MAP2K4

MAP3K5 phosphorylates and activates MAP2K3 (MKK3) and MAP2K6 (MKK6) (Ichijo et al. 1997, Takekawa et al. 2005), which act as p38 MAPK kinases, as well as MAP2K4 (SEK1) (Ichijo et al. 1997, Matsuura et al. 2002), which, together with MAP2K7 (MKK7), acts as a JNK kinase.

MKK3 and MKK6 phosphorylate and activate p38 MAPK alpha (MAPK14) and beta (MAPK11) (Raingeaud et al. 1996), enabling p38 MAPKs to phosphorylate and activate MAPKAPK2 (MK2) and MAPKAPK3 (MK3) (Ben-Levy et al. 1995, Clifton et al. 1996, McLaughlin et al. 1996, Sithanandam et al. 1996, Meng et al. 2002, Lukas et al. 2004, White et al. 2007), as well as MAPKAPK5 (PRAK) (New et al. 1998 and 2003, Sun et al. 2007).

Phosphorylation of JNKs (MAPK8, MAPK9 and MAPK10) by MAP3K5-activated MAP2K4 (Deacon and Blank 1997, Fleming et al. 2000) allows JNKs to migrate to the nucleus (Mizukami et al. 1997) where they phosphorylate JUN. Phosphorylated JUN binds FOS phosphorylated by ERK1 or ERK2, downstream of activated RAS (Okazaki and Sagata 1995, Murphy et al. 2002), forming the activated protein 1 (AP-1) complex (FOS:JUN heterodimer) (Glover and Harrison 1995, Ainbinder et al. 1997).

Activation of p38 MAPKs and JNKs downstream of MAP3K5 (ASK1) ultimately converges on transcriptional regulation of CDKN2A locus. In dividing cells, nucleosomes bound to the CDKN2A locus are trimethylated on lysine residue 28 of histone H3 (HIST1H3A) by the Polycomb repressor complex 2 (PRC2), creating the H3K27Me3 (Me3K-28-HIST1H3A) mark (Bracken et al. 2007, Kotake et al. 2007). The expression of Polycomb constituents of PRC2 (Kuzmichev et al. 2002) - EZH2, EED and SUZ12 - and thereby formation of the PRC2, is positively regulated in growing cells by E2F1, E2F2 and E2F3 (Weinmann et al. 2001, Bracken et al. 2003). H3K27Me3 mark serves as a docking site for the Polycomb repressor complex 1 (PRC1) that contains BMI1 (PCGF4) and is therefore named PRC1.4, leading to the repression of transcription of p16-INK4A and p14-ARF from the CDKN2A locus, where PCR1.4 mediated repression of p14-ARF transcription in humans may be context dependent (Voncken et al. 2005, Dietrich et al. 2007, Agherbi et al. 2009, Gao et al. 2012). MAPKAPK2 and MAPKAPK3, activated downstream of the MAP3K5-p38 MAPK cascade, phosphorylate BMI1 of the PRC1.4 complex, leading to dissociation of PRC1.4 complex from the CDKN2A locus and upregulation of p14-ARF transcription (Voncken et al. 2005). AP-1 transcription factor, formed as a result of MAP3K5-JNK signaling, as well as RAS signaling, binds the promoter of KDM6B (JMJD3) gene and stimulates KDM6B expression. KDM6B is a histone demethylase that removes H3K27Me3 mark i.e. demethylates lysine K28 of HIST1H3A, thereby preventing PRC1.4 binding to the CDKN2A locus and allowing transcription of p16-INK4A (Agger et al. 2009, Barradas et al. 2009, Lin et al. 2012).

p16-INK4A inhibits phosphorylation-mediated inactivation of RB family members by CDK4 and CDK6, leading to cell cycle arrest (Serrano et al. 1993). p14-ARF inhibits MDM2-mediated degradation of TP53 (p53) (Zhang et al. 1998), which also contributes to cell cycle arrest in cells undergoing oxidative stress. In addition, phosphorylation of TP53 by MAPKAPK5 (PRAK) activated downstream of MAP3K5-p38 MAPK signaling, activates TP53 and contributes to cellular senescence (Sun et al. 2007)

The JNKs are encoded by at least three genes: JNK1/SAPK-gamma, JNK2/SAPK-alpha and JNK3/ SAPK-beta. The first two are ubiquitously expressed, whereas the JNK3 protein is found mainly in brain and to a lesser extent in heart and testes. As a result of alternative gene splicing all cells express distinct active forms of JNK from 46 to 55 kDa in size. Sequence alignment of these different products shows homologies of >80%. In spite of this similarity, the multiple JNK isoforms differ in their ability to bind and phosphorylate different target proteins, thus leading to the distinctive cellular processes: induction of apoptosis, or enhancment of cell survival, or proliferation.

Activation of JNKs is mediated by activated TAK1 which phosphorylates two dual specificity enzymes MKK4 (MAPK kinase 4) and MKK7(MAPK kinase 7)

In the absence of extra-cellular signals, cytosolic MAP3K8 (TPL2) is held inactive in the complex with ABIN2 (TNIP2) and NFkB p105 (NFKB1) (Beinke S et al., 2003; Waterfield MR et al., 2003; Lang V et al., 2004). This interaction stabilizes MAP3K8 (TPL2) but also prevents MAP3K8 and NFkB from activating their downstream signaling cascades by inhibiting the kinase activity of MAP3K8 and the proteolysis of NFkB precursor protein p105. Upon activation of MAP3K8 by various stimuli (such as LPS, TNF-alpha, and IL-1 beta), IKBKB phosphorylates NFkB p105 (NFKB1) at Ser927 and Ser932, which trigger p105 proteasomal degradation and releases MAP3K8 from the complex (Beinke S et al., 2003, 2004; Roget K et al., 2012). Simultaneously, MAP3K8 is activated by auto- and/or transphosphorylation (Gantke T et al. 2011; Yang HT et al. 2012). The released active MAP3K8 phosphorylates its substrates, MAP2Ks. The free MAP3K8, however, is also unstable and is targeted for proteasome-mediated degradation, thus restricting prolonged activation of MAP3K8 (TPL2) and its downstream signaling pathways (Waterfield MR et al. 2003; Cho J et al., 2005). Furthermore, partially degraded NFkB p105 (NFKB1) into p50 can dimerize with other NFkB family members to regulate the transcription of target genes.

MAP3K8 activity is thought to regulate the dynamics of transcription factors that control an expression of diverse genes involved in growth, differentiation, and inflammation. Suppressing the MAP3K8 kinase activity with selective inhibitors, such as C8-chloronaphthyridine-3-carbonitrile, caused a significant reduction in TNFalpha production in LPS- and IL-1beta-induced both primary human monocytes and human blood (Hall JP et al. 2007). Similar results have been reported for mouse LPS-stimulated RAW264.7 cells (Hirata K et al. 2010). Moreover, LPS-stimulated macrophages derived from Map3k8 knockout mice secreted lower levels of pro-inflammatory cytokines such as TNFalpha, Cox2, Pge2 and CXCL1 (Dumitru CD et al. 2000; Eliopoulos AG et al. 2002). Additionally, bone marrow-derived dendritic cells (BMDCs) and macrophages from Map3k8 knockout mice showed significantly lower expression of IL-1beta in response to LPS, poly IC and LPS/MDP (Mielke et al., 2009). However, several other studies seem to contradict these findings and Map3k8 deficiency in mice has been also reported to enhance pro-inflammatory profiles. Map3k8 deficiency in LPS-stimulated macrophages was associated with an increase in nitric oxide synthase 2 (NOS2) expression (López-Peláez et al., 2011). Similarly, expression of IRAK-M, whose function is to compete with IL-1R-associated kinase (IRAK) family of kinases, was decreased in Map3k8-/- macrophages while levels of TNF and IL6 were elevated (Zacharioudaki et al., 2009). Moreover, significantly higher inflammation level was observed in 12-O-tetradecanoylphorbol-13-acetate (TPA)-treated Map3k8-/- mouse skin compared to WT skin (DeCicco-Skinner K. et al., 2011). Additionally, MAP3K8 activity is associated with NFkB inflammatory pathway. High levels of active p65 NFkB were observed in the nucleus of Map3k8 -/- mouse keratinocytes that dramatically increased within 15-30 minutes of TPA treatment. Similarly, increased p65 NFkB was observed in Map3k8-deficient BMDC both basally and after stimulation with LPS when compared to wild type controls (Mielke et al., 2009). The data opposes the findings that Map3k8-deficient mouse embryo fibroblasts and human Jurkat T cells with kinase domain-deficient protein have a reduction in NFkB activation but only when certain stimuli are administered (Lin et al., 1999; Das S et al., 2005). Thus, it is possible that whether MAP3K8 serves more of a pro-inflammatory or anti-inflammatory role may depend on cell- or tissue type and on stimuli (LPS vs. TPA, etc.) (Mielke et al., 2009; DeCicco-Skinner K. et al., 2012).

MAP3K8 has been also studied in the context of carcinogenesis, however the physiological role of MAP3K8 in the etiology of human cancers is also convoluted (Vougioukalaki M et al., 2011; DeCicco-Skinner K. et al., 2012)