Title: The Regulation of AP-1 Activity by Mitogen-activated Protein Kinases
Abstract: AP-1 is a sequence-specific transcriptional activator composed of members of the Jun and Fos families (for review see Ref. 1Angel P. Karin M. Biochim. Biophys. Acta. 1991; 1072: 129-157Crossref PubMed Scopus (3256) Google Scholar). These proteins, which belong to the bZIP group of DNA binding proteins (for review see Ref. 2Johnson P.F. McKnight S.L. Annu. Rev. Biochem. 1989; 58: 799-839Crossref PubMed Scopus (829) Google Scholar), associate to form a variety of homo- and heterodimers that bind to a common site(1Angel P. Karin M. Biochim. Biophys. Acta. 1991; 1072: 129-157Crossref PubMed Scopus (3256) Google Scholar). First identified by its role in human metallothionein IIA gene regulation(3Lee W. Haslinger A. Karin M. Tjian R. Nature. 1987; 325: 368-372Crossref PubMed Scopus (484) Google Scholar), AP-1 was also found as a transcription factor that mediates gene induction by the phorbol ester tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA) and hence the name TRE1 1The abbreviations used are: TRETPA (12-O-tetradecanoylphorbol-13-acetate) response elementMAPKmitogen-activated protein kinaseSREserum response elementTCFternary complex factorCBPCREB binding proteinJAKJanus kinaseERKextracellular stimulus responsive kinaseATFactivating transcription factorJNKJun N-terminal kinaseCREBcAMP response element binding proteinFRKFos regulating kinaseMEKMAPK or ERK kinase. 1The abbreviations used are: TRETPA (12-O-tetradecanoylphorbol-13-acetate) response elementMAPKmitogen-activated protein kinaseSREserum response elementTCFternary complex factorCBPCREB binding proteinJAKJanus kinaseERKextracellular stimulus responsive kinaseATFactivating transcription factorJNKJun N-terminal kinaseCREBcAMP response element binding proteinFRKFos regulating kinaseMEKMAPK or ERK kinase.(TPA response element) for its recognition site(4Angel P. Imagawa M. Chiu R. Stein B. Imbra R.J. Rahmsdorf H.J. Jonat C. Herrlich P. Karin M. Cell. 1987; 49: 729-739Abstract Full Text PDF PubMed Scopus (2152) Google Scholar). Following its discovery, AP-1 activity was found to be induced by many other stimuli, including growth factors, cytokines, T cell activators, neurotransmitters, and UV irradiation (1Angel P. Karin M. Biochim. Biophys. Acta. 1991; 1072: 129-157Crossref PubMed Scopus (3256) Google Scholar). Several mechanisms are involved in induction of AP-1 activity and may be classified as those that increase the abundance of AP-1 components and those that stimulate their activity. A complete discussion of all the mechanisms that regulate AP-1 activity, either positively or negatively, is beyond the scope of this minireview, and the reader is referred to an earlier, more comprehensive review(1Angel P. Karin M. Biochim. Biophys. Acta. 1991; 1072: 129-157Crossref PubMed Scopus (3256) Google Scholar). The present review focuses on the role of mitogen-activated protein kinases (MAPKs) in regulation of AP-1 activity. The regulation and functions of these important signal-transducing enzymes were recently reviewed(5Marshall C.J. Cell. 1995; 80: 179-185Abstract Full Text PDF PubMed Scopus (4225) Google Scholar, 6Cobb M. Goldsmith E.J. J. Biol. Chem. 1995; 270: 14843-14846Abstract Full Text Full Text PDF PubMed Scopus (1659) Google Scholar). TPA (12-O-tetradecanoylphorbol-13-acetate) response element mitogen-activated protein kinase serum response element ternary complex factor CREB binding protein Janus kinase extracellular stimulus responsive kinase activating transcription factor Jun N-terminal kinase cAMP response element binding protein Fos regulating kinase MAPK or ERK kinase. TPA (12-O-tetradecanoylphorbol-13-acetate) response element mitogen-activated protein kinase serum response element ternary complex factor CREB binding protein Janus kinase extracellular stimulus responsive kinase activating transcription factor Jun N-terminal kinase cAMP response element binding protein Fos regulating kinase MAPK or ERK kinase. The reader should be aware that although AP-1 DNA binding activity can be conveniently measured by electrophoretic mobility shift or footprinting assays, changes in AP-1 DNA binding activity do not mirror the transcriptional activity of this complex factor. Therefore, when dealing with AP-1, it is critical to measure its ability to activate transcription of an AP-1-dependent reporter gene. A useful promoter for such experiments is that of the human collagenase gene (4Angel P. Imagawa M. Chiu R. Stein B. Imbra R.J. Rahmsdorf H.J. Jonat C. Herrlich P. Karin M. Cell. 1987; 49: 729-739Abstract Full Text PDF PubMed Scopus (2152) Google Scholar). The reasons for this discrepancy are several. First and foremost, several proteins can form complexes that bind to AP-1 sites. These proteins, however, differ considerably in their ability to activate transcription of target genes. For instance, both c-Fos and Fra-1 form stable heterodimers with any of the Jun proteins, and these heterodimers have similar DNA binding activities and specificities, yet c-Fos has a potent transactivation domain that is absent from the smaller Fra-1 protein(7Suzuki T. Okuno H. Yoshida T. Endo T. Nishina H. Iba H. Nucleic Acids Res. 1992; 19: 5537-5542Crossref Scopus (194) Google Scholar). Second, phosphorylation at specific sites enhances the transactivating potential of several AP-1 proteins, including c-Jun and c-Fos, without having any effect on their DNA binding activities(8Smeal T. Binetruy B. Mercola D. Grover-Bardwick A. Heidecker G. Rapp U.R. Karin M. Mol. Cell. Biol. 1992; 12: 3507-3513Crossref PubMed Scopus (305) Google Scholar, 9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar). Most of the genes that encode AP-1 components behave as “immediate-early” genes, i.e. genes whose transcription is rapidly induced, independently of de novo protein synthesis, following cell stimulation. Among these, the regulation of c-fos and c-jun transcription is best understood. Several cis elements mediate c-fos induction in response to a diverse spectrum of extracellular stimuli (reviewed in Ref. 10Treisman R. Trends Biochem. Sci. 1992; 17: 423-426Abstract Full Text PDF PubMed Scopus (349) Google Scholar). A cAMP response element mediates c-fos induction in response to neurotransmitters and polypeptide hormones which, by using either cAMP or Ca2+ as second messengers, activated either protein kinase A or calmodulin-dependent protein kinases, respectively(11Sheng M.E. Thompson M.A. Greenberg M.E. Science. 1994; 252: 1427-1430Crossref Scopus (1279) Google Scholar). A serum response element (SRE) mediates c-fos induction by growth factors, cytokines, and other stimuli that activate MAPKs(10Treisman R. Trends Biochem. Sci. 1992; 17: 423-426Abstract Full Text PDF PubMed Scopus (349) Google Scholar), and a Sis-inducible enhancer mediates induction by stimuli that activate the JAK group of protein kinases (12Darnell Jr., J.E. Kerr I.M. Stark G.R. Science. 1994; 264: 1415-1421Crossref PubMed Scopus (4974) Google Scholar). Given this complexity, it is not surprising that c-fos transcription is rapidly induced in response to almost any imaginable extracellular stimulus (Fig. 1A). The SRE is recognized by the serum response factor, whose binding results in recruitment of the ternary complex factor (TCF), which cannot bind to the SRE by itself(10Treisman R. Trends Biochem. Sci. 1992; 17: 423-426Abstract Full Text PDF PubMed Scopus (349) Google Scholar). Following mitogenic stimulation, Elk-1, one of several candidate TCFs(13Treisman R. Curr. Opin. Genet. & Dev. 1994; 4: 96-101Crossref PubMed Scopus (619) Google Scholar), is rapidly phosphorylated, most likely by members of the ERK group of MAPKs(14Gille H. Sharrocks A. Shaw P. Nature. 1992; 358: 414-417Crossref PubMed Scopus (814) Google Scholar, 15Marais R. Wynne J. Treisman R. Cell. 1993; 73: 381-393Abstract Full Text PDF PubMed Scopus (1104) Google Scholar). Phosphorylation of Elk-1 was reported to facilitate formation of the ternary complex composed of itself, the serum response factor, and the SRE (14Gille H. Sharrocks A. Shaw P. Nature. 1992; 358: 414-417Crossref PubMed Scopus (814) Google Scholar) and to stimulate its ability to activate transcription, without affecting its DNA binding properties(15Marais R. Wynne J. Treisman R. Cell. 1993; 73: 381-393Abstract Full Text PDF PubMed Scopus (1104) Google Scholar). Since in vivo the SRE appears to be constitutively occupied(10Treisman R. Trends Biochem. Sci. 1992; 17: 423-426Abstract Full Text PDF PubMed Scopus (349) Google Scholar), increased Elk-1 transcriptional activity is a more likely mechanism by which ERK activation causes c-fos induction. The sites at which Elk-1 is phosphorylated are clustered within its C-terminal activation domain and are conserved in other candidate TCFs, such as SAP-1(13Treisman R. Curr. Opin. Genet. & Dev. 1994; 4: 96-101Crossref PubMed Scopus (619) Google Scholar). Since the SRE also mediates c-fos induction in response to stimuli such as UV irradiation(16Buscher M. Rahmsdorf H.J. Liftin M. Karin M. Herrlich P. Oncogene. 1988; 3: 301-311PubMed Google Scholar), which has only a marginal effect on ERK activity(17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar), it is possible that Elk-1 or other TCFs are also phosphorylated by other MAPKs (see below). Nevertheless, ERK activation leads to elevated AP-1 activity via c-fos induction. This results in increased synthesis of c-Fos, which upon translocation to the nucleus combines with pre-existing Jun proteins to form AP-1 dimers that are more stable than those formed by Jun proteins alone(18Smeal T. Angel P. Meek J. Karin M. Genes & Dev. 1989; 3: 2091-2100Crossref PubMed Scopus (185) Google Scholar). Increased stability results in higher levels of AP-1 DNA binding activity because it shifts the equilibrium toward dimer formation, which is essential for DNA binding. By comparison with c-fos, the c-jun promoter is somewhat simpler, and most of its inducers operate through one major cis element, the c-jun TRE (Fig. 1B). This TRE differs from the consensus TRE sequence by 1-base pair insertion(19Angel P. Hattori K. Smeal T. Karin M. Cell. 1988; 55: 875-885Abstract Full Text PDF PubMed Scopus (990) Google Scholar), and due to this subtle change it is more efficiently recognized by c-Jun·ATF2 heterodimers than by conventional AP-1 complexes(20van Dam H. Duyndam M. Rottier R. Bosch A. de Vries-Smits L. Herrlich P. Zantema A. Angel P. van der Eb A.J. EMBO J. 1993; 12: 479-487Crossref PubMed Scopus (342) Google Scholar). Unlike c-Jun, ATF2 is a constitutively expressed protein. However, despite its inducible expression, most cell types contain some c-Jun protein prior to their stimulation. Like the c-fos SRE, the c-jun TRE is constitutively occupied in vivo(21Rozek D. Pfeifer G.P. Mol. Cell. Biol. 1993; 13: 5490-5499Crossref PubMed Scopus (115) Google Scholar). Following exposure to stimuli that activate members of the JNK group of MAPKs(22Dérijard B. Hibi M. Wu I.-H. Barrett T. Su B. Deng T. Karin M. Davis R.J. Cell. 1994; 76: 1-20Abstract Full Text PDF PubMed Scopus (2953) Google Scholar), both c-Jun (23Devary Y. Gottlieb R. Smeal T. Bauskin A.R. Ben-Neriah Y. Karin M. Cell. 1992; 71: 1081-1091Abstract Full Text PDF PubMed Scopus (795) Google Scholar) and ATF2 (24Gupta S. Campbell D. Derijard B. Davis R.J. Science. 1995; 267: 389-393Crossref PubMed Scopus (1336) Google Scholar) are rapidly phosphorylated. The constitutive occupancy of the c-jun TRE indicates that this phosphorylation occurs while the proteins are bound to the c-jun promoter. Similarly, in the case of c-fos, Elk-1 must be phosphorylated while bound to DNA. Phosphorylation of c-Jun and ATF2 stimulates their ability to activate transcription, thereby leading to c-jun induction. Thus, part of the increase in AP-1 activity in response to JNK-activating stimuli (such as tumor necrosis factor α, UV irradiation) is due to increased c-Jun synthesis and possibly c-Fos synthesis (as the JNKs may also phosphorylate and activate Elk-1; see below). Another part of the increase in AP-1 activity is due to c-Jun phosphorylation. The activities of both pre-existing and newly synthesized AP-1 components are modulated through their phosphorylation. So far, this form of posttranslational control was demonstrated for c-Jun, c-Fos, and ATF2, but it is likely that other Jun and Fos proteins are similarly regulated. In the case of c-Jun, phosphorylation at a cluster of sites located next to its basic region inhibits DNA binding by c-Jun homodimers but not by c-Jun·c-Fos heterodimers(25Boyle W.J. Smeal T. Defize L.H.K. Angel P. Woodgett J.R. Karin M. Hunter T. Cell. 1991; 64: 573-584Abstract Full Text PDF PubMed Scopus (852) Google Scholar). On the other hand, phosphorylation of c-Jun at Ser-73 and Ser-63, located within its transactivation domain, potentiates its ability to activate transcription as either a homodimer (26Pulverer B.J. Kyriakis J.M. Avruch J. Nikolakaki E. Woodgett J.R. Nature. 1991; 353: 670-674Crossref PubMed Scopus (1191) Google Scholar, 27Smeal T. Binetruy B. Mercola D.A. Birrer M. Karin M. Nature. 1991; 354: 494-496Crossref PubMed Scopus (697) Google Scholar) or a heterodimer with c-Fos(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar). These residues, which do not affect DNA binding activities, are phosphorylated by the newly discovered members of the MAPK family, the Jun kinases or JNKs(22Dérijard B. Hibi M. Wu I.-H. Barrett T. Su B. Deng T. Karin M. Davis R.J. Cell. 1994; 76: 1-20Abstract Full Text PDF PubMed Scopus (2953) Google Scholar, 28Hibi M. Lin A. Smeal T. Minden A. Karin M. Genes & Dev. 1993; 7: 2135-2148Crossref PubMed Scopus (1708) Google Scholar). So far, the JNKs are the only protein kinases found to efficiently phosphorylate the N-terminal sites of c-Jun. Interestingly, neither ERK1 nor ERK2 phosphorylates the N-terminal stimulatory sites of c-Jun and instead phosphorylate one of the inhibitory sites located next to the C-terminal DNA binding domain (17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar, 29Chou S. Baichval V. Ferrell Jr., J.E. Mol. Biol. Cell. 1992; 3: 1117-1180Crossref PubMed Scopus (59) Google Scholar). Using an altered specificity mutant of c-Jun that is phosphorylated by protein kinase A instead of JNK, phosphorylation of Ser-73 (and Ser-63 to a lesser extent) was demonstrated to be directly responsible for potentiating the transactivation function(30Smeal T. Hibi M. Karin M. EMBO J. 1994; 13: 6006-6010Crossref PubMed Scopus (84) Google Scholar). Phosphorylation may potentiate c-Jun transcriptional activity through recruitment of CREB binding protein (CBP), a protein that was originally identified by virtue of its binding to phospho-CREB, another bZIP transcription factor that is activated by protein kinase A(31Arias J. Alberts A.S. Brindle P. Claret F.X. Smeal T. Karin M. Feramisco J. Montminy M. Nature. 1994; 370: 226-229Crossref PubMed Scopus (680) Google Scholar, 32Kwok R.P. Lundblad J.R. Chrivia J.C. Richards J.P. Bachinger H.P. Brennan R.G. Roberts S.G. Green M.R. Goodman R.H. Nature. 1994; 370: 223-226Crossref PubMed Scopus (1280) Google Scholar). Following phosphorylation of its N-terminal sites, but not the C-terminal sites, c-Jun can bind CBP and CBP can potentiate its ability to activate transcription(31Arias J. Alberts A.S. Brindle P. Claret F.X. Smeal T. Karin M. Feramisco J. Montminy M. Nature. 1994; 370: 226-229Crossref PubMed Scopus (680) Google Scholar). CBP is postulated to connect the phosphorylated activation domains of CREB or c-Jun to the basal transcriptional machinery. Interestingly, the sequence surrounding the N-terminal phosphoacceptors of c-Jun is also conserved in the C-terminal activation domain of c-Fos (33Sutherland J.A. Cook A. Bannister A.J. Kouzarides T. Genes & Dev. 1992; 6: 1810-1819Crossref PubMed Scopus (99) Google Scholar), suggesting that phosphorylation at Thr-232, the homolog of Ser-73 of c-Jun, potentiates c-Fos transcriptional activity. This prediction had turned out to be correct. However, despite the similarity between the two phosphoacceptor sites, Thr-232 of c-Fos is not phosphorylated by either JNK1 or JNK2 but by a novel 88-kDa MAPK termed FRK(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar). Like the ERKs and the JNKs, FRK is a proline-directed kinase, whose activity is rapidly stimulated in response to Ha-Ras activation by growth factors. Although the mechanism by which phosphorylation at Thr-232 stimulates c-Fos transcriptional activity is not clear, in the context of a c-Jun·c-Fos heterodimer, phosphorylation of each protein makes a similar contribution to stimulation of transcriptional activity, suggesting that both activation domains interact with the transcriptional machinery. A similar situation may apply for c-Jun·ATF2 heterodimers, as ATF2 phosphorylation at Thr-63 and Thr-71 within its N-terminal activation domain was recently shown to stimulate its transcriptional activity (22Dérijard B. Hibi M. Wu I.-H. Barrett T. Su B. Deng T. Karin M. Davis R.J. Cell. 1994; 76: 1-20Abstract Full Text PDF PubMed Scopus (2953) Google Scholar). Like c-Jun, ATF2 is also phosphorylated by the JNKs(24Gupta S. Campbell D. Derijard B. Davis R.J. Science. 1995; 267: 389-393Crossref PubMed Scopus (1336) Google Scholar). Transactivation by ATF2 is also potentiated upon binding of Rb or E1A, probably through recruitment of additional activation domains to the DNA-bound ATF2 dimer(34Kim S.J. Wagner S. Liu F. O'Reilly M.A. Robbins P.D. Green M.R. Nature. 1992; 358: 331-334Crossref PubMed Scopus (212) Google Scholar). Both E1A and Rb act in concert with phosphorylation of ATF2(24Gupta S. Campbell D. Derijard B. Davis R.J. Science. 1995; 267: 389-393Crossref PubMed Scopus (1336) Google Scholar). Although E1A can induce c-jun transcription(20van Dam H. Duyndam M. Rottier R. Bosch A. de Vries-Smits L. Herrlich P. Zantema A. Angel P. van der Eb A.J. EMBO J. 1993; 12: 479-487Crossref PubMed Scopus (342) Google Scholar), it represses AP-1 activity(35Offringa R. Gebel S. van Dam H. Timmers M. Smits A. Zwart R. Stein B. Bos J.L. van der Eb A. Herrlich P. Cell. 1990; 62: 527-538Abstract Full Text PDF PubMed Scopus (99) Google Scholar). This repression could be mediated through competition for CBP, which is very similar to the p300 E1A binding protein(36Arany F. Sellers W.R. Livingstone D.M. Eickner R. Cell. 1994; 97: 799-800Abstract Full Text PDF Scopus (370) Google Scholar). Indeed, it was recently shown that p300 and CBP are functionally interchangeable and that E1A can inhibit the coactivation function of both factors(37Arany Z. Newsome D. Oldread E. Livingston D.M. Eckner R. Nature. 1995; 374: 81-84Crossref PubMed Scopus (490) Google Scholar, 38Lundblad J.R. Kwok R.P.S. Laurance M.E. Harter M.L. Goodman R.H. Nature. 1995; 374: 85-88Crossref PubMed Scopus (531) Google Scholar). As described above, three different types of MAPKs, the ERKs, the JNKs, and FRKs, contribute to induction of AP-1 activity in response to a diverse array of extracellular stimuli. It is of considerable interest that each of these types of MAPKs is affecting AP-1 activity through phosphorylation of a different substrate (Fig. 2). While the ERKs phosphorylate TCF/Elk-1 and thereby induce c-Fos synthesis, they do not phosphorylate c-Jun or c-Fos on sites that potentiate their transcriptional activities(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar, 17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar, 29Chou S. Baichval V. Ferrell Jr., J.E. Mol. Biol. Cell. 1992; 3: 1117-1180Crossref PubMed Scopus (59) Google Scholar). In addition, the ERKs do not appear to be involved in ATF2 phosphorylation (24Gupta S. Campbell D. Derijard B. Davis R.J. Science. 1995; 267: 389-393Crossref PubMed Scopus (1336) Google Scholar). The JNKs, on the other hand, phosphorylate the stimulatory sites of c-Jun and ATF2 but do not phosphorylate c-Fos(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar, 24Gupta S. Campbell D. Derijard B. Davis R.J. Science. 1995; 267: 389-393Crossref PubMed Scopus (1336) Google Scholar, 28Hibi M. Lin A. Smeal T. Minden A. Karin M. Genes & Dev. 1993; 7: 2135-2148Crossref PubMed Scopus (1708) Google Scholar). The JNKs are also capable of phosphorylation and activation of TCF/Elk-1, suggesting they may be involved in c-fos induction under certain circumstances.2 2M. Cavigelli and M. Karin, unpublished results. FRK, so far, is only known to affect c-Fos activity(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar). These results clearly indicate that MAPKs are highly specific in their choice of substrates and do not phosphorylate just any Ser or Thr residue that is followed by a Pro, as previously assumed. The molecular mechanisms underlying this high degree of substrate specificity are being explored with the Jun-JNK interaction as a paradigm. Efficient phosphorylation by the JNKs requires a docking site located between residues 30 and 60 of c-Jun(28Hibi M. Lin A. Smeal T. Minden A. Karin M. Genes & Dev. 1993; 7: 2135-2148Crossref PubMed Scopus (1708) Google Scholar, 39Adler V. Franklin C.C. Kraft S. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 5341-5345Crossref PubMed Scopus (138) Google Scholar). In vitro, this site mediates binding of c-Jun to the JNKs, and although c-Jun·JNK complexes have not yet been detected in living cells, the integrity of the docking site is essential for phosphorylation and stimulation of c-Jun activity(28Hibi M. Lin A. Smeal T. Minden A. Karin M. Genes & Dev. 1993; 7: 2135-2148Crossref PubMed Scopus (1708) Google Scholar). The docking site is not the only feature of c-Jun that ensures efficient phosphorylation by the JNKs, because in vitro JunB also binds to the JNKs through a similar region but is not phosphorylated by them.3 3T. Deng, T. Kallunki, and M. Karin, unpublished results. JunB is not phosphorylated by the JNKs because its homologs of Ser-63 and Ser-73 of c-Jun are not followed by prolines. Once prolines are inserted after these serines in JunB, the resulting variant becomes JNK-responsive.4 4T. Deng and M. Karin, unpublished results. In addition to the docking site and Pro at the P+1 position, efficient phosphorylation of Jun proteins by the JNKs requires specific residues flanking the phosphoacceptor site. These residues, however, are not a part of the docking site and do not affect JNK binding.3 When the binding of the two human JNKs to c-Jun was compared, JNK2 was found to bind much better than JNK1(40Kallunki T. Su B. Tsigelny I. Sluss H.K. Dérijard B. Moore G. Davis R. Karin M. Genes & Dev. 1994; 8: 2996-3007Crossref PubMed Scopus (593) Google Scholar). Consequently, the Km of JNK2 toward c-Jun is lower than the Km of JNK1 toward c-Jun and its Vm is higher(40Kallunki T. Su B. Tsigelny I. Sluss H.K. Dérijard B. Moore G. Davis R. Karin M. Genes & Dev. 1994; 8: 2996-3007Crossref PubMed Scopus (593) Google Scholar). The catalytic properties of JNK2 measured with other substrates are not considerably different from those of JNK1, and JNK1 may be the more effective kinase for other substrates. The basis for the higher affinity of JNK2 toward c-Jun was traced to a small region of approximately 20 residues located near its catalytic pocket(40Kallunki T. Su B. Tsigelny I. Sluss H.K. Dérijard B. Moore G. Davis R. Karin M. Genes & Dev. 1994; 8: 2996-3007Crossref PubMed Scopus (593) Google Scholar). This region, which is variable among all MAPKs, is not a part of the catalytic pocket itself. Most likely, it is the element of JNK2 that interacts with the docking sites on c-Jun, as illustrated in Fig. 3. In addition to the differences in substrate specificities, the three types of MAPKs that affect AP-1 activity differ in their responses to extracellular stimuli. The ERKs are most efficiently stimulated by growth factors and phorbol esters(6Cobb M. Goldsmith E.J. J. Biol. Chem. 1995; 270: 14843-14846Abstract Full Text Full Text PDF PubMed Scopus (1659) Google Scholar, 17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar, 41Kyriakis J.M. Banerjee P. Nikolakaki E. Dai T. Rubie E.A. Ahmad M.F. Avruch J. Woodgett J.R. Nature. 1994; 369: 156-160Crossref PubMed Scopus (2411) Google Scholar), whereas FRK responds to growth factors but not to phorbol esters(9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar). Neither FRK nor ERK activities are considerably stimulated by exposure to UV irradiation or tumor necrosis factor, stimuli that are effective for JNK activation (9Deng T. Karin M. Nature. 1994; 371: 171-175Crossref PubMed Scopus (318) Google Scholar, 17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar, 14Gille H. Sharrocks A. Shaw P. Nature. 1992; 358: 414-417Crossref PubMed Scopus (814) Google Scholar). Compared with the FRKs and the ERKs, JNK activity is modestly stimulated by growth factors(17Minden A. Lin A. Smeal T. Dérijard B. Cobb M. Davis R. Karin M. Mol. Cell. Biol. 1994; 14: 6683-6688Crossref PubMed Scopus (436) Google Scholar, 38Lundblad J.R. Kwok R.P.S. Laurance M.E. Harter M.L. Goodman R.H. Nature. 1995; 374: 85-88Crossref PubMed Scopus (531) Google Scholar, 41Kyriakis J.M. Banerjee P. Nikolakaki E. Dai T. Rubie E.A. Ahmad M.F. Avruch J. Woodgett J.R. Nature. 1994; 369: 156-160Crossref PubMed Scopus (2411) Google Scholar). The largest increases in JNK activity are observed after UV irradiation (22Dérijard B. Hibi M. Wu I.-H. Barrett T. Su B. Deng T. Karin M. Davis R.J. Cell. 1994; 76: 1-20Abstract Full Text PDF PubMed Scopus (2953) Google Scholar, 28Hibi M. Lin A. Smeal T. Minden A. Karin M. Genes & Dev. 1993; 7: 2135-2148Crossref PubMed Scopus (1708) Google Scholar) or costimulatory activation of T cells(40Kallunki T. Su B. Tsigelny I. Sluss H.K. Dérijard B. Moore G. Davis R. Karin M. Genes & Dev. 1994; 8: 2996-3007Crossref PubMed Scopus (593) Google Scholar). Although all three types of kinases are stimulated in response to Ras activation(19Angel P. Hattori K. Smeal T. Karin M. Cell. 1988; 55: 875-885Abstract Full Text PDF PubMed Scopus (990) Google Scholar, 42Minden A. Lin A. McMahon M. Lange-Carter C. Derijard B. Davis R.J. Johnson G.L. Karin M. Science. 1994; 226: 1719-1723Crossref Scopus (1011) Google Scholar, 43Robbins D.J. Cheng M. Zhen E. Vanderbilt C.A. Feig L.A. Cobb M.H. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 6924-6928Crossref PubMed Scopus (148) Google Scholar, 44Thomas S.M. De Marco M. D'Arcangelo G. Halegoua S. Brugge J.S. Cell. 1992; 68: 1031-1040Abstract Full Text PDF PubMed Scopus (503) Google Scholar), the JNKs also respond to Ras-independent signals(42Minden A. Lin A. McMahon M. Lange-Carter C. Derijard B. Davis R.J. Johnson G.L. Karin M. Science. 1994; 226: 1719-1723Crossref Scopus (1011) Google Scholar). However, even Ras activation affects ERK and JNK through different kinase cascades (Fig. 4). The major pathway leading from Ras to ERK is based on the Ras-mediated recruitment of Raf-1 to the plasma membrane(45Leevers S.J. Paterson H.F. Marshall C.J. Nature. 1994; 369: 411-414Crossref PubMed Scopus (881) Google Scholar). This results in activation of Raf-1, a Ser/Thr kinase that phosphorylates and activates the dual specificity kinases MEK1 and MEK2 (46Dent P. Hasar W. Haystead T. Vincent L. Roberts T. Sturgill T. Science. 1992; 254: 1404-1407Crossref Scopus (498) Google Scholar). The latter are responsible for phosphorylation and activation of the ERKs(47Crews C.G. Alessandrini A. Erickson R.L. Science. 1992; 258: 478-480Crossref PubMed Scopus (738) Google Scholar). Ras also activates MEKK1, a Ser/Thr kinase unrelated in its primary structure to Raf(48Lange-Carter C.A. Pleiman C. Gardner A. Blumer K. Johnson G. Science. 1993; 260: 315-319Crossref PubMed Scopus (873) Google Scholar, 49Lange-Carter C.A. Johnson G.L. Science. 1994; 265: 1458-1461Crossref PubMed Scopus (295) Google Scholar). So far, no direct interaction between Ras and MEKK1 has been observed, and it is not clear how it is activated by Ras. Although in vitro MEKK1 is an efficient MEK activator(48Lange-Carter C.A. Pleiman C. Gardner A. Blumer K. Johnson G. Science. 1993; 260: 315-319Crossref PubMed Scopus (873) Google Scholar, 49Lange-Carter C.A. Johnson G.L. Science. 1994; 265: 1458-1461Crossref PubMed Scopus (295) Google Scholar), in vivo it mostly activates JNKK1 (also called SEK1 or MKK4), a dual specificity kinase that phosphorylates and activates the JNKs(42Minden A. Lin A. McMahon M. Lange-Carter C. Derijard B. Davis R.J. Johnson G.L. Karin M. Science. 1994; 226: 1719-1723Crossref Scopus (1011) Google Scholar, 50Sanchez I. Hughes R.T. Mayer B.J. Yee K. Woodgett J.R. Arrach J. Kyrriakes J.M. Zon L.I. Nature. 1994; 372: 794-798Crossref PubMed Scopus (916) Google Scholar, 51Dérijard B. Raingeaud J. Barrett T. Wu I.H. Han J. Ulevitch R.J. Davis R.J. Science. 1995; 267: 682-685Crossref PubMed Scopus (1407) Google Scholar, 52Lin A. Minden A. Martinetto H. Claret F.X. Lange-Carter C. Mercurio F. Johnson G.L. Karin M. Science. 1995; 268: 286-290Crossref PubMed Scopus (708) Google Scholar). In the future it would be of interest to identify the factors that contribute to limiting the specificity of MEKK1 action and confine it to the JNK activation cascade. Raf-1, on the other hand, has only a marginal effect on JNKK1 activity, and none of the MEKs can activate the JNKs (42Minden A. Lin A. McMahon M. Lange-Carter C. Derijard B. Davis R.J. Johnson G.L. Karin M. Science. 1994; 226: 1719-1723Crossref Scopus (1011) Google Scholar, 52Lin A. Minden A. Martinetto H. Claret F.X. Lange-Carter C. Mercurio F. Johnson G.L. Karin M. Science. 1995; 268: 286-290Crossref PubMed Scopus (708) Google Scholar). Likewise, JNKK1 does not activate the ERKs(51Dérijard B. Raingeaud J. Barrett T. Wu I.H. Han J. Ulevitch R.J. Davis R.J. Science. 1995; 267: 682-685Crossref PubMed Scopus (1407) Google Scholar, 52Lin A. Minden A. Martinetto H. Claret F.X. Lange-Carter C. Mercurio F. Johnson G.L. Karin M. Science. 1995; 268: 286-290Crossref PubMed Scopus (708) Google Scholar). Thus, the kinase cascades responsible for ERK and JNK activation are distinct and exhibit very little cross-talk (Fig. 4). While a great deal remains to be learned about the mechanisms that contribute to the regulation of AP-1 activity and most of the important target genes whose expression is modulated by the different forms of AP-1 are yet to be identified, quite a lot has been revealed so far by focusing on this transcription factor and its response to extracellular stimuli. Most importantly, the investigation of AP-1 regulation had revealed some of the general mechanisms by which protein phosphorylation modulates transcription factor activity (reviewed in Ref. 53Hunter T. Karin M. Cell. 1992; 70: 375-387Abstract Full Text PDF PubMed Scopus (1118) Google Scholar) and the strategies used by cell surface receptors to communicate with the nucleus (reviewed in Ref. 54Hill C.S. Treisman R. Cell. 1995; 80: 199-211Abstract Full Text PDF PubMed Scopus (1195) Google Scholar). In addition to the identification of important AP-1 target genes that will explain the physiological functions of the different forms of this transcription factor, a major challenge for the future is understanding the mechanisms that confer biological specificity to the actions of protein kinases and transcription factors. It is clear that even generic and ubiquitous signaling proteins like the components of AP-1 and the MAPK cascades can be involved in highly specific biological responses. I thank P. Alford for preparation of the manuscript and Drs. T. Kallunki and B. Su for preparation of the figures.