Role of Mitogen-Activated Protein Kinase Kinase 4 in Cancer

Home , MAP2K4, MAP kinase kinase kinase, Protein kinase

AJ Whitmarsh1 and RJ Davis2

1Faculty of Life Sciences, University of Manchester, Manchester, UK and 2Howard Hughes Medical Institute, Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA

Mitogen-activated protein (MAP) kinase kinase 4 of the extracellular signal-regulated kinase (ERK), (MKK4) is a component of stress activated MAP kinase ERK5, c-Jun N-terminal kinase (JNK) and p38 (Chang signaling modules.It directly phosphorylates and activates and Karin, 2001). the c-Jun N-terminal kinase (JNK) and p38 families of The JNK and p38 MAP kinases are collectively MAP kinases in response to environmental stress, pro- referred to as stress-activated MAP kinases. They are inflammatory cytokines and developmental cues.MKK4 is activated in response to a variety of environmental ubiquitously expressed and the targeted deletion of the stresses and pro-inflammatory cytokines, and also play Mkk4 gene in mice results in early embryonic lethality. important roles in development (Davis, 2000; Kyriakis Further studies in mice have indicated a role for MKK4 in and Avruch, 2001). A number of MKKs can phosphor- liver formation, the immune system and cardiac hyper- ylate and activate JNK and p38. MKK3 and MKK6 trophy.In humans, it is reported that loss of function activate p38, MKK7 activates JNK, whereas MKK4 mutations in the MKK4 gene are found in approximately can activate both JNK and p38 (Davis, 2000; Kyriakis 5% of tumors from a variety of tissues, suggesting it may and Avruch, 2001) (Figure 1a). The JNK and p38 have a tumor suppression function.Furthermore, MKK4 pathways are implicated in tumor suppression (Kennedy has been identified as a suppressor of metastasis of and Davis, 2003; Bulavin and Fornace Jr, 2004) and this prostate and ovarian cancers.However, the role of MKK4 is supported by the presence of loss of function in cancer development appears complex as other studies mutations in the MKK4 gene in approximately 5% of support a pro-oncogenic role for MKK4 and JNK.Here human tumors from a variety of tissues (Teng et al., we review the biochemical and functional properties of 1997; Su et al., 1998). However, it is also reported that MKK4 and discuss the likely mechanisms by which it may MKK4 and JNK can participate in tumor formation regulate the steps leading to the formation of cancers. suggesting a more complex role for this pathway in Oncogene (2007) 26, 3172–3184. doi:10.1038/sj.onc.1210410 tumor development (Kennedy and Davis, 2003; Wang et al., 2004). Keywords: MAP kinase; MKK4; JNK; p38; tumor In this review, we describe the properties of MKK4 suppressor and examine how this protein kinase and its downstream targets contribute to controlling the development of cancers.

Introduction Biochemical properties of MKK4 The mitogen-activated protein (MAP) kinase signaling Cloning, structure and tissue distribution of MKK4 pathways are important mediators of cellular responses MKK4 was first identified in screens for novel MKK to extracellular signals that include growth factors, family members in Xenopus laevis and termed XMEK2 hormones, cytokines and environmental stresses (Chang (Yashar et al., 1993). Subsequently the Drosophila, and Karin, 2001). These pathways are evolutionarily mouse and human homologs were cloned and named conserved among eukaryotes and feature a triple kinase DMKK4, stress-activate protein kinase/extracellular- cascade comprised of the MAP kinase which is signal-regulated protein kinase kinase-1 and MKK4, phosphorylated and activated by a MAP kinase kinase respectively (Sanchez et al., 1994; De´ rijard et al., 1995; (MKK), which itself is phosphorylated and activated by Lin et al., 1995; Han et al., 1998). The human MKK4 a MAP kinase kinase kinase (MKKK) (Chang and gene is located on chromosome 17 and encodes a protein Karin, 2001). In mammals, four distinct MAP kinase of 399 amino acids (De´ rijard et al., 1995; Yoshida et al., pathways have been identified that lead to the activation 1999). Overall the mammalian MKK family share about 40% homology within their catalytic domains and Correspondence: Professor RJ Davis, Howard Hughes Medical MKK4 is most similar to MKK7 in this region (50% Institute, Program in Molecular Medicine, University of Massachu- setts Medical School, 373 Plantation Street, Worcester, MA 01605, identity; Tournier et al., 1997; Cuenda, 2000). The USA. catalytic domains of MKKs, like other Ser/Thr kinases, E-mail: [email protected] contain 11 subdomains (Hanks et al., 1988). The crystal MKK4 in cancer AJ Whitmarsh and RJ Davis 3173 a binding determinant is the domain for versatile docking Ras and Rho-family GTPases (DVD) located in the C-terminus of MKKs (Xia et al., 1998; Takekawa et al., 2005; Figure 1b). The interactions between MKKs and the upstream and downstream kinases of the cascade appear to be critical for efficient MEKK MLK TAK1 ASK1 TPL2 signal transfer through MAP kinase pathways (Xia et al., 1998; Ho et al., 2003; Takekawa et al., 2005). MKK4 mRNA is ubiquitously expressed in adult MKK7MKK4 MKK3 MKK6 mouse and human tissue with the highest levels of expression in brain, in particular in the cerebral cortex, hypothalamus, hippocampus and cerebellum (Sanchez et al., 1994; De´ rijard et al., 1995; Carboni et al., 1997; JNK p38 Lee et al., 1999). In early embryogenesis in mice (up to embryonic day 10 (E10)), Mkk4 transcripts are confined to the central nervous system. Later, starting at E12, Cyclin D1 MAPKAPK2 Mkk4 becomes highly expressed in the developing liver c-Jun Bax p53 p53 Bcl2 coincident with a period of active differentiation and an Bax increase in liver size (Lee et al., 1999). Subcellular 14-3-3 Bim AR CDC25 RXRα Bmf localization studies demonstrate that MKK4 protein is RARα Bcl-XL mainly found in the cytoplasm, although some nuclear AR Mcl-1 localization has been detected (Tournier et al., 1999; Coffey et al., 2000). 257 261 b S-I-A-K-T 1 94 399 D KD DVD MKK4 activation of JNK and p38 MAP kinases 39 45 364 387 MKK4 is unique among the mammalian MKK family Figure 1 (a) Mammalian stress-activated MAP kinase pathways. in its ability to phosphorylate and activate two MAP JNK is activated by phosphorylation by MKK4 and MKK7 kinase groups: JNK and p38 (De´ rijard et al., 1995; Lin whereas p38 is activated by MKK4, MKK3 and MKK6. These et al., 1995). MKK3 and MKK6 are specific for p38, MKKs can be activated by many different MKKKs depending on whereas MKK7 is a specific JNK activator (Davis, 2000; the stimulus. JNK and p38 phosphorylate many known regulators of tumorigenesis. (b) Domain structure of MKK4. The kinase Kyriakis and Avruch, 2001). MKK4 activates all the domain (KD) of MKK4 contains eleven subdomains. MKK4 is mammalian JNK isoforms (JNK1, JNK2 and JNK3) activated by phosphorylation of the Ser and Thr residues (in bold) and a subset of p38 isoforms (p38a, p38b) (De´ rijard within the S-I-A-K-T motif located between subdomains VII and et al., 1995; Lin et al., 1995; Jiang et al., 1996). VIII. At the N-terminus there is a D-domain (D) motif for binding to JNK and p38, and at the C-terminus a DVD domain that All MAP kinases are activated by phosphorylation of mediates interactions with various MKKKs. The numbers refer to the Thr and Tyr residues of a Thr-X-Tyr motif located amino acids in human MKK4. within kinase subdomain VIII (Chang and Karin, 2001). Both residues need to be phosphorylated for full activation of the MAP kinase (Chang and Karin, 2001). However, it was observed in vitro that MKK4 structures of MEK1 and MEK2, the MKKs in the ERK preferentially phosphorylated the Tyr residue on JNK pathway, demonstrate that MKKs fold into a small b- (Sanchez et al., 1994; Lin et al., 1995), whereas the stranded N-terminal lobe and a larger helical C-terminal second JNK activator, MKK7, preferentially targeted lobe (Ohren et al., 2004). The adenosine triphosphate the Thr residue (Lawler et al., 1998). These observations binding site is located in the cleft formed between the led to the hypothesis that JNK isoforms are activated two lobes and is surrounded by residues that are synergistically by MKK4 and MKK7 (Lawler et al., conserved between the MKK family members (Cuenda, 1998; Fleming et al., 2000; Lisnock et al., 2000). Some 2000; Ohren et al., 2004). in vivo evidence to support this model has come from MKK4, similar to the other MKKs, also contains studies using mouse embryonic stem (ES) cells and docking sites for both upstream and downstream mouse embryonic fibroblasts (MEFs) that feature components of the JNK and p38 signaling cascades targeted deletions of the Mkk4 and Mkk7 genes (Xia et al., 1998; Ho et al., 2003; Takekawa et al., 2005; (Tournier et al., 2001; Wada et al., 2001; Kishimoto Figure 1b). At the N-terminus there is a D-domain type et al., 2003). These studies also demonstrated that docking site that binds to JNK and p38 (Ho et al., distinct stimuli might differentially utilize MKK4 and 2003). In addition to MKKs, D-domain docking sites MKK7. For example, the activation of JNK in response that bind to MAP kinases are found in many proteins to the pro-inflammatory cytokines tumor necrosis involved in MAP kinase signaling including MAP factor-a (TNFa) and interleukin-1 (IL-1) was almost kinase phosphatases, substrates, and scaffold or adaptor completely abolished in Mkk7À/À MEFs, but reduced proteins (Sharrocks et al., 2000). The region encom- to around 50% activation in Mkk4À/À cells (Tournier passing the D-domain may also participate in the et al., 2001). This indicates that MKK7 is essential for binding of MKKKs to MKKs, although the major JNK activation by these cytokines, whereas MKK4

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3174 contributes to optimal JNK activation. This is sup- put forward whereby there are sequential bipartite ported by studies demonstrating that TNFa and IL-1 interactions between MEKK1 and MKK4, and preferentially activate MKK7 in vitro and in cells (Finch MKK4 and JNK (Xia et al., 1998). It is proposed that et al., 1997; Lawler et al., 1997; Moriguchi et al., 1997; MEKK1 binds to MKK4 and phosphorylates it Tournier et al., 2001). In contrast, it was observed in resulting in the dissociation of the activated MKK4 Mkk4À/À and Mkk7À/À ES cells and MEFs that both from MEKK1, thereby allowing it to interact with and MKK4 and MKK7 make similar contributions to JNK phosphorylate JNK (Xia et al., 1998). activation in response to a number of environmental In addition to direct interactions between the protein stresses including UV radiation, heat shock and osmotic kinase components of MAP kinase signaling cascades, shock (Tournier et al., 2001; Kishimoto et al., 2003). these pathways can be regulated by their association Although the studies described above provided strong with scaffold proteins that can act to colocalize the genetic evidence to support the role of MKK4 as a JNK components of the cascade and facilitate their activation activator in vivo, it had been less clear whether MKK4 (Morrison and Davis, 2003). A number of scaffold could regulate p38 activity in vivo. The Drosophila proteins interact with MKK4 and promote JNK homolog, DMKK4, has been shown to target JNK but activation including JNK-interacting protein-3 (JIP3) not p38 (Han et al., 1998), whereas in Mkk4À/À murine and JIP4 isoforms, plenty of SH3s (POSH), and b- ES cells there was no defect in p38 activation in response arrestin-2 (McDonald et al., 2000; Xu et al., 2003b; to a number of stresses (Yang et al., 1997b; Ganiatsas Whitmarsh, 2006). JIP3 and POSH play important roles et al., 1998). Meanwhile, experiments using Mkk4À/À in brain development, neuronal trafficking and apopto- MEFs have provided conflicting evidence as to whether sis (Kelkar et al., 2003; Xu et al., 2003a, b; Ha et al., there is a defect in p38 activation in response to TNFa 2005; Kim et al., 2005), although how these functions and IL-1 (Ganiatsas et al., 1998; Tournier et al., 2001; relate to their MKK4-JNK scaffolding role is not fully Brancho et al., 2003). A role for MKK4 in UV understood. Similarly, the physiological role of the b- radiation-induced activation of p38 in MEFs lacking arrestin-2 complex with the JNK pathway is unclear both p38-specific MKKs, MKK3 and MKK6, has been (McDonald et al., 2000). MKK4 is therefore a demonstrated (Brancho et al., 2003). In these cells some component of multiple signaling complexes, some of UV-induced p38 activity remains and this can be which are cell type dependent, and these may participate suppressed by reducing the level of MKK4 by small in distinct cellular processes. interfering RNA (siRNA) (Brancho et al., 2003). Unlike the preferential Tyr phosphorylation of JNK by mammalian MKK4, p38 is phosphorylated equally well by MKK4 on both the activating Tyr and Thr residues Roles of MKK4 in vivo (Brancho et al., 2003). Mice with a targeted deletion of the Mkk4 gene die during embryogenesis providing clear evidence that Mechanism of MKK4 activation MKK4 has distinct functions in cells that cannot be MKKKs activate MKK4 by phosphorylating the Ser/ complemented by other MKKs (Yang et al., 1997b; Thr residues in the Ser-Ile-Ala-Lys-Thr motif located in Ganiatsas et al., 1998). The Mkk4À/À embryos die the T-loop of the kinase domain (KD) between between E11.5 and E13.5 and display anemia and severe subdomains VII and VIII (Cuenda, 2000; Kyriakis and hemorrhaging in the liver (Ganiatsas et al., 1998; Avruch, 2001). Many MKKKs are capable of phos- Nishina et al., 1999). They also display impaired liver phorylating and activating MKK4 including members formation and abnormal hepatogenesis, which corre- of the mitogen and extracellular-regulated kinase kinase lates with significant apoptosis of liver cells (Ganiatsas (MEKK) and mixed-lineage kinase (MLK) families, as et al., 1998; Nishina et al., 1999). well as apoptosis signal-regulated kinase-1 (ASK1), Further transgenic studies in mice have uncovered transforming growth factor-b (TGFb)-activated kinase- potentially important roles for MKK4 in the immune 1 (TAK1) and Tpl2 (Cuenda, 2000; Kyriakis and system and in the heart. The precise role of MKK4 in Avruch, 2001) (Figure 1a). Different MKKKs have the immune system is unclear as separate studies in different specificities for the MKKs. For example, Mkk4À/À mouse chimeras developed using recombina- MEKK1 and TAK1 phosphorylate both MKK4 and tion-activating gene (RAG)2 blastocyst complementa- MKK7, whereas MEKK4 appears to preferentially tion came to different conclusions. One group reported activate MKK4 (Kyriakis and Avruch, 2001; Takekawa that Mkk4 ablation impaired the development of both et al., 2005). This is consistent with the ability of B- and T-cell lineages and that the mice had a greatly MEKK1 to bind to the C-terminal DVD sites on both reduced thymus size (Nishina et al., 1997a, b). Specific MKK4 and MKK7, whereas MEKK4 specifically defects included reduced numbers of CD4 þ /CD8 þ interacts with MKK4 (Takekawa et al., 2005). In double-positive (DP) immature thymocytes, increased addition to colocalizing the MKKKs with MKK4, these sensitivity to apoptosis triggered by CD95 or CD3 interactions may alter the conformation of MKK4 to crosslinking of both the DP thymocytes and peripheral enhance the accessibility of the Ser/Thr residues within T cells, and no transition from pro-B to pre-B cells in the the T-loop (Takekawa et al., 2005). A model of signal bone marrow (Nishina et al., 1997a, b). In contrast, a transmission via MEKK1, MKK4 and JNK has been second group found no evidence that MKK4 was

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3175 required for the development of T and B lymphocytes or derived from pancreas, breast, colon and testis (Teng for protecting thymocytes from cell death (Swat et al., et al., 1997). Of these five intragenic mutations, four 1998). Instead, the mice exhibited lymphadenopathy resulted in the inability of MKK4 to phosphorylate and polyclonal B- and T-cell expansions suggesting JNK in vitro indicating that they were loss of function an important role for MKK4 in maintaining peripheral mutations (Teng et al., 1997). Indeed, the mutations lymphoid homeostasis (Swat et al., 1998). The discre- result in either the premature termination of the protein pancies between the studies could be owing to dif- product, thereby eliminating critical protein kinase ferences in the abilities of the homozygous Mkk4À/À subdomains, or amino-acid substitutions in functionally ES clones being used to reconstitute B- and T-cell important residues that are conserved among the MKK lineages in the context of RAG2-deficient blastocyst family members (Teng et al., 1997). A role for MKK4 complementation. The generation of mice with specific as a tumor suppressor is supported by experiments inactivation of Mkk4 in T and B lymphocytes should demonstrating that a dominant-negative mutant of help to resolve these issues. MKK4 promoted ES cell transformation and enhanced The MKK4 signaling pathway may also contribute to the tumorigenicity of ES cells injected into athymic nude cardiac hypertrophy, a process involving alterations in mice (Cazillis et al., 2004). the morphology of cardiomyocytes and the extracellular Since the original report investigating MKK4 muta- matrix to compensate for systolic wall stress caused by tions in cancer cell lines (Teng et al., 1997), a number of an increased workload. If this occurs over a prolonged studies have reported loss-of-function mutations in period it leads to wall thickening of the heart, chamber MKK4 at a fairly consistent rate (B5%) across a wide dilation and myocardial dysfunction. The expression of spectrum of primary cancers including those of the a dominant-negative mutant form of MKK4 in the rat pancreas, bile duct, breast, prostate and ovary (Su et al., heart leads to reduced JNK activity and a reduced 1998, 2002; Kim et al., 2001; Xin et al., 2004; Nakayama hypertrophic response following pressure overload, et al., 2006). Missense and nonsense mutations in the implicating the MKK4–JNK pathway as a key pathway MKK4 gene have also been identified in lung tumors controlling cardiac hypertrophy (Choukroun et al., in large-scale tumor screens (COSMIC: http://www. 1999). However, the expression of dominant-negative sanger.ac.uk/genetics/CGP/cosmic/). In pancreatic can- mutants has the potential to affect multiple pathways; cers there was a correlation between the loss of MKK4 therefore, further investigations are required including protein and shorter survival times, with the MKK4- the conditional deletion of the Mkk4 gene in mouse positive carcinomas carrying half the risk of death cardiomyocytes. compared to MKK4-negative carcinomas (Xin et al., 2004). It is likely that MKK4 may participate in a distinct tumor suppressive signaling pathway owing to the presence of coexistent mutations in other tumor MKK4 in cancer suppressors including p53, BRCA2, DPC4 and p16INK4a (Su et al., 1998). Over the past decade, a number of studies have supported a role for MKK4 in regulating steps in the development of cancers. Many studies propose that MKK4 as a metastasis suppressor MKK4 is a tumor suppressor and a suppressor of meta- While loss of function of MKK4 may play a role in the stasis, whereas other studies support a pro-oncogenic formation of some primary tumors, it may also be role for MKK4. This suggests a complex role for MKK4 linked with more advanced stages of cancer progression. and its downstream targets JNK and p38 in cancer The impaired expression of MKK4 in prostate and development. ovarian tumors appears to promote their metastasis (Yoshida et al., 1999; Yamada et al., 2002), while MKK4 mutations in cancer cells reduced MKK4 mRNA levels have been reported in The loss or inactivation of tumor suppressor genes breast cancer to brain metastases (Stark et al., 2005). In promotes cancer formation and progression and can normal prostate tissue there are high levels of MKK4 occur through loss of heterozygosity (LOH) or follow- protein expression in the epithelial compartment but not ing homozygous gene deletion. The human MKK4 gene in the stromal compartment, whereas in neoplastic is located on chromosome 17p11.2 and lies centromeric prostate tissues the levels of MKK4 were reduced and to the p53 tumor suppressor gene (Yoshida et al., 1999). there was in inverse relationship between the reduction This arm of chromosome 17 is one of the most of MKK4 expression and its metastatic potential (Kim frequently deleted in humans. A role for MKK4 as a et al., 2001). Experiments using the highly metastatic rat tumor suppressor was first proposed by Teng et al. prostate cancer cell line AT6.1 (which lacks MKK4 (1997). Two tumor cell lines derived from pancreatic expression) as a model system have demonstrated that carcinoma and lung carcinoma were identified which the overexpression of MKK4 significantly reduces their harbored homozygous deletions that eliminated coding metastatic ability (Yoshida et al., 1999). Severe com- portions of the MKK4 locus (Teng et al., 1997). In bined immunodeficient (SCID) mice were injected with addition, 88 cancer cell lines that had been pre-screened either the parental AT6.1 cells, AT6.1 cells overexpres- for LOH were investigated and two nonsense mutations sing MKK4 (AT6.1-Mkk4) or cells overexpressing a and three missense mutations were identified in cell lines kinase-inactive mutant of MKK4 (AT6.1-Mkk4(KR))

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3176 and examined for lung metastases. The AT6.1-Mkk4 ectopic expression of the p38 activator MKK6 also mice displayed fewer macroscopic lung metastases and suppressed metastasis whereas expression of MKK7 did survived longer than the mice injected with the parental not, suggesting that p38, rather than JNK, may be the line or the AT6.1-Mkk4(KR) cells, indicating that the relevant MKK4 target in ovarian cancer (Hickson et al., kinase activity of MKK4 is essential for its metastasis 2006; Figure 2). suppression function (Yoshida et al., 1999; Vander Taken together, these studies suggest that MKK4 Griend et al., 2005) (Figure 2). Micrometastatic foci functions as a metastasis suppressor and may utilize were detected in the lungs of AT6.1-Mkk4 mice distinct MAP kinases depending on the environmental suggesting that cells that have escaped from the primary context. It is likely that in different tissues and organs tumor are growth inhibited in the lung (Yoshida et al., there may be distinct stimuli that dictate which MAP 1999). A specific role for MKK4 in metastasis suppres- kinase pathway is targeted by MKK4. sion is further supported by the fact that the growth rate of the primary tumor was not affected (Yoshida et al., 1999). Pro-oncogenic role of MKK4 Further complementation experiments demonstrated Although there is an increasing body of evidence to that the ectopic expression of the second JNK activator support a role for MKK4 in tumor or metastasis MKK7, but not of the p38 activator MKK6, could suppression, there are also studies pointing to a pro- suppress metastasis by inhibiting the ability of AT6.1 oncogenic role for MKK4. It is reported that in breast cells to colonize the lung (Vander Griend et al., 2005) and pancreatic cancer cell lines that lack endogenous (Figure 2). This suggests that the JNK pathway, rather MKK4, the ectopic expression of MKK4 stimulates cell than the p38 pathway, is mediating the metastasis proliferation and invasion (Wang et al., 2004). Con- suppression by MKK4. Currently no mutations in the versely, the knock down of MKK4 expression by siRNA MKK7 gene in human tumors have been reported. in the MKK4-positive breast cancer cell line MDA-MB- MKK4 protein expression is also reduced in ovarian 231 results in decreased anchorage-independent growth, metastatic tissues compared to normal ovarian epithelial increased susceptibility to apoptosis upon serum starva- cells (Yamada et al., 2002). To demonstrate a potential tion and suppressed tumor growth in a mouse xenograft role for MKK4 in metastasis suppression in ovarian model (Wang et al., 2004). Additional evidence for a cancer, a similar complementation approach was used as role of MKK4 in cell proliferation comes from the that described above for examining prostate cancer demonstration that the expression of a dominant- metastasis. MKK4 was ectopically expressed in the negative mutant of MKK4 in H1299 non-small-cell human ovarian cell line SKOV3ip.1, which lacks lung cancer (NSCLC) cells cooperated with the inhibi- endogenous MKK4 expression, and injected into SCID tion of the phosphatidylinositol 3-kinase signaling mice. This led to a significant decrease in overt pathway to block cell proliferation and reduce the size metastatic implants on a number of tissues and organs of H1299 NSCLC xenograft tumors (Lee et al., 2005), compared to parental cells and increased the life span of whereas the overexpression of a constitutively active the mice by 70% (Yamada et al., 2002). In contrast to mutant of MKK4 in human bronchial epithelial cell the prostate metastasis model discussed above, the lines increased their proliferation and invasive properties (Khatlani et al., 2006). A recent study using a pancreatic cancer cell line PL5 (Panc 4.03) featuring the targeted disruption of the Prostate Ovary MKK4 gene has provided further support for a pro- oncogenic function of MKK4 (Cunningham et al., 2006). Intravenous injection of the parental cells or MKK4 MKK4 MKK4 þ /À cells into mice led to numerous lung metastases, whereas the mice injected with the MKK4À/À cells had very few lung metastases (Cunning- ham et al., 2006). When the mice were injected MKK7 JNK p38 MKK6 subcutaneously with MKK4À/À cells, the resulting tumors had a longer tumor volume doubling rate compared to mice injected with the parental or MKK4 þ /À cells, suggesting that MKK4 promotes Lung metastasis tumor growth (Cunningham et al., 2006). JNK activation, but not p38 activation, was compromised in the Figure 2 MKK4 signaling pathways play distinct roles in MKK4À/À cells indicating that MKK4 may be prima- metastasis to lung from different tissues. The injection of mice with the metastatic prostate cancer cell line AT6.1 or the ovarian rily acting through JNK to promote tumor growth cancer cell line SKOV3ip.1 overexpressing MKK4 significantly (Cunningham et al., 2006). reduces their metastatic ability compared to the parental cells Taken together, these studies suggest that the role of suggesting a metastasis suppressor role for MKK4. Based on MKK4 in regulating cancer development is highly complementation analysis with other MKKs in the JNK (MKK7) and p38 (MKK6) pathways the metastasis suppressor effect of context dependent and can vary according to tissue MKK4 in prostate is mediated by JNK whereas in ovary it is type, the environmental conditions and by interactions mediated by p38. with other intracellular signaling pathways.

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3177 Downstream targets of MKK4 in cancer reported in several cancer cell lines (Kennedy and Davis, 2003). In a Drosophila model of tumor forma- The protein kinase activity of MKK4 is required for its tion, Ras and JNK have been demonstrated to be various reported roles in cancers. The major phosphor- cooperative (Uhlirova et al., 2005), whereas in mice ylation targets of MKK4 are the MAP kinases JNK and carrying c-Jun with mutated JNK phosphorylation sites, p38, suggesting they are likely to be required for c-Fos-induced osteosarcomas and skin tumors in mediating the effects of alterations in MKK4 expression response to constitutive Ras activation are reduced or activity. Not surprisingly, considering the complexity (Behrens et al., 2000). A further example has been of MKK4 involvement in cancer, there appears to be demonstrated in the Apc(Min) mouse model of intesti- parallel complexities in the role of JNK in cellular nal cancer where the loss of c-Jun N-terminal phos- transformation and tumor development (Kennedy and phorylation leads to reduced tumor number and Davis, 2003). The role of p38 in cancer has received less increased lifespan (Nateri et al., 2005). The proposed attention, but there is an increasing body of evidence mechanism involves phosphorylated c-Jun forming suggesting that it participates in tumor suppression transcriptionally active complexes with TCF4 and b- (Bulavin and Fornace Jr, 2004). In this section, we catenin (Nateri et al., 2005). Interestingly in liver, JNK discuss how these MAP kinase pathways may regulate promotes chemically induced hepatocarcinogenesis in- cancer development downstream of MKK4. dependently of c-Jun phosphorylation, suggesting other important targets exist (Eferl et al., 2003; Sakurai et al., 2006). Role of JNK in cancer Nuclear hormone receptors may also represent Over the past decade many studies have implicated JNK important targets of the JNK pathway in controlling in promoting cellular transformation by oncogenes cancer development. There is evidence that retinoid including Ras, c-fos, Met and BCR-Abl, as well as receptors can suppress tumorigenesis and that the loss of epidermal growth factor (Smeal et al., 1991; Rodrigues specific receptors promotes carcinogenesis in many et al., 1997; Bost et al., 1999; Behrens et al., 2000; Xiao tissues (Altucci and Gronemeyer, 2001). Retinoid acid and Lang, 2000; Hess et al., 2002). Ras is activated by receptor a and retinoid X receptor a (RXRa) are mutation in approximately 30% of human cancers and phosphorylated by JNK leading to inhibition of retinoic cooperates with the oncogene c-Jun, encoding a acid (RA)-mediated transcriptional events (Adam-Stitah transcription factor target of JNK, to enhance cellular et al., 1999; Lee et al., 2000; Mann et al., 2005; Srinivas transformation (Schutte et al., 1989; Johnson et al., et al., 2005). Surprisingly, RXRa is also directly 1996). Indeed, Ras is unable to transform c-Jun null phosphorylated by MKK4 on Tyr residues leading to fibroblasts (Johnson et al., 1996). The role of JNK the suppression of RA-mediated transcription (Lee phosphorylation of c-Jun in cellular transformation is et al., 2000). Currently, RXRa represents the only less clear. Ras induces the phosphorylation of the JNK MKK4 substrate identified apart from the JNK and p38 sites in c-Jun and these sites have been demonstrated to MAP kinases. These data suggest that the MKK4-JNK be required for efficient co-transformation activity with pathway antagonizes the tumor suppression function of Ras (Smeal et al., 1991; Derijard et al., 1994). Moreover, RA signaling and may therefore promote tumorigenesis. the tumor suppressor p16INK4a is proposed to interfere While the studies described above support a role for with Ras-c-Jun induced cell transformation by binding the JNK pathway in tumorigenesis, there is also an to JNK and inhibiting its phosphorylation of c-Jun increasing body of evidence that JNK suppresses tumor (Choi et al., 2005), while fibroblasts harboring c-Jun development. The intravenous injection into athymic mutated at the JNK phosphorylation sites are resistant mice of Ras transformed wild type and JNK-null cells to Ras-induced transformation (Behrens et al., 2000). led to the formation of tumors in both sets of mice but a However, the interpretation of the latter result is greatly increased tumor burden in the mice injected with complicated by evidence that the transformation-resis- JNK-null cells (Kennedy et al., 2003). These mice tant phenotype caused by the removal of the JNK sites displayed an increased number of tumor nodules and can be reversed by mutating the C-terminal negative increased tumor size. This study supports a model regulatory phosphorylation site of c-Jun that is targeted whereby JNK suppresses tumor formation induced by by glycogen synthase kinase-3 (Bost et al., 2001). In oncogenic Ras (Kennedy et al., 2003). A probable addition, in some cell types ERK MAP kinase, rather mechanism by which MKK4 and JNK act as tumor than JNK, may mediate Ras-induced c-Jun phospho- suppressors is through apoptosis. Indeed, in the Ras- rylation (Pulverer et al., 1991; Leppa et al., 1998). induced JNK-null tumors very few apoptotic cells were Furthermore, studies using JNK-null fibroblasts demon- detected compared to the wild-type tumors (Kennedy strated only a modest decrease in growth on soft agar et al., 2003). The JNK pathway is a well characterized compared to wild-type cells in response to oncogenic mediator of apoptosis (Davis, 2000). In fibroblasts it is Ras, suggesting a minor contribution of JNK to Ras- required for stress-induced cytochrome c release from induced transformation in these cells (Kennedy et al., mitochondria, a key driver of apoptosis (Tournier et al., 2003). Taken together these studies indicate a context- 2000). Further work to elucidate the mechanisms dependent role for JNK in cellular transformation. involved has demonstrated a JNK-dependent require- JNK is also proposed to play a role in tumor ment for the proapoptotic Bcl2 family members Bax and development and high levels of JNK activity are Bak in cytochrome c release (Lei et al., 2002). The

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3178 regulation of Bax and Bak may occur by JNK a JNK phosphorylation of two additional Bcl2 family proteins, Bcl2 Bim and Bmf (Lei and Davis, 2003; Putcha et al., 2003; c-Jun Bcl-XL Figure 3a). This leads to their release from sequestration Bim/Bmf in dynein and myosin V complexes and the promotion Mcl-1 of Bax and Bak activation by an unknown mechanism 14-3-3 TGFβ1 (Lei and Davis, 2003). Recently, it was reported that Bax Bax itself may be a direct target of JNK and p38 signaling pathways and that these pathways promote Tumor Bax translocation to the mitochondria before apoptosis Cyt c development (Kim et al., 2006). JNK may also contribute to Bax translocation to the mitochondria through the phosphorylation of members of the 14-3-3 family. Bax is Apoptosis normally sequestered in the cytoplasm by binding to 14- 3-3 but upon stress JNK phosphorylates 14-3-3 leading b p38 to the dissociation of Bax and its translocation to the mitochondria (Tsuruta et al., 2004). In addition to the MAPKAPK2 phosphorylation of proapoptotic Bcl2 family members, JNK phosphorylates and inactivates the anti-apoptotic Bcl2 family proteins Bcl2, Bcl-XL and Mcl-1, thus CyclinD1/CDK4 CDC25 p53 further contributing to Bax activation and apoptosis (Maundrell et al., 1997; Yamamoto et al., 1999; Fan et al., 2000; Deng et al., 2001; Inoshita et al., 2002; Figure 3a). Apoptosis Recently, it has been proposed that JNK may act as a G1/S G2/M tumor suppressor by regulating the autocrine expression of TGF-b1 (Ventura et al, 2004; Figure 3a). Increased expression levels of TGF-b1 have previously been Tumor growth reported to contribute to cancer progression and the inhibition of TGF-b1 signaling blocks tumor growth in Figure 3 The JNK and p38 signaling pathways contribute to mice injected with Ras transformed cells (Ventura et al, tumor suppression via several mechanisms. (a) JNK promotes Bax activation and apoptosis via direct phosphorylation and also 2004; Bierie and Moses, 2006). JNK-null fibroblasts through the phosphorylation of the proapoptotic Bcl2 family display increased TGF-b1 levels compared with wild- members Bim and Bmf. JNK also phosphorylates and inactivates type cells and this correlates with increased invasive of the antiapoptotic family members Bcl2, Bcl-XL and Mcl-1, as behavior and proliferation (Ventura et al, 2004). This well as inhibiting cytoplasmic sequestration of Bax by 14-3-3. In effect was attributable to a distal promoter region in addition, JNK inhibits TGFb1 gene transcription via c-Jun. (b) p38 negatively regulates cyclin D1 both by reducing gene transcription the TGF-1b gene that binds to c-Jun and represses and by promoting protein instability and thereby blocking the G1/ transcription by recruiting the histone deacetylase S transition. The p38 pathway, via MAPKAPK2, also down- HDAC3 in wild-type cells, but has reduced binding to regulates the activity of CDC25 family members thereby inhibiting c-Jun and HDAC3 in JNK-null cells. This suggests that cell cycle progression. The tumor suppressor p53 is a direct target of p38 phosphorylation, which promotes its stability and transcrip- JNK phosphorylation of c-Jun is required for c-Jun/ tional activation leading to cell cycle arrest and apoptosis. AP-1 recruitment to the TGF-1b promoter. Conversely, SMAD3/SMAD4 recruitment to the promoter is enhanced in the JNK-null cells and may contribute to the auto-induction of TGF-1b gene expression (Ventura this model JNK1 has a tumor suppressor role and that et al., 2004). However, multiple levels of crosstalk exist JNK2 promotes tumorigenesis. Further support for a between the TGF-b and JNK signaling pathways. For specific role of JNK1 in tumor suppression comes from example, JNK can also positively contribute to TGF-b- the demonstration that it is required for tumor induced gene expression via regulation of AP-1 activity surveillance by the immune system (Gao et al., 2005), and JNK can target SMAD2 and SMAD3 for whereas a preferential role for JNK2 in tumorigenesis is phosphorylation leading to the upregulation of TGF- supported by studies in glioblastoma, prostate, and lung b-responsive genes (Mori et al., 2004). carcinoma cell lines (Bost et al., 1999; Potapova et al., The different isoforms of JNK may have distinct roles 2000; Yang et al., 2003; Cui et al., 2006). The molecular in tumorigenesis. Although the ubiquitously expressed mechanism underlying these opposing actions of JNK1 JNK1 and JNK2 display significant functional redun- and JNK2 are unclear as unique substrates for dancy in many cellular processes, there is evidence that individual JNK isoforms have not been uncovered. It they have distinct roles in skin tumor development. has been proposed, based on experiments using JNK1- JNK1-null mice display enhanced skin tumor develop- or JNK2-deficient fibroblasts, that JNK1 and JNK2 ment in response to phorbol ester whereas there is differentially regulate c-Jun stability and transcriptional suppression of skin tumorigenesis in JNK2-null mice activity depending upon their activation state (Saba- (Chen et al., 2001; She et al., 2002). This suggests that in pathy et al., 2004). However, an alternative explanation

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3179 is based on the observation that JNK1 activity increases overexpressed (Bulavin et al., 2002), whereas in Wip1- in the absence of JNK2 protein, but not upon inhibition null mice, p38 activity and p16INK4a levels are increased of JNK2 activity (Jaeschke et al., 2006). Using a and the mice are protected from the early onset of chemical genetic approach it was demonstrated that mammary tumors in response to Ras, Neu and ErbB2 JNK1 and JNK2 are both positive regulators of c-Jun, oncogenes (Bulavin et al., 2004; Demidov et al., 2006). rather than being antagonistic (Jaeschke et al., 2006). It The addition of a pharmacological inhibitor of p38 is possible, therefore, that the increased tumor suppres- activity to the Wip1-null mice reduced p16INK4a expression observed in JNK2-null mice could be owing to sion and led to the formation of mammary tumors increased JNK1 activity, rather than owing to the loss of (Bulavin et al., 2004), whereas mice bearing a constitu- JNK2. tively active form of the p38 activator MKK6 suppressed The third JNK isoform, JNK3, is mainly expressed in the tumor-prone phenotype of mice overexpressing Wip1 brain and testis and has functions distinct from JNK1 in mammary epithelium (Demidov et al., 2006). and JNK2 (Yang et al., 1997a). One study has reported The members of the CDC25 protein phosphatase that 10 out of 19 human brain tumors that were family are also targets of p38. CDC25A is overexpressed examined contained mutations in the JNK3 gene in over 40% of primary human breast cancers and (Yoshida et al., 2001). Although this suggests that activates CDK2 by removing inhibitory phosphoryla- JNK3, like MKK4, may also be a tumor suppressor tion sites (Bartek and Lukas, 2001). Osmotic stress leads gene, further studies are required to demonstrate a direct to the phosphorylation of CDC25A on Ser75 resulting effect of JNK3 loss on brain tumor development. in its degradation and thereby suppressing CDC25A activity and potentially promoting tumor suppression (Goloudina et al., 2003). This effect is mediated by the Role of p38 in cancer p38 pathway, although it is unclear which protein kinase The role of members of the p38 family in cancer is less is responsible for directly phosphorylating this site in well established, but there is increasing evidence that p38 CDC25A (Goloudina et al., 2003). The p38 pathway can may act as a tumor suppressor (Bulavin and Fornace Jr, also negatively regulate the G2/Mcell cycle transition by 2004). For example, MEFs lacking the p38 activators phosphorylating the other members of this protein MKK3 and MKK6 display defects in serum starvation- phosphatase family. CDC25B is phosphorylated at induced growth arrest and the subcutaneous injection of Ser309 and CDC25C at Ser216, which triggers their SV40-large-T-antigen immortalized Mkk3À/ÀMkk6À/À binding and sequestration by 14-3-3 proteins and fibroblasts into athymic mice results in a significantly thereby prevents activation of the Cdc2, an important increased tumor burden compared to the injection of kinase that drives the G2/Mtransition (Bulavin et al., wild-type cells (Brancho et al., 2003). p38 has been 2001). Initially, it was proposed that p38 might directly demonstrated to negatively regulate cell cycle progres- phosphorylate CDC25B and CDC25C (Bulavin et al., sion and proliferation through several mechanisms. 2001), but recent studies supported by genetic studies in Targets of p38 signaling include cyclin D1, CDC25 and yeast, indicate that MAPK-activated protein kinase-2 p53 (Bulavin and Fornace Jr, 2004; Figure 3b). Cyclin (MAPKAPK2), a kinase downstream of p38, is respon- D1 collaborates with cyclin-dependent kinase 4 (CDK4) sible for the phosphorylations in vivo (Lopez-Aviles to regulate the G1/S cell cycle transition and the loss of et al., 2005; Manke et al., 2005) (Figure 3b). These regulation of cyclin D1 expression levels can contribute studies suggest that p38 can block cell cycle progression to tumor formation (Ortega et al., 2002). Cyclin D1 via interactions with distinct members of the CDC25 levels can be regulated by p38 both at the level of gene family and thereby contribute to tumor suppression. transcription and post-translationally (Lavoie et al., The tumor suppressor p53 plays a central role in many 1996; Casanovas et al., 2000; Brancho et al., 2003). p38 cellular events including regulating cell cycle check- negatively regulates the Cyclin D1 promoter (Lavoie points, apoptosis and genomic stability (Vogelstein et al., et al., 1996; Brancho et al., 2003), potentially via 2000). In response to stress, particularly genotoxic stress, phosphorylation and stabilization of the HBP1 repressor p53 is stabilized by dissociation from ubiquitin ligases, protein (Xiu et al., 2003), whereas the cyclin D1 protein such as murine double minute 2 (MDM2), and by Ser/ is phosphorylated at Thr286 by p38 under certain stress Thr phosphorylation at multiple sites which also conditions and this leads to its proteasome-dependent contribute to its transcriptional activity (Vogelstein degradation (Casanovas et al., 2000). The expression of a et al., 2000). The activation of the p38 pathway leads cyclin D1 mutant that cannot be phosphorylated at to enhanced p53 transcriptional activity, which can lead Thr286 promotes cellular transformation of fibroblasts to cell cycle arrest and apoptosis (Bulavin et al., 1999; and tumor growth in mice (Alt et al., 2000). p38 also Takekawa et al., 2000; Figure 3b). p38 phosphorylates indirectly affects cyclin D1/CDK4 activity through human p53 at Ser33 and Ser46 in response to several upregulating the expression of the cyclin-dependent stresses and may indirectly lead to p53 phosphorylation kinase inhibitors p16Ink4a and p19Arf (Bulavin et al., at Ser392 via the activation of casein kinase-2 (Bulavin 2004). p38 regulation of these cyclin-dependent kinase et al., 1999; Bulavin and Fornace Jr, 2004). The inhibitors is controlled by p53-induced phosphatase-1 phosphorylation of Ser33 and Ser46 of p53 is required (Wip1), although the precise mechanism is unclear for the recruitment of the prolyl isomerase Pin1 which (Bulavin et al., 2004). It is reported that there is contributes to p53 stability and function (Zacchi et al., decreased p38 activity in human cancers where Wip1 is 2002; Zheng et al., 2002). Interestingly, JNK can also

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3180 phosphorylate p53 and regulate its stability but it recent genetic studies in mice have provided important remains unclear how JNK contributes to speciﬁc p53 information on their in vivo functions. There is increas- functions (Buschmann et al., 2001). ing evidence to support an important role for MKK4 Similar to the case of p53, both p38 and JNK can pathways in controlling cancer development in humans. phosphorylate the androgen receptor (AR) at Ser650 and A function for MKK4 as a tumor suppressor is promote its nuclear export in prostate cancer cells thereby supported by the fact that the MKK4 gene is mutated reducing AR-mediated transcription (Gioeli et al., 2006). at a frequency of 5% in a number of cancer types and It was also demonstrated that decreasing MKK4 or that complementation experiments support a role in the MKK6 expression levels by siRNA led to increased AR suppression of metastasis from prostate and ovary. In nuclear accumulation and transactivation. (Gioeli et al., contrast, studies in other cell types suggest that MKK4 2006). As the AR plays a key role in the progression of and JNK participate in tumor promotion. To more prostate cancer, then loss of function mutants of MKK4 directly address the role of MKK4 in tumor suppression may hypersensitize the AR to androgen and promote the or promotion, appropriate mouse models featuring androgen-independent diseased state. tissue-speciﬁc Mkk4 deletion or mutations of the The studies discussed here indicate a potential tumor Mkk4 gene will need to be generated and characterized. suppressor role for p38. However, there is also limited These studies will need to be correlated with similar evidence that p38 contributes to lung metastasis of studies to address which branches of the MKK4 tumor cells. A recent study using mice heterozygous for pathway are relevant in particular cancer types. A full the p38a isoform found that these mice had markedly understanding of the role of MKK4 and its downstream fewer colonies of tumor cells in lungs in an in vivo targets will be required for designing strategies for metastasis assay (Matsuo et al., 2006). Therefore p38, tumor therapy using small molecule inhibitors of like JNK, may play important roles in both tumor MKK4 or the JNK and p38 MAP kinases. suppression and oncogenesis downstream of MKK4. Acknowledgements

Concluding remarks We thank A Sharrocks and S-H Yang for comments on the manuscript. AJW is a Lister Institute-Jenner Research Fellow. The signaling pathways featuring MKK4, JNK and p38 RJD is an Investigator of the Howard Hughes Medical have been studied intensely at the molecular level, and Institute.

References

Adam-Stitah S, Penna L, Chambon P, Rochette-Egly C. human tumors abrogates p53 tumor-suppressor activity. (1999). Hyperphosphorylation of the retinoid X receptor Nat Genet 31: 210–215. a by activated c-Jun NH2-terminal kinase. J Biol Chem 274: Bulavin DV, Fornace Jr AJ. (2004). p38 MAP kinase’s emer- 18932–18941. ging role as a tumor suppressor. Adv Cancer Res 92: 95–118. Alt JR, Cleveland JL, Hannink M, Diehl JA. (2000). Bulavin DV, Higashimoto Y, Gaarde WA, Basrur V, Phosphorylation-dependent regulation of cyclin D1 nuclear Potapova O, Appella E et al. (2001). Initiation of a G2/M export and cyclin D1-dependent cellular transformation. checkpoint after ultraviolet radiation requires p38 kinase. Genes Dev 14: 3102–3114. Nature 411: 102–107. Altucci L, Gronemeyer H. (2001). The promise of retinoids to Bulavin DV, Phillips C, Nannenga B, Timofeev O, Donehower ﬁght against cancer. Nat Rev Cancer 1: 181–193. LA, Anderson CW et al. (2004). Inactivation of the wip1 Bartek J, Lukas J. (2001). Pathways governing G1/S transition phosphatase inhibits mammary tumorigenesis through p38 and their response to DNA damage. FEBS Lett 490: 117–122. MAPK-mediated activation of the Ink4a/Arf pathway. Nat Behrens A, Jochum W, Sibilia M, Wagner EF. (2000). Genet 36: 343–350. Oncogenic transformation by ras and fos is mediated by Bulavin DV, Saito S, Hollander MC, Sakaguchi K, c-Jun N-terminal phosphorylation. Oncogene 19: 2657–2663. Anderson CW, Appella E et al. (1999). Phosphorylation Bierie B, Moses HL. (2006). Tumour microenvironment: of human p53 by p38 kinase coordinates N-terminal TGFbeta: the molecular Jekyll and Hyde of cancer. Nat phosphorylation and apoptosis in response to UV radiation. Rev Cancer 6: 506–520. EMBO J 18: 6845–6854. Bost F, Caron L, Vial E, Montreau N, Marchetti I, Dejong V Buschmann T, Potapova O, Bar-Shira A, Ivanov VN, et al. (2001). The defective transforming phenotype of c-Jun Fuchs SY, Henderson S et al. (2001). Jun NH2-terminal Ala(63/73) is rescued by mutation of the C-terminal kinase phosphorylation of p53 on Thr-81 is important for phosphorylation site. Oncogene 20: 7425–7429. p53 stabilization and transcriptional activities in response to Bost F, McKay R, Bost M, Potapova O, Dean NM, Mercola D. stress. Mol Cell Biol 21: 2743–2754. (1999). The Jun kinase 2 isoform is preferentially required Carboni L, Tacconi S, Carletti R, Bettini E, Ferraguti F. for epidermal growth factor-induced transformation (1997). Localization of the messenger RNA for the c-Jun of human A549 lung carcinoma cells. Mol Cell Biol 19: NH2-terminal kinase kinase in the adult and developing 1938–1949. rat brain: an in situ hybridization study. Neuroscience 80: Brancho D, Tanaka N, Jaeschke A, Ventura JJ, Kelkar N, 147–160. Tanaka Y et al. (2003). Mechanism of p38 MAP kinase Casanovas O, Miro F, Estanyol JM, Itarte E, Agell N, activation in vivo. Genes Dev 17: 1969–1978. Bachs O. (2000). Osmotic stress regulates the stability of Bulavin DV, Demidov ON, Saito S, Kauraniemi P, Phillips C, cyclin D1 in a p38SAPK2-dependent manner. J Biol Chem Amundson SA et al. (2002). Ampliﬁcation of PPM1D in 275: 35091–35097.

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3181 Cazillis M, Bringuier AF, Delautier D, Buisine M, Bernuau D, Ganiatsas S, Kwee L, Fujiwara Y, Perkins A, Ikeda T, Gespach C et al. (2004). Disruption of MKK4 signaling Labow MA et al. (1998). SEK1 deficiency reveals mitogen- reveals its tumor-suppressor role in embryonic stem cells. activated protein kinase cascade crossregulation and leads Oncogene 23: 4735–4744. to abnormal hepatogenesis. Proc Natl Acad Sci USA 95: Chang L, Karin M. (2001). Mammalian MAP kinase signaling 6881–6886. cascades. Nature 410: 37–40. Gao Y, Tao J, Li MO, Chi H, Henegariu O, Kaech SM et al. Chen N, Nomura M, She QB, Ma WY, Bode AM, Wang L (2005). JNK1 is essential for CD8+ T cell-mediated tumor et al. (2001). Suppression of skin tumorigenesis in c-Jun immune surveillance. J Immunol 175: 5783–5789. NH(2)-terminal kinase-2-deficient mice. Cancer Res 61: Gioeli D, Black BE, Gordon V, Spencer A, Kesler CT, 3908–3912. Eblen ST et al. (2006). Stress kinase signaling regulates Choi BY, Choi HS, Ko K, Cho YY, Zhu F, Kang BS et al. androgen receptor phosphorylation, transcription, and (2005). The tumor suppressor p16INK4a prevents cell localization. Mol Endocrinol 20: 503–515. transformation through inhibition of c-Jun phosphorylation Goloudina A, Yamaguchi H, Chervyakava DB, Appella E, and AP-1 activity. Nat Struct Mol Biol 12: 699–707. Fornace Jr AJ, Bulavin DV. (2003). Regulation of human Choukroun G, Hajjar R, Fry S, del Monte F, Haq S, Cdc25A stability by serine phosphorylation is not sufficient Guerrero JL et al. (1999). Regulation of cardiac hypertro- to activate a S-phase checkpoint. Cell Cycle 2: 473–478. phy in vivo by the stress-activated protein kinases/ Ha HY, Cho IH, Lee KW, Lee KW, Song JY, Kim KS et al. c-JunNH(2)-terminal kinases. J Clin Invest 104: 391–398. (2005). The axon guidance defect of the telencephalic Coffey ET, Hongisto V, Dickens M, Davis RJ, Courtney MJ. commissures of the JSAP1-deficient brain was partially (2000). Dual roles for c-Jun N-terminal kinase in develop- rescued by the transgenic expression of JIP1. Dev Biol 277: mental and stress responses in cerebellar granule neurons. 184–199. J Neurosci 20: 7602–7613. Han ZS, Enslen H, Hu X, Meng X, Wu IH, Barrett T et al. Cuenda A. (2000). Mitogen-activated protein kinase kinase 4 (1998). A conserved p38 mitogen-activated protein kinase (MKK4). Int J Biochem Cell Biol 32: 581–587. pathway regulates Drosophila immunity gene expression. Cui J, Han SY, Wang C, Su W, Harshyne L, Holgado- Mol Cell Biol 18: 3527–3539. Madruga M et al. (2006). c-Jun NH2-terminal kinase 2a2 Hanks SK, Quinn AN, Hunter T. (1988). The protein kinase promotes the tumorigenicity of human glioblastoma cells. family: conserved features and deduced phylogeny of the Cancer Res 66: 10024–10031. catalytic domains. Science 241: 42–52. Cunningham SC, Gallmeier E, Hucl T, Dezentje DA, Hess P, Pihan G, Sawyers CL, Flavell RA, Davis RJ. (2002). Calhoun ES, Falco G et al. (2006). Targeted deletion of Survival signaling mediated by c-Jun NH(2)-terminal kinase MKK4 in cancer cells: a detrimental phenotype manifests as in transformed B lymphoblasts. Nat Genet 32: 201–205. decreased experimental metastasis and suggests a counter- Hickson JA, Huo D, Vander Griend DJ, Lin A, Rinker- weight to the evolution of tumor-suppressor loss. Cancer Schaeffer CW, Yamada SD. (2006). The p38 kinases MKK4 Res 66: 5560–5564. and MKK6 suppress metastatic colonization in human Davis RJ. (2000). Signal transduction by the JNK group of ovarian carcinoma. Cancer Res 66: 2264–2270. MAP kinases. Cell 103: 239–252. Ho DT, Bardwell J, Abdollahi M, Bardwell L. (2003). A Demidov ON, Kek C, Shreeram S, Timofeev O, Fornace Jr docking site in MKK4 mediates high affinity binding to AJ, Appella E et al. (2006). The role of the MKK6/p38 JNK MAPKs and competes with similar docking sites in MAPK pathway in Wip1-dependent regulation of ErB2- JNK substrates. J Biol Chem 278: 32662–32672. driven mammary gland tumorigenesis. Oncogene (in press). Inoshita S, Takeda K, Hatai T, Terada Y, Sano M, Hata J Deng X, Xiao L, Lang W, Gao F, Ruvolo P, May WS, Jr. et al. (2002). Phosphorylation and inactivation of myeloid (2001). Novel role for JNK as a stress-activated Bcl2 kinase. cell leukemia 1 by JNK in response to oxidative stress. J Biol J Biol Chem 276: 23681–23688. Chem 277: 43730–43734. Derijard B, Hibi M, Wu IH, Barrett T, Su B, Deng T et al. Jaeschke A, Karasarides M, Ventura J-J, Ehrhardt A, Zhang C, (1994). JNK1: a protein kinase stimulated by UV light and Flavell RA et al. (2006). JNK2 is a positive regulator of the Ha-Ras that binds and phosphorylates the c-Jun activation c-Jun transcription factor. Mol Cell 23: 899–911. domain. Cell 76: 1025–1037. Jiang Y, Chen C, Li Z, Guo W, Gegner JA, Lin S et al. (1996). De´ rijard B, Raingeaud J, Barrett T, Wu I-H, Han J, Characterization of the structure and function of a new Ulevitch RJ et al. (1995). Independent human MAP kinase mitogen-activated protein kinase (p38beta). J Biol Chem signal transduction pathways defined by MEK and MKK 271: 17920–17926. isoforms. Science 267: 682–685. Johnson R, Spiegelman B, Hanahan D, Wisdom R. (1996). Eferl R, Ricci R, Kenner L, Zenz R, David JP, Rath M et al. Cellular transformation and malignancy induced by ras (2003). Liver tumor development. c-Jun antagonizes the requires c-jun. Mol Cell Biol 16: 4504–4511. proapoptotic activity of p53. Cell 112: 181–192. Kelkar N, Delmotte MH, Weston CR, Barrett T, Fan M, Goodwin M, Vu T, Brantley-Finley C, Gaarde WA, Sheppard BJ, Flavell RA et al. (2003). Morphogenesis Chambers TC. (2000). Vinblastine-induced phosphorylation of the telencephalic commissure requires scaffold protein of Bcl-2 and Bcl-XL is mediated by JNK and occurs in JNK-interacting protein 3 (JIP3). Proc Natl Acad Sci USA parallel with inactivation of the Raf-1/MEK/ERK cascade. 100: 9843–9848. J Biol Chem 275: 29980–29985. Kennedy NJ, Davis RJ. (2003). Role of JNK in tumor Finch A, Holland P, Cooper J, Saklatvala J, Kracht M. (1997). development. Cell Cycle 2: 199–201. Selective activation of JNK/SAPK by interleukin-1 in rabbit Kennedy NJ, Sluss HK, Jones SN, Bar-Sagi D, Flavell RA, liver is mediated by MKK7. FEBS Lett 418: 144–148. Davis RJ. (2003). Suppression of Ras-stimulated Fleming Y, Armstrong CG, Morrice N, Paterson A, Goedert M, transformation by the JNK signal transduction pathway. Cohen P. (2000). Synergistic activation of stress-activated Genes Dev 17: 629–637. protein kinase 1/c-Jun N-terminal kinase (SAPK1/JNK) Khatlani TS, Wislez M, Sun M, Srinivas H, Iwanaga K, Ma L isoforms by mitogen-activated protein kinase kinase 4 et al. (2006). c-Jun N-terminal kinase is activated in (MKK4) and MKK7. Biochem J 352: 145–154. non-small-cell ling cancer and promotes neoplastic trans-

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3182 formation in human bronchial epithelial cells. Oncogene the CDC25 phosphatase by the stress-activated Srk1 kinase (in press). in fission yeast. Mol Cell 17: 49–59. Kim BJ, Ryu SW, Song BJ. (2006). JNK-and p38 kinase- Manke IA, Nguyen A, Lim D, Stewart MQ, Elia AE, mediated phosphorylation of Bax leads to its activation and Yaffe MB. (2005). MAPKAP kinase-2 is a cell cycle mitochondrial translocation and to apoptosis of human checkpoint kinase that regulates the G2/Mtransition and hepatoma HepG2 cells. J Biol Chem 281: 21256–21265. S phase progression in response to UV irradiation. Mol Cell Kim GH, Park E, Han JK. (2005). The assembly of POSH- 17: 37–48. JNK regulates Xenopus anterior neural development. Dev Mann KK, Padovani AMS, Guo Q, Colosimo AL, Lee HY, Biol 286: 256–269. Kurie JM et al. (2005). Arsenic trioxide inhibits nuclear Kim HL, Vander Griend DJ, Yang X, Benson DA, Dubauskas Z, receptor function via SEK1/JNK-mediated RXRa Yoshida BA et al. (2001). Mitogen-activated protein kinase phosphorylation. J Clin Invest 115: 2924–2933. kinase 4 metastasis suppressor gene expression is inversely Matsuo Y, Amano S, Furuya M, Namiki K, Sakurai K, related to histological pattern in advancing human prostatic Nishiyama M et al. (2006). Involvement of p38alpha cancers. Cancer Res 61: 2833–2837. mitogen-activated protein kinase in lung metastasis of Kishimoto H, Nakagawa K, Watanabe T, Kitagawa D, tumor cells. J Biol Chem 281: 36767–36775. Momose H, Seo J et al. (2003). Different properties of Maundrell K, Antonsson B, Magnenat E, Camps M, Muda SEK1 and MKK7 in dual phosphorylation of stress-induced M, Chabert C et al. (1997). Bcl-2 undergoes phosphoryla- activated protein kinase SAPK/JNK in embryonic stem tion by c-Jun N-terminal kinase/stress-activated protein cells. J Biol Chem 278: 16595–16601. kinases in the presence of the constitutively active GTP- Kyriakis JM, Avruch J. (2001). Mammalian mitogen-activated binding protein Rac1. J Biol Chem 272: 25238–25242. protein kinase signal transduction pathways activated by McDonald PH, Chow C-W, Miller WE, Laporte SA, Field ME, stress and inflammation. Physiol Rev 81: 807–869. Lin F-T et al. (2000). arrestin-2: a receptor-regulated Lavoie JN, L’Allemain G, Brunet A, Muller R, Pouyssegur J. MAPK scaffold for the activation of JNK3. Science 290: (1996). Cyclin D1 expression is regulated positively by p42/ 1574–1577. p44MAPK and negatively by the p38/HOG-MAPK path- Mori S, Matsuzaki K, Yoshida K, Furukawa F, Tahashi Y, way. J Biol Chem 271: 20608–20616. Yamagata H et al. (2004). TGF-beta and HGF transmit the Lawler S, Cuenda A, Goedert M, Cohen P. (1997). SKK4, a signals through JNK-dependent Smad2/3 phosphorylation novel activator of stress-activated protein kinase-1 (SAPK1/ of the linker regions. Oncogene 23: 7416–7429. JNK). FEBS Lett 414: 153–158. Moriguchi T, Toyoshima F, Masuyama N, Hanafusa H, Lawler S, Fleming Y, Goedert M, Cohen P. (1998). Synergistic Gotoh Y, Nishida E. (1997). A novel SAPK/JNK kinase, activation of SAPK1/JNK1 by two MAP kinase kinases MKK7, stimulated by TNFa and cellular stresses. EMBO J in vitro. Curr Biol 8: 1387–1390. 16: 7045–7053. Lee HY, Oh SH, Suh YA, Baek JH, Papadimitrakopoulou V, Morrison DK, Davis RJ. (2003). Regulation of MAP kinase Huang S et al. (2005). Response of non-small cell lung signaling modules by scaffold proteins in mammals. Annu cancer cells to the inhibitors of phosphatidylinositol Rev Cell Dev Biol 19: 91–118. 3-kinase/Akt- and MAPK kinase 4/c-Jun NH2-terminal Nakayama K, Nakayama N, Davidson B, Katabuchi H, kinase pathways: an effective therapeutic strategy for lung Kurman RJ, Velculescu VE et al. (2006). Homozygous cancer. Clin Cancer Res 11: 6065–6074. deletion of MKK4 in ovarian serous carcinoma. Cancer Biol Lee HY, Suh YA, Robinson MJ, Clifford JL, Hong WK, Ther 5: 630–634. Woodgett JR et al. (2000). Stress pathway activation induces Nateri AS, Spencer-Dene B, Behrens A. (2005). Interaction of phosphorylation of retinoid X receptor. J Biol Chem 275: phosphorylated c-Jun with TCF4 regulates intestinal cancer 32193–32199. development. Nature 437: 281–285. Lee J-K, Hwang W-S, Lee Y-D, Han P-L. (1999). Dynamic Nishina H, Bachmann M, Oliveiria-dos-Santos AJ, expression of SEK1 suggests multiple roles of the gene Kozieradzki I, Fischer KD, Odermatt B et al. (1997a). during emryogenesis and in adult brain of mice. Mol Brain Impaired CD28-mediated interleukin 2 production and Res 66: 133–140. proliferation in stress kinase SAPK/ERK1 kinase (SEK1)/ Lei K, Davis RJ. (2003). JNK phosphorylation of Bim-related mitogen-activated protein kinase kinase 4 (MKK4)-deficient members of the Bcl2 family induces Bax-dependent apop- T lymphocytes. J Exp Med 186: 941–953. tosis. Proc Natl Acad Sci USA 100: 2432–2437. Nishina H, Fischer KD, Radvanyi L, Shahinian A, Hakem R, Lei K, Nimnual A, Zong WX, Kennedy NJ, Flavell RA, Rubie EA et al. (1997b). Stress-signaling kinase SEK1 Thompson CB et al. (2002). The Bax subfamily of protects thymocytes from apoptosis mediated by CD95 and Bcl2-related proteins is essential for apoptotic signal CD3. Nature 385: 350–353. transduction by c-Jun NH(2)-terminal kinase. Mol Cell Biol Nishina H, Vaz C, Billia P, Nghiem M, Sasaki T, De la Pompa JL 22: 4929–4942. et al. (1999). Defective liver formation and liver cell apoptosis Leppa S, Saffrich R, Ansorge W, Bohmann D. (1998). in mice lacking the stress signaling kinase SEK1/MKK4. Differential regulation of c-Jun by ERK and JNK during Development 126: 505–516. PC12 cell differentiation. EMBO J 17: 4404–4413. Ohren JF, Chen H, Pavlovsky A, Whitehead C, Zhang E, Lin A, Minden A, Martinetto H, Claret F-X, Lange-Carter C, Kuffa P et al. (2004). Structures of human MAP kinase Mercurio F et al. (1995). Identification of a dual specificity kinase 1 (MEK1) and MEK2 describe novel noncompetitive kinase that activates the Jun kinases and p38-Mpk2. Science kinase inhibition. Nat Struct Mol Biol 11: 1192–1197. 268: 286–290. Ortega S, Malumbres M, Barbacid M. (2002). Cyclin Lisnock J-M, Griffin P, Calaycay J, Frantz B, Parsons J, D-dependent kinases, INK4 inhibitors and cancer. Biochim O’Keefe SJ et al. (2000). Activation of JNK3a1 requires Biophys Acta 1602: 73–87. both MKK4 and MKK7: kinetic characterization of in vitro Potapova O, Gorospe M, Bost F, Dean NM, Gaarde WA, phosphorylation JNK3a1. Biochemistry 39: 3141–3148. Mercola D et al. (2000). c-Jun N-terminal kinase is essential Lopez-Aviles S, Grande M, Gonzalez M, Helgesen AL, for growth of human T98G glioblastoma cells. J Biol Chem Alemany V, Sanchez-Piris M et al. (2005). Inactivation of 275: 24767–24775.

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3183 PulvererBJ,KyriakisJM,AvruchJ,NikolakakiE,WoodgettJR. Teng DH, Perry III WL, Hogan JK, Baumgard M, Bell R, (1991). Phosphorylation of c-Jun mediated by MAP kinases. Berry S et al. (1997). Human mitogen-activated protein Nature 353: 670–674. kinase 4 as a candidate tumor suppressor. Cancer Res 57: Putcha GV, Le S, Frank S, Besirli CG, Clark K, Chu B et al. 4177–4182. (2003). JNK-mediated BIMphosphorylation potentiates Tournier C, Dong C, Turner TK, Jones SN, Flavell RA, BAX-dependent apoptosis. Neuron 38: 899–914. Davis RJ. (2001). MKK7 is an essential component of the Rodrigues GA, Park M, Schlessinger J. (1997). Activation of JNK signal transduction pathway activated by proinflam- the JNK pathway is essential for transformation by the Met matory cytokines. Genes Dev 15: 1419–1426. oncogene. EMBO J 16: 2634–2645. Tournier C, Hess P, Yang DD, Xu J, Turner TK, Nimnual A Sabapathy K, Hochedlinger K, Nam SY, Bauer A, Karin M, et al. (2000). Requirement of JNK for stress-induced Wagner EF. (2004). Distinct roles for JNK1 and JNK2 in activation of the cytochrome c-mediated death pathway. regulating JNK activity and c-Jun-dependent cell prolifera- Science 288: 870–874. tion. Mol Cell 15: 713–725. Tournier C, Whitmarsh AJ, Cavanagh J, Barrett T, Davis RJ. Sakurai T, Maeda S, Chang L, Karin M. (2006). Loss of (1997). Mitogen-activated protein kinase kinase 7 is an hepatic NF-kB activity enhances chemical hepatocarcino- activator of the c-Jun NH2-terminal kinase. Proc Natl Acad genesis through sustained c-Jun N-terminal kinase Sci USA 94: 7337–7342. 1 activation. Proc Natl Acad Sci USA 103: 10544–10551. Tournier C, Whitmarsh AJ, Cavanagh J, Barrett T, Davis RJ. Sanchez I, Hughes RT, Mayer BJ, Yee K, Woodgett JR, (1999). The MKK7 gene encodes a group of c-Jun Avruch J et al. (1994). Role of SAPK/ERK kinase-1 in the NH2-terminal kinase kinases. Mol Cell Biol 19: 1569–1581. stress-activated pathway regulating the transcription factor Tsuruta F, Sunayama J, Mori Y, Hattori S, Shimizu S, c-Jun. Nature 372: 794–798. Tsujimoto Y et al. (2004). JNK promotes Bax transloca- Schutte J, Minna JD, Birrer MJ. (1989). Deregulated expres- tion to mitochondria through phosphorylation of 14-3-3 sion of human c-jun transforms primary rat embryo cells in proteins. EMBO J 23: 1889–1899. cooperation with an activated c-Ha-ras gene and transforms Uhlirova M, Jasper H, Bohmann D. (2005). Non-cell- rat-1a cells as a single gene. Proc Natl Acad Sci USA 86: autonomous induction of tissue overgrowth by JNK/Ras 2257–2261. cooperation in a Drosophila tumor model. Proc Natl Acad Sharrocks AD, Yang S-H, Galanis A. (2000). Docking Sci USA 102: 13123–13128. domains and substrate-specificity determination for MAP Vander Griend DJ, Kocherginsky M, Hickson JA, kinases. Trends Biochem Sci 25: 448–453. Stadler WM, Lin A, Rinker-Schaeffer CW. (2005). She QB, Chen N, Bode AM, Flavell RA, Dong Z. (2002). Suppression of metastatic colonization by the context- Deficiency of c-Jun NH(2)terminal kinase-1 in mice dependent activation of the c-Jun NH2-terminal enhances skin tumor development by 12-O-tetradecanoyl- kinase kinases JNKK1/MKK4 and MKK7. Cancer Res phorbol-13-acetate. Cancer Res 62: 1343–1348. 65: 10984–10991. Smeal T, Binetruy B, Mercola DA, Birrer M, Karin M. (1991). Ventura JJ, Kennedy NJ, Flavell RA, Davis RJ. (2004). Oncogenic and transcriptional cooperation with Ha-Ras JNK regulates autocrine expression of TGF-b1. Mol Cell 15: requires phosphorylation of c-Jun on serines 63 and 73. 269–278. Nature 354: 494–496. Vogelstein B, Lane D, Levine AJ. (2000). Surfing the p53 Srinivas H, Juroske DM, Kalyankrishna S, Cody DD, network. Nature 408: 307–310. Price RE, Xu XC et al. (2005). c-Jun N-terminal kinase Wada T, Nakagawa K, Watanabe T, Nishitai G, Seo J, contributes to aberrant retinoid signaling in lung cancer cells Kishimoto H et al. (2001). Impaired synergistic activation of by phosphorylating and inducing proteosomal degradation stress-activated protein kinase SAPK/JNK in mouse of retinoic acid receptor a. Mol Cell Biol 25: 1054–1069. embryonic stem cells lacking SEK1/MKK4. J Biol Chem Stark AM, Tongers K, Maass N, Mehdorn HM, Held- 276: 30892–30897. Feindt J. (2005). Reduced metastasis-suppressor gene Wang L, Pan Y, Dai JL. (2004). Evidence of MKK4 mRNA-expression in breast cancer brain metastases. pro-oncogenic activity in breast and pancreatic tumors. J Cancer Res Clin Oncol 131: 191–198. Oncogene 23: 5978–5985. Su GH, Hilgers W, Shekher MC, Tang DJ, Yeo CJ, Whitmarsh AJ. (2006). The JIP family of MAP kinase scaffold Hruban RH et al. (1998). Alterations in pancreatic, proteins. Biochem Soc Trans 34: 828–832. biliary, and breast carcinomers support MKK4 as a Xia Y, Wu Z, Su B, Murray B, Karin M. (1998). JNKK1 genetically targeted tumor suppressor gene. Cancer Res 58: organizes a MAP kinase module through specific and 2339–2342. sequential interactions with upstream and downstream Su GH, Song JJ, Repasky EA, Schutte M, Kern SE. (2002). components mediated by its amino-terminal extension. Mutation rate of MAP2K4/MKK4 in breast carcinoma. Genes Dev 12: 3369–3381. Hum Mutat 19: 81. Xiao L, Lang W. (2000). A dominant role for the c-Jun Swat W, Fujikawa K, Ganiatsas S, Yang D, Xavier RJ, NH2-terminal kinase in oncogenic ras-induced morphologic Harris NL et al. (1998). SEK1/MKK4 is required for transformation of human lung carcinoma cells. Cancer Res maintenance of a normal peripheral lymphoid compartment 60: 400–408. but not for lymphocyte development. Immunity 8: 625–634. Xin W, Yun KJ, Ricci F, Zahurak M, Qiu W, Su GH et al. Takekawa M, Adachi M, Nakahata A, Nakayama I, Itoh F, (2004). MAP2K4/MKK4 expression in pancreatic cancer: Tsukuda H et al. (2000). p53-inducible wip1 phosphatase genetic validation of immunohistochemistry and relation- mediates a negative feedback regulation of p38 MAPK-p53 ship to disease course. Clin Cancer Res 10: 8516–8520. signaling in response to UV radiation. EMBO J 19: Xiu M, Kim J, Sampson E, Huang CY, Davis RJ, Paulson KE 6517–6526. et al. (2003). The transcriptional repressor HBP1 is a target Takekawa M, Tatebayashi K, Saito H. (2005). Conserved of the p38 mitogen-activated protein kinase pathway in cell docking site is essential for activation of mammalian MAP cycle regulation. Mol Cell Biol 23: 8890–8901. kinase kinases by specific MAP kinase kinase kinases. Mol Xu P, Yoshioka K, Yoshimura D, Tominaga Y, Nishioka T, Cell 18: 295–306. Ito M et al. (2003a). In vitro development of mouse

Oncogene MKK4 in cancer AJ Whitmarsh and RJ Davis 3184 embryonic stem cells lacking JNK/stress-activated protein Yang YM, Bost F, Charbono W, Dean N, McKay R, Rhim JS kinase-associated protein 1 (JSAP1) scaffold protein re- et al. (2003). c-Jun NH(2)-terminal kinase mediates pro- vealed its requirement during early embryonic neurogenesis. liferation and tumor growth of human prostate carcinoma. J Biol Chem 278: 48422–48433. Clin Cancer Res 9: 391–401. Xu Z, Kukekov NV, Greene LA. (2003b). POSH acts as a Yashar BM, Kelley C, Yee K, Errede B, Zon LI. (1993). Novel scaffold for a multiprotein complex that mediates JNK Members of the mitogen-activated protein kinase activator activation in apoptosis. EMBO J 22: 252–261. family in Xenopus laevis. Mol Cell Biol 13: 5738–5748. Yamada SD, Hickson JA, Hrobowski Y, Vander Griend DJ, Yoshida BA, Dubauskas Z, Chekmareva MA, Christiano TR, Benson D, Montag A et al. (2002). Mitogen-activated Stadler WM, Rinker-Schaeffer CW. (1999). Mitogen- protein kinase kinase 4 (MKK4) acts as a metastasis activated protein kinase kinase 4/stress-activated protein/ suppressor gene in human ovarian carcinoma. Cancer Res ERK kinase 1 (MKK4/SEK1), a prostate cancer metastasis 62: 6717–6723. suppressor gene encoded by human chromosome 17. Cancer Yamamoto K, Ichijo H, Korsmeyer SJ. (1999). BCL-2 is Res 59: 5483–5487. phosphorylated and inactivated by an ASK1/Jun Yoshida S, Fukino K, Harada H, Nagai H, Imoto I, Inazawa J N-terminal protein kinase pathway normally activated at et al. (2001). The c-Jun NH2-terminal kinase3 (JNK3) G(2)/M. Mol Cell Biol 19: 8469–8478. gene: genomic structure, chromosomal assignment, and Yang DD, Kuan CY, Whitmarsh AJ, Rincon M, Zheng TS, loss of expression in brain tumors. J Hum Genet 46: Davis RJ et al. (1997a). Absence of excitotoxicity-induced 182–187. apoptosis in the hippocampus of mice lacking the Jnk3 gene. Zacchi P, Gostissa M, Uchida T, Salvagno C, Avolio E, Nature 389: 865–870. Volinia S et al. (2002). The prolyl isomerase Pin1 reveals a Yang DD, Tournier C, Wysk M, Lu HT, Xu J, Davis RJ et al. mechanism to control p53 functions after genotoxic insults. (1997b). Targeted disruption of the MKK4 gene causes Nature 419: 853–857. embryonic death, inhibition of c-Jun NH2-terminal kinase Zheng H, You H, Zhou XZ, Murray SA, Uchida T, Wulf G activation, and defects in AP-1 transcriptional activity. Proc et al. (2002). The prolyl isomerase Pin1 is a regulator of p53 Natl Acad Sci USA 94: 3004–3009. in genotoxic response. Nature 419: 849–853.

Oncogene