During infection of Escherichia coli, bacteriophage T4 usurps the host transcriptional machinery, redirecting it to the expression of early, middle, and late phage genes. This machinery is driven by E. coli RNA polymerase, which, like all bacterial polymerases, is composed of a core of subunits (beta, beta', alpha1, alpha2, and omega) that have RNA synthesizing activity and a specificity factor (sigma). The sigma protein identifies the start of transcription by recognizing and binding to sequence elements within promoter DNA. During exponential growth, the primary sigma of E. coli is sigma70, which, like all primary sigmas, is composed of four regions. Sigma70 recognizes DNA elements around positions -10 and -35 of host promoter DNA, using residues in its central portion (regions 2 and 3) and C-terminal portion (region 4), respectively. In addition, residues within region 4 must also interact with a structure within core polymerase, called the beta-flap, to position sigma70 region 4 so it can contact the -35 DNA. T4 takes over E. coli RNA polymerase through the action of phage-encoded factors that interact with polymerase and change its specificity for promoter DNA. Early T4 promoters, which have -10 and -35 elements that are similar to that of the host, are recognized by sigma70 regions 2 and 4, respectively. However, although T4 middle promoters have an excellent match to the sigma70 -10 element, they have a phage element (a MotA box) centered at -30 rather than the sigma70 -35 element. Two T4-encoded proteins, a DNA-binding activator (MotA) and a T4-encoded co-activator (AsiA), are required to activate the middle promoters. AsiA alone inhibits transcription from a large class of E. coli promoters by binding to and structurally remodeling sigma70 region 4, preventing its interaction with the -35 element and with the beta-flap. In addition to its inhibitory activity, the AsiA-induced remodeling allows the N-terminal domain of MotA (MotANTD) to bind to the C-terminus of sigma70 and the C-terminal domain of MotA (MotACTD) to bind to the MotA box. This process is called sigma appropriation. Using multiple available structures (E. coli RNA polymerase, the Thermus aquaticus RNA polymerase/DNA open complex, AsiA /sigma70 Region 4, MotANTD, and MotACTD), the Molsoft ICM program, and extensive biochemical evidence indicating the position of MotA relative to the DNA, we have been developing a structure-based model for sigma appropriation. Determining the structure of a proteinDNA complex can be difficult, particularly if the protein, like MotA, does not bind tightly to the DNA, if there are no homologous proteins from which the DNA binding can be inferred, and/or if only portions of the protein can be crystallized. If the protein comprises just a part of a large multi-subunit complex, other complications can arise such as the complex being too large for NMR studies, or it is not possible to obtain the amounts of protein and nucleic acids needed for crystallographic analyses. We have developed a technique that can be used to map the position of an activator protein relative to the DNA within a large transcription complex. Using this technique, we have determined the position of MotA on the DNA from data generated using activator proteins that had been conjugated at specific residues with the chemical cleaving reagent, iron bromoacetamidobenzyl-EDTA (FeBABE). These analyses have been combined with 3-D models of the available structures of portions of the activator protein and B-form DNA to obtain a 3-D picture of the protein relative to the DNA. Finally, we have used the Molsoft program to refine the position, revealing the architecture of the proteinDNA within the transcription complex. This technique is applicable to many large protein/DNA complexes whose architecture cannot be solved by typical structural analyses.
