Oncogene (2001) 20, 2438 ± 2452 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

Close encounters of many kinds: Fos-Jun interactions that mediate transcription regulatory speci®city

Yurii Chinenov1 and Tom K Kerppola*,1

1Howard Hughes Medical Institute, University of Michigan Medical School Ann Arbor, Michigan, MI 48109-0650, USA

Fos and Jun family proteins regulate the expression of a Wang et al., 1992). However, only a few of the genes myriad of genes in a variety of tissues and cell types. that mediate the essential functions of speci®c Fos and This functional versatility emerges from their interac- Jun family members have been identi®ed (Bakin and tions with related bZIP proteins and with structurally Curran, 1999; Fu et al., 2000) unrelated transcription factors. These interactions at Fos and Jun family proteins function as dimeric composite regulatory elements produce nucleoprotein transcription factors that bind to AP-1 regulatory complexes with high sequence-speci®city and regulatory elements in the promoter and enhancer regions of selectivity. Several general principles including binding numerous mammalian genes (Curran and Franza, cooperativity and conformational adaptability have 1988). Jun proteins form both homodimers and emerged from studies of regulatory complexes containing heterodimers with Fos proteins, whereas Fos proteins Fos-Jun family proteins. The structural properties of do not form homodimers and require heterodimeriza- Fos-Jun family proteins including opposite orientations tion to bind DNA. The DNA-binding and dimerization of heterodimer binding and the ability to bend DNA can domains among di€erent family members are highly contribute to the assembly and functions of such conserved and di€erent members of the Fos and Jun complexes. The cooperative recruitment of transcription families have similar DNA-binding and dimerization factors, coactivators and chromatin remodeling factors to speci®cities. promoter and enhancer regions generates multiprotein In vitro, dimers formed by Fos and Jun bind with the transcription regulatory complexes with cell- and highest anity to an asymmetric heptanucleotide stimulus-speci®c transcriptional activities. The gene- recognition sequence TGA(C/G)TCA (AP-1) and with speci®c architecture of these complexes can mediate the slightly lower anity, to a symmetric octanucleotide selective control of transcriptional activity. Oncogene TGACGTCA (CRE) (Nakabeppu et al., 1988; (2001) 20, 2438 ± 2452. Rauscher et al., 1988). The AP-1 site is a ubiquitous regulatory element that is found in a wide range of Keywords: transcription factor cooperativity; nucleo- promoter and enhancer regions. Since the AP-1 site protein complex architecture; transcriptional synergy; and variants thereof occur with a high frequency in the composite regulatory elements; orientation of hetero- genome, it is unlikely that Fos-Jun family proteins dimer binding regulate all genes that contain AP-1 recognition sequences. Conversely, many genes that are bona ®de regulatory targets of Fos-Jun family proteins do not contain consensus AP-1 recognition sequences within Regulatory functions of Fos and Jun family transcription their control regions. In natural promoter and factors enhancer regions, the sequences of AP-1 regulatory elements often deviate from the optimal recognition Members of the Fos and Jun protein families (Curran sequence. This variation in recognition sequences may and Teich, 1982; Maki et al., 1987) participate in the contribute to the di€erential functions of di€erent Fos- regulation of a variety of cellular processes including Jun family dimers at various regulatory elements cell proliferation, di€erentiation, apoptosis and onco- (McBride and Nemer, 1998) The weaker binding genesis. Members of this family (Fos, Fra-1, Fra-2, anities of Fos-Jun family members at these non- FosB, Jun, JunB and JunD) are widely expressed in consensus recognition sites may also impose a require- diverse cell types and tissues. The results of gene ment for interactions with other transcription factors. knockout experiments indicate that Fos and Jun family The functions of Fos-Jun family proteins depend on proteins have both overlapping functions as well as the speci®c cell type in which they are expressed. Only unique roles that cannot be compensated for by other a small subset of all potential regulatory targets is family members (Grigoriadis et al., 1994; Hilberg et al., controlled by Fos-Jun family proteins in any particular 1993; Johnson et al., 1992, 1993; Schorpp-Kistner et cell type. The activities and regulatory targets of Fos- al., 1999; Schreiber et al., 2000; Thepot et al., 2000; Jun family proteins are also a€ected by the speci®c signals that elicit their expression. Thus, the functions of Fos-Jun family proteins must be mediated by *Correspondence: TK Kerppola mechanisms that depend on the cellular context in Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2439 which they are expressed. Several mechanisms that may progression of di€erent homo- and heterodimeric Fos contribute to the cell type speci®city of Fos-Jun family and Jun family protein complexes is thought to proteins can be envisioned. These include di€erential contribute to the time-dependent induction of di€erent post-translational modi®cations, selective dimerization early and late genes in di€erent cell types activated by between di€erent family members and interactions with di€erent stimuli. other regulatory protein. The ®rst two mechanisms modulate the activities of Fos-Jun proteins, but they have mostly indirect e€ects on selection of the genes Interaction between Fos-Jun proteins and other that are regulated by Fos-Jun family proteins in a bZIP family proteins particular cell type. Interactions with other transcrip- tion factors can modify the regulatory speci®cities of Fos-Jun family proteins are members of a large group Fos-Jun family proteins in a cell or tissue speci®c of transcription factors (the bZIP family) containing a manner. Thus, Fos-Jun proteins have to be considered highly conserved basic region involved in DNA binding in the broad context of dynamically changing protein- and a heptad repeat of leucine residues, the leucine protein interactions on and o€ DNA. zipper, required for dimerization (Landschulz et al., Interactions between various Fos-Jun family members 1988). All bZIP proteins form dimeric complexes and more than 50 di€erent proteins have been reported through the leucine zipper that juxtapose the two basic (Table 1). Fos-Jun interacting proteins can be sub- regions to form a contiguous DNA-contact interface in divided into four groups: (1) structurally related basic which each monomer interacts with the major groove region ± leucine zipper proteins; (2) unrelated DNA of one half-site (Ellenberger et al., 1992; Fujii et al., binding proteins; (3) transcriptional coactivators that do 2000; Glover and Harrison, 1995; Schumacher et al., not bind DNA directly; and (4) structural components 2000). Since members of di€erent bZIP protein of the nucleus. In many cases, either the functional subfamilies exhibit distinct DNA binding speci®cities, signi®cance or the structural basis of the interaction dimerization between Fos-Jun and other bZIP proteins remains under investigation. Nevertheless, it is clear that expands the repertoire of binding sites for Fos-Jun interactions among many structurally divergent protein family proteins to include sequences composed of families can contribute to the functional speci®city of di€erent half-sites. Diversi®cation of binding speci®- Fos-Jun family proteins. In this review, we focus on cities through the formation of cross-family dimers is a interactions between Fos-Jun family members and other common characteristic of bZIP proteins. Cross-family DNA binding proteins that can in¯uence the regulatory dimerization between Fos-Jun proteins and various speci®cities of Fos-Jun family proteins. members of the ATF, C/EBP, Maf and NF-E2 (CNC) proteins (Table 1) has been observed.

Interactions among Fos and Jun family members ATF proteins The repertoire of Fos-Jun proteins in a given cell is Several members of the ATF group of bZIP proteins subject to changes in response to various extracellular form heterodimers with Fos and Jun family proteins. stimuli. Through dimerization mediated by the leucine The ATF2 and ATF3 proteins preferentially interact zipper, the seven Fos-Jun family members can form 18 with CRE (TGACGTCA) rather than AP-1 sites. di€erent homo and heterodimers. The number of Heterodimerization between Fos-Jun and these pro- detectable Fos-Jun dimers varies among di€erent cell teins can target Fos-Jun proteins to CRE-like sites (Hai types (Kovary and Bravo, 1991; Lallemand et al., 1997; and Curran, 1991). Many of the promoters that are McCabe et al., 1995; Sonnenberg et al., 1989). controlled by heterodimers formed by ATF2 with Fos- Quiescent ®broblasts contain mainly homodimers Jun family proteins contain asymmetric binding sites formed by Jun and JunD, but following serum (TTACCTCA in the Jun promoter, TGACATAG in stimulation, heterodimers formed ®rst by Fos and the b-interferon promoter and CTCAGTCA in the FosB and later by Fra-1 with Jun and JunB become adenovirus E2A promoter). Neither Fos-Jun nor Jun- the predominant AP-1 binding proteins (Kovary and Jun bind eciently to these sites (Du et al., 1993; van Bravo, 1992; Lallemand et al., 1997). In exponentially Dam et al., 1993). Since Jun and ATF proteins can be growing ®broblasts, heterodimers formed by Fra-2 activated by di€erent families of protein-kinases (Jnk with Jun and JunD are the predominant AP-1 binding and p38 SAPK respectively) dimerization between complexes (Lallemand et al., 1997). In the mouse brain, these two protein families may mediate the integration metrazole-induced seizures cause a similar progression of signals from di€erent signal transduction pathways of di€erent heterodimers formed ®rst by Fos (Sonnen- (van Dam et al., 1998). berg et al., 1989). In undi€erentiated osteoblasts all seven members of Fos-Jun family are detectable but MAF and NF-E2 (CNC) family proteins during di€erentiation JunD, Fra-1 and Fra-2 become the predominant AP-1 binding proteins (McCabe et al., Dimerization of Fos-Jun proteins with members of the 1996). Similarly, during keratinocyte di€erentiation the Maf and NF-E2 (CNC) families further expands the repertoire of Fos-Jun family member undergoes several range of Fos-Jun protein target sites. Maf and NF-E2 temporally regulated shifts (Eckert et al., 1997). This (CNC) proteins form homo and heterodimers with

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2440 Table 1 Proteins that can interact with Fos-Jun family members Protein family Jun JunB JunD Fos FosB Fra-1 Fra-2

Jun Y Y Y Y Y Y Y JunB Y Y Y Y Y Y Y JunD Y Y Y Y Y Y Y Fos Y Y Y N N N N Fra-1 Y Y Y N N N N Fra-2 Y Y Y N N N N ATF-2 ATF/bZIP Y Y Kerppola and Curran, 1993 ATF-3/LRF-1 ATF/bZIP Y Y Hai and Curran, 1991 ATF-4 ATF/bZIP Y Y Y Hai and Curran, 1991; Kerppola and Curran, 1993 ATFa ATF/bZIP Y De Graeve et al., 1999 B-ATF ATF/bZIP Y Y Echlin et al., 2000 cMaf Maf/bZIP Y Y N Y Kataoka et al., 1994b; Li et al., 1999 MafB MafbZIP N Y Kataoka et al., 1994a MafA Maf/bZIP Y Y Benkhelifa et al., 1998 Maf G/F/K Maf/bZIP Y Y Kataoka et al., 1995; 1996 Nrl Maf/bZIP Y Y Kerppola and Curran, 1994b Nrf-1 CNC/bZIP Y Y Y Venugopal and Jaiswal, 1998 Nrf-2 CNC/bZIP Y Y Venugopal and Jaiswal, 1998 NF-IL6 (C/EBP b) C/EBP/bZIP Y Y Hsu et al., 1994 CHOP (GADD153) C/EBP/bZIP Y Y Y Ubeda et al., 1999 Meq bZIP Y Qian et al., 1996 p21SNFT bZIP Y Y Iacobelli et al., 2000 MyoD bHLH Y Li et al., 1992 Myogenin bHLH Y Li et al., 1992 USF-1 bHLH N N N Y N Pognonec et al., 1997 MITF bHLH Y Sato et al., 1999 Elf-1 ETS Y Y Y Bassuk and Leiden, 1995 PU.1 ETS Y Y Y Bassuk and Leiden, 1995 Ets-1 ETS Y Y Y Bassuk and Leiden, 1995; Basuyaux et al., 1997 Ets-2 ETS Y Y Basuyaux et al., 1997 Fli-1 ETS Y Y Y Bassuk and Leiden, 1995 Erg ETS Y Y Buttice et al., 1996 SMAD3 SMAD Y Y Y Zhang et al., 1998 SMAD4 SMAD Y Y Y Liberati et al., 1999 NFAT1 (NFATC2, NFATp) NFAT Y Y Y Y Jain et al., 1993a; Chen et al., 1998 NFAT2 (NFATC1, NFATc) NFAT Wolfe et al., 1997 NFAT3 (NFATC4) NFAT Y Hoey et al., 1995 NFAT4 (NFATC3, NFATx) NFAT Y Hoey et al., 1995 P65 NF-kB Rel Y Y Stein et al., 1993 SP1 SP Y Chen and Chang, 2000 GATA-2 GATA Y Y Kawana et al., 1995 STAT Y Zhang et al., 1999 TBP Y Y Metz et al., 1994 Oct-1 Ullman et al., 1993 Glucocorticoid Receptor Y Y Diamond et al., 1990; Kerppola et al., 1993 SPBP Y N Kirstein et al., 1996 pRB Y Nead et al., 1998 CBP/p300 Y Y Bannister and Kouzarides, 1995 Jab1 Y Y Claret et al., 1996 SRC-1 Y Y Lee et al., 1998 alpha-NAC Y Moreau et al., 1998 QM/Jif-1 Y Monteclaro and Vogt, 1993 SMRT Y Y Lee et al., 2000a p202 Y Y Min et al., 1996 ASC-2 (PRIP) Y Y Lee et al., 2000b Menin N N Y N N N N Agarwal et al., 1999 JDP-2 Y Aronheim et al., 1997 HTLV Tax Baranger et al., 1995 HBV pX Y Perini et al., 1999 BEF Y Virbasius et al., 1999 Ubinuclein Y Aho et al., 2000

slightly di€erent binding speci®cities. Maf homodimers members requires an atypical non-a-helical basic region recognize palindromic binding sites consisting of and an ancillary DNA binding region that mediates TGCtgaC half-sites (capital letters indicate the most major groove contacts with the extended recognition conserved base pairs) (Kataoka et al., 1994b; Kerppola elements (Dlakic et al., 2001). Fos-Jun proteins can and Curran, 1994b). Recognition of the exceptionally interact with at least six members of the Maf family (c- long (13 ± 14 base pair) binding sites by Maf family Maf, MafB, NRL, MafG, MafF and MafK) two

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2441 proteins from the NF ± E2 (CNC) family (Nrf1 and to di€erent signal transduction pathways. Their Nrf2) (Kataoka et al., 1994a,b; Kerppola and Curran, concerted action at composite regulatory elements can 1994a; Venugopal and Jaiswal, 1998). Fos-Maf and therefore integrate multiple extracellular signals at a Jun-NRL dimers preferentially interact with recogni- speci®c promoter. tion sites composed of AP-1 (TGAC) and Maf The three-dimensional structures of two multiprotein (TGCtgaC) half-sites (Kerppola and Curran, 1994b). transcription factor complexes at composite regulatory Heterodimers formed between small Maf proteins elements have been solved including the Fos-Jun- (MafF, MafG and MafK) and the p45 NF-E2 NFAT-ARRE2 and MATa-MCM-STE6 complexes preferentially bind recognition sequences composed of (Chen et al., 1998; Tan and Richmond, 1998). Several TGCtgaC and (a/g)TGAC half-sites (Kataoka et al., features common to these complexes may also apply to 1994a,b). JunB and Maf synergistically activate the IL4 other higher-order complexes and can explain coop- promoter during T-helper cell di€erentiation (Li et al., erative DNA binding by transcription factors at 1999). However, since the IL4 promoter contains both composite regulatory elements. The interactions be- a weak AP1 site and a Maf binding site, the synergistic tween cooperating DNA binding proteins frequently activation may re¯ect cooperation between two homo- involve regions in a close proximity to DNA. Thus, the dimers bound to separate sites. DNA binding domains alone can be sucient for Heterodimers formed by Jun and Nrf proteins cooperative complex formation at composite regulatory participate in the regulation of class II detoxi®cation elements. Transcription factor binding to adjacent sites enzymes in response to xenobiotics and phenolic can create an uninterrupted protein-DNA interface antioxidants through interactions with antioxidant extending across both recognition elements, thereby response elements (ARE) found in the promoter increasing the speci®city and anity of DNA binding. regions of the NAD(P)H:quinone oxidoreductase, The DNA and protein conformations are often altered heme oxygenase-1, g-glutamylcysteine synthetase and to form the protein-protein and protein-DNA interac- glutathione S-transferase A2 genes (Alam et al., 2000; tion interfaces. These conformational rearrangements Jaiswal, 2000; Moinova and Mulcahy, 1999; Nguyen et may contribute to the selectivity of multiprotein al., 2000; Rushmore and Pickett, 1990). The ARE complex formation. sequence (a/gTGACnnnGC) is similar to that recog- Cooperative DNA binding frequently requires a nized by p45 NF ± E2 ± Maf heterodimers. In nuclear precise spacing and orientation of the individual extracts, Jun, Nrf1, Nrf2 and small Maf proteins can binding sites that comprise a composite regulatory all bind to the ARE site. Since neither Nrf1 nor Nrf2 element. Since AP-1 binding sites contain symmetrical form homodimers and Jun generally does not interact half-sites, Fos-Jun heterodimers can potentially bind to with ARE alone, they have been proposed to form regulatory elements in two opposite orientations heterodimers at this site (Moinova and Mulcahy, 1999; (Figure 1). Both orientations are observed in the Fos- Venugopal and Jaiswal, 1998). Jun-AP-1 crystal (Glover and Harrison, 1995). In solution, sequences ¯anking the core AP-1 recognition element determine the preferred orientation of Fos-Jun Fos-Jun interactions with structurally unrelated binding (Leonard et al., 1997; Rajaram and Kerppola, DNA-binding proteins 1997; Leonard and Kerppola, 1998; Ramirez-Carrozzi and Kerppola, 2001b). Fos-Jun heterodimers that bind Synergistic interactions with non-bZIP DNA-binding in opposite orientations present di€erent surfaces for proteins can further increase the combinatorial poten- interactions with proteins that bind to adjacent tial of Fos-Jun dependent transcription regulation. regulatory elements. The preferred orientation of Fos- Several transcription factor families including NFAT, Jun binding can therefore in¯uence interactions with Ets, Smad and bHLH can activate or repress proteins bound to adjacent recognition sites and transcription in conjunction with Fos-Jun family thereby control the eciency of transcription activation proteins by binding to regulatory elements adjacent (Chytil et al., 1998; Kerppola, 1998; Ramirez-Carrozzi to AP-1 sites. One of the mechanisms underlying this and Kerppola, 2001a). synergy involves protein-protein interactions between Fos-Jun family members and other transcription Fos-Jun-NFAT complexes factors at composite regulatory elements. Composite regulatory elements function as a single unit and The DNA binding domain of NFAT family proteins is possess activities over and above those of their distantly related to that of Rel family proteins and constituent binding sites. Multiprotein complexes adopts a structure similar to the s-type immunoglobu- formed at composite regulatory elements are more lin fold (Chen et al., 1998). Unlike Rel proteins, NFAT stable than the complexes formed by individual typically binds DNA as a monomer. NFAT family transcription factors at their respective recognition proteins recognize a GGAAaa sequence elements (Jain sites. The cooperative interaction can also extend the et al., 1993a; McCa€rey et al., 1993; Rao et al., 1997). recognition speci®city of the complex to regulatory NFAT proteins are involved in the regulation of a elements that are not bound by the individual proteins. variety of genes that are induced during T-cell The multiple transcription factors that bind to activation (Shaw et al., 1988; reviewed in Rao et al., composite regulatory elements often mediate responses 1997). The promoters of many of these genes contain

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2442 adjacent NFAT and AP-1 binding sites (Rao et al., proteins do not appear to interact with each other in 1997, Table 2). The cooperative binding of Fos-Jun the absence of DNA. The Fos-Jun-NFAT1 ternary and NFAT proteins to these sites greatly enhances the complex bound to the ARRE2 site in the IL-2 DNA binding anity and the transcriptional activity promoter (Table 2) spans a 15 bp composite site in of the ternary complex compared to either NFAT or which all base pairs including the intervening TT linker Fos-Jun alone (Jain et al., 1993a; McCa€rey et al., are contacted by either Fos-Jun or NFAT1 (Chen et 1993; Peterson et al., 1996). NFAT and Fos-Jun family al., 1998). The cooperative interaction between Fos-Jun

Figure 1 Fos-Jun heterodimer bind to an asymmetric AP1 site (TGAGTCA) in two orientations (Glover and Harrison, 1995). The bZIP domain of Jun is colored blue and the bZIP domain of Fos is colored red. The DNA sugar-phosphate backbone is colored yellow. The sequence of the AP-1 recognition element is shown below the structures. The two structures represent opposite orientations of Fos-Jun binding to the same DNA sequence, and re¯ect rotation of the heterodimer by approximately 1808 about the dimer axis

Table 2 Composite AP1/NFAT regulatory elements in promoter and enhancer regions IL-2 ARRE2 7280 AGGAAAaac - - TGTTCA Jain et al., 1993b 7135 AGGAAAaat - - GAAGGTA Rao et al., 1997 790 TTGAAAta - - - TGTGTAA Chen et al., 1998

GM-CSF enhancer GM330 cGGAGGccc - - TGAGTCA Rao et al., 1997 GM420 tGGAAAga - - TGACATCA GM550 aGGAAAgca - - AGAGTCA

IL-3 595 tGGAAAct - GTAGCTCAr 576 Duncliffe et al., 1997

IL-4 human 769 tGGAAAttt - TCGTTACAr Rooney et al., 1995 mouse 768 tGGAAAtt - TTATTACAr

IL5 human 7115 tGGAAAcattTAGTTTCA Stranick et al., 1997

CTLA4 human 7195 tGGAAAAt - - - GTACTCAr Perkins et al., 1996 mouse 7207 tGGAAAAt - - - GTATTCAr TNFa human 789 GGAGAAACCCATGAGCTCA Tsai et al., 1996

The sequences of NFAT and CRE sites were aligned and introduced gaps are indicated by dashes. The sequences of AP1/CRE sites are shown in boldface, the sequence of NFAT binding site is underlined. r indicated that sequence is shown for non-coding strand

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2443 and NFAT1 induces a DNA bend of approximately required to bring NFAT1 and Fos-Jun together as they 208 toward the interaction interface (Chen et al., 1998; would not contact each other on straight DNA. Both Diebold et al., 1998) (Figure 2a). This DNA bend is Fos-Jun and NFAT undergo conformational changes

Figure 2 Comparison of the structure of the Fos-Jun-NFAT1-ARRE2 complex with the structures of Fos-Jun-AP-1 and NFAT2- ARRE2 complexes. (a) Structure of the Fos-Jun-NFAT1-ARRE2 complex (Chen et al., 1998). The bZIP domain of Fos is shown in red, Jun in blue, and NFAT1 in green. The DNA binding site for NFAT1 (GGAAAA) is colored orange and the DNA binding site for AP-1 (TGTTTCA) is shown in purple. The white line shows the DNA helix axis. (b) Superposition of the bZIP domains of Fos and Jun in the complex with NFAT1 (red and blue respectively) with Fos and Jun bound to the AP-1 site (Glover and Harrison, 1995) (purple and cyan respectively, marked with asterisks) show that the leucine zipper is tilted toward NFAT1, shown in grey) in the Fos-Jun-NFAT1-ARRE2 complex. (c) Superposition of NFAT1 in the Fos-Jun-NFAT1-ARRE2 complex with NFAT2 in the NFAT2-ARRE2 complex (Zhou et al., 1998)

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2444 to form the interaction interface (Figure 2). The leucine Ets and Fos-Jun proteins do not appear to require zipper of the Fos-Jun heterodimer is tilted by prior DNA binding. DNA-independent interactions approximately 158 toward NFAT1 (Figure 2b). The between Elf-1, Pu.1, Ets-1, Ets-2, Erg, Fli-1 and Jun conformation of NFAT in the ternary complex may proteins (cJun, JunD and JunB) have been detected in also change relative to the binary complex (Figure 2c). vitro by co-immunoprecipitation and GST pull down NFAT1 forms an extensive interaction interface with assays. In stimulated human T cells antibodies against the Fos-Jun heterodimer involving one face of the all seven members of Fos-Jun protein family co- leucine zipper. The interaction between NFAT1 and precipitated endogenous Elf-1 (Bassuk and Leiden, Fos-Jun is asymmetric and requires a speci®c orienta- 1995). Whereas Jun family proteins can also interact tion of Fos-Jun binding (Chen et al., 1995, 1998; with Elf-1 in vitro, no direct interaction between Elf-1 Kerppola, 1998). Three amino acid residues at the and Fos proteins (cFos, Fra-1 and Fra-2) has been amino-terminal ends of the leucine zippers of Fos and observed, suggesting that this interaction is mediated Jun are the principal determinants of the asymmetric by Jun. The interaction between Jun and Elf-1 requires interactions with NFAT1 (Diebold et al., 1998). The the basic region of Jun and the Ets domain of Elf-1 other amino acid residues in the leucine zippers and the (Bassuk and Leiden, 1995; Basuyaux et al., 1997). hinge regions that contact NFAT1 in the crystal Replacement of the basic region of Jun by that of Fos structure can be exchanged between Fos and Jun with abolishes its interaction with Elf-1 and reduces no e€ect on the asymmetric interaction with NFAT1 transcription activation of a reporter construct contain- (Diebold et al., 1998). Since these amino acid residues ing multiple composite AP-1/ETS elements (Bassuk are not conserved between Fos and Jun, the amino and Leiden, 1995). Interactions with Ets family acid residues at the amino-terminal ends of the leucine proteins are not unique to Fos-Jun family members, zippers must constitute the primary interaction inter- but have been observed also for Maf family proteins in face with NFAT1. the absence of DNA (Sieweke et al., 1996). Despite the The Fos-Jun-NFAT complex is a convenient model functional synergy and physical contacts between Fos- for analysis of the dynamics of nucleoprotein complex Jun and Ets family proteins, no cooperative complex assembly and reorganization. Multicomponent nucleo- between these proteins has been observed at any of the protein complexes may form either through random composite elements examined. collisions or through an ordered pathway. Cooperative Composite AP-1/ETS regulatory elements from DNA binding by Fos-Jun and NFAT requires a various promoters display some similarity in sequence speci®c orientation of heterodimer binding. Since organization (Table 3). Simultaneous binding of Fos- Fos-Jun heterodimers can bind to many composite Jun and Ets proteins could create a continuous binding regulatory elements in both orientations, heterodimers interface in the major grove spanning the 13 ± 14 bound in the reverse orientation must be reoriented nucleotide composite site. The interactions between subsequent to NFAT binding. Fos-Jun heterodimers Fos-Jun and Ets proteins in such a complex could could be reoriented either through stochastic dissocia- involve the basic region of Jun and the C-terminal tion and re-binding; or through a de®ned reorientation portion of the DNA-binding helix of the Ets domain pathway. Studies of the dynamics of Fos-Jun hetero- (Figure 3). This region is well conserved in most Ets dimer reorientation by NFAT1 in solution suggest the proteins, which might account for the apparent existence of a pathway for Fos-Jun reorientation in promiscuity of Fos-Jun interactions with Ets proteins. association with the composite regulatory element Since the ETS binding site is asymmetric, the (Ramirez-Carrozzi and Kerppola, 2001a). interactions with Ets proteins might also be a€ected by the orientation of Fos-Jun binding. Nonetheless, simultaneous binding by Fos-Jun and Ets family Fos-Jun interactions with Ets family proteins proteins has not been detected to date. Following the discovery of synergistic activation of the polyoma virus enhancer through a composite AP-1/ Fos-Jun interactions with Smad family proteins ETS site (Martin et al., 1988) both regulatory interactions between ETS and AP-1 binding sites in In contrast to NFAT and Ets family proteins, which various promoters as well as physical contacts between can bind DNA and activate transcription at indepen- Fos-Jun and Ets family proteins have been observed dent regulatory elements, Smad proteins bind DNA (Bassuk and Leiden, 1995; Graves and Petersen, 1998). with low anity and speci®city in the absence of The Ets proteins comprise a large protein family cooperating binding partners. Therefore Smad proteins characterized by a conserved 80 ± 90 amino acid rely on interactions with other DNA binding proteins DNA binding domain termed the Ets domain. The to target them to speci®c regulatory elements (Chen et Ets domain recognizes a short sequence element al., 1996; Derynck et al., 1998; Shi et al., 1998). Smad containing a GGAA/T tetranucleotide core (Graves proteins play a central role in TGFb induced signal and Petersen, 1998). This domain folds into a compact transduction. Binding of TGFb to a membrane surface helix ± turn ± helix like structure with a single helix that receptor activates its intracellular Ser/Thr kinase forms base speci®c contacts in the major groove of activity, resulting in phosphorylation of associated DNA (Kodandapani et al., 1996) (Figure 3). In contast Smad2 and Smad3. The phosphorylated Smad proteins to the Fos-Jun-NFAT complex, interactions between dissociate from the receptor and interact with

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2445

Figure 3 Superposition of the Fos-Jun heterodimer and the ETS binding domain of Pu.1 (Kodandapani et al., 1996) on the AP1/ ETS composite site from the polyoma virus enhancer. The bZIP domain of Fos is shown in red, Jun in blue, and the ETS domain of Pu.1 in green. The DNA binding helix of the ETS domain potentially involved in protein-protein contacts with the basic region of Jun is colored white. The sugar-phosphate backbone of DNA is shown in yellow

Table 3 Composite AP1/ETS regulatory elements in promoter and enhancer regions GMCSF promoter 755caTTAATCA------TTTCCTtaa Wang et al., 1994 Macrophage 767aaTGTGTCA------TTTCCTtt Wu et al., 1994 Scavenger receptor Moulton et al., 1994 r Distal enhancer aaTGACTAA------TTTCCTtt 74500 r uPA human gaTGACCTCA ------TTTCCTcc 71976 Nerlov et al., 1992 r mouse TGACCTCA ------TTTCCT 72400 D'Orazio et al., 1997 human 75350 gaTGATTCA------CTTCCCtt mouse 76871 ggTGATTCA------CTTCCTtt r Polyomavirus enhancer agTTAGTCA------CTTCCTgc 5131 Martin et al., 1988 r MMP9 ttTGACTCAg------CTTCCTct 7542 Gum et al., 1996 hIL-3 7303gcTGAGTCAgg------CTTCCCCTTCCTgc Gottschalk et al., 1993 CD11c 752gcTGACAATctt ------CTTCCTtcccc Noti et al., 1996 r RANTES agTGAGCTCA tcag ------TTTCCTtc 7208 Boehlk et al., 2000 r hTNFa caTGAGCTCA tctgg -----AGGAAGcg 7119 Kramer et al., 1995 TIMP1 7856gaTGAGTAAt g c gtcc ---AGGAAGCC Logan et al., 1996 r MMP1 cTGACTCAtgctttataa CATCCTct 795 Gutman and Wasylyk, 1990

The sequences of Ets and CRE sites were aligned and introduced gaps are indicated by dashes. The sequences of AP1/CRE sites are shown in boldface, the sequences of Ets binding sites are underlined. r indicated that sequence is shown for non-coding strand

cytoplasmic Smad4. Upon co-translocation to the an amino-terminal MH1 domain that is responsible for nucleus, these complexes interact with composite DNA-binding and a carboxyl-terminal MH2 domain regulatory elements in association with various DNA involved in transcription activation and oligomeriza- binding partners, including FAST-1, TEF3 and Jun tion. The DNA-binding MH1 domain recognizes a (Chen et al., 1996; Hua et al., 1998; Zhang et al., 1998). tetranucleotide GTCT recognition element employing a Smad proteins contain two highly conserved domains; unique b-hairpin DNA-binding motif (Shi et al., 1998).

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2446 Smad proteins participate in the regulation of many substitutions in Smad3 that in¯uence the interaction genes including junB,c-jun, Smad7 and collagenase-1 in with Jun (L43A, L43S, K46A in Smad4) also a€ect response to TGFb family signaling molecules (Brodin DNA binding or nuclear localization (L43S), a more et al., 2000; Jonk et al., 1998; Wong et al., 1999; Zhang general e€ect of these substitutions on protein et al., 1998). In some of these promoters, the SMAD conformation cannot be excluded (Jones and Kern, binding sites are adjacent to AP-1 or CRE sites (Table 2000). A similar interaction between ATF2 and Smad4 4). Fos-Jun family members and Smad3/Smad4 display involves the bZIP domain of ATF2 (Sano et al., 1999). strong transcriptional synergy at these sites in response Replacement of a single leucine in the leucine zipper of to TGFb. This synergy may be mediated by physical ATF2 by a valine interferes with ATF2 homodimeriza- interactions between Smad3 and c-Jun (Zhang et al., tion and abolishes Smad 4-ATF2 interactions in vitro, 1998). As in the case of NFAT and Ets family proteins, suggesting that the Smad binding determinant in ATF2 Smad-Jun interactions involve the DNA-binding is a€ected by the dimerization state (Sano et al., 1999). domains of both proteins. Site-directed mutagenesis Superposition of Smad3 and Fos-Jun on a composite of the MH1 domain as well as analysis of naturally binding site from the collagenase promoter shows that occurring mutants suggest that residues involved in the two proteins bound to adjacent sites would clash in DNA-binding by Smad proteins as well as basic the major groove. This suggests that major conforma- residues adjacent to the DNA contact interface are tional changes would be necessary to accommodate important for the interaction with Jun (Figure 4) (Qing simultaneous binding by both proteins. Nonetheless, as et al., 2000). However, since most of the amino acid in the case of Fos-Jun interactions with Ets proteins,

Table 4 Composite AP1/SMAD regulatory elements in promoter and enhancer regions MMP1 (collagenase I) aaAGCCAGAGGTTGCTGACTCActr 773 Zhang et al., 1998 SMAD7 7337 gcGTCTAGACggccacgTGACGAGgc Brodin et al., 2000 germline Ig a 7121 AGACcacaggccAGACaTGACGTGgg Zhang and Derynck, 2000

The sequences of SMADEts and CRE sites were aligned and introduced gaps are indicated by dashes. The sequences of AP1/CRE sites are shown in boldface, the sequences of SMAD binding sites are underlined. r indicated that sequence is shown for the non-coding strand

Figure 4 Superposition of the Fos-Jun heterodimer and the MH1 DNA binding domain of Smad3 (Shi et al., 1998) on the AP1/ Smad composite element from the collagenase promoter. The bZIP domains of Fos and Jun are colored as in Figure 3. The MH1 domain of Smad3 is shown in green. The DNA binding b-hairpin in MH1 domain is shown in white. Simultaneous binding by both protein results in a steric clash between the DNA binding b-hairpin of Smad3 and the basic region in the heterodimer (indicated by a circle)

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2447 no ternary complex between Fos-Jun and Smad family interdependent and require the concerted function of proteins has been observed. all of the proteins in the complex (Robertson et al., Transcription factors from the STAT, bHLH and 1995). The multiprotein regulatory complexes can steroid hormone receptor families also display func- coordinate several steps in transcription activation, tional interactions with members of the Fos-Jun family including covalent histone modi®cation, nucleosome (Diamond et al., 1990; Li et al., 1992; Pognonec et al., remodeling and pre-initiation complex assembly (Aga- 1997; Zhang et al., 1999) (Table 1). The molecular lioti et al., 2000; Spicuglia et al., 2000). mechanisms that mediate these interactions remain The assembly of transcription regulatory complexes under investigation, and in several cases direct physical can be facilitated both by cooperative binding of interactions have been identi®ed although no coopera- multiple proteins to the regulatory region as well as by tive complex has been observed at the composite changes in DNA and chromatin structure. DNA regulatory elements examined. Other potential mechan- bending can contribute to the assembly of regulatory isms of cooperativity between Fos-Jun family proteins complexes by facilitating interactions between proteins and other transcription factors include assisted folding that bind to separate recognition elements or by of DNA-binding domain of Fos-Jun, which is dis- altering the conformations of protein-nucleic acid ordered in solution in the absence of DNA. Alter- complexes. Fos and Jun family proteins induce natively the formation of speci®c heterodimers could be opposite directions of DNA bending (Kerppola and stabilized, providing a mechanism for the selective Curran, 1991, 1993; Rajaram and Kerppola, 1997). dimerization between speci®c Fos-Jun family members The opposite DNA bending properties control the in the cell. These proteins may also bind DNA orientation of Fos-Jun heterodimer binding, and can independently, but cooperate in recruitment or activa- in¯uence the transcriptional activities of Fos-Jun tion of additional components of the transcription heterodimers at composite regulatory elements (Leo- complex. nard and Kerppola, 1998; Ramirez-Carrozzi and Interactions between transcription factors can also Kerppola, 2001a,b). They can also contribute to the mediate negative interference between di€erent signal- e€ects of the sequence of the DNA binding site on the ing pathways. The inhibitory interactions between Fos- transcriptional activities of Fos-Jun family proteins Jun family and steroid hormone receptor superfamily (Lefstin and Yamamoto, 1998; Ramirez-Carrozzi and members have been investigated in greatest detail Kerppola, 2001a,b). (Diamond et al., 1990; Konig et al., 1992; Schule et al., 1990). There are likely to be many mechanisms The TCRa enhancer involved in the complex regulatory relationships between these transcription factor families (Caelles et Several well-studied transcription regulatory complexes al., 1997; Kerppola et al., 1993; Lee et al., 2000b). Such assembled over promoter and enhancer regions illus- negative interactions may be required to coordinate the trate these principles. Assembly of such a complex at e€ects of di€erent signal transduction pathways and to the TCRa enhancer involves at least four di€erent balance responses to competing environmental signals. transcription factors including PEB2a, Ets-1, LEF-1 Thus, there are numerous mechanisms whereby tran- and ATF2 or CREB (Giese et al., 1995; Mayall et al., scription factors can in¯uence the functions of each 1997). LEF-1 induces a large DNA bend at the center other, and further studies of the interactions involving of the 70 base pair enhancer region which facilitates Fos-Jun family proteins are required to identify the long-range interactions between ATF2 or CREB and mechanisms that mediate their functional interrelation- Ets-1 proteins. Binding of CREB to the CRE site ships. disrupts the nucleosomal organization around the TCRa enhancer based on micrococal nuclease digestion and DNAase I hyper-sensitivity (Mayall et al., 1997). Multiprotein transcription regulatory complexes at promoter and enhacer regions The IFN-b promoter Multiprotein transcription regulatory complexes Activation of IFN-b transcription in response to viral formed at promoters and enhancers can involve infection is mediated by a multiprotein regulatory transcription factors bound to both simple and complex at a 55-bp enhancer region. Viral infection composite regulatory elements, coactivators and chro- results in cooperative assembly of a multiprotein matin remodeling factors. Assembly of such higher- enhanceosome complex containing ATF2-Jun, IRF order complexes often requires a precise spatial proteins, NF-kB p50 ± p65, HMGI(Y) and p300/CBP arrangement of transcription regulatory elements and (Du et al., 1993; Merika et al., 1998; Thanos and is accompanied by changes in DNA structure and Maniatis, 1995). Binding of HMGI(Y) to several sites chromatin organization (Agalioti et al., 2000; Giese et within the enhancer modi®es DNA structure and al., 1995; Hernandez-Munain et al., 1998). Mutation of facilitates interactions among the proteins bound to even a single regulatory element can disrupt assembly separate recognition sites (Yie et al., 1999). Recruit- of the multiprotein complex and alter transcription ment of CBP/p300 is thought to be mediated by regulation. Consequently, the functions of individual multidentate interactions with several of the transcrip- transcription regulatory proteins in such complexes are tion factors in the complex.

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2448 Assembly of the IFN-b enhanceosome requires a ATF2/Jun heterodimer and forms a composite element ®xed spatial arrangement of the binding sites within the with the k3-NFAT site (Table 2 and Figure 5). There enhancer region. This likely re¯ects the requirement for are two ETS binding sites upstream (7117 NFAT/ a precise alignment of the interaction surfaces between ETS) and downstream (784 ETS/Elk) from the CRE, the individual proteins in the complex. Transcription a downstream NFAT/ETS binding site (776 NFAT/ activation by the IFN-b enhancer also requires a ETS) and a downstream SP1 binding site (Figure 5) speci®c orientation of ATF2-Jun heterodimer binding (Tsai et al., 1996a,b, 2000; Falvo et al., 2000a, c). In (Falvo et al., 2000b). Cooperative activation of IFN-b di€erent cell types activated by di€erent extracellular transcription requires interactions between the bZIP stimuli such as viral infection, LPS or T cell activation, domain of ATF-2 and the DNA-binding domain of distinct multiprotein transcription regulatory com- IRF3. When ATF2-Jun heterodimers bind in the plexes assemble at this promoter. The CRE, the opposite orientation, they fail to recruit IRF3, and 7117 NFAT/ETS, k3-NFAT, 776 NFAT/ETS and do not facilitate the assembly of a functional the downstream SP1 element are required for inducer- enhancesome (Falvo et al., 2000b). Substitution of speci®c transcription of TNFa in macrophages, mono- the ATF2-Jun binding site in the IFN-b enhancer by cytes, T and B cells (Table 2). Mutation of any of these similar regulatory elements from the c-jun promoter or binding sites either abolishes or signi®cantly reduces the urokinase plasminogen activator enhancer resulted LPS inducibility. In LPS-stimulated monocytes, ATF2- in orientation-dependent function. Therefore, regula- Jun, Ets-1, Elk-1, SP1 can all bind to the TNFa tion of these genes may also require a speci®c promoter based on chromatin immunoprecipitation orientation of ATF2-Jun binding. assays (Tsai et al., 2000). Thus, activation of the TNFa promoter likely requires simultaneous occupancy of the promoter by multiple transcription factors from the The TNFa promoter Fos-Jun, Ets and SP1 families. The TNFa promoter is be regulated by multiprotein In T cells activated by ionophore treatment, a complexes similar to that described at the IFN-b di€erent set of transcription factors is involved in the promoter (Falvo et al., 2000a; Tsai et al., 1996b; 2000). activation of the TNFa enhancer. ATF2-Jun is The region involved in the regulation of TNFa constitutively bound to the enhancer and NFAT binds transcription contains a CRE binding site, which binds to multiple sites in the promoter in response to

Figure 5 Promoter and enhancer regions regulated by multiprotein transcription regulatory complexes. The nucleotide sequences and transcription factor binding sites in the TCR-a enhancer (Giese et al., 1995) the INF-b promoter (Du et al., 1993; Merika et al., 1998; Thanos and Maniatis, 1995), the TNF-a enhancer (Falvo et al., 2000c; Tsai et al., 2000), and the uPA enhancer (Nerlov et al., 1992) are shown. Stimulus-speci®c transcription factor interacting with TNF-a enhancer are indicated with red arrows for LPS stimulation and with blue arrows for calcium signaling

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2449 ionophore stimulation (Tsai et al., 1996b). The down- respond in a purposeful manner to the virtually in®nite stream 776 NFAT/ETS site as well as the 7117 variety of environmental signals that contribute to NFAT/ETS site interact with NFAT rather than Ets long-term changes in cellular phenotype. This challenge proteins. In the case of virus stimulation of T cells, SP1 is particularly evident for transcription factors such as is inducibly recruited. Thus, di€erent extracellular those in the Fos-Jun family that are expressed in a signals can stimulate the assembly of di€erent variety of di€erent cell types and that respond to regulatory protein complexes with di€erent transcrip- diverse extracellular signals. These proteins therefore tional activities at the same regulatory region in provide useful model systems for investigation of di€erent cell types (Falvo et al., 2000c; Tsai et al., mechanisms of speci®city in transcription regulation. 2000). The studies to date have demonstrated the critical role that interactions between di€erent transcription factor families play in control of the regulatory The uPA enhancer speci®cities of Fos-Jun family proteins. There are likely Regulation of urokinase plasminogen activator (uPA) to be numerous mechanisms whereby interactions transcription depends of an enhancer region located between di€erent transcription factors families in¯u- approximately 2000 bp upstream of the start site ence regulatory speci®city. Future studies of the (Nerlov et al., 1992). The uPA enhancer contains a various transcription factor interactions are required composite CRE/ETS site and an AP1 site separated by to identify the full range of strategies used by the cell a 74 base pair sequence that is essential for synergy to control regulatory speci®city. Most studies of between the CRE/ETS and AP-1 sites. This sequence transcription factor interactions thus far have been contains binding sites that can interact with the limited to analysis of complex formation in vitro and homeodomain protein PREP and the POU protein studies of regulatory interactions in transfected cells. Oct1 (De Cesare et al., 1997; Palazzolo et al., 2000). These studies must be extended to de®ne the mechan- Although the sequence separating the CRE/ETS and isms that are responsible for synergistic interactions AP-1 sites does not activate transcription alone, between di€erent combinations of transcription factors mutations in the binding sites in this region reduce at multiple promoter and enhancer regions. Future the activity of the uPA enhancer. The Ets, Fos-Jun and studies should also address the selectivity and dynamics Oct-1 proteins also synergistically activate the human of transcription factor interactions in living cells and pro®llagrin promoter, suggesting that regulatory inter- extend the functional analysis to investigation of the actions among these proteins may represent a common roles of endogenous proteins under physiologically mechanism for the integration of regulatory signals relevant conditions. The progress over the past few (Jang et al., 2000). The mechanisms of cooperation years in this ®eld justi®es the optimistic view that we among Oct-1, Fos-Jun and Ets proteins remain to be will attain an understanding of the principal mechan- elucidated. No direct interactions between Fos-Jun isms that mediate the regulatory speci®city of Fos-Jun proteins and Oct-1 or between Ets proteins and Oct-1 family proteins under physiological conditions in the have been reported. Thus, additional studies of this near future. and other promoter regions are required to elucidate how Fos and Jun family proteins elicit transcriptional responses in concert with many other transcription regulatory proteins. Acknowledgments We apologize to the many people whose work could not be cited due to space limitations. We thank Mensur Dlakic, Perspectives Asya Grinberg and Nirmala Rajaram for helpful discus- sions during preparation of the manuscript and Ann E Goldfeld (Harvard Medical School), Andy Sharroks The selective regulation of each of more than 30 000 (Manchester University), Laura McCabe (Michigan State genes in the human genome is a demanding task for University), Helen Moinova (Case Western Reserve Uni- the regulatory machinery of the cell. Even more versity), Moshe Yaniv (Institute Pasteur) and Thomas daunting is the requirement that each gene must Herdegen (University of Kiel) for comments.

References

Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T and Alam J, Wicks C, Stewart D, Gong P, Touchard C, Otterbein Thanos D. (2000). Cell, 103, 667 ± 678. S, Choi AM, Burow ME and Tou J. (2000). J. Biol. Chem., Agarwal SK, Guru SC, Heppner C, Erdos MR, Collins RM, 275, 27694 ± 27702. Park SY, Saggar S, Chandrasekharappa SC, Collins FS, Aronheim A, Zandi E, Hennemann H, Elledge SJ and Karin Spiegel AM, Marx SJ and Burns AL. (1999). Cell, 96, M. (1997). Mol. Cell. Biol., 17, 3094 ± 3102. 143 ± 152. Bakin AV and Curran T. (1999). Science, 283, 387 ± 390. Aho S, Buisson M, Pajunen T, Ryoo YW, Giot JF, Gru€at Bannister AJ and Kouzarides T. (1995). EMBO J., 14, 4758 ± H, Sergeant A and Uitto J. (2000). J. Cell. Biol., 148, 4762. 1165 ± 1176.

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2450 Baranger AM, Palmer CR, Hamm MK, Giebler HD, Fu SL, Waha A and Vogt PK. (2000). Oncogene, 19, 3537 ± Brauweiler A, Nyborg JK and Schepartz A. (1995). 3545. Nature, 376, 606 ± 608. Fujii Y, Shimizu T, Toda T, Yanagida M and Hakoshima T. Bassuk AG and Leiden JM. (1995). Immunity, 3, 223 ± 237. (2000). Nat. Struct. Biol., 7, 889 ± 893. Basuyaux JP, Ferreira E, Stehelin D and Buttice G. (1997). J. Giese K, Kingsley C, Kirshner JR and Grosschedl R. (1995). Biol. Chem., 272, 26188 ± 26195. Genes Dev., 9, 995 ± 1008. Benkhelifa S, Provot S, Lecoq O, Pouponnot C, Calothy G Glover JN and Harrison SC. (1995). Nature, 373, 257 ± 261. and Felder-Schmittbuhl MP. (1998). Oncogene, 17, 247 ± Gottschalk LR, Giannola DM and Emerson SG. (1993). J. 254. Exp. Med., 178, 1681 ± 1692. Boehlk S, Fessele S, Mojaat A, Miyamoto NG, Werner T, Graves BJ and Petersen JM. (1998). Adv. Cancer Res., 75, 1± Nelson EL, Schlondor€ D and Nelson PJ. (2000). Eur. J. 55. Immunol., 30, 1102 ± 1112. Grigoriadis AE, Wang ZQ, Cecchini MG, Hofstetter W, Brodin G, Ahgren A, ten Dijke P, Heldin CH and Heuchel R. Felix R, Fleisch HA and Wagner EF. (1994). Science, 266, (2000). J. Biol. Chem., 275, 29023 ± 29030. 443 ± 448. Buttice G, Duterque-Coquillaud M, Basuyaux JP, Carrere S, Gum R, Lengyel E, Juarez J, Chen JH, Sato H, Seiki M and Kurkinen M and Stehelin D. (1996). Oncogene, 13, 2297 ± Boyd D. (1996). J. Biol. Chem., 271, 10672 ± 10680. 2306. Gutman A and Wasylyk B. (1990). EMBO J., 9, 2241 ± 2246. Caelles C, Gonzalez-Sancho JM and Munoz A. (1997). Genes Hai T and Curran T. (1991). Proc. Natl. Acad. Sci. USA, 88, Dev., 11, 3351 ± 3364. 3720 ± 3724. Chen BK and Chang WC. (2000). Proc. Natl. Acad. Sci. Hernandez-Munain C, Roberts JL and Krangel MS. (1998). USA, 97, 10406 ± 10411. Mol. Cell. Biol., 18, 3223 ± 3233. Chen L, Glover JN, Hogan PG, Rao A and Harrison SC. Hilberg F, Aguzzi A, Howells N and Wagner EF. (1993). (1998). Nature, 392, 42 ± 48. Nature, 365, 179 ± 181. Chen L, Oakley MG, Glover JN, Jain J, Dervan PB, Hogan Hoey T, Sun YL, Williamson K and Xu X. (1995). Immunity, PG, Rao A and Verdine GL. (1995). Curr. Biol., 5, 882 ± 2, 461 ± 472. 889. Hsu W, Kerppola TK, Chen PL, Curran T and Chen-Kiang Chen X, Rubock MJ and Whitman M. (1996). Nature, 383, S. (1994). Mol. Cell. Biol., 14, 268 ± 276. 691 ± 696. Hua X, Liu X, Ansari DO and Lodish HF. (1998). Genes Chytil M, Peterson BR, Erlanson DA and Verdine GL. Dev., 12, 3084 ± 3095. (1998). Proc. Natl. Acad. Sci. USA, 95, 14076 ± 14081. Iacobelli M, Wachsman W and McGuire KL. (2000). J. Claret FX, Hibi M, Dhut S, Toda T and Karin M. (1996). Immunol., 165, 860 ± 868. Nature, 383, 453 ± 457. Jain J, McCa€rey PG, Miner Z, Kerppola TK, Lambert JN, Curran T and Teich NM. (1982). J. Virol., 42, 114 ± 122. Verdine GL, Curran T and Rao A. (1993a). Nature, 365, Curran T and Franza BR. (1988). Cell, 55, 395 ± 397. 352 ± 355. De Cesare D, Palazzolo M, Berthelsen J and Blasi F. (1997). Jain J, Miner Z and Rao A. (1993b). J. Immunol., 151, 837 ± J. Biol. Chem., 272, 23921 ± 23929. 848. De Graeve F, Bahr A, Sabapathy KT, Hauss C, Wagner EF, Jaiswal AK. (2000). Free Radic. Biol. Med., 29, 254 ± 262. Kedinger C and Chatton B. (1999). Oncogene, 18, 3491 ± Jang SI, Karaman-Jurukovska N, Morasso MI, Steinert PM 3500. and Markova NG. (2000). J. Biol. Chem., 275, 15295 ± Derynck R, Zhang Y and Feng XH. (1998). Cell, 95, 737 ± 15304. 740. Johnson RS, Spiegelman BM and Papaioannou V. (1992). Diamond MI, Miner JN, Yoshinaga SK and Yamamoto KR. Cell, 71, 577 ± 586. (1990). Science, 249, 1266 ± 1272. Johnson RS, van Lingen B, Papaioannou VE and Spiegel- Diebold RJ, Rajaram N, Leonard DA and Kerppola TK. man BM. (1993). Genes Dev., 7, 1309 ± 1317. (1998). Proc. Natl. Acad. Sci. USA, 95, 7915 ± 7920. Jones JB and Kern SE. (2000). Nucleic Acids Res., 28, 2363 ± Dlakic M, Grinberg AV, Leonard DA and Kerppola TK. 2368. (2001). EMBO J., 20, 1±14. Jonk LJ, Itoh S, Heldin CH, ten Dijke P and Kruijer W. D'Orazio D, Besser D, Marksitzer R, Kunz C, Hume DA, (1998). J. Biol. Chem., 273, 21145 ± 21152. Kiefer B and Nagamine Y. (1997). Gene, 201, 179 ± 187. KataokaK,FujiwaraKT,NodaMandNishizawaM. Du W, Thanos D and Maniatis T. (1993). Cell, 74, 887 ± 898. (1994a). Mol. Cell. Biol., 14, 7581 ± 7591. Duncli€e KN, Bert AG, Vadas MA and Cockerill PN. Kataoka K, Igarashi K, Itoh K, Fujiwara KT, Noda M, (1997). Immunity, 6, 175 ± 185. Yamamoto M and Nishizawa M. (1995). Mol. Cell. Biol., Echlin DR, Tae HJ, Mitin N and Taparowsky EJ. (2000). 15, 2180 ± 2190. Oncogene, 19, 1752 ± 1763. Kataoka K, Noda M and Nishizawa M. (1994b). Mol. Cell. Eckert RL, Crish JF, Banks EB and Welter JF. (1997). J. Biol., 14, 700 ± 712. Invest. Dermatol., 109, 501 ± 509. Kataoka K, Noda M and Nishizawa M. (1996). Oncogene, Ellenberger TE, Brandl CJ, Struhl K and Harrison SC. 12, 53 ± 62. (1992). Cell, 71, 1223 ± 1237. Kawana M, Lee ME, Quertermous EE and Quertermous T. Falvo JV, Brinkman BM, Tsytsykova AV, Tsai EY, Yao TP, (1995). Mol. Cell. Biol., 15, 4225 ± 4231. Kung AL and Goldfeld AE. (2000a). Proc. Natl. Acad. Kerppola TK. (1998). Structure, 6, 549 ± 554. Sci. USA, 97, 3925 ± 3929. Kerppola TK and Curran T. (1991). Cell, 66, 317 ± 326. Falvo JV, Parekh BS, Lin CH, Fraenkel E and Maniatis T. Kerppola TK and Curran T. (1993). Mol. Cell. Biol., 13, (2000b). Mol. Cell. Biol., 20, 4814 ± 4825. 5479 ± 5489. Falvo JV, Uglialoro AM, Brinkman BM, Merika M, Parekh Kerppola TK and Curran T. (1994a). Oncogene, 9, 3149 ± BS,TsaiEY,KingHC,MorielliAD,PeraltaEG, 3158. Maniatis T, Thanos D and Goldfeld AE. (2000c). Mol. Kerppola TK and Curran T. (1994b). Oncogene, 9, 675 ± 684. Cell. Biol., 20, 2239 ± 2247.

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2451 KerppolaTK,LukDandCurranT.(1993).Mol. Cell. Biol., Monteclaro FS and Vogt PK. (1993). Proc. Natl. Acad. Sci. 13, 3782 ± 3791. USA, 90, 6726 ± 6730. KirsteinM,SanzL,QuinonesS,MoscatJ,Diaz-MecoMT Moreau A, Yotov WV, Glorieux FH and St-Arnaud R. and Saus J. (1996). J. Biol. Chem., 271, 18231 ± 18236. (1998). Mol. Cell. Biol., 18, 1312 ± 1321. Kodandapani R, Pio F, Ni CZ, Piccialli G, Klemsz M, Moulton KS, Semple K, Wu H and Glass CK. (1994). Mol. McKercher S, Maki RA and Ely KR. (1996). Nature, 380, Cell. Biol., 14, 4408 ± 4418. 456 ± 460. Nakabeppu Y, Ryder K and Nathans D. (1988). Cell, 55, Konig H, Ponta H, Rahmsdorf HJ and Herrlich P. (1992). 907 ± 915. EMBO J., 11, 2241 ± 2246. Nead MA, Baglia LA, Antinore MJ, Ludlow JW and Kovary K and Bravo R. (1991). Mol. Cell. Biol., 11, 2451 ± McCance DJ. (1998). EMBO J., 17, 2342 ± 2352. 2459. Nerlov C, De Cesare D, Pergola F, Caracciolo A, Blasi F, Kovary K and Bravo R. (1992). Mol. Cell. Biol., 12, 5015 ± JohnsenMandVerdeP.(1992).EMBO J., 11, 4573 ± 5023. 4582. Kramer B, Wiegmann K and Kronke M. (1995). J. Biol. Nguyen T, Huang HC and Pickett CB. (2000). J. Biol. Chem., Chem., 270, 6577 ± 6583. 275, 15466 ± 15473. Lallemand D, Spyrou G, Yaniv M and Pfarr CM. (1997). Noti JD, Reinemann C and Petrus MN. (1996). Mol. Oncogene, 14, 819 ± 830. Immunol., 33, 115 ± 127. Landschulz WH, Johnson PF and McKnight SL. (1988). Palazzolo M, Berthelsen J, De Cesare D and Blasi F. (2000). Science, 240, 1759 ± 1764. Eur. J. Biochem., 267, 5427 ± 5437. Lee SK, Kim HJ, Na SY, Kim TS, Choi HS, Im SY and Lee Perini G, Oetjen E and Green MR. (1999). J. Biol. Chem., JW. (1998). J. Biol. Chem., 273, 16651 ± 16654. 274, 13970 ± 13977. Lee SK, Kim JH, Lee YC, Cheong J and Lee JW. (2000a). J. Perkins D, Wang Z, Donovan C, He H, Mark D, Guan G, Biol. Chem., 275, 12470 ± 12474. Wang Y, Walunas T, Bluestone J and Listman J. (1996). J. Lee SK, Na SY, Jung SY, Choi JE, Jhun BH, Cheong J, Immunol., 156, 4154 ± 4159. Meltzer PS, Lee YC and Lee JW. (2000b). Mol. Peterson BR, Sun LJ and Verdine GL. (1996). Proc. Natl. Endocrinol., 14, 915 ± 925. Acad.Sci.USA,93, 13671 ± 13676. Lefstin JA and Yamamoto KR. (1998). Nature, 392, 885 ± Pognonec P, Boulukos KE, Aperlo C, Fujimoto M, Ariga H, 888. Nomoto A and Kato H. (1997). Oncogene, 14, 2091 ± 2098. Leonard DA and Kerppola TK. (1998). Nat. Struct. Biol., 5, Qian Z, Brunovskis P, Lee L, Vogt PK and Kung HJ. (1996). 877 ± 881. J. Virol., 70, 7161 ± 7170. Leonard DA, Rajaram N and Kerppola TK. (1997). Proc. Qing J, Zhang Y and Derynck R. (2000). J. Biol. Chem., 275, Natl. Sci. USA, 13, 4913 ± 4918. 38802 ± 38812. Li B, Tournier C, Davis RJ and Flavell RA. (1999). EMBO J, Rajaram N and Kerppola TK. (1997). EMBO J., 16, 2917 ± 18, 420 ± 432. 2925. Li L, Chambard JC, Karin M and Olson EN. (1992). Genes Ramirez-Carrozzi VR and Kerppola TK. (2001a). Proc. Dev., 6, 676 ± 689. Natl. Acad. Sci. USA, 98, 4893 ± 4898. Liberati NT, Datto MB, Frederick JP, Shen X, Wong C, Ramirez-Carrozzi VR and Kerppola TK. (2001b). J. Mol. Rougier-Chapman EM and Wang XF. (1999). Proc. Natl. Biol., 305, 411 ± 427. Acad. Sci. USA, 96, 4844 ± 4849. Rao A, Luo C and Hogan PG. (1997). Annu.Rev.Immunol., Logan SK, Garabedian MJ, Campbell CE and Werb Z. 15, 707 ± 747. (1996). J. Biol. Chem., 271, 774 ± 782. Rauscher FJD, Voulalas PJ, Franza Jr, BR and Curran T. Maki Y, Bos TJ, Davis C, Starbuck M and Vogt PK. (1987). (1988). Genes Dev., 2, 1687 ± 1699. Proc. Natl. Acad. Sci. USA, 84, 2848 ± 2852. Robertson LM, Kerppola TK, Vendrell M, Luk D, Smeyne Martin ME, Piette J, Yaniv M, Tang WJ and Folk WR. RJ, Bocchiaro C, Morgan JI and Curran T. (1995). (1988). Proc. Natl. Acad. Sci. USA, 85, 5839 ± 5843. Neuron, 14, 241 ± 252. Mayall TP, Sheridan PL, Montminy MR and Jones KA. Rooney JW, Hoey T and Glimcher LH. (1995). Immunity, 2, (1997). Genes Dev., 11, 887 ± 899. 473 ± 483. McBride K and Nemer M. (1998). Mol. Cell. Biol., 18, 5073 ± Rushmore TH and Pickett CB. (1990). J. Biol. Chem., 265, 5081. 14648 ± 14653. McCabe LR, Kockz M, Lian J and Stein G. (1995). Exp. Cell Sano Y, Harada J, Tashiro S, Gotoh-Mandeville R, Res., 218, 255 ± 262. Maekawa T and Ishii S. (1999). J. Biol. Chem., 274, McCabe LR, Banerjee C, Kundu R, Harrison RJ, Dobner 8949 ± 8957. PR, Stein JL, Lian JB and Stein GS. (1996). Endocrinol- SatoM,MoriiE,Takebayashi-SuzukiK,YasuiN,OchiT, ogy, 137, 4398 ± 4408. Kitamura Y and Nomura S. (1999). Biochem. Biophys. McCa€rey PG, Luo C, Kerppola TK, Jain J, Badalian TM, Res. Commun., 254, 384 ± 387. Ho AM, Burgeon E, Lane WS, Lambert JN, Curran T, Schorpp-Kistner M, Wang ZQ, Angel P and Wagner EF. Verdine GV, Rao A and Hogan PG. (1993). Science, 262, (1999). EMBO J, 18, 934 ± 948. 750 ± 754. Schreiber M, Wang Z, Jochum W, Fetka I, Elliott C and Merika M, Williams AJ, Chen G, Collins T and Thanos D. Wagner EF. (2000). Development, 127, 4937 ± 4948. (1998). Mol. Cell., 1, 277 ± 287. Schule R, Umesono K, Mangelsdorf DJ, Bolado J, Pike JW Metz R, Bannister AJ, Sutherland JA, Hagemeier C, and Evans RM. (1990). Cell, 61, 497 ± 504. O'Rourke EC, Cook A, Bravo R and Kouzarides T. Schumacher MA, Goodman RH and Brennan RG. (2000). J. (1994). Mol. Cell. Biol., 14, 6021 ± 6029. Biol. Chem., 275, 35242 ± 35247. Min W, Ghosh S and Lengyel P. (1996). Mol. Cell. Biol., 16, ShawJP,UtzPJ,DurandDB,TooleJJ,EmmelEAand 359 ± 368. Crabtree GR. (1988). Science, 241, 202 ± 205. Moinova HR and Mulcahy RT. (1999). Biochem. Biophys. Shi Y, Wang YF, Jayaraman L, Yang H, Massague J and Res. Commun., 261, 661 ± 668. Pavletich NP. (1998). Cell, 94, 585 ± 594.

Oncogene Protein-protein interactions of Fos-Jun family members Y Chinenov and TK Kerppola 2452 Sieweke MH, Tekotte H, Frampton J and Graf T. (1996). van Dam H, Huguier S, Kooistra K, Baguet J, Vial E, van Cell, 85, 49 ± 60. der Eb AJ, Herrlich P, Angel P and Castellazzi M. (1998). Sonnenberg JL, Macgregor-Leon PF, Curran T and Morgan Genes. Dev., 12, 1227 ± 1239. JI. (1989). Neuron, 3, 359 ± 365. Venugopal R and Jaiswal AK. (1998). Oncogene, 17, 3145 ± Spicuglia S, Payet D, Tripathi RK, Rameil P, Verthuy C, 3156. Imbert J, Ferrier P and Hempel WM. (2000). EMBO J., Virbasius CM, Wagner S and Green MR. (1999). Mol. Cell., 19, 2034 ± 2045. 4, 219 ± 228. Stein B, Baldwin AS, Ballard DW, Greene WC, Angel P and Wang CY, Bassuk AG, Boise LH, Thompson CB, Bravo R Herrlich P. (1993). EMBO J., 12, 3897 ± 3891. and Leiden JM. (1994). Mol. Cell. Biol., 14, 1153 ± 1159. Stranick KS, Zambas DN, Uss AS, Egan RW, Billah MM Wang ZQ, Ovitt C, Grigoriadis AE, Mohle-Steinlein U, and Umland SP. (1997). J. Biol. Chem., 272, 16453 ± Ruther U and Wagner EF. (1992). Nature, 360, 741 ± 745. 16465. Wolfe SA, Zhou P, Dotsch V, Chen L, You A, Ho SN, Tan S and Richmond TJ. (1998). Nature, 391, 660 ± 666. Crabtree GR, Wagner G and Verdine GL. (1997). Nature, Thanos D and Maniatis T. (1995). Cell, 83, 1091 ± 1100. 385, 172 ± 176. Thepot D, Weitzman JB, Barra J, Segretain D, Stinnakre Wong C, Rougier-Chapman EM, Frederick JP, Datto MB, MG, Babinet C and Yaniv M. (2000). Development, 127, Liberati NT, Li JM and Wang XF. (1999). Mol. Cell. Biol., 143 ± 153. 19, 1821 ± 1830. Tsai EY, Jain J, Pesavento PA, Rao A and Goldfeld AE. Wu H, Moulton K, Horvai A, Parik S and Glass CK. (1994). (1996a). Mol. Cell. Biol., 16, 459 ± 467. Mol. Cell. Biol., 14, 2129 ± 2139. Tsai EY, Yie J, Thanos D and Goldfeld AE. (1996b). Mol. YieJ,MerikaM,MunshiN,ChenGandThanosD.(1999). Cell. Biol., 16, 5232 ± 5244. EMBO J., 18, 3074 ± 3089. Tsai EY, Falvo JV, Tsytsykova AV, Barczak AK, Reimold Zhang X, Wrzeszczynska MH, Horvath CM and Darnell JE, AM,GlimcherLH,FentonMJ,GordonDC,DunnIF Jr. (1999). Mol. Cell. Biol., 19, 7138 ± 7146. and Goldfeld AE. (2000). Mol. Cell. Biol., 20, 6084 ± 6094. Zhang Y and Derynck R. (2000). J. Biol. Chem., 275, 16979 ± Ubeda M, Vallejo M and Habener JF. (1999). Mol. Cell. 16985. Biol., 19, 7589 ± 7599. Zhang Y, Feng XH and Derynck R. (1998). Nature, 394, UllmanKS,NorthropJP,AdmonAandCrabtreeGR. 909 ± 913. (1993). Genes. Dev., 7, 188 ± 196. Zhou P, Sun LJ, Dotsch V, Wagner G and Verdine GL. van Dam H, Duyndam M, Rottier R, Bosch A, de Vries- (1998). Cell, 92, 687 ± 696. Smits L, Herrlich P, Zantema A, Angel P and van der Eb AJ. (1993). EMBO J., 12, 479 ± 487.

Oncogene