Figure 1. Known steps in replicative and non-replicative (repair) pathways of Mu transposition.

The transposase MuA, in the presence of E. coli protein HU, first introduces single-stranded cleavages at the 3′ ends. With assistance from MuB, the 3′OHs at the cleaved ends are transferred by MuA to phosphodiester bonds spaced 5 bp apart in the target [3], [4]. The resultant branched strand transfer intermediate is processed alternately. During the lytic cycle, Mu is inserted in the chromosome, the target is also in the chromosome, so the purple flanking DNA is continuous with the orange target; transposition is intramolecular. The target OHs found in the strand transfer intermediate are used as primers to replicate Mu (green lines). ClpX, IF2-2 and other uncharacterized factors are required for disassembly of the transpososome followed by assembly of the PriA restart primosome on the Mu ends [5]. During integration of infecting Mu, the purple flanking DNA on the incoming Mu genome is non-covalently joined to itself via phage N protein; transposition into the chromosome target is intermolecular [9], [10], [11]. The branched strand transfer intermediate is resolved/repaired by MuA_Nuc-mediated resection of the flap DNA [13], [14]. ClpX is required for this reaction. The remaining gaps are thought to be filled by host enzymes.