λ Integrase and the λ Int Family

This chapter starts with Allan Campbell's insightful proposal for the pathway by which the chromosome of bacteriophage λ is integrated into and excised from the chromosome of its Escherichia coli host. The chapter discusses different levels of λ Int family complexity. There are presently four classes of integron integrases, IntI1 to -4, sharing approximately 50% identity. In contrast to the IntI1 integrase, which has not suggested any striking deviations from its cousins, the integron recombination targets, attI1 and attC , present several new variations on λ Int family themes. Recombination between att sites with identical 7-bp spacer regions is not any more efficient than that for two sites differing at five positions. Recent evidence suggests that λ Int may have gone even further in evolving a dependence on arm binding. Whereas full Int binds very poorly to core-type DNA sites, C65 (lacking the N-terminal domain) binds very well. The amino acid and nucleotide residues responsible for distinction have been variously identified by genetic selections, construction of chimeric integrases (via recombination or site-directed mutagenesis), and alteration of core sites. Superimposing the crystal structure of the catalytic domain of vaccinia virus topoisomerase on the coordinates of the previously reported HP1 and Cre recombinase structures, it is apparent that the order and topology of the secondary and tertiary structural elements are strikingly similar. A unified topological mechanism of site-specific recombination by λ integrase family members has remained elusive until fairly recently, in contrast to the well-characterized mechanisms for several enzymes of the resolvase/invertase family.