Multi-Domain Metalloproteins

Amann, Barbara

Abstract

Many proteins are built up as arrays of structural units often called domains. Among the most abundant of these domains encoded in the human and other eukaryotic genomes are those domains that are structurally organized around a bound metal ion, most frequently zinc(li). These metal-binding domains are particularly common in proteins that interacts with nucleic acids. The most common and welt-characterized metal-binding domains are the """"""""TFIIIA-type zinc finger"""""""" domains. Proteins containing these arrays bind to specific DNA sequences and regulated gene expression. Because of their structures, arrays of these domains have proven to be extraordinarily useful for designing DNA binding proteins with desired specificities. While naturally occurring amino acids have been varied extensively to discover novel DNA-binding specificities, the use of unnatural amino acids have not been explored. Recently developed synthetic methods make incorporation of unnatural amino acids into such proteins feasible. These studies will expand the utility of these proteins which are currently under development in the private sector as probes and, potentially, therapeutic agents. Other classes of metal-binding domains binds single stranded DNA and RNA. The generation of designed arrays of these domains provides a rational strategy for the design of sequence specific single stranded nucleic acid binding proteins using many of the same principles that have been used for the TFlllA-tike zinc finger domains. These designed proteins should be useful for targeting specific single stranded RNA and DNA sequences. Other classes of metal-binding domains play specific roles in controlling expression in the nervous system. However, the structural basis for this function has not been well established. The recent determination of the three-dimensional structure of one such zinc-binding domain will facilitate further structural and functional characterization of these proteins.