Abstract
The Tn3-deletion method [Davies and Hutchison, Nucleic Acids Res. 19, 5731-5738, (1991)] was used to sequence a 9.4 kb DNA fragment. Transpositional 'warm' spots were not a limiting factor but a 935 bp 'cold' spot was completed using a synthetic oligonucleotide primer. Two hundred and twenty three miniTn3 insertion sites from three sequencing projects were aligned and a 19 bp asymmetric consensus site was identified. There is no absolute sequence requirement at any position in this consensus, so insertion occurs promiscuously (approximately 37% of sites are potential targets). In our sequencing projects, multiply targeted sites always closely matched the consensus, although not all close matches were targeted frequently. The 935 bp cold spot showed no unusual features when analysed with the consensus sequence. The consensus can be used to accurately predict likely insertion sites in a new sequence. Synthetic oligonucleotides based on the consensus and a known hot spot for Tn3 were mutagenised. These sequences were not hot spots in our vectors, suggesting that the primary sequence alone is not sufficient to create an insertional hot spot. We conclude that some other factor, such as DNA secondary structure, also plays an important role in target site selection for the transposon Tn3.