Asymmetric recognition of nonconsensus AP-1 sites by Fos-Jun and Jun-Jun influences transcriptional cooperativity with NFAT1

Abstract

Many regulatory elements in eukaryotic promoters do not correspond to optimal recognition sequences for the transcription factors that regulate promoter function by binding to the elements. The sequence of the binding site may influence the structural and functional properties of regulatory protein complexes. Fos-Jun heterodimers were found to bind nonconsensus AP-1 sites in a preferred orientation. Oriented Fos-Jun heterodimer binding was attributed to nonidentical recognition of the two half-sites by Fos and Jun. Jun bound preferentially to the consensus half-site, whereas Fos was able to bind nonconsensus half-sites. The orientation of heterodimer binding affected the transcriptional cooperativity of Fos-Jun-NFAT1 complexes at composite regulatory elements in mammalian cells. Jun dimerization with Fos versus ATF2 caused it to bind opposite half-sites at nonconsensus AP-1 elements. Similarly, ATF2 bound to opposite half-sites in Fos-ATF2-NFAT1 and ATF2-Jun-NFAT1 complexes. The orientations of nonconsensus AP-1 sites within composite regulatory elements affected the cooperativity of Fos-Jun as well as Jun-Jun binding with NFAT1. Since Jun homodimers cannot bind to AP-1 sites in a preferred orientation, the effects of the orientations of nonconsensus AP-1 sites on the stabilities of Jun-Jun-NFAT1 complexes are likely to be due to asymmetric conformational changes in the two subunits of the homodimer. Nonconsensus AP-1 site orientation also affected the synergy of transcription activation between Jun homodimers and NFAT1 at composite regulatory elements. The asymmetric recognition of nonconsensus AP-1 sites can therefore influence the transcriptional activities of Fos and Jun both through effects on the orientation of heterodimer binding and through differential conformational changes in the two subunits of the dimer.

FIG. 1.

Effects of base substitutions within the consensus AP-1 site on the preferred orientation of Fos-Jun heterodimer binding. The orientation preferences of Fos-Jun heterodimers at AP-1 sites containing symmetry-related base substitutions were determined using the gelFRET approach. In this approach, heterodimers labeled with an acceptor fluorophore on either Fos or Jun were incubated with oligonucleotides labeled on opposite ends with a donor fluorophore, and the complexes were analyzed on a polyacrylamide gel. The relative efficiencies of energy transfer from opposite ends (end preferences, calculated as described in Materials and Methods) at AP-1 sites containing the base substitutions given below the graph are shown for heterodimers labeled on Fos (filled bars) and Jun (open bars). Standard deviations from three or more independent experiments, except at the NC−2C and NC+2G sites, are shown. A high end preference value indicates that the labeled subunit favors binding to the left half-site, whereas a low end preference value indicates a preference for the right half site (arrow on right). The numbering of the base pairs in the AP-1 site is indicated above the sequence. The diagrams below the graph depict the preferred orientations of Fos-Jun heterodimer binding at the different binding sites. Since the subunit that binds to the left half-site at the TGACTCA site can contact the central guanine on the lower strand, it is convenient to consider the central base on the lower strand to be part of the left half-site and the central base on the upper strand to be part of the right half-site. Base substitutions are shown in lowercase, and their positions are marked by an X in the diagrams.

FIG. 2.

Combined effects of contacts to the central base pair and asymmetric base substitutions on the orientation of heterodimer binding. (A) Comparison of the orientation preferences of Fos-Jun heterodimers at the binding sites indicated below the bars. The end preferences of Fos-Jun heterodimers at AP-1 sites with a central CG base pair (NC sites) and at AP-1 sites with a central GC base pair (NG sites) were compared. Diagrams below the graph depict the orientation preferences of Fos-Jun heterodimers at AP-1 sites containing the indicated base substitutions. (B) Comparison of the orientation preferences of Fos-JunRI heterodimers and FosRI-Jun heterodimers at the binding sites indicated below the bars. Filled bars show end preferences for heterodimers labeled on Fos or FosRI, and open bars show end preferences for heterodimers labeled on Jun or JunRI. Standard deviations from three or more independent experiments are shown for complexes at the NC, NC−4A, and NC+4T sites.

FIG. 3.

Effects of the preferred orientation of heterodimer binding on transcription activation by Fos-Jun-NFAT1 complexes. (A) Diagrams illustrate the preferred orientations of FosRI-Jun and Fos-JunRI heterodimer binding at the NG-Renilla and NC-firefly reporter genes. Double-headed arrows indicate the heterodimer orientation that favors cooperative DNA binding with NFAT1 (8, 35, 36). The two promoters were linked to different reporter genes that can be assayed in parallel in the same cell extract (Renilla and firefly luciferase genes). (B) Activation of NG-Renilla (filled bars) and NC-firefly (open bars) reporter gene transcription by Fos-Jun, FosRI-Jun, or Fos-JunRI in the presence or absence of NFAT1. Data shown are averages and standard deviations from four parallel transfection experiments for each complex. Each transfection mixture contained equal amounts of NG-Renilla and NC-firefly reporter plasmids, together with expression vectors encoding the proteins indicated below the bars. (C) Stimulation of transcription by NFAT1 in the presence of different heterodimers. Enhancement by NFAT1 (NFAT1 stimulation) of the transcriptional activity of each promoter was calculated based on the ratio between its transcriptional activities in the presence of a given heterodimer, indicated at the bottom of the figure, with and without NFAT1. Stimulation of promoters linked to Renilla luciferase is shown in the upper panels (solid bars), and stimulation of promoters linked to firefly luciferase is shown in the center panels (open bars). The ratio between the effects of NFAT1 on transcription of the Renilla and firefly reporter genes in the presence of a given heterodimer is shown in the lower panels (striped bars). Data are representative of four independent experiments in which the absolute levels of NFAT1 stimulation were variable but the relative effects of NFAT1 on the activities of heterodimers with opposite orientation preferences were reproducible.

FIG. 4.

Effects of base substitutions and interactions with NFAT1 on the binding orientations of ATF2 heterodimers with Fos and Jun. (A) gelFRET analysis of the binding orientations of Fos-Jun, Fos-ATF2, and Jun-ATF2 heterodimers in the absence and the presence of NFAT1. The proteins indicated above the lanes were incubated with NC site oligonucleotides labeled with fluorescein at either the left (L) or the right (R) end, and the complexes were separated by polyacrylamide gel electrophoresis. The gel was scanned using 488-nm laser excitation, and the emissions from donor (green) and acceptor (red) fluorophores were measured at each position in the gel and were superimposed to produce the image (29, 34). (B) Comparison of the effects of asymmetric base substitutions on the preferred orientations of Fos-Jun, Fos-ATF2, and Jun-ATF2 binding. The end preferences of heterodimers, alone or in complex with NFAT1, were measured at the binding sites given below the bars. The subunit labeled with Texas Red is indicated by the subscript TR. Standard deviations from three or more independent experiments, except for dimers formed by ATF2 at the NC+4C and NC+3G sites, are shown. Diagrams below the graphs depict the orientations of Fos-Jun, Fos-ATF2, and Jun-ATF2 heterodimers in quaternary complexes with NFAT1.

FIG. 5.

Effects of nonconsensus AP-1 site orientation within composite regulatory elements on the efficiency of competition for Fos-Jun-NFAT1 and Jun-Jun-NFAT1 complexes. (A) Comparison of the efficiencies of competition by oligonucleotide competitors for Fos-Jun-NFAT1 and Fos-Jun complexes. Different concentrations (1, 5, or 25 μM) of the competitor oligonucleotides indicated above the lanes containing the same AP-1 site in opposite orientations relative to the NFAT site were incubated with Fos, Jun, a limiting concentration of NFAT1, and the NC site oligonucleotide labeled with fluorescein. The complexes were separated by gel electrophoresis, and the fluorescence emission at each position in the gel was measured using a fluorescence imager. (B) Comparison of the relative efficiencies of competition for Fos-Jun-NFAT1 complexes (upper graph) and Fos-Jun complexes (lower graph) by NC−4A and NG+4T competitor oligonucleotides. The relative amounts of complexes formed in the presence of different competitors are plotted as fractions of the amounts of these complexes formed in the absence of competitors. Diagrams below the graphs depict the asymmetric recognition of the NC−4A and NG+4T sites by Fos-Jun heterodimers. The double-headed arrow indicates the preferential configuration for interactions with NFAT1. (C) Relative efficiencies of competition for NFAT1 by NC−4A and NG+4T competitor oligonucleotides. (D) Relative efficiencies of competition for Jun-Jun-NFAT1 complexes (upper graph) and Jun homodimers (lower graph) by NC−4A and NG+4T competitor oligonucleotides. Standard deviations from three or more independent experiments are shown. Diagrams below the graphs depict the asymmetric recognition of the NC−4A and NG+4T sites by Jun homodimers and their differential interactions with NFAT1.

FIG. 6.

Effects of nonconsensus AP-1 site orientation within composite regulatory elements on the stabilities of Fos-Jun-NFAT1 and Jun-Jun-NFAT1 complexes. Shown is a comparison of the dissociation rates of Fos-Jun-NFAT1 (left panels) and Jun-Jun-NFAT1 (right panels) at the NC−4A and NG+4T sites containing the same AP-1 site in opposite orientations relative to the NFAT site. Jun was labeled with Texas Red, and the oligonucleotides were labeled with fluorescein. Changes in fluorescence emissions from fluorescein (open symbols) and Texas Red (filled symbols) were monitored after addition of an excess of competitor DNA to Fos-Jun-NFAT1 and Jun-Jun-NFAT1 complexes at the binding sites shown to the right of each graph. Changes in fluorescence were normalized to the same range to allow comparison of the rates. Data for each complex were fitted to a first-order exponential function (R < 0.98 for all complexes), and the half-life (t_1/2) was calculated from the best fit. Diagrams indicate the influence of the asymmetric recognition of AP-1 sites on interactions with NFAT1 (double-headed arrows).

FIG. 7.

Effects of nonconsensus AP-1 site orientation within composite regulatory elements on the transcriptional synergy between Jun homodimers and NFAT1. (A) Diagrams illustrate asymmetric recognition of the nonconsensus AP-1 sites within the composite regulatory elements in the NG+4T-Renilla and NC−4A-firefly reporter genes. The promoters contain the same AP-1 site in opposite orientations relative to the NFAT site. A double-headed arrow indicates the AP-1 site orientation that favors cooperative interactions between Jun homodimers and NFAT1 based on the results from oligonucleotide competition and dissociation analyses (Fig. 5 and 6). (B) Comparison of transcription activation by Jun alone and by Jun with NFAT1 at composite regulatory elements containing nonconsensus AP-1 sites in opposite orientations (NG+4T and NC−4A) linked to Renilla (filled bars) and firefly (open bars) reporters. The reporter constructs indicated in each vertical set of graphs were cotransfected into cells with expression vectors encoding the proteins indicated below the graphs, and reporter gene activities were assayed in the same cell extract. Data shown are averages and standard deviations from four parallel transfection experiments for each complex. (C) Effects of NFAT1 on the transcriptional activities of the NG+4T and NC−4A promoters in the presence of Jun homodimers. Stimulation by NFAT1 was calculated based on the ratio between the transcriptional activities of the reporter genes in the presence of Jun and NFAT1 and activities in the presence of Jun alone. Data are representative of five independent experiments, two of which were performed using the NG+4T and NC−4A reporters, and three of which were performed using the NC+4T and NC−4A reporters.