Phosphorylation and regulation of choline kinase from Saccharomyces cerevisiae by protein kinase A

The CKI1-encoded choline kinase (ATP:choline phosphotransferase, EC 2.7.1.32) from Saccharomyces cerevisiae was phosphorylated in vivo on multiple serine residues. Activation of protein kinase A activity in vivo resulted in a transient increase in the phosphorylation of choline kinase. This phosphorylation was accompanied by a stimulation in choline kinase activity. In vitro, protein kinase A phosphorylated choline kinase on a serine residue with a stoichiometry (0.44 mol of phosphate/mol of choline kinase) consistent with one phosphorylation site/choline kinase subunit. The major phosphopeptide derived from the enzyme phosphorylated in vitro by protein kinase A was common to one of the major phosphopeptides derived from the enzyme phosphorylated in vivo. Protein kinase A activity was dose- and time-dependent and dependent on the concentrations of ATP (Km 2.1 microM) and choline kinase (Km 0.12 microM). Phosphorylation of choline kinase with protein kinase A resulted in a stimulation (1.9-fold) in choline kinase activity whereas alkaline phosphatase treatment of choline kinase resulted in a 60% decrease in choline kinase activity. The mechanism of the protein kinase A-mediated stimulation in choline kinase activity involved an increase in the apparent Vmax values with respect to ATP (2.6-fold) and choline (2.7-fold). Overall, the results reported here were consistent with the conclusion that choline kinase was regulated by protein kinase A phosphorylation.     

Troglitazone inhibits angiotensin II-induced extracellular signal-regulated kinase 1/2 nuclear translocation and activation in vascular smooth muscle cells.

The thiazolidinedione troglitazone inhibits angiotensin II-induced extracellular signal-regulated kinase 1/2 mitogen-activated protein kinase activity in vascular smooth muscle cells. Activation of extracellular signal-regulated kinase 1/2 by angiotensin II is a multistep process involving both its phosphorylation by mitogen-activated protein kinase extracellular signal-regulated kinase kinase in the cytoplasm and a subsequent translocation to the nucleus. The cytoplasmic activation of extracellular signal-regulated kinase 1/2 in vascular smooth muscle cells proceeds through the protein kinase Czeta --> mitogen-activated protein kinase extracellular signal-regulated kinase kinase --> extracellular signal-regulated kinase pathway. Troglitazone did not affect the angiotensin II-induced activation of protein kinase Czeta or its downstream signaling kinases extracellular signal-regulated kinase 1/2 in the cytosol. In contrast, angiotensin II-induced activation of protein kinase Czeta and extracellular signal-regulated kinase 1/2 in the nucleus were both inhibited by troglitazone. Nuclear translocation of extracellular signal-regulated kinase 1/2 induced by angiotensin II was completely blocked by troglitazone. Protein kinase Czeta, however, did not translocate upon angiotensin II stimulation. Troglitazone, therefore, inhibits both angiotensin II-induced nuclear translocation of extracellular signal-regulated kinase 1/2 and the nuclear activity of its upstream signaling kinase protein kinase Czeta. Since extracellular signal-regulated kinase 1/2 nuclear translocation may be a critical signaling step for multiple growth factors that stimulate vascular smooth muscle cells proliferation and migration, troglitazone may provide a new therapeutical approach for the prevention and treatment of atherosclerosis and restenosis.     

Evidence that phosphorylase kinase exhibits phosphatidylinositol kinase activity

Phosphorylase kinase phosphorylates the pure phospholipid phosphatidylinositol. Furthermore, it catalyzed phosphatidylinositol 4-phosphate formation using as substrate phosphatidylinositol that is associated with an isolated trypsin-treated Ca2+-transport adenosinetriphosphatase (ATPase) preparation from skeletal muscle sarcoplasmic reticulum. On this basis a fast and easy assay was developed that allows one to follow the phosphatidylinositol kinase activity during a standard phosphorylase kinase preparation. Both activities are enriched in parallel approximately to the same degree. Neither chromatography on DEAE-cellulose nor that on hydroxyapatite in the presence of 1 M KCl separates phosphatidylinositol kinase from phosphorylase kinase. The presence of a lipid kinase, phosphatidylinositol kinase, in phosphorylase kinase is not a general phenomenon; diacylglycerol kinase can be easily separated from phosphorylase kinase. Polyclonal anti-phosphorylase kinase antibodies as well as a monoclonal antibody directed specifically against the alpha subunit of phosphorylase kinase immunoprecipitate both phosphorylase kinase and phosphatidylinositol kinase.     

Abscisic Acid Activates a 48-Kilodalton Protein Kinase in Guard Cell Protoplasts

A 49- and a 46-kD Ca2+-independent protein kinase and a 53-kD Ca2+-dependent protein kinase were detected in Vicia faba guard cell protoplasts (GCPs) by an in-gel protein kinase assay using myelin basic protein as a substrate. A 48-kD protein kinase designated as abscisic acid (ABA)-responsive protein kinase (ABR kinase) appeared when GCPs were treated with ABA. The activation of ABR kinase was suppressed by the protein kinase inhibitor staurosporine, indicating that a putative activator protein kinase phosphorylates and activates ABR kinase. The treatment of GCPs with 1,2-bis(o-aminophenoxy)ethan-N,N,N',N'-tetraacetic acid, a calcium chelator, suppressed the activation of ABR kinase, suggesting that an influx of extracellular Ca2+ is required for the activation. Staurosporine and K-252a inhibited both the activity of ABR kinase and the stomatal closure induced by ABA treatment of V. faba epidermal peels. These results suggest that ABR kinase and its activator kinase may consist of a protein kinase cascade in a signal transduction pathway linking ABA perception to stomatal closure. The mobility of the 53-kD Ca2+-dependent protein kinase in sodium dodecyl sulfate-polyacrylamide gel was shifted upon Ca2+ binding to the enzyme, thus exhibiting the characteristics of a Ca2+-dependent or calmodulin-like domain protein kinase. This kinase may be the activator of ABR kinase.     

Light activates a 46-kDa MAP kinase-like protein kinase in soybean cell culture

Light induced rapid and transient activation of a 46-kDa protein kinase in soybean photomixotrophic cell culture. This kinase was designated as LAP kinase (light signal-activated protein kinase). Activation of LAP kinase in response to light was associated with tyrosine phosphorylation of the kinase, and treatment of the kinase with protein tyrosine phosphatase abolished its activity. The LAP kinase efficiently phosphorylated myelin basic protein and histone, but did not phosphorylate casein. Phospho-amino acid analysis indicated that the LAP kinase was a serine/threonine protein kinase. These results indicated that the LAP kinase is related to the MAP kinase family of protein kinases.     

Ca(2+)- and protein kinase C-dependent stimulation of mitogen-activated protein kinase in detergent-skinned vascular smooth muscle

Protein kinase C and mitogen-activated protein (MAP) kinase are expressed in all smooth muscle cells and believed to be important in several physiologically relevant properties of this muscle. Our goal was to determine if protein kinase C and MAP kinase are activated by a simple increase in cellular Ca(2+) and to determine if protein kinase C is an upstream activator of MAP kinase. These studies were performed in the Triton X-100 detergent-skinned preparation of the swine carotid artery, which allows control of the intracellular environment without influence from membrane or receptor-mediated modulation. The p42 and p44 isoforms of MAP kinase were activated in a concentration-dependent fashion by an increase in Ca2+. This was shown by in-the-gel kinase assay and direct measurement of MAP kinase phosphotransferase activity. Protein kinase C was also activated by an increase in Ca2+, as shown by a novel assay that measures total active protein kinase C in the tissue. Inhibition of protein kinase C activity completely abolished MAP kinase activity. Additionally, inhibition of Ca2+/calmodulin-dependent protein kinase II (CaM kinase II) also abolished MAP kinase activity. Using intact swine carotid arteries, we showed p42 and p44 MAP kinase to be activated by both histamine and phorbol dibutyrate, but only the p42 isoform was calcium-sensitive. Our results suggest that a Ca(2+)-dependent isoform of protein kinase C and CaM kinase II are upstream activators of MAP kinase in the swine carotid artery.     

Activation of the MAP kinase pathway by the protein kinase raf.

Both MAP kinases and the protein kinase p74raf-1 are activated by many growth factors in a c-ras-dependent manner and by oncogenic p21ras. We were therefore interested in determining the relationship between MAP kinases and raf. The MAP kinase ERK2 is activated by expression of oncogenically activated raf, independently of cellular ras. Overexpressed p74raf-1 potentiates activation of ERK2 by EGF and TPA. MAP kinase kinase inactivated by phosphatase 2A treatment is phosphorylated and reactivated by incubation with p74raf-1 immunoprecipitated from phorbol ester-treated cells. We conclude that raf protein kinase is upstream of MAP kinases and is either a MAP kinase kinase kinase or a MAP kinase kinase kinase kinase.     

Biosynthesis and posttranslational modifications of protein kinase C in human breast cancer cells

Several forms of protein kinase C with molecular masses of 74-, 77-, and 80-kDa were detected in subcellular fractions of human breast cancer MDA-MB-231 cells which express the alpha-type protein kinase C. Several lines of evidence indicated that the 74-kDa is the precursor of the 77- and 80-kDa protein kinase C forms. (i) Pulse-labeling experiments revealed that protein kinase C is synthesized on membranes as a 74-kDa protein that can be chased into the 77- and the 80-kDa protein kinase C forms. (ii) The primary translation product of protein kinase C displayed an apparent molecular size of 74-kDa as determined by in vitro translation of poly(A)+ RNA from MDA-MB-231 cells. (iii) Incubation with serine/threonine-specific protein phosphatases (potato acid phosphatase and phosphatase 1 or 2A) resulted in the complete dephosphorylation of the 77-kDa to the 74-kDa protein kinase C form. Protein kinase C appears to be synthesized in membranes as an unphosphorylated and presumably inactive 74-kDa form that is converted into the active 77- and 80-kDa protein kinase C by post-translational modification involving at least two phosphorylation steps. The first phosphorylation is probably achieved by a specific, yet unidentified, "protein kinase C kinase" since the 74-kDa protein kinase C species did not undergo autophosphorylation and was neither a substrate for the purified protein kinase C, S6 kinase, phosphorylase kinase, casein kinase II, nor for the catalytic subunit of cAMP-dependent protein kinase. Except for phosphorylase kinase and the catalytic subunit of the cAMP-dependent protein kinase, phosphorylation of the 77-kDa protein kinase C form with purified protein kinase C (autophosphorylation), S6 kinase or casein kinase II shifted the molecular mass of the 77-kDa protein kinase C to 80-kDa. Prolonged exposure of MDA-MB-231 cells to phorbol 12-myristate 13-acetate not only leads to a complete down-regulation of protein kinase C activity but also to an accumulation of 74-kDa protein kinase C due to a retarded conversion of the 74-kDa into the 77- and 80-kDa protein kinase C forms in these cells. Our data indicate that tumor promoters additionally interfere with the posttranslational processing that converts the 74-kDa protein kinase C precursor into the 77- and 80-kDa forms of the enzyme.     

MAP kinase activator from insulin-stimulated skeletal muscle is a protein threonine/tyrosine kinase.

A 'MAP kinase activator' was purified several thousand-fold from insulin-stimulated rabbit skeletal muscle, which resembled the 'activator' from nerve growth factor-stimulated PC12 cells in that it could be inactivated by incubation with protein phosphatase 2A, but not by protein tyrosine phosphatases and its apparent molecular mass was 45-50 kDa. In the presence of MgATP, 'MAP kinase activator' converted the normal 'wild-type' 42 kDa MAP kinase from an inactive dephosphorylated form to the fully active diphosphorylated species. Phosphorylation occurred on the same threonine and tyrosine residues which are phosphorylated in vivo in response to growth factors or phorbol esters. A mutant MAP kinase produced by changing a lysine at the active centre to arginine was phosphorylated in an identical manner by the 'MAP kinase activator', but no activity was generated. The results demonstrate that 'MAP kinase activator' is a protein kinase (MAP kinase kinase) and not a protein that stimulates the autophosphorylation of MAP kinase. MAP kinase kinase is the first established example of a protein kinase that can phosphorylate an exogenous protein on threonine as well as tyrosine residues.     

Phosphorylation of caldesmon

The phosphorylation of caldesmon was studied to determine if kinase activity reflected either an endogenous kinase or caldesmon itself. Titration of kinase activity with calmodulin yielded maximum activity at substoichiometric ratios of calmodulin/caldesmon. The sites of phosphorylation on caldesmon for calcium/calmodulin-dependent protein kinase II and endogenous kinase were the same, but distinct from protein kinase C sites. Phosphorylation in the presence of Ca2+ and calmodulin resulted in a subsequent increase of endogenous kinase activity in the absence of Ca2+. These results suggest that caldesmon is not a protein kinase and that kinase activity in caldesmon preparations is due to calcium/calmodulin-dependent protein kinase II.     

Effect of phorbol esters on cytosolic protein kinase C content and activity in the human monoblastoid U937 cell

12-O-Tetradecanoylphorbol-13-acetate (TPA) stimulates the human monoblastoid U937 cell to differentiate into a mature monocyte/macrophage-like cell. Since TPA may produce cellular responses by activating protein kinase C, the effects of TPA on kinase activity in the U937 cell were investigated. Brief exposures (less than or equal to 60 min) to TPA dramatically diminished protein kinase C-dependent phosphorylation of histone and endogenous substrates. However, using a peptide substrate corresponding to residues 720-737 of protein kinase C-epsilon, Ca2(+)-, phospholipid-, and diacylglycerol-dependent kinase activity was reduced only modestly after exposure to TPA. This phospholipid-dependent kinase activity coeluted on DEAE chromatography with protein kinase C. Examination of cytosolic protein kinase C content by Western blot analysis demonstrated a moderate decline in kinase content after TPA treatment. The decline was due primarily to loss of an 80-kDa species with preservation of a 76-kDa protein. The immunoreactive 76-kDa protein observed after TPA treatment comigrated on DEAE chromatography with the kinase activity phosphorylating the protein kinase C-epsilon peptide and had an elution profile similar to protein kinase C derived from untreated cells. Using antisera recognizing the catalytic and regulatory domains of the kinase, no evidence for proteolytic degradation of protein kinase C was observed. Although incubation of extracts from vehicle and TPA-treated cells inhibited the activity of partially purified protein kinase C, the degree of inhibition was similar in the two extracts. These findings suggest that TPA markedly diminishes protein kinase C-dependent phosphorylation of histone and endogenous substrates in part by altering kinase substrate specificity. These observations provide evidence for a novel post-translational process that can modulate protein kinase C-dependent phosphorylation.     

The reverse in-gel kinase assay to profile physiological kinase substrates

Elucidating kinase-substrate relationships is critical for understanding how phosphorylation affects signal transduction and regulatory cascades. Using the alpha catalytic subunit of protein kinase CK2 (CK2alpha) as a paradigm, we developed an in-gel method to facilitate identification of physiologic kinase substrates. In this approach, the roles of kinase and substrate in a classic in-gel kinase assay are reversed. In the reverse in-gel kinase assay (RIKA), a kinase is copolymerized in a denaturing polyacrylamide gel used to resolve a tissue or cell protein extract. Restoration of kinase activity and substrate structure followed by an in situ kinase reaction and mass spectrometric analyses results in identification of potential kinase substrates. We demonstrate that this method can be used to profile both known and novel human and mouse substrates of CK2alpha and cAMP-dependent protein kinase (PKA). Using widely available straightforward technology, the RIKA has the potential to facilitate discovery of physiologic kinase substrates in any biological system.     

Influence of choline and ethanolamine administration on choline and ethanolamine phosphorylating activities of mouse liver and kidney.

The administration of ethanolamine to adult male mice resulted in a significant increase in ethanolamine kinase activity in liver and kidney. Similarly, choline administration resulted in a significant increase in choline kinase activity in liver and kidney. The administration of ethanolamine resulted in enhancement of choline kinase activity concomitantly with ethanolamine kinase activity in liver and kidney. The administration of choline, however, did not result in any significant increase in ethanolamine kinase activity in liver or kidney. Cycloheximide administration along with choline-ethanolamine prevented the increase in kinase activity in liver and kidney. The results obtained have been discussed in relation to the regulatory role of choline kinase and ethanolamine kinase by de novo synthesis in response to enhanced substrate concentration, the secondary nature of choline kinase induction on ethanolamine administration, and possible distinction between choline kinase and ethanolamine kinase.     

Phosphorylation of the yeast choline kinase by protein kinase C. Identification of Ser25 and Ser30 as major sites of phosphorylation

The Saccharomyces cerevisiae CKI1-encoded choline kinase catalyzes the committed step in phosphatidylcholine synthesis via the Kennedy pathway. The enzyme is phosphorylated on multiple serine residues, and some of this phosphorylation is mediated by protein kinase A. In this work we examined the hypothesis that choline kinase is also phosphorylated by protein kinase C. Using choline kinase as a substrate, protein kinase C activity was dose- and time-dependent and dependent on the concentrations of choline kinase (K(m) = 27 microg/ml) and ATP (K(m) = 15 microM). This phosphorylation, which occurred on a serine residue, was accompanied by a 1.6-fold stimulation of choline kinase activity. The synthetic peptide SRSSSQRRHS (V5max/K(m) = 17.5 mm(-1) micromol min(-1) mg(-1)) that contains the protein kinase C motif for Ser25 was a substrate for protein kinase C. A Ser25 to Ala (S25A) mutation in choline kinase resulted in a 60% decrease in protein kinase C phosphorylation of the enzyme. Phosphopeptide mapping analysis of the S25A mutant enzyme confirmed that Ser25 was a protein kinase C target site. In vivo the S25A mutation correlated with a decrease (55%) in phosphatidylcholine synthesis via the Kennedy pathway, whereas an S25D phosphorylation site mimic correlated with an increase (44%) in phosphatidylcholine synthesis. Although the S25A (protein kinase C site) mutation did not affect the phosphorylation of choline kinase by protein kinase A, the S30A (protein kinase A site) mutation caused a 46% reduction in enzyme phosphorylation by protein kinase C. A choline kinase synthetic peptide (SQRRHSLTRQ) containing Ser30 was a substrate (V(max)/K(m) = 3.0 mm(-1) micromol min(-1) mg(-1)) for protein kinase C. Comparison of phosphopeptide maps of the wild type and S30A mutant choline kinase enzymes phosphorylated by protein kinase C confirmed that Ser30 was also a target site for protein kinase C.     

MAP kinases: universal multi-purpose signaling tools

MAP (mitogen-activated protein) kinases are serine/threonine protein kinases and mediate intracellular phosphorylation events linking various extracellular signals to different cellular targets. MAP kinase, MAP kinase kinase and MAP kinase kinase kinase are functional protein kinase units that are conserved in several signal transduction pathways in animals and yeasts. Isolation of all three components was also shown in plants and suggests conservation of a protein kinase module in all eukaryotic cells. In plants, MAP kinase modules appear to be involved in ethylene signaling and auxin-induced cell proliferation. Therefore, coupling of different extracellular signals to different physiological responses is mediated by MAP kinase cascades and appears to have evolved from a single prototypical protein kinase module which has been adapted to the specific requirements of different organisms.     

Stimulation by growth hormone of MAP kinase activity in 3T3-F442A fibroblasts.

We have previously demonstrated that growth hormone (GH) promotes an increase in tyrosine kinase activity associated with the GH receptor. To gain insight into the role of GH-dependent tyrosine kinase activity in signaling by GH, we investigated the possibility that GH might stimulate MAP kinase, a serine/threonine/tyrosine kinase thought to be a common element in tyrosine kinase-initiated response cascades. Treatment of 3T3-F442A fibroblasts with 100 ng/ml GH results in a 3-6-fold increase in the ability of cell-free extracts to phosphorylate MAP-2 and myelin basic protein. GH-stimulated kinase activity is unaffected by heparin, H7, or cAMP-dependent protein kinase inhibitor peptide, partially reduced by staurosporin and inhibited by fluoride and calcium ions, indicating that the kinase is not protein kinase C or A, casein kinase, or a calcium/calmodulin-dependent protein kinase. Based on gel permeation chromatography, the molecular mass of the GH-stimulated MAP kinase is approximately kDa. Furthermore, anti-phosphotyrosine antibodies revealed the GH-dependent appearance of two phosphotyrosine-containing proteins in cell-free lysates of GH-treated cells that co-migrate with proteins recognized by anti-MAP kinase antibodies. The GH-dependent increase in MAP kinase activity displays a biphasic time course and is dependent on the concentration of GH applied to the cells. GH-dependent MAP kinase activity, partially purified by Mono-Q chromatography, is inactivated by treatment with alkaline phosphatase. Addition of H7 to the cells prior to the addition of GH has no effect, whereas addition of H8 increases MAP kinase activity in control cells with no effect in GH-treated cells, indicating that protein kinase C is unlikely to be an intermediary in the GH-dependent stimulation of MAP kinase activity. These findings indicate that signaling by GH in 3T3-F443A cells may, at least in part, utilize a kinase cascade similar to those that have been proposed for other membrane receptors with associated tyrosine kinase activity.     

The relationship between calmodulin binding and phosphorylation of smooth muscle myosin kinase by the catalytic subunit of 3':5' cAMP-dependent protein kinase

Smooth muscle myosin light chain kinase, a calmodulin-dependent enzyme, binds 1 mol of calmodulin/mol of kinase in the presence of calcium (Adelstein, R. S., and Klee, C. B. (1981) J. Biol. Chem. 256, in press. This enzyme is a substrate for cAMP-dependent protein kinase whether or not calmodulin is bound. When calmodulin is not bound to myosin kinase, protein kinase incorporates phosphate into two sites in myosin kinase. Under these circumstances, phosphorylation markedly lowers the rate of myosin kinase activity. The decrease in myosin kinase activity is due to a 10-20-fold increase in the amount of calmodulin necessary for 50% activation of kinase activity. The effect of phosphorylation on the activity of myosin kinase can be reversed by dephosphorylation using a purified phosphatase (Pato, M. D., and Adelstein, R. S. (1980) J. Biol. Chem. 255, 6535-6538) isolated from smooth muscle. When calmodulin is bound to myosin kinase, phosphate is incorporated into a single site with no effect on myosin kinase activity. The presence of at least two sites that can be phosphorylated in myosin kinase was confirmed by tryptic digestion of denatured myosin kinase.     

Phosphagen kinase evolution. Expression in echinoderms

Arginine kinase and creatine kinase that catalyze the transfer of a phosphate group between ATP and arginine and creatine, respectively, play an important role in cellular energetics. In contrast to most animals which exhibit a single phosphagen kinase activity (creatine kinase in chordates and arginine kinase in protostomians), echinoderms exhibit both arginine kinase and creatine kinase activities, sometimes in the same tissue. In contrast to chordates in which creatine kinases are dimers (consisting of two subunits of 40 kDa) and protostomians in which arginine kinases are usually monomers (40 kDa), echinoids contain specific phosphagen kinases: a dimeric arginine kinase (consisting of two subunits of 42 kDa) in eggs and a monomeric creatine kinase (145 kDa) in sperm. We have examined echinoderms from the five existing classes (echinoids, asteroids, ophiuroids, holothurians and crinoids) for the expression of these specific phosphagen kinases in different tissues. Gel filtration was used to determine the molecular masses of the native enzymes. Antibodies specific for arginine kinase or for creatine kinase were used to characterize the subunit composition of arginine kinase and creatine kinase after SDS/PAGE and transfer. In all echinoderms analyzed, arginine kinase always occurred as an enzyme of about 81 kDa consisting of two subunits of 42 kDa and creatine kinase as a monomeric enzyme of 140-155 kDa. The occurrence in echinoderms of both phosphagen kinases with distinct specificities and specific molecular structures is discussed from both a developmental and evolutionary point of view.     

Salicylic acid activates a 48-kD MAP kinase in tobacco.

The involvement of phosphorylation/dephosphorylation in the salicylic acid (SA) signal transduction pathway leading to pathogenesis-related gene induction has previously been demonstrated using kinase and phosphatase inhibitors. Here, we show that in tobacco suspension cells, SA induced a rapid and transient activation of a 48-kD kinase that uses myelin basic protein as a substrate. This kinase is called the p48 SIP kinase (for SA-Induced Protein kinase). Biologically active analogs of SA, which induce pathogenesis-related genes and enhanced resistance, also activated this kinase, whereas inactive analogs did not. Phosphorylation of a tyrosine residue(s) in the SIP kinase was associated with its activation. The SIP kinase was purified to homogeneity from SA-treated tobacco suspension culture cells. The purified SIP kinase is strongly phosphorylated on a tyrosine residue(s), and treatment with either protein tyrosine or serine/threonine phosphatases abolished its activity. Using primers corresponding to the sequences of internal tryptic peptides, we cloned the SIP kinase gene. Analysis of the SIP kinase sequence indicates that it belongs to the MAP kinase family and that it is distinct from the other plant MAP kinases previously implicated in stress responses, suggesting that different members of the MAP kinase family are activated by different stresses.