Nucleic Acids Research Advance Access published online on August 15, 2007
Nucleic Acids Research, doi:10.1093/nar/gkm604
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Nucleic Acid Enzymes |
A randomized library approach to identifying functional lox site domains for the Cre recombinase
University of Colorado, Department of Molecular, Cellular and Developmental Biology, 347 UCB, Boulder, CO 80309-0347
*To whom correspondence should be addressed. Tel: 303-492-7606; Fax: 303-492-8907; Email: sheren{at}colorado.edu
Received May 31, 2007. Revised July 9, 2007. Accepted July 25, 2007.
The bacteriophage P1 Cre/loxP site-specific recombination system is a useful tool in a number of genetic engineering processes. The Cre recombinase has been shown to act on DNA sequences that vary considerably from that of its bacteriophage recognition sequence, loxP. However, little is known about the sequence requirements for functional lox-like sequences. In this study, we have implemented a randomized library approach to identify the sequence characteristics of functional lox site domains. We created a randomized spacer library and a randomized arm library, and then tested them for recombination in vivo and in vitro. Results from the spacer library show that, while there is great plasticity, identity between spacer pairs is the most important factor influencing function, especially in in vitro reactions. The presence of one completely randomized arm in a functional loxP recombination reaction revealed that only three wild-type loxP arms are necessary for successful recombination in Cre-expressing bacteria, and that there are nucleotide preferences at the first three and last three positions of the randomized arm for the most efficiently recombined sequences. Finally, we found that in vitro Cre recombination reactions are much more stringent for evaluating which sequences can support efficient recombination compared to the 294-CRE system.