Published online 7 December 2004
Nucleic Acids Research, Vol. 32 No. 21 © Oxford University Press 2004; all rights reserved
Predicting oligonucleotide-directed mutagenesis failures in protein engineering
1 Department of Computer Science, 2 Department of Chemistry, 3 Department of Biomedical Engineering and 4 Department of Molecular Biology and Biochemistry, University of California, Irvine, CA 92697-2025, USA
* To whom correspondence should be addressed. Tel: +1 949 824 5566; Fax: +1 949 824 9920; Email: gweiss{at}uci.edu
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors
Received August 28, 2004; Revised November 3, 2004; Accepted November 15, 2004
Protein engineering uses oligonucleotide-directed mutagenesis to modify DNA sequences through a two-step process of hybridization and enzymatic synthesis. Inefficient reactions confound attempts to introduce mutations, especially for the construction of vast combinatorial protein libraries. This paper applied computational approaches to the problem of inefficient mutagenesis. Several results implicated oligonucleotide annealing to non-target sites, termed ‘cross-hybridization’, as a significant contributor to mutagenesis reaction failures. Test oligonucleotides demonstrated control over reaction outcomes. A novel cross-hybridization score, quickly computable for any plasmid and oligonucleotide mixture, directly correlated with yields of deleterious mutagenesis side products. Cross-hybridization was confirmed conclusively by partial incorporation of an oligonucleotide at a predicted cross-hybridization site, and by modification of putative template secondary structure to control cross-hybridization. Even in low concentrations, cross-hybridizing species in mixtures poisoned reactions. These results provide a basis for improved mutagenesis efficiencies and increased diversities of cognate protein libraries.