Evolution of DNA restriction endonucleases and the creation of new endonucleases for DNA manipulation.

Genetic engineering, essential to the progress of all current biological research, would be impossible without a class of catalysts termed restriction enzymes. These allow the cutting and pasting of segments of DNA to test hypotheses and to create novel routes to vital healthcare products such as insulin. Arguably, it is fair to say that nothing in modern biology, biotechnology and biomedical research would be possible without them. The number of restriction enzymes known is large and many have been commercialised. However new enzymes are still being sought and new uses for them are still being invented. We have recently solved the structure of the first restriction enzyme to be discovered and purified in 1968. These "Type I" restriction enzymes have not so far had impact on healthcare, research or wealth creation despite their abundance in natural populations of bacteria and their impact on slowing down the transfer of DNA encoding, for example, antibiotic resistance and the creation of "superbugs". However our new structure shows how Type I restriction enzymes can be converted into valuable products for research and healthcare. Since thousands of Type I restriction enzymes are known to exist, this ability to convert them promises a veritable bonanza of new and useful restriction enzymes. We propose to perform this conversion process on the Type I restriction enzymes. As well as aiding new research these new enzymes may well prove valuable new tools for gene targeting and editing; a relatively new area studying the genetics of complex organisms particularly with an emphasis on understanding and eventually treating human genetic diseases.

Restriction endonucleases have been the foundation of molecular biology since the 70's. Even today when sophisticated designer nucleases such as site-specific Zinc finger nucleases (ZFNs), TAL effector nucleases (TALENs) can be created for gene targeting and restriction enzymes are considered to be simple tools, restriction enzymes are still used at some stage in almost every molecular biology/biotechnology project for DNA manipulation. New restriction enzymes are still being sought to expand the repertoire of DNA targets and new uses discovered.

The Type I restriction enzymes cut DNA on either side of their target (invariably a bipartite sequence such as AACN6GTGC) but at variable distances (usually thousands of bp distant). The random cutting sites preclude commercial use of these enzymes. The Type IIB cut at defined short distances on either side of their bipartite target (e.g. BcgI cuts |10/12|CGAN6TGC|12/10|) and this specificity makes them commercially valuable.

We have recently determined the first structures of the Type I restriction enzymes. These show how domain deletion and fusion can be achieved in three-D space to create Type IIB (and IIG) restriction enzymes from Type I restriction enzymes. The number of Type IIB/IIG restriction enzymes known is currently small (<30 target specificities and most commercially available). The number of known and putative Type I restriction enzymes in sequenced genomes is far greater (>1000).

Thus the question we wish to ask is whether hundreds of Type I restriction enzymes can be easily converted into commercially viable Type IIB/IIG enzymes with new target specificities?

The Type I and IIB enzymes additionally prefer DNA substrates with two copies of the target sequence, thus the proposed new enzymes should demonstrate activity on DNA containing adjacent copies of the target in much the same manner as the ZFNs and TALENs. Therefore they may be functional in gene targeting as well as genetic engineering.

Our new enzymes may well be able to complement the so-called designer nucleases such as Zinc finger (ZFN) nucleases, TALEN nucleases and meganucleases which have attracted great attention recently due to their ability to target specific DNA loci in eukaryotic genomes. Although these designer nucleases were highlighted in Nature Methods as their "Method of the Year for 2011", several difficulties in their creation and application were noted as still to be overcome before they could reach their full potential, e.g. their hit-or-miss activity (usually a "miss") and a preference for G-rich targets. These difficulties largely arise because the enzymes are being constructed almost from "scratch". By relying upon the restriction enzyme structures already naturally available and honed by hundreds of millions of years of evolution, we hope to avoid or minimise this problem to make a new set of useful tools. Essentially we propose to use natural biobricks (i.e. protein domains) for the creation of modular designer nucleases. This work will impact upon all scientists and clinicians currently attempting to develop gene therapy for the treatment of genetic diseases. Ultimately the public will benefit from this project as new diagnostic and clinical procedures come into use.

Site-specific nucleases are highly commercial entities and numerous companies selling restriction enzymes and latterly the designer nucleases exist in many countries. The best companies involved in restriction enzymes are New England Biolabs (Nobel Laureate Sir Rich Roberts, Research Director) and Fermentas (Vilnius, Lithuania) while many major biochemical suppliers (Roche etc) also market these enzymes. ZFNs are licensed by Sangamo Biosciences (Nobel Laureate Sir Aaron Klug on science advisory board) solely to Sigma-Aldrich while the newer TALENs are currently being intensely studied academically but have also very recently become commercially available (Cellectis). The enzyme suppliers' main markets are not only academic but also the major pharmaceutical and biotechnology companies. Successful completion of this project will generate novel IP, licensing opportunities and products with commercial potential in genetic engineering and gene targeting.