Growth Factors and Mitogen-Activated Protein Kinases
Mammalian cells respond to external stimuli by activation of a variety of signal transduction pathways, which culminate in stereotypical responses, such as proliferation, growth arrest, hypertrophy, differentiation, or apoptosis. In vertebrates the actions of many stimuli resulting in proliferative or hypertrophic growth converge on a set of cellular kinase cascades, which are collectively called the mitogen-activated protein (MAP) kinase cascades. These MAP kinases have been implicated in vascular smooth muscle cell proliferation and hypertrophy, responses that are central to the pathophysiology of hypertension. In this review, we will examine how proliferative and hypertrophic stimuli activate these MAP kinase cascades, what are the consequences of that activation on gene expression, and how do these signals drive the cell into one of the stereotypical responses noted above.
- MAP kinases
- stress-activated protein kinases
- G-protein-coupled receptors
- angiotensin II
- growth factors
- Ang II = angiotensin II
- CREB = cAMP response element binding protein
- EGF = epidermal growth factor
- ERK = extracellular signal-regulated kinase
- ET = endothelin
- Grb2 = growth factor receptor bound protein 2
- JNK = c-Jun N-terminal kinase
- MAP = mitogen-activated protein
- MAPKAP = mitogen-activated protein kinase activated protein kinase
- MKP = mitogen-activated protein kinase phosphatase
- PDGF = platelet-derived growth factor
- PI = phosphatidylinositol
- PKC = protein kinase C
- SAPK = stress-activated protein kinase
- SH = Src homology
- SOS = son of sevenless
- TNF = tumor necrosis factor
Three MAP kinase cascades have been characterized in mammalian cells (Fig 1). The most well characterized of these is the ERK cascade. Treatment of vascular smooth muscle cells with growth factors that induce proliferation or hypertrophy, such as EGF, PDGF, endothelin, vasopressin, or angiotensin II, activate this pathway. Numerous studies have demonstrated that the ERK cascade is critical to the mitogenic response, to cellular differentiation, and, in some cells, to the induction of hypertrophy. The two other MAP kinase cascades are activated by cellular stress. One of these cascades is called the p54 MAP kinase, the SAPK or JNK pathway. Cellular stresses such as oxidant stress, reperfusion of ischemic tissue, cell stretch, shear stress, or exposure to inflammatory cytokines, TNF-α or interleukin-1β, or to vasoactive peptides including angiotensin II and endothelin activate this pathway.
p38 is the MAP kinase in the third cascade and is the second stress-response MAP kinase. p38 is activated by many of the same stimuli that activate the SAPK/JNK pathway. While these stress-response pathways do not appear to be involved in the transduction of mitogenic signals, they are clearly relevant to the understanding of vascular wall pathology since they transduce growth inhibitory or apoptotic signals and may be responsible, in part, for induction of the hypertrophic phenotype in various types of cells.
When activated, all three MAP kinases have transcription factors as their primary targets. It is these transcription factors that regulate the induction of sets of genes which determine, in large part, the ultimate biological response of the cell. While much has been learned of the pathways controlling the MAP kinases and the transcription factors regulated by them, it is only recently that we have begun to define the pathophysiological importance of these cascades.
Kinases and Substrates of the ERK Pathway
ERK-1 and ERK-2 were initially identified as two protein kinases that became phosphorylated on tyrosine in response to insulin and other growth factors.1 This pathway can be activated by stimulation of receptors with intrinsic tyrosine kinase activity or G-protein-coupled receptors.2 A number of agents important for control of smooth muscle cell hypertrophy and proliferation, vascular resistance, and platelet aggregation activate these kinases. These agonists include PDGF,2 Ang II,3 ET,4–6⇓⇓ thromboxane A2,7 prostaglandin H2,7 prostaglandin F2a,8 thrombin,4 norepinephrine,9 and acetylcholine10.
The cascade of cellular proteins that results in sequential activation in the ERK pathway has been established most extensively in cases where the initiating event is binding of a ligand, such as PDGF, to its receptor which has tyrosine kinase activity. A PDGF dimer binds to two receptors, forming a receptor dimer.11 The dimerized receptors, with intrinsic tyrosine kinase activity, then “autophosphorylate” or, more correctly, cross-phosphorylate several tyrosine residues within the receptor.12 One phosphotyrosine residue within the kinase domain enhances catalytic activity of the receptor’s tyrosine kinase. Several other phosphotyrosine residues outside the kinase domain act as docking sites for signal transducing molecules (Fig 2). These signal transducing molecules interact with the various phosphotyrosine residues on the receptor via SH2 domains, first identified in the nonreceptor tyrosine kinase, Src. These SH2 domains are composed of approximately 100 amino acids and interact with phosphotyrosine residues but not unphosphorylated tyrosine residues. When the cell is in a resting state, a cytosolic protein, Grb2, which contains an SH2 domain, binds a guanine nucleotide exchange protein, mSOS (the mammalian homolog to Drosophila SOS) through the two SH3 domains of Grb2 and proline-rich sequences of mSOS.13 When the PDGF receptor is tyrosine phosphorylated, the Grb2-mSOS complex is recruited to the receptor when the SH2 domain of Grb2 interacts with a specific region of the receptor, which contains a phosphorylated tyrosine. This recruitment of mSOS to the membrane allows mSOS to interact with the small molecular weight GTP-binding (G) protein, Ras, and converts Ras from the GDP-bound (inactive) to the GTP-bound (active) state.
Ras-GTP then in turn recruits Raf kinase family members (Raf-1, A-Raf, and B-Raf).14 Raf-1 binds to Ras via a region in its N terminus, residues 51 to 131, contained in the CR 1 region (conserved region 1, which is conserved in mammalian, Drosophila, and Caenorhabditis elegans Raf homologs).14 Raf, when translocated to the membrane, is constitutively active but its full activation depends on additional factors. Targeting of Raf-1 to the membrane by attaching a membrane localization signal (CAAX box plus a polybasic domain) is sufficient to activate c-Raf-1, even in the presence of dominant negative Ras.15,16⇓ However, c-Raf-1 is not fully activated in this setting unless a growth factor is also added. Furthermore, coincubation of purified GTP-loaded Ras with c-Raf-1 fails to activate the latter, consistent with the need for an additional growth factor-activated cofactor to activate c-Raf-1.
The most likely cofactors/activators of c-Raf-1 are the 14-3-3 proteins.17,18⇓ When 14-3-3 is injected into Xenopus oocytes, c-Raf-1 becomes activated. Activation of c-Raf-1 in a yeast yeast expressing Ras and c-Raf-1 is dependent on yeast 14-3-3 proteins. The β isoform of 14-3-3, identified in a yeast two-hybrid screen with c-Raf-1 as the “bait,” associates with the N-terminal regulatory domain of c-Raf-1.19 Each 14-3-3 molecule can bind two c-Raf-1 molecules.20–22⇓⇓ The 14-3-3 protein appears to allow c-Raf-1 molecules to membrane by effectively when they are brought to the cell membrane by GTP-bound Ras. Oligomerization of c-Raf-1 is critical for activation of c-Raf-1 kinase activity and may bring about cross-phosphorylation of one c-Raf-1 molecule by the other.22 There remains controversy as to whether oligomerization is sufficient to activate Raf-1, in which case the role of Ras-GTP is to recruit c-Raf so that oligomers can form, or whether Ras-GTP remains necessary after the formation of oligomers.20,21⇓ Activation presumably occurs when the inhibitory amino-terminal regulatory domain of c-Raf-1, which normally masks the kinase domain, swings out of the way of the kinase domain, freeing it to interact with its target.23 The importance of this regulatory domain of c-Raf-1 is best seen with truncation mutants of the protein lacking the regulatory domain, such as BXB-Raf, which are highly transforming.23,24⇓
Activated Raf can then phosphorylate the dual specificity kinases, MEK1a on Ser-218 and Ser-222 or MEK1b and MEK2 on comparable conserved residues within kinase subdomain VIII. These MEKs have high substrate specificity for their downstream targets, ERK1 and ERK2. MEKs phosphorylate ERKs on Thr-183 and Tyr-185, in a TEY sequence. The ERKs, in turn, are proline-directed kinases that phosphorylate substrates with a PX(S/T)P sequence. ERKs activate other serine/threonine kinases, the p90rsk isoforms RSK1, RSK2, and RSK3.25,26⇓ ERKs also phosphorylate cytosolic phospholipase A27,28⇓2 and the EGF receptor.29 Recently two ERK substrates have been identified and named MAP kinase-interacting serine/threonine kinase 1 and 2 (Mnk1 and Mnk2).30,31⇓ Mnk1 phosphorylates eukaryotic initiation factor-4E (also known as eukaryotic initiation factor-4a) at Ser-209. Eukaryotic initiation factor-4E is a translation initiation factor that binds the 7-methylguanosine cap on all eukaryotic mRNAs. This protein plays an important role in the regulation of translation in mammalian cells.32 When phosphorylated eukaryotic initiation factor-4E has enhanced affinity for the 5′ cap.33
Regulation of Transcription by the ERKs
On activation, a portion of the ERK proteins translocate to the nucleus.34,35⇓ In general, if a growth factor leads to prolonged activation of the ERKs, translocation of the ERKs to the nucleus is more pronounced, and this correlates with induction of mitogenesis.34,36⇓
ERKs phosphorylate and activate at least two transcription factors, c-Myc and Elk-1.37–39⇓⇓ Elk-1 and the related SAP-1 and SAP-2 are ternary complex factors, which form a complex with serum response factor and together bind to the promoter of a number of genes including c-fos that contain the serum response element.40 Elk-1, SAP-1, and SAP-2 contain an Ets DNA-binding domain.41,42⇓ Phosphorylation of each of these proteins promotes ternary complex formation and enhances transcriptional activating activity, which leads to increased production of c-fos mRNA.38 These activated transcription factors play critical roles in the induction of immediate early genes and in the mitogenic response. The RSK protein kinase family, which are substrates of the ERKs, also translocate to the nucleus when cells are exposed to activators of the c-Raf-1/ERK cascade.43 One family member, RSK2, may also play an important role in immediate early gene induction and mitogenesis since it appears to phosphorylate CREB at Ser-133, a critical residue for activation of CREB and for expression of c-fos in response to some growth factors.44 The existence of Ets domain-containing transcription factors in Drosophila, PointedP2 and Yan, that are also ERK substrates suggests that many other mammalian transcription factors will be identified which are regulated by this pathway.38 Identifying these substrates, how they are regulated by the ERKs, and what role they play in mitogenesis and hypertrophy will likely offer important insight into these pathophysiological responses of the vascular wall.
G-protein Receptor Activation of the ERK Pathway
Many of the agonists that activate the ERK pathway do not interact with receptors with intrinsic tyrosine kinase activity. Rather, agonists such as Ang II, ET, and prostaglandin F2a interact with receptors that are coupled to heterotrimeric GTP-binding proteins. The processes responsible for activation of the ERKs by these receptors have been less well established, although recent progress in this area has provided a great deal of insight. In some cell types (eg, Rat 1 cells), activation of the ERK pathway by G-protein-coupled receptors may involve activation of receptor tyrosine kinases (e.g., EGF receptor) via an ill-defined crosstalk mechanism.45 In most cells, however, activation of the ERK pathway does not appear to use this mechanism.
In the cases of many agonists, including Ang II and ET-1, the ERK pathway can be activated by both Ras-dependent and Ras-independent mechanisms with different mechanisms operant in different cell types (Fig 2). For those cells using Ras, the activation by ligands with heterotrimeric G-protein-coupled receptors may proceed via α- or βγ-subunits, depending on the type of receptor. Lefkowitz and co-workers46 demonstrated that stimulation of both the α1B-adrenergic receptor, which couples to Gq/11, or the α2A-adrenergic receptor, which couples to G1, culminates in activation of the ERKs via a Ras-dependent mechanism; however, the proximal mechanisms utilized by the receptors differ. For the Gi-coupled receptor, expression of a βγ-subunit sequestering protein derived from the β-adrenergic receptor kinase blocked agonist-induced Ras, c-Raf-1, and ERK activation, suggesting that βγ mediated activation of the cascade.46 These data support earlier work showing that overexpression of βγ- subunits, but not constitutively active αi subunits, activated the ERKs, and that expression of a dominant negative Ras prevented that activation, placing Ras downstream of βγ. For the Gq/11-coupled receptor, the βγ sequestering β-adrenergic receptor kinase peptide does not prevent ERK activation, suggesting that αq mediates the activation of Ras.
How do βγ and αq activate Ras? In 1991 we reported that ET, vasopressin, and Ang II signaled via phospholipase C to enhance tyrosine phosphorylation of a number of substrates in rat renal mesangial cells. Activation of protein kinase C was not necessary for this activation of tyrosine phosphorylation. With this observation it was clear that tyrosine phosphorylation was likely to be an important component of the signaling pathways of these three agonists whose interactions with cells were mediated via G protein-coupled receptors.47 Src was the first identified tyrosine kinase, which was activated in response to ET-1. Simonson and co-workers subsequently reported that induction of c-fos in response to ET-1 requires Src and that Ras is downstream of Src in this ET-1 signaling pathway to the ERKs.47a
With the confirmation that Src was critical for linking G-protein-coupled receptors to Ras and ERK activation, two major questions remained: (1) How do βγ or αq activate Src and (2) How does Src activate Ras? A working model (Fig 2) is emerging whereby βγ and αq-mediated activation of Src and the rest of the cascade appear to converge very proximally at activation of phospholipase Cβ, and the remainder of the pathway to ERK activation may be shared by many types of receptors in many types of cells. Activation of phospholipase Cβ generates an increase in intracellular calcium concentration. This calcium transient then leads to tyrosine phosphorylation of a nonreceptor tyrosine kinase, Pyk2. Phosphorylation of residue Tyr-402 of Pyk2 allows it to form a complex with Src via the latter’s SH2 domain and results in the activation of Src. That these two tyrosine kinases play critical roles in activation of Ras and the ERK cascade is suggested by inhibition of ERK activation in response to occupation of the α2A-adrenergic receptor or the lysophosphatidic acid receptor (Gi-linked), or α1B-adrenergic receptors or bradykinin receptors (Gq-linked) by overexpression of Csk (the C-terminal Src kinase that phosphorylates and inactivates Src) or overexpression of dominant interfering mutants of Src or Pyk2.47b
Once Src is activated, the Src-induced activation of Ras employs many of the same signaling molecules used by the growth factor receptors with intrinsic tyrosine kinase activity. After activation of Pyk2 and Src, the adaptor protein Shc associates with Src, Pyk2, or both, depending on the cell type, via interaction of the SH2 domain of Shc with phosphotyrosine residues of the kinases. After binding to the kinase(s), Shc then is phosphorylated on a tyrosine residue, which recruits the Grb2/mSOS complex via interaction of the SH2 domain of Grb2 with the phosphotyrosine residue of Shc. Alternatively, Grb2/mSOS may be recruited directly by Pyk2 without the intermediary, Shc. As with the growth factor receptors, association of the Shc/Grb2/mSOS complex with Src serves to recruit the guanine nucleotide exchange factor to the membrane (and thus to Ras) since Src is membrane associated via N-terminal myristylation.
Ras-Independent Activation of the ERKs
The predominant Ras-independent pathway to ERK activation appears to involve PKC-dependent activation of Raf-1 and, in many cells, uses classical PKC isoforms that can be downregulated by prolonged incubation of cells with high concentrations of phorbol esters. Activation of this PKC-dependent pathway is more often associated with Gq-linked receptors than with Gi-linked receptors and has been demonstrated with receptors including the ET-1 receptor, the vasopressin receptor, the α1B-adrenergic receptor, and the m1-muscarinic receptor. This pathway appears to be activated by αq and not βγ since expression of a constitutively active mutant of αq (but not αi) in COS cells activates the c-Raf-1/ERK cascade. Furthermore, Lefkowitz and co-workers46 found that activation of the c-Raf-1/ERK cascade by the Gq-coupled α1B-adrenergic receptor, and the m1-muscarinic receptor were not blocked by expression of the β-adrenergic receptor kinase peptide, a dominant negative form of Ras, or by protein tyrosine kinase inhibitors, but were inhibited by dominant negative c-Raf-1 or by PKC depletion with prolonged phorbol ester treatment. Thus it appears that Gq-coupled receptors can also utilize αq to activate phosphoinositide hydrolysis via a Ras-independent mechanism, leading to PKC activation and, subsequently, activation of c-Raf-1 (Fig 2).
The mechanism of PKC-induced activation of the Raf-1/ ERK cascade is not clear. Expression of a constitutively active mutant of PKCδ, or overexpression of wild-type PKCε activates the ERK cascade, and the activation is Ras-independent, but c-Raf-1-dependent, suggesting PKCs might activate c-Raf-1 directly.48 PKC isoforms have been proposed to be direct activators of c-Raf-1. c-Raf-1 can be phosphorylated by PKCs in vitro at Ser-499, a residue within the kinase domain, and this phosphorylation is associated with activation of c-Raf-1. However, mutation of Ser-499 does not prevent activation of c-Raf-1 in insect cells when coexpressed with Ras and Src or activation of c-Raf-1 in response to a number of stimuli. It appears that phosphorylation of Raf-1 by PKCs may increase c-Raf-1 autokinase (or autophosphorylating) activity, but not its kinase activity directed at its physiological substrate, MEK1.
In summary, the site of action of the PKCs, that can be downregulated by prolonged phorbol ester administration in the Gq-activated Ras-independent pathway to activation of the Raf-1/ERK cascade, remains unclear at this time, but the activation is likely not to be via direct phosphorylation of c-Raf-1. Understanding this Ras-independent pathway to ERK activation may be important in the vasculature since Ang II-mediated activation of the ERKs and enhancement of protein synthesis in vascular smooth muscle cells appears to proceed via this pathway.3
The role of the novel isozymes of PKC, PKCζ and PKCλ, in signaling by G protein-coupled receptors are unclear. These isoforms are not downregulated by prolonged exposure of cells to phorbol esters, making it difficult to study their role in ligand-induced activation of the ERK cascade. A constitutively active mutant of PKCζ is capable of activating the ERK (and the SAPK) cascade, and a dominant negative mutant inhibits serum-induced activation of the ERKs. Furthermore, Berk and colleagues3 propose that PKCζfunctions as a MEK kinase and is responsible for activation of the ERK pathway in response to Ang II. Although these data are intriguing, the importance of the novel PKC isozymes in vascular signaling remains to be determined.
PI-3 kinases may also be involved in the signaling cascade from G protein-linked receptors to Ras.49 One type of PI-3 kinase is activated by growth factors. This PI-3 kinase consists of a heterodimer of a p110 catalytic subunit and a p85 adapter protein, which mediates binding to growth factor receptors and to IRS-1 via its SH2 domain. PI3Kγ does not appear to interact with p85 subunits, but can be activated in vitro by either α- or βγ-subunits of G-proteins. βγ-subunits recruit PI3Kγ to the membrane where its substrates are located. PI3kγ then activates Src or a Src-like kinase, which then triggers tyrosine phosphorylation of Shc, its association with Grb2/mSOS, and activation of Ras and the c-Raf-1/ERK cascade. These studies of the role of PI3Kγ were performed with the Gi-linked m2 muscarinic receptor.
There may be an important role for the arachidonic acid metabolic pathways in Ang II-mediated activation of ERK1/2 kinase and mitogenesis. It has been reported that the 12-lipoxygenase pathway of arachidonate metabolism plays a critical role in the hypertrophic effects of Ang II in vascular smooth muscle cells.50 In Chinese hamster ovary cells transfected with the Ang II receptor, Ang II induces a biphasic increase in ERK1/2 kinase activity.51 Inhibitors of the 12-lipoxygenase pathway inhibit mitogenesis as well as the sustained late peak of ERK1/2 activation. Furthermore the 12-lipoxygenase product, 12-hydroxyeicosatetraenoic acid, increases ERK1/2 activity and increases cell proliferation.
Hyperplasia, Hypertrophy, and Interaction of the ERK Cascade With the Cell Cycle
Expression of constitutively active forms of Ras, c-Raf-1, or MEK-1 are sufficient to activate mitogenesis in fibroblasts.52,53⇓ Furthermore, expression of dominant inhibitory mutants of Ras, c-Raf-1, and MEK-1 block growth factor-induced mitogenesis. These data confirm the critical importance of this cascade in the mitogenic response. Furthermore, since active MEK-1 is sufficient to induce mitogenesis, it is likely that the ERKs are the key downstream kinases promoting cell cycle progression rather than components of a divergent pathway activated by Ras or c-Raf-1. Sustained, as opposed to transient, activation of the ERKs appears to be required for many cells to pass the G1 restriction point and to enter S phase, in which cellular DNA is replicated.54,55⇓ Although ERK activation is linked to the cell cycle, it had not been clear where the c-Raf-1/ERK pathway might interact with the cell cycle machinery. Expression of the D-type cyclins, which are the regulatory (activating) subunits for the cyclin-dependent kinase 4 and 6 (cdk4 and cdk6) catalytic subunits, appear to control the early stages of the transition toward S phase. Levels of D type cyclins rise in response to growth factor stimulation.56 A critical link between signal transduction and the cell cycle has recently been suggested by the finding that expression of dominant inhibitory mutants of MEK-1 or ERK-1 or expression of the MKP-1, which dephosphorylates and inactivates the ERKs (see below), inhibited growth factor-dependent expression of cyclin D1. Expression of a constitutively active mutant of MEK-1 or various Raf constructs increased cyclin D1 expression.57–59⇓⇓ Although the mechanisms are not known, activation of the Raf/ERK cascade also correlates with increased expression of cyclin E (the regulatory subunit of cdk2) and decreased expression of the cyclin-dependent kinase inhibitor p27Kip1. These data suggest the Raf/ERK cascade may interact at multiple sites to induce cell cycle progression.
Surprisingly, in NIH3T3 cells, the ability of various Raf proteins to induce proliferation is inversely correlated with their ability to activate the ERKs. Raf proteins which induce proliferation, only weakly activate the ERKs, and strong activators of the ERKs can induce cell cycle arrest in G1. The ability of the Raf constructs to induce cell cycle arrest is highly correlated with their ability to induce expression of the cyclin-dependent kinase inhibitor, p21Cip1, which inhibits activity of the cyclin D-cdk4 and cyclin E-cdk2 complexes.58,59⇓ These intriguing data suggest a mechanism that may explain the ability of the c-Raf-1/ERK cascade to signal either mitogenesis or cell cycle arrest and differentiation and suggest that this decision depends, at least in part, on the strength of activation of the pathway.
Clearly, the decision of a cell to enter the cell cycle or undergo cell cycle arrest and differentiation also depends on the cell type since transgenic mice expressing constitutively active Ras in the skin show enhanced differentiation of keratinocytes whereas those expressing Ras in the pancreas and liver show enhanced proliferation. In the heart, hypertrophic cardiomyopathy may involve activation of the ERK pathway since transgenic mice expressing constitutively active Ras in their left ventricle develop cardiac hypertrophy and diastolic dysfunction.60 On the other hand, when the hypertrophic response is evaluated by monitoring reporter gene expression that is under control of the atrial natriuretic factor or myosin light chain-2 kinase promoters in cardiac cells, constitutively active MEK does not induce gene expression and inhibits expression that is induced by the hypertrophic stimulus, phenylephrine.61
Turning the ERK Pathway Off
The link between the ERK pathway and cell growth strongly suggests that cells must have regulated mechanisms for turning the ERK pathway off.62 Cells use at least two mechanisms to turn off the ERKs, both of which limit the mitogenic response. Incubation of activated ERKs with either serine/threonine or tyrosine phosphatases in vitro inactivates the kinases. Dual specificity phosphatases (acting on both Ser/Thr and Tyr residues), the MKPs, probably inactivate the ERKs in vivo.63,64⇓ Regulation of MKP-1 activity appears to be predominantly at the level of transcription and marked increases in MKP-1 mRNA are seen within the first 60 minutes after exposure of cells to mitogens. Overexpression of MKP-1 inhibits Ras- and serum-induced DNA synthesis, implicating MKP-1 as a major negative modulator of the mitogenic response to agonists which act via Ras.63 Ang II induces the expression of MKP-1 in vascular smooth muscle cells.65 Antisense oligonucleotides to MKP-1 inhibit MKP-1 expression and result in prolongation of the Ang II-induced ERK1/2 activation, implicating MKP-1 as a dominant ERK1/2 phosphatase in vivo.65
The second mechanism that limits ERK activation involves the cAMP-dependent protein kinase, protein kinase A. Increases in cellular cAMP levels inhibit growth factor-induced proliferation of vascular smooth muscle cells. cAMP-activates protein kinase A, which antagonizes PDGF-induced ERK activation in human arterial smooth muscle cells without affecting receptor tyrosine phosphorylation or phospholipase C activation.66 This inhibition is due, at least in part, to prevention of Ras-dependent activation of c-Raf due to inhibition of the coupling of GTP-loaded Ras to c-Raf-1.67,68⇓ Protein kinase A phosphorylates c-Raf-1 on Ser-43 within the regulatory domain of the protein. This phosphorylation reduces the affinity with which c-Raf-1 binds to Ras.67 Protein kinase A also activates Rap1, a small molecular weight G-protein in the Ras superfamily, which inhibits activation of Raf-1.69 Other mechanisms may play a role in the antagonism of cAMP on the response to Ang II. Takahashi et al70 have reported that cAMP antagonizes the hypertrophic response of vascular smooth muscle cells to Ang II without affecting Ras and MAP kinase activation.70 Given the pleotrophic effects of cAMP, this could be explained by other signaling pathways or other effects of cAMP on transcription factors.
Some drugs also inhibit the ERK pathway. Heparin is a potent inhibitor of vascular smooth muscle proliferation.71,72⇓ Heparin inhibits PDGF-induced tyrosine phosphosphorylation and activation of ERK1/2 at later time points without altering phosphorylation of the PDGF receptor or early activation of ERK1/2.73 Heparin inhibits PDGF-induced phosphorylation of c-Raf-1 and does not stimulate tyrosine phosphatases.73
Novel insulin-sensitizing agents, thiazolidinediones, have been demonstrated to inhibit insulin-, EGF- and basic fibroblast growth factor-induced growth of vascular smooth muscle cells.74 Troglitazone, a member of this family of agents, and related agents attenuate hypertension in a number of animal models.75,76⇓ Troglitazone inhibits basic fibroblast growth factor-induced proliferation and c-fos induction and inhibits neointimal thickening after aortic balloon injury. The effect on c-fos gene induction was associated with inhibition of Elk-1 activation, which was apparently due to inhibition of ERK1/2 phosphorylation of Elk-1.77 However, there was no inhibition of ERK1/2 kinase activity, placing the inhibition of Elk-1 phosphorylation at a site distal to the ERKs. This group also demonstrated that troglitazone inhibits Ang II-induced DNA synthesis and migration in vascular smooth muscle cells.78 As distinct from the situation described above with basic fibroblast growth factor, troglitazone inhibits ERK1/2 kinase activation by Ang II.
Recently, progress has been made in the design of drugs to inhibit components of the MAP kinase cascades. One compound, PD98059, relatively specifically inhibits MEK-1, blocking ERK activation in response to many stimuli. Although this and other agents have been largely used to dissect the contribution of the MAPK pathways to a specific biological response in cells in culture, the potential for the use of agents such as this in the therapy of vascular disease is substantial.
Stress Response MAP Kinase Cascades
Kinases and Substrates of the SAPK/JNK and p38 Cascades
Two other families of MAP kinases are minimally activated by classic growth factors but potently activated by cellular stresses and by agonists that play important roles in hypertension including Ang II, ET-1, and α-adrenergic agents. One family has been designated SAPKs (which has 54- and 46-kD isoforms encoded by at least three genes), since they were activated by cellular stress, or JNK, based on the ability of the kinase to phosphorylate the amino terminus of c-Jun.79–83⇓⇓⇓⇓ The other family includes isoforms of p38, the mammalian homolog of HOG-1, a yeast kinase involved in the response to osmolar stress. Like ERK1/2, the SAPKs and p38 are proline directed and require phosphorylation on both tyrosine and threonine residues for activation.81 Unlike the TEY motif of the ERKs, the SAPKs contain a TPY motif and p38 a TGY motif within kinase subdomain VIII which, when phosphorylated, activates the kinases. Overall, there is 40% to 50% identity in the catalytic domains when comparing the ERKs, SAPKs, and p38.83
p38 and the SAPKs are activated by many of the same cellular stresses including inflammatory cytokines (TNF-α and interleukin-1β), heat shock, osmolar stress, ultraviolet and ionizing radiation, cell stretch, and shear stress.79,80,81⇓⇓ However, differences in activation patterns do exist. For example, we found that the SAPKs are not activated by ischemia alone, but are markedly activated by reperfusion of ischemic kidney.84,85⇓ In contrast, p38 is activated during the ischemic phase in kidney and heart but activity declines during reperfusion.84–86⇓⇓
The SAPKs are also activated in the arterial wall of aortic, carotid, and femoral arteries by acute hypertension in rats whether the hypertension is caused by Ang II or phenylephrine infusion.87 The vasoactive peptides, Ang II, in liver epithelial cells, and endothelin, in airway smooth muscle cells and glomerular mesangial cells, activate the SAPKs and, of note, SAPK activation is greater than ERK activation.88–90⇓⇓ These data raise the possibility that the SAPKs may be a major signaling arm of the vasoactive peptides and suggest that these kinases might play a role in the hypertrophic adaptation of myocytes, vascular smooth muscle cells, and renal mesangial cells.82
The signaling cascades resulting in SAPK/JNK and p38 activation have direct parallels with the ERK cascade, but the cascades are relatively insulated from one another since the MEKs that activate one MAPK are much less effective at activating the others (Fig 1). Although there may be some cross-activation of, for example, ERK1 or p38 by SEK1 (SAPK/ERK kinase-1; also known as MKK4, one of the MEKs upstream of the SAPKs) in cells overexpressing SEK1, at more physiological levels of expression, there appears to be minimal cross-talk.91 92⇓ Additional evidence for the specificity of SEK1/MKK4 is the observation that a dominant negative construct of SEK1 blocks SAPK activation but does not alter agonist-induced ERK activation.91 Similarly, MKK3 and MKK6 are relatively specific activators of p38.92
Upstream of SEK1/MKK4 in the SAPK cascade is MEKK1.93 Although originally named because of its ability to activate MEK1, MEKK1 is a much more specific activator of SEK1/MKK4 than it is of MEK1.94 MEKK1 is not the only activator of SEK1/MKK4 and the SAPK cascade. Multiple protein kinases from different families are capable of activating SEK1/MKK4 and thus function as MEKKs.80,81⇓ This diversity of upstream activators probably reflects the incredible diversity of stimuli that coverge at activation of the SAPKs. MEKKs in the p38 cascade have not been clearly identified.
The signaling components upstream of the MEKK level are also likely to be quite different, depending on the stimulus (eg, osmolar stress versus inflammatory cytokines). However, for the vasculature and its adaptation to hypertension, one family of kinases highly conserved throughout evolution is likely to play a critical role. We and Bokoch and co-workers95,96⇓ first demonstrated that two kinases from the Sterile20 (Ste20) family (so named because yeast with mutations in the STE20 gene do not mate normally in response to pheromone) could activate the SAPK and p38 cascades.95,96⇓ One of these kinases, germinal center kinase, is a member of a growing subfamily of Ste20-like kinases, many of which activate the SAPKs. Although the physiological role of most of these kinases is not known since activators have not been identified, germinal center kinase probably plays a role in SAPK activation by TNF-α and possibly other inflammatory cytokines.96 The other Ste20 family member likely to play a role in hypertension is the p21 activated kinase. p21 activated kinase 1 is regulated by small G-proteins of the Ras superfamily, Rac1 and Cdc42Hs. The p21 activated kinases are also regulated by some heterotrimetric G-protein-coupled receptors, making them ideal candidates for regulating SAPK and, possibly, p38 activation by the vasoactive peptides.
Targets of the Stress Response MAP Kinases
Like the ERKs, after phosphorylation and activation, the SAPKs and p38 translocate to the nucleus where they phosphorylate and activate several targets. Transcription factors are major targets of the SAPKs and p38. The SAPKs phosphorylate c-Jun on Ser-63 and Ser-73, two residues within the transcriptional activation domain, and this enhances transactivating activity of c-Jun. The SAPKs also phosphorylate the transcription factor ATF-2. Like c-Jun, ATF-2 contains an N-terminal transcriptional activation domain and phosphorylation by the SAPKs (or p38) at Thr-69 and Thr-71 within this domain enhances the transcriptional activating activity of ATF-2. That the SAPKs are physiologically relevant ATF-2 kinases is suggested by the observation that the SAPKs are the predominant ATF-2 transactivation domain kinases in postischemic kidney.85 ATF-2 can form homodimers or heterodimers with other members of its family, ATF-3 and CREB, or with c-Jun or NF-kB, suggesting it may play a role in the activation of transcription from many promoters. For example, a c-Jun/ ATF-2 dimer appears to control induction of c-jun in response to cellular stresses, and it is likely that the SAPKs transduce this signal by phosphorylating both transcription factors. In addition, interleukin-1-induced transcription of the gene encoding the adhesion molecule, E-selectin, is regulated, at least in part, by a c-Jun/ATF-2 dimer acting at the NF-ELAM1 site of the E-selectin promoter.
Like the ERKs, the SAPKs and p38 phosphorylate the ternary complex factor, Elk-1, within the C-terminal activation domain which enhances ternary complex formation, DNA binding, and transcriptional activating activity of Elk-1.97 Because all three MAP kinases activate Elk-1 and because ternary complex formation at the SRE of the c-fos promoter controls, at least in part, c-fos induction, it is not surprising that c-fos is induced in response to such a wide variety of stimuli.
The gene encoding collagenase is regulated by an AP-1 site within the promoter, which binds a c-Jun/c-Fos dimer. Because the SAPKs activate c-Jun and Elk-1, the latter leading to increased production of c-Fos, stimuli that recruit the SAPKs are potent inducers of the gene encoding collagenase.82
Although the SAPKs and p38 share many substrates, their substrate specificity is not identical since c-Jun is a SAPK (but not p38) target and the transcription factors SAP-1 and CHOP, a transcription factor involved in stress-induced G1 arrest, are p38 (but not SAPK targets).97 p38 also phosphorylates and activates MAPKAP kinase-2 and -3. Like the ERKs, which control activation of CREB via activation of the protein kinase, RSK2, the p38 pathway may control activation of CREB and the related ATF-1 via MAPKAP kinase-2.98 MAPKAP kinase-2 phosphorylates CREB at Ser-133, increasing transcriptional activating activity of CREB.98 CREB is phosphorylated on Ser-133 after cellular stress. It is likely that this is mediated via activation of p38 and its target, MAPKAP kinase-2, since the p38 inhibitor, SB203580, markedly inhibits CREB phosphorylation and activation.98
MAPKAP kinase-2 also phosphorylates the small heat shock protein, Hsp25/Hsp27, whose phosphorylation is a prominent part of the response to cytokines, cellular stress (ATP depletion, heat shock, etc.), and some growth factors. While the physiological significance of the phosphorylation is unknown and the function of Hsp25/HSP27 is not clear, it likely plays a protective role in the cell.
p38 has other substrates with roles in signal transduction and the response to stress. p38 plays an important role in the aggregation of platelets in response to a variety of agonists since pretreatment of platelets with SB203580 inhibits aggregation.99,100⇓ This has been postulated to be related to phosphorylation by p38 of a critical residue (Ser-505) of cytosolic phospholipase A2, a residue that is also phosphorylated by the ERKs.100 Phosphorylation of this residue by the ERKs has been reported to increase the activity of cytosolic phospholipase A2.
Biological Effects of SAPK/p38 Activation
Although insights into the biological function of the SAPKs and p38 can be gained by identifying downstream targets, this approach has limitations. For example, although the SAPKs and p38 clearly play a role in c-jun and c-fos induction, these IE genes are induced in response to such a wide variety of stimuli that it is difficult to draw any conclusions about the specific biological responses initiated by the kinases. These must be examined directly.
Based on a series of studies, it is now clear that activation of the SAPK and/or p38 cascades can trigger apoptosis or programmed cell death.101 A wide array of stimuli, including withdrawal of nerve growth factor from PC12 cells, oxidant stress, ionizing radiation, and TNF-α, induce apoptosis in susceptible cells. Because expression of dominant inhibitory mutants of components of the SAPK or p38 pathways can block apoptosis, and expression of constitutively active components can induce apoptosis, it is likely that these pathways transduce critical signals in the response to certain apoptotic stimuli. In PC12 cells, expression of constitutively active MEK1, the immediate upstream activator of the ERKs, prevents apoptosis induced by nerve growth factor withdrawal, suggesting that the decision to initiate apoptosis or not may depend on the balance between anti-apoptotic signals transduced by the ERK cascade and pro-apoptotic signals transduced by the SAPK/p38 cascades. c-Raf-1 may be involved as well since it phosphorylates and inactivates the pro-apoptotic Bcl-2 family member, Bad.102,103⇓
The role of apoptosis in the pathogenesis of or adaptation to hypertension is only starting to be explored. Increasing evidence suggests that apoptosis may be a prominent component of ischemia-induced cardiomyocyte cell death and in the progression of idiopathic, ischemic, and hypertensive cardiomyopathy.104 It will be critical to the understanding of these and other processes, such as vascular remodeling and the evolution of the atherosclerotic plaque, to determine the roles played by the SAPKs, p38, and the ERKs in their development.
It seems likely that the SAPKs and p38 will also play a prominent role in the induction of the hypertrophic phenotype in susceptible cells. Although relatively little work has been done in vascular smooth muscle cells, studies using neonatal rat ventricular myocytes have demonstrated that expression of constitutively active MEKK1 induces an increase in myocyte size and transcriptional changes characteristic of the hypertrophic response (increased promoter activity of the "marker" genes, atrial natriuretic factor, β-myosin heavy chain, and skeletal muscle α-actin).105 It is likely, however, that the SAPK cascade is not solely responsible for full induction of the hypertrophic response, since another marker of the hypertrophic response, an increase in organization of the myofilaments, was not seen after transfection of MEKK1105
In some cells, activation of the p38 or SAPK cascade does not induce mitogenesis, apoptosis, or hypertrophy, but causes cell cycle arrest at the G1/S transition.106 p38 may induce cell cycle arrest by inhibiting the induction of G1 cyclins.57 These data suggest that the stress response kinase pathways might also force cells out of the cell cycle, toward a postmitotic, differentiated phenotype. The balance between ERK activation and p38 or SAPK activation may help determine, in part, whether the cell decides to respond to a particular stimulus by undergoing cell cycle arrest/differentiation or by reentering the cell cycle. Thus, the cell’s response to agents, such as Ang II and ET-1, which can induce either the hypertrophic phenotype or mitogenesis, may depend on the balance between activation of these pathways. Determining how the cell makes this decision is clearly one of the most important challenges in the field.
Many of the components of the MAP kinase pathways from cell surface receptors or cellular stress to the nucleus have now been identified. Although certain components remain unclear (eg, the MEKKs in the p38 pathway), and these need to be identified, it is our opinion that the critical issues to be addressed in hypertension research are 2-fold: (1) determine what are the proximal components that transduce the signals from relevant G-protein-coupled receptors (Ang II, endothelin, etc.) and stress (eg, cell stretch, etc.) to activation of stress-response kinase pathways, and (2) determine the biological-responses regulated by activation of these kinases and how those responses might be altered. Many questions need to be answered, including how do growth factors cause mitogenesis in one cell and hypertrophy in another, what role do the ERKs, SAPKs, and p38 play in hypertrophy, and how does the cell decide to respond to SAPK/p38 activation by undergoing apoptosis versus cell cycle arrest versus hypertrophy? Although these and other questions can initially be examined in vitro, ultimately the observations made in cultured cells will need to be confirmed in the intact animal. For this purpose, strategies likely to produce answers include expressing dominant interfering mutants of a MAPK pathway under the control of inducible promoters in an organ or tissue specific manner, or the use of better pathway-specific pharmacological inhibitors. With these complementary approaches, we may be able to dissect the role of these pathways in the progression of hypertensive cardiovascular disease and to define therapeutic strategies to alter that progression.
This work was supported by US Public Health Service Grants DK50282 (to T.F.), and MERIT Award DK39773, DK38452, and NS10828 (to J.V.B.). Dr. Force is an Established Investigator of the American Heart Association.
- Received September 24, 1997.
- Revision received October 22, 1997.
- Accepted November 3, 1997.
Ray LB, Sturgill TW. Insulin-stimulated microtubule-associated protein kinase is phosphorylated on tyrosine and threonine in vivo. Proc Natl Acad Sci U S A. 1988; 85 : 3753 –3757.
Denhardt DT. Signal-transducing protein phosphorylation cascades mediated by Ras/Rho proteins in the mammalian cell: the potential for multiplex signalling. Biochem J. 1996; 318 : 729 –747.
Liao D-F, Monia B, Dean N, Berk BC. Protein kinase C-ζ mediates angiotensin II activation of ERK1/2 in vascular smooth muscle cells. J Biol Chem. 1997; 272 : 6146 –6150.
Kyriakis JM, Force TL, Rapp UR, Bonventre JV, Avruch J. Mitogen regulation of c-Raf-1 protein kinase activity toward mitogen-activated protein kinase kinase. J Biol Chem. 1993; 268 : 16009 –16019.
Wang Y, Simonson MS, Pouyssegur J, Dunn MJ. Endothelin rapidly stimulates mitogen-activated protein kinase activity in rat mesangial cells. Biochem J. 1992; 287 : 589 –594.
Morinelli TA, Zhang L-M, Newman WH, Meier KE. Thromboxane A2/prostaglandin H2-stimulated mitogenesis of coronary artery smooth muscle cells involves activation of mitogen-activated protein kinase and S6 kinase. J Biol Chem. 1994; 269 : 5693 –5698.
Watanabe T. Waga I, Honda Z-i, Kurokawa K, Shimizu T. Prostaglandin F2α stimulates formation of p21ras-GTP complex and mitogen-activated protein kinase in NIH-3T3 cells via Gq-protein-coupled pathway. J Biol Chem. 1995; 270 : 8984 –8990.
Muthalif MM, Benter IF, Uddin MR, Malik KU. Calcium/calmodulin-dependent protein kinase IIa mediates activation of mitogen-activated protein kinase and cytosolic phospholipase A2 in norepinephrine-induced arachidonic acid release in rabbit aortic smooth muscle cells. J Biol Chem. 1996; 271 : 30149 –30157.
Stokoe D, Macdonald SG, Cadwallader K, Symons M, Hancock JF. Activation of Raf as a result of recruitment to the plasma membrane. Science. 1994; 264 : 1463 –1467.
Zhao Y, Bjorbaek C, Weremowicz S, Morton C, Moller DE. RSK3 encodes a novel pp90rsk isoform with a unique N-terminal sequence: growth factor stimulated kinase function and nuclear translocation. Mol Cell Biol. 1995; 15 : 4353 –4363.
Nemenoff RA, Winitz S, Qian N-X, Van Putten V, Johnson GL, Heasley LE. Phosphorylation and activation of a high molecular weight form of phospholipase A2 by p42 microtubule-associated protein 2 kinase and protein kinase C. J Biol Chem. 1993; 268 : 1960 –1964.
Takishima K, Griswold-Prenner I, Ingebritsen T, Rosner MR. Epidermal growth factor (EGF) receptor T669 peptide kinase from 3T3-L1 cells is an EGF-stimulated ‘MAP’ kinase. Proc Natl Acad Sci U S A. 1991; 88 : 2520 –2524.
Fukunaga R, Hunter T. Mnk1, a new MAP kinase-activated protein kinase, isolated by a novel expression screening method for identifying protein kinase substrates. EMBO J. 1997; 16 : 1921 –1933.
Waskiewicz AJ, Flynn A, Proud CG, Cooper JA. Mitogen-activated protein kinases activate the serine/threonine kinases Mnk1 and Mnk2. EMBO J. 1997; 16 : 1909 –1920.
Minich WB, Balasta ML, Goss DJ, Rhoads RE. Chromatographic resolution of in vivo phosphorylated and nonphosphorylated eukaryotic translation initiation factor eIF-4E. Proc Natl Acad Sci U S A. 1994; 91 : 7668 –7672.
Lenormand P, Sardet C, Pages G, L’Allemain G, Brunet A, Pouyssegur J. Growth factors induce nuclear translocation of MAP kinases (p42 and p44) but not of their activator MAP kinase kinase (p45) in fibroblasts. J Cell Biol. 1993; 122 : 1079 –1088.
Gonzalez F, Seth A, Raden D, Bowman D, Fay F, Davis R. Serum-induced translocation of mitogen-activated protein kinase to the cell surface ruffling membrance and the nucleus. J Cell Biol. 1993; 122 : 1089 –1101.
Seger R, Krebs EG. The MAPK signaling cascade. FASEB J. 1995; 9 : 726 –735.
Chen R-H, Sarnecki C, Blenis J. Nuclear localization and regulation of erk-and rsk-encoded protein kinases. Mol Cell Biol. 1992; 12 : 915 –927.
Xing J, Ginty DD, Greenberg ME. Coupling of the RAS-MAPK pathways to gene activation by RSK2, a growth factor-regulated CREB kinase. Science. 1996; 273 : 959 –960.
Della Rocca GJ, van Biesen T, Daaka Y, Luttrell DK, Luttrel LM, Lefkowitz RJ. Ras-dependent mitogen-activated protein kinase activation by G protein-coupled receptors. J Biol Chem. 1997; 272 : 19125 –19132.
Force T, Kyriakis JM, Avruch J, Bonventre JV. Endothelin, vasopressin, and angiotensin II enhance tyrosine phosphorylation by protein kinase C-dependent and -independent pathways in glomerular mesangial cells. J Biol Chem. 1991; 266 : 6650 –6656.
Herman WH, Simonson MS. Nuclear signaling by endothelin-1. J Biol Chem . 1995; 270 : 11654 –11661.
Ueda Y, Hirai S, Osada S, Suzuki A, Mizuno K, Ohno S. Protein kinase C delta activates the MEK-ERK pathway in a manner independent of Ras and dependent on Raf. J Biol Chem. 1996; 271 : 23512 –23519.
Lopez-Ilasaca M, Crespo P, Pellici Giuseppe P, Gutkind Silvio J, Wetzker R. Linkage of G protein-coupled receptors to the MAPK signaling pathway through P1 3-kinase. Science. 1997; 275 : 394 –397.
Natarajan R, Gonzales N, Lanting L, Nadler JL. Role of the lipoxygenase pathway in angiotensin II-induced vascular smooth muscle hypertrophy. Hypertension. 1994; 23 : 142 –147.
Mansour S, Matten W, Hermann A, Candia J, Rong S, Fukasawa K, Woude G, Ahn N. Transformation of mammalian cells by constitutively active MAP kinase kinase. Science. 1994; 265 : 966 –970.
Pages G, Lenormand P, L’Allemain G, Chambard J-C, Meloche S, Pouyssegur J. Mitogen-activated protein kinases p42mapk and p44mapk are required for fibroblast proliferation. Proc Natl Acad Sci U S A. 1993; 90 : 8319 –8323.
Lavoie JN, L’Allemain G, Brunet A, Muller R, Pouyssegur J. Cyclin D1 expression is regulated positively by the p42/44 MAPK and negatively by the p38/HOG MAPK pathway. J Biol Chem. 1996; 271 : 20608 –20616.
Woods D, Parry D, Cherwinski H, Bosch E, Lees E, McMahon M. Raf-induced proliferation or cell cycle arrest is determined by the level of Raf activity with arrest mediated by p21Cip1. Mol Cell Biol. 1997; 17 : 5598 –5611.
Sewing A, Wiseman B, Lloyd AC, Land H. High-intensity Raf signal causes cell cycle arrest mediated by p21Cip1. Mol Cell Biol. 1997; 17 : 5588 –5597.
Hunter JJ, Tanaka N, Rockman HA, Ross J Jr, Chien KR. Ventricular expression of the MLC-2v-ras fusion gene induces cardiac hypertrophy and selective diastolic dysfunction in transgenic mice. J Biol Chem. 1995; 270 : 23173 –23178.
Thorburn J, Carlson M, Mansour SJ, Chien KR, Ahn NG, Throburn A. Inhibition of a signaling pathway in cardiac muscle cells by active mitogen activated protein kinase kinase. Mol Biol Cell. 1995; 6 : 1479 –1490.
Cobb MH, Goldsmith EJ. How MAP kinases are regulated. J Biol Chem. 1995; 270 : 14843 –14846.
Sun H, Tonks NK, Bar-Sagi D. Inhibition of Ras-induced DNA synthesis by expression of the phosphatase MKP-1. Science. 1994; 266 : 285 –288.
Duff JL, Monia BP, Berk BC. Mitogen-activated protein (MAP) kinase is regulated by the MAP kinase phosphatase (MKP-1) in vascular smooth muscle cells: effect of actinomycin D and antisense olionucleotides. J Biol Chem. 1995; 270 : 7161 –7166.
Graves LM, Bornfeldt KE, Raines EW, Potts BC, Macdonald SG, Ross R, Krebs EG. Protein kinase A antagonizes platelet-derived growth factor-induced signaling by mitogen-activated protein kinase in human arterial smooth muscle cells. Proc Natl Acad Sci U S A. 1993; 90 : 10300 –10304.
Wu J, Dent P, Jelinek T, Wolfman A, Weber MJ, Sturgill TW. Inhibition of the EGF-activated MAP kinase signaling pathway by adenosine 3′,5′-monophosphate. Science. 1993; 262 : 1065 –1069.
Cook SJ, McCormick F. Inhibition by cAMP of Ras-dependent activation of Raf. Science. 1993; 262 : 1069 –1072.
Dubey RK, Zhang HY, Reddy SR, Boegehold MA, Kotchen TA. Pioglitazone attenuates hypertension and inhibits growth of renal arteriolar smooth muscle in rats. Am J Physiol. 1993; R726 –R732.
Zhang H, Reddy S, Kotchen T. Antihypertensive effect of pioglitazone is not invariably associated with increased insulin sensitivity. Hypertension. 1994; 24 : 106 –110.
Kyriakis JM, Avruch J. Sounding the alarm: protein kinase cascades activated by stress and inflammation. J Biol Chem. 1996; 271 : 24313 –24316.
Force T, Pombo CM, Avruch JA, Bonventre JV, Kyriakis JM. Stress-activated protein kinases in cardiovascular disease. Circ Res. 1996; 78 : 947 –953.
Pombo CP, Bonventre JV, Avruch J, Woodgett JR, Kyriakis JM, Force T. The stress-activated protein kinases (SAPKs) are major c-Jun amino-terminal kinases activated by ischemia and reperfusion. J Biol Chem. 1994; 269 : 26546 –26551.
Morooka H, Bonventre JV, Pombo CM, Kyriakis JM, Force T, Ischemia and reperfusion enhance ATF-2 and c-Jun binding to cAMP response elements and to an AP-1 binding site from the c-jun promoter. J Biol Chem. 1995; 270 : 30084 –30092.
Yin T, Sandhu G, Wolfgand CD, Burrier T, Webb RL, Rigel DF, Hai T, Whelan J. Tissue-specific pattern of stress kinase activation in ischemic/reperfused heart and kidney. J Biol Chem. 1997; 272 : 19943 –19950.
Zohn IE, Yu H, Li X, Cox AD, Earp HS. Angiotensin II stimulates calcium-dependent activation of c-Jun N-terminal kinase. Mol Cell Biol. 1995; 15 : 6160 –6168.
Shapiro PS, Evans JN, Davis RJ, Posada JA. The seven-transmembrane-spanning receptors for endothelin and thrombin cause proliferation of airway smooth muscle cells and activation of the extracellular regulated kinase and c-Jun NH2-terminal kinase groups of mitogen-activated protein kinases. J Biol Chem. 1996; 271 : 5750 –5754.
Derijard B, Raingeaud J, Barrett T, Wu IH, Han J, Ulevitch RJ, Davis RJ. Independent human MAP kinase signal transduction pathways defined by MEK and MKK isoforms. Science. 1995; 267 : 682 –685
Lange-Carter CA, Johnson GL. Ras-dependent growth factor regulation of MEK kinase in PC12 cells. Science. 1994; 265 : 1458 –1461.
Zhang S, Han J, Sells MA, Chernoff J, Knaus UG, Ulevitch RJ, Bokoch GM. Rho family GTPases regulate p38 MAP kinase through the downstream mediator Pak1. J Biol Chem. 1995; 270 : 23934 –23936.
Whitmarsh AJ, Yang S-H, Su M-S, Sharrocks AD, Davis RJ. Role of p38 and JNK mitogen-activated protein kinases in activation of ternary complex factors. Mol Cell Biol. 1997; 17 : 2360 –2371.
Saklatvala J, Rawlinson L, Waller RJ, Sarsfield S, Lee JC, Morton LF, Barnes MJ, Farndale RW. Role for p38 mitogen-activated protein kinase in platelet aggregation caused by collagen or a thromboxane analogue. J Biol Chem. 1996; 271 : 6586 –6589.
Kramer RM, Roberts EF, Um SL, Borsch-Haubold AG, Watson SP, Fisher MJ, Jakubowski JA. p38 mitogen-activated protein kinase phosphorylates cytosolic phospholipase A2 in thrombin-stimulated platelets. J Biol Chem. 1996; 271 : 27723 –27729.
Xia Z, Dickens M, Raingeaud J, Davis RJ, Greenberg ME. Opposing effects of ERK and JNK-p38 MAP kinases on apoptosis. Science. 1995; 270 : 1326 –1331.
Thorburn J, Xu S, Thorburn A. MAP kinase- and Rho-dependent signals interact to regulate gene expression but not actin morphology in cardiac muscle cells. EMBO J. 1997; 16 : 1888 –1900.
Molnar A, Theodoras AM, Zon LI, Kyriakis JM. Cdc42Hs, but not Racl, inhibits serum-stimulated cell cycle pregression at G1/S through a mechanism requiring p38/RK. J Biol Chem. 1997; 272 : 13229 –13235.