Integration host factor (IHF) is a protein that binds to
Integration host factor (IHF)1 was discovered as a protein required for site-specific recombination of bacteriophage (1). It is a small (20 kDa) basic heterodimer and member of the HU-like family of DNA binding proteins. These proteins have two domains, a helix-turn-helix motif involved in protein-protein dimerization interactions and an arm composed of two antiparallel -sheets that bind to and bend DNA (2, 3). Unlike other HU-like proteins, IHF binds to DNA with a high degree of sequence-directed specificity (4). How IHF achieves sequence specificity has been a challenging question since its discovery.The HЈ site of attP from bacteriophage is one of the best characterized IHF binding sites. The HЈ site is defined as the 34 base pairs protected from DNase I digestion by IHF (4). Alignment of the HЈ site and other known IHF binding sites revealed three conserved sequence elements (5). The WATCAR element (where A, C, G, and T are the standard nucleotide bases, W is A or T, and R is A or G) forms the core of the consensus. It is separated from a second element to its 3Ј side, TTR, by four base pairs that are not conserved among IHF binding sites. The third element, a poly(dAT) tract of 4 -6 base pairs, is found in a subset of the known IHF binding sites. It is located 5Ј to the WATCAR element but is separated from it by approximately eight base pairs. The bases between the poly(dAT) and the WATCAR element also appear not to be conserved among IHF binding sites.The validity of the consensus sequence was confirmed in a genetic study in which base pair substitutions that disrupt IHF binding were isolated within each of these elements (6). One of these mutant HЈ sites, which contains a T to A change at the center position of the TTR element, HЈ44A, was used in a genetic selection to find substitution mutants within IHF that allow it to bind this variant (7). The IHF mutants isolated replaced glutamic acid 44 of the -subunit with a glycine (E44G), lysine (E44K), or valine (E44V). A model based on the crystal structure of HU from Bacillus stearothermophilus predicted that position 44 of the -subunit was in the center of a small -sheet, which is highly conserved among IHF proteins from different species (8). It connects the dimerization domain and the DNA-binding antiparallel -sheets (3). Thus, this -sheet is implicated in specific DNA binding to the TTR element (6). The recent IHF-HЈ site cocrystal structure confirms the conclusions of the studies detailed above and reveals amino acid side chains that contact DNA (2). In this structure, only one amino acid (-Arg 46 ) of this -sheet makes a hydrogen bond with a DNA base of the TTR element. The majority of contacts made by IHF with the HЈ site are to the phosphates and riboses of the DNA backbone. Thus, it appears that IHF recognizes the HЈ site sequence largely through indirect sequence-dependent variation of the DNA backbone conformation.In this report, we examine the importance of the bases of the TTR e...