Molecular Architecture and Dynamics of Transcription Regulatory Protein Complexes

Tom K. W. Kerppola, Ph.D. Associate Investigator, University of Michigan Medical School Biography...

Summary: Tom Kerppola investigates the assembly and structural organization of transcription factor complexes in an effort to elucidate mechanisms that contribute to the specificity of mammalian gene regulation.

Transcription Factor Interactions Control Gene Expression
The appropriate growth and differentiation of mammalian cells require selective regulation of the expression of each gene in the genome. The independent control of individual transcription units is achieved through interactions among multiple transcription regulatory proteins at each promoter and enhancer. The participation of multiple transcription regulatory proteins is necessary for the specification of unique sites within the genome, whether the regulation occurs at the level of nuclear compartmentalization, chromatin remodeling, coactivator recruitment, or assembly of the transcription machinery. Transcription factors can interact either prior to DNA binding or upon binding to the same regulatory region. These interactions can be regulated both independent of DNA binding and depending on the DNA regulatory region at which they function. Our research is directed toward understanding the assembly and structural organization of transcription regulatory protein complexes.

The Orientation of Fos-Jun Heterodimer Binding
Many mammalian transcription regulatory proteins bind DNA as heterodimeric complexes. Such heterodimers frequently recognize palindromic DNA sequence elements with symmetrical half-sites. These heterodimers can potentially bind to their recognition elements in either of two opposite orientations. Opposite orientations of heterodimer binding can alter interactions with transcription factors that bind to adjacent regulatory elements. We have used heterodimers formed by the proto-oncogene transcription factors Fos and Jun as a model system to investigate the structural basis and functional significance of the orientation of heterodimer binding. To quantify the fraction of heterodimers bound in each orientation, we developed a novel gel-based fluorescence resonance energy transfer assay (gelFRET). We found that Fos-Jun heterodimers bind in opposite orientations to different AP-1 recognition sequences. Remarkably, amino acid residues and base pairs outside the DNA contact interface observed in the Fos–Jun–AP-1 crystal structure influence the orientation of heterodimer binding. Amino acid residues of opposite charge on Fos and on Jun induce opposite directions of DNA bending, and differences in the bending propensities of sequences flanking the AP-1 site determine the preferred orientation of heterodimer binding. Long-range electrostatic interactions are therefore an important determinant of the architecture of nucleoprotein complexes.

Dynamics of Fos-Jun-NFAT1 Complexes
The assembly of multiprotein transcription regulatory complexes by proteins that bind to the same regulatory region may occur through random collisions or facilitated pathways. We have used the Fos-Jun-NFAT1 complex as a model system to investigate the pathway for complex assembly at composite regulatory elements. To investigate the dynamics of transcription complex assembly and conformational isomerization, we developed a multicolor fluorescence resonance energy transfer assay. Cooperative DNA binding by Fos-Jun and NFAT1 requires a specific orientation of heterodimer binding. NFAT1 was able to reverse the orientation of Fos-Jun binding. Remarkably, the orientation of Fos-Jun binding could also be reversed in the presence of protein or DNA competitors. Fos-Jun therefore remain associated with DNA during heterodimer reorientation. At some binding sites, Fos-Jun reorientation was faster than heterodimer dissociation. Fos-Jun heterodimer reorientation in association with DNA can therefore facilitate Fos-Jun-NFAT1 complex assembly at composite regulatory elements.

Functional Consequences of Fos-Jun Binding Orientation
The preferred orientation of heterodimer binding can influence interactions with transcription factors that bind to adjacent DNA sequences. We have investigated the functional consequences of Fos-Jun orientation preference on cooperative interactions with NFAT1 at composite regulatory elements. We developed a multitemplate in vitro transcription assay to compare the transcriptional activities of heterodimers on templates that favor opposite binding orientations. The preferred orientation of Fos-Jun binding affected the stability of Fos-Jun-NFAT1 complexes. Single–base pair substitutions that shifted the orientation of heterodimer binding in opposite directions also had opposite effects on Fos-Jun-NFAT1 complex stability. Mutations in Fos and Jun that reversed the effects of the base substitutions on heterodimer orientation also reversed their effects on Fos-Jun-NFAT1 complex stability. The preferred orientation of Fos-Jun binding also affected the transcriptional activity of Fos-Jun-NFAT1 complexes. Heterodimers that favored the orientation of binding required for interaction with NFAT1 exhibited higher transcriptional activity in cooperation with NFAT1 than heterodimers that favored the opposite binding orientation. Relocation of the NFAT recognition element to the opposite side of the AP-1 site reversed the relative transcriptional activities of the heterodimers. The preferred orientation of heterodimer binding can therefore influence the transcriptional activity and promoter selectivity of Fos-Jun-NFAT1 complexes.

Conformational Divergence Between Maf and Other bZIP Family Proteins
Members of a DNA-binding protein family generally adopt a common secondary structure and contact DNA in a similar manner. We have investigated the conformations of an unusual group of bZIP family proteins related to the proto-oncogene c-maf. These proteins contain an ancillary DNA-binding region on the amino-terminal side of the bZIP domain and recognize DNA sequence elements that are almost twice as long as elements recognized by other bZIP proteins. We have used circular dichroism, protease sensitivity, fluorescence anisotropy, and DNA contact mapping to investigate Maf conformations at different DNA-binding sites. The results of these studies demonstrate that DNA binding by Maf is coupled to conformational changes that vary at different DNA sequence elements. In contrast to other bZIP family proteins, the basic region of Maf does not adopt a fully a-helical conformation upon DNA binding. The atypical structure of the Maf basic region facilitates the conformational change required for recognition of the extended DNA-binding site. The DNA contacts of Maf also differ from those observed in complexes formed by other bZIP proteins. Molecular modeling and threading analysis suggest that the ancillary DNA-binding region forms a helical bundle that contacts the extended DNA recognition sequence. Thus both the secondary structure and the DNA contact interface of Maf differ from those of other bZIP family proteins. DNA-binding proteins that are members of the same protein family based on sequence similarity can therefore differ in the structural basis of DNA recognition.

Last updated: May 25, 2001