GCN4 Leucine Zipper and Homodimer

Donald Auer, Aaron Downs, and David Marcey, 1997

© David Marcey, 1997

Contents:

References

O'Shea K. Erin, Klemm D. Juli, Kim S. Peter, Alber Tom. X-Ray Structure of the GCN4 Leucine Zipper, A two-Stranded, Parallel Coiled Coil. Science, Vol 254. 539-534, 1991.

Elenberger E. Thomas, Brandl J. Christopher, Struhl Kevin, Harrison C. Stephen. The GCN4 Basic Region Leucine Zipper Binds DNA as a Dimer of Uninterrupted Alpha Helices: Crystal Structure of the Protein-DNA Complex. Cell, Vol 71. 1223-1237, 1992.

Konig Peter, Richmond J. Timothy. The X-Ray Structure of the GCN4-bZIP to ATF/CREB Site DNA Shows the Complex Depends on DNA Flexibility. Journal of Molecular Biology, Vol. . 139-153, 1993.

I. Introduction

 Many gene regulatory proteins form dimers in their recognition and binding of DNA; these protein structures achieve strong, specific binding by contacting two half sites on opposing sides of the DNA. Most of these proteins have regions responsible for dimerization, and for DNA recognition and interaction. The leucine zipper motif contains two alpha helices from two monomers which form a coiled coil and continue beyond the point of physical dimerization to contact the major groove of DNA. Leucine zipper proteins can form either homodimers or heterodimers. There are many types of leucine zipper proteins in organisms as diverse as eukaryotes and yeast that regulate the expression of a wide array of genes. Whether or not a particular homodimer or heterodimer forms depends upon the particular hydrophobic side chains of each monomer and the interactions between these side chains. The GCN4 homodimer is a yeast transcriptional activator of the basic-region-leucine-zipper, bZIP motif, responsible for the general control of amino acid biosynthesis. The intact and complete protein is composed of a total of 281 residues which forms a coiled coil <> over the monomers' carboxy-termini 33 residues, and continues beyond the point of dimerization towards the N-terminus and the protein's 25 residue basic region. The interface of the coiled coil dimer is aligned perpendicular to the DNA axis. The coiled coil orients each basic region over its respective DNA binding half-site, and these basic regions are secured in the major groove by positively charged and polar residues which provide hydrogen bonds to unesterfied oxygens of the phosphodiester backbone; it is these bonds which align the basic region side chains for a series of base-specific contacts.

II. GCN4 Dimerization

 Dimerization is controlled by the leucine zipper which is a conserved sequence motif of the protein. This peptide is a dimer of parallel, amphipathic alpha helices that forms a shallow left-handed coiled coil. The alpha helices dimerize across the 32 residues of their carboxy-termini, and form a crossing angle of 18 degrees <>. The dimer is around 45 angstroms long and 30 angstroms wide. The alpha helices wrap around each other and form a one-fourth turn of a left-handed supercoil. The formation of this parallel coiled coil is dependent upon hydrophobic residues spaced every four and then three residues apart in the primary sequence; this spacing is known as the heptad repeat and is found in other alpha helical proteins such as keratin and fibrinogen. The GCN4 coiled coil is much shorter than these classical coiled coils. Classical coils have functions as diverse as protein recognition, motility, and cell structure, while the GCN4 coil is responsible for dimerization alone. The helical repeat is 3.5 residues per turn, as compared to the 3.6 residues per turn in a regular alpha helix. This helical repeat allows the interaction pattern of side chains between the helices to repeat integrally every 7 residues. Therefore, stability of this dimer is dependent upon the hydrophobic sidechains, Leu5, Leu12, Leu19, and Leu26, Val9, Asn16, Val23, and Val30 of one helix packing partly adjacent to the corresponding sidechains of the second helix <>. The distinctive hydrogen bond formed between the polar Asn16 side chains <> may actually have a destabilizing function. These polar side chains bury themselves within the protein, disrupting interhelical ion pairing, and packing less tightly against adjacent layers than Val or Met side chains at position (a) of the heptad repeat. Therefore, Asn16 may aid in the formation of the parallel nature of the coiled coil characteristic of the bZIP class. An anti-parallel arrangement would cause the polar Asn16 substituents to pack against non-polar residues: an energetically unfavorable and destabilizing effect. This destabilization could control the concentration of functional dimers in vivo and thus control the affinity of the protein for DNA. Positive and negative residues generally alternate throughout the helices, producing a net charge around zero; this alternation enables the formation of inter- and intrahelical ion pairing. Interhelical ion pairs are formed between Lys15 <> and Glu20' <>, Glu22 and Lys27', and between Glu22' and Lys27 <>. Intrahelical ion pairing is evident between Lys8 <> and Glu11 <> and between Glu22 <> and Arg25 <>. These intra- and interhelical ion pairings add to the rigidity of the dimer and the individual alpha helices.

III. GCN4 Dimer and DNA Interaction

 Click here to see GCN4 Dimer and DNA interaction.

The DNA-binding element is a recognition module that is unfolded in solution, and commands a stable conformation when bound to DNA; this model is known as the induced helical fork model. This basic region, which composes 25 residues projecting towards the amino-terminal, passes through the major groove of the DNA binding site and continues beyond DNA interaction. The GCN4 interaction with DNA <> , displayed previously, stabilizes the protein complementary to this binding site.

The pseudo-symmetric Ap-1 site of DNA is the functional target of GCN4 in vivo. The DNA bound to GCN4 across the AP-1 site is straight and B-form across the region that is contacted by the protein dimer. The left half-site monomer contacts the DNA through the following side chains: Arg249; Lys246; Arg245; Arg243; Ser242; Arg241; Arg240; Ala239; Ala238; Thr236; Asn235; Arg234 <> . The right half-site monomer contacts the DNA bases through the following side chains: Arg249; Lys246; Arg245; Arg243; Ser242; Arg241; Ala239; Ala238; Thr236; Arg234; Asn235; Arg232.

The GCN4 dimer has also been demonstrated to bind in vitro to the symmetric ATF/CREB site with the same affinity as that for the AP-1 site. The additional G-C base pair at the center of the ATF/CREB site creates a difference in the geometric relationship between the two half-site TCAT sequences. This difference corresponds to relative displacements of the DNA groups available for GCN4 interaction of roughly 4 angstroms for the bases and 7 angstroms for the phosphates. Therefore, unlike protein binding in the AP-1 site, DNA flexibility plays a predominant role in the preservation of protein contacts with the ATF/CREB site. The DNA fragment is bent symmetrically by 20 degrees at its center toward the leucine zipper basic region. This DNA flexing and unwinding situates the two half-sites in relatively the same orientation with respect to each other as in the AP-1 site. Also, the two structural DNA forms possess similar energies in regard to one another. Several basic region residues make contacts to the ATF/CREB half-sites of DNA bases and phosphate and phosphate oxygens; 13 of 16 consecutive amino acids from each monomer make these contacts with DNA bases.

IV. Conclusions

 An understanding of the bZIP motif is important, considering the prevalence of its relative structure amongst a diverse array of transcriptional activators. Muscle proteins, alpha-keratin, bacterial surface proteins, intermediate filaments, laminins, dynein, tumor suppressors, and oncogene products are a number of proteins incorporating the leucine zipper into their structure. Thus, the structure of the GCN4 coiled coil can be used to model interactions in heterodimers and other parallel two-stranded coiled coils. The GCN4 and ATF/CREB interaction suggests that DNA bending can be induced by bZIP binding. In this ATF/CREB site complex, DNA bending is required in order that the two DNA half-sites have an alignment close to that for the AP-1 and DNA interaction. Thus, DNA flexing is a direct result of the dimeric interaction. Besides the AP-1 and ATF/CREB binding sites, more divergent sequences are recognized by other members of the bZIP class: C/EMP recognizes the sequence (ATTGCGCAAT); TAF-1 and EMBP-1 recognizes (CACGTGGC); and Zta recognizes ((TG(A/T)GCAAG)). Also, like GCN4 and the AP-1 and ATF/CREB sites, other bZIP proteins can bind with comparable affinity a second DNA binding sequence. For example, Fos/Jun can bind to the AP-1 site; Zta can bind to the AP-1 site; and TAF-1 and EmBP-1 can bind to ATF/CREB like sequences.

V. Future Studies

 The alpha helices of the GCN4 protein homodimer are thought to pass beyond the site of DNA interaction and continue in a straight helical path towards the N-terminal. However, recent studies have indicated, but not proved with crystal structure, that the helices may actually kink and wrap around the path of the major groove.

Useful Links

1 Crystal structures of these complexes and others mentioned can be found in the Brookhaven Protein Databank.

2 Images of the GCN4 leucine zipper.

3 Leucine zipper coiled coil structure.

4 The leucine zipper and eukaryotic gene regulatory proteins.

References

O'Shea K. Erin, Klemm D. Juli, Kim S. Peter, Alber Tom. X-Ray Structure of the GCN4 Leucine Zipper, A two-Stranded, Parallel Coiled Coil. Science, Vol 254. 539-534, 1991.

Elenberger E. Thomas, Brandl J. Christopher, Struhl Kevin, Harrison C. Stephen. The GCN4 Basic Region Leucine Zipper Binds DNA as a Dimer of Uninterrupted Alpha Helices: Crystal Structure of the Protein-DNA Complex. Cell, Vol 71. 1223-1237, 1992.

Konig Peter, Richmond J. Timothy. The X-Ray Structure of the GCN4-bZIP to ATF/CREB Site DNA Shows the Complex Depends on DNA Flexibility. Journal of Molecular Biology, Vol. . 139-153, 1993.

BACK TO KENYON'S HYPERACTIVE MACROMOLECULE PAGE
Comments or Suggestions to:David Marcey at marcey@kenyon.edu