A number of different DNA-binding motifs contain one or more zinc atoms whose function is to stabilize the modular structure of the domain by coordinating with amino acids, usually cysteine or histidine, in the appropriate spatial orientation. The only common feature of these domains, of which six have been identified to date, is the presence of zinc. Otherwise they are structurally diverse.
The most ubiquitous of these structures is the "zinc finger" found in many eukaryotic transcription factors. The structure of a typical zinc finger consists of a well-defined a-helix packed against two b-strands arranged in a hairpin structure. The finger structure is stabilized by the tetrahedral coordination of zinc to two closely spaced cysteine residues in the b-strands and two closely spaced histidine residues at the C-terminus of the helical region. In addition the structure is further stabilized by a hydrophobic pocket in the interior of the structure.
When bound to DNA, the finger is directed into the major groove such that base-specific contacts are made by the proximal end of the a-helical region with contacting successive amino acids being separated by approximately one helical turn. The sugar-phosphate backbone is contacted by one of the zinc-coordinated histidine residues. A single zinc finger generally contacts three successive base-pairs in DNA, usually in a single-strand of the duplex. However many zinc-finger proteins contain multiple contiguous fingers separated by a short linker. In such examples synergy can occur between adjacent zinc fingers so that each specify four bp subsites overlapping by one base-pair. In proteins with multiple DNA-binding zinc fingers, the structure of the first finger often differs from the canonical structure. For example, the first finger of the SWI5 protein is preceded by a short a-helix and contains an additional b-strand preceding the normal b hairpin loop. In proteins with multiple zinc fingers, some of the fingers may not be involved in DNA binding but instead can act to mediate protein-protein contacts.
Another DNA-binding motif which contains zinc is exemplified by DNA-binding domain of the steroid receptors. In this fold zinc is coordinated by four cysteine residues. In this case, the path of the polypeptide backbone between the proximal and distal pairs of coordinating cysteines is distinct from that of the TFIIIA type of finger. In the receptor motif, the proximal cysteine residues, as in zinc fingers, reside in the b-sheet structures. By contrast, although the histidine residues involved in zinc coordination are often contained within an a-helix, the corresponding distal cysteine are not. This difference is directly related to DNA binding. The DNA binding domains of the glucocorticoid and oestrogen receptors contain two C-C—C-C coordination domains, the first of which is of similar extent to the TFIIIA fingers whereas the second is significantly shorter. Unlike the TFIIIA fingers both pairs of distal cysteines are immediately followed by an a-helical region. The second a-helix forms direct hydrophobic contacts with the first so that the two fingers and the two a-helices together form a single structural domain, the ‘double-loop-zinc-helix’. This domain also includes a region that specifies the dimerization contacts. The determinants for sequence specific recognition lie within the a-helix immediately distal to the first cysteine tetrad. For each receptor the DNA recognition domains bind as dimers to a palindromic sequence containing three conserved bases on each side. But the separation between the conserved trimers differs for different receptors. This means that the rotational orientation of the individual components of the dimer must also be different and be determined by the position of the dimerization contacts relative to the a-helix responsible for sequence recognition.
In a third type of zinc-containing DNA binding domain, typified by the GAL4 protein, the zinc-stabilized structure is connected by a flexible region of polypeptide chain to a short amphipathic a-helix which forms hydrophobic contacts with another monomer in dimer formation. In GAL4, as with other zinc-containing motifs, direct contact with the bases is mediated in the major groove by amino acids contained within a short a-helical region. However, most of the contacts to the sugar-phosphate backbone are effected by amino acids in the more extended chain. The GAL4 protein, like the steroid hormone receptors, binds to its DNA recognition site as a dimer with the two monomeric units being held together by a short coiled-coil of two parallel a-helices.
