IS1 is an insertion element first identified in Escherichia coli (for a review, see reference 28). IS1 is only 768 bp in length and has imperfect terminal inverted repeats (IR) of about 30 bp (19,30). IS1 transposes to a new site and generates a duplication of a target sequence of 9 bp (in most cases) or 8 bp (in some cases) (13,14,17,29). IS1 has two open reading frames (ORFs), insA and BЈ-insB, which are required for transposing itself (18,25,26). insB is in the Ϫ1 frame with respect to insA. The Ϫ1 frameshifting occurs during translation at the AAAAAAC (A 6 C) sequence in the overlapping region between insA and BЈ-insB. The resulting transframe protein, InsA-BЈ-B, is IS1 transposase (10, 37).IS1 generates the InsA protein from insA unless translational frameshifting occurs. The InsA protein, which has the ability to bind to terminal IR sequences (43, 52, 54), inhibits transposition of IS1 (24, 53). IS1 with a 1-bp insertion in the A 6 C sequence, causing the two frames insA and BЈ-insB to be in the same frame, produces transposase without frameshifting and can thus transpose at a very high frequency (37, 39). These facts suggest that the InsB segment in IS1 transposase forms a catalytic domain. Many elements with structural features similar to those seen in IS1 have been identified in various bacteria and their plasmids (for a review, see reference 28). Most of them are highly homologous to IS1 and to one another (more than 90% sequence identity), whereas a few, such as IS1(NuXi) from Shigella dysenteriae, have low homology, about 55% (31).Transposases encoded by IS3 family elements and retroviral integrases have been found to have three acidic amino acid residues, D, D, and E, constituting a motif called the D-D-E motif, at positions where a polypeptide segment 35 amino acids long is usually present between the second D and third E residues (11, 23). The D-D-E motif has also been found in transposases encoded by IS630, Tn7, Tn552, and others and in integrases encoded by retrotransposons (7,34). Transposases encoded by IS10 and phage Mu also have the D-D-E motif, but a longer polypeptide segment of 130 or 55 amino acids is present between the second D and the E (4, 5, 49). The polypeptide segment with the D-D-E motif forms a catalytic domain in which the three acidic amino acid residues have an essential role in capturing Mg 2ϩ and other divalent cations (4). Tertiary structures of the catalytic core domains of human immunodeficiency virus type 1 (HIV-1) integrase and phage Mu transposase (MuA) and full-length Tn5 transposase have been determined by X-ray crystallography (6, 9, 35). These proteins characteristically have three  sheets in tandem in the segment with the D-D-E motif, such that the first D residue is present in the first  sheet.The transposase encoded by IS1 has been thought not to belong to the family of proteins with the D-D-E motif but to the phage integrase (Int) family (44) because IS1 transposase has the H-R-Y motif, which is conserved in all the Int family proteins (1, 3). In this study,...