but cannot activate the ERK pathway (50). When overexpressed, J25 Ras acts as a dominant inhibitor of the Ras-ERK pathway (3). The GEC clone that overexpresses J25 Ras tended to have higher basal ERK phosphorylation, but after incubation with complement, there was only a small and insignificant increase in ERK phosphorylation, compared with parental (untransfected) GEC (Fig. 3, A and C). However, PMA-induced activation of ERK was unaffected by J25 Ras expression (Fig. 3, B and C).

    We then addressed the role of PKC in ERK activation. Previously, we demonstrated that prolonged incubation of GEC with a high dose of the PKC agonist PMA downregulated PKC activity by 100% (10). Depletion of PKC did not affect complement-induced activation of ERK significantly, although there was a downward trend, compared with untreated cells. However, PKC depletion abolished PMA-induced ERK activation completely (Fig. 3, DF). Therefore, complement-induced ERK activation proceeds via Ras, whereas PMA, as expected, activates ERK via PKC.

    ERK activation by complement leads to phosphorylation of substrates. In the next series of experiments, we examined potential targets of activated ERK. The components of the ERK cascade are localized in the cytoplasm, but on activation, ERK may remain active in the cytoplasm or translocate to the nucleus (31). Substrates of ERK may include transcription factors, cytoplasmic proteins, or protein kinases. First, we examined ERK-mediated phosphorylation of the nuclear transcription factor Elk-1 on serine383 (31). Despite inducing ERK phosphorylation, incubation of GEC with complement did not lead to detectable phosphorylation of Elk-1 (Fig. 4A). In contrast, PMA and EGF induced phosphorylation of both ERK and Elk-1 (Fig. 4, B and C).

    cPLA2 is a cytoplasmic protein and contains an ERK phosphorylation site at serine505 (18, 22). Incubation of GEC with complement induced phosphorylation of cPLA2 at serine505 (Fig. 5A). Similarly, PMA and EGF induced cPLA2 phosphorylation (Fig. 5, B and C). To verify that the complement-induced serine505 phosphorylation of cPLA2 observed in cultured GEC is relevant to in vivo pathophysiology, we also assessed cPLA2 phosphorylation in PHN. By analogy, cPLA2 phosphorylation was increased in glomeruli of rats with PHN, compared with normal rats (Fig. 5, D and E).

    MAPKAPK-2 is regarded primarily as a substrate of p38 kinase but is also reported to be a substrate of ERK (31). We measured activation of MAPKAPK-2 by monitoring threonine222 phosphorylation. Complement stimulated MAPKAPK-2 phosphorylation, and the increase in phosphorylation was partially inhibited in the presence of PD-98059, a specific inhibitor of MEK (Fig. 6, A and B) (15, 16). The p38 kinase inhibitor, SB-203580, also blocked complement-induced MAPKAPK-2 phosphorylation, and inhibition appeared greater compared with PD-98059 (Fig. 6C). Thus activation of MAPKAPK-2 by complement is dependent on both ERK and p38 kinase.

    Complement-induced activation of ERK is dependent on an intact F-actin cytoskeleton. Transmission of intracellular signals may be dependent on compartmentalization of mediators, which requires an intact cytoskeleton, specifically, the actin filament network (11). Previously, we showed that in resting GEC, F-actin is distributed in a cortical pattern (11). To determine the effect of complement on the actin cytoskeleton, we quantified F-actin content in GEC by rhodamine-phalloidin labeling of F-actin and normalizing the values for Alexa-Fluro-488-conjugated DNase1-labeled G-actin. Incubation of GEC with complement for 2 h tended to increase the F/G actin ratio slightly (109 ± 9% of control; 5 experiments), but the increase was not statistically significant.

    Cytochalasin D and latrunculin B are membrane-permeant inhibitors of actin polymerization that act by distinct mechanisms (6, 43). Incubation of GEC with cytochalasin D (20 μM) or latrunculin B (1 μM) for 30 min resulted in a partial dissolution of the cortical F-actin structure (11). In GEC that had been preincubated with cytochalasin D or latrunculin B, complement-induced ERK activation was attenuated significantly (Fig. 7, A and B). Phosphorylation of the ezrin-radixin-moesin family of proteins is required for cross-linking of actin to the plasma membrane (46), and such phosphorylation can be induced by treatment of cells with calyculin A, a serine/threonine protein phosphatase inhibitor (which inhibits phosphatases 1 and 2). Treatment of many cell lines with calyculin A results in condensation of actin filaments at the plasma membrane (2, 35). Preincubation of GEC with calyculin A also inhibited the complement-induced increase in ERK phosphorylation (Fig. 7C).

    Previously, we demonstrated that the transactivation of EGF-R by C5b-9 occurred efficiently even if F-actin had been disrupted (11). To further define the site of action of cytochalasin D and latrunculin B within the ERK pathway, we studied PMA- and EGF-induced phosphorylation of ERK. Both PMA and EGF increased ERK phosphorylation, and these increases were blocked by cytochalasin D and latrunculin B (Fig. 7D). PMA activates the ERK pathway via PKC/Raf (independently of Ras), while EGF-induced ERK activation is PKC independent and Ras/Raf dependent (10). Thus the PMA and EGF results together indicate that an intact cytoskeleton is required for efficient activation of the Raf-MEK-ERK cascade.

    In the next set of experiments, we examined whether cytochalasin D and latrunculin B affected ERK phosphorylation by MEK, the kinase just upstream of ERK. We employed a clone of GEC that stably overexpresses constitutively active MEK (R4F-MEK) (3, 25). Steady-state ERK phosphorylation in the unstimulated R4F-MEK-expressing clone is significantly greater, compared with parental GEC (Fig. 7E). Treatment of R4F-MEK-expressing cells with cytochalasin D or latrunculin B markedly reduced ERK phosphorylation (Fig. 7F), by 91 ± 4 and 97 ± 1%, respectively (P < 0.005, n = 3). Thus an intact actin cytoskeleton is required for efficient phosphorylation of ERK by MEK. However, we cannot exclude that the two more proximal steps in the ERK cascade (i.e., Raf or MEK activation) might also be dependent on an intact actin cytoskeleton.

    In an earlier study, we demonstrated that complement-induced activation of JNK is dependent on arachidonic acid release and production of superoxide (32). In contrast to the activation of ERK, cytochalasin D, latrunculin B, and calyculin A did not inhibit complement-induced JNK activation (Fig. 8, A and B). H2O2 is a potent activator of JNK in GEC (32). By analogy to complement, pretreatment of GEC with cytochalasin D, latrunculin B, or calyculin A did not inhibit JNK activation by H2O2 (Fig. 8, C and E). Actually, the three drugs appeared to independently increase JNK activation in resting and stimulated GEC (Fig. 8).

    Activation or inhibition of RhoA blocks complement-mediated ERK activation. Rho GTPases are known to stabilize actin filaments and induce stress fiber formation in various cells, and these effects are often mediated by Rho-associated kinase (ROCK) (4). Previously, we reported that in GEC, stable expression of a constitutively active RhoA mutant (L63RhoA) resulted in increased actin polymerization, as reflected by the appearance of stress fibers superimposed on the cortical distribution of F-actin (11). Moreover, the L63RhoA-transfected GEC showed a relative resistance to depolymerization of cortical actin by latrunculin B, compared with Neo GEC (11). Basal ERK phosphorylation was slightly, but significantly, greater in GEC that overexpress L63RhoA, compared with Neo (Fig. 9, A and B). However, complement was not able to stimulate ERK phosphorylation in the L63RhoA-expressing cells (Fig. 9, A and B). In another study, it was reported that complement activates RhoA in GEC in vivo (52). Preincubation of Neo GEC with the Rho-associated kinase inhibitor Y-27632 (15, 47), blocked complement-induced ERK phosphorylation (Fig. 9, C and D).

    Roles of ERK and cytoskeleton in complement-mediated GEC injury. This series of experiments addressed the potential functional roles of ERK and the cytoskeleton in complement-mediated GEC injury. In these experiments, GEC were sensitized with antibody and incubated with serially increasing doses of complement that induced minimal to moderate cell lysis at either 40 min or 18 h. This protocol would allow for C5b-9 to stimulate ERK activation, but with increasing incubation time and complement dose, a portion of the cells will undergo lysis. Incubation of GEC with complement in the presence of cytochalasin D or latrunculin B for 18 h enhanced lysis compared with complement alone (Fig. 10). These effects of cytochalasin D and latrunculin B were not, however, evident in brief incubations (NS for 40 min), demonstrated earlier (11). Complement-mediated cytolysis was not affected during blockade of the ERK pathway by PD-98059 for 40 min or for 18 h (Fig. 11, A and B). In contrast, complement-mediated cytolysis (both acute and chronic incubations) was enhanced in the GEC that overexpress R4F-MEK, compared with Neo GEC (Fig. 11, C and D). Furthermore, concurrent treatment of the R4F-MEK-expressing cells with PD-98059 reduced the levels of complement lysis to those observed in Neo GEC (Fig. 11D), indicating that the greater cytotoxic effect of complement in the R4F-MEK-expressing GEC was actually dependent on the activation of ERK by R4F-MEK. These results indicate that complement-induced ERK activation does not directly modulate complement-dependent cytotoxicity, but cytotoxicity is exacerbated by chronic, constitutive ERK activation.

    DISCUSSION

    This study elucidates the mechanisms of C5b-9-mediated activation of the ERK cascade. ERK activation proceeded via Ras and was dependent on cytoskeletal remodelling. Activated ERK phosphorylated two distinct substrates, and chronic constitutive activation of the ERK pathway exacerbated complement-dependent cytotoxicity. In keeping with earlier results, we demonstrate that in cultured GEC, C5b-9 induced ERK threonine202/tyrosine204 phosphorylation (Fig. 1), which correlates with activation (10). Furthermore, we show that ERK is phosphorylated in glomeruli of rats with PHN (i.e., C5b-9-dependent GEC injury in vivo), and the localization of phospho-ERK is consistent with GEC (Figs. 1 and 2). In previous studies, we demonstrated that assembly of C5b-9 in the GEC plasma membrane (in culture and in vivo) induced transactivation of receptor tyrosine kinases, including EGF-R (13). EGF-R transactivation resulted in activation of phospholipase C-1 and PKC, as well as binding of the adaptor protein, Grb2, which links receptor tyrosine kinases to Ras (11, 13). To determine the role of Ras and PKC in complement-mediated ERK activation, we monitored ERK phosphorylation in GEC that had been stably transfected with a dominant-inhibitory mutant of Ras, and in GEC depleted of PKC. These experiments showed that the ERK pathway was activated predominantly via Ras (Fig. 3). Our results are distinct from those in K562 and COS-7 cells, where complement-induced ERK activation occurred via PKC (21).

    The protocol employed in the present study did not result in quantitative changes in F-actin following incubation with complement, but in another study (44), GEC injury by complement was associated with loss of actin stress fibers and focal contacts. In vivo, GEC contain F-actin as a layer at the base of the foot processes, and the actin cytoskeleton is important in the maintenance of cell architecture. Previously, we showed that drugs that disassemble the actin cytoskeleton (cytochalasin D and latrunculin B) did not affect complement-mediated activation of EGF-R but inhibited activation of phospholipase C-1 and downstream signaling, including activation of PKC and cPLA2 (11). The present study demonstrates that complement-induced ERK activation was inhibited by cytochalasin D and latrunculin B (Fig. 7, A and B). Both drugs also inhibited ERK activation by EGF (which occurs via the Ras and Raf) and PKC (which occurs via Raf, but not Ras) (Figs. 3 and 7D). Thus an intact actin cytoskeleton is required for sequential activation of Raf, MEK, and ERK. Furthermore, cytochalasin D and latrunculin B also blocked ERK phosphorylation in GEC that stably overexpress R4F-MEK (Fig. 7F). This result implies that efficient phosphorylation of ERK by MEK is dependent on an intact actin cytoskeleton. The actions of cytochalasin D and latrunculin B on ERK phosphorylation were similar, although cytochalasin D and latrunculins induce actin cytoskeleton depolymerization through different mechanisms. Cytochalasin D binds to the barbed (growing end) of actin filaments and prevents actin filament formation or leads to disruption of actively turning over actin stress fibers. Latrunculins sequester G-actin monomers preventing actin polymerization and effectively disrupt both actin stress fibers, as well as cortical actin filaments, which are more resistant to cytochalasin D (23).

    Similar to the actin-depolymerizing drugs, condensation and stabilization of actin filaments at the cell periphery near the plasma membrane by calyculin A inhibited the complement-induced increase in ERK phosphorylation (Fig. 7C). Stable expression of L63RhoA also attenuated ERK phosphorylation by complement (Fig. 9), and this effect is associated with enhanced stress fiber formation (11). L63RhoA may have acted by a mechanism analogous to calyculin A, as constitutively active RhoA was reported to induce phosphorylation of ezrin-radixin-moesin proteins (4). Similar to RhoA activation, inhibition of Rho-associated kinase (a downstream effector of RhoA) reduced ERK activation (Fig. 9). Together, the results suggest that both disassembly and stabilization of the actin cytoskeleton can reduce ERK phosphorylation, implying that ERK activation is dependent on cytoskeletal remodelling. Finally, the cytoskeleton-altering drugs had disparate effects on the activation of the JNK pathway in GEC (Fig. 8). Thus there are major differences in the role of the cytoskeleton in facilitating activation of pathways by complement, and pharmacological disassembly of the actin filament network does not exert a general inhibitory effect on all signaling pathways.

    Our study, which revealed an important role for the actin cytoskeleton in complement signaling, is in keeping with studies in other systems. For example, in response to insulin treatment, actin filament disassembly blocked activation of ERK and p38 kinase, but not insulin receptor autophosphorylation, phoshatidylinositol 3-kinase, or S6 kinase (42, 45). Moreover, binding of Shc to the insulin receptor was not affected, but binding of Grb2 to Shc was disrupted (45). However, cytochalasin D did not affect ERK activation by lysophosphatidic acid, bombesin, or platelet-derived growth factor (39). In addition, the actin cytoskeleton may be important in cell cycle progression, transcription of serum-inducible genes, and induction of nitric oxide synthase (51). It may seem unusual that calyculin A and L63RhoA expression (which facilitate actin polymerization) should produce effects similar to those of cytochalasin D or latrunculin B (which depolymerize the cytoskeleton) and inhibition of Rho-associated kinase. All of these treatments were inhibitory to ERK activation in the present study, and parallel effects of increasing actin polymerization and depolymerization have been reported in other systems. For example, F-actin polymerization with the drug jasplakinolide and depolymerization with latrunculin B or cytochalasin D both inhibited insulin-stimulated glucose uptake in adipocytes (20), lipopolysaccharide-mediated production of reactive oxygen species in monocytes (36), accumulation of phosphatidylinositol 3,4,5-trisphosphate in response to a chemotactic stimulus in neutrophils (48), and apoptosis in airway epithelial cells (49).

    To address the functional role of complement-mediated ERK activation in GEC, we examined substrates of ERK, which may point to potential involvement in cellular functions (31). Complement did not induce serine383 phosphorylation of the transcription factor Elk-1, a potential nuclear target of ERK (Fig. 4). As expected, EGF, a well-known activator of ERK, did actually induce Elk-1 phosphorylation. Perhaps EGF was able to activate/recruit additional factors that are required for ERK to phosphorylate nuclear transcription factors, while complement was not effective. However, complement induced phosphorylation of MAPKAPK-2, which was, in part, dependent on the ERK pathway (as well as p38 kinase; Fig. 6). MAPKAPK-2 is more typically a substrate of p38 kinase, but by analogy to GEC, ERK-dependent activation of MAPKAPK-2 has been reported in neutrophils (7). Under basal conditions, MAPKAPK-2 is believed to be located in the nucleus, whereas on activation, MAPKAPK-2 can phosphorylate nuclear transcription factors, or be exported into the cytoplasm, where it can activate other substrates, including Hsp27 (27). Our result supports a role for complement-mediated ERK activation in transcription, as well as in cytoplasmic pathways. Moreover, complement stimulated phosphorylation of cPLA2, a cytoplasmic ERK target on serine505 (Fig. 5), a well-defined ERK phosphorylation site (18, 22). Interestingly, although cPLA2 is a substrate for ERK in GEC, the complement-induced increase in cPLA2 catalytic activity in these cells did not require ERK activation (10) but was dependent on PKC. Phosphorylation of the cPLA2 serine505 site contributes to EGF-mediated release of arachidonic acid in GEC (10) and correlates with agonist-stimulated cPLA2 activity in other cells (22). Further study will be required to elucidate additional ERK substrates.

    In earlier studies, we demonstrated that complement-induced activation of JNK and p38 mitogen-activated protein kinases was cytoprotective in GEC, i.e., that JNK and p38 limited or restricted complement cytotoxicity (1, 32). The mechanism of protection by p38 was, at least in part, dependent on Hsp27, a substrate of MAPKAPK-2 (1). This study suggests that the role of ERK in GEC injury may be more complex. Inhibition of the complement-induced activation of the ERK cascade in GEC did not affect complement cytotoxicity significantly (Fig. 11). However, a cytoprotective role for ERK via MAPKAPK-2 and Hsp27 cannot be excluded because MAPKAP-2 activation was only partially dependent on ERK (Fig. 6). Chronic (but not brief) incubation with complement in the presence of cytochalasin D or latrunculin B increased complement lysis (Fig. 10), but the combination of ERK inhibition and cytoskeletal disassembly did not increase complement lysis beyond that observed with cytoskeletal disassembly alone (data not shown). In contrast, complement cytotoxicity was enhanced in GEC that stably express constitutively active MEK (Fig. 11). The kinetics of ERK activation are likely to be different between the two conditions. Previously, we demonstrated that during chronic incubation with complement, ERK phosphorylation was enhanced within 20 min and returned toward basal levels by 5 h (13). In contrast, in the GEC that stably express R4F-MEK, ERK remains activated continuously (25). Thus complement-dependent cytotoxicity was exacerbated by chronic, constitutive ERK activation. In another study, ERK inhibition in K562 and COS-7 cells enhanced complement lysis (measured as trypan blue uptake) (21). A recent study in GEC (33) showed that complement induced DNA damage and increased expression of the cell cycle-regulatory genes, p21 and GADD45 (growth-arrest DNA damage-45). In this study, inhibition of MEK exacerbated DNA damage and reduced levels of p21 and GADD45. The authors concluded that ERK activation protects GEC from DNA damage via regulation of p21 and GADD45 genes. However, although these authors showed that complement did not induce apoptosis, cytolytic injury (e.g., LDH release) was not examined. Furthermore, the same investigators demonstrated earlier that C5b-9 can induce DNA synthesis (although not proliferation) in GEC (40). Thus the effect of C5b-9 and ERK on DNA regulation and cell injury will require additional study. With the advent of drugs that modulate protein kinase pathways, a better understanding of the activation of protein kinases by C5b-9 will provide insights into novel targets for therapy of glomerular diseases.

    GRANTS

    This work was supported by Research Grants from the Canadian Institutes of Health Research and the Kidney Foundation of Canada. A. V. Cybulsky and T. Takano hold Scholarships from the Fonds de la Recherche en Santé du Québec. K. Bijian was awarded a Fellowship from the McGill University Health Centre Research Institute.

    FOOTNOTES

    The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

    REFERENCES

    Aoudjit L, Stanciu M, Li H, Lemay S, and Takano T. p38 Mitogen-activated protein kinase protects glomerular epithelial cells from complement-mediated cell injury. Am J Physiol Renal Physiol 285: F765F774, 2003.

    Berridge MJ, Lipp P, and Bootman MD. Signal transduction. The calcium entry pas de deux. Science 287: 16041605, 2000.

    Bijian K, Takano T, Papillon J, Khadir A, and Cybulsky AV. Extracellular matrix regulates glomerular epithelial cell survival and proliferation. Am J Physiol Renal Physiol 286: F255F266, 2004.

    Bishop AL and Hall A. Rho GTPases and their effector proteins. Biochem J 348: 241255, 2000.

    Coers W, Reivinen J, Miettinen A, Huitema S, Vos JT, Salant DJ, and Weening JJ. Characterization of a rat glomerular visceral epithelial cell line. Exp Nephrol 4: 184192, 1996.

    Cooper JA. Effects of cytochalasin and phalloidin on actin. J Cell Biol 105: 14731478, 1987.

    Coxon PY, Rane MJ, Uriarte S, Powell DW, Singh S, Butt W, Chen Q, and McLeish KR. MAPK-activated protein kinase-2 participates in p38 MAPK-dependent and ERK-dependent functions in human neutrophils. Cell Signal 15: 9931001, 2003.

    Cybulsky AV, Foster MH, Quigg RJ, and Salant DJ. Immunologic mechanisms of glomerular disease. In: The Kidney: Physiology and Pathophysiology (3rd ed.), edited by Seldin DW and Giebisch G. Philadelphia, PA: Lippincott-Raven, 2000, p. 26452697.

    Cybulsky AV, Monge JC, Papillon J, and McTavish AJ. Complement C5b-9 activates cytosolic phospholipase A2 in glomerular epithelial cells. Am J Physiol Renal Fluid Electrolyte Physiol 269: F739F749, 1995.

    Cybulsky AV, Papillon J, and McTavish AJ. Complement activates phospholipases and protein kinases in glomerular epithelial cells. Kidney Int 54: 360372, 1998.

    Cybulsky AV, Takano T, Papillon J, Khadir A, Bijian K, and Le Berre L. The actin cytoskeleton facilitates complement-mediated activation of cytosolic phospholipase A2. Am J Physiol Renal Physiol 286: F466F476, 2004.

    Cybulsky AV, Takano T, Papillon J, Khadir A, Liu J, and Peng H. Complement C5b-9 membrane attack complex increases expression of endoplasmic reticulum stress proteins in glomerular epithelial cells. J Biol Chem 277: 4134241351, 2002.

    Cybulsky AV, Takano T, Papillon J, and McTavish AJ. Complement C5b-9 induces receptor tyrosine kinase transactivation in glomerular epithelial cells. Am J Pathol 155: 17011711, 1999.

    Cybulsky AV, Takano T, Papillon J, and McTavish AJ. Complement-induced phospholipase A2 activation in experimental membranous nephropathy. Kidney Int 57: 10521062, 2000.

    Davies SP, Reddy H, Caivano M, and Cohen P. Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351: 95105, 2000.

    Dudley DT, Pang L, Decker SJ, Bridges AJ, and Saltiel AR. A synthetic inhibitor of the mitogen-activated protein kinase cascade. Proc Natl Acad Sci USA 92: 76867689, 1995.

    Heymann W, Hackel DB, Harwood S, Wilson SG, and Hunter JL. Production of nephrotic syndrome in rats by Freund’s adjuvants and rat kidney suspensions. J Am Soc Nephrol 11: 183188, 2000.

    Hirabayashi T and Shimizu T. Localization and regulation of cytosolic phospholipase A2. Biochim Biophys Acta 1488: 124138, 2000.

    Jou TS and Nelson WJ. Effects of regulated expression of mutant RhoA and Rac1 small GTPases on the development of epithelial (MDCK) cell polarity. J Cell Biol 142: 85100, 1998.

    Kanzaki M and Pessin JE. Insulin-stimulated GLUT4 translocation in adipocytes is dependent upon cortical actin remodeling. J Biol Chem 276: 4243642444, 2001.

    Kraus S, Seger R, and Fishelson Z. Involvement of the ERK mitogen-activated protein kinase in cell resistance to complement-mediated lysis. Clin Exp Immunol 123: 366374, 2001.

    Leslie CC. Properties and regulation of cytosolic phospholipase A2. J Biol Chem 272: 1670916712, 1997.

    Lunn JA, Wong H, Rozengurt E, and Walsh JH. Requirement of cortical actin organization for bombesin, endothelin, and EGF receptor internalization. Am J Physiol Cell Physiol 279: C2019C2027, 2000.

    Malarkey K, Belham CM, Paul A, Graham A, McLees A, Scott PH, and Plevin R. The regulation of tyrosine kinase signalling pathways by growth factor and G-protein-coupled receptors. Biochem J 309: 361375, 1995.

    Mansour SJ, Matten WT, Hermann AS, Candia JM, Rong S, Fukasawa K, Vande Woude GF, and Ahn NG. Transformation of mammalian cells by constitutively active MAP kinase kinase. Science 265: 966970, 1994.

    Marshall CJ. Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell 80: 179185, 1995.

    Meng W, Swenson LL, Fitzgibbon MJ, Hayakawa K, Ter Haar E, Behrens AE, Fulghum JR, and Lippke JA. Structure of mitogen-activated protein kinase-activated protein (MAPKAP) kinase 2 suggests a bifunctional switch that couples kinase activation with nuclear export. J Biol Chem 277: 3740137405, 2002.

    Morgan BP. Effects of the membrane attack complex of complement on nucleated cells. Curr Top Microbiol Immunol 178: 115140, 1992.

    Nicholson-Weller A and Halperin JA. Membrane signaling by complement C5b-9, the membrane attack complex. Immunol Res 12: 244257, 1993.

    Panesar M, Papillon J, McTavish AJ, and Cybulsky AV. Activation of phospholipase A2 by complement C5b-9 in glomerular epithelial cells. J Immunol 159: 35843594, 1997.

    Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K, and Cobb MH. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev 22: 153183, 2001.

    Peng H, Takano T, Papillon J, Bijian K, Khadir A, and Cybulsky AV. Complement activates the c-Jun N-terminal kinase/stress-activated protein kinase in glomerular epithelial cells. J Immunol 169: 25942601, 2002.

    Pippin JW, Durvasula R, Petermann A, Hiromura K, Couser WG, and Shankland SJ. DNA damage is a novel response to sublytic complement C5b-9-induced injury in podocytes. J Clin Invest 111: 877885, 2003.

    Raucher D, Stauffer T, Chen W, Shen K, Guo S, York JD, Sheetz MP, and Meyer T. Phosphatidylinositol 4,5-bisphosphate functions as a second messenger that regulates cytoskeleton-plasma membrane adhesion. Cell 100: 221228, 2000.

    Rosado JA and Sage SO. A role for the actin cytoskeleton in the initiation and maintenance of store-mediated calcium entry in human platelets. Trends Cardiovasc Med 10: 327332, 2000.

    Rossol M, Gartner D, and Hauschildt S. Diverse regulation of microfilament assembly, production of TNF-, and reactive oxygen intermediates by actin modulating substances and inhibitors of ADP-ribosylation in human monocytes stimulated with LPS. Cell Motil Cytoskeleton 48: 96108, 2001.

    Rus HG, Niculescu FI, and Shin ML. Role of the C5b-9 complement complex in cell cycle and apoptosis. Immunol Rev 180: 4955, 2001.

    Seger R and Krebs EG. The MAPK signaling cascade. FASEB J 9: 726735, 1995.

    Seufferlein T, Withers DJ, Mann D, and Rozengurt E. Dissociation of mitogen-activated protein kinase activation from p125 focal adhesion kinase tyrosine phosphorylation in Swiss 3T3 cells stimulated by bombesin, lysophosphatidic acid, and platelet-derived growth factor. Mol Biol Cell 7: 18651875, 1996.

    Shankland SJ, Pippin JW, and Couser WG. Complement (C5b-9) induces glomerular epithelial cell DNA synthesis but not proliferation in vitro. Kidney Int 56: 538548, 1999.

    Shin ML, Rus HG, and Niculescu FI. Membrane attack by complement: assembly and biology of terminal complement complexes. In: Biomembranes, edited by Lee AG. Greenwich, CT: JAI, 1996, p. 119146.

    Sotiropoulos A, Gineitis D, Copeland J, and Treisman R. Signal-regulated activation of serum response factor is mediated by changes in actin dynamics. Cell 98: 159169, 1999.

    Spector I, Shochet NR, Blasberger D, and Kashman Y. Latrunculinsnovel marine macrolides that disrupt microfilament organization and affect cell growth: I. Comparison with cytochalasin D. Cell Motil Cytoskeleton 13: 127144, 1989.

    Topham PS, Haydar SA, Kuphal R, Lightfoot JD, and Salant DJ. Complement-mediated injury reversibly disrupts glomerular epithelial cell actin microfilaments and focal adhesions. Kidney Int 55: 17631775, 1999.

    Tsakiridis T, Bergman A, Somwar R, Taha C, Aktories K, Cruz TF, Klip A, and Downey GP. Actin filaments facilitate insulin activation of the src and collagen homologous/mitogen-activated protein kinase pathway leading to DNA synthesis and c-fos expression. J Biol Chem 273: 2832228331, 1998.

    Tsukita S and Yonemura S. Cortical actin organization: lessons from ERM (ezrin/radixin/moesin) proteins. J Biol Chem 274: 3450734510, 1999.

    Uehata M, Ishizaki T, Satoh H, Ono T, Kawahara T, Morishita T, Tamakawa H, Yamagami K, Inui J, Maekawa M, and Narumiya S. Calcium sensitization of smooth muscle mediated by a Rho-associated protein kinase in hypertension. Nature 389: 990994, 1997.

    Wang F, Herzmark P, Weiner OD, Srinivasan S, Servant G, and Bourne HR. Lipid products of PI(3)Ks maintain persistent cell polarity and directed motility in neutrophils. Nat Cell Biol 4: 513518, 2002.

    White SR, Williams P, Wojcik KR, Sun S, Hiemstra PS, Rabe KF, and Dorscheid DR. Initiation of apoptosis by actin cytoskeletal derangement in human airway epithelial cells. Am J Respir Cell Mol Biol 24: 282294, 2001.

    Winkler DG, Johnson JC, Cooper JA, and Vojtek AB. Identification and characterization of mutations in Ha-Ras that selectively decrease binding to cRaf-1. J Biol Chem 272: 2440224409, 1997.

    Zeng C and Morrison AR. Disruption of the actin cytoskeleton regulates cytokine-induced iNOS expression. Am J Physiol Cell Physiol 281: C932C940, 2001.

    Zhang H, Li H, Aoudjit L, Lemay S, and Takano T. Role of Rho-family of small GTPases in glomerular epithelial cells (Abstract). J Am Soc Nephrol 15: 14A, 2004.