December 4-7, 2001
Friiberghs Manor, Örsundsbro and Stockholm University
by Aaron Klug
MRC Laboratory of Molecular Biology Hills
Road, Cambridge CB2 2QH, UK
Gene expression is controlled in the first instance at the level of transcription, when the DNA coding region of a gene is transcribed into messenger RNA. This control is exercised by protein transcription factors which bind specifically to the regulatory DNA sequences of the genes, namely, promoters usually found close to the coding region and gene-specific activating sequences further upstream. These interactions create an initiation complex to which is recruited the enzyme RNA polymerase which carries out the transcription.
If we could artificially manipulate the system, we could switch genes on or off, or change the rate of the expression, for example, of those involved in a developmental pathway, or in a disease process.
One approach to the artificial control of gene expression is to try to modify or adapt natural transcription factors, containing potent activation or repression domains, but engineered with new DNA-binding specificities so as to recognise specifically a desired stretch of DNA sequence. These engineered proteins would be targeted to rare DNA-binding sites associated only, or primarily, with the target gene. For this purpose, we have focussed on a particular class of transcription factors, whose DNA binding domains are built out of small structural units called zinc fingers, arranged in tandem, which I discovered some years ago, and which have a simple mode of interaction with DNA.
Each zinc finger (Zf) motif is about thirty amino acids long and forms a compact, independently folded zinc-containing mini-domain, used repeatedly in a modular fashion to achieve sequence-specific recognition of DNA. Zf motifs of the TFIIIA type have turned up in hundreds of proteins, and they appear to be the most widely used of all types of DNA-binding domains. This design may truly be called modular, since the multiply repeated domains all have the same structural framework, but can achieve chemical distinctiveness through variations in certain key amino acid residues.
The zinc finger uses a zinc ion, held by a pair of histidine and a pair of cysteine residues, to stabilise the packing of an antiparallel b-sheet against an a-helix (the "recognition helix"). This helix contacts three adjacent bases (a triplet) in the DNA. Thus the mode of DNA recognition is principally a one-to one interaction between three amino acids in the recognition helix and the nucleotide bases. Moreover, because the fingers function as independent modules, fingers with different triplet specificities can be combined to give specific recognition of longer DNA sequences. The combinatorial design gives great versatility.
For this reason, the zinc finger motifs are ideal natural building blocks for the de novo design of proteins for recognising any given sequence of DNA. The first protein engineering experiments showed that it is possible to alter rationally the DNA-binding characteristics of individual zinc fingers. It became possible to propose some rules relating amino acids on the recognition a-helix to corresponding bases in the bound DNA sequence.
A more efficient alternative to this rational design of proteins with new specificities is the isolation of desirable mutants from a large pool or library of variants. A powerful method of selecting such proteins is the cloning of peptides or protein domains as fusions to the minor coat protein (pIII) of the linear bacteriophage fd, which leads to their expression on the tip of the capsid. Phage displaying the peptides of interest can then be affinity purified by binding to the desired DNA target and then amplified for use in further rounds of selection. The DNA of the selected phage can be sequenced and cloned and hence the amino acid sequence of the particular zinc finger peptide deduced.
The first practical demonstration of the use of this zinc finger technology was made 6 years ago when we described a proof-of-principle experiment in which a three-finger DNA binding domain was engineered and used to down-regulate a leukaemia oncogene in a mouse cell line. Since then, phage display has been used in a number of laboratories, including our own, to select zinc finger peptides against a variety of DNA targets. Examples of the results on the repression and activation of various genes will be described.
In the last few years, a good deal of effort has also gone into improving the specificity of interaction of zinc finger constructs with long tracts of DNA. This has been done, for example, by fusing together two separate zinc finger peptides , each of which is designed to recognise independent DNA sites a certain distance apart. This kind of DNA pattern would have a very low frequency of occurrence, and so constitutes a virtually unique target even in a large genome, as is highly desirable in practical applications.
Printer Friendly 
Comments & Questions
Tell a Friend 
