Mitogen-Activated Protein Kinase and Cytoskeleton in Mitogenic Signal Transduction Eisuke Nishida and Yukiko Gotoh Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, Tokyo 113, Japan

I. Introduction

Quiescent (GO/Gl) cells, the growth of which has been arrested by incubation in serum-free medium, do not begin to proliferate unless growth factors (GFs) are given. The GF stimulation of cells triggers a variety of cascades of biochemical and physiological reactions in the cells, resulting eventually in the entry into S phase, initiation of DNA synthesis. Although many possibly crucial cellular events occurring between the GF-membrane receptor binding reaction and the commitment of cell into initiation of the cell cycle have been described, including a marked increase in tyrosine and serinehhreonine phosphorylations of cellular proteins, acceleration of phosphoinositide turnover, activation of N a + / H + exchange, resulting in a rise in pH, a rise in intracellular concentration of Ca2+,and so on, the mechanism by which GFs initiate cell growth is largely unknown (Pardee, 1989). One of primary and common responses to many GF-induced transmembrane signals is the rapid activation of transcription of the proto-oncogene c-fos and other early-response genes (Greenberg and Ziff, 1984) and de nouo synthesis of some proteins is shown to be required for the reentry of the cells into the cell cycle. Thus, the primarily important point for mitogenic signal transduction resides in the signaling pathways from the cell membrane to the nucleus. The cytoskeleton comprises microtubules (MT), actin filaments, and intermediate filaments. These three types of protein filaments exist in cytoplasm, outside nucleus, and inside the cell. For example, most cytoplasmic MTs originate from centrosome near nucleus, extending Inrernutionul Review


Cyrology. Vol. 138

21 1

Copyright Q 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.



radially near plasma membrane in interphase cells (Fig. 1). These structures are likely to play a role in determining not only cell shape but also the localization and orientation of intracellular organelles, such as endoplasmic reticulum, lysosome, Golgi apparatus, and mitochondria, and are presumed to be involved in a variety of intracellular transport systems. An important feature of cytoskeleton is its dramatic change in architecture or shape during mitosis (M phase). The cytoplasmic MT network disappears and the mitotic spindle forms (Fig. 1). In cultured cells the actin stress fibers existing in interphase cells disappear and the contractile ring consisting of actin, myosin, and others appears during telophase. The mitotic spindle is the specialized apparatus where chromosomes segregated. The contractile ring then functions to divide the cell into two. Although recent studies on cytoskeleton in relation to cell motility, cell shape, and many other cellular physiological functions have grealy increased our understanding of functions of the cytoskeleton, relatively little attention has been paid to possible involvement of the cytoskeleton in the mitogenic signal transduction. The first half of this article focuses on the GF-induced actin cytoskeletal reorganization (Section 11) and the possible functions of cytoplasmic MTs in the GF-signaling pathway (Section 111). Mitogen-activated protein (MAP) kinase is a serinehhreonine-specific protein kinase, the activation and phosphorylation of which are induced by a variety of mitogens. MAP kinase was discovered, as it can utilize MAP (microtubule-associated protein) 2, one of the best known MTassociated proteins, as an excellent substrate in vitro. Sturgill and coworkers found MAP kinase as an insulin-stimulated protein kinase in 3T3-Ll adipocytes that can phosphorylate MAP 2 in vitro (Ray and Sturgill, 1987; 1988a), and we found it as a protein kinase that is commonly activated in human and rat fibroblastic cells by a variety of mitogens including epiderminal growth factor (EGF), platelet-derived growth factor (PDGF), insulin-like growth factor-I (IGF-I), fibroblast growth factor (FGF), 12-0-tetradecanoylphorbol13-acetate (TPA), and fetal calf serum (FCS) (Hoshi et al., 1988, 1989). At least two MAP kinases (43 kDa and 41 kDa) with similar characteristics exist in mammalian cultured cells (Gotoh et al., 1990a,b). Recently, we have found that a Xenopus 42-kDa MAP kinase is tyrosine-phosphorylated and activated during M phase of oocyte cell cycles and is involved in the interphase-M phase transition of MT dynamics (Gotoh et al., 1991a). Thus, we can use the term “MAP kinase” as it stands for MAP 2 kinase, mitogen-activated protein kinase and M phase (or maturation)activated protein kinase. The latter half of this article (Sections IV and V) describes identification, properties, activation mechanism, and functions of MAP kinase.

FIG. 1 Interphase-M phase transition of microtubule arrays. Microtubules were formed in vitro by incubating an M phase extract (upper) or an interphase extract (lower) with purified centrosomes and visualized by the immunofluorescence staining with anti-tubulin antibody as described by Gotoh et a / . (1991a). The M phase and the interphase extracts were prepared from mature Xenopus oocytes (metaphase 11) and activated Xenopus eggs, respectively. The typical cell stainings of the mammalian mitotic spindle (upper) and the cytoplasmic microtubule networks (lower) are shown in the insets.



II. Cytoskeletal Reorganization Elicited by Growth Factors A. Induction of Membrane Rufflings as One of Common Cellular Responses t o Growth Factors

When cultured cells are transformed by pp60""" or related oncogenes, the actin stress fibers are dissoluted (Pollack et al., 1975) and the plasma membrane ruffles. The membrane ruffles are highly dynamic structures consisting of many actin filaments. This dramatic change in the actin architecture, accompanied by cell rounding, is one of the most conspicuous features of the transformed cells. The similar change in actin cytoskeleton has repeatedly been observed in cultured cells stimulated by a variety of GFs, such as PDGF, EGF, and insulin (Bockus and Stiles, 1984; Brunk et al., 1976; Chinkers et al., 1979, 1981; Connolly et al., 1984; Goshima et al., 1984; Kadowaki et al., 1986; Mellstrom et al., 1983; Miyata et al., 1988a). Among the changes observed (disorganization or dissolution of the actin stress fibers, occurrence of membrane rufflings, and concentration of actin filaments in the ruffles), the formation of membrane ruffles is easily observed under the phase-contrast microscope without any laborious experimental procedures. Therefore, we can relatively easily answer questions of what stimuli trigger induction of membrane rufflings, how and when membrane rufflings begin and subside, and how various conditions affect the induction of the rufflings. Human KB cells exhibit rapid induction of striking membrane ruffling in response to GFs (Fig. 2, Goshima et al., 1984; Kadowaki et al., 1986), and they possess a relatively large number of EGF receptors, IGF-I receptors and insulin receptors (Kadowaki et al., 1986), so the KB cells are suitable for analyses of the GF-induced membrane ruffling. The detailed studies with monoclonal antibodies against each of these GF receptors have confirmed that EGF, insluin, and IGF-I elicit membrane ruffling through their binding to their corresponding receptors (Kadowaki et al., 1986). Further, tyrosine phosphorylation mediated by the receptor-associated tyrosine kinase is shown to be required for the ligand-induced membrane ruffling formation, as specific antibodies against the tyrosine kinase domain of the receptors block the induction of ruffling (Izumi et al., 1988). Thus, the membrane ruffling is one of intracellular events mediated by G F receptors, which possess the tyrosine kinase activity. Cell staining with rhodamine-labeled phalloidin, which binds specifically to filamentous actin, demonstrated that after GF stimulation, the actin stress fibers disappear and actin filaments become co-localized with membrane ruffling regions. Interestingly, insulin- or IGF-I-induced membrane ruffles are morphologically different from EGF-induced ones; insulin- or



IGF-I-induced ruffles are visible as dark and irregular rims under a phasecontrast microscope, whereas EGF-induced ruffles are visible as curtainlike folds or rod-like spikes (Fig. 2). In this context, it is noteworthy that in KB cells insulin and IGF-I elicit tyrosine phosphorylation of similar sets of cellular proteins, whereas EGF elicits tyrosine phosphorylation of some overlapping and some different cellular proteins (Kadowaki et af., 1987). Therefore, insulin receptors and IGF-I receptors may transmit almost identical signals, whereas EGF receptors transmit similar but different signals. The induction of ruffling membrane is rapid and transient. We have developed the computer-aided image processing method to visualize and quantitate the cell motility and cell morphology (Miyata et af., 1988a). The quantitative analysis by the programmed trace mode of the AVEC system has shown that in KB cells the motion of the membrane rufflings is observed within 2 minutes, reaches the maximum level within 4-8 minutes,

Phase contrast



+IGF-I (10 nM)

+EGF (10 nM)

FIG. 2 Membrane ruffling induced by IGF-I and EGF. KB cells were treated with IGF-I or

EGF for 4 minutes at 37°C and then fixed and stained with rhodamine-phalloidin (right panel). Phalloidin stains filamentous actin. Phase-contrast micrographs (left panel) are also shown.



and then rapidly declines within 10-20 minutes after the addition of GFs. The rapid and transient induction of membrane ruffling is commonly observed in rat fibroblast 3Yl cells and mouse Balb/c 3T3 cells stimulated by EGF, insulin, or IGF-I, although the time course is slightly different from that in KB cells. The induction of membrane ruffling is always accompanied by disorganization or diminution of the actin stress fibers and apparent concentration of actin filaments in membrane ruffling regions. Thus, the rapid actin cytoskeletal reorganization, which results in the conspicuous morphological change such as the induction of membrane ruffling, is one of common cellular responses to GFs. It is naturally thought that the rapid and reversible reorganization of cytoskeleton as well as cell membrane may be concerned with the stimulation of pinocytosis (fluid-phase endocytosis). Indeed, a number of reports described the stimulation of pinocytosis in cells stimulated by GFs, for example, A431 cells by EGF (Haigler et al., 1979), SW-8 cells by EGF (Wiley and Cunningham, 1982), 3T3-Ll adipocytes by insulin (Gibbs et al., 1986), and KB cells by EGF, insulin, and IGF-I (Miyata et al., 1988b). We found further that the fluid-phase exocytosis is also stimulated by EGF, insulin, and IGF-I in KB cells (Miyata et al., 1988b). We then proposed that the stimulation of fluid-phase influx and efflux (pinocytosis and fluid-phase exocytosis), like induction of membrane ruffling, may constitute one of the common early cellular responses to GFs. Although the functional and physiological significance of the stimulation of fluidphase influx and efflux as well as the induction of membrane ruffling in response to GFs is not clearly understood at present, it is interesting that H-vas proteins also stimulate membrane ruffling and pinocytosis (Bar-Sagi and Feramisco, 1986). A possible coupling may be proposed between the mitogenic activity and these early cellular responses. However, the coupling, if any, may not be straightforward as EGF inhibits, rather than stimulates, multiplication of KB cells in a dose-dependent manner (Koyasu et al., 1988),whereas EGF induces the stimulation of the fluid-phase influx and efflux and the induction of membrane ruffling. B. Regulation by lntracellular Ca2+ and cAMP Concentrations

Intracellular concentrations of Ca2+ and cAMP play regulatory roles in many cellular physiological functions. The induction of membrane ruffling and the stimulation of fluid-phase endocytosis and exocytosis are both sensitive to increased concentrations of Ca2+and cAMP (Miyata et al., 1989a). An increase in intracellular Ca2 concentration by treatment with a calcium ionophore (A23187) or an increase in intracellular cAMP level +



by treatment with dibutyryl cAMP or an adenylate cyclase activator (forskolin) completely inhibits the EGF-, insulin-, or IGF-I-induced formation of ruffling membranes in KB cells. Interestingly, the stimulation of fluid-phase influx and efflux by these GFs is also completely inhibited by either of these treatments, suggesting that the induction of membrane ruffling and the enhancement of influx and efflux have a common regulatory mechanism(s) involving intracellular concentrations of Ca2+ and CAMP. The rise in intracellular Ca2+ or cAMP level, however, does not inhibit the EGF or IGF-I binding to the cells nor subsequent tyrosine phosphorylation of their receptors (Miyata et al., 1989a; Y. Miyata and E. Nishida, unpublished observations). Therefore, Ca and/or CAMP-sensitive intracellular ractions exist downstream from the receptor kinase activation in the process of these cellular responses. Because EGF increases the intracellular concentation of Ca2+in KB cells, it is possible that Ca2+ plays a role in the negative feedback of these EGF-induced early cellular responses. Considering the fact that the ruffling membrane is accompanied by reorganization of actin cytoskeleton, the rise in Ca2+or cAMP may regulate the functions of cytoskeletal proteins presumably responsible for these early cellular responses. C. Protein Kinase C-Dependent and -Independent Pathways

The role of protein kinase C (PKC) as a modulator of GF stimulation has been the subject of considerable importance and interest (Nishizuka, 1984, 1986). Further, many efforts have been directed to studies on “cross talk” betwen PKC and EGF receptors (Schlessinger, 1986; Davis, 1988). Because the induction of ruffling membrane seems a powerful tool for probing the acute effects of various GFs, it is of great interest to examine the role of PKC in this GF-signaling pathway that results in induction of membrane ruffling (Miyata et al., 1989b). To analyze the possible role of PKC, we modulated the intracellular distribution and state of PKC by treating KB cells with 4-phorbol-12,13dibutyrate (PDBu), one of the potent tumor promoters. After a short-term PDBu treatment (100 ng/ml PDBu for 30 minutes), about 80% of the PKC was found in the membrane fraction (the PKC-activated state). Prolonged treatment of KB cells with PDBu (200 ng/ml PDBu for 15 hours) induced nearly complete depletion (down-regulation) of the PKC activity in both the cytosolic and membrane fractions (the PKC-depleted state). In the PKC-activated cells, EGF, insulin, and IGF-I no longer induced membrane ruffling. In these cells the high affinity EGF binding was lost, and EGF neither induced tyrosine phosphorylation of the EGF receptors nor increased the intracellular Ca2+concentration. In the PKC-depleted cells,





+IGF-I (10 nM)

+EGF (10 nM)



Down-Regulated FIG. 3 EGF does not induce membrane ruffling in PKC-depleted cells. KB cells with (right) or without (left) prior prolonged PDBu treatment were incubated with IGF-I or EGF for 4 minutes and then stained with rhodamine-phalloidin. IGF-I could induce membrane ruffling both in control and PKC-depleted cells, but EGF could induce membrane rufRing only in control cells.

insulin and IGF-I induced membrane ruffling while EGF did not elicit membrane ruffling at all (Fig. 3). In these PKC-depleted cells, EGF binding to the cells, EGF-induced tyrosine phosphorylation of the receptors, and EGF-induced increase in intracellular Ca2' level occurred normally. These results suggest that activatable PKC is required for the induction of membrane ruffling by EGF but not for the induction by insulin or IGF-I. The experiments with two selective kinase inhibitors, H-7 and H-8 (Hidaka et al., 1984), confirmed this conclusion. Thus, H-7 and H-8, two potent inhibitors for PKC (H-7 is severalfold more potent than H-8),



inhibited the EGF-induced membrane ruffling (H-7 was severalfold more potent than H-8 in this case, too), whereas they did not inhibit the insulin/ IGF-l-induced membrane ruffling at all. These results taken together indicate that the membrane ruffling involving cytoskeletal reorganization can be induced by GFs through two different signal transduction pathways; one is PKC-dependent and the other is PKC-independent. Although PKC mediates the EGF-induced membrane ruffling, the PKC activatin is not sufficient for the membrane ruffling induction in KB cells because TPA or PDBu treatment alone does not elicit ruffling. We propose that PKC plays dual roles in the EGF-induced membrane ruffling. The PKC activation induced by EGF, as evidenced by demonstration of EGFinduced PKC translocation to the membrane (Miyata e? al., 1989b), is necessary for transmission of the EGF signal to induction of membrane ruffling, but the activated PKC may in turn inhibit the duration of EGF effect. This may account for the transient nature of the ruffling. It is likely that PKC plays a negative feedback role in the EGF signaling.

111. Regulation of Mitogenic Signal Transduction by Cytoplasmic Microtubules A. Microtubule Disruption and Initiation of DNA Synthesis

MT cytoskeleton is required for segregation of chromosomes during M phase of cell cycles and is therefore essential for cell division. It has been known that MT-disrupting agents or anti-MT drugs, such as colchicine, vinblastine, and nocodazole, block the cell cycle of proliferating cells at M phase by interfering with the function and integrity of spindle MTs during M phase, indicating the positive role of MTs in propelling the cell cycle. However, in quiescent (arrested at GO/G I phase) cells, disruption of cytoplasmic MTs by treatment of the cells with the anti-MT drugs induces the exit of the cells from the quiescent state. Vasiliev e? al. (1971) found that colcemid stimulated the initiation of DNA synthesis in primary cultures of density-arrested mouse embryo fibroblast-like cells in the presence of 10% bovine serum. Many reports, then, established that as for confluently cultured, well-arrested fibroblastic cells, colchicine and other anti-MT drugs potentiate the effect of GFs on initiation of DNA synthesis (Otto, 1987). Subsequently, Crossin and Carney (1981a,b) demonstrated that colchicine treatment alone, in the absence of serum GFs, is capable of initiating DNA synthesis in secondary cultures of mammalian and avian fibroblasts. Although their experiments were controlled adequately in that



the target of colchicine’s action was only MTs, the stimulation of thymidine incorporation (DNA synthesis) by colchicine was no more than 2-3fold relative to untreated cells, partly because the cultures might not be completely rendered quiescent. To analyze the mechanism by which MT disruption could function as a mitogen, it was required to identify cell lines more responsive to mitogenic action of anti-MT drugs. We found that a rat embryonic fibroblastic cell line 3Y 1, which can be easily rendered quiescent by incubating confluent cultures in serum-free medium, is highly responsive to colchicine and other anti-MT drugs in the absence of GFs (Shinohara et al., 1989). The maximal stimulation of DNA synthesis by these drugs attained 25-35% of that induced by 10% fetal calf serum and was comparable to that induced by any of individual peptide GFs such as EGF. Tubulin biosynthesis is noted for its autoregulatory system in which the elevated level of free tubulin reduces synthesis of tubulin (Ben-Ze’ev er al., 1979; Cleveland et al., 1981; Yen et al., 1988), indicating that cells have a “sensor” for free tubulin level. However, an elevated level of free tubulin is not required for initiation of DNA synthesis induced by the antiMT drugs because incubation of quiescent 3Y 1 cells with high concentrations of vinblastine or those of vinblastine plus colcemid, which formed large tubulin paracrystals concomitant with disruption of cytoplasmic MTs without increasing free tubulin level, initiated DNA synthesis efficiently (Shinohara er al., 1989). It is MT disruption per se that acts as a mitogen. Moreover, it was confirmed that colchicine treatment did not induce secretion of any mitogenic factors from 3Y1 cells. Thus, disruption of normal cytoplasmic MT networks in quiescent cells may elicit intracellular signals leading to cell proliferation. In other words, the cytoplasmic MTs may function to prevent the cells from exiting from the quiescent state. We can assume that MT disruption induces intracellular events that also occur in normal mitogenic signaling pathways evoked by G F stimulations. How is the MT disruption-induced pathway related to the normal mitogenic signaling pathway? From what point do both pathways converge? We first directed our attention to MAP kinase, which is shown to be rapidly activated in common by various growth factors (Sections I, IV, and V). Our data have shown that colchicine treatment of quiescent 3Y1 cells induces MAP kinase activation through phosphorylation reactions (Shinohara-Gotoh er al., 1991). It is likely that MT disruption, like G F stimulations, induces activation of MAP kinase activator(s)that can induce phosphorylation and activatin of MAP kinase. Because the chain reaction of phorphorylation cascades begins with activation of the receptorassociated tyrosine kinase in the case of GF stimulations and because the MAP kinase activation is one of early cellular responses (within 13 minutes of GF stimulations), it is reasonably assumed that MT disruption



may mimic the GF stimulations from the starting point of the GF signaling pathway, that is, the receptor activation. B. Regulation of lntracellular Dynamics of Growth Factor Receptors by Cytoplasmic Microtubules

The GF such as EGF induces clustering of its receptors into coated vesicles, which are then changed into endosomes (Fig. 4), and the endosomes are internalized and fused into lysosomes (Willingham and Pastan, 1982). This ligand-induced endocytosis or cellular routing of GF receptors may be tightly coupled to tyrosine kinase activity of the receptors (Glenney et al., 1988; Honegger et al., 1987), which is shown to be indispensable for the mitogenic signal transduction (Chen et al., 1987; Moolenaar et al., 1988). It is well known that Golgi apparatus, lysosomes, and endoplasmic reticulum, which are located in the perinuclear regions at MT-organizing centers (centrosomes), are dispersed by disruption of cytoplasmic MTs (Moskalewski et al., 1975; Terasaki et al., 1986; Swanson et al., 1987;

FIG. 4 EGF-induced endosomes. KB cells were incubated with (right) or without (left) EGF (10 nM) for 30 minutes, and the EGF receptors were visualized by immunofluorescence staining with anti-EGF receptor antibody (528 IgG). EGF-induced endosomes are clearly

visible in EGF-treated cells (right).



Matteoni and Kreis, 1987). Further, translocation and clustering of endosomes and lysosomes depend on cytoplasmic MTs (Matteoni and Kreis, 1987). We hypothesized that MT disruption induces changes in both the oligomeric state and the steady state cellular routing of G F receptors, which lead to the enhancement of their tyrosine kinase activity. To examine this hypothesis, we chose A431 cells, which express extremely large amounts of EGF receptors. Our results (Y. Gotoh and E. Nishida, unpublished data) are the following: (1) MT disruption by treatment with anti-MT drugs induces the formation of endosome-like intracellular aggregates containing EGF receptors in a reversible manner. (2) the oligomerization (or dimerization) of the EGF receptors occurs in the colchicine-induced, endosome-like aggregates. (3) The tyrosine kinase activity of the intracellular membrane fraction in the colchicine-treated cells is greatly enhanced. Thus, MT disruption in A43 1 cells induces intracellular accumulation of endosome-like vesicles, which is accompanied by the oligomeriaation and activation of EGF receptors. This demonstrates that dimerization and tyrosine kinase activation of EGF receptors can occur without EGF in cells, being consistent with the idea that dimerization of EGF receptors (which is stabilized by EGF binding in the case of normal EGF stimulation) is sufficient for tyrosine kinase activation (Schlessinger, 1988). There are two explanations for the MT disruption-induced change in intracellular dynamics or distribution of EGF receptors. First, the mobility of receptors in cell membranes may be restricted by membrane-underlying cytoplasmic MTs, so the receptors that are released from the restriction by MT disruption may freely interact with one another to oligomerize. Second, the ligand-independent, constitutive routing of endosomes containing oligomerized EGF receptors may depend on integrity of cytoplasmic MTs, the disruption of which may prevent the cellular routing, leading to accumulation of endosomes inside cells. We favor the latter, as our data concerning the effect of MT disruption on the dynamics of EGF and its receptors have shown that the degradative pathway (translocation pathway to lysosomes) of the internalized EGF and EGF receptor in EGFtreated cells depends on cytoplasmic MTs (Y. Gotoh and E. Nishida, unpublished results). It is certain that intracellular distribution, oligomeric state, and tyrosine kinase activity of GF receptors depend on cytoplasmic arrays of MTs. This reveals a previously unidentified function of cytoplasmic MTs, regulation of intracellular GF receptor dynamics. We propose a working hypothesis that cytoplasmic MTs negatively regulate the initiation of cell proliferation, that is, keep the cells quiescent, by preventing ligand-independent, spontaneous oligomerization and activation of GF receptors.



IV. Mitogen-Activated Protein Kinase: A MitogenActivated Serine/Threonine Kinase That Phosphorylates Microtubule-Associated Protein 2 in Vitro A. Mitogenic Stimulation of Phosphorylation of MAP-like Proteins

Phosphorylation and dephosphorylation reactions play crucial regulatory roles in many cellular functions. Recent studies have established the importance of protein phosphorylation reactions in the mitogenic signal transduction, partly because the tyrosine kinase activity many G F receptors possess has been shown to be indispensable for initiation of cell growth (Yarden and Ullrich, 1988; Schlessinger, 1988) and because the receptor for tumor promoters such as TPA was identified as PKC, a phospholipiddependent serinekhreonine kinase, and PKC has been shown to be activated by diacylglycerol generated by the enhanced phosphatidylinositol turnover in response to various GF stimulations (Nishizuka, 1984, 1986). Identification of important substrate proteins for tyrosine kinase has been one of critical subjects because many oncogene products as well as GF receptors have the tyrosine kinase activity. However, serinekhreonine phosphorylation is also thought to play critical roles, as a burst in serine/ threonine phosphorylation occurs after mitogenic stimulation of cells and many regulatory mechanisms involving serinekhreonine phosphorylation are known. The existence of the proto-oncogene products that themselves are serinehhreonine kinases, such as Raf-1 and c-mos, further lends support to the importance of serinelthreonine phosphorylations in the mitogenic signal transduction. In 1985 Sat0 and co-workers reported that treatment of quiescent rat fibroblastic 3Yl cells with EGF, insulin, TPA, or fresh fetal calf serum stimulates markedly the phosphorylation of cytoskeleton-associated 350kDa and 300-kDa proteins, which can be immunoprecipitated with antibodies raised against brain high molecular weight MT-associated proteins. They further showed that the phosphorylated form of these high molecular weight proteins is translocated into the U 1-ribonucleoprotein region of the nucleus (Sato et al., 1986a,b). The marked enhancement of the phosphorylation of the 350-kDa and 300-kDa proteins occurs also in the 3Y1 cells stimulated by other GFs, such as PDGF, IGF-I, and FGF (Nishida et al., 1988). In quiescent 3Y 1 cells, any of these GFs or phorbol esters can act individually as a mitogen, that is, incubation of the cells with one of these factors results in the entry of the cells into S phase (Sato et al., 1985;



Nishida et al., 1988). The enhanced phosphorylation of both proteins always occurs on almost exclusively serine residues and the immunoprecipitated proteins from 32P-labeledcells stimulated by these mitogens produce similar phosphopeptide mapping patterns irrespective of the stimulant (Nishida et al., 1988), suggesting that a wide variety of GFs may elicit the phosphorylation of the same sites (serine residues) on the 350-kDa and 300-kDa proteins. We thought that a certain serinehhreonine-specific protein kinase that is responsible for the phosphorylation is commonly activated upon stimulation of the cells with these various GFs and phorbol esters. One of possible candidates might be PKC, but it has been shown that in fibroblastic cells, EGF, IGF-I, and insulin can induce no or little activation of PKC. Then, a possibility has risen that a previously unidentified mitogen-activated serine-threonine kinase exists. B. Universal Activation of MAP Kinase by Mitogenic Stimuli

MAP 2 is one of the best characterized MAPSand is a major high molecular weight MAP in mammalian and avian brain. MAP 2 is a long, filamentous, heat-stable protein and can serve as a good substrate in uitro for various kinases including CAMP-dependent protein kinase, Ca/calmodulin-dependent protein kinase 11, PKC, and the receptor-associated tyrosine kinases (Sloboda et al., 1975; Vallee, 1980; Yamamoto et al., 1985; Tsuyama et al., 1986; Akiyama et al., 1986; Kadowaki et al., 1985; Nishida et al., 1987). We have used MAP 2 as an exogenous substrate (MAP 2 is not expressed in fibroblastic cells) to detect activation of the kinase activity in extracts from mammalian quiescent fibroblastic cells stimulated by mitogenic factors. Then, we found that treatment of the cells with EGF, PDGF, FGF, insulin, IGF-I, TPA, PDBu, or fetal calf serum activates rapidly the kinase activity toward MAP 2 (Hoshi et al., 1988). MAP 2 that was phosphorylated by the kinase activity induced by different stimuli gave similar phosphopeptide mapping patterns. The kinase activity induced by any of these mitogenic factors was not active toward histone or casein in uitro and phosphorylated serine (and slightly threonine) residues of MAP 2. This kinase was clearly distinct from S6 kinase(s), which were known to be commonly activated by various GFs. From these results, we suggested that different mitogenic factors activate in common a previously unidentified serinelthreonine-specific protein kinase in quiescent fibroblastic cells (Hoshi et al., 1988). We have then isolated and characterized this kinase, which is activated by EGF or TPA treatment of the mammalian fibroblastic cells (Hoshi et al., 1989; Gotoh et al., 1990a,b; see Section V). Independently, Ray and Sturgill (1987) described an insulin-stimulated, serine/



threonine-specific protein kinase that phosphorylated MAP 2 in uitro in 3T3-Ll adipocytes. Subsequently, they partially purified their kinase and showed that the kinase preferentially phosphorylates MAP 2 in uitro (Ray and Sturgill, 1988a) and that their kinase may be a major tyrosine kinase target protein, pp42 (Rossomando et al., 1989), which has been shown to be commonly phosphorylated on tyrosine residues upon stimulation with a variety of mitogenic factors (Cooper et al., 1982, 1984). These results together with later studies (see Section V) indicate that our group and Sturgill’s group described the same or similar kinase. Both groups called this kinase MAP 2 kinase at first. But because it is evident that MAP 2 is not a sole substrate for this kinase, this kinase has been called MAP kinase as it stands for mitogen-activated protein kinase. Our recent study, however, revealed that this kinase is activated also in M phase of the cell cycle (see Section V), indicating that MAP kinase stands for M phase-activated protein kinase as well. But we propose that the term “MAP” of MAP kinase should be regarded as a meaningless term because its in uiuo substrates have not yet been fully identified and the activation and function of the kinase may not be limited to GO/G1 or M phase. It is remarkable that those stimuli that initiate the DNA synthesis in quiescent cells are able to induce activation of MAP kinase. In addition to the aforementioned GFs and phorbol esters such as PDGF, EGF, and TPA, another kind of mitogenic factor, including bombesin and bradykinin, can activate MAP kinase (Y. Gotoh and E. Nishida, unpublished data). There are PKC-dependent and -independent pathways in MAP kinase activation (Hoshi et al., 1989). Furthermore, as mentioned in Section 111, disruption of cytoplasmic MTs by the anti-MT drug treatment induces activation of MAP kinase as well as initiation of DNA synthesis in quiescent fibroblastic cells (Shinohara et al., 1989; Shinohara-Gotoh et al., 1991). Another remarkable feature of MAP kinase is the rapidity of the activation. In quiescent cultures of various mammalian fibroblastic cells, MAP kinase is activated half-maximally within 1 minute after G F simulations, is maximally activated within 3-5 minutes, and then gradually decreases (Fig. 5; Gotoh et al., 1990a). Sturgill and co-workers presented evidence that insulin- or EGF-stimulated MAP kinase is phosphorylated on tyrosine and threonine residues (Ray and Sturgill, 1988b)and that the MAP kinase is only active when both residues are phosphorylated (Anderson et al., 1990). Thus, MAP kinase is characterized as a serine-threonine-specific kinase that is commonly and rapidly activated by a variety of GFs through phosphorylation on tyrosine and threonine residues. Moreover, it has been reported that MAP kinase can phosphorylate and activate S6 kinase I1 (Sturgill et af., 1988) and 70K S6 kinase (Gregory et al., 1989), although a recent report argued that MAP kinase does not activate or phosphorylate



1000 I




! ll







do o;






Time (min)

FIG. 5 Rapid activation of MAP kinase after mitogenic stimulation of mammalian fibroblastic cells. Quiescent rat fibroblastic 3Y1 cells were treated with 10 nM EGF (Oh100 ng/ml TPA (O),or 10 nM IGF-I (A) for indicated times, and then MAP kinase activity of the cell extracts toward MAP 2 was determined. The kinase activity is shown in an arbitrary unit.

70 K S6 kinase (Ballou et al., 1991). MAP kinase is therefore thought to have a critical role in a network of protein kinases in mitogenic signal transduction. The mechanism by which MAP kinase is activated remains to be unraveled. The discovery of the enhanced phosphorylation of the 350-kDa and 300-kDa proteins (MAP-like proteins) in mitogen-stimulatedcells triggered our studies on MAP kinase, and a correlation was found among three biological phenomena, the phosphorylation of these proteins, the activation of MAP kinase, and the initiation of DNA synthesis, in Balb/c 3T3 cells stimulated by TPA and/or insulin (Kawakami et al., 1991). This might suggest the physiological significance of MAP kinase, but it is still uncertain whether MAP kinase is responsible for the phosphorylation of these high molecular weight proteins. Identification of target proteins of MAP kinase is a critical step for our understanding of the function of MAP kinase in GO/Gl phase. In addition to mitogenic factors, nerve GF (NGF), a differentiating factor, has been shown to activate MAP kinase in PC12 cells (Gotoh et al., 1990a; Miyasaka et al., 1990).Biochemical characterization has shown that NGF and EGF activate the same MAP kinases (43-kDa and 41-kDa) in PC12 cells as do a variety of mitogens in fibroblastic cells (Gotoh et al., 1990a). Furthermore, it has been reported that MAP kinase is activated in adrenal chromaffin cells in response to stimuli that induce catecholamine secretion (Ely et al., 1990)and that mitogenic stimulationofT lymphocytes is also accompanied by activation of MAP kinase (Nel et al., 1990). MAP



kinase may play an important role in many different signal transduction pathways.

V. Function and Properties of Mitogen-Activated Protein Kinase A. Activation of MAP Kinase during M Phase

During the cell cycle a burst in protein phosphorylation and kinase activation occurs most notably in M phase (Maller et al., 1977; Karsenti et al., 1987; Cicirelli et al., 1988; Pelech et al., 1988; Sanghera et al., 1990b) and may be important in tiggering the dramatic reorganization of the cell, including cytoskeletal reorganization, nuclear envelope breakdown, and chromosome condensation. We inquired whether MAP kinase is activated during M phase. Immature Xenopus oocytes which are naturally arrested at the G2/M border in first meiotic prophase maturate upon progesterone treatment, and after the meiotic cell cycles the oocytes are matured and are arrested at the second metaphase. Mature oocytes (unfertilized eggs) are released from arrest in metaphase I1 upon fertilization or elecrtrical activation. The fertilized eggs then undergo early embryonic cell cycles. These processes are characterized by repetition of M phase and interphase and, therefore, would be suitable for study on the analysis of M phase. Because rnyelin basic protein (MBP) and MAP 2 are major in uitro substrates of MAP kinase (Hoshi et al., 1989; Anderson et al., 1990; Ahn et al., 1990, and see later), we measured the kinase activities toward MAP 2 and MBP in cell extracts obtained from various stages. We found that both kinase activities oscillate coincidentally and peak at M phase during the meiotic and the first mitotic cell cycles (Gotoh et al., 1991a). The major M phase-activated MAP 2/MBP kinase has been purified (molecular weight 42 K) from mature oocytes and identified as MAP kinase (Gotoh et al., 1991a). Analyses with 32P-labeledoocytes and eggs gave us an interesting finding that Xenopus MAP kinase is phosphorylated on tyrosine and serinelthreonine residues in M phase and the phosphorylation correlates with activation of MAP kinase as a kinase (Gotoh et al., 1991a). Xenopus MAP kinase might correspond to the 42 K phosphotyrosine-containing protein that has been detected in Xenopus mature oocytes (Cooper, 1989). A 44 K MBP kinase that has been purified from starfish maturing oocytes (Sanghera et al., 1990b) may be a starfish MAP kinase or a related enzyme. M phase-promoting factor (MPF), which consists of p34cdc2and cyclin B and has a strong histone H1 kinase activity, is a central control element



of the onset of M phase (Murray and Kirschner, 1989;Maller, 1990;Nurse, 1990). Therefore, it is important to see the relationship between MPF and MAP kinase in M phase. As mentioned, MAP kinase activation in GO/Gl phase is triggered by extracellular stimuli, such as GFs, and then it is a critical question what triggers the activation of MAP kinase in M phase. MPF likely acts as an upstream activator of MAP kinase, as we observed that activation of MAP kinase lags slightly behind activation of the histone HI kinase activity (Gotoh et al., 1991a). But it is also certain that there should exist M phase kinases upstream, or independent, of MPF. To address these questions, we injected purified MPF or purified MAP kinase into immature Xenopus oocytes. The MAP kinase was activated dramatically after injection of MPF, whereas the histone H1 kinase was not activated at all by injection of MAP kinase (Gotoh e ta l., 1991b).Moreover, in cell-free extracts prepared from Xenopus interphase eggs, the addition of purified MPF resulted in a time-dependent activation of MAP kinase, whereas no activation of the histone H1 kinase occurred after the addition of purified MAP kinase (Gotoh et al., 1991b). Thus, activation of MAP kinase in M phase occurs under the control of MPF.

6. Properties of MAP Kinase

We purified the Xenopus M phase-activated MAP kinase from unfertilized (mature) eggs by sequential chromatography to apparent homogeneity (Gotoh et al., 1991a). In the final purification step on Mono Q, the elution of the only detectable band on silver-stained SDS-polyacrylamide gel (42kDa) coincided completely with the elution of the MBP/MAP 2-phosphorylating activity. Moreover, that the 42-kDa protein is the MAP kinase itself has been confirmed by the kinase detection assay within gels after SDS-PAGE (Fig. 6). This assay method is a modification (Gotoh er al., 1990a,b) of the method originally described by Kameshita and Fujisawa (1989). In our assay the kinase preparation is electrophoresed in an SDSpolyacrylamide gel containing the substrate protein MBP, the kinase is then renatured in the gel, the gel is incubated with [cz-~*P]ATP, and the MBP in the gel is phosphorylated by the renatured kinase there. Then, the MAP kinase polypeptide can be detected after SDS-PAGE by the MBP kinase activity. The MAP kinase activity in the kinase detection assay within gels correlates well with the ordinary kinase assay in test tube. For example, Xenopus MAP kniase is inactivated by phosphatase 2A treatment (Gotoh er a!., 1991a), and this inactive form of MAP kinase is not detectable at all in the kinase detection assay within gels (Y.Gotoh, T. Yamashita, and E. Nishida, unpublished data).



FIG. 6 Purified Xenopus M phase MAP kinase. Xenopus MAP kinase (42 kDa) was purified as described by Gotoh et a/. (1991a), and analyzed by the silver staining (left)and the kinase detection assay within gels containing MBP (right) after SDS-PAGE.

Isolation of mammalian mitogen-activated MAP kinase from rat fibroblastic cells stimulated by EGF has been carried out by sequential chromatography (Gotoh et al., 1990a). In these chromatographic steps, MAP 2 kinase activity and MBP kinase activity behaved almost identically, confirming that MAP 2 and MBP can serve as good in uitro substrates for MAP kinase. The isolated MAP kinase was eluted in gel filtration as a single peak corresponding to an apparent molecular weight of 40-45 K. The kinase detection assay within MBP-containing polyacrylamide gels after SDS-PAGE revealed that two kinase polypeptides, the apparent molecular weights on SDS-PAGE of which were 43 K and 41 K, were detected in the purified mammalian MAP kinase fraction (Gotoh el al., 1990a). These two kinase polypeptides can be detected in MBP- or MAP 2-containing gels but not detected at all in histone- or casein-containing gels. Moreover, these two kinases are major MBP/MAP 2-phosphorylating enzymes that can be detected in extracts obtained from mitogen-stimulated cells but not in extracts obtained from quiescent cells (Gotoh et al., 1990a, Fig. 7). Furthermore, they both were commonly observed in cells stimulated with various GFs and phorbol esters (Fig. 8). Because in these kinase detection assays, the obtained cell extracts were boiled in SDS-sample buffer immediately after cell extraction, the lower molecular weight species (41 K) may not be a degradation product of the higher molecular weight species (43 K). Considereing the fact that the apparent molecular weight of the native form of MAP kinase is estimated to be 40-45K in gel



FIG. 7 Two MAP kinases (43 K and 41 K) are activated in mitogen-stimulated mammalian fibroblastic cells. Extracts obtained from unstimulated (lane a) and EGF-stimulated (10 nM EGF, 5 minutes, lane b) 3Y1 cells were analyzed by the kinase detection assay within MBPcontaining gels as described by Gotoh et a / . (1990a). Lane c shows the purified mammalian MAP kinases (43 K and 41 K) detected by the kinases detection assay within gels.

FIG. 8 Two MAP kinases (43 K and 41 K)are activated commonly by a variety of growth factors. Rat 3Y1 cells were treated with EGF (10 nM), TPA (100 ng/ml), IGF-I (I0 nM), PDGF (10 ng/ml), or FGF (50 ng/ml). Extracts were subjected to the kinase detection assay within polyacrylamide gels containing MBP.



filtration chromatography, these results suggest that each of these two polypeptides may be a monomeric protein having a kinase activity and can be identified as MAP kinase. Therefore, at least two similar but distinguishable MAP kinases are present in mammalian cells. Krebs and coworkers also described activation of two MAP kinases, MBP/MAP 2 kinases 1 and 2, in EGF-stimulated Swiss 3T3 cells, the apparent molecular weights of which were 41 K and 44 K, respectively (Ahn et al., 1990, 1991). These two kinases may correspond to our two MAP kinases. Most recently, Cobb and co-workers (Boulton et al., 1991) reported purification of rat insulin-activaed MAP kinase (called ERK- I), the apparent molecular weight of which was 43 K, which may correspond to the higher molecular weight species of MAP kinases. They reported the partial amino acid sequence of ERK-1 based on its cDNA cloning (Boulton et al., 1990). Sturgill and co-workers first presented evidence that their MAP kinase (pp42), which may correspond to the lower molecular weight species of MAP kinases, is phosphorylated on tyrosine and threonine residues and that the kinase is only active when both residues are phosphorylated (Anderson et al., 1990).Our recent data (unpublished) obtained with several anti-MAP kinase antibodies have shown that both 41 K and43 K MAP kinases become phosphorylated on tyrosine and threonine residues upon treatment with a variety of mitogenic factors and become active as a kinase (Fig. 8). The detailed biochemical and enzymatic characterization of Xenopus M phase-activated MAP kinase and mammalian mitogen-activated MAP kinase has revealed that the Xenopus kinase is very similar to the mammalian kinase in substrate specificity, chromatographic behavior, optimal ionic conditions for kinase acitivity (e.g., 20 mM Mg2+,2 mM Mn2+, pH -8), sensitivity to various inhibitors, apparent molecular weight on gel filtration, and others (Gotoh e f al., 1991a). These properties together with tyrosine phosphorylation accompanying the kinase activation can distinguish MAP kinase from other known serinekhreonine kinases. It has recently been revealed that both mammalian MAP kinase and starfish MAP kinase phosphorylate specifically The9’ residue (-Thr-Pro-Arg-Thr9’Pro-Pro-Pro-) of MBP (Erickson el al., 1990; Sanghera et al., 1990a), but the consensus sequence for phosphorylation by MAP kinase still remains to be determined. We carried out molecular cloning of Xenopus MAP kinase by screening a Xenopus cDNA library with two oligonucleotide probes, which were synthesized according to partial amino acid sequences of Xenopus MAP kinase. An obtained clone contained an open reading frame, which encodes a protein consisting of 361 amino acid residues with a calculated molecularmass of41.3 kDa(Gotohetal., 1991b).Thepredictedaminoacid sequence has all the characteristics of serinekhreonine kinases (Hanks et al., 1988). The alignment of Xenopus MAP kinase with ERK-1 (43 K



rat insulin-activated MAP kinase) reveals that both sequences are 84% identical. This extensive sequence similarity between mammalian and amphibian MAP kinases may imply the significance of MAP kinase'in fundamental biological functions. In addition, more than 50% sequence identity exists between MAP kinase and three yeast kinases (Gotoh et al., 1991b; Boulton et al., 1990), two of which are the budding yeast kinases (FUS3 and KSSl) mediating the yeast response to pheromones (Elion et al., 1990; Courchesne et af., 1989) and one of which is the fission yeast kinase (spkl + ) conferring staurosporine resistance when carried on multicopy plasmids and existing upstream of S pombe AP-1 (pap1 ) (Toda et af., 1991). However, it remains to be determined whether these kinases are yeast MAP kinase or not. As MAP kinase activation is always accompanied by its phosphorylation on both tyrosine and serinelthreonine residues either in GOlGl phase or in M phase, elucidation of the mechanisms by which GF receptors or MPF activates MAP kinase is the subject of great interest. The simplest idea may be that in either GOlGl or M phase there exist two kinds of MAP kinase kinases, which phosphorylate tyrosine and serinelthreonine residues of MAP kinase, respectively. However, the recent demonstration of the existence of kinases with specificity for both tyrosine and serinel threonine (Stern et al., 1991; Howell et al., 1991; Ben-David et al., 1991) suggests the possibility that a single kinase may be sufficient for MAP kinase activation. Krebs and co-workers have proposed several other possibilities based on their results that a 60-kDa component that is derived from EGF-stimulated cells and is eluted as a single peak on Mono S chromatography can activate MAP kinase (Ahn et al., 1991). Of their proposed possibilities, a possibility that the MAP kinase activator may not be a kinase at all but may induce changes in MAP kinase that allow autophosphorylation is an attractive hypothesis not previously assumed. In line with this hypothesis, we found that the bacterially expressed Xenopus MAP kinase is capable of undergoing autophosphorylation on tyrosine (unpublished data). Further studies are required for elucidation of the mechanism of MAP kinase activation during GO/Gl and M phases.


C. Possible Functions of MAP Kinase during the Cell Cycle

During the transition from interphase to M phase, there is the dramatic reorganization of the cell, which may be triggered by phosphorylation mediated directly or indirectly by p34cdc2or MPF. We first focused our attention on MT reorganization (Fig. 1) as a possible target event of MAP kinase in M phase. Karsenti and co-workers reported an in uitro system that reflected in uiuo



organization and dynamics of MTs (Verde et al., 1990). Using Xenopus egg extracts and mammalian centrosomes, they showed that MTs nucleated by the added centrosomes in interphase extracts grow faster and are longer at steady state than those in metaphase I1 extracts (Fig. 1). Belmont et al. (1990)revealed that MTs formed in M phase extracts are more dynamic and change into the shrinking phase more frequently than those in interphase extracts. Karsenti and co-workers further showed that the addition of purified MPF to interphase extracts can induce the interphase-M phase transition of MT dynamics in this in uitro cell-free system (Verde el al., 1990). Interestingly, we found that the addition of purified Xenopus MAP kinase to the interphase extracts can induce the same or similar changes in MT dynamics, that is, MTs in interphase extracts change into the M phase type after the addition of MAP kinase (Gotoh et al., 1991a). This effect of MAP kinase was not seen when K252a, which strongly inhibited MAP kinase activity, was added with MAP kinase. This clearly indicates that the MAP kinase-induced transition of MT dynamics is mediated by MAP kinase-catalyzed phosphorylation events. Identification of substrates of MAP kinase in this in vitro system must shed light on the study pursuing in uiuo targets of MAP kinase at M phase. It should be pointed out that these results obtained from the in uitro system of Xenopus egg extracts not only reveal an M phase function of MAP kinase but also provide the important implications concerning regulation of M phase cellular events by MPF. As described in Section V,A, in this in vitro system MAP kinase can be activated by the addition of MPF while no activation of MPF occurs after the addition of MAP kinase. Therefore, the effect of MPF on MT dynamics may be mediated, at least partly, by activation of MAP kinase. Thus, M phase MAP kinase acts as an intermediate between MPF and the interphase-M phase transition of MT organization (Fig. 9). There exists the kinase cascade, which controls M phase events, downstream of MPF. It is of great importance to determine whether other M phase cellular events, such as nuclear envelope breakdown and chromosome condensation, are also mediated by other kinases, including MAP kinase, or are directly regulated by phosphorylations catalyzed by MPF alone. MAP kinase was originally found as a kinase that is activated in GO/G1 phase by various mitogenic factors (see Section IV,B) and has been assumed to have a critical role in mitogenic signal transduction, constituting


- -

MAP Kinase Activator


MAP Kinase * InterphaseIM Transition Microtubulephase Dynamics

FIG. 9 A simple model illustrating the role of MAP kinase in M phase.




part of a network of protein kinases and phosphorylation reactions downstream of GF receptor-associated tyrosine kinase. It remains to be studied whether or not MAP kinase activation is essential to the GF-signaling pathway. Moreover, because we do not know the targets of MAP kinase in GO/Gl phase except for S6 kinase I1 homolog, identification of substrate proteins is an urgent necessity to reveal GO/Gl functions of MAP kinase.

VI. Conclusions

We have dealt with cytoskeleton, actin filaments, MTs, and a newly found kinase, MAP kinase, with reference to mitogenic signal transduction. Actin filaments and cytoplasmic MTs are dynamic by nature and the turnover or exchange of the subunit proteins in polymers occurs rapidly in every phase of the cell cycle. Nevertheless, integrity of these cytoskeletal structures in nonproliferating state of cells appears to function to keep the cells quiescent and tend to lose as the cells enter proliferating state. Thus, relatively static structures of actin filaments, that is, stress fibers, disappear and relatively dynamic structures, ruffling membranes, appear upon mitogenic stimulation of the quiescent cells. Positive roles of this actin reorganization in the transmission of growth signals remain to be shown. In the quiescent cells, cytoplasmic MTs seem to act as a negative regulator of GF receptors by preventing them from ligand-independent activation. The cytoplasmic MT networks may ensure that the quiescent cells do not begin to proliferate unless GFs are given. The precise mechanism by which MTs regulate intracellular GF receptor dynamics should be clarified. MAP kinase has been found as a serinehhreonine kinase that is commonly activated in quiescent (GO/Gl) cells by a variety of mitogenic factors, including peptide GFs and tumor promoters, and later shown to be activated also in M phase. In both GO/GI and M phases, MAP kinase activation is accompanied by its tyrosine phosphorylation. It is noteworthy that MAP kinase activation in M phase is under the control of MPF (p34cdc2/cyclinB complex). Because ~34'~''is a master control element in the GI-S transition as well as in the G2-M transition of yeast cell cycles and because p34cdc2or related molecules could be also involved in the G1-S transition of mammalian cell cycles, it is of great interest to see whether MAP kinase is activated in S phase under the control of p34cdC2 or related molecules. In the cell cycles of Xenopus oocytes and eggs, only M phase activation of MAP kinase has been observed, and the S phase (interphase) activation cannot be observed. But this does not rule out the possibility of the S phase activation of MAP kinase in somatic cell cycles



because there may be the essential difference between somatic cell cycles and early embryonic cell cycles in the role of ~ 3 4 as' reflected ~ ~ ~ in the existence of G1 and G2 phases in the former cell cycles. MAP kinase may function not only as a mitogenic signal transducer but also as a cell cycle regulator. Finally, it might be incidental, but interesting, that MAP kinase, which has been found by its ability to phosphorylate MAP 2 in uitro, plays a role in the interphase-M phase transition of MT dynamics presumably by phosphorylating MAPS in uiuo at M phase.

Acknowledgments The authors thank many collaborators of our laboratory for unpublished data and drawings used here.

References Ahn, N. G., Weiel, J. E., Chan, C. P., and Krebs, E. G. (1990). J . Biol. Chem. 265, 11487-1 1494. Ahn, N. G . , Seger, R.. Bratlien, R. L.,Diltz, C . D., Tonks, N. K., and Krebs, E. G. (1991). J . Biol. Chem. 266, 4220-4227. Akiyama, T., Nishida, E., Ishida, J., Saji, N., Ogawara, H., Hoshi, M., Miyata, Y., and Sakai, H.(1986). J. Biol. Chem. 261, 15648-15651. Anderson, N . G., Maller, J. L., Tonks, N. K., and Sturgill, T. W. (1990). Nature (London) 343, 651-653. Ballou, L . M.. Luther, H . , and Thomas, G. (1991). Nature (London) 349, 348-350. Bar-Sagi, D., and Feramisco, J. R. (1986). Science 233, 1061-1068. Belmont, L. D., Hyman, A. A., Sawin, K. E., and Mitchison, T. J. (1990). Cell (Cambridge, Mass.) 62, 579-589. Ben-David. Y . , Litwin, K., Tannock, L., Bernstin, A., and Pawson, T. (1991). EMBO J . 10, 317-325. Ben-Ze'ev, A., Farmer, S. R., and Penman. S. (1979). Cell(Cambridge, Mass.) 17,319-325. Bockus, U . , and Stile, C. D. (1984). Exp. Cell Res. 153, 186-197. Boulton, T . G., Yancopoulos, G. D., Gregory, J. S.. Slaughter, C., Moomaw, C., Hus, J., and Cobb, M. H. (1990). Science 249, 64-67. Boulton, T. G . , Gregory. J. S., and Cobb, M. H . (1991). Biochemistry 30, 278-286. Brunk, U.,Schellens, J., and Westermark, B. (1976). Exp. Cell Res. 103, 295-302. Chen, W. S., Lazer, C. S., Poenie. M., Tsien, R. Y.,Gill, G. N., and Rosenfeld, M . G . (1987). Nature (London) 328, 820-823. Chinkers, M., McKanna, J. A., and Cohen, S. (1979). J . Biol. Chem. 83, 260-265. Chinkers, M.,McKanna, J. A., and Cohen, S. (1981). J. Biol. Chem. 88, 422-429. Cicirelli, M. F., Pelech, S. L., and Krebs, E. G. (1988). J . Biol. Chem. 263, 2009-2019. Cleveland, D. W., Lopata, M. A., Sherline, P., and Kirschner, M. W. (1981). Cell (Cambridge, Mass.) 25, 537-546. Connolly, J. L., Green, S. A., and Greene, L . A. (1984). J. Cell Biol. 98, 457-465. Cooper, J. A. (1989). Mol. Cell. Biol. 9, 3143-3147.



Cooper, J. A., Bowen-Pope, D. F., Raines, E., Ross, R., and Hunter, T. (1982). Cell (Cambridge, Muss.) 31, 263-273. Cooper, J. A., Sefton, B. M., and Hunter, T. (1984). Mol. Cell. Biol. 4, 30-37. Courchesne, W. E., Kunisawa, R., and Thorner, J. (1989). Cell (Cambridge, Muss.) 58, 1107-1119. Crossin, K. L., and Carney, D. H. (1981a). Cell (Cambridge, Muss.) 23, 61-71. Crossin, K. L., and Carney, D. H. (1981b). Cell (Cambridge. Muss.) 27, 341-350. Davis, R. J. (1988). J. Biol. Chem. 263, 9462-9469. Elion, E. A., Grisafi, P. L., and Fink, G. R. (1990). Cell (Cambridge, Mass.) 60,649-664. Ely, C. M., Oddie, K. M., Litz, J. S., Rossomando, A. J., Kanner, S. B., Sturgill, T. W., and Parsons, S. J. (1990). J . Biol. Chem. 110, 731-742. Erickson, A. K., Payne, D. M., Martino, P. A., Rossomando, A. J., Shabanowitz,J., Weber, M., Hunt, D. F., and Sturgill, T. W. (1990). J. Biol. Chem. 265, 19728-19735. Gibbs, E. M., Lienhard, G. E., Appleman, J. R., Lane, M. D., and Frost, S. C. (1986). J . Biol. Chem. 261, 3944-395 1. Glenney, J. R., Chen. W. S., Lazer, C. S., Walton, G. M., Zokas, L. M., Rosenfeld, M. G., and Gill, G. N. (1988). Cell (Cambridge, Muss.) 52, 675-684. Goshima, K., Masuda, A., and Owaribe, K. (1984). J. Cell Biol. 98, 801-809. Gotoh, Y., Nishida, E., Yamashita, T., Hoshi, M., Kawakami, M., and Sakai, H. (1990a). Eur. J . Biochem. 193,661-669. Gotoh, Y., Nishida, E., and Sakai, H. (1990b). Eur. J . Biochem. 193, 671-674. Gotoh, Y., Nishida, E., Matsuda, S., Shiina. N., Kosako, H., Shiokawa, K., Akiyama, T., Ohta, K., and Sakai, H. (1991a). Nature (London) 349, 251-254. Gotoh, Y., Moriyama, K., Matsuda, S., Okumura, E., Kishimoto, T., Kawasaki, H., Suzuki, K., Yahara, I., Sakai. H., and Nishida, E. (1991b). EMBO J. 10, 2661-2668. Greenberg, M. E., and Ziff, E. B. (1984). Nature (London) 311,433-438. Gregory, J. S., Boulton, T. G., Sang, B.-C., and Cobb, M. H. (1989). J. Biol. Chem. 264, 18397- 18401. Haigler, H. T., McKanna, J. A., and Cohen, S. (1979). J. Cell Biol. 83, 82-90. Hanks, S. K., Quinn, A. M., and Hunter, T. (1988). Science 241, 42-52. Hidaka, H., Inagaki, M., Kawamato, S., and Sasaki, Y. (1984). Biochemistry23,5036-5041. Honegger, A. M., Dull, T. J., Felder, S., Van Obberghen, E., Bellot, F., Szapary, D., Schmidt, A., Ullrich, A., and Schlessinger, J. (1987). Cell(Cambridge,Mass.)51, 199-209. Hoshi, M., Nishida, E., and Sakai, H. (1988). J. Biol. Chem. 263, 5396-5401. Hoshi, M., Nishida, E., and Sakai, H. (1989). Eur. J . Biochem. 184, 477-486. Howell, B. W., Afar, D. E. H., Lew, J., Douville, E. M. J., Icely, P. L. E., Gray, D. A., and Bell, J. C. (1991). Mol. Cell. Biol. 11, 568-572. Izumi, T., Saeki, Y., Akanuma, Y., Takaku, F., and Kasuga, M. (1988). J. Biol. Chem. 263, 10386- 10393. Kadowaki, T., Fujita-Yamaguchi, Y., Nishida, E., Takaku, F., Akiyama, T., Kathuria, S., Akanuma, Y., and Kasuga, M. (1985). J . Biol. Chem. 260,4016-4020. Kadowaki, T., Koyasu, S., Nishida, E., Sakai, H.. Takaku, F., Yahara, I., and Kasuga, M. (1986). J . Biol. Chem. 261, 16141-16147. Kadowaki, T., Koyasu, S., Nishida, E.,Tobe, K., Izumi, T.,Takaku, F., Sakai, H., Yahara, I., and Kasuga, M. (1987). J . Biol. Chem. 262, 7342-7350. Kameshita, I., and Fujisawa, H. (1989). Anal. Biochem. 183, 139-143. Karsenti, E., Brave, R., and Kirschner, M. W. (1987). Deu. Biol. 119, 442-453. Kawakami, M., Nishida, E., Tobe, K., Hoshi, M., Kadowaki, T., Kasuga, M., and Sakai, H. (1991). Exp. Cell Res. 193, 120-126. Koyasu, S., Kadowaki, T., Nishida, E., Tobe, K., Abe, E., Kasuga, M., Sakai, H., and Yahara, I. (1988). Exp. Cell Res. 176, 107-116.



Maller, J. L. (1990). Biochemistry 29, 3157-3166. Maller, J. L., Wu, M., and Gerhart, J. C. (1977). Deu. Biol. 99, 489-501. Matteoni, R.. and Kreis, T. E. (1987). J. Cell Biol. 105, 1253-1265. Mellstrom, K.. Hoglund. A.-S., Nister, M.. Heldin, C. H . , Westermark, B., and Lindberg, U. (1983). J. Muscle Res. Cell Motil. 4, 589-609. Miyasaka, T., Chao, M., Sherline, P., and Saltiel. A. (1990). J. Biol. Chem. 265,4730-4735. Miyata, Y., Nishida. E., and Sakai, H. (1988a). Exp. Cell Res. 175, 286-297. Miyata, Y., Hoshi, M., Koyasu, S., Kadowaki, T., Kasuga, M., Yahara, I., Nishida. E., and Sakai, H. (1988b). Exp. Cell Res. 178, 73-83. Miyata, Y., Nishida, E., Koyasu, S., Yahara, I., and Sakai, H. (1989a). Exp. Cell Res. 181, 454-462. Miyata, Y., Nishida, E., Koyasu, S., Yahara, I., and Sakai, H . (1989b). J . Biol. Chem. 264, 15565-15568. Moolenaar, W. H., Bierman. A. J., Tilly, B. C., Verlaan, I., Defize, L . H . K., Honegger, A., Ullrich, A., and Schlessinger, J. (1988). EMBO J . 7, 707-710. Moskalewski, S . , Thyberg, J., Lohmander. S., and Friberg, U. (1975). Exp. Cell Res. 95, 440-454. Murray, A. W.. and Kirschner, M. W. (1989). Science 246, 614-621. Nel, A.. Hanekom, C., Rheeder, A., Williams, K., Pollack, S., Katz, R.,and Landreth, G. (1990). J. Immunol. 144, 2683-2689. Nishida, E., Hoshi, M., Miyata, Y., Sakai. H.. Kadowaki, T., Kasuga, M., Saijo, S., Ogawara, H., and Akiyama, T. (1987). J. B i d . Chem. 262, 16200-16204. Nishida, E., Tobe, K.. Kadowaki, T., Kasuga, M., Sato, C., and Sakai, H. (1988). Cell Struct. Funct. 13, 417-423. Nishizuka, Y. (1984). Nature (London) 308, 693-698. Nishizuka, Y. (1986). Science 233, 305-312. Nurse, P. (1990). Nature (London) 344,503-508. Otto, A. M. (1987). Int. Rev. Cytol. 109, 113-158. Pardee, A. B. (1989). Science 246, 603-608. Pelech, S. L., Tombes, R. M., Meijer, L., and Krebs, E. G. (1988). Deu. Biol. 130, 28-36. Pollack, R., Osborn, M., and Weber, K. (1975). Proc. Natl. Acad. Sci. U . S . A .72,994-998. Ray, L. B., and Sturgill, T . W. (1987). Proc. Natl. Acad. Sci. U . S . A . 84, 1502-1506. Ray, L. B., and Sturgill, T. W. (1988a). J. Biol. Chem. 263, 12721-12727. Ray, L. B., and Sturgill, T. W. (1988b). Proc. Natl. Acad. Sci. U.S.A. 85, 3753-3757. Rossomando, A., Payne, D. M., Weber, M., and Sturgill, T. W. (1989). Proc. Natl. Acad. Sci. U.S.A.86, 6940-6943. Sanghera, J. S, Aebersold, R., Morrison, H. D., Bures, E. J., and Pelech, S. L. (1990a). FEES Lett. 273, 223-226. Sanghera, J. S., Paddon, H. B., Bader, S. A., and Pelech, S. L. (1990b). J . Biol. Chem. 265, 52-57. Sato, C.. Nishizawa, K., Nakayama, T., and Kobayashi, T. (1985). J. CellBiol. 100,748-753. Sato, C., Nishizawa, K., Nakayama, T., Nose, K., Takasaki, Y ., Hirose, S., and Nakamura, H. (1986a). Proc. Natl. Acad. Sci. U.S.A. 83, 7287-7291. Sato, C., Nishizawa, K., Nakayama, T., Hirai, R.,and Nakamura, H. (1986b). Exp. Cell Res. 167, 281-287. Schlessinger, J. (1986). J. Cell B i d . 103, 2067-2072. Schlessinger, J. (1988). Biochemistry 27, 31 19-3123. Shinohara, Y., Nishida, E., and Sakai, H. (1989). Eur. J. Biochem. 183, 275-280. Shinohara-Gotoh, Y., Nishida, E., Hoshi, M., and Sakai, H. (1991). Exp. Cell Res. 193, 161- 166. Sloboda, R. D., Rudolph, S. A., Rosenbaum, J. L.. and Greengard, P. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 177-181.



Stern, D. F., Zheng, P., Beidler, D. R., and Zerillo, C. (1991). Mol. Cell. Biol. 11,987-1001. Sturgill, T. W., Ray, L. B., Erickson, E., and Maller, J. L. (1988). Nature (London)334, 715-718. Swanson, J. A., Bushnell, A., and Silverstein, S . C. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 1921-1925. Terasaki, M., Chen, L. B., and Fujiwara, K. (1986). J . CeNBiol. 103, 1557-1568. Toda, T., Shimanuki, M., and Yanagida, M. (1991). Genes Deu. 5, 60-73. Tsuyama, S., Bramblett, G. T., Huang, K.-P., and Flavin, M. (1986). J. Biol. Chern. 261, 41 10-4 1 16. Vallee, R. B. (1980). Proc. Natl. Acad. Sci. U.S.A. 77,3206-3210. Vasiliev, J. M., Gelfand, I. M., and Gueltstein, V. I. (1971). Proc. Natl. Acad. Sci. U.S.A. 68,977-979. Verde, F., Labbe, J.-C., Doree, M., and Karsenti, E. (1990). Nature (London)343,233-238. Wiley, H. S., and Cunningham, D. D. (1982). J . Cell Biochern. 19, 383-394. Willingham, M. C., and Pastan, I. (1982). J. Cell Biol. 94, 207-212. Yamamoto, H., Fukunaga, K., Goto, S., Tanaka, E., and Miyamoto, E. (1985). J. Neurochern. 44,759-767. Yarden, Y., and Ullrich, A. (1988). Biochemistry 27, 3113-3119. Yen, T. J., Machlin, P. S.,and Cleveland, D. W. (1988). Nature (London) 334, 580-585.

Mitogen-activated protein kinase and cytoskeleton in mitogenic signal transduction.

Mitogen-Activated Protein Kinase and Cytoskeleton in Mitogenic Signal Transduction Eisuke Nishida and Yukiko Gotoh Department of Biophysics and Bioche...
2MB Sizes 0 Downloads 0 Views